{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.338078291814947,
  "eval_steps": 3000,
  "global_step": 54000,
  "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
    },
    {
      "entropy": 5.655107164382935,
      "epoch": 2.076413602214314,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004571366264229162,
      "loss": 5.0675,
      "mean_token_accuracy": 0.2149159789085388,
      "num_tokens": 48109813.0,
      "step": 21005
    },
    {
      "entropy": 5.640252590179443,
      "epoch": 2.0769078687228153,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004571159045436807,
      "loss": 5.0063,
      "mean_token_accuracy": 0.2218811556696892,
      "num_tokens": 48120736.0,
      "step": 21010
    },
    {
      "entropy": 5.630234146118164,
      "epoch": 2.0774021352313166,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004570951781843657,
      "loss": 4.9748,
      "mean_token_accuracy": 0.22196894884109497,
      "num_tokens": 48131631.0,
      "step": 21015
    },
    {
      "entropy": 5.573506116867065,
      "epoch": 2.0778964017398183,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045707444734548116,
      "loss": 4.9171,
      "mean_token_accuracy": 0.22053711712360383,
      "num_tokens": 48141662.0,
      "step": 21020
    },
    {
      "entropy": 5.551257705688476,
      "epoch": 2.0783906682483195,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004570537120275369,
      "loss": 4.9712,
      "mean_token_accuracy": 0.22225781977176667,
      "num_tokens": 48152752.0,
      "step": 21025
    },
    {
      "entropy": 5.687500429153443,
      "epoch": 2.078884934756821,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000457032972231043,
      "loss": 5.0249,
      "mean_token_accuracy": 0.21508511006832123,
      "num_tokens": 48163510.0,
      "step": 21030
    },
    {
      "entropy": 5.64520378112793,
      "epoch": 2.079379201265322,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004570122279565098,
      "loss": 5.0406,
      "mean_token_accuracy": 0.21083574146032333,
      "num_tokens": 48174241.0,
      "step": 21035
    },
    {
      "entropy": 5.641775465011596,
      "epoch": 2.0798734677738238,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045699147920444757,
      "loss": 5.0043,
      "mean_token_accuracy": 0.2137088805437088,
      "num_tokens": 48186062.0,
      "step": 21040
    },
    {
      "entropy": 5.650915813446045,
      "epoch": 2.080367734282325,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004569707259753667,
      "loss": 4.8833,
      "mean_token_accuracy": 0.22385343611240388,
      "num_tokens": 48197108.0,
      "step": 21045
    },
    {
      "entropy": 5.55268349647522,
      "epoch": 2.0808620007908263,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004569499682697778,
      "loss": 4.9311,
      "mean_token_accuracy": 0.23262353092432023,
      "num_tokens": 48208780.0,
      "step": 21050
    },
    {
      "entropy": 5.617620944976807,
      "epoch": 2.081356267299328,
      "grad_norm": 1.0,
      "learning_rate": 0.0004569292060881914,
      "loss": 4.9797,
      "mean_token_accuracy": 0.22262774258852006,
      "num_tokens": 48223164.0,
      "step": 21055
    },
    {
      "entropy": 5.609575939178467,
      "epoch": 2.0818505338078293,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045690843943111824,
      "loss": 4.9233,
      "mean_token_accuracy": 0.22118592113256455,
      "num_tokens": 48234402.0,
      "step": 21060
    },
    {
      "entropy": 5.6509024143219,
      "epoch": 2.0823448003163305,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045688766829906936,
      "loss": 4.9891,
      "mean_token_accuracy": 0.21137000918388366,
      "num_tokens": 48245147.0,
      "step": 21065
    },
    {
      "entropy": 5.655809688568115,
      "epoch": 2.0828390668248318,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004568668926925556,
      "loss": 5.0576,
      "mean_token_accuracy": 0.20728397071361543,
      "num_tokens": 48256787.0,
      "step": 21070
    },
    {
      "entropy": 5.632467603683471,
      "epoch": 2.0833333333333335,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000456846112612088,
      "loss": 4.9902,
      "mean_token_accuracy": 0.20940980911254883,
      "num_tokens": 48267648.0,
      "step": 21075
    },
    {
      "entropy": 5.612430000305176,
      "epoch": 2.0838275998418347,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045682532805817796,
      "loss": 4.9991,
      "mean_token_accuracy": 0.21526008993387222,
      "num_tokens": 48279274.0,
      "step": 21080
    },
    {
      "entropy": 5.636698818206787,
      "epoch": 2.084321866350336,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004568045390313365,
      "loss": 4.9994,
      "mean_token_accuracy": 0.218184894323349,
      "num_tokens": 48290849.0,
      "step": 21085
    },
    {
      "entropy": 5.645367527008057,
      "epoch": 2.0848161328588377,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045678374553207525,
      "loss": 5.0725,
      "mean_token_accuracy": 0.2104203835129738,
      "num_tokens": 48301271.0,
      "step": 21090
    },
    {
      "entropy": 5.6737768173217775,
      "epoch": 2.085310399367339,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045676294756090565,
      "loss": 5.0029,
      "mean_token_accuracy": 0.20916156172752381,
      "num_tokens": 48311869.0,
      "step": 21095
    },
    {
      "entropy": 5.667511796951294,
      "epoch": 2.0858046658758402,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004567421451183393,
      "loss": 5.006,
      "mean_token_accuracy": 0.21990291327238082,
      "num_tokens": 48324907.0,
      "step": 21100
    },
    {
      "entropy": 5.621210098266602,
      "epoch": 2.0862989323843415,
      "grad_norm": 1.0,
      "learning_rate": 0.000456721338204888,
      "loss": 4.9521,
      "mean_token_accuracy": 0.22186561971902846,
      "num_tokens": 48335887.0,
      "step": 21105
    },
    {
      "entropy": 5.645153665542603,
      "epoch": 2.086793198892843,
      "grad_norm": 1.125,
      "learning_rate": 0.0004567005268210636,
      "loss": 4.9644,
      "mean_token_accuracy": 0.21500124037265778,
      "num_tokens": 48348190.0,
      "step": 21110
    },
    {
      "entropy": 5.699085903167725,
      "epoch": 2.0872874654013445,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045667971096737805,
      "loss": 5.126,
      "mean_token_accuracy": 0.20489919036626816,
      "num_tokens": 48360113.0,
      "step": 21115
    },
    {
      "entropy": 5.700635099411011,
      "epoch": 2.0877817319098457,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004566588906443434,
      "loss": 5.0363,
      "mean_token_accuracy": 0.21095264106988906,
      "num_tokens": 48371463.0,
      "step": 21120
    },
    {
      "entropy": 5.640934181213379,
      "epoch": 2.088275998418347,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004566380658524719,
      "loss": 5.1143,
      "mean_token_accuracy": 0.20196683555841446,
      "num_tokens": 48383969.0,
      "step": 21125
    },
    {
      "entropy": 5.627220726013183,
      "epoch": 2.0887702649268487,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004566172365922758,
      "loss": 4.9472,
      "mean_token_accuracy": 0.22313515990972518,
      "num_tokens": 48395551.0,
      "step": 21130
    },
    {
      "entropy": 5.674346256256103,
      "epoch": 2.08926453143535,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004565964028642675,
      "loss": 4.9938,
      "mean_token_accuracy": 0.21910232901573182,
      "num_tokens": 48408154.0,
      "step": 21135
    },
    {
      "entropy": 5.563052320480347,
      "epoch": 2.089758797943851,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004565755646689595,
      "loss": 4.9448,
      "mean_token_accuracy": 0.22279797941446305,
      "num_tokens": 48420060.0,
      "step": 21140
    },
    {
      "entropy": 5.642939472198487,
      "epoch": 2.0902530644523525,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004565547220068644,
      "loss": 5.0458,
      "mean_token_accuracy": 0.21619906425476074,
      "num_tokens": 48432347.0,
      "step": 21145
    },
    {
      "entropy": 5.777932024002075,
      "epoch": 2.090747330960854,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004565338748784951,
      "loss": 5.075,
      "mean_token_accuracy": 0.20275344401597978,
      "num_tokens": 48443177.0,
      "step": 21150
    },
    {
      "entropy": 5.718689489364624,
      "epoch": 2.0912415974693555,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004565130232843642,
      "loss": 5.0102,
      "mean_token_accuracy": 0.2108388900756836,
      "num_tokens": 48453605.0,
      "step": 21155
    },
    {
      "entropy": 5.620134067535401,
      "epoch": 2.0917358639778567,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004564921672249848,
      "loss": 5.0401,
      "mean_token_accuracy": 0.21297027468681334,
      "num_tokens": 48465364.0,
      "step": 21160
    },
    {
      "entropy": 5.612469291687011,
      "epoch": 2.0922301304863584,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045647130670086996,
      "loss": 4.9706,
      "mean_token_accuracy": 0.21598804444074632,
      "num_tokens": 48477255.0,
      "step": 21165
    },
    {
      "entropy": 5.652618885040283,
      "epoch": 2.0927243969948597,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004564504417125328,
      "loss": 5.0123,
      "mean_token_accuracy": 0.2177692875266075,
      "num_tokens": 48487551.0,
      "step": 21170
    },
    {
      "entropy": 5.708688831329345,
      "epoch": 2.093218663503361,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045642957226048667,
      "loss": 4.9853,
      "mean_token_accuracy": 0.21612523198127748,
      "num_tokens": 48500422.0,
      "step": 21175
    },
    {
      "entropy": 5.5957526683807375,
      "epoch": 2.093712930011862,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045640869834524486,
      "loss": 4.9725,
      "mean_token_accuracy": 0.22203195840120316,
      "num_tokens": 48511588.0,
      "step": 21180
    },
    {
      "entropy": 5.648880910873413,
      "epoch": 2.094207196520364,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004563878199673209,
      "loss": 5.0548,
      "mean_token_accuracy": 0.2171315848827362,
      "num_tokens": 48522700.0,
      "step": 21185
    },
    {
      "entropy": 5.743068981170654,
      "epoch": 2.094701463028865,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004563669371272285,
      "loss": 5.0624,
      "mean_token_accuracy": 0.2143555298447609,
      "num_tokens": 48534253.0,
      "step": 21190
    },
    {
      "entropy": 5.600001096725464,
      "epoch": 2.0951957295373664,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045634604982548125,
      "loss": 4.9632,
      "mean_token_accuracy": 0.21917160153388976,
      "num_tokens": 48545543.0,
      "step": 21195
    },
    {
      "entropy": 5.674845027923584,
      "epoch": 2.095689996045868,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045632515806259306,
      "loss": 5.0638,
      "mean_token_accuracy": 0.2094424247741699,
      "num_tokens": 48557433.0,
      "step": 21200
    },
    {
      "entropy": 5.687759304046631,
      "epoch": 2.0961842625543694,
      "grad_norm": 1.0,
      "learning_rate": 0.00045630426183907786,
      "loss": 5.0386,
      "mean_token_accuracy": 0.21329042911529542,
      "num_tokens": 48570271.0,
      "step": 21205
    },
    {
      "entropy": 5.641349458694458,
      "epoch": 2.0966785290628707,
      "grad_norm": 1.125,
      "learning_rate": 0.0004562833611554497,
      "loss": 4.9875,
      "mean_token_accuracy": 0.21511641442775725,
      "num_tokens": 48582376.0,
      "step": 21210
    },
    {
      "entropy": 5.650782632827759,
      "epoch": 2.097172795571372,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004562624560122227,
      "loss": 5.0296,
      "mean_token_accuracy": 0.21544295847415923,
      "num_tokens": 48595188.0,
      "step": 21215
    },
    {
      "entropy": 5.598039293289185,
      "epoch": 2.0976670620798736,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004562415464099112,
      "loss": 4.9283,
      "mean_token_accuracy": 0.2194436475634575,
      "num_tokens": 48606093.0,
      "step": 21220
    },
    {
      "entropy": 5.646231746673584,
      "epoch": 2.098161328588375,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004562206323490295,
      "loss": 5.0503,
      "mean_token_accuracy": 0.21346221268177032,
      "num_tokens": 48618163.0,
      "step": 21225
    },
    {
      "entropy": 5.700385427474975,
      "epoch": 2.098655595096876,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045619971383009216,
      "loss": 5.0371,
      "mean_token_accuracy": 0.22046100199222565,
      "num_tokens": 48629771.0,
      "step": 21230
    },
    {
      "entropy": 5.66089391708374,
      "epoch": 2.0991498616053774,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004561787908536138,
      "loss": 4.9957,
      "mean_token_accuracy": 0.21111521124839783,
      "num_tokens": 48640409.0,
      "step": 21235
    },
    {
      "entropy": 5.598044681549072,
      "epoch": 2.099644128113879,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045615786342010896,
      "loss": 4.9993,
      "mean_token_accuracy": 0.21732162386178971,
      "num_tokens": 48652223.0,
      "step": 21240
    },
    {
      "entropy": 5.6727148532867435,
      "epoch": 2.1001383946223804,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045613693153009264,
      "loss": 4.8977,
      "mean_token_accuracy": 0.22795259654521943,
      "num_tokens": 48663667.0,
      "step": 21245
    },
    {
      "entropy": 5.7089094638824465,
      "epoch": 2.1006326611308817,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004561159951840797,
      "loss": 5.0557,
      "mean_token_accuracy": 0.21293465346097945,
      "num_tokens": 48674691.0,
      "step": 21250
    },
    {
      "entropy": 5.634056138992309,
      "epoch": 2.1011269276393834,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004560950543825852,
      "loss": 5.0149,
      "mean_token_accuracy": 0.21330103725194932,
      "num_tokens": 48686888.0,
      "step": 21255
    },
    {
      "entropy": 5.630914258956909,
      "epoch": 2.1016211941478846,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045607410912612426,
      "loss": 5.0201,
      "mean_token_accuracy": 0.2149464949965477,
      "num_tokens": 48698666.0,
      "step": 21260
    },
    {
      "entropy": 5.647128582000732,
      "epoch": 2.102115460656386,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004560531594152121,
      "loss": 4.9529,
      "mean_token_accuracy": 0.22177106738090516,
      "num_tokens": 48709793.0,
      "step": 21265
    },
    {
      "entropy": 5.583790683746338,
      "epoch": 2.102609727164887,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004560322052503642,
      "loss": 4.8893,
      "mean_token_accuracy": 0.2306046724319458,
      "num_tokens": 48720835.0,
      "step": 21270
    },
    {
      "entropy": 5.71058087348938,
      "epoch": 2.103103993673389,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004560112466320959,
      "loss": 5.1491,
      "mean_token_accuracy": 0.20738973319530488,
      "num_tokens": 48732990.0,
      "step": 21275
    },
    {
      "entropy": 5.6254298210144045,
      "epoch": 2.10359826018189,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045599028356092293,
      "loss": 4.9547,
      "mean_token_accuracy": 0.21950895786285402,
      "num_tokens": 48744985.0,
      "step": 21280
    },
    {
      "entropy": 5.580562400817871,
      "epoch": 2.1040925266903914,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004559693160373608,
      "loss": 4.9273,
      "mean_token_accuracy": 0.22839030921459197,
      "num_tokens": 48755328.0,
      "step": 21285
    },
    {
      "entropy": 5.622778415679932,
      "epoch": 2.1045867931988926,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004559483440619255,
      "loss": 4.9624,
      "mean_token_accuracy": 0.21852560639381408,
      "num_tokens": 48766666.0,
      "step": 21290
    },
    {
      "entropy": 5.666731595993042,
      "epoch": 2.1050810597073943,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045592736763513285,
      "loss": 5.0115,
      "mean_token_accuracy": 0.21371037065982817,
      "num_tokens": 48779134.0,
      "step": 21295
    },
    {
      "entropy": 5.676667594909668,
      "epoch": 2.1055753262158956,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004559063867574989,
      "loss": 5.0321,
      "mean_token_accuracy": 0.2083706811070442,
      "num_tokens": 48791338.0,
      "step": 21300
    },
    {
      "entropy": 5.6195532321929935,
      "epoch": 2.106069592724397,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00045588540142953973,
      "loss": 4.9497,
      "mean_token_accuracy": 0.22276730239391326,
      "num_tokens": 48802711.0,
      "step": 21305
    },
    {
      "entropy": 5.6214611530303955,
      "epoch": 2.1065638592328986,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004558644116517717,
      "loss": 4.9737,
      "mean_token_accuracy": 0.2244958609342575,
      "num_tokens": 48815219.0,
      "step": 21310
    },
    {
      "entropy": 5.604186773300171,
      "epoch": 2.1070581257414,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045584341742471095,
      "loss": 4.9991,
      "mean_token_accuracy": 0.21294087320566177,
      "num_tokens": 48826075.0,
      "step": 21315
    },
    {
      "entropy": 5.590073490142823,
      "epoch": 2.107552392249901,
      "grad_norm": 1.125,
      "learning_rate": 0.0004558224187488742,
      "loss": 4.9039,
      "mean_token_accuracy": 0.2194947436451912,
      "num_tokens": 48837375.0,
      "step": 21320
    },
    {
      "entropy": 5.66591854095459,
      "epoch": 2.1080466587584024,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045580141562477785,
      "loss": 5.0362,
      "mean_token_accuracy": 0.2112525761127472,
      "num_tokens": 48849205.0,
      "step": 21325
    },
    {
      "entropy": 5.6284441471099855,
      "epoch": 2.108540925266904,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045578040805293864,
      "loss": 4.9894,
      "mean_token_accuracy": 0.22381118386983873,
      "num_tokens": 48861700.0,
      "step": 21330
    },
    {
      "entropy": 5.708299684524536,
      "epoch": 2.1090351917754053,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045575939603387336,
      "loss": 5.1444,
      "mean_token_accuracy": 0.20869310647249223,
      "num_tokens": 48873510.0,
      "step": 21335
    },
    {
      "entropy": 5.711788177490234,
      "epoch": 2.1095294582839066,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004557383795680988,
      "loss": 5.0152,
      "mean_token_accuracy": 0.21011894941329956,
      "num_tokens": 48885011.0,
      "step": 21340
    },
    {
      "entropy": 5.6726264476776125,
      "epoch": 2.1100237247924083,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045571735865613215,
      "loss": 4.9547,
      "mean_token_accuracy": 0.2200231209397316,
      "num_tokens": 48895767.0,
      "step": 21345
    },
    {
      "entropy": 5.649470043182373,
      "epoch": 2.1105179913009096,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045569633329849043,
      "loss": 5.0896,
      "mean_token_accuracy": 0.2028256833553314,
      "num_tokens": 48906616.0,
      "step": 21350
    },
    {
      "entropy": 5.673553133010865,
      "epoch": 2.111012257809411,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004556753034956909,
      "loss": 5.0197,
      "mean_token_accuracy": 0.21206318438053132,
      "num_tokens": 48916841.0,
      "step": 21355
    },
    {
      "entropy": 5.645083284378051,
      "epoch": 2.111506524317912,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004556542692482508,
      "loss": 4.9765,
      "mean_token_accuracy": 0.2180880293250084,
      "num_tokens": 48928798.0,
      "step": 21360
    },
    {
      "entropy": 5.608558654785156,
      "epoch": 2.112000790826414,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004556332305566877,
      "loss": 4.965,
      "mean_token_accuracy": 0.22064916044473648,
      "num_tokens": 48940235.0,
      "step": 21365
    },
    {
      "entropy": 5.645768165588379,
      "epoch": 2.112495057334915,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004556121874215191,
      "loss": 4.9745,
      "mean_token_accuracy": 0.21920240223407744,
      "num_tokens": 48951718.0,
      "step": 21370
    },
    {
      "entropy": 5.643590927124023,
      "epoch": 2.1129893238434163,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004555911398432627,
      "loss": 4.9577,
      "mean_token_accuracy": 0.2176239550113678,
      "num_tokens": 48963215.0,
      "step": 21375
    },
    {
      "entropy": 5.718164396286011,
      "epoch": 2.1134835903519176,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045557008782243626,
      "loss": 5.0684,
      "mean_token_accuracy": 0.2136334538459778,
      "num_tokens": 48975761.0,
      "step": 21380
    },
    {
      "entropy": 5.641120100021363,
      "epoch": 2.1139778568604193,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004555490313595575,
      "loss": 4.9746,
      "mean_token_accuracy": 0.22332814633846282,
      "num_tokens": 48987192.0,
      "step": 21385
    },
    {
      "entropy": 5.580223798751831,
      "epoch": 2.1144721233689205,
      "grad_norm": 1.0,
      "learning_rate": 0.0004555279704551447,
      "loss": 4.934,
      "mean_token_accuracy": 0.21804456561803817,
      "num_tokens": 48999503.0,
      "step": 21390
    },
    {
      "entropy": 5.631662130355835,
      "epoch": 2.114966389877422,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004555069051097158,
      "loss": 5.0112,
      "mean_token_accuracy": 0.21067973375320434,
      "num_tokens": 49010046.0,
      "step": 21395
    },
    {
      "entropy": 5.651467227935791,
      "epoch": 2.115460656385923,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045548583532378903,
      "loss": 5.0007,
      "mean_token_accuracy": 0.22020939290523528,
      "num_tokens": 49021028.0,
      "step": 21400
    },
    {
      "entropy": 5.703426933288574,
      "epoch": 2.1159549228944248,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004554647610978827,
      "loss": 5.1131,
      "mean_token_accuracy": 0.21216829121112823,
      "num_tokens": 49034522.0,
      "step": 21405
    },
    {
      "entropy": 5.730159282684326,
      "epoch": 2.116449189402926,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004554436824325152,
      "loss": 5.0587,
      "mean_token_accuracy": 0.21673103272914887,
      "num_tokens": 49045034.0,
      "step": 21410
    },
    {
      "entropy": 5.621622896194458,
      "epoch": 2.1169434559114273,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045542259932820515,
      "loss": 5.0185,
      "mean_token_accuracy": 0.21675456464290618,
      "num_tokens": 49056350.0,
      "step": 21415
    },
    {
      "entropy": 5.647717761993408,
      "epoch": 2.117437722419929,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004554015117854712,
      "loss": 5.0369,
      "mean_token_accuracy": 0.21650834232568741,
      "num_tokens": 49067362.0,
      "step": 21420
    },
    {
      "entropy": 5.644987106323242,
      "epoch": 2.1179319889284303,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004553804198048321,
      "loss": 5.0012,
      "mean_token_accuracy": 0.2159202739596367,
      "num_tokens": 49078067.0,
      "step": 21425
    },
    {
      "entropy": 5.595560121536255,
      "epoch": 2.1184262554369315,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045535932338680655,
      "loss": 4.9989,
      "mean_token_accuracy": 0.2183430403470993,
      "num_tokens": 49090205.0,
      "step": 21430
    },
    {
      "entropy": 5.588483667373657,
      "epoch": 2.118920521945433,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004553382225319138,
      "loss": 4.978,
      "mean_token_accuracy": 0.21822975426912308,
      "num_tokens": 49101134.0,
      "step": 21435
    },
    {
      "entropy": 5.68566484451294,
      "epoch": 2.1194147884539345,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004553171172406728,
      "loss": 5.0115,
      "mean_token_accuracy": 0.21091445833444594,
      "num_tokens": 49112282.0,
      "step": 21440
    },
    {
      "entropy": 5.637170743942261,
      "epoch": 2.1199090549624358,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004552960075136027,
      "loss": 4.9647,
      "mean_token_accuracy": 0.21583682149648667,
      "num_tokens": 49123228.0,
      "step": 21445
    },
    {
      "entropy": 5.6031959533691404,
      "epoch": 2.120403321470937,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045527489335122284,
      "loss": 4.9797,
      "mean_token_accuracy": 0.2167018696665764,
      "num_tokens": 49134769.0,
      "step": 21450
    },
    {
      "entropy": 5.559924745559693,
      "epoch": 2.1208975879794387,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045525377475405266,
      "loss": 4.9023,
      "mean_token_accuracy": 0.23155401200056075,
      "num_tokens": 49146096.0,
      "step": 21455
    },
    {
      "entropy": 5.6534138202667235,
      "epoch": 2.12139185448794,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045523265172261155,
      "loss": 5.0058,
      "mean_token_accuracy": 0.21082258373498916,
      "num_tokens": 49156829.0,
      "step": 21460
    },
    {
      "entropy": 5.630233192443848,
      "epoch": 2.1218861209964412,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045521152425741934,
      "loss": 5.0498,
      "mean_token_accuracy": 0.2165581241250038,
      "num_tokens": 49167364.0,
      "step": 21465
    },
    {
      "entropy": 5.696925115585327,
      "epoch": 2.1223803875049425,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045519039235899573,
      "loss": 5.041,
      "mean_token_accuracy": 0.2169329822063446,
      "num_tokens": 49178283.0,
      "step": 21470
    },
    {
      "entropy": 5.768008708953857,
      "epoch": 2.122874654013444,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045516925602786043,
      "loss": 5.0906,
      "mean_token_accuracy": 0.21194446980953216,
      "num_tokens": 49190466.0,
      "step": 21475
    },
    {
      "entropy": 5.638028001785278,
      "epoch": 2.1233689205219455,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004551481152645335,
      "loss": 4.9731,
      "mean_token_accuracy": 0.21756818592548371,
      "num_tokens": 49202601.0,
      "step": 21480
    },
    {
      "entropy": 5.531018924713135,
      "epoch": 2.1238631870304467,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045512697006953504,
      "loss": 4.8465,
      "mean_token_accuracy": 0.22627948820590973,
      "num_tokens": 49214228.0,
      "step": 21485
    },
    {
      "entropy": 5.55597448348999,
      "epoch": 2.124357453538948,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045510582044338513,
      "loss": 4.9667,
      "mean_token_accuracy": 0.21313206255435943,
      "num_tokens": 49225495.0,
      "step": 21490
    },
    {
      "entropy": 5.644659328460693,
      "epoch": 2.1248517200474497,
      "grad_norm": 1.0625,
      "learning_rate": 0.000455084666386604,
      "loss": 5.0046,
      "mean_token_accuracy": 0.21568795293569565,
      "num_tokens": 49237348.0,
      "step": 21495
    },
    {
      "entropy": 5.683156824111938,
      "epoch": 2.125345986555951,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045506350789971227,
      "loss": 5.0465,
      "mean_token_accuracy": 0.20792530477046967,
      "num_tokens": 49247691.0,
      "step": 21500
    },
    {
      "entropy": 5.590653610229492,
      "epoch": 2.1258402530644522,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004550423449832302,
      "loss": 4.9784,
      "mean_token_accuracy": 0.22258145064115525,
      "num_tokens": 49259191.0,
      "step": 21505
    },
    {
      "entropy": 5.666344118118286,
      "epoch": 2.126334519572954,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004550211776376787,
      "loss": 4.9652,
      "mean_token_accuracy": 0.2176661431789398,
      "num_tokens": 49271760.0,
      "step": 21510
    },
    {
      "entropy": 5.670879697799682,
      "epoch": 2.126828786081455,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045500000586357807,
      "loss": 5.0843,
      "mean_token_accuracy": 0.21370499283075334,
      "num_tokens": 49283918.0,
      "step": 21515
    },
    {
      "entropy": 5.639508056640625,
      "epoch": 2.1273230525899565,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045497882966144955,
      "loss": 4.926,
      "mean_token_accuracy": 0.22255134731531143,
      "num_tokens": 49294313.0,
      "step": 21520
    },
    {
      "entropy": 5.571183586120606,
      "epoch": 2.1278173190984577,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004549576490318138,
      "loss": 4.9522,
      "mean_token_accuracy": 0.21616321057081223,
      "num_tokens": 49305213.0,
      "step": 21525
    },
    {
      "entropy": 5.6119873046875,
      "epoch": 2.1283115856069594,
      "grad_norm": 1.078125,
      "learning_rate": 0.000454936463975192,
      "loss": 4.9436,
      "mean_token_accuracy": 0.22240554094314574,
      "num_tokens": 49317245.0,
      "step": 21530
    },
    {
      "entropy": 5.719055843353272,
      "epoch": 2.1288058521154607,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004549152744921053,
      "loss": 5.0733,
      "mean_token_accuracy": 0.20729273706674575,
      "num_tokens": 49328400.0,
      "step": 21535
    },
    {
      "entropy": 5.740843725204468,
      "epoch": 2.129300118623962,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004548940805830749,
      "loss": 5.0843,
      "mean_token_accuracy": 0.20744290500879287,
      "num_tokens": 49341094.0,
      "step": 21540
    },
    {
      "entropy": 5.635720109939575,
      "epoch": 2.129794385132463,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045487288224862224,
      "loss": 4.9791,
      "mean_token_accuracy": 0.21727656275033952,
      "num_tokens": 49352472.0,
      "step": 21545
    },
    {
      "entropy": 5.573558235168457,
      "epoch": 2.130288651640965,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045485167948926876,
      "loss": 4.9664,
      "mean_token_accuracy": 0.21765517443418503,
      "num_tokens": 49362432.0,
      "step": 21550
    },
    {
      "entropy": 5.596670818328858,
      "epoch": 2.130782918149466,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000454830472305536,
      "loss": 4.98,
      "mean_token_accuracy": 0.22480798065662383,
      "num_tokens": 49373857.0,
      "step": 21555
    },
    {
      "entropy": 5.667988204956055,
      "epoch": 2.1312771846579674,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004548092606979458,
      "loss": 5.043,
      "mean_token_accuracy": 0.2161017045378685,
      "num_tokens": 49387545.0,
      "step": 21560
    },
    {
      "entropy": 5.662838745117187,
      "epoch": 2.131771451166469,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045478804466701984,
      "loss": 5.0131,
      "mean_token_accuracy": 0.21660160422325134,
      "num_tokens": 49398273.0,
      "step": 21565
    },
    {
      "entropy": 5.650055360794068,
      "epoch": 2.1322657176749704,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045476682421328015,
      "loss": 4.99,
      "mean_token_accuracy": 0.21125394552946092,
      "num_tokens": 49408662.0,
      "step": 21570
    },
    {
      "entropy": 5.668147087097168,
      "epoch": 2.1327599841834717,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045474559933724866,
      "loss": 5.0997,
      "mean_token_accuracy": 0.21337754875421525,
      "num_tokens": 49421189.0,
      "step": 21575
    },
    {
      "entropy": 5.687134552001953,
      "epoch": 2.133254250691973,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045472437003944753,
      "loss": 5.0561,
      "mean_token_accuracy": 0.2158006325364113,
      "num_tokens": 49432878.0,
      "step": 21580
    },
    {
      "entropy": 5.68028712272644,
      "epoch": 2.1337485172004746,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004547031363203991,
      "loss": 5.0673,
      "mean_token_accuracy": 0.21448256969451904,
      "num_tokens": 49444058.0,
      "step": 21585
    },
    {
      "entropy": 5.646247100830078,
      "epoch": 2.134242783708976,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045468189818062554,
      "loss": 5.0012,
      "mean_token_accuracy": 0.20970974415540694,
      "num_tokens": 49456602.0,
      "step": 21590
    },
    {
      "entropy": 5.670156526565552,
      "epoch": 2.134737050217477,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004546606556206494,
      "loss": 5.0061,
      "mean_token_accuracy": 0.21191326081752776,
      "num_tokens": 49467287.0,
      "step": 21595
    },
    {
      "entropy": 5.641109991073608,
      "epoch": 2.135231316725979,
      "grad_norm": 1.125,
      "learning_rate": 0.0004546394086409933,
      "loss": 4.956,
      "mean_token_accuracy": 0.21748731136322022,
      "num_tokens": 49478662.0,
      "step": 21600
    },
    {
      "entropy": 5.627913570404052,
      "epoch": 2.13572558323448,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045461815724217984,
      "loss": 5.0391,
      "mean_token_accuracy": 0.2139389246702194,
      "num_tokens": 49489365.0,
      "step": 21605
    },
    {
      "entropy": 5.689600324630737,
      "epoch": 2.1362198497429814,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045459690142473193,
      "loss": 5.0673,
      "mean_token_accuracy": 0.20628269016742706,
      "num_tokens": 49501318.0,
      "step": 21610
    },
    {
      "entropy": 5.660380744934082,
      "epoch": 2.1367141162514827,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004545756411891723,
      "loss": 5.0481,
      "mean_token_accuracy": 0.20921465903520584,
      "num_tokens": 49512138.0,
      "step": 21615
    },
    {
      "entropy": 5.689916038513184,
      "epoch": 2.1372083827599844,
      "grad_norm": 1.125,
      "learning_rate": 0.0004545543765360241,
      "loss": 4.9918,
      "mean_token_accuracy": 0.21490930914878845,
      "num_tokens": 49523325.0,
      "step": 21620
    },
    {
      "entropy": 5.611585664749145,
      "epoch": 2.1377026492684856,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045453310746581036,
      "loss": 4.9349,
      "mean_token_accuracy": 0.2220182940363884,
      "num_tokens": 49535382.0,
      "step": 21625
    },
    {
      "entropy": 5.604008436203003,
      "epoch": 2.138196915776987,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004545118339790543,
      "loss": 4.9168,
      "mean_token_accuracy": 0.22366950511932374,
      "num_tokens": 49546151.0,
      "step": 21630
    },
    {
      "entropy": 5.614528560638428,
      "epoch": 2.138691182285488,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004544905560762793,
      "loss": 4.9959,
      "mean_token_accuracy": 0.220633564889431,
      "num_tokens": 49558617.0,
      "step": 21635
    },
    {
      "entropy": 5.63063497543335,
      "epoch": 2.13918544879399,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045446927375800876,
      "loss": 4.9419,
      "mean_token_accuracy": 0.2289588764309883,
      "num_tokens": 49568494.0,
      "step": 21640
    },
    {
      "entropy": 5.635165452957153,
      "epoch": 2.139679715302491,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045444798702476625,
      "loss": 4.9937,
      "mean_token_accuracy": 0.22235831916332244,
      "num_tokens": 49579763.0,
      "step": 21645
    },
    {
      "entropy": 5.597065401077271,
      "epoch": 2.1401739818109924,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004544266958770755,
      "loss": 4.9354,
      "mean_token_accuracy": 0.21802193820476531,
      "num_tokens": 49591631.0,
      "step": 21650
    },
    {
      "entropy": 5.650187253952026,
      "epoch": 2.1406682483194936,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045440540031546004,
      "loss": 5.0101,
      "mean_token_accuracy": 0.21588210612535477,
      "num_tokens": 49603458.0,
      "step": 21655
    },
    {
      "entropy": 5.628285360336304,
      "epoch": 2.1411625148279954,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045438410034044397,
      "loss": 4.9664,
      "mean_token_accuracy": 0.22002998888492584,
      "num_tokens": 49614076.0,
      "step": 21660
    },
    {
      "entropy": 5.652142477035523,
      "epoch": 2.1416567813364966,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045436279595255125,
      "loss": 5.0768,
      "mean_token_accuracy": 0.21290842443704605,
      "num_tokens": 49624982.0,
      "step": 21665
    },
    {
      "entropy": 5.630671644210816,
      "epoch": 2.142151047844998,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004543414871523059,
      "loss": 4.9537,
      "mean_token_accuracy": 0.21931060701608657,
      "num_tokens": 49635897.0,
      "step": 21670
    },
    {
      "entropy": 5.628613138198853,
      "epoch": 2.1426453143534996,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004543201739402321,
      "loss": 4.9686,
      "mean_token_accuracy": 0.21775830537080765,
      "num_tokens": 49647023.0,
      "step": 21675
    },
    {
      "entropy": 5.638038444519043,
      "epoch": 2.143139580862001,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045429885631685424,
      "loss": 5.0576,
      "mean_token_accuracy": 0.20432451069355012,
      "num_tokens": 49658578.0,
      "step": 21680
    },
    {
      "entropy": 5.653888893127442,
      "epoch": 2.143633847370502,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004542775342826967,
      "loss": 4.9749,
      "mean_token_accuracy": 0.21984683275222777,
      "num_tokens": 49670978.0,
      "step": 21685
    },
    {
      "entropy": 5.578224086761475,
      "epoch": 2.1441281138790034,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045425620783828396,
      "loss": 4.9981,
      "mean_token_accuracy": 0.21594302207231522,
      "num_tokens": 49682553.0,
      "step": 21690
    },
    {
      "entropy": 5.64458818435669,
      "epoch": 2.144622380387505,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004542348769841407,
      "loss": 4.9842,
      "mean_token_accuracy": 0.22028686702251435,
      "num_tokens": 49694589.0,
      "step": 21695
    },
    {
      "entropy": 5.650266981124878,
      "epoch": 2.1451166468960063,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045421354172079173,
      "loss": 5.0279,
      "mean_token_accuracy": 0.21618122905492781,
      "num_tokens": 49706530.0,
      "step": 21700
    },
    {
      "entropy": 5.607585000991821,
      "epoch": 2.1456109134045076,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045419220204876175,
      "loss": 5.0033,
      "mean_token_accuracy": 0.21740746796131133,
      "num_tokens": 49718524.0,
      "step": 21705
    },
    {
      "entropy": 5.639693546295166,
      "epoch": 2.1461051799130093,
      "grad_norm": 1.125,
      "learning_rate": 0.00045417085796857596,
      "loss": 5.0068,
      "mean_token_accuracy": 0.2152506470680237,
      "num_tokens": 49729555.0,
      "step": 21710
    },
    {
      "entropy": 5.6691991806030275,
      "epoch": 2.1465994464215106,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045414950948075906,
      "loss": 5.0494,
      "mean_token_accuracy": 0.21223363876342774,
      "num_tokens": 49739708.0,
      "step": 21715
    },
    {
      "entropy": 5.681228256225586,
      "epoch": 2.147093712930012,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045412815658583657,
      "loss": 5.0476,
      "mean_token_accuracy": 0.21409208327531815,
      "num_tokens": 49750997.0,
      "step": 21720
    },
    {
      "entropy": 5.602482318878174,
      "epoch": 2.147587979438513,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004541067992843336,
      "loss": 4.9879,
      "mean_token_accuracy": 0.22401541471481323,
      "num_tokens": 49761508.0,
      "step": 21725
    },
    {
      "entropy": 5.644294548034668,
      "epoch": 2.148082245947015,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045408543757677555,
      "loss": 4.9027,
      "mean_token_accuracy": 0.22139550894498825,
      "num_tokens": 49772821.0,
      "step": 21730
    },
    {
      "entropy": 5.556284666061401,
      "epoch": 2.148576512455516,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000454064071463688,
      "loss": 4.9096,
      "mean_token_accuracy": 0.2241299644112587,
      "num_tokens": 49783932.0,
      "step": 21735
    },
    {
      "entropy": 5.6272820949554445,
      "epoch": 2.1490707789640173,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004540427009455965,
      "loss": 4.9905,
      "mean_token_accuracy": 0.21326961368322372,
      "num_tokens": 49795126.0,
      "step": 21740
    },
    {
      "entropy": 5.627553033828735,
      "epoch": 2.1495650454725186,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004540213260230267,
      "loss": 4.9976,
      "mean_token_accuracy": 0.21134135872125626,
      "num_tokens": 49806110.0,
      "step": 21745
    },
    {
      "entropy": 5.71685562133789,
      "epoch": 2.1500593119810203,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004539999466965046,
      "loss": 5.1265,
      "mean_token_accuracy": 0.20695742219686508,
      "num_tokens": 49818879.0,
      "step": 21750
    },
    {
      "entropy": 5.661255693435669,
      "epoch": 2.1505535784895216,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045397856296655603,
      "loss": 4.9667,
      "mean_token_accuracy": 0.21782406121492387,
      "num_tokens": 49828986.0,
      "step": 21755
    },
    {
      "entropy": 5.647045564651489,
      "epoch": 2.151047844998023,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045395717483370703,
      "loss": 5.032,
      "mean_token_accuracy": 0.21559846997261048,
      "num_tokens": 49841987.0,
      "step": 21760
    },
    {
      "entropy": 5.610520267486573,
      "epoch": 2.1515421115065245,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045393578229848373,
      "loss": 5.0292,
      "mean_token_accuracy": 0.21143033355474472,
      "num_tokens": 49854151.0,
      "step": 21765
    },
    {
      "entropy": 5.660473823547363,
      "epoch": 2.152036378015026,
      "grad_norm": 1.125,
      "learning_rate": 0.00045391438536141257,
      "loss": 5.0439,
      "mean_token_accuracy": 0.21628856807947158,
      "num_tokens": 49866531.0,
      "step": 21770
    },
    {
      "entropy": 5.6451061248779295,
      "epoch": 2.152530644523527,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045389298402301965,
      "loss": 4.9544,
      "mean_token_accuracy": 0.22130872011184693,
      "num_tokens": 49877737.0,
      "step": 21775
    },
    {
      "entropy": 5.731892061233521,
      "epoch": 2.1530249110320283,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004538715782838316,
      "loss": 5.0497,
      "mean_token_accuracy": 0.20574968457221984,
      "num_tokens": 49889214.0,
      "step": 21780
    },
    {
      "entropy": 5.6572154521942135,
      "epoch": 2.15351917754053,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045385016814437494,
      "loss": 5.0213,
      "mean_token_accuracy": 0.2188189819455147,
      "num_tokens": 49900293.0,
      "step": 21785
    },
    {
      "entropy": 5.709380769729615,
      "epoch": 2.1540134440490313,
      "grad_norm": 1.0,
      "learning_rate": 0.00045382875360517645,
      "loss": 5.0974,
      "mean_token_accuracy": 0.2023204281926155,
      "num_tokens": 49912193.0,
      "step": 21790
    },
    {
      "entropy": 5.705944442749024,
      "epoch": 2.1545077105575325,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004538073346667628,
      "loss": 4.9875,
      "mean_token_accuracy": 0.22009722292423248,
      "num_tokens": 49922780.0,
      "step": 21795
    },
    {
      "entropy": 5.597604417800904,
      "epoch": 2.155001977066034,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045378591132966105,
      "loss": 4.958,
      "mean_token_accuracy": 0.21570774167776108,
      "num_tokens": 49934551.0,
      "step": 21800
    },
    {
      "entropy": 5.6567998886108395,
      "epoch": 2.1554962435745355,
      "grad_norm": 1.0,
      "learning_rate": 0.0004537644835943982,
      "loss": 5.0059,
      "mean_token_accuracy": 0.21579796373844146,
      "num_tokens": 49945779.0,
      "step": 21805
    },
    {
      "entropy": 5.584072160720825,
      "epoch": 2.1559905100830368,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004537430514615012,
      "loss": 4.8754,
      "mean_token_accuracy": 0.22672225683927535,
      "num_tokens": 49957127.0,
      "step": 21810
    },
    {
      "entropy": 5.611971473693847,
      "epoch": 2.156484776591538,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045372161493149746,
      "loss": 4.969,
      "mean_token_accuracy": 0.21689440160989762,
      "num_tokens": 49968679.0,
      "step": 21815
    },
    {
      "entropy": 5.627978277206421,
      "epoch": 2.1569790431000397,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004537001740049142,
      "loss": 4.9784,
      "mean_token_accuracy": 0.21968361586332322,
      "num_tokens": 49979924.0,
      "step": 21820
    },
    {
      "entropy": 5.6861762523651125,
      "epoch": 2.157473309608541,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000453678728682279,
      "loss": 5.0066,
      "mean_token_accuracy": 0.21664040088653563,
      "num_tokens": 49991318.0,
      "step": 21825
    },
    {
      "entropy": 5.4896852493286135,
      "epoch": 2.1579675761170423,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004536572789641193,
      "loss": 4.9092,
      "mean_token_accuracy": 0.22200099974870682,
      "num_tokens": 50003096.0,
      "step": 21830
    },
    {
      "entropy": 5.569819021224975,
      "epoch": 2.1584618426255435,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045363582485096295,
      "loss": 4.9627,
      "mean_token_accuracy": 0.21801090687513353,
      "num_tokens": 50013953.0,
      "step": 21835
    },
    {
      "entropy": 5.646777963638305,
      "epoch": 2.1589561091340452,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004536143663433375,
      "loss": 4.973,
      "mean_token_accuracy": 0.22065008133649827,
      "num_tokens": 50025696.0,
      "step": 21840
    },
    {
      "entropy": 5.666622304916382,
      "epoch": 2.1594503756425465,
      "grad_norm": 1.125,
      "learning_rate": 0.00045359290344177094,
      "loss": 4.9782,
      "mean_token_accuracy": 0.21890347748994826,
      "num_tokens": 50036880.0,
      "step": 21845
    },
    {
      "entropy": 5.624815130233765,
      "epoch": 2.1599446421510478,
      "grad_norm": 1.125,
      "learning_rate": 0.0004535714361467913,
      "loss": 5.0353,
      "mean_token_accuracy": 0.21481664776802062,
      "num_tokens": 50048746.0,
      "step": 21850
    },
    {
      "entropy": 5.588080024719238,
      "epoch": 2.1604389086595495,
      "grad_norm": 1.125,
      "learning_rate": 0.0004535499644589266,
      "loss": 4.9632,
      "mean_token_accuracy": 0.22443701028823854,
      "num_tokens": 50060711.0,
      "step": 21855
    },
    {
      "entropy": 5.652092313766479,
      "epoch": 2.1609331751680507,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045352848837870496,
      "loss": 4.9539,
      "mean_token_accuracy": 0.21888331025838853,
      "num_tokens": 50071878.0,
      "step": 21860
    },
    {
      "entropy": 5.694894409179687,
      "epoch": 2.161427441676552,
      "grad_norm": 1.046875,
      "learning_rate": 0.000453507007906655,
      "loss": 5.022,
      "mean_token_accuracy": 0.21179139465093613,
      "num_tokens": 50084600.0,
      "step": 21865
    },
    {
      "entropy": 5.5414904117584225,
      "epoch": 2.1619217081850532,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004534855230433048,
      "loss": 4.8916,
      "mean_token_accuracy": 0.22710329592227935,
      "num_tokens": 50095814.0,
      "step": 21870
    },
    {
      "entropy": 5.599045896530152,
      "epoch": 2.162415974693555,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045346403378918314,
      "loss": 5.0222,
      "mean_token_accuracy": 0.21904779374599456,
      "num_tokens": 50108390.0,
      "step": 21875
    },
    {
      "entropy": 5.631582307815552,
      "epoch": 2.162910241202056,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004534425401448185,
      "loss": 5.0602,
      "mean_token_accuracy": 0.21365441679954528,
      "num_tokens": 50118939.0,
      "step": 21880
    },
    {
      "entropy": 5.595835494995117,
      "epoch": 2.1634045077105575,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004534210421107397,
      "loss": 4.9555,
      "mean_token_accuracy": 0.22360922396183014,
      "num_tokens": 50129510.0,
      "step": 21885
    },
    {
      "entropy": 5.595914840698242,
      "epoch": 2.1638987742190587,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004533995396874756,
      "loss": 4.9831,
      "mean_token_accuracy": 0.22330474853515625,
      "num_tokens": 50140611.0,
      "step": 21890
    },
    {
      "entropy": 5.641322612762451,
      "epoch": 2.1643930407275604,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045337803287555525,
      "loss": 5.0296,
      "mean_token_accuracy": 0.21516034752130508,
      "num_tokens": 50152214.0,
      "step": 21895
    },
    {
      "entropy": 5.74531397819519,
      "epoch": 2.1648873072360617,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004533565216755075,
      "loss": 5.1367,
      "mean_token_accuracy": 0.20491723269224166,
      "num_tokens": 50164201.0,
      "step": 21900
    },
    {
      "entropy": 5.626381874084473,
      "epoch": 2.165381573744563,
      "grad_norm": 1.125,
      "learning_rate": 0.0004533350060878615,
      "loss": 4.9743,
      "mean_token_accuracy": 0.22156866043806075,
      "num_tokens": 50176515.0,
      "step": 21905
    },
    {
      "entropy": 5.664905643463134,
      "epoch": 2.1658758402530642,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004533134861131469,
      "loss": 5.0193,
      "mean_token_accuracy": 0.21336598098278045,
      "num_tokens": 50187624.0,
      "step": 21910
    },
    {
      "entropy": 5.517172145843506,
      "epoch": 2.166370106761566,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004532919617518929,
      "loss": 4.8562,
      "mean_token_accuracy": 0.22781997472047805,
      "num_tokens": 50199549.0,
      "step": 21915
    },
    {
      "entropy": 5.676529216766357,
      "epoch": 2.166864373270067,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045327043300462887,
      "loss": 5.044,
      "mean_token_accuracy": 0.21444866359233855,
      "num_tokens": 50210920.0,
      "step": 21920
    },
    {
      "entropy": 5.626539039611816,
      "epoch": 2.1673586397785685,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004532488998718846,
      "loss": 4.9536,
      "mean_token_accuracy": 0.2275352358818054,
      "num_tokens": 50221335.0,
      "step": 21925
    },
    {
      "entropy": 5.608596229553223,
      "epoch": 2.16785290628707,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045322736235418973,
      "loss": 5.0392,
      "mean_token_accuracy": 0.21377147734165192,
      "num_tokens": 50235138.0,
      "step": 21930
    },
    {
      "entropy": 5.648913717269897,
      "epoch": 2.1683471727955714,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045320582045207405,
      "loss": 4.9659,
      "mean_token_accuracy": 0.22183896601200104,
      "num_tokens": 50246454.0,
      "step": 21935
    },
    {
      "entropy": 5.683821344375611,
      "epoch": 2.1688414393040727,
      "grad_norm": 1.125,
      "learning_rate": 0.0004531842741660675,
      "loss": 5.0109,
      "mean_token_accuracy": 0.2130260467529297,
      "num_tokens": 50257591.0,
      "step": 21940
    },
    {
      "entropy": 5.660471963882446,
      "epoch": 2.169335705812574,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004531627234967003,
      "loss": 4.9927,
      "mean_token_accuracy": 0.21873451322317122,
      "num_tokens": 50268557.0,
      "step": 21945
    },
    {
      "entropy": 5.691157293319702,
      "epoch": 2.1698299723210757,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004531411684445024,
      "loss": 5.1197,
      "mean_token_accuracy": 0.20806851536035537,
      "num_tokens": 50280617.0,
      "step": 21950
    },
    {
      "entropy": 5.603302145004273,
      "epoch": 2.170324238829577,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045311960901000403,
      "loss": 4.9579,
      "mean_token_accuracy": 0.21671452820301057,
      "num_tokens": 50292071.0,
      "step": 21955
    },
    {
      "entropy": 5.634258031845093,
      "epoch": 2.170818505338078,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004530980451937358,
      "loss": 4.9869,
      "mean_token_accuracy": 0.21541567295789718,
      "num_tokens": 50303389.0,
      "step": 21960
    },
    {
      "entropy": 5.635688400268554,
      "epoch": 2.17131277184658,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004530764769962279,
      "loss": 5.05,
      "mean_token_accuracy": 0.22100289762020112,
      "num_tokens": 50315088.0,
      "step": 21965
    },
    {
      "entropy": 5.651118183135987,
      "epoch": 2.171807038355081,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004530549044180111,
      "loss": 4.9636,
      "mean_token_accuracy": 0.220535808801651,
      "num_tokens": 50326177.0,
      "step": 21970
    },
    {
      "entropy": 5.600054979324341,
      "epoch": 2.1723013048635824,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045303332745961605,
      "loss": 4.9686,
      "mean_token_accuracy": 0.22103646248579026,
      "num_tokens": 50337333.0,
      "step": 21975
    },
    {
      "entropy": 5.671171712875366,
      "epoch": 2.1727955713720837,
      "grad_norm": 1.125,
      "learning_rate": 0.00045301174612157345,
      "loss": 4.998,
      "mean_token_accuracy": 0.21949263662099838,
      "num_tokens": 50348406.0,
      "step": 21980
    },
    {
      "entropy": 5.616403341293335,
      "epoch": 2.1732898378805854,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004529901604044143,
      "loss": 4.9361,
      "mean_token_accuracy": 0.22730042040348053,
      "num_tokens": 50360057.0,
      "step": 21985
    },
    {
      "entropy": 5.671643400192261,
      "epoch": 2.1737841043890866,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004529685703086697,
      "loss": 4.9916,
      "mean_token_accuracy": 0.21968017667531967,
      "num_tokens": 50370355.0,
      "step": 21990
    },
    {
      "entropy": 5.643956518173217,
      "epoch": 2.174278370897588,
      "grad_norm": 1.0,
      "learning_rate": 0.00045294697583487056,
      "loss": 5.0082,
      "mean_token_accuracy": 0.21507970839738846,
      "num_tokens": 50382193.0,
      "step": 21995
    },
    {
      "entropy": 5.644825220108032,
      "epoch": 2.174772637406089,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045292537698354824,
      "loss": 4.9657,
      "mean_token_accuracy": 0.215004625916481,
      "num_tokens": 50394383.0,
      "step": 22000
    },
    {
      "entropy": 5.612366771697998,
      "epoch": 2.175266903914591,
      "grad_norm": 1.09375,
      "learning_rate": 0.000452903773755234,
      "loss": 4.9316,
      "mean_token_accuracy": 0.22849385142326356,
      "num_tokens": 50405006.0,
      "step": 22005
    },
    {
      "entropy": 5.679098176956177,
      "epoch": 2.175761170423092,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004528821661504594,
      "loss": 5.0745,
      "mean_token_accuracy": 0.21529259234666825,
      "num_tokens": 50416940.0,
      "step": 22010
    },
    {
      "entropy": 5.637016820907593,
      "epoch": 2.1762554369315934,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004528605541697558,
      "loss": 5.0043,
      "mean_token_accuracy": 0.2183789148926735,
      "num_tokens": 50428506.0,
      "step": 22015
    },
    {
      "entropy": 5.6875998973846436,
      "epoch": 2.176749703440095,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045283893781365513,
      "loss": 5.0433,
      "mean_token_accuracy": 0.2110864281654358,
      "num_tokens": 50440421.0,
      "step": 22020
    },
    {
      "entropy": 5.668367862701416,
      "epoch": 2.1772439699485964,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004528173170826889,
      "loss": 5.0475,
      "mean_token_accuracy": 0.21114738285541534,
      "num_tokens": 50451819.0,
      "step": 22025
    },
    {
      "entropy": 5.775361347198486,
      "epoch": 2.1777382364570976,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004527956919773891,
      "loss": 5.1532,
      "mean_token_accuracy": 0.20011237263679504,
      "num_tokens": 50464382.0,
      "step": 22030
    },
    {
      "entropy": 5.707822370529175,
      "epoch": 2.178232502965599,
      "grad_norm": 1.125,
      "learning_rate": 0.0004527740624982877,
      "loss": 5.0219,
      "mean_token_accuracy": 0.21936631798744202,
      "num_tokens": 50474773.0,
      "step": 22035
    },
    {
      "entropy": 5.619495916366577,
      "epoch": 2.1787267694741006,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045275242864591683,
      "loss": 4.9489,
      "mean_token_accuracy": 0.22284717559814454,
      "num_tokens": 50485626.0,
      "step": 22040
    },
    {
      "entropy": 5.630292558670044,
      "epoch": 2.179221035982602,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045273079042080854,
      "loss": 4.9471,
      "mean_token_accuracy": 0.21731263399124146,
      "num_tokens": 50496130.0,
      "step": 22045
    },
    {
      "entropy": 5.629637765884399,
      "epoch": 2.179715302491103,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045270914782349534,
      "loss": 5.0385,
      "mean_token_accuracy": 0.2181318446993828,
      "num_tokens": 50506578.0,
      "step": 22050
    },
    {
      "entropy": 5.667500448226929,
      "epoch": 2.1802095689996044,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045268750085450946,
      "loss": 5.1049,
      "mean_token_accuracy": 0.20530329048633575,
      "num_tokens": 50517398.0,
      "step": 22055
    },
    {
      "entropy": 5.594801425933838,
      "epoch": 2.180703835508106,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004526658495143835,
      "loss": 4.9024,
      "mean_token_accuracy": 0.2297630116343498,
      "num_tokens": 50528577.0,
      "step": 22060
    },
    {
      "entropy": 5.636479663848877,
      "epoch": 2.1811981020166074,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004526441938036501,
      "loss": 4.9684,
      "mean_token_accuracy": 0.22255425304174423,
      "num_tokens": 50539118.0,
      "step": 22065
    },
    {
      "entropy": 5.655508041381836,
      "epoch": 2.1816923685251086,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045262253372284193,
      "loss": 5.0353,
      "mean_token_accuracy": 0.21686965376138687,
      "num_tokens": 50549934.0,
      "step": 22070
    },
    {
      "entropy": 5.637511587142944,
      "epoch": 2.1821866350336103,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045260086927249193,
      "loss": 4.9842,
      "mean_token_accuracy": 0.21996564269065857,
      "num_tokens": 50560661.0,
      "step": 22075
    },
    {
      "entropy": 5.648717403411865,
      "epoch": 2.1826809015421116,
      "grad_norm": 1.046875,
      "learning_rate": 0.000452579200453133,
      "loss": 5.0454,
      "mean_token_accuracy": 0.2059583455324173,
      "num_tokens": 50573304.0,
      "step": 22080
    },
    {
      "entropy": 5.677945375442505,
      "epoch": 2.183175168050613,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004525575272652981,
      "loss": 5.0395,
      "mean_token_accuracy": 0.20937439054250717,
      "num_tokens": 50584166.0,
      "step": 22085
    },
    {
      "entropy": 5.724975967407227,
      "epoch": 2.183669434559114,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004525358497095206,
      "loss": 5.0181,
      "mean_token_accuracy": 0.20789888352155686,
      "num_tokens": 50596425.0,
      "step": 22090
    },
    {
      "entropy": 5.649881219863891,
      "epoch": 2.184163701067616,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045251416778633354,
      "loss": 4.9518,
      "mean_token_accuracy": 0.22508205324411393,
      "num_tokens": 50607483.0,
      "step": 22095
    },
    {
      "entropy": 5.628113698959351,
      "epoch": 2.184657967576117,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045249248149627046,
      "loss": 4.9361,
      "mean_token_accuracy": 0.22333286255598067,
      "num_tokens": 50619655.0,
      "step": 22100
    },
    {
      "entropy": 5.662592887878418,
      "epoch": 2.1851522340846183,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045247079083986485,
      "loss": 5.0399,
      "mean_token_accuracy": 0.20970261842012405,
      "num_tokens": 50631323.0,
      "step": 22105
    },
    {
      "entropy": 5.607187652587891,
      "epoch": 2.18564650059312,
      "grad_norm": 1.125,
      "learning_rate": 0.0004524490958176502,
      "loss": 4.941,
      "mean_token_accuracy": 0.22389119267463684,
      "num_tokens": 50642931.0,
      "step": 22110
    },
    {
      "entropy": 5.558671379089356,
      "epoch": 2.1861407671016213,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004524273964301603,
      "loss": 4.9033,
      "mean_token_accuracy": 0.22713505625724792,
      "num_tokens": 50655396.0,
      "step": 22115
    },
    {
      "entropy": 5.611291694641113,
      "epoch": 2.1866350336101226,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045240569267792887,
      "loss": 5.0253,
      "mean_token_accuracy": 0.2204742968082428,
      "num_tokens": 50667264.0,
      "step": 22120
    },
    {
      "entropy": 5.571619606018066,
      "epoch": 2.187129300118624,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045238398456148995,
      "loss": 4.926,
      "mean_token_accuracy": 0.22592018246650697,
      "num_tokens": 50679073.0,
      "step": 22125
    },
    {
      "entropy": 5.659052467346191,
      "epoch": 2.1876235666271255,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045236227208137744,
      "loss": 5.0146,
      "mean_token_accuracy": 0.21489075124263762,
      "num_tokens": 50689554.0,
      "step": 22130
    },
    {
      "entropy": 5.719157457351685,
      "epoch": 2.188117833135627,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004523405552381256,
      "loss": 5.0364,
      "mean_token_accuracy": 0.2110719934105873,
      "num_tokens": 50702419.0,
      "step": 22135
    },
    {
      "entropy": 5.638813066482544,
      "epoch": 2.188612099644128,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004523188340322685,
      "loss": 4.9633,
      "mean_token_accuracy": 0.21964765042066575,
      "num_tokens": 50713016.0,
      "step": 22140
    },
    {
      "entropy": 5.672059011459351,
      "epoch": 2.1891063661526293,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004522971084643406,
      "loss": 5.0405,
      "mean_token_accuracy": 0.2096104294061661,
      "num_tokens": 50723936.0,
      "step": 22145
    },
    {
      "entropy": 5.672735214233398,
      "epoch": 2.189600632661131,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045227537853487627,
      "loss": 4.9725,
      "mean_token_accuracy": 0.2199035704135895,
      "num_tokens": 50734339.0,
      "step": 22150
    },
    {
      "entropy": 5.710371828079223,
      "epoch": 2.1900948991696323,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045225364424441016,
      "loss": 5.08,
      "mean_token_accuracy": 0.2040013149380684,
      "num_tokens": 50746262.0,
      "step": 22155
    },
    {
      "entropy": 5.675328302383423,
      "epoch": 2.1905891656781336,
      "grad_norm": 0.984375,
      "learning_rate": 0.000452231905593477,
      "loss": 4.9785,
      "mean_token_accuracy": 0.2201933890581131,
      "num_tokens": 50758218.0,
      "step": 22160
    },
    {
      "entropy": 5.737107276916504,
      "epoch": 2.191083432186635,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004522101625826113,
      "loss": 5.0744,
      "mean_token_accuracy": 0.20993250906467437,
      "num_tokens": 50769263.0,
      "step": 22165
    },
    {
      "entropy": 5.564160299301148,
      "epoch": 2.1915776986951365,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004521884152123482,
      "loss": 4.8647,
      "mean_token_accuracy": 0.23291980475187302,
      "num_tokens": 50780049.0,
      "step": 22170
    },
    {
      "entropy": 5.601339101791382,
      "epoch": 2.192071965203638,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004521666634832227,
      "loss": 5.003,
      "mean_token_accuracy": 0.21563053429126738,
      "num_tokens": 50791817.0,
      "step": 22175
    },
    {
      "entropy": 5.619451665878296,
      "epoch": 2.192566231712139,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004521449073957697,
      "loss": 4.993,
      "mean_token_accuracy": 0.21952337473630906,
      "num_tokens": 50802669.0,
      "step": 22180
    },
    {
      "entropy": 5.675236701965332,
      "epoch": 2.1930604982206408,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045212314695052447,
      "loss": 5.0103,
      "mean_token_accuracy": 0.2183389037847519,
      "num_tokens": 50813920.0,
      "step": 22185
    },
    {
      "entropy": 5.692987298965454,
      "epoch": 2.193554764729142,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045210138214802227,
      "loss": 5.058,
      "mean_token_accuracy": 0.21045937687158583,
      "num_tokens": 50825828.0,
      "step": 22190
    },
    {
      "entropy": 5.672288417816162,
      "epoch": 2.1940490312376433,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045207961298879875,
      "loss": 5.0624,
      "mean_token_accuracy": 0.20568755269050598,
      "num_tokens": 50837425.0,
      "step": 22195
    },
    {
      "entropy": 5.681703996658325,
      "epoch": 2.1945432977461445,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004520578394733892,
      "loss": 5.0068,
      "mean_token_accuracy": 0.22216037660837173,
      "num_tokens": 50848720.0,
      "step": 22200
    },
    {
      "entropy": 5.595181703567505,
      "epoch": 2.1950375642546462,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045203606160232936,
      "loss": 4.9615,
      "mean_token_accuracy": 0.21896006017923356,
      "num_tokens": 50860304.0,
      "step": 22205
    },
    {
      "entropy": 5.657786798477173,
      "epoch": 2.1955318307631475,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004520142793761549,
      "loss": 5.0203,
      "mean_token_accuracy": 0.20914703905582427,
      "num_tokens": 50872099.0,
      "step": 22210
    },
    {
      "entropy": 5.765751028060913,
      "epoch": 2.1960260972716488,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004519924927954017,
      "loss": 5.1041,
      "mean_token_accuracy": 0.2108405977487564,
      "num_tokens": 50884116.0,
      "step": 22215
    },
    {
      "entropy": 5.646667432785034,
      "epoch": 2.1965203637801505,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004519707018606059,
      "loss": 4.9214,
      "mean_token_accuracy": 0.22320738732814788,
      "num_tokens": 50896293.0,
      "step": 22220
    },
    {
      "entropy": 5.611848640441894,
      "epoch": 2.1970146302886517,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045194890657230315,
      "loss": 4.9138,
      "mean_token_accuracy": 0.22903822958469391,
      "num_tokens": 50908633.0,
      "step": 22225
    },
    {
      "entropy": 5.630171203613282,
      "epoch": 2.197508896797153,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045192710693103,
      "loss": 4.9479,
      "mean_token_accuracy": 0.22072465270757674,
      "num_tokens": 50919524.0,
      "step": 22230
    },
    {
      "entropy": 5.640546226501465,
      "epoch": 2.1980031633056543,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004519053029373224,
      "loss": 5.0196,
      "mean_token_accuracy": 0.21351983845233918,
      "num_tokens": 50930157.0,
      "step": 22235
    },
    {
      "entropy": 5.726377964019775,
      "epoch": 2.198497429814156,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000451883494591717,
      "loss": 5.0899,
      "mean_token_accuracy": 0.21191806495189666,
      "num_tokens": 50942474.0,
      "step": 22240
    },
    {
      "entropy": 5.709763622283935,
      "epoch": 2.1989916963226572,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045186168189475026,
      "loss": 5.0668,
      "mean_token_accuracy": 0.20783522427082063,
      "num_tokens": 50954847.0,
      "step": 22245
    },
    {
      "entropy": 5.638669538497925,
      "epoch": 2.1994859628311585,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004518398648469586,
      "loss": 5.0494,
      "mean_token_accuracy": 0.21514272540807725,
      "num_tokens": 50966446.0,
      "step": 22250
    },
    {
      "entropy": 5.584299850463867,
      "epoch": 2.1999802293396598,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045181804344887887,
      "loss": 4.993,
      "mean_token_accuracy": 0.22143091708421708,
      "num_tokens": 50977949.0,
      "step": 22255
    },
    {
      "entropy": 5.681093311309814,
      "epoch": 2.2004744958481615,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045179621770104786,
      "loss": 5.0329,
      "mean_token_accuracy": 0.21312574595212935,
      "num_tokens": 50989297.0,
      "step": 22260
    },
    {
      "entropy": 5.6770096778869625,
      "epoch": 2.2009687623566627,
      "grad_norm": 1.0,
      "learning_rate": 0.00045177438760400244,
      "loss": 5.0522,
      "mean_token_accuracy": 0.21251516938209533,
      "num_tokens": 51001616.0,
      "step": 22265
    },
    {
      "entropy": 5.680920505523682,
      "epoch": 2.201463028865164,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004517525531582797,
      "loss": 5.0431,
      "mean_token_accuracy": 0.2150268405675888,
      "num_tokens": 51012870.0,
      "step": 22270
    },
    {
      "entropy": 5.623628616333008,
      "epoch": 2.2019572953736652,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045173071436441665,
      "loss": 4.9845,
      "mean_token_accuracy": 0.22150683552026748,
      "num_tokens": 51025368.0,
      "step": 22275
    },
    {
      "entropy": 5.657191133499145,
      "epoch": 2.202451561882167,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004517088712229507,
      "loss": 4.9628,
      "mean_token_accuracy": 0.218977253139019,
      "num_tokens": 51035964.0,
      "step": 22280
    },
    {
      "entropy": 5.6696672439575195,
      "epoch": 2.202945828390668,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045168702373441904,
      "loss": 4.9824,
      "mean_token_accuracy": 0.21080860495567322,
      "num_tokens": 51048313.0,
      "step": 22285
    },
    {
      "entropy": 5.625257873535157,
      "epoch": 2.2034400948991695,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045166517189935924,
      "loss": 4.9829,
      "mean_token_accuracy": 0.2228904828429222,
      "num_tokens": 51060503.0,
      "step": 22290
    },
    {
      "entropy": 5.637003993988037,
      "epoch": 2.203934361407671,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045164331571830875,
      "loss": 4.9003,
      "mean_token_accuracy": 0.2190896451473236,
      "num_tokens": 51070865.0,
      "step": 22295
    },
    {
      "entropy": 5.664429616928101,
      "epoch": 2.2044286279161724,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045162145519180534,
      "loss": 5.0023,
      "mean_token_accuracy": 0.21651459187269212,
      "num_tokens": 51081158.0,
      "step": 22300
    },
    {
      "entropy": 5.578027629852295,
      "epoch": 2.2049228944246737,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004515995903203867,
      "loss": 4.8849,
      "mean_token_accuracy": 0.23054795563220978,
      "num_tokens": 51090661.0,
      "step": 22305
    },
    {
      "entropy": 5.57456865310669,
      "epoch": 2.205417160933175,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004515777211045908,
      "loss": 4.9788,
      "mean_token_accuracy": 0.21539676636457444,
      "num_tokens": 51101825.0,
      "step": 22310
    },
    {
      "entropy": 5.610644674301147,
      "epoch": 2.2059114274416767,
      "grad_norm": 1.125,
      "learning_rate": 0.00045155584754495546,
      "loss": 5.015,
      "mean_token_accuracy": 0.21632042080163955,
      "num_tokens": 51114546.0,
      "step": 22315
    },
    {
      "entropy": 5.686025190353393,
      "epoch": 2.206405693950178,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004515339696420189,
      "loss": 4.971,
      "mean_token_accuracy": 0.21759041547775268,
      "num_tokens": 51124888.0,
      "step": 22320
    },
    {
      "entropy": 5.5615733623504635,
      "epoch": 2.206899960458679,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004515120873963194,
      "loss": 4.9021,
      "mean_token_accuracy": 0.22509758025407792,
      "num_tokens": 51135892.0,
      "step": 22325
    },
    {
      "entropy": 5.647225475311279,
      "epoch": 2.207394226967181,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004514902008083951,
      "loss": 5.014,
      "mean_token_accuracy": 0.21233924627304077,
      "num_tokens": 51147513.0,
      "step": 22330
    },
    {
      "entropy": 5.666854476928711,
      "epoch": 2.207888493475682,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004514683098787845,
      "loss": 5.0193,
      "mean_token_accuracy": 0.21820505261421203,
      "num_tokens": 51159514.0,
      "step": 22335
    },
    {
      "entropy": 5.685096216201782,
      "epoch": 2.2083827599841834,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004514464146080261,
      "loss": 5.0654,
      "mean_token_accuracy": 0.2144334778189659,
      "num_tokens": 51171245.0,
      "step": 22340
    },
    {
      "entropy": 5.624295997619629,
      "epoch": 2.2088770264926847,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045142451499665846,
      "loss": 4.9905,
      "mean_token_accuracy": 0.22204509228467942,
      "num_tokens": 51182763.0,
      "step": 22345
    },
    {
      "entropy": 5.677038860321045,
      "epoch": 2.2093712930011864,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045140261104522044,
      "loss": 5.0217,
      "mean_token_accuracy": 0.21245643645524978,
      "num_tokens": 51193811.0,
      "step": 22350
    },
    {
      "entropy": 5.617627048492432,
      "epoch": 2.2098655595096877,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004513807027542508,
      "loss": 4.9256,
      "mean_token_accuracy": 0.22626098096370698,
      "num_tokens": 51203963.0,
      "step": 22355
    },
    {
      "entropy": 5.609482717514038,
      "epoch": 2.210359826018189,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004513587901242886,
      "loss": 4.9067,
      "mean_token_accuracy": 0.2241803675889969,
      "num_tokens": 51215128.0,
      "step": 22360
    },
    {
      "entropy": 5.580700445175171,
      "epoch": 2.21085409252669,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045133687315587263,
      "loss": 4.9432,
      "mean_token_accuracy": 0.22594606280326843,
      "num_tokens": 51226773.0,
      "step": 22365
    },
    {
      "entropy": 5.594133567810059,
      "epoch": 2.211348359035192,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004513149518495424,
      "loss": 4.9854,
      "mean_token_accuracy": 0.2172035649418831,
      "num_tokens": 51238226.0,
      "step": 22370
    },
    {
      "entropy": 5.656217956542969,
      "epoch": 2.211842625543693,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045129302620583683,
      "loss": 5.0008,
      "mean_token_accuracy": 0.22127577513456345,
      "num_tokens": 51249096.0,
      "step": 22375
    },
    {
      "entropy": 5.680662488937378,
      "epoch": 2.2123368920521944,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045127109622529563,
      "loss": 4.9951,
      "mean_token_accuracy": 0.2232608452439308,
      "num_tokens": 51259709.0,
      "step": 22380
    },
    {
      "entropy": 5.5715508460998535,
      "epoch": 2.212831158560696,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045124916190845797,
      "loss": 4.976,
      "mean_token_accuracy": 0.22865558713674544,
      "num_tokens": 51271787.0,
      "step": 22385
    },
    {
      "entropy": 5.673215627670288,
      "epoch": 2.2133254250691974,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045122722325586373,
      "loss": 5.0271,
      "mean_token_accuracy": 0.21401347666978837,
      "num_tokens": 51284093.0,
      "step": 22390
    },
    {
      "entropy": 5.706920099258423,
      "epoch": 2.2138196915776986,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045120528026805235,
      "loss": 5.0269,
      "mean_token_accuracy": 0.21827742755413054,
      "num_tokens": 51295659.0,
      "step": 22395
    },
    {
      "entropy": 5.6794092655181885,
      "epoch": 2.2143139580862,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045118333294556377,
      "loss": 5.0468,
      "mean_token_accuracy": 0.21142876744270325,
      "num_tokens": 51307960.0,
      "step": 22400
    },
    {
      "entropy": 5.581096935272217,
      "epoch": 2.2148082245947016,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004511613812889379,
      "loss": 4.9061,
      "mean_token_accuracy": 0.22600414454936982,
      "num_tokens": 51319366.0,
      "step": 22405
    },
    {
      "entropy": 5.583076333999633,
      "epoch": 2.215302491103203,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045113942529871473,
      "loss": 4.9469,
      "mean_token_accuracy": 0.22197468727827072,
      "num_tokens": 51332445.0,
      "step": 22410
    },
    {
      "entropy": 5.605283975601196,
      "epoch": 2.215796757611704,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004511174649754343,
      "loss": 4.9155,
      "mean_token_accuracy": 0.2230609431862831,
      "num_tokens": 51342900.0,
      "step": 22415
    },
    {
      "entropy": 5.6048516750335695,
      "epoch": 2.2162910241202054,
      "grad_norm": 1.03125,
      "learning_rate": 0.000451095500319637,
      "loss": 4.9838,
      "mean_token_accuracy": 0.22243473082780837,
      "num_tokens": 51354437.0,
      "step": 22420
    },
    {
      "entropy": 5.594480228424072,
      "epoch": 2.216785290628707,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000451073531331863,
      "loss": 4.9069,
      "mean_token_accuracy": 0.2228037804365158,
      "num_tokens": 51364630.0,
      "step": 22425
    },
    {
      "entropy": 5.649523544311523,
      "epoch": 2.2172795571372084,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004510515580126528,
      "loss": 5.0741,
      "mean_token_accuracy": 0.2083497703075409,
      "num_tokens": 51376547.0,
      "step": 22430
    },
    {
      "entropy": 5.700908088684082,
      "epoch": 2.2177738236457096,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000451029580362547,
      "loss": 5.0363,
      "mean_token_accuracy": 0.2199498012661934,
      "num_tokens": 51388704.0,
      "step": 22435
    },
    {
      "entropy": 5.811529207229614,
      "epoch": 2.2182680901542113,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004510075983820862,
      "loss": 5.1978,
      "mean_token_accuracy": 0.19749512374401093,
      "num_tokens": 51400884.0,
      "step": 22440
    },
    {
      "entropy": 5.640939378738404,
      "epoch": 2.2187623566627126,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045098561207181114,
      "loss": 4.9041,
      "mean_token_accuracy": 0.23067692071199417,
      "num_tokens": 51412946.0,
      "step": 22445
    },
    {
      "entropy": 5.662036371231079,
      "epoch": 2.219256623171214,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004509636214322628,
      "loss": 5.0517,
      "mean_token_accuracy": 0.21512101888656615,
      "num_tokens": 51425860.0,
      "step": 22450
    },
    {
      "entropy": 5.649212741851807,
      "epoch": 2.219750889679715,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004509416264639821,
      "loss": 4.9906,
      "mean_token_accuracy": 0.21654520481824874,
      "num_tokens": 51438812.0,
      "step": 22455
    },
    {
      "entropy": 5.695921993255615,
      "epoch": 2.220245156188217,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004509196271675101,
      "loss": 5.0964,
      "mean_token_accuracy": 0.2097064048051834,
      "num_tokens": 51449881.0,
      "step": 22460
    },
    {
      "entropy": 5.608725595474243,
      "epoch": 2.220739422696718,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045089762354338784,
      "loss": 4.9545,
      "mean_token_accuracy": 0.21869312226772308,
      "num_tokens": 51460403.0,
      "step": 22465
    },
    {
      "entropy": 5.690785884857178,
      "epoch": 2.2212336892052194,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045087561559215683,
      "loss": 5.001,
      "mean_token_accuracy": 0.2102307140827179,
      "num_tokens": 51471414.0,
      "step": 22470
    },
    {
      "entropy": 5.602718734741211,
      "epoch": 2.221727955713721,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045085360331435835,
      "loss": 4.9478,
      "mean_token_accuracy": 0.22205538749694825,
      "num_tokens": 51482855.0,
      "step": 22475
    },
    {
      "entropy": 5.702400398254395,
      "epoch": 2.2222222222222223,
      "grad_norm": 1.046875,
      "learning_rate": 0.000450831586710534,
      "loss": 5.0625,
      "mean_token_accuracy": 0.21329119354486464,
      "num_tokens": 51495363.0,
      "step": 22480
    },
    {
      "entropy": 5.653773546218872,
      "epoch": 2.2227164887307236,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045080956578122535,
      "loss": 4.972,
      "mean_token_accuracy": 0.21495070457458496,
      "num_tokens": 51506678.0,
      "step": 22485
    },
    {
      "entropy": 5.654488945007325,
      "epoch": 2.223210755239225,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045078754052697406,
      "loss": 5.0138,
      "mean_token_accuracy": 0.21856178045272828,
      "num_tokens": 51518596.0,
      "step": 22490
    },
    {
      "entropy": 5.709525537490845,
      "epoch": 2.2237050217477266,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045076551094832206,
      "loss": 5.1159,
      "mean_token_accuracy": 0.20502690076828003,
      "num_tokens": 51531292.0,
      "step": 22495
    },
    {
      "entropy": 5.6593663692474365,
      "epoch": 2.224199288256228,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045074347704581114,
      "loss": 4.9644,
      "mean_token_accuracy": 0.22537117600440978,
      "num_tokens": 51542507.0,
      "step": 22500
    },
    {
      "entropy": 5.657744312286377,
      "epoch": 2.224693554764729,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045072143881998345,
      "loss": 4.9882,
      "mean_token_accuracy": 0.2155720293521881,
      "num_tokens": 51553895.0,
      "step": 22505
    },
    {
      "entropy": 5.705785608291626,
      "epoch": 2.2251878212732303,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004506993962713812,
      "loss": 5.0807,
      "mean_token_accuracy": 0.2111979305744171,
      "num_tokens": 51566533.0,
      "step": 22510
    },
    {
      "entropy": 5.688817834854126,
      "epoch": 2.225682087781732,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045067734940054644,
      "loss": 5.0,
      "mean_token_accuracy": 0.21248604208230973,
      "num_tokens": 51576308.0,
      "step": 22515
    },
    {
      "entropy": 5.6117274284362795,
      "epoch": 2.2261763542902333,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045065529820802165,
      "loss": 4.9447,
      "mean_token_accuracy": 0.2170528531074524,
      "num_tokens": 51589844.0,
      "step": 22520
    },
    {
      "entropy": 5.629724407196045,
      "epoch": 2.2266706207987346,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045063324269434924,
      "loss": 4.9311,
      "mean_token_accuracy": 0.22700488716363906,
      "num_tokens": 51601685.0,
      "step": 22525
    },
    {
      "entropy": 5.672740507125854,
      "epoch": 2.227164887307236,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004506111828600719,
      "loss": 4.9934,
      "mean_token_accuracy": 0.22276042997837067,
      "num_tokens": 51613258.0,
      "step": 22530
    },
    {
      "entropy": 5.6977897644042965,
      "epoch": 2.2276591538157375,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004505891187057322,
      "loss": 5.0285,
      "mean_token_accuracy": 0.22136635035276414,
      "num_tokens": 51624088.0,
      "step": 22535
    },
    {
      "entropy": 5.651932334899902,
      "epoch": 2.228153420324239,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004505670502318729,
      "loss": 4.9792,
      "mean_token_accuracy": 0.2240745395421982,
      "num_tokens": 51636617.0,
      "step": 22540
    },
    {
      "entropy": 5.689702939987183,
      "epoch": 2.22864768683274,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045054497743903687,
      "loss": 5.0451,
      "mean_token_accuracy": 0.20711152553558348,
      "num_tokens": 51647777.0,
      "step": 22545
    },
    {
      "entropy": 5.652089786529541,
      "epoch": 2.2291419533412418,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004505229003277673,
      "loss": 5.0521,
      "mean_token_accuracy": 0.20854245573282243,
      "num_tokens": 51660505.0,
      "step": 22550
    },
    {
      "entropy": 5.702612781524659,
      "epoch": 2.229636219849743,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045050081889860713,
      "loss": 4.9966,
      "mean_token_accuracy": 0.22267340123653412,
      "num_tokens": 51672296.0,
      "step": 22555
    },
    {
      "entropy": 5.649834060668946,
      "epoch": 2.2301304863582443,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045047873315209945,
      "loss": 5.0137,
      "mean_token_accuracy": 0.22388536632061004,
      "num_tokens": 51684437.0,
      "step": 22560
    },
    {
      "entropy": 5.583037185668945,
      "epoch": 2.2306247528667456,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004504566430887878,
      "loss": 4.8438,
      "mean_token_accuracy": 0.2330440565943718,
      "num_tokens": 51694919.0,
      "step": 22565
    },
    {
      "entropy": 5.60652174949646,
      "epoch": 2.2311190193752473,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045043454870921553,
      "loss": 4.9681,
      "mean_token_accuracy": 0.22372659891843796,
      "num_tokens": 51705273.0,
      "step": 22570
    },
    {
      "entropy": 5.668671989440918,
      "epoch": 2.2316132858837485,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004504124500139261,
      "loss": 5.0047,
      "mean_token_accuracy": 0.21230462938547134,
      "num_tokens": 51717048.0,
      "step": 22575
    },
    {
      "entropy": 5.7057915210723875,
      "epoch": 2.23210755239225,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004503903470034632,
      "loss": 5.0706,
      "mean_token_accuracy": 0.21053809225559234,
      "num_tokens": 51729208.0,
      "step": 22580
    },
    {
      "entropy": 5.588547468185425,
      "epoch": 2.2326018189007515,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045036823967837066,
      "loss": 4.9378,
      "mean_token_accuracy": 0.22529369443655015,
      "num_tokens": 51740393.0,
      "step": 22585
    },
    {
      "entropy": 5.661834716796875,
      "epoch": 2.2330960854092528,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004503461280391921,
      "loss": 5.0047,
      "mean_token_accuracy": 0.2206682503223419,
      "num_tokens": 51751044.0,
      "step": 22590
    },
    {
      "entropy": 5.686704683303833,
      "epoch": 2.233590351917754,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004503240120864716,
      "loss": 5.0039,
      "mean_token_accuracy": 0.21298899501562119,
      "num_tokens": 51761969.0,
      "step": 22595
    },
    {
      "entropy": 5.644410610198975,
      "epoch": 2.2340846184262553,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004503018918207532,
      "loss": 5.0276,
      "mean_token_accuracy": 0.21174830794334412,
      "num_tokens": 51774185.0,
      "step": 22600
    },
    {
      "entropy": 5.692815256118775,
      "epoch": 2.234578884934757,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004502797672425811,
      "loss": 4.9822,
      "mean_token_accuracy": 0.21953933089971542,
      "num_tokens": 51786155.0,
      "step": 22605
    },
    {
      "entropy": 5.666292905807495,
      "epoch": 2.2350731514432582,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045025763835249957,
      "loss": 5.0219,
      "mean_token_accuracy": 0.2182562306523323,
      "num_tokens": 51797125.0,
      "step": 22610
    },
    {
      "entropy": 5.718952798843384,
      "epoch": 2.2355674179517595,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004502355051510528,
      "loss": 5.1062,
      "mean_token_accuracy": 0.21430083513259887,
      "num_tokens": 51809046.0,
      "step": 22615
    },
    {
      "entropy": 5.703240919113159,
      "epoch": 2.2360616844602608,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045021336763878555,
      "loss": 5.152,
      "mean_token_accuracy": 0.2022416114807129,
      "num_tokens": 51821857.0,
      "step": 22620
    },
    {
      "entropy": 5.711896800994873,
      "epoch": 2.2365559509687625,
      "grad_norm": 1.3125,
      "learning_rate": 0.00045019122581624224,
      "loss": 4.9533,
      "mean_token_accuracy": 0.22200523167848588,
      "num_tokens": 51833504.0,
      "step": 22625
    },
    {
      "entropy": 5.612013530731201,
      "epoch": 2.2370502174772637,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045016907968396764,
      "loss": 5.0253,
      "mean_token_accuracy": 0.21867602914571763,
      "num_tokens": 51844166.0,
      "step": 22630
    },
    {
      "entropy": 5.665346956253051,
      "epoch": 2.237544483985765,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004501469292425064,
      "loss": 5.0142,
      "mean_token_accuracy": 0.21681543290615082,
      "num_tokens": 51855674.0,
      "step": 22635
    },
    {
      "entropy": 5.618327236175537,
      "epoch": 2.2380387504942667,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045012477449240363,
      "loss": 4.9752,
      "mean_token_accuracy": 0.2187501758337021,
      "num_tokens": 51866128.0,
      "step": 22640
    },
    {
      "entropy": 5.651944017410278,
      "epoch": 2.238533017002768,
      "grad_norm": 1.125,
      "learning_rate": 0.00045010261543420416,
      "loss": 5.0691,
      "mean_token_accuracy": 0.21898249834775924,
      "num_tokens": 51877036.0,
      "step": 22645
    },
    {
      "entropy": 5.659620761871338,
      "epoch": 2.2390272835112692,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004500804520684532,
      "loss": 4.9838,
      "mean_token_accuracy": 0.2199044182896614,
      "num_tokens": 51888331.0,
      "step": 22650
    },
    {
      "entropy": 5.702177047729492,
      "epoch": 2.2395215500197705,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045005828439569594,
      "loss": 5.0271,
      "mean_token_accuracy": 0.21263622343540192,
      "num_tokens": 51899843.0,
      "step": 22655
    },
    {
      "entropy": 5.743517255783081,
      "epoch": 2.240015816528272,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045003611241647783,
      "loss": 5.0493,
      "mean_token_accuracy": 0.2091672897338867,
      "num_tokens": 51911429.0,
      "step": 22660
    },
    {
      "entropy": 5.648055839538574,
      "epoch": 2.2405100830367735,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045001393613134413,
      "loss": 5.0148,
      "mean_token_accuracy": 0.2187858358025551,
      "num_tokens": 51923154.0,
      "step": 22665
    },
    {
      "entropy": 5.691600704193116,
      "epoch": 2.2410043495452747,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004499917555408405,
      "loss": 5.0305,
      "mean_token_accuracy": 0.21180993020534516,
      "num_tokens": 51935126.0,
      "step": 22670
    },
    {
      "entropy": 5.666724061965942,
      "epoch": 2.241498616053776,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044996957064551244,
      "loss": 5.0398,
      "mean_token_accuracy": 0.21079792976379394,
      "num_tokens": 51947629.0,
      "step": 22675
    },
    {
      "entropy": 5.6925272941589355,
      "epoch": 2.2419928825622777,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004499473814459059,
      "loss": 5.0296,
      "mean_token_accuracy": 0.21798262149095535,
      "num_tokens": 51957631.0,
      "step": 22680
    },
    {
      "entropy": 5.693587589263916,
      "epoch": 2.242487149070779,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044992518794256657,
      "loss": 5.0359,
      "mean_token_accuracy": 0.21122840642929078,
      "num_tokens": 51969854.0,
      "step": 22685
    },
    {
      "entropy": 5.718819236755371,
      "epoch": 2.24298141557928,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044990299013604045,
      "loss": 5.0334,
      "mean_token_accuracy": 0.2175116330385208,
      "num_tokens": 51980501.0,
      "step": 22690
    },
    {
      "entropy": 5.657420539855957,
      "epoch": 2.243475682087782,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044988078802687373,
      "loss": 5.0243,
      "mean_token_accuracy": 0.21538246124982835,
      "num_tokens": 51991336.0,
      "step": 22695
    },
    {
      "entropy": 5.675856304168701,
      "epoch": 2.243969948596283,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004498585816156124,
      "loss": 5.0755,
      "mean_token_accuracy": 0.2148443877696991,
      "num_tokens": 52003215.0,
      "step": 22700
    },
    {
      "entropy": 5.712303209304809,
      "epoch": 2.2444642151047844,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004498363709028029,
      "loss": 5.0391,
      "mean_token_accuracy": 0.21391084790229797,
      "num_tokens": 52014772.0,
      "step": 22705
    },
    {
      "entropy": 5.674064254760742,
      "epoch": 2.2449584816132857,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044981415588899153,
      "loss": 5.0393,
      "mean_token_accuracy": 0.21104245334863664,
      "num_tokens": 52025808.0,
      "step": 22710
    },
    {
      "entropy": 5.699333715438843,
      "epoch": 2.2454527481217874,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004497919365747248,
      "loss": 5.0535,
      "mean_token_accuracy": 0.21469468325376512,
      "num_tokens": 52037397.0,
      "step": 22715
    },
    {
      "entropy": 5.596655178070068,
      "epoch": 2.2459470146302887,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044976971296054936,
      "loss": 4.9709,
      "mean_token_accuracy": 0.21980248540639877,
      "num_tokens": 52048874.0,
      "step": 22720
    },
    {
      "entropy": 5.684652280807495,
      "epoch": 2.24644128113879,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004497474850470118,
      "loss": 5.0851,
      "mean_token_accuracy": 0.20508341044187545,
      "num_tokens": 52061180.0,
      "step": 22725
    },
    {
      "entropy": 5.694707584381104,
      "epoch": 2.2469355476472916,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000449725252834659,
      "loss": 4.998,
      "mean_token_accuracy": 0.21377017050981523,
      "num_tokens": 52073400.0,
      "step": 22730
    },
    {
      "entropy": 5.674159908294678,
      "epoch": 2.247429814155793,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044970301632403796,
      "loss": 5.0473,
      "mean_token_accuracy": 0.221113421022892,
      "num_tokens": 52084737.0,
      "step": 22735
    },
    {
      "entropy": 5.618103933334351,
      "epoch": 2.247924080664294,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044968077551569554,
      "loss": 4.9825,
      "mean_token_accuracy": 0.22225635796785354,
      "num_tokens": 52096044.0,
      "step": 22740
    },
    {
      "entropy": 5.681001615524292,
      "epoch": 2.2484183471727954,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000449658530410179,
      "loss": 4.9761,
      "mean_token_accuracy": 0.22113857120275499,
      "num_tokens": 52108701.0,
      "step": 22745
    },
    {
      "entropy": 5.713064575195313,
      "epoch": 2.248912613681297,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004496362810080354,
      "loss": 5.023,
      "mean_token_accuracy": 0.21937201619148256,
      "num_tokens": 52119797.0,
      "step": 22750
    },
    {
      "entropy": 5.645633506774902,
      "epoch": 2.2494068801897984,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004496140273098123,
      "loss": 5.0743,
      "mean_token_accuracy": 0.21006072610616683,
      "num_tokens": 52132557.0,
      "step": 22755
    },
    {
      "entropy": 5.566066265106201,
      "epoch": 2.2499011466982997,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044959176931605704,
      "loss": 4.8886,
      "mean_token_accuracy": 0.22179346680641174,
      "num_tokens": 52144058.0,
      "step": 22760
    },
    {
      "entropy": 5.666570663452148,
      "epoch": 2.250395413206801,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044956950702731714,
      "loss": 5.0118,
      "mean_token_accuracy": 0.21774857342243195,
      "num_tokens": 52155432.0,
      "step": 22765
    },
    {
      "entropy": 5.630380344390869,
      "epoch": 2.2508896797153026,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004495472404441403,
      "loss": 4.9704,
      "mean_token_accuracy": 0.2209709405899048,
      "num_tokens": 52166925.0,
      "step": 22770
    },
    {
      "entropy": 5.5793393611907955,
      "epoch": 2.251383946223804,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004495249695670742,
      "loss": 4.9814,
      "mean_token_accuracy": 0.218789741396904,
      "num_tokens": 52177452.0,
      "step": 22775
    },
    {
      "entropy": 5.633113813400269,
      "epoch": 2.251878212732305,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004495026943966669,
      "loss": 4.954,
      "mean_token_accuracy": 0.2157249018549919,
      "num_tokens": 52189288.0,
      "step": 22780
    },
    {
      "entropy": 5.624559020996093,
      "epoch": 2.2523724792408064,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044948041493346615,
      "loss": 4.9546,
      "mean_token_accuracy": 0.22229027152061462,
      "num_tokens": 52199963.0,
      "step": 22785
    },
    {
      "entropy": 5.620898818969726,
      "epoch": 2.252866745749308,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004494581311780202,
      "loss": 5.0203,
      "mean_token_accuracy": 0.2140634089708328,
      "num_tokens": 52210727.0,
      "step": 22790
    },
    {
      "entropy": 5.6376947402954105,
      "epoch": 2.2533610122578094,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044943584313087705,
      "loss": 5.0304,
      "mean_token_accuracy": 0.21278845220804216,
      "num_tokens": 52222211.0,
      "step": 22795
    },
    {
      "entropy": 5.672648859024048,
      "epoch": 2.2538552787663106,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044941355079258517,
      "loss": 5.0482,
      "mean_token_accuracy": 0.21027215272188188,
      "num_tokens": 52234453.0,
      "step": 22800
    },
    {
      "entropy": 5.625530529022217,
      "epoch": 2.2543495452748123,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004493912541636928,
      "loss": 4.9959,
      "mean_token_accuracy": 0.2218208685517311,
      "num_tokens": 52246324.0,
      "step": 22805
    },
    {
      "entropy": 5.694216775894165,
      "epoch": 2.2548438117833136,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004493689532447486,
      "loss": 5.0522,
      "mean_token_accuracy": 0.21249908804893494,
      "num_tokens": 52258223.0,
      "step": 22810
    },
    {
      "entropy": 5.65999698638916,
      "epoch": 2.255338078291815,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004493466480363011,
      "loss": 5.0456,
      "mean_token_accuracy": 0.21982511579990388,
      "num_tokens": 52270066.0,
      "step": 22815
    },
    {
      "entropy": 5.765982580184937,
      "epoch": 2.255832344800316,
      "grad_norm": 1.125,
      "learning_rate": 0.00044932433853889903,
      "loss": 5.0688,
      "mean_token_accuracy": 0.20999107807874678,
      "num_tokens": 52281366.0,
      "step": 22820
    },
    {
      "entropy": 5.663675928115845,
      "epoch": 2.256326611308818,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004493020247530912,
      "loss": 5.0454,
      "mean_token_accuracy": 0.21232940405607223,
      "num_tokens": 52292109.0,
      "step": 22825
    },
    {
      "entropy": 5.586491489410401,
      "epoch": 2.256820877817319,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004492797066794264,
      "loss": 4.8799,
      "mean_token_accuracy": 0.23107425421476363,
      "num_tokens": 52302707.0,
      "step": 22830
    },
    {
      "entropy": 5.647696781158447,
      "epoch": 2.2573151443258204,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044925738431845385,
      "loss": 4.9872,
      "mean_token_accuracy": 0.21742101162672042,
      "num_tokens": 52313715.0,
      "step": 22835
    },
    {
      "entropy": 5.681366491317749,
      "epoch": 2.257809410834322,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044923505767072267,
      "loss": 5.0068,
      "mean_token_accuracy": 0.20868510901927947,
      "num_tokens": 52325612.0,
      "step": 22840
    },
    {
      "entropy": 5.562087535858154,
      "epoch": 2.2583036773428233,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044921272673678197,
      "loss": 4.8918,
      "mean_token_accuracy": 0.22701400518417358,
      "num_tokens": 52338233.0,
      "step": 22845
    },
    {
      "entropy": 5.617707061767578,
      "epoch": 2.2587979438513246,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004491903915171812,
      "loss": 4.9528,
      "mean_token_accuracy": 0.2214539334177971,
      "num_tokens": 52348792.0,
      "step": 22850
    },
    {
      "entropy": 5.650212144851684,
      "epoch": 2.259292210359826,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044916805201246975,
      "loss": 5.0561,
      "mean_token_accuracy": 0.21108245551586152,
      "num_tokens": 52359334.0,
      "step": 22855
    },
    {
      "entropy": 5.696758460998535,
      "epoch": 2.2597864768683276,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044914570822319723,
      "loss": 5.0529,
      "mean_token_accuracy": 0.21030161529779434,
      "num_tokens": 52370600.0,
      "step": 22860
    },
    {
      "entropy": 5.6942384243011475,
      "epoch": 2.260280743376829,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004491233601499132,
      "loss": 5.039,
      "mean_token_accuracy": 0.21550527215003967,
      "num_tokens": 52382783.0,
      "step": 22865
    },
    {
      "entropy": 5.59625473022461,
      "epoch": 2.26077500988533,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044910100779316757,
      "loss": 4.9117,
      "mean_token_accuracy": 0.22785827666521072,
      "num_tokens": 52393372.0,
      "step": 22870
    },
    {
      "entropy": 5.635334825515747,
      "epoch": 2.261269276393832,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004490786511535102,
      "loss": 4.9636,
      "mean_token_accuracy": 0.22134226262569429,
      "num_tokens": 52405181.0,
      "step": 22875
    },
    {
      "entropy": 5.645710754394531,
      "epoch": 2.261763542902333,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044905629023149084,
      "loss": 4.9791,
      "mean_token_accuracy": 0.21486060172319413,
      "num_tokens": 52417371.0,
      "step": 22880
    },
    {
      "entropy": 5.6763176918029785,
      "epoch": 2.2622578094108343,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004490339250276598,
      "loss": 5.0323,
      "mean_token_accuracy": 0.21425150632858275,
      "num_tokens": 52428712.0,
      "step": 22885
    },
    {
      "entropy": 5.629108667373657,
      "epoch": 2.2627520759193356,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004490115555425672,
      "loss": 5.0536,
      "mean_token_accuracy": 0.2128271698951721,
      "num_tokens": 52439764.0,
      "step": 22890
    },
    {
      "entropy": 5.617012786865234,
      "epoch": 2.263246342427837,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004489891817767634,
      "loss": 5.0416,
      "mean_token_accuracy": 0.21095010936260222,
      "num_tokens": 52451409.0,
      "step": 22895
    },
    {
      "entropy": 5.697663116455078,
      "epoch": 2.2637406089363385,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044896680373079874,
      "loss": 5.0069,
      "mean_token_accuracy": 0.21348000317811966,
      "num_tokens": 52461471.0,
      "step": 22900
    },
    {
      "entropy": 5.69527907371521,
      "epoch": 2.26423487544484,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044894442140522365,
      "loss": 5.0133,
      "mean_token_accuracy": 0.2141559302806854,
      "num_tokens": 52472693.0,
      "step": 22905
    },
    {
      "entropy": 5.639054965972901,
      "epoch": 2.264729141953341,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004489220348005889,
      "loss": 4.9204,
      "mean_token_accuracy": 0.2180305004119873,
      "num_tokens": 52483449.0,
      "step": 22910
    },
    {
      "entropy": 5.601035928726196,
      "epoch": 2.265223408461843,
      "grad_norm": 1.1875,
      "learning_rate": 0.000448899643917445,
      "loss": 5.0083,
      "mean_token_accuracy": 0.22308041155338287,
      "num_tokens": 52494048.0,
      "step": 22915
    },
    {
      "entropy": 5.6666422367095945,
      "epoch": 2.265717674970344,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004488772487563429,
      "loss": 5.032,
      "mean_token_accuracy": 0.21332575529813766,
      "num_tokens": 52506224.0,
      "step": 22920
    },
    {
      "entropy": 5.677220726013184,
      "epoch": 2.2662119414788453,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004488548493178336,
      "loss": 5.0059,
      "mean_token_accuracy": 0.209632109105587,
      "num_tokens": 52517157.0,
      "step": 22925
    },
    {
      "entropy": 5.6694670677185055,
      "epoch": 2.2667062079873466,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044883244560246793,
      "loss": 5.0707,
      "mean_token_accuracy": 0.21487941592931747,
      "num_tokens": 52528826.0,
      "step": 22930
    },
    {
      "entropy": 5.663610506057739,
      "epoch": 2.2672004744958483,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004488100376107972,
      "loss": 4.8953,
      "mean_token_accuracy": 0.23146772831678392,
      "num_tokens": 52540068.0,
      "step": 22935
    },
    {
      "entropy": 5.607417631149292,
      "epoch": 2.2676947410043495,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004487876253433726,
      "loss": 4.9508,
      "mean_token_accuracy": 0.21942494660615922,
      "num_tokens": 52552611.0,
      "step": 22940
    },
    {
      "entropy": 5.620614004135132,
      "epoch": 2.268189007512851,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004487652088007454,
      "loss": 5.0042,
      "mean_token_accuracy": 0.21663637608289718,
      "num_tokens": 52564532.0,
      "step": 22945
    },
    {
      "entropy": 5.721346759796143,
      "epoch": 2.2686832740213525,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004487427879834671,
      "loss": 5.0491,
      "mean_token_accuracy": 0.20971158444881438,
      "num_tokens": 52576110.0,
      "step": 22950
    },
    {
      "entropy": 5.690198564529419,
      "epoch": 2.2691775405298538,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044872036289208926,
      "loss": 4.9938,
      "mean_token_accuracy": 0.21658954322338103,
      "num_tokens": 52586890.0,
      "step": 22955
    },
    {
      "entropy": 5.603496980667114,
      "epoch": 2.269671807038355,
      "grad_norm": 1.125,
      "learning_rate": 0.0004486979335271635,
      "loss": 4.958,
      "mean_token_accuracy": 0.2178856536746025,
      "num_tokens": 52598597.0,
      "step": 22960
    },
    {
      "entropy": 5.5988301753997805,
      "epoch": 2.2701660735468563,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044867549988924176,
      "loss": 4.9729,
      "mean_token_accuracy": 0.22163741886615754,
      "num_tokens": 52609361.0,
      "step": 22965
    },
    {
      "entropy": 5.725370979309082,
      "epoch": 2.270660340055358,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044865306197887573,
      "loss": 5.0648,
      "mean_token_accuracy": 0.21067475378513337,
      "num_tokens": 52620954.0,
      "step": 22970
    },
    {
      "entropy": 5.725204229354858,
      "epoch": 2.2711546065638593,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044863061979661745,
      "loss": 5.0845,
      "mean_token_accuracy": 0.21283821612596512,
      "num_tokens": 52631944.0,
      "step": 22975
    },
    {
      "entropy": 5.699615526199341,
      "epoch": 2.2716488730723605,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004486081733430189,
      "loss": 5.0351,
      "mean_token_accuracy": 0.2083725020289421,
      "num_tokens": 52642517.0,
      "step": 22980
    },
    {
      "entropy": 5.703259134292603,
      "epoch": 2.2721431395808622,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044858572261863245,
      "loss": 5.0818,
      "mean_token_accuracy": 0.21242880821228027,
      "num_tokens": 52653654.0,
      "step": 22985
    },
    {
      "entropy": 5.679076337814331,
      "epoch": 2.2726374060893635,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004485632676240102,
      "loss": 5.0493,
      "mean_token_accuracy": 0.21610891371965407,
      "num_tokens": 52666068.0,
      "step": 22990
    },
    {
      "entropy": 5.685547161102295,
      "epoch": 2.2731316725978647,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004485408083597047,
      "loss": 4.9965,
      "mean_token_accuracy": 0.21779437810182573,
      "num_tokens": 52677797.0,
      "step": 22995
    },
    {
      "entropy": 5.64984712600708,
      "epoch": 2.273625939106366,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004485183448262684,
      "loss": 4.9811,
      "mean_token_accuracy": 0.21992648392915726,
      "num_tokens": 52688947.0,
      "step": 23000
    },
    {
      "entropy": 5.645264434814453,
      "epoch": 2.2741202056148677,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004484958770242538,
      "loss": 5.0236,
      "mean_token_accuracy": 0.20975972563028336,
      "num_tokens": 52700357.0,
      "step": 23005
    },
    {
      "entropy": 5.73111662864685,
      "epoch": 2.274614472123369,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044847340495421384,
      "loss": 5.0983,
      "mean_token_accuracy": 0.20657340288162232,
      "num_tokens": 52712602.0,
      "step": 23010
    },
    {
      "entropy": 5.712519025802612,
      "epoch": 2.2751087386318702,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044845092861670114,
      "loss": 4.9916,
      "mean_token_accuracy": 0.21676507592201233,
      "num_tokens": 52723480.0,
      "step": 23015
    },
    {
      "entropy": 5.636447954177856,
      "epoch": 2.2756030051403715,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004484284480122688,
      "loss": 4.9926,
      "mean_token_accuracy": 0.21316518634557724,
      "num_tokens": 52735418.0,
      "step": 23020
    },
    {
      "entropy": 5.602292776107788,
      "epoch": 2.276097271648873,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044840596314146955,
      "loss": 4.9408,
      "mean_token_accuracy": 0.22402605265378953,
      "num_tokens": 52746308.0,
      "step": 23025
    },
    {
      "entropy": 5.581596803665161,
      "epoch": 2.2765915381573745,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004483834740048568,
      "loss": 4.8915,
      "mean_token_accuracy": 0.22750741392374038,
      "num_tokens": 52756834.0,
      "step": 23030
    },
    {
      "entropy": 5.624752378463745,
      "epoch": 2.2770858046658757,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004483609806029837,
      "loss": 4.9974,
      "mean_token_accuracy": 0.21336622536182404,
      "num_tokens": 52768398.0,
      "step": 23035
    },
    {
      "entropy": 5.702380847930908,
      "epoch": 2.277580071174377,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044833848293640354,
      "loss": 5.072,
      "mean_token_accuracy": 0.21296582221984864,
      "num_tokens": 52780062.0,
      "step": 23040
    },
    {
      "entropy": 5.664226913452149,
      "epoch": 2.2780743376828787,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044831598100566984,
      "loss": 4.9702,
      "mean_token_accuracy": 0.2221826732158661,
      "num_tokens": 52791760.0,
      "step": 23045
    },
    {
      "entropy": 5.627329730987549,
      "epoch": 2.27856860419138,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044829347481133613,
      "loss": 4.9294,
      "mean_token_accuracy": 0.22126214653253556,
      "num_tokens": 52802638.0,
      "step": 23050
    },
    {
      "entropy": 5.611829090118408,
      "epoch": 2.2790628706998812,
      "grad_norm": 1.15625,
      "learning_rate": 0.000448270964353956,
      "loss": 5.01,
      "mean_token_accuracy": 0.2156575709581375,
      "num_tokens": 52813590.0,
      "step": 23055
    },
    {
      "entropy": 5.640258693695069,
      "epoch": 2.279557137208383,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004482484496340833,
      "loss": 5.0378,
      "mean_token_accuracy": 0.22053225785493852,
      "num_tokens": 52824177.0,
      "step": 23060
    },
    {
      "entropy": 5.781703758239746,
      "epoch": 2.280051403716884,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004482259306522719,
      "loss": 5.1116,
      "mean_token_accuracy": 0.2061627835035324,
      "num_tokens": 52834637.0,
      "step": 23065
    },
    {
      "entropy": 5.720599126815796,
      "epoch": 2.2805456702253855,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044820340740907565,
      "loss": 5.0572,
      "mean_token_accuracy": 0.21326611042022706,
      "num_tokens": 52847893.0,
      "step": 23070
    },
    {
      "entropy": 5.682198810577392,
      "epoch": 2.2810399367338867,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004481808799050488,
      "loss": 4.9588,
      "mean_token_accuracy": 0.2172829195857048,
      "num_tokens": 52859339.0,
      "step": 23075
    },
    {
      "entropy": 5.679493618011475,
      "epoch": 2.2815342032423884,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044815834814074544,
      "loss": 5.0303,
      "mean_token_accuracy": 0.2108192890882492,
      "num_tokens": 52872405.0,
      "step": 23080
    },
    {
      "entropy": 5.629469728469848,
      "epoch": 2.2820284697508897,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004481358121167198,
      "loss": 4.9003,
      "mean_token_accuracy": 0.23121637105941772,
      "num_tokens": 52882187.0,
      "step": 23085
    },
    {
      "entropy": 5.616191720962524,
      "epoch": 2.282522736259391,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044811327183352627,
      "loss": 4.9466,
      "mean_token_accuracy": 0.22572716176509858,
      "num_tokens": 52894004.0,
      "step": 23090
    },
    {
      "entropy": 5.749456310272217,
      "epoch": 2.2830170027678927,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004480907272917195,
      "loss": 5.0437,
      "mean_token_accuracy": 0.20853720754384994,
      "num_tokens": 52904749.0,
      "step": 23095
    },
    {
      "entropy": 5.656525135040283,
      "epoch": 2.283511269276394,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004480681784918539,
      "loss": 5.0072,
      "mean_token_accuracy": 0.21528094857931138,
      "num_tokens": 52915794.0,
      "step": 23100
    },
    {
      "entropy": 5.652636623382568,
      "epoch": 2.284005535784895,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044804562543448436,
      "loss": 4.9765,
      "mean_token_accuracy": 0.2114195078611374,
      "num_tokens": 52926695.0,
      "step": 23105
    },
    {
      "entropy": 5.6036334991455075,
      "epoch": 2.2844998022933964,
      "grad_norm": 1.125,
      "learning_rate": 0.0004480230681201655,
      "loss": 4.9479,
      "mean_token_accuracy": 0.220542511343956,
      "num_tokens": 52937824.0,
      "step": 23110
    },
    {
      "entropy": 5.610993337631226,
      "epoch": 2.284994068801898,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044800050654945233,
      "loss": 4.9992,
      "mean_token_accuracy": 0.21658325493335723,
      "num_tokens": 52949174.0,
      "step": 23115
    },
    {
      "entropy": 5.679161787033081,
      "epoch": 2.2854883353103994,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044797794072289987,
      "loss": 5.0644,
      "mean_token_accuracy": 0.21063904017210006,
      "num_tokens": 52960675.0,
      "step": 23120
    },
    {
      "entropy": 5.668533182144165,
      "epoch": 2.2859826018189007,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044795537064106335,
      "loss": 5.0083,
      "mean_token_accuracy": 0.2149379551410675,
      "num_tokens": 52971705.0,
      "step": 23125
    },
    {
      "entropy": 5.6634509563446045,
      "epoch": 2.2864768683274024,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044793279630449777,
      "loss": 5.0223,
      "mean_token_accuracy": 0.21304048597812653,
      "num_tokens": 52983025.0,
      "step": 23130
    },
    {
      "entropy": 5.68214225769043,
      "epoch": 2.2869711348359036,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004479102177137586,
      "loss": 5.012,
      "mean_token_accuracy": 0.2077503815293312,
      "num_tokens": 52993810.0,
      "step": 23135
    },
    {
      "entropy": 5.661705827713012,
      "epoch": 2.287465401344405,
      "grad_norm": 1.0,
      "learning_rate": 0.0004478876348694012,
      "loss": 5.0089,
      "mean_token_accuracy": 0.20985692441463472,
      "num_tokens": 53005902.0,
      "step": 23140
    },
    {
      "entropy": 5.675730276107788,
      "epoch": 2.287959667852906,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004478650477719812,
      "loss": 5.0274,
      "mean_token_accuracy": 0.2128356322646141,
      "num_tokens": 53017511.0,
      "step": 23145
    },
    {
      "entropy": 5.612431049346924,
      "epoch": 2.2884539343614074,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004478424564220542,
      "loss": 4.901,
      "mean_token_accuracy": 0.22601800560951232,
      "num_tokens": 53028514.0,
      "step": 23150
    },
    {
      "entropy": 5.669909334182739,
      "epoch": 2.288948200869909,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044781986082017595,
      "loss": 4.9637,
      "mean_token_accuracy": 0.2181425303220749,
      "num_tokens": 53039338.0,
      "step": 23155
    },
    {
      "entropy": 5.641277074813843,
      "epoch": 2.2894424673784104,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044779726096690227,
      "loss": 4.9509,
      "mean_token_accuracy": 0.22108167856931688,
      "num_tokens": 53050877.0,
      "step": 23160
    },
    {
      "entropy": 5.693188142776489,
      "epoch": 2.2899367338869117,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004477746568627892,
      "loss": 5.0062,
      "mean_token_accuracy": 0.21315236538648605,
      "num_tokens": 53062819.0,
      "step": 23165
    },
    {
      "entropy": 5.685944843292236,
      "epoch": 2.2904310003954134,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044775204850839277,
      "loss": 5.0654,
      "mean_token_accuracy": 0.21257565170526505,
      "num_tokens": 53073486.0,
      "step": 23170
    },
    {
      "entropy": 5.594070053100586,
      "epoch": 2.2909252669039146,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004477294359042692,
      "loss": 4.9725,
      "mean_token_accuracy": 0.22101456969976424,
      "num_tokens": 53084499.0,
      "step": 23175
    },
    {
      "entropy": 5.638562965393066,
      "epoch": 2.291419533412416,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004477068190509747,
      "loss": 5.0478,
      "mean_token_accuracy": 0.21086382120847702,
      "num_tokens": 53097181.0,
      "step": 23180
    },
    {
      "entropy": 5.72480092048645,
      "epoch": 2.291913799920917,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044768419794906556,
      "loss": 5.0834,
      "mean_token_accuracy": 0.21605592221021652,
      "num_tokens": 53108586.0,
      "step": 23185
    },
    {
      "entropy": 5.7527566909790036,
      "epoch": 2.292408066429419,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044766157259909853,
      "loss": 4.9605,
      "mean_token_accuracy": 0.21318284273147584,
      "num_tokens": 53119481.0,
      "step": 23190
    },
    {
      "entropy": 5.685546493530273,
      "epoch": 2.29290233293792,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004476389430016299,
      "loss": 4.9683,
      "mean_token_accuracy": 0.22081106305122375,
      "num_tokens": 53129482.0,
      "step": 23195
    },
    {
      "entropy": 5.576179647445679,
      "epoch": 2.2933965994464214,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044761630915721645,
      "loss": 4.9851,
      "mean_token_accuracy": 0.21864856630563737,
      "num_tokens": 53142311.0,
      "step": 23200
    },
    {
      "entropy": 5.632884359359741,
      "epoch": 2.293890865954923,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004475936710664151,
      "loss": 4.9642,
      "mean_token_accuracy": 0.2128441110253334,
      "num_tokens": 53153485.0,
      "step": 23205
    },
    {
      "entropy": 5.681209373474121,
      "epoch": 2.2943851324634243,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004475710287297826,
      "loss": 4.9987,
      "mean_token_accuracy": 0.20964257121086122,
      "num_tokens": 53164121.0,
      "step": 23210
    },
    {
      "entropy": 5.67168493270874,
      "epoch": 2.2948793989719256,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004475483821478761,
      "loss": 4.9976,
      "mean_token_accuracy": 0.21431410014629365,
      "num_tokens": 53174949.0,
      "step": 23215
    },
    {
      "entropy": 5.598780012130737,
      "epoch": 2.295373665480427,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044752573132125263,
      "loss": 4.9489,
      "mean_token_accuracy": 0.22340573817491532,
      "num_tokens": 53186762.0,
      "step": 23220
    },
    {
      "entropy": 5.700349569320679,
      "epoch": 2.2958679319889286,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004475030762504694,
      "loss": 4.9925,
      "mean_token_accuracy": 0.21599723398685455,
      "num_tokens": 53198261.0,
      "step": 23225
    },
    {
      "entropy": 5.575061702728272,
      "epoch": 2.29636219849743,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004474804169360836,
      "loss": 4.8972,
      "mean_token_accuracy": 0.22586987018585206,
      "num_tokens": 53209143.0,
      "step": 23230
    },
    {
      "entropy": 5.71795163154602,
      "epoch": 2.296856465005931,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00044745775337865293,
      "loss": 5.1495,
      "mean_token_accuracy": 0.208558090031147,
      "num_tokens": 53222619.0,
      "step": 23235
    },
    {
      "entropy": 5.637172365188599,
      "epoch": 2.297350731514433,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044743508557873473,
      "loss": 4.9373,
      "mean_token_accuracy": 0.22620487660169603,
      "num_tokens": 53233829.0,
      "step": 23240
    },
    {
      "entropy": 5.666169357299805,
      "epoch": 2.297844998022934,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044741241353688665,
      "loss": 5.0242,
      "mean_token_accuracy": 0.21004287749528885,
      "num_tokens": 53244248.0,
      "step": 23245
    },
    {
      "entropy": 5.686219501495361,
      "epoch": 2.2983392645314353,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004473897372536665,
      "loss": 5.0422,
      "mean_token_accuracy": 0.21702573001384734,
      "num_tokens": 53255638.0,
      "step": 23250
    },
    {
      "entropy": 5.680190181732177,
      "epoch": 2.2988335310399366,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000447367056729632,
      "loss": 5.0147,
      "mean_token_accuracy": 0.21081238836050034,
      "num_tokens": 53265798.0,
      "step": 23255
    },
    {
      "entropy": 5.664991044998169,
      "epoch": 2.2993277975484383,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004473443719653412,
      "loss": 5.0007,
      "mean_token_accuracy": 0.21793916076421738,
      "num_tokens": 53278001.0,
      "step": 23260
    },
    {
      "entropy": 5.687960386276245,
      "epoch": 2.2998220640569396,
      "grad_norm": 1.125,
      "learning_rate": 0.00044732168296135217,
      "loss": 5.0361,
      "mean_token_accuracy": 0.2097020074725151,
      "num_tokens": 53289742.0,
      "step": 23265
    },
    {
      "entropy": 5.696910858154297,
      "epoch": 2.300316330565441,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044729898971822296,
      "loss": 4.9705,
      "mean_token_accuracy": 0.22231234163045882,
      "num_tokens": 53301285.0,
      "step": 23270
    },
    {
      "entropy": 5.583860111236572,
      "epoch": 2.300810597073942,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004472762922365118,
      "loss": 4.9139,
      "mean_token_accuracy": 0.2226989209651947,
      "num_tokens": 53312883.0,
      "step": 23275
    },
    {
      "entropy": 5.658785343170166,
      "epoch": 2.301304863582444,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004472535905167773,
      "loss": 5.064,
      "mean_token_accuracy": 0.2056912139058113,
      "num_tokens": 53325215.0,
      "step": 23280
    },
    {
      "entropy": 5.664778375625611,
      "epoch": 2.301799130090945,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004472308845595776,
      "loss": 5.067,
      "mean_token_accuracy": 0.20703037679195405,
      "num_tokens": 53337104.0,
      "step": 23285
    },
    {
      "entropy": 5.673450040817261,
      "epoch": 2.3022933965994463,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004472081743654715,
      "loss": 5.0412,
      "mean_token_accuracy": 0.21587115675210952,
      "num_tokens": 53348097.0,
      "step": 23290
    },
    {
      "entropy": 5.654085445404053,
      "epoch": 2.3027876631079476,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044718545993501754,
      "loss": 4.9345,
      "mean_token_accuracy": 0.22569898813962935,
      "num_tokens": 53358405.0,
      "step": 23295
    },
    {
      "entropy": 5.654303073883057,
      "epoch": 2.3032819296164493,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004471627412687745,
      "loss": 5.0298,
      "mean_token_accuracy": 0.21355980932712554,
      "num_tokens": 53370012.0,
      "step": 23300
    },
    {
      "entropy": 5.747406721115112,
      "epoch": 2.3037761961249505,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044714001836730145,
      "loss": 5.0996,
      "mean_token_accuracy": 0.20933543741703034,
      "num_tokens": 53383214.0,
      "step": 23305
    },
    {
      "entropy": 5.7046712875366214,
      "epoch": 2.304270462633452,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044711729123115716,
      "loss": 5.0465,
      "mean_token_accuracy": 0.2133384644985199,
      "num_tokens": 53395672.0,
      "step": 23310
    },
    {
      "entropy": 5.642041635513306,
      "epoch": 2.3047647291419535,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044709455986090084,
      "loss": 4.9777,
      "mean_token_accuracy": 0.21674280017614364,
      "num_tokens": 53407577.0,
      "step": 23315
    },
    {
      "entropy": 5.625307273864746,
      "epoch": 2.305258995650455,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004470718242570916,
      "loss": 4.9526,
      "mean_token_accuracy": 0.21659355461597443,
      "num_tokens": 53417772.0,
      "step": 23320
    },
    {
      "entropy": 5.700246143341064,
      "epoch": 2.305753262158956,
      "grad_norm": 1.0,
      "learning_rate": 0.0004470490844202889,
      "loss": 5.0148,
      "mean_token_accuracy": 0.2123020887374878,
      "num_tokens": 53429675.0,
      "step": 23325
    },
    {
      "entropy": 5.650335359573364,
      "epoch": 2.3062475286674573,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004470263403510519,
      "loss": 5.0329,
      "mean_token_accuracy": 0.2086232915520668,
      "num_tokens": 53442352.0,
      "step": 23330
    },
    {
      "entropy": 5.7242063045501705,
      "epoch": 2.306741795175959,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044700359204994034,
      "loss": 5.1417,
      "mean_token_accuracy": 0.20519371181726456,
      "num_tokens": 53453710.0,
      "step": 23335
    },
    {
      "entropy": 5.709506177902222,
      "epoch": 2.3072360616844603,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044698083951751365,
      "loss": 4.9712,
      "mean_token_accuracy": 0.22754202634096146,
      "num_tokens": 53464764.0,
      "step": 23340
    },
    {
      "entropy": 5.646165990829468,
      "epoch": 2.3077303281929615,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004469580827543318,
      "loss": 4.9402,
      "mean_token_accuracy": 0.22145799398422242,
      "num_tokens": 53476495.0,
      "step": 23345
    },
    {
      "entropy": 5.673996543884277,
      "epoch": 2.3082245947014632,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004469353217609542,
      "loss": 5.0252,
      "mean_token_accuracy": 0.2099227175116539,
      "num_tokens": 53488613.0,
      "step": 23350
    },
    {
      "entropy": 5.682986927032471,
      "epoch": 2.3087188612099645,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004469125565379411,
      "loss": 5.0955,
      "mean_token_accuracy": 0.20858614146709442,
      "num_tokens": 53501851.0,
      "step": 23355
    },
    {
      "entropy": 5.679410076141357,
      "epoch": 2.3092131277184658,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004468897870858525,
      "loss": 5.028,
      "mean_token_accuracy": 0.2097813829779625,
      "num_tokens": 53514325.0,
      "step": 23360
    },
    {
      "entropy": 5.695396757125854,
      "epoch": 2.309707394226967,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044686701340524836,
      "loss": 4.9831,
      "mean_token_accuracy": 0.21775872856378556,
      "num_tokens": 53525717.0,
      "step": 23365
    },
    {
      "entropy": 5.64153642654419,
      "epoch": 2.3102016607354687,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004468442354966891,
      "loss": 5.0172,
      "mean_token_accuracy": 0.21389828473329545,
      "num_tokens": 53536488.0,
      "step": 23370
    },
    {
      "entropy": 5.729271745681762,
      "epoch": 2.31069592724397,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000446821453360735,
      "loss": 5.1216,
      "mean_token_accuracy": 0.20298956483602523,
      "num_tokens": 53547230.0,
      "step": 23375
    },
    {
      "entropy": 5.714764213562011,
      "epoch": 2.3111901937524713,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044679866699794644,
      "loss": 5.1282,
      "mean_token_accuracy": 0.20588886886835098,
      "num_tokens": 53558673.0,
      "step": 23380
    },
    {
      "entropy": 5.6543221950531,
      "epoch": 2.311684460260973,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004467758764088841,
      "loss": 4.9744,
      "mean_token_accuracy": 0.22005589604377745,
      "num_tokens": 53570088.0,
      "step": 23385
    },
    {
      "entropy": 5.657515144348144,
      "epoch": 2.3121787267694742,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004467530815941085,
      "loss": 5.0254,
      "mean_token_accuracy": 0.21949983686208724,
      "num_tokens": 53582115.0,
      "step": 23390
    },
    {
      "entropy": 5.6302672863006595,
      "epoch": 2.3126729932779755,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044673028255418047,
      "loss": 5.0289,
      "mean_token_accuracy": 0.2136667862534523,
      "num_tokens": 53594719.0,
      "step": 23395
    },
    {
      "entropy": 5.5587951183319095,
      "epoch": 2.3131672597864767,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004467074792896608,
      "loss": 4.8575,
      "mean_token_accuracy": 0.22768252491950988,
      "num_tokens": 53606453.0,
      "step": 23400
    },
    {
      "entropy": 5.6474823474884035,
      "epoch": 2.313661526294978,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044668467180111054,
      "loss": 4.9929,
      "mean_token_accuracy": 0.21750637143850327,
      "num_tokens": 53617038.0,
      "step": 23405
    },
    {
      "entropy": 5.74325008392334,
      "epoch": 2.3141557928034797,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044666186008909077,
      "loss": 5.0369,
      "mean_token_accuracy": 0.21823037713766097,
      "num_tokens": 53627230.0,
      "step": 23410
    },
    {
      "entropy": 5.715768671035766,
      "epoch": 2.314650059311981,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044663904415416245,
      "loss": 5.0794,
      "mean_token_accuracy": 0.21024925857782364,
      "num_tokens": 53638247.0,
      "step": 23415
    },
    {
      "entropy": 5.681694650650025,
      "epoch": 2.3151443258204822,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004466162239968871,
      "loss": 4.9906,
      "mean_token_accuracy": 0.22165091037750245,
      "num_tokens": 53648643.0,
      "step": 23420
    },
    {
      "entropy": 5.626164436340332,
      "epoch": 2.315638592328984,
      "grad_norm": 1.125,
      "learning_rate": 0.0004465933996178261,
      "loss": 4.9363,
      "mean_token_accuracy": 0.22904857397079467,
      "num_tokens": 53659715.0,
      "step": 23425
    },
    {
      "entropy": 5.608146524429321,
      "epoch": 2.316132858837485,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044657057101754073,
      "loss": 4.9828,
      "mean_token_accuracy": 0.21807103157043456,
      "num_tokens": 53671119.0,
      "step": 23430
    },
    {
      "entropy": 5.651777696609497,
      "epoch": 2.3166271253459865,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044654773819659277,
      "loss": 4.981,
      "mean_token_accuracy": 0.2197079047560692,
      "num_tokens": 53682145.0,
      "step": 23435
    },
    {
      "entropy": 5.704486513137818,
      "epoch": 2.3171213918544877,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044652490115554386,
      "loss": 4.9816,
      "mean_token_accuracy": 0.22056100964546205,
      "num_tokens": 53693882.0,
      "step": 23440
    },
    {
      "entropy": 5.649781131744385,
      "epoch": 2.3176156583629894,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004465020598949557,
      "loss": 5.024,
      "mean_token_accuracy": 0.2093565508723259,
      "num_tokens": 53706010.0,
      "step": 23445
    },
    {
      "entropy": 5.646624660491943,
      "epoch": 2.3181099248714907,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004464792144153903,
      "loss": 5.0455,
      "mean_token_accuracy": 0.21391244232654572,
      "num_tokens": 53717657.0,
      "step": 23450
    },
    {
      "entropy": 5.66133680343628,
      "epoch": 2.318604191379992,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004464563647174096,
      "loss": 5.0017,
      "mean_token_accuracy": 0.21722687929868698,
      "num_tokens": 53729181.0,
      "step": 23455
    },
    {
      "entropy": 5.642555475234985,
      "epoch": 2.3190984578884937,
      "grad_norm": 1.125,
      "learning_rate": 0.0004464335108015757,
      "loss": 4.9774,
      "mean_token_accuracy": 0.2229529544711113,
      "num_tokens": 53740864.0,
      "step": 23460
    },
    {
      "entropy": 5.62421989440918,
      "epoch": 2.319592724396995,
      "grad_norm": 1.125,
      "learning_rate": 0.0004464106526684509,
      "loss": 4.9826,
      "mean_token_accuracy": 0.2158585622906685,
      "num_tokens": 53751726.0,
      "step": 23465
    },
    {
      "entropy": 5.7025621891021725,
      "epoch": 2.320086990905496,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004463877903185974,
      "loss": 5.0303,
      "mean_token_accuracy": 0.21454910188913345,
      "num_tokens": 53763629.0,
      "step": 23470
    },
    {
      "entropy": 5.665783929824829,
      "epoch": 2.3205812574139975,
      "grad_norm": 1.125,
      "learning_rate": 0.0004463649237525777,
      "loss": 4.9996,
      "mean_token_accuracy": 0.22046219259500505,
      "num_tokens": 53774814.0,
      "step": 23475
    },
    {
      "entropy": 5.697757053375244,
      "epoch": 2.321075523922499,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004463420529709543,
      "loss": 5.0092,
      "mean_token_accuracy": 0.21670455634593963,
      "num_tokens": 53786298.0,
      "step": 23480
    },
    {
      "entropy": 5.674049949645996,
      "epoch": 2.3215697904310004,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044631917797428985,
      "loss": 5.0142,
      "mean_token_accuracy": 0.2241431936621666,
      "num_tokens": 53798509.0,
      "step": 23485
    },
    {
      "entropy": 5.700702762603759,
      "epoch": 2.3220640569395017,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000446296298763147,
      "loss": 5.0851,
      "mean_token_accuracy": 0.21019182801246644,
      "num_tokens": 53810689.0,
      "step": 23490
    },
    {
      "entropy": 5.699579668045044,
      "epoch": 2.3225583234480034,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004462734153380886,
      "loss": 5.0496,
      "mean_token_accuracy": 0.21225131899118424,
      "num_tokens": 53822432.0,
      "step": 23495
    },
    {
      "entropy": 5.626414775848389,
      "epoch": 2.3230525899565047,
      "grad_norm": 1.125,
      "learning_rate": 0.00044625052769967766,
      "loss": 4.9796,
      "mean_token_accuracy": 0.2215825840830803,
      "num_tokens": 53833701.0,
      "step": 23500
    },
    {
      "entropy": 5.636937665939331,
      "epoch": 2.323546856465006,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044622763584847716,
      "loss": 4.9609,
      "mean_token_accuracy": 0.21911939531564711,
      "num_tokens": 53844413.0,
      "step": 23505
    },
    {
      "entropy": 5.608329153060913,
      "epoch": 2.324041122973507,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044620473978505023,
      "loss": 4.9216,
      "mean_token_accuracy": 0.22223889976739883,
      "num_tokens": 53856392.0,
      "step": 23510
    },
    {
      "entropy": 5.712033796310425,
      "epoch": 2.324535389482009,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044618183950996017,
      "loss": 5.0053,
      "mean_token_accuracy": 0.22014512866735458,
      "num_tokens": 53867938.0,
      "step": 23515
    },
    {
      "entropy": 5.586626720428467,
      "epoch": 2.32502965599051,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004461589350237702,
      "loss": 4.8715,
      "mean_token_accuracy": 0.22319066822528838,
      "num_tokens": 53878153.0,
      "step": 23520
    },
    {
      "entropy": 5.56695442199707,
      "epoch": 2.3255239224990114,
      "grad_norm": 1.21875,
      "learning_rate": 0.000446136026327044,
      "loss": 4.9501,
      "mean_token_accuracy": 0.21799801886081696,
      "num_tokens": 53889090.0,
      "step": 23525
    },
    {
      "entropy": 5.6354179859161375,
      "epoch": 2.3260181890075127,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044611311342034494,
      "loss": 4.9755,
      "mean_token_accuracy": 0.2236956238746643,
      "num_tokens": 53900459.0,
      "step": 23530
    },
    {
      "entropy": 5.659226036071777,
      "epoch": 2.3265124555160144,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004460901963042368,
      "loss": 4.9699,
      "mean_token_accuracy": 0.2204741895198822,
      "num_tokens": 53912815.0,
      "step": 23535
    },
    {
      "entropy": 5.590457439422607,
      "epoch": 2.3270067220245156,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044606727497928324,
      "loss": 4.9324,
      "mean_token_accuracy": 0.22062502503395082,
      "num_tokens": 53924387.0,
      "step": 23540
    },
    {
      "entropy": 5.714711856842041,
      "epoch": 2.327500988533017,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044604434944604826,
      "loss": 5.0222,
      "mean_token_accuracy": 0.21449391394853592,
      "num_tokens": 53936147.0,
      "step": 23545
    },
    {
      "entropy": 5.671673679351807,
      "epoch": 2.327995255041518,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004460214197050956,
      "loss": 5.046,
      "mean_token_accuracy": 0.2137337163090706,
      "num_tokens": 53948102.0,
      "step": 23550
    },
    {
      "entropy": 5.673119926452637,
      "epoch": 2.32848952155002,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004459984857569896,
      "loss": 5.057,
      "mean_token_accuracy": 0.20863454788923264,
      "num_tokens": 53959721.0,
      "step": 23555
    },
    {
      "entropy": 5.623241710662842,
      "epoch": 2.328983788058521,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044597554760229433,
      "loss": 4.9887,
      "mean_token_accuracy": 0.21970128118991852,
      "num_tokens": 53970302.0,
      "step": 23560
    },
    {
      "entropy": 5.731866025924683,
      "epoch": 2.3294780545670224,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000445952605241574,
      "loss": 5.0808,
      "mean_token_accuracy": 0.21380094438791275,
      "num_tokens": 53982198.0,
      "step": 23565
    },
    {
      "entropy": 5.738956451416016,
      "epoch": 2.329972321075524,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044592965867539306,
      "loss": 5.1036,
      "mean_token_accuracy": 0.2059256210923195,
      "num_tokens": 53994443.0,
      "step": 23570
    },
    {
      "entropy": 5.647835969924927,
      "epoch": 2.3304665875840254,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044590670790431604,
      "loss": 4.9714,
      "mean_token_accuracy": 0.21688535660505295,
      "num_tokens": 54005666.0,
      "step": 23575
    },
    {
      "entropy": 5.64765567779541,
      "epoch": 2.3309608540925266,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004458837529289074,
      "loss": 4.9358,
      "mean_token_accuracy": 0.22210220396518707,
      "num_tokens": 54016173.0,
      "step": 23580
    },
    {
      "entropy": 5.601264095306396,
      "epoch": 2.331455120601028,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004458607937497321,
      "loss": 5.0732,
      "mean_token_accuracy": 0.20985782891511917,
      "num_tokens": 54027743.0,
      "step": 23585
    },
    {
      "entropy": 5.724059152603149,
      "epoch": 2.3319493871095296,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004458378303673546,
      "loss": 5.0903,
      "mean_token_accuracy": 0.20857136994600295,
      "num_tokens": 54039804.0,
      "step": 23590
    },
    {
      "entropy": 5.691845417022705,
      "epoch": 2.332443653618031,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004458148627823401,
      "loss": 5.0481,
      "mean_token_accuracy": 0.21689510196447373,
      "num_tokens": 54050202.0,
      "step": 23595
    },
    {
      "entropy": 5.6657263278961185,
      "epoch": 2.332937920126532,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044579189099525334,
      "loss": 4.9699,
      "mean_token_accuracy": 0.22159582376480103,
      "num_tokens": 54061088.0,
      "step": 23600
    },
    {
      "entropy": 5.572190523147583,
      "epoch": 2.333432186635034,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044576891500665967,
      "loss": 4.9267,
      "mean_token_accuracy": 0.2253502279520035,
      "num_tokens": 54072524.0,
      "step": 23605
    },
    {
      "entropy": 5.638398504257202,
      "epoch": 2.333926453143535,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004457459348171242,
      "loss": 5.0064,
      "mean_token_accuracy": 0.21764014661312103,
      "num_tokens": 54083265.0,
      "step": 23610
    },
    {
      "entropy": 5.6630456924438475,
      "epoch": 2.3344207196520363,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044572295042721215,
      "loss": 4.9368,
      "mean_token_accuracy": 0.22668247520923615,
      "num_tokens": 54093848.0,
      "step": 23615
    },
    {
      "entropy": 5.609566688537598,
      "epoch": 2.3349149861605376,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044569996183748913,
      "loss": 4.9565,
      "mean_token_accuracy": 0.23079534918069838,
      "num_tokens": 54104595.0,
      "step": 23620
    },
    {
      "entropy": 5.561442708969116,
      "epoch": 2.3354092526690393,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044567696904852057,
      "loss": 4.888,
      "mean_token_accuracy": 0.23076927959918975,
      "num_tokens": 54115577.0,
      "step": 23625
    },
    {
      "entropy": 5.648702239990234,
      "epoch": 2.3359035191775406,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000445653972060872,
      "loss": 4.9591,
      "mean_token_accuracy": 0.2203268140554428,
      "num_tokens": 54125579.0,
      "step": 23630
    },
    {
      "entropy": 5.656605386734009,
      "epoch": 2.336397785686042,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044563097087510934,
      "loss": 5.0627,
      "mean_token_accuracy": 0.2127949818968773,
      "num_tokens": 54137895.0,
      "step": 23635
    },
    {
      "entropy": 5.701214599609375,
      "epoch": 2.3368920521945435,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044560796549179826,
      "loss": 5.0682,
      "mean_token_accuracy": 0.21045654118061066,
      "num_tokens": 54149423.0,
      "step": 23640
    },
    {
      "entropy": 5.773826313018799,
      "epoch": 2.337386318703045,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004455849559115048,
      "loss": 5.0396,
      "mean_token_accuracy": 0.21631911545991897,
      "num_tokens": 54160743.0,
      "step": 23645
    },
    {
      "entropy": 5.683167362213135,
      "epoch": 2.337880585211546,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004455619421347949,
      "loss": 5.0643,
      "mean_token_accuracy": 0.20993875861167907,
      "num_tokens": 54171517.0,
      "step": 23650
    },
    {
      "entropy": 5.645554542541504,
      "epoch": 2.3383748517200473,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044553892416223485,
      "loss": 4.9375,
      "mean_token_accuracy": 0.22406360507011414,
      "num_tokens": 54182959.0,
      "step": 23655
    },
    {
      "entropy": 5.669104433059692,
      "epoch": 2.3388691182285486,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044551590199439075,
      "loss": 5.0388,
      "mean_token_accuracy": 0.2129897266626358,
      "num_tokens": 54194577.0,
      "step": 23660
    },
    {
      "entropy": 5.693298959732056,
      "epoch": 2.3393633847370503,
      "grad_norm": 1.046875,
      "learning_rate": 0.000445492875631829,
      "loss": 5.1436,
      "mean_token_accuracy": 0.20494206696748735,
      "num_tokens": 54207423.0,
      "step": 23665
    },
    {
      "entropy": 5.668057632446289,
      "epoch": 2.3398576512455516,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004454698450751161,
      "loss": 4.997,
      "mean_token_accuracy": 0.2173857346177101,
      "num_tokens": 54218776.0,
      "step": 23670
    },
    {
      "entropy": 5.700232219696045,
      "epoch": 2.340351917754053,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004454468103248186,
      "loss": 4.9257,
      "mean_token_accuracy": 0.2222027063369751,
      "num_tokens": 54229508.0,
      "step": 23675
    },
    {
      "entropy": 5.682994985580445,
      "epoch": 2.3408461842625545,
      "grad_norm": 1.0625,
      "learning_rate": 0.000445423771381503,
      "loss": 5.0962,
      "mean_token_accuracy": 0.20942182540893556,
      "num_tokens": 54240546.0,
      "step": 23680
    },
    {
      "entropy": 5.630633687973022,
      "epoch": 2.341340450771056,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004454007282457364,
      "loss": 5.0337,
      "mean_token_accuracy": 0.2196759507060051,
      "num_tokens": 54252486.0,
      "step": 23685
    },
    {
      "entropy": 5.723714447021484,
      "epoch": 2.341834717279557,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044537768091808516,
      "loss": 5.0117,
      "mean_token_accuracy": 0.2188166096806526,
      "num_tokens": 54264240.0,
      "step": 23690
    },
    {
      "entropy": 5.703940200805664,
      "epoch": 2.3423289837880583,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004453546293991168,
      "loss": 4.9894,
      "mean_token_accuracy": 0.22072886675596237,
      "num_tokens": 54275894.0,
      "step": 23695
    },
    {
      "entropy": 5.685349178314209,
      "epoch": 2.34282325029656,
      "grad_norm": 1.0625,
      "learning_rate": 0.000445331573689398,
      "loss": 5.0326,
      "mean_token_accuracy": 0.21345318853855133,
      "num_tokens": 54287697.0,
      "step": 23700
    },
    {
      "entropy": 5.60703935623169,
      "epoch": 2.3433175168050613,
      "grad_norm": 1.15625,
      "learning_rate": 0.000445308513789496,
      "loss": 4.9336,
      "mean_token_accuracy": 0.22863015681505203,
      "num_tokens": 54300327.0,
      "step": 23705
    },
    {
      "entropy": 5.700958299636841,
      "epoch": 2.3438117833135625,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044528544969997823,
      "loss": 5.0808,
      "mean_token_accuracy": 0.21131146103143691,
      "num_tokens": 54311949.0,
      "step": 23710
    },
    {
      "entropy": 5.658927869796753,
      "epoch": 2.3443060498220643,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044526238142141196,
      "loss": 4.9735,
      "mean_token_accuracy": 0.2162339985370636,
      "num_tokens": 54322134.0,
      "step": 23715
    },
    {
      "entropy": 5.702368259429932,
      "epoch": 2.3448003163305655,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044523930895436475,
      "loss": 5.0888,
      "mean_token_accuracy": 0.20286719352006913,
      "num_tokens": 54334579.0,
      "step": 23720
    },
    {
      "entropy": 5.670440196990967,
      "epoch": 2.3452945828390668,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044521623229940407,
      "loss": 5.0098,
      "mean_token_accuracy": 0.21451695561408995,
      "num_tokens": 54346624.0,
      "step": 23725
    },
    {
      "entropy": 5.649648237228393,
      "epoch": 2.345788849347568,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044519315145709765,
      "loss": 4.976,
      "mean_token_accuracy": 0.21944692879915237,
      "num_tokens": 54356455.0,
      "step": 23730
    },
    {
      "entropy": 5.696085119247437,
      "epoch": 2.3462831158560697,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004451700664280133,
      "loss": 5.1123,
      "mean_token_accuracy": 0.20624695271253585,
      "num_tokens": 54368141.0,
      "step": 23735
    },
    {
      "entropy": 5.689579248428345,
      "epoch": 2.346777382364571,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000445146977212719,
      "loss": 5.0373,
      "mean_token_accuracy": 0.213021881878376,
      "num_tokens": 54379304.0,
      "step": 23740
    },
    {
      "entropy": 5.657140445709229,
      "epoch": 2.3472716488730723,
      "grad_norm": 1.125,
      "learning_rate": 0.0004451238838117826,
      "loss": 4.9722,
      "mean_token_accuracy": 0.22089823484420776,
      "num_tokens": 54391462.0,
      "step": 23745
    },
    {
      "entropy": 5.629985094070435,
      "epoch": 2.347765915381574,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004451007862257723,
      "loss": 4.9447,
      "mean_token_accuracy": 0.22952927350997926,
      "num_tokens": 54402325.0,
      "step": 23750
    },
    {
      "entropy": 5.604015254974366,
      "epoch": 2.3482601818900752,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004450776844552562,
      "loss": 4.9186,
      "mean_token_accuracy": 0.21776869148015976,
      "num_tokens": 54413874.0,
      "step": 23755
    },
    {
      "entropy": 5.632415914535523,
      "epoch": 2.3487544483985765,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004450545785008027,
      "loss": 5.0363,
      "mean_token_accuracy": 0.2186850130558014,
      "num_tokens": 54426266.0,
      "step": 23760
    },
    {
      "entropy": 5.6446596622467045,
      "epoch": 2.3492487149070778,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044503146836298026,
      "loss": 5.0133,
      "mean_token_accuracy": 0.21799371540546417,
      "num_tokens": 54436682.0,
      "step": 23765
    },
    {
      "entropy": 5.658512258529663,
      "epoch": 2.3497429814155795,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044500835404235727,
      "loss": 4.9994,
      "mean_token_accuracy": 0.2183884397149086,
      "num_tokens": 54447462.0,
      "step": 23770
    },
    {
      "entropy": 5.726205682754516,
      "epoch": 2.3502372479240807,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004449852355395024,
      "loss": 5.0401,
      "mean_token_accuracy": 0.22163018137216567,
      "num_tokens": 54460116.0,
      "step": 23775
    },
    {
      "entropy": 5.715434265136719,
      "epoch": 2.350731514432582,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044496211285498434,
      "loss": 5.004,
      "mean_token_accuracy": 0.22352475076913833,
      "num_tokens": 54471016.0,
      "step": 23780
    },
    {
      "entropy": 5.62493371963501,
      "epoch": 2.3512257809410833,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000444938985989372,
      "loss": 4.9289,
      "mean_token_accuracy": 0.22665604054927826,
      "num_tokens": 54482410.0,
      "step": 23785
    },
    {
      "entropy": 5.599335050582885,
      "epoch": 2.351720047449585,
      "grad_norm": 1.125,
      "learning_rate": 0.00044491585494323415,
      "loss": 4.9813,
      "mean_token_accuracy": 0.21639845222234727,
      "num_tokens": 54495555.0,
      "step": 23790
    },
    {
      "entropy": 5.721944856643677,
      "epoch": 2.3522143139580862,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044489271971713995,
      "loss": 5.0256,
      "mean_token_accuracy": 0.2180335521697998,
      "num_tokens": 54505658.0,
      "step": 23795
    },
    {
      "entropy": 5.575951004028321,
      "epoch": 2.3527085804665875,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044486958031165856,
      "loss": 4.9517,
      "mean_token_accuracy": 0.22421502321958542,
      "num_tokens": 54518633.0,
      "step": 23800
    },
    {
      "entropy": 5.714538192749023,
      "epoch": 2.3532028469750887,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044484643672735907,
      "loss": 5.1438,
      "mean_token_accuracy": 0.2094655692577362,
      "num_tokens": 54531451.0,
      "step": 23805
    },
    {
      "entropy": 5.6757584571838375,
      "epoch": 2.3536971134835905,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004448232889648109,
      "loss": 4.9542,
      "mean_token_accuracy": 0.2257176533341408,
      "num_tokens": 54542798.0,
      "step": 23810
    },
    {
      "entropy": 5.639838695526123,
      "epoch": 2.3541913799920917,
      "grad_norm": 1.125,
      "learning_rate": 0.0004448001370245835,
      "loss": 5.057,
      "mean_token_accuracy": 0.2142079919576645,
      "num_tokens": 54554551.0,
      "step": 23815
    },
    {
      "entropy": 5.697571659088135,
      "epoch": 2.354685646500593,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044477698090724635,
      "loss": 5.0607,
      "mean_token_accuracy": 0.21112409085035325,
      "num_tokens": 54564711.0,
      "step": 23820
    },
    {
      "entropy": 5.709876680374146,
      "epoch": 2.3551799130090947,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044475382061336913,
      "loss": 5.0627,
      "mean_token_accuracy": 0.2137662723660469,
      "num_tokens": 54576096.0,
      "step": 23825
    },
    {
      "entropy": 5.665984678268432,
      "epoch": 2.355674179517596,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004447306561435215,
      "loss": 5.0475,
      "mean_token_accuracy": 0.21065954566001893,
      "num_tokens": 54587939.0,
      "step": 23830
    },
    {
      "entropy": 5.699896287918091,
      "epoch": 2.356168446026097,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044470748749827354,
      "loss": 4.9516,
      "mean_token_accuracy": 0.2186424434185028,
      "num_tokens": 54598470.0,
      "step": 23835
    },
    {
      "entropy": 5.708815240859986,
      "epoch": 2.3566627125345985,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044468431467819505,
      "loss": 5.007,
      "mean_token_accuracy": 0.21999007910490037,
      "num_tokens": 54608840.0,
      "step": 23840
    },
    {
      "entropy": 5.669919538497925,
      "epoch": 2.3571569790431,
      "grad_norm": 1.140625,
      "learning_rate": 0.000444661137683856,
      "loss": 4.9963,
      "mean_token_accuracy": 0.21669238656759263,
      "num_tokens": 54619319.0,
      "step": 23845
    },
    {
      "entropy": 5.633040189743042,
      "epoch": 2.3576512455516014,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004446379565158268,
      "loss": 4.9947,
      "mean_token_accuracy": 0.2215055838227272,
      "num_tokens": 54630410.0,
      "step": 23850
    },
    {
      "entropy": 5.717750406265258,
      "epoch": 2.3581455120601027,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004446147711746774,
      "loss": 5.1063,
      "mean_token_accuracy": 0.20812556147575378,
      "num_tokens": 54642376.0,
      "step": 23855
    },
    {
      "entropy": 5.681838417053223,
      "epoch": 2.3586397785686044,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004445915816609784,
      "loss": 4.9934,
      "mean_token_accuracy": 0.21400044113397598,
      "num_tokens": 54654378.0,
      "step": 23860
    },
    {
      "entropy": 5.651501703262329,
      "epoch": 2.3591340450771057,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044456838797530006,
      "loss": 4.9813,
      "mean_token_accuracy": 0.21630917638540267,
      "num_tokens": 54664879.0,
      "step": 23865
    },
    {
      "entropy": 5.640066814422608,
      "epoch": 2.359628311585607,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004445451901182132,
      "loss": 5.0012,
      "mean_token_accuracy": 0.21393912732601167,
      "num_tokens": 54677251.0,
      "step": 23870
    },
    {
      "entropy": 5.6589438915252686,
      "epoch": 2.360122578094108,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004445219880902884,
      "loss": 4.9709,
      "mean_token_accuracy": 0.21162965893745422,
      "num_tokens": 54688827.0,
      "step": 23875
    },
    {
      "entropy": 5.613009691238403,
      "epoch": 2.36061684460261,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004444987818920963,
      "loss": 5.0434,
      "mean_token_accuracy": 0.21211196333169938,
      "num_tokens": 54701046.0,
      "step": 23880
    },
    {
      "entropy": 5.723282146453857,
      "epoch": 2.361111111111111,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004444755715242078,
      "loss": 5.0838,
      "mean_token_accuracy": 0.20903788954019548,
      "num_tokens": 54713254.0,
      "step": 23885
    },
    {
      "entropy": 5.6945147037506105,
      "epoch": 2.3616053776196124,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044445235698719406,
      "loss": 4.9467,
      "mean_token_accuracy": 0.22054114192724228,
      "num_tokens": 54723734.0,
      "step": 23890
    },
    {
      "entropy": 5.712226533889771,
      "epoch": 2.362099644128114,
      "grad_norm": 1.140625,
      "learning_rate": 0.000444429138281626,
      "loss": 5.0303,
      "mean_token_accuracy": 0.21615202277898787,
      "num_tokens": 54735080.0,
      "step": 23895
    },
    {
      "entropy": 5.598332309722901,
      "epoch": 2.3625939106366154,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044440591540807496,
      "loss": 4.9297,
      "mean_token_accuracy": 0.2242228150367737,
      "num_tokens": 54745164.0,
      "step": 23900
    },
    {
      "entropy": 5.6097643852233885,
      "epoch": 2.3630881771451167,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000444382688367112,
      "loss": 5.0331,
      "mean_token_accuracy": 0.21686645448207856,
      "num_tokens": 54757350.0,
      "step": 23905
    },
    {
      "entropy": 5.641051244735718,
      "epoch": 2.363582443653618,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044435945715930864,
      "loss": 4.9195,
      "mean_token_accuracy": 0.22130585908889772,
      "num_tokens": 54768290.0,
      "step": 23910
    },
    {
      "entropy": 5.699190378189087,
      "epoch": 2.364076710162119,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004443362217852364,
      "loss": 5.0611,
      "mean_token_accuracy": 0.2084745392203331,
      "num_tokens": 54779053.0,
      "step": 23915
    },
    {
      "entropy": 5.779437780380249,
      "epoch": 2.364570976670621,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004443129822454668,
      "loss": 5.1514,
      "mean_token_accuracy": 0.2049174815416336,
      "num_tokens": 54789787.0,
      "step": 23920
    },
    {
      "entropy": 5.718475675582885,
      "epoch": 2.365065243179122,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044428973854057164,
      "loss": 5.0492,
      "mean_token_accuracy": 0.2186028927564621,
      "num_tokens": 54802036.0,
      "step": 23925
    },
    {
      "entropy": 5.681513977050781,
      "epoch": 2.3655595096876234,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044426649067112257,
      "loss": 5.0329,
      "mean_token_accuracy": 0.21332904398441316,
      "num_tokens": 54811765.0,
      "step": 23930
    },
    {
      "entropy": 5.630385494232177,
      "epoch": 2.366053776196125,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004442432386376916,
      "loss": 4.9873,
      "mean_token_accuracy": 0.22059177309274675,
      "num_tokens": 54823057.0,
      "step": 23935
    },
    {
      "entropy": 5.65553994178772,
      "epoch": 2.3665480427046264,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044421998244085066,
      "loss": 5.012,
      "mean_token_accuracy": 0.21475279778242112,
      "num_tokens": 54834070.0,
      "step": 23940
    },
    {
      "entropy": 5.628724002838135,
      "epoch": 2.3670423092131276,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044419672208117204,
      "loss": 5.0118,
      "mean_token_accuracy": 0.21337459087371827,
      "num_tokens": 54845913.0,
      "step": 23945
    },
    {
      "entropy": 5.670712232589722,
      "epoch": 2.367536575721629,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004441734575592276,
      "loss": 5.0297,
      "mean_token_accuracy": 0.21595293581485747,
      "num_tokens": 54856721.0,
      "step": 23950
    },
    {
      "entropy": 5.7325849533081055,
      "epoch": 2.3680308422301306,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044415018887559,
      "loss": 5.0404,
      "mean_token_accuracy": 0.2144642159342766,
      "num_tokens": 54868652.0,
      "step": 23955
    },
    {
      "entropy": 5.637701940536499,
      "epoch": 2.368525108738632,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004441269160308315,
      "loss": 4.9334,
      "mean_token_accuracy": 0.22737201899290085,
      "num_tokens": 54880753.0,
      "step": 23960
    },
    {
      "entropy": 5.570465612411499,
      "epoch": 2.369019375247133,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044410363902552466,
      "loss": 4.9095,
      "mean_token_accuracy": 0.23207402974367142,
      "num_tokens": 54891724.0,
      "step": 23965
    },
    {
      "entropy": 5.697908544540406,
      "epoch": 2.369513641755635,
      "grad_norm": 1.125,
      "learning_rate": 0.000444080357860242,
      "loss": 5.0384,
      "mean_token_accuracy": 0.21368204802274704,
      "num_tokens": 54902516.0,
      "step": 23970
    },
    {
      "entropy": 5.6927961826324465,
      "epoch": 2.370007908264136,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044405707253555626,
      "loss": 5.0506,
      "mean_token_accuracy": 0.2173882320523262,
      "num_tokens": 54914392.0,
      "step": 23975
    },
    {
      "entropy": 5.6504237174987795,
      "epoch": 2.3705021747726374,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004440337830520404,
      "loss": 4.9179,
      "mean_token_accuracy": 0.22457821816205978,
      "num_tokens": 54925209.0,
      "step": 23980
    },
    {
      "entropy": 5.676028394699097,
      "epoch": 2.3709964412811386,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044401048941026725,
      "loss": 4.9438,
      "mean_token_accuracy": 0.21309418082237244,
      "num_tokens": 54935972.0,
      "step": 23985
    },
    {
      "entropy": 5.641994619369507,
      "epoch": 2.3714907077896403,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004439871916108098,
      "loss": 4.936,
      "mean_token_accuracy": 0.22319420129060746,
      "num_tokens": 54945631.0,
      "step": 23990
    },
    {
      "entropy": 5.651434469223022,
      "epoch": 2.3719849742981416,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044396388965424124,
      "loss": 5.0168,
      "mean_token_accuracy": 0.21790778189897536,
      "num_tokens": 54955734.0,
      "step": 23995
    },
    {
      "entropy": 5.67791519165039,
      "epoch": 2.372479240806643,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004439405835411348,
      "loss": 5.0563,
      "mean_token_accuracy": 0.2126620590686798,
      "num_tokens": 54968506.0,
      "step": 24000
    },
    {
      "epoch": 2.372479240806643,
      "eval_entropy": 5.495205940795561,
      "eval_loss": 5.071153163909912,
      "eval_mean_token_accuracy": 0.22140243299222961,
      "eval_num_tokens": 54968506.0,
      "eval_runtime": 20.5129,
      "eval_samples_per_second": 1585.002,
      "eval_steps_per_second": 198.168,
      "step": 24000
    },
    {
      "entropy": 5.71533899307251,
      "epoch": 2.3729735073151446,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004439172732720637,
      "loss": 4.9582,
      "mean_token_accuracy": 0.22417230010032654,
      "num_tokens": 54979809.0,
      "step": 24005
    },
    {
      "entropy": 5.727009201049805,
      "epoch": 2.373467773823646,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004438939588476016,
      "loss": 5.1081,
      "mean_token_accuracy": 0.21223565191030502,
      "num_tokens": 54991276.0,
      "step": 24010
    },
    {
      "entropy": 5.665151500701905,
      "epoch": 2.373962040332147,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044387064026832183,
      "loss": 5.0002,
      "mean_token_accuracy": 0.2188743531703949,
      "num_tokens": 55002320.0,
      "step": 24015
    },
    {
      "entropy": 5.561293268203736,
      "epoch": 2.3744563068406483,
      "grad_norm": 1.125,
      "learning_rate": 0.00044384731753479813,
      "loss": 4.8679,
      "mean_token_accuracy": 0.23687697649002076,
      "num_tokens": 55012995.0,
      "step": 24020
    },
    {
      "entropy": 5.645366764068603,
      "epoch": 2.37495057334915,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044382399064760424,
      "loss": 5.0469,
      "mean_token_accuracy": 0.21571362167596816,
      "num_tokens": 55025342.0,
      "step": 24025
    },
    {
      "entropy": 5.602294635772705,
      "epoch": 2.3754448398576513,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000443800659607314,
      "loss": 5.0054,
      "mean_token_accuracy": 0.22263902872800828,
      "num_tokens": 55036925.0,
      "step": 24030
    },
    {
      "entropy": 5.722566080093384,
      "epoch": 2.3759391063661526,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004437773244145013,
      "loss": 5.0077,
      "mean_token_accuracy": 0.20871272683143616,
      "num_tokens": 55048913.0,
      "step": 24035
    },
    {
      "entropy": 5.665289068222046,
      "epoch": 2.376433372874654,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004437539850697403,
      "loss": 4.9962,
      "mean_token_accuracy": 0.21763528138399124,
      "num_tokens": 55059654.0,
      "step": 24040
    },
    {
      "entropy": 5.740592002868652,
      "epoch": 2.3769276393831555,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044373064157360504,
      "loss": 5.0654,
      "mean_token_accuracy": 0.21257323473691941,
      "num_tokens": 55071077.0,
      "step": 24045
    },
    {
      "entropy": 5.709898567199707,
      "epoch": 2.377421905891657,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004437072939266698,
      "loss": 5.0051,
      "mean_token_accuracy": 0.2170957773923874,
      "num_tokens": 55082659.0,
      "step": 24050
    },
    {
      "entropy": 5.6168194770812985,
      "epoch": 2.377916172400158,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044368394212950896,
      "loss": 5.0642,
      "mean_token_accuracy": 0.21388001441955568,
      "num_tokens": 55093938.0,
      "step": 24055
    },
    {
      "entropy": 5.714694356918335,
      "epoch": 2.3784104389086593,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044366058618269707,
      "loss": 5.0343,
      "mean_token_accuracy": 0.21535336524248122,
      "num_tokens": 55105491.0,
      "step": 24060
    },
    {
      "entropy": 5.63551139831543,
      "epoch": 2.378904705417161,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004436372260868085,
      "loss": 4.8939,
      "mean_token_accuracy": 0.22774088829755784,
      "num_tokens": 55116941.0,
      "step": 24065
    },
    {
      "entropy": 5.7483419418334964,
      "epoch": 2.3793989719256623,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044361386184241805,
      "loss": 5.0487,
      "mean_token_accuracy": 0.21291780322790146,
      "num_tokens": 55128294.0,
      "step": 24070
    },
    {
      "entropy": 5.75245213508606,
      "epoch": 2.3798932384341636,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044359049345010044,
      "loss": 5.0508,
      "mean_token_accuracy": 0.21437966227531433,
      "num_tokens": 55139781.0,
      "step": 24075
    },
    {
      "entropy": 5.680056238174439,
      "epoch": 2.3803875049426653,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004435671209104305,
      "loss": 4.9903,
      "mean_token_accuracy": 0.21878640949726105,
      "num_tokens": 55151517.0,
      "step": 24080
    },
    {
      "entropy": 5.594570970535278,
      "epoch": 2.3808817714511665,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004435437442239833,
      "loss": 4.9795,
      "mean_token_accuracy": 0.22098229974508285,
      "num_tokens": 55163029.0,
      "step": 24085
    },
    {
      "entropy": 5.5961638450622555,
      "epoch": 2.381376037959668,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004435203633913337,
      "loss": 4.9403,
      "mean_token_accuracy": 0.22853411138057708,
      "num_tokens": 55174262.0,
      "step": 24090
    },
    {
      "entropy": 5.643753385543823,
      "epoch": 2.381870304468169,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044349697841305715,
      "loss": 4.9852,
      "mean_token_accuracy": 0.22451517283916472,
      "num_tokens": 55185047.0,
      "step": 24095
    },
    {
      "entropy": 5.673589468002319,
      "epoch": 2.3823645709766708,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044347358928972883,
      "loss": 4.9989,
      "mean_token_accuracy": 0.21840925216674806,
      "num_tokens": 55196396.0,
      "step": 24100
    },
    {
      "entropy": 5.670108318328857,
      "epoch": 2.382858837485172,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004434501960219239,
      "loss": 5.0419,
      "mean_token_accuracy": 0.21966529339551927,
      "num_tokens": 55208774.0,
      "step": 24105
    },
    {
      "entropy": 5.695929670333863,
      "epoch": 2.3833531039936733,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004434267986102181,
      "loss": 5.0177,
      "mean_token_accuracy": 0.20808807462453843,
      "num_tokens": 55219946.0,
      "step": 24110
    },
    {
      "entropy": 5.673986101150513,
      "epoch": 2.383847370502175,
      "grad_norm": 1.171875,
      "learning_rate": 0.000443403397055187,
      "loss": 5.0752,
      "mean_token_accuracy": 0.21460306644439697,
      "num_tokens": 55231926.0,
      "step": 24115
    },
    {
      "entropy": 5.635173320770264,
      "epoch": 2.3843416370106763,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044337999135740614,
      "loss": 4.9659,
      "mean_token_accuracy": 0.21592362374067306,
      "num_tokens": 55245240.0,
      "step": 24120
    },
    {
      "entropy": 5.733977317810059,
      "epoch": 2.3848359035191775,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004433565815174514,
      "loss": 5.0389,
      "mean_token_accuracy": 0.21592750251293183,
      "num_tokens": 55256862.0,
      "step": 24125
    },
    {
      "entropy": 5.653119230270386,
      "epoch": 2.3853301700276788,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044333316753589864,
      "loss": 5.0002,
      "mean_token_accuracy": 0.2151127353310585,
      "num_tokens": 55268157.0,
      "step": 24130
    },
    {
      "entropy": 5.680840682983399,
      "epoch": 2.3858244365361805,
      "grad_norm": 1.125,
      "learning_rate": 0.0004433097494133238,
      "loss": 5.0645,
      "mean_token_accuracy": 0.21176636070013047,
      "num_tokens": 55279580.0,
      "step": 24135
    },
    {
      "entropy": 5.763088130950928,
      "epoch": 2.3863187030446817,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044328632715030303,
      "loss": 5.2072,
      "mean_token_accuracy": 0.19810367822647096,
      "num_tokens": 55292067.0,
      "step": 24140
    },
    {
      "entropy": 5.669901418685913,
      "epoch": 2.386812969553183,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044326290074741257,
      "loss": 4.9832,
      "mean_token_accuracy": 0.2228647217154503,
      "num_tokens": 55303338.0,
      "step": 24145
    },
    {
      "entropy": 5.675422048568725,
      "epoch": 2.3873072360616843,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004432394702052286,
      "loss": 5.0609,
      "mean_token_accuracy": 0.2061280280351639,
      "num_tokens": 55315165.0,
      "step": 24150
    },
    {
      "entropy": 5.703721380233764,
      "epoch": 2.387801502570186,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044321603552432753,
      "loss": 5.0451,
      "mean_token_accuracy": 0.20892222225666046,
      "num_tokens": 55327928.0,
      "step": 24155
    },
    {
      "entropy": 5.652971887588501,
      "epoch": 2.3882957690786872,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044319259670528595,
      "loss": 4.9635,
      "mean_token_accuracy": 0.22193489521741866,
      "num_tokens": 55338438.0,
      "step": 24160
    },
    {
      "entropy": 5.613698482513428,
      "epoch": 2.3887900355871885,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044316915374868033,
      "loss": 4.9408,
      "mean_token_accuracy": 0.22139861881732942,
      "num_tokens": 55349513.0,
      "step": 24165
    },
    {
      "entropy": 5.621834754943848,
      "epoch": 2.3892843020956898,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044314570665508746,
      "loss": 5.0477,
      "mean_token_accuracy": 0.21473981589078903,
      "num_tokens": 55362498.0,
      "step": 24170
    },
    {
      "entropy": 5.698979997634888,
      "epoch": 2.3897785686041915,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004431222554250841,
      "loss": 5.0195,
      "mean_token_accuracy": 0.21215529441833497,
      "num_tokens": 55374065.0,
      "step": 24175
    },
    {
      "entropy": 5.660956478118896,
      "epoch": 2.3902728351126927,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044309880005924726,
      "loss": 4.9878,
      "mean_token_accuracy": 0.22342018634080887,
      "num_tokens": 55385175.0,
      "step": 24180
    },
    {
      "entropy": 5.728198623657226,
      "epoch": 2.390767101621194,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044307534055815375,
      "loss": 5.0876,
      "mean_token_accuracy": 0.20970371067523957,
      "num_tokens": 55395457.0,
      "step": 24185
    },
    {
      "entropy": 5.683646154403687,
      "epoch": 2.3912613681296957,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044305187692238087,
      "loss": 5.0155,
      "mean_token_accuracy": 0.2163003697991371,
      "num_tokens": 55406341.0,
      "step": 24190
    },
    {
      "entropy": 5.695096635818482,
      "epoch": 2.391755634638197,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044302840915250575,
      "loss": 4.9782,
      "mean_token_accuracy": 0.2140378847718239,
      "num_tokens": 55417893.0,
      "step": 24195
    },
    {
      "entropy": 5.685885334014893,
      "epoch": 2.392249901146698,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004430049372491056,
      "loss": 5.0688,
      "mean_token_accuracy": 0.2129547417163849,
      "num_tokens": 55428286.0,
      "step": 24200
    },
    {
      "entropy": 5.575006437301636,
      "epoch": 2.3927441676551995,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044298146121275804,
      "loss": 4.9365,
      "mean_token_accuracy": 0.22150723934173583,
      "num_tokens": 55439861.0,
      "step": 24205
    },
    {
      "entropy": 5.7483823776245115,
      "epoch": 2.393238434163701,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004429579810440404,
      "loss": 5.021,
      "mean_token_accuracy": 0.20894015282392503,
      "num_tokens": 55450971.0,
      "step": 24210
    },
    {
      "entropy": 5.7034608840942385,
      "epoch": 2.3937327006722025,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004429344967435305,
      "loss": 4.9999,
      "mean_token_accuracy": 0.20725676268339158,
      "num_tokens": 55460584.0,
      "step": 24215
    },
    {
      "entropy": 5.758437728881836,
      "epoch": 2.3942269671807037,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044291100831180577,
      "loss": 5.1205,
      "mean_token_accuracy": 0.20620004683732987,
      "num_tokens": 55471576.0,
      "step": 24220
    },
    {
      "entropy": 5.714835214614868,
      "epoch": 2.3947212336892054,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004428875157494442,
      "loss": 5.0255,
      "mean_token_accuracy": 0.22211101055145263,
      "num_tokens": 55483807.0,
      "step": 24225
    },
    {
      "entropy": 5.667826223373413,
      "epoch": 2.3952155001977067,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044286401905702376,
      "loss": 5.0163,
      "mean_token_accuracy": 0.21918080300092696,
      "num_tokens": 55495130.0,
      "step": 24230
    },
    {
      "entropy": 5.6276264667510985,
      "epoch": 2.395709766706208,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004428405182351224,
      "loss": 4.9821,
      "mean_token_accuracy": 0.22084650844335557,
      "num_tokens": 55506460.0,
      "step": 24235
    },
    {
      "entropy": 5.731813430786133,
      "epoch": 2.396204033214709,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004428170132843182,
      "loss": 5.0826,
      "mean_token_accuracy": 0.2139906570315361,
      "num_tokens": 55518401.0,
      "step": 24240
    },
    {
      "entropy": 5.707881593704224,
      "epoch": 2.396698299723211,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044279350420518956,
      "loss": 4.9812,
      "mean_token_accuracy": 0.22103473246097566,
      "num_tokens": 55530403.0,
      "step": 24245
    },
    {
      "entropy": 5.715621566772461,
      "epoch": 2.397192566231712,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044276999099831463,
      "loss": 5.0247,
      "mean_token_accuracy": 0.21752341240644454,
      "num_tokens": 55541422.0,
      "step": 24250
    },
    {
      "entropy": 5.670212650299073,
      "epoch": 2.3976868327402134,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044274647366427186,
      "loss": 4.9733,
      "mean_token_accuracy": 0.21887525618076326,
      "num_tokens": 55553798.0,
      "step": 24255
    },
    {
      "entropy": 5.688078594207764,
      "epoch": 2.398181099248715,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044272295220363977,
      "loss": 5.0155,
      "mean_token_accuracy": 0.21859420984983444,
      "num_tokens": 55565970.0,
      "step": 24260
    },
    {
      "entropy": 5.695979499816895,
      "epoch": 2.3986753657572164,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004426994266169971,
      "loss": 5.0617,
      "mean_token_accuracy": 0.21257923990488053,
      "num_tokens": 55576385.0,
      "step": 24265
    },
    {
      "entropy": 5.587963104248047,
      "epoch": 2.3991696322657177,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004426758969049225,
      "loss": 4.939,
      "mean_token_accuracy": 0.23063060343265535,
      "num_tokens": 55588752.0,
      "step": 24270
    },
    {
      "entropy": 5.612295913696289,
      "epoch": 2.399663898774219,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004426523630679949,
      "loss": 5.0014,
      "mean_token_accuracy": 0.21514350175857544,
      "num_tokens": 55599348.0,
      "step": 24275
    },
    {
      "entropy": 5.654560184478759,
      "epoch": 2.4001581652827206,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004426288251067931,
      "loss": 4.9699,
      "mean_token_accuracy": 0.21450434178113936,
      "num_tokens": 55610168.0,
      "step": 24280
    },
    {
      "entropy": 5.681350469589233,
      "epoch": 2.400652431791222,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044260528302189617,
      "loss": 5.0326,
      "mean_token_accuracy": 0.21115672290325166,
      "num_tokens": 55622487.0,
      "step": 24285
    },
    {
      "entropy": 5.7031796932220455,
      "epoch": 2.401146698299723,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004425817368138833,
      "loss": 5.0498,
      "mean_token_accuracy": 0.21588754653930664,
      "num_tokens": 55634486.0,
      "step": 24290
    },
    {
      "entropy": 5.694023323059082,
      "epoch": 2.4016409648082244,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004425581864833337,
      "loss": 5.0459,
      "mean_token_accuracy": 0.2158758357167244,
      "num_tokens": 55646642.0,
      "step": 24295
    },
    {
      "entropy": 5.602871036529541,
      "epoch": 2.402135231316726,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004425346320308267,
      "loss": 4.8949,
      "mean_token_accuracy": 0.23538341969251633,
      "num_tokens": 55657921.0,
      "step": 24300
    },
    {
      "entropy": 5.639219951629639,
      "epoch": 2.4026294978252274,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004425110734569418,
      "loss": 5.0163,
      "mean_token_accuracy": 0.2221133142709732,
      "num_tokens": 55669414.0,
      "step": 24305
    },
    {
      "entropy": 5.635937023162842,
      "epoch": 2.4031237643337287,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044248751076225853,
      "loss": 4.9571,
      "mean_token_accuracy": 0.22063554972410201,
      "num_tokens": 55681854.0,
      "step": 24310
    },
    {
      "entropy": 5.68591160774231,
      "epoch": 2.40361803084223,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044246394394735646,
      "loss": 4.9307,
      "mean_token_accuracy": 0.2212206169962883,
      "num_tokens": 55692458.0,
      "step": 24315
    },
    {
      "entropy": 5.698038959503174,
      "epoch": 2.4041122973507316,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044244037301281535,
      "loss": 5.0132,
      "mean_token_accuracy": 0.21802736669778824,
      "num_tokens": 55703676.0,
      "step": 24320
    },
    {
      "entropy": 5.684928321838379,
      "epoch": 2.404606563859233,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004424167979592152,
      "loss": 5.0129,
      "mean_token_accuracy": 0.21255262196063995,
      "num_tokens": 55714372.0,
      "step": 24325
    },
    {
      "entropy": 5.686886548995972,
      "epoch": 2.405100830367734,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004423932187871358,
      "loss": 4.9882,
      "mean_token_accuracy": 0.21610285341739655,
      "num_tokens": 55724342.0,
      "step": 24330
    },
    {
      "entropy": 5.694574356079102,
      "epoch": 2.405595096876236,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004423696354971572,
      "loss": 5.0925,
      "mean_token_accuracy": 0.20919658243656158,
      "num_tokens": 55736272.0,
      "step": 24335
    },
    {
      "entropy": 5.6718299865722654,
      "epoch": 2.406089363384737,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004423460480898596,
      "loss": 4.9731,
      "mean_token_accuracy": 0.2244173288345337,
      "num_tokens": 55747174.0,
      "step": 24340
    },
    {
      "entropy": 5.62099986076355,
      "epoch": 2.4065836298932384,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004423224565658233,
      "loss": 4.913,
      "mean_token_accuracy": 0.22494356483221054,
      "num_tokens": 55759676.0,
      "step": 24345
    },
    {
      "entropy": 5.635448694229126,
      "epoch": 2.4070778964017396,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004422988609256286,
      "loss": 5.0307,
      "mean_token_accuracy": 0.21425362378358842,
      "num_tokens": 55772140.0,
      "step": 24350
    },
    {
      "entropy": 5.645157527923584,
      "epoch": 2.4075721629102413,
      "grad_norm": 1.078125,
      "learning_rate": 0.000442275261169856,
      "loss": 4.9538,
      "mean_token_accuracy": 0.21871390640735627,
      "num_tokens": 55783639.0,
      "step": 24355
    },
    {
      "entropy": 5.768876409530639,
      "epoch": 2.4080664294187426,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044225165729908604,
      "loss": 5.0911,
      "mean_token_accuracy": 0.20983131527900695,
      "num_tokens": 55795867.0,
      "step": 24360
    },
    {
      "entropy": 5.632832050323486,
      "epoch": 2.408560695927244,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044222804931389934,
      "loss": 4.9332,
      "mean_token_accuracy": 0.2237125262618065,
      "num_tokens": 55807528.0,
      "step": 24365
    },
    {
      "entropy": 5.615286827087402,
      "epoch": 2.4090549624357456,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004422044372148766,
      "loss": 4.9563,
      "mean_token_accuracy": 0.22373713105916976,
      "num_tokens": 55817547.0,
      "step": 24370
    },
    {
      "entropy": 5.586380481719971,
      "epoch": 2.409549228944247,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044218082100259883,
      "loss": 4.9415,
      "mean_token_accuracy": 0.22736895680427552,
      "num_tokens": 55828474.0,
      "step": 24375
    },
    {
      "entropy": 5.672614288330078,
      "epoch": 2.410043495452748,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044215720067764696,
      "loss": 4.955,
      "mean_token_accuracy": 0.22223177850246428,
      "num_tokens": 55838485.0,
      "step": 24380
    },
    {
      "entropy": 5.65007438659668,
      "epoch": 2.4105377619612494,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000442133576240602,
      "loss": 5.0316,
      "mean_token_accuracy": 0.21676303893327714,
      "num_tokens": 55850219.0,
      "step": 24385
    },
    {
      "entropy": 5.635620832443237,
      "epoch": 2.411032028469751,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004421099476920451,
      "loss": 4.9112,
      "mean_token_accuracy": 0.22755282521247863,
      "num_tokens": 55860241.0,
      "step": 24390
    },
    {
      "entropy": 5.717881584167481,
      "epoch": 2.4115262949782523,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004420863150325575,
      "loss": 5.1245,
      "mean_token_accuracy": 0.20093853622674943,
      "num_tokens": 55871201.0,
      "step": 24395
    },
    {
      "entropy": 5.658941268920898,
      "epoch": 2.4120205614867536,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004420626782627208,
      "loss": 4.9322,
      "mean_token_accuracy": 0.2284773901104927,
      "num_tokens": 55881758.0,
      "step": 24400
    },
    {
      "entropy": 5.6776713848114015,
      "epoch": 2.412514827995255,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044203903738311616,
      "loss": 5.0156,
      "mean_token_accuracy": 0.2147562250494957,
      "num_tokens": 55892931.0,
      "step": 24405
    },
    {
      "entropy": 5.688210725784302,
      "epoch": 2.4130090945037566,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044201539239432534,
      "loss": 5.0018,
      "mean_token_accuracy": 0.21955032348632814,
      "num_tokens": 55903376.0,
      "step": 24410
    },
    {
      "entropy": 5.745237922668457,
      "epoch": 2.413503361012258,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044199174329693,
      "loss": 5.0514,
      "mean_token_accuracy": 0.21398818194866182,
      "num_tokens": 55913821.0,
      "step": 24415
    },
    {
      "entropy": 5.662148809432983,
      "epoch": 2.413997627520759,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044196809009151183,
      "loss": 5.0211,
      "mean_token_accuracy": 0.2126387059688568,
      "num_tokens": 55925085.0,
      "step": 24420
    },
    {
      "entropy": 5.6319129943847654,
      "epoch": 2.4144918940292603,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004419444327786527,
      "loss": 5.0225,
      "mean_token_accuracy": 0.21670317649841309,
      "num_tokens": 55937572.0,
      "step": 24425
    },
    {
      "entropy": 5.667691946029663,
      "epoch": 2.414986160537762,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004419207713589347,
      "loss": 5.0024,
      "mean_token_accuracy": 0.21289613842964172,
      "num_tokens": 55948890.0,
      "step": 24430
    },
    {
      "entropy": 5.720224857330322,
      "epoch": 2.4154804270462633,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004418971058329398,
      "loss": 5.0668,
      "mean_token_accuracy": 0.20685085654258728,
      "num_tokens": 55959319.0,
      "step": 24435
    },
    {
      "entropy": 5.696013164520264,
      "epoch": 2.4159746935547646,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044187343620125016,
      "loss": 5.0437,
      "mean_token_accuracy": 0.2142469361424446,
      "num_tokens": 55970922.0,
      "step": 24440
    },
    {
      "entropy": 5.671346664428711,
      "epoch": 2.4164689600632663,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044184976246444815,
      "loss": 5.024,
      "mean_token_accuracy": 0.21596957445144654,
      "num_tokens": 55982610.0,
      "step": 24445
    },
    {
      "entropy": 5.705566930770874,
      "epoch": 2.4169632265717675,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044182608462311606,
      "loss": 4.9343,
      "mean_token_accuracy": 0.2240268647670746,
      "num_tokens": 55994393.0,
      "step": 24450
    },
    {
      "entropy": 5.624672889709473,
      "epoch": 2.417457493080269,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044180240267783637,
      "loss": 4.9046,
      "mean_token_accuracy": 0.22389164716005325,
      "num_tokens": 56005705.0,
      "step": 24455
    },
    {
      "entropy": 5.700733327865601,
      "epoch": 2.41795175958877,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044177871662919184,
      "loss": 5.0153,
      "mean_token_accuracy": 0.21178597807884217,
      "num_tokens": 56016942.0,
      "step": 24460
    },
    {
      "entropy": 5.7080597400665285,
      "epoch": 2.4184460260972718,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044175502647776486,
      "loss": 5.0256,
      "mean_token_accuracy": 0.2120076298713684,
      "num_tokens": 56028495.0,
      "step": 24465
    },
    {
      "entropy": 5.7025776386260985,
      "epoch": 2.418940292605773,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044173133222413845,
      "loss": 5.0848,
      "mean_token_accuracy": 0.21206338852643966,
      "num_tokens": 56039654.0,
      "step": 24470
    },
    {
      "entropy": 5.664964008331299,
      "epoch": 2.4194345591142743,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004417076338688954,
      "loss": 4.9924,
      "mean_token_accuracy": 0.22174919098615647,
      "num_tokens": 56051392.0,
      "step": 24475
    },
    {
      "entropy": 5.622236728668213,
      "epoch": 2.419928825622776,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044168393141261864,
      "loss": 4.9181,
      "mean_token_accuracy": 0.2277354270219803,
      "num_tokens": 56062574.0,
      "step": 24480
    },
    {
      "entropy": 5.652249479293824,
      "epoch": 2.4204230921312773,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044166022485589136,
      "loss": 4.926,
      "mean_token_accuracy": 0.2230381190776825,
      "num_tokens": 56072637.0,
      "step": 24485
    },
    {
      "entropy": 5.641043376922608,
      "epoch": 2.4209173586397785,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044163651419929667,
      "loss": 4.958,
      "mean_token_accuracy": 0.22054929435253143,
      "num_tokens": 56083573.0,
      "step": 24490
    },
    {
      "entropy": 5.664576673507691,
      "epoch": 2.42141162514828,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044161279944341785,
      "loss": 5.0039,
      "mean_token_accuracy": 0.2141207218170166,
      "num_tokens": 56094833.0,
      "step": 24495
    },
    {
      "entropy": 5.6440824508667,
      "epoch": 2.4219058916567815,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004415890805888383,
      "loss": 4.9622,
      "mean_token_accuracy": 0.2187026172876358,
      "num_tokens": 56107750.0,
      "step": 24500
    },
    {
      "entropy": 5.65316333770752,
      "epoch": 2.4224001581652828,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044156535763614167,
      "loss": 4.9588,
      "mean_token_accuracy": 0.22000904381275177,
      "num_tokens": 56118815.0,
      "step": 24505
    },
    {
      "entropy": 5.654229021072387,
      "epoch": 2.422894424673784,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004415416305859113,
      "loss": 5.0134,
      "mean_token_accuracy": 0.212288998067379,
      "num_tokens": 56129120.0,
      "step": 24510
    },
    {
      "entropy": 5.619678735733032,
      "epoch": 2.4233886911822857,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044151789943873095,
      "loss": 4.9459,
      "mean_token_accuracy": 0.22148673683404924,
      "num_tokens": 56140293.0,
      "step": 24515
    },
    {
      "entropy": 5.746534442901611,
      "epoch": 2.423882957690787,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004414941641951845,
      "loss": 5.0538,
      "mean_token_accuracy": 0.21042693853378297,
      "num_tokens": 56151942.0,
      "step": 24520
    },
    {
      "entropy": 5.659108591079712,
      "epoch": 2.4243772241992882,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044147042485585575,
      "loss": 4.9833,
      "mean_token_accuracy": 0.21781952828168868,
      "num_tokens": 56163088.0,
      "step": 24525
    },
    {
      "entropy": 5.6389707088470455,
      "epoch": 2.4248714907077895,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004414466814213287,
      "loss": 5.018,
      "mean_token_accuracy": 0.21804199665784835,
      "num_tokens": 56174879.0,
      "step": 24530
    },
    {
      "entropy": 5.715752124786377,
      "epoch": 2.425365757216291,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004414229338921875,
      "loss": 5.0431,
      "mean_token_accuracy": 0.217880317568779,
      "num_tokens": 56187407.0,
      "step": 24535
    },
    {
      "entropy": 5.6570347309112545,
      "epoch": 2.4258600237247925,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044139918226901627,
      "loss": 4.9189,
      "mean_token_accuracy": 0.22698651999235153,
      "num_tokens": 56198830.0,
      "step": 24540
    },
    {
      "entropy": 5.746195363998413,
      "epoch": 2.4263542902332937,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044137542655239945,
      "loss": 5.0939,
      "mean_token_accuracy": 0.20447674244642258,
      "num_tokens": 56210779.0,
      "step": 24545
    },
    {
      "entropy": 5.647099781036377,
      "epoch": 2.426848556741795,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004413516667429212,
      "loss": 4.9903,
      "mean_token_accuracy": 0.21783885806798936,
      "num_tokens": 56222258.0,
      "step": 24550
    },
    {
      "entropy": 5.606891298294068,
      "epoch": 2.4273428232502967,
      "grad_norm": 1.25,
      "learning_rate": 0.0004413279028411662,
      "loss": 4.9558,
      "mean_token_accuracy": 0.22344013154506684,
      "num_tokens": 56232834.0,
      "step": 24555
    },
    {
      "entropy": 5.616108560562134,
      "epoch": 2.427837089758798,
      "grad_norm": 1.203125,
      "learning_rate": 0.000441304134847719,
      "loss": 4.9295,
      "mean_token_accuracy": 0.2182505488395691,
      "num_tokens": 56244517.0,
      "step": 24560
    },
    {
      "entropy": 5.757014894485474,
      "epoch": 2.4283313562672992,
      "grad_norm": 1.125,
      "learning_rate": 0.00044128036276316424,
      "loss": 5.1029,
      "mean_token_accuracy": 0.21037830859422685,
      "num_tokens": 56256629.0,
      "step": 24565
    },
    {
      "entropy": 5.736163139343262,
      "epoch": 2.4288256227758005,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044125658658808676,
      "loss": 5.1353,
      "mean_token_accuracy": 0.20654115080833435,
      "num_tokens": 56268992.0,
      "step": 24570
    },
    {
      "entropy": 5.5915507793426515,
      "epoch": 2.429319889284302,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004412328063230715,
      "loss": 4.9362,
      "mean_token_accuracy": 0.2265341490507126,
      "num_tokens": 56281588.0,
      "step": 24575
    },
    {
      "entropy": 5.74658522605896,
      "epoch": 2.4298141557928035,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004412090219687034,
      "loss": 5.1159,
      "mean_token_accuracy": 0.20606843531131744,
      "num_tokens": 56292587.0,
      "step": 24580
    },
    {
      "entropy": 5.6768317222595215,
      "epoch": 2.4303084223013047,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044118523352556756,
      "loss": 5.0018,
      "mean_token_accuracy": 0.21250355392694473,
      "num_tokens": 56304873.0,
      "step": 24585
    },
    {
      "entropy": 5.646546459197998,
      "epoch": 2.4308026888098064,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004411614409942492,
      "loss": 4.9677,
      "mean_token_accuracy": 0.2191954031586647,
      "num_tokens": 56316048.0,
      "step": 24590
    },
    {
      "entropy": 5.637232160568237,
      "epoch": 2.4312969553183077,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044113764437533356,
      "loss": 4.9827,
      "mean_token_accuracy": 0.2216390401124954,
      "num_tokens": 56328185.0,
      "step": 24595
    },
    {
      "entropy": 5.7336516857147215,
      "epoch": 2.431791221826809,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004411138436694062,
      "loss": 5.1007,
      "mean_token_accuracy": 0.20844763070344924,
      "num_tokens": 56341533.0,
      "step": 24600
    },
    {
      "entropy": 5.8027015209198,
      "epoch": 2.43228548833531,
      "grad_norm": 1.25,
      "learning_rate": 0.00044109003887705245,
      "loss": 5.1631,
      "mean_token_accuracy": 0.20699408352375032,
      "num_tokens": 56351610.0,
      "step": 24605
    },
    {
      "entropy": 5.7029906749725345,
      "epoch": 2.432779754843812,
      "grad_norm": 1.125,
      "learning_rate": 0.000441066229998858,
      "loss": 5.0258,
      "mean_token_accuracy": 0.21874250769615172,
      "num_tokens": 56361822.0,
      "step": 24610
    },
    {
      "entropy": 5.6800535202026365,
      "epoch": 2.433274021352313,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044104241703540854,
      "loss": 5.0228,
      "mean_token_accuracy": 0.21563208401203154,
      "num_tokens": 56373432.0,
      "step": 24615
    },
    {
      "entropy": 5.661668300628662,
      "epoch": 2.4337682878608144,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004410185999872898,
      "loss": 4.9928,
      "mean_token_accuracy": 0.2187672510743141,
      "num_tokens": 56385086.0,
      "step": 24620
    },
    {
      "entropy": 5.720230960845948,
      "epoch": 2.434262554369316,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044099477885508774,
      "loss": 4.9921,
      "mean_token_accuracy": 0.21573259085416793,
      "num_tokens": 56395963.0,
      "step": 24625
    },
    {
      "entropy": 5.725389528274536,
      "epoch": 2.4347568208778174,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044097095363938843,
      "loss": 5.063,
      "mean_token_accuracy": 0.20798081457614898,
      "num_tokens": 56407685.0,
      "step": 24630
    },
    {
      "entropy": 5.680238246917725,
      "epoch": 2.4352510873863187,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004409471243407779,
      "loss": 4.9871,
      "mean_token_accuracy": 0.21958784610033036,
      "num_tokens": 56418570.0,
      "step": 24635
    },
    {
      "entropy": 5.720429992675781,
      "epoch": 2.43574535389482,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004409232909598424,
      "loss": 5.0455,
      "mean_token_accuracy": 0.21304434835910796,
      "num_tokens": 56430312.0,
      "step": 24640
    },
    {
      "entropy": 5.7441215991973875,
      "epoch": 2.4362396204033216,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004408994534971682,
      "loss": 5.0993,
      "mean_token_accuracy": 0.2052801087498665,
      "num_tokens": 56441847.0,
      "step": 24645
    },
    {
      "entropy": 5.688012886047363,
      "epoch": 2.436733886911823,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044087561195334165,
      "loss": 5.0019,
      "mean_token_accuracy": 0.21972954124212266,
      "num_tokens": 56453567.0,
      "step": 24650
    },
    {
      "entropy": 5.689534759521484,
      "epoch": 2.437228153420324,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004408517663289493,
      "loss": 5.0168,
      "mean_token_accuracy": 0.21641867160797118,
      "num_tokens": 56464726.0,
      "step": 24655
    },
    {
      "entropy": 5.664986419677734,
      "epoch": 2.4377224199288254,
      "grad_norm": 1.125,
      "learning_rate": 0.0004408279166245778,
      "loss": 5.0102,
      "mean_token_accuracy": 0.21572286784648895,
      "num_tokens": 56475989.0,
      "step": 24660
    },
    {
      "entropy": 5.713948059082031,
      "epoch": 2.438216686437327,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044080406284081387,
      "loss": 5.0057,
      "mean_token_accuracy": 0.21967438310384751,
      "num_tokens": 56487782.0,
      "step": 24665
    },
    {
      "entropy": 5.675281667709351,
      "epoch": 2.4387109529458284,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004407802049782442,
      "loss": 5.0207,
      "mean_token_accuracy": 0.215204681456089,
      "num_tokens": 56500529.0,
      "step": 24670
    },
    {
      "entropy": 5.586161518096924,
      "epoch": 2.4392052194543297,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004407563430374558,
      "loss": 4.9606,
      "mean_token_accuracy": 0.21975905299186707,
      "num_tokens": 56512411.0,
      "step": 24675
    },
    {
      "entropy": 5.650100803375244,
      "epoch": 2.439699485962831,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004407324770190357,
      "loss": 4.8867,
      "mean_token_accuracy": 0.2270260453224182,
      "num_tokens": 56522983.0,
      "step": 24680
    },
    {
      "entropy": 5.6379321098327635,
      "epoch": 2.4401937524713326,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044070860692357086,
      "loss": 5.039,
      "mean_token_accuracy": 0.21439638137817382,
      "num_tokens": 56534380.0,
      "step": 24685
    },
    {
      "entropy": 5.634187793731689,
      "epoch": 2.440688018979834,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044068473275164863,
      "loss": 5.0215,
      "mean_token_accuracy": 0.21479888409376144,
      "num_tokens": 56546003.0,
      "step": 24690
    },
    {
      "entropy": 5.631953477859497,
      "epoch": 2.441182285488335,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004406608545038563,
      "loss": 4.9154,
      "mean_token_accuracy": 0.22390684485435486,
      "num_tokens": 56556006.0,
      "step": 24695
    },
    {
      "entropy": 5.641700744628906,
      "epoch": 2.441676551996837,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004406369721807812,
      "loss": 4.9543,
      "mean_token_accuracy": 0.22098313719034196,
      "num_tokens": 56567569.0,
      "step": 24700
    },
    {
      "entropy": 5.626941204071045,
      "epoch": 2.442170818505338,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004406130857830109,
      "loss": 5.0479,
      "mean_token_accuracy": 0.21779407560825348,
      "num_tokens": 56580143.0,
      "step": 24705
    },
    {
      "entropy": 5.669708108901977,
      "epoch": 2.4426650850138394,
      "grad_norm": 1.03125,
      "learning_rate": 0.000440589195311133,
      "loss": 4.959,
      "mean_token_accuracy": 0.22387165278196336,
      "num_tokens": 56591169.0,
      "step": 24710
    },
    {
      "entropy": 5.672441005706787,
      "epoch": 2.4431593515223406,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044056530076573515,
      "loss": 5.0078,
      "mean_token_accuracy": 0.21476769596338272,
      "num_tokens": 56602556.0,
      "step": 24715
    },
    {
      "entropy": 5.674004983901978,
      "epoch": 2.4436536180308424,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004405414021474053,
      "loss": 4.9768,
      "mean_token_accuracy": 0.2264782518148422,
      "num_tokens": 56614854.0,
      "step": 24720
    },
    {
      "entropy": 5.639186668395996,
      "epoch": 2.4441478845393436,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004405174994567313,
      "loss": 4.9198,
      "mean_token_accuracy": 0.2222625955939293,
      "num_tokens": 56626004.0,
      "step": 24725
    },
    {
      "entropy": 5.777587413787842,
      "epoch": 2.444642151047845,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044049359269430105,
      "loss": 5.049,
      "mean_token_accuracy": 0.21429750472307205,
      "num_tokens": 56636373.0,
      "step": 24730
    },
    {
      "entropy": 5.635915517807007,
      "epoch": 2.4451364175563466,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004404696818607028,
      "loss": 4.8996,
      "mean_token_accuracy": 0.2281292513012886,
      "num_tokens": 56648414.0,
      "step": 24735
    },
    {
      "entropy": 5.665419960021973,
      "epoch": 2.445630684064848,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044044576695652463,
      "loss": 4.9957,
      "mean_token_accuracy": 0.21772778332233428,
      "num_tokens": 56660430.0,
      "step": 24740
    },
    {
      "entropy": 5.649163293838501,
      "epoch": 2.446124950573349,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000440421847982355,
      "loss": 4.9943,
      "mean_token_accuracy": 0.2177042081952095,
      "num_tokens": 56671133.0,
      "step": 24745
    },
    {
      "entropy": 5.695397520065308,
      "epoch": 2.4466192170818504,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004403979249387822,
      "loss": 5.0717,
      "mean_token_accuracy": 0.21322779953479767,
      "num_tokens": 56682788.0,
      "step": 24750
    },
    {
      "entropy": 5.659640216827393,
      "epoch": 2.447113483590352,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004403739978263948,
      "loss": 5.0185,
      "mean_token_accuracy": 0.21997392773628235,
      "num_tokens": 56693832.0,
      "step": 24755
    },
    {
      "entropy": 5.705827283859253,
      "epoch": 2.4476077500988533,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044035006664578145,
      "loss": 5.0638,
      "mean_token_accuracy": 0.2102128252387047,
      "num_tokens": 56705915.0,
      "step": 24760
    },
    {
      "entropy": 5.669723033905029,
      "epoch": 2.4481020166073546,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044032613139753077,
      "loss": 4.9867,
      "mean_token_accuracy": 0.22068042755126954,
      "num_tokens": 56717461.0,
      "step": 24765
    },
    {
      "entropy": 5.610535907745361,
      "epoch": 2.4485962831158563,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004403021920822316,
      "loss": 4.8569,
      "mean_token_accuracy": 0.23334282338619233,
      "num_tokens": 56728695.0,
      "step": 24770
    },
    {
      "entropy": 5.649751901626587,
      "epoch": 2.4490905496243576,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004402782487004728,
      "loss": 4.9543,
      "mean_token_accuracy": 0.21636142134666442,
      "num_tokens": 56740520.0,
      "step": 24775
    },
    {
      "entropy": 5.580807638168335,
      "epoch": 2.449584816132859,
      "grad_norm": 1.125,
      "learning_rate": 0.00044025430125284353,
      "loss": 4.9017,
      "mean_token_accuracy": 0.22548197507858275,
      "num_tokens": 56751071.0,
      "step": 24780
    },
    {
      "entropy": 5.698151874542236,
      "epoch": 2.45007908264136,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004402303497399329,
      "loss": 5.0729,
      "mean_token_accuracy": 0.2098132073879242,
      "num_tokens": 56764759.0,
      "step": 24785
    },
    {
      "entropy": 5.734591054916382,
      "epoch": 2.4505733491498614,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044020639416232984,
      "loss": 5.0457,
      "mean_token_accuracy": 0.21618868857622148,
      "num_tokens": 56775592.0,
      "step": 24790
    },
    {
      "entropy": 5.698403835296631,
      "epoch": 2.451067615658363,
      "grad_norm": 1.203125,
      "learning_rate": 0.000440182434520624,
      "loss": 5.1041,
      "mean_token_accuracy": 0.20755320489406587,
      "num_tokens": 56786860.0,
      "step": 24795
    },
    {
      "entropy": 5.683164501190186,
      "epoch": 2.4515618821668643,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004401584708154045,
      "loss": 5.0408,
      "mean_token_accuracy": 0.21671999394893646,
      "num_tokens": 56798741.0,
      "step": 24800
    },
    {
      "entropy": 5.69742636680603,
      "epoch": 2.4520561486753656,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044013450304726114,
      "loss": 4.9792,
      "mean_token_accuracy": 0.2199311926960945,
      "num_tokens": 56810612.0,
      "step": 24805
    },
    {
      "entropy": 5.62714638710022,
      "epoch": 2.4525504151838673,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004401105312167833,
      "loss": 5.014,
      "mean_token_accuracy": 0.218637552857399,
      "num_tokens": 56822169.0,
      "step": 24810
    },
    {
      "entropy": 5.639250183105469,
      "epoch": 2.4530446816923686,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004400865553245608,
      "loss": 4.9609,
      "mean_token_accuracy": 0.22391692399978638,
      "num_tokens": 56833873.0,
      "step": 24815
    },
    {
      "entropy": 5.734986257553101,
      "epoch": 2.45353894820087,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044006257537118344,
      "loss": 4.9183,
      "mean_token_accuracy": 0.22980219423770903,
      "num_tokens": 56844616.0,
      "step": 24820
    },
    {
      "entropy": 5.688593244552612,
      "epoch": 2.454033214709371,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004400385913572412,
      "loss": 5.0114,
      "mean_token_accuracy": 0.2106737896800041,
      "num_tokens": 56855877.0,
      "step": 24825
    },
    {
      "entropy": 5.648899364471435,
      "epoch": 2.454527481217873,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044001460328332387,
      "loss": 5.0072,
      "mean_token_accuracy": 0.21450023353099823,
      "num_tokens": 56868103.0,
      "step": 24830
    },
    {
      "entropy": 5.708245468139649,
      "epoch": 2.455021747726374,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004399906111500218,
      "loss": 5.1085,
      "mean_token_accuracy": 0.2047686383128166,
      "num_tokens": 56878957.0,
      "step": 24835
    },
    {
      "entropy": 5.773703861236572,
      "epoch": 2.4555160142348753,
      "grad_norm": 1.078125,
      "learning_rate": 0.000439966614957925,
      "loss": 5.0306,
      "mean_token_accuracy": 0.20389980673789979,
      "num_tokens": 56891915.0,
      "step": 24840
    },
    {
      "entropy": 5.667133378982544,
      "epoch": 2.456010280743377,
      "grad_norm": 1.125,
      "learning_rate": 0.0004399426147076239,
      "loss": 4.9881,
      "mean_token_accuracy": 0.21946462541818618,
      "num_tokens": 56902322.0,
      "step": 24845
    },
    {
      "entropy": 5.682002878189087,
      "epoch": 2.4565045472518783,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000439918610399709,
      "loss": 5.0609,
      "mean_token_accuracy": 0.20221312195062638,
      "num_tokens": 56913560.0,
      "step": 24850
    },
    {
      "entropy": 5.711883974075318,
      "epoch": 2.4569988137603795,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043989460203477053,
      "loss": 4.9977,
      "mean_token_accuracy": 0.2135155528783798,
      "num_tokens": 56924235.0,
      "step": 24855
    },
    {
      "entropy": 5.649257755279541,
      "epoch": 2.457493080268881,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043987058961339937,
      "loss": 4.863,
      "mean_token_accuracy": 0.2276879444718361,
      "num_tokens": 56935257.0,
      "step": 24860
    },
    {
      "entropy": 5.664929914474487,
      "epoch": 2.4579873467773825,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043984657313618616,
      "loss": 5.0196,
      "mean_token_accuracy": 0.2208859518170357,
      "num_tokens": 56945896.0,
      "step": 24865
    },
    {
      "entropy": 5.737520027160644,
      "epoch": 2.4584816132858838,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004398225526037216,
      "loss": 5.1315,
      "mean_token_accuracy": 0.20247652977705002,
      "num_tokens": 56957274.0,
      "step": 24870
    },
    {
      "entropy": 5.650135183334351,
      "epoch": 2.458975879794385,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043979852801659673,
      "loss": 4.9619,
      "mean_token_accuracy": 0.21924443542957306,
      "num_tokens": 56969647.0,
      "step": 24875
    },
    {
      "entropy": 5.710012817382813,
      "epoch": 2.4594701463028867,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004397744993754025,
      "loss": 4.9641,
      "mean_token_accuracy": 0.21612230837345123,
      "num_tokens": 56981443.0,
      "step": 24880
    },
    {
      "entropy": 5.642097759246826,
      "epoch": 2.459964412811388,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043975046668072996,
      "loss": 4.95,
      "mean_token_accuracy": 0.22610472589731218,
      "num_tokens": 56992482.0,
      "step": 24885
    },
    {
      "entropy": 5.711119270324707,
      "epoch": 2.4604586793198893,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043972642993317043,
      "loss": 5.0128,
      "mean_token_accuracy": 0.2154277890920639,
      "num_tokens": 57003172.0,
      "step": 24890
    },
    {
      "entropy": 5.719520568847656,
      "epoch": 2.4609529458283905,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043970238913331513,
      "loss": 5.1169,
      "mean_token_accuracy": 0.20815230458974837,
      "num_tokens": 57014763.0,
      "step": 24895
    },
    {
      "entropy": 5.6418785572052,
      "epoch": 2.4614472123368922,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043967834428175565,
      "loss": 5.0154,
      "mean_token_accuracy": 0.22428970485925676,
      "num_tokens": 57025564.0,
      "step": 24900
    },
    {
      "entropy": 5.723076915740966,
      "epoch": 2.4619414788453935,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043965429537908316,
      "loss": 4.9932,
      "mean_token_accuracy": 0.21742018610239028,
      "num_tokens": 57036121.0,
      "step": 24905
    },
    {
      "entropy": 5.700960540771485,
      "epoch": 2.4624357453538948,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004396302424258896,
      "loss": 5.1308,
      "mean_token_accuracy": 0.20497939735651016,
      "num_tokens": 57048759.0,
      "step": 24910
    },
    {
      "entropy": 5.633072376251221,
      "epoch": 2.462930011862396,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004396061854227665,
      "loss": 4.9381,
      "mean_token_accuracy": 0.2216528758406639,
      "num_tokens": 57060791.0,
      "step": 24915
    },
    {
      "entropy": 5.644799470901489,
      "epoch": 2.4634242783708977,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043958212437030566,
      "loss": 4.9601,
      "mean_token_accuracy": 0.2201666057109833,
      "num_tokens": 57070788.0,
      "step": 24920
    },
    {
      "entropy": 5.6648184299469,
      "epoch": 2.463918544879399,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000439558059269099,
      "loss": 5.0443,
      "mean_token_accuracy": 0.21208992153406142,
      "num_tokens": 57082008.0,
      "step": 24925
    },
    {
      "entropy": 5.676169109344483,
      "epoch": 2.4644128113879002,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004395339901197386,
      "loss": 4.9721,
      "mean_token_accuracy": 0.21825537979602813,
      "num_tokens": 57093081.0,
      "step": 24930
    },
    {
      "entropy": 5.622377252578735,
      "epoch": 2.4649070778964015,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004395099169228166,
      "loss": 4.9276,
      "mean_token_accuracy": 0.2246924877166748,
      "num_tokens": 57103294.0,
      "step": 24935
    },
    {
      "entropy": 5.616104984283448,
      "epoch": 2.465401344404903,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043948583967892495,
      "loss": 4.9064,
      "mean_token_accuracy": 0.22889331430196763,
      "num_tokens": 57114509.0,
      "step": 24940
    },
    {
      "entropy": 5.7534528255462645,
      "epoch": 2.4658956109134045,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004394617583886563,
      "loss": 5.0622,
      "mean_token_accuracy": 0.2054104909300804,
      "num_tokens": 57126481.0,
      "step": 24945
    },
    {
      "entropy": 5.692492961883545,
      "epoch": 2.4663898774219057,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004394376730526028,
      "loss": 4.9816,
      "mean_token_accuracy": 0.2149360477924347,
      "num_tokens": 57138235.0,
      "step": 24950
    },
    {
      "entropy": 5.672536420822143,
      "epoch": 2.4668841439304074,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043941358367135706,
      "loss": 4.9983,
      "mean_token_accuracy": 0.21777955293655396,
      "num_tokens": 57150731.0,
      "step": 24955
    },
    {
      "entropy": 5.674402952194214,
      "epoch": 2.4673784104389087,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043938949024551155,
      "loss": 5.0217,
      "mean_token_accuracy": 0.21622692346572875,
      "num_tokens": 57162729.0,
      "step": 24960
    },
    {
      "entropy": 5.706616067886353,
      "epoch": 2.46787267694741,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043936539277565916,
      "loss": 5.0164,
      "mean_token_accuracy": 0.2167980119585991,
      "num_tokens": 57173855.0,
      "step": 24965
    },
    {
      "entropy": 5.641604089736939,
      "epoch": 2.4683669434559112,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043934129126239256,
      "loss": 4.9634,
      "mean_token_accuracy": 0.2234649747610092,
      "num_tokens": 57184829.0,
      "step": 24970
    },
    {
      "entropy": 5.696771335601807,
      "epoch": 2.468861209964413,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004393171857063048,
      "loss": 5.0423,
      "mean_token_accuracy": 0.21251803934574126,
      "num_tokens": 57195679.0,
      "step": 24975
    },
    {
      "entropy": 5.730101108551025,
      "epoch": 2.469355476472914,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043929307610798867,
      "loss": 4.9894,
      "mean_token_accuracy": 0.2182790756225586,
      "num_tokens": 57208022.0,
      "step": 24980
    },
    {
      "entropy": 5.630070972442627,
      "epoch": 2.4698497429814155,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004392689624680375,
      "loss": 4.8999,
      "mean_token_accuracy": 0.2281287968158722,
      "num_tokens": 57218699.0,
      "step": 24985
    },
    {
      "entropy": 5.64049654006958,
      "epoch": 2.470344009489917,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004392448447870443,
      "loss": 5.0569,
      "mean_token_accuracy": 0.21426043808460235,
      "num_tokens": 57231333.0,
      "step": 24990
    },
    {
      "entropy": 5.748660516738892,
      "epoch": 2.4708382759984184,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004392207230656024,
      "loss": 5.0221,
      "mean_token_accuracy": 0.20917847156524658,
      "num_tokens": 57242748.0,
      "step": 24995
    },
    {
      "entropy": 5.719303369522095,
      "epoch": 2.4713325425069197,
      "grad_norm": 1.125,
      "learning_rate": 0.0004391965973043053,
      "loss": 4.9774,
      "mean_token_accuracy": 0.22064861357212068,
      "num_tokens": 57255438.0,
      "step": 25000
    },
    {
      "entropy": 5.751277685165405,
      "epoch": 2.471826809015421,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004391724675037464,
      "loss": 5.1046,
      "mean_token_accuracy": 0.21224838346242905,
      "num_tokens": 57267472.0,
      "step": 25005
    },
    {
      "entropy": 5.692398405075073,
      "epoch": 2.4723210755239227,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043914833366451934,
      "loss": 4.9938,
      "mean_token_accuracy": 0.2160089671611786,
      "num_tokens": 57277921.0,
      "step": 25010
    },
    {
      "entropy": 5.678887605667114,
      "epoch": 2.472815342032424,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004391241957872178,
      "loss": 5.0084,
      "mean_token_accuracy": 0.21960539072752,
      "num_tokens": 57290207.0,
      "step": 25015
    },
    {
      "entropy": 5.675688886642456,
      "epoch": 2.473309608540925,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043910005387243564,
      "loss": 4.9531,
      "mean_token_accuracy": 0.22220274806022644,
      "num_tokens": 57300599.0,
      "step": 25020
    },
    {
      "entropy": 5.677533960342407,
      "epoch": 2.473803875049427,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043907590792076665,
      "loss": 5.0513,
      "mean_token_accuracy": 0.2088640108704567,
      "num_tokens": 57313035.0,
      "step": 25025
    },
    {
      "entropy": 5.673757648468017,
      "epoch": 2.474298141557928,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004390517579328048,
      "loss": 4.9993,
      "mean_token_accuracy": 0.219614477455616,
      "num_tokens": 57325141.0,
      "step": 25030
    },
    {
      "entropy": 5.720339393615722,
      "epoch": 2.4747924080664294,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004390276039091443,
      "loss": 5.0427,
      "mean_token_accuracy": 0.21940211504697799,
      "num_tokens": 57336824.0,
      "step": 25035
    },
    {
      "entropy": 5.764863777160644,
      "epoch": 2.4752866745749307,
      "grad_norm": 1.296875,
      "learning_rate": 0.00043900344585037933,
      "loss": 5.1136,
      "mean_token_accuracy": 0.20562119036912918,
      "num_tokens": 57347078.0,
      "step": 25040
    },
    {
      "entropy": 5.656171035766602,
      "epoch": 2.475780941083432,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043897928375710417,
      "loss": 4.9607,
      "mean_token_accuracy": 0.21789245903491974,
      "num_tokens": 57359320.0,
      "step": 25045
    },
    {
      "entropy": 5.734521436691284,
      "epoch": 2.4762752075919336,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004389551176299132,
      "loss": 5.0274,
      "mean_token_accuracy": 0.2110027402639389,
      "num_tokens": 57370642.0,
      "step": 25050
    },
    {
      "entropy": 5.656056022644043,
      "epoch": 2.476769474100435,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004389309474694008,
      "loss": 5.0008,
      "mean_token_accuracy": 0.22700024098157884,
      "num_tokens": 57382082.0,
      "step": 25055
    },
    {
      "entropy": 5.653369331359864,
      "epoch": 2.477263740608936,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004389067732761617,
      "loss": 5.0867,
      "mean_token_accuracy": 0.20839382112026214,
      "num_tokens": 57393591.0,
      "step": 25060
    },
    {
      "entropy": 5.70035948753357,
      "epoch": 2.477758007117438,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043888259505079063,
      "loss": 4.948,
      "mean_token_accuracy": 0.21773191690444946,
      "num_tokens": 57404999.0,
      "step": 25065
    },
    {
      "entropy": 5.688670587539673,
      "epoch": 2.478252273625939,
      "grad_norm": 1.125,
      "learning_rate": 0.0004388584127938822,
      "loss": 5.0351,
      "mean_token_accuracy": 0.21979344636201859,
      "num_tokens": 57416467.0,
      "step": 25070
    },
    {
      "entropy": 5.712506532669067,
      "epoch": 2.4787465401344404,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004388342265060314,
      "loss": 5.0771,
      "mean_token_accuracy": 0.21515004932880402,
      "num_tokens": 57427201.0,
      "step": 25075
    },
    {
      "entropy": 5.744004344940185,
      "epoch": 2.4792408066429417,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043881003618783326,
      "loss": 5.0616,
      "mean_token_accuracy": 0.21530888378620147,
      "num_tokens": 57438345.0,
      "step": 25080
    },
    {
      "entropy": 5.6976001262664795,
      "epoch": 2.4797350731514434,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004387858418398828,
      "loss": 5.0412,
      "mean_token_accuracy": 0.21416350901126863,
      "num_tokens": 57449818.0,
      "step": 25085
    },
    {
      "entropy": 5.720158958435059,
      "epoch": 2.4802293396599446,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004387616434627753,
      "loss": 5.0168,
      "mean_token_accuracy": 0.21614874005317689,
      "num_tokens": 57461961.0,
      "step": 25090
    },
    {
      "entropy": 5.631670188903809,
      "epoch": 2.480723606168446,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043873744105710585,
      "loss": 4.9702,
      "mean_token_accuracy": 0.22089387327432633,
      "num_tokens": 57473422.0,
      "step": 25095
    },
    {
      "entropy": 5.682746267318725,
      "epoch": 2.4812178726769476,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043871323462347,
      "loss": 4.955,
      "mean_token_accuracy": 0.22464104145765304,
      "num_tokens": 57484618.0,
      "step": 25100
    },
    {
      "entropy": 5.701671743392945,
      "epoch": 2.481712139185449,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004386890241624631,
      "loss": 4.8593,
      "mean_token_accuracy": 0.22736164182424545,
      "num_tokens": 57493521.0,
      "step": 25105
    },
    {
      "entropy": 5.645185327529907,
      "epoch": 2.48220640569395,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004386648096746809,
      "loss": 5.0164,
      "mean_token_accuracy": 0.2136268898844719,
      "num_tokens": 57505693.0,
      "step": 25110
    },
    {
      "entropy": 5.664259815216065,
      "epoch": 2.4827006722024514,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043864059116071897,
      "loss": 4.9656,
      "mean_token_accuracy": 0.22260957956314087,
      "num_tokens": 57517734.0,
      "step": 25115
    },
    {
      "entropy": 5.73551664352417,
      "epoch": 2.483194938710953,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004386163686211731,
      "loss": 5.0363,
      "mean_token_accuracy": 0.21635823398828508,
      "num_tokens": 57529393.0,
      "step": 25120
    },
    {
      "entropy": 5.625054407119751,
      "epoch": 2.4836892052194544,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043859214205663915,
      "loss": 4.9253,
      "mean_token_accuracy": 0.2234334886074066,
      "num_tokens": 57541109.0,
      "step": 25125
    },
    {
      "entropy": 5.727078247070312,
      "epoch": 2.4841834717279556,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043856791146771327,
      "loss": 5.0817,
      "mean_token_accuracy": 0.20471252351999283,
      "num_tokens": 57551869.0,
      "step": 25130
    },
    {
      "entropy": 5.683696126937866,
      "epoch": 2.4846777382364573,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004385436768549912,
      "loss": 4.9889,
      "mean_token_accuracy": 0.22505471408367156,
      "num_tokens": 57564174.0,
      "step": 25135
    },
    {
      "entropy": 5.74558391571045,
      "epoch": 2.4851720047449586,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043851943821906944,
      "loss": 5.0668,
      "mean_token_accuracy": 0.20998207479715347,
      "num_tokens": 57575110.0,
      "step": 25140
    },
    {
      "entropy": 5.682272624969483,
      "epoch": 2.48566627125346,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043849519556054406,
      "loss": 4.9846,
      "mean_token_accuracy": 0.21560257077217101,
      "num_tokens": 57587143.0,
      "step": 25145
    },
    {
      "entropy": 5.637385606765747,
      "epoch": 2.486160537761961,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043847094888001156,
      "loss": 4.9552,
      "mean_token_accuracy": 0.22274481505155563,
      "num_tokens": 57598828.0,
      "step": 25150
    },
    {
      "entropy": 5.690423154830933,
      "epoch": 2.486654804270463,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004384466981780683,
      "loss": 5.0314,
      "mean_token_accuracy": 0.2129915952682495,
      "num_tokens": 57609598.0,
      "step": 25155
    },
    {
      "entropy": 5.699505805969238,
      "epoch": 2.487149070778964,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043842244345531094,
      "loss": 4.9869,
      "mean_token_accuracy": 0.22090262770652772,
      "num_tokens": 57620702.0,
      "step": 25160
    },
    {
      "entropy": 5.712781238555908,
      "epoch": 2.4876433372874653,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043839818471233614,
      "loss": 5.0375,
      "mean_token_accuracy": 0.20699797868728637,
      "num_tokens": 57632996.0,
      "step": 25165
    },
    {
      "entropy": 5.696783638000488,
      "epoch": 2.4881376037959666,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004383739219497406,
      "loss": 5.0244,
      "mean_token_accuracy": 0.21355998665094375,
      "num_tokens": 57644222.0,
      "step": 25170
    },
    {
      "entropy": 5.6758842945098875,
      "epoch": 2.4886318703044683,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043834965516812124,
      "loss": 5.0003,
      "mean_token_accuracy": 0.21898068189620973,
      "num_tokens": 57656399.0,
      "step": 25175
    },
    {
      "entropy": 5.722968053817749,
      "epoch": 2.4891261368129696,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043832538436807506,
      "loss": 5.0546,
      "mean_token_accuracy": 0.20792030394077302,
      "num_tokens": 57668229.0,
      "step": 25180
    },
    {
      "entropy": 5.63116340637207,
      "epoch": 2.489620403321471,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043830110955019903,
      "loss": 4.9653,
      "mean_token_accuracy": 0.22337630093097688,
      "num_tokens": 57679828.0,
      "step": 25185
    },
    {
      "entropy": 5.693574285507202,
      "epoch": 2.490114669829972,
      "grad_norm": 1.125,
      "learning_rate": 0.00043827683071509044,
      "loss": 5.0277,
      "mean_token_accuracy": 0.215861676633358,
      "num_tokens": 57692248.0,
      "step": 25190
    },
    {
      "entropy": 5.707590341567993,
      "epoch": 2.490608936338474,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043825254786334646,
      "loss": 5.0198,
      "mean_token_accuracy": 0.2157030999660492,
      "num_tokens": 57704123.0,
      "step": 25195
    },
    {
      "entropy": 5.60373821258545,
      "epoch": 2.491103202846975,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043822826099556444,
      "loss": 4.8995,
      "mean_token_accuracy": 0.22218241095542907,
      "num_tokens": 57715867.0,
      "step": 25200
    },
    {
      "entropy": 5.723360395431518,
      "epoch": 2.4915974693554763,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043820397011234193,
      "loss": 5.092,
      "mean_token_accuracy": 0.20710570514202117,
      "num_tokens": 57727551.0,
      "step": 25205
    },
    {
      "entropy": 5.67418270111084,
      "epoch": 2.492091735863978,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004381796752142764,
      "loss": 4.9519,
      "mean_token_accuracy": 0.22568920403718948,
      "num_tokens": 57738872.0,
      "step": 25210
    },
    {
      "entropy": 5.647845268249512,
      "epoch": 2.4925860023724793,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004381553763019655,
      "loss": 4.9112,
      "mean_token_accuracy": 0.22385506629943847,
      "num_tokens": 57748323.0,
      "step": 25215
    },
    {
      "entropy": 5.669284343719482,
      "epoch": 2.4930802688809806,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043813107337600717,
      "loss": 5.0508,
      "mean_token_accuracy": 0.21331439465284346,
      "num_tokens": 57759697.0,
      "step": 25220
    },
    {
      "entropy": 5.672144079208374,
      "epoch": 2.493574535389482,
      "grad_norm": 0.984375,
      "learning_rate": 0.000438106766436999,
      "loss": 4.9767,
      "mean_token_accuracy": 0.21207967400550842,
      "num_tokens": 57772532.0,
      "step": 25225
    },
    {
      "entropy": 5.664194393157959,
      "epoch": 2.4940688018979835,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043808245548553916,
      "loss": 4.9353,
      "mean_token_accuracy": 0.22585006207227706,
      "num_tokens": 57783895.0,
      "step": 25230
    },
    {
      "entropy": 5.59704213142395,
      "epoch": 2.494563068406485,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043805814052222554,
      "loss": 4.9468,
      "mean_token_accuracy": 0.22482881993055343,
      "num_tokens": 57795566.0,
      "step": 25235
    },
    {
      "entropy": 5.658424997329712,
      "epoch": 2.495057334914986,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004380338215476564,
      "loss": 5.0347,
      "mean_token_accuracy": 0.21452810615301132,
      "num_tokens": 57807792.0,
      "step": 25240
    },
    {
      "entropy": 5.706170511245728,
      "epoch": 2.4955516014234878,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043800949856242995,
      "loss": 4.9884,
      "mean_token_accuracy": 0.22332818508148194,
      "num_tokens": 57817771.0,
      "step": 25245
    },
    {
      "entropy": 5.722303628921509,
      "epoch": 2.496045867931989,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004379851715671445,
      "loss": 5.0563,
      "mean_token_accuracy": 0.21444186419248581,
      "num_tokens": 57828442.0,
      "step": 25250
    },
    {
      "entropy": 5.750983953475952,
      "epoch": 2.4965401344404903,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043796084056239864,
      "loss": 5.1014,
      "mean_token_accuracy": 0.21133596450090408,
      "num_tokens": 57840537.0,
      "step": 25255
    },
    {
      "entropy": 5.724470090866089,
      "epoch": 2.4970344009489915,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043793650554879075,
      "loss": 5.1406,
      "mean_token_accuracy": 0.20724169611930848,
      "num_tokens": 57851645.0,
      "step": 25260
    },
    {
      "entropy": 5.665840148925781,
      "epoch": 2.4975286674574932,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004379121665269196,
      "loss": 4.9294,
      "mean_token_accuracy": 0.2209942474961281,
      "num_tokens": 57863812.0,
      "step": 25265
    },
    {
      "entropy": 5.7336629867553714,
      "epoch": 2.4980229339659945,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043788782349738387,
      "loss": 5.1079,
      "mean_token_accuracy": 0.20944133549928665,
      "num_tokens": 57876472.0,
      "step": 25270
    },
    {
      "entropy": 5.677082538604736,
      "epoch": 2.4985172004744958,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004378634764607824,
      "loss": 4.9399,
      "mean_token_accuracy": 0.2182406485080719,
      "num_tokens": 57887905.0,
      "step": 25275
    },
    {
      "entropy": 5.649707269668579,
      "epoch": 2.4990114669829975,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004378391254177141,
      "loss": 4.9401,
      "mean_token_accuracy": 0.2210611656308174,
      "num_tokens": 57899084.0,
      "step": 25280
    },
    {
      "entropy": 5.589944458007812,
      "epoch": 2.4995057334914987,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043781477036877807,
      "loss": 4.9785,
      "mean_token_accuracy": 0.22290801107883454,
      "num_tokens": 57910134.0,
      "step": 25285
    },
    {
      "entropy": 5.663058519363403,
      "epoch": 2.5,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004377904113145735,
      "loss": 4.9929,
      "mean_token_accuracy": 0.21774454414844513,
      "num_tokens": 57921957.0,
      "step": 25290
    },
    {
      "entropy": 5.618829393386841,
      "epoch": 2.5004942665085013,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004377660482556995,
      "loss": 4.9295,
      "mean_token_accuracy": 0.22694969028234482,
      "num_tokens": 57932924.0,
      "step": 25295
    },
    {
      "entropy": 5.636366271972657,
      "epoch": 2.5009885330170025,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043774168119275553,
      "loss": 4.9893,
      "mean_token_accuracy": 0.22189912348985671,
      "num_tokens": 57945275.0,
      "step": 25300
    },
    {
      "entropy": 5.652057218551636,
      "epoch": 2.5014827995255042,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043771731012634085,
      "loss": 5.0053,
      "mean_token_accuracy": 0.2223360762000084,
      "num_tokens": 57957933.0,
      "step": 25305
    },
    {
      "entropy": 5.6775623798370365,
      "epoch": 2.5019770660340055,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004376929350570551,
      "loss": 4.9973,
      "mean_token_accuracy": 0.21329478174448013,
      "num_tokens": 57970401.0,
      "step": 25310
    },
    {
      "entropy": 5.692291402816773,
      "epoch": 2.5024713325425068,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000437668555985498,
      "loss": 5.021,
      "mean_token_accuracy": 0.21095018088817596,
      "num_tokens": 57981433.0,
      "step": 25315
    },
    {
      "entropy": 5.686311054229736,
      "epoch": 2.5029655990510085,
      "grad_norm": 1.25,
      "learning_rate": 0.0004376441729122691,
      "loss": 4.9935,
      "mean_token_accuracy": 0.21841758340597153,
      "num_tokens": 57992288.0,
      "step": 25320
    },
    {
      "entropy": 5.720408010482788,
      "epoch": 2.5034598655595097,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004376197858379684,
      "loss": 5.0795,
      "mean_token_accuracy": 0.21744418293237686,
      "num_tokens": 58004156.0,
      "step": 25325
    },
    {
      "entropy": 5.653982639312744,
      "epoch": 2.503954132068011,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004375953947631957,
      "loss": 4.986,
      "mean_token_accuracy": 0.2212711200118065,
      "num_tokens": 58014949.0,
      "step": 25330
    },
    {
      "entropy": 5.653584718704224,
      "epoch": 2.5044483985765122,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043757099968855104,
      "loss": 4.9569,
      "mean_token_accuracy": 0.2144590884447098,
      "num_tokens": 58025664.0,
      "step": 25335
    },
    {
      "entropy": 5.664389610290527,
      "epoch": 2.504942665085014,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004375466006146346,
      "loss": 5.0251,
      "mean_token_accuracy": 0.2219024956226349,
      "num_tokens": 58037550.0,
      "step": 25340
    },
    {
      "entropy": 5.68461046218872,
      "epoch": 2.505436931593515,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004375221975420465,
      "loss": 4.9586,
      "mean_token_accuracy": 0.21920741051435472,
      "num_tokens": 58048292.0,
      "step": 25345
    },
    {
      "entropy": 5.643983602523804,
      "epoch": 2.5059311981020165,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004374977904713872,
      "loss": 4.962,
      "mean_token_accuracy": 0.2220889449119568,
      "num_tokens": 58060471.0,
      "step": 25350
    },
    {
      "entropy": 5.649974632263183,
      "epoch": 2.506425464610518,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000437473379403257,
      "loss": 4.9751,
      "mean_token_accuracy": 0.22036201506853104,
      "num_tokens": 58072205.0,
      "step": 25355
    },
    {
      "entropy": 5.694737291336059,
      "epoch": 2.5069197311190194,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043744896433825655,
      "loss": 4.9933,
      "mean_token_accuracy": 0.21448899209499359,
      "num_tokens": 58083789.0,
      "step": 25360
    },
    {
      "entropy": 5.659192180633545,
      "epoch": 2.5074139976275207,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004374245452769863,
      "loss": 5.1092,
      "mean_token_accuracy": 0.21911460906267166,
      "num_tokens": 58096064.0,
      "step": 25365
    },
    {
      "entropy": 5.747098922729492,
      "epoch": 2.507908264136022,
      "grad_norm": 1.125,
      "learning_rate": 0.00043740012222004697,
      "loss": 5.0569,
      "mean_token_accuracy": 0.21643559485673905,
      "num_tokens": 58106935.0,
      "step": 25370
    },
    {
      "entropy": 5.681591892242432,
      "epoch": 2.5084025306445237,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043737569516803955,
      "loss": 4.9867,
      "mean_token_accuracy": 0.21928021162748337,
      "num_tokens": 58119204.0,
      "step": 25375
    },
    {
      "entropy": 5.649598169326782,
      "epoch": 2.508896797153025,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043735126412156484,
      "loss": 4.9528,
      "mean_token_accuracy": 0.22029024511575698,
      "num_tokens": 58132006.0,
      "step": 25380
    },
    {
      "entropy": 5.7143261432647705,
      "epoch": 2.509391063661526,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004373268290812238,
      "loss": 5.071,
      "mean_token_accuracy": 0.2092391163110733,
      "num_tokens": 58143845.0,
      "step": 25385
    },
    {
      "entropy": 5.582916975021362,
      "epoch": 2.509885330170028,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043730239004761756,
      "loss": 4.8689,
      "mean_token_accuracy": 0.23230449259281158,
      "num_tokens": 58154838.0,
      "step": 25390
    },
    {
      "entropy": 5.742804098129272,
      "epoch": 2.510379596678529,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043727794702134743,
      "loss": 5.0717,
      "mean_token_accuracy": 0.20775987654924394,
      "num_tokens": 58165446.0,
      "step": 25395
    },
    {
      "entropy": 5.589677000045777,
      "epoch": 2.5108738631870304,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043725350000301455,
      "loss": 4.905,
      "mean_token_accuracy": 0.23176924586296083,
      "num_tokens": 58177815.0,
      "step": 25400
    },
    {
      "entropy": 5.682651424407959,
      "epoch": 2.5113681296955317,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004372290489932204,
      "loss": 5.0204,
      "mean_token_accuracy": 0.21828600913286209,
      "num_tokens": 58190296.0,
      "step": 25405
    },
    {
      "entropy": 5.6945267677307125,
      "epoch": 2.511862396204033,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004372045939925665,
      "loss": 4.9706,
      "mean_token_accuracy": 0.220380200445652,
      "num_tokens": 58200293.0,
      "step": 25410
    },
    {
      "entropy": 5.624544477462768,
      "epoch": 2.5123566627125347,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004371801350016545,
      "loss": 5.0297,
      "mean_token_accuracy": 0.2108832135796547,
      "num_tokens": 58212263.0,
      "step": 25415
    },
    {
      "entropy": 5.714477109909057,
      "epoch": 2.512850929221036,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043715567202108597,
      "loss": 5.0573,
      "mean_token_accuracy": 0.21203925013542174,
      "num_tokens": 58222961.0,
      "step": 25420
    },
    {
      "entropy": 5.758030271530151,
      "epoch": 2.5133451957295376,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004371312050514626,
      "loss": 5.0238,
      "mean_token_accuracy": 0.21647415310144424,
      "num_tokens": 58233514.0,
      "step": 25425
    },
    {
      "entropy": 5.7024109840393065,
      "epoch": 2.513839462238039,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043710673409338665,
      "loss": 4.9479,
      "mean_token_accuracy": 0.21910342425107956,
      "num_tokens": 58243637.0,
      "step": 25430
    },
    {
      "entropy": 5.626990509033203,
      "epoch": 2.51433372874654,
      "grad_norm": 1.0,
      "learning_rate": 0.0004370822591474598,
      "loss": 4.9569,
      "mean_token_accuracy": 0.2250988468527794,
      "num_tokens": 58256603.0,
      "step": 25435
    },
    {
      "entropy": 5.645711088180542,
      "epoch": 2.5148279952550414,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004370577802142842,
      "loss": 4.9227,
      "mean_token_accuracy": 0.2226131945848465,
      "num_tokens": 58268112.0,
      "step": 25440
    },
    {
      "entropy": 5.761875915527344,
      "epoch": 2.5153222617635427,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043703329729446213,
      "loss": 5.0695,
      "mean_token_accuracy": 0.21268550008535386,
      "num_tokens": 58281026.0,
      "step": 25445
    },
    {
      "entropy": 5.667680311203003,
      "epoch": 2.5158165282720444,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004370088103885958,
      "loss": 5.0439,
      "mean_token_accuracy": 0.2166520208120346,
      "num_tokens": 58292749.0,
      "step": 25450
    },
    {
      "entropy": 5.714810562133789,
      "epoch": 2.5163107947805456,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004369843194972876,
      "loss": 5.0583,
      "mean_token_accuracy": 0.20796494781970978,
      "num_tokens": 58304019.0,
      "step": 25455
    },
    {
      "entropy": 5.73217191696167,
      "epoch": 2.516805061289047,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043695982462114,
      "loss": 5.0695,
      "mean_token_accuracy": 0.21553469747304915,
      "num_tokens": 58315998.0,
      "step": 25460
    },
    {
      "entropy": 5.692540884017944,
      "epoch": 2.5172993277975486,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043693532576075556,
      "loss": 5.023,
      "mean_token_accuracy": 0.20998884737491608,
      "num_tokens": 58327839.0,
      "step": 25465
    },
    {
      "entropy": 5.674967956542969,
      "epoch": 2.51779359430605,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000436910822916737,
      "loss": 4.954,
      "mean_token_accuracy": 0.2182467356324196,
      "num_tokens": 58339847.0,
      "step": 25470
    },
    {
      "entropy": 5.647035026550293,
      "epoch": 2.518287860814551,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000436886316089687,
      "loss": 4.9248,
      "mean_token_accuracy": 0.2262007251381874,
      "num_tokens": 58351727.0,
      "step": 25475
    },
    {
      "entropy": 5.66247067451477,
      "epoch": 2.5187821273230524,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043686180528020865,
      "loss": 4.9856,
      "mean_token_accuracy": 0.21872608214616776,
      "num_tokens": 58362519.0,
      "step": 25480
    },
    {
      "entropy": 5.730219888687134,
      "epoch": 2.519276393831554,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004368372904889047,
      "loss": 5.0622,
      "mean_token_accuracy": 0.21947843730449676,
      "num_tokens": 58373747.0,
      "step": 25485
    },
    {
      "entropy": 5.653955364227295,
      "epoch": 2.5197706603400554,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004368127717163783,
      "loss": 4.954,
      "mean_token_accuracy": 0.2212348386645317,
      "num_tokens": 58385196.0,
      "step": 25490
    },
    {
      "entropy": 5.706541204452515,
      "epoch": 2.5202649268485566,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004367882489632327,
      "loss": 5.0396,
      "mean_token_accuracy": 0.21207271218299867,
      "num_tokens": 58397005.0,
      "step": 25495
    },
    {
      "entropy": 5.595873212814331,
      "epoch": 2.5207591933570583,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043676372223007095,
      "loss": 4.9326,
      "mean_token_accuracy": 0.219038662314415,
      "num_tokens": 58407685.0,
      "step": 25500
    },
    {
      "entropy": 5.627429723739624,
      "epoch": 2.5212534598655596,
      "grad_norm": 1.125,
      "learning_rate": 0.0004367391915174966,
      "loss": 4.9829,
      "mean_token_accuracy": 0.21679268628358841,
      "num_tokens": 58419250.0,
      "step": 25505
    },
    {
      "entropy": 5.707195520401001,
      "epoch": 2.521747726374061,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004367146568261131,
      "loss": 5.0047,
      "mean_token_accuracy": 0.2134971722960472,
      "num_tokens": 58431135.0,
      "step": 25510
    },
    {
      "entropy": 5.723687648773193,
      "epoch": 2.522241992882562,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004366901181565239,
      "loss": 5.0656,
      "mean_token_accuracy": 0.20971050709486008,
      "num_tokens": 58443865.0,
      "step": 25515
    },
    {
      "entropy": 5.706615781784057,
      "epoch": 2.5227362593910634,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043666557550933266,
      "loss": 5.0543,
      "mean_token_accuracy": 0.22006092220544815,
      "num_tokens": 58456484.0,
      "step": 25520
    },
    {
      "entropy": 5.702159357070923,
      "epoch": 2.523230525899565,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004366410288851433,
      "loss": 5.0323,
      "mean_token_accuracy": 0.21417780816555024,
      "num_tokens": 58468384.0,
      "step": 25525
    },
    {
      "entropy": 5.639272880554199,
      "epoch": 2.5237247924080664,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043661647828455944,
      "loss": 4.993,
      "mean_token_accuracy": 0.2178630754351616,
      "num_tokens": 58479703.0,
      "step": 25530
    },
    {
      "entropy": 5.800481224060059,
      "epoch": 2.524219058916568,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004365919237081852,
      "loss": 5.1211,
      "mean_token_accuracy": 0.21218721717596054,
      "num_tokens": 58491313.0,
      "step": 25535
    },
    {
      "entropy": 5.737855529785156,
      "epoch": 2.5247133254250693,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043656736515662457,
      "loss": 5.0477,
      "mean_token_accuracy": 0.20464641749858856,
      "num_tokens": 58503323.0,
      "step": 25540
    },
    {
      "entropy": 5.705129671096802,
      "epoch": 2.5252075919335706,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004365428026304816,
      "loss": 4.9724,
      "mean_token_accuracy": 0.21658896058797836,
      "num_tokens": 58514935.0,
      "step": 25545
    },
    {
      "entropy": 5.850637292861938,
      "epoch": 2.525701858442072,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004365182361303607,
      "loss": 5.1643,
      "mean_token_accuracy": 0.20337425619363786,
      "num_tokens": 58525724.0,
      "step": 25550
    },
    {
      "entropy": 5.738964223861695,
      "epoch": 2.526196124950573,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043649366565686615,
      "loss": 5.051,
      "mean_token_accuracy": 0.22097738534212114,
      "num_tokens": 58536868.0,
      "step": 25555
    },
    {
      "entropy": 5.65815281867981,
      "epoch": 2.526690391459075,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043646909121060237,
      "loss": 5.0212,
      "mean_token_accuracy": 0.21731212139129638,
      "num_tokens": 58547285.0,
      "step": 25560
    },
    {
      "entropy": 5.699366235733033,
      "epoch": 2.527184657967576,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004364445127921738,
      "loss": 5.0371,
      "mean_token_accuracy": 0.21405023336410522,
      "num_tokens": 58559960.0,
      "step": 25565
    },
    {
      "entropy": 5.69747748374939,
      "epoch": 2.5276789244760773,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004364199304021853,
      "loss": 4.9999,
      "mean_token_accuracy": 0.213905668258667,
      "num_tokens": 58571229.0,
      "step": 25570
    },
    {
      "entropy": 5.7575304985046385,
      "epoch": 2.528173190984579,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004363953440412414,
      "loss": 5.0656,
      "mean_token_accuracy": 0.21360839009284974,
      "num_tokens": 58580988.0,
      "step": 25575
    },
    {
      "entropy": 5.655761671066284,
      "epoch": 2.5286674574930803,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000436370753709947,
      "loss": 5.0321,
      "mean_token_accuracy": 0.2207318514585495,
      "num_tokens": 58592547.0,
      "step": 25580
    },
    {
      "entropy": 5.679642963409424,
      "epoch": 2.5291617240015816,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004363461594089071,
      "loss": 5.035,
      "mean_token_accuracy": 0.22493955194950105,
      "num_tokens": 58604820.0,
      "step": 25585
    },
    {
      "entropy": 5.676651287078857,
      "epoch": 2.529655990510083,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043632156113872656,
      "loss": 4.9571,
      "mean_token_accuracy": 0.22722918689250945,
      "num_tokens": 58615635.0,
      "step": 25590
    },
    {
      "entropy": 5.70601487159729,
      "epoch": 2.5301502570185845,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043629695890001064,
      "loss": 4.9909,
      "mean_token_accuracy": 0.22076619863510133,
      "num_tokens": 58628799.0,
      "step": 25595
    },
    {
      "entropy": 5.6048609733581545,
      "epoch": 2.530644523527086,
      "grad_norm": 1.0,
      "learning_rate": 0.0004362723526933645,
      "loss": 4.993,
      "mean_token_accuracy": 0.22045634984970092,
      "num_tokens": 58641514.0,
      "step": 25600
    },
    {
      "entropy": 5.662329816818238,
      "epoch": 2.531138790035587,
      "grad_norm": 1.125,
      "learning_rate": 0.0004362477425193934,
      "loss": 4.9425,
      "mean_token_accuracy": 0.2210210844874382,
      "num_tokens": 58651719.0,
      "step": 25605
    },
    {
      "entropy": 5.737444591522217,
      "epoch": 2.5316330565440888,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004362231283787029,
      "loss": 4.969,
      "mean_token_accuracy": 0.22187156975269318,
      "num_tokens": 58663032.0,
      "step": 25610
    },
    {
      "entropy": 5.589449739456176,
      "epoch": 2.53212732305259,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004361985102718984,
      "loss": 4.925,
      "mean_token_accuracy": 0.22426300197839738,
      "num_tokens": 58674188.0,
      "step": 25615
    },
    {
      "entropy": 5.618265151977539,
      "epoch": 2.5326215895610913,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043617388819958554,
      "loss": 4.899,
      "mean_token_accuracy": 0.22599843442440032,
      "num_tokens": 58684892.0,
      "step": 25620
    },
    {
      "entropy": 5.642139911651611,
      "epoch": 2.5331158560695926,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043614926216237,
      "loss": 4.9328,
      "mean_token_accuracy": 0.22733301818370819,
      "num_tokens": 58695520.0,
      "step": 25625
    },
    {
      "entropy": 5.675538015365601,
      "epoch": 2.5336101225780943,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004361246321608577,
      "loss": 4.9281,
      "mean_token_accuracy": 0.22020948827266693,
      "num_tokens": 58706579.0,
      "step": 25630
    },
    {
      "entropy": 5.7443122386932375,
      "epoch": 2.5341043890865955,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004360999981956544,
      "loss": 5.0068,
      "mean_token_accuracy": 0.21650769412517548,
      "num_tokens": 58717292.0,
      "step": 25635
    },
    {
      "entropy": 5.677191209793091,
      "epoch": 2.534598655595097,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043607536026736603,
      "loss": 5.0438,
      "mean_token_accuracy": 0.2113322362303734,
      "num_tokens": 58728761.0,
      "step": 25640
    },
    {
      "entropy": 5.652059698104859,
      "epoch": 2.5350929221035985,
      "grad_norm": 1.265625,
      "learning_rate": 0.000436050718376599,
      "loss": 4.9072,
      "mean_token_accuracy": 0.22349171787500383,
      "num_tokens": 58739494.0,
      "step": 25645
    },
    {
      "entropy": 5.744930028915405,
      "epoch": 2.5355871886120998,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004360260725239592,
      "loss": 5.1085,
      "mean_token_accuracy": 0.2070629432797432,
      "num_tokens": 58750119.0,
      "step": 25650
    },
    {
      "entropy": 5.684302043914795,
      "epoch": 2.536081455120601,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043600142271005304,
      "loss": 4.9649,
      "mean_token_accuracy": 0.21406439393758775,
      "num_tokens": 58760438.0,
      "step": 25655
    },
    {
      "entropy": 5.736627435684204,
      "epoch": 2.5365757216291023,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000435976768935487,
      "loss": 5.0679,
      "mean_token_accuracy": 0.21018319129943847,
      "num_tokens": 58771484.0,
      "step": 25660
    },
    {
      "entropy": 5.704580783843994,
      "epoch": 2.5370699881376035,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004359521112008674,
      "loss": 5.006,
      "mean_token_accuracy": 0.21156415790319444,
      "num_tokens": 58782716.0,
      "step": 25665
    },
    {
      "entropy": 5.7397631168365475,
      "epoch": 2.5375642546461052,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004359274495068009,
      "loss": 5.0599,
      "mean_token_accuracy": 0.21154440939426422,
      "num_tokens": 58793557.0,
      "step": 25670
    },
    {
      "entropy": 5.674383735656738,
      "epoch": 2.5380585211546065,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004359027838538941,
      "loss": 4.9864,
      "mean_token_accuracy": 0.21530407965183257,
      "num_tokens": 58804225.0,
      "step": 25675
    },
    {
      "entropy": 5.639667129516601,
      "epoch": 2.538552787663108,
      "grad_norm": 1.25,
      "learning_rate": 0.000435878114242754,
      "loss": 4.9411,
      "mean_token_accuracy": 0.2246530368924141,
      "num_tokens": 58814368.0,
      "step": 25680
    },
    {
      "entropy": 5.663958740234375,
      "epoch": 2.5390470541716095,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043585344067398725,
      "loss": 4.9559,
      "mean_token_accuracy": 0.21983493268489837,
      "num_tokens": 58825925.0,
      "step": 25685
    },
    {
      "entropy": 5.76571831703186,
      "epoch": 2.5395413206801107,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004358287631482009,
      "loss": 5.0445,
      "mean_token_accuracy": 0.21155012249946595,
      "num_tokens": 58837334.0,
      "step": 25690
    },
    {
      "entropy": 5.692574501037598,
      "epoch": 2.540035587188612,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043580408166600206,
      "loss": 4.9865,
      "mean_token_accuracy": 0.21738438904285431,
      "num_tokens": 58848045.0,
      "step": 25695
    },
    {
      "entropy": 5.693753385543824,
      "epoch": 2.5405298536971133,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043577939622799786,
      "loss": 4.9784,
      "mean_token_accuracy": 0.22172187864780427,
      "num_tokens": 58858186.0,
      "step": 25700
    },
    {
      "entropy": 5.650910472869873,
      "epoch": 2.541024120205615,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004357547068347955,
      "loss": 4.9349,
      "mean_token_accuracy": 0.22337903082370758,
      "num_tokens": 58870102.0,
      "step": 25705
    },
    {
      "entropy": 5.7014459609985355,
      "epoch": 2.5415183867141162,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004357300134870025,
      "loss": 5.0221,
      "mean_token_accuracy": 0.21377768516540527,
      "num_tokens": 58881246.0,
      "step": 25710
    },
    {
      "entropy": 5.747487783432007,
      "epoch": 2.5420126532226175,
      "grad_norm": 1.125,
      "learning_rate": 0.0004357053161852262,
      "loss": 5.1051,
      "mean_token_accuracy": 0.20898186415433884,
      "num_tokens": 58893722.0,
      "step": 25715
    },
    {
      "entropy": 5.662268590927124,
      "epoch": 2.542506919731119,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004356806149300742,
      "loss": 4.8671,
      "mean_token_accuracy": 0.22970073372125627,
      "num_tokens": 58904703.0,
      "step": 25720
    },
    {
      "entropy": 5.591051292419434,
      "epoch": 2.5430011862396205,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004356559097221541,
      "loss": 4.9201,
      "mean_token_accuracy": 0.2281045600771904,
      "num_tokens": 58916626.0,
      "step": 25725
    },
    {
      "entropy": 5.670955991744995,
      "epoch": 2.5434954527481217,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043563120056207375,
      "loss": 5.0447,
      "mean_token_accuracy": 0.2172209396958351,
      "num_tokens": 58927943.0,
      "step": 25730
    },
    {
      "entropy": 5.798449897766114,
      "epoch": 2.543989719256623,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004356064874504409,
      "loss": 5.108,
      "mean_token_accuracy": 0.2179997131228447,
      "num_tokens": 58940086.0,
      "step": 25735
    },
    {
      "entropy": 5.745757341384888,
      "epoch": 2.5444839857651247,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004355817703878637,
      "loss": 5.0036,
      "mean_token_accuracy": 0.21982791572809218,
      "num_tokens": 58952280.0,
      "step": 25740
    },
    {
      "entropy": 5.746413612365723,
      "epoch": 2.544978252273626,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004355570493749498,
      "loss": 5.02,
      "mean_token_accuracy": 0.21654891967773438,
      "num_tokens": 58963726.0,
      "step": 25745
    },
    {
      "entropy": 5.703045654296875,
      "epoch": 2.545472518782127,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004355323244123077,
      "loss": 5.0227,
      "mean_token_accuracy": 0.21851530969142913,
      "num_tokens": 58974941.0,
      "step": 25750
    },
    {
      "entropy": 5.62224440574646,
      "epoch": 2.545966785290629,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043550759550054554,
      "loss": 5.0017,
      "mean_token_accuracy": 0.21759624779224396,
      "num_tokens": 58985234.0,
      "step": 25755
    },
    {
      "entropy": 5.66828031539917,
      "epoch": 2.54646105179913,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043548286264027155,
      "loss": 4.9271,
      "mean_token_accuracy": 0.22616045773029328,
      "num_tokens": 58996781.0,
      "step": 25760
    },
    {
      "entropy": 5.6900678157806395,
      "epoch": 2.5469553183076314,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043545812583209426,
      "loss": 4.9707,
      "mean_token_accuracy": 0.21759846359491347,
      "num_tokens": 59008044.0,
      "step": 25765
    },
    {
      "entropy": 5.674028921127319,
      "epoch": 2.5474495848161327,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043543338507662227,
      "loss": 5.0349,
      "mean_token_accuracy": 0.21647845655679704,
      "num_tokens": 59021579.0,
      "step": 25770
    },
    {
      "entropy": 5.739804267883301,
      "epoch": 2.547943851324634,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004354086403744641,
      "loss": 5.0306,
      "mean_token_accuracy": 0.2109982892870903,
      "num_tokens": 59033541.0,
      "step": 25775
    },
    {
      "entropy": 5.722770166397095,
      "epoch": 2.5484381178331357,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004353838917262284,
      "loss": 4.9837,
      "mean_token_accuracy": 0.2161448985338211,
      "num_tokens": 59044067.0,
      "step": 25780
    },
    {
      "entropy": 5.586360311508178,
      "epoch": 2.548932384341637,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004353591391325241,
      "loss": 4.916,
      "mean_token_accuracy": 0.22428347617387773,
      "num_tokens": 59056758.0,
      "step": 25785
    },
    {
      "entropy": 5.60769214630127,
      "epoch": 2.5494266508501386,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043533438259396015,
      "loss": 4.9176,
      "mean_token_accuracy": 0.2264774352312088,
      "num_tokens": 59067681.0,
      "step": 25790
    },
    {
      "entropy": 5.714349031448364,
      "epoch": 2.54992091735864,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043530962211114547,
      "loss": 5.0235,
      "mean_token_accuracy": 0.21930799633264542,
      "num_tokens": 59079148.0,
      "step": 25795
    },
    {
      "entropy": 5.734426403045655,
      "epoch": 2.550415183867141,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004352848576846893,
      "loss": 5.0806,
      "mean_token_accuracy": 0.20495122522115708,
      "num_tokens": 59090597.0,
      "step": 25800
    },
    {
      "entropy": 5.674501705169678,
      "epoch": 2.5509094503756424,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043526008931520067,
      "loss": 4.9236,
      "mean_token_accuracy": 0.22471970170736313,
      "num_tokens": 59102445.0,
      "step": 25805
    },
    {
      "entropy": 5.755542469024658,
      "epoch": 2.5514037168841437,
      "grad_norm": 1.1875,
      "learning_rate": 0.000435235317003289,
      "loss": 5.0397,
      "mean_token_accuracy": 0.20923540145158767,
      "num_tokens": 59113470.0,
      "step": 25810
    },
    {
      "entropy": 5.606893825531006,
      "epoch": 2.5518979833926454,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043521054074956367,
      "loss": 4.8829,
      "mean_token_accuracy": 0.23383215069770813,
      "num_tokens": 59124370.0,
      "step": 25815
    },
    {
      "entropy": 5.666479063034058,
      "epoch": 2.5523922499011467,
      "grad_norm": 1.125,
      "learning_rate": 0.00043518576055463423,
      "loss": 5.0753,
      "mean_token_accuracy": 0.2185169130563736,
      "num_tokens": 59135783.0,
      "step": 25820
    },
    {
      "entropy": 5.7129518508911135,
      "epoch": 2.552886516409648,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004351609764191102,
      "loss": 5.0039,
      "mean_token_accuracy": 0.21676357537508012,
      "num_tokens": 59146729.0,
      "step": 25825
    },
    {
      "entropy": 5.769175434112549,
      "epoch": 2.5533807829181496,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004351361883436014,
      "loss": 5.0573,
      "mean_token_accuracy": 0.21704217940568923,
      "num_tokens": 59159413.0,
      "step": 25830
    },
    {
      "entropy": 5.664879179000854,
      "epoch": 2.553875049426651,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004351113963287173,
      "loss": 4.9581,
      "mean_token_accuracy": 0.21388407051563263,
      "num_tokens": 59171013.0,
      "step": 25835
    },
    {
      "entropy": 5.697875928878784,
      "epoch": 2.554369315935152,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043508660037506813,
      "loss": 5.0322,
      "mean_token_accuracy": 0.21326106935739517,
      "num_tokens": 59182571.0,
      "step": 25840
    },
    {
      "entropy": 5.593610382080078,
      "epoch": 2.5548635824436534,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004350618004832638,
      "loss": 4.9256,
      "mean_token_accuracy": 0.22768286764621734,
      "num_tokens": 59194310.0,
      "step": 25845
    },
    {
      "entropy": 5.66645188331604,
      "epoch": 2.555357848952155,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004350369966539143,
      "loss": 5.0036,
      "mean_token_accuracy": 0.21655056178569793,
      "num_tokens": 59206656.0,
      "step": 25850
    },
    {
      "entropy": 5.727938604354859,
      "epoch": 2.5558521154606564,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043501218888762985,
      "loss": 5.0911,
      "mean_token_accuracy": 0.2082471579313278,
      "num_tokens": 59219180.0,
      "step": 25855
    },
    {
      "entropy": 5.662927913665771,
      "epoch": 2.5563463819691576,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004349873771850208,
      "loss": 4.9513,
      "mean_token_accuracy": 0.22282288372516632,
      "num_tokens": 59230879.0,
      "step": 25860
    },
    {
      "entropy": 5.764781665802002,
      "epoch": 2.5568406484776594,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004349625615466974,
      "loss": 5.0593,
      "mean_token_accuracy": 0.20757858604192733,
      "num_tokens": 59242283.0,
      "step": 25865
    },
    {
      "entropy": 5.644295024871826,
      "epoch": 2.5573349149861606,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043493774197327014,
      "loss": 4.9293,
      "mean_token_accuracy": 0.22780680954456328,
      "num_tokens": 59253655.0,
      "step": 25870
    },
    {
      "entropy": 5.726610994338989,
      "epoch": 2.557829181494662,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043491291846534965,
      "loss": 5.0008,
      "mean_token_accuracy": 0.20999871492385863,
      "num_tokens": 59266600.0,
      "step": 25875
    },
    {
      "entropy": 5.754388904571533,
      "epoch": 2.558323448003163,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043488809102354653,
      "loss": 5.0967,
      "mean_token_accuracy": 0.21101153790950775,
      "num_tokens": 59277422.0,
      "step": 25880
    },
    {
      "entropy": 5.629184818267822,
      "epoch": 2.558817714511665,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043486325964847163,
      "loss": 4.9423,
      "mean_token_accuracy": 0.21830690950155257,
      "num_tokens": 59289179.0,
      "step": 25885
    },
    {
      "entropy": 5.612477397918701,
      "epoch": 2.559311981020166,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043483842434073567,
      "loss": 4.8893,
      "mean_token_accuracy": 0.2338917076587677,
      "num_tokens": 59299692.0,
      "step": 25890
    },
    {
      "entropy": 5.692693471908569,
      "epoch": 2.5598062475286674,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004348135851009497,
      "loss": 4.9793,
      "mean_token_accuracy": 0.21635424643754958,
      "num_tokens": 59310797.0,
      "step": 25895
    },
    {
      "entropy": 5.617331600189209,
      "epoch": 2.560300514037169,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004347887419297247,
      "loss": 4.9835,
      "mean_token_accuracy": 0.22083178162574768,
      "num_tokens": 59323167.0,
      "step": 25900
    },
    {
      "entropy": 5.633772134780884,
      "epoch": 2.5607947805456703,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043476389482767193,
      "loss": 4.9929,
      "mean_token_accuracy": 0.21701076328754426,
      "num_tokens": 59335956.0,
      "step": 25905
    },
    {
      "entropy": 5.747253942489624,
      "epoch": 2.5612890470541716,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004347390437954026,
      "loss": 5.0466,
      "mean_token_accuracy": 0.2073051318526268,
      "num_tokens": 59348090.0,
      "step": 25910
    },
    {
      "entropy": 5.684022998809814,
      "epoch": 2.561783313562673,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004347141888335279,
      "loss": 5.0037,
      "mean_token_accuracy": 0.21830187141895294,
      "num_tokens": 59360276.0,
      "step": 25915
    },
    {
      "entropy": 5.703080129623413,
      "epoch": 2.562277580071174,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004346893299426595,
      "loss": 5.0497,
      "mean_token_accuracy": 0.20764821767807007,
      "num_tokens": 59372049.0,
      "step": 25920
    },
    {
      "entropy": 5.7351963996887205,
      "epoch": 2.562771846579676,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043466446712340866,
      "loss": 5.0018,
      "mean_token_accuracy": 0.21832403242588044,
      "num_tokens": 59384017.0,
      "step": 25925
    },
    {
      "entropy": 5.791358566284179,
      "epoch": 2.563266113088177,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004346396003763873,
      "loss": 5.0794,
      "mean_token_accuracy": 0.21252351999282837,
      "num_tokens": 59395966.0,
      "step": 25930
    },
    {
      "entropy": 5.655357217788696,
      "epoch": 2.563760379596679,
      "grad_norm": 1.125,
      "learning_rate": 0.0004346147297022069,
      "loss": 4.9902,
      "mean_token_accuracy": 0.21827693283557892,
      "num_tokens": 59405815.0,
      "step": 25935
    },
    {
      "entropy": 5.684198904037475,
      "epoch": 2.56425464610518,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043458985510147946,
      "loss": 4.982,
      "mean_token_accuracy": 0.2167527660727501,
      "num_tokens": 59416662.0,
      "step": 25940
    },
    {
      "entropy": 5.737367773056031,
      "epoch": 2.5647489126136813,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004345649765748168,
      "loss": 5.0327,
      "mean_token_accuracy": 0.2193744108080864,
      "num_tokens": 59427143.0,
      "step": 25945
    },
    {
      "entropy": 5.598809051513672,
      "epoch": 2.5652431791221826,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043454009412283097,
      "loss": 4.999,
      "mean_token_accuracy": 0.21687404662370682,
      "num_tokens": 59439657.0,
      "step": 25950
    },
    {
      "entropy": 5.670913648605347,
      "epoch": 2.565737445630684,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043451520774613405,
      "loss": 5.0094,
      "mean_token_accuracy": 0.21599899530410765,
      "num_tokens": 59452313.0,
      "step": 25955
    },
    {
      "entropy": 5.692306184768677,
      "epoch": 2.5662317121391856,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043449031744533824,
      "loss": 4.9485,
      "mean_token_accuracy": 0.21973504722118378,
      "num_tokens": 59462412.0,
      "step": 25960
    },
    {
      "entropy": 5.709722423553467,
      "epoch": 2.566725978647687,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043446542322105595,
      "loss": 5.0095,
      "mean_token_accuracy": 0.21773470491170882,
      "num_tokens": 59474784.0,
      "step": 25965
    },
    {
      "entropy": 5.642324352264405,
      "epoch": 2.567220245156188,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004344405250738994,
      "loss": 4.9024,
      "mean_token_accuracy": 0.22559424340724946,
      "num_tokens": 59487006.0,
      "step": 25970
    },
    {
      "entropy": 5.6816877841949465,
      "epoch": 2.56771451166469,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043441562300448134,
      "loss": 5.003,
      "mean_token_accuracy": 0.2187163144350052,
      "num_tokens": 59498625.0,
      "step": 25975
    },
    {
      "entropy": 5.655596542358398,
      "epoch": 2.568208778173191,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004343907170134142,
      "loss": 4.8907,
      "mean_token_accuracy": 0.22329788655042648,
      "num_tokens": 59509257.0,
      "step": 25980
    },
    {
      "entropy": 5.720489645004273,
      "epoch": 2.5687030446816923,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004343658071013106,
      "loss": 5.1033,
      "mean_token_accuracy": 0.20779407769441605,
      "num_tokens": 59520876.0,
      "step": 25985
    },
    {
      "entropy": 5.677516174316406,
      "epoch": 2.5691973111901936,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043434089326878356,
      "loss": 4.956,
      "mean_token_accuracy": 0.2148314267396927,
      "num_tokens": 59532047.0,
      "step": 25990
    },
    {
      "entropy": 5.66168909072876,
      "epoch": 2.5696915776986953,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004343159755164457,
      "loss": 5.0194,
      "mean_token_accuracy": 0.2183456763625145,
      "num_tokens": 59543831.0,
      "step": 25995
    },
    {
      "entropy": 5.62736439704895,
      "epoch": 2.5701858442071965,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004342910538449102,
      "loss": 4.9133,
      "mean_token_accuracy": 0.22225775420665742,
      "num_tokens": 59555298.0,
      "step": 26000
    },
    {
      "entropy": 5.773473501205444,
      "epoch": 2.570680110715698,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043426612825479003,
      "loss": 5.0033,
      "mean_token_accuracy": 0.21374738216400146,
      "num_tokens": 59566204.0,
      "step": 26005
    },
    {
      "entropy": 5.726179361343384,
      "epoch": 2.5711743772241995,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004342411987466984,
      "loss": 5.0858,
      "mean_token_accuracy": 0.20720468759536742,
      "num_tokens": 59577505.0,
      "step": 26010
    },
    {
      "entropy": 5.701810455322265,
      "epoch": 2.5716686437327008,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004342162653212486,
      "loss": 5.0562,
      "mean_token_accuracy": 0.21396088004112243,
      "num_tokens": 59589430.0,
      "step": 26015
    },
    {
      "entropy": 5.679551410675049,
      "epoch": 2.572162910241202,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004341913279790539,
      "loss": 4.9349,
      "mean_token_accuracy": 0.22740085870027543,
      "num_tokens": 59600328.0,
      "step": 26020
    },
    {
      "entropy": 5.701921367645264,
      "epoch": 2.5726571767497033,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004341663867207279,
      "loss": 4.9977,
      "mean_token_accuracy": 0.21352200508117675,
      "num_tokens": 59611643.0,
      "step": 26025
    },
    {
      "entropy": 5.705938386917114,
      "epoch": 2.5731514432582046,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043414144154688405,
      "loss": 5.0844,
      "mean_token_accuracy": 0.21820371747016906,
      "num_tokens": 59622586.0,
      "step": 26030
    },
    {
      "entropy": 5.752703094482422,
      "epoch": 2.5736457097667063,
      "grad_norm": 1.125,
      "learning_rate": 0.00043411649245813613,
      "loss": 5.0765,
      "mean_token_accuracy": 0.21112840622663498,
      "num_tokens": 59632719.0,
      "step": 26035
    },
    {
      "entropy": 5.672464799880982,
      "epoch": 2.5741399762752075,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004340915394550977,
      "loss": 4.9416,
      "mean_token_accuracy": 0.22603845745325088,
      "num_tokens": 59643788.0,
      "step": 26040
    },
    {
      "entropy": 5.6741680145263675,
      "epoch": 2.5746342427837092,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004340665825383828,
      "loss": 4.989,
      "mean_token_accuracy": 0.22299239039421082,
      "num_tokens": 59655355.0,
      "step": 26045
    },
    {
      "entropy": 5.73401141166687,
      "epoch": 2.5751285092922105,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043404162170860525,
      "loss": 5.1005,
      "mean_token_accuracy": 0.21177371144294738,
      "num_tokens": 59667099.0,
      "step": 26050
    },
    {
      "entropy": 5.692948341369629,
      "epoch": 2.5756227758007118,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004340166569663792,
      "loss": 5.0571,
      "mean_token_accuracy": 0.20790252089500427,
      "num_tokens": 59679514.0,
      "step": 26055
    },
    {
      "entropy": 5.822658920288086,
      "epoch": 2.576117042309213,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043399168831231865,
      "loss": 5.1207,
      "mean_token_accuracy": 0.21273076832294463,
      "num_tokens": 59691569.0,
      "step": 26060
    },
    {
      "entropy": 5.759627771377564,
      "epoch": 2.5766113088177143,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004339667157470379,
      "loss": 5.051,
      "mean_token_accuracy": 0.22027494609355927,
      "num_tokens": 59702841.0,
      "step": 26065
    },
    {
      "entropy": 5.669215250015259,
      "epoch": 2.577105575326216,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004339417392711513,
      "loss": 4.9789,
      "mean_token_accuracy": 0.22170305699110032,
      "num_tokens": 59714834.0,
      "step": 26070
    },
    {
      "entropy": 5.716326189041138,
      "epoch": 2.5775998418347172,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004339167588852733,
      "loss": 5.0488,
      "mean_token_accuracy": 0.2122848302125931,
      "num_tokens": 59726863.0,
      "step": 26075
    },
    {
      "entropy": 5.706021928787232,
      "epoch": 2.5780941083432185,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004338917745900183,
      "loss": 5.0646,
      "mean_token_accuracy": 0.21499428451061248,
      "num_tokens": 59738952.0,
      "step": 26080
    },
    {
      "entropy": 5.655021858215332,
      "epoch": 2.57858837485172,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004338667863860011,
      "loss": 4.9797,
      "mean_token_accuracy": 0.21834299713373184,
      "num_tokens": 59750445.0,
      "step": 26085
    },
    {
      "entropy": 5.689514589309693,
      "epoch": 2.5790826413602215,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004338417942738362,
      "loss": 5.013,
      "mean_token_accuracy": 0.21743761748075485,
      "num_tokens": 59762447.0,
      "step": 26090
    },
    {
      "entropy": 5.61665940284729,
      "epoch": 2.5795769078687227,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004338167982541386,
      "loss": 4.9199,
      "mean_token_accuracy": 0.22988397032022476,
      "num_tokens": 59773122.0,
      "step": 26095
    },
    {
      "entropy": 5.750516128540039,
      "epoch": 2.580071174377224,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043379179832752306,
      "loss": 5.0509,
      "mean_token_accuracy": 0.21873348951339722,
      "num_tokens": 59784355.0,
      "step": 26100
    },
    {
      "entropy": 5.6588757038116455,
      "epoch": 2.5805654408857257,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043376679449460463,
      "loss": 4.92,
      "mean_token_accuracy": 0.23175878673791886,
      "num_tokens": 59796529.0,
      "step": 26105
    },
    {
      "entropy": 5.735718202590943,
      "epoch": 2.581059707394227,
      "grad_norm": 1.34375,
      "learning_rate": 0.00043374178675599853,
      "loss": 4.9888,
      "mean_token_accuracy": 0.2125225469470024,
      "num_tokens": 59807510.0,
      "step": 26110
    },
    {
      "entropy": 5.71296534538269,
      "epoch": 2.5815539739027282,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043371677511231977,
      "loss": 5.0102,
      "mean_token_accuracy": 0.21287502199411393,
      "num_tokens": 59819355.0,
      "step": 26115
    },
    {
      "entropy": 5.680662393569946,
      "epoch": 2.58204824041123,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004336917595641838,
      "loss": 4.959,
      "mean_token_accuracy": 0.2135409876704216,
      "num_tokens": 59830780.0,
      "step": 26120
    },
    {
      "entropy": 5.658672571182251,
      "epoch": 2.582542506919731,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004336667401122058,
      "loss": 5.0338,
      "mean_token_accuracy": 0.21758261620998381,
      "num_tokens": 59842031.0,
      "step": 26125
    },
    {
      "entropy": 5.671532249450683,
      "epoch": 2.5830367734282325,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004336417167570015,
      "loss": 5.015,
      "mean_token_accuracy": 0.2202451393008232,
      "num_tokens": 59852788.0,
      "step": 26130
    },
    {
      "entropy": 5.671050500869751,
      "epoch": 2.5835310399367337,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043361668949918627,
      "loss": 4.9977,
      "mean_token_accuracy": 0.22320355772972106,
      "num_tokens": 59865618.0,
      "step": 26135
    },
    {
      "entropy": 5.721094131469727,
      "epoch": 2.5840253064452354,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004335916583393759,
      "loss": 4.9959,
      "mean_token_accuracy": 0.21211876720190048,
      "num_tokens": 59877053.0,
      "step": 26140
    },
    {
      "entropy": 5.665718698501587,
      "epoch": 2.5845195729537367,
      "grad_norm": 1.125,
      "learning_rate": 0.0004335666232781861,
      "loss": 4.9263,
      "mean_token_accuracy": 0.22258886843919753,
      "num_tokens": 59888730.0,
      "step": 26145
    },
    {
      "entropy": 5.683137845993042,
      "epoch": 2.585013839462238,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004335415843162328,
      "loss": 5.021,
      "mean_token_accuracy": 0.21481456905603408,
      "num_tokens": 59900255.0,
      "step": 26150
    },
    {
      "entropy": 5.693092346191406,
      "epoch": 2.5855081059707397,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043351654145413197,
      "loss": 5.041,
      "mean_token_accuracy": 0.21829826682806014,
      "num_tokens": 59910705.0,
      "step": 26155
    },
    {
      "entropy": 5.620993041992188,
      "epoch": 2.586002372479241,
      "grad_norm": 1.0,
      "learning_rate": 0.0004334914946924996,
      "loss": 4.8975,
      "mean_token_accuracy": 0.22369939982891082,
      "num_tokens": 59923199.0,
      "step": 26160
    },
    {
      "entropy": 5.709030675888061,
      "epoch": 2.586496638987742,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043346644403195186,
      "loss": 4.9835,
      "mean_token_accuracy": 0.21240891814231871,
      "num_tokens": 59934098.0,
      "step": 26165
    },
    {
      "entropy": 5.707919883728027,
      "epoch": 2.5869909054962434,
      "grad_norm": 1.0,
      "learning_rate": 0.000433441389473105,
      "loss": 5.0178,
      "mean_token_accuracy": 0.2137576088309288,
      "num_tokens": 59945566.0,
      "step": 26170
    },
    {
      "entropy": 5.770123863220215,
      "epoch": 2.5874851720047447,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004334163310165754,
      "loss": 5.0685,
      "mean_token_accuracy": 0.21106477230787277,
      "num_tokens": 59956354.0,
      "step": 26175
    },
    {
      "entropy": 5.6716385841369625,
      "epoch": 2.5879794385132464,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043339126866297943,
      "loss": 5.0263,
      "mean_token_accuracy": 0.21532163619995118,
      "num_tokens": 59967413.0,
      "step": 26180
    },
    {
      "entropy": 5.704101276397705,
      "epoch": 2.5884737050217477,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004333662024129337,
      "loss": 5.0138,
      "mean_token_accuracy": 0.2163545846939087,
      "num_tokens": 59977936.0,
      "step": 26185
    },
    {
      "entropy": 5.7335728168487545,
      "epoch": 2.5889679715302494,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004333411322670549,
      "loss": 5.0148,
      "mean_token_accuracy": 0.21911537498235703,
      "num_tokens": 59989566.0,
      "step": 26190
    },
    {
      "entropy": 5.7011682987213135,
      "epoch": 2.5894622380387506,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043331605822595964,
      "loss": 5.0091,
      "mean_token_accuracy": 0.2202601134777069,
      "num_tokens": 60000339.0,
      "step": 26195
    },
    {
      "entropy": 5.66443099975586,
      "epoch": 2.589956504547252,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043329098029026484,
      "loss": 4.9437,
      "mean_token_accuracy": 0.2179483577609062,
      "num_tokens": 60012903.0,
      "step": 26200
    },
    {
      "entropy": 5.735072898864746,
      "epoch": 2.590450771055753,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043326589846058724,
      "loss": 5.0431,
      "mean_token_accuracy": 0.22036902308464051,
      "num_tokens": 60024135.0,
      "step": 26205
    },
    {
      "entropy": 5.6460906028747555,
      "epoch": 2.5909450375642544,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043324081273754403,
      "loss": 4.999,
      "mean_token_accuracy": 0.2155824527144432,
      "num_tokens": 60036314.0,
      "step": 26210
    },
    {
      "entropy": 5.7402331829071045,
      "epoch": 2.591439304072756,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043321572312175234,
      "loss": 4.9841,
      "mean_token_accuracy": 0.21894258856773377,
      "num_tokens": 60047337.0,
      "step": 26215
    },
    {
      "entropy": 5.764920091629028,
      "epoch": 2.5919335705812574,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043319062961382924,
      "loss": 5.0496,
      "mean_token_accuracy": 0.2134683459997177,
      "num_tokens": 60057875.0,
      "step": 26220
    },
    {
      "entropy": 5.639710235595703,
      "epoch": 2.5924278370897587,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043316553221439214,
      "loss": 5.0187,
      "mean_token_accuracy": 0.21831004321575165,
      "num_tokens": 60069577.0,
      "step": 26225
    },
    {
      "entropy": 5.71601128578186,
      "epoch": 2.5929221035982604,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043314043092405836,
      "loss": 5.0633,
      "mean_token_accuracy": 0.21288186460733413,
      "num_tokens": 60081744.0,
      "step": 26230
    },
    {
      "entropy": 5.772866678237915,
      "epoch": 2.5934163701067616,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004331153257434455,
      "loss": 5.0589,
      "mean_token_accuracy": 0.21265597343444825,
      "num_tokens": 60092837.0,
      "step": 26235
    },
    {
      "entropy": 5.681906890869141,
      "epoch": 2.593910636615263,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004330902166731711,
      "loss": 5.0064,
      "mean_token_accuracy": 0.21188703924417496,
      "num_tokens": 60104512.0,
      "step": 26240
    },
    {
      "entropy": 5.638290977478027,
      "epoch": 2.594404903123764,
      "grad_norm": 1.125,
      "learning_rate": 0.0004330651037138529,
      "loss": 4.9741,
      "mean_token_accuracy": 0.22064238637685776,
      "num_tokens": 60115652.0,
      "step": 26245
    },
    {
      "entropy": 5.734539890289307,
      "epoch": 2.594899169632266,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004330399868661085,
      "loss": 5.0116,
      "mean_token_accuracy": 0.22068501859903336,
      "num_tokens": 60127834.0,
      "step": 26250
    },
    {
      "entropy": 5.73403959274292,
      "epoch": 2.595393436140767,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000433014866130556,
      "loss": 5.033,
      "mean_token_accuracy": 0.21219731271266937,
      "num_tokens": 60140200.0,
      "step": 26255
    },
    {
      "entropy": 5.703223133087159,
      "epoch": 2.5958877026492684,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043298974150781326,
      "loss": 5.0727,
      "mean_token_accuracy": 0.21635280102491378,
      "num_tokens": 60150506.0,
      "step": 26260
    },
    {
      "entropy": 5.73765115737915,
      "epoch": 2.59638196915777,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043296461299849835,
      "loss": 5.1302,
      "mean_token_accuracy": 0.2085862398147583,
      "num_tokens": 60162107.0,
      "step": 26265
    },
    {
      "entropy": 5.775042867660522,
      "epoch": 2.5968762356662713,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043293948060322944,
      "loss": 5.0807,
      "mean_token_accuracy": 0.21121090799570083,
      "num_tokens": 60174435.0,
      "step": 26270
    },
    {
      "entropy": 5.776816701889038,
      "epoch": 2.5973705021747726,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004329143443226249,
      "loss": 5.1135,
      "mean_token_accuracy": 0.2144237220287323,
      "num_tokens": 60186319.0,
      "step": 26275
    },
    {
      "entropy": 5.729561376571655,
      "epoch": 2.597864768683274,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004328892041573029,
      "loss": 5.0506,
      "mean_token_accuracy": 0.21685020625591278,
      "num_tokens": 60197981.0,
      "step": 26280
    },
    {
      "entropy": 5.644644737243652,
      "epoch": 2.598359035191775,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000432864060107882,
      "loss": 4.9111,
      "mean_token_accuracy": 0.22440594732761382,
      "num_tokens": 60209086.0,
      "step": 26285
    },
    {
      "entropy": 5.672186231613159,
      "epoch": 2.598853301700277,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043283891217498074,
      "loss": 4.9808,
      "mean_token_accuracy": 0.22676471173763274,
      "num_tokens": 60221040.0,
      "step": 26290
    },
    {
      "entropy": 5.628039836883545,
      "epoch": 2.599347568208778,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004328137603592177,
      "loss": 4.8972,
      "mean_token_accuracy": 0.22752701938152314,
      "num_tokens": 60232709.0,
      "step": 26295
    },
    {
      "entropy": 5.692705202102661,
      "epoch": 2.59984183471728,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004327886046612117,
      "loss": 5.0103,
      "mean_token_accuracy": 0.21013389825820922,
      "num_tokens": 60244539.0,
      "step": 26300
    },
    {
      "entropy": 5.701635313034058,
      "epoch": 2.600336101225781,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004327634450815817,
      "loss": 4.9883,
      "mean_token_accuracy": 0.22222006767988206,
      "num_tokens": 60255124.0,
      "step": 26305
    },
    {
      "entropy": 5.792408800125122,
      "epoch": 2.6008303677342823,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043273828162094624,
      "loss": 5.1414,
      "mean_token_accuracy": 0.20262690931558608,
      "num_tokens": 60267819.0,
      "step": 26310
    },
    {
      "entropy": 5.710248613357544,
      "epoch": 2.6013246342427836,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004327131142799247,
      "loss": 5.0455,
      "mean_token_accuracy": 0.21518499553203582,
      "num_tokens": 60279694.0,
      "step": 26315
    },
    {
      "entropy": 5.7176896095275875,
      "epoch": 2.601818900751285,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004326879430591361,
      "loss": 5.0485,
      "mean_token_accuracy": 0.21125290989875795,
      "num_tokens": 60291215.0,
      "step": 26320
    },
    {
      "entropy": 5.6756532192230225,
      "epoch": 2.6023131672597866,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043266276795919954,
      "loss": 4.962,
      "mean_token_accuracy": 0.21455197185277938,
      "num_tokens": 60303639.0,
      "step": 26325
    },
    {
      "entropy": 5.694013166427612,
      "epoch": 2.602807433768288,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004326375889807345,
      "loss": 4.9921,
      "mean_token_accuracy": 0.21615112721920013,
      "num_tokens": 60315290.0,
      "step": 26330
    },
    {
      "entropy": 5.696866655349732,
      "epoch": 2.603301700276789,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004326124061243603,
      "loss": 5.0093,
      "mean_token_accuracy": 0.21491447389125823,
      "num_tokens": 60325391.0,
      "step": 26335
    },
    {
      "entropy": 5.681035947799683,
      "epoch": 2.603795966785291,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004325872193906965,
      "loss": 4.9303,
      "mean_token_accuracy": 0.2281200110912323,
      "num_tokens": 60335758.0,
      "step": 26340
    },
    {
      "entropy": 5.720879220962525,
      "epoch": 2.604290233293792,
      "grad_norm": 1.125,
      "learning_rate": 0.00043256202878036264,
      "loss": 4.9956,
      "mean_token_accuracy": 0.2181279554963112,
      "num_tokens": 60347365.0,
      "step": 26345
    },
    {
      "entropy": 5.629631662368775,
      "epoch": 2.6047844998022933,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043253683429397843,
      "loss": 4.8947,
      "mean_token_accuracy": 0.2234208807349205,
      "num_tokens": 60359222.0,
      "step": 26350
    },
    {
      "entropy": 5.71719970703125,
      "epoch": 2.6052787663107946,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043251163593216365,
      "loss": 5.0933,
      "mean_token_accuracy": 0.205734683573246,
      "num_tokens": 60370730.0,
      "step": 26355
    },
    {
      "entropy": 5.638969278335571,
      "epoch": 2.6057730328192963,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043248643369553813,
      "loss": 4.9467,
      "mean_token_accuracy": 0.22514139711856843,
      "num_tokens": 60381955.0,
      "step": 26360
    },
    {
      "entropy": 5.756603097915649,
      "epoch": 2.6062672993277975,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000432461227584722,
      "loss": 5.063,
      "mean_token_accuracy": 0.20892783850431443,
      "num_tokens": 60395331.0,
      "step": 26365
    },
    {
      "entropy": 5.6695075035095215,
      "epoch": 2.606761565836299,
      "grad_norm": 1.125,
      "learning_rate": 0.0004324360176003352,
      "loss": 4.9334,
      "mean_token_accuracy": 0.22383388578891755,
      "num_tokens": 60407179.0,
      "step": 26370
    },
    {
      "entropy": 5.62222728729248,
      "epoch": 2.6072558323448005,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004324108037429979,
      "loss": 5.0021,
      "mean_token_accuracy": 0.21616304516792298,
      "num_tokens": 60418377.0,
      "step": 26375
    },
    {
      "entropy": 5.709018659591675,
      "epoch": 2.607750098853302,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004323855860133305,
      "loss": 4.9863,
      "mean_token_accuracy": 0.21497878581285476,
      "num_tokens": 60428802.0,
      "step": 26380
    },
    {
      "entropy": 5.695616579055786,
      "epoch": 2.608244365361803,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004323603644119532,
      "loss": 5.0368,
      "mean_token_accuracy": 0.2188412845134735,
      "num_tokens": 60440445.0,
      "step": 26385
    },
    {
      "entropy": 5.69155650138855,
      "epoch": 2.6087386318703043,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043233513893948654,
      "loss": 4.9803,
      "mean_token_accuracy": 0.21946085542440413,
      "num_tokens": 60451040.0,
      "step": 26390
    },
    {
      "entropy": 5.706317377090454,
      "epoch": 2.609232898378806,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004323099095965511,
      "loss": 5.0782,
      "mean_token_accuracy": 0.21186997890472412,
      "num_tokens": 60463820.0,
      "step": 26395
    },
    {
      "entropy": 5.713330984115601,
      "epoch": 2.6097271648873073,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004322846763837674,
      "loss": 4.9887,
      "mean_token_accuracy": 0.2237784370779991,
      "num_tokens": 60474285.0,
      "step": 26400
    },
    {
      "entropy": 5.649153089523315,
      "epoch": 2.6102214313958085,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043225943930175626,
      "loss": 4.9235,
      "mean_token_accuracy": 0.23180877715349196,
      "num_tokens": 60484475.0,
      "step": 26405
    },
    {
      "entropy": 5.683869695663452,
      "epoch": 2.6107156979043102,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004322341983511386,
      "loss": 4.9295,
      "mean_token_accuracy": 0.22741284668445588,
      "num_tokens": 60494182.0,
      "step": 26410
    },
    {
      "entropy": 5.673253679275513,
      "epoch": 2.6112099644128115,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043220895353253516,
      "loss": 4.9893,
      "mean_token_accuracy": 0.22148875445127486,
      "num_tokens": 60505455.0,
      "step": 26415
    },
    {
      "entropy": 5.6130575180053714,
      "epoch": 2.6117042309213128,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043218370484656717,
      "loss": 4.9325,
      "mean_token_accuracy": 0.22093128114938737,
      "num_tokens": 60518040.0,
      "step": 26420
    },
    {
      "entropy": 5.706735229492187,
      "epoch": 2.612198497429814,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043215845229385567,
      "loss": 5.029,
      "mean_token_accuracy": 0.2197766959667206,
      "num_tokens": 60528896.0,
      "step": 26425
    },
    {
      "entropy": 5.718177747726441,
      "epoch": 2.6126927639383153,
      "grad_norm": 1.125,
      "learning_rate": 0.00043213319587502173,
      "loss": 4.9286,
      "mean_token_accuracy": 0.22685479372739792,
      "num_tokens": 60539505.0,
      "step": 26430
    },
    {
      "entropy": 5.66402759552002,
      "epoch": 2.613187030446817,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004321079355906869,
      "loss": 4.9976,
      "mean_token_accuracy": 0.22045601308345794,
      "num_tokens": 60550544.0,
      "step": 26435
    },
    {
      "entropy": 5.752495527267456,
      "epoch": 2.6136812969553183,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043208267144147244,
      "loss": 5.0892,
      "mean_token_accuracy": 0.20932816863059997,
      "num_tokens": 60561697.0,
      "step": 26440
    },
    {
      "entropy": 5.650682497024536,
      "epoch": 2.61417556346382,
      "grad_norm": 1.0,
      "learning_rate": 0.00043205740342799995,
      "loss": 4.9071,
      "mean_token_accuracy": 0.2291514277458191,
      "num_tokens": 60572817.0,
      "step": 26445
    },
    {
      "entropy": 5.6928071022033695,
      "epoch": 2.6146698299723212,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043203213155089095,
      "loss": 4.9964,
      "mean_token_accuracy": 0.21810807138681412,
      "num_tokens": 60584679.0,
      "step": 26450
    },
    {
      "entropy": 5.6843626499176025,
      "epoch": 2.6151640964808225,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004320068558107671,
      "loss": 5.0098,
      "mean_token_accuracy": 0.21317216604948044,
      "num_tokens": 60594871.0,
      "step": 26455
    },
    {
      "entropy": 5.634895849227905,
      "epoch": 2.6156583629893237,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043198157620825023,
      "loss": 5.016,
      "mean_token_accuracy": 0.22486073076725005,
      "num_tokens": 60605626.0,
      "step": 26460
    },
    {
      "entropy": 5.645549631118774,
      "epoch": 2.616152629497825,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043195629274396237,
      "loss": 5.0038,
      "mean_token_accuracy": 0.22918743789196014,
      "num_tokens": 60617802.0,
      "step": 26465
    },
    {
      "entropy": 5.738967418670654,
      "epoch": 2.6166468960063267,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043193100541852526,
      "loss": 5.0601,
      "mean_token_accuracy": 0.21122468411922454,
      "num_tokens": 60629044.0,
      "step": 26470
    },
    {
      "entropy": 5.6645190715789795,
      "epoch": 2.617141162514828,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000431905714232561,
      "loss": 4.9691,
      "mean_token_accuracy": 0.22353244870901107,
      "num_tokens": 60639729.0,
      "step": 26475
    },
    {
      "entropy": 5.696509981155396,
      "epoch": 2.6176354290233292,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043188041918669197,
      "loss": 5.0588,
      "mean_token_accuracy": 0.2144579604268074,
      "num_tokens": 60650499.0,
      "step": 26480
    },
    {
      "entropy": 5.784479379653931,
      "epoch": 2.618129695531831,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004318551202815402,
      "loss": 5.1156,
      "mean_token_accuracy": 0.2068298026919365,
      "num_tokens": 60661453.0,
      "step": 26485
    },
    {
      "entropy": 5.724158763885498,
      "epoch": 2.618623962040332,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043182981751772816,
      "loss": 4.9935,
      "mean_token_accuracy": 0.21910915672779083,
      "num_tokens": 60672748.0,
      "step": 26490
    },
    {
      "entropy": 5.653290700912476,
      "epoch": 2.6191182285488335,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004318045108958783,
      "loss": 4.9718,
      "mean_token_accuracy": 0.22034027725458144,
      "num_tokens": 60683679.0,
      "step": 26495
    },
    {
      "entropy": 5.649635457992554,
      "epoch": 2.6196124950573347,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043177920041661313,
      "loss": 4.9374,
      "mean_token_accuracy": 0.2231195628643036,
      "num_tokens": 60696988.0,
      "step": 26500
    },
    {
      "entropy": 5.716681289672851,
      "epoch": 2.6201067615658364,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004317538860805553,
      "loss": 5.1158,
      "mean_token_accuracy": 0.21333537697792054,
      "num_tokens": 60709817.0,
      "step": 26505
    },
    {
      "entropy": 5.753416681289673,
      "epoch": 2.6206010280743377,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043172856788832755,
      "loss": 5.0394,
      "mean_token_accuracy": 0.21691515445709228,
      "num_tokens": 60720941.0,
      "step": 26510
    },
    {
      "entropy": 5.770007085800171,
      "epoch": 2.621095294582839,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043170324584055275,
      "loss": 5.1285,
      "mean_token_accuracy": 0.208763886988163,
      "num_tokens": 60731873.0,
      "step": 26515
    },
    {
      "entropy": 5.6864594459533695,
      "epoch": 2.6215895610913407,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043167791993785375,
      "loss": 5.0071,
      "mean_token_accuracy": 0.22098004668951035,
      "num_tokens": 60744361.0,
      "step": 26520
    },
    {
      "entropy": 5.733385515213013,
      "epoch": 2.622083827599842,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004316525901808536,
      "loss": 5.052,
      "mean_token_accuracy": 0.2154487207531929,
      "num_tokens": 60756111.0,
      "step": 26525
    },
    {
      "entropy": 5.733601903915405,
      "epoch": 2.622578094108343,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043162725657017544,
      "loss": 4.9063,
      "mean_token_accuracy": 0.22673411220312117,
      "num_tokens": 60766602.0,
      "step": 26530
    },
    {
      "entropy": 5.731798601150513,
      "epoch": 2.6230723606168445,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043160191910644253,
      "loss": 5.031,
      "mean_token_accuracy": 0.2158122882246971,
      "num_tokens": 60778670.0,
      "step": 26535
    },
    {
      "entropy": 5.73510332107544,
      "epoch": 2.6235666271253457,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004315765777902781,
      "loss": 5.0937,
      "mean_token_accuracy": 0.21073362827301026,
      "num_tokens": 60790026.0,
      "step": 26540
    },
    {
      "entropy": 5.724614810943604,
      "epoch": 2.6240608936338474,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004315512326223055,
      "loss": 5.0712,
      "mean_token_accuracy": 0.21171680986881256,
      "num_tokens": 60801325.0,
      "step": 26545
    },
    {
      "entropy": 5.733736658096314,
      "epoch": 2.6245551601423487,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004315258836031483,
      "loss": 5.0531,
      "mean_token_accuracy": 0.2118220865726471,
      "num_tokens": 60813421.0,
      "step": 26550
    },
    {
      "entropy": 5.750084924697876,
      "epoch": 2.6250494266508504,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004315005307334301,
      "loss": 5.0483,
      "mean_token_accuracy": 0.20678311735391616,
      "num_tokens": 60826447.0,
      "step": 26555
    },
    {
      "entropy": 5.7129754543304445,
      "epoch": 2.6255436931593517,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004314751740137746,
      "loss": 5.0007,
      "mean_token_accuracy": 0.21546009182929993,
      "num_tokens": 60837663.0,
      "step": 26560
    },
    {
      "entropy": 5.733322811126709,
      "epoch": 2.626037959667853,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043144981344480553,
      "loss": 5.0084,
      "mean_token_accuracy": 0.21074688732624053,
      "num_tokens": 60849040.0,
      "step": 26565
    },
    {
      "entropy": 5.674941349029541,
      "epoch": 2.626532226176354,
      "grad_norm": 1.125,
      "learning_rate": 0.00043142444902714676,
      "loss": 5.0626,
      "mean_token_accuracy": 0.21186541318893432,
      "num_tokens": 60862112.0,
      "step": 26570
    },
    {
      "entropy": 5.660215759277344,
      "epoch": 2.6270264926848554,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043139908076142224,
      "loss": 4.9906,
      "mean_token_accuracy": 0.21630873531103134,
      "num_tokens": 60872822.0,
      "step": 26575
    },
    {
      "entropy": 5.724995803833008,
      "epoch": 2.627520759193357,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004313737086482562,
      "loss": 5.0735,
      "mean_token_accuracy": 0.20913240760564805,
      "num_tokens": 60884374.0,
      "step": 26580
    },
    {
      "entropy": 5.719517993927002,
      "epoch": 2.6280150257018584,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004313483326882725,
      "loss": 5.0546,
      "mean_token_accuracy": 0.2164910092949867,
      "num_tokens": 60895490.0,
      "step": 26585
    },
    {
      "entropy": 5.62080864906311,
      "epoch": 2.6285092922103597,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004313229528820957,
      "loss": 4.9081,
      "mean_token_accuracy": 0.22872539460659028,
      "num_tokens": 60906190.0,
      "step": 26590
    },
    {
      "entropy": 5.708775281906128,
      "epoch": 2.6290035587188614,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043129756923034995,
      "loss": 5.0291,
      "mean_token_accuracy": 0.21680515855550767,
      "num_tokens": 60918466.0,
      "step": 26595
    },
    {
      "entropy": 5.650860691070557,
      "epoch": 2.6294978252273626,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004312721817336597,
      "loss": 4.9135,
      "mean_token_accuracy": 0.22271667122840882,
      "num_tokens": 60930202.0,
      "step": 26600
    },
    {
      "entropy": 5.699853038787841,
      "epoch": 2.629992091735864,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043124679039264964,
      "loss": 5.0346,
      "mean_token_accuracy": 0.21889408230781554,
      "num_tokens": 60942089.0,
      "step": 26605
    },
    {
      "entropy": 5.710665321350097,
      "epoch": 2.630486358244365,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043122139520794425,
      "loss": 4.961,
      "mean_token_accuracy": 0.21609577685594558,
      "num_tokens": 60954493.0,
      "step": 26610
    },
    {
      "entropy": 5.654888439178467,
      "epoch": 2.630980624752867,
      "grad_norm": 1.3125,
      "learning_rate": 0.00043119599618016824,
      "loss": 5.014,
      "mean_token_accuracy": 0.2144119545817375,
      "num_tokens": 60965888.0,
      "step": 26615
    },
    {
      "entropy": 5.667346096038818,
      "epoch": 2.631474891261368,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004311705933099466,
      "loss": 4.954,
      "mean_token_accuracy": 0.21972678303718568,
      "num_tokens": 60976414.0,
      "step": 26620
    },
    {
      "entropy": 5.660156106948852,
      "epoch": 2.6319691577698694,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004311451865979041,
      "loss": 4.9472,
      "mean_token_accuracy": 0.2212633639574051,
      "num_tokens": 60986353.0,
      "step": 26625
    },
    {
      "entropy": 5.686699295043946,
      "epoch": 2.632463424278371,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043111977604466575,
      "loss": 4.9894,
      "mean_token_accuracy": 0.2276598870754242,
      "num_tokens": 60996916.0,
      "step": 26630
    },
    {
      "entropy": 5.590903663635254,
      "epoch": 2.6329576907868724,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004310943616508567,
      "loss": 4.935,
      "mean_token_accuracy": 0.22646029591560363,
      "num_tokens": 61008741.0,
      "step": 26635
    },
    {
      "entropy": 5.754882335662842,
      "epoch": 2.6334519572953736,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004310689434171022,
      "loss": 5.1152,
      "mean_token_accuracy": 0.20967260897159576,
      "num_tokens": 61019988.0,
      "step": 26640
    },
    {
      "entropy": 5.749897718429565,
      "epoch": 2.633946223803875,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043104352134402734,
      "loss": 5.0295,
      "mean_token_accuracy": 0.21297657936811448,
      "num_tokens": 61031667.0,
      "step": 26645
    },
    {
      "entropy": 5.729730749130249,
      "epoch": 2.6344404903123766,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043101809543225777,
      "loss": 4.9932,
      "mean_token_accuracy": 0.2207876205444336,
      "num_tokens": 61042203.0,
      "step": 26650
    },
    {
      "entropy": 5.665720987319946,
      "epoch": 2.634934756820878,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043099266568241885,
      "loss": 4.9682,
      "mean_token_accuracy": 0.2185746029019356,
      "num_tokens": 61054243.0,
      "step": 26655
    },
    {
      "entropy": 5.6023646831512455,
      "epoch": 2.635429023329379,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043096723209513606,
      "loss": 4.837,
      "mean_token_accuracy": 0.23610788583755493,
      "num_tokens": 61064068.0,
      "step": 26660
    },
    {
      "entropy": 5.606233787536621,
      "epoch": 2.635923289837881,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004309417946710352,
      "loss": 4.9263,
      "mean_token_accuracy": 0.2320251926779747,
      "num_tokens": 61075273.0,
      "step": 26665
    },
    {
      "entropy": 5.717097520828247,
      "epoch": 2.636417556346382,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000430916353410742,
      "loss": 5.028,
      "mean_token_accuracy": 0.21700633466243743,
      "num_tokens": 61086536.0,
      "step": 26670
    },
    {
      "entropy": 5.671091985702515,
      "epoch": 2.6369118228548833,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004308909083148824,
      "loss": 4.9949,
      "mean_token_accuracy": 0.2183769166469574,
      "num_tokens": 61097567.0,
      "step": 26675
    },
    {
      "entropy": 5.679510927200317,
      "epoch": 2.6374060893633846,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004308654593840822,
      "loss": 4.9601,
      "mean_token_accuracy": 0.21929265111684798,
      "num_tokens": 61108379.0,
      "step": 26680
    },
    {
      "entropy": 5.618005895614624,
      "epoch": 2.637900355871886,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043084000661896757,
      "loss": 4.9705,
      "mean_token_accuracy": 0.22360235303640366,
      "num_tokens": 61119715.0,
      "step": 26685
    },
    {
      "entropy": 5.718747329711914,
      "epoch": 2.6383946223803876,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004308145500201645,
      "loss": 5.0811,
      "mean_token_accuracy": 0.2129189118742943,
      "num_tokens": 61131810.0,
      "step": 26690
    },
    {
      "entropy": 5.6651510238647464,
      "epoch": 2.638888888888889,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043078908958829945,
      "loss": 4.9469,
      "mean_token_accuracy": 0.22582807540893554,
      "num_tokens": 61142771.0,
      "step": 26695
    },
    {
      "entropy": 5.640940237045288,
      "epoch": 2.6393831553973905,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043076362532399855,
      "loss": 4.9916,
      "mean_token_accuracy": 0.21778007298707963,
      "num_tokens": 61154105.0,
      "step": 26700
    },
    {
      "entropy": 5.707059526443482,
      "epoch": 2.639877421905892,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004307381572278882,
      "loss": 4.9124,
      "mean_token_accuracy": 0.2288718491792679,
      "num_tokens": 61164574.0,
      "step": 26705
    },
    {
      "entropy": 5.730502605438232,
      "epoch": 2.640371688414393,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004307126853005952,
      "loss": 5.0712,
      "mean_token_accuracy": 0.21152730286121368,
      "num_tokens": 61177085.0,
      "step": 26710
    },
    {
      "entropy": 5.698912286758423,
      "epoch": 2.6408659549228943,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004306872095427459,
      "loss": 4.961,
      "mean_token_accuracy": 0.22628767937421798,
      "num_tokens": 61188313.0,
      "step": 26715
    },
    {
      "entropy": 5.616076183319092,
      "epoch": 2.6413602214313956,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004306617299549672,
      "loss": 4.9214,
      "mean_token_accuracy": 0.22424285411834716,
      "num_tokens": 61199726.0,
      "step": 26720
    },
    {
      "entropy": 5.704315519332885,
      "epoch": 2.6418544879398973,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004306362465378857,
      "loss": 4.9895,
      "mean_token_accuracy": 0.21507151424884796,
      "num_tokens": 61211861.0,
      "step": 26725
    },
    {
      "entropy": 5.699096059799194,
      "epoch": 2.6423487544483986,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043061075929212836,
      "loss": 4.9743,
      "mean_token_accuracy": 0.21925762742757798,
      "num_tokens": 61222464.0,
      "step": 26730
    },
    {
      "entropy": 5.724076890945435,
      "epoch": 2.6428430209569,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004305852682183222,
      "loss": 4.9692,
      "mean_token_accuracy": 0.21709523350000381,
      "num_tokens": 61233678.0,
      "step": 26735
    },
    {
      "entropy": 5.662415981292725,
      "epoch": 2.6433372874654015,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004305597733170944,
      "loss": 5.0318,
      "mean_token_accuracy": 0.21709883213043213,
      "num_tokens": 61244636.0,
      "step": 26740
    },
    {
      "entropy": 5.748428106307983,
      "epoch": 2.643831553973903,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043053427458907196,
      "loss": 4.9959,
      "mean_token_accuracy": 0.21096169501543044,
      "num_tokens": 61256334.0,
      "step": 26745
    },
    {
      "entropy": 5.741091728210449,
      "epoch": 2.644325820482404,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004305087720348823,
      "loss": 5.0252,
      "mean_token_accuracy": 0.2193573459982872,
      "num_tokens": 61268614.0,
      "step": 26750
    },
    {
      "entropy": 5.7390929698944095,
      "epoch": 2.6448200869909053,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043048326565515267,
      "loss": 5.0753,
      "mean_token_accuracy": 0.21022985875606537,
      "num_tokens": 61280220.0,
      "step": 26755
    },
    {
      "entropy": 5.715899705886841,
      "epoch": 2.645314353499407,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043045775545051055,
      "loss": 4.9253,
      "mean_token_accuracy": 0.2259628102183342,
      "num_tokens": 61290796.0,
      "step": 26760
    },
    {
      "entropy": 5.690618515014648,
      "epoch": 2.6458086200079083,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004304322414215836,
      "loss": 5.0318,
      "mean_token_accuracy": 0.21538103222846985,
      "num_tokens": 61303645.0,
      "step": 26765
    },
    {
      "entropy": 5.662274408340454,
      "epoch": 2.6463028865164095,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043040672356899946,
      "loss": 5.0058,
      "mean_token_accuracy": 0.21806153655052185,
      "num_tokens": 61316641.0,
      "step": 26770
    },
    {
      "entropy": 5.721528482437134,
      "epoch": 2.6467971530249113,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043038120189338566,
      "loss": 5.1016,
      "mean_token_accuracy": 0.20748539716005326,
      "num_tokens": 61329314.0,
      "step": 26775
    },
    {
      "entropy": 5.742600202560425,
      "epoch": 2.6472914195334125,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043035567639537026,
      "loss": 5.0899,
      "mean_token_accuracy": 0.21027597934007644,
      "num_tokens": 61340847.0,
      "step": 26780
    },
    {
      "entropy": 5.8133955001831055,
      "epoch": 2.6477856860419138,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043033014707558113,
      "loss": 5.0638,
      "mean_token_accuracy": 0.21224619895219804,
      "num_tokens": 61351875.0,
      "step": 26785
    },
    {
      "entropy": 5.71115140914917,
      "epoch": 2.648279952550415,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004303046139346462,
      "loss": 5.0384,
      "mean_token_accuracy": 0.22300745993852616,
      "num_tokens": 61364441.0,
      "step": 26790
    },
    {
      "entropy": 5.691520071029663,
      "epoch": 2.6487742190589163,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004302790769731938,
      "loss": 4.9942,
      "mean_token_accuracy": 0.21748262643814087,
      "num_tokens": 61375084.0,
      "step": 26795
    },
    {
      "entropy": 5.743958139419556,
      "epoch": 2.649268485567418,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043025353619185195,
      "loss": 5.0712,
      "mean_token_accuracy": 0.20926353633403777,
      "num_tokens": 61386036.0,
      "step": 26800
    },
    {
      "entropy": 5.728154611587525,
      "epoch": 2.6497627520759193,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043022799159124904,
      "loss": 5.0144,
      "mean_token_accuracy": 0.22576939314603806,
      "num_tokens": 61396006.0,
      "step": 26805
    },
    {
      "entropy": 5.72993049621582,
      "epoch": 2.650257018584421,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043020244317201344,
      "loss": 5.0867,
      "mean_token_accuracy": 0.21387850046157836,
      "num_tokens": 61406453.0,
      "step": 26810
    },
    {
      "entropy": 5.715651893615723,
      "epoch": 2.6507512850929222,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004301768909347737,
      "loss": 5.0761,
      "mean_token_accuracy": 0.21549927890300752,
      "num_tokens": 61418271.0,
      "step": 26815
    },
    {
      "entropy": 5.79262957572937,
      "epoch": 2.6512455516014235,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004301513348801583,
      "loss": 5.0159,
      "mean_token_accuracy": 0.21253803372383118,
      "num_tokens": 61430575.0,
      "step": 26820
    },
    {
      "entropy": 5.630987977981567,
      "epoch": 2.6517398181099248,
      "grad_norm": 1.015625,
      "learning_rate": 0.000430125775008796,
      "loss": 4.9287,
      "mean_token_accuracy": 0.22457814514636992,
      "num_tokens": 61442336.0,
      "step": 26825
    },
    {
      "entropy": 5.71291675567627,
      "epoch": 2.652234084618426,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043010021132131564,
      "loss": 4.9984,
      "mean_token_accuracy": 0.21855925470590593,
      "num_tokens": 61453945.0,
      "step": 26830
    },
    {
      "entropy": 5.7469401359558105,
      "epoch": 2.6527283511269277,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000430074643818346,
      "loss": 4.9856,
      "mean_token_accuracy": 0.2194122329354286,
      "num_tokens": 61463877.0,
      "step": 26835
    },
    {
      "entropy": 5.6636425971984865,
      "epoch": 2.653222617635429,
      "grad_norm": 1.3125,
      "learning_rate": 0.00043004907250051606,
      "loss": 4.9188,
      "mean_token_accuracy": 0.22524941563606263,
      "num_tokens": 61473278.0,
      "step": 26840
    },
    {
      "entropy": 5.71568193435669,
      "epoch": 2.6537168841439303,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004300234973684548,
      "loss": 5.0838,
      "mean_token_accuracy": 0.2134735956788063,
      "num_tokens": 61486038.0,
      "step": 26845
    },
    {
      "entropy": 5.6928071022033695,
      "epoch": 2.654211150652432,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004299979184227916,
      "loss": 5.0253,
      "mean_token_accuracy": 0.2130693405866623,
      "num_tokens": 61497722.0,
      "step": 26850
    },
    {
      "entropy": 5.684412240982056,
      "epoch": 2.6547054171609332,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042997233566415547,
      "loss": 4.9746,
      "mean_token_accuracy": 0.2159498155117035,
      "num_tokens": 61508791.0,
      "step": 26855
    },
    {
      "entropy": 5.687562656402588,
      "epoch": 2.6551996836694345,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042994674909317584,
      "loss": 4.9765,
      "mean_token_accuracy": 0.22060172110795975,
      "num_tokens": 61520047.0,
      "step": 26860
    },
    {
      "entropy": 5.73206057548523,
      "epoch": 2.6556939501779357,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042992115871048214,
      "loss": 4.9717,
      "mean_token_accuracy": 0.21753401458263397,
      "num_tokens": 61531005.0,
      "step": 26865
    },
    {
      "entropy": 5.678264045715332,
      "epoch": 2.6561882166864375,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042989556451670396,
      "loss": 5.0158,
      "mean_token_accuracy": 0.21903164237737655,
      "num_tokens": 61542087.0,
      "step": 26870
    },
    {
      "entropy": 5.673075771331787,
      "epoch": 2.6566824831949387,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004298699665124708,
      "loss": 5.0813,
      "mean_token_accuracy": 0.21386438012123107,
      "num_tokens": 61553575.0,
      "step": 26875
    },
    {
      "entropy": 5.761148929595947,
      "epoch": 2.65717674970344,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042984436469841245,
      "loss": 5.0811,
      "mean_token_accuracy": 0.21132880002260207,
      "num_tokens": 61564833.0,
      "step": 26880
    },
    {
      "entropy": 5.661766624450683,
      "epoch": 2.6576710162119417,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042981875907515874,
      "loss": 4.8688,
      "mean_token_accuracy": 0.23008145838975907,
      "num_tokens": 61575721.0,
      "step": 26885
    },
    {
      "entropy": 5.736316537857055,
      "epoch": 2.658165282720443,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042979314964333953,
      "loss": 5.0207,
      "mean_token_accuracy": 0.21607204526662827,
      "num_tokens": 61587022.0,
      "step": 26890
    },
    {
      "entropy": 5.7123880863189695,
      "epoch": 2.658659549228944,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004297675364035848,
      "loss": 4.9777,
      "mean_token_accuracy": 0.21809093058109283,
      "num_tokens": 61597610.0,
      "step": 26895
    },
    {
      "entropy": 5.8117307186126705,
      "epoch": 2.6591538157374455,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004297419193565247,
      "loss": 5.0879,
      "mean_token_accuracy": 0.20330984741449357,
      "num_tokens": 61608254.0,
      "step": 26900
    },
    {
      "entropy": 5.7257812976837155,
      "epoch": 2.659648082245947,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042971629850278924,
      "loss": 4.9901,
      "mean_token_accuracy": 0.2176758050918579,
      "num_tokens": 61619937.0,
      "step": 26905
    },
    {
      "entropy": 5.6658965110778805,
      "epoch": 2.6601423487544484,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042969067384300905,
      "loss": 5.0247,
      "mean_token_accuracy": 0.2169333964586258,
      "num_tokens": 61630460.0,
      "step": 26910
    },
    {
      "entropy": 5.6873921871185305,
      "epoch": 2.6606366152629497,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042966504537781416,
      "loss": 4.9606,
      "mean_token_accuracy": 0.21972622275352477,
      "num_tokens": 61640855.0,
      "step": 26915
    },
    {
      "entropy": 5.779351615905762,
      "epoch": 2.6611308817714514,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004296394131078351,
      "loss": 5.054,
      "mean_token_accuracy": 0.21082686334848405,
      "num_tokens": 61652405.0,
      "step": 26920
    },
    {
      "entropy": 5.722966814041138,
      "epoch": 2.6616251482799527,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042961377703370265,
      "loss": 4.957,
      "mean_token_accuracy": 0.22248823195695877,
      "num_tokens": 61664379.0,
      "step": 26925
    },
    {
      "entropy": 5.679649591445923,
      "epoch": 2.662119414788454,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004295881371560472,
      "loss": 4.9709,
      "mean_token_accuracy": 0.22487534284591676,
      "num_tokens": 61675259.0,
      "step": 26930
    },
    {
      "entropy": 5.644587135314941,
      "epoch": 2.662613681296955,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042956249347549964,
      "loss": 4.991,
      "mean_token_accuracy": 0.21793091744184495,
      "num_tokens": 61687035.0,
      "step": 26935
    },
    {
      "entropy": 5.755333471298218,
      "epoch": 2.6631079478054565,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004295368459926907,
      "loss": 5.0248,
      "mean_token_accuracy": 0.2154269203543663,
      "num_tokens": 61698082.0,
      "step": 26940
    },
    {
      "entropy": 5.762499904632568,
      "epoch": 2.663602214313958,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042951119470825146,
      "loss": 5.0072,
      "mean_token_accuracy": 0.2109287738800049,
      "num_tokens": 61708533.0,
      "step": 26945
    },
    {
      "entropy": 5.725760889053345,
      "epoch": 2.6640964808224594,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042948553962281286,
      "loss": 5.0377,
      "mean_token_accuracy": 0.210419762134552,
      "num_tokens": 61719301.0,
      "step": 26950
    },
    {
      "entropy": 5.651941204071045,
      "epoch": 2.664590747330961,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004294598807370059,
      "loss": 4.9649,
      "mean_token_accuracy": 0.21873128563165664,
      "num_tokens": 61732225.0,
      "step": 26955
    },
    {
      "entropy": 5.625168895721435,
      "epoch": 2.6650850138394624,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000429434218051462,
      "loss": 4.9576,
      "mean_token_accuracy": 0.2251772940158844,
      "num_tokens": 61743901.0,
      "step": 26960
    },
    {
      "entropy": 5.678807878494263,
      "epoch": 2.6655792803479637,
      "grad_norm": 1.125,
      "learning_rate": 0.0004294085515668123,
      "loss": 4.9848,
      "mean_token_accuracy": 0.2220032051205635,
      "num_tokens": 61756047.0,
      "step": 26965
    },
    {
      "entropy": 5.7197283744812015,
      "epoch": 2.666073546856465,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042938288128368835,
      "loss": 4.9756,
      "mean_token_accuracy": 0.22080026268959047,
      "num_tokens": 61767784.0,
      "step": 26970
    },
    {
      "entropy": 5.7354528427124025,
      "epoch": 2.666567813364966,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042935720720272156,
      "loss": 5.081,
      "mean_token_accuracy": 0.214602030813694,
      "num_tokens": 61778915.0,
      "step": 26975
    },
    {
      "entropy": 5.637601613998413,
      "epoch": 2.667062079873468,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004293315293245435,
      "loss": 4.9918,
      "mean_token_accuracy": 0.2192279428243637,
      "num_tokens": 61791046.0,
      "step": 26980
    },
    {
      "entropy": 5.717722654342651,
      "epoch": 2.667556346381969,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042930584764978597,
      "loss": 4.9755,
      "mean_token_accuracy": 0.2235797256231308,
      "num_tokens": 61802135.0,
      "step": 26985
    },
    {
      "entropy": 5.780620098114014,
      "epoch": 2.6680506128904704,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042928016217908053,
      "loss": 5.039,
      "mean_token_accuracy": 0.22073888927698135,
      "num_tokens": 61815077.0,
      "step": 26990
    },
    {
      "entropy": 5.665139770507812,
      "epoch": 2.668544879398972,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004292544729130592,
      "loss": 4.9312,
      "mean_token_accuracy": 0.22228024899959564,
      "num_tokens": 61826031.0,
      "step": 26995
    },
    {
      "entropy": 5.661031484603882,
      "epoch": 2.6690391459074734,
      "grad_norm": 1.125,
      "learning_rate": 0.00042922877985235395,
      "loss": 4.9553,
      "mean_token_accuracy": 0.22762898355722427,
      "num_tokens": 61836372.0,
      "step": 27000
    },
    {
      "epoch": 2.6690391459074734,
      "eval_entropy": 5.4216988277318,
      "eval_loss": 5.053977012634277,
      "eval_mean_token_accuracy": 0.2238519115843163,
      "eval_num_tokens": 61836372.0,
      "eval_runtime": 20.548,
      "eval_samples_per_second": 1582.299,
      "eval_steps_per_second": 197.83,
      "step": 27000
    },
    {
      "entropy": 5.6321619033813475,
      "epoch": 2.6695334124159746,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042920308299759677,
      "loss": 4.9372,
      "mean_token_accuracy": 0.21940778344869613,
      "num_tokens": 61847829.0,
      "step": 27005
    },
    {
      "entropy": 5.690559816360474,
      "epoch": 2.670027678924476,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004291773823494198,
      "loss": 5.0168,
      "mean_token_accuracy": 0.22404668033123015,
      "num_tokens": 61859074.0,
      "step": 27010
    },
    {
      "entropy": 5.759458208084107,
      "epoch": 2.6705219454329776,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042915167790845535,
      "loss": 5.1186,
      "mean_token_accuracy": 0.20576048642396927,
      "num_tokens": 61868596.0,
      "step": 27015
    },
    {
      "entropy": 5.664997100830078,
      "epoch": 2.671016211941479,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004291259696753356,
      "loss": 5.0191,
      "mean_token_accuracy": 0.2199859455227852,
      "num_tokens": 61879173.0,
      "step": 27020
    },
    {
      "entropy": 5.679371070861817,
      "epoch": 2.67151047844998,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004291002576506932,
      "loss": 4.9739,
      "mean_token_accuracy": 0.22460290789604187,
      "num_tokens": 61889784.0,
      "step": 27025
    },
    {
      "entropy": 5.7657787799835205,
      "epoch": 2.672004744958482,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042907454183516055,
      "loss": 5.0206,
      "mean_token_accuracy": 0.22319450974464417,
      "num_tokens": 61900295.0,
      "step": 27030
    },
    {
      "entropy": 5.688719415664673,
      "epoch": 2.672499011466983,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004290488222293702,
      "loss": 4.9609,
      "mean_token_accuracy": 0.2239790141582489,
      "num_tokens": 61911938.0,
      "step": 27035
    },
    {
      "entropy": 5.678220272064209,
      "epoch": 2.6729932779754844,
      "grad_norm": 1.25,
      "learning_rate": 0.00042902309883395497,
      "loss": 4.9982,
      "mean_token_accuracy": 0.221888168156147,
      "num_tokens": 61922609.0,
      "step": 27040
    },
    {
      "entropy": 5.640492963790893,
      "epoch": 2.6734875444839856,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004289973716495476,
      "loss": 4.9691,
      "mean_token_accuracy": 0.2174384132027626,
      "num_tokens": 61932744.0,
      "step": 27045
    },
    {
      "entropy": 5.66418776512146,
      "epoch": 2.673981810992487,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042897164067678094,
      "loss": 4.9754,
      "mean_token_accuracy": 0.2229044497013092,
      "num_tokens": 61944099.0,
      "step": 27050
    },
    {
      "entropy": 5.740688467025757,
      "epoch": 2.6744760775009886,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042894590591628813,
      "loss": 5.0711,
      "mean_token_accuracy": 0.21727485209703445,
      "num_tokens": 61955241.0,
      "step": 27055
    },
    {
      "entropy": 5.615402412414551,
      "epoch": 2.67497034400949,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042892016736870213,
      "loss": 4.9487,
      "mean_token_accuracy": 0.22464238107204437,
      "num_tokens": 61966578.0,
      "step": 27060
    },
    {
      "entropy": 5.6774450778961185,
      "epoch": 2.6754646105179916,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042889442503465616,
      "loss": 4.9419,
      "mean_token_accuracy": 0.23150235265493393,
      "num_tokens": 61977628.0,
      "step": 27065
    },
    {
      "entropy": 5.7972900390625,
      "epoch": 2.675958877026493,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042886867891478344,
      "loss": 5.1034,
      "mean_token_accuracy": 0.20824008584022521,
      "num_tokens": 61988904.0,
      "step": 27070
    },
    {
      "entropy": 5.674889039993286,
      "epoch": 2.676453143534994,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004288429290097173,
      "loss": 5.0568,
      "mean_token_accuracy": 0.21449539214372634,
      "num_tokens": 62000977.0,
      "step": 27075
    },
    {
      "entropy": 5.711830425262451,
      "epoch": 2.6769474100434953,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042881717532009125,
      "loss": 5.0263,
      "mean_token_accuracy": 0.2151870533823967,
      "num_tokens": 62011023.0,
      "step": 27080
    },
    {
      "entropy": 5.726917314529419,
      "epoch": 2.6774416765519966,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042879141784653887,
      "loss": 5.0237,
      "mean_token_accuracy": 0.21477681398391724,
      "num_tokens": 62022679.0,
      "step": 27085
    },
    {
      "entropy": 5.7148909091949465,
      "epoch": 2.6779359430604983,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042876565658969375,
      "loss": 4.9404,
      "mean_token_accuracy": 0.21711788773536683,
      "num_tokens": 62033922.0,
      "step": 27090
    },
    {
      "entropy": 5.72533106803894,
      "epoch": 2.6784302095689996,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042873989155018955,
      "loss": 5.0837,
      "mean_token_accuracy": 0.2145394802093506,
      "num_tokens": 62045511.0,
      "step": 27095
    },
    {
      "entropy": 5.591947984695435,
      "epoch": 2.678924476077501,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004287141227286602,
      "loss": 4.8805,
      "mean_token_accuracy": 0.22692678719758988,
      "num_tokens": 62055113.0,
      "step": 27100
    },
    {
      "entropy": 5.784708738327026,
      "epoch": 2.6794187425860025,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004286883501257396,
      "loss": 5.1242,
      "mean_token_accuracy": 0.2087639257311821,
      "num_tokens": 62067808.0,
      "step": 27105
    },
    {
      "entropy": 5.727280378341675,
      "epoch": 2.679913009094504,
      "grad_norm": 1.125,
      "learning_rate": 0.00042866257374206174,
      "loss": 4.9489,
      "mean_token_accuracy": 0.21731794625520706,
      "num_tokens": 62078266.0,
      "step": 27110
    },
    {
      "entropy": 5.682216453552246,
      "epoch": 2.680407275603005,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042863679357826076,
      "loss": 5.0477,
      "mean_token_accuracy": 0.21350272744894028,
      "num_tokens": 62091612.0,
      "step": 27115
    },
    {
      "entropy": 5.68894100189209,
      "epoch": 2.6809015421115063,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004286110096349708,
      "loss": 5.0197,
      "mean_token_accuracy": 0.21286636888980864,
      "num_tokens": 62104198.0,
      "step": 27120
    },
    {
      "entropy": 5.6859534740447994,
      "epoch": 2.681395808620008,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004285852219128261,
      "loss": 4.9473,
      "mean_token_accuracy": 0.21675826758146285,
      "num_tokens": 62115155.0,
      "step": 27125
    },
    {
      "entropy": 5.736457204818725,
      "epoch": 2.6818900751285093,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004285594304124612,
      "loss": 5.0592,
      "mean_token_accuracy": 0.210584457218647,
      "num_tokens": 62126555.0,
      "step": 27130
    },
    {
      "entropy": 5.715244388580322,
      "epoch": 2.6823843416370106,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042853363513451056,
      "loss": 5.0208,
      "mean_token_accuracy": 0.21914572566747664,
      "num_tokens": 62138976.0,
      "step": 27135
    },
    {
      "entropy": 5.735333967208862,
      "epoch": 2.6828786081455123,
      "grad_norm": 1.0,
      "learning_rate": 0.00042850783607960855,
      "loss": 5.0834,
      "mean_token_accuracy": 0.21475236117839813,
      "num_tokens": 62150453.0,
      "step": 27140
    },
    {
      "entropy": 5.653365278244019,
      "epoch": 2.6833728746540135,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042848203324839,
      "loss": 4.861,
      "mean_token_accuracy": 0.229047591984272,
      "num_tokens": 62162342.0,
      "step": 27145
    },
    {
      "entropy": 5.690463733673096,
      "epoch": 2.683867141162515,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004284562266414896,
      "loss": 5.0187,
      "mean_token_accuracy": 0.2196005180478096,
      "num_tokens": 62173850.0,
      "step": 27150
    },
    {
      "entropy": 5.621695899963379,
      "epoch": 2.684361407671016,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042843041625954223,
      "loss": 4.934,
      "mean_token_accuracy": 0.21622993797063828,
      "num_tokens": 62185410.0,
      "step": 27155
    },
    {
      "entropy": 5.68964376449585,
      "epoch": 2.6848556741795178,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004284046021031828,
      "loss": 4.9582,
      "mean_token_accuracy": 0.22273259609937668,
      "num_tokens": 62197620.0,
      "step": 27160
    },
    {
      "entropy": 5.671229362487793,
      "epoch": 2.685349940688019,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004283787841730464,
      "loss": 4.9842,
      "mean_token_accuracy": 0.2170150339603424,
      "num_tokens": 62209435.0,
      "step": 27165
    },
    {
      "entropy": 5.675285863876343,
      "epoch": 2.6858442071965203,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042835296246976806,
      "loss": 5.0029,
      "mean_token_accuracy": 0.21671825647354126,
      "num_tokens": 62219948.0,
      "step": 27170
    },
    {
      "entropy": 5.682850408554077,
      "epoch": 2.686338473705022,
      "grad_norm": 1.125,
      "learning_rate": 0.0004283271369939831,
      "loss": 5.0134,
      "mean_token_accuracy": 0.21759459525346755,
      "num_tokens": 62230299.0,
      "step": 27175
    },
    {
      "entropy": 5.704958391189575,
      "epoch": 2.6868327402135233,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042830130774632687,
      "loss": 4.9981,
      "mean_token_accuracy": 0.22276640087366104,
      "num_tokens": 62241952.0,
      "step": 27180
    },
    {
      "entropy": 5.741428518295288,
      "epoch": 2.6873270067220245,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004282754747274345,
      "loss": 5.0279,
      "mean_token_accuracy": 0.21613580882549285,
      "num_tokens": 62253200.0,
      "step": 27185
    },
    {
      "entropy": 5.729916143417358,
      "epoch": 2.6878212732305258,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042824963793794174,
      "loss": 4.998,
      "mean_token_accuracy": 0.22157507091760636,
      "num_tokens": 62264203.0,
      "step": 27190
    },
    {
      "entropy": 5.746968364715576,
      "epoch": 2.688315539739027,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042822379737848424,
      "loss": 4.9953,
      "mean_token_accuracy": 0.22050076276063918,
      "num_tokens": 62276296.0,
      "step": 27195
    },
    {
      "entropy": 5.719486045837402,
      "epoch": 2.6888098062475287,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004281979530496974,
      "loss": 5.0324,
      "mean_token_accuracy": 0.2144046738743782,
      "num_tokens": 62287827.0,
      "step": 27200
    },
    {
      "entropy": 5.803039026260376,
      "epoch": 2.68930407275603,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004281721049522172,
      "loss": 5.0617,
      "mean_token_accuracy": 0.21033957302570344,
      "num_tokens": 62299451.0,
      "step": 27205
    },
    {
      "entropy": 5.682394123077392,
      "epoch": 2.6897983392645317,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042814625308667945,
      "loss": 5.0075,
      "mean_token_accuracy": 0.21834331303834914,
      "num_tokens": 62311508.0,
      "step": 27210
    },
    {
      "entropy": 5.698611497879028,
      "epoch": 2.690292605773033,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042812039745372004,
      "loss": 4.972,
      "mean_token_accuracy": 0.2263050377368927,
      "num_tokens": 62323279.0,
      "step": 27215
    },
    {
      "entropy": 5.788854694366455,
      "epoch": 2.6907868722815342,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042809453805397525,
      "loss": 5.0956,
      "mean_token_accuracy": 0.21407433450222016,
      "num_tokens": 62335897.0,
      "step": 27220
    },
    {
      "entropy": 5.7665125846862795,
      "epoch": 2.6912811387900355,
      "grad_norm": 1.125,
      "learning_rate": 0.000428068674888081,
      "loss": 5.0471,
      "mean_token_accuracy": 0.2097582072019577,
      "num_tokens": 62347090.0,
      "step": 27225
    },
    {
      "entropy": 5.747049331665039,
      "epoch": 2.6917754052985368,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042804280795667363,
      "loss": 5.0117,
      "mean_token_accuracy": 0.21165463626384734,
      "num_tokens": 62358211.0,
      "step": 27230
    },
    {
      "entropy": 5.721274042129517,
      "epoch": 2.6922696718070385,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004280169372603894,
      "loss": 4.9942,
      "mean_token_accuracy": 0.22276009619235992,
      "num_tokens": 62368628.0,
      "step": 27235
    },
    {
      "entropy": 5.6463953971862795,
      "epoch": 2.6927639383155397,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004279910627998647,
      "loss": 4.9992,
      "mean_token_accuracy": 0.21832842826843263,
      "num_tokens": 62380633.0,
      "step": 27240
    },
    {
      "entropy": 5.764396619796753,
      "epoch": 2.693258204824041,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042796518457573625,
      "loss": 5.0752,
      "mean_token_accuracy": 0.21331270188093185,
      "num_tokens": 62392615.0,
      "step": 27245
    },
    {
      "entropy": 5.73046178817749,
      "epoch": 2.6937524713325427,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004279393025886405,
      "loss": 4.995,
      "mean_token_accuracy": 0.22365687787532806,
      "num_tokens": 62403087.0,
      "step": 27250
    },
    {
      "entropy": 5.765312576293946,
      "epoch": 2.694246737841044,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042791341683921417,
      "loss": 5.0216,
      "mean_token_accuracy": 0.21563545614480972,
      "num_tokens": 62415487.0,
      "step": 27255
    },
    {
      "entropy": 5.673682975769043,
      "epoch": 2.694741004349545,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004278875273280941,
      "loss": 4.9836,
      "mean_token_accuracy": 0.22069939225912094,
      "num_tokens": 62428123.0,
      "step": 27260
    },
    {
      "entropy": 5.695256948471069,
      "epoch": 2.6952352708580465,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004278616340559171,
      "loss": 5.0031,
      "mean_token_accuracy": 0.21745419055223464,
      "num_tokens": 62440250.0,
      "step": 27265
    },
    {
      "entropy": 5.6742844581604,
      "epoch": 2.695729537366548,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042783573702332014,
      "loss": 4.9823,
      "mean_token_accuracy": 0.2162162706255913,
      "num_tokens": 62451500.0,
      "step": 27270
    },
    {
      "entropy": 5.730587768554687,
      "epoch": 2.6962238038750495,
      "grad_norm": 1.28125,
      "learning_rate": 0.00042780983623094037,
      "loss": 5.0067,
      "mean_token_accuracy": 0.21840197890996932,
      "num_tokens": 62461839.0,
      "step": 27275
    },
    {
      "entropy": 5.715597820281983,
      "epoch": 2.6967180703835507,
      "grad_norm": 1.125,
      "learning_rate": 0.0004277839316794149,
      "loss": 4.9779,
      "mean_token_accuracy": 0.21947042793035507,
      "num_tokens": 62473650.0,
      "step": 27280
    },
    {
      "entropy": 5.718701601028442,
      "epoch": 2.6972123368920524,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042775802336938106,
      "loss": 5.028,
      "mean_token_accuracy": 0.2158135011792183,
      "num_tokens": 62484024.0,
      "step": 27285
    },
    {
      "entropy": 5.729912519454956,
      "epoch": 2.6977066034005537,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004277321113014762,
      "loss": 4.9996,
      "mean_token_accuracy": 0.21877060383558272,
      "num_tokens": 62495958.0,
      "step": 27290
    },
    {
      "entropy": 5.71972427368164,
      "epoch": 2.698200869909055,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042770619547633767,
      "loss": 5.1237,
      "mean_token_accuracy": 0.2141130343079567,
      "num_tokens": 62508474.0,
      "step": 27295
    },
    {
      "entropy": 5.6921507835388185,
      "epoch": 2.698695136417556,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042768027589460295,
      "loss": 5.0077,
      "mean_token_accuracy": 0.22028692364692687,
      "num_tokens": 62520281.0,
      "step": 27300
    },
    {
      "entropy": 5.766227388381958,
      "epoch": 2.6991894029260575,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004276543525569098,
      "loss": 5.0208,
      "mean_token_accuracy": 0.21635464131832122,
      "num_tokens": 62532242.0,
      "step": 27305
    },
    {
      "entropy": 5.725044107437133,
      "epoch": 2.699683669434559,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042762842546389593,
      "loss": 5.0193,
      "mean_token_accuracy": 0.22090674340724945,
      "num_tokens": 62544416.0,
      "step": 27310
    },
    {
      "entropy": 5.763671112060547,
      "epoch": 2.7001779359430604,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042760249461619914,
      "loss": 5.0479,
      "mean_token_accuracy": 0.2149918332695961,
      "num_tokens": 62555932.0,
      "step": 27315
    },
    {
      "entropy": 5.619484186172485,
      "epoch": 2.700672202451562,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042757656001445727,
      "loss": 4.9282,
      "mean_token_accuracy": 0.22598726600408553,
      "num_tokens": 62567322.0,
      "step": 27320
    },
    {
      "entropy": 5.665970754623413,
      "epoch": 2.7011664689600634,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004275506216593083,
      "loss": 4.9663,
      "mean_token_accuracy": 0.21986768543720245,
      "num_tokens": 62577491.0,
      "step": 27325
    },
    {
      "entropy": 5.695775556564331,
      "epoch": 2.7016607354685647,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004275246795513904,
      "loss": 5.024,
      "mean_token_accuracy": 0.21667778640985488,
      "num_tokens": 62589912.0,
      "step": 27330
    },
    {
      "entropy": 5.661094427108765,
      "epoch": 2.702155001977066,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004274987336913418,
      "loss": 5.0186,
      "mean_token_accuracy": 0.21415269523859023,
      "num_tokens": 62601571.0,
      "step": 27335
    },
    {
      "entropy": 5.726788759231567,
      "epoch": 2.702649268485567,
      "grad_norm": 1.125,
      "learning_rate": 0.0004274727840798007,
      "loss": 5.0127,
      "mean_token_accuracy": 0.2140343889594078,
      "num_tokens": 62615006.0,
      "step": 27340
    },
    {
      "entropy": 5.785875368118286,
      "epoch": 2.703143534994069,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042744683071740524,
      "loss": 5.0779,
      "mean_token_accuracy": 0.21749793440103532,
      "num_tokens": 62626678.0,
      "step": 27345
    },
    {
      "entropy": 5.63812108039856,
      "epoch": 2.70363780150257,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004274208736047944,
      "loss": 4.9538,
      "mean_token_accuracy": 0.21945013105869293,
      "num_tokens": 62637819.0,
      "step": 27350
    },
    {
      "entropy": 5.617285966873169,
      "epoch": 2.7041320680110714,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042739491274260615,
      "loss": 4.9047,
      "mean_token_accuracy": 0.23025340288877488,
      "num_tokens": 62649705.0,
      "step": 27355
    },
    {
      "entropy": 5.647848844528198,
      "epoch": 2.704626334519573,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042736894813147954,
      "loss": 4.9305,
      "mean_token_accuracy": 0.22029455453157426,
      "num_tokens": 62661695.0,
      "step": 27360
    },
    {
      "entropy": 5.832144117355346,
      "epoch": 2.7051206010280744,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042734297977205325,
      "loss": 5.1102,
      "mean_token_accuracy": 0.2035825729370117,
      "num_tokens": 62674102.0,
      "step": 27365
    },
    {
      "entropy": 5.764290285110474,
      "epoch": 2.7056148675365757,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042731700766496586,
      "loss": 5.0598,
      "mean_token_accuracy": 0.21395417898893357,
      "num_tokens": 62685562.0,
      "step": 27370
    },
    {
      "entropy": 5.7593817710876465,
      "epoch": 2.706109134045077,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004272910318108565,
      "loss": 5.0122,
      "mean_token_accuracy": 0.21911872625350953,
      "num_tokens": 62696653.0,
      "step": 27375
    },
    {
      "entropy": 5.7076108932495115,
      "epoch": 2.7066034005535786,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042726505221036415,
      "loss": 5.0007,
      "mean_token_accuracy": 0.21944233328104018,
      "num_tokens": 62708760.0,
      "step": 27380
    },
    {
      "entropy": 5.735338830947876,
      "epoch": 2.70709766706208,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004272390688641279,
      "loss": 5.0109,
      "mean_token_accuracy": 0.21540824323892593,
      "num_tokens": 62718548.0,
      "step": 27385
    },
    {
      "entropy": 5.667323970794678,
      "epoch": 2.707591933570581,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042721308177278696,
      "loss": 4.991,
      "mean_token_accuracy": 0.21974425315856932,
      "num_tokens": 62730792.0,
      "step": 27390
    },
    {
      "entropy": 5.796203708648681,
      "epoch": 2.708086200079083,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042718709093698054,
      "loss": 5.0484,
      "mean_token_accuracy": 0.21394769251346588,
      "num_tokens": 62741528.0,
      "step": 27395
    },
    {
      "entropy": 5.679714298248291,
      "epoch": 2.708580466587584,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042716109635734806,
      "loss": 4.9552,
      "mean_token_accuracy": 0.22015808075666427,
      "num_tokens": 62751650.0,
      "step": 27400
    },
    {
      "entropy": 5.699984455108643,
      "epoch": 2.7090747330960854,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042713509803452904,
      "loss": 5.0076,
      "mean_token_accuracy": 0.22100409269332885,
      "num_tokens": 62763248.0,
      "step": 27405
    },
    {
      "entropy": 5.703722667694092,
      "epoch": 2.7095689996045866,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042710909596916307,
      "loss": 5.0159,
      "mean_token_accuracy": 0.2154308408498764,
      "num_tokens": 62775696.0,
      "step": 27410
    },
    {
      "entropy": 5.673343896865845,
      "epoch": 2.7100632661130883,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042708309016188965,
      "loss": 4.9252,
      "mean_token_accuracy": 0.22757147997617722,
      "num_tokens": 62786564.0,
      "step": 27415
    },
    {
      "entropy": 5.682564640045166,
      "epoch": 2.7105575326215896,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004270570806133486,
      "loss": 5.0075,
      "mean_token_accuracy": 0.21976580917835237,
      "num_tokens": 62799235.0,
      "step": 27420
    },
    {
      "entropy": 5.650778722763062,
      "epoch": 2.711051799130091,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004270310673241799,
      "loss": 4.9843,
      "mean_token_accuracy": 0.2160615235567093,
      "num_tokens": 62809605.0,
      "step": 27425
    },
    {
      "entropy": 5.74376654624939,
      "epoch": 2.7115460656385926,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004270050502950233,
      "loss": 4.9801,
      "mean_token_accuracy": 0.2216406598687172,
      "num_tokens": 62820900.0,
      "step": 27430
    },
    {
      "entropy": 5.6701117038726805,
      "epoch": 2.712040332147094,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004269790295265189,
      "loss": 4.9245,
      "mean_token_accuracy": 0.22991884350776673,
      "num_tokens": 62831761.0,
      "step": 27435
    },
    {
      "entropy": 5.787686109542847,
      "epoch": 2.712534598655595,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042695300501930674,
      "loss": 5.0762,
      "mean_token_accuracy": 0.21061451137065887,
      "num_tokens": 62842512.0,
      "step": 27440
    },
    {
      "entropy": 5.710083389282227,
      "epoch": 2.7130288651640964,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004269269767740271,
      "loss": 5.027,
      "mean_token_accuracy": 0.2196846142411232,
      "num_tokens": 62853495.0,
      "step": 27445
    },
    {
      "entropy": 5.605766439437867,
      "epoch": 2.7135231316725976,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042690094479132023,
      "loss": 4.9321,
      "mean_token_accuracy": 0.23107286542654037,
      "num_tokens": 62864524.0,
      "step": 27450
    },
    {
      "entropy": 5.760519409179688,
      "epoch": 2.7140173981810993,
      "grad_norm": 1.25,
      "learning_rate": 0.00042687490907182665,
      "loss": 5.1222,
      "mean_token_accuracy": 0.2090625986456871,
      "num_tokens": 62874522.0,
      "step": 27455
    },
    {
      "entropy": 5.7070318222045895,
      "epoch": 2.7145116646896006,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042684886961618664,
      "loss": 4.9887,
      "mean_token_accuracy": 0.2167014554142952,
      "num_tokens": 62885477.0,
      "step": 27460
    },
    {
      "entropy": 5.703550863265991,
      "epoch": 2.715005931198102,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042682282642504096,
      "loss": 5.026,
      "mean_token_accuracy": 0.2221037268638611,
      "num_tokens": 62896140.0,
      "step": 27465
    },
    {
      "entropy": 5.773190498352051,
      "epoch": 2.7155001977066036,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042679677949903017,
      "loss": 5.0945,
      "mean_token_accuracy": 0.20963554531335832,
      "num_tokens": 62907707.0,
      "step": 27470
    },
    {
      "entropy": 5.77922568321228,
      "epoch": 2.715994464215105,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042677072883879496,
      "loss": 5.1012,
      "mean_token_accuracy": 0.21329701840877532,
      "num_tokens": 62919209.0,
      "step": 27475
    },
    {
      "entropy": 5.622571516036987,
      "epoch": 2.716488730723606,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042674467444497644,
      "loss": 4.8583,
      "mean_token_accuracy": 0.24185421615839003,
      "num_tokens": 62931971.0,
      "step": 27480
    },
    {
      "entropy": 5.704189586639404,
      "epoch": 2.7169829972321073,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004267186163182153,
      "loss": 5.0457,
      "mean_token_accuracy": 0.21980664283037185,
      "num_tokens": 62944246.0,
      "step": 27485
    },
    {
      "entropy": 5.746811866760254,
      "epoch": 2.717477263740609,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004266925544591527,
      "loss": 5.0946,
      "mean_token_accuracy": 0.21023415327072142,
      "num_tokens": 62956973.0,
      "step": 27490
    },
    {
      "entropy": 5.788717269897461,
      "epoch": 2.7179715302491103,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004266664888684297,
      "loss": 5.0947,
      "mean_token_accuracy": 0.20888681411743165,
      "num_tokens": 62968590.0,
      "step": 27495
    },
    {
      "entropy": 5.721470165252685,
      "epoch": 2.7184657967576116,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042664041954668753,
      "loss": 5.0168,
      "mean_token_accuracy": 0.22157567143440246,
      "num_tokens": 62980382.0,
      "step": 27500
    },
    {
      "entropy": 5.706231212615966,
      "epoch": 2.7189600632661133,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004266143464945675,
      "loss": 4.9909,
      "mean_token_accuracy": 0.22012377232313157,
      "num_tokens": 62990432.0,
      "step": 27505
    },
    {
      "entropy": 5.711324644088745,
      "epoch": 2.7194543297746145,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042658826971271103,
      "loss": 5.0253,
      "mean_token_accuracy": 0.2160892218351364,
      "num_tokens": 63003341.0,
      "step": 27510
    },
    {
      "entropy": 5.75231556892395,
      "epoch": 2.719948596283116,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042656218920175963,
      "loss": 5.041,
      "mean_token_accuracy": 0.21142753213644028,
      "num_tokens": 63015527.0,
      "step": 27515
    },
    {
      "entropy": 5.764517736434937,
      "epoch": 2.720442862791617,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000426536104962355,
      "loss": 5.0161,
      "mean_token_accuracy": 0.21887631416320802,
      "num_tokens": 63026995.0,
      "step": 27520
    },
    {
      "entropy": 5.708220815658569,
      "epoch": 2.7209371293001188,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004265100169951385,
      "loss": 4.9369,
      "mean_token_accuracy": 0.22315262705087663,
      "num_tokens": 63037156.0,
      "step": 27525
    },
    {
      "entropy": 5.742822456359863,
      "epoch": 2.72143139580862,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004264839253007521,
      "loss": 5.0324,
      "mean_token_accuracy": 0.22105981260538102,
      "num_tokens": 63047513.0,
      "step": 27530
    },
    {
      "entropy": 5.725811862945557,
      "epoch": 2.7219256623171213,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042645782987983763,
      "loss": 4.9568,
      "mean_token_accuracy": 0.21601782292127608,
      "num_tokens": 63059968.0,
      "step": 27535
    },
    {
      "entropy": 5.701229429244995,
      "epoch": 2.722419928825623,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042643173073303713,
      "loss": 4.9914,
      "mean_token_accuracy": 0.21586666405200958,
      "num_tokens": 63071514.0,
      "step": 27540
    },
    {
      "entropy": 5.7435227394104,
      "epoch": 2.7229141953341243,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042640562786099245,
      "loss": 5.0207,
      "mean_token_accuracy": 0.22378179281949998,
      "num_tokens": 63082810.0,
      "step": 27545
    },
    {
      "entropy": 5.732158660888672,
      "epoch": 2.7234084618426255,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042637952126434593,
      "loss": 5.0142,
      "mean_token_accuracy": 0.21673722416162491,
      "num_tokens": 63093324.0,
      "step": 27550
    },
    {
      "entropy": 5.679854679107666,
      "epoch": 2.723902728351127,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004263534109437397,
      "loss": 5.0279,
      "mean_token_accuracy": 0.21593224555253981,
      "num_tokens": 63104834.0,
      "step": 27555
    },
    {
      "entropy": 5.676044273376465,
      "epoch": 2.724396994859628,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042632729689981604,
      "loss": 5.0315,
      "mean_token_accuracy": 0.2163797065615654,
      "num_tokens": 63116625.0,
      "step": 27560
    },
    {
      "entropy": 5.755739688873291,
      "epoch": 2.7248912613681298,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042630117913321734,
      "loss": 5.0223,
      "mean_token_accuracy": 0.20924390703439713,
      "num_tokens": 63127616.0,
      "step": 27565
    },
    {
      "entropy": 5.667864418029785,
      "epoch": 2.725385527876631,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042627505764458626,
      "loss": 4.9971,
      "mean_token_accuracy": 0.22559669315814973,
      "num_tokens": 63140546.0,
      "step": 27570
    },
    {
      "entropy": 5.627913331985473,
      "epoch": 2.7258797943851327,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004262489324345652,
      "loss": 4.9217,
      "mean_token_accuracy": 0.22269585877656936,
      "num_tokens": 63151693.0,
      "step": 27575
    },
    {
      "entropy": 5.691790771484375,
      "epoch": 2.726374060893634,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004262228035037971,
      "loss": 4.9542,
      "mean_token_accuracy": 0.2124678075313568,
      "num_tokens": 63162569.0,
      "step": 27580
    },
    {
      "entropy": 5.58673620223999,
      "epoch": 2.7268683274021353,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004261966708529244,
      "loss": 4.8702,
      "mean_token_accuracy": 0.23437179028987884,
      "num_tokens": 63174073.0,
      "step": 27585
    },
    {
      "entropy": 5.732895803451538,
      "epoch": 2.7273625939106365,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042617053448259026,
      "loss": 4.9972,
      "mean_token_accuracy": 0.2205752432346344,
      "num_tokens": 63184752.0,
      "step": 27590
    },
    {
      "entropy": 5.714764642715454,
      "epoch": 2.7278568604191378,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004261443943934375,
      "loss": 4.9309,
      "mean_token_accuracy": 0.22033771723508835,
      "num_tokens": 63197682.0,
      "step": 27595
    },
    {
      "entropy": 5.775260734558105,
      "epoch": 2.7283511269276395,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042611825058610917,
      "loss": 5.0766,
      "mean_token_accuracy": 0.20358372628688812,
      "num_tokens": 63208868.0,
      "step": 27600
    },
    {
      "entropy": 5.752489328384399,
      "epoch": 2.7288453934361407,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004260921030612483,
      "loss": 5.1009,
      "mean_token_accuracy": 0.21696856319904329,
      "num_tokens": 63220418.0,
      "step": 27605
    },
    {
      "entropy": 5.716807413101196,
      "epoch": 2.729339659944642,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004260659518194985,
      "loss": 4.9885,
      "mean_token_accuracy": 0.21934777945280076,
      "num_tokens": 63231431.0,
      "step": 27610
    },
    {
      "entropy": 5.6594061851501465,
      "epoch": 2.7298339264531437,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004260397968615027,
      "loss": 5.0042,
      "mean_token_accuracy": 0.20967702865600585,
      "num_tokens": 63243994.0,
      "step": 27615
    },
    {
      "entropy": 5.64479432106018,
      "epoch": 2.730328192961645,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004260136381879045,
      "loss": 4.9387,
      "mean_token_accuracy": 0.22758468836545945,
      "num_tokens": 63255414.0,
      "step": 27620
    },
    {
      "entropy": 5.725132131576538,
      "epoch": 2.7308224594701462,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042598747579934736,
      "loss": 5.0932,
      "mean_token_accuracy": 0.21382007002830505,
      "num_tokens": 63266289.0,
      "step": 27625
    },
    {
      "entropy": 5.726867628097534,
      "epoch": 2.7313167259786475,
      "grad_norm": 1.125,
      "learning_rate": 0.00042596130969647494,
      "loss": 4.9151,
      "mean_token_accuracy": 0.22526416033506394,
      "num_tokens": 63276497.0,
      "step": 27630
    },
    {
      "entropy": 5.711026906967163,
      "epoch": 2.731810992487149,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004259351398799308,
      "loss": 4.9706,
      "mean_token_accuracy": 0.22336022257804872,
      "num_tokens": 63286892.0,
      "step": 27635
    },
    {
      "entropy": 5.713762664794922,
      "epoch": 2.7323052589956505,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042590896635035885,
      "loss": 5.0342,
      "mean_token_accuracy": 0.21670519560575485,
      "num_tokens": 63299978.0,
      "step": 27640
    },
    {
      "entropy": 5.7279047012329105,
      "epoch": 2.7327995255041517,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042588278910840296,
      "loss": 5.0202,
      "mean_token_accuracy": 0.21966993063688278,
      "num_tokens": 63311916.0,
      "step": 27645
    },
    {
      "entropy": 5.705901765823365,
      "epoch": 2.7332937920126534,
      "grad_norm": 1.234375,
      "learning_rate": 0.000425856608154707,
      "loss": 4.9734,
      "mean_token_accuracy": 0.21916033774614335,
      "num_tokens": 63322994.0,
      "step": 27650
    },
    {
      "entropy": 5.760623168945313,
      "epoch": 2.7337880585211547,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004258304234899151,
      "loss": 5.0448,
      "mean_token_accuracy": 0.21260327398777007,
      "num_tokens": 63334708.0,
      "step": 27655
    },
    {
      "entropy": 5.688549900054932,
      "epoch": 2.734282325029656,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004258042351146713,
      "loss": 4.9476,
      "mean_token_accuracy": 0.2194727376103401,
      "num_tokens": 63346823.0,
      "step": 27660
    },
    {
      "entropy": 5.734207534790039,
      "epoch": 2.734776591538157,
      "grad_norm": 1.25,
      "learning_rate": 0.00042577804302961994,
      "loss": 5.0023,
      "mean_token_accuracy": 0.21713496893644332,
      "num_tokens": 63357677.0,
      "step": 27665
    },
    {
      "entropy": 5.625634479522705,
      "epoch": 2.7352708580466585,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042575184723540546,
      "loss": 4.8937,
      "mean_token_accuracy": 0.22304207533597947,
      "num_tokens": 63369685.0,
      "step": 27670
    },
    {
      "entropy": 5.75112099647522,
      "epoch": 2.73576512455516,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042572564773267203,
      "loss": 5.0485,
      "mean_token_accuracy": 0.2096169710159302,
      "num_tokens": 63380340.0,
      "step": 27675
    },
    {
      "entropy": 5.676293516159058,
      "epoch": 2.7362593910636615,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004256994445220643,
      "loss": 5.019,
      "mean_token_accuracy": 0.22390236854553222,
      "num_tokens": 63390868.0,
      "step": 27680
    },
    {
      "entropy": 5.690142059326172,
      "epoch": 2.736753657572163,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004256732376042269,
      "loss": 4.9667,
      "mean_token_accuracy": 0.21894970983266832,
      "num_tokens": 63401669.0,
      "step": 27685
    },
    {
      "entropy": 5.712251663208008,
      "epoch": 2.7372479240806644,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004256470269798044,
      "loss": 4.9844,
      "mean_token_accuracy": 0.22077011168003083,
      "num_tokens": 63412593.0,
      "step": 27690
    },
    {
      "entropy": 5.732409811019897,
      "epoch": 2.7377421905891657,
      "grad_norm": 1.25,
      "learning_rate": 0.00042562081264944175,
      "loss": 5.0261,
      "mean_token_accuracy": 0.22626218795776368,
      "num_tokens": 63423952.0,
      "step": 27695
    },
    {
      "entropy": 5.708841228485108,
      "epoch": 2.738236457097667,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042559459461378367,
      "loss": 5.0361,
      "mean_token_accuracy": 0.21374601870775223,
      "num_tokens": 63435152.0,
      "step": 27700
    },
    {
      "entropy": 5.684716606140137,
      "epoch": 2.738730723606168,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042556837287347524,
      "loss": 4.9,
      "mean_token_accuracy": 0.22808788418769838,
      "num_tokens": 63445977.0,
      "step": 27705
    },
    {
      "entropy": 5.747796487808228,
      "epoch": 2.73922499011467,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042554214742916154,
      "loss": 5.0846,
      "mean_token_accuracy": 0.2097993165254593,
      "num_tokens": 63456790.0,
      "step": 27710
    },
    {
      "entropy": 5.659936618804932,
      "epoch": 2.739719256623171,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004255159182814876,
      "loss": 4.9858,
      "mean_token_accuracy": 0.22480814158916473,
      "num_tokens": 63467775.0,
      "step": 27715
    },
    {
      "entropy": 5.789885997772217,
      "epoch": 2.7402135231316724,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004254896854310987,
      "loss": 5.0897,
      "mean_token_accuracy": 0.21030840873718262,
      "num_tokens": 63479445.0,
      "step": 27720
    },
    {
      "entropy": 5.7877874851226805,
      "epoch": 2.740707789640174,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004254634488786402,
      "loss": 5.0954,
      "mean_token_accuracy": 0.2054578110575676,
      "num_tokens": 63489938.0,
      "step": 27725
    },
    {
      "entropy": 5.7030946731567385,
      "epoch": 2.7412020561486754,
      "grad_norm": 1.25,
      "learning_rate": 0.00042543720862475747,
      "loss": 4.9892,
      "mean_token_accuracy": 0.22576913088560105,
      "num_tokens": 63501509.0,
      "step": 27730
    },
    {
      "entropy": 5.674745893478393,
      "epoch": 2.7416963226571767,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042541096467009617,
      "loss": 4.9697,
      "mean_token_accuracy": 0.2223714455962181,
      "num_tokens": 63512222.0,
      "step": 27735
    },
    {
      "entropy": 5.769401264190674,
      "epoch": 2.742190589165678,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042538471701530174,
      "loss": 5.1019,
      "mean_token_accuracy": 0.21478029191493989,
      "num_tokens": 63524841.0,
      "step": 27740
    },
    {
      "entropy": 5.724522686004638,
      "epoch": 2.7426848556741796,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042535846566102,
      "loss": 4.9753,
      "mean_token_accuracy": 0.2166510969400406,
      "num_tokens": 63536167.0,
      "step": 27745
    },
    {
      "entropy": 5.718362951278687,
      "epoch": 2.743179122182681,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004253322106078966,
      "loss": 4.9903,
      "mean_token_accuracy": 0.22181790620088576,
      "num_tokens": 63547774.0,
      "step": 27750
    },
    {
      "entropy": 5.748501920700074,
      "epoch": 2.743673388691182,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042530595185657756,
      "loss": 5.1123,
      "mean_token_accuracy": 0.21061017364263535,
      "num_tokens": 63560712.0,
      "step": 27755
    },
    {
      "entropy": 5.70149416923523,
      "epoch": 2.744167655199684,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042527968940770874,
      "loss": 5.0517,
      "mean_token_accuracy": 0.21655175983905792,
      "num_tokens": 63572587.0,
      "step": 27760
    },
    {
      "entropy": 5.733349609375,
      "epoch": 2.744661921708185,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042525342326193624,
      "loss": 5.0736,
      "mean_token_accuracy": 0.21250307261943818,
      "num_tokens": 63584694.0,
      "step": 27765
    },
    {
      "entropy": 5.7263007164001465,
      "epoch": 2.7451561882166864,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00042522715341990625,
      "loss": 5.019,
      "mean_token_accuracy": 0.21938451379537582,
      "num_tokens": 63597894.0,
      "step": 27770
    },
    {
      "entropy": 5.748881387710571,
      "epoch": 2.7456504547251877,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000425200879882265,
      "loss": 5.0575,
      "mean_token_accuracy": 0.21837457716464997,
      "num_tokens": 63608873.0,
      "step": 27775
    },
    {
      "entropy": 5.757216548919677,
      "epoch": 2.7461447212336894,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004251746026496588,
      "loss": 5.0588,
      "mean_token_accuracy": 0.2139126256108284,
      "num_tokens": 63621013.0,
      "step": 27780
    },
    {
      "entropy": 5.7875627994537355,
      "epoch": 2.7466389877421906,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042514832172273404,
      "loss": 5.1158,
      "mean_token_accuracy": 0.2072718247771263,
      "num_tokens": 63633438.0,
      "step": 27785
    },
    {
      "entropy": 5.7316060066223145,
      "epoch": 2.747133254250692,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042512203710213733,
      "loss": 4.9515,
      "mean_token_accuracy": 0.21874932050704957,
      "num_tokens": 63645032.0,
      "step": 27790
    },
    {
      "entropy": 5.6924913883209225,
      "epoch": 2.7476275207591936,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004250957487885151,
      "loss": 4.9616,
      "mean_token_accuracy": 0.22638950645923614,
      "num_tokens": 63657747.0,
      "step": 27795
    },
    {
      "entropy": 5.674510860443116,
      "epoch": 2.748121787267695,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004250694567825143,
      "loss": 4.9478,
      "mean_token_accuracy": 0.22383185625076293,
      "num_tokens": 63669472.0,
      "step": 27800
    },
    {
      "entropy": 5.67832670211792,
      "epoch": 2.748616053776196,
      "grad_norm": 1.125,
      "learning_rate": 0.00042504316108478153,
      "loss": 4.994,
      "mean_token_accuracy": 0.22650742083787917,
      "num_tokens": 63681498.0,
      "step": 27805
    },
    {
      "entropy": 5.618740463256836,
      "epoch": 2.7491103202846974,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042501686169596367,
      "loss": 4.9103,
      "mean_token_accuracy": 0.23110314160585405,
      "num_tokens": 63692628.0,
      "step": 27810
    },
    {
      "entropy": 5.712465238571167,
      "epoch": 2.7496045867931986,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042499055861670783,
      "loss": 4.9829,
      "mean_token_accuracy": 0.21796264201402665,
      "num_tokens": 63704732.0,
      "step": 27815
    },
    {
      "entropy": 5.746215915679931,
      "epoch": 2.7500988533017003,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004249642518476609,
      "loss": 5.0526,
      "mean_token_accuracy": 0.2119926020503044,
      "num_tokens": 63716346.0,
      "step": 27820
    },
    {
      "entropy": 5.724862098693848,
      "epoch": 2.7505931198102016,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042493794138947013,
      "loss": 5.041,
      "mean_token_accuracy": 0.21886069029569627,
      "num_tokens": 63728320.0,
      "step": 27825
    },
    {
      "entropy": 5.678915309906006,
      "epoch": 2.7510873863187033,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042491162724278276,
      "loss": 5.0623,
      "mean_token_accuracy": 0.21261305660009383,
      "num_tokens": 63740211.0,
      "step": 27830
    },
    {
      "entropy": 5.663030052185059,
      "epoch": 2.7515816528272046,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042488530940824616,
      "loss": 4.9094,
      "mean_token_accuracy": 0.23074459731578828,
      "num_tokens": 63751706.0,
      "step": 27835
    },
    {
      "entropy": 5.727924108505249,
      "epoch": 2.752075919335706,
      "grad_norm": 1.125,
      "learning_rate": 0.0004248589878865076,
      "loss": 5.0302,
      "mean_token_accuracy": 0.21577073335647584,
      "num_tokens": 63764045.0,
      "step": 27840
    },
    {
      "entropy": 5.685721063613892,
      "epoch": 2.752570185844207,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004248326626782147,
      "loss": 4.9309,
      "mean_token_accuracy": 0.22513374388217927,
      "num_tokens": 63774148.0,
      "step": 27845
    },
    {
      "entropy": 5.719576454162597,
      "epoch": 2.7530644523527084,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042480633378401503,
      "loss": 4.9662,
      "mean_token_accuracy": 0.22026122212409974,
      "num_tokens": 63785202.0,
      "step": 27850
    },
    {
      "entropy": 5.718375110626221,
      "epoch": 2.75355871886121,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004247800012045564,
      "loss": 4.9834,
      "mean_token_accuracy": 0.21291326135396957,
      "num_tokens": 63796990.0,
      "step": 27855
    },
    {
      "entropy": 5.6479393482208256,
      "epoch": 2.7540529853697113,
      "grad_norm": 1.25,
      "learning_rate": 0.00042475366494048645,
      "loss": 4.8845,
      "mean_token_accuracy": 0.23210216015577317,
      "num_tokens": 63807201.0,
      "step": 27860
    },
    {
      "entropy": 5.681387138366699,
      "epoch": 2.7545472518782126,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042472732499245316,
      "loss": 4.9824,
      "mean_token_accuracy": 0.2256856828927994,
      "num_tokens": 63818805.0,
      "step": 27865
    },
    {
      "entropy": 5.6817117691040036,
      "epoch": 2.7550415183867143,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042470098136110444,
      "loss": 4.9553,
      "mean_token_accuracy": 0.2223185494542122,
      "num_tokens": 63829444.0,
      "step": 27870
    },
    {
      "entropy": 5.642939281463623,
      "epoch": 2.7555357848952156,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042467463404708826,
      "loss": 4.9741,
      "mean_token_accuracy": 0.22321929037570953,
      "num_tokens": 63840868.0,
      "step": 27875
    },
    {
      "entropy": 5.683885383605957,
      "epoch": 2.756030051403717,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042464828305105283,
      "loss": 5.014,
      "mean_token_accuracy": 0.21687053591012956,
      "num_tokens": 63851992.0,
      "step": 27880
    },
    {
      "entropy": 5.707004547119141,
      "epoch": 2.756524317912218,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042462192837364654,
      "loss": 5.0033,
      "mean_token_accuracy": 0.2164152294397354,
      "num_tokens": 63863362.0,
      "step": 27885
    },
    {
      "entropy": 5.747110176086426,
      "epoch": 2.75701858442072,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042459557001551764,
      "loss": 5.0662,
      "mean_token_accuracy": 0.21289311349391937,
      "num_tokens": 63875712.0,
      "step": 27890
    },
    {
      "entropy": 5.786071157455444,
      "epoch": 2.757512850929221,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042456920797731443,
      "loss": 4.9801,
      "mean_token_accuracy": 0.22067447006702423,
      "num_tokens": 63887253.0,
      "step": 27895
    },
    {
      "entropy": 5.757207155227661,
      "epoch": 2.7580071174377223,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042454284225968553,
      "loss": 5.0872,
      "mean_token_accuracy": 0.21548418402671815,
      "num_tokens": 63898328.0,
      "step": 27900
    },
    {
      "entropy": 5.642959880828857,
      "epoch": 2.758501383946224,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004245164728632795,
      "loss": 4.8913,
      "mean_token_accuracy": 0.2313155472278595,
      "num_tokens": 63908080.0,
      "step": 27905
    },
    {
      "entropy": 5.704622268676758,
      "epoch": 2.7589956504547253,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000424490099788745,
      "loss": 5.0515,
      "mean_token_accuracy": 0.21184180229902266,
      "num_tokens": 63918860.0,
      "step": 27910
    },
    {
      "entropy": 5.756003046035767,
      "epoch": 2.7594899169632265,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004244637230367309,
      "loss": 4.9634,
      "mean_token_accuracy": 0.2202967509627342,
      "num_tokens": 63929601.0,
      "step": 27915
    },
    {
      "entropy": 5.691011381149292,
      "epoch": 2.759984183471728,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000424437342607886,
      "loss": 5.0293,
      "mean_token_accuracy": 0.22072141468524933,
      "num_tokens": 63941170.0,
      "step": 27920
    },
    {
      "entropy": 5.666725301742554,
      "epoch": 2.760478449980229,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004244109585028593,
      "loss": 4.9531,
      "mean_token_accuracy": 0.2271920010447502,
      "num_tokens": 63951907.0,
      "step": 27925
    },
    {
      "entropy": 5.740157270431519,
      "epoch": 2.7609727164887308,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042438457072229986,
      "loss": 5.0047,
      "mean_token_accuracy": 0.21590171456336976,
      "num_tokens": 63963140.0,
      "step": 27930
    },
    {
      "entropy": 5.723110103607178,
      "epoch": 2.761466982997232,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042435817926685684,
      "loss": 4.9989,
      "mean_token_accuracy": 0.21543813049793242,
      "num_tokens": 63974445.0,
      "step": 27935
    },
    {
      "entropy": 5.678072547912597,
      "epoch": 2.7619612495057337,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004243317841371793,
      "loss": 4.9575,
      "mean_token_accuracy": 0.22454602420330047,
      "num_tokens": 63985463.0,
      "step": 27940
    },
    {
      "entropy": 5.7105320453643795,
      "epoch": 2.762455516014235,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042430538533391683,
      "loss": 5.0283,
      "mean_token_accuracy": 0.2175595283508301,
      "num_tokens": 63997120.0,
      "step": 27945
    },
    {
      "entropy": 5.686862421035767,
      "epoch": 2.7629497825227363,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042427898285771865,
      "loss": 5.0107,
      "mean_token_accuracy": 0.2157677009701729,
      "num_tokens": 64009195.0,
      "step": 27950
    },
    {
      "entropy": 5.735896158218384,
      "epoch": 2.7634440490312375,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004242525767092344,
      "loss": 5.0222,
      "mean_token_accuracy": 0.22532825022935868,
      "num_tokens": 64020915.0,
      "step": 27955
    },
    {
      "entropy": 5.68587851524353,
      "epoch": 2.763938315539739,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004242261668891135,
      "loss": 4.9561,
      "mean_token_accuracy": 0.22005146294832229,
      "num_tokens": 64031187.0,
      "step": 27960
    },
    {
      "entropy": 5.731714105606079,
      "epoch": 2.7644325820482405,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004241997533980059,
      "loss": 5.0154,
      "mean_token_accuracy": 0.21325055062770842,
      "num_tokens": 64043045.0,
      "step": 27965
    },
    {
      "entropy": 5.842581462860108,
      "epoch": 2.7649268485567418,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042417333623656113,
      "loss": 5.1242,
      "mean_token_accuracy": 0.20680923759937286,
      "num_tokens": 64054030.0,
      "step": 27970
    },
    {
      "entropy": 5.732256889343262,
      "epoch": 2.765421115065243,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042414691540542917,
      "loss": 5.0079,
      "mean_token_accuracy": 0.2215494245290756,
      "num_tokens": 64064179.0,
      "step": 27975
    },
    {
      "entropy": 5.761953067779541,
      "epoch": 2.7659153815737447,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042412049090525996,
      "loss": 5.084,
      "mean_token_accuracy": 0.20787105560302735,
      "num_tokens": 64075812.0,
      "step": 27980
    },
    {
      "entropy": 5.6603254795074465,
      "epoch": 2.766409648082246,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004240940627367035,
      "loss": 4.9696,
      "mean_token_accuracy": 0.22319453358650207,
      "num_tokens": 64086848.0,
      "step": 27985
    },
    {
      "entropy": 5.7114259719848635,
      "epoch": 2.7669039145907472,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042406763090040996,
      "loss": 5.0122,
      "mean_token_accuracy": 0.21582835614681245,
      "num_tokens": 64098879.0,
      "step": 27990
    },
    {
      "entropy": 5.712428855895996,
      "epoch": 2.7673981810992485,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042404119539702956,
      "loss": 5.0428,
      "mean_token_accuracy": 0.21064504683017732,
      "num_tokens": 64109946.0,
      "step": 27995
    },
    {
      "entropy": 5.755137920379639,
      "epoch": 2.76789244760775,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042401475622721264,
      "loss": 5.0895,
      "mean_token_accuracy": 0.2120663657784462,
      "num_tokens": 64122219.0,
      "step": 28000
    },
    {
      "entropy": 5.715014171600342,
      "epoch": 2.7683867141162515,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004239883133916097,
      "loss": 4.9842,
      "mean_token_accuracy": 0.22325509488582612,
      "num_tokens": 64132950.0,
      "step": 28005
    },
    {
      "entropy": 5.708972215652466,
      "epoch": 2.7688809806247527,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004239618668908711,
      "loss": 5.0632,
      "mean_token_accuracy": 0.21418076604604722,
      "num_tokens": 64145060.0,
      "step": 28010
    },
    {
      "entropy": 5.755795526504516,
      "epoch": 2.7693752471332544,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004239354167256474,
      "loss": 5.0755,
      "mean_token_accuracy": 0.21114394813776016,
      "num_tokens": 64157294.0,
      "step": 28015
    },
    {
      "entropy": 5.756480884552002,
      "epoch": 2.7698695136417557,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042390896289658936,
      "loss": 5.0399,
      "mean_token_accuracy": 0.21127624064683914,
      "num_tokens": 64169001.0,
      "step": 28020
    },
    {
      "entropy": 5.67667555809021,
      "epoch": 2.770363780150257,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004238825054043477,
      "loss": 4.9483,
      "mean_token_accuracy": 0.22188242077827453,
      "num_tokens": 64180525.0,
      "step": 28025
    },
    {
      "entropy": 5.665124416351318,
      "epoch": 2.7708580466587582,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042385604424957337,
      "loss": 4.9586,
      "mean_token_accuracy": 0.21906013190746307,
      "num_tokens": 64192398.0,
      "step": 28030
    },
    {
      "entropy": 5.681071615219116,
      "epoch": 2.77135231316726,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004238295794329172,
      "loss": 5.0252,
      "mean_token_accuracy": 0.21208575516939163,
      "num_tokens": 64203485.0,
      "step": 28035
    },
    {
      "entropy": 5.683232355117798,
      "epoch": 2.771846579675761,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042380311095503037,
      "loss": 4.9904,
      "mean_token_accuracy": 0.2218636080622673,
      "num_tokens": 64214226.0,
      "step": 28040
    },
    {
      "entropy": 5.682394742965698,
      "epoch": 2.7723408461842625,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042377663881656394,
      "loss": 4.9494,
      "mean_token_accuracy": 0.2221979171037674,
      "num_tokens": 64225873.0,
      "step": 28045
    },
    {
      "entropy": 5.843520593643189,
      "epoch": 2.772835112692764,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042375016301816903,
      "loss": 5.1755,
      "mean_token_accuracy": 0.20077353864908218,
      "num_tokens": 64238023.0,
      "step": 28050
    },
    {
      "entropy": 5.76641960144043,
      "epoch": 2.7733293792012654,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004237236835604971,
      "loss": 5.0489,
      "mean_token_accuracy": 0.20876119732856752,
      "num_tokens": 64248533.0,
      "step": 28055
    },
    {
      "entropy": 5.739803314208984,
      "epoch": 2.7738236457097667,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004236972004441994,
      "loss": 5.0834,
      "mean_token_accuracy": 0.21748532354831696,
      "num_tokens": 64260539.0,
      "step": 28060
    },
    {
      "entropy": 5.745120477676392,
      "epoch": 2.774317912218268,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042367071366992753,
      "loss": 5.0059,
      "mean_token_accuracy": 0.21207655370235443,
      "num_tokens": 64271280.0,
      "step": 28065
    },
    {
      "entropy": 5.742461681365967,
      "epoch": 2.774812178726769,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042364422323833313,
      "loss": 5.0601,
      "mean_token_accuracy": 0.2093740686774254,
      "num_tokens": 64282728.0,
      "step": 28070
    },
    {
      "entropy": 5.72702169418335,
      "epoch": 2.775306445235271,
      "grad_norm": 1.125,
      "learning_rate": 0.00042361772915006763,
      "loss": 5.0518,
      "mean_token_accuracy": 0.2110229402780533,
      "num_tokens": 64294728.0,
      "step": 28075
    },
    {
      "entropy": 5.721891117095947,
      "epoch": 2.775800711743772,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004235912314057831,
      "loss": 5.0908,
      "mean_token_accuracy": 0.21324075907468795,
      "num_tokens": 64305873.0,
      "step": 28080
    },
    {
      "entropy": 5.672266817092895,
      "epoch": 2.776294978252274,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004235647300061311,
      "loss": 4.9361,
      "mean_token_accuracy": 0.22254360765218734,
      "num_tokens": 64318892.0,
      "step": 28085
    },
    {
      "entropy": 5.633163261413574,
      "epoch": 2.776789244760775,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004235382249517637,
      "loss": 4.931,
      "mean_token_accuracy": 0.22110871076583863,
      "num_tokens": 64328918.0,
      "step": 28090
    },
    {
      "entropy": 5.677274465560913,
      "epoch": 2.7772835112692764,
      "grad_norm": 1.125,
      "learning_rate": 0.00042351171624333293,
      "loss": 4.9759,
      "mean_token_accuracy": 0.22169163674116135,
      "num_tokens": 64338670.0,
      "step": 28095
    },
    {
      "entropy": 5.732836580276489,
      "epoch": 2.7777777777777777,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042348520388149093,
      "loss": 4.9802,
      "mean_token_accuracy": 0.21526568979024888,
      "num_tokens": 64350503.0,
      "step": 28100
    },
    {
      "entropy": 5.674016427993775,
      "epoch": 2.778272044286279,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004234586878668899,
      "loss": 4.9269,
      "mean_token_accuracy": 0.22605169266462327,
      "num_tokens": 64361176.0,
      "step": 28105
    },
    {
      "entropy": 5.713886785507202,
      "epoch": 2.7787663107947806,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004234321682001821,
      "loss": 5.0496,
      "mean_token_accuracy": 0.2157477542757988,
      "num_tokens": 64373397.0,
      "step": 28110
    },
    {
      "entropy": 5.672181940078735,
      "epoch": 2.779260577303282,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042340564488201987,
      "loss": 4.9796,
      "mean_token_accuracy": 0.21026417911052703,
      "num_tokens": 64384910.0,
      "step": 28115
    },
    {
      "entropy": 5.648859119415283,
      "epoch": 2.779754843811783,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004233791179130559,
      "loss": 4.9853,
      "mean_token_accuracy": 0.21526231616735458,
      "num_tokens": 64396787.0,
      "step": 28120
    },
    {
      "entropy": 5.695188283920288,
      "epoch": 2.780249110320285,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004233525872939424,
      "loss": 5.0516,
      "mean_token_accuracy": 0.2135879412293434,
      "num_tokens": 64407911.0,
      "step": 28125
    },
    {
      "entropy": 5.803978490829468,
      "epoch": 2.780743376828786,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042332605302533246,
      "loss": 5.0805,
      "mean_token_accuracy": 0.2148550420999527,
      "num_tokens": 64420697.0,
      "step": 28130
    },
    {
      "entropy": 5.709131813049316,
      "epoch": 2.7812376433372874,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042329951510787847,
      "loss": 4.9422,
      "mean_token_accuracy": 0.22601468563079835,
      "num_tokens": 64431675.0,
      "step": 28135
    },
    {
      "entropy": 5.701867866516113,
      "epoch": 2.7817319098457887,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042327297354223346,
      "loss": 5.005,
      "mean_token_accuracy": 0.21770417243242263,
      "num_tokens": 64443398.0,
      "step": 28140
    },
    {
      "entropy": 5.672602653503418,
      "epoch": 2.7822261763542904,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004232464283290502,
      "loss": 5.0216,
      "mean_token_accuracy": 0.21968140006065368,
      "num_tokens": 64455306.0,
      "step": 28145
    },
    {
      "entropy": 5.676162910461426,
      "epoch": 2.7827204428627916,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004232198794689819,
      "loss": 4.9635,
      "mean_token_accuracy": 0.22292641699314117,
      "num_tokens": 64466475.0,
      "step": 28150
    },
    {
      "entropy": 5.667986869812012,
      "epoch": 2.783214709371293,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004231933269626815,
      "loss": 4.9788,
      "mean_token_accuracy": 0.2228396490216255,
      "num_tokens": 64476729.0,
      "step": 28155
    },
    {
      "entropy": 5.766189002990723,
      "epoch": 2.7837089758797946,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004231667708108024,
      "loss": 5.071,
      "mean_token_accuracy": 0.2161652848124504,
      "num_tokens": 64489403.0,
      "step": 28160
    },
    {
      "entropy": 5.681957054138183,
      "epoch": 2.784203242388296,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042314021101399763,
      "loss": 4.9879,
      "mean_token_accuracy": 0.2309042751789093,
      "num_tokens": 64501000.0,
      "step": 28165
    },
    {
      "entropy": 5.701148557662964,
      "epoch": 2.784697508896797,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004231136475729207,
      "loss": 4.9749,
      "mean_token_accuracy": 0.22036862820386888,
      "num_tokens": 64513504.0,
      "step": 28170
    },
    {
      "entropy": 5.643229722976685,
      "epoch": 2.7851917754052984,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042308708048822513,
      "loss": 4.9419,
      "mean_token_accuracy": 0.22409293204545974,
      "num_tokens": 64525522.0,
      "step": 28175
    },
    {
      "entropy": 5.706642532348633,
      "epoch": 2.7856860419137996,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004230605097605644,
      "loss": 5.0417,
      "mean_token_accuracy": 0.22110858857631682,
      "num_tokens": 64537261.0,
      "step": 28180
    },
    {
      "entropy": 5.702304029464722,
      "epoch": 2.7861803084223014,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042303393539059204,
      "loss": 4.9407,
      "mean_token_accuracy": 0.2263340100646019,
      "num_tokens": 64548527.0,
      "step": 28185
    },
    {
      "entropy": 5.625427198410034,
      "epoch": 2.7866745749308026,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000423007357378962,
      "loss": 4.9274,
      "mean_token_accuracy": 0.22905777841806413,
      "num_tokens": 64560196.0,
      "step": 28190
    },
    {
      "entropy": 5.787480115890503,
      "epoch": 2.7871688414393043,
      "grad_norm": 1.203125,
      "learning_rate": 0.000422980775726328,
      "loss": 5.048,
      "mean_token_accuracy": 0.22112140953540801,
      "num_tokens": 64570828.0,
      "step": 28195
    },
    {
      "entropy": 5.7775331974029545,
      "epoch": 2.7876631079478056,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004229541904333439,
      "loss": 5.0968,
      "mean_token_accuracy": 0.2063877373933792,
      "num_tokens": 64581845.0,
      "step": 28200
    },
    {
      "entropy": 5.721597576141358,
      "epoch": 2.788157374456307,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042292760150066386,
      "loss": 5.0367,
      "mean_token_accuracy": 0.21710703074932097,
      "num_tokens": 64593105.0,
      "step": 28205
    },
    {
      "entropy": 5.682293891906738,
      "epoch": 2.788651640964808,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042290100892894184,
      "loss": 5.0053,
      "mean_token_accuracy": 0.21532258987426758,
      "num_tokens": 64605075.0,
      "step": 28210
    },
    {
      "entropy": 5.711063861846924,
      "epoch": 2.7891459074733094,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042287441271883204,
      "loss": 5.0281,
      "mean_token_accuracy": 0.21748075783252716,
      "num_tokens": 64617091.0,
      "step": 28215
    },
    {
      "entropy": 5.751304006576538,
      "epoch": 2.789640173981811,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004228478128709888,
      "loss": 4.9436,
      "mean_token_accuracy": 0.22211069017648696,
      "num_tokens": 64627870.0,
      "step": 28220
    },
    {
      "entropy": 5.695521068572998,
      "epoch": 2.7901344404903123,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004228212093860664,
      "loss": 5.012,
      "mean_token_accuracy": 0.21488795280456544,
      "num_tokens": 64639583.0,
      "step": 28225
    },
    {
      "entropy": 5.706600952148437,
      "epoch": 2.7906287069988136,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004227946022647192,
      "loss": 4.9617,
      "mean_token_accuracy": 0.22071884125471114,
      "num_tokens": 64650477.0,
      "step": 28230
    },
    {
      "entropy": 5.741800260543823,
      "epoch": 2.7911229735073153,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000422767991507602,
      "loss": 5.0007,
      "mean_token_accuracy": 0.21872826367616655,
      "num_tokens": 64660994.0,
      "step": 28235
    },
    {
      "entropy": 5.728321647644043,
      "epoch": 2.7916172400158166,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004227413771153692,
      "loss": 5.0278,
      "mean_token_accuracy": 0.2242534875869751,
      "num_tokens": 64672348.0,
      "step": 28240
    },
    {
      "entropy": 5.714672565460205,
      "epoch": 2.792111506524318,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042271475908867563,
      "loss": 5.0026,
      "mean_token_accuracy": 0.21088871508836746,
      "num_tokens": 64683352.0,
      "step": 28245
    },
    {
      "entropy": 5.75800518989563,
      "epoch": 2.792605773032819,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004226881374281761,
      "loss": 5.0601,
      "mean_token_accuracy": 0.2142708867788315,
      "num_tokens": 64695429.0,
      "step": 28250
    },
    {
      "entropy": 5.660664653778076,
      "epoch": 2.793100039541321,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004226615121345254,
      "loss": 4.9222,
      "mean_token_accuracy": 0.22085565626621245,
      "num_tokens": 64706438.0,
      "step": 28255
    },
    {
      "entropy": 5.701246881484986,
      "epoch": 2.793594306049822,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004226348832083787,
      "loss": 4.9852,
      "mean_token_accuracy": 0.21242036670446396,
      "num_tokens": 64717991.0,
      "step": 28260
    },
    {
      "entropy": 5.717717409133911,
      "epoch": 2.7940885725583233,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042260825065039095,
      "loss": 5.0494,
      "mean_token_accuracy": 0.21643056720495224,
      "num_tokens": 64730196.0,
      "step": 28265
    },
    {
      "entropy": 5.668759489059449,
      "epoch": 2.794582839066825,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004225816144612173,
      "loss": 5.0206,
      "mean_token_accuracy": 0.2109068140387535,
      "num_tokens": 64742977.0,
      "step": 28270
    },
    {
      "entropy": 5.721907281875611,
      "epoch": 2.7950771055753263,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042255497464151307,
      "loss": 5.0349,
      "mean_token_accuracy": 0.20828558057546614,
      "num_tokens": 64754101.0,
      "step": 28275
    },
    {
      "entropy": 5.693812847137451,
      "epoch": 2.7955713720838276,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042252833119193345,
      "loss": 5.0168,
      "mean_token_accuracy": 0.21898158490657807,
      "num_tokens": 64766611.0,
      "step": 28280
    },
    {
      "entropy": 5.685315895080566,
      "epoch": 2.796065638592329,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042250168411313413,
      "loss": 5.0099,
      "mean_token_accuracy": 0.22030934691429138,
      "num_tokens": 64778764.0,
      "step": 28285
    },
    {
      "entropy": 5.697000217437744,
      "epoch": 2.7965599051008305,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042247503340577046,
      "loss": 4.9787,
      "mean_token_accuracy": 0.2200830966234207,
      "num_tokens": 64790208.0,
      "step": 28290
    },
    {
      "entropy": 5.7213341236114506,
      "epoch": 2.797054171609332,
      "grad_norm": 1.125,
      "learning_rate": 0.0004224483790704981,
      "loss": 5.0441,
      "mean_token_accuracy": 0.2125803902745247,
      "num_tokens": 64801692.0,
      "step": 28295
    },
    {
      "entropy": 5.743679189682007,
      "epoch": 2.797548438117833,
      "grad_norm": 1.3125,
      "learning_rate": 0.00042242172110797273,
      "loss": 4.9338,
      "mean_token_accuracy": 0.2200103297829628,
      "num_tokens": 64811919.0,
      "step": 28300
    },
    {
      "entropy": 5.687502098083496,
      "epoch": 2.7980427046263348,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042239505951885013,
      "loss": 4.9622,
      "mean_token_accuracy": 0.21980877816677094,
      "num_tokens": 64821679.0,
      "step": 28305
    },
    {
      "entropy": 5.63314118385315,
      "epoch": 2.798536971134836,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042236839430378624,
      "loss": 4.9946,
      "mean_token_accuracy": 0.2218866840004921,
      "num_tokens": 64833295.0,
      "step": 28310
    },
    {
      "entropy": 5.6306657791137695,
      "epoch": 2.7990312376433373,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042234172546343695,
      "loss": 4.9019,
      "mean_token_accuracy": 0.2301711618900299,
      "num_tokens": 64844970.0,
      "step": 28315
    },
    {
      "entropy": 5.683224439620972,
      "epoch": 2.7995255041518385,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042231505299845836,
      "loss": 4.9013,
      "mean_token_accuracy": 0.22618651241064072,
      "num_tokens": 64855496.0,
      "step": 28320
    },
    {
      "entropy": 5.668324136734009,
      "epoch": 2.80001977066034,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042228837690950664,
      "loss": 4.9604,
      "mean_token_accuracy": 0.2243376851081848,
      "num_tokens": 64866836.0,
      "step": 28325
    },
    {
      "entropy": 5.659656143188476,
      "epoch": 2.8005140371688415,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042226169719723793,
      "loss": 4.9949,
      "mean_token_accuracy": 0.22212968468666078,
      "num_tokens": 64879105.0,
      "step": 28330
    },
    {
      "entropy": 5.734632158279419,
      "epoch": 2.8010083036773428,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004222350138623087,
      "loss": 5.0069,
      "mean_token_accuracy": 0.22199305891990662,
      "num_tokens": 64891227.0,
      "step": 28335
    },
    {
      "entropy": 5.6966410160064695,
      "epoch": 2.8015025701858445,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042220832690537525,
      "loss": 5.0107,
      "mean_token_accuracy": 0.21483013331890105,
      "num_tokens": 64904070.0,
      "step": 28340
    },
    {
      "entropy": 5.720845794677734,
      "epoch": 2.8019968366943457,
      "grad_norm": 1.25,
      "learning_rate": 0.0004221816363270941,
      "loss": 4.9349,
      "mean_token_accuracy": 0.22258009463548661,
      "num_tokens": 64915087.0,
      "step": 28345
    },
    {
      "entropy": 5.612367773056031,
      "epoch": 2.802491103202847,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042215494212812193,
      "loss": 4.8931,
      "mean_token_accuracy": 0.23271571099758148,
      "num_tokens": 64926163.0,
      "step": 28350
    },
    {
      "entropy": 5.647972822189331,
      "epoch": 2.8029853697113483,
      "grad_norm": 1.25,
      "learning_rate": 0.00042212824430911533,
      "loss": 4.9321,
      "mean_token_accuracy": 0.22908053249120713,
      "num_tokens": 64936513.0,
      "step": 28355
    },
    {
      "entropy": 5.618739175796509,
      "epoch": 2.8034796362198495,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004221015428707311,
      "loss": 4.8712,
      "mean_token_accuracy": 0.23348571807146073,
      "num_tokens": 64947232.0,
      "step": 28360
    },
    {
      "entropy": 5.732770156860352,
      "epoch": 2.8039739027283512,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000422074837813626,
      "loss": 5.0486,
      "mean_token_accuracy": 0.21315110474824905,
      "num_tokens": 64958622.0,
      "step": 28365
    },
    {
      "entropy": 5.757647466659546,
      "epoch": 2.8044681692368525,
      "grad_norm": 1.125,
      "learning_rate": 0.00042204812913845726,
      "loss": 5.0167,
      "mean_token_accuracy": 0.21504825204610825,
      "num_tokens": 64969975.0,
      "step": 28370
    },
    {
      "entropy": 5.727634763717651,
      "epoch": 2.8049624357453538,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042202141684588156,
      "loss": 4.9243,
      "mean_token_accuracy": 0.21933132112026216,
      "num_tokens": 64981195.0,
      "step": 28375
    },
    {
      "entropy": 5.734689950942993,
      "epoch": 2.8054567022538555,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004219947009365563,
      "loss": 5.0746,
      "mean_token_accuracy": 0.2178567960858345,
      "num_tokens": 64992265.0,
      "step": 28380
    },
    {
      "entropy": 5.68792953491211,
      "epoch": 2.8059509687623567,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004219679814111386,
      "loss": 5.035,
      "mean_token_accuracy": 0.21525471061468124,
      "num_tokens": 65003613.0,
      "step": 28385
    },
    {
      "entropy": 5.666007852554321,
      "epoch": 2.806445235270858,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004219412582702857,
      "loss": 4.9468,
      "mean_token_accuracy": 0.22464779317378997,
      "num_tokens": 65014768.0,
      "step": 28390
    },
    {
      "entropy": 5.724464797973633,
      "epoch": 2.8069395017793592,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042191453151465506,
      "loss": 4.9971,
      "mean_token_accuracy": 0.21822245121002198,
      "num_tokens": 65026425.0,
      "step": 28395
    },
    {
      "entropy": 5.721783542633057,
      "epoch": 2.807433768287861,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042188780114490413,
      "loss": 5.0094,
      "mean_token_accuracy": 0.22101344168186188,
      "num_tokens": 65037624.0,
      "step": 28400
    },
    {
      "entropy": 5.724779033660889,
      "epoch": 2.807928034796362,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004218610671616905,
      "loss": 5.0953,
      "mean_token_accuracy": 0.21858241707086562,
      "num_tokens": 65048659.0,
      "step": 28405
    },
    {
      "entropy": 5.674625015258789,
      "epoch": 2.8084223013048635,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042183432956567184,
      "loss": 4.9247,
      "mean_token_accuracy": 0.2311659649014473,
      "num_tokens": 65059084.0,
      "step": 28410
    },
    {
      "entropy": 5.690136861801148,
      "epoch": 2.808916567813365,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042180758835750595,
      "loss": 4.9554,
      "mean_token_accuracy": 0.22586261928081514,
      "num_tokens": 65071525.0,
      "step": 28415
    },
    {
      "entropy": 5.699678087234497,
      "epoch": 2.8094108343218664,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004217808435378506,
      "loss": 5.0428,
      "mean_token_accuracy": 0.2175804167985916,
      "num_tokens": 65082738.0,
      "step": 28420
    },
    {
      "entropy": 5.692951965332031,
      "epoch": 2.8099051008303677,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004217540951073637,
      "loss": 5.0342,
      "mean_token_accuracy": 0.2167535498738289,
      "num_tokens": 65094988.0,
      "step": 28425
    },
    {
      "entropy": 5.7198585033416744,
      "epoch": 2.810399367338869,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004217273430667033,
      "loss": 5.0132,
      "mean_token_accuracy": 0.21366462111473083,
      "num_tokens": 65106412.0,
      "step": 28430
    },
    {
      "entropy": 5.718555450439453,
      "epoch": 2.8108936338473702,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004217005874165274,
      "loss": 5.0023,
      "mean_token_accuracy": 0.22496997117996215,
      "num_tokens": 65117756.0,
      "step": 28435
    },
    {
      "entropy": 5.713331604003907,
      "epoch": 2.811387900355872,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004216738281574944,
      "loss": 5.0557,
      "mean_token_accuracy": 0.2145448699593544,
      "num_tokens": 65129368.0,
      "step": 28440
    },
    {
      "entropy": 5.739205932617187,
      "epoch": 2.811882166864373,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042164706529026255,
      "loss": 5.0483,
      "mean_token_accuracy": 0.21423470824956894,
      "num_tokens": 65140553.0,
      "step": 28445
    },
    {
      "entropy": 5.653905963897705,
      "epoch": 2.812376433372875,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004216202988154901,
      "loss": 4.945,
      "mean_token_accuracy": 0.22129978984594345,
      "num_tokens": 65152047.0,
      "step": 28450
    },
    {
      "entropy": 5.706011486053467,
      "epoch": 2.812870699881376,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004215935287338354,
      "loss": 4.9786,
      "mean_token_accuracy": 0.21933094561100006,
      "num_tokens": 65164001.0,
      "step": 28455
    },
    {
      "entropy": 5.72477240562439,
      "epoch": 2.8133649663898774,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004215667550459573,
      "loss": 4.9901,
      "mean_token_accuracy": 0.2267170399427414,
      "num_tokens": 65175949.0,
      "step": 28460
    },
    {
      "entropy": 5.708213758468628,
      "epoch": 2.8138592328983787,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004215399777525142,
      "loss": 4.9995,
      "mean_token_accuracy": 0.22296242713928222,
      "num_tokens": 65186725.0,
      "step": 28465
    },
    {
      "entropy": 5.733828592300415,
      "epoch": 2.81435349940688,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000421513196854165,
      "loss": 4.9819,
      "mean_token_accuracy": 0.21910453736782073,
      "num_tokens": 65198489.0,
      "step": 28470
    },
    {
      "entropy": 5.636229848861694,
      "epoch": 2.8148477659153817,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004214864123515684,
      "loss": 4.9111,
      "mean_token_accuracy": 0.222913259267807,
      "num_tokens": 65210572.0,
      "step": 28475
    },
    {
      "entropy": 5.693554401397705,
      "epoch": 2.815342032423883,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042145962424538336,
      "loss": 5.0121,
      "mean_token_accuracy": 0.21850177198648452,
      "num_tokens": 65221785.0,
      "step": 28480
    },
    {
      "entropy": 5.749634599685669,
      "epoch": 2.815836298932384,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004214328325362687,
      "loss": 5.0471,
      "mean_token_accuracy": 0.21214486807584762,
      "num_tokens": 65233998.0,
      "step": 28485
    },
    {
      "entropy": 5.744606971740723,
      "epoch": 2.816330565440886,
      "grad_norm": 1.125,
      "learning_rate": 0.0004214060372248838,
      "loss": 5.03,
      "mean_token_accuracy": 0.20983162075281142,
      "num_tokens": 65243907.0,
      "step": 28490
    },
    {
      "entropy": 5.729289293289185,
      "epoch": 2.816824831949387,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042137923831188763,
      "loss": 4.9929,
      "mean_token_accuracy": 0.2183765560388565,
      "num_tokens": 65254555.0,
      "step": 28495
    },
    {
      "entropy": 5.740906715393066,
      "epoch": 2.8173190984578884,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042135243579793946,
      "loss": 4.9995,
      "mean_token_accuracy": 0.21524958610534667,
      "num_tokens": 65265751.0,
      "step": 28500
    },
    {
      "entropy": 5.670943021774292,
      "epoch": 2.8178133649663897,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042132562968369864,
      "loss": 4.9317,
      "mean_token_accuracy": 0.225890851020813,
      "num_tokens": 65275839.0,
      "step": 28505
    },
    {
      "entropy": 5.691793775558471,
      "epoch": 2.8183076314748914,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004212988199698247,
      "loss": 4.9525,
      "mean_token_accuracy": 0.2192242607474327,
      "num_tokens": 65287796.0,
      "step": 28510
    },
    {
      "entropy": 5.71829628944397,
      "epoch": 2.8188018979833926,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042127200665697715,
      "loss": 5.0301,
      "mean_token_accuracy": 0.22356114238500596,
      "num_tokens": 65299014.0,
      "step": 28515
    },
    {
      "entropy": 5.706383609771729,
      "epoch": 2.819296164491894,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042124518974581536,
      "loss": 4.9748,
      "mean_token_accuracy": 0.22318002879619597,
      "num_tokens": 65309830.0,
      "step": 28520
    },
    {
      "entropy": 5.754660701751709,
      "epoch": 2.8197904310003956,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004212183692369994,
      "loss": 5.0589,
      "mean_token_accuracy": 0.21097784191370011,
      "num_tokens": 65321921.0,
      "step": 28525
    },
    {
      "entropy": 5.685256147384644,
      "epoch": 2.820284697508897,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004211915451311887,
      "loss": 4.9744,
      "mean_token_accuracy": 0.22360786199569702,
      "num_tokens": 65334175.0,
      "step": 28530
    },
    {
      "entropy": 5.720682430267334,
      "epoch": 2.820778964017398,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004211647174290434,
      "loss": 5.065,
      "mean_token_accuracy": 0.21811071187257766,
      "num_tokens": 65345460.0,
      "step": 28535
    },
    {
      "entropy": 5.680964231491089,
      "epoch": 2.8212732305258994,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042113788613122333,
      "loss": 4.9472,
      "mean_token_accuracy": 0.2261755034327507,
      "num_tokens": 65357252.0,
      "step": 28540
    },
    {
      "entropy": 5.716677188873291,
      "epoch": 2.821767497034401,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004211110512383887,
      "loss": 5.1166,
      "mean_token_accuracy": 0.21755707412958145,
      "num_tokens": 65370195.0,
      "step": 28545
    },
    {
      "entropy": 5.6663566589355465,
      "epoch": 2.8222617635429024,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004210842127511995,
      "loss": 4.9633,
      "mean_token_accuracy": 0.22314319014549255,
      "num_tokens": 65382531.0,
      "step": 28550
    },
    {
      "entropy": 5.77213487625122,
      "epoch": 2.8227560300514036,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004210573706703159,
      "loss": 5.0771,
      "mean_token_accuracy": 0.2115390807390213,
      "num_tokens": 65393829.0,
      "step": 28555
    },
    {
      "entropy": 5.741116046905518,
      "epoch": 2.8232502965599053,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004210305249963984,
      "loss": 4.9816,
      "mean_token_accuracy": 0.21960900574922562,
      "num_tokens": 65404479.0,
      "step": 28560
    },
    {
      "entropy": 5.748715925216675,
      "epoch": 2.8237445630684066,
      "grad_norm": 1.25,
      "learning_rate": 0.0004210036757301073,
      "loss": 5.0473,
      "mean_token_accuracy": 0.21575185358524324,
      "num_tokens": 65415468.0,
      "step": 28565
    },
    {
      "entropy": 5.624507570266724,
      "epoch": 2.824238829576908,
      "grad_norm": 1.125,
      "learning_rate": 0.0004209768228721032,
      "loss": 4.9035,
      "mean_token_accuracy": 0.22806870490312575,
      "num_tokens": 65425994.0,
      "step": 28570
    },
    {
      "entropy": 5.636254167556762,
      "epoch": 2.824733096085409,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042094996642304657,
      "loss": 5.0252,
      "mean_token_accuracy": 0.2171539694070816,
      "num_tokens": 65438394.0,
      "step": 28575
    },
    {
      "entropy": 5.691539669036866,
      "epoch": 2.8252273625939104,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004209231063835981,
      "loss": 5.0509,
      "mean_token_accuracy": 0.2207360938191414,
      "num_tokens": 65450248.0,
      "step": 28580
    },
    {
      "entropy": 5.7311779975891115,
      "epoch": 2.825721629102412,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042089624275441865,
      "loss": 5.0152,
      "mean_token_accuracy": 0.21828146427869796,
      "num_tokens": 65460803.0,
      "step": 28585
    },
    {
      "entropy": 5.70726752281189,
      "epoch": 2.8262158956109134,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004208693755361689,
      "loss": 5.0096,
      "mean_token_accuracy": 0.22166830003261567,
      "num_tokens": 65471899.0,
      "step": 28590
    },
    {
      "entropy": 5.705237531661988,
      "epoch": 2.826710162119415,
      "grad_norm": 1.125,
      "learning_rate": 0.0004208425047295099,
      "loss": 5.0283,
      "mean_token_accuracy": 0.21661674827337266,
      "num_tokens": 65483767.0,
      "step": 28595
    },
    {
      "entropy": 5.677091407775879,
      "epoch": 2.8272044286279163,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004208156303351027,
      "loss": 4.8931,
      "mean_token_accuracy": 0.22830077707767488,
      "num_tokens": 65494859.0,
      "step": 28600
    },
    {
      "entropy": 5.784985876083374,
      "epoch": 2.8276986951364176,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042078875235360835,
      "loss": 5.0958,
      "mean_token_accuracy": 0.2109665244817734,
      "num_tokens": 65506411.0,
      "step": 28605
    },
    {
      "entropy": 5.713237619400024,
      "epoch": 2.828192961644919,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042076187078568804,
      "loss": 4.9581,
      "mean_token_accuracy": 0.2189013510942459,
      "num_tokens": 65516655.0,
      "step": 28610
    },
    {
      "entropy": 5.566547203063965,
      "epoch": 2.82868722815342,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004207349856320031,
      "loss": 4.9543,
      "mean_token_accuracy": 0.22132923305034638,
      "num_tokens": 65529305.0,
      "step": 28615
    },
    {
      "entropy": 5.696669054031372,
      "epoch": 2.829181494661922,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042070809689321497,
      "loss": 5.0035,
      "mean_token_accuracy": 0.21742099970579148,
      "num_tokens": 65542129.0,
      "step": 28620
    },
    {
      "entropy": 5.705373668670655,
      "epoch": 2.829675761170423,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000420681204569985,
      "loss": 4.9519,
      "mean_token_accuracy": 0.23366078436374665,
      "num_tokens": 65554240.0,
      "step": 28625
    },
    {
      "entropy": 5.675036144256592,
      "epoch": 2.8301700276789243,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042065430866297486,
      "loss": 4.9295,
      "mean_token_accuracy": 0.22743548154830934,
      "num_tokens": 65564994.0,
      "step": 28630
    },
    {
      "entropy": 5.67653169631958,
      "epoch": 2.830664294187426,
      "grad_norm": 1.125,
      "learning_rate": 0.0004206274091728461,
      "loss": 4.9536,
      "mean_token_accuracy": 0.2186505004763603,
      "num_tokens": 65575986.0,
      "step": 28635
    },
    {
      "entropy": 5.753972864151001,
      "epoch": 2.8311585606959273,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004206005061002604,
      "loss": 5.0734,
      "mean_token_accuracy": 0.21588441282510756,
      "num_tokens": 65587792.0,
      "step": 28640
    },
    {
      "entropy": 5.798526096343994,
      "epoch": 2.8316528272044286,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004205735994458798,
      "loss": 5.1731,
      "mean_token_accuracy": 0.19955829977989198,
      "num_tokens": 65599845.0,
      "step": 28645
    },
    {
      "entropy": 5.745310211181641,
      "epoch": 2.83214709371293,
      "grad_norm": 1.25,
      "learning_rate": 0.0004205466892103659,
      "loss": 5.042,
      "mean_token_accuracy": 0.2107723817229271,
      "num_tokens": 65611796.0,
      "step": 28650
    },
    {
      "entropy": 5.816731405258179,
      "epoch": 2.8326413602214315,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042051977539438103,
      "loss": 5.0931,
      "mean_token_accuracy": 0.2105873018503189,
      "num_tokens": 65622319.0,
      "step": 28655
    },
    {
      "entropy": 5.6564515113830565,
      "epoch": 2.833135626729933,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042049285799858707,
      "loss": 4.9615,
      "mean_token_accuracy": 0.22197322249412538,
      "num_tokens": 65634403.0,
      "step": 28660
    },
    {
      "entropy": 5.660837078094483,
      "epoch": 2.833629893238434,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00042046593702364614,
      "loss": 5.0188,
      "mean_token_accuracy": 0.21811422556638718,
      "num_tokens": 65644719.0,
      "step": 28665
    },
    {
      "entropy": 5.851485824584961,
      "epoch": 2.8341241597469358,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004204390124702207,
      "loss": 5.1012,
      "mean_token_accuracy": 0.2078378066420555,
      "num_tokens": 65657083.0,
      "step": 28670
    },
    {
      "entropy": 5.6949550151824955,
      "epoch": 2.834618426255437,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004204120843389729,
      "loss": 5.0136,
      "mean_token_accuracy": 0.2129308670759201,
      "num_tokens": 65667579.0,
      "step": 28675
    },
    {
      "entropy": 5.676826858520508,
      "epoch": 2.8351126927639383,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042038515263056537,
      "loss": 4.9859,
      "mean_token_accuracy": 0.22246946394443512,
      "num_tokens": 65679066.0,
      "step": 28680
    },
    {
      "entropy": 5.7206175327301025,
      "epoch": 2.8356069592724396,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004203582173456604,
      "loss": 5.0724,
      "mean_token_accuracy": 0.21377201229333878,
      "num_tokens": 65690785.0,
      "step": 28685
    },
    {
      "entropy": 5.7132669448852536,
      "epoch": 2.836101225780941,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042033127848492086,
      "loss": 5.0242,
      "mean_token_accuracy": 0.2218717157840729,
      "num_tokens": 65701613.0,
      "step": 28690
    },
    {
      "entropy": 5.716776418685913,
      "epoch": 2.8365954922894425,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004203043360490093,
      "loss": 4.9931,
      "mean_token_accuracy": 0.2194090321660042,
      "num_tokens": 65713146.0,
      "step": 28695
    },
    {
      "entropy": 5.7508848190307615,
      "epoch": 2.837089758797944,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042027739003858855,
      "loss": 5.0349,
      "mean_token_accuracy": 0.20797315686941148,
      "num_tokens": 65724216.0,
      "step": 28700
    },
    {
      "entropy": 5.6577417850494385,
      "epoch": 2.8375840253064455,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004202504404543214,
      "loss": 4.9389,
      "mean_token_accuracy": 0.22201081216335297,
      "num_tokens": 65735357.0,
      "step": 28705
    },
    {
      "entropy": 5.713968229293823,
      "epoch": 2.8380782918149468,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004202234872968709,
      "loss": 4.9774,
      "mean_token_accuracy": 0.2148330807685852,
      "num_tokens": 65747679.0,
      "step": 28710
    },
    {
      "entropy": 5.7436014175415036,
      "epoch": 2.838572558323448,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004201965305669001,
      "loss": 4.9842,
      "mean_token_accuracy": 0.22125888019800186,
      "num_tokens": 65759126.0,
      "step": 28715
    },
    {
      "entropy": 5.700781011581421,
      "epoch": 2.8390668248319493,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042016957026507206,
      "loss": 5.0466,
      "mean_token_accuracy": 0.21566016525030135,
      "num_tokens": 65771240.0,
      "step": 28720
    },
    {
      "entropy": 5.6683799743652346,
      "epoch": 2.8395610913404505,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004201426063920501,
      "loss": 4.9293,
      "mean_token_accuracy": 0.22560853362083436,
      "num_tokens": 65783323.0,
      "step": 28725
    },
    {
      "entropy": 5.68269910812378,
      "epoch": 2.8400553578489522,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042011563894849756,
      "loss": 4.9356,
      "mean_token_accuracy": 0.22480877935886384,
      "num_tokens": 65794165.0,
      "step": 28730
    },
    {
      "entropy": 5.674124622344971,
      "epoch": 2.8405496243574535,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004200886679350778,
      "loss": 4.9124,
      "mean_token_accuracy": 0.22809515595436097,
      "num_tokens": 65804541.0,
      "step": 28735
    },
    {
      "entropy": 5.659866905212402,
      "epoch": 2.8410438908659548,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004200616933524542,
      "loss": 5.0128,
      "mean_token_accuracy": 0.22012083679437638,
      "num_tokens": 65816986.0,
      "step": 28740
    },
    {
      "entropy": 5.696641969680786,
      "epoch": 2.8415381573744565,
      "grad_norm": 1.125,
      "learning_rate": 0.0004200347152012905,
      "loss": 4.9945,
      "mean_token_accuracy": 0.21558830440044402,
      "num_tokens": 65828711.0,
      "step": 28745
    },
    {
      "entropy": 5.758247423171997,
      "epoch": 2.8420324238829577,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004200077334822503,
      "loss": 5.0161,
      "mean_token_accuracy": 0.21532185077667237,
      "num_tokens": 65840525.0,
      "step": 28750
    },
    {
      "entropy": 5.6997987747192385,
      "epoch": 2.842526690391459,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004199807481959973,
      "loss": 4.9295,
      "mean_token_accuracy": 0.22420274168252946,
      "num_tokens": 65851663.0,
      "step": 28755
    },
    {
      "entropy": 5.679914236068726,
      "epoch": 2.8430209568999603,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004199537593431955,
      "loss": 4.9793,
      "mean_token_accuracy": 0.22218133360147477,
      "num_tokens": 65863337.0,
      "step": 28760
    },
    {
      "entropy": 5.8040605068206785,
      "epoch": 2.843515223408462,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004199267669245087,
      "loss": 5.1321,
      "mean_token_accuracy": 0.20283611118793488,
      "num_tokens": 65875581.0,
      "step": 28765
    },
    {
      "entropy": 5.735746097564697,
      "epoch": 2.8440094899169632,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041989977094060096,
      "loss": 5.0031,
      "mean_token_accuracy": 0.22016289979219436,
      "num_tokens": 65886577.0,
      "step": 28770
    },
    {
      "entropy": 5.678605365753174,
      "epoch": 2.8445037564254645,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004198727713921364,
      "loss": 5.0002,
      "mean_token_accuracy": 0.2172711044549942,
      "num_tokens": 65897914.0,
      "step": 28775
    },
    {
      "entropy": 5.683418846130371,
      "epoch": 2.844998022933966,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004198457682797791,
      "loss": 4.9898,
      "mean_token_accuracy": 0.22387681752443314,
      "num_tokens": 65909781.0,
      "step": 28780
    },
    {
      "entropy": 5.702006578445435,
      "epoch": 2.8454922894424675,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041981876160419353,
      "loss": 4.9734,
      "mean_token_accuracy": 0.2218906596302986,
      "num_tokens": 65921107.0,
      "step": 28785
    },
    {
      "entropy": 5.733965587615967,
      "epoch": 2.8459865559509687,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041979175136604385,
      "loss": 5.0112,
      "mean_token_accuracy": 0.22336645275354386,
      "num_tokens": 65931166.0,
      "step": 28790
    },
    {
      "entropy": 5.781806898117066,
      "epoch": 2.84648082245947,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004197647375659947,
      "loss": 5.1202,
      "mean_token_accuracy": 0.21038123667240144,
      "num_tokens": 65943835.0,
      "step": 28795
    },
    {
      "entropy": 5.671651411056518,
      "epoch": 2.8469750889679717,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041973772020471054,
      "loss": 4.9619,
      "mean_token_accuracy": 0.22450049072504044,
      "num_tokens": 65956961.0,
      "step": 28800
    },
    {
      "entropy": 5.752721166610717,
      "epoch": 2.847469355476473,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041971069928285596,
      "loss": 5.0441,
      "mean_token_accuracy": 0.2167324036359787,
      "num_tokens": 65967774.0,
      "step": 28805
    },
    {
      "entropy": 5.671965646743774,
      "epoch": 2.847963621984974,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041968367480109577,
      "loss": 4.9885,
      "mean_token_accuracy": 0.22039819210767747,
      "num_tokens": 65978318.0,
      "step": 28810
    },
    {
      "entropy": 5.722200489044189,
      "epoch": 2.848457888493476,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004196566467600947,
      "loss": 4.9553,
      "mean_token_accuracy": 0.21982679665088653,
      "num_tokens": 65989763.0,
      "step": 28815
    },
    {
      "entropy": 5.65638108253479,
      "epoch": 2.848952155001977,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041962961516051777,
      "loss": 4.9264,
      "mean_token_accuracy": 0.23013641983270644,
      "num_tokens": 66001117.0,
      "step": 28820
    },
    {
      "entropy": 5.728678178787232,
      "epoch": 2.8494464215104784,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004196025800030298,
      "loss": 5.0121,
      "mean_token_accuracy": 0.21587574779987334,
      "num_tokens": 66012340.0,
      "step": 28825
    },
    {
      "entropy": 5.666395044326782,
      "epoch": 2.8499406880189797,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004195755412882959,
      "loss": 4.9953,
      "mean_token_accuracy": 0.22075920701026916,
      "num_tokens": 66023151.0,
      "step": 28830
    },
    {
      "entropy": 5.601254415512085,
      "epoch": 2.850434954527481,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041954849901698133,
      "loss": 4.8612,
      "mean_token_accuracy": 0.23114654123783113,
      "num_tokens": 66033399.0,
      "step": 28835
    },
    {
      "entropy": 5.649206924438476,
      "epoch": 2.8509292210359827,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041952145318975114,
      "loss": 4.9776,
      "mean_token_accuracy": 0.22748736888170243,
      "num_tokens": 66045817.0,
      "step": 28840
    },
    {
      "entropy": 5.784841823577881,
      "epoch": 2.851423487544484,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004194944038072709,
      "loss": 5.1172,
      "mean_token_accuracy": 0.2099556192755699,
      "num_tokens": 66058696.0,
      "step": 28845
    },
    {
      "entropy": 5.747118377685547,
      "epoch": 2.8519177540529856,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041946735087020586,
      "loss": 4.9526,
      "mean_token_accuracy": 0.22156989723443984,
      "num_tokens": 66070197.0,
      "step": 28850
    },
    {
      "entropy": 5.6451598644256595,
      "epoch": 2.852412020561487,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004194402943792216,
      "loss": 4.8259,
      "mean_token_accuracy": 0.23361034393310548,
      "num_tokens": 66080250.0,
      "step": 28855
    },
    {
      "entropy": 5.693163728713989,
      "epoch": 2.852906287069988,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004194132343349837,
      "loss": 5.0386,
      "mean_token_accuracy": 0.21631951183080672,
      "num_tokens": 66090957.0,
      "step": 28860
    },
    {
      "entropy": 5.68308916091919,
      "epoch": 2.8534005535784894,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004193861707381578,
      "loss": 4.9467,
      "mean_token_accuracy": 0.22356856167316436,
      "num_tokens": 66103042.0,
      "step": 28865
    },
    {
      "entropy": 5.667127418518066,
      "epoch": 2.8538948200869907,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041935910358940974,
      "loss": 4.9841,
      "mean_token_accuracy": 0.2217377617955208,
      "num_tokens": 66113832.0,
      "step": 28870
    },
    {
      "entropy": 5.635734128952026,
      "epoch": 2.8543890865954924,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004193320328894053,
      "loss": 4.9029,
      "mean_token_accuracy": 0.22974541932344436,
      "num_tokens": 66126268.0,
      "step": 28875
    },
    {
      "entropy": 5.700114774703979,
      "epoch": 2.8548833531039937,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004193049586388105,
      "loss": 4.9179,
      "mean_token_accuracy": 0.21951544731855394,
      "num_tokens": 66137590.0,
      "step": 28880
    },
    {
      "entropy": 5.602246332168579,
      "epoch": 2.855377619612495,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041927788083829127,
      "loss": 4.9436,
      "mean_token_accuracy": 0.22722191512584686,
      "num_tokens": 66149110.0,
      "step": 28885
    },
    {
      "entropy": 5.6901427745819095,
      "epoch": 2.8558718861209966,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004192507994885137,
      "loss": 4.9249,
      "mean_token_accuracy": 0.22700337320566177,
      "num_tokens": 66160116.0,
      "step": 28890
    },
    {
      "entropy": 5.764993190765381,
      "epoch": 2.856366152629498,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004192237145901443,
      "loss": 5.064,
      "mean_token_accuracy": 0.20774249583482743,
      "num_tokens": 66170743.0,
      "step": 28895
    },
    {
      "entropy": 5.7725286960601805,
      "epoch": 2.856860419137999,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000419196626143849,
      "loss": 5.0084,
      "mean_token_accuracy": 0.2114725425839424,
      "num_tokens": 66182280.0,
      "step": 28900
    },
    {
      "entropy": 5.745388793945312,
      "epoch": 2.8573546856465004,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041916953415029447,
      "loss": 5.0446,
      "mean_token_accuracy": 0.21913896054029464,
      "num_tokens": 66193387.0,
      "step": 28905
    },
    {
      "entropy": 5.703210258483887,
      "epoch": 2.857848952155002,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004191424386101469,
      "loss": 4.9916,
      "mean_token_accuracy": 0.2136000171303749,
      "num_tokens": 66206180.0,
      "step": 28910
    },
    {
      "entropy": 5.651580667495727,
      "epoch": 2.8583432186635034,
      "grad_norm": 1.203125,
      "learning_rate": 0.000419115339524073,
      "loss": 4.9346,
      "mean_token_accuracy": 0.22482164353132247,
      "num_tokens": 66217495.0,
      "step": 28915
    },
    {
      "entropy": 5.745947933197021,
      "epoch": 2.8588374851720046,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004190882368927394,
      "loss": 5.0219,
      "mean_token_accuracy": 0.21022023856639863,
      "num_tokens": 66229107.0,
      "step": 28920
    },
    {
      "entropy": 5.7303589344024655,
      "epoch": 2.8593317516805064,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004190611307168128,
      "loss": 4.9677,
      "mean_token_accuracy": 0.2268611565232277,
      "num_tokens": 66240480.0,
      "step": 28925
    },
    {
      "entropy": 5.630237007141114,
      "epoch": 2.8598260181890076,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041903402099696007,
      "loss": 4.9349,
      "mean_token_accuracy": 0.23011812418699265,
      "num_tokens": 66251785.0,
      "step": 28930
    },
    {
      "entropy": 5.676721525192261,
      "epoch": 2.860320284697509,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004190069077338481,
      "loss": 4.9099,
      "mean_token_accuracy": 0.22742154598236083,
      "num_tokens": 66262557.0,
      "step": 28935
    },
    {
      "entropy": 5.7323627948760985,
      "epoch": 2.86081455120601,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004189797909281438,
      "loss": 5.0518,
      "mean_token_accuracy": 0.21916950047016143,
      "num_tokens": 66272931.0,
      "step": 28940
    },
    {
      "entropy": 5.705012798309326,
      "epoch": 2.8613088177145114,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041895267058051434,
      "loss": 4.9839,
      "mean_token_accuracy": 0.22632209211587906,
      "num_tokens": 66283550.0,
      "step": 28945
    },
    {
      "entropy": 5.709415864944458,
      "epoch": 2.861803084223013,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041892554669162687,
      "loss": 5.0379,
      "mean_token_accuracy": 0.21380527019500734,
      "num_tokens": 66294740.0,
      "step": 28950
    },
    {
      "entropy": 5.6588582515716555,
      "epoch": 2.8622973507315144,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041889841926214856,
      "loss": 4.9103,
      "mean_token_accuracy": 0.22401730120182037,
      "num_tokens": 66306241.0,
      "step": 28955
    },
    {
      "entropy": 5.765018367767334,
      "epoch": 2.862791617240016,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041887128829274685,
      "loss": 5.0233,
      "mean_token_accuracy": 0.21215690672397614,
      "num_tokens": 66320006.0,
      "step": 28960
    },
    {
      "entropy": 5.746566438674927,
      "epoch": 2.8632858837485173,
      "grad_norm": 1.125,
      "learning_rate": 0.00041884415378408917,
      "loss": 4.9836,
      "mean_token_accuracy": 0.21629102081060408,
      "num_tokens": 66330810.0,
      "step": 28965
    },
    {
      "entropy": 5.721741676330566,
      "epoch": 2.8637801502570186,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004188170157368429,
      "loss": 5.0164,
      "mean_token_accuracy": 0.21850817948579787,
      "num_tokens": 66341858.0,
      "step": 28970
    },
    {
      "entropy": 5.681955099105835,
      "epoch": 2.86427441676552,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004187898741516758,
      "loss": 4.9948,
      "mean_token_accuracy": 0.2210603728890419,
      "num_tokens": 66353175.0,
      "step": 28975
    },
    {
      "entropy": 5.747458124160767,
      "epoch": 2.864768683274021,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004187627290292555,
      "loss": 5.0063,
      "mean_token_accuracy": 0.22123963087797166,
      "num_tokens": 66364245.0,
      "step": 28980
    },
    {
      "entropy": 5.653993940353393,
      "epoch": 2.865262949782523,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041873558037024974,
      "loss": 5.0233,
      "mean_token_accuracy": 0.21779378801584243,
      "num_tokens": 66376793.0,
      "step": 28985
    },
    {
      "entropy": 5.740129041671753,
      "epoch": 2.865757216291024,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004187084281753264,
      "loss": 5.052,
      "mean_token_accuracy": 0.21087127029895783,
      "num_tokens": 66388130.0,
      "step": 28990
    },
    {
      "entropy": 5.799789762496948,
      "epoch": 2.8662514827995254,
      "grad_norm": 1.25,
      "learning_rate": 0.00041868127244515336,
      "loss": 5.0263,
      "mean_token_accuracy": 0.21852745562791825,
      "num_tokens": 66400784.0,
      "step": 28995
    },
    {
      "entropy": 5.717172908782959,
      "epoch": 2.866745749308027,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041865411318039883,
      "loss": 4.9999,
      "mean_token_accuracy": 0.21946692913770677,
      "num_tokens": 66412159.0,
      "step": 29000
    },
    {
      "entropy": 5.685863590240478,
      "epoch": 2.8672400158165283,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041862695038173085,
      "loss": 4.9638,
      "mean_token_accuracy": 0.22451932430267335,
      "num_tokens": 66423471.0,
      "step": 29005
    },
    {
      "entropy": 5.758540678024292,
      "epoch": 2.8677342823250296,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004185997840498175,
      "loss": 5.1052,
      "mean_token_accuracy": 0.2064451366662979,
      "num_tokens": 66436889.0,
      "step": 29010
    },
    {
      "entropy": 5.760631465911866,
      "epoch": 2.868228548833531,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004185726141853272,
      "loss": 5.0448,
      "mean_token_accuracy": 0.2159798413515091,
      "num_tokens": 66448157.0,
      "step": 29015
    },
    {
      "entropy": 5.6205979347229,
      "epoch": 2.8687228153420326,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004185454407889284,
      "loss": 4.8841,
      "mean_token_accuracy": 0.23127532452344896,
      "num_tokens": 66459877.0,
      "step": 29020
    },
    {
      "entropy": 5.739146757125854,
      "epoch": 2.869217081850534,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004185182638612895,
      "loss": 4.9615,
      "mean_token_accuracy": 0.22201238125562667,
      "num_tokens": 66469648.0,
      "step": 29025
    },
    {
      "entropy": 5.727696895599365,
      "epoch": 2.869711348359035,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041849108340307893,
      "loss": 5.0718,
      "mean_token_accuracy": 0.2139240935444832,
      "num_tokens": 66480878.0,
      "step": 29030
    },
    {
      "entropy": 5.796301221847534,
      "epoch": 2.870205614867537,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041846389941496557,
      "loss": 5.105,
      "mean_token_accuracy": 0.21072110831737517,
      "num_tokens": 66492099.0,
      "step": 29035
    },
    {
      "entropy": 5.760681295394898,
      "epoch": 2.870699881376038,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041843671189761795,
      "loss": 5.017,
      "mean_token_accuracy": 0.2149985611438751,
      "num_tokens": 66503194.0,
      "step": 29040
    },
    {
      "entropy": 5.728110122680664,
      "epoch": 2.8711941478845393,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000418409520851705,
      "loss": 5.0422,
      "mean_token_accuracy": 0.21155631691217422,
      "num_tokens": 66513393.0,
      "step": 29045
    },
    {
      "entropy": 5.698626947402954,
      "epoch": 2.8716884143930406,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004183823262778955,
      "loss": 4.9814,
      "mean_token_accuracy": 0.2319273978471756,
      "num_tokens": 66524042.0,
      "step": 29050
    },
    {
      "entropy": 5.790049457550049,
      "epoch": 2.8721826809015423,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004183551281768586,
      "loss": 5.0373,
      "mean_token_accuracy": 0.21906707286834717,
      "num_tokens": 66534111.0,
      "step": 29055
    },
    {
      "entropy": 5.753539371490478,
      "epoch": 2.8726769474100435,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041832792654926333,
      "loss": 5.023,
      "mean_token_accuracy": 0.2183144897222519,
      "num_tokens": 66545605.0,
      "step": 29060
    },
    {
      "entropy": 5.665373802185059,
      "epoch": 2.873171213918545,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041830072139577886,
      "loss": 4.9891,
      "mean_token_accuracy": 0.22244725823402406,
      "num_tokens": 66556579.0,
      "step": 29065
    },
    {
      "entropy": 5.701911211013794,
      "epoch": 2.8736654804270465,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041827351271707435,
      "loss": 4.9961,
      "mean_token_accuracy": 0.21557383239269257,
      "num_tokens": 66570262.0,
      "step": 29070
    },
    {
      "entropy": 5.7843729019165036,
      "epoch": 2.8741597469355478,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041824630051381917,
      "loss": 5.0551,
      "mean_token_accuracy": 0.20418720990419387,
      "num_tokens": 66581159.0,
      "step": 29075
    },
    {
      "entropy": 5.726523542404175,
      "epoch": 2.874654013444049,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00041821908478668283,
      "loss": 4.9726,
      "mean_token_accuracy": 0.22645521759986878,
      "num_tokens": 66591641.0,
      "step": 29080
    },
    {
      "entropy": 5.6469262599945065,
      "epoch": 2.8751482799525503,
      "grad_norm": 1.25,
      "learning_rate": 0.00041819186553633476,
      "loss": 5.0307,
      "mean_token_accuracy": 0.21355478018522261,
      "num_tokens": 66603737.0,
      "step": 29085
    },
    {
      "entropy": 5.717840766906738,
      "epoch": 2.8756425464610516,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041816464276344463,
      "loss": 5.0134,
      "mean_token_accuracy": 0.21755453646183015,
      "num_tokens": 66615874.0,
      "step": 29090
    },
    {
      "entropy": 5.7800170421600345,
      "epoch": 2.8761368129695533,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041813741646868204,
      "loss": 5.061,
      "mean_token_accuracy": 0.21498048901557923,
      "num_tokens": 66627297.0,
      "step": 29095
    },
    {
      "entropy": 5.773064422607422,
      "epoch": 2.8766310794780545,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004181101866527168,
      "loss": 5.0619,
      "mean_token_accuracy": 0.21054091304540634,
      "num_tokens": 66638342.0,
      "step": 29100
    },
    {
      "entropy": 5.734130620956421,
      "epoch": 2.8771253459865562,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004180829533162188,
      "loss": 4.9469,
      "mean_token_accuracy": 0.22385955303907396,
      "num_tokens": 66648599.0,
      "step": 29105
    },
    {
      "entropy": 5.618639898300171,
      "epoch": 2.8776196124950575,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004180557164598579,
      "loss": 4.9273,
      "mean_token_accuracy": 0.23167620301246644,
      "num_tokens": 66660850.0,
      "step": 29110
    },
    {
      "entropy": 5.641179752349854,
      "epoch": 2.8781138790035588,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004180284760843042,
      "loss": 4.9814,
      "mean_token_accuracy": 0.2163483127951622,
      "num_tokens": 66672577.0,
      "step": 29115
    },
    {
      "entropy": 5.793479347229004,
      "epoch": 2.87860814551206,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041800123219022786,
      "loss": 5.0295,
      "mean_token_accuracy": 0.2161676749587059,
      "num_tokens": 66683720.0,
      "step": 29120
    },
    {
      "entropy": 5.791900873184204,
      "epoch": 2.8791024120205613,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004179739847782989,
      "loss": 5.0808,
      "mean_token_accuracy": 0.21394957602024078,
      "num_tokens": 66696492.0,
      "step": 29125
    },
    {
      "entropy": 5.710444974899292,
      "epoch": 2.879596678529063,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004179467338491878,
      "loss": 5.0106,
      "mean_token_accuracy": 0.21962936520576476,
      "num_tokens": 66707203.0,
      "step": 29130
    },
    {
      "entropy": 5.740445709228515,
      "epoch": 2.8800909450375642,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004179194794035649,
      "loss": 5.0131,
      "mean_token_accuracy": 0.21750188171863555,
      "num_tokens": 66717433.0,
      "step": 29135
    },
    {
      "entropy": 5.655262231826782,
      "epoch": 2.8805852115460655,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041789222144210066,
      "loss": 4.8804,
      "mean_token_accuracy": 0.23399192243814468,
      "num_tokens": 66729452.0,
      "step": 29140
    },
    {
      "entropy": 5.670908784866333,
      "epoch": 2.881079478054567,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041786495996546546,
      "loss": 5.0382,
      "mean_token_accuracy": 0.22025166898965837,
      "num_tokens": 66741360.0,
      "step": 29145
    },
    {
      "entropy": 5.706701469421387,
      "epoch": 2.8815737445630685,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004178376949743302,
      "loss": 4.9795,
      "mean_token_accuracy": 0.22138807773590088,
      "num_tokens": 66752585.0,
      "step": 29150
    },
    {
      "entropy": 5.707409954071045,
      "epoch": 2.8820680110715697,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004178104264693655,
      "loss": 4.9748,
      "mean_token_accuracy": 0.2266859456896782,
      "num_tokens": 66763898.0,
      "step": 29155
    },
    {
      "entropy": 5.692349195480347,
      "epoch": 2.882562277580071,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000417783154451242,
      "loss": 4.8676,
      "mean_token_accuracy": 0.23227767795324325,
      "num_tokens": 66774272.0,
      "step": 29160
    },
    {
      "entropy": 5.695982503890991,
      "epoch": 2.8830565440885727,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041775587892063095,
      "loss": 4.9596,
      "mean_token_accuracy": 0.21961992383003234,
      "num_tokens": 66784943.0,
      "step": 29165
    },
    {
      "entropy": 5.776537847518921,
      "epoch": 2.883550810597074,
      "grad_norm": 1.09375,
      "learning_rate": 0.000417728599878203,
      "loss": 5.0795,
      "mean_token_accuracy": 0.21163523495197295,
      "num_tokens": 66796449.0,
      "step": 29170
    },
    {
      "entropy": 5.665265464782715,
      "epoch": 2.8840450771055752,
      "grad_norm": 1.125,
      "learning_rate": 0.0004177013173246294,
      "loss": 5.0173,
      "mean_token_accuracy": 0.2173009067773819,
      "num_tokens": 66809936.0,
      "step": 29175
    },
    {
      "entropy": 5.778957271575928,
      "epoch": 2.884539343614077,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004176740312605813,
      "loss": 5.0411,
      "mean_token_accuracy": 0.21382412314414978,
      "num_tokens": 66820846.0,
      "step": 29180
    },
    {
      "entropy": 5.769250297546387,
      "epoch": 2.885033610122578,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004176467416867298,
      "loss": 5.0269,
      "mean_token_accuracy": 0.21041557043790818,
      "num_tokens": 66832919.0,
      "step": 29185
    },
    {
      "entropy": 5.704727125167847,
      "epoch": 2.8855278766310795,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041761944860374637,
      "loss": 4.9243,
      "mean_token_accuracy": 0.22857650965452195,
      "num_tokens": 66844944.0,
      "step": 29190
    },
    {
      "entropy": 5.695947074890137,
      "epoch": 2.8860221431395807,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004175921520123023,
      "loss": 5.0035,
      "mean_token_accuracy": 0.21847164779901504,
      "num_tokens": 66857987.0,
      "step": 29195
    },
    {
      "entropy": 5.6469746112823485,
      "epoch": 2.886516409648082,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004175648519130693,
      "loss": 4.957,
      "mean_token_accuracy": 0.22026453018188477,
      "num_tokens": 66869100.0,
      "step": 29200
    },
    {
      "entropy": 5.670305919647217,
      "epoch": 2.8870106761565837,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041753754830671876,
      "loss": 4.9572,
      "mean_token_accuracy": 0.21612282395362853,
      "num_tokens": 66881168.0,
      "step": 29205
    },
    {
      "entropy": 5.7087959289550785,
      "epoch": 2.887504942665085,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004175102411939224,
      "loss": 5.0211,
      "mean_token_accuracy": 0.21840193271636962,
      "num_tokens": 66894080.0,
      "step": 29210
    },
    {
      "entropy": 5.70404314994812,
      "epoch": 2.8879992091735867,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041748293057535207,
      "loss": 4.9977,
      "mean_token_accuracy": 0.22230230569839476,
      "num_tokens": 66906155.0,
      "step": 29215
    },
    {
      "entropy": 5.706106042861938,
      "epoch": 2.888493475682088,
      "grad_norm": 1.375,
      "learning_rate": 0.00041745561645167953,
      "loss": 4.9379,
      "mean_token_accuracy": 0.2300686612725258,
      "num_tokens": 66915745.0,
      "step": 29220
    },
    {
      "entropy": 5.764196538925171,
      "epoch": 2.888987742190589,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004174282988235767,
      "loss": 5.0162,
      "mean_token_accuracy": 0.21629587113857268,
      "num_tokens": 66927722.0,
      "step": 29225
    },
    {
      "entropy": 5.699118614196777,
      "epoch": 2.8894820086990904,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004174009776917156,
      "loss": 5.0217,
      "mean_token_accuracy": 0.21865702867507936,
      "num_tokens": 66939063.0,
      "step": 29230
    },
    {
      "entropy": 5.698572063446045,
      "epoch": 2.8899762752075917,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004173736530567684,
      "loss": 4.9024,
      "mean_token_accuracy": 0.2304903194308281,
      "num_tokens": 66950596.0,
      "step": 29235
    },
    {
      "entropy": 5.704935121536255,
      "epoch": 2.8904705417160934,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004173463249194072,
      "loss": 4.9592,
      "mean_token_accuracy": 0.21991574317216872,
      "num_tokens": 66961836.0,
      "step": 29240
    },
    {
      "entropy": 5.75808687210083,
      "epoch": 2.8909648082245947,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041731899328030443,
      "loss": 5.1039,
      "mean_token_accuracy": 0.20811564326286316,
      "num_tokens": 66976064.0,
      "step": 29245
    },
    {
      "entropy": 5.659255075454712,
      "epoch": 2.891459074733096,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004172916581401322,
      "loss": 4.8925,
      "mean_token_accuracy": 0.23715227246284484,
      "num_tokens": 66987572.0,
      "step": 29250
    },
    {
      "entropy": 5.6431841373443605,
      "epoch": 2.8919533412415976,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004172643194995632,
      "loss": 4.9195,
      "mean_token_accuracy": 0.23159011602401733,
      "num_tokens": 66998668.0,
      "step": 29255
    },
    {
      "entropy": 5.686891222000122,
      "epoch": 2.892447607750099,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004172369773592699,
      "loss": 5.0316,
      "mean_token_accuracy": 0.2183586537837982,
      "num_tokens": 67010140.0,
      "step": 29260
    },
    {
      "entropy": 5.673285484313965,
      "epoch": 2.8929418742586,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004172096317199248,
      "loss": 4.8691,
      "mean_token_accuracy": 0.23667492270469664,
      "num_tokens": 67021727.0,
      "step": 29265
    },
    {
      "entropy": 5.699222564697266,
      "epoch": 2.8934361407671014,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004171822825822007,
      "loss": 4.9889,
      "mean_token_accuracy": 0.22674315124750138,
      "num_tokens": 67033327.0,
      "step": 29270
    },
    {
      "entropy": 5.742250394821167,
      "epoch": 2.893930407275603,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004171549299467704,
      "loss": 5.0526,
      "mean_token_accuracy": 0.21962883025407792,
      "num_tokens": 67044827.0,
      "step": 29275
    },
    {
      "entropy": 5.725108861923218,
      "epoch": 2.8944246737841044,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004171275738143068,
      "loss": 4.9147,
      "mean_token_accuracy": 0.22060459703207017,
      "num_tokens": 67056042.0,
      "step": 29280
    },
    {
      "entropy": 5.722684574127197,
      "epoch": 2.8949189402926057,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041710021418548283,
      "loss": 4.9968,
      "mean_token_accuracy": 0.22067476361989974,
      "num_tokens": 67068346.0,
      "step": 29285
    },
    {
      "entropy": 5.712063932418824,
      "epoch": 2.8954132068011074,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004170728510609716,
      "loss": 4.9199,
      "mean_token_accuracy": 0.22658058255910873,
      "num_tokens": 67078282.0,
      "step": 29290
    },
    {
      "entropy": 5.693100690841675,
      "epoch": 2.8959074733096086,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041704548444144605,
      "loss": 4.9965,
      "mean_token_accuracy": 0.21355136185884477,
      "num_tokens": 67089958.0,
      "step": 29295
    },
    {
      "entropy": 5.738266849517823,
      "epoch": 2.89640173981811,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041701811432757964,
      "loss": 5.0119,
      "mean_token_accuracy": 0.2174461379647255,
      "num_tokens": 67101672.0,
      "step": 29300
    },
    {
      "entropy": 5.689097547531128,
      "epoch": 2.896896006326611,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004169907407200456,
      "loss": 4.9399,
      "mean_token_accuracy": 0.2195773258805275,
      "num_tokens": 67113334.0,
      "step": 29305
    },
    {
      "entropy": 5.77873125076294,
      "epoch": 2.897390272835113,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041696336361951727,
      "loss": 5.0262,
      "mean_token_accuracy": 0.2111230507493019,
      "num_tokens": 67123880.0,
      "step": 29310
    },
    {
      "entropy": 5.718956518173218,
      "epoch": 2.897884539343614,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004169359830266682,
      "loss": 5.0008,
      "mean_token_accuracy": 0.21950687915086747,
      "num_tokens": 67134966.0,
      "step": 29315
    },
    {
      "entropy": 5.675970983505249,
      "epoch": 2.8983788058521154,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004169085989421719,
      "loss": 4.9608,
      "mean_token_accuracy": 0.22214278429746628,
      "num_tokens": 67145407.0,
      "step": 29320
    },
    {
      "entropy": 5.692808389663696,
      "epoch": 2.898873072360617,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004168812113667021,
      "loss": 4.9868,
      "mean_token_accuracy": 0.2174471840262413,
      "num_tokens": 67156201.0,
      "step": 29325
    },
    {
      "entropy": 5.65477180480957,
      "epoch": 2.8993673388691183,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041685382030093245,
      "loss": 4.9765,
      "mean_token_accuracy": 0.2209431856870651,
      "num_tokens": 67168851.0,
      "step": 29330
    },
    {
      "entropy": 5.586418724060058,
      "epoch": 2.8998616053776196,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041682642574553683,
      "loss": 4.8427,
      "mean_token_accuracy": 0.23947807550430297,
      "num_tokens": 67180559.0,
      "step": 29335
    },
    {
      "entropy": 5.734999465942383,
      "epoch": 2.900355871886121,
      "grad_norm": 1.125,
      "learning_rate": 0.00041679902770118914,
      "loss": 5.0154,
      "mean_token_accuracy": 0.21532484292984008,
      "num_tokens": 67192494.0,
      "step": 29340
    },
    {
      "entropy": 5.705340003967285,
      "epoch": 2.900850138394622,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041677162616856337,
      "loss": 4.9741,
      "mean_token_accuracy": 0.22153246104717256,
      "num_tokens": 67203846.0,
      "step": 29345
    },
    {
      "entropy": 5.73865761756897,
      "epoch": 2.901344404903124,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041674422114833357,
      "loss": 4.9761,
      "mean_token_accuracy": 0.22285375297069548,
      "num_tokens": 67214528.0,
      "step": 29350
    },
    {
      "entropy": 5.745752811431885,
      "epoch": 2.901838671411625,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004167168126411739,
      "loss": 5.035,
      "mean_token_accuracy": 0.21285356879234313,
      "num_tokens": 67225798.0,
      "step": 29355
    },
    {
      "entropy": 5.805123901367187,
      "epoch": 2.902332937920127,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004166894006477587,
      "loss": 5.1185,
      "mean_token_accuracy": 0.21292482018470765,
      "num_tokens": 67237245.0,
      "step": 29360
    },
    {
      "entropy": 5.754734039306641,
      "epoch": 2.902827204428628,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004166619851687623,
      "loss": 5.0175,
      "mean_token_accuracy": 0.21246568262577056,
      "num_tokens": 67249741.0,
      "step": 29365
    },
    {
      "entropy": 5.767901468276977,
      "epoch": 2.9033214709371293,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041663456620485913,
      "loss": 5.0714,
      "mean_token_accuracy": 0.21601364612579346,
      "num_tokens": 67261452.0,
      "step": 29370
    },
    {
      "entropy": 5.728582048416138,
      "epoch": 2.9038157374456306,
      "grad_norm": 1.0,
      "learning_rate": 0.0004166071437567236,
      "loss": 5.0075,
      "mean_token_accuracy": 0.216458822786808,
      "num_tokens": 67273420.0,
      "step": 29375
    },
    {
      "entropy": 5.752809047698975,
      "epoch": 2.904310003954132,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004165797178250303,
      "loss": 5.0556,
      "mean_token_accuracy": 0.21285313069820405,
      "num_tokens": 67284853.0,
      "step": 29380
    },
    {
      "entropy": 5.65129976272583,
      "epoch": 2.9048042704626336,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041655228841045404,
      "loss": 4.872,
      "mean_token_accuracy": 0.23664825558662414,
      "num_tokens": 67295956.0,
      "step": 29385
    },
    {
      "entropy": 5.706582689285279,
      "epoch": 2.905298536971135,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041652485551366955,
      "loss": 5.0078,
      "mean_token_accuracy": 0.2241780012845993,
      "num_tokens": 67307150.0,
      "step": 29390
    },
    {
      "entropy": 5.681396055221557,
      "epoch": 2.905792803479636,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004164974191353516,
      "loss": 4.8797,
      "mean_token_accuracy": 0.22783144414424897,
      "num_tokens": 67318012.0,
      "step": 29395
    },
    {
      "entropy": 5.677178525924683,
      "epoch": 2.906287069988138,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004164699792761753,
      "loss": 4.9724,
      "mean_token_accuracy": 0.22274019867181777,
      "num_tokens": 67331335.0,
      "step": 29400
    },
    {
      "entropy": 5.762203931808472,
      "epoch": 2.906781336496639,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041644253593681543,
      "loss": 5.0354,
      "mean_token_accuracy": 0.2137201175093651,
      "num_tokens": 67342595.0,
      "step": 29405
    },
    {
      "entropy": 5.706224632263184,
      "epoch": 2.9072756030051403,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041641508911794724,
      "loss": 4.9878,
      "mean_token_accuracy": 0.22095876187086105,
      "num_tokens": 67355620.0,
      "step": 29410
    },
    {
      "entropy": 5.722298336029053,
      "epoch": 2.9077698695136416,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000416387638820246,
      "loss": 4.9858,
      "mean_token_accuracy": 0.21738834232091903,
      "num_tokens": 67366600.0,
      "step": 29415
    },
    {
      "entropy": 5.73576250076294,
      "epoch": 2.9082641360221433,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041636018504438683,
      "loss": 5.015,
      "mean_token_accuracy": 0.21352139115333557,
      "num_tokens": 67377122.0,
      "step": 29420
    },
    {
      "entropy": 5.698271083831787,
      "epoch": 2.9087584025306445,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004163327277910452,
      "loss": 5.0549,
      "mean_token_accuracy": 0.21910945177078248,
      "num_tokens": 67387572.0,
      "step": 29425
    },
    {
      "entropy": 5.680827045440674,
      "epoch": 2.909252669039146,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041630526706089657,
      "loss": 4.9124,
      "mean_token_accuracy": 0.2211971253156662,
      "num_tokens": 67399608.0,
      "step": 29430
    },
    {
      "entropy": 5.772811841964722,
      "epoch": 2.9097469355476475,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041627780285461645,
      "loss": 5.0569,
      "mean_token_accuracy": 0.20994145572185516,
      "num_tokens": 67410144.0,
      "step": 29435
    },
    {
      "entropy": 5.725733041763306,
      "epoch": 2.910241202056149,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041625033517288045,
      "loss": 4.9767,
      "mean_token_accuracy": 0.2218368634581566,
      "num_tokens": 67422777.0,
      "step": 29440
    },
    {
      "entropy": 5.739725732803345,
      "epoch": 2.91073546856465,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004162228640163643,
      "loss": 5.0708,
      "mean_token_accuracy": 0.21369231045246123,
      "num_tokens": 67434226.0,
      "step": 29445
    },
    {
      "entropy": 5.803880596160889,
      "epoch": 2.9112297350731513,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004161953893857438,
      "loss": 5.0736,
      "mean_token_accuracy": 0.22120934575796128,
      "num_tokens": 67445316.0,
      "step": 29450
    },
    {
      "entropy": 5.765415287017822,
      "epoch": 2.9117240015816526,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004161679112816948,
      "loss": 5.0449,
      "mean_token_accuracy": 0.21660278737545013,
      "num_tokens": 67456973.0,
      "step": 29455
    },
    {
      "entropy": 5.66401891708374,
      "epoch": 2.9122182680901543,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041614042970489326,
      "loss": 4.8786,
      "mean_token_accuracy": 0.2321219265460968,
      "num_tokens": 67468134.0,
      "step": 29460
    },
    {
      "entropy": 5.735815334320068,
      "epoch": 2.9127125345986555,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041611294465601525,
      "loss": 5.062,
      "mean_token_accuracy": 0.20952952057123184,
      "num_tokens": 67479973.0,
      "step": 29465
    },
    {
      "entropy": 5.732098054885864,
      "epoch": 2.9132068011071572,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041608545613573694,
      "loss": 5.0028,
      "mean_token_accuracy": 0.2230370044708252,
      "num_tokens": 67491469.0,
      "step": 29470
    },
    {
      "entropy": 5.724022054672242,
      "epoch": 2.9137010676156585,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004160579641447345,
      "loss": 5.0854,
      "mean_token_accuracy": 0.21310922652482986,
      "num_tokens": 67502911.0,
      "step": 29475
    },
    {
      "entropy": 5.820222473144531,
      "epoch": 2.9141953341241598,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041603046868368427,
      "loss": 5.1069,
      "mean_token_accuracy": 0.2115452527999878,
      "num_tokens": 67513958.0,
      "step": 29480
    },
    {
      "entropy": 5.76347713470459,
      "epoch": 2.914689600632661,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041600296975326267,
      "loss": 5.0072,
      "mean_token_accuracy": 0.22098410427570342,
      "num_tokens": 67524894.0,
      "step": 29485
    },
    {
      "entropy": 5.733859825134277,
      "epoch": 2.9151838671411623,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004159754673541461,
      "loss": 4.9302,
      "mean_token_accuracy": 0.2239797070622444,
      "num_tokens": 67536256.0,
      "step": 29490
    },
    {
      "entropy": 5.690882682800293,
      "epoch": 2.915678133649664,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041594796148701116,
      "loss": 4.9735,
      "mean_token_accuracy": 0.22068268656730652,
      "num_tokens": 67548064.0,
      "step": 29495
    },
    {
      "entropy": 5.733793497085571,
      "epoch": 2.9161724001581653,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004159204521525345,
      "loss": 5.002,
      "mean_token_accuracy": 0.21627732664346694,
      "num_tokens": 67559530.0,
      "step": 29500
    },
    {
      "entropy": 5.787374687194824,
      "epoch": 2.9166666666666665,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004158929393513929,
      "loss": 5.0942,
      "mean_token_accuracy": 0.2177201732993126,
      "num_tokens": 67571633.0,
      "step": 29505
    },
    {
      "entropy": 5.727038717269897,
      "epoch": 2.9171609331751682,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041586542308426306,
      "loss": 4.9668,
      "mean_token_accuracy": 0.21750555634498597,
      "num_tokens": 67583778.0,
      "step": 29510
    },
    {
      "entropy": 5.6886971950531,
      "epoch": 2.9176551996836695,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041583790335182196,
      "loss": 4.9204,
      "mean_token_accuracy": 0.2265894040465355,
      "num_tokens": 67595803.0,
      "step": 29515
    },
    {
      "entropy": 5.68999171257019,
      "epoch": 2.9181494661921707,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041581038015474673,
      "loss": 4.9927,
      "mean_token_accuracy": 0.21940671950578688,
      "num_tokens": 67606030.0,
      "step": 29520
    },
    {
      "entropy": 5.773620367050171,
      "epoch": 2.918643732700672,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041578285349371414,
      "loss": 5.0966,
      "mean_token_accuracy": 0.21141974925994872,
      "num_tokens": 67617632.0,
      "step": 29525
    },
    {
      "entropy": 5.768137788772583,
      "epoch": 2.9191379992091737,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004157553233694016,
      "loss": 4.9726,
      "mean_token_accuracy": 0.22305114716291427,
      "num_tokens": 67627954.0,
      "step": 29530
    },
    {
      "entropy": 5.709478759765625,
      "epoch": 2.919632265717675,
      "grad_norm": 1.125,
      "learning_rate": 0.0004157277897824862,
      "loss": 4.9754,
      "mean_token_accuracy": 0.2175694853067398,
      "num_tokens": 67639180.0,
      "step": 29535
    },
    {
      "entropy": 5.679262018203735,
      "epoch": 2.9201265322261762,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004157002527336454,
      "loss": 5.0656,
      "mean_token_accuracy": 0.21360366940498351,
      "num_tokens": 67651834.0,
      "step": 29540
    },
    {
      "entropy": 5.703747892379761,
      "epoch": 2.920620798734678,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041567271222355657,
      "loss": 4.9665,
      "mean_token_accuracy": 0.22111957669258117,
      "num_tokens": 67663263.0,
      "step": 29545
    },
    {
      "entropy": 5.750272464752197,
      "epoch": 2.921115065243179,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004156451682528971,
      "loss": 4.9912,
      "mean_token_accuracy": 0.21966143399477006,
      "num_tokens": 67673970.0,
      "step": 29550
    },
    {
      "entropy": 5.844485092163086,
      "epoch": 2.9216093317516805,
      "grad_norm": 1.25,
      "learning_rate": 0.00041561762082234467,
      "loss": 5.085,
      "mean_token_accuracy": 0.21227838546037675,
      "num_tokens": 67685668.0,
      "step": 29555
    },
    {
      "entropy": 5.7038818359375,
      "epoch": 2.9221035982601817,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000415590069932577,
      "loss": 4.9207,
      "mean_token_accuracy": 0.23047964721918107,
      "num_tokens": 67696169.0,
      "step": 29560
    },
    {
      "entropy": 5.7554121017456055,
      "epoch": 2.9225978647686834,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004155625155842719,
      "loss": 5.0363,
      "mean_token_accuracy": 0.20834168195724487,
      "num_tokens": 67709211.0,
      "step": 29565
    },
    {
      "entropy": 5.6942955493927006,
      "epoch": 2.9230921312771847,
      "grad_norm": 1.171875,
      "learning_rate": 0.000415534957778107,
      "loss": 5.073,
      "mean_token_accuracy": 0.21190536618232728,
      "num_tokens": 67720967.0,
      "step": 29570
    },
    {
      "entropy": 5.7143402099609375,
      "epoch": 2.923586397785686,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004155073965147604,
      "loss": 4.9244,
      "mean_token_accuracy": 0.22879833728075027,
      "num_tokens": 67733099.0,
      "step": 29575
    },
    {
      "entropy": 5.725629758834839,
      "epoch": 2.9240806642941877,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041547983179491,
      "loss": 4.9964,
      "mean_token_accuracy": 0.22062858641147615,
      "num_tokens": 67744739.0,
      "step": 29580
    },
    {
      "entropy": 5.71390118598938,
      "epoch": 2.924574930802689,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041545226361923394,
      "loss": 5.0524,
      "mean_token_accuracy": 0.21765577793121338,
      "num_tokens": 67756944.0,
      "step": 29585
    },
    {
      "entropy": 5.759599828720093,
      "epoch": 2.92506919731119,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041542469198841043,
      "loss": 5.0872,
      "mean_token_accuracy": 0.2140345796942711,
      "num_tokens": 67769018.0,
      "step": 29590
    },
    {
      "entropy": 5.749445295333862,
      "epoch": 2.9255634638196915,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004153971169031177,
      "loss": 4.9802,
      "mean_token_accuracy": 0.22066487073898317,
      "num_tokens": 67780164.0,
      "step": 29595
    },
    {
      "entropy": 5.750535154342652,
      "epoch": 2.9260577303281927,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004153695383640341,
      "loss": 5.0121,
      "mean_token_accuracy": 0.21774566620588304,
      "num_tokens": 67790841.0,
      "step": 29600
    },
    {
      "entropy": 5.729867696762085,
      "epoch": 2.9265519968366944,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041534195637183807,
      "loss": 5.0055,
      "mean_token_accuracy": 0.2202119156718254,
      "num_tokens": 67802397.0,
      "step": 29605
    },
    {
      "entropy": 5.741932487487793,
      "epoch": 2.9270462633451957,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004153143709272082,
      "loss": 5.0324,
      "mean_token_accuracy": 0.22103935480117798,
      "num_tokens": 67814079.0,
      "step": 29610
    },
    {
      "entropy": 5.77334508895874,
      "epoch": 2.927540529853697,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000415286782030823,
      "loss": 5.0724,
      "mean_token_accuracy": 0.2096322998404503,
      "num_tokens": 67825377.0,
      "step": 29615
    },
    {
      "entropy": 5.687475776672363,
      "epoch": 2.9280347963621987,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041525918968336116,
      "loss": 4.9743,
      "mean_token_accuracy": 0.22541534453630446,
      "num_tokens": 67838277.0,
      "step": 29620
    },
    {
      "entropy": 5.764729309082031,
      "epoch": 2.9285290628707,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041523159388550155,
      "loss": 5.0215,
      "mean_token_accuracy": 0.21633100956678392,
      "num_tokens": 67849720.0,
      "step": 29625
    },
    {
      "entropy": 5.731776762008667,
      "epoch": 2.929023329379201,
      "grad_norm": 1.125,
      "learning_rate": 0.00041520399463792297,
      "loss": 4.9545,
      "mean_token_accuracy": 0.22855816781520844,
      "num_tokens": 67860994.0,
      "step": 29630
    },
    {
      "entropy": 5.693594598770142,
      "epoch": 2.9295175958877024,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041517639194130435,
      "loss": 5.0656,
      "mean_token_accuracy": 0.21111013740301132,
      "num_tokens": 67874445.0,
      "step": 29635
    },
    {
      "entropy": 5.7159360408782955,
      "epoch": 2.930011862396204,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041514878579632477,
      "loss": 4.9581,
      "mean_token_accuracy": 0.22365018129348754,
      "num_tokens": 67886789.0,
      "step": 29640
    },
    {
      "entropy": 5.79415192604065,
      "epoch": 2.9305061289047054,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004151211762036632,
      "loss": 5.0244,
      "mean_token_accuracy": 0.22314428687095642,
      "num_tokens": 67898839.0,
      "step": 29645
    },
    {
      "entropy": 5.712095546722412,
      "epoch": 2.9310003954132067,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041509356316399916,
      "loss": 5.0443,
      "mean_token_accuracy": 0.21280768662691116,
      "num_tokens": 67909419.0,
      "step": 29650
    },
    {
      "entropy": 5.690729808807373,
      "epoch": 2.9314946619217084,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041506594667801164,
      "loss": 5.003,
      "mean_token_accuracy": 0.2170872151851654,
      "num_tokens": 67922011.0,
      "step": 29655
    },
    {
      "entropy": 5.807738971710205,
      "epoch": 2.9319889284302096,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041503832674638,
      "loss": 5.1503,
      "mean_token_accuracy": 0.20964345186948777,
      "num_tokens": 67934445.0,
      "step": 29660
    },
    {
      "entropy": 5.801150846481323,
      "epoch": 2.932483194938711,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004150107033697839,
      "loss": 5.0834,
      "mean_token_accuracy": 0.21272073090076446,
      "num_tokens": 67944914.0,
      "step": 29665
    },
    {
      "entropy": 5.711522579193115,
      "epoch": 2.932977461447212,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004149830765489028,
      "loss": 5.0123,
      "mean_token_accuracy": 0.21578700840473175,
      "num_tokens": 67955375.0,
      "step": 29670
    },
    {
      "entropy": 5.7317736625671385,
      "epoch": 2.933471727955714,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041495544628441627,
      "loss": 4.9973,
      "mean_token_accuracy": 0.2219604343175888,
      "num_tokens": 67968453.0,
      "step": 29675
    },
    {
      "entropy": 5.775078678131104,
      "epoch": 2.933965994464215,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041492781257700405,
      "loss": 5.0291,
      "mean_token_accuracy": 0.2121727705001831,
      "num_tokens": 67979099.0,
      "step": 29680
    },
    {
      "entropy": 5.756981468200683,
      "epoch": 2.9344602609727164,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041490017542734594,
      "loss": 4.9867,
      "mean_token_accuracy": 0.2151431143283844,
      "num_tokens": 67989399.0,
      "step": 29685
    },
    {
      "entropy": 5.686401987075806,
      "epoch": 2.934954527481218,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041487253483612173,
      "loss": 4.9161,
      "mean_token_accuracy": 0.2286037415266037,
      "num_tokens": 68001255.0,
      "step": 29690
    },
    {
      "entropy": 5.721219635009765,
      "epoch": 2.9354487939897194,
      "grad_norm": 1.125,
      "learning_rate": 0.0004148448908040116,
      "loss": 5.0214,
      "mean_token_accuracy": 0.21575491577386857,
      "num_tokens": 68012844.0,
      "step": 29695
    },
    {
      "entropy": 5.660803365707397,
      "epoch": 2.9359430604982206,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004148172433316954,
      "loss": 5.0028,
      "mean_token_accuracy": 0.21902500689029694,
      "num_tokens": 68024986.0,
      "step": 29700
    },
    {
      "entropy": 5.722166681289673,
      "epoch": 2.936437327006722,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004147895924198533,
      "loss": 4.9772,
      "mean_token_accuracy": 0.2258203700184822,
      "num_tokens": 68036081.0,
      "step": 29705
    },
    {
      "entropy": 5.746487092971802,
      "epoch": 2.936931593515223,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041476193806916557,
      "loss": 5.0139,
      "mean_token_accuracy": 0.21699535101652145,
      "num_tokens": 68049054.0,
      "step": 29710
    },
    {
      "entropy": 5.756295967102051,
      "epoch": 2.937425860023725,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004147342802803124,
      "loss": 5.0417,
      "mean_token_accuracy": 0.21462388038635255,
      "num_tokens": 68060440.0,
      "step": 29715
    },
    {
      "entropy": 5.696961212158203,
      "epoch": 2.937920126532226,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004147066190539744,
      "loss": 5.0171,
      "mean_token_accuracy": 0.22197043150663376,
      "num_tokens": 68072260.0,
      "step": 29720
    },
    {
      "entropy": 5.7434087753295895,
      "epoch": 2.938414393040728,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004146789543908318,
      "loss": 4.9783,
      "mean_token_accuracy": 0.2180328756570816,
      "num_tokens": 68084737.0,
      "step": 29725
    },
    {
      "entropy": 5.723550844192505,
      "epoch": 2.938908659549229,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004146512862915653,
      "loss": 4.971,
      "mean_token_accuracy": 0.22655564397573472,
      "num_tokens": 68097146.0,
      "step": 29730
    },
    {
      "entropy": 5.797867250442505,
      "epoch": 2.9394029260577303,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004146236147568554,
      "loss": 5.0813,
      "mean_token_accuracy": 0.21524348855018616,
      "num_tokens": 68108158.0,
      "step": 29735
    },
    {
      "entropy": 5.650473928451538,
      "epoch": 2.9398971925662316,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004145959397873831,
      "loss": 4.9514,
      "mean_token_accuracy": 0.22025123238563538,
      "num_tokens": 68120242.0,
      "step": 29740
    },
    {
      "entropy": 5.725341749191284,
      "epoch": 2.940391459074733,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041456826138382884,
      "loss": 5.0677,
      "mean_token_accuracy": 0.20995645523071288,
      "num_tokens": 68131372.0,
      "step": 29745
    },
    {
      "entropy": 5.761231756210327,
      "epoch": 2.9408857255832346,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004145405795468738,
      "loss": 5.0355,
      "mean_token_accuracy": 0.22340237945318223,
      "num_tokens": 68142828.0,
      "step": 29750
    },
    {
      "entropy": 5.7412913799285885,
      "epoch": 2.941379992091736,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004145128942771988,
      "loss": 5.0723,
      "mean_token_accuracy": 0.2066093862056732,
      "num_tokens": 68154398.0,
      "step": 29755
    },
    {
      "entropy": 5.727962589263916,
      "epoch": 2.941874258600237,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000414485205575485,
      "loss": 5.0702,
      "mean_token_accuracy": 0.21236569583415985,
      "num_tokens": 68166645.0,
      "step": 29760
    },
    {
      "entropy": 5.767404222488404,
      "epoch": 2.942368525108739,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041445751344241344,
      "loss": 4.9935,
      "mean_token_accuracy": 0.21915703415870666,
      "num_tokens": 68176800.0,
      "step": 29765
    },
    {
      "entropy": 5.7513994693756105,
      "epoch": 2.94286279161724,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041442981787866547,
      "loss": 5.0248,
      "mean_token_accuracy": 0.22235556244850158,
      "num_tokens": 68187933.0,
      "step": 29770
    },
    {
      "entropy": 5.772892999649048,
      "epoch": 2.9433570581257413,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004144021188849223,
      "loss": 5.1152,
      "mean_token_accuracy": 0.2070513352751732,
      "num_tokens": 68199925.0,
      "step": 29775
    },
    {
      "entropy": 5.669220209121704,
      "epoch": 2.9438513246342426,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004143744164618655,
      "loss": 4.9512,
      "mean_token_accuracy": 0.22645019143819808,
      "num_tokens": 68211313.0,
      "step": 29780
    },
    {
      "entropy": 5.726499795913696,
      "epoch": 2.9443455911427443,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041434671061017634,
      "loss": 5.0455,
      "mean_token_accuracy": 0.2142818436026573,
      "num_tokens": 68223148.0,
      "step": 29785
    },
    {
      "entropy": 5.711609363555908,
      "epoch": 2.9448398576512456,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004143190013305364,
      "loss": 4.9969,
      "mean_token_accuracy": 0.22075555771589278,
      "num_tokens": 68234369.0,
      "step": 29790
    },
    {
      "entropy": 5.775344753265381,
      "epoch": 2.945334124159747,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041429128862362756,
      "loss": 4.9984,
      "mean_token_accuracy": 0.21884723305702208,
      "num_tokens": 68243903.0,
      "step": 29795
    },
    {
      "entropy": 5.754987096786499,
      "epoch": 2.9458283906682485,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004142635724901313,
      "loss": 4.9937,
      "mean_token_accuracy": 0.21506194025278091,
      "num_tokens": 68254997.0,
      "step": 29800
    },
    {
      "entropy": 5.646459341049194,
      "epoch": 2.94632265717675,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004142358529307296,
      "loss": 4.9432,
      "mean_token_accuracy": 0.22475767284631729,
      "num_tokens": 68266203.0,
      "step": 29805
    },
    {
      "entropy": 5.760909748077393,
      "epoch": 2.946816923685251,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004142081299461044,
      "loss": 4.9987,
      "mean_token_accuracy": 0.22593790292739868,
      "num_tokens": 68275930.0,
      "step": 29810
    },
    {
      "entropy": 5.760689353942871,
      "epoch": 2.9473111901937523,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004141804035369375,
      "loss": 5.0057,
      "mean_token_accuracy": 0.22065238207578658,
      "num_tokens": 68287468.0,
      "step": 29815
    },
    {
      "entropy": 5.702890872955322,
      "epoch": 2.9478054567022536,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004141526737039111,
      "loss": 4.9535,
      "mean_token_accuracy": 0.2214091032743454,
      "num_tokens": 68299151.0,
      "step": 29820
    },
    {
      "entropy": 5.7437779903411865,
      "epoch": 2.9482997232107553,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004141249404477074,
      "loss": 5.0733,
      "mean_token_accuracy": 0.20754510313272476,
      "num_tokens": 68310874.0,
      "step": 29825
    },
    {
      "entropy": 5.758662700653076,
      "epoch": 2.9487939897192565,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004140972037690085,
      "loss": 4.9881,
      "mean_token_accuracy": 0.2190428927540779,
      "num_tokens": 68323449.0,
      "step": 29830
    },
    {
      "entropy": 5.703635358810425,
      "epoch": 2.9492882562277583,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041406946366849686,
      "loss": 5.0106,
      "mean_token_accuracy": 0.22204073369503022,
      "num_tokens": 68334799.0,
      "step": 29835
    },
    {
      "entropy": 5.650119876861572,
      "epoch": 2.9497825227362595,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041404172014685483,
      "loss": 4.9752,
      "mean_token_accuracy": 0.22216772437095642,
      "num_tokens": 68348019.0,
      "step": 29840
    },
    {
      "entropy": 5.835838270187378,
      "epoch": 2.9502767892447608,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004140139732047649,
      "loss": 5.0325,
      "mean_token_accuracy": 0.21513766646385193,
      "num_tokens": 68360969.0,
      "step": 29845
    },
    {
      "entropy": 5.781921434402466,
      "epoch": 2.950771055753262,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004139862228429097,
      "loss": 5.0348,
      "mean_token_accuracy": 0.2189636528491974,
      "num_tokens": 68373245.0,
      "step": 29850
    },
    {
      "entropy": 5.694391107559204,
      "epoch": 2.9512653222617633,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041395846906197174,
      "loss": 4.9618,
      "mean_token_accuracy": 0.22050405740737916,
      "num_tokens": 68385278.0,
      "step": 29855
    },
    {
      "entropy": 5.716791868209839,
      "epoch": 2.951759588770265,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041393071186263396,
      "loss": 4.976,
      "mean_token_accuracy": 0.22127410620450974,
      "num_tokens": 68396624.0,
      "step": 29860
    },
    {
      "entropy": 5.715279388427734,
      "epoch": 2.9522538552787663,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004139029512455791,
      "loss": 4.9748,
      "mean_token_accuracy": 0.21673957705497743,
      "num_tokens": 68408656.0,
      "step": 29865
    },
    {
      "entropy": 5.710910654067993,
      "epoch": 2.9527481217872675,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004138751872114902,
      "loss": 4.9287,
      "mean_token_accuracy": 0.2239289700984955,
      "num_tokens": 68421150.0,
      "step": 29870
    },
    {
      "entropy": 5.795772886276245,
      "epoch": 2.9532423882957692,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041384741976105004,
      "loss": 5.0193,
      "mean_token_accuracy": 0.2085531085729599,
      "num_tokens": 68432446.0,
      "step": 29875
    },
    {
      "entropy": 5.724555015563965,
      "epoch": 2.9537366548042705,
      "grad_norm": 1.125,
      "learning_rate": 0.00041381964889494185,
      "loss": 5.0491,
      "mean_token_accuracy": 0.2138671189546585,
      "num_tokens": 68443508.0,
      "step": 29880
    },
    {
      "entropy": 5.7088744163513185,
      "epoch": 2.9542309213127718,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004137918746138487,
      "loss": 5.0235,
      "mean_token_accuracy": 0.21643654257059097,
      "num_tokens": 68454425.0,
      "step": 29885
    },
    {
      "entropy": 5.717440795898438,
      "epoch": 2.954725187821273,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041376409691845395,
      "loss": 5.0,
      "mean_token_accuracy": 0.2169597178697586,
      "num_tokens": 68467759.0,
      "step": 29890
    },
    {
      "entropy": 5.738907146453857,
      "epoch": 2.9552194543297747,
      "grad_norm": 1.328125,
      "learning_rate": 0.00041373631580944093,
      "loss": 4.9924,
      "mean_token_accuracy": 0.22265892326831818,
      "num_tokens": 68478447.0,
      "step": 29895
    },
    {
      "entropy": 5.772379779815674,
      "epoch": 2.955713720838276,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000413708531287493,
      "loss": 5.0894,
      "mean_token_accuracy": 0.21027482002973558,
      "num_tokens": 68489056.0,
      "step": 29900
    },
    {
      "entropy": 5.775525379180908,
      "epoch": 2.9562079873467773,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004136807433532936,
      "loss": 5.0801,
      "mean_token_accuracy": 0.21107235848903655,
      "num_tokens": 68500697.0,
      "step": 29905
    },
    {
      "entropy": 5.756763601303101,
      "epoch": 2.956702253855279,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004136529520075265,
      "loss": 4.9901,
      "mean_token_accuracy": 0.21894254833459853,
      "num_tokens": 68510594.0,
      "step": 29910
    },
    {
      "entropy": 5.708130311965943,
      "epoch": 2.9571965203637802,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041362515725087523,
      "loss": 5.031,
      "mean_token_accuracy": 0.2129152312874794,
      "num_tokens": 68521293.0,
      "step": 29915
    },
    {
      "entropy": 5.676908922195435,
      "epoch": 2.9576907868722815,
      "grad_norm": 1.125,
      "learning_rate": 0.00041359735908402356,
      "loss": 4.9576,
      "mean_token_accuracy": 0.2282138541340828,
      "num_tokens": 68532380.0,
      "step": 29920
    },
    {
      "entropy": 5.721704626083374,
      "epoch": 2.9581850533807827,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041356955750765535,
      "loss": 4.9898,
      "mean_token_accuracy": 0.22226012051105498,
      "num_tokens": 68543510.0,
      "step": 29925
    },
    {
      "entropy": 5.752847003936767,
      "epoch": 2.9586793198892845,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004135417525224546,
      "loss": 5.0811,
      "mean_token_accuracy": 0.2115839973092079,
      "num_tokens": 68556191.0,
      "step": 29930
    },
    {
      "entropy": 5.803689908981323,
      "epoch": 2.9591735863977857,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004135139441291052,
      "loss": 5.1284,
      "mean_token_accuracy": 0.21017372608184814,
      "num_tokens": 68568225.0,
      "step": 29935
    },
    {
      "entropy": 5.830569839477539,
      "epoch": 2.959667852906287,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004134861323282913,
      "loss": 5.0542,
      "mean_token_accuracy": 0.2075226843357086,
      "num_tokens": 68580040.0,
      "step": 29940
    },
    {
      "entropy": 5.724698066711426,
      "epoch": 2.9601621194147887,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041345831712069696,
      "loss": 4.9787,
      "mean_token_accuracy": 0.21417305916547774,
      "num_tokens": 68590338.0,
      "step": 29945
    },
    {
      "entropy": 5.716367387771607,
      "epoch": 2.96065638592329,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041343049850700673,
      "loss": 5.032,
      "mean_token_accuracy": 0.21398171931505203,
      "num_tokens": 68601153.0,
      "step": 29950
    },
    {
      "entropy": 5.741646385192871,
      "epoch": 2.961150652431791,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041340267648790464,
      "loss": 5.0177,
      "mean_token_accuracy": 0.21677428632974624,
      "num_tokens": 68612665.0,
      "step": 29955
    },
    {
      "entropy": 5.732406806945801,
      "epoch": 2.9616449189402925,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004133748510640752,
      "loss": 4.9535,
      "mean_token_accuracy": 0.22279123812913895,
      "num_tokens": 68624229.0,
      "step": 29960
    },
    {
      "entropy": 5.762173461914062,
      "epoch": 2.9621391854487937,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041334702223620306,
      "loss": 5.0416,
      "mean_token_accuracy": 0.21539605855941774,
      "num_tokens": 68635811.0,
      "step": 29965
    },
    {
      "entropy": 5.697637367248535,
      "epoch": 2.9626334519572954,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041331919000497263,
      "loss": 4.9627,
      "mean_token_accuracy": 0.2201569750905037,
      "num_tokens": 68647518.0,
      "step": 29970
    },
    {
      "entropy": 5.791468477249145,
      "epoch": 2.9631277184657967,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041329135437106866,
      "loss": 5.1573,
      "mean_token_accuracy": 0.21058289855718612,
      "num_tokens": 68658739.0,
      "step": 29975
    },
    {
      "entropy": 5.696962356567383,
      "epoch": 2.9636219849742984,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041326351533517604,
      "loss": 4.9507,
      "mean_token_accuracy": 0.23010381013154985,
      "num_tokens": 68670317.0,
      "step": 29980
    },
    {
      "entropy": 5.692986488342285,
      "epoch": 2.9641162514827997,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004132356728979794,
      "loss": 4.9767,
      "mean_token_accuracy": 0.22743176370859147,
      "num_tokens": 68681584.0,
      "step": 29985
    },
    {
      "entropy": 5.6973104000091555,
      "epoch": 2.964610517991301,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041320782706016375,
      "loss": 4.9477,
      "mean_token_accuracy": 0.2273889109492302,
      "num_tokens": 68694296.0,
      "step": 29990
    },
    {
      "entropy": 5.626551961898803,
      "epoch": 2.965104784499802,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004131799778224142,
      "loss": 4.8942,
      "mean_token_accuracy": 0.22775945514440538,
      "num_tokens": 68705281.0,
      "step": 29995
    },
    {
      "entropy": 5.673750972747802,
      "epoch": 2.9655990510083035,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041315212518541565,
      "loss": 4.8712,
      "mean_token_accuracy": 0.2273642435669899,
      "num_tokens": 68716146.0,
      "step": 30000
    },
    {
      "epoch": 2.9655990510083035,
      "eval_entropy": 5.471667922232277,
      "eval_loss": 5.042595863342285,
      "eval_mean_token_accuracy": 0.2259625170283652,
      "eval_num_tokens": 68716146.0,
      "eval_runtime": 20.5162,
      "eval_samples_per_second": 1584.749,
      "eval_steps_per_second": 198.136,
      "step": 30000
    },
    {
      "entropy": 5.669720983505249,
      "epoch": 2.966093317516805,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004131242691498534,
      "loss": 4.9859,
      "mean_token_accuracy": 0.2254968464374542,
      "num_tokens": 68728812.0,
      "step": 30005
    },
    {
      "entropy": 5.704223203659057,
      "epoch": 2.9665875840253064,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041309640971641276,
      "loss": 5.0302,
      "mean_token_accuracy": 0.2114323526620865,
      "num_tokens": 68741283.0,
      "step": 30010
    },
    {
      "entropy": 5.733886480331421,
      "epoch": 2.9670818505338077,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000413068546885779,
      "loss": 4.936,
      "mean_token_accuracy": 0.22926866114139557,
      "num_tokens": 68752701.0,
      "step": 30015
    },
    {
      "entropy": 5.7278026103973385,
      "epoch": 2.9675761170423094,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004130406806586375,
      "loss": 4.9257,
      "mean_token_accuracy": 0.23022190034389495,
      "num_tokens": 68763133.0,
      "step": 30020
    },
    {
      "entropy": 5.692061138153076,
      "epoch": 2.9680703835508107,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000413012811035674,
      "loss": 4.973,
      "mean_token_accuracy": 0.2238582506775856,
      "num_tokens": 68774692.0,
      "step": 30025
    },
    {
      "entropy": 5.715055847167969,
      "epoch": 2.968564650059312,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004129849380175738,
      "loss": 5.0143,
      "mean_token_accuracy": 0.21372091323137282,
      "num_tokens": 68786233.0,
      "step": 30030
    },
    {
      "entropy": 5.6613853931427,
      "epoch": 2.969058916567813,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004129570616050227,
      "loss": 4.9752,
      "mean_token_accuracy": 0.21903624832630159,
      "num_tokens": 68799674.0,
      "step": 30035
    },
    {
      "entropy": 5.74995789527893,
      "epoch": 2.969553183076315,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041292918179870665,
      "loss": 5.0269,
      "mean_token_accuracy": 0.21365875899791717,
      "num_tokens": 68811163.0,
      "step": 30040
    },
    {
      "entropy": 5.763164901733399,
      "epoch": 2.970047449584816,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004129012985993112,
      "loss": 4.9906,
      "mean_token_accuracy": 0.21980062872171402,
      "num_tokens": 68822781.0,
      "step": 30045
    },
    {
      "entropy": 5.722312927246094,
      "epoch": 2.9705417160933174,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041287341200752235,
      "loss": 5.0219,
      "mean_token_accuracy": 0.22174839079380035,
      "num_tokens": 68834908.0,
      "step": 30050
    },
    {
      "entropy": 5.649223279953003,
      "epoch": 2.971035982601819,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041284552202402634,
      "loss": 4.8753,
      "mean_token_accuracy": 0.2298997536301613,
      "num_tokens": 68844253.0,
      "step": 30055
    },
    {
      "entropy": 5.666108131408691,
      "epoch": 2.9715302491103204,
      "grad_norm": 1.140625,
      "learning_rate": 0.000412817628649509,
      "loss": 4.9491,
      "mean_token_accuracy": 0.22055847495794295,
      "num_tokens": 68856132.0,
      "step": 30060
    },
    {
      "entropy": 5.740057849884034,
      "epoch": 2.9720245156188216,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041278973188465655,
      "loss": 5.0096,
      "mean_token_accuracy": 0.2179216966032982,
      "num_tokens": 68867536.0,
      "step": 30065
    },
    {
      "entropy": 5.662638092041016,
      "epoch": 2.972518782127323,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004127618317301555,
      "loss": 4.9999,
      "mean_token_accuracy": 0.22032577097415923,
      "num_tokens": 68880544.0,
      "step": 30070
    },
    {
      "entropy": 5.70892014503479,
      "epoch": 2.973013048635824,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041273392818669193,
      "loss": 5.0849,
      "mean_token_accuracy": 0.21786096394062043,
      "num_tokens": 68890755.0,
      "step": 30075
    },
    {
      "entropy": 5.789656496047973,
      "epoch": 2.973507315144326,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041270602125495227,
      "loss": 5.0741,
      "mean_token_accuracy": 0.20461636632680893,
      "num_tokens": 68902280.0,
      "step": 30080
    },
    {
      "entropy": 5.8169135570526125,
      "epoch": 2.974001581652827,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041267811093562325,
      "loss": 5.0686,
      "mean_token_accuracy": 0.21582219302654265,
      "num_tokens": 68913674.0,
      "step": 30085
    },
    {
      "entropy": 5.787605285644531,
      "epoch": 2.974495848161329,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041265019722939127,
      "loss": 5.0144,
      "mean_token_accuracy": 0.22127629220485687,
      "num_tokens": 68925129.0,
      "step": 30090
    },
    {
      "entropy": 5.703379726409912,
      "epoch": 2.97499011466983,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041262228013694307,
      "loss": 4.9295,
      "mean_token_accuracy": 0.22230359315872192,
      "num_tokens": 68935157.0,
      "step": 30095
    },
    {
      "entropy": 5.6842700958251955,
      "epoch": 2.9754843811783314,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041259435965896555,
      "loss": 4.9811,
      "mean_token_accuracy": 0.2140813171863556,
      "num_tokens": 68946752.0,
      "step": 30100
    },
    {
      "entropy": 5.741587734222412,
      "epoch": 2.9759786476868326,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004125664357961453,
      "loss": 4.9725,
      "mean_token_accuracy": 0.2192420557141304,
      "num_tokens": 68958078.0,
      "step": 30105
    },
    {
      "entropy": 5.778346347808838,
      "epoch": 2.976472914195334,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041253850854916946,
      "loss": 5.0358,
      "mean_token_accuracy": 0.21373041868209838,
      "num_tokens": 68970384.0,
      "step": 30110
    },
    {
      "entropy": 5.71354489326477,
      "epoch": 2.9769671807038356,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041251057791872497,
      "loss": 4.9784,
      "mean_token_accuracy": 0.22338918894529342,
      "num_tokens": 68983209.0,
      "step": 30115
    },
    {
      "entropy": 5.800860261917114,
      "epoch": 2.977461447212337,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041248264390549884,
      "loss": 5.1121,
      "mean_token_accuracy": 0.2100605845451355,
      "num_tokens": 68994710.0,
      "step": 30120
    },
    {
      "entropy": 5.755162239074707,
      "epoch": 2.977955713720838,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041245470651017846,
      "loss": 5.0564,
      "mean_token_accuracy": 0.21620704382658004,
      "num_tokens": 69005457.0,
      "step": 30125
    },
    {
      "entropy": 5.732906341552734,
      "epoch": 2.97844998022934,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041242676573345096,
      "loss": 4.9446,
      "mean_token_accuracy": 0.22765863239765166,
      "num_tokens": 69018379.0,
      "step": 30130
    },
    {
      "entropy": 5.778547525405884,
      "epoch": 2.978944246737841,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004123988215760036,
      "loss": 5.0539,
      "mean_token_accuracy": 0.20956070423126222,
      "num_tokens": 69029608.0,
      "step": 30135
    },
    {
      "entropy": 5.732872104644775,
      "epoch": 2.9794385132463423,
      "grad_norm": 1.234375,
      "learning_rate": 0.000412370874038524,
      "loss": 5.0022,
      "mean_token_accuracy": 0.2188788741827011,
      "num_tokens": 69040341.0,
      "step": 30140
    },
    {
      "entropy": 5.725385618209839,
      "epoch": 2.9799327797548436,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041234292312169964,
      "loss": 5.0606,
      "mean_token_accuracy": 0.21333449631929396,
      "num_tokens": 69050814.0,
      "step": 30145
    },
    {
      "entropy": 5.79220061302185,
      "epoch": 2.9804270462633453,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004123149688262179,
      "loss": 5.0961,
      "mean_token_accuracy": 0.2071923092007637,
      "num_tokens": 69063087.0,
      "step": 30150
    },
    {
      "entropy": 5.7737102031707765,
      "epoch": 2.9809213127718466,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004122870111527668,
      "loss": 5.0789,
      "mean_token_accuracy": 0.21390679329633713,
      "num_tokens": 69075342.0,
      "step": 30155
    },
    {
      "entropy": 5.778729104995728,
      "epoch": 2.981415579280348,
      "grad_norm": 1.125,
      "learning_rate": 0.00041225905010203385,
      "loss": 5.0641,
      "mean_token_accuracy": 0.20892024040222168,
      "num_tokens": 69087616.0,
      "step": 30160
    },
    {
      "entropy": 5.820131778717041,
      "epoch": 2.9819098457888495,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041223108567470696,
      "loss": 5.008,
      "mean_token_accuracy": 0.21184780299663544,
      "num_tokens": 69099816.0,
      "step": 30165
    },
    {
      "entropy": 5.6977824687957765,
      "epoch": 2.982404112297351,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041220311787147407,
      "loss": 4.87,
      "mean_token_accuracy": 0.2336866959929466,
      "num_tokens": 69111806.0,
      "step": 30170
    },
    {
      "entropy": 5.706397294998169,
      "epoch": 2.982898378805852,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004121751466930232,
      "loss": 5.0203,
      "mean_token_accuracy": 0.21648335307836533,
      "num_tokens": 69123479.0,
      "step": 30175
    },
    {
      "entropy": 5.810094738006592,
      "epoch": 2.9833926453143533,
      "grad_norm": 1.125,
      "learning_rate": 0.0004121471721400424,
      "loss": 5.0997,
      "mean_token_accuracy": 0.20706073343753814,
      "num_tokens": 69134541.0,
      "step": 30180
    },
    {
      "entropy": 5.766567373275757,
      "epoch": 2.983886911822855,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041211919421322,
      "loss": 4.9664,
      "mean_token_accuracy": 0.2215376913547516,
      "num_tokens": 69144765.0,
      "step": 30185
    },
    {
      "entropy": 5.788903522491455,
      "epoch": 2.9843811783313563,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004120912129132441,
      "loss": 5.0182,
      "mean_token_accuracy": 0.21651130467653273,
      "num_tokens": 69156761.0,
      "step": 30190
    },
    {
      "entropy": 5.713674116134643,
      "epoch": 2.9848754448398576,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004120632282408031,
      "loss": 4.9835,
      "mean_token_accuracy": 0.21852971613407135,
      "num_tokens": 69167968.0,
      "step": 30195
    },
    {
      "entropy": 5.741098546981812,
      "epoch": 2.9853697113483593,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004120352401965854,
      "loss": 5.024,
      "mean_token_accuracy": 0.22322219759225845,
      "num_tokens": 69178356.0,
      "step": 30200
    },
    {
      "entropy": 5.773768615722656,
      "epoch": 2.9858639778568605,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041200724878127953,
      "loss": 4.944,
      "mean_token_accuracy": 0.2318148359656334,
      "num_tokens": 69188777.0,
      "step": 30205
    },
    {
      "entropy": 5.734551000595093,
      "epoch": 2.986358244365362,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041197925399557407,
      "loss": 4.9571,
      "mean_token_accuracy": 0.21552761197090148,
      "num_tokens": 69199166.0,
      "step": 30210
    },
    {
      "entropy": 5.665743064880371,
      "epoch": 2.986852510873863,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004119512558401578,
      "loss": 4.9346,
      "mean_token_accuracy": 0.22371985614299775,
      "num_tokens": 69211062.0,
      "step": 30215
    },
    {
      "entropy": 5.769170808792114,
      "epoch": 2.9873467773823643,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041192325431571933,
      "loss": 5.0439,
      "mean_token_accuracy": 0.2166271388530731,
      "num_tokens": 69222263.0,
      "step": 30220
    },
    {
      "entropy": 5.779414510726928,
      "epoch": 2.987841043890866,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041189524942294753,
      "loss": 5.0417,
      "mean_token_accuracy": 0.22066823840141297,
      "num_tokens": 69234888.0,
      "step": 30225
    },
    {
      "entropy": 5.783027505874633,
      "epoch": 2.9883353103993673,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004118672411625314,
      "loss": 5.0521,
      "mean_token_accuracy": 0.21986494064331055,
      "num_tokens": 69244618.0,
      "step": 30230
    },
    {
      "entropy": 5.742040395736694,
      "epoch": 2.988829576907869,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041183922953515987,
      "loss": 4.9693,
      "mean_token_accuracy": 0.2201568901538849,
      "num_tokens": 69254687.0,
      "step": 30235
    },
    {
      "entropy": 5.76533055305481,
      "epoch": 2.9893238434163703,
      "grad_norm": 1.109375,
      "learning_rate": 0.000411811214541522,
      "loss": 5.0597,
      "mean_token_accuracy": 0.21636332869529723,
      "num_tokens": 69267029.0,
      "step": 30240
    },
    {
      "entropy": 5.783213186264038,
      "epoch": 2.9898181099248715,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004117831961823072,
      "loss": 5.0108,
      "mean_token_accuracy": 0.22210270315408706,
      "num_tokens": 69278001.0,
      "step": 30245
    },
    {
      "entropy": 5.720000600814819,
      "epoch": 2.9903123764333728,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004117551744582045,
      "loss": 4.97,
      "mean_token_accuracy": 0.22349727749824524,
      "num_tokens": 69290591.0,
      "step": 30250
    },
    {
      "entropy": 5.757104778289795,
      "epoch": 2.990806642941874,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004117271493699032,
      "loss": 5.0966,
      "mean_token_accuracy": 0.20326961874961852,
      "num_tokens": 69301977.0,
      "step": 30255
    },
    {
      "entropy": 5.711636400222778,
      "epoch": 2.9913009094503757,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041169912091809294,
      "loss": 4.9987,
      "mean_token_accuracy": 0.2179024785757065,
      "num_tokens": 69313457.0,
      "step": 30260
    },
    {
      "entropy": 5.728415298461914,
      "epoch": 2.991795175958877,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000411671089103463,
      "loss": 5.0011,
      "mean_token_accuracy": 0.22736626714468003,
      "num_tokens": 69324860.0,
      "step": 30265
    },
    {
      "entropy": 5.774936580657959,
      "epoch": 2.9922894424673783,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004116430539267031,
      "loss": 5.0295,
      "mean_token_accuracy": 0.21638217568397522,
      "num_tokens": 69335609.0,
      "step": 30270
    },
    {
      "entropy": 5.732158231735229,
      "epoch": 2.99278370897588,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041161501538850286,
      "loss": 5.0364,
      "mean_token_accuracy": 0.21488066464662553,
      "num_tokens": 69346952.0,
      "step": 30275
    },
    {
      "entropy": 5.619992256164551,
      "epoch": 2.9932779754843812,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000411586973489552,
      "loss": 4.9046,
      "mean_token_accuracy": 0.22518128603696824,
      "num_tokens": 69357875.0,
      "step": 30280
    },
    {
      "entropy": 5.7201330184936525,
      "epoch": 2.9937722419928825,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041155892823054046,
      "loss": 4.9935,
      "mean_token_accuracy": 0.21628873497247697,
      "num_tokens": 69370371.0,
      "step": 30285
    },
    {
      "entropy": 5.791064262390137,
      "epoch": 2.9942665085013838,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004115308796121582,
      "loss": 5.0754,
      "mean_token_accuracy": 0.2135318323969841,
      "num_tokens": 69381567.0,
      "step": 30290
    },
    {
      "entropy": 5.746912574768066,
      "epoch": 2.9947607750098855,
      "grad_norm": 1.046875,
      "learning_rate": 0.000411502827635095,
      "loss": 4.9866,
      "mean_token_accuracy": 0.2135758727788925,
      "num_tokens": 69393559.0,
      "step": 30295
    },
    {
      "entropy": 5.788520574569702,
      "epoch": 2.9952550415183867,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041147477230004103,
      "loss": 5.0068,
      "mean_token_accuracy": 0.21810808032751083,
      "num_tokens": 69404727.0,
      "step": 30300
    },
    {
      "entropy": 5.70776891708374,
      "epoch": 2.995749308026888,
      "grad_norm": 1.359375,
      "learning_rate": 0.00041144671360768655,
      "loss": 4.9672,
      "mean_token_accuracy": 0.22071568965911864,
      "num_tokens": 69414919.0,
      "step": 30305
    },
    {
      "entropy": 5.727727317810059,
      "epoch": 2.9962435745353897,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004114186515587218,
      "loss": 4.9832,
      "mean_token_accuracy": 0.22871017307043076,
      "num_tokens": 69426362.0,
      "step": 30310
    },
    {
      "entropy": 5.742846918106079,
      "epoch": 2.996737841043891,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041139058615383703,
      "loss": 5.0182,
      "mean_token_accuracy": 0.2199365481734276,
      "num_tokens": 69438641.0,
      "step": 30315
    },
    {
      "entropy": 5.716010522842407,
      "epoch": 2.997232107552392,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004113625173937226,
      "loss": 5.0065,
      "mean_token_accuracy": 0.21939396113157272,
      "num_tokens": 69449237.0,
      "step": 30320
    },
    {
      "entropy": 5.73662052154541,
      "epoch": 2.9977263740608935,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004113344452790691,
      "loss": 4.9886,
      "mean_token_accuracy": 0.21753095388412474,
      "num_tokens": 69460423.0,
      "step": 30325
    },
    {
      "entropy": 5.701431226730347,
      "epoch": 2.9982206405693947,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041130636981056716,
      "loss": 4.8831,
      "mean_token_accuracy": 0.23606034368276596,
      "num_tokens": 69470999.0,
      "step": 30330
    },
    {
      "entropy": 5.818048429489136,
      "epoch": 2.9987149070778965,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004112782909889074,
      "loss": 5.0929,
      "mean_token_accuracy": 0.2087991252541542,
      "num_tokens": 69481517.0,
      "step": 30335
    },
    {
      "entropy": 5.724060297012329,
      "epoch": 2.9992091735863977,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004112502088147804,
      "loss": 5.0474,
      "mean_token_accuracy": 0.21982107311487198,
      "num_tokens": 69493387.0,
      "step": 30340
    },
    {
      "entropy": 5.7181077003479,
      "epoch": 2.9997034400948994,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041122212328887725,
      "loss": 4.9586,
      "mean_token_accuracy": 0.2208357572555542,
      "num_tokens": 69504715.0,
      "step": 30345
    },
    {
      "entropy": 5.776590871810913,
      "epoch": 3.0001977066034007,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041119403441188874,
      "loss": 5.1058,
      "mean_token_accuracy": 0.21024317294359207,
      "num_tokens": 69515558.0,
      "step": 30350
    },
    {
      "entropy": 5.751272296905517,
      "epoch": 3.000691973111902,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004111659421845058,
      "loss": 5.0018,
      "mean_token_accuracy": 0.2176106244325638,
      "num_tokens": 69528195.0,
      "step": 30355
    },
    {
      "entropy": 5.825820446014404,
      "epoch": 3.001186239620403,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041113784660741954,
      "loss": 4.9738,
      "mean_token_accuracy": 0.22079409509897233,
      "num_tokens": 69539565.0,
      "step": 30360
    },
    {
      "entropy": 5.67646017074585,
      "epoch": 3.001680506128905,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004111097476813212,
      "loss": 4.9149,
      "mean_token_accuracy": 0.22978126704692842,
      "num_tokens": 69550827.0,
      "step": 30365
    },
    {
      "entropy": 5.7710155010223385,
      "epoch": 3.002174772637406,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041108164540690196,
      "loss": 5.0288,
      "mean_token_accuracy": 0.2232765629887581,
      "num_tokens": 69564205.0,
      "step": 30370
    },
    {
      "entropy": 5.7783668518066404,
      "epoch": 3.0026690391459074,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041105353978485303,
      "loss": 4.9542,
      "mean_token_accuracy": 0.22511908560991287,
      "num_tokens": 69577353.0,
      "step": 30375
    },
    {
      "entropy": 5.690641641616821,
      "epoch": 3.0031633056544087,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041102543081586603,
      "loss": 4.9159,
      "mean_token_accuracy": 0.23622947186231613,
      "num_tokens": 69588401.0,
      "step": 30380
    },
    {
      "entropy": 5.721901607513428,
      "epoch": 3.0036575721629104,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004109973185006322,
      "loss": 4.9785,
      "mean_token_accuracy": 0.21664253175258635,
      "num_tokens": 69599966.0,
      "step": 30385
    },
    {
      "entropy": 5.7363073348999025,
      "epoch": 3.0041518386714117,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041096920283984335,
      "loss": 4.9843,
      "mean_token_accuracy": 0.2260678917169571,
      "num_tokens": 69611264.0,
      "step": 30390
    },
    {
      "entropy": 5.754953193664551,
      "epoch": 3.004646105179913,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000410941083834191,
      "loss": 4.9665,
      "mean_token_accuracy": 0.22258009165525436,
      "num_tokens": 69621638.0,
      "step": 30395
    },
    {
      "entropy": 5.722536039352417,
      "epoch": 3.005140371688414,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004109129614843669,
      "loss": 4.9037,
      "mean_token_accuracy": 0.23678981512784958,
      "num_tokens": 69633236.0,
      "step": 30400
    },
    {
      "entropy": 5.750876998901367,
      "epoch": 3.005634638196916,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004108848357910628,
      "loss": 4.9396,
      "mean_token_accuracy": 0.22600589990615844,
      "num_tokens": 69644672.0,
      "step": 30405
    },
    {
      "entropy": 5.705188512802124,
      "epoch": 3.006128904705417,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041085670675497065,
      "loss": 4.9554,
      "mean_token_accuracy": 0.21679213792085647,
      "num_tokens": 69655580.0,
      "step": 30410
    },
    {
      "entropy": 5.716627836227417,
      "epoch": 3.0066231712139184,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004108285743767825,
      "loss": 4.9892,
      "mean_token_accuracy": 0.2191089302301407,
      "num_tokens": 69668692.0,
      "step": 30415
    },
    {
      "entropy": 5.792183256149292,
      "epoch": 3.00711743772242,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041080043865719033,
      "loss": 5.0086,
      "mean_token_accuracy": 0.21841350048780442,
      "num_tokens": 69679885.0,
      "step": 30420
    },
    {
      "entropy": 5.737840270996093,
      "epoch": 3.0076117042309214,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004107722995968863,
      "loss": 5.0035,
      "mean_token_accuracy": 0.22233931273221968,
      "num_tokens": 69691395.0,
      "step": 30425
    },
    {
      "entropy": 5.706387090682983,
      "epoch": 3.0081059707394227,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004107441571965627,
      "loss": 4.8508,
      "mean_token_accuracy": 0.2240563616156578,
      "num_tokens": 69701777.0,
      "step": 30430
    },
    {
      "entropy": 5.735692358016967,
      "epoch": 3.008600237247924,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004107160114569117,
      "loss": 4.9424,
      "mean_token_accuracy": 0.22495780140161514,
      "num_tokens": 69712922.0,
      "step": 30435
    },
    {
      "entropy": 5.71577262878418,
      "epoch": 3.0090945037564256,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004106878623786258,
      "loss": 4.9109,
      "mean_token_accuracy": 0.2244071438908577,
      "num_tokens": 69724342.0,
      "step": 30440
    },
    {
      "entropy": 5.675990772247315,
      "epoch": 3.009588770264927,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004106597099623974,
      "loss": 4.9392,
      "mean_token_accuracy": 0.22462610751390458,
      "num_tokens": 69735650.0,
      "step": 30445
    },
    {
      "entropy": 5.708612108230591,
      "epoch": 3.010083036773428,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004106315542089191,
      "loss": 4.9087,
      "mean_token_accuracy": 0.22759891748428346,
      "num_tokens": 69746745.0,
      "step": 30450
    },
    {
      "entropy": 5.757959461212158,
      "epoch": 3.0105773032819294,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004106033951188835,
      "loss": 4.8961,
      "mean_token_accuracy": 0.21829326897859574,
      "num_tokens": 69759315.0,
      "step": 30455
    },
    {
      "entropy": 5.702524137496948,
      "epoch": 3.011071569790431,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041057523269298336,
      "loss": 4.9422,
      "mean_token_accuracy": 0.22083195745944978,
      "num_tokens": 69769883.0,
      "step": 30460
    },
    {
      "entropy": 5.6981323719024655,
      "epoch": 3.0115658362989324,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004105470669319115,
      "loss": 4.9464,
      "mean_token_accuracy": 0.21629173755645753,
      "num_tokens": 69781449.0,
      "step": 30465
    },
    {
      "entropy": 5.644127559661865,
      "epoch": 3.0120601028074336,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004105188978363608,
      "loss": 4.888,
      "mean_token_accuracy": 0.23074255734682084,
      "num_tokens": 69793952.0,
      "step": 30470
    },
    {
      "entropy": 5.721765232086182,
      "epoch": 3.0125543693159353,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004104907254070241,
      "loss": 4.9462,
      "mean_token_accuracy": 0.2289516508579254,
      "num_tokens": 69805705.0,
      "step": 30475
    },
    {
      "entropy": 5.790915441513062,
      "epoch": 3.0130486358244366,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004104625496445946,
      "loss": 5.0247,
      "mean_token_accuracy": 0.21664580553770066,
      "num_tokens": 69817298.0,
      "step": 30480
    },
    {
      "entropy": 5.778224658966065,
      "epoch": 3.013542902332938,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041043437054976524,
      "loss": 5.0177,
      "mean_token_accuracy": 0.21752094477415085,
      "num_tokens": 69829280.0,
      "step": 30485
    },
    {
      "entropy": 5.687246465682984,
      "epoch": 3.014037168841439,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004104061881232295,
      "loss": 4.8783,
      "mean_token_accuracy": 0.23266586661338806,
      "num_tokens": 69841500.0,
      "step": 30490
    },
    {
      "entropy": 5.707006168365479,
      "epoch": 3.014531435349941,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041037800236568044,
      "loss": 4.9361,
      "mean_token_accuracy": 0.22687130868434907,
      "num_tokens": 69852651.0,
      "step": 30495
    },
    {
      "entropy": 5.693068170547486,
      "epoch": 3.015025701858442,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004103498132778116,
      "loss": 4.9396,
      "mean_token_accuracy": 0.22818387299776077,
      "num_tokens": 69864746.0,
      "step": 30500
    },
    {
      "entropy": 5.630860376358032,
      "epoch": 3.0155199683669434,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004103216208603163,
      "loss": 4.8812,
      "mean_token_accuracy": 0.22865380644798278,
      "num_tokens": 69875696.0,
      "step": 30505
    },
    {
      "entropy": 5.702639722824097,
      "epoch": 3.0160142348754446,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004102934251138881,
      "loss": 4.8617,
      "mean_token_accuracy": 0.23673107028007506,
      "num_tokens": 69887363.0,
      "step": 30510
    },
    {
      "entropy": 5.779938697814941,
      "epoch": 3.0165085013839463,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00041026522603922075,
      "loss": 4.9582,
      "mean_token_accuracy": 0.2271068811416626,
      "num_tokens": 69898561.0,
      "step": 30515
    },
    {
      "entropy": 5.723837566375733,
      "epoch": 3.0170027678924476,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004102370236370077,
      "loss": 5.0142,
      "mean_token_accuracy": 0.2168318122625351,
      "num_tokens": 69910297.0,
      "step": 30520
    },
    {
      "entropy": 5.692807197570801,
      "epoch": 3.017497034400949,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041020881790794305,
      "loss": 5.0053,
      "mean_token_accuracy": 0.21866068094968796,
      "num_tokens": 69922435.0,
      "step": 30525
    },
    {
      "entropy": 5.693126630783081,
      "epoch": 3.0179913009094506,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041018060885272047,
      "loss": 4.9317,
      "mean_token_accuracy": 0.23414943516254424,
      "num_tokens": 69935151.0,
      "step": 30530
    },
    {
      "entropy": 5.764499044418335,
      "epoch": 3.018485567417952,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041015239647203386,
      "loss": 4.9479,
      "mean_token_accuracy": 0.22107988744974136,
      "num_tokens": 69945915.0,
      "step": 30535
    },
    {
      "entropy": 5.692897939682007,
      "epoch": 3.018979833926453,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041012418076657737,
      "loss": 4.9177,
      "mean_token_accuracy": 0.2252020761370659,
      "num_tokens": 69957016.0,
      "step": 30540
    },
    {
      "entropy": 5.604632949829101,
      "epoch": 3.0194741004349543,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041009596173704506,
      "loss": 4.855,
      "mean_token_accuracy": 0.23118189871311187,
      "num_tokens": 69967863.0,
      "step": 30545
    },
    {
      "entropy": 5.737838077545166,
      "epoch": 3.019968366943456,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041006773938413114,
      "loss": 5.0247,
      "mean_token_accuracy": 0.21551278680562974,
      "num_tokens": 69978901.0,
      "step": 30550
    },
    {
      "entropy": 5.699930858612061,
      "epoch": 3.0204626334519573,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004100395137085299,
      "loss": 4.9193,
      "mean_token_accuracy": 0.22240727245807648,
      "num_tokens": 69990829.0,
      "step": 30555
    },
    {
      "entropy": 5.745898771286011,
      "epoch": 3.0209568999604586,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004100112847109357,
      "loss": 4.9086,
      "mean_token_accuracy": 0.22435375303030014,
      "num_tokens": 70002130.0,
      "step": 30560
    },
    {
      "entropy": 5.730222940444946,
      "epoch": 3.0214511664689603,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004099830523920429,
      "loss": 4.9499,
      "mean_token_accuracy": 0.22606257796287538,
      "num_tokens": 70012913.0,
      "step": 30565
    },
    {
      "entropy": 5.706943464279175,
      "epoch": 3.0219454329774615,
      "grad_norm": 1.125,
      "learning_rate": 0.00040995481675254616,
      "loss": 4.9487,
      "mean_token_accuracy": 0.22607666850090027,
      "num_tokens": 70024689.0,
      "step": 30570
    },
    {
      "entropy": 5.719606685638428,
      "epoch": 3.022439699485963,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004099265777931398,
      "loss": 4.9423,
      "mean_token_accuracy": 0.22446005046367645,
      "num_tokens": 70036153.0,
      "step": 30575
    },
    {
      "entropy": 5.69336953163147,
      "epoch": 3.022933965994464,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040989833551451885,
      "loss": 4.9637,
      "mean_token_accuracy": 0.220648455619812,
      "num_tokens": 70049190.0,
      "step": 30580
    },
    {
      "entropy": 5.685654592514038,
      "epoch": 3.0234282325029658,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040987008991737786,
      "loss": 4.873,
      "mean_token_accuracy": 0.22690671533346177,
      "num_tokens": 70061391.0,
      "step": 30585
    },
    {
      "entropy": 5.623353338241577,
      "epoch": 3.023922499011467,
      "grad_norm": 1.25,
      "learning_rate": 0.0004098418410024118,
      "loss": 4.8562,
      "mean_token_accuracy": 0.23547518402338027,
      "num_tokens": 70072811.0,
      "step": 30590
    },
    {
      "entropy": 5.666262531280518,
      "epoch": 3.0244167655199683,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040981358877031546,
      "loss": 4.8785,
      "mean_token_accuracy": 0.22842076420783997,
      "num_tokens": 70083776.0,
      "step": 30595
    },
    {
      "entropy": 5.720788764953613,
      "epoch": 3.0249110320284696,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004097853332217839,
      "loss": 4.982,
      "mean_token_accuracy": 0.2208441063761711,
      "num_tokens": 70095720.0,
      "step": 30600
    },
    {
      "entropy": 5.743113279342651,
      "epoch": 3.0254052985369713,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004097570743575123,
      "loss": 4.9791,
      "mean_token_accuracy": 0.21792996525764466,
      "num_tokens": 70106315.0,
      "step": 30605
    },
    {
      "entropy": 5.7307572841644285,
      "epoch": 3.0258995650454725,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040972881217819573,
      "loss": 5.0005,
      "mean_token_accuracy": 0.21487215161323547,
      "num_tokens": 70118018.0,
      "step": 30610
    },
    {
      "entropy": 5.721570348739624,
      "epoch": 3.026393831553974,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004097005466845295,
      "loss": 4.9355,
      "mean_token_accuracy": 0.22338203489780425,
      "num_tokens": 70130543.0,
      "step": 30615
    },
    {
      "entropy": 5.684346342086792,
      "epoch": 3.0268880980624755,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004096722778772089,
      "loss": 4.8827,
      "mean_token_accuracy": 0.2322327584028244,
      "num_tokens": 70142582.0,
      "step": 30620
    },
    {
      "entropy": 5.715225124359131,
      "epoch": 3.0273823645709768,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040964400575692946,
      "loss": 4.9909,
      "mean_token_accuracy": 0.22311019599437715,
      "num_tokens": 70155002.0,
      "step": 30625
    },
    {
      "entropy": 5.730599641799927,
      "epoch": 3.027876631079478,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040961573032438647,
      "loss": 4.9741,
      "mean_token_accuracy": 0.22599612027406693,
      "num_tokens": 70167354.0,
      "step": 30630
    },
    {
      "entropy": 5.731669950485229,
      "epoch": 3.0283708975879793,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004095874515802757,
      "loss": 5.004,
      "mean_token_accuracy": 0.2211572930216789,
      "num_tokens": 70178561.0,
      "step": 30635
    },
    {
      "entropy": 5.733244562149048,
      "epoch": 3.028865164096481,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040955916952529277,
      "loss": 4.9485,
      "mean_token_accuracy": 0.22722513228654861,
      "num_tokens": 70189954.0,
      "step": 30640
    },
    {
      "entropy": 5.779803133010864,
      "epoch": 3.0293594306049823,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040953088416013325,
      "loss": 5.0162,
      "mean_token_accuracy": 0.2126403883099556,
      "num_tokens": 70200595.0,
      "step": 30645
    },
    {
      "entropy": 5.71193265914917,
      "epoch": 3.0298536971134835,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004095025954854933,
      "loss": 4.9687,
      "mean_token_accuracy": 0.22694027721881865,
      "num_tokens": 70211757.0,
      "step": 30650
    },
    {
      "entropy": 5.564568901062012,
      "epoch": 3.0303479636219848,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004094743035020685,
      "loss": 4.8302,
      "mean_token_accuracy": 0.23605324029922486,
      "num_tokens": 70223404.0,
      "step": 30655
    },
    {
      "entropy": 5.778880643844604,
      "epoch": 3.0308422301304865,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040944600821055506,
      "loss": 5.0585,
      "mean_token_accuracy": 0.2120603233575821,
      "num_tokens": 70234022.0,
      "step": 30660
    },
    {
      "entropy": 5.758217716217041,
      "epoch": 3.0313364966389877,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004094177096116489,
      "loss": 4.9662,
      "mean_token_accuracy": 0.22400532066822051,
      "num_tokens": 70245279.0,
      "step": 30665
    },
    {
      "entropy": 5.701322078704834,
      "epoch": 3.031830763147489,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040938940770604624,
      "loss": 4.9467,
      "mean_token_accuracy": 0.22491933703422545,
      "num_tokens": 70256561.0,
      "step": 30670
    },
    {
      "entropy": 5.6540955066680905,
      "epoch": 3.0323250296559907,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040936110249444323,
      "loss": 4.8885,
      "mean_token_accuracy": 0.23269078135490417,
      "num_tokens": 70268535.0,
      "step": 30675
    },
    {
      "entropy": 5.668256855010986,
      "epoch": 3.032819296164492,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004093327939775364,
      "loss": 4.9167,
      "mean_token_accuracy": 0.23604007065296173,
      "num_tokens": 70279681.0,
      "step": 30680
    },
    {
      "entropy": 5.721025037765503,
      "epoch": 3.0333135626729932,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004093044821560219,
      "loss": 4.9056,
      "mean_token_accuracy": 0.228829962015152,
      "num_tokens": 70290796.0,
      "step": 30685
    },
    {
      "entropy": 5.6644690990447994,
      "epoch": 3.0338078291814945,
      "grad_norm": 1.25,
      "learning_rate": 0.00040927616703059636,
      "loss": 4.8918,
      "mean_token_accuracy": 0.22568174302577973,
      "num_tokens": 70302254.0,
      "step": 30690
    },
    {
      "entropy": 5.636654281616211,
      "epoch": 3.034302095689996,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004092478486019562,
      "loss": 4.8565,
      "mean_token_accuracy": 0.2362230360507965,
      "num_tokens": 70313022.0,
      "step": 30695
    },
    {
      "entropy": 5.73512864112854,
      "epoch": 3.0347963621984975,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040921952687079824,
      "loss": 4.9285,
      "mean_token_accuracy": 0.22421761453151703,
      "num_tokens": 70325356.0,
      "step": 30700
    },
    {
      "entropy": 5.7651876449584964,
      "epoch": 3.0352906287069987,
      "grad_norm": 1.046875,
      "learning_rate": 0.000409191201837819,
      "loss": 5.0296,
      "mean_token_accuracy": 0.21503649204969405,
      "num_tokens": 70336622.0,
      "step": 30705
    },
    {
      "entropy": 5.735172748565674,
      "epoch": 3.0357848952155,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040916287350371546,
      "loss": 4.9538,
      "mean_token_accuracy": 0.22120559960603714,
      "num_tokens": 70347799.0,
      "step": 30710
    },
    {
      "entropy": 5.678066825866699,
      "epoch": 3.0362791617240017,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040913454186918436,
      "loss": 4.941,
      "mean_token_accuracy": 0.2258310079574585,
      "num_tokens": 70358918.0,
      "step": 30715
    },
    {
      "entropy": 5.759299659729004,
      "epoch": 3.036773428232503,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040910620693492277,
      "loss": 5.0472,
      "mean_token_accuracy": 0.2138002946972847,
      "num_tokens": 70371853.0,
      "step": 30720
    },
    {
      "entropy": 5.749831008911133,
      "epoch": 3.037267694741004,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004090778687016276,
      "loss": 4.9944,
      "mean_token_accuracy": 0.21727119088172914,
      "num_tokens": 70383696.0,
      "step": 30725
    },
    {
      "entropy": 5.698311758041382,
      "epoch": 3.037761961249506,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040904952716999617,
      "loss": 4.89,
      "mean_token_accuracy": 0.22516567260026932,
      "num_tokens": 70395651.0,
      "step": 30730
    },
    {
      "entropy": 5.747574424743652,
      "epoch": 3.038256227758007,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004090211823407255,
      "loss": 4.9688,
      "mean_token_accuracy": 0.21962679475545882,
      "num_tokens": 70408621.0,
      "step": 30735
    },
    {
      "entropy": 5.638271903991699,
      "epoch": 3.0387504942665085,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000408992834214513,
      "loss": 4.8554,
      "mean_token_accuracy": 0.2269839808344841,
      "num_tokens": 70419627.0,
      "step": 30740
    },
    {
      "entropy": 5.629597234725952,
      "epoch": 3.0392447607750097,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004089644827920559,
      "loss": 4.9226,
      "mean_token_accuracy": 0.22708497196435928,
      "num_tokens": 70431795.0,
      "step": 30745
    },
    {
      "entropy": 5.676336431503296,
      "epoch": 3.0397390272835114,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040893612807405184,
      "loss": 4.8802,
      "mean_token_accuracy": 0.2393552169203758,
      "num_tokens": 70444086.0,
      "step": 30750
    },
    {
      "entropy": 5.764472818374633,
      "epoch": 3.0402332937920127,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004089077700611982,
      "loss": 4.9548,
      "mean_token_accuracy": 0.22660592049360276,
      "num_tokens": 70454844.0,
      "step": 30755
    },
    {
      "entropy": 5.756230783462525,
      "epoch": 3.040727560300514,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040887940875419267,
      "loss": 5.0214,
      "mean_token_accuracy": 0.2166493207216263,
      "num_tokens": 70466003.0,
      "step": 30760
    },
    {
      "entropy": 5.7397068500518795,
      "epoch": 3.041221826809015,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004088510441537329,
      "loss": 4.9237,
      "mean_token_accuracy": 0.2283734545111656,
      "num_tokens": 70476586.0,
      "step": 30765
    },
    {
      "entropy": 5.763432550430298,
      "epoch": 3.041716093317517,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040882267626051664,
      "loss": 5.0275,
      "mean_token_accuracy": 0.21383546739816667,
      "num_tokens": 70488561.0,
      "step": 30770
    },
    {
      "entropy": 5.766893148422241,
      "epoch": 3.042210359826018,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004087943050752419,
      "loss": 4.9621,
      "mean_token_accuracy": 0.2150772914290428,
      "num_tokens": 70499454.0,
      "step": 30775
    },
    {
      "entropy": 5.7595781803131105,
      "epoch": 3.0427046263345194,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004087659305986064,
      "loss": 4.9924,
      "mean_token_accuracy": 0.21298387497663498,
      "num_tokens": 70510880.0,
      "step": 30780
    },
    {
      "entropy": 5.696375942230224,
      "epoch": 3.043198892843021,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040873755283130824,
      "loss": 4.984,
      "mean_token_accuracy": 0.21819486916065217,
      "num_tokens": 70523199.0,
      "step": 30785
    },
    {
      "entropy": 5.691677093505859,
      "epoch": 3.0436931593515224,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004087091717740456,
      "loss": 4.905,
      "mean_token_accuracy": 0.22182126939296723,
      "num_tokens": 70535124.0,
      "step": 30790
    },
    {
      "entropy": 5.7096093654632565,
      "epoch": 3.0441874258600237,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040868078742751655,
      "loss": 4.9031,
      "mean_token_accuracy": 0.2305453196167946,
      "num_tokens": 70547543.0,
      "step": 30795
    },
    {
      "entropy": 5.713780689239502,
      "epoch": 3.044681692368525,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040865239979241946,
      "loss": 4.9243,
      "mean_token_accuracy": 0.23029208034276963,
      "num_tokens": 70559788.0,
      "step": 30800
    },
    {
      "entropy": 5.701771068572998,
      "epoch": 3.0451759588770266,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004086240088694525,
      "loss": 4.9106,
      "mean_token_accuracy": 0.2341471329331398,
      "num_tokens": 70570303.0,
      "step": 30805
    },
    {
      "entropy": 5.648312044143677,
      "epoch": 3.045670225385528,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040859561465931426,
      "loss": 4.8764,
      "mean_token_accuracy": 0.22893944382667542,
      "num_tokens": 70580924.0,
      "step": 30810
    },
    {
      "entropy": 5.708470487594605,
      "epoch": 3.046164491894029,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040856721716270316,
      "loss": 4.9363,
      "mean_token_accuracy": 0.22932678759098052,
      "num_tokens": 70592557.0,
      "step": 30815
    },
    {
      "entropy": 5.611740922927856,
      "epoch": 3.046658758402531,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004085388163803178,
      "loss": 4.8157,
      "mean_token_accuracy": 0.24518696665763856,
      "num_tokens": 70604374.0,
      "step": 30820
    },
    {
      "entropy": 5.727024888992309,
      "epoch": 3.047153024911032,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004085104123128568,
      "loss": 4.9496,
      "mean_token_accuracy": 0.22042208164930344,
      "num_tokens": 70616380.0,
      "step": 30825
    },
    {
      "entropy": 5.714840364456177,
      "epoch": 3.0476472914195334,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040848200496101895,
      "loss": 4.9957,
      "mean_token_accuracy": 0.22494620382785796,
      "num_tokens": 70628297.0,
      "step": 30830
    },
    {
      "entropy": 5.783543109893799,
      "epoch": 3.0481415579280347,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004084535943255031,
      "loss": 4.9733,
      "mean_token_accuracy": 0.21780431419610977,
      "num_tokens": 70639815.0,
      "step": 30835
    },
    {
      "entropy": 5.756053066253662,
      "epoch": 3.0486358244365364,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040842518040700815,
      "loss": 4.9816,
      "mean_token_accuracy": 0.22455343306064607,
      "num_tokens": 70651283.0,
      "step": 30840
    },
    {
      "entropy": 5.683791589736939,
      "epoch": 3.0491300909450376,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000408396763206233,
      "loss": 4.9262,
      "mean_token_accuracy": 0.22923647463321686,
      "num_tokens": 70662465.0,
      "step": 30845
    },
    {
      "entropy": 5.7987658977508545,
      "epoch": 3.049624357453539,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004083683427238768,
      "loss": 5.0357,
      "mean_token_accuracy": 0.21786613166332244,
      "num_tokens": 70674301.0,
      "step": 30850
    },
    {
      "entropy": 5.735969161987304,
      "epoch": 3.05011862396204,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004083399189606387,
      "loss": 4.9514,
      "mean_token_accuracy": 0.2262736976146698,
      "num_tokens": 70685423.0,
      "step": 30855
    },
    {
      "entropy": 5.72317476272583,
      "epoch": 3.050612890470542,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004083114919172179,
      "loss": 4.8616,
      "mean_token_accuracy": 0.2286628469824791,
      "num_tokens": 70697043.0,
      "step": 30860
    },
    {
      "entropy": 5.759106111526489,
      "epoch": 3.051107156979043,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004082830615943137,
      "loss": 4.9866,
      "mean_token_accuracy": 0.22315319031476974,
      "num_tokens": 70708585.0,
      "step": 30865
    },
    {
      "entropy": 5.684482765197754,
      "epoch": 3.0516014234875444,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004082546279926256,
      "loss": 4.9163,
      "mean_token_accuracy": 0.22303294837474824,
      "num_tokens": 70720703.0,
      "step": 30870
    },
    {
      "entropy": 5.750705528259277,
      "epoch": 3.052095689996046,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040822619111285284,
      "loss": 4.9663,
      "mean_token_accuracy": 0.21766726672649384,
      "num_tokens": 70731146.0,
      "step": 30875
    },
    {
      "entropy": 5.708006858825684,
      "epoch": 3.0525899565045473,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040819775095569515,
      "loss": 4.9143,
      "mean_token_accuracy": 0.22630273401737214,
      "num_tokens": 70742393.0,
      "step": 30880
    },
    {
      "entropy": 5.708527946472168,
      "epoch": 3.0530842230130486,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004081693075218522,
      "loss": 4.8347,
      "mean_token_accuracy": 0.23410673439502716,
      "num_tokens": 70753419.0,
      "step": 30885
    },
    {
      "entropy": 5.650593614578247,
      "epoch": 3.05357848952155,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040814086081202364,
      "loss": 4.8916,
      "mean_token_accuracy": 0.2321733519434929,
      "num_tokens": 70765445.0,
      "step": 30890
    },
    {
      "entropy": 5.689175701141357,
      "epoch": 3.0540727560300516,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040811241082690914,
      "loss": 4.9457,
      "mean_token_accuracy": 0.2226109653711319,
      "num_tokens": 70776598.0,
      "step": 30895
    },
    {
      "entropy": 5.7555629253387455,
      "epoch": 3.054567022538553,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004080839575672087,
      "loss": 4.9641,
      "mean_token_accuracy": 0.22415139079093932,
      "num_tokens": 70787982.0,
      "step": 30900
    },
    {
      "entropy": 5.709179830551148,
      "epoch": 3.055061289047054,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040805550103362225,
      "loss": 4.9129,
      "mean_token_accuracy": 0.2233087420463562,
      "num_tokens": 70799080.0,
      "step": 30905
    },
    {
      "entropy": 5.655936431884766,
      "epoch": 3.0555555555555554,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040802704122684994,
      "loss": 4.9157,
      "mean_token_accuracy": 0.22856585830450057,
      "num_tokens": 70809831.0,
      "step": 30910
    },
    {
      "entropy": 5.653885746002198,
      "epoch": 3.056049822064057,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040799857814759167,
      "loss": 5.0101,
      "mean_token_accuracy": 0.22028934955596924,
      "num_tokens": 70821623.0,
      "step": 30915
    },
    {
      "entropy": 5.8056083679199215,
      "epoch": 3.0565440885725583,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004079701117965477,
      "loss": 5.0566,
      "mean_token_accuracy": 0.2126554310321808,
      "num_tokens": 70833875.0,
      "step": 30920
    },
    {
      "entropy": 5.70546932220459,
      "epoch": 3.0570383550810596,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004079416421744185,
      "loss": 4.9442,
      "mean_token_accuracy": 0.22270976305007933,
      "num_tokens": 70845287.0,
      "step": 30925
    },
    {
      "entropy": 5.772080707550049,
      "epoch": 3.0575326215895613,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040791316928190423,
      "loss": 4.9529,
      "mean_token_accuracy": 0.2258085533976555,
      "num_tokens": 70856294.0,
      "step": 30930
    },
    {
      "entropy": 5.684677934646606,
      "epoch": 3.0580268880980626,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00040788469311970525,
      "loss": 4.9689,
      "mean_token_accuracy": 0.22408422976732253,
      "num_tokens": 70867490.0,
      "step": 30935
    },
    {
      "entropy": 5.683466863632202,
      "epoch": 3.058521154606564,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040785621368852234,
      "loss": 4.9624,
      "mean_token_accuracy": 0.22475160360336305,
      "num_tokens": 70877991.0,
      "step": 30940
    },
    {
      "entropy": 5.731028413772583,
      "epoch": 3.059015421115065,
      "grad_norm": 1.25,
      "learning_rate": 0.00040782773098905594,
      "loss": 4.9831,
      "mean_token_accuracy": 0.22464986145496368,
      "num_tokens": 70888439.0,
      "step": 30945
    },
    {
      "entropy": 5.696441507339477,
      "epoch": 3.059509687623567,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004077992450220068,
      "loss": 4.8854,
      "mean_token_accuracy": 0.22986168116331102,
      "num_tokens": 70898331.0,
      "step": 30950
    },
    {
      "entropy": 5.711409044265747,
      "epoch": 3.060003954132068,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004077707557880755,
      "loss": 4.9086,
      "mean_token_accuracy": 0.22759510725736617,
      "num_tokens": 70908221.0,
      "step": 30955
    },
    {
      "entropy": 5.709689378738403,
      "epoch": 3.0604982206405693,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040774226328796315,
      "loss": 4.8946,
      "mean_token_accuracy": 0.2283681094646454,
      "num_tokens": 70919720.0,
      "step": 30960
    },
    {
      "entropy": 5.77891354560852,
      "epoch": 3.0609924871490706,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004077137675223704,
      "loss": 4.9567,
      "mean_token_accuracy": 0.22452180385589598,
      "num_tokens": 70931322.0,
      "step": 30965
    },
    {
      "entropy": 5.675710010528564,
      "epoch": 3.0614867536575723,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004076852684919985,
      "loss": 4.9593,
      "mean_token_accuracy": 0.22506803125143052,
      "num_tokens": 70943124.0,
      "step": 30970
    },
    {
      "entropy": 5.694900941848755,
      "epoch": 3.0619810201660735,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004076567661975484,
      "loss": 4.9241,
      "mean_token_accuracy": 0.22845420837402344,
      "num_tokens": 70954986.0,
      "step": 30975
    },
    {
      "entropy": 5.721801900863648,
      "epoch": 3.062475286674575,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004076282606397211,
      "loss": 4.8757,
      "mean_token_accuracy": 0.226080784201622,
      "num_tokens": 70965500.0,
      "step": 30980
    },
    {
      "entropy": 5.701119232177734,
      "epoch": 3.0629695531830765,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004075997518192182,
      "loss": 4.9116,
      "mean_token_accuracy": 0.2215220585465431,
      "num_tokens": 70976717.0,
      "step": 30985
    },
    {
      "entropy": 5.635010719299316,
      "epoch": 3.0634638196915778,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004075712397367408,
      "loss": 4.9337,
      "mean_token_accuracy": 0.227948597073555,
      "num_tokens": 70987286.0,
      "step": 30990
    },
    {
      "entropy": 5.7938155174255375,
      "epoch": 3.063958086200079,
      "grad_norm": 1.3125,
      "learning_rate": 0.00040754272439299025,
      "loss": 5.0353,
      "mean_token_accuracy": 0.21077050715684892,
      "num_tokens": 70998710.0,
      "step": 30995
    },
    {
      "entropy": 5.73235764503479,
      "epoch": 3.0644523527085803,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040751420578866816,
      "loss": 4.9833,
      "mean_token_accuracy": 0.22203432619571686,
      "num_tokens": 71010725.0,
      "step": 31000
    },
    {
      "entropy": 5.754951620101929,
      "epoch": 3.064946619217082,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004074856839244761,
      "loss": 5.0228,
      "mean_token_accuracy": 0.22167063057422637,
      "num_tokens": 71021723.0,
      "step": 31005
    },
    {
      "entropy": 5.691845464706421,
      "epoch": 3.0654408857255833,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004074571588011156,
      "loss": 4.9123,
      "mean_token_accuracy": 0.2324998214840889,
      "num_tokens": 71032744.0,
      "step": 31010
    },
    {
      "entropy": 5.673231983184815,
      "epoch": 3.0659351522340845,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00040742863041928854,
      "loss": 4.8764,
      "mean_token_accuracy": 0.2278897985816002,
      "num_tokens": 71044296.0,
      "step": 31015
    },
    {
      "entropy": 5.719087553024292,
      "epoch": 3.066429418742586,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040740009877969654,
      "loss": 5.0061,
      "mean_token_accuracy": 0.21975739896297455,
      "num_tokens": 71054636.0,
      "step": 31020
    },
    {
      "entropy": 5.7545655250549315,
      "epoch": 3.0669236852510875,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004073715638830415,
      "loss": 4.9887,
      "mean_token_accuracy": 0.22613947242498397,
      "num_tokens": 71067350.0,
      "step": 31025
    },
    {
      "entropy": 5.762490272521973,
      "epoch": 3.0674179517595888,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040734302573002565,
      "loss": 4.9801,
      "mean_token_accuracy": 0.2222036749124527,
      "num_tokens": 71078951.0,
      "step": 31030
    },
    {
      "entropy": 5.686424493789673,
      "epoch": 3.06791221826809,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004073144843213507,
      "loss": 4.9131,
      "mean_token_accuracy": 0.22868626415729523,
      "num_tokens": 71090039.0,
      "step": 31035
    },
    {
      "entropy": 5.66706395149231,
      "epoch": 3.0684064847765917,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040728593965771895,
      "loss": 4.8971,
      "mean_token_accuracy": 0.2343684107065201,
      "num_tokens": 71101203.0,
      "step": 31040
    },
    {
      "entropy": 5.712617826461792,
      "epoch": 3.068900751285093,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040725739173983253,
      "loss": 4.8894,
      "mean_token_accuracy": 0.22742980122566223,
      "num_tokens": 71112720.0,
      "step": 31045
    },
    {
      "entropy": 5.766223001480102,
      "epoch": 3.0693950177935942,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004072288405683938,
      "loss": 5.0112,
      "mean_token_accuracy": 0.22089194357395173,
      "num_tokens": 71123126.0,
      "step": 31050
    },
    {
      "entropy": 5.707275629043579,
      "epoch": 3.0698892843020955,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040720028614410497,
      "loss": 4.8771,
      "mean_token_accuracy": 0.22817520648241044,
      "num_tokens": 71135555.0,
      "step": 31055
    },
    {
      "entropy": 5.709377336502075,
      "epoch": 3.070383550810597,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004071717284676687,
      "loss": 4.8917,
      "mean_token_accuracy": 0.22726430594921113,
      "num_tokens": 71147304.0,
      "step": 31060
    },
    {
      "entropy": 5.749471807479859,
      "epoch": 3.0708778173190985,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004071431675397874,
      "loss": 4.9561,
      "mean_token_accuracy": 0.21819481998682022,
      "num_tokens": 71157564.0,
      "step": 31065
    },
    {
      "entropy": 5.6463275909423825,
      "epoch": 3.0713720838275997,
      "grad_norm": 1.125,
      "learning_rate": 0.00040711460336116364,
      "loss": 4.8295,
      "mean_token_accuracy": 0.23938505500555038,
      "num_tokens": 71168994.0,
      "step": 31070
    },
    {
      "entropy": 5.760703754425049,
      "epoch": 3.0718663503361014,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040708603593250016,
      "loss": 4.9544,
      "mean_token_accuracy": 0.2271053224802017,
      "num_tokens": 71180736.0,
      "step": 31075
    },
    {
      "entropy": 5.617252826690674,
      "epoch": 3.0723606168446027,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040705746525449964,
      "loss": 4.8698,
      "mean_token_accuracy": 0.23434925824403763,
      "num_tokens": 71193187.0,
      "step": 31080
    },
    {
      "entropy": 5.682850313186646,
      "epoch": 3.072854883353104,
      "grad_norm": 1.125,
      "learning_rate": 0.00040702889132786507,
      "loss": 4.8653,
      "mean_token_accuracy": 0.23569113165140151,
      "num_tokens": 71204294.0,
      "step": 31085
    },
    {
      "entropy": 5.708966016769409,
      "epoch": 3.0733491498616052,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004070003141532994,
      "loss": 4.9333,
      "mean_token_accuracy": 0.22191777229309081,
      "num_tokens": 71215259.0,
      "step": 31090
    },
    {
      "entropy": 5.6914191246032715,
      "epoch": 3.073843416370107,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004069717337315053,
      "loss": 4.936,
      "mean_token_accuracy": 0.2264552891254425,
      "num_tokens": 71227008.0,
      "step": 31095
    },
    {
      "entropy": 5.70731029510498,
      "epoch": 3.074337682878608,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040694315006318616,
      "loss": 4.9518,
      "mean_token_accuracy": 0.22575569599866868,
      "num_tokens": 71238190.0,
      "step": 31100
    },
    {
      "entropy": 5.781340312957764,
      "epoch": 3.0748319493871095,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040691456314904506,
      "loss": 4.9909,
      "mean_token_accuracy": 0.21723149269819259,
      "num_tokens": 71250534.0,
      "step": 31105
    },
    {
      "entropy": 5.689663505554199,
      "epoch": 3.0753262158956107,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040688597298978526,
      "loss": 4.9054,
      "mean_token_accuracy": 0.2318237841129303,
      "num_tokens": 71261432.0,
      "step": 31110
    },
    {
      "entropy": 5.7156758308410645,
      "epoch": 3.0758204824041124,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004068573795861101,
      "loss": 4.9962,
      "mean_token_accuracy": 0.21474476158618927,
      "num_tokens": 71272090.0,
      "step": 31115
    },
    {
      "entropy": 5.68539342880249,
      "epoch": 3.0763147489126137,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004068287829387229,
      "loss": 4.9342,
      "mean_token_accuracy": 0.2241128757596016,
      "num_tokens": 71284696.0,
      "step": 31120
    },
    {
      "entropy": 5.683704566955567,
      "epoch": 3.076809015421115,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040680018304832717,
      "loss": 4.9132,
      "mean_token_accuracy": 0.22966473549604416,
      "num_tokens": 71295701.0,
      "step": 31125
    },
    {
      "entropy": 5.723582983016968,
      "epoch": 3.0773032819296167,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040677157991562653,
      "loss": 4.9353,
      "mean_token_accuracy": 0.23452252000570298,
      "num_tokens": 71306798.0,
      "step": 31130
    },
    {
      "entropy": 5.703930521011353,
      "epoch": 3.077797548438118,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004067429735413245,
      "loss": 4.9366,
      "mean_token_accuracy": 0.22511770725250244,
      "num_tokens": 71319171.0,
      "step": 31135
    },
    {
      "entropy": 5.7279328346252445,
      "epoch": 3.078291814946619,
      "grad_norm": 1.15625,
      "learning_rate": 0.000406714363926125,
      "loss": 4.9478,
      "mean_token_accuracy": 0.22067981362342834,
      "num_tokens": 71331479.0,
      "step": 31140
    },
    {
      "entropy": 5.669634246826172,
      "epoch": 3.0787860814551204,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040668575107073165,
      "loss": 4.9178,
      "mean_token_accuracy": 0.23055005073547363,
      "num_tokens": 71343044.0,
      "step": 31145
    },
    {
      "entropy": 5.810487413406372,
      "epoch": 3.079280347963622,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040665713497584836,
      "loss": 4.9931,
      "mean_token_accuracy": 0.2229029580950737,
      "num_tokens": 71354552.0,
      "step": 31150
    },
    {
      "entropy": 5.713155937194824,
      "epoch": 3.0797746144721234,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040662851564217915,
      "loss": 4.9011,
      "mean_token_accuracy": 0.23371504247188568,
      "num_tokens": 71366201.0,
      "step": 31155
    },
    {
      "entropy": 5.70947380065918,
      "epoch": 3.0802688809806247,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040659989307042816,
      "loss": 4.8986,
      "mean_token_accuracy": 0.2202175334095955,
      "num_tokens": 71377966.0,
      "step": 31160
    },
    {
      "entropy": 5.710884380340576,
      "epoch": 3.080763147489126,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040657126726129924,
      "loss": 4.9721,
      "mean_token_accuracy": 0.2262341558933258,
      "num_tokens": 71388927.0,
      "step": 31165
    },
    {
      "entropy": 5.702050399780274,
      "epoch": 3.0812574139976276,
      "grad_norm": 1.125,
      "learning_rate": 0.0004065426382154968,
      "loss": 4.9549,
      "mean_token_accuracy": 0.22521415650844573,
      "num_tokens": 71401728.0,
      "step": 31170
    },
    {
      "entropy": 5.810238265991211,
      "epoch": 3.081751680506129,
      "grad_norm": 1.140625,
      "learning_rate": 0.000406514005933725,
      "loss": 5.0944,
      "mean_token_accuracy": 0.2099160596728325,
      "num_tokens": 71413694.0,
      "step": 31175
    },
    {
      "entropy": 5.7336677551269535,
      "epoch": 3.08224594701463,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004064853704166883,
      "loss": 4.9104,
      "mean_token_accuracy": 0.2298359453678131,
      "num_tokens": 71425485.0,
      "step": 31180
    },
    {
      "entropy": 5.731386661529541,
      "epoch": 3.082740213523132,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000406456731665091,
      "loss": 4.9518,
      "mean_token_accuracy": 0.21932245939970016,
      "num_tokens": 71437320.0,
      "step": 31185
    },
    {
      "entropy": 5.7461425304412845,
      "epoch": 3.083234480031633,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040642808967963787,
      "loss": 4.9479,
      "mean_token_accuracy": 0.2205512046813965,
      "num_tokens": 71448948.0,
      "step": 31190
    },
    {
      "entropy": 5.694533634185791,
      "epoch": 3.0837287465401344,
      "grad_norm": 1.125,
      "learning_rate": 0.00040639944446103324,
      "loss": 4.8874,
      "mean_token_accuracy": 0.2287810727953911,
      "num_tokens": 71460997.0,
      "step": 31195
    },
    {
      "entropy": 5.672366380691528,
      "epoch": 3.0842230130486357,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040637079600998186,
      "loss": 4.8923,
      "mean_token_accuracy": 0.22718246877193451,
      "num_tokens": 71472174.0,
      "step": 31200
    },
    {
      "entropy": 5.735339498519897,
      "epoch": 3.0847172795571374,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004063421443271885,
      "loss": 4.9658,
      "mean_token_accuracy": 0.22571628391742707,
      "num_tokens": 71484306.0,
      "step": 31205
    },
    {
      "entropy": 5.721938848495483,
      "epoch": 3.0852115460656386,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00040631348941335814,
      "loss": 4.906,
      "mean_token_accuracy": 0.22412378937005997,
      "num_tokens": 71494866.0,
      "step": 31210
    },
    {
      "entropy": 5.737693786621094,
      "epoch": 3.08570581257414,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004062848312691955,
      "loss": 4.9307,
      "mean_token_accuracy": 0.22584894597530364,
      "num_tokens": 71505591.0,
      "step": 31215
    },
    {
      "entropy": 5.718345594406128,
      "epoch": 3.086200079082641,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040625616989540565,
      "loss": 4.9472,
      "mean_token_accuracy": 0.22469822615385054,
      "num_tokens": 71517651.0,
      "step": 31220
    },
    {
      "entropy": 5.787763833999634,
      "epoch": 3.086694345591143,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004062275052926936,
      "loss": 4.9671,
      "mean_token_accuracy": 0.21957968026399613,
      "num_tokens": 71529301.0,
      "step": 31225
    },
    {
      "entropy": 5.609581422805786,
      "epoch": 3.087188612099644,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004061988374617646,
      "loss": 4.8004,
      "mean_token_accuracy": 0.23956407755613326,
      "num_tokens": 71539971.0,
      "step": 31230
    },
    {
      "entropy": 5.697654819488525,
      "epoch": 3.0876828786081454,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004061701664033239,
      "loss": 5.0008,
      "mean_token_accuracy": 0.2199922189116478,
      "num_tokens": 71551905.0,
      "step": 31235
    },
    {
      "entropy": 5.729603624343872,
      "epoch": 3.088177145116647,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004061414921180767,
      "loss": 4.923,
      "mean_token_accuracy": 0.22901807427406312,
      "num_tokens": 71562945.0,
      "step": 31240
    },
    {
      "entropy": 5.716499996185303,
      "epoch": 3.0886714116251484,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040611281460672846,
      "loss": 5.0207,
      "mean_token_accuracy": 0.2128499686717987,
      "num_tokens": 71574429.0,
      "step": 31245
    },
    {
      "entropy": 5.717672634124756,
      "epoch": 3.0891656781336496,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004060841338699847,
      "loss": 4.9806,
      "mean_token_accuracy": 0.22163566052913666,
      "num_tokens": 71584827.0,
      "step": 31250
    },
    {
      "entropy": 5.75890007019043,
      "epoch": 3.089659944642151,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040605544990855074,
      "loss": 4.8915,
      "mean_token_accuracy": 0.22389666587114335,
      "num_tokens": 71595729.0,
      "step": 31255
    },
    {
      "entropy": 5.7560326099395756,
      "epoch": 3.0901542111506526,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040602676272313257,
      "loss": 4.9406,
      "mean_token_accuracy": 0.2305128663778305,
      "num_tokens": 71606912.0,
      "step": 31260
    },
    {
      "entropy": 5.706454801559448,
      "epoch": 3.090648477659154,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040599807231443566,
      "loss": 4.9866,
      "mean_token_accuracy": 0.22579252421855928,
      "num_tokens": 71618706.0,
      "step": 31265
    },
    {
      "entropy": 5.703484725952149,
      "epoch": 3.091142744167655,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004059693786831658,
      "loss": 4.9061,
      "mean_token_accuracy": 0.22688717991113663,
      "num_tokens": 71630826.0,
      "step": 31270
    },
    {
      "entropy": 5.767189979553223,
      "epoch": 3.0916370106761564,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000405940681830029,
      "loss": 5.0019,
      "mean_token_accuracy": 0.22507247775793077,
      "num_tokens": 71643914.0,
      "step": 31275
    },
    {
      "entropy": 5.726987314224243,
      "epoch": 3.092131277184658,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000405911981755731,
      "loss": 4.9029,
      "mean_token_accuracy": 0.2315288871526718,
      "num_tokens": 71655523.0,
      "step": 31280
    },
    {
      "entropy": 5.6731593132019045,
      "epoch": 3.0926255436931593,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004058832784609779,
      "loss": 4.9094,
      "mean_token_accuracy": 0.23002844154834748,
      "num_tokens": 71667423.0,
      "step": 31285
    },
    {
      "entropy": 5.659725856781006,
      "epoch": 3.0931198102016606,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040585457194647595,
      "loss": 4.9119,
      "mean_token_accuracy": 0.22417925596237182,
      "num_tokens": 71678116.0,
      "step": 31290
    },
    {
      "entropy": 5.745096063613891,
      "epoch": 3.0936140767101623,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040582586221293127,
      "loss": 4.9962,
      "mean_token_accuracy": 0.2172016754746437,
      "num_tokens": 71690109.0,
      "step": 31295
    },
    {
      "entropy": 5.776556539535522,
      "epoch": 3.0941083432186636,
      "grad_norm": 1.125,
      "learning_rate": 0.00040579714926105,
      "loss": 5.0262,
      "mean_token_accuracy": 0.21753207445144654,
      "num_tokens": 71702640.0,
      "step": 31300
    },
    {
      "entropy": 5.717349672317505,
      "epoch": 3.094602609727165,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004057684330915386,
      "loss": 4.9112,
      "mean_token_accuracy": 0.2259405493736267,
      "num_tokens": 71714537.0,
      "step": 31305
    },
    {
      "entropy": 5.685066080093383,
      "epoch": 3.095096876235666,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040573971370510344,
      "loss": 5.0178,
      "mean_token_accuracy": 0.2217104747891426,
      "num_tokens": 71724901.0,
      "step": 31310
    },
    {
      "entropy": 5.759277057647705,
      "epoch": 3.095591142744168,
      "grad_norm": 1.0625,
      "learning_rate": 0.000405710991102451,
      "loss": 4.994,
      "mean_token_accuracy": 0.2186857581138611,
      "num_tokens": 71736640.0,
      "step": 31315
    },
    {
      "entropy": 5.780472993850708,
      "epoch": 3.096085409252669,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00040568226528428795,
      "loss": 4.9923,
      "mean_token_accuracy": 0.21885482370853424,
      "num_tokens": 71747193.0,
      "step": 31320
    },
    {
      "entropy": 5.7577821731567385,
      "epoch": 3.0965796757611703,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000405653536251321,
      "loss": 5.0427,
      "mean_token_accuracy": 0.21533706337213515,
      "num_tokens": 71759073.0,
      "step": 31325
    },
    {
      "entropy": 5.762305116653442,
      "epoch": 3.097073942269672,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004056248040042566,
      "loss": 4.9603,
      "mean_token_accuracy": 0.21935534626245498,
      "num_tokens": 71770486.0,
      "step": 31330
    },
    {
      "entropy": 5.761210250854492,
      "epoch": 3.0975682087781733,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040559606854380186,
      "loss": 5.0241,
      "mean_token_accuracy": 0.22060358971357347,
      "num_tokens": 71781536.0,
      "step": 31335
    },
    {
      "entropy": 5.743462181091308,
      "epoch": 3.0980624752866746,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040556732987066357,
      "loss": 5.009,
      "mean_token_accuracy": 0.22493914514780045,
      "num_tokens": 71792521.0,
      "step": 31340
    },
    {
      "entropy": 5.705568647384643,
      "epoch": 3.098556741795176,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004055385879855487,
      "loss": 4.9938,
      "mean_token_accuracy": 0.22277334332466125,
      "num_tokens": 71804150.0,
      "step": 31345
    },
    {
      "entropy": 5.647363138198853,
      "epoch": 3.0990510083036775,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040550984288916436,
      "loss": 4.9196,
      "mean_token_accuracy": 0.23266882002353667,
      "num_tokens": 71815842.0,
      "step": 31350
    },
    {
      "entropy": 5.8149501323699955,
      "epoch": 3.099545274812179,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040548109458221756,
      "loss": 4.9558,
      "mean_token_accuracy": 0.2228508174419403,
      "num_tokens": 71827243.0,
      "step": 31355
    },
    {
      "entropy": 5.775880241394043,
      "epoch": 3.10003954132068,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004054523430654156,
      "loss": 5.0069,
      "mean_token_accuracy": 0.2256786346435547,
      "num_tokens": 71839424.0,
      "step": 31360
    },
    {
      "entropy": 5.739609336853027,
      "epoch": 3.1005338078291813,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040542358833946577,
      "loss": 5.0157,
      "mean_token_accuracy": 0.21733000725507737,
      "num_tokens": 71851709.0,
      "step": 31365
    },
    {
      "entropy": 5.7281341552734375,
      "epoch": 3.101028074337683,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004053948304050755,
      "loss": 4.9285,
      "mean_token_accuracy": 0.225421042740345,
      "num_tokens": 71862841.0,
      "step": 31370
    },
    {
      "entropy": 5.7415464401245115,
      "epoch": 3.1015223408461843,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040536606926295203,
      "loss": 4.935,
      "mean_token_accuracy": 0.23016288727521897,
      "num_tokens": 71873104.0,
      "step": 31375
    },
    {
      "entropy": 5.792458152770996,
      "epoch": 3.1020166073546855,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040533730491380316,
      "loss": 5.0455,
      "mean_token_accuracy": 0.21659953147172928,
      "num_tokens": 71884809.0,
      "step": 31380
    },
    {
      "entropy": 5.693872165679932,
      "epoch": 3.102510873863187,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004053085373583363,
      "loss": 4.9467,
      "mean_token_accuracy": 0.2260803371667862,
      "num_tokens": 71896625.0,
      "step": 31385
    },
    {
      "entropy": 5.682625436782837,
      "epoch": 3.1030051403716885,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004052797665972593,
      "loss": 4.9256,
      "mean_token_accuracy": 0.22938040345907212,
      "num_tokens": 71908628.0,
      "step": 31390
    },
    {
      "entropy": 5.757524728775024,
      "epoch": 3.1034994068801898,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00040525099263127976,
      "loss": 4.9654,
      "mean_token_accuracy": 0.22198040932416915,
      "num_tokens": 71919971.0,
      "step": 31395
    },
    {
      "entropy": 5.764160108566284,
      "epoch": 3.103993673388691,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040522221546110556,
      "loss": 5.0166,
      "mean_token_accuracy": 0.22115676254034042,
      "num_tokens": 71931240.0,
      "step": 31400
    },
    {
      "entropy": 5.701149129867554,
      "epoch": 3.1044879398971927,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040519343508744473,
      "loss": 4.9414,
      "mean_token_accuracy": 0.22984751611948012,
      "num_tokens": 71942735.0,
      "step": 31405
    },
    {
      "entropy": 5.688599395751953,
      "epoch": 3.104982206405694,
      "grad_norm": 1.25,
      "learning_rate": 0.00040516465151100515,
      "loss": 4.9084,
      "mean_token_accuracy": 0.24144310802221297,
      "num_tokens": 71953685.0,
      "step": 31410
    },
    {
      "entropy": 5.7322924613952635,
      "epoch": 3.1054764729141953,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00040513586473249496,
      "loss": 4.9649,
      "mean_token_accuracy": 0.22066458761692048,
      "num_tokens": 71966061.0,
      "step": 31415
    },
    {
      "entropy": 5.7660572052001955,
      "epoch": 3.1059707394226965,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004051070747526223,
      "loss": 4.9936,
      "mean_token_accuracy": 0.2141634553670883,
      "num_tokens": 71978002.0,
      "step": 31420
    },
    {
      "entropy": 5.677839040756226,
      "epoch": 3.1064650059311982,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00040507828157209546,
      "loss": 4.8687,
      "mean_token_accuracy": 0.23158190548419952,
      "num_tokens": 71988504.0,
      "step": 31425
    },
    {
      "entropy": 5.722964239120484,
      "epoch": 3.1069592724396995,
      "grad_norm": 1.125,
      "learning_rate": 0.00040504948519162276,
      "loss": 5.0031,
      "mean_token_accuracy": 0.2152043864130974,
      "num_tokens": 72000900.0,
      "step": 31430
    },
    {
      "entropy": 5.793111658096313,
      "epoch": 3.1074535389482008,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004050206856119124,
      "loss": 5.0191,
      "mean_token_accuracy": 0.21697353422641755,
      "num_tokens": 72011134.0,
      "step": 31435
    },
    {
      "entropy": 5.7023913860321045,
      "epoch": 3.1079478054567025,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004049918828336732,
      "loss": 4.9119,
      "mean_token_accuracy": 0.22637180387973785,
      "num_tokens": 72022784.0,
      "step": 31440
    },
    {
      "entropy": 5.673501110076904,
      "epoch": 3.1084420719652037,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004049630768576133,
      "loss": 4.8836,
      "mean_token_accuracy": 0.22919871658086777,
      "num_tokens": 72033353.0,
      "step": 31445
    },
    {
      "entropy": 5.744571208953857,
      "epoch": 3.108936338473705,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040493426768444164,
      "loss": 4.9521,
      "mean_token_accuracy": 0.2181500419974327,
      "num_tokens": 72044182.0,
      "step": 31450
    },
    {
      "entropy": 5.797312307357788,
      "epoch": 3.1094306049822062,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004049054553148668,
      "loss": 5.0151,
      "mean_token_accuracy": 0.21082063317298888,
      "num_tokens": 72056365.0,
      "step": 31455
    },
    {
      "entropy": 5.724458694458008,
      "epoch": 3.109924871490708,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004048766397495977,
      "loss": 4.9187,
      "mean_token_accuracy": 0.22703131139278412,
      "num_tokens": 72066580.0,
      "step": 31460
    },
    {
      "entropy": 5.643708944320679,
      "epoch": 3.110419137999209,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040484782098934303,
      "loss": 4.9641,
      "mean_token_accuracy": 0.22813219726085662,
      "num_tokens": 72077731.0,
      "step": 31465
    },
    {
      "entropy": 5.7252216815948485,
      "epoch": 3.1109134045077105,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004048189990348119,
      "loss": 4.9782,
      "mean_token_accuracy": 0.21828000098466874,
      "num_tokens": 72088785.0,
      "step": 31470
    },
    {
      "entropy": 5.751339101791382,
      "epoch": 3.1114076710162117,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040479017388671317,
      "loss": 4.927,
      "mean_token_accuracy": 0.23021653443574905,
      "num_tokens": 72100235.0,
      "step": 31475
    },
    {
      "entropy": 5.645326852798462,
      "epoch": 3.1119019375247134,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040476134554575613,
      "loss": 4.8912,
      "mean_token_accuracy": 0.2246718406677246,
      "num_tokens": 72112124.0,
      "step": 31480
    },
    {
      "entropy": 5.717307138442993,
      "epoch": 3.1123962040332147,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004047325140126497,
      "loss": 4.9369,
      "mean_token_accuracy": 0.22012407034635545,
      "num_tokens": 72124138.0,
      "step": 31485
    },
    {
      "entropy": 5.732042026519776,
      "epoch": 3.112890470541716,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040470367928810343,
      "loss": 4.9369,
      "mean_token_accuracy": 0.22572326064109802,
      "num_tokens": 72134330.0,
      "step": 31490
    },
    {
      "entropy": 5.824021339416504,
      "epoch": 3.1133847370502177,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040467484137282644,
      "loss": 5.068,
      "mean_token_accuracy": 0.2133909910917282,
      "num_tokens": 72146999.0,
      "step": 31495
    },
    {
      "entropy": 5.745410680770874,
      "epoch": 3.113879003558719,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040464600026752834,
      "loss": 4.9349,
      "mean_token_accuracy": 0.2273472413420677,
      "num_tokens": 72158648.0,
      "step": 31500
    },
    {
      "entropy": 5.7780516147613525,
      "epoch": 3.11437327006722,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004046171559729184,
      "loss": 4.9525,
      "mean_token_accuracy": 0.2263324126601219,
      "num_tokens": 72169621.0,
      "step": 31505
    },
    {
      "entropy": 5.679072475433349,
      "epoch": 3.1148675365757215,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004045883084897064,
      "loss": 4.8808,
      "mean_token_accuracy": 0.22987501323223114,
      "num_tokens": 72179928.0,
      "step": 31510
    },
    {
      "entropy": 5.6903337955474855,
      "epoch": 3.115361803084223,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040455945781860185,
      "loss": 4.9067,
      "mean_token_accuracy": 0.22699038833379745,
      "num_tokens": 72190699.0,
      "step": 31515
    },
    {
      "entropy": 5.709006929397583,
      "epoch": 3.1158560695927244,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004045306039603146,
      "loss": 4.9792,
      "mean_token_accuracy": 0.22325705289840697,
      "num_tokens": 72202893.0,
      "step": 31520
    },
    {
      "entropy": 5.685155963897705,
      "epoch": 3.1163503361012257,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040450174691555434,
      "loss": 5.0271,
      "mean_token_accuracy": 0.21683308631181716,
      "num_tokens": 72214614.0,
      "step": 31525
    },
    {
      "entropy": 5.685437250137329,
      "epoch": 3.116844602609727,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004044728866850311,
      "loss": 4.939,
      "mean_token_accuracy": 0.22852666079998016,
      "num_tokens": 72225882.0,
      "step": 31530
    },
    {
      "entropy": 5.782245111465454,
      "epoch": 3.1173388691182287,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004044440232694547,
      "loss": 4.9572,
      "mean_token_accuracy": 0.22150728106498718,
      "num_tokens": 72236264.0,
      "step": 31535
    },
    {
      "entropy": 5.740688943862915,
      "epoch": 3.11783313562673,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004044151566695353,
      "loss": 5.0272,
      "mean_token_accuracy": 0.2189900115132332,
      "num_tokens": 72247380.0,
      "step": 31540
    },
    {
      "entropy": 5.73533992767334,
      "epoch": 3.118327402135231,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040438628688598287,
      "loss": 4.8638,
      "mean_token_accuracy": 0.230111263692379,
      "num_tokens": 72259141.0,
      "step": 31545
    },
    {
      "entropy": 5.6774030208587645,
      "epoch": 3.118821668643733,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004043574139195078,
      "loss": 4.8268,
      "mean_token_accuracy": 0.23159618824720382,
      "num_tokens": 72269549.0,
      "step": 31550
    },
    {
      "entropy": 5.673881196975708,
      "epoch": 3.119315935152234,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040432853777082025,
      "loss": 4.9053,
      "mean_token_accuracy": 0.23007372319698333,
      "num_tokens": 72281589.0,
      "step": 31555
    },
    {
      "entropy": 5.7159655570983885,
      "epoch": 3.1198102016607354,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004042996584406305,
      "loss": 5.0175,
      "mean_token_accuracy": 0.21938789784908294,
      "num_tokens": 72294239.0,
      "step": 31560
    },
    {
      "entropy": 5.690757465362549,
      "epoch": 3.1203044681692367,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040427077592964923,
      "loss": 4.9886,
      "mean_token_accuracy": 0.22104758769273758,
      "num_tokens": 72306153.0,
      "step": 31565
    },
    {
      "entropy": 5.71517162322998,
      "epoch": 3.1207987346777384,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004042418902385868,
      "loss": 4.9349,
      "mean_token_accuracy": 0.2226513773202896,
      "num_tokens": 72317270.0,
      "step": 31570
    },
    {
      "entropy": 5.721489095687867,
      "epoch": 3.1212930011862396,
      "grad_norm": 1.125,
      "learning_rate": 0.00040421300136815375,
      "loss": 4.9838,
      "mean_token_accuracy": 0.22252339124679565,
      "num_tokens": 72329073.0,
      "step": 31575
    },
    {
      "entropy": 5.691077470779419,
      "epoch": 3.121787267694741,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004041841093190609,
      "loss": 4.9224,
      "mean_token_accuracy": 0.22589288651943207,
      "num_tokens": 72339677.0,
      "step": 31580
    },
    {
      "entropy": 5.776186132431031,
      "epoch": 3.1222815342032426,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004041552140920189,
      "loss": 4.9716,
      "mean_token_accuracy": 0.22918739765882493,
      "num_tokens": 72350527.0,
      "step": 31585
    },
    {
      "entropy": 5.759203195571899,
      "epoch": 3.122775800711744,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004041263156877386,
      "loss": 4.9799,
      "mean_token_accuracy": 0.219963239133358,
      "num_tokens": 72362740.0,
      "step": 31590
    },
    {
      "entropy": 5.710923910140991,
      "epoch": 3.123270067220245,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040409741410693084,
      "loss": 4.9462,
      "mean_token_accuracy": 0.21731142103672027,
      "num_tokens": 72372705.0,
      "step": 31595
    },
    {
      "entropy": 5.596602964401245,
      "epoch": 3.1237643337287464,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040406850935030677,
      "loss": 4.8619,
      "mean_token_accuracy": 0.23526178449392318,
      "num_tokens": 72385085.0,
      "step": 31600
    },
    {
      "entropy": 5.6815074443817135,
      "epoch": 3.124258600237248,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040403960141857724,
      "loss": 4.9771,
      "mean_token_accuracy": 0.22717182487249374,
      "num_tokens": 72397017.0,
      "step": 31605
    },
    {
      "entropy": 5.746268367767334,
      "epoch": 3.1247528667457494,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004040106903124536,
      "loss": 4.9338,
      "mean_token_accuracy": 0.221175517141819,
      "num_tokens": 72409378.0,
      "step": 31610
    },
    {
      "entropy": 5.754933214187622,
      "epoch": 3.1252471332542506,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004039817760326469,
      "loss": 5.0087,
      "mean_token_accuracy": 0.21484584361314774,
      "num_tokens": 72420585.0,
      "step": 31615
    },
    {
      "entropy": 5.806163024902344,
      "epoch": 3.125741399762752,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040395285857986845,
      "loss": 5.0827,
      "mean_token_accuracy": 0.21263491958379746,
      "num_tokens": 72431779.0,
      "step": 31620
    },
    {
      "entropy": 5.745228576660156,
      "epoch": 3.1262356662712536,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004039239379548298,
      "loss": 4.9075,
      "mean_token_accuracy": 0.22435821890830993,
      "num_tokens": 72441421.0,
      "step": 31625
    },
    {
      "entropy": 5.774961614608765,
      "epoch": 3.126729932779755,
      "grad_norm": 1.25,
      "learning_rate": 0.0004038950141582422,
      "loss": 4.968,
      "mean_token_accuracy": 0.22007061243057252,
      "num_tokens": 72452664.0,
      "step": 31630
    },
    {
      "entropy": 5.68954610824585,
      "epoch": 3.127224199288256,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004038660871908173,
      "loss": 4.8621,
      "mean_token_accuracy": 0.23072129487991333,
      "num_tokens": 72464008.0,
      "step": 31635
    },
    {
      "entropy": 5.773107481002808,
      "epoch": 3.1277184657967574,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040383715705326665,
      "loss": 5.0476,
      "mean_token_accuracy": 0.21385151892900467,
      "num_tokens": 72475782.0,
      "step": 31640
    },
    {
      "entropy": 5.704550647735596,
      "epoch": 3.128212732305259,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004038082237463019,
      "loss": 4.975,
      "mean_token_accuracy": 0.2204524338245392,
      "num_tokens": 72487770.0,
      "step": 31645
    },
    {
      "entropy": 5.653398656845093,
      "epoch": 3.1287069988137604,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040377928727063485,
      "loss": 4.8925,
      "mean_token_accuracy": 0.2310608670115471,
      "num_tokens": 72499336.0,
      "step": 31650
    },
    {
      "entropy": 5.708260726928711,
      "epoch": 3.1292012653222616,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040375034762697736,
      "loss": 4.937,
      "mean_token_accuracy": 0.223634672164917,
      "num_tokens": 72510658.0,
      "step": 31655
    },
    {
      "entropy": 5.721911668777466,
      "epoch": 3.1296955318307633,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004037214048160415,
      "loss": 4.9633,
      "mean_token_accuracy": 0.21866945475339888,
      "num_tokens": 72522617.0,
      "step": 31660
    },
    {
      "entropy": 5.684777450561524,
      "epoch": 3.1301897983392646,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004036924588385389,
      "loss": 4.8519,
      "mean_token_accuracy": 0.23900388777256013,
      "num_tokens": 72532829.0,
      "step": 31665
    },
    {
      "entropy": 5.720617961883545,
      "epoch": 3.130684064847766,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040366350969518196,
      "loss": 4.9315,
      "mean_token_accuracy": 0.2335021197795868,
      "num_tokens": 72543556.0,
      "step": 31670
    },
    {
      "entropy": 5.79214825630188,
      "epoch": 3.131178331356267,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004036345573866827,
      "loss": 4.9969,
      "mean_token_accuracy": 0.21669473201036454,
      "num_tokens": 72554545.0,
      "step": 31675
    },
    {
      "entropy": 5.803208446502685,
      "epoch": 3.131672597864769,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040360560191375335,
      "loss": 5.0471,
      "mean_token_accuracy": 0.21242739707231523,
      "num_tokens": 72566158.0,
      "step": 31680
    },
    {
      "entropy": 5.758175325393677,
      "epoch": 3.13216686437327,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004035766432771063,
      "loss": 5.0666,
      "mean_token_accuracy": 0.21780523657798767,
      "num_tokens": 72578229.0,
      "step": 31685
    },
    {
      "entropy": 5.6706194400787355,
      "epoch": 3.1326611308817713,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040354768147745387,
      "loss": 4.8682,
      "mean_token_accuracy": 0.23317491561174392,
      "num_tokens": 72589237.0,
      "step": 31690
    },
    {
      "entropy": 5.750687026977539,
      "epoch": 3.133155397390273,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004035187165155085,
      "loss": 4.9702,
      "mean_token_accuracy": 0.22496314197778702,
      "num_tokens": 72600531.0,
      "step": 31695
    },
    {
      "entropy": 5.710323143005371,
      "epoch": 3.1336496638987743,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004034897483919827,
      "loss": 4.9053,
      "mean_token_accuracy": 0.22635679543018342,
      "num_tokens": 72611861.0,
      "step": 31700
    },
    {
      "entropy": 5.732558298110962,
      "epoch": 3.1341439304072756,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004034607771075893,
      "loss": 4.9887,
      "mean_token_accuracy": 0.22363526821136476,
      "num_tokens": 72624142.0,
      "step": 31705
    },
    {
      "entropy": 5.702888154983521,
      "epoch": 3.134638196915777,
      "grad_norm": 1.3125,
      "learning_rate": 0.00040343180266304074,
      "loss": 4.901,
      "mean_token_accuracy": 0.22720416337251664,
      "num_tokens": 72634995.0,
      "step": 31710
    },
    {
      "entropy": 5.768390750885009,
      "epoch": 3.1351324634242785,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040340282505904997,
      "loss": 5.0145,
      "mean_token_accuracy": 0.2214521810412407,
      "num_tokens": 72646798.0,
      "step": 31715
    },
    {
      "entropy": 5.74534592628479,
      "epoch": 3.13562672993278,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040337384429632974,
      "loss": 4.9645,
      "mean_token_accuracy": 0.22634658962488174,
      "num_tokens": 72657938.0,
      "step": 31720
    },
    {
      "entropy": 5.786575651168823,
      "epoch": 3.136120996441281,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000403344860375593,
      "loss": 5.0207,
      "mean_token_accuracy": 0.21638917922973633,
      "num_tokens": 72671069.0,
      "step": 31725
    },
    {
      "entropy": 5.732324123382568,
      "epoch": 3.1366152629497823,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004033158732975529,
      "loss": 4.9656,
      "mean_token_accuracy": 0.2240966945886612,
      "num_tokens": 72682695.0,
      "step": 31730
    },
    {
      "entropy": 5.732345533370972,
      "epoch": 3.137109529458284,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004032868830629223,
      "loss": 5.0004,
      "mean_token_accuracy": 0.2207697555422783,
      "num_tokens": 72694514.0,
      "step": 31735
    },
    {
      "entropy": 5.686329936981201,
      "epoch": 3.1376037959667853,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040325788967241453,
      "loss": 4.9014,
      "mean_token_accuracy": 0.22417863160371782,
      "num_tokens": 72705427.0,
      "step": 31740
    },
    {
      "entropy": 5.655508327484131,
      "epoch": 3.1380980624752866,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040322889312674275,
      "loss": 4.8744,
      "mean_token_accuracy": 0.22913722395896913,
      "num_tokens": 72716877.0,
      "step": 31745
    },
    {
      "entropy": 5.685426902770996,
      "epoch": 3.1385923289837883,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040319989342662036,
      "loss": 4.9027,
      "mean_token_accuracy": 0.22637013345956802,
      "num_tokens": 72727927.0,
      "step": 31750
    },
    {
      "entropy": 5.776641845703125,
      "epoch": 3.1390865954922895,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040317089057276067,
      "loss": 5.0313,
      "mean_token_accuracy": 0.22064234018325807,
      "num_tokens": 72739645.0,
      "step": 31755
    },
    {
      "entropy": 5.759417247772217,
      "epoch": 3.139580862000791,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004031418845658772,
      "loss": 5.0366,
      "mean_token_accuracy": 0.20906257778406143,
      "num_tokens": 72750961.0,
      "step": 31760
    },
    {
      "entropy": 5.826264142990112,
      "epoch": 3.140075128509292,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040311287540668346,
      "loss": 5.015,
      "mean_token_accuracy": 0.21465834081172944,
      "num_tokens": 72763599.0,
      "step": 31765
    },
    {
      "entropy": 5.737165355682373,
      "epoch": 3.1405693950177938,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040308386309589314,
      "loss": 4.8999,
      "mean_token_accuracy": 0.22906113564968109,
      "num_tokens": 72775807.0,
      "step": 31770
    },
    {
      "entropy": 5.730438756942749,
      "epoch": 3.141063661526295,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004030548476342199,
      "loss": 4.9335,
      "mean_token_accuracy": 0.22640907615423203,
      "num_tokens": 72786161.0,
      "step": 31775
    },
    {
      "entropy": 5.735044145584107,
      "epoch": 3.1415579280347963,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040302582902237755,
      "loss": 4.9932,
      "mean_token_accuracy": 0.22479147762060164,
      "num_tokens": 72798211.0,
      "step": 31780
    },
    {
      "entropy": 5.780780792236328,
      "epoch": 3.1420521945432975,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040299680726107995,
      "loss": 4.9966,
      "mean_token_accuracy": 0.21778926104307175,
      "num_tokens": 72809187.0,
      "step": 31785
    },
    {
      "entropy": 5.709589052200317,
      "epoch": 3.1425464610517992,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040296778235104105,
      "loss": 4.9373,
      "mean_token_accuracy": 0.23076942861080169,
      "num_tokens": 72820486.0,
      "step": 31790
    },
    {
      "entropy": 5.6492979526519775,
      "epoch": 3.1430407275603005,
      "grad_norm": 1.25,
      "learning_rate": 0.0004029387542929748,
      "loss": 4.8851,
      "mean_token_accuracy": 0.23109911531209945,
      "num_tokens": 72831442.0,
      "step": 31795
    },
    {
      "entropy": 5.68577618598938,
      "epoch": 3.1435349940688018,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040290972308759535,
      "loss": 4.892,
      "mean_token_accuracy": 0.22983373403549195,
      "num_tokens": 72841902.0,
      "step": 31800
    },
    {
      "entropy": 5.641751766204834,
      "epoch": 3.1440292605773035,
      "grad_norm": 1.25,
      "learning_rate": 0.00040288068873561687,
      "loss": 4.8203,
      "mean_token_accuracy": 0.24102201759815217,
      "num_tokens": 72852327.0,
      "step": 31805
    },
    {
      "entropy": 5.751206731796264,
      "epoch": 3.1445235270858047,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004028516512377537,
      "loss": 4.9664,
      "mean_token_accuracy": 0.21753532737493514,
      "num_tokens": 72862643.0,
      "step": 31810
    },
    {
      "entropy": 5.682747793197632,
      "epoch": 3.145017793594306,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040282261059471997,
      "loss": 4.9484,
      "mean_token_accuracy": 0.2258808732032776,
      "num_tokens": 72874081.0,
      "step": 31815
    },
    {
      "entropy": 5.713838005065918,
      "epoch": 3.1455120601028073,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040279356680723025,
      "loss": 4.9666,
      "mean_token_accuracy": 0.2238702654838562,
      "num_tokens": 72885472.0,
      "step": 31820
    },
    {
      "entropy": 5.770482158660888,
      "epoch": 3.146006326611309,
      "grad_norm": 1.125,
      "learning_rate": 0.0004027645198759988,
      "loss": 4.927,
      "mean_token_accuracy": 0.23033130764961243,
      "num_tokens": 72896626.0,
      "step": 31825
    },
    {
      "entropy": 5.743639373779297,
      "epoch": 3.1465005931198102,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004027354698017405,
      "loss": 4.9388,
      "mean_token_accuracy": 0.23063793033361435,
      "num_tokens": 72907782.0,
      "step": 31830
    },
    {
      "entropy": 5.693716192245484,
      "epoch": 3.1469948596283115,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040270641658516986,
      "loss": 4.9009,
      "mean_token_accuracy": 0.22122326195240022,
      "num_tokens": 72919024.0,
      "step": 31835
    },
    {
      "entropy": 5.725043535232544,
      "epoch": 3.147489126136813,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004026773602270015,
      "loss": 4.9725,
      "mean_token_accuracy": 0.22053425014019012,
      "num_tokens": 72930576.0,
      "step": 31840
    },
    {
      "entropy": 5.730515813827514,
      "epoch": 3.1479833926453145,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040264830072795027,
      "loss": 5.0028,
      "mean_token_accuracy": 0.21909455060958863,
      "num_tokens": 72942489.0,
      "step": 31845
    },
    {
      "entropy": 5.789901971817017,
      "epoch": 3.1484776591538157,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004026192380887311,
      "loss": 5.0128,
      "mean_token_accuracy": 0.2121368795633316,
      "num_tokens": 72954209.0,
      "step": 31850
    },
    {
      "entropy": 5.840720796585083,
      "epoch": 3.148971925662317,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004025901723100588,
      "loss": 5.0419,
      "mean_token_accuracy": 0.22017311602830886,
      "num_tokens": 72965939.0,
      "step": 31855
    },
    {
      "entropy": 5.680883121490479,
      "epoch": 3.1494661921708187,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004025611033926484,
      "loss": 4.8969,
      "mean_token_accuracy": 0.23203047066926957,
      "num_tokens": 72977114.0,
      "step": 31860
    },
    {
      "entropy": 5.716052389144897,
      "epoch": 3.14996045867932,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040253203133721515,
      "loss": 5.0123,
      "mean_token_accuracy": 0.22163487523794173,
      "num_tokens": 72989726.0,
      "step": 31865
    },
    {
      "entropy": 5.697899341583252,
      "epoch": 3.150454725187821,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004025029561444742,
      "loss": 4.9529,
      "mean_token_accuracy": 0.22133278399705886,
      "num_tokens": 73000055.0,
      "step": 31870
    },
    {
      "entropy": 5.668669605255127,
      "epoch": 3.1509489916963225,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040247387781514065,
      "loss": 4.9303,
      "mean_token_accuracy": 0.22347757518291472,
      "num_tokens": 73010769.0,
      "step": 31875
    },
    {
      "entropy": 5.786803245544434,
      "epoch": 3.151443258204824,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040244479634992997,
      "loss": 5.0038,
      "mean_token_accuracy": 0.22169877737760543,
      "num_tokens": 73022032.0,
      "step": 31880
    },
    {
      "entropy": 5.720450973510742,
      "epoch": 3.1519375247133254,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040241571174955755,
      "loss": 4.9419,
      "mean_token_accuracy": 0.21972451806068422,
      "num_tokens": 73033877.0,
      "step": 31885
    },
    {
      "entropy": 5.683818674087524,
      "epoch": 3.1524317912218267,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004023866240147389,
      "loss": 4.8947,
      "mean_token_accuracy": 0.22308821529150008,
      "num_tokens": 73045670.0,
      "step": 31890
    },
    {
      "entropy": 5.650382328033447,
      "epoch": 3.152926057730328,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004023575331461894,
      "loss": 4.9318,
      "mean_token_accuracy": 0.2268572822213173,
      "num_tokens": 73056879.0,
      "step": 31895
    },
    {
      "entropy": 5.713074779510498,
      "epoch": 3.1534203242388297,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040232843914462493,
      "loss": 4.9806,
      "mean_token_accuracy": 0.21902749985456466,
      "num_tokens": 73068422.0,
      "step": 31900
    },
    {
      "entropy": 5.746976041793824,
      "epoch": 3.153914590747331,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004022993420107611,
      "loss": 4.8892,
      "mean_token_accuracy": 0.22467874139547347,
      "num_tokens": 73079883.0,
      "step": 31905
    },
    {
      "entropy": 5.737526369094849,
      "epoch": 3.154408857255832,
      "grad_norm": 1.328125,
      "learning_rate": 0.00040227024174531364,
      "loss": 4.9535,
      "mean_token_accuracy": 0.22818573713302612,
      "num_tokens": 73090832.0,
      "step": 31910
    },
    {
      "entropy": 5.703965139389038,
      "epoch": 3.154903123764334,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004022411383489985,
      "loss": 4.8958,
      "mean_token_accuracy": 0.23206118643283843,
      "num_tokens": 73102154.0,
      "step": 31915
    },
    {
      "entropy": 5.666731357574463,
      "epoch": 3.155397390272835,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004022120318225317,
      "loss": 4.9717,
      "mean_token_accuracy": 0.22813572138547897,
      "num_tokens": 73113530.0,
      "step": 31920
    },
    {
      "entropy": 5.77650556564331,
      "epoch": 3.1558916567813364,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040218292216662903,
      "loss": 5.0284,
      "mean_token_accuracy": 0.220971742272377,
      "num_tokens": 73124379.0,
      "step": 31925
    },
    {
      "entropy": 5.727950954437256,
      "epoch": 3.1563859232898377,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040215380938200693,
      "loss": 4.98,
      "mean_token_accuracy": 0.22099777907133103,
      "num_tokens": 73136406.0,
      "step": 31930
    },
    {
      "entropy": 5.797395038604736,
      "epoch": 3.1568801897983394,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040212469346938127,
      "loss": 5.0074,
      "mean_token_accuracy": 0.21926964521408082,
      "num_tokens": 73149302.0,
      "step": 31935
    },
    {
      "entropy": 5.7296020030975345,
      "epoch": 3.1573744563068407,
      "grad_norm": 1.25,
      "learning_rate": 0.0004020955744294685,
      "loss": 4.9673,
      "mean_token_accuracy": 0.21609448343515397,
      "num_tokens": 73160462.0,
      "step": 31940
    },
    {
      "entropy": 5.68043270111084,
      "epoch": 3.157868722815342,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040206645226298484,
      "loss": 4.8763,
      "mean_token_accuracy": 0.231180277466774,
      "num_tokens": 73171849.0,
      "step": 31945
    },
    {
      "entropy": 5.699740219116211,
      "epoch": 3.1583629893238436,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004020373269706468,
      "loss": 4.9097,
      "mean_token_accuracy": 0.22628739476203918,
      "num_tokens": 73182132.0,
      "step": 31950
    },
    {
      "entropy": 5.7765709400177006,
      "epoch": 3.158857255832345,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004020081985531706,
      "loss": 5.0026,
      "mean_token_accuracy": 0.21831277012825012,
      "num_tokens": 73194527.0,
      "step": 31955
    },
    {
      "entropy": 5.700388860702515,
      "epoch": 3.159351522340846,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004019790670112733,
      "loss": 4.9256,
      "mean_token_accuracy": 0.22562604248523713,
      "num_tokens": 73206104.0,
      "step": 31960
    },
    {
      "entropy": 5.685428857803345,
      "epoch": 3.1598457888493474,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040194993234567107,
      "loss": 4.9418,
      "mean_token_accuracy": 0.22462059706449508,
      "num_tokens": 73217665.0,
      "step": 31965
    },
    {
      "entropy": 5.718275880813598,
      "epoch": 3.160340055357849,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040192079455708083,
      "loss": 4.9629,
      "mean_token_accuracy": 0.2210962027311325,
      "num_tokens": 73229677.0,
      "step": 31970
    },
    {
      "entropy": 5.762109375,
      "epoch": 3.1608343218663504,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040189165364621943,
      "loss": 4.9928,
      "mean_token_accuracy": 0.21976627707481383,
      "num_tokens": 73240806.0,
      "step": 31975
    },
    {
      "entropy": 5.689442348480225,
      "epoch": 3.1613285883748516,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004018625096138036,
      "loss": 4.8893,
      "mean_token_accuracy": 0.2234673798084259,
      "num_tokens": 73252016.0,
      "step": 31980
    },
    {
      "entropy": 5.701041412353516,
      "epoch": 3.161822854883353,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040183336246055033,
      "loss": 4.9373,
      "mean_token_accuracy": 0.22504760771989823,
      "num_tokens": 73262790.0,
      "step": 31985
    },
    {
      "entropy": 5.800733804702759,
      "epoch": 3.1623171213918546,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004018042121871767,
      "loss": 5.0346,
      "mean_token_accuracy": 0.21860115230083466,
      "num_tokens": 73275424.0,
      "step": 31990
    },
    {
      "entropy": 5.716345930099488,
      "epoch": 3.162811387900356,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040177505879439984,
      "loss": 4.927,
      "mean_token_accuracy": 0.2281204491853714,
      "num_tokens": 73286790.0,
      "step": 31995
    },
    {
      "entropy": 5.7645610809326175,
      "epoch": 3.163305654408857,
      "grad_norm": 1.125,
      "learning_rate": 0.0004017459022829368,
      "loss": 5.0627,
      "mean_token_accuracy": 0.21259057223796846,
      "num_tokens": 73298876.0,
      "step": 32000
    },
    {
      "entropy": 5.733574295043946,
      "epoch": 3.163799920917359,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040171674265350487,
      "loss": 4.9594,
      "mean_token_accuracy": 0.22140556573867798,
      "num_tokens": 73310635.0,
      "step": 32005
    },
    {
      "entropy": 5.7724817276000975,
      "epoch": 3.16429418742586,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004016875799068215,
      "loss": 5.0216,
      "mean_token_accuracy": 0.21486473232507705,
      "num_tokens": 73321998.0,
      "step": 32010
    },
    {
      "entropy": 5.697901487350464,
      "epoch": 3.1647884539343614,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004016584140436039,
      "loss": 4.9225,
      "mean_token_accuracy": 0.23045843243598937,
      "num_tokens": 73332979.0,
      "step": 32015
    },
    {
      "entropy": 5.7338958263397215,
      "epoch": 3.1652827204428626,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004016292450645698,
      "loss": 4.9886,
      "mean_token_accuracy": 0.21328095942735673,
      "num_tokens": 73344405.0,
      "step": 32020
    },
    {
      "entropy": 5.731864786148071,
      "epoch": 3.1657769869513643,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040160007297043655,
      "loss": 4.9695,
      "mean_token_accuracy": 0.22475258857011796,
      "num_tokens": 73356291.0,
      "step": 32025
    },
    {
      "entropy": 5.761966848373413,
      "epoch": 3.1662712534598656,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040157089776192184,
      "loss": 4.9771,
      "mean_token_accuracy": 0.22883784621953965,
      "num_tokens": 73368299.0,
      "step": 32030
    },
    {
      "entropy": 5.76413140296936,
      "epoch": 3.166765519968367,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040154171943974345,
      "loss": 5.0274,
      "mean_token_accuracy": 0.22084034979343414,
      "num_tokens": 73380527.0,
      "step": 32035
    },
    {
      "entropy": 5.744593286514283,
      "epoch": 3.167259786476868,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040151253800461913,
      "loss": 4.9732,
      "mean_token_accuracy": 0.22335683554410934,
      "num_tokens": 73392547.0,
      "step": 32040
    },
    {
      "entropy": 5.6888103008270265,
      "epoch": 3.16775405298537,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004014833534572667,
      "loss": 4.9066,
      "mean_token_accuracy": 0.23269703090190888,
      "num_tokens": 73404293.0,
      "step": 32045
    },
    {
      "entropy": 5.7477500438690186,
      "epoch": 3.168248319493871,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004014541657984043,
      "loss": 4.8751,
      "mean_token_accuracy": 0.22786087840795516,
      "num_tokens": 73414581.0,
      "step": 32050
    },
    {
      "entropy": 5.715932130813599,
      "epoch": 3.1687425860023724,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040142497502874963,
      "loss": 5.0292,
      "mean_token_accuracy": 0.2114308312535286,
      "num_tokens": 73426485.0,
      "step": 32055
    },
    {
      "entropy": 5.7287070751190186,
      "epoch": 3.169236852510874,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040139578114902104,
      "loss": 5.0479,
      "mean_token_accuracy": 0.21972786039113998,
      "num_tokens": 73438481.0,
      "step": 32060
    },
    {
      "entropy": 5.782337951660156,
      "epoch": 3.1697311190193753,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040136658415993663,
      "loss": 4.9421,
      "mean_token_accuracy": 0.21798247396945952,
      "num_tokens": 73448552.0,
      "step": 32065
    },
    {
      "entropy": 5.727556324005127,
      "epoch": 3.1702253855278766,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004013373840622146,
      "loss": 4.9239,
      "mean_token_accuracy": 0.2203081354498863,
      "num_tokens": 73459354.0,
      "step": 32070
    },
    {
      "entropy": 5.72588267326355,
      "epoch": 3.170719652036378,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00040130818085657333,
      "loss": 4.9402,
      "mean_token_accuracy": 0.2236169919371605,
      "num_tokens": 73470718.0,
      "step": 32075
    },
    {
      "entropy": 5.768819093704224,
      "epoch": 3.1712139185448796,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004012789745437313,
      "loss": 5.0525,
      "mean_token_accuracy": 0.21579015851020814,
      "num_tokens": 73483133.0,
      "step": 32080
    },
    {
      "entropy": 5.77171745300293,
      "epoch": 3.171708185053381,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004012497651244068,
      "loss": 4.9521,
      "mean_token_accuracy": 0.2193336695432663,
      "num_tokens": 73494675.0,
      "step": 32085
    },
    {
      "entropy": 5.714121675491333,
      "epoch": 3.172202451561882,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040122055259931856,
      "loss": 4.9367,
      "mean_token_accuracy": 0.23067618310451507,
      "num_tokens": 73505412.0,
      "step": 32090
    },
    {
      "entropy": 5.63411054611206,
      "epoch": 3.172696718070384,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040119133696918515,
      "loss": 4.8395,
      "mean_token_accuracy": 0.23873535841703414,
      "num_tokens": 73515835.0,
      "step": 32095
    },
    {
      "entropy": 5.72991418838501,
      "epoch": 3.173190984578885,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004011621182347252,
      "loss": 4.9432,
      "mean_token_accuracy": 0.22705233693122864,
      "num_tokens": 73526943.0,
      "step": 32100
    },
    {
      "entropy": 5.752754259109497,
      "epoch": 3.1736852510873863,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040113289639665765,
      "loss": 4.9572,
      "mean_token_accuracy": 0.22403370589017868,
      "num_tokens": 73540076.0,
      "step": 32105
    },
    {
      "entropy": 5.8096246242523195,
      "epoch": 3.1741795175958876,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040110367145570125,
      "loss": 5.0193,
      "mean_token_accuracy": 0.22183199524879454,
      "num_tokens": 73552265.0,
      "step": 32110
    },
    {
      "entropy": 5.724063158035278,
      "epoch": 3.1746737841043893,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000401074443412575,
      "loss": 4.946,
      "mean_token_accuracy": 0.22147433012723922,
      "num_tokens": 73562581.0,
      "step": 32115
    },
    {
      "entropy": 5.695050048828125,
      "epoch": 3.1751680506128905,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00040104521226799777,
      "loss": 4.9361,
      "mean_token_accuracy": 0.23346880376338958,
      "num_tokens": 73573478.0,
      "step": 32120
    },
    {
      "entropy": 5.7909619331359865,
      "epoch": 3.175662317121392,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040101597802268883,
      "loss": 4.9589,
      "mean_token_accuracy": 0.2238469123840332,
      "num_tokens": 73584639.0,
      "step": 32125
    },
    {
      "entropy": 5.760508489608765,
      "epoch": 3.176156583629893,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004009867406773673,
      "loss": 4.9987,
      "mean_token_accuracy": 0.22034768164157867,
      "num_tokens": 73596084.0,
      "step": 32130
    },
    {
      "entropy": 5.710902690887451,
      "epoch": 3.1766508501383948,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040095750023275243,
      "loss": 4.9774,
      "mean_token_accuracy": 0.22693269550800324,
      "num_tokens": 73606551.0,
      "step": 32135
    },
    {
      "entropy": 5.736982583999634,
      "epoch": 3.177145116646896,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004009282566895635,
      "loss": 4.8987,
      "mean_token_accuracy": 0.23246588110923766,
      "num_tokens": 73616134.0,
      "step": 32140
    },
    {
      "entropy": 5.6890191555023195,
      "epoch": 3.1776393831553973,
      "grad_norm": 1.125,
      "learning_rate": 0.0004008990100485199,
      "loss": 4.8554,
      "mean_token_accuracy": 0.22833207845687867,
      "num_tokens": 73626813.0,
      "step": 32145
    },
    {
      "entropy": 5.675510358810425,
      "epoch": 3.1781336496638986,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040086976031034104,
      "loss": 4.9646,
      "mean_token_accuracy": 0.2279843270778656,
      "num_tokens": 73638205.0,
      "step": 32150
    },
    {
      "entropy": 5.8437971591949465,
      "epoch": 3.1786279161724003,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040084050747574666,
      "loss": 5.0818,
      "mean_token_accuracy": 0.21092552691698074,
      "num_tokens": 73652276.0,
      "step": 32155
    },
    {
      "entropy": 5.785452890396118,
      "epoch": 3.1791221826809015,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004008112515454562,
      "loss": 5.0645,
      "mean_token_accuracy": 0.21881006956100463,
      "num_tokens": 73665516.0,
      "step": 32160
    },
    {
      "entropy": 5.743360948562622,
      "epoch": 3.179616449189403,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004007819925201895,
      "loss": 4.9202,
      "mean_token_accuracy": 0.23202859461307526,
      "num_tokens": 73675783.0,
      "step": 32165
    },
    {
      "entropy": 5.76856837272644,
      "epoch": 3.1801107156979045,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004007527304006663,
      "loss": 5.0024,
      "mean_token_accuracy": 0.21402398943901063,
      "num_tokens": 73687853.0,
      "step": 32170
    },
    {
      "entropy": 5.697969388961792,
      "epoch": 3.1806049822064058,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040072346518760634,
      "loss": 4.908,
      "mean_token_accuracy": 0.23285412788391113,
      "num_tokens": 73699189.0,
      "step": 32175
    },
    {
      "entropy": 5.807528257369995,
      "epoch": 3.181099248714907,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040069419688172967,
      "loss": 5.0841,
      "mean_token_accuracy": 0.21684761196374894,
      "num_tokens": 73711003.0,
      "step": 32180
    },
    {
      "entropy": 5.785439777374267,
      "epoch": 3.1815935152234083,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040066492548375627,
      "loss": 5.0305,
      "mean_token_accuracy": 0.21365318894386293,
      "num_tokens": 73723311.0,
      "step": 32185
    },
    {
      "entropy": 5.8026378631591795,
      "epoch": 3.18208778173191,
      "grad_norm": 1.25,
      "learning_rate": 0.00040063565099440616,
      "loss": 5.0145,
      "mean_token_accuracy": 0.2226735144853592,
      "num_tokens": 73734155.0,
      "step": 32190
    },
    {
      "entropy": 5.678349542617798,
      "epoch": 3.1825820482404112,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004006063734143996,
      "loss": 4.9263,
      "mean_token_accuracy": 0.22843949794769286,
      "num_tokens": 73745289.0,
      "step": 32195
    },
    {
      "entropy": 5.6801502227783205,
      "epoch": 3.1830763147489125,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004005770927444567,
      "loss": 4.9553,
      "mean_token_accuracy": 0.2269407406449318,
      "num_tokens": 73756955.0,
      "step": 32200
    },
    {
      "entropy": 5.735558128356933,
      "epoch": 3.183570581257414,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004005478089852979,
      "loss": 4.9433,
      "mean_token_accuracy": 0.21973614245653153,
      "num_tokens": 73768676.0,
      "step": 32205
    },
    {
      "entropy": 5.694872236251831,
      "epoch": 3.1840648477659155,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004005185221376435,
      "loss": 4.932,
      "mean_token_accuracy": 0.2317672148346901,
      "num_tokens": 73779505.0,
      "step": 32210
    },
    {
      "entropy": 5.710838317871094,
      "epoch": 3.1845591142744167,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040048923220221397,
      "loss": 4.9809,
      "mean_token_accuracy": 0.22104528546333313,
      "num_tokens": 73790785.0,
      "step": 32215
    },
    {
      "entropy": 5.747603845596314,
      "epoch": 3.185053380782918,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040045993917972985,
      "loss": 4.9421,
      "mean_token_accuracy": 0.22288586050271988,
      "num_tokens": 73802689.0,
      "step": 32220
    },
    {
      "entropy": 5.745974349975586,
      "epoch": 3.1855476472914197,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004004306430709118,
      "loss": 4.9492,
      "mean_token_accuracy": 0.22333626300096512,
      "num_tokens": 73815029.0,
      "step": 32225
    },
    {
      "entropy": 5.769425582885742,
      "epoch": 3.186041913799921,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040040134387648045,
      "loss": 4.9817,
      "mean_token_accuracy": 0.22192592471837996,
      "num_tokens": 73825509.0,
      "step": 32230
    },
    {
      "entropy": 5.77627534866333,
      "epoch": 3.1865361803084222,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004003720415971566,
      "loss": 4.97,
      "mean_token_accuracy": 0.22519028931856155,
      "num_tokens": 73836014.0,
      "step": 32235
    },
    {
      "entropy": 5.686406850814819,
      "epoch": 3.1870304468169235,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004003427362336611,
      "loss": 4.9045,
      "mean_token_accuracy": 0.22717816829681398,
      "num_tokens": 73847044.0,
      "step": 32240
    },
    {
      "entropy": 5.700253057479858,
      "epoch": 3.187524713325425,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040031342778671476,
      "loss": 4.9533,
      "mean_token_accuracy": 0.22426968663930893,
      "num_tokens": 73859748.0,
      "step": 32245
    },
    {
      "entropy": 5.6867982864379885,
      "epoch": 3.1880189798339265,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004002841162570387,
      "loss": 4.9176,
      "mean_token_accuracy": 0.2298780471086502,
      "num_tokens": 73870450.0,
      "step": 32250
    },
    {
      "entropy": 5.783068704605102,
      "epoch": 3.1885132463424277,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040025480164535405,
      "loss": 5.0382,
      "mean_token_accuracy": 0.2143593117594719,
      "num_tokens": 73882533.0,
      "step": 32255
    },
    {
      "entropy": 5.756817245483399,
      "epoch": 3.1890075128509294,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040022548395238173,
      "loss": 4.9896,
      "mean_token_accuracy": 0.2241203874349594,
      "num_tokens": 73894247.0,
      "step": 32260
    },
    {
      "entropy": 5.734116744995117,
      "epoch": 3.1895017793594307,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004001961631788431,
      "loss": 4.9377,
      "mean_token_accuracy": 0.23192739635705947,
      "num_tokens": 73905246.0,
      "step": 32265
    },
    {
      "entropy": 5.7124857902526855,
      "epoch": 3.189996045867932,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040016683932545937,
      "loss": 4.9291,
      "mean_token_accuracy": 0.2254141613841057,
      "num_tokens": 73915906.0,
      "step": 32270
    },
    {
      "entropy": 5.689729261398315,
      "epoch": 3.190490312376433,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000400137512392952,
      "loss": 4.958,
      "mean_token_accuracy": 0.21993598639965056,
      "num_tokens": 73928629.0,
      "step": 32275
    },
    {
      "entropy": 5.8052743434906,
      "epoch": 3.190984578884935,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040010818238204256,
      "loss": 5.0152,
      "mean_token_accuracy": 0.2224835366010666,
      "num_tokens": 73940474.0,
      "step": 32280
    },
    {
      "entropy": 5.714233779907227,
      "epoch": 3.191478845393436,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040007884929345224,
      "loss": 4.9006,
      "mean_token_accuracy": 0.22905313670635224,
      "num_tokens": 73951636.0,
      "step": 32285
    },
    {
      "entropy": 5.6802863597869875,
      "epoch": 3.1919731119019374,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004000495131279029,
      "loss": 4.8975,
      "mean_token_accuracy": 0.2281394973397255,
      "num_tokens": 73963253.0,
      "step": 32290
    },
    {
      "entropy": 5.651004266738892,
      "epoch": 3.1924673784104387,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004000201738861162,
      "loss": 4.8854,
      "mean_token_accuracy": 0.22926275432109833,
      "num_tokens": 73974215.0,
      "step": 32295
    },
    {
      "entropy": 5.716013336181641,
      "epoch": 3.1929616449189404,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003999908315688137,
      "loss": 4.9956,
      "mean_token_accuracy": 0.22173261046409606,
      "num_tokens": 73986127.0,
      "step": 32300
    },
    {
      "entropy": 5.7708740234375,
      "epoch": 3.1934559114274417,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00039996148617671743,
      "loss": 4.9636,
      "mean_token_accuracy": 0.22691153585910798,
      "num_tokens": 73996998.0,
      "step": 32305
    },
    {
      "entropy": 5.757706356048584,
      "epoch": 3.193950177935943,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003999321377105492,
      "loss": 4.9446,
      "mean_token_accuracy": 0.2277391731739044,
      "num_tokens": 74009432.0,
      "step": 32310
    },
    {
      "entropy": 5.67088041305542,
      "epoch": 3.1944444444444446,
      "grad_norm": 1.15625,
      "learning_rate": 0.000399902786171031,
      "loss": 4.9283,
      "mean_token_accuracy": 0.2227945163846016,
      "num_tokens": 74021242.0,
      "step": 32315
    },
    {
      "entropy": 5.738876914978027,
      "epoch": 3.194938710952946,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039987343155888484,
      "loss": 4.9714,
      "mean_token_accuracy": 0.2153807118535042,
      "num_tokens": 74032865.0,
      "step": 32320
    },
    {
      "entropy": 5.717230653762817,
      "epoch": 3.195432977461447,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000399844073874833,
      "loss": 4.9181,
      "mean_token_accuracy": 0.23045008927583693,
      "num_tokens": 74044505.0,
      "step": 32325
    },
    {
      "entropy": 5.727366733551025,
      "epoch": 3.1959272439699484,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039981471311959746,
      "loss": 4.9665,
      "mean_token_accuracy": 0.21949754357337953,
      "num_tokens": 74056551.0,
      "step": 32330
    },
    {
      "entropy": 5.744142961502075,
      "epoch": 3.19642151047845,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039978534929390066,
      "loss": 4.9351,
      "mean_token_accuracy": 0.22213992327451706,
      "num_tokens": 74066746.0,
      "step": 32335
    },
    {
      "entropy": 5.733849954605103,
      "epoch": 3.1969157769869514,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039975598239846494,
      "loss": 4.9941,
      "mean_token_accuracy": 0.22086290419101715,
      "num_tokens": 74078493.0,
      "step": 32340
    },
    {
      "entropy": 5.746189832687378,
      "epoch": 3.1974100434954527,
      "grad_norm": 1.25,
      "learning_rate": 0.00039972661243401267,
      "loss": 4.9133,
      "mean_token_accuracy": 0.22525355815887452,
      "num_tokens": 74089484.0,
      "step": 32345
    },
    {
      "entropy": 5.7634727478027346,
      "epoch": 3.1979043100039544,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003996972394012664,
      "loss": 4.9648,
      "mean_token_accuracy": 0.22273753434419633,
      "num_tokens": 74101038.0,
      "step": 32350
    },
    {
      "entropy": 5.806222295761108,
      "epoch": 3.1983985765124556,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003996678633009486,
      "loss": 5.0473,
      "mean_token_accuracy": 0.21717166155576706,
      "num_tokens": 74112708.0,
      "step": 32355
    },
    {
      "entropy": 5.7628068923950195,
      "epoch": 3.198892843020957,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003996384841337822,
      "loss": 4.9921,
      "mean_token_accuracy": 0.21744083166122435,
      "num_tokens": 74123899.0,
      "step": 32360
    },
    {
      "entropy": 5.680123519897461,
      "epoch": 3.199387109529458,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003996091019004896,
      "loss": 4.9554,
      "mean_token_accuracy": 0.2233150541782379,
      "num_tokens": 74136025.0,
      "step": 32365
    },
    {
      "entropy": 5.7972798347473145,
      "epoch": 3.19988137603796,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003995797166017938,
      "loss": 5.1058,
      "mean_token_accuracy": 0.21531136631965636,
      "num_tokens": 74149140.0,
      "step": 32370
    },
    {
      "entropy": 5.793102025985718,
      "epoch": 3.200375642546461,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039955032823841766,
      "loss": 4.9598,
      "mean_token_accuracy": 0.22249247133731842,
      "num_tokens": 74160849.0,
      "step": 32375
    },
    {
      "entropy": 5.746348476409912,
      "epoch": 3.2008699090549624,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039952093681108416,
      "loss": 4.9997,
      "mean_token_accuracy": 0.22070683985948564,
      "num_tokens": 74171678.0,
      "step": 32380
    },
    {
      "entropy": 5.665137910842896,
      "epoch": 3.2013641755634636,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003994915423205162,
      "loss": 4.9248,
      "mean_token_accuracy": 0.22837164103984833,
      "num_tokens": 74183000.0,
      "step": 32385
    },
    {
      "entropy": 5.812749290466309,
      "epoch": 3.2018584420719653,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003994621447674371,
      "loss": 5.0308,
      "mean_token_accuracy": 0.21924428641796112,
      "num_tokens": 74195495.0,
      "step": 32390
    },
    {
      "entropy": 5.752200603485107,
      "epoch": 3.2023527085804666,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003994327441525699,
      "loss": 4.983,
      "mean_token_accuracy": 0.2202370971441269,
      "num_tokens": 74206023.0,
      "step": 32395
    },
    {
      "entropy": 5.729682350158692,
      "epoch": 3.202846975088968,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003994033404766378,
      "loss": 4.993,
      "mean_token_accuracy": 0.2224993422627449,
      "num_tokens": 74217943.0,
      "step": 32400
    },
    {
      "entropy": 5.7644007205963135,
      "epoch": 3.203341241597469,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003993739337403643,
      "loss": 5.0169,
      "mean_token_accuracy": 0.21626025140285493,
      "num_tokens": 74228142.0,
      "step": 32405
    },
    {
      "entropy": 5.76966233253479,
      "epoch": 3.203835508105971,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003993445239444728,
      "loss": 5.0015,
      "mean_token_accuracy": 0.22228449136018752,
      "num_tokens": 74239692.0,
      "step": 32410
    },
    {
      "entropy": 5.78749713897705,
      "epoch": 3.204329774614472,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003993151110896866,
      "loss": 4.9779,
      "mean_token_accuracy": 0.21892784386873246,
      "num_tokens": 74251010.0,
      "step": 32415
    },
    {
      "entropy": 5.672426223754883,
      "epoch": 3.2048240411229734,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003992856951767294,
      "loss": 4.8657,
      "mean_token_accuracy": 0.22561863511800767,
      "num_tokens": 74262579.0,
      "step": 32420
    },
    {
      "entropy": 5.710741901397705,
      "epoch": 3.205318307631475,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00039925627620632486,
      "loss": 4.8851,
      "mean_token_accuracy": 0.23000517338514329,
      "num_tokens": 74272833.0,
      "step": 32425
    },
    {
      "entropy": 5.783285427093506,
      "epoch": 3.2058125741399763,
      "grad_norm": 1.125,
      "learning_rate": 0.0003992268541791967,
      "loss": 5.0206,
      "mean_token_accuracy": 0.2168182298541069,
      "num_tokens": 74284565.0,
      "step": 32430
    },
    {
      "entropy": 5.655234861373901,
      "epoch": 3.2063068406484776,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003991974290960685,
      "loss": 4.8547,
      "mean_token_accuracy": 0.2417011260986328,
      "num_tokens": 74295473.0,
      "step": 32435
    },
    {
      "entropy": 5.8008466243743895,
      "epoch": 3.206801107156979,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00039916800095766443,
      "loss": 5.0512,
      "mean_token_accuracy": 0.20862850546836853,
      "num_tokens": 74305688.0,
      "step": 32440
    },
    {
      "entropy": 5.708681583404541,
      "epoch": 3.2072953736654806,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039913856976470823,
      "loss": 4.9252,
      "mean_token_accuracy": 0.22708481699228286,
      "num_tokens": 74318268.0,
      "step": 32445
    },
    {
      "entropy": 5.771266269683838,
      "epoch": 3.207789640173982,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003991091355179239,
      "loss": 5.0295,
      "mean_token_accuracy": 0.2138544112443924,
      "num_tokens": 74330310.0,
      "step": 32450
    },
    {
      "entropy": 5.74018874168396,
      "epoch": 3.208283906682483,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003990796982180356,
      "loss": 5.0061,
      "mean_token_accuracy": 0.22078711092472075,
      "num_tokens": 74342040.0,
      "step": 32455
    },
    {
      "entropy": 5.832875204086304,
      "epoch": 3.208778173190985,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039905025786576755,
      "loss": 5.0533,
      "mean_token_accuracy": 0.2120827093720436,
      "num_tokens": 74354502.0,
      "step": 32460
    },
    {
      "entropy": 5.820251560211181,
      "epoch": 3.209272439699486,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003990208144618438,
      "loss": 4.9692,
      "mean_token_accuracy": 0.2199249669909477,
      "num_tokens": 74366100.0,
      "step": 32465
    },
    {
      "entropy": 5.846017742156983,
      "epoch": 3.2097667062079873,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00039899136800698883,
      "loss": 5.0609,
      "mean_token_accuracy": 0.20613151043653488,
      "num_tokens": 74377627.0,
      "step": 32470
    },
    {
      "entropy": 5.682811641693116,
      "epoch": 3.2102609727164886,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000398961918501927,
      "loss": 4.9458,
      "mean_token_accuracy": 0.22553450912237166,
      "num_tokens": 74389689.0,
      "step": 32475
    },
    {
      "entropy": 5.7804991722106935,
      "epoch": 3.2107552392249903,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003989324659473827,
      "loss": 5.0777,
      "mean_token_accuracy": 0.20669523924589156,
      "num_tokens": 74399902.0,
      "step": 32480
    },
    {
      "entropy": 5.790475654602051,
      "epoch": 3.2112495057334915,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039890301034408056,
      "loss": 5.0386,
      "mean_token_accuracy": 0.21036529839038848,
      "num_tokens": 74410935.0,
      "step": 32485
    },
    {
      "entropy": 5.755534696578979,
      "epoch": 3.211743772241993,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003988735516927451,
      "loss": 4.9262,
      "mean_token_accuracy": 0.2284962147474289,
      "num_tokens": 74421850.0,
      "step": 32490
    },
    {
      "entropy": 5.719126033782959,
      "epoch": 3.212238038750494,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039884408999410107,
      "loss": 4.9513,
      "mean_token_accuracy": 0.22695187628269195,
      "num_tokens": 74434104.0,
      "step": 32495
    },
    {
      "entropy": 5.772160816192627,
      "epoch": 3.212732305258996,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003988146252488732,
      "loss": 5.033,
      "mean_token_accuracy": 0.21827645599842072,
      "num_tokens": 74446738.0,
      "step": 32500
    },
    {
      "entropy": 5.871346950531006,
      "epoch": 3.213226571767497,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003987851574577863,
      "loss": 5.0991,
      "mean_token_accuracy": 0.21547407060861587,
      "num_tokens": 74458202.0,
      "step": 32505
    },
    {
      "entropy": 5.795695209503174,
      "epoch": 3.2137208382759983,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003987556866215654,
      "loss": 5.0258,
      "mean_token_accuracy": 0.22029586732387543,
      "num_tokens": 74470349.0,
      "step": 32510
    },
    {
      "entropy": 5.705416631698609,
      "epoch": 3.2142151047845,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003987262127409354,
      "loss": 4.9093,
      "mean_token_accuracy": 0.22899301946163178,
      "num_tokens": 74482569.0,
      "step": 32515
    },
    {
      "entropy": 5.73810305595398,
      "epoch": 3.2147093712930013,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003986967358166213,
      "loss": 4.9386,
      "mean_token_accuracy": 0.2188004046678543,
      "num_tokens": 74495425.0,
      "step": 32520
    },
    {
      "entropy": 5.775293922424316,
      "epoch": 3.2152036378015025,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039866725584934834,
      "loss": 5.0254,
      "mean_token_accuracy": 0.21586144268512725,
      "num_tokens": 74506429.0,
      "step": 32525
    },
    {
      "entropy": 5.808415269851684,
      "epoch": 3.215697904310004,
      "grad_norm": 1.265625,
      "learning_rate": 0.00039863777283984174,
      "loss": 5.0123,
      "mean_token_accuracy": 0.22003390938043593,
      "num_tokens": 74517426.0,
      "step": 32530
    },
    {
      "entropy": 5.784921789169312,
      "epoch": 3.2161921708185055,
      "grad_norm": 1.328125,
      "learning_rate": 0.00039860828678882674,
      "loss": 4.9941,
      "mean_token_accuracy": 0.2211372196674347,
      "num_tokens": 74528777.0,
      "step": 32535
    },
    {
      "entropy": 5.725776243209839,
      "epoch": 3.2166864373270068,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003985787976970287,
      "loss": 4.9587,
      "mean_token_accuracy": 0.21525656878948213,
      "num_tokens": 74540384.0,
      "step": 32540
    },
    {
      "entropy": 5.7518346309661865,
      "epoch": 3.217180703835508,
      "grad_norm": 1.203125,
      "learning_rate": 0.000398549305565173,
      "loss": 4.8943,
      "mean_token_accuracy": 0.22654535323381425,
      "num_tokens": 74551698.0,
      "step": 32545
    },
    {
      "entropy": 5.750669860839844,
      "epoch": 3.2176749703440093,
      "grad_norm": 1.125,
      "learning_rate": 0.00039851981039398527,
      "loss": 4.9645,
      "mean_token_accuracy": 0.22653490155935288,
      "num_tokens": 74563467.0,
      "step": 32550
    },
    {
      "entropy": 5.656029796600341,
      "epoch": 3.218169236852511,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039849031218419114,
      "loss": 4.9057,
      "mean_token_accuracy": 0.23221013098955154,
      "num_tokens": 74575341.0,
      "step": 32555
    },
    {
      "entropy": 5.669033527374268,
      "epoch": 3.2186635033610123,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000398460810936516,
      "loss": 4.8297,
      "mean_token_accuracy": 0.23788087517023088,
      "num_tokens": 74587706.0,
      "step": 32560
    },
    {
      "entropy": 5.746804523468017,
      "epoch": 3.2191577698695135,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039843130665168585,
      "loss": 4.9628,
      "mean_token_accuracy": 0.22396714836359025,
      "num_tokens": 74599310.0,
      "step": 32565
    },
    {
      "entropy": 5.75776515007019,
      "epoch": 3.2196520363780152,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039840179933042644,
      "loss": 5.0728,
      "mean_token_accuracy": 0.2160360500216484,
      "num_tokens": 74610436.0,
      "step": 32570
    },
    {
      "entropy": 5.835077810287475,
      "epoch": 3.2201463028865165,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039837228897346357,
      "loss": 4.9722,
      "mean_token_accuracy": 0.21848828345537186,
      "num_tokens": 74621788.0,
      "step": 32575
    },
    {
      "entropy": 5.785645771026611,
      "epoch": 3.2206405693950177,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039834277558152326,
      "loss": 4.9865,
      "mean_token_accuracy": 0.22078637927770614,
      "num_tokens": 74633345.0,
      "step": 32580
    },
    {
      "entropy": 5.730307722091675,
      "epoch": 3.221134835903519,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003983132591553315,
      "loss": 5.0041,
      "mean_token_accuracy": 0.22205347716808319,
      "num_tokens": 74646217.0,
      "step": 32585
    },
    {
      "entropy": 5.751239347457886,
      "epoch": 3.2216291024120207,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003982837396956145,
      "loss": 4.9803,
      "mean_token_accuracy": 0.22730713188648224,
      "num_tokens": 74657935.0,
      "step": 32590
    },
    {
      "entropy": 5.738717842102051,
      "epoch": 3.222123368920522,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00039825421720309835,
      "loss": 5.0015,
      "mean_token_accuracy": 0.22322916090488434,
      "num_tokens": 74668547.0,
      "step": 32595
    },
    {
      "entropy": 5.809048748016357,
      "epoch": 3.2226176354290232,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003982246916785093,
      "loss": 5.0402,
      "mean_token_accuracy": 0.21314360499382018,
      "num_tokens": 74680013.0,
      "step": 32600
    },
    {
      "entropy": 5.766875171661377,
      "epoch": 3.223111901937525,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003981951631225738,
      "loss": 4.9479,
      "mean_token_accuracy": 0.22897373735904694,
      "num_tokens": 74689927.0,
      "step": 32605
    },
    {
      "entropy": 5.7854045867919925,
      "epoch": 3.223606168446026,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003981656315360181,
      "loss": 4.9853,
      "mean_token_accuracy": 0.22221137583255768,
      "num_tokens": 74701854.0,
      "step": 32610
    },
    {
      "entropy": 5.713367986679077,
      "epoch": 3.2241004349545275,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039813609691956883,
      "loss": 4.9554,
      "mean_token_accuracy": 0.22234144508838655,
      "num_tokens": 74713974.0,
      "step": 32615
    },
    {
      "entropy": 5.820211362838745,
      "epoch": 3.2245947014630287,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003981065592739524,
      "loss": 5.0454,
      "mean_token_accuracy": 0.21344489455223084,
      "num_tokens": 74725367.0,
      "step": 32620
    },
    {
      "entropy": 5.812423181533814,
      "epoch": 3.2250889679715304,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039807701859989554,
      "loss": 4.9683,
      "mean_token_accuracy": 0.21789056658744813,
      "num_tokens": 74736801.0,
      "step": 32625
    },
    {
      "entropy": 5.703966760635376,
      "epoch": 3.2255832344800317,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039804747489812495,
      "loss": 4.9757,
      "mean_token_accuracy": 0.22358919084072112,
      "num_tokens": 74747949.0,
      "step": 32630
    },
    {
      "entropy": 5.6846458435058596,
      "epoch": 3.226077500988533,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003980179281693674,
      "loss": 4.9441,
      "mean_token_accuracy": 0.22936786115169525,
      "num_tokens": 74759985.0,
      "step": 32635
    },
    {
      "entropy": 5.728292417526245,
      "epoch": 3.2265717674970342,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039798837841434966,
      "loss": 4.9242,
      "mean_token_accuracy": 0.22488351613283158,
      "num_tokens": 74770981.0,
      "step": 32640
    },
    {
      "entropy": 5.729997301101685,
      "epoch": 3.227066034005536,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039795882563379876,
      "loss": 4.9031,
      "mean_token_accuracy": 0.23429159224033355,
      "num_tokens": 74783120.0,
      "step": 32645
    },
    {
      "entropy": 5.6912806034088135,
      "epoch": 3.227560300514037,
      "grad_norm": 1.25,
      "learning_rate": 0.0003979292698284417,
      "loss": 4.8996,
      "mean_token_accuracy": 0.2270379051566124,
      "num_tokens": 74793492.0,
      "step": 32650
    },
    {
      "entropy": 5.6635195255279545,
      "epoch": 3.2280545670225385,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039789971099900545,
      "loss": 4.9517,
      "mean_token_accuracy": 0.22720779478549957,
      "num_tokens": 74805333.0,
      "step": 32655
    },
    {
      "entropy": 5.716989755630493,
      "epoch": 3.2285488335310397,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003978701491462174,
      "loss": 4.9151,
      "mean_token_accuracy": 0.22994406521320343,
      "num_tokens": 74816269.0,
      "step": 32660
    },
    {
      "entropy": 5.755623626708984,
      "epoch": 3.2290431000395414,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003978405842708045,
      "loss": 4.931,
      "mean_token_accuracy": 0.23005022704601288,
      "num_tokens": 74826901.0,
      "step": 32665
    },
    {
      "entropy": 5.735284757614136,
      "epoch": 3.2295373665480427,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039781101637349417,
      "loss": 4.9718,
      "mean_token_accuracy": 0.2257653221487999,
      "num_tokens": 74837373.0,
      "step": 32670
    },
    {
      "entropy": 5.748017406463623,
      "epoch": 3.230031633056544,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039778144545501374,
      "loss": 4.94,
      "mean_token_accuracy": 0.224665567278862,
      "num_tokens": 74848609.0,
      "step": 32675
    },
    {
      "entropy": 5.70569953918457,
      "epoch": 3.2305258995650457,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003977518715160908,
      "loss": 4.9623,
      "mean_token_accuracy": 0.22923813313245772,
      "num_tokens": 74859953.0,
      "step": 32680
    },
    {
      "entropy": 5.731884527206421,
      "epoch": 3.231020166073547,
      "grad_norm": 1.390625,
      "learning_rate": 0.00039772229455745285,
      "loss": 4.9716,
      "mean_token_accuracy": 0.22389605045318603,
      "num_tokens": 74872173.0,
      "step": 32685
    },
    {
      "entropy": 5.717567777633667,
      "epoch": 3.231514432582048,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039769271457982733,
      "loss": 4.9196,
      "mean_token_accuracy": 0.22607224732637404,
      "num_tokens": 74884060.0,
      "step": 32690
    },
    {
      "entropy": 5.747975778579712,
      "epoch": 3.2320086990905494,
      "grad_norm": 1.25,
      "learning_rate": 0.0003976631315839419,
      "loss": 4.9599,
      "mean_token_accuracy": 0.2199810415506363,
      "num_tokens": 74895660.0,
      "step": 32695
    },
    {
      "entropy": 5.751423120498657,
      "epoch": 3.232502965599051,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003976335455705245,
      "loss": 4.9819,
      "mean_token_accuracy": 0.22002631276845933,
      "num_tokens": 74907228.0,
      "step": 32700
    },
    {
      "entropy": 5.72869119644165,
      "epoch": 3.2329972321075524,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003976039565403029,
      "loss": 5.0019,
      "mean_token_accuracy": 0.21891435533761977,
      "num_tokens": 74919062.0,
      "step": 32705
    },
    {
      "entropy": 5.694114112854004,
      "epoch": 3.2334914986160537,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00039757436449400484,
      "loss": 4.9281,
      "mean_token_accuracy": 0.2266903728246689,
      "num_tokens": 74930551.0,
      "step": 32710
    },
    {
      "entropy": 5.754921388626099,
      "epoch": 3.2339857651245554,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003975447694323585,
      "loss": 4.9184,
      "mean_token_accuracy": 0.23049383908510207,
      "num_tokens": 74941219.0,
      "step": 32715
    },
    {
      "entropy": 5.777109813690186,
      "epoch": 3.2344800316330566,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003975151713560918,
      "loss": 4.9921,
      "mean_token_accuracy": 0.22407320737838746,
      "num_tokens": 74951857.0,
      "step": 32720
    },
    {
      "entropy": 5.6656012535095215,
      "epoch": 3.234974298141558,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003974855702659328,
      "loss": 4.8247,
      "mean_token_accuracy": 0.22974827140569687,
      "num_tokens": 74963807.0,
      "step": 32725
    },
    {
      "entropy": 5.720166206359863,
      "epoch": 3.235468564650059,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039745596616260983,
      "loss": 4.8998,
      "mean_token_accuracy": 0.23068222403526306,
      "num_tokens": 74975646.0,
      "step": 32730
    },
    {
      "entropy": 5.740240812301636,
      "epoch": 3.235962831158561,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003974263590468511,
      "loss": 5.0263,
      "mean_token_accuracy": 0.22296654284000397,
      "num_tokens": 74987197.0,
      "step": 32735
    },
    {
      "entropy": 5.704717397689819,
      "epoch": 3.236457097667062,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003973967489193849,
      "loss": 4.9732,
      "mean_token_accuracy": 0.22435937225818633,
      "num_tokens": 74999020.0,
      "step": 32740
    },
    {
      "entropy": 5.751628398895264,
      "epoch": 3.2369513641755634,
      "grad_norm": 1.359375,
      "learning_rate": 0.00039736713578093975,
      "loss": 4.993,
      "mean_token_accuracy": 0.2288254350423813,
      "num_tokens": 75010889.0,
      "step": 32745
    },
    {
      "entropy": 5.766255283355713,
      "epoch": 3.2374456306840647,
      "grad_norm": 1.171875,
      "learning_rate": 0.000397337519632244,
      "loss": 5.0046,
      "mean_token_accuracy": 0.21778091341257094,
      "num_tokens": 75021934.0,
      "step": 32750
    },
    {
      "entropy": 5.673018169403076,
      "epoch": 3.2379398971925664,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039730790047402635,
      "loss": 4.8945,
      "mean_token_accuracy": 0.23094050586223602,
      "num_tokens": 75032337.0,
      "step": 32755
    },
    {
      "entropy": 5.721152639389038,
      "epoch": 3.2384341637010676,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00039727827830701533,
      "loss": 4.8914,
      "mean_token_accuracy": 0.23109899014234542,
      "num_tokens": 75043094.0,
      "step": 32760
    },
    {
      "entropy": 5.724946355819702,
      "epoch": 3.238928430209569,
      "grad_norm": 1.28125,
      "learning_rate": 0.00039724865313193976,
      "loss": 4.8977,
      "mean_token_accuracy": 0.23170656710863113,
      "num_tokens": 75055108.0,
      "step": 32765
    },
    {
      "entropy": 5.734191274642944,
      "epoch": 3.23942269671807,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003972190249495283,
      "loss": 4.939,
      "mean_token_accuracy": 0.22555171996355056,
      "num_tokens": 75065388.0,
      "step": 32770
    },
    {
      "entropy": 5.726159143447876,
      "epoch": 3.239916963226572,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003971893937605099,
      "loss": 4.9634,
      "mean_token_accuracy": 0.22535429298877716,
      "num_tokens": 75077390.0,
      "step": 32775
    },
    {
      "entropy": 5.75620698928833,
      "epoch": 3.240411229735073,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039715975956561343,
      "loss": 4.9439,
      "mean_token_accuracy": 0.22031983733177185,
      "num_tokens": 75088795.0,
      "step": 32780
    },
    {
      "entropy": 5.75298228263855,
      "epoch": 3.2409054962435744,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00039713012236556797,
      "loss": 4.9954,
      "mean_token_accuracy": 0.22003937661647796,
      "num_tokens": 75100336.0,
      "step": 32785
    },
    {
      "entropy": 5.775192070007324,
      "epoch": 3.241399762752076,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003971004821611025,
      "loss": 5.0218,
      "mean_token_accuracy": 0.21449678540229797,
      "num_tokens": 75111639.0,
      "step": 32790
    },
    {
      "entropy": 5.743088912963867,
      "epoch": 3.2418940292605773,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00039707083895294625,
      "loss": 4.8871,
      "mean_token_accuracy": 0.2298685908317566,
      "num_tokens": 75122006.0,
      "step": 32795
    },
    {
      "entropy": 5.787891149520874,
      "epoch": 3.2423882957690786,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00039704119274182835,
      "loss": 4.9963,
      "mean_token_accuracy": 0.2201235368847847,
      "num_tokens": 75132406.0,
      "step": 32800
    },
    {
      "entropy": 5.742186117172241,
      "epoch": 3.24288256227758,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003970115435284783,
      "loss": 4.958,
      "mean_token_accuracy": 0.22476572543382645,
      "num_tokens": 75145444.0,
      "step": 32805
    },
    {
      "entropy": 5.730851173400879,
      "epoch": 3.2433768287860816,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039698189131362526,
      "loss": 4.8953,
      "mean_token_accuracy": 0.23222496658563613,
      "num_tokens": 75156069.0,
      "step": 32810
    },
    {
      "entropy": 5.740574073791504,
      "epoch": 3.243871095294583,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039695223609799885,
      "loss": 4.9888,
      "mean_token_accuracy": 0.219991435110569,
      "num_tokens": 75167075.0,
      "step": 32815
    },
    {
      "entropy": 5.711857509613037,
      "epoch": 3.244365361803084,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003969225778823285,
      "loss": 4.9103,
      "mean_token_accuracy": 0.2285653531551361,
      "num_tokens": 75178464.0,
      "step": 32820
    },
    {
      "entropy": 5.660870981216431,
      "epoch": 3.244859628311586,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003968929166673438,
      "loss": 4.8386,
      "mean_token_accuracy": 0.23088498264551163,
      "num_tokens": 75189803.0,
      "step": 32825
    },
    {
      "entropy": 5.7452253818511965,
      "epoch": 3.245353894820087,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039686325245377437,
      "loss": 5.025,
      "mean_token_accuracy": 0.21920556277036668,
      "num_tokens": 75201207.0,
      "step": 32830
    },
    {
      "entropy": 5.792042589187622,
      "epoch": 3.2458481613285883,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003968335852423501,
      "loss": 5.054,
      "mean_token_accuracy": 0.21513331085443496,
      "num_tokens": 75214092.0,
      "step": 32835
    },
    {
      "entropy": 5.743435192108154,
      "epoch": 3.2463424278370896,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003968039150338007,
      "loss": 4.9291,
      "mean_token_accuracy": 0.22192394882440566,
      "num_tokens": 75226139.0,
      "step": 32840
    },
    {
      "entropy": 5.741968822479248,
      "epoch": 3.2468366943455913,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039677424182885606,
      "loss": 4.9042,
      "mean_token_accuracy": 0.22605155706405639,
      "num_tokens": 75238273.0,
      "step": 32845
    },
    {
      "entropy": 5.753855609893799,
      "epoch": 3.2473309608540926,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003967445656282462,
      "loss": 5.0336,
      "mean_token_accuracy": 0.22025871872901917,
      "num_tokens": 75250356.0,
      "step": 32850
    },
    {
      "entropy": 5.775371742248535,
      "epoch": 3.247825227362594,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003967148864327012,
      "loss": 4.9573,
      "mean_token_accuracy": 0.22411278486251832,
      "num_tokens": 75260329.0,
      "step": 32855
    },
    {
      "entropy": 5.776127099990845,
      "epoch": 3.248319493871095,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003966852042429511,
      "loss": 5.0166,
      "mean_token_accuracy": 0.21955184787511825,
      "num_tokens": 75271405.0,
      "step": 32860
    },
    {
      "entropy": 5.787915515899658,
      "epoch": 3.248813760379597,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039665551905972607,
      "loss": 5.1031,
      "mean_token_accuracy": 0.2119460552930832,
      "num_tokens": 75283033.0,
      "step": 32865
    },
    {
      "entropy": 5.789566373825073,
      "epoch": 3.249308026888098,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003966258308837564,
      "loss": 4.9739,
      "mean_token_accuracy": 0.22148305624723436,
      "num_tokens": 75293582.0,
      "step": 32870
    },
    {
      "entropy": 5.737289905548096,
      "epoch": 3.2498022933965993,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003965961397157724,
      "loss": 4.956,
      "mean_token_accuracy": 0.2272091254591942,
      "num_tokens": 75306500.0,
      "step": 32875
    },
    {
      "entropy": 5.722474002838135,
      "epoch": 3.250296559905101,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039656644555650456,
      "loss": 4.9392,
      "mean_token_accuracy": 0.2249496951699257,
      "num_tokens": 75318562.0,
      "step": 32880
    },
    {
      "entropy": 5.737403106689453,
      "epoch": 3.2507908264136023,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039653674840668326,
      "loss": 5.0121,
      "mean_token_accuracy": 0.22013545632362366,
      "num_tokens": 75330211.0,
      "step": 32885
    },
    {
      "entropy": 5.698039960861206,
      "epoch": 3.2512850929221035,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003965070482670391,
      "loss": 4.9095,
      "mean_token_accuracy": 0.23064723908901213,
      "num_tokens": 75341244.0,
      "step": 32890
    },
    {
      "entropy": 5.7085212707519535,
      "epoch": 3.251779359430605,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003964773451383027,
      "loss": 4.9129,
      "mean_token_accuracy": 0.22117076218128204,
      "num_tokens": 75353381.0,
      "step": 32895
    },
    {
      "entropy": 5.709394454956055,
      "epoch": 3.2522736259391065,
      "grad_norm": 1.3125,
      "learning_rate": 0.00039644763902120473,
      "loss": 4.9209,
      "mean_token_accuracy": 0.22609392702579498,
      "num_tokens": 75364777.0,
      "step": 32900
    },
    {
      "entropy": 5.751534843444825,
      "epoch": 3.252767892447608,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039641792991647605,
      "loss": 4.9617,
      "mean_token_accuracy": 0.2263875976204872,
      "num_tokens": 75375677.0,
      "step": 32905
    },
    {
      "entropy": 5.775531387329101,
      "epoch": 3.253262158956109,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003963882178248475,
      "loss": 4.984,
      "mean_token_accuracy": 0.2281189888715744,
      "num_tokens": 75387925.0,
      "step": 32910
    },
    {
      "entropy": 5.639747571945191,
      "epoch": 3.2537564254646103,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003963585027470498,
      "loss": 4.8202,
      "mean_token_accuracy": 0.23912526071071624,
      "num_tokens": 75399000.0,
      "step": 32915
    },
    {
      "entropy": 5.65747103691101,
      "epoch": 3.254250691973112,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003963287846838142,
      "loss": 4.8877,
      "mean_token_accuracy": 0.22225039303302765,
      "num_tokens": 75409930.0,
      "step": 32920
    },
    {
      "entropy": 5.742747735977173,
      "epoch": 3.2547449584816133,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039629906363587177,
      "loss": 4.9957,
      "mean_token_accuracy": 0.21618987619876862,
      "num_tokens": 75421621.0,
      "step": 32925
    },
    {
      "entropy": 5.825606393814087,
      "epoch": 3.2552392249901145,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00039626933960395345,
      "loss": 5.0765,
      "mean_token_accuracy": 0.21737864911556243,
      "num_tokens": 75434339.0,
      "step": 32930
    },
    {
      "entropy": 5.758001470565796,
      "epoch": 3.2557334914986162,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039623961258879057,
      "loss": 5.0345,
      "mean_token_accuracy": 0.223114912211895,
      "num_tokens": 75446720.0,
      "step": 32935
    },
    {
      "entropy": 5.674108171463013,
      "epoch": 3.2562277580071175,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003962098825911144,
      "loss": 4.9203,
      "mean_token_accuracy": 0.22712511867284774,
      "num_tokens": 75459354.0,
      "step": 32940
    },
    {
      "entropy": 5.775084209442139,
      "epoch": 3.2567220245156188,
      "grad_norm": 1.296875,
      "learning_rate": 0.00039618014961165633,
      "loss": 5.0434,
      "mean_token_accuracy": 0.21416785269975663,
      "num_tokens": 75471704.0,
      "step": 32945
    },
    {
      "entropy": 5.829787111282348,
      "epoch": 3.25721629102412,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003961504136511478,
      "loss": 5.0631,
      "mean_token_accuracy": 0.2210068702697754,
      "num_tokens": 75483145.0,
      "step": 32950
    },
    {
      "entropy": 5.778584766387939,
      "epoch": 3.2577105575326217,
      "grad_norm": 1.28125,
      "learning_rate": 0.00039612067471032024,
      "loss": 4.9467,
      "mean_token_accuracy": 0.223041108250618,
      "num_tokens": 75493678.0,
      "step": 32955
    },
    {
      "entropy": 5.717800235748291,
      "epoch": 3.258204824041123,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003960909327899053,
      "loss": 4.9549,
      "mean_token_accuracy": 0.2232361763715744,
      "num_tokens": 75504581.0,
      "step": 32960
    },
    {
      "entropy": 5.652930831909179,
      "epoch": 3.2586990905496243,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003960611878906347,
      "loss": 4.863,
      "mean_token_accuracy": 0.23400799334049224,
      "num_tokens": 75515116.0,
      "step": 32965
    },
    {
      "entropy": 5.663910436630249,
      "epoch": 3.259193357058126,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003960314400132401,
      "loss": 4.8933,
      "mean_token_accuracy": 0.2326750412583351,
      "num_tokens": 75527001.0,
      "step": 32970
    },
    {
      "entropy": 5.769656085968018,
      "epoch": 3.2596876235666272,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003960016891584532,
      "loss": 4.9491,
      "mean_token_accuracy": 0.22817399501800537,
      "num_tokens": 75537641.0,
      "step": 32975
    },
    {
      "entropy": 5.754283857345581,
      "epoch": 3.2601818900751285,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003959719353270061,
      "loss": 4.9454,
      "mean_token_accuracy": 0.22727811932563782,
      "num_tokens": 75551225.0,
      "step": 32980
    },
    {
      "entropy": 5.660560035705567,
      "epoch": 3.2606761565836297,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039594217851963054,
      "loss": 4.9102,
      "mean_token_accuracy": 0.2328321620821953,
      "num_tokens": 75563112.0,
      "step": 32985
    },
    {
      "entropy": 5.7021783828735355,
      "epoch": 3.2611704230921315,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039591241873705864,
      "loss": 4.8914,
      "mean_token_accuracy": 0.2300262987613678,
      "num_tokens": 75573709.0,
      "step": 32990
    },
    {
      "entropy": 5.7520205020904545,
      "epoch": 3.2616646896006327,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003958826559800225,
      "loss": 4.9877,
      "mean_token_accuracy": 0.22024189531803132,
      "num_tokens": 75585269.0,
      "step": 32995
    },
    {
      "entropy": 5.747878551483154,
      "epoch": 3.262158956109134,
      "grad_norm": 1.125,
      "learning_rate": 0.0003958528902492543,
      "loss": 4.9928,
      "mean_token_accuracy": 0.21585804224014282,
      "num_tokens": 75597084.0,
      "step": 33000
    },
    {
      "epoch": 3.262158956109134,
      "eval_entropy": 5.481827690621877,
      "eval_loss": 5.033544063568115,
      "eval_mean_token_accuracy": 0.2278480052343154,
      "eval_num_tokens": 75597084.0,
      "eval_runtime": 20.2975,
      "eval_samples_per_second": 1601.823,
      "eval_steps_per_second": 200.271,
      "step": 33000
    },
    {
      "entropy": 5.675453424453735,
      "epoch": 3.2626532226176352,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039582312154548614,
      "loss": 4.9086,
      "mean_token_accuracy": 0.23547551929950714,
      "num_tokens": 75608948.0,
      "step": 33005
    },
    {
      "entropy": 5.777303457260132,
      "epoch": 3.263147489126137,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039579334986945055,
      "loss": 4.9603,
      "mean_token_accuracy": 0.22497835606336594,
      "num_tokens": 75619845.0,
      "step": 33010
    },
    {
      "entropy": 5.782904767990113,
      "epoch": 3.263641755634638,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003957635752218797,
      "loss": 5.0213,
      "mean_token_accuracy": 0.21278902143239975,
      "num_tokens": 75632116.0,
      "step": 33015
    },
    {
      "entropy": 5.7099076271057125,
      "epoch": 3.2641360221431395,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003957337976035063,
      "loss": 4.938,
      "mean_token_accuracy": 0.21921972036361695,
      "num_tokens": 75643675.0,
      "step": 33020
    },
    {
      "entropy": 5.776952028274536,
      "epoch": 3.2646302886516407,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003957040170150626,
      "loss": 4.9691,
      "mean_token_accuracy": 0.2205577239394188,
      "num_tokens": 75654451.0,
      "step": 33025
    },
    {
      "entropy": 5.692875719070434,
      "epoch": 3.2651245551601424,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003956742334572815,
      "loss": 4.874,
      "mean_token_accuracy": 0.2325876161456108,
      "num_tokens": 75666260.0,
      "step": 33030
    },
    {
      "entropy": 5.735413074493408,
      "epoch": 3.2656188216686437,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003956444469308955,
      "loss": 4.9913,
      "mean_token_accuracy": 0.21995376348495482,
      "num_tokens": 75678785.0,
      "step": 33035
    },
    {
      "entropy": 5.712336826324463,
      "epoch": 3.266113088177145,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003956146574366374,
      "loss": 4.9176,
      "mean_token_accuracy": 0.2280648395419121,
      "num_tokens": 75690500.0,
      "step": 33040
    },
    {
      "entropy": 5.655271720886231,
      "epoch": 3.2666073546856467,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003955848649752399,
      "loss": 4.9077,
      "mean_token_accuracy": 0.23271434605121613,
      "num_tokens": 75703616.0,
      "step": 33045
    },
    {
      "entropy": 5.693547916412354,
      "epoch": 3.267101621194148,
      "grad_norm": 1.28125,
      "learning_rate": 0.00039555506954743613,
      "loss": 4.9197,
      "mean_token_accuracy": 0.22507739067077637,
      "num_tokens": 75713996.0,
      "step": 33050
    },
    {
      "entropy": 5.78652925491333,
      "epoch": 3.267595887702649,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039552527115395896,
      "loss": 5.0318,
      "mean_token_accuracy": 0.21904132068157195,
      "num_tokens": 75726746.0,
      "step": 33055
    },
    {
      "entropy": 5.806911563873291,
      "epoch": 3.2680901542111505,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039549546979554133,
      "loss": 5.053,
      "mean_token_accuracy": 0.21448031067848206,
      "num_tokens": 75738441.0,
      "step": 33060
    },
    {
      "entropy": 5.715211391448975,
      "epoch": 3.268584420719652,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003954656654729166,
      "loss": 5.0144,
      "mean_token_accuracy": 0.21815508604049683,
      "num_tokens": 75749592.0,
      "step": 33065
    },
    {
      "entropy": 5.822201442718506,
      "epoch": 3.2690786872281534,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039543585818681764,
      "loss": 5.1018,
      "mean_token_accuracy": 0.21020036637783052,
      "num_tokens": 75761668.0,
      "step": 33070
    },
    {
      "entropy": 5.782623291015625,
      "epoch": 3.2695729537366547,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003954060479379779,
      "loss": 4.9893,
      "mean_token_accuracy": 0.2198414087295532,
      "num_tokens": 75773277.0,
      "step": 33075
    },
    {
      "entropy": 5.759593868255616,
      "epoch": 3.2700672202451564,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003953762347271308,
      "loss": 4.9709,
      "mean_token_accuracy": 0.22093071788549423,
      "num_tokens": 75783888.0,
      "step": 33080
    },
    {
      "entropy": 5.687453079223633,
      "epoch": 3.2705614867536577,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039534641855500964,
      "loss": 4.9454,
      "mean_token_accuracy": 0.22809427678585054,
      "num_tokens": 75796157.0,
      "step": 33085
    },
    {
      "entropy": 5.6389994621276855,
      "epoch": 3.271055753262159,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00039531659942234785,
      "loss": 4.8417,
      "mean_token_accuracy": 0.23880015909671784,
      "num_tokens": 75807523.0,
      "step": 33090
    },
    {
      "entropy": 5.7750657081604,
      "epoch": 3.27155001977066,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039528677732987913,
      "loss": 5.0272,
      "mean_token_accuracy": 0.21138565689325334,
      "num_tokens": 75819383.0,
      "step": 33095
    },
    {
      "entropy": 5.747149229049683,
      "epoch": 3.272044286279162,
      "grad_norm": 1.109375,
      "learning_rate": 0.000395256952278337,
      "loss": 5.0245,
      "mean_token_accuracy": 0.22631698101758957,
      "num_tokens": 75830957.0,
      "step": 33100
    },
    {
      "entropy": 5.757021808624268,
      "epoch": 3.272538552787663,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039522712426845507,
      "loss": 4.9517,
      "mean_token_accuracy": 0.22000519633293153,
      "num_tokens": 75841260.0,
      "step": 33105
    },
    {
      "entropy": 5.725371980667115,
      "epoch": 3.2730328192961644,
      "grad_norm": 1.25,
      "learning_rate": 0.00039519729330096735,
      "loss": 4.9818,
      "mean_token_accuracy": 0.22631653696298598,
      "num_tokens": 75852803.0,
      "step": 33110
    },
    {
      "entropy": 5.768563175201416,
      "epoch": 3.273527085804666,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003951674593766076,
      "loss": 4.98,
      "mean_token_accuracy": 0.2222690165042877,
      "num_tokens": 75863989.0,
      "step": 33115
    },
    {
      "entropy": 5.693273305892944,
      "epoch": 3.2740213523131674,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003951376224961096,
      "loss": 4.9023,
      "mean_token_accuracy": 0.23370018750429153,
      "num_tokens": 75875381.0,
      "step": 33120
    },
    {
      "entropy": 5.68121337890625,
      "epoch": 3.2745156188216686,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003951077826602075,
      "loss": 4.8729,
      "mean_token_accuracy": 0.23197812288999559,
      "num_tokens": 75887204.0,
      "step": 33125
    },
    {
      "entropy": 5.7156453132629395,
      "epoch": 3.27500988533017,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003950779398696353,
      "loss": 4.9791,
      "mean_token_accuracy": 0.21795835793018342,
      "num_tokens": 75898666.0,
      "step": 33130
    },
    {
      "entropy": 5.716331720352173,
      "epoch": 3.275504151838671,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00039504809412512724,
      "loss": 4.9281,
      "mean_token_accuracy": 0.22574625313282012,
      "num_tokens": 75910151.0,
      "step": 33135
    },
    {
      "entropy": 5.771676158905029,
      "epoch": 3.275998418347173,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003950182454274173,
      "loss": 4.9861,
      "mean_token_accuracy": 0.228470116853714,
      "num_tokens": 75920533.0,
      "step": 33140
    },
    {
      "entropy": 5.76217942237854,
      "epoch": 3.276492684855674,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039498839377723996,
      "loss": 4.9486,
      "mean_token_accuracy": 0.23305534273386003,
      "num_tokens": 75932335.0,
      "step": 33145
    },
    {
      "entropy": 5.782145261764526,
      "epoch": 3.2769869513641754,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003949585391753295,
      "loss": 4.9945,
      "mean_token_accuracy": 0.2259720340371132,
      "num_tokens": 75943741.0,
      "step": 33150
    },
    {
      "entropy": 5.676116895675659,
      "epoch": 3.277481217872677,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003949286816224204,
      "loss": 4.9493,
      "mean_token_accuracy": 0.22634295821189881,
      "num_tokens": 75955755.0,
      "step": 33155
    },
    {
      "entropy": 5.744834852218628,
      "epoch": 3.2779754843811784,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003948988211192471,
      "loss": 4.9046,
      "mean_token_accuracy": 0.22474804371595383,
      "num_tokens": 75967057.0,
      "step": 33160
    },
    {
      "entropy": 5.710540342330932,
      "epoch": 3.2784697508896796,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003948689576665443,
      "loss": 4.9605,
      "mean_token_accuracy": 0.22550025582313538,
      "num_tokens": 75978367.0,
      "step": 33165
    },
    {
      "entropy": 5.787054347991943,
      "epoch": 3.278964017398181,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039483909126504643,
      "loss": 5.0541,
      "mean_token_accuracy": 0.21333929151296616,
      "num_tokens": 75990738.0,
      "step": 33170
    },
    {
      "entropy": 5.762073278427124,
      "epoch": 3.2794582839066826,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003948092219154883,
      "loss": 5.0134,
      "mean_token_accuracy": 0.2165161356329918,
      "num_tokens": 76003058.0,
      "step": 33175
    },
    {
      "entropy": 5.726857662200928,
      "epoch": 3.279952550415184,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003947793496186048,
      "loss": 4.9224,
      "mean_token_accuracy": 0.23334858417510987,
      "num_tokens": 76013743.0,
      "step": 33180
    },
    {
      "entropy": 5.750691270828247,
      "epoch": 3.280446816923685,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003947494743751307,
      "loss": 4.9865,
      "mean_token_accuracy": 0.22257972061634063,
      "num_tokens": 76024346.0,
      "step": 33185
    },
    {
      "entropy": 5.774991703033447,
      "epoch": 3.280941083432187,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039471959618580095,
      "loss": 4.9903,
      "mean_token_accuracy": 0.21920051127672197,
      "num_tokens": 76035444.0,
      "step": 33190
    },
    {
      "entropy": 5.692640542984009,
      "epoch": 3.281435349940688,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003946897150513506,
      "loss": 4.9292,
      "mean_token_accuracy": 0.23168145567178727,
      "num_tokens": 76047363.0,
      "step": 33195
    },
    {
      "entropy": 5.730627727508545,
      "epoch": 3.2819296164491893,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003946598309725146,
      "loss": 4.9916,
      "mean_token_accuracy": 0.216161872446537,
      "num_tokens": 76059581.0,
      "step": 33200
    },
    {
      "entropy": 5.69808406829834,
      "epoch": 3.2824238829576906,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003946299439500283,
      "loss": 4.8981,
      "mean_token_accuracy": 0.23362579047679902,
      "num_tokens": 76071455.0,
      "step": 33205
    },
    {
      "entropy": 5.67490062713623,
      "epoch": 3.2829181494661923,
      "grad_norm": 1.125,
      "learning_rate": 0.0003946000539846267,
      "loss": 4.8946,
      "mean_token_accuracy": 0.23215041905641556,
      "num_tokens": 76083794.0,
      "step": 33210
    },
    {
      "entropy": 5.768623161315918,
      "epoch": 3.2834124159746936,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003945701610770453,
      "loss": 4.9929,
      "mean_token_accuracy": 0.21989970356225969,
      "num_tokens": 76095690.0,
      "step": 33215
    },
    {
      "entropy": 5.67642560005188,
      "epoch": 3.283906682483195,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003945402652280193,
      "loss": 4.9053,
      "mean_token_accuracy": 0.2241143152117729,
      "num_tokens": 76106730.0,
      "step": 33220
    },
    {
      "entropy": 5.688515472412109,
      "epoch": 3.2844009489916965,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003945103664382844,
      "loss": 4.972,
      "mean_token_accuracy": 0.22458657175302504,
      "num_tokens": 76117333.0,
      "step": 33225
    },
    {
      "entropy": 5.680511713027954,
      "epoch": 3.284895215500198,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003944804647085759,
      "loss": 4.9014,
      "mean_token_accuracy": 0.23284486681222916,
      "num_tokens": 76129060.0,
      "step": 33230
    },
    {
      "entropy": 5.72352819442749,
      "epoch": 3.285389482008699,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003944505600396294,
      "loss": 4.9088,
      "mean_token_accuracy": 0.23243329674005508,
      "num_tokens": 76140724.0,
      "step": 33235
    },
    {
      "entropy": 5.66750955581665,
      "epoch": 3.2858837485172003,
      "grad_norm": 1.125,
      "learning_rate": 0.0003944206524321805,
      "loss": 4.9024,
      "mean_token_accuracy": 0.23183085918426513,
      "num_tokens": 76153021.0,
      "step": 33240
    },
    {
      "entropy": 5.717838907241822,
      "epoch": 3.286378015025702,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039439074188696516,
      "loss": 4.9592,
      "mean_token_accuracy": 0.2249625727534294,
      "num_tokens": 76165759.0,
      "step": 33245
    },
    {
      "entropy": 5.78043007850647,
      "epoch": 3.2868722815342033,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003943608284047189,
      "loss": 4.958,
      "mean_token_accuracy": 0.22408254444599152,
      "num_tokens": 76178645.0,
      "step": 33250
    },
    {
      "entropy": 5.751153230667114,
      "epoch": 3.2873665480427046,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003943309119861779,
      "loss": 4.9169,
      "mean_token_accuracy": 0.22611104398965837,
      "num_tokens": 76189320.0,
      "step": 33255
    },
    {
      "entropy": 5.798932027816773,
      "epoch": 3.287860814551206,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003943009926320778,
      "loss": 5.0551,
      "mean_token_accuracy": 0.21849936693906785,
      "num_tokens": 76200672.0,
      "step": 33260
    },
    {
      "entropy": 5.726233243942261,
      "epoch": 3.2883550810597075,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003942710703431548,
      "loss": 4.9406,
      "mean_token_accuracy": 0.22550822645425797,
      "num_tokens": 76211035.0,
      "step": 33265
    },
    {
      "entropy": 5.7367957592010494,
      "epoch": 3.288849347568209,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003942411451201451,
      "loss": 4.9764,
      "mean_token_accuracy": 0.22294462025165557,
      "num_tokens": 76222262.0,
      "step": 33270
    },
    {
      "entropy": 5.7552320003509525,
      "epoch": 3.28934361407671,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039421121696378466,
      "loss": 4.9852,
      "mean_token_accuracy": 0.2240929961204529,
      "num_tokens": 76234346.0,
      "step": 33275
    },
    {
      "entropy": 5.731439447402954,
      "epoch": 3.2898378805852113,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003941812858748098,
      "loss": 4.9527,
      "mean_token_accuracy": 0.22616081386804582,
      "num_tokens": 76244536.0,
      "step": 33280
    },
    {
      "entropy": 5.686013698577881,
      "epoch": 3.290332147093713,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003941513518539568,
      "loss": 4.9642,
      "mean_token_accuracy": 0.22632805854082108,
      "num_tokens": 76256080.0,
      "step": 33285
    },
    {
      "entropy": 5.725487089157104,
      "epoch": 3.2908264136022143,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039412141490196206,
      "loss": 4.9289,
      "mean_token_accuracy": 0.2282807484269142,
      "num_tokens": 76266741.0,
      "step": 33290
    },
    {
      "entropy": 5.728825044631958,
      "epoch": 3.2913206801107155,
      "grad_norm": 1.3125,
      "learning_rate": 0.000394091475019562,
      "loss": 4.9459,
      "mean_token_accuracy": 0.21818799674510955,
      "num_tokens": 76279099.0,
      "step": 33295
    },
    {
      "entropy": 5.750980424880981,
      "epoch": 3.2918149466192173,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003940615322074932,
      "loss": 5.0788,
      "mean_token_accuracy": 0.21638076901435851,
      "num_tokens": 76290692.0,
      "step": 33300
    },
    {
      "entropy": 5.761463069915772,
      "epoch": 3.2923092131277185,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039403158646649227,
      "loss": 4.9571,
      "mean_token_accuracy": 0.22315480262041093,
      "num_tokens": 76302003.0,
      "step": 33305
    },
    {
      "entropy": 5.698015832901001,
      "epoch": 3.2928034796362198,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003940016377972958,
      "loss": 4.9532,
      "mean_token_accuracy": 0.22567558586597442,
      "num_tokens": 76313255.0,
      "step": 33310
    },
    {
      "entropy": 5.689161682128907,
      "epoch": 3.293297746144721,
      "grad_norm": 1.296875,
      "learning_rate": 0.00039397168620064065,
      "loss": 4.8999,
      "mean_token_accuracy": 0.23211898505687714,
      "num_tokens": 76323154.0,
      "step": 33315
    },
    {
      "entropy": 5.8088874340057375,
      "epoch": 3.2937920126532227,
      "grad_norm": 1.5,
      "learning_rate": 0.00039394173167726346,
      "loss": 5.0019,
      "mean_token_accuracy": 0.21445233672857283,
      "num_tokens": 76333168.0,
      "step": 33320
    },
    {
      "entropy": 5.75108494758606,
      "epoch": 3.294286279161724,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003939117742279013,
      "loss": 5.015,
      "mean_token_accuracy": 0.2187953218817711,
      "num_tokens": 76344916.0,
      "step": 33325
    },
    {
      "entropy": 5.795640182495117,
      "epoch": 3.2947805456702253,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000393881813853291,
      "loss": 4.9796,
      "mean_token_accuracy": 0.22268164455890654,
      "num_tokens": 76357124.0,
      "step": 33330
    },
    {
      "entropy": 5.836868190765381,
      "epoch": 3.295274812178727,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039385185055416965,
      "loss": 5.0408,
      "mean_token_accuracy": 0.2126067817211151,
      "num_tokens": 76368784.0,
      "step": 33335
    },
    {
      "entropy": 5.808990240097046,
      "epoch": 3.2957690786872282,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039382188433127437,
      "loss": 5.0089,
      "mean_token_accuracy": 0.21562146097421647,
      "num_tokens": 76379864.0,
      "step": 33340
    },
    {
      "entropy": 5.759326934814453,
      "epoch": 3.2962633451957295,
      "grad_norm": 1.25,
      "learning_rate": 0.0003937919151853422,
      "loss": 4.9744,
      "mean_token_accuracy": 0.22397873550653458,
      "num_tokens": 76392184.0,
      "step": 33345
    },
    {
      "entropy": 5.75353856086731,
      "epoch": 3.2967576117042308,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003937619431171107,
      "loss": 4.9318,
      "mean_token_accuracy": 0.22035236805677413,
      "num_tokens": 76404312.0,
      "step": 33350
    },
    {
      "entropy": 5.7463438510894775,
      "epoch": 3.2972518782127325,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003937319681273168,
      "loss": 5.0381,
      "mean_token_accuracy": 0.21764319092035295,
      "num_tokens": 76414351.0,
      "step": 33355
    },
    {
      "entropy": 5.762243366241455,
      "epoch": 3.2977461447212337,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003937019902166981,
      "loss": 5.0076,
      "mean_token_accuracy": 0.2170119896531105,
      "num_tokens": 76425851.0,
      "step": 33360
    },
    {
      "entropy": 5.734449005126953,
      "epoch": 3.298240411229735,
      "grad_norm": 1.234375,
      "learning_rate": 0.000393672009385992,
      "loss": 4.9321,
      "mean_token_accuracy": 0.22990116178989412,
      "num_tokens": 76437669.0,
      "step": 33365
    },
    {
      "entropy": 5.71902027130127,
      "epoch": 3.2987346777382367,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003936420256359361,
      "loss": 4.8922,
      "mean_token_accuracy": 0.22655837535858153,
      "num_tokens": 76449589.0,
      "step": 33370
    },
    {
      "entropy": 5.794484663009643,
      "epoch": 3.299228944246738,
      "grad_norm": 1.21875,
      "learning_rate": 0.000393612038967268,
      "loss": 5.0596,
      "mean_token_accuracy": 0.21757402271032333,
      "num_tokens": 76462143.0,
      "step": 33375
    },
    {
      "entropy": 5.764831686019898,
      "epoch": 3.2997232107552392,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039358204938072527,
      "loss": 4.9633,
      "mean_token_accuracy": 0.22035585790872575,
      "num_tokens": 76473731.0,
      "step": 33380
    },
    {
      "entropy": 5.754424047470093,
      "epoch": 3.3002174772637405,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039355205687704575,
      "loss": 4.9619,
      "mean_token_accuracy": 0.2258112460374832,
      "num_tokens": 76484287.0,
      "step": 33385
    },
    {
      "entropy": 5.687608242034912,
      "epoch": 3.3007117437722417,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003935220614569672,
      "loss": 4.937,
      "mean_token_accuracy": 0.22870177626609803,
      "num_tokens": 76495940.0,
      "step": 33390
    },
    {
      "entropy": 5.694515609741211,
      "epoch": 3.3012060102807435,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039349206312122765,
      "loss": 4.8959,
      "mean_token_accuracy": 0.2354155421257019,
      "num_tokens": 76507890.0,
      "step": 33395
    },
    {
      "entropy": 5.764904499053955,
      "epoch": 3.3017002767892447,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003934620618705649,
      "loss": 4.9568,
      "mean_token_accuracy": 0.22027647495269775,
      "num_tokens": 76519654.0,
      "step": 33400
    },
    {
      "entropy": 5.742703199386597,
      "epoch": 3.302194543297746,
      "grad_norm": 1.15625,
      "learning_rate": 0.000393432057705717,
      "loss": 4.9372,
      "mean_token_accuracy": 0.22233319133520127,
      "num_tokens": 76532026.0,
      "step": 33405
    },
    {
      "entropy": 5.707004404067993,
      "epoch": 3.3026888098062477,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003934020506274222,
      "loss": 4.9021,
      "mean_token_accuracy": 0.23226378262043,
      "num_tokens": 76543629.0,
      "step": 33410
    },
    {
      "entropy": 5.750660800933838,
      "epoch": 3.303183076314749,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003933720406364185,
      "loss": 4.98,
      "mean_token_accuracy": 0.22413599640130996,
      "num_tokens": 76554069.0,
      "step": 33415
    },
    {
      "entropy": 5.707640981674194,
      "epoch": 3.30367734282325,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039334202773344437,
      "loss": 4.925,
      "mean_token_accuracy": 0.22701105773448943,
      "num_tokens": 76566148.0,
      "step": 33420
    },
    {
      "entropy": 5.7383016586303714,
      "epoch": 3.3041716093317515,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039331201191923794,
      "loss": 4.9396,
      "mean_token_accuracy": 0.22432364970445634,
      "num_tokens": 76577208.0,
      "step": 33425
    },
    {
      "entropy": 5.675359153747559,
      "epoch": 3.304665875840253,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003932819931945375,
      "loss": 4.9212,
      "mean_token_accuracy": 0.23862544000148772,
      "num_tokens": 76588532.0,
      "step": 33430
    },
    {
      "entropy": 5.795589780807495,
      "epoch": 3.3051601423487544,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003932519715600819,
      "loss": 5.0207,
      "mean_token_accuracy": 0.22104718685150146,
      "num_tokens": 76600054.0,
      "step": 33435
    },
    {
      "entropy": 5.768060398101807,
      "epoch": 3.3056544088572557,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039322194701660936,
      "loss": 4.9818,
      "mean_token_accuracy": 0.2276659607887268,
      "num_tokens": 76612571.0,
      "step": 33440
    },
    {
      "entropy": 5.761131000518799,
      "epoch": 3.3061486753657574,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039319191956485854,
      "loss": 5.0091,
      "mean_token_accuracy": 0.21776643097400666,
      "num_tokens": 76625435.0,
      "step": 33445
    },
    {
      "entropy": 5.698491907119751,
      "epoch": 3.3066429418742587,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003931618892055682,
      "loss": 4.9059,
      "mean_token_accuracy": 0.23080936074256897,
      "num_tokens": 76635529.0,
      "step": 33450
    },
    {
      "entropy": 5.658594942092895,
      "epoch": 3.30713720838276,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039313185593947713,
      "loss": 4.8371,
      "mean_token_accuracy": 0.2367470696568489,
      "num_tokens": 76645819.0,
      "step": 33455
    },
    {
      "entropy": 5.822653532028198,
      "epoch": 3.307631474891261,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039310181976732404,
      "loss": 5.1158,
      "mean_token_accuracy": 0.20834122002124786,
      "num_tokens": 76657803.0,
      "step": 33460
    },
    {
      "entropy": 5.838195419311523,
      "epoch": 3.308125741399763,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00039307178068984783,
      "loss": 5.083,
      "mean_token_accuracy": 0.2120375454425812,
      "num_tokens": 76670260.0,
      "step": 33465
    },
    {
      "entropy": 5.709773540496826,
      "epoch": 3.308620007908264,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003930417387077875,
      "loss": 4.9226,
      "mean_token_accuracy": 0.22853083312511444,
      "num_tokens": 76682591.0,
      "step": 33470
    },
    {
      "entropy": 5.724830913543701,
      "epoch": 3.3091142744167654,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003930116938218821,
      "loss": 4.9237,
      "mean_token_accuracy": 0.22352764159440994,
      "num_tokens": 76694147.0,
      "step": 33475
    },
    {
      "entropy": 5.750931358337402,
      "epoch": 3.309608540925267,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039298164603287063,
      "loss": 5.0383,
      "mean_token_accuracy": 0.2250906527042389,
      "num_tokens": 76706325.0,
      "step": 33480
    },
    {
      "entropy": 5.78615140914917,
      "epoch": 3.3101028074337684,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003929515953414926,
      "loss": 4.9835,
      "mean_token_accuracy": 0.22102261781692506,
      "num_tokens": 76719574.0,
      "step": 33485
    },
    {
      "entropy": 5.724516344070435,
      "epoch": 3.3105970739422697,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003929215417484868,
      "loss": 4.8978,
      "mean_token_accuracy": 0.2268466144800186,
      "num_tokens": 76731820.0,
      "step": 33490
    },
    {
      "entropy": 5.656566381454468,
      "epoch": 3.311091340450771,
      "grad_norm": 1.25,
      "learning_rate": 0.00039289148525459283,
      "loss": 4.855,
      "mean_token_accuracy": 0.23334207832813264,
      "num_tokens": 76742849.0,
      "step": 33495
    },
    {
      "entropy": 5.791948509216309,
      "epoch": 3.3115856069592726,
      "grad_norm": 1.265625,
      "learning_rate": 0.00039286142586055005,
      "loss": 5.038,
      "mean_token_accuracy": 0.21924878358840943,
      "num_tokens": 76754747.0,
      "step": 33500
    },
    {
      "entropy": 5.717371129989624,
      "epoch": 3.312079873467774,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000392831363567098,
      "loss": 4.892,
      "mean_token_accuracy": 0.2289115607738495,
      "num_tokens": 76766738.0,
      "step": 33505
    },
    {
      "entropy": 5.7241425037384035,
      "epoch": 3.312574139976275,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00039280129837497595,
      "loss": 4.9818,
      "mean_token_accuracy": 0.21793031096458435,
      "num_tokens": 76777457.0,
      "step": 33510
    },
    {
      "entropy": 5.7857160568237305,
      "epoch": 3.3130684064847764,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039277123028492376,
      "loss": 5.0349,
      "mean_token_accuracy": 0.21183947026729583,
      "num_tokens": 76788335.0,
      "step": 33515
    },
    {
      "entropy": 5.801105308532715,
      "epoch": 3.313562672993278,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003927411592976811,
      "loss": 4.9453,
      "mean_token_accuracy": 0.22345901578664779,
      "num_tokens": 76800471.0,
      "step": 33520
    },
    {
      "entropy": 5.724969816207886,
      "epoch": 3.3140569395017794,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00039271108541398753,
      "loss": 4.955,
      "mean_token_accuracy": 0.2281900614500046,
      "num_tokens": 76811792.0,
      "step": 33525
    },
    {
      "entropy": 5.70270938873291,
      "epoch": 3.3145512060102806,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003926810086345831,
      "loss": 4.89,
      "mean_token_accuracy": 0.23091378062963486,
      "num_tokens": 76824332.0,
      "step": 33530
    },
    {
      "entropy": 5.766416597366333,
      "epoch": 3.315045472518782,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039265092896020767,
      "loss": 4.9819,
      "mean_token_accuracy": 0.21932831704616546,
      "num_tokens": 76835712.0,
      "step": 33535
    },
    {
      "entropy": 5.689514923095703,
      "epoch": 3.3155397390272836,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000392620846391601,
      "loss": 4.9061,
      "mean_token_accuracy": 0.2360375180840492,
      "num_tokens": 76847915.0,
      "step": 33540
    },
    {
      "entropy": 5.720696353912354,
      "epoch": 3.316034005535785,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039259076092950333,
      "loss": 4.9936,
      "mean_token_accuracy": 0.2188990220427513,
      "num_tokens": 76859982.0,
      "step": 33545
    },
    {
      "entropy": 5.695939064025879,
      "epoch": 3.316528272044286,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003925606725746547,
      "loss": 4.8999,
      "mean_token_accuracy": 0.23364005088806153,
      "num_tokens": 76870743.0,
      "step": 33550
    },
    {
      "entropy": 5.629205989837646,
      "epoch": 3.317022538552788,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003925305813277952,
      "loss": 4.8683,
      "mean_token_accuracy": 0.23170077204704284,
      "num_tokens": 76882118.0,
      "step": 33555
    },
    {
      "entropy": 5.729952144622803,
      "epoch": 3.317516805061289,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039250048718966524,
      "loss": 4.9697,
      "mean_token_accuracy": 0.22129811197519303,
      "num_tokens": 76893658.0,
      "step": 33560
    },
    {
      "entropy": 5.769687747955322,
      "epoch": 3.3180110715697904,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039247039016100514,
      "loss": 4.9496,
      "mean_token_accuracy": 0.22413306832313537,
      "num_tokens": 76904974.0,
      "step": 33565
    },
    {
      "entropy": 5.730953741073608,
      "epoch": 3.3185053380782916,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039244029024255506,
      "loss": 4.9608,
      "mean_token_accuracy": 0.22415735125541686,
      "num_tokens": 76915972.0,
      "step": 33570
    },
    {
      "entropy": 5.74026837348938,
      "epoch": 3.3189996045867933,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003924101874350557,
      "loss": 4.9468,
      "mean_token_accuracy": 0.22654052823781967,
      "num_tokens": 76927397.0,
      "step": 33575
    },
    {
      "entropy": 5.783735275268555,
      "epoch": 3.3194938710952946,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003923800817392475,
      "loss": 4.9561,
      "mean_token_accuracy": 0.21487514227628707,
      "num_tokens": 76939744.0,
      "step": 33580
    },
    {
      "entropy": 5.814383697509766,
      "epoch": 3.319988137603796,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039234997315587105,
      "loss": 5.0672,
      "mean_token_accuracy": 0.21589359641075134,
      "num_tokens": 76951747.0,
      "step": 33585
    },
    {
      "entropy": 5.769013071060181,
      "epoch": 3.3204824041122976,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003923198616856671,
      "loss": 4.9672,
      "mean_token_accuracy": 0.2254791423678398,
      "num_tokens": 76961734.0,
      "step": 33590
    },
    {
      "entropy": 5.746283340454101,
      "epoch": 3.320976670620799,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00039228974732937637,
      "loss": 4.9895,
      "mean_token_accuracy": 0.21665629297494887,
      "num_tokens": 76972566.0,
      "step": 33595
    },
    {
      "entropy": 5.768842458724976,
      "epoch": 3.3214709371293,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003922596300877396,
      "loss": 4.9882,
      "mean_token_accuracy": 0.2165379822254181,
      "num_tokens": 76983446.0,
      "step": 33600
    },
    {
      "entropy": 5.765625953674316,
      "epoch": 3.3219652036378013,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003922295099614978,
      "loss": 4.9431,
      "mean_token_accuracy": 0.22241671085357667,
      "num_tokens": 76994292.0,
      "step": 33605
    },
    {
      "entropy": 5.7250487327575685,
      "epoch": 3.322459470146303,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039219938695139186,
      "loss": 4.9628,
      "mean_token_accuracy": 0.22445331364870072,
      "num_tokens": 77007095.0,
      "step": 33610
    },
    {
      "entropy": 5.737336111068726,
      "epoch": 3.3229537366548043,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003921692610581628,
      "loss": 4.9334,
      "mean_token_accuracy": 0.22273040413856507,
      "num_tokens": 77017358.0,
      "step": 33615
    },
    {
      "entropy": 5.6836442947387695,
      "epoch": 3.3234480031633056,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003921391322825517,
      "loss": 4.8863,
      "mean_token_accuracy": 0.2275211066007614,
      "num_tokens": 77028794.0,
      "step": 33620
    },
    {
      "entropy": 5.66890287399292,
      "epoch": 3.3239422696718073,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039210900062529986,
      "loss": 4.9292,
      "mean_token_accuracy": 0.22593703269958496,
      "num_tokens": 77040795.0,
      "step": 33625
    },
    {
      "entropy": 5.76408281326294,
      "epoch": 3.3244365361803085,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003920788660871483,
      "loss": 4.9807,
      "mean_token_accuracy": 0.22772539108991624,
      "num_tokens": 77051810.0,
      "step": 33630
    },
    {
      "entropy": 5.759575891494751,
      "epoch": 3.32493080268881,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039204872866883863,
      "loss": 5.0466,
      "mean_token_accuracy": 0.22512495666742324,
      "num_tokens": 77064504.0,
      "step": 33635
    },
    {
      "entropy": 5.779351663589478,
      "epoch": 3.325425069197311,
      "grad_norm": 1.15625,
      "learning_rate": 0.000392018588371112,
      "loss": 4.9768,
      "mean_token_accuracy": 0.2178046301007271,
      "num_tokens": 77075163.0,
      "step": 33640
    },
    {
      "entropy": 5.77309217453003,
      "epoch": 3.3259193357058123,
      "grad_norm": 1.3125,
      "learning_rate": 0.00039198844519470994,
      "loss": 4.9773,
      "mean_token_accuracy": 0.22501225173473358,
      "num_tokens": 77086871.0,
      "step": 33645
    },
    {
      "entropy": 5.691672945022583,
      "epoch": 3.326413602214314,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000391958299140374,
      "loss": 4.8848,
      "mean_token_accuracy": 0.2329185351729393,
      "num_tokens": 77097285.0,
      "step": 33650
    },
    {
      "entropy": 5.650037813186645,
      "epoch": 3.3269078687228153,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003919281502088457,
      "loss": 4.8477,
      "mean_token_accuracy": 0.23862815350294114,
      "num_tokens": 77107035.0,
      "step": 33655
    },
    {
      "entropy": 5.774787855148316,
      "epoch": 3.3274021352313166,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003918979984008668,
      "loss": 4.9793,
      "mean_token_accuracy": 0.21611535251140596,
      "num_tokens": 77119449.0,
      "step": 33660
    },
    {
      "entropy": 5.744748592376709,
      "epoch": 3.3278964017398183,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00039186784371717903,
      "loss": 4.9779,
      "mean_token_accuracy": 0.22698270678520202,
      "num_tokens": 77130508.0,
      "step": 33665
    },
    {
      "entropy": 5.758039951324463,
      "epoch": 3.3283906682483195,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003918376861585242,
      "loss": 4.9699,
      "mean_token_accuracy": 0.21827473789453505,
      "num_tokens": 77142080.0,
      "step": 33670
    },
    {
      "entropy": 5.791952133178711,
      "epoch": 3.328884934756821,
      "grad_norm": 1.265625,
      "learning_rate": 0.00039180752572564405,
      "loss": 4.9971,
      "mean_token_accuracy": 0.22108243703842162,
      "num_tokens": 77152366.0,
      "step": 33675
    },
    {
      "entropy": 5.71120662689209,
      "epoch": 3.329379201265322,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003917773624192808,
      "loss": 5.0435,
      "mean_token_accuracy": 0.21778865307569503,
      "num_tokens": 77164984.0,
      "step": 33680
    },
    {
      "entropy": 5.781825399398803,
      "epoch": 3.3298734677738238,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003917471962401762,
      "loss": 5.0512,
      "mean_token_accuracy": 0.2145717367529869,
      "num_tokens": 77176870.0,
      "step": 33685
    },
    {
      "entropy": 5.727587509155273,
      "epoch": 3.330367734282325,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003917170271890724,
      "loss": 4.8823,
      "mean_token_accuracy": 0.23652447909116744,
      "num_tokens": 77188376.0,
      "step": 33690
    },
    {
      "entropy": 5.741088819503784,
      "epoch": 3.3308620007908263,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003916868552667118,
      "loss": 4.9261,
      "mean_token_accuracy": 0.22716683596372605,
      "num_tokens": 77198556.0,
      "step": 33695
    },
    {
      "entropy": 5.712704372406006,
      "epoch": 3.331356267299328,
      "grad_norm": 1.296875,
      "learning_rate": 0.00039165668047383634,
      "loss": 4.9995,
      "mean_token_accuracy": 0.22532976418733597,
      "num_tokens": 77209409.0,
      "step": 33700
    },
    {
      "entropy": 5.731375885009766,
      "epoch": 3.3318505338078293,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003916265028111886,
      "loss": 4.9882,
      "mean_token_accuracy": 0.22507862895727157,
      "num_tokens": 77219986.0,
      "step": 33705
    },
    {
      "entropy": 5.742857646942139,
      "epoch": 3.3323448003163305,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039159632227951066,
      "loss": 4.9876,
      "mean_token_accuracy": 0.22230661511421204,
      "num_tokens": 77231380.0,
      "step": 33710
    },
    {
      "entropy": 5.69151201248169,
      "epoch": 3.3328390668248318,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000391566138879545,
      "loss": 4.9834,
      "mean_token_accuracy": 0.2237701430916786,
      "num_tokens": 77241765.0,
      "step": 33715
    },
    {
      "entropy": 5.7134480476379395,
      "epoch": 3.3333333333333335,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003915359526120344,
      "loss": 4.992,
      "mean_token_accuracy": 0.2184361070394516,
      "num_tokens": 77256091.0,
      "step": 33720
    },
    {
      "entropy": 5.804354429244995,
      "epoch": 3.3338275998418347,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003915057634777213,
      "loss": 5.0206,
      "mean_token_accuracy": 0.2190558835864067,
      "num_tokens": 77267084.0,
      "step": 33725
    },
    {
      "entropy": 5.811003684997559,
      "epoch": 3.334321866350336,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003914755714773483,
      "loss": 5.0064,
      "mean_token_accuracy": 0.2206461176276207,
      "num_tokens": 77278435.0,
      "step": 33730
    },
    {
      "entropy": 5.734785461425782,
      "epoch": 3.3348161328588377,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003914453766116581,
      "loss": 4.9843,
      "mean_token_accuracy": 0.21770980805158616,
      "num_tokens": 77290971.0,
      "step": 33735
    },
    {
      "entropy": 5.719811487197876,
      "epoch": 3.335310399367339,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003914151788813937,
      "loss": 4.9396,
      "mean_token_accuracy": 0.22535114288330077,
      "num_tokens": 77302704.0,
      "step": 33740
    },
    {
      "entropy": 5.736298704147339,
      "epoch": 3.3358046658758402,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039138497828729773,
      "loss": 4.8979,
      "mean_token_accuracy": 0.23244884014129638,
      "num_tokens": 77314107.0,
      "step": 33745
    },
    {
      "entropy": 5.790796136856079,
      "epoch": 3.3362989323843415,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003913547748301133,
      "loss": 4.9569,
      "mean_token_accuracy": 0.22317126095294954,
      "num_tokens": 77324040.0,
      "step": 33750
    },
    {
      "entropy": 5.666530466079712,
      "epoch": 3.336793198892843,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003913245685105834,
      "loss": 4.9065,
      "mean_token_accuracy": 0.23847365975379944,
      "num_tokens": 77336471.0,
      "step": 33755
    },
    {
      "entropy": 5.766614103317261,
      "epoch": 3.3372874654013445,
      "grad_norm": 1.3125,
      "learning_rate": 0.00039129435932945094,
      "loss": 5.0058,
      "mean_token_accuracy": 0.21859643161296843,
      "num_tokens": 77346947.0,
      "step": 33760
    },
    {
      "entropy": 5.727384328842163,
      "epoch": 3.3377817319098457,
      "grad_norm": 1.25,
      "learning_rate": 0.0003912641472874593,
      "loss": 4.8949,
      "mean_token_accuracy": 0.2340430051088333,
      "num_tokens": 77357995.0,
      "step": 33765
    },
    {
      "entropy": 5.711845111846924,
      "epoch": 3.338275998418347,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003912339323853516,
      "loss": 4.9217,
      "mean_token_accuracy": 0.22435169368982316,
      "num_tokens": 77368910.0,
      "step": 33770
    },
    {
      "entropy": 5.801765012741089,
      "epoch": 3.3387702649268487,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039120371462387105,
      "loss": 5.0073,
      "mean_token_accuracy": 0.22015341371297836,
      "num_tokens": 77381010.0,
      "step": 33775
    },
    {
      "entropy": 5.802192831039429,
      "epoch": 3.33926453143535,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039117349400376116,
      "loss": 5.0438,
      "mean_token_accuracy": 0.21507867872714997,
      "num_tokens": 77392182.0,
      "step": 33780
    },
    {
      "entropy": 5.819150400161743,
      "epoch": 3.339758797943851,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039114327052576525,
      "loss": 5.0453,
      "mean_token_accuracy": 0.21451151072978974,
      "num_tokens": 77403574.0,
      "step": 33785
    },
    {
      "entropy": 5.636699295043945,
      "epoch": 3.3402530644523525,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00039111304419062686,
      "loss": 4.846,
      "mean_token_accuracy": 0.23021390438079833,
      "num_tokens": 77414941.0,
      "step": 33790
    },
    {
      "entropy": 5.7312331199646,
      "epoch": 3.340747330960854,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003910828149990897,
      "loss": 5.0167,
      "mean_token_accuracy": 0.21681069284677507,
      "num_tokens": 77426361.0,
      "step": 33795
    },
    {
      "entropy": 5.774086141586304,
      "epoch": 3.3412415974693555,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039105258295189717,
      "loss": 5.0181,
      "mean_token_accuracy": 0.22102180272340774,
      "num_tokens": 77438366.0,
      "step": 33800
    },
    {
      "entropy": 5.742863655090332,
      "epoch": 3.3417358639778567,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00039102234804979307,
      "loss": 4.9391,
      "mean_token_accuracy": 0.23176540285348893,
      "num_tokens": 77448662.0,
      "step": 33805
    },
    {
      "entropy": 5.773565483093262,
      "epoch": 3.3422301304863584,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003909921102935213,
      "loss": 4.9115,
      "mean_token_accuracy": 0.23041762858629228,
      "num_tokens": 77459806.0,
      "step": 33810
    },
    {
      "entropy": 5.732771110534668,
      "epoch": 3.3427243969948597,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039096186968382555,
      "loss": 5.0315,
      "mean_token_accuracy": 0.21806803792715074,
      "num_tokens": 77470571.0,
      "step": 33815
    },
    {
      "entropy": 5.7407310009002686,
      "epoch": 3.343218663503361,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003909316262214498,
      "loss": 4.9568,
      "mean_token_accuracy": 0.23083746284246445,
      "num_tokens": 77483365.0,
      "step": 33820
    },
    {
      "entropy": 5.75578122138977,
      "epoch": 3.343712930011862,
      "grad_norm": 1.28125,
      "learning_rate": 0.00039090137990713817,
      "loss": 4.9077,
      "mean_token_accuracy": 0.22607372999191283,
      "num_tokens": 77493710.0,
      "step": 33825
    },
    {
      "entropy": 5.699822187423706,
      "epoch": 3.344207196520364,
      "grad_norm": 1.25,
      "learning_rate": 0.0003908711307416345,
      "loss": 4.9138,
      "mean_token_accuracy": 0.22897327244281768,
      "num_tokens": 77505153.0,
      "step": 33830
    },
    {
      "entropy": 5.792786169052124,
      "epoch": 3.344701463028865,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003908408787256831,
      "loss": 5.0652,
      "mean_token_accuracy": 0.21909737586975098,
      "num_tokens": 77516370.0,
      "step": 33835
    },
    {
      "entropy": 5.8154134273529055,
      "epoch": 3.3451957295373664,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003908106238600281,
      "loss": 5.0285,
      "mean_token_accuracy": 0.2240742027759552,
      "num_tokens": 77526660.0,
      "step": 33840
    },
    {
      "entropy": 5.666768550872803,
      "epoch": 3.345689996045868,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039078036614541376,
      "loss": 4.8717,
      "mean_token_accuracy": 0.2325726181268692,
      "num_tokens": 77538112.0,
      "step": 33845
    },
    {
      "entropy": 5.698589134216308,
      "epoch": 3.3461842625543694,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003907501055825845,
      "loss": 4.934,
      "mean_token_accuracy": 0.22542956322431565,
      "num_tokens": 77548807.0,
      "step": 33850
    },
    {
      "entropy": 5.700926065444946,
      "epoch": 3.3466785290628707,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039071984217228464,
      "loss": 4.9606,
      "mean_token_accuracy": 0.2278685450553894,
      "num_tokens": 77559803.0,
      "step": 33855
    },
    {
      "entropy": 5.77211627960205,
      "epoch": 3.347172795571372,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039068957591525873,
      "loss": 4.9621,
      "mean_token_accuracy": 0.21957990825176238,
      "num_tokens": 77570973.0,
      "step": 33860
    },
    {
      "entropy": 5.744198131561279,
      "epoch": 3.3476670620798736,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003906593068122513,
      "loss": 4.9946,
      "mean_token_accuracy": 0.22218303829431535,
      "num_tokens": 77583614.0,
      "step": 33865
    },
    {
      "entropy": 5.726515007019043,
      "epoch": 3.348161328588375,
      "grad_norm": 1.328125,
      "learning_rate": 0.000390629034864007,
      "loss": 4.9969,
      "mean_token_accuracy": 0.22657445669174195,
      "num_tokens": 77595086.0,
      "step": 33870
    },
    {
      "entropy": 5.793664932250977,
      "epoch": 3.348655595096876,
      "grad_norm": 1.3125,
      "learning_rate": 0.00039059876007127044,
      "loss": 4.9873,
      "mean_token_accuracy": 0.21717930734157562,
      "num_tokens": 77605492.0,
      "step": 33875
    },
    {
      "entropy": 5.730284833908081,
      "epoch": 3.349149861605378,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003905684824347865,
      "loss": 4.9515,
      "mean_token_accuracy": 0.2239883467555046,
      "num_tokens": 77618123.0,
      "step": 33880
    },
    {
      "entropy": 5.737873649597168,
      "epoch": 3.349644128113879,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003905382019553,
      "loss": 5.0057,
      "mean_token_accuracy": 0.21829362511634826,
      "num_tokens": 77630733.0,
      "step": 33885
    },
    {
      "entropy": 5.770794248580932,
      "epoch": 3.3501383946223804,
      "grad_norm": 1.25,
      "learning_rate": 0.00039050791863355573,
      "loss": 5.0041,
      "mean_token_accuracy": 0.22246245592832564,
      "num_tokens": 77641337.0,
      "step": 33890
    },
    {
      "entropy": 5.805846309661865,
      "epoch": 3.3506326611308817,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039047763247029876,
      "loss": 5.0197,
      "mean_token_accuracy": 0.22011965662240982,
      "num_tokens": 77652567.0,
      "step": 33895
    },
    {
      "entropy": 5.774141550064087,
      "epoch": 3.351126927639383,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003904473434662741,
      "loss": 5.0468,
      "mean_token_accuracy": 0.21931514292955398,
      "num_tokens": 77664146.0,
      "step": 33900
    },
    {
      "entropy": 5.829627990722656,
      "epoch": 3.3516211941478846,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003904170516222269,
      "loss": 4.9892,
      "mean_token_accuracy": 0.22041493356227876,
      "num_tokens": 77675406.0,
      "step": 33905
    },
    {
      "entropy": 5.744736528396606,
      "epoch": 3.352115460656386,
      "grad_norm": 1.25,
      "learning_rate": 0.0003903867569389024,
      "loss": 4.9894,
      "mean_token_accuracy": 0.22463519871234894,
      "num_tokens": 77686846.0,
      "step": 33910
    },
    {
      "entropy": 5.749107456207275,
      "epoch": 3.352609727164887,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003903564594170457,
      "loss": 4.9517,
      "mean_token_accuracy": 0.22118581533432008,
      "num_tokens": 77697136.0,
      "step": 33915
    },
    {
      "entropy": 5.853708791732788,
      "epoch": 3.353103993673389,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039032615905740226,
      "loss": 5.0276,
      "mean_token_accuracy": 0.21709294766187667,
      "num_tokens": 77708056.0,
      "step": 33920
    },
    {
      "entropy": 5.761302375793457,
      "epoch": 3.35359826018189,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003902958558607173,
      "loss": 4.9957,
      "mean_token_accuracy": 0.22170864343643187,
      "num_tokens": 77718582.0,
      "step": 33925
    },
    {
      "entropy": 5.728082466125488,
      "epoch": 3.3540925266903914,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003902655498277366,
      "loss": 4.9693,
      "mean_token_accuracy": 0.22448395192623138,
      "num_tokens": 77729388.0,
      "step": 33930
    },
    {
      "entropy": 5.7912506580352785,
      "epoch": 3.3545867931988926,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039023524095920534,
      "loss": 4.8706,
      "mean_token_accuracy": 0.2272852048277855,
      "num_tokens": 77739952.0,
      "step": 33935
    },
    {
      "entropy": 5.731177234649659,
      "epoch": 3.3550810597073943,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003902049292558693,
      "loss": 4.956,
      "mean_token_accuracy": 0.22535153180360795,
      "num_tokens": 77752415.0,
      "step": 33940
    },
    {
      "entropy": 5.673904705047607,
      "epoch": 3.3555753262158956,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00039017461471847417,
      "loss": 4.8776,
      "mean_token_accuracy": 0.22828159034252166,
      "num_tokens": 77764102.0,
      "step": 33945
    },
    {
      "entropy": 5.7191397190094,
      "epoch": 3.356069592724397,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003901442973477656,
      "loss": 4.9582,
      "mean_token_accuracy": 0.22674656957387923,
      "num_tokens": 77774908.0,
      "step": 33950
    },
    {
      "entropy": 5.76264271736145,
      "epoch": 3.3565638592328986,
      "grad_norm": 1.25,
      "learning_rate": 0.0003901139771444894,
      "loss": 5.0158,
      "mean_token_accuracy": 0.21967628598213196,
      "num_tokens": 77785913.0,
      "step": 33955
    },
    {
      "entropy": 5.729533672332764,
      "epoch": 3.3570581257414,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003900836541093916,
      "loss": 4.9711,
      "mean_token_accuracy": 0.22423508763313293,
      "num_tokens": 77797539.0,
      "step": 33960
    },
    {
      "entropy": 5.759144401550293,
      "epoch": 3.357552392249901,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000390053328243218,
      "loss": 4.9992,
      "mean_token_accuracy": 0.22237349003553392,
      "num_tokens": 77809777.0,
      "step": 33965
    },
    {
      "entropy": 5.749133014678955,
      "epoch": 3.3580466587584024,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003900229995467147,
      "loss": 4.8975,
      "mean_token_accuracy": 0.23006992638111115,
      "num_tokens": 77820401.0,
      "step": 33970
    },
    {
      "entropy": 5.719423770904541,
      "epoch": 3.358540925266904,
      "grad_norm": 1.25,
      "learning_rate": 0.0003899926680206278,
      "loss": 4.9569,
      "mean_token_accuracy": 0.2166220024228096,
      "num_tokens": 77832235.0,
      "step": 33975
    },
    {
      "entropy": 5.7361043930053714,
      "epoch": 3.3590351917754053,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038996233366570335,
      "loss": 4.9909,
      "mean_token_accuracy": 0.22359827011823655,
      "num_tokens": 77843938.0,
      "step": 33980
    },
    {
      "entropy": 5.769420003890991,
      "epoch": 3.3595294582839066,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038993199648268775,
      "loss": 4.9789,
      "mean_token_accuracy": 0.2152797371149063,
      "num_tokens": 77853921.0,
      "step": 33985
    },
    {
      "entropy": 5.719605922698975,
      "epoch": 3.3600237247924083,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00038990165647232713,
      "loss": 4.8426,
      "mean_token_accuracy": 0.23854149281978607,
      "num_tokens": 77864661.0,
      "step": 33990
    },
    {
      "entropy": 5.845148324966431,
      "epoch": 3.3605179913009096,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003898713136353679,
      "loss": 5.1118,
      "mean_token_accuracy": 0.2067774310708046,
      "num_tokens": 77875743.0,
      "step": 33995
    },
    {
      "entropy": 5.712846136093139,
      "epoch": 3.361012257809411,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003898409679725566,
      "loss": 4.9555,
      "mean_token_accuracy": 0.2252158537507057,
      "num_tokens": 77888575.0,
      "step": 34000
    },
    {
      "entropy": 5.883233499526978,
      "epoch": 3.361506524317912,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003898106194846396,
      "loss": 5.1055,
      "mean_token_accuracy": 0.21185964792966844,
      "num_tokens": 77901212.0,
      "step": 34005
    },
    {
      "entropy": 5.728261423110962,
      "epoch": 3.362000790826414,
      "grad_norm": 1.3125,
      "learning_rate": 0.00038978026817236365,
      "loss": 4.9393,
      "mean_token_accuracy": 0.23533645868301392,
      "num_tokens": 77911905.0,
      "step": 34010
    },
    {
      "entropy": 5.7659642696380615,
      "epoch": 3.362495057334915,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038974991403647536,
      "loss": 4.9994,
      "mean_token_accuracy": 0.21642355918884276,
      "num_tokens": 77923710.0,
      "step": 34015
    },
    {
      "entropy": 5.74726939201355,
      "epoch": 3.3629893238434163,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00038971955707772126,
      "loss": 4.9566,
      "mean_token_accuracy": 0.2257274091243744,
      "num_tokens": 77934501.0,
      "step": 34020
    },
    {
      "entropy": 5.746541500091553,
      "epoch": 3.3634835903519176,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003896891972968482,
      "loss": 4.9232,
      "mean_token_accuracy": 0.22666034251451492,
      "num_tokens": 77944512.0,
      "step": 34025
    },
    {
      "entropy": 5.738111877441407,
      "epoch": 3.3639778568604193,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003896588346946033,
      "loss": 4.924,
      "mean_token_accuracy": 0.23085123896598816,
      "num_tokens": 77955253.0,
      "step": 34030
    },
    {
      "entropy": 5.6593996524810795,
      "epoch": 3.3644721233689205,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00038962846927173306,
      "loss": 4.8685,
      "mean_token_accuracy": 0.22945576310157775,
      "num_tokens": 77967193.0,
      "step": 34035
    },
    {
      "entropy": 5.765934610366822,
      "epoch": 3.364966389877422,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003895981010289848,
      "loss": 5.048,
      "mean_token_accuracy": 0.21797843724489213,
      "num_tokens": 77977105.0,
      "step": 34040
    },
    {
      "entropy": 5.748804426193237,
      "epoch": 3.365460656385923,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038956772996710554,
      "loss": 4.9799,
      "mean_token_accuracy": 0.2239413544535637,
      "num_tokens": 77989517.0,
      "step": 34045
    },
    {
      "entropy": 5.785615539550781,
      "epoch": 3.3659549228944248,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038953735608684216,
      "loss": 4.9963,
      "mean_token_accuracy": 0.21782848834991456,
      "num_tokens": 78000797.0,
      "step": 34050
    },
    {
      "entropy": 5.665162897109985,
      "epoch": 3.366449189402926,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003895069793889422,
      "loss": 4.8355,
      "mean_token_accuracy": 0.23953462094068528,
      "num_tokens": 78012470.0,
      "step": 34055
    },
    {
      "entropy": 5.7289807319641115,
      "epoch": 3.3669434559114273,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003894765998741528,
      "loss": 4.926,
      "mean_token_accuracy": 0.22437241077423095,
      "num_tokens": 78023774.0,
      "step": 34060
    },
    {
      "entropy": 5.734833192825318,
      "epoch": 3.367437722419929,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003894462175432212,
      "loss": 5.0059,
      "mean_token_accuracy": 0.21929133236408233,
      "num_tokens": 78035181.0,
      "step": 34065
    },
    {
      "entropy": 5.751251316070556,
      "epoch": 3.3679319889284303,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00038941583239689495,
      "loss": 4.9713,
      "mean_token_accuracy": 0.22184754461050032,
      "num_tokens": 78045728.0,
      "step": 34070
    },
    {
      "entropy": 5.838360118865967,
      "epoch": 3.3684262554369315,
      "grad_norm": 1.25,
      "learning_rate": 0.00038938544443592146,
      "loss": 5.0033,
      "mean_token_accuracy": 0.2181825593113899,
      "num_tokens": 78056675.0,
      "step": 34075
    },
    {
      "entropy": 5.721235990524292,
      "epoch": 3.368920521945433,
      "grad_norm": 1.25,
      "learning_rate": 0.0003893550536610483,
      "loss": 4.8889,
      "mean_token_accuracy": 0.2346413776278496,
      "num_tokens": 78067686.0,
      "step": 34080
    },
    {
      "entropy": 5.666837215423584,
      "epoch": 3.3694147884539345,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000389324660073023,
      "loss": 4.9535,
      "mean_token_accuracy": 0.23075952678918837,
      "num_tokens": 78078063.0,
      "step": 34085
    },
    {
      "entropy": 5.770784854888916,
      "epoch": 3.3699090549624358,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003892942636725934,
      "loss": 5.0622,
      "mean_token_accuracy": 0.21937436014413833,
      "num_tokens": 78088729.0,
      "step": 34090
    },
    {
      "entropy": 5.678276491165161,
      "epoch": 3.370403321470937,
      "grad_norm": 1.25,
      "learning_rate": 0.0003892638644605072,
      "loss": 4.8836,
      "mean_token_accuracy": 0.2351837381720543,
      "num_tokens": 78100468.0,
      "step": 34095
    },
    {
      "entropy": 5.749306583404541,
      "epoch": 3.3708975879794387,
      "grad_norm": 1.28125,
      "learning_rate": 0.00038923346243751217,
      "loss": 4.9831,
      "mean_token_accuracy": 0.2261546343564987,
      "num_tokens": 78111911.0,
      "step": 34100
    },
    {
      "entropy": 5.661248970031738,
      "epoch": 3.37139185448794,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00038920305760435626,
      "loss": 4.8768,
      "mean_token_accuracy": 0.22817135453224183,
      "num_tokens": 78122411.0,
      "step": 34105
    },
    {
      "entropy": 5.771761846542359,
      "epoch": 3.3718861209964412,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003891726499617874,
      "loss": 5.0468,
      "mean_token_accuracy": 0.21325188279151916,
      "num_tokens": 78134313.0,
      "step": 34110
    },
    {
      "entropy": 5.663639879226684,
      "epoch": 3.3723803875049425,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003891422395105536,
      "loss": 4.7984,
      "mean_token_accuracy": 0.23698732256889343,
      "num_tokens": 78144753.0,
      "step": 34115
    },
    {
      "entropy": 5.6623725414276125,
      "epoch": 3.372874654013444,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038911182625140305,
      "loss": 4.8766,
      "mean_token_accuracy": 0.22556447982788086,
      "num_tokens": 78154772.0,
      "step": 34120
    },
    {
      "entropy": 5.68983006477356,
      "epoch": 3.3733689205219455,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038908141018508387,
      "loss": 4.9448,
      "mean_token_accuracy": 0.23074347972869874,
      "num_tokens": 78166876.0,
      "step": 34125
    },
    {
      "entropy": 5.7253423690795895,
      "epoch": 3.3738631870304467,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00038905099131234433,
      "loss": 4.9218,
      "mean_token_accuracy": 0.2382257714867592,
      "num_tokens": 78177530.0,
      "step": 34130
    },
    {
      "entropy": 5.721208095550537,
      "epoch": 3.3743574535389484,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038902056963393265,
      "loss": 4.9201,
      "mean_token_accuracy": 0.2268524572253227,
      "num_tokens": 78187986.0,
      "step": 34135
    },
    {
      "entropy": 5.707645750045776,
      "epoch": 3.3748517200474497,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038899014515059735,
      "loss": 4.9532,
      "mean_token_accuracy": 0.22383246272802354,
      "num_tokens": 78199874.0,
      "step": 34140
    },
    {
      "entropy": 5.691944217681884,
      "epoch": 3.375345986555951,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003889597178630867,
      "loss": 4.9312,
      "mean_token_accuracy": 0.2287371724843979,
      "num_tokens": 78210995.0,
      "step": 34145
    },
    {
      "entropy": 5.713322114944458,
      "epoch": 3.3758402530644522,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003889292877721493,
      "loss": 4.91,
      "mean_token_accuracy": 0.22524206042289735,
      "num_tokens": 78221792.0,
      "step": 34150
    },
    {
      "entropy": 5.653506183624268,
      "epoch": 3.3763345195729535,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038889885487853384,
      "loss": 4.914,
      "mean_token_accuracy": 0.23960447311401367,
      "num_tokens": 78232799.0,
      "step": 34155
    },
    {
      "entropy": 5.790098381042481,
      "epoch": 3.376828786081455,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003888684191829889,
      "loss": 4.9957,
      "mean_token_accuracy": 0.2183838203549385,
      "num_tokens": 78243784.0,
      "step": 34160
    },
    {
      "entropy": 5.733755207061767,
      "epoch": 3.3773230525899565,
      "grad_norm": 1.25,
      "learning_rate": 0.00038883798068626305,
      "loss": 4.8787,
      "mean_token_accuracy": 0.23317979872226716,
      "num_tokens": 78255161.0,
      "step": 34165
    },
    {
      "entropy": 5.674816513061524,
      "epoch": 3.3778173190984577,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038880753938910523,
      "loss": 4.9094,
      "mean_token_accuracy": 0.2264370948076248,
      "num_tokens": 78266462.0,
      "step": 34170
    },
    {
      "entropy": 5.724034690856934,
      "epoch": 3.3783115856069594,
      "grad_norm": 1.34375,
      "learning_rate": 0.00038877709529226446,
      "loss": 4.9781,
      "mean_token_accuracy": 0.21809536963701248,
      "num_tokens": 78278136.0,
      "step": 34175
    },
    {
      "entropy": 5.7751367568969725,
      "epoch": 3.3788058521154607,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003887466483964893,
      "loss": 4.9866,
      "mean_token_accuracy": 0.21861551254987716,
      "num_tokens": 78290046.0,
      "step": 34180
    },
    {
      "entropy": 5.784373760223389,
      "epoch": 3.379300118623962,
      "grad_norm": 1.125,
      "learning_rate": 0.00038871619870252904,
      "loss": 4.9504,
      "mean_token_accuracy": 0.22399170994758605,
      "num_tokens": 78301636.0,
      "step": 34185
    },
    {
      "entropy": 5.7526853561401365,
      "epoch": 3.379794385132463,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038868574621113265,
      "loss": 4.9532,
      "mean_token_accuracy": 0.21787270456552504,
      "num_tokens": 78313326.0,
      "step": 34190
    },
    {
      "entropy": 5.752616882324219,
      "epoch": 3.380288651640965,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003886552909230492,
      "loss": 4.9943,
      "mean_token_accuracy": 0.21607912629842757,
      "num_tokens": 78325214.0,
      "step": 34195
    },
    {
      "entropy": 5.721040821075439,
      "epoch": 3.380782918149466,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000388624832839028,
      "loss": 4.9227,
      "mean_token_accuracy": 0.22896347045898438,
      "num_tokens": 78336461.0,
      "step": 34200
    },
    {
      "entropy": 5.739862871170044,
      "epoch": 3.3812771846579674,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00038859437195981833,
      "loss": 4.9979,
      "mean_token_accuracy": 0.22157906442880632,
      "num_tokens": 78347852.0,
      "step": 34205
    },
    {
      "entropy": 5.715168809890747,
      "epoch": 3.381771451166469,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00038856390828616945,
      "loss": 4.9565,
      "mean_token_accuracy": 0.2184998258948326,
      "num_tokens": 78358736.0,
      "step": 34210
    },
    {
      "entropy": 5.847535943984985,
      "epoch": 3.3822657176749704,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003885334418188308,
      "loss": 5.1115,
      "mean_token_accuracy": 0.213736130297184,
      "num_tokens": 78370842.0,
      "step": 34215
    },
    {
      "entropy": 5.719734477996826,
      "epoch": 3.3827599841834717,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038850297255855186,
      "loss": 4.8558,
      "mean_token_accuracy": 0.22974586635828018,
      "num_tokens": 78382159.0,
      "step": 34220
    },
    {
      "entropy": 5.758401536941529,
      "epoch": 3.383254250691973,
      "grad_norm": 1.359375,
      "learning_rate": 0.00038847250050608217,
      "loss": 5.067,
      "mean_token_accuracy": 0.21461165696382523,
      "num_tokens": 78392130.0,
      "step": 34225
    },
    {
      "entropy": 5.717522811889649,
      "epoch": 3.3837485172004746,
      "grad_norm": 1.25,
      "learning_rate": 0.00038844202566217146,
      "loss": 4.9555,
      "mean_token_accuracy": 0.2256264492869377,
      "num_tokens": 78403271.0,
      "step": 34230
    },
    {
      "entropy": 5.760998725891113,
      "epoch": 3.384242783708976,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038841154802756925,
      "loss": 4.9809,
      "mean_token_accuracy": 0.21933382898569107,
      "num_tokens": 78414621.0,
      "step": 34235
    },
    {
      "entropy": 5.854096603393555,
      "epoch": 3.384737050217477,
      "grad_norm": 1.0,
      "learning_rate": 0.00038838106760302537,
      "loss": 5.0974,
      "mean_token_accuracy": 0.2113494783639908,
      "num_tokens": 78427708.0,
      "step": 34240
    },
    {
      "entropy": 5.745160102844238,
      "epoch": 3.385231316725979,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038835058438928965,
      "loss": 4.9329,
      "mean_token_accuracy": 0.2261657029390335,
      "num_tokens": 78438736.0,
      "step": 34245
    },
    {
      "entropy": 5.708537960052491,
      "epoch": 3.38572558323448,
      "grad_norm": 1.3125,
      "learning_rate": 0.000388320098387112,
      "loss": 4.9668,
      "mean_token_accuracy": 0.2280779853463173,
      "num_tokens": 78449118.0,
      "step": 34250
    },
    {
      "entropy": 5.756624555587768,
      "epoch": 3.3862198497429814,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00038828960959724225,
      "loss": 5.0186,
      "mean_token_accuracy": 0.2169206976890564,
      "num_tokens": 78460027.0,
      "step": 34255
    },
    {
      "entropy": 5.727454996109008,
      "epoch": 3.3867141162514827,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038825911802043057,
      "loss": 4.9018,
      "mean_token_accuracy": 0.22845005989074707,
      "num_tokens": 78471251.0,
      "step": 34260
    },
    {
      "entropy": 5.750979423522949,
      "epoch": 3.3872083827599844,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00038822862365742704,
      "loss": 4.9936,
      "mean_token_accuracy": 0.22444596588611604,
      "num_tokens": 78482042.0,
      "step": 34265
    },
    {
      "entropy": 5.675802326202392,
      "epoch": 3.3877026492684856,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003881981265089818,
      "loss": 4.8917,
      "mean_token_accuracy": 0.22728991210460664,
      "num_tokens": 78492941.0,
      "step": 34270
    },
    {
      "entropy": 5.760136318206787,
      "epoch": 3.388196915776987,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003881676265758451,
      "loss": 4.9747,
      "mean_token_accuracy": 0.21787781566381453,
      "num_tokens": 78506108.0,
      "step": 34275
    },
    {
      "entropy": 5.766173553466797,
      "epoch": 3.388691182285488,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038813712385876714,
      "loss": 4.9463,
      "mean_token_accuracy": 0.22834778130054473,
      "num_tokens": 78517784.0,
      "step": 34280
    },
    {
      "entropy": 5.7463782787322994,
      "epoch": 3.38918544879399,
      "grad_norm": 1.328125,
      "learning_rate": 0.00038810661835849845,
      "loss": 4.9921,
      "mean_token_accuracy": 0.22301958352327347,
      "num_tokens": 78528795.0,
      "step": 34285
    },
    {
      "entropy": 5.692702054977417,
      "epoch": 3.389679715302491,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038807611007578947,
      "loss": 4.8962,
      "mean_token_accuracy": 0.2358813464641571,
      "num_tokens": 78540724.0,
      "step": 34290
    },
    {
      "entropy": 5.823853397369385,
      "epoch": 3.3901739818109924,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003880455990113905,
      "loss": 5.0575,
      "mean_token_accuracy": 0.21508000642061234,
      "num_tokens": 78552569.0,
      "step": 34295
    },
    {
      "entropy": 5.752160930633545,
      "epoch": 3.3906682483194936,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003880150851660524,
      "loss": 4.9983,
      "mean_token_accuracy": 0.22020193338394164,
      "num_tokens": 78564785.0,
      "step": 34300
    },
    {
      "entropy": 5.737940406799316,
      "epoch": 3.3911625148279954,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00038798456854052555,
      "loss": 4.9843,
      "mean_token_accuracy": 0.22239601612091064,
      "num_tokens": 78576323.0,
      "step": 34305
    },
    {
      "entropy": 5.802496290206909,
      "epoch": 3.3916567813364966,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00038795404913556083,
      "loss": 5.0406,
      "mean_token_accuracy": 0.21609234064817429,
      "num_tokens": 78586978.0,
      "step": 34310
    },
    {
      "entropy": 5.751192998886109,
      "epoch": 3.392151047844998,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038792352695190895,
      "loss": 5.0012,
      "mean_token_accuracy": 0.22484201490879058,
      "num_tokens": 78597591.0,
      "step": 34315
    },
    {
      "entropy": 5.71071228981018,
      "epoch": 3.3926453143534996,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003878930019903209,
      "loss": 4.9229,
      "mean_token_accuracy": 0.225471656024456,
      "num_tokens": 78608189.0,
      "step": 34320
    },
    {
      "entropy": 5.779983472824097,
      "epoch": 3.393139580862001,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003878624742515474,
      "loss": 5.0123,
      "mean_token_accuracy": 0.22660823315382003,
      "num_tokens": 78620170.0,
      "step": 34325
    },
    {
      "entropy": 5.783258056640625,
      "epoch": 3.393633847370502,
      "grad_norm": 1.25,
      "learning_rate": 0.00038783194373633956,
      "loss": 4.9791,
      "mean_token_accuracy": 0.22277851551771163,
      "num_tokens": 78632423.0,
      "step": 34330
    },
    {
      "entropy": 5.73965163230896,
      "epoch": 3.3941281138790034,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038780141044544836,
      "loss": 4.9949,
      "mean_token_accuracy": 0.22563783079385757,
      "num_tokens": 78644138.0,
      "step": 34335
    },
    {
      "entropy": 5.747307682037354,
      "epoch": 3.394622380387505,
      "grad_norm": 1.078125,
      "learning_rate": 0.000387770874379625,
      "loss": 4.9996,
      "mean_token_accuracy": 0.2226667121052742,
      "num_tokens": 78655659.0,
      "step": 34340
    },
    {
      "entropy": 5.647861051559448,
      "epoch": 3.3951166468960063,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003877403355396206,
      "loss": 4.9355,
      "mean_token_accuracy": 0.22966020852327346,
      "num_tokens": 78665625.0,
      "step": 34345
    },
    {
      "entropy": 5.755460166931153,
      "epoch": 3.3956109134045076,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003877097939261865,
      "loss": 4.9871,
      "mean_token_accuracy": 0.22006723731756211,
      "num_tokens": 78677253.0,
      "step": 34350
    },
    {
      "entropy": 5.793676900863647,
      "epoch": 3.3961051799130093,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038767924954007387,
      "loss": 4.9635,
      "mean_token_accuracy": 0.22454598397016526,
      "num_tokens": 78688574.0,
      "step": 34355
    },
    {
      "entropy": 5.745223140716552,
      "epoch": 3.3965994464215106,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003876487023820343,
      "loss": 4.9089,
      "mean_token_accuracy": 0.23061376959085464,
      "num_tokens": 78700537.0,
      "step": 34360
    },
    {
      "entropy": 5.681247329711914,
      "epoch": 3.397093712930012,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00038761815245281925,
      "loss": 4.8828,
      "mean_token_accuracy": 0.2323605790734291,
      "num_tokens": 78710257.0,
      "step": 34365
    },
    {
      "entropy": 5.729528570175171,
      "epoch": 3.397587979438513,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038758759975318006,
      "loss": 4.9196,
      "mean_token_accuracy": 0.22853223234415054,
      "num_tokens": 78722503.0,
      "step": 34370
    },
    {
      "entropy": 5.733088397979737,
      "epoch": 3.398082245947015,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003875570442838686,
      "loss": 4.9795,
      "mean_token_accuracy": 0.22297899872064592,
      "num_tokens": 78734476.0,
      "step": 34375
    },
    {
      "entropy": 5.839818811416626,
      "epoch": 3.398576512455516,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003875264860456362,
      "loss": 5.0198,
      "mean_token_accuracy": 0.22072531133890153,
      "num_tokens": 78745685.0,
      "step": 34380
    },
    {
      "entropy": 5.812339353561401,
      "epoch": 3.3990707789640173,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038749592503923484,
      "loss": 5.0204,
      "mean_token_accuracy": 0.22023228853940963,
      "num_tokens": 78758977.0,
      "step": 34385
    },
    {
      "entropy": 5.717302942276001,
      "epoch": 3.399565045472519,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003874653612654164,
      "loss": 4.9407,
      "mean_token_accuracy": 0.22151706963777543,
      "num_tokens": 78770702.0,
      "step": 34390
    },
    {
      "entropy": 5.739876079559326,
      "epoch": 3.4000593119810203,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003874347947249325,
      "loss": 4.9821,
      "mean_token_accuracy": 0.2196206584572792,
      "num_tokens": 78781763.0,
      "step": 34395
    },
    {
      "entropy": 5.770134162902832,
      "epoch": 3.4005535784895216,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00038740422541853523,
      "loss": 4.9801,
      "mean_token_accuracy": 0.2222976878285408,
      "num_tokens": 78792341.0,
      "step": 34400
    },
    {
      "entropy": 5.737035751342773,
      "epoch": 3.401047844998023,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038737365334697656,
      "loss": 4.9322,
      "mean_token_accuracy": 0.22510787844657898,
      "num_tokens": 78803947.0,
      "step": 34405
    },
    {
      "entropy": 5.678115749359131,
      "epoch": 3.401542111506524,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00038734307851100863,
      "loss": 4.8218,
      "mean_token_accuracy": 0.23743830472230912,
      "num_tokens": 78815503.0,
      "step": 34410
    },
    {
      "entropy": 5.744970846176147,
      "epoch": 3.402036378015026,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003873125009113835,
      "loss": 4.986,
      "mean_token_accuracy": 0.21829708367586137,
      "num_tokens": 78826206.0,
      "step": 34415
    },
    {
      "entropy": 5.757236957550049,
      "epoch": 3.402530644523527,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038728192054885346,
      "loss": 4.9265,
      "mean_token_accuracy": 0.22752766460180282,
      "num_tokens": 78837472.0,
      "step": 34420
    },
    {
      "entropy": 5.805315256118774,
      "epoch": 3.4030249110320283,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003872513374241707,
      "loss": 5.0915,
      "mean_token_accuracy": 0.2133657917380333,
      "num_tokens": 78848005.0,
      "step": 34425
    },
    {
      "entropy": 5.755751466751098,
      "epoch": 3.40351917754053,
      "grad_norm": 1.25,
      "learning_rate": 0.00038722075153808764,
      "loss": 4.9407,
      "mean_token_accuracy": 0.223907034099102,
      "num_tokens": 78859577.0,
      "step": 34430
    },
    {
      "entropy": 5.769728469848633,
      "epoch": 3.4040134440490313,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003871901628913568,
      "loss": 4.989,
      "mean_token_accuracy": 0.22358810752630234,
      "num_tokens": 78871527.0,
      "step": 34435
    },
    {
      "entropy": 5.739281797409058,
      "epoch": 3.4045077105575325,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003871595714847304,
      "loss": 4.9777,
      "mean_token_accuracy": 0.22732713520526887,
      "num_tokens": 78882557.0,
      "step": 34440
    },
    {
      "entropy": 5.738371181488037,
      "epoch": 3.405001977066034,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00038712897731896126,
      "loss": 4.9446,
      "mean_token_accuracy": 0.22249909490346909,
      "num_tokens": 78893191.0,
      "step": 34445
    },
    {
      "entropy": 5.758450698852539,
      "epoch": 3.4054962435745355,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003870983803948017,
      "loss": 4.9865,
      "mean_token_accuracy": 0.22425025403499604,
      "num_tokens": 78904397.0,
      "step": 34450
    },
    {
      "entropy": 5.768943786621094,
      "epoch": 3.4059905100830368,
      "grad_norm": 1.125,
      "learning_rate": 0.0003870677807130047,
      "loss": 4.9867,
      "mean_token_accuracy": 0.22557110637426375,
      "num_tokens": 78916765.0,
      "step": 34455
    },
    {
      "entropy": 5.783790493011475,
      "epoch": 3.406484776591538,
      "grad_norm": 1.21875,
      "learning_rate": 0.000387037178274323,
      "loss": 4.9613,
      "mean_token_accuracy": 0.21932793408632278,
      "num_tokens": 78929745.0,
      "step": 34460
    },
    {
      "entropy": 5.786370515823364,
      "epoch": 3.4069790431000397,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038700657307950926,
      "loss": 4.9499,
      "mean_token_accuracy": 0.22982837110757828,
      "num_tokens": 78941158.0,
      "step": 34465
    },
    {
      "entropy": 5.691815614700317,
      "epoch": 3.407473309608541,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038697596512931643,
      "loss": 4.9147,
      "mean_token_accuracy": 0.22486531883478164,
      "num_tokens": 78953465.0,
      "step": 34470
    },
    {
      "entropy": 5.7444963455200195,
      "epoch": 3.4079675761170423,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003869453544244975,
      "loss": 4.9709,
      "mean_token_accuracy": 0.2220073387026787,
      "num_tokens": 78964841.0,
      "step": 34475
    },
    {
      "entropy": 5.760384178161621,
      "epoch": 3.4084618426255435,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038691474096580545,
      "loss": 4.9807,
      "mean_token_accuracy": 0.22266217917203904,
      "num_tokens": 78975173.0,
      "step": 34480
    },
    {
      "entropy": 5.739780139923096,
      "epoch": 3.4089561091340452,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003868841247539934,
      "loss": 4.9677,
      "mean_token_accuracy": 0.2278680145740509,
      "num_tokens": 78986020.0,
      "step": 34485
    },
    {
      "entropy": 5.729197072982788,
      "epoch": 3.4094503756425465,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003868535057898146,
      "loss": 4.9487,
      "mean_token_accuracy": 0.2223416492342949,
      "num_tokens": 78997906.0,
      "step": 34490
    },
    {
      "entropy": 5.822689771652222,
      "epoch": 3.4099446421510478,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00038682288407402215,
      "loss": 5.0211,
      "mean_token_accuracy": 0.21067382842302323,
      "num_tokens": 79009322.0,
      "step": 34495
    },
    {
      "entropy": 5.68965163230896,
      "epoch": 3.4104389086595495,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003867922596073694,
      "loss": 4.8915,
      "mean_token_accuracy": 0.23269434571266173,
      "num_tokens": 79020940.0,
      "step": 34500
    },
    {
      "entropy": 5.772840309143066,
      "epoch": 3.4109331751680507,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003867616323906098,
      "loss": 5.0165,
      "mean_token_accuracy": 0.22040231227874757,
      "num_tokens": 79033002.0,
      "step": 34505
    },
    {
      "entropy": 5.682721948623657,
      "epoch": 3.411427441676552,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038673100242449666,
      "loss": 4.8826,
      "mean_token_accuracy": 0.23146036714315416,
      "num_tokens": 79044129.0,
      "step": 34510
    },
    {
      "entropy": 5.685306692123413,
      "epoch": 3.4119217081850532,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003867003697097835,
      "loss": 4.9396,
      "mean_token_accuracy": 0.23317753970623017,
      "num_tokens": 79054961.0,
      "step": 34515
    },
    {
      "entropy": 5.775302600860596,
      "epoch": 3.412415974693555,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003866697342472239,
      "loss": 4.9874,
      "mean_token_accuracy": 0.21745218336582184,
      "num_tokens": 79066136.0,
      "step": 34520
    },
    {
      "entropy": 5.715613746643067,
      "epoch": 3.412910241202056,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003866390960375716,
      "loss": 4.9742,
      "mean_token_accuracy": 0.21906571388244628,
      "num_tokens": 79077821.0,
      "step": 34525
    },
    {
      "entropy": 5.722734498977661,
      "epoch": 3.4134045077105575,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003866084550815801,
      "loss": 4.9336,
      "mean_token_accuracy": 0.2199051097035408,
      "num_tokens": 79087554.0,
      "step": 34530
    },
    {
      "entropy": 5.8052552223205565,
      "epoch": 3.4138987742190587,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00038657781138000337,
      "loss": 5.0504,
      "mean_token_accuracy": 0.22023759484291078,
      "num_tokens": 79098434.0,
      "step": 34535
    },
    {
      "entropy": 5.769784832000733,
      "epoch": 3.4143930407275604,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00038654716493359515,
      "loss": 4.9599,
      "mean_token_accuracy": 0.22111113965511323,
      "num_tokens": 79108962.0,
      "step": 34540
    },
    {
      "entropy": 5.662703180313111,
      "epoch": 3.4148873072360617,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003865165157431093,
      "loss": 4.9399,
      "mean_token_accuracy": 0.2254787415266037,
      "num_tokens": 79119592.0,
      "step": 34545
    },
    {
      "entropy": 5.722964334487915,
      "epoch": 3.415381573744563,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003864858638093,
      "loss": 4.9078,
      "mean_token_accuracy": 0.22666240334510804,
      "num_tokens": 79130170.0,
      "step": 34550
    },
    {
      "entropy": 5.752389621734619,
      "epoch": 3.4158758402530642,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00038645520913292117,
      "loss": 4.9647,
      "mean_token_accuracy": 0.22595714628696442,
      "num_tokens": 79140949.0,
      "step": 34555
    },
    {
      "entropy": 5.750032234191894,
      "epoch": 3.416370106761566,
      "grad_norm": 1.34375,
      "learning_rate": 0.00038642455171472676,
      "loss": 4.9558,
      "mean_token_accuracy": 0.21651601791381836,
      "num_tokens": 79151508.0,
      "step": 34560
    },
    {
      "entropy": 5.740336894989014,
      "epoch": 3.416864373270067,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003863938915554711,
      "loss": 4.9761,
      "mean_token_accuracy": 0.22562813460826875,
      "num_tokens": 79163062.0,
      "step": 34565
    },
    {
      "entropy": 5.781988668441772,
      "epoch": 3.4173586397785685,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003863632286559085,
      "loss": 4.9492,
      "mean_token_accuracy": 0.22952550053596496,
      "num_tokens": 79174025.0,
      "step": 34570
    },
    {
      "entropy": 5.718377637863159,
      "epoch": 3.41785290628707,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003863325630167932,
      "loss": 4.8672,
      "mean_token_accuracy": 0.23368196040391923,
      "num_tokens": 79185169.0,
      "step": 34575
    },
    {
      "entropy": 5.7356866836547855,
      "epoch": 3.4183471727955714,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00038630189463887955,
      "loss": 4.9626,
      "mean_token_accuracy": 0.2266966313123703,
      "num_tokens": 79196503.0,
      "step": 34580
    },
    {
      "entropy": 5.738524389266968,
      "epoch": 3.4188414393040727,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038627122352292206,
      "loss": 4.9726,
      "mean_token_accuracy": 0.21931502372026443,
      "num_tokens": 79207887.0,
      "step": 34585
    },
    {
      "entropy": 5.715691423416137,
      "epoch": 3.419335705812574,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038624054966967516,
      "loss": 4.9127,
      "mean_token_accuracy": 0.22652419060468673,
      "num_tokens": 79219339.0,
      "step": 34590
    },
    {
      "entropy": 5.744101619720459,
      "epoch": 3.4198299723210757,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038620987307989347,
      "loss": 4.9965,
      "mean_token_accuracy": 0.21834992617368698,
      "num_tokens": 79231454.0,
      "step": 34595
    },
    {
      "entropy": 5.808569431304932,
      "epoch": 3.420324238829577,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038617919375433176,
      "loss": 4.966,
      "mean_token_accuracy": 0.22571472376585006,
      "num_tokens": 79241761.0,
      "step": 34600
    },
    {
      "entropy": 5.787451457977295,
      "epoch": 3.420818505338078,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038614851169374456,
      "loss": 4.9325,
      "mean_token_accuracy": 0.22021746039390563,
      "num_tokens": 79252272.0,
      "step": 34605
    },
    {
      "entropy": 5.733857202529907,
      "epoch": 3.42131277184658,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003861178268988867,
      "loss": 4.9092,
      "mean_token_accuracy": 0.22506607919931412,
      "num_tokens": 79263612.0,
      "step": 34610
    },
    {
      "entropy": 5.6819069385528564,
      "epoch": 3.421807038355081,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003860871393705131,
      "loss": 4.9031,
      "mean_token_accuracy": 0.22698323726654052,
      "num_tokens": 79275361.0,
      "step": 34615
    },
    {
      "entropy": 5.786751937866211,
      "epoch": 3.4223013048635824,
      "grad_norm": 1.3125,
      "learning_rate": 0.00038605644910937856,
      "loss": 4.9472,
      "mean_token_accuracy": 0.22897537648677826,
      "num_tokens": 79285277.0,
      "step": 34620
    },
    {
      "entropy": 5.607639026641846,
      "epoch": 3.4227955713720837,
      "grad_norm": 1.25,
      "learning_rate": 0.00038602575611623827,
      "loss": 4.8124,
      "mean_token_accuracy": 0.23914988487958908,
      "num_tokens": 79296713.0,
      "step": 34625
    },
    {
      "entropy": 5.741404914855957,
      "epoch": 3.4232898378805854,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000385995060391847,
      "loss": 4.9703,
      "mean_token_accuracy": 0.21990007609128953,
      "num_tokens": 79308363.0,
      "step": 34630
    },
    {
      "entropy": 5.7628120422363285,
      "epoch": 3.4237841043890866,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038596436193696015,
      "loss": 5.0281,
      "mean_token_accuracy": 0.21933696717023848,
      "num_tokens": 79319626.0,
      "step": 34635
    },
    {
      "entropy": 5.733559322357178,
      "epoch": 3.424278370897588,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003859336607523327,
      "loss": 4.9364,
      "mean_token_accuracy": 0.2214723452925682,
      "num_tokens": 79330628.0,
      "step": 34640
    },
    {
      "entropy": 5.701859998703003,
      "epoch": 3.424772637406089,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00038590295683872005,
      "loss": 4.9285,
      "mean_token_accuracy": 0.2218850940465927,
      "num_tokens": 79342323.0,
      "step": 34645
    },
    {
      "entropy": 5.684063196182251,
      "epoch": 3.425266903914591,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003858722501968774,
      "loss": 4.8474,
      "mean_token_accuracy": 0.23856300562620164,
      "num_tokens": 79354108.0,
      "step": 34650
    },
    {
      "entropy": 5.72578239440918,
      "epoch": 3.425761170423092,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038584154082756036,
      "loss": 4.9855,
      "mean_token_accuracy": 0.2311626449227333,
      "num_tokens": 79365494.0,
      "step": 34655
    },
    {
      "entropy": 5.635528802871704,
      "epoch": 3.4262554369315934,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000385810828731524,
      "loss": 4.8032,
      "mean_token_accuracy": 0.23252518624067306,
      "num_tokens": 79376881.0,
      "step": 34660
    },
    {
      "entropy": 5.7045622825622555,
      "epoch": 3.4267497034400947,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003857801139095242,
      "loss": 4.905,
      "mean_token_accuracy": 0.22882604449987412,
      "num_tokens": 79388154.0,
      "step": 34665
    },
    {
      "entropy": 5.810230112075805,
      "epoch": 3.4272439699485964,
      "grad_norm": 1.375,
      "learning_rate": 0.00038574939636231635,
      "loss": 4.9903,
      "mean_token_accuracy": 0.22079915851354598,
      "num_tokens": 79398995.0,
      "step": 34670
    },
    {
      "entropy": 5.750847053527832,
      "epoch": 3.4277382364570976,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003857186760906562,
      "loss": 4.9885,
      "mean_token_accuracy": 0.22457312494516374,
      "num_tokens": 79410697.0,
      "step": 34675
    },
    {
      "entropy": 5.724606370925903,
      "epoch": 3.428232502965599,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038568795309529947,
      "loss": 4.9884,
      "mean_token_accuracy": 0.22242140173912048,
      "num_tokens": 79421862.0,
      "step": 34680
    },
    {
      "entropy": 5.786133861541748,
      "epoch": 3.4287267694741006,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00038565722737700185,
      "loss": 4.9758,
      "mean_token_accuracy": 0.22239427864551545,
      "num_tokens": 79432610.0,
      "step": 34685
    },
    {
      "entropy": 5.7748418807983395,
      "epoch": 3.429221035982602,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00038562649893651924,
      "loss": 4.9847,
      "mean_token_accuracy": 0.22598233222961425,
      "num_tokens": 79443386.0,
      "step": 34690
    },
    {
      "entropy": 5.760905742645264,
      "epoch": 3.429715302491103,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003855957677746077,
      "loss": 5.0068,
      "mean_token_accuracy": 0.22393664121627807,
      "num_tokens": 79454334.0,
      "step": 34695
    },
    {
      "entropy": 5.783784341812134,
      "epoch": 3.4302095689996044,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003855650338920231,
      "loss": 4.9631,
      "mean_token_accuracy": 0.2238968789577484,
      "num_tokens": 79465489.0,
      "step": 34700
    },
    {
      "entropy": 5.787978219985962,
      "epoch": 3.430703835508106,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00038553429728952144,
      "loss": 5.0,
      "mean_token_accuracy": 0.21619585901498795,
      "num_tokens": 79477293.0,
      "step": 34705
    },
    {
      "entropy": 5.804285764694214,
      "epoch": 3.4311981020166074,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038550355796785897,
      "loss": 5.0288,
      "mean_token_accuracy": 0.2129867345094681,
      "num_tokens": 79488995.0,
      "step": 34710
    },
    {
      "entropy": 5.7826683044433596,
      "epoch": 3.4316923685251086,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038547281592779184,
      "loss": 5.0605,
      "mean_token_accuracy": 0.2194013148546219,
      "num_tokens": 79502757.0,
      "step": 34715
    },
    {
      "entropy": 5.791015577316284,
      "epoch": 3.4321866350336103,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003854420711700763,
      "loss": 4.9899,
      "mean_token_accuracy": 0.2170246124267578,
      "num_tokens": 79515220.0,
      "step": 34720
    },
    {
      "entropy": 5.817648601531983,
      "epoch": 3.4326809015421116,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00038541132369546865,
      "loss": 5.0579,
      "mean_token_accuracy": 0.2135082110762596,
      "num_tokens": 79527577.0,
      "step": 34725
    },
    {
      "entropy": 5.739390754699707,
      "epoch": 3.433175168050613,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003853805735047253,
      "loss": 4.9404,
      "mean_token_accuracy": 0.2223578944802284,
      "num_tokens": 79539136.0,
      "step": 34730
    },
    {
      "entropy": 5.674014663696289,
      "epoch": 3.433669434559114,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038534982059860274,
      "loss": 4.8612,
      "mean_token_accuracy": 0.23533019423484802,
      "num_tokens": 79550123.0,
      "step": 34735
    },
    {
      "entropy": 5.7918602466583256,
      "epoch": 3.434163701067616,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038531906497785753,
      "loss": 5.0347,
      "mean_token_accuracy": 0.21710706800222396,
      "num_tokens": 79563072.0,
      "step": 34740
    },
    {
      "entropy": 5.759722948074341,
      "epoch": 3.434657967576117,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038528830664324613,
      "loss": 4.9417,
      "mean_token_accuracy": 0.226544289290905,
      "num_tokens": 79573901.0,
      "step": 34745
    },
    {
      "entropy": 5.666585493087768,
      "epoch": 3.4351522340846183,
      "grad_norm": 1.125,
      "learning_rate": 0.00038525754559552526,
      "loss": 4.9366,
      "mean_token_accuracy": 0.2247488260269165,
      "num_tokens": 79585384.0,
      "step": 34750
    },
    {
      "entropy": 5.716826725006103,
      "epoch": 3.43564650059312,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003852267818354517,
      "loss": 5.0202,
      "mean_token_accuracy": 0.22154323309659957,
      "num_tokens": 79597277.0,
      "step": 34755
    },
    {
      "entropy": 5.802966642379761,
      "epoch": 3.4361407671016213,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003851960153637822,
      "loss": 5.0554,
      "mean_token_accuracy": 0.21783553212881088,
      "num_tokens": 79609797.0,
      "step": 34760
    },
    {
      "entropy": 5.781498289108276,
      "epoch": 3.4366350336101226,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038516524618127364,
      "loss": 4.9479,
      "mean_token_accuracy": 0.2301510751247406,
      "num_tokens": 79621247.0,
      "step": 34765
    },
    {
      "entropy": 5.801448583602905,
      "epoch": 3.437129300118624,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003851344742886829,
      "loss": 5.0314,
      "mean_token_accuracy": 0.21347317695617676,
      "num_tokens": 79632892.0,
      "step": 34770
    },
    {
      "entropy": 5.744549942016602,
      "epoch": 3.4376235666271255,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038510369968676705,
      "loss": 5.0135,
      "mean_token_accuracy": 0.21615123450756074,
      "num_tokens": 79644426.0,
      "step": 34775
    },
    {
      "entropy": 5.753431797027588,
      "epoch": 3.438117833135627,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003850729223762831,
      "loss": 4.8864,
      "mean_token_accuracy": 0.2305927187204361,
      "num_tokens": 79656414.0,
      "step": 34780
    },
    {
      "entropy": 5.730092525482178,
      "epoch": 3.438612099644128,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003850421423579882,
      "loss": 5.0015,
      "mean_token_accuracy": 0.2225076213479042,
      "num_tokens": 79667244.0,
      "step": 34785
    },
    {
      "entropy": 5.755971050262451,
      "epoch": 3.4391063661526293,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038501135963263947,
      "loss": 4.9714,
      "mean_token_accuracy": 0.22137348353862762,
      "num_tokens": 79677837.0,
      "step": 34790
    },
    {
      "entropy": 5.802629280090332,
      "epoch": 3.439600632661131,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038498057420099425,
      "loss": 5.0401,
      "mean_token_accuracy": 0.21932614594697952,
      "num_tokens": 79689500.0,
      "step": 34795
    },
    {
      "entropy": 5.707279443740845,
      "epoch": 3.4400948991696323,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038494978606380984,
      "loss": 4.912,
      "mean_token_accuracy": 0.22836938947439195,
      "num_tokens": 79701621.0,
      "step": 34800
    },
    {
      "entropy": 5.661640787124634,
      "epoch": 3.4405891656781336,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003849189952218436,
      "loss": 4.9678,
      "mean_token_accuracy": 0.23213227242231368,
      "num_tokens": 79712635.0,
      "step": 34805
    },
    {
      "entropy": 5.778584051132202,
      "epoch": 3.441083432186635,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003848882016758531,
      "loss": 5.0474,
      "mean_token_accuracy": 0.2108338937163353,
      "num_tokens": 79723626.0,
      "step": 34810
    },
    {
      "entropy": 5.796139669418335,
      "epoch": 3.4415776986951365,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038485740542659573,
      "loss": 5.0061,
      "mean_token_accuracy": 0.21480755805969237,
      "num_tokens": 79734983.0,
      "step": 34815
    },
    {
      "entropy": 5.7980384349823,
      "epoch": 3.442071965203638,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003848266064748292,
      "loss": 4.9841,
      "mean_token_accuracy": 0.21694866865873336,
      "num_tokens": 79746767.0,
      "step": 34820
    },
    {
      "entropy": 5.75792236328125,
      "epoch": 3.442566231712139,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003847958048213111,
      "loss": 4.979,
      "mean_token_accuracy": 0.2227740541100502,
      "num_tokens": 79759062.0,
      "step": 34825
    },
    {
      "entropy": 5.684417247772217,
      "epoch": 3.4430604982206408,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00038476500046679915,
      "loss": 4.8955,
      "mean_token_accuracy": 0.23331020772457123,
      "num_tokens": 79769962.0,
      "step": 34830
    },
    {
      "entropy": 5.675603294372559,
      "epoch": 3.443554764729142,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038473419341205115,
      "loss": 4.9387,
      "mean_token_accuracy": 0.2238091304898262,
      "num_tokens": 79781279.0,
      "step": 34835
    },
    {
      "entropy": 5.784378719329834,
      "epoch": 3.4440490312376433,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000384703383657825,
      "loss": 4.9987,
      "mean_token_accuracy": 0.2203410357236862,
      "num_tokens": 79792406.0,
      "step": 34840
    },
    {
      "entropy": 5.810380649566651,
      "epoch": 3.4445432977461445,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003846725712048786,
      "loss": 5.0264,
      "mean_token_accuracy": 0.2184491127729416,
      "num_tokens": 79804150.0,
      "step": 34845
    },
    {
      "entropy": 5.767418003082275,
      "epoch": 3.4450375642546462,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003846417560539699,
      "loss": 4.8993,
      "mean_token_accuracy": 0.22999347150325775,
      "num_tokens": 79814267.0,
      "step": 34850
    },
    {
      "entropy": 5.720052433013916,
      "epoch": 3.4455318307631475,
      "grad_norm": 1.234375,
      "learning_rate": 0.000384610938205857,
      "loss": 4.9326,
      "mean_token_accuracy": 0.22880630940198898,
      "num_tokens": 79824887.0,
      "step": 34855
    },
    {
      "entropy": 5.748180389404297,
      "epoch": 3.4460260972716488,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038458011766129795,
      "loss": 5.0013,
      "mean_token_accuracy": 0.21667182445526123,
      "num_tokens": 79836840.0,
      "step": 34860
    },
    {
      "entropy": 5.735796070098877,
      "epoch": 3.4465203637801505,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00038454929442105105,
      "loss": 4.9175,
      "mean_token_accuracy": 0.2268517166376114,
      "num_tokens": 79846876.0,
      "step": 34865
    },
    {
      "entropy": 5.756755256652832,
      "epoch": 3.4470146302886517,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003845184684858746,
      "loss": 5.0007,
      "mean_token_accuracy": 0.22102858126163483,
      "num_tokens": 79859271.0,
      "step": 34870
    },
    {
      "entropy": 5.740095090866089,
      "epoch": 3.447508896797153,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038448763985652674,
      "loss": 5.0517,
      "mean_token_accuracy": 0.21844163089990615,
      "num_tokens": 79871938.0,
      "step": 34875
    },
    {
      "entropy": 5.780250406265258,
      "epoch": 3.4480031633056543,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003844568085337659,
      "loss": 5.0164,
      "mean_token_accuracy": 0.22304982095956802,
      "num_tokens": 79883202.0,
      "step": 34880
    },
    {
      "entropy": 5.712198066711426,
      "epoch": 3.448497429814156,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038442597451835076,
      "loss": 4.9434,
      "mean_token_accuracy": 0.22576947808265685,
      "num_tokens": 79895794.0,
      "step": 34885
    },
    {
      "entropy": 5.746598196029663,
      "epoch": 3.4489916963226572,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003843951378110396,
      "loss": 4.9377,
      "mean_token_accuracy": 0.229790398478508,
      "num_tokens": 79906548.0,
      "step": 34890
    },
    {
      "entropy": 5.75026650428772,
      "epoch": 3.4494859628311585,
      "grad_norm": 1.171875,
      "learning_rate": 0.000384364298412591,
      "loss": 4.9978,
      "mean_token_accuracy": 0.22277528196573257,
      "num_tokens": 79918116.0,
      "step": 34895
    },
    {
      "entropy": 5.7345386981964115,
      "epoch": 3.4499802293396598,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038433345632376376,
      "loss": 4.9927,
      "mean_token_accuracy": 0.22818876802921295,
      "num_tokens": 79930791.0,
      "step": 34900
    },
    {
      "entropy": 5.723190498352051,
      "epoch": 3.4504744958481615,
      "grad_norm": 8.625,
      "learning_rate": 0.00038430261154531655,
      "loss": 4.9472,
      "mean_token_accuracy": 0.23318518102169036,
      "num_tokens": 79942359.0,
      "step": 34905
    },
    {
      "entropy": 5.709157657623291,
      "epoch": 3.4509687623566627,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003842717640780081,
      "loss": 4.9187,
      "mean_token_accuracy": 0.22305017709732056,
      "num_tokens": 79954424.0,
      "step": 34910
    },
    {
      "entropy": 5.7242498874664305,
      "epoch": 3.451463028865164,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00038424091392259737,
      "loss": 4.925,
      "mean_token_accuracy": 0.2246979370713234,
      "num_tokens": 79965704.0,
      "step": 34915
    },
    {
      "entropy": 5.750075435638427,
      "epoch": 3.4519572953736652,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003842100610798431,
      "loss": 4.9391,
      "mean_token_accuracy": 0.2288699373602867,
      "num_tokens": 79977891.0,
      "step": 34920
    },
    {
      "entropy": 5.7055689811706545,
      "epoch": 3.452451561882167,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038417920555050447,
      "loss": 4.9765,
      "mean_token_accuracy": 0.22124988734722137,
      "num_tokens": 79988709.0,
      "step": 34925
    },
    {
      "entropy": 5.718805599212646,
      "epoch": 3.452945828390668,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00038414834733534046,
      "loss": 4.9537,
      "mean_token_accuracy": 0.2270994558930397,
      "num_tokens": 80000349.0,
      "step": 34930
    },
    {
      "entropy": 5.702279472351075,
      "epoch": 3.4534400948991695,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003841174864351101,
      "loss": 4.846,
      "mean_token_accuracy": 0.2380574494600296,
      "num_tokens": 80011317.0,
      "step": 34935
    },
    {
      "entropy": 5.75683069229126,
      "epoch": 3.453934361407671,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038408662285057267,
      "loss": 5.0259,
      "mean_token_accuracy": 0.22479131817817688,
      "num_tokens": 80023471.0,
      "step": 34940
    },
    {
      "entropy": 5.73680009841919,
      "epoch": 3.4544286279161724,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038405575658248737,
      "loss": 4.9617,
      "mean_token_accuracy": 0.22336150854825973,
      "num_tokens": 80035937.0,
      "step": 34945
    },
    {
      "entropy": 5.748815107345581,
      "epoch": 3.4549228944246737,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003840248876316135,
      "loss": 4.9493,
      "mean_token_accuracy": 0.22325362265110016,
      "num_tokens": 80048435.0,
      "step": 34950
    },
    {
      "entropy": 5.707042121887207,
      "epoch": 3.455417160933175,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003839940159987106,
      "loss": 4.9586,
      "mean_token_accuracy": 0.22586894780397415,
      "num_tokens": 80059675.0,
      "step": 34955
    },
    {
      "entropy": 5.821738910675049,
      "epoch": 3.4559114274416767,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038396314168453793,
      "loss": 5.0079,
      "mean_token_accuracy": 0.2158147782087326,
      "num_tokens": 80071232.0,
      "step": 34960
    },
    {
      "entropy": 5.731034660339356,
      "epoch": 3.456405693950178,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003839322646898551,
      "loss": 4.9208,
      "mean_token_accuracy": 0.22737734764814377,
      "num_tokens": 80083331.0,
      "step": 34965
    },
    {
      "entropy": 5.752064895629883,
      "epoch": 3.456899960458679,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00038390138501542164,
      "loss": 5.0157,
      "mean_token_accuracy": 0.22142302393913268,
      "num_tokens": 80095883.0,
      "step": 34970
    },
    {
      "entropy": 5.688836336135864,
      "epoch": 3.457394226967181,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003838705026619972,
      "loss": 4.9597,
      "mean_token_accuracy": 0.22674014717340468,
      "num_tokens": 80108098.0,
      "step": 34975
    },
    {
      "entropy": 5.698457908630371,
      "epoch": 3.457888493475682,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003838396176303416,
      "loss": 4.9841,
      "mean_token_accuracy": 0.21836493611335756,
      "num_tokens": 80121184.0,
      "step": 34980
    },
    {
      "entropy": 5.750385761260986,
      "epoch": 3.4583827599841834,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003838087299212144,
      "loss": 4.9165,
      "mean_token_accuracy": 0.2225114330649376,
      "num_tokens": 80132862.0,
      "step": 34985
    },
    {
      "entropy": 5.762396144866943,
      "epoch": 3.4588770264926847,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038377783953537565,
      "loss": 4.9431,
      "mean_token_accuracy": 0.22454853951931,
      "num_tokens": 80143756.0,
      "step": 34990
    },
    {
      "entropy": 5.737108278274536,
      "epoch": 3.4593712930011864,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003837469464735851,
      "loss": 4.9601,
      "mean_token_accuracy": 0.2255374625325203,
      "num_tokens": 80154570.0,
      "step": 34995
    },
    {
      "entropy": 5.6966629981994625,
      "epoch": 3.4598655595096877,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038371605073660284,
      "loss": 4.955,
      "mean_token_accuracy": 0.22564540058374405,
      "num_tokens": 80165968.0,
      "step": 35000
    },
    {
      "entropy": 5.695196151733398,
      "epoch": 3.460359826018189,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00038368515232518885,
      "loss": 4.945,
      "mean_token_accuracy": 0.21703900247812272,
      "num_tokens": 80177584.0,
      "step": 35005
    },
    {
      "entropy": 5.727974557876587,
      "epoch": 3.4608540925266906,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003836542512401033,
      "loss": 4.9547,
      "mean_token_accuracy": 0.22055343985557557,
      "num_tokens": 80190113.0,
      "step": 35010
    },
    {
      "entropy": 5.802620506286621,
      "epoch": 3.461348359035192,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003836233474821063,
      "loss": 5.0157,
      "mean_token_accuracy": 0.21482030004262925,
      "num_tokens": 80201074.0,
      "step": 35015
    },
    {
      "entropy": 5.73555736541748,
      "epoch": 3.461842625543693,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038359244105195807,
      "loss": 4.9886,
      "mean_token_accuracy": 0.23028539270162582,
      "num_tokens": 80212248.0,
      "step": 35020
    },
    {
      "entropy": 5.730514478683472,
      "epoch": 3.4623368920521944,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038356153195041886,
      "loss": 4.9006,
      "mean_token_accuracy": 0.2240366071462631,
      "num_tokens": 80223151.0,
      "step": 35025
    },
    {
      "entropy": 5.699669742584229,
      "epoch": 3.462831158560696,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00038353062017824923,
      "loss": 4.8857,
      "mean_token_accuracy": 0.2262493446469307,
      "num_tokens": 80234978.0,
      "step": 35030
    },
    {
      "entropy": 5.727032613754273,
      "epoch": 3.4633254250691974,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003834997057362096,
      "loss": 4.9275,
      "mean_token_accuracy": 0.22904649376869202,
      "num_tokens": 80246297.0,
      "step": 35035
    },
    {
      "entropy": 5.7936502456665036,
      "epoch": 3.4638196915776986,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003834687886250602,
      "loss": 5.0045,
      "mean_token_accuracy": 0.2158038854598999,
      "num_tokens": 80258628.0,
      "step": 35040
    },
    {
      "entropy": 5.728713321685791,
      "epoch": 3.4643139580862,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003834378688455618,
      "loss": 4.9198,
      "mean_token_accuracy": 0.2268444687128067,
      "num_tokens": 80270694.0,
      "step": 35045
    },
    {
      "entropy": 5.694925785064697,
      "epoch": 3.4648082245947016,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003834069463984751,
      "loss": 4.9636,
      "mean_token_accuracy": 0.2226157933473587,
      "num_tokens": 80282032.0,
      "step": 35050
    },
    {
      "entropy": 5.7082291603088375,
      "epoch": 3.465302491103203,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003833760212845605,
      "loss": 4.9478,
      "mean_token_accuracy": 0.2250010445713997,
      "num_tokens": 80294033.0,
      "step": 35055
    },
    {
      "entropy": 5.7644435405731205,
      "epoch": 3.465796757611704,
      "grad_norm": 1.375,
      "learning_rate": 0.00038334509350457913,
      "loss": 4.8608,
      "mean_token_accuracy": 0.2316405475139618,
      "num_tokens": 80304273.0,
      "step": 35060
    },
    {
      "entropy": 5.663561201095581,
      "epoch": 3.4662910241202054,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003833141630592916,
      "loss": 4.8957,
      "mean_token_accuracy": 0.2306319624185562,
      "num_tokens": 80314049.0,
      "step": 35065
    },
    {
      "entropy": 5.797319793701172,
      "epoch": 3.466785290628707,
      "grad_norm": 1.28125,
      "learning_rate": 0.00038328322994945883,
      "loss": 5.0381,
      "mean_token_accuracy": 0.21499863266944885,
      "num_tokens": 80324632.0,
      "step": 35070
    },
    {
      "entropy": 5.724928522109986,
      "epoch": 3.4672795571372084,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038325229417584194,
      "loss": 4.9054,
      "mean_token_accuracy": 0.23125911951065065,
      "num_tokens": 80335824.0,
      "step": 35075
    },
    {
      "entropy": 5.704902839660645,
      "epoch": 3.4677738236457096,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038322135573920176,
      "loss": 4.9264,
      "mean_token_accuracy": 0.21805291026830673,
      "num_tokens": 80346537.0,
      "step": 35080
    },
    {
      "entropy": 5.746596193313598,
      "epoch": 3.4682680901542113,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003831904146402994,
      "loss": 4.9137,
      "mean_token_accuracy": 0.2273559555411339,
      "num_tokens": 80358446.0,
      "step": 35085
    },
    {
      "entropy": 5.738199472427368,
      "epoch": 3.4687623566627126,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000383159470879896,
      "loss": 4.9341,
      "mean_token_accuracy": 0.23073399513959886,
      "num_tokens": 80370219.0,
      "step": 35090
    },
    {
      "entropy": 5.739638805389404,
      "epoch": 3.469256623171214,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003831285244587529,
      "loss": 4.9711,
      "mean_token_accuracy": 0.22299767881631852,
      "num_tokens": 80380886.0,
      "step": 35095
    },
    {
      "entropy": 5.703467273712159,
      "epoch": 3.469750889679715,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003830975753776314,
      "loss": 4.8964,
      "mean_token_accuracy": 0.2301630988717079,
      "num_tokens": 80391171.0,
      "step": 35100
    },
    {
      "entropy": 5.728148698806763,
      "epoch": 3.470245156188217,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003830666236372927,
      "loss": 4.9317,
      "mean_token_accuracy": 0.22347177118062972,
      "num_tokens": 80403527.0,
      "step": 35105
    },
    {
      "entropy": 5.78545298576355,
      "epoch": 3.470739422696718,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038303566923849827,
      "loss": 5.0382,
      "mean_token_accuracy": 0.218366439640522,
      "num_tokens": 80414992.0,
      "step": 35110
    },
    {
      "entropy": 5.7094153881073,
      "epoch": 3.4712336892052194,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00038300471218200965,
      "loss": 4.9533,
      "mean_token_accuracy": 0.22169215381145477,
      "num_tokens": 80426167.0,
      "step": 35115
    },
    {
      "entropy": 5.74866418838501,
      "epoch": 3.471727955713721,
      "grad_norm": 1.125,
      "learning_rate": 0.00038297375246858835,
      "loss": 4.9499,
      "mean_token_accuracy": 0.22430207431316376,
      "num_tokens": 80438063.0,
      "step": 35120
    },
    {
      "entropy": 5.6901782035827635,
      "epoch": 3.4722222222222223,
      "grad_norm": 1.125,
      "learning_rate": 0.000382942790098996,
      "loss": 4.9381,
      "mean_token_accuracy": 0.21947703808546065,
      "num_tokens": 80449898.0,
      "step": 35125
    },
    {
      "entropy": 5.756030941009522,
      "epoch": 3.4727164887307236,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038291182507399425,
      "loss": 4.9534,
      "mean_token_accuracy": 0.2271957129240036,
      "num_tokens": 80461623.0,
      "step": 35130
    },
    {
      "entropy": 5.738477277755737,
      "epoch": 3.473210755239225,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038288085739434486,
      "loss": 4.9054,
      "mean_token_accuracy": 0.23444311916828156,
      "num_tokens": 80472627.0,
      "step": 35135
    },
    {
      "entropy": 5.708876037597657,
      "epoch": 3.4737050217477266,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003828498870608096,
      "loss": 4.9317,
      "mean_token_accuracy": 0.22621617913246156,
      "num_tokens": 80483494.0,
      "step": 35140
    },
    {
      "entropy": 5.682329225540161,
      "epoch": 3.474199288256228,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038281891407415046,
      "loss": 4.8978,
      "mean_token_accuracy": 0.22756169736385345,
      "num_tokens": 80495122.0,
      "step": 35145
    },
    {
      "entropy": 5.656877088546753,
      "epoch": 3.474693554764729,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038278793843512935,
      "loss": 4.9143,
      "mean_token_accuracy": 0.23086659908294677,
      "num_tokens": 80506639.0,
      "step": 35150
    },
    {
      "entropy": 5.769659233093262,
      "epoch": 3.4751878212732303,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003827569601445081,
      "loss": 4.9489,
      "mean_token_accuracy": 0.22331925928592683,
      "num_tokens": 80517897.0,
      "step": 35155
    },
    {
      "entropy": 5.784477424621582,
      "epoch": 3.475682087781732,
      "grad_norm": 1.109375,
      "learning_rate": 0.000382725979203049,
      "loss": 4.9704,
      "mean_token_accuracy": 0.21801871806383133,
      "num_tokens": 80530112.0,
      "step": 35160
    },
    {
      "entropy": 5.706895637512207,
      "epoch": 3.4761763542902333,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038269499561151407,
      "loss": 4.912,
      "mean_token_accuracy": 0.2316989481449127,
      "num_tokens": 80541811.0,
      "step": 35165
    },
    {
      "entropy": 5.716243505477905,
      "epoch": 3.4766706207987346,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038266400937066555,
      "loss": 4.9591,
      "mean_token_accuracy": 0.2291096955537796,
      "num_tokens": 80553107.0,
      "step": 35170
    },
    {
      "entropy": 5.798077535629273,
      "epoch": 3.477164887307236,
      "grad_norm": 1.25,
      "learning_rate": 0.00038263302048126567,
      "loss": 4.9985,
      "mean_token_accuracy": 0.2228512942790985,
      "num_tokens": 80562916.0,
      "step": 35175
    },
    {
      "entropy": 5.718015813827515,
      "epoch": 3.4776591538157375,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003826020289440768,
      "loss": 4.9283,
      "mean_token_accuracy": 0.22203734666109085,
      "num_tokens": 80573179.0,
      "step": 35180
    },
    {
      "entropy": 5.670910835266113,
      "epoch": 3.478153420324239,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038257103475986134,
      "loss": 4.958,
      "mean_token_accuracy": 0.2293822631239891,
      "num_tokens": 80583494.0,
      "step": 35185
    },
    {
      "entropy": 5.661175966262817,
      "epoch": 3.47864768683274,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038254003792938176,
      "loss": 4.973,
      "mean_token_accuracy": 0.22705525606870652,
      "num_tokens": 80595926.0,
      "step": 35190
    },
    {
      "entropy": 5.685512447357178,
      "epoch": 3.4791419533412418,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038250903845340057,
      "loss": 4.9416,
      "mean_token_accuracy": 0.23341089338064194,
      "num_tokens": 80607449.0,
      "step": 35195
    },
    {
      "entropy": 5.749845504760742,
      "epoch": 3.479636219849743,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003824780363326803,
      "loss": 4.9502,
      "mean_token_accuracy": 0.2240004763007164,
      "num_tokens": 80619201.0,
      "step": 35200
    },
    {
      "entropy": 5.757830858230591,
      "epoch": 3.4801304863582443,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038244703156798364,
      "loss": 5.042,
      "mean_token_accuracy": 0.22113657593727112,
      "num_tokens": 80632068.0,
      "step": 35205
    },
    {
      "entropy": 5.741551685333252,
      "epoch": 3.4806247528667456,
      "grad_norm": 1.125,
      "learning_rate": 0.0003824160241600734,
      "loss": 4.9329,
      "mean_token_accuracy": 0.22343134135007858,
      "num_tokens": 80643569.0,
      "step": 35210
    },
    {
      "entropy": 5.790683555603027,
      "epoch": 3.4811190193752473,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038238501410971226,
      "loss": 4.9462,
      "mean_token_accuracy": 0.2252105489373207,
      "num_tokens": 80656000.0,
      "step": 35215
    },
    {
      "entropy": 5.7622706413269045,
      "epoch": 3.4816132858837485,
      "grad_norm": 1.390625,
      "learning_rate": 0.00038235400141766314,
      "loss": 4.9308,
      "mean_token_accuracy": 0.22718267142772675,
      "num_tokens": 80667473.0,
      "step": 35220
    },
    {
      "entropy": 5.811671543121338,
      "epoch": 3.48210755239225,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038232298608468894,
      "loss": 5.0724,
      "mean_token_accuracy": 0.22198492735624314,
      "num_tokens": 80679322.0,
      "step": 35225
    },
    {
      "entropy": 5.84144139289856,
      "epoch": 3.4826018189007515,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003822919681115526,
      "loss": 5.0844,
      "mean_token_accuracy": 0.21785397976636886,
      "num_tokens": 80690907.0,
      "step": 35230
    },
    {
      "entropy": 5.696099138259887,
      "epoch": 3.4830960854092528,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003822609474990172,
      "loss": 4.8696,
      "mean_token_accuracy": 0.22997491657733918,
      "num_tokens": 80702012.0,
      "step": 35235
    },
    {
      "entropy": 5.826912450790405,
      "epoch": 3.483590351917754,
      "grad_norm": 1.15625,
      "learning_rate": 0.000382229924247846,
      "loss": 5.074,
      "mean_token_accuracy": 0.2123814344406128,
      "num_tokens": 80714283.0,
      "step": 35240
    },
    {
      "entropy": 5.714649724960327,
      "epoch": 3.4840846184262553,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003821988983588018,
      "loss": 4.9085,
      "mean_token_accuracy": 0.23351623117923737,
      "num_tokens": 80724964.0,
      "step": 35245
    },
    {
      "entropy": 5.786600160598755,
      "epoch": 3.484578884934757,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038216786983264817,
      "loss": 4.9504,
      "mean_token_accuracy": 0.2237883150577545,
      "num_tokens": 80735934.0,
      "step": 35250
    },
    {
      "entropy": 5.800012063980103,
      "epoch": 3.4850731514432582,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038213683867014824,
      "loss": 5.0574,
      "mean_token_accuracy": 0.21366148740053176,
      "num_tokens": 80746988.0,
      "step": 35255
    },
    {
      "entropy": 5.797828722000122,
      "epoch": 3.4855674179517595,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038210580487206563,
      "loss": 5.0306,
      "mean_token_accuracy": 0.20996270328760147,
      "num_tokens": 80759300.0,
      "step": 35260
    },
    {
      "entropy": 5.699312305450439,
      "epoch": 3.486061684460261,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003820747684391636,
      "loss": 4.993,
      "mean_token_accuracy": 0.2224402606487274,
      "num_tokens": 80771055.0,
      "step": 35265
    },
    {
      "entropy": 5.6758167266845705,
      "epoch": 3.4865559509687625,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003820437293722054,
      "loss": 4.9029,
      "mean_token_accuracy": 0.22777575999498367,
      "num_tokens": 80782803.0,
      "step": 35270
    },
    {
      "entropy": 5.770343017578125,
      "epoch": 3.4870502174772637,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003820126876719551,
      "loss": 4.9619,
      "mean_token_accuracy": 0.22241358309984208,
      "num_tokens": 80793833.0,
      "step": 35275
    },
    {
      "entropy": 5.753967905044556,
      "epoch": 3.487544483985765,
      "grad_norm": 1.203125,
      "learning_rate": 0.000381981643339176,
      "loss": 4.9435,
      "mean_token_accuracy": 0.22572194784879684,
      "num_tokens": 80804800.0,
      "step": 35280
    },
    {
      "entropy": 5.757425785064697,
      "epoch": 3.4880387504942663,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003819505963746318,
      "loss": 4.968,
      "mean_token_accuracy": 0.217978572845459,
      "num_tokens": 80815934.0,
      "step": 35285
    },
    {
      "entropy": 5.77924075126648,
      "epoch": 3.488533017002768,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003819195467790864,
      "loss": 4.9724,
      "mean_token_accuracy": 0.22822180092334748,
      "num_tokens": 80826971.0,
      "step": 35290
    },
    {
      "entropy": 5.752380752563477,
      "epoch": 3.4890272835112692,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003818884945533036,
      "loss": 5.0099,
      "mean_token_accuracy": 0.217777518928051,
      "num_tokens": 80839847.0,
      "step": 35295
    },
    {
      "entropy": 5.768237686157226,
      "epoch": 3.4895215500197705,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003818574396980472,
      "loss": 4.9864,
      "mean_token_accuracy": 0.2221311792731285,
      "num_tokens": 80852937.0,
      "step": 35300
    },
    {
      "entropy": 5.741956281661987,
      "epoch": 3.490015816528272,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003818263822140812,
      "loss": 4.9034,
      "mean_token_accuracy": 0.23298288583755494,
      "num_tokens": 80862978.0,
      "step": 35305
    },
    {
      "entropy": 5.725862455368042,
      "epoch": 3.4905100830367735,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038179532210216974,
      "loss": 4.9389,
      "mean_token_accuracy": 0.23028623610734938,
      "num_tokens": 80874446.0,
      "step": 35310
    },
    {
      "entropy": 5.6892200946807865,
      "epoch": 3.4910043495452747,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003817642593630767,
      "loss": 4.9993,
      "mean_token_accuracy": 0.22696565091609955,
      "num_tokens": 80886289.0,
      "step": 35315
    },
    {
      "entropy": 5.735862731933594,
      "epoch": 3.491498616053776,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038173319399756626,
      "loss": 4.982,
      "mean_token_accuracy": 0.22164596617221832,
      "num_tokens": 80897363.0,
      "step": 35320
    },
    {
      "entropy": 5.780193519592285,
      "epoch": 3.4919928825622777,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003817021260064028,
      "loss": 4.9572,
      "mean_token_accuracy": 0.2191171392798424,
      "num_tokens": 80907850.0,
      "step": 35325
    },
    {
      "entropy": 5.797368860244751,
      "epoch": 3.492487149070779,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003816710553903504,
      "loss": 4.983,
      "mean_token_accuracy": 0.2207235127687454,
      "num_tokens": 80918512.0,
      "step": 35330
    },
    {
      "entropy": 5.773538112640381,
      "epoch": 3.49298141557928,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003816399821501736,
      "loss": 4.9529,
      "mean_token_accuracy": 0.2301558569073677,
      "num_tokens": 80929550.0,
      "step": 35335
    },
    {
      "entropy": 5.703844118118286,
      "epoch": 3.493475682087782,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038160890628663663,
      "loss": 4.9444,
      "mean_token_accuracy": 0.22311123013496398,
      "num_tokens": 80941129.0,
      "step": 35340
    },
    {
      "entropy": 5.794006252288819,
      "epoch": 3.493969948596283,
      "grad_norm": 1.140625,
      "learning_rate": 0.000381577827800504,
      "loss": 4.927,
      "mean_token_accuracy": 0.22684859335422516,
      "num_tokens": 80951804.0,
      "step": 35345
    },
    {
      "entropy": 5.724021100997925,
      "epoch": 3.4944642151047844,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003815467466925402,
      "loss": 4.9124,
      "mean_token_accuracy": 0.23357467353343964,
      "num_tokens": 80962556.0,
      "step": 35350
    },
    {
      "entropy": 5.767390775680542,
      "epoch": 3.4949584816132857,
      "grad_norm": 1.25,
      "learning_rate": 0.00038151566296351,
      "loss": 5.0804,
      "mean_token_accuracy": 0.21496041715145112,
      "num_tokens": 80974550.0,
      "step": 35355
    },
    {
      "entropy": 5.7617134094238285,
      "epoch": 3.4954527481217874,
      "grad_norm": 1.28125,
      "learning_rate": 0.000381484576614178,
      "loss": 4.9379,
      "mean_token_accuracy": 0.22010361552238464,
      "num_tokens": 80987033.0,
      "step": 35360
    },
    {
      "entropy": 5.767470264434815,
      "epoch": 3.4959470146302887,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038145348764530887,
      "loss": 4.9004,
      "mean_token_accuracy": 0.22782645374536514,
      "num_tokens": 80998216.0,
      "step": 35365
    },
    {
      "entropy": 5.718341302871704,
      "epoch": 3.49644128113879,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038142239605766737,
      "loss": 4.9394,
      "mean_token_accuracy": 0.2257169231772423,
      "num_tokens": 81010276.0,
      "step": 35370
    },
    {
      "entropy": 5.756587266921997,
      "epoch": 3.4969355476472916,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00038139130185201845,
      "loss": 4.9468,
      "mean_token_accuracy": 0.22149068266153335,
      "num_tokens": 81020892.0,
      "step": 35375
    },
    {
      "entropy": 5.699207973480225,
      "epoch": 3.497429814155793,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038136020502912705,
      "loss": 4.8832,
      "mean_token_accuracy": 0.2350930616259575,
      "num_tokens": 81032390.0,
      "step": 35380
    },
    {
      "entropy": 5.699143886566162,
      "epoch": 3.497924080664294,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003813291055897581,
      "loss": 4.9153,
      "mean_token_accuracy": 0.23069553673267365,
      "num_tokens": 81043433.0,
      "step": 35385
    },
    {
      "entropy": 5.7259584903717045,
      "epoch": 3.4984183471727954,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003812980035346766,
      "loss": 4.9178,
      "mean_token_accuracy": 0.22666165083646775,
      "num_tokens": 81054798.0,
      "step": 35390
    },
    {
      "entropy": 5.7428895950317385,
      "epoch": 3.498912613681297,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003812668988646477,
      "loss": 5.0001,
      "mean_token_accuracy": 0.22207673043012618,
      "num_tokens": 81067746.0,
      "step": 35395
    },
    {
      "entropy": 5.7249915599823,
      "epoch": 3.4994068801897984,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003812357915804366,
      "loss": 4.987,
      "mean_token_accuracy": 0.22018432915210723,
      "num_tokens": 81079328.0,
      "step": 35400
    },
    {
      "entropy": 5.7407608985900875,
      "epoch": 3.4999011466982997,
      "grad_norm": 1.25,
      "learning_rate": 0.0003812046816828086,
      "loss": 4.9613,
      "mean_token_accuracy": 0.224355486035347,
      "num_tokens": 81091447.0,
      "step": 35405
    },
    {
      "entropy": 5.81247091293335,
      "epoch": 3.5003954132068014,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000381173569172529,
      "loss": 5.026,
      "mean_token_accuracy": 0.21899720877408982,
      "num_tokens": 81103614.0,
      "step": 35410
    },
    {
      "entropy": 5.810672998428345,
      "epoch": 3.5008896797153026,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000381142454050363,
      "loss": 5.0718,
      "mean_token_accuracy": 0.21583905071020126,
      "num_tokens": 81115387.0,
      "step": 35415
    },
    {
      "entropy": 5.75110502243042,
      "epoch": 3.501383946223804,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038111133631707624,
      "loss": 4.9771,
      "mean_token_accuracy": 0.21944610029459,
      "num_tokens": 81128063.0,
      "step": 35420
    },
    {
      "entropy": 5.777230930328369,
      "epoch": 3.501878212732305,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003810802159734342,
      "loss": 4.97,
      "mean_token_accuracy": 0.2271800383925438,
      "num_tokens": 81139137.0,
      "step": 35425
    },
    {
      "entropy": 5.765392732620239,
      "epoch": 3.5023724792408064,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003810490930202023,
      "loss": 5.0755,
      "mean_token_accuracy": 0.21693194806575775,
      "num_tokens": 81151896.0,
      "step": 35430
    },
    {
      "entropy": 5.850603723526001,
      "epoch": 3.502866745749308,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003810179674581463,
      "loss": 4.9531,
      "mean_token_accuracy": 0.22099511772394181,
      "num_tokens": 81162843.0,
      "step": 35435
    },
    {
      "entropy": 5.834975957870483,
      "epoch": 3.5033610122578094,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038098683928803186,
      "loss": 5.1118,
      "mean_token_accuracy": 0.2126134902238846,
      "num_tokens": 81176316.0,
      "step": 35440
    },
    {
      "entropy": 5.808986520767212,
      "epoch": 3.5038552787663106,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038095570851062467,
      "loss": 4.9647,
      "mean_token_accuracy": 0.22293055951595306,
      "num_tokens": 81187956.0,
      "step": 35445
    },
    {
      "entropy": 5.759843444824218,
      "epoch": 3.5043495452748123,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038092457512669067,
      "loss": 4.9466,
      "mean_token_accuracy": 0.22627142518758775,
      "num_tokens": 81199575.0,
      "step": 35450
    },
    {
      "entropy": 5.694977140426635,
      "epoch": 3.5048438117833136,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003808934391369957,
      "loss": 4.9139,
      "mean_token_accuracy": 0.23169865757226943,
      "num_tokens": 81211959.0,
      "step": 35455
    },
    {
      "entropy": 5.698625659942627,
      "epoch": 3.505338078291815,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003808623005423057,
      "loss": 4.9064,
      "mean_token_accuracy": 0.23203559964895248,
      "num_tokens": 81222358.0,
      "step": 35460
    },
    {
      "entropy": 5.759253978729248,
      "epoch": 3.505832344800316,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038083115934338667,
      "loss": 4.9458,
      "mean_token_accuracy": 0.22519572973251342,
      "num_tokens": 81233674.0,
      "step": 35465
    },
    {
      "entropy": 5.698000907897949,
      "epoch": 3.506326611308818,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038080001554100467,
      "loss": 4.9226,
      "mean_token_accuracy": 0.22892085164785386,
      "num_tokens": 81243885.0,
      "step": 35470
    },
    {
      "entropy": 5.822749662399292,
      "epoch": 3.506820877817319,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003807688691359259,
      "loss": 5.0762,
      "mean_token_accuracy": 0.2177318200469017,
      "num_tokens": 81255869.0,
      "step": 35475
    },
    {
      "entropy": 5.764692592620849,
      "epoch": 3.5073151443258204,
      "grad_norm": 1.25,
      "learning_rate": 0.0003807377201289165,
      "loss": 4.9497,
      "mean_token_accuracy": 0.21617780327796937,
      "num_tokens": 81266746.0,
      "step": 35480
    },
    {
      "entropy": 5.761414575576782,
      "epoch": 3.507809410834322,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003807065685207429,
      "loss": 4.9545,
      "mean_token_accuracy": 0.22384707778692245,
      "num_tokens": 81277356.0,
      "step": 35485
    },
    {
      "entropy": 5.741801691055298,
      "epoch": 3.5083036773428233,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003806754143121712,
      "loss": 5.0136,
      "mean_token_accuracy": 0.22078490257263184,
      "num_tokens": 81288926.0,
      "step": 35490
    },
    {
      "entropy": 5.805824565887451,
      "epoch": 3.5087979438513246,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038064425750396793,
      "loss": 4.9728,
      "mean_token_accuracy": 0.2262909695506096,
      "num_tokens": 81300990.0,
      "step": 35495
    },
    {
      "entropy": 5.712761068344117,
      "epoch": 3.509292210359826,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00038061309809689955,
      "loss": 4.9184,
      "mean_token_accuracy": 0.22160707861185075,
      "num_tokens": 81312964.0,
      "step": 35500
    },
    {
      "entropy": 5.766635847091675,
      "epoch": 3.5097864768683276,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038058193609173256,
      "loss": 4.9668,
      "mean_token_accuracy": 0.22672232538461684,
      "num_tokens": 81324196.0,
      "step": 35505
    },
    {
      "entropy": 5.7676207542419435,
      "epoch": 3.510280743376829,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003805507714892336,
      "loss": 4.974,
      "mean_token_accuracy": 0.21922513097524643,
      "num_tokens": 81334723.0,
      "step": 35510
    },
    {
      "entropy": 5.786857080459595,
      "epoch": 3.51077500988533,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00038051960429016923,
      "loss": 5.0075,
      "mean_token_accuracy": 0.21894041150808335,
      "num_tokens": 81346126.0,
      "step": 35515
    },
    {
      "entropy": 5.721186876296997,
      "epoch": 3.511269276393832,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003804884344953063,
      "loss": 4.9474,
      "mean_token_accuracy": 0.23070349544286728,
      "num_tokens": 81358093.0,
      "step": 35520
    },
    {
      "entropy": 5.736390161514282,
      "epoch": 3.511763542902333,
      "grad_norm": 1.328125,
      "learning_rate": 0.00038045726210541144,
      "loss": 5.0046,
      "mean_token_accuracy": 0.22015675157308578,
      "num_tokens": 81369734.0,
      "step": 35525
    },
    {
      "entropy": 5.7293352603912355,
      "epoch": 3.5122578094108343,
      "grad_norm": 1.25,
      "learning_rate": 0.0003804260871212516,
      "loss": 4.978,
      "mean_token_accuracy": 0.2226536124944687,
      "num_tokens": 81381474.0,
      "step": 35530
    },
    {
      "entropy": 5.672225761413574,
      "epoch": 3.5127520759193356,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003803949095435937,
      "loss": 4.9062,
      "mean_token_accuracy": 0.22413829416036607,
      "num_tokens": 81394661.0,
      "step": 35535
    },
    {
      "entropy": 5.843664169311523,
      "epoch": 3.513246342427837,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003803637293732046,
      "loss": 5.0532,
      "mean_token_accuracy": 0.21841167658567429,
      "num_tokens": 81405591.0,
      "step": 35540
    },
    {
      "entropy": 5.810409355163574,
      "epoch": 3.5137406089363385,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00038033254661085146,
      "loss": 5.0665,
      "mean_token_accuracy": 0.2227519080042839,
      "num_tokens": 81417561.0,
      "step": 35545
    },
    {
      "entropy": 5.783941268920898,
      "epoch": 3.51423487544484,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00038030136125730135,
      "loss": 5.0315,
      "mean_token_accuracy": 0.21669332385063172,
      "num_tokens": 81430502.0,
      "step": 35550
    },
    {
      "entropy": 5.728549194335938,
      "epoch": 3.5147291419533415,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003802701733133214,
      "loss": 4.9938,
      "mean_token_accuracy": 0.22475088387727737,
      "num_tokens": 81441717.0,
      "step": 35555
    },
    {
      "entropy": 5.705388307571411,
      "epoch": 3.515223408461843,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038023898277967885,
      "loss": 4.9107,
      "mean_token_accuracy": 0.22867615520954132,
      "num_tokens": 81453282.0,
      "step": 35560
    },
    {
      "entropy": 5.782865810394287,
      "epoch": 3.515717674970344,
      "grad_norm": 1.234375,
      "learning_rate": 0.000380207789657141,
      "loss": 4.9541,
      "mean_token_accuracy": 0.22169853001832962,
      "num_tokens": 81463582.0,
      "step": 35565
    },
    {
      "entropy": 5.727762174606323,
      "epoch": 3.5162119414788453,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003801765939464752,
      "loss": 4.9464,
      "mean_token_accuracy": 0.2283109560608864,
      "num_tokens": 81475490.0,
      "step": 35570
    },
    {
      "entropy": 5.705303192138672,
      "epoch": 3.5167062079873466,
      "grad_norm": 1.125,
      "learning_rate": 0.0003801453956484489,
      "loss": 4.9067,
      "mean_token_accuracy": 0.23532119989395142,
      "num_tokens": 81486939.0,
      "step": 35575
    },
    {
      "entropy": 5.718761920928955,
      "epoch": 3.5172004744958483,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003801141947638295,
      "loss": 4.9649,
      "mean_token_accuracy": 0.2252896249294281,
      "num_tokens": 81498260.0,
      "step": 35580
    },
    {
      "entropy": 5.773679971694946,
      "epoch": 3.5176947410043495,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00038008299129338463,
      "loss": 4.907,
      "mean_token_accuracy": 0.23305947929620743,
      "num_tokens": 81508788.0,
      "step": 35585
    },
    {
      "entropy": 5.806025457382202,
      "epoch": 3.518189007512851,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038005178523788184,
      "loss": 5.0104,
      "mean_token_accuracy": 0.2194717526435852,
      "num_tokens": 81520500.0,
      "step": 35590
    },
    {
      "entropy": 5.682344055175781,
      "epoch": 3.5186832740213525,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038002057659808884,
      "loss": 4.9584,
      "mean_token_accuracy": 0.22543362230062486,
      "num_tokens": 81531617.0,
      "step": 35595
    },
    {
      "entropy": 5.773967218399048,
      "epoch": 3.5191775405298538,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003799893653747735,
      "loss": 4.9574,
      "mean_token_accuracy": 0.2260918140411377,
      "num_tokens": 81542814.0,
      "step": 35600
    },
    {
      "entropy": 5.734518814086914,
      "epoch": 3.519671807038355,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00037995815156870334,
      "loss": 4.876,
      "mean_token_accuracy": 0.23281973004341125,
      "num_tokens": 81552551.0,
      "step": 35605
    },
    {
      "entropy": 5.777448606491089,
      "epoch": 3.5201660735468563,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037992693518064635,
      "loss": 5.0863,
      "mean_token_accuracy": 0.2107955053448677,
      "num_tokens": 81564576.0,
      "step": 35610
    },
    {
      "entropy": 5.725781965255737,
      "epoch": 3.520660340055358,
      "grad_norm": 1.109375,
      "learning_rate": 0.00037989571621137053,
      "loss": 4.9532,
      "mean_token_accuracy": 0.22887909561395645,
      "num_tokens": 81575639.0,
      "step": 35615
    },
    {
      "entropy": 5.822841787338257,
      "epoch": 3.5211546065638593,
      "grad_norm": 1.3125,
      "learning_rate": 0.00037986449466164383,
      "loss": 5.0014,
      "mean_token_accuracy": 0.217158342897892,
      "num_tokens": 81587012.0,
      "step": 35620
    },
    {
      "entropy": 5.779533815383911,
      "epoch": 3.5216488730723605,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003798332705322343,
      "loss": 4.96,
      "mean_token_accuracy": 0.22823356986045837,
      "num_tokens": 81598194.0,
      "step": 35625
    },
    {
      "entropy": 5.758492136001587,
      "epoch": 3.5221431395808622,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003798020438239101,
      "loss": 4.9816,
      "mean_token_accuracy": 0.22224068343639375,
      "num_tokens": 81608702.0,
      "step": 35630
    },
    {
      "entropy": 5.7276102066040036,
      "epoch": 3.5226374060893635,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003797708145374392,
      "loss": 4.9516,
      "mean_token_accuracy": 0.23077966868877411,
      "num_tokens": 81619598.0,
      "step": 35635
    },
    {
      "entropy": 5.762475299835205,
      "epoch": 3.5231316725978647,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003797395826735901,
      "loss": 4.9752,
      "mean_token_accuracy": 0.21124812215566635,
      "num_tokens": 81631928.0,
      "step": 35640
    },
    {
      "entropy": 5.852632761001587,
      "epoch": 3.523625939106366,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037970834823313106,
      "loss": 5.1006,
      "mean_token_accuracy": 0.21089707762002946,
      "num_tokens": 81644553.0,
      "step": 35645
    },
    {
      "entropy": 5.854278612136841,
      "epoch": 3.5241202056148673,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037967711121683037,
      "loss": 5.0691,
      "mean_token_accuracy": 0.21828970164060593,
      "num_tokens": 81656633.0,
      "step": 35650
    },
    {
      "entropy": 5.771139764785767,
      "epoch": 3.524614472123369,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003796458716254565,
      "loss": 4.9555,
      "mean_token_accuracy": 0.22815842628479005,
      "num_tokens": 81669287.0,
      "step": 35655
    },
    {
      "entropy": 5.686197233200073,
      "epoch": 3.5251087386318702,
      "grad_norm": 1.296875,
      "learning_rate": 0.00037961462945977794,
      "loss": 4.8741,
      "mean_token_accuracy": 0.23084161430597305,
      "num_tokens": 81680298.0,
      "step": 35660
    },
    {
      "entropy": 5.76054539680481,
      "epoch": 3.525603005140372,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003795833847205633,
      "loss": 5.0344,
      "mean_token_accuracy": 0.21372692734003068,
      "num_tokens": 81691088.0,
      "step": 35665
    },
    {
      "entropy": 5.752209186553955,
      "epoch": 3.526097271648873,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003795521374085811,
      "loss": 4.9945,
      "mean_token_accuracy": 0.2204019919037819,
      "num_tokens": 81701791.0,
      "step": 35670
    },
    {
      "entropy": 5.827968597412109,
      "epoch": 3.5265915381573745,
      "grad_norm": 1.234375,
      "learning_rate": 0.00037952088752460025,
      "loss": 5.0776,
      "mean_token_accuracy": 0.21598095297813416,
      "num_tokens": 81713682.0,
      "step": 35675
    },
    {
      "entropy": 5.706695461273194,
      "epoch": 3.5270858046658757,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003794896350693893,
      "loss": 4.8397,
      "mean_token_accuracy": 0.24375058561563492,
      "num_tokens": 81724857.0,
      "step": 35680
    },
    {
      "entropy": 5.773783016204834,
      "epoch": 3.527580071174377,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000379458380043717,
      "loss": 4.9768,
      "mean_token_accuracy": 0.2226027324795723,
      "num_tokens": 81737074.0,
      "step": 35685
    },
    {
      "entropy": 5.813046407699585,
      "epoch": 3.5280743376828787,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003794271224483524,
      "loss": 5.0737,
      "mean_token_accuracy": 0.21293605268001556,
      "num_tokens": 81749679.0,
      "step": 35690
    },
    {
      "entropy": 5.760980939865112,
      "epoch": 3.52856860419138,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037939586228406445,
      "loss": 4.9703,
      "mean_token_accuracy": 0.22602469921112062,
      "num_tokens": 81761165.0,
      "step": 35695
    },
    {
      "entropy": 5.706920576095581,
      "epoch": 3.5290628706998812,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00037936459955162204,
      "loss": 4.8574,
      "mean_token_accuracy": 0.23544945418834687,
      "num_tokens": 81773289.0,
      "step": 35700
    },
    {
      "entropy": 5.75397424697876,
      "epoch": 3.529557137208383,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003793333342517943,
      "loss": 4.9215,
      "mean_token_accuracy": 0.22247510999441147,
      "num_tokens": 81785021.0,
      "step": 35705
    },
    {
      "entropy": 5.697779321670533,
      "epoch": 3.530051403716884,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003793020663853503,
      "loss": 4.9263,
      "mean_token_accuracy": 0.22515626400709152,
      "num_tokens": 81796111.0,
      "step": 35710
    },
    {
      "entropy": 5.7483972072601315,
      "epoch": 3.5305456702253855,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003792707959530594,
      "loss": 4.9355,
      "mean_token_accuracy": 0.228161758184433,
      "num_tokens": 81806927.0,
      "step": 35715
    },
    {
      "entropy": 5.801257753372193,
      "epoch": 3.5310399367338867,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003792395229556907,
      "loss": 5.0787,
      "mean_token_accuracy": 0.21583347469568254,
      "num_tokens": 81819942.0,
      "step": 35720
    },
    {
      "entropy": 5.734513521194458,
      "epoch": 3.5315342032423884,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003792082473940135,
      "loss": 4.9661,
      "mean_token_accuracy": 0.2264382690191269,
      "num_tokens": 81831192.0,
      "step": 35725
    },
    {
      "entropy": 5.761072444915771,
      "epoch": 3.5320284697508897,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003791769692687973,
      "loss": 4.9924,
      "mean_token_accuracy": 0.22412968873977662,
      "num_tokens": 81844684.0,
      "step": 35730
    },
    {
      "entropy": 5.767040729522705,
      "epoch": 3.532522736259391,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037914568858081146,
      "loss": 5.0117,
      "mean_token_accuracy": 0.22279942631721497,
      "num_tokens": 81855379.0,
      "step": 35735
    },
    {
      "entropy": 5.745207977294922,
      "epoch": 3.5330170027678927,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003791144053308256,
      "loss": 4.8374,
      "mean_token_accuracy": 0.22777983099222182,
      "num_tokens": 81867560.0,
      "step": 35740
    },
    {
      "entropy": 5.708595609664917,
      "epoch": 3.533511269276394,
      "grad_norm": 1.234375,
      "learning_rate": 0.00037908311951960916,
      "loss": 4.9713,
      "mean_token_accuracy": 0.22204873710870743,
      "num_tokens": 81879298.0,
      "step": 35745
    },
    {
      "entropy": 5.743827295303345,
      "epoch": 3.534005535784895,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003790518311479318,
      "loss": 4.9634,
      "mean_token_accuracy": 0.22549499273300172,
      "num_tokens": 81891323.0,
      "step": 35750
    },
    {
      "entropy": 5.7458302021026615,
      "epoch": 3.5344998022933964,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003790205402165633,
      "loss": 4.8837,
      "mean_token_accuracy": 0.23330256938934327,
      "num_tokens": 81901324.0,
      "step": 35755
    },
    {
      "entropy": 5.794938850402832,
      "epoch": 3.534994068801898,
      "grad_norm": 1.359375,
      "learning_rate": 0.00037898924672627346,
      "loss": 4.9932,
      "mean_token_accuracy": 0.2228091165423393,
      "num_tokens": 81911740.0,
      "step": 35760
    },
    {
      "entropy": 5.702045249938965,
      "epoch": 3.5354883353103994,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037895795067783184,
      "loss": 4.8832,
      "mean_token_accuracy": 0.22579411715269088,
      "num_tokens": 81922685.0,
      "step": 35765
    },
    {
      "entropy": 5.736795854568482,
      "epoch": 3.5359826018189007,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003789266520720086,
      "loss": 4.9696,
      "mean_token_accuracy": 0.2302746444940567,
      "num_tokens": 81935822.0,
      "step": 35770
    },
    {
      "entropy": 5.722377061843872,
      "epoch": 3.5364768683274024,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003788953509095736,
      "loss": 4.9319,
      "mean_token_accuracy": 0.22888694405555726,
      "num_tokens": 81948420.0,
      "step": 35775
    },
    {
      "entropy": 5.764716863632202,
      "epoch": 3.5369711348359036,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003788640471912968,
      "loss": 4.9771,
      "mean_token_accuracy": 0.2236185759305954,
      "num_tokens": 81959740.0,
      "step": 35780
    },
    {
      "entropy": 5.74137692451477,
      "epoch": 3.537465401344405,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037883274091794837,
      "loss": 4.9328,
      "mean_token_accuracy": 0.22990221232175828,
      "num_tokens": 81971334.0,
      "step": 35785
    },
    {
      "entropy": 5.77884726524353,
      "epoch": 3.537959667852906,
      "grad_norm": 1.359375,
      "learning_rate": 0.00037880143209029844,
      "loss": 4.9489,
      "mean_token_accuracy": 0.2228084087371826,
      "num_tokens": 81981740.0,
      "step": 35790
    },
    {
      "entropy": 5.734438467025757,
      "epoch": 3.5384539343614074,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003787701207091171,
      "loss": 5.006,
      "mean_token_accuracy": 0.21568506211042404,
      "num_tokens": 81992987.0,
      "step": 35795
    },
    {
      "entropy": 5.715889883041382,
      "epoch": 3.538948200869909,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037873880677517476,
      "loss": 4.9496,
      "mean_token_accuracy": 0.22788602411746978,
      "num_tokens": 82003590.0,
      "step": 35800
    },
    {
      "entropy": 5.8146388053894045,
      "epoch": 3.5394424673784104,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003787074902892416,
      "loss": 5.0225,
      "mean_token_accuracy": 0.21453812718391418,
      "num_tokens": 82014894.0,
      "step": 35805
    },
    {
      "entropy": 5.7466412544250485,
      "epoch": 3.5399367338869117,
      "grad_norm": 1.3125,
      "learning_rate": 0.00037867617125208813,
      "loss": 4.9165,
      "mean_token_accuracy": 0.22870483249425888,
      "num_tokens": 82025214.0,
      "step": 35810
    },
    {
      "entropy": 5.702399587631225,
      "epoch": 3.5404310003954134,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003786448496644848,
      "loss": 4.93,
      "mean_token_accuracy": 0.22937235981225967,
      "num_tokens": 82038866.0,
      "step": 35815
    },
    {
      "entropy": 5.792431306838989,
      "epoch": 3.5409252669039146,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000378613525527202,
      "loss": 5.0414,
      "mean_token_accuracy": 0.22038553655147552,
      "num_tokens": 82051114.0,
      "step": 35820
    },
    {
      "entropy": 5.709439516067505,
      "epoch": 3.541419533412416,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037858219884101047,
      "loss": 4.9661,
      "mean_token_accuracy": 0.2258375257253647,
      "num_tokens": 82063091.0,
      "step": 35825
    },
    {
      "entropy": 5.749315595626831,
      "epoch": 3.541913799920917,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003785508696066808,
      "loss": 4.9332,
      "mean_token_accuracy": 0.22747773081064224,
      "num_tokens": 82073363.0,
      "step": 35830
    },
    {
      "entropy": 5.827824115753174,
      "epoch": 3.542408066429419,
      "grad_norm": 1.34375,
      "learning_rate": 0.00037851953782498367,
      "loss": 5.0215,
      "mean_token_accuracy": 0.2127931922674179,
      "num_tokens": 82083477.0,
      "step": 35835
    },
    {
      "entropy": 5.7536694526672365,
      "epoch": 3.54290233293792,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003784882034966898,
      "loss": 4.8898,
      "mean_token_accuracy": 0.22668828219175338,
      "num_tokens": 82094738.0,
      "step": 35840
    },
    {
      "entropy": 5.756086730957032,
      "epoch": 3.5433965994464214,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037845686662257,
      "loss": 4.9593,
      "mean_token_accuracy": 0.22838023751974107,
      "num_tokens": 82105769.0,
      "step": 35845
    },
    {
      "entropy": 5.776323509216309,
      "epoch": 3.543890865954923,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003784255272033954,
      "loss": 5.0026,
      "mean_token_accuracy": 0.21800482869148255,
      "num_tokens": 82117644.0,
      "step": 35850
    },
    {
      "entropy": 5.696103239059449,
      "epoch": 3.5443851324634243,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037839418523993667,
      "loss": 4.9072,
      "mean_token_accuracy": 0.2297068014740944,
      "num_tokens": 82129716.0,
      "step": 35855
    },
    {
      "entropy": 5.643079662322998,
      "epoch": 3.5448793989719256,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000378362840732965,
      "loss": 4.8428,
      "mean_token_accuracy": 0.23724465072155,
      "num_tokens": 82142378.0,
      "step": 35860
    },
    {
      "entropy": 5.690996694564819,
      "epoch": 3.545373665480427,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037833149368325143,
      "loss": 4.8713,
      "mean_token_accuracy": 0.2325006291270256,
      "num_tokens": 82152893.0,
      "step": 35865
    },
    {
      "entropy": 5.7738429546356205,
      "epoch": 3.5458679319889286,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000378300144091567,
      "loss": 5.0153,
      "mean_token_accuracy": 0.22508428245782852,
      "num_tokens": 82165813.0,
      "step": 35870
    },
    {
      "entropy": 5.734207105636597,
      "epoch": 3.54636219849743,
      "grad_norm": 1.171875,
      "learning_rate": 0.000378268791958683,
      "loss": 4.9498,
      "mean_token_accuracy": 0.21678674221038818,
      "num_tokens": 82177641.0,
      "step": 35875
    },
    {
      "entropy": 5.791926050186158,
      "epoch": 3.546856465005931,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003782374372853708,
      "loss": 4.9406,
      "mean_token_accuracy": 0.2200591668486595,
      "num_tokens": 82187594.0,
      "step": 35880
    },
    {
      "entropy": 5.6983870506286625,
      "epoch": 3.547350731514433,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003782060800724016,
      "loss": 4.9594,
      "mean_token_accuracy": 0.21996378004550934,
      "num_tokens": 82199177.0,
      "step": 35885
    },
    {
      "entropy": 5.790413522720337,
      "epoch": 3.547844998022934,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003781747203205468,
      "loss": 4.9671,
      "mean_token_accuracy": 0.22168762236833572,
      "num_tokens": 82211176.0,
      "step": 35890
    },
    {
      "entropy": 5.7486106872558596,
      "epoch": 3.5483392645314353,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037814335803057773,
      "loss": 4.875,
      "mean_token_accuracy": 0.23064166754484178,
      "num_tokens": 82222767.0,
      "step": 35895
    },
    {
      "entropy": 5.727271366119385,
      "epoch": 3.5488335310399366,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003781119932032662,
      "loss": 4.9858,
      "mean_token_accuracy": 0.22085098028182984,
      "num_tokens": 82234060.0,
      "step": 35900
    },
    {
      "entropy": 5.773919725418091,
      "epoch": 3.549327797548438,
      "grad_norm": 1.296875,
      "learning_rate": 0.00037808062583938367,
      "loss": 4.9548,
      "mean_token_accuracy": 0.22422117441892625,
      "num_tokens": 82244910.0,
      "step": 35905
    },
    {
      "entropy": 5.790648698806763,
      "epoch": 3.5498220640569396,
      "grad_norm": 1.125,
      "learning_rate": 0.0003780492559397017,
      "loss": 5.0498,
      "mean_token_accuracy": 0.21800085604190828,
      "num_tokens": 82257730.0,
      "step": 35910
    },
    {
      "entropy": 5.724204254150391,
      "epoch": 3.550316330565441,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003780178835049921,
      "loss": 4.9007,
      "mean_token_accuracy": 0.2358308032155037,
      "num_tokens": 82268688.0,
      "step": 35915
    },
    {
      "entropy": 5.741845989227295,
      "epoch": 3.5508105970739425,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003779865085360265,
      "loss": 5.0002,
      "mean_token_accuracy": 0.2221138969063759,
      "num_tokens": 82279562.0,
      "step": 35920
    },
    {
      "entropy": 5.753812503814697,
      "epoch": 3.551304863582444,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003779551310335768,
      "loss": 4.9078,
      "mean_token_accuracy": 0.23089702427387238,
      "num_tokens": 82289945.0,
      "step": 35925
    },
    {
      "entropy": 5.8040056228637695,
      "epoch": 3.551799130090945,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037792375099841505,
      "loss": 4.9844,
      "mean_token_accuracy": 0.22936297953128815,
      "num_tokens": 82301878.0,
      "step": 35930
    },
    {
      "entropy": 5.774952220916748,
      "epoch": 3.5522933965994463,
      "grad_norm": 1.234375,
      "learning_rate": 0.00037789236843131297,
      "loss": 5.0059,
      "mean_token_accuracy": 0.22326663881540298,
      "num_tokens": 82313565.0,
      "step": 35935
    },
    {
      "entropy": 5.746735429763794,
      "epoch": 3.5527876631079476,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003778609833330426,
      "loss": 4.994,
      "mean_token_accuracy": 0.21841957569122314,
      "num_tokens": 82324375.0,
      "step": 35940
    },
    {
      "entropy": 5.828620910644531,
      "epoch": 3.5532819296164493,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003778295957043761,
      "loss": 5.1064,
      "mean_token_accuracy": 0.20658788084983826,
      "num_tokens": 82335642.0,
      "step": 35945
    },
    {
      "entropy": 5.738768482208252,
      "epoch": 3.5537761961249505,
      "grad_norm": 1.25,
      "learning_rate": 0.0003777982055460857,
      "loss": 4.8742,
      "mean_token_accuracy": 0.23628106564283372,
      "num_tokens": 82345847.0,
      "step": 35950
    },
    {
      "entropy": 5.830169248580932,
      "epoch": 3.554270462633452,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003777668128589434,
      "loss": 5.0507,
      "mean_token_accuracy": 0.21960015296936036,
      "num_tokens": 82358249.0,
      "step": 35955
    },
    {
      "entropy": 5.746911525726318,
      "epoch": 3.5547647291419535,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003777354176437216,
      "loss": 4.9596,
      "mean_token_accuracy": 0.22475355714559556,
      "num_tokens": 82369996.0,
      "step": 35960
    },
    {
      "entropy": 5.789019298553467,
      "epoch": 3.555258995650455,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003777040199011926,
      "loss": 4.9181,
      "mean_token_accuracy": 0.22842899560928345,
      "num_tokens": 82381569.0,
      "step": 35965
    },
    {
      "entropy": 5.749431037902832,
      "epoch": 3.555753262158956,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037767261963212866,
      "loss": 4.9924,
      "mean_token_accuracy": 0.21827427744865419,
      "num_tokens": 82392343.0,
      "step": 35970
    },
    {
      "entropy": 5.692380571365357,
      "epoch": 3.5562475286674573,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037764121683730247,
      "loss": 4.9927,
      "mean_token_accuracy": 0.2241477847099304,
      "num_tokens": 82403681.0,
      "step": 35975
    },
    {
      "entropy": 5.789081001281739,
      "epoch": 3.556741795175959,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037760981151748645,
      "loss": 5.0202,
      "mean_token_accuracy": 0.21497927457094193,
      "num_tokens": 82415773.0,
      "step": 35980
    },
    {
      "entropy": 5.723096466064453,
      "epoch": 3.5572360616844603,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003775784036734531,
      "loss": 4.8679,
      "mean_token_accuracy": 0.23354502469301225,
      "num_tokens": 82426636.0,
      "step": 35985
    },
    {
      "entropy": 5.836662340164184,
      "epoch": 3.5577303281929615,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003775469933059751,
      "loss": 5.0604,
      "mean_token_accuracy": 0.214945450425148,
      "num_tokens": 82438546.0,
      "step": 35990
    },
    {
      "entropy": 5.730599355697632,
      "epoch": 3.5582245947014632,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037751558041582514,
      "loss": 4.9732,
      "mean_token_accuracy": 0.223971027135849,
      "num_tokens": 82451911.0,
      "step": 35995
    },
    {
      "entropy": 5.769511651992798,
      "epoch": 3.5587188612099645,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000377484165003776,
      "loss": 4.895,
      "mean_token_accuracy": 0.22900905460119247,
      "num_tokens": 82464691.0,
      "step": 36000
    },
    {
      "epoch": 3.5587188612099645,
      "eval_entropy": 5.508558938306782,
      "eval_loss": 5.021881103515625,
      "eval_mean_token_accuracy": 0.22951890462952199,
      "eval_num_tokens": 82464691.0,
      "eval_runtime": 20.5404,
      "eval_samples_per_second": 1582.879,
      "eval_steps_per_second": 197.903,
      "step": 36000
    },
    {
      "entropy": 5.740245532989502,
      "epoch": 3.5592131277184658,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037745274707060055,
      "loss": 4.9868,
      "mean_token_accuracy": 0.22039039433002472,
      "num_tokens": 82476498.0,
      "step": 36005
    },
    {
      "entropy": 5.679426479339599,
      "epoch": 3.559707394226967,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037742132661707157,
      "loss": 4.9214,
      "mean_token_accuracy": 0.2336275190114975,
      "num_tokens": 82488434.0,
      "step": 36010
    },
    {
      "entropy": 5.725425863265992,
      "epoch": 3.5602016607354683,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037738990364396207,
      "loss": 4.8866,
      "mean_token_accuracy": 0.22621850967407225,
      "num_tokens": 82500277.0,
      "step": 36015
    },
    {
      "entropy": 5.796396636962891,
      "epoch": 3.56069592724397,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037735847815204503,
      "loss": 5.014,
      "mean_token_accuracy": 0.2212940573692322,
      "num_tokens": 82513107.0,
      "step": 36020
    },
    {
      "entropy": 5.76483907699585,
      "epoch": 3.5611901937524713,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00037732705014209366,
      "loss": 4.9792,
      "mean_token_accuracy": 0.2203971266746521,
      "num_tokens": 82524967.0,
      "step": 36025
    },
    {
      "entropy": 5.723723840713501,
      "epoch": 3.561684460260973,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003772956196148809,
      "loss": 4.9577,
      "mean_token_accuracy": 0.2198776811361313,
      "num_tokens": 82536721.0,
      "step": 36030
    },
    {
      "entropy": 5.781989431381225,
      "epoch": 3.5621787267694742,
      "grad_norm": 1.125,
      "learning_rate": 0.00037726418657118004,
      "loss": 4.9882,
      "mean_token_accuracy": 0.22244239151477813,
      "num_tokens": 82549386.0,
      "step": 36035
    },
    {
      "entropy": 5.763472843170166,
      "epoch": 3.5626729932779755,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037723275101176436,
      "loss": 4.8728,
      "mean_token_accuracy": 0.23617226481437684,
      "num_tokens": 82560207.0,
      "step": 36040
    },
    {
      "entropy": 5.762670230865479,
      "epoch": 3.5631672597864767,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003772013129374071,
      "loss": 5.0191,
      "mean_token_accuracy": 0.21680087298154832,
      "num_tokens": 82570484.0,
      "step": 36045
    },
    {
      "entropy": 5.721595001220703,
      "epoch": 3.563661526294978,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003771698723488817,
      "loss": 4.9253,
      "mean_token_accuracy": 0.2233380541205406,
      "num_tokens": 82581833.0,
      "step": 36050
    },
    {
      "entropy": 5.661393451690674,
      "epoch": 3.5641557928034797,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037713842924696155,
      "loss": 4.8701,
      "mean_token_accuracy": 0.23219945877790452,
      "num_tokens": 82593351.0,
      "step": 36055
    },
    {
      "entropy": 5.753630208969116,
      "epoch": 3.564650059311981,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003771069836324203,
      "loss": 5.0005,
      "mean_token_accuracy": 0.22054840326309205,
      "num_tokens": 82606644.0,
      "step": 36060
    },
    {
      "entropy": 5.716038179397583,
      "epoch": 3.5651443258204822,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003770755355060313,
      "loss": 5.0098,
      "mean_token_accuracy": 0.22876815646886825,
      "num_tokens": 82618779.0,
      "step": 36065
    },
    {
      "entropy": 5.773155450820923,
      "epoch": 3.565638592328984,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003770440848685683,
      "loss": 4.9249,
      "mean_token_accuracy": 0.22824518829584123,
      "num_tokens": 82629968.0,
      "step": 36070
    },
    {
      "entropy": 5.745453691482544,
      "epoch": 3.566132858837485,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037701263172080514,
      "loss": 4.9566,
      "mean_token_accuracy": 0.23086657077074052,
      "num_tokens": 82641586.0,
      "step": 36075
    },
    {
      "entropy": 5.830714082717895,
      "epoch": 3.5666271253459865,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037698117606351525,
      "loss": 5.0395,
      "mean_token_accuracy": 0.21885924637317658,
      "num_tokens": 82651901.0,
      "step": 36080
    },
    {
      "entropy": 5.70860276222229,
      "epoch": 3.5671213918544877,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003769497178974726,
      "loss": 4.9103,
      "mean_token_accuracy": 0.22742430716753007,
      "num_tokens": 82663055.0,
      "step": 36085
    },
    {
      "entropy": 5.640104532241821,
      "epoch": 3.5676156583629894,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003769182572234512,
      "loss": 4.8747,
      "mean_token_accuracy": 0.23550799787044524,
      "num_tokens": 82674709.0,
      "step": 36090
    },
    {
      "entropy": 5.739613437652588,
      "epoch": 3.5681099248714907,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003768867940422248,
      "loss": 4.8719,
      "mean_token_accuracy": 0.23496613949537276,
      "num_tokens": 82685215.0,
      "step": 36095
    },
    {
      "entropy": 5.791907596588135,
      "epoch": 3.568604191379992,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003768553283545674,
      "loss": 5.0157,
      "mean_token_accuracy": 0.22023407071828843,
      "num_tokens": 82697473.0,
      "step": 36100
    },
    {
      "entropy": 5.699418020248413,
      "epoch": 3.5690984578884937,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003768238601612532,
      "loss": 4.9234,
      "mean_token_accuracy": 0.2325322523713112,
      "num_tokens": 82708033.0,
      "step": 36105
    },
    {
      "entropy": 5.752238988876343,
      "epoch": 3.569592724396995,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003767923894630562,
      "loss": 4.9212,
      "mean_token_accuracy": 0.22852040380239486,
      "num_tokens": 82719410.0,
      "step": 36110
    },
    {
      "entropy": 5.804840469360352,
      "epoch": 3.570086990905496,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00037676091626075056,
      "loss": 4.8861,
      "mean_token_accuracy": 0.22700747698545456,
      "num_tokens": 82730648.0,
      "step": 36115
    },
    {
      "entropy": 5.77003664970398,
      "epoch": 3.5705812574139975,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003767294405551107,
      "loss": 4.9699,
      "mean_token_accuracy": 0.2223377823829651,
      "num_tokens": 82742041.0,
      "step": 36120
    },
    {
      "entropy": 5.75500111579895,
      "epoch": 3.571075523922499,
      "grad_norm": 1.25,
      "learning_rate": 0.0003766979623469108,
      "loss": 4.987,
      "mean_token_accuracy": 0.21900109946727753,
      "num_tokens": 82754083.0,
      "step": 36125
    },
    {
      "entropy": 5.780793952941894,
      "epoch": 3.5715697904310004,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003766664816369252,
      "loss": 5.0392,
      "mean_token_accuracy": 0.21500103175640106,
      "num_tokens": 82766735.0,
      "step": 36130
    },
    {
      "entropy": 5.841730451583862,
      "epoch": 3.5720640569395017,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037663499842592845,
      "loss": 5.0888,
      "mean_token_accuracy": 0.21570731252431868,
      "num_tokens": 82778498.0,
      "step": 36135
    },
    {
      "entropy": 5.879758358001709,
      "epoch": 3.5725583234480034,
      "grad_norm": 1.34375,
      "learning_rate": 0.00037660351271469496,
      "loss": 5.0675,
      "mean_token_accuracy": 0.2123200163245201,
      "num_tokens": 82789927.0,
      "step": 36140
    },
    {
      "entropy": 5.826226234436035,
      "epoch": 3.5730525899565047,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003765720245039993,
      "loss": 4.9625,
      "mean_token_accuracy": 0.214798441529274,
      "num_tokens": 82800846.0,
      "step": 36145
    },
    {
      "entropy": 5.729911041259766,
      "epoch": 3.573546856465006,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037654053379461606,
      "loss": 4.9664,
      "mean_token_accuracy": 0.22246886789798737,
      "num_tokens": 82812912.0,
      "step": 36150
    },
    {
      "entropy": 5.793826770782471,
      "epoch": 3.574041122973507,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003765090405873199,
      "loss": 5.0186,
      "mean_token_accuracy": 0.22350654602050782,
      "num_tokens": 82824186.0,
      "step": 36155
    },
    {
      "entropy": 5.732223701477051,
      "epoch": 3.5745353894820084,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003764775448828856,
      "loss": 4.9872,
      "mean_token_accuracy": 0.22499649971723557,
      "num_tokens": 82835599.0,
      "step": 36160
    },
    {
      "entropy": 5.729511404037476,
      "epoch": 3.57502965599051,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037644604668208805,
      "loss": 4.8683,
      "mean_token_accuracy": 0.23070819079875945,
      "num_tokens": 82846272.0,
      "step": 36165
    },
    {
      "entropy": 5.7895670413970945,
      "epoch": 3.5755239224990114,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003764145459857019,
      "loss": 4.9833,
      "mean_token_accuracy": 0.22185556292533876,
      "num_tokens": 82858245.0,
      "step": 36170
    },
    {
      "entropy": 5.78285436630249,
      "epoch": 3.576018189007513,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00037638304279450225,
      "loss": 4.9801,
      "mean_token_accuracy": 0.22269493639469146,
      "num_tokens": 82869951.0,
      "step": 36175
    },
    {
      "entropy": 5.738373279571533,
      "epoch": 3.5765124555160144,
      "grad_norm": 1.3125,
      "learning_rate": 0.00037635153710926394,
      "loss": 4.8996,
      "mean_token_accuracy": 0.23027793914079667,
      "num_tokens": 82880489.0,
      "step": 36180
    },
    {
      "entropy": 5.749619817733764,
      "epoch": 3.5770067220245156,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037632002893076217,
      "loss": 5.0165,
      "mean_token_accuracy": 0.2174355834722519,
      "num_tokens": 82892188.0,
      "step": 36185
    },
    {
      "entropy": 5.728027057647705,
      "epoch": 3.577500988533017,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000376288518259772,
      "loss": 4.9204,
      "mean_token_accuracy": 0.23277533203363418,
      "num_tokens": 82903504.0,
      "step": 36190
    },
    {
      "entropy": 5.69014163017273,
      "epoch": 3.577995255041518,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003762570050970685,
      "loss": 4.8779,
      "mean_token_accuracy": 0.22655326277017593,
      "num_tokens": 82914396.0,
      "step": 36195
    },
    {
      "entropy": 5.695727014541626,
      "epoch": 3.57848952155002,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000376225489443427,
      "loss": 4.8795,
      "mean_token_accuracy": 0.22872186601161956,
      "num_tokens": 82925759.0,
      "step": 36200
    },
    {
      "entropy": 5.727814674377441,
      "epoch": 3.578983788058521,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003761939712996227,
      "loss": 4.9219,
      "mean_token_accuracy": 0.2267654374241829,
      "num_tokens": 82937474.0,
      "step": 36205
    },
    {
      "entropy": 5.780701684951782,
      "epoch": 3.5794780545670224,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003761624506664311,
      "loss": 5.0169,
      "mean_token_accuracy": 0.21450553387403487,
      "num_tokens": 82948686.0,
      "step": 36210
    },
    {
      "entropy": 5.805887126922608,
      "epoch": 3.579972321075524,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003761309275446274,
      "loss": 4.9984,
      "mean_token_accuracy": 0.2257796049118042,
      "num_tokens": 82960099.0,
      "step": 36215
    },
    {
      "entropy": 5.742119455337525,
      "epoch": 3.5804665875840254,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003760994019349873,
      "loss": 4.9336,
      "mean_token_accuracy": 0.2234094113111496,
      "num_tokens": 82971415.0,
      "step": 36220
    },
    {
      "entropy": 5.76347918510437,
      "epoch": 3.5809608540925266,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003760678738382862,
      "loss": 5.0082,
      "mean_token_accuracy": 0.21401462107896804,
      "num_tokens": 82983581.0,
      "step": 36225
    },
    {
      "entropy": 5.764483880996704,
      "epoch": 3.581455120601028,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003760363432552997,
      "loss": 4.968,
      "mean_token_accuracy": 0.2259003698825836,
      "num_tokens": 82994817.0,
      "step": 36230
    },
    {
      "entropy": 5.738733863830566,
      "epoch": 3.5819493871095296,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003760048101868035,
      "loss": 4.8705,
      "mean_token_accuracy": 0.2328740105032921,
      "num_tokens": 83004620.0,
      "step": 36235
    },
    {
      "entropy": 5.744269227981567,
      "epoch": 3.582443653618031,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00037597327463357336,
      "loss": 4.9715,
      "mean_token_accuracy": 0.22100413888692855,
      "num_tokens": 83015257.0,
      "step": 36240
    },
    {
      "entropy": 5.654111433029175,
      "epoch": 3.582937920126532,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003759417365963849,
      "loss": 4.8207,
      "mean_token_accuracy": 0.2398465797305107,
      "num_tokens": 83026765.0,
      "step": 36245
    },
    {
      "entropy": 5.75731258392334,
      "epoch": 3.583432186635034,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037591019607601407,
      "loss": 4.9815,
      "mean_token_accuracy": 0.2226743295788765,
      "num_tokens": 83038893.0,
      "step": 36250
    },
    {
      "entropy": 5.677341747283935,
      "epoch": 3.583926453143535,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037587865307323683,
      "loss": 4.9459,
      "mean_token_accuracy": 0.2274229571223259,
      "num_tokens": 83050615.0,
      "step": 36255
    },
    {
      "entropy": 5.781007528305054,
      "epoch": 3.5844207196520363,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037584710758882897,
      "loss": 5.0057,
      "mean_token_accuracy": 0.22131242007017135,
      "num_tokens": 83062459.0,
      "step": 36260
    },
    {
      "entropy": 5.831990385055542,
      "epoch": 3.5849149861605376,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00037581555962356666,
      "loss": 5.056,
      "mean_token_accuracy": 0.21695964485406877,
      "num_tokens": 83073563.0,
      "step": 36265
    },
    {
      "entropy": 5.791493892669678,
      "epoch": 3.585409252669039,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00037578400917822604,
      "loss": 5.0137,
      "mean_token_accuracy": 0.21658323109149932,
      "num_tokens": 83084831.0,
      "step": 36270
    },
    {
      "entropy": 5.727228403091431,
      "epoch": 3.5859035191775406,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000375752456253583,
      "loss": 4.9179,
      "mean_token_accuracy": 0.2277284160256386,
      "num_tokens": 83096504.0,
      "step": 36275
    },
    {
      "entropy": 5.727636623382568,
      "epoch": 3.586397785686042,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000375720900850414,
      "loss": 4.9922,
      "mean_token_accuracy": 0.22366588711738586,
      "num_tokens": 83108312.0,
      "step": 36280
    },
    {
      "entropy": 5.7168800830841064,
      "epoch": 3.5868920521945435,
      "grad_norm": 1.25,
      "learning_rate": 0.0003756893429694952,
      "loss": 4.8958,
      "mean_token_accuracy": 0.23572079986333846,
      "num_tokens": 83119040.0,
      "step": 36285
    },
    {
      "entropy": 5.742487573623658,
      "epoch": 3.587386318703045,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00037565778261160296,
      "loss": 4.8485,
      "mean_token_accuracy": 0.2396532416343689,
      "num_tokens": 83129803.0,
      "step": 36290
    },
    {
      "entropy": 5.742160034179688,
      "epoch": 3.587880585211546,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003756262197775137,
      "loss": 4.9295,
      "mean_token_accuracy": 0.22827044129371643,
      "num_tokens": 83142364.0,
      "step": 36295
    },
    {
      "entropy": 5.721851491928101,
      "epoch": 3.5883748517200473,
      "grad_norm": 1.375,
      "learning_rate": 0.00037559465446800364,
      "loss": 4.9721,
      "mean_token_accuracy": 0.2236535966396332,
      "num_tokens": 83152813.0,
      "step": 36300
    },
    {
      "entropy": 5.856607675552368,
      "epoch": 3.5888691182285486,
      "grad_norm": 1.3125,
      "learning_rate": 0.00037556308668384964,
      "loss": 5.1089,
      "mean_token_accuracy": 0.21373351961374282,
      "num_tokens": 83164052.0,
      "step": 36305
    },
    {
      "entropy": 5.756061792373657,
      "epoch": 3.5893633847370503,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037553151642582807,
      "loss": 4.9208,
      "mean_token_accuracy": 0.22529462575912476,
      "num_tokens": 83176236.0,
      "step": 36310
    },
    {
      "entropy": 5.727978754043579,
      "epoch": 3.5898576512455516,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003754999436947156,
      "loss": 4.979,
      "mean_token_accuracy": 0.22029863744974137,
      "num_tokens": 83187173.0,
      "step": 36315
    },
    {
      "entropy": 5.708658361434937,
      "epoch": 3.590351917754053,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003754683684912889,
      "loss": 4.9648,
      "mean_token_accuracy": 0.2277854710817337,
      "num_tokens": 83199628.0,
      "step": 36320
    },
    {
      "entropy": 5.797570037841797,
      "epoch": 3.5908461842625545,
      "grad_norm": 1.359375,
      "learning_rate": 0.00037543679081632474,
      "loss": 4.9796,
      "mean_token_accuracy": 0.22740527093410492,
      "num_tokens": 83211748.0,
      "step": 36325
    },
    {
      "entropy": 5.696346712112427,
      "epoch": 3.591340450771056,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003754052106706,
      "loss": 4.8934,
      "mean_token_accuracy": 0.2308681309223175,
      "num_tokens": 83223078.0,
      "step": 36330
    },
    {
      "entropy": 5.761122512817383,
      "epoch": 3.591834717279557,
      "grad_norm": 1.25,
      "learning_rate": 0.00037537362805489143,
      "loss": 4.9436,
      "mean_token_accuracy": 0.22822879254817963,
      "num_tokens": 83234046.0,
      "step": 36335
    },
    {
      "entropy": 5.763217115402222,
      "epoch": 3.5923289837880583,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003753420429699761,
      "loss": 4.987,
      "mean_token_accuracy": 0.2167970821261406,
      "num_tokens": 83244121.0,
      "step": 36340
    },
    {
      "entropy": 5.795695495605469,
      "epoch": 3.59282325029656,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003753104554166309,
      "loss": 4.9655,
      "mean_token_accuracy": 0.2181445837020874,
      "num_tokens": 83254182.0,
      "step": 36345
    },
    {
      "entropy": 5.716990041732788,
      "epoch": 3.5933175168050613,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003752788653956329,
      "loss": 4.9216,
      "mean_token_accuracy": 0.22640663981437684,
      "num_tokens": 83264527.0,
      "step": 36350
    },
    {
      "entropy": 5.810683536529541,
      "epoch": 3.5938117833135625,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003752472729077593,
      "loss": 4.9842,
      "mean_token_accuracy": 0.21407381892204286,
      "num_tokens": 83275186.0,
      "step": 36355
    },
    {
      "entropy": 5.832469272613525,
      "epoch": 3.5943060498220643,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003752156779537873,
      "loss": 5.08,
      "mean_token_accuracy": 0.21350655406713487,
      "num_tokens": 83286919.0,
      "step": 36360
    },
    {
      "entropy": 5.744521188735962,
      "epoch": 3.5948003163305655,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037518408053449404,
      "loss": 4.9434,
      "mean_token_accuracy": 0.21720613539218903,
      "num_tokens": 83298112.0,
      "step": 36365
    },
    {
      "entropy": 5.723997592926025,
      "epoch": 3.5952945828390668,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003751524806506568,
      "loss": 5.0129,
      "mean_token_accuracy": 0.22148964405059815,
      "num_tokens": 83308978.0,
      "step": 36370
    },
    {
      "entropy": 5.774290704727173,
      "epoch": 3.595788849347568,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037512087830305306,
      "loss": 5.0604,
      "mean_token_accuracy": 0.21661423593759538,
      "num_tokens": 83320780.0,
      "step": 36375
    },
    {
      "entropy": 5.825651788711548,
      "epoch": 3.5962831158560697,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003750892734924601,
      "loss": 5.0326,
      "mean_token_accuracy": 0.21498379707336426,
      "num_tokens": 83332070.0,
      "step": 36380
    },
    {
      "entropy": 5.786178779602051,
      "epoch": 3.596777382364571,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003750576662196556,
      "loss": 4.9959,
      "mean_token_accuracy": 0.22339484691619874,
      "num_tokens": 83345506.0,
      "step": 36385
    },
    {
      "entropy": 5.840700149536133,
      "epoch": 3.5972716488730723,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003750260564854169,
      "loss": 5.0732,
      "mean_token_accuracy": 0.21451534181833268,
      "num_tokens": 83357097.0,
      "step": 36390
    },
    {
      "entropy": 5.726877880096436,
      "epoch": 3.597765915381574,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003749944442905217,
      "loss": 4.871,
      "mean_token_accuracy": 0.23418209701776505,
      "num_tokens": 83369088.0,
      "step": 36395
    },
    {
      "entropy": 5.780043506622315,
      "epoch": 3.5982601818900752,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003749628296357477,
      "loss": 4.9943,
      "mean_token_accuracy": 0.22435135245323182,
      "num_tokens": 83381249.0,
      "step": 36400
    },
    {
      "entropy": 5.7276922225952145,
      "epoch": 3.5987544483985765,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037493121252187256,
      "loss": 4.9486,
      "mean_token_accuracy": 0.22232571691274644,
      "num_tokens": 83392000.0,
      "step": 36405
    },
    {
      "entropy": 5.70304913520813,
      "epoch": 3.5992487149070778,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003748995929496741,
      "loss": 4.8728,
      "mean_token_accuracy": 0.23157589584589006,
      "num_tokens": 83403841.0,
      "step": 36410
    },
    {
      "entropy": 5.761758661270141,
      "epoch": 3.599742981415579,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037486797091993015,
      "loss": 4.9907,
      "mean_token_accuracy": 0.22523076832294464,
      "num_tokens": 83415114.0,
      "step": 36415
    },
    {
      "entropy": 5.7278139114379885,
      "epoch": 3.6002372479240807,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037483634643341866,
      "loss": 5.0262,
      "mean_token_accuracy": 0.22027490586042403,
      "num_tokens": 83428158.0,
      "step": 36420
    },
    {
      "entropy": 5.83130841255188,
      "epoch": 3.600731514432582,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003748047194909175,
      "loss": 4.9998,
      "mean_token_accuracy": 0.22703043669462203,
      "num_tokens": 83439535.0,
      "step": 36425
    },
    {
      "entropy": 5.783418846130371,
      "epoch": 3.6012257809410837,
      "grad_norm": 1.234375,
      "learning_rate": 0.00037477309009320477,
      "loss": 4.9918,
      "mean_token_accuracy": 0.2218065306544304,
      "num_tokens": 83451215.0,
      "step": 36430
    },
    {
      "entropy": 5.808160209655762,
      "epoch": 3.601720047449585,
      "grad_norm": 1.25,
      "learning_rate": 0.00037474145824105854,
      "loss": 4.9839,
      "mean_token_accuracy": 0.21818913370370865,
      "num_tokens": 83461750.0,
      "step": 36435
    },
    {
      "entropy": 5.701957130432129,
      "epoch": 3.6022143139580862,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037470982393525697,
      "loss": 4.9387,
      "mean_token_accuracy": 0.2263297602534294,
      "num_tokens": 83473790.0,
      "step": 36440
    },
    {
      "entropy": 5.719630432128906,
      "epoch": 3.6027085804665875,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037467818717657825,
      "loss": 4.9481,
      "mean_token_accuracy": 0.2241119235754013,
      "num_tokens": 83485470.0,
      "step": 36445
    },
    {
      "entropy": 5.784329843521118,
      "epoch": 3.6032028469750887,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003746465479658007,
      "loss": 4.952,
      "mean_token_accuracy": 0.2236907035112381,
      "num_tokens": 83496007.0,
      "step": 36450
    },
    {
      "entropy": 5.792724657058716,
      "epoch": 3.6036971134835905,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003746149063037026,
      "loss": 5.0507,
      "mean_token_accuracy": 0.218854358792305,
      "num_tokens": 83508620.0,
      "step": 36455
    },
    {
      "entropy": 5.737733459472656,
      "epoch": 3.6041913799920917,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003745832621910623,
      "loss": 4.9369,
      "mean_token_accuracy": 0.23154576867818832,
      "num_tokens": 83520323.0,
      "step": 36460
    },
    {
      "entropy": 5.66882004737854,
      "epoch": 3.604685646500593,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037455161562865833,
      "loss": 4.9012,
      "mean_token_accuracy": 0.2287164568901062,
      "num_tokens": 83532471.0,
      "step": 36465
    },
    {
      "entropy": 5.720019054412842,
      "epoch": 3.6051799130090947,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003745199666172692,
      "loss": 4.9186,
      "mean_token_accuracy": 0.2306458532810211,
      "num_tokens": 83544156.0,
      "step": 36470
    },
    {
      "entropy": 5.7970428466796875,
      "epoch": 3.605674179517596,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037448831515767353,
      "loss": 4.9487,
      "mean_token_accuracy": 0.22554730772972106,
      "num_tokens": 83556321.0,
      "step": 36475
    },
    {
      "entropy": 5.759917783737182,
      "epoch": 3.606168446026097,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003744566612506497,
      "loss": 4.9429,
      "mean_token_accuracy": 0.2262440949678421,
      "num_tokens": 83566604.0,
      "step": 36480
    },
    {
      "entropy": 5.739758920669556,
      "epoch": 3.6066627125345985,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037442500489697666,
      "loss": 4.9451,
      "mean_token_accuracy": 0.22738515585660934,
      "num_tokens": 83578349.0,
      "step": 36485
    },
    {
      "entropy": 5.754388999938965,
      "epoch": 3.6071569790431,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003743933460974331,
      "loss": 4.9611,
      "mean_token_accuracy": 0.22614455968141556,
      "num_tokens": 83589198.0,
      "step": 36490
    },
    {
      "entropy": 5.838378810882569,
      "epoch": 3.6076512455516014,
      "grad_norm": 1.125,
      "learning_rate": 0.00037436168485279777,
      "loss": 5.04,
      "mean_token_accuracy": 0.21892516762018205,
      "num_tokens": 83602076.0,
      "step": 36495
    },
    {
      "entropy": 5.772137975692749,
      "epoch": 3.6081455120601027,
      "grad_norm": 1.296875,
      "learning_rate": 0.00037433002116384957,
      "loss": 4.9256,
      "mean_token_accuracy": 0.22084573954343795,
      "num_tokens": 83613340.0,
      "step": 36500
    },
    {
      "entropy": 5.730703067779541,
      "epoch": 3.6086397785686044,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037429835503136744,
      "loss": 4.9938,
      "mean_token_accuracy": 0.22658517807722092,
      "num_tokens": 83624715.0,
      "step": 36505
    },
    {
      "entropy": 5.704571723937988,
      "epoch": 3.6091340450771057,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003742666864561304,
      "loss": 5.012,
      "mean_token_accuracy": 0.22628515362739562,
      "num_tokens": 83635833.0,
      "step": 36510
    },
    {
      "entropy": 5.7348085880279545,
      "epoch": 3.609628311585607,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003742350154389174,
      "loss": 4.9124,
      "mean_token_accuracy": 0.22448546290397645,
      "num_tokens": 83647281.0,
      "step": 36515
    },
    {
      "entropy": 5.803982067108154,
      "epoch": 3.610122578094108,
      "grad_norm": 1.234375,
      "learning_rate": 0.00037420334198050775,
      "loss": 4.9975,
      "mean_token_accuracy": 0.220235675573349,
      "num_tokens": 83658745.0,
      "step": 36520
    },
    {
      "entropy": 5.795662927627563,
      "epoch": 3.6106168446026095,
      "grad_norm": 1.25,
      "learning_rate": 0.0003741716660816804,
      "loss": 4.961,
      "mean_token_accuracy": 0.21857482343912124,
      "num_tokens": 83671580.0,
      "step": 36525
    },
    {
      "entropy": 5.7819664478302,
      "epoch": 3.611111111111111,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037413998774321475,
      "loss": 4.9926,
      "mean_token_accuracy": 0.2225994125008583,
      "num_tokens": 83682667.0,
      "step": 36530
    },
    {
      "entropy": 5.736459112167358,
      "epoch": 3.6116053776196124,
      "grad_norm": 1.3125,
      "learning_rate": 0.00037410830696588994,
      "loss": 4.8967,
      "mean_token_accuracy": 0.23010231852531432,
      "num_tokens": 83693276.0,
      "step": 36535
    },
    {
      "entropy": 5.764947509765625,
      "epoch": 3.612099644128114,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037407662375048544,
      "loss": 5.0086,
      "mean_token_accuracy": 0.22436152398586273,
      "num_tokens": 83705304.0,
      "step": 36540
    },
    {
      "entropy": 5.740032386779785,
      "epoch": 3.6125939106366154,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037404493809778063,
      "loss": 4.9766,
      "mean_token_accuracy": 0.22539893686771392,
      "num_tokens": 83717166.0,
      "step": 36545
    },
    {
      "entropy": 5.747815895080566,
      "epoch": 3.6130881771451167,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00037401325000855495,
      "loss": 4.9098,
      "mean_token_accuracy": 0.23014338612556456,
      "num_tokens": 83727322.0,
      "step": 36550
    },
    {
      "entropy": 5.766996717453003,
      "epoch": 3.613582443653618,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037398155948358805,
      "loss": 4.9244,
      "mean_token_accuracy": 0.2272268533706665,
      "num_tokens": 83738721.0,
      "step": 36555
    },
    {
      "entropy": 5.712638425827026,
      "epoch": 3.614076710162119,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003739498665236593,
      "loss": 4.9392,
      "mean_token_accuracy": 0.2322855442762375,
      "num_tokens": 83750622.0,
      "step": 36560
    },
    {
      "entropy": 5.795522165298462,
      "epoch": 3.614570976670621,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003739181711295485,
      "loss": 5.0189,
      "mean_token_accuracy": 0.21720549762248992,
      "num_tokens": 83761755.0,
      "step": 36565
    },
    {
      "entropy": 5.692834186553955,
      "epoch": 3.615065243179122,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037388647330203545,
      "loss": 4.8672,
      "mean_token_accuracy": 0.23607228547334672,
      "num_tokens": 83772335.0,
      "step": 36570
    },
    {
      "entropy": 5.710109901428223,
      "epoch": 3.6155595096876234,
      "grad_norm": 1.125,
      "learning_rate": 0.0003738547730418999,
      "loss": 4.9937,
      "mean_token_accuracy": 0.2242111384868622,
      "num_tokens": 83784949.0,
      "step": 36575
    },
    {
      "entropy": 5.740633916854859,
      "epoch": 3.616053776196125,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003738230703499213,
      "loss": 4.973,
      "mean_token_accuracy": 0.222284696996212,
      "num_tokens": 83795555.0,
      "step": 36580
    },
    {
      "entropy": 5.757212352752686,
      "epoch": 3.6165480427046264,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00037379136522688,
      "loss": 4.9287,
      "mean_token_accuracy": 0.22794184237718582,
      "num_tokens": 83806682.0,
      "step": 36585
    },
    {
      "entropy": 5.769719362258911,
      "epoch": 3.6170423092131276,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037375965767355584,
      "loss": 4.9435,
      "mean_token_accuracy": 0.22778161913156508,
      "num_tokens": 83818485.0,
      "step": 36590
    },
    {
      "entropy": 5.639184093475341,
      "epoch": 3.617536575721629,
      "grad_norm": 1.25,
      "learning_rate": 0.0003737279476907287,
      "loss": 4.873,
      "mean_token_accuracy": 0.23780063539743423,
      "num_tokens": 83829757.0,
      "step": 36595
    },
    {
      "entropy": 5.725658893585205,
      "epoch": 3.6180308422301306,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037369623527917876,
      "loss": 4.9775,
      "mean_token_accuracy": 0.22660067081451415,
      "num_tokens": 83841827.0,
      "step": 36600
    },
    {
      "entropy": 5.669194507598877,
      "epoch": 3.618525108738632,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003736645204396861,
      "loss": 4.8136,
      "mean_token_accuracy": 0.23405515104532243,
      "num_tokens": 83852471.0,
      "step": 36605
    },
    {
      "entropy": 5.736182832717896,
      "epoch": 3.619019375247133,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003736328031730309,
      "loss": 4.9097,
      "mean_token_accuracy": 0.23059954792261123,
      "num_tokens": 83864322.0,
      "step": 36610
    },
    {
      "entropy": 5.789630556106568,
      "epoch": 3.619513641755635,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003736010834799936,
      "loss": 5.0008,
      "mean_token_accuracy": 0.22216322124004365,
      "num_tokens": 83875286.0,
      "step": 36615
    },
    {
      "entropy": 5.709139823913574,
      "epoch": 3.620007908264136,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003735693613613542,
      "loss": 4.8747,
      "mean_token_accuracy": 0.23301409780979157,
      "num_tokens": 83887794.0,
      "step": 36620
    },
    {
      "entropy": 5.766672945022583,
      "epoch": 3.6205021747726374,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003735376368178932,
      "loss": 4.9641,
      "mean_token_accuracy": 0.2260563775897026,
      "num_tokens": 83898644.0,
      "step": 36625
    },
    {
      "entropy": 5.761818790435791,
      "epoch": 3.6209964412811386,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003735059098503911,
      "loss": 5.0192,
      "mean_token_accuracy": 0.2149218052625656,
      "num_tokens": 83909944.0,
      "step": 36630
    },
    {
      "entropy": 5.75932240486145,
      "epoch": 3.6214907077896403,
      "grad_norm": 1.25,
      "learning_rate": 0.0003734741804596284,
      "loss": 4.9809,
      "mean_token_accuracy": 0.22384252846240998,
      "num_tokens": 83920928.0,
      "step": 36635
    },
    {
      "entropy": 5.725000858306885,
      "epoch": 3.6219849742981416,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003734424486463855,
      "loss": 4.8849,
      "mean_token_accuracy": 0.23618270009756087,
      "num_tokens": 83933519.0,
      "step": 36640
    },
    {
      "entropy": 5.808859968185425,
      "epoch": 3.622479240806643,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037341071441144316,
      "loss": 5.0408,
      "mean_token_accuracy": 0.22432909905910492,
      "num_tokens": 83946933.0,
      "step": 36645
    },
    {
      "entropy": 5.735810136795044,
      "epoch": 3.6229735073151446,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003733789777555819,
      "loss": 4.9351,
      "mean_token_accuracy": 0.2275382786989212,
      "num_tokens": 83957858.0,
      "step": 36650
    },
    {
      "entropy": 5.730549383163452,
      "epoch": 3.623467773823646,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037334723867958256,
      "loss": 4.9213,
      "mean_token_accuracy": 0.23089100867509843,
      "num_tokens": 83968363.0,
      "step": 36655
    },
    {
      "entropy": 5.751165103912354,
      "epoch": 3.623962040332147,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003733154971842259,
      "loss": 4.9118,
      "mean_token_accuracy": 0.23166271150112153,
      "num_tokens": 83978719.0,
      "step": 36660
    },
    {
      "entropy": 5.709116411209107,
      "epoch": 3.6244563068406483,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037328375327029274,
      "loss": 4.9116,
      "mean_token_accuracy": 0.225031578540802,
      "num_tokens": 83990195.0,
      "step": 36665
    },
    {
      "entropy": 5.719848680496216,
      "epoch": 3.6249505733491496,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037325200693856406,
      "loss": 5.0005,
      "mean_token_accuracy": 0.21995290517807006,
      "num_tokens": 84002068.0,
      "step": 36670
    },
    {
      "entropy": 5.713297176361084,
      "epoch": 3.6254448398576513,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003732202581898206,
      "loss": 4.9513,
      "mean_token_accuracy": 0.2229781463742256,
      "num_tokens": 84013405.0,
      "step": 36675
    },
    {
      "entropy": 5.812728071212769,
      "epoch": 3.6259391063661526,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003731885070248437,
      "loss": 5.0576,
      "mean_token_accuracy": 0.2140640303492546,
      "num_tokens": 84025650.0,
      "step": 36680
    },
    {
      "entropy": 5.712250995635986,
      "epoch": 3.6264333728746543,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037315675344441426,
      "loss": 4.8682,
      "mean_token_accuracy": 0.23593227863311766,
      "num_tokens": 84036753.0,
      "step": 36685
    },
    {
      "entropy": 5.833469343185425,
      "epoch": 3.6269276393831555,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003731249974493135,
      "loss": 5.0357,
      "mean_token_accuracy": 0.22362569719552994,
      "num_tokens": 84049082.0,
      "step": 36690
    },
    {
      "entropy": 5.787313032150268,
      "epoch": 3.627421905891657,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037309323904032254,
      "loss": 4.9753,
      "mean_token_accuracy": 0.21836165487766265,
      "num_tokens": 84059471.0,
      "step": 36695
    },
    {
      "entropy": 5.728457403182984,
      "epoch": 3.627916172400158,
      "grad_norm": 1.25,
      "learning_rate": 0.0003730614782182226,
      "loss": 4.9877,
      "mean_token_accuracy": 0.2323715940117836,
      "num_tokens": 84071475.0,
      "step": 36700
    },
    {
      "entropy": 5.754220294952392,
      "epoch": 3.6284104389086593,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003730297149837952,
      "loss": 4.9933,
      "mean_token_accuracy": 0.2199402153491974,
      "num_tokens": 84082465.0,
      "step": 36705
    },
    {
      "entropy": 5.746408700942993,
      "epoch": 3.628904705417161,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003729979493378215,
      "loss": 4.9136,
      "mean_token_accuracy": 0.22645459920167924,
      "num_tokens": 84093845.0,
      "step": 36710
    },
    {
      "entropy": 5.798669099807739,
      "epoch": 3.6293989719256623,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00037296618128108316,
      "loss": 4.9607,
      "mean_token_accuracy": 0.2252052053809166,
      "num_tokens": 84105865.0,
      "step": 36715
    },
    {
      "entropy": 5.748604822158813,
      "epoch": 3.6298932384341636,
      "grad_norm": 1.25,
      "learning_rate": 0.00037293441081436154,
      "loss": 4.9699,
      "mean_token_accuracy": 0.2211304634809494,
      "num_tokens": 84116376.0,
      "step": 36720
    },
    {
      "entropy": 5.6996846199035645,
      "epoch": 3.6303875049426653,
      "grad_norm": 1.25,
      "learning_rate": 0.0003729026379384381,
      "loss": 4.8853,
      "mean_token_accuracy": 0.23264809101819992,
      "num_tokens": 84126806.0,
      "step": 36725
    },
    {
      "entropy": 5.780883121490478,
      "epoch": 3.6308817714511665,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003728708626540948,
      "loss": 4.9801,
      "mean_token_accuracy": 0.2275504469871521,
      "num_tokens": 84136750.0,
      "step": 36730
    },
    {
      "entropy": 5.7800546169281,
      "epoch": 3.631376037959668,
      "grad_norm": 1.234375,
      "learning_rate": 0.00037283908496211295,
      "loss": 5.0488,
      "mean_token_accuracy": 0.21009242236614228,
      "num_tokens": 84148614.0,
      "step": 36735
    },
    {
      "entropy": 5.821066570281983,
      "epoch": 3.631870304468169,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003728073048632744,
      "loss": 5.0246,
      "mean_token_accuracy": 0.21641982793807985,
      "num_tokens": 84160347.0,
      "step": 36740
    },
    {
      "entropy": 5.784425497055054,
      "epoch": 3.6323645709766708,
      "grad_norm": 1.25,
      "learning_rate": 0.00037277552235836104,
      "loss": 5.0054,
      "mean_token_accuracy": 0.22381488382816314,
      "num_tokens": 84172240.0,
      "step": 36745
    },
    {
      "entropy": 5.791566562652588,
      "epoch": 3.632858837485172,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037274373744815466,
      "loss": 4.9852,
      "mean_token_accuracy": 0.22311032712459564,
      "num_tokens": 84184050.0,
      "step": 36750
    },
    {
      "entropy": 5.724491453170776,
      "epoch": 3.6333531039936733,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003727119501334371,
      "loss": 4.9491,
      "mean_token_accuracy": 0.2268210083246231,
      "num_tokens": 84194661.0,
      "step": 36755
    },
    {
      "entropy": 5.745427799224854,
      "epoch": 3.633847370502175,
      "grad_norm": 1.234375,
      "learning_rate": 0.00037268016041499055,
      "loss": 4.9577,
      "mean_token_accuracy": 0.22544464468955994,
      "num_tokens": 84205677.0,
      "step": 36760
    },
    {
      "entropy": 5.789601802825928,
      "epoch": 3.6343416370106763,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003726483682935968,
      "loss": 4.9446,
      "mean_token_accuracy": 0.22020726054906845,
      "num_tokens": 84217342.0,
      "step": 36765
    },
    {
      "entropy": 5.7664214134216305,
      "epoch": 3.6348359035191775,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037261657377003803,
      "loss": 4.978,
      "mean_token_accuracy": 0.2184204414486885,
      "num_tokens": 84229637.0,
      "step": 36770
    },
    {
      "entropy": 5.760247993469238,
      "epoch": 3.6353301700276788,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003725847768450964,
      "loss": 5.0226,
      "mean_token_accuracy": 0.22189818173646927,
      "num_tokens": 84241156.0,
      "step": 36775
    },
    {
      "entropy": 5.831021165847778,
      "epoch": 3.63582443653618,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003725529775195542,
      "loss": 4.9953,
      "mean_token_accuracy": 0.22493703812360763,
      "num_tokens": 84253172.0,
      "step": 36780
    },
    {
      "entropy": 5.759899091720581,
      "epoch": 3.6363187030446817,
      "grad_norm": 1.25,
      "learning_rate": 0.00037252117579419355,
      "loss": 4.9107,
      "mean_token_accuracy": 0.22540205270051955,
      "num_tokens": 84265349.0,
      "step": 36785
    },
    {
      "entropy": 5.659424304962158,
      "epoch": 3.636812969553183,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037248937166979687,
      "loss": 4.8948,
      "mean_token_accuracy": 0.22992639392614364,
      "num_tokens": 84275931.0,
      "step": 36790
    },
    {
      "entropy": 5.740987730026245,
      "epoch": 3.6373072360616847,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003724575651471465,
      "loss": 4.9773,
      "mean_token_accuracy": 0.22295117676258086,
      "num_tokens": 84286502.0,
      "step": 36795
    },
    {
      "entropy": 5.836364650726319,
      "epoch": 3.637801502570186,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003724257562270249,
      "loss": 5.0298,
      "mean_token_accuracy": 0.210837984085083,
      "num_tokens": 84297748.0,
      "step": 36800
    },
    {
      "entropy": 5.780948972702026,
      "epoch": 3.6382957690786872,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00037239394491021463,
      "loss": 5.0167,
      "mean_token_accuracy": 0.22192592322826385,
      "num_tokens": 84309733.0,
      "step": 36805
    },
    {
      "entropy": 5.799102735519409,
      "epoch": 3.6387900355871885,
      "grad_norm": 1.265625,
      "learning_rate": 0.00037236213119749823,
      "loss": 5.0421,
      "mean_token_accuracy": 0.2177009865641594,
      "num_tokens": 84320606.0,
      "step": 36810
    },
    {
      "entropy": 5.776124715805054,
      "epoch": 3.6392843020956898,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037233031508965823,
      "loss": 4.9362,
      "mean_token_accuracy": 0.22778684943914412,
      "num_tokens": 84331149.0,
      "step": 36815
    },
    {
      "entropy": 5.794047880172729,
      "epoch": 3.6397785686041915,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003722984965874774,
      "loss": 4.9779,
      "mean_token_accuracy": 0.22784479409456254,
      "num_tokens": 84343016.0,
      "step": 36820
    },
    {
      "entropy": 5.7488563537597654,
      "epoch": 3.6402728351126927,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003722666756917385,
      "loss": 5.0115,
      "mean_token_accuracy": 0.2191008672118187,
      "num_tokens": 84354846.0,
      "step": 36825
    },
    {
      "entropy": 5.762816095352173,
      "epoch": 3.640767101621194,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003722348524032242,
      "loss": 4.9667,
      "mean_token_accuracy": 0.2260129541158676,
      "num_tokens": 84366698.0,
      "step": 36830
    },
    {
      "entropy": 5.772870302200317,
      "epoch": 3.6412613681296957,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037220302672271756,
      "loss": 4.9675,
      "mean_token_accuracy": 0.2290659725666046,
      "num_tokens": 84378186.0,
      "step": 36835
    },
    {
      "entropy": 5.796053743362426,
      "epoch": 3.641755634638197,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00037217119865100127,
      "loss": 5.038,
      "mean_token_accuracy": 0.21514128297567367,
      "num_tokens": 84390301.0,
      "step": 36840
    },
    {
      "entropy": 5.723567056655884,
      "epoch": 3.642249901146698,
      "grad_norm": 1.265625,
      "learning_rate": 0.00037213936818885853,
      "loss": 4.9809,
      "mean_token_accuracy": 0.22278282195329666,
      "num_tokens": 84401134.0,
      "step": 36845
    },
    {
      "entropy": 5.750906276702881,
      "epoch": 3.6427441676551995,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037210753533707214,
      "loss": 4.9521,
      "mean_token_accuracy": 0.23028752356767654,
      "num_tokens": 84411734.0,
      "step": 36850
    },
    {
      "entropy": 5.8296699047088625,
      "epoch": 3.643238434163701,
      "grad_norm": 1.328125,
      "learning_rate": 0.00037207570009642536,
      "loss": 5.0533,
      "mean_token_accuracy": 0.2112957000732422,
      "num_tokens": 84422952.0,
      "step": 36855
    },
    {
      "entropy": 5.767107009887695,
      "epoch": 3.6437327006722025,
      "grad_norm": 1.25,
      "learning_rate": 0.0003720438624677014,
      "loss": 4.9075,
      "mean_token_accuracy": 0.23190695196390151,
      "num_tokens": 84433048.0,
      "step": 36860
    },
    {
      "entropy": 5.752732419967652,
      "epoch": 3.6442269671807037,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003720120224516832,
      "loss": 4.9811,
      "mean_token_accuracy": 0.22363260090351106,
      "num_tokens": 84443318.0,
      "step": 36865
    },
    {
      "entropy": 5.704191446304321,
      "epoch": 3.6447212336892054,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003719801800491543,
      "loss": 4.947,
      "mean_token_accuracy": 0.22531927824020387,
      "num_tokens": 84453441.0,
      "step": 36870
    },
    {
      "entropy": 5.716620969772339,
      "epoch": 3.6452155001977067,
      "grad_norm": 1.234375,
      "learning_rate": 0.00037194833526089797,
      "loss": 4.9678,
      "mean_token_accuracy": 0.22396662533283235,
      "num_tokens": 84465290.0,
      "step": 36875
    },
    {
      "entropy": 5.80748610496521,
      "epoch": 3.645709766706208,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003719164880876975,
      "loss": 5.0044,
      "mean_token_accuracy": 0.22186266928911208,
      "num_tokens": 84476482.0,
      "step": 36880
    },
    {
      "entropy": 5.757922983169555,
      "epoch": 3.646204033214709,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003718846385303363,
      "loss": 4.8988,
      "mean_token_accuracy": 0.2323881357908249,
      "num_tokens": 84488641.0,
      "step": 36885
    },
    {
      "entropy": 5.704709720611572,
      "epoch": 3.646698299723211,
      "grad_norm": 1.25,
      "learning_rate": 0.0003718527865895981,
      "loss": 5.0056,
      "mean_token_accuracy": 0.22506581842899323,
      "num_tokens": 84501967.0,
      "step": 36890
    },
    {
      "entropy": 5.7624289989471436,
      "epoch": 3.647192566231712,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003718209322662662,
      "loss": 4.9682,
      "mean_token_accuracy": 0.21895689368247986,
      "num_tokens": 84513982.0,
      "step": 36895
    },
    {
      "entropy": 5.796509742736816,
      "epoch": 3.6476868327402134,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003717890755611244,
      "loss": 4.9274,
      "mean_token_accuracy": 0.22604838907718658,
      "num_tokens": 84523395.0,
      "step": 36900
    },
    {
      "entropy": 5.71971492767334,
      "epoch": 3.648181099248715,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037175721647495635,
      "loss": 4.9223,
      "mean_token_accuracy": 0.23434487134218215,
      "num_tokens": 84536019.0,
      "step": 36905
    },
    {
      "entropy": 5.719699096679688,
      "epoch": 3.6486753657572164,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003717253550085458,
      "loss": 4.9581,
      "mean_token_accuracy": 0.22978586405515672,
      "num_tokens": 84546882.0,
      "step": 36910
    },
    {
      "entropy": 5.787475728988648,
      "epoch": 3.6491696322657177,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037169349116267643,
      "loss": 4.9953,
      "mean_token_accuracy": 0.22336759120225907,
      "num_tokens": 84560195.0,
      "step": 36915
    },
    {
      "entropy": 5.794946718215942,
      "epoch": 3.649663898774219,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00037166162493813224,
      "loss": 5.0513,
      "mean_token_accuracy": 0.2212998613715172,
      "num_tokens": 84573165.0,
      "step": 36920
    },
    {
      "entropy": 5.738963603973389,
      "epoch": 3.65015816528272,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00037162975633569704,
      "loss": 4.8244,
      "mean_token_accuracy": 0.23829384297132492,
      "num_tokens": 84584871.0,
      "step": 36925
    },
    {
      "entropy": 5.790006494522094,
      "epoch": 3.650652431791222,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037159788535615486,
      "loss": 5.0284,
      "mean_token_accuracy": 0.21869076788425446,
      "num_tokens": 84596122.0,
      "step": 36930
    },
    {
      "entropy": 5.764120960235596,
      "epoch": 3.651146698299723,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037156601200028975,
      "loss": 4.9929,
      "mean_token_accuracy": 0.22429582178592683,
      "num_tokens": 84607432.0,
      "step": 36935
    },
    {
      "entropy": 5.732975912094116,
      "epoch": 3.651640964808225,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003715341362688857,
      "loss": 4.8822,
      "mean_token_accuracy": 0.22737666964530945,
      "num_tokens": 84618440.0,
      "step": 36940
    },
    {
      "entropy": 5.699876165390014,
      "epoch": 3.652135231316726,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003715022581627269,
      "loss": 4.8787,
      "mean_token_accuracy": 0.234268656373024,
      "num_tokens": 84629173.0,
      "step": 36945
    },
    {
      "entropy": 5.707715606689453,
      "epoch": 3.6526294978252274,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003714703776825976,
      "loss": 4.8548,
      "mean_token_accuracy": 0.22917697429656983,
      "num_tokens": 84639770.0,
      "step": 36950
    },
    {
      "entropy": 5.744900131225586,
      "epoch": 3.6531237643337287,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003714384948292821,
      "loss": 4.9788,
      "mean_token_accuracy": 0.22067993879318237,
      "num_tokens": 84651405.0,
      "step": 36955
    },
    {
      "entropy": 5.719801235198974,
      "epoch": 3.65361803084223,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003714066096035646,
      "loss": 4.8434,
      "mean_token_accuracy": 0.23817473649978638,
      "num_tokens": 84663136.0,
      "step": 36960
    },
    {
      "entropy": 5.731724977493286,
      "epoch": 3.6541122973507316,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003713747220062296,
      "loss": 4.8956,
      "mean_token_accuracy": 0.22873106449842454,
      "num_tokens": 84675432.0,
      "step": 36965
    },
    {
      "entropy": 5.712669610977173,
      "epoch": 3.654606563859233,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003713428320380614,
      "loss": 4.9072,
      "mean_token_accuracy": 0.22904399633407593,
      "num_tokens": 84686961.0,
      "step": 36970
    },
    {
      "entropy": 5.722235250473022,
      "epoch": 3.655100830367734,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003713109396998447,
      "loss": 4.8738,
      "mean_token_accuracy": 0.23667887449264527,
      "num_tokens": 84697316.0,
      "step": 36975
    },
    {
      "entropy": 5.758138847351074,
      "epoch": 3.655595096876236,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037127904499236386,
      "loss": 5.0446,
      "mean_token_accuracy": 0.21933169215917586,
      "num_tokens": 84709228.0,
      "step": 36980
    },
    {
      "entropy": 5.781071138381958,
      "epoch": 3.656089363384737,
      "grad_norm": 1.234375,
      "learning_rate": 0.00037124714791640363,
      "loss": 4.9407,
      "mean_token_accuracy": 0.22102712988853454,
      "num_tokens": 84721300.0,
      "step": 36985
    },
    {
      "entropy": 5.780516338348389,
      "epoch": 3.6565836298932384,
      "grad_norm": 1.296875,
      "learning_rate": 0.00037121524847274855,
      "loss": 4.9554,
      "mean_token_accuracy": 0.2265076071023941,
      "num_tokens": 84732264.0,
      "step": 36990
    },
    {
      "entropy": 5.7190204620361325,
      "epoch": 3.6570778964017396,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003711833466621835,
      "loss": 4.9215,
      "mean_token_accuracy": 0.22608231455087663,
      "num_tokens": 84743636.0,
      "step": 36995
    },
    {
      "entropy": 5.714638423919678,
      "epoch": 3.6575721629102413,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00037115144248549324,
      "loss": 4.927,
      "mean_token_accuracy": 0.22801250070333481,
      "num_tokens": 84754696.0,
      "step": 37000
    },
    {
      "entropy": 5.840717077255249,
      "epoch": 3.6580664294187426,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037111953594346256,
      "loss": 4.991,
      "mean_token_accuracy": 0.22281605303287505,
      "num_tokens": 84766026.0,
      "step": 37005
    },
    {
      "entropy": 5.833309268951416,
      "epoch": 3.658560695927244,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037108762703687634,
      "loss": 4.9702,
      "mean_token_accuracy": 0.2319195955991745,
      "num_tokens": 84777265.0,
      "step": 37010
    },
    {
      "entropy": 5.718849849700928,
      "epoch": 3.6590549624357456,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003710557157665196,
      "loss": 4.9278,
      "mean_token_accuracy": 0.22437967658042907,
      "num_tokens": 84787715.0,
      "step": 37015
    },
    {
      "entropy": 5.788674068450928,
      "epoch": 3.659549228944247,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003710238021331773,
      "loss": 5.0279,
      "mean_token_accuracy": 0.2233757957816124,
      "num_tokens": 84798874.0,
      "step": 37020
    },
    {
      "entropy": 5.79263858795166,
      "epoch": 3.660043495452748,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003709918861376346,
      "loss": 5.049,
      "mean_token_accuracy": 0.2145603269338608,
      "num_tokens": 84810072.0,
      "step": 37025
    },
    {
      "entropy": 5.81492805480957,
      "epoch": 3.6605377619612494,
      "grad_norm": 1.3125,
      "learning_rate": 0.00037095996778067664,
      "loss": 4.9497,
      "mean_token_accuracy": 0.22933973520994186,
      "num_tokens": 84820831.0,
      "step": 37030
    },
    {
      "entropy": 5.72659010887146,
      "epoch": 3.6610320284697506,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003709280470630886,
      "loss": 4.8383,
      "mean_token_accuracy": 0.23568262606859208,
      "num_tokens": 84832329.0,
      "step": 37035
    },
    {
      "entropy": 5.780866861343384,
      "epoch": 3.6615262949782523,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003708961239856557,
      "loss": 4.9859,
      "mean_token_accuracy": 0.2191890448331833,
      "num_tokens": 84844239.0,
      "step": 37040
    },
    {
      "entropy": 5.74135308265686,
      "epoch": 3.6620205614867536,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037086419854916327,
      "loss": 4.9537,
      "mean_token_accuracy": 0.22821117639541627,
      "num_tokens": 84855939.0,
      "step": 37045
    },
    {
      "entropy": 5.74976077079773,
      "epoch": 3.6625148279952553,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003708322707543967,
      "loss": 4.9398,
      "mean_token_accuracy": 0.22210075408220292,
      "num_tokens": 84867170.0,
      "step": 37050
    },
    {
      "entropy": 5.757513380050659,
      "epoch": 3.6630090945037566,
      "grad_norm": 1.375,
      "learning_rate": 0.00037080034060214134,
      "loss": 4.9644,
      "mean_token_accuracy": 0.22816435843706132,
      "num_tokens": 84877986.0,
      "step": 37055
    },
    {
      "entropy": 5.708131408691406,
      "epoch": 3.663503361012258,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003707684080931827,
      "loss": 4.8703,
      "mean_token_accuracy": 0.23605380207300186,
      "num_tokens": 84888795.0,
      "step": 37060
    },
    {
      "entropy": 5.696499824523926,
      "epoch": 3.663997627520759,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00037073647322830657,
      "loss": 4.9281,
      "mean_token_accuracy": 0.23496715873479843,
      "num_tokens": 84900569.0,
      "step": 37065
    },
    {
      "entropy": 5.76435604095459,
      "epoch": 3.6644918940292603,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003707045360082982,
      "loss": 5.0075,
      "mean_token_accuracy": 0.2246442422270775,
      "num_tokens": 84912018.0,
      "step": 37070
    },
    {
      "entropy": 5.836646318435669,
      "epoch": 3.664986160537762,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003706725964339434,
      "loss": 4.9977,
      "mean_token_accuracy": 0.21550026386976243,
      "num_tokens": 84923088.0,
      "step": 37075
    },
    {
      "entropy": 5.756434488296509,
      "epoch": 3.6654804270462633,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000370640654506028,
      "loss": 4.9261,
      "mean_token_accuracy": 0.23003214001655578,
      "num_tokens": 84933838.0,
      "step": 37080
    },
    {
      "entropy": 5.717120599746704,
      "epoch": 3.6659746935547646,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003706087102253375,
      "loss": 4.9039,
      "mean_token_accuracy": 0.22759612202644347,
      "num_tokens": 84945208.0,
      "step": 37085
    },
    {
      "entropy": 5.651730394363403,
      "epoch": 3.6664689600632663,
      "grad_norm": 1.109375,
      "learning_rate": 0.000370576763592658,
      "loss": 4.8715,
      "mean_token_accuracy": 0.2374713197350502,
      "num_tokens": 84956224.0,
      "step": 37090
    },
    {
      "entropy": 5.781054735183716,
      "epoch": 3.6669632265717675,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003705448146087752,
      "loss": 5.0098,
      "mean_token_accuracy": 0.2258610710501671,
      "num_tokens": 84967507.0,
      "step": 37095
    },
    {
      "entropy": 5.738318157196045,
      "epoch": 3.667457493080269,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037051286327447527,
      "loss": 4.9162,
      "mean_token_accuracy": 0.23021315634250641,
      "num_tokens": 84979578.0,
      "step": 37100
    },
    {
      "entropy": 5.843438291549683,
      "epoch": 3.66795175958877,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003704809095905439,
      "loss": 5.0936,
      "mean_token_accuracy": 0.21126451194286347,
      "num_tokens": 84990875.0,
      "step": 37105
    },
    {
      "entropy": 5.731909227371216,
      "epoch": 3.6684460260972718,
      "grad_norm": 1.34375,
      "learning_rate": 0.00037044895355776747,
      "loss": 4.8881,
      "mean_token_accuracy": 0.2333762988448143,
      "num_tokens": 85001787.0,
      "step": 37110
    },
    {
      "entropy": 5.774100637435913,
      "epoch": 3.668940292605773,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000370416995176932,
      "loss": 5.0295,
      "mean_token_accuracy": 0.2164655178785324,
      "num_tokens": 85015189.0,
      "step": 37115
    },
    {
      "entropy": 5.7636645317077635,
      "epoch": 3.6694345591142743,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003703850344488236,
      "loss": 4.9601,
      "mean_token_accuracy": 0.22757385969161986,
      "num_tokens": 85026267.0,
      "step": 37120
    },
    {
      "entropy": 5.72706937789917,
      "epoch": 3.669928825622776,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037035307137422855,
      "loss": 4.9252,
      "mean_token_accuracy": 0.23113259971141814,
      "num_tokens": 85038296.0,
      "step": 37125
    },
    {
      "entropy": 5.73038535118103,
      "epoch": 3.6704230921312773,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037032110595393313,
      "loss": 4.9701,
      "mean_token_accuracy": 0.2260370820760727,
      "num_tokens": 85050518.0,
      "step": 37130
    },
    {
      "entropy": 5.829587984085083,
      "epoch": 3.6709173586397785,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00037028913818872375,
      "loss": 4.9584,
      "mean_token_accuracy": 0.2186059445142746,
      "num_tokens": 85062002.0,
      "step": 37135
    },
    {
      "entropy": 5.807325029373169,
      "epoch": 3.67141162514828,
      "grad_norm": 1.265625,
      "learning_rate": 0.00037025716807938676,
      "loss": 4.9969,
      "mean_token_accuracy": 0.22650517225265504,
      "num_tokens": 85072818.0,
      "step": 37140
    },
    {
      "entropy": 5.677710151672363,
      "epoch": 3.6719058916567815,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003702251956267087,
      "loss": 4.8968,
      "mean_token_accuracy": 0.23530446141958236,
      "num_tokens": 85083747.0,
      "step": 37145
    },
    {
      "entropy": 5.734698104858398,
      "epoch": 3.6724001581652828,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000370193220831476,
      "loss": 4.9097,
      "mean_token_accuracy": 0.23229578137397766,
      "num_tokens": 85095546.0,
      "step": 37150
    },
    {
      "entropy": 5.779178285598755,
      "epoch": 3.672894424673784,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00037016124369447533,
      "loss": 4.9641,
      "mean_token_accuracy": 0.21975258886814117,
      "num_tokens": 85105891.0,
      "step": 37155
    },
    {
      "entropy": 5.797160243988037,
      "epoch": 3.6733886911822857,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003701292642164933,
      "loss": 4.9734,
      "mean_token_accuracy": 0.21880849003791808,
      "num_tokens": 85118378.0,
      "step": 37160
    },
    {
      "entropy": 5.797689008712768,
      "epoch": 3.673882957690787,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00037009728239831665,
      "loss": 5.023,
      "mean_token_accuracy": 0.21879465878009796,
      "num_tokens": 85130111.0,
      "step": 37165
    },
    {
      "entropy": 5.801438522338867,
      "epoch": 3.6743772241992882,
      "grad_norm": 1.375,
      "learning_rate": 0.0003700652982407321,
      "loss": 5.0113,
      "mean_token_accuracy": 0.21810525357723237,
      "num_tokens": 85140037.0,
      "step": 37170
    },
    {
      "entropy": 5.744135761260987,
      "epoch": 3.6748714907077895,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003700333117445264,
      "loss": 4.8982,
      "mean_token_accuracy": 0.23409973382949828,
      "num_tokens": 85152398.0,
      "step": 37175
    },
    {
      "entropy": 5.776797533035278,
      "epoch": 3.6753657572162908,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037000132291048655,
      "loss": 5.0128,
      "mean_token_accuracy": 0.21632560938596726,
      "num_tokens": 85164793.0,
      "step": 37180
    },
    {
      "entropy": 5.7823755741119385,
      "epoch": 3.6758600237247925,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00036996933173939935,
      "loss": 5.021,
      "mean_token_accuracy": 0.21646941155195237,
      "num_tokens": 85176291.0,
      "step": 37185
    },
    {
      "entropy": 5.745422744750977,
      "epoch": 3.6763542902332937,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003699373382320519,
      "loss": 4.9706,
      "mean_token_accuracy": 0.2210451766848564,
      "num_tokens": 85188706.0,
      "step": 37190
    },
    {
      "entropy": 5.805277013778687,
      "epoch": 3.6768485567417954,
      "grad_norm": 1.25,
      "learning_rate": 0.00036990534238923124,
      "loss": 5.0256,
      "mean_token_accuracy": 0.22470012158155442,
      "num_tokens": 85200771.0,
      "step": 37195
    },
    {
      "entropy": 5.7734715938568115,
      "epoch": 3.6773428232502967,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036987334421172434,
      "loss": 5.0088,
      "mean_token_accuracy": 0.22506896406412125,
      "num_tokens": 85211209.0,
      "step": 37200
    },
    {
      "entropy": 5.74662709236145,
      "epoch": 3.677837089758798,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036984134370031846,
      "loss": 4.9055,
      "mean_token_accuracy": 0.2316187709569931,
      "num_tokens": 85222370.0,
      "step": 37205
    },
    {
      "entropy": 5.730053043365478,
      "epoch": 3.6783313562672992,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003698093408558009,
      "loss": 4.8965,
      "mean_token_accuracy": 0.2215649291872978,
      "num_tokens": 85234210.0,
      "step": 37210
    },
    {
      "entropy": 5.779282426834106,
      "epoch": 3.6788256227758005,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003697773356789588,
      "loss": 4.958,
      "mean_token_accuracy": 0.23065546303987502,
      "num_tokens": 85245506.0,
      "step": 37215
    },
    {
      "entropy": 5.768721628189087,
      "epoch": 3.679319889284302,
      "grad_norm": 1.3125,
      "learning_rate": 0.00036974532817057946,
      "loss": 4.9315,
      "mean_token_accuracy": 0.2294473320245743,
      "num_tokens": 85256904.0,
      "step": 37220
    },
    {
      "entropy": 5.733954048156738,
      "epoch": 3.6798141557928035,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003697133183314505,
      "loss": 4.9462,
      "mean_token_accuracy": 0.2240203708410263,
      "num_tokens": 85268660.0,
      "step": 37225
    },
    {
      "entropy": 5.738315296173096,
      "epoch": 3.6803084223013047,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003696813061623591,
      "loss": 4.9875,
      "mean_token_accuracy": 0.22494334876537322,
      "num_tokens": 85280292.0,
      "step": 37230
    },
    {
      "entropy": 5.806431484222412,
      "epoch": 3.6808026888098064,
      "grad_norm": 1.265625,
      "learning_rate": 0.00036964929166409294,
      "loss": 4.9737,
      "mean_token_accuracy": 0.22737062424421312,
      "num_tokens": 85292029.0,
      "step": 37235
    },
    {
      "entropy": 5.704042959213257,
      "epoch": 3.6812969553183077,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036961727483743957,
      "loss": 4.8614,
      "mean_token_accuracy": 0.22255279272794723,
      "num_tokens": 85304350.0,
      "step": 37240
    },
    {
      "entropy": 5.72512583732605,
      "epoch": 3.681791221826809,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036958525568318645,
      "loss": 4.9615,
      "mean_token_accuracy": 0.2307597354054451,
      "num_tokens": 85315646.0,
      "step": 37245
    },
    {
      "entropy": 5.847604465484619,
      "epoch": 3.68228548833531,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003695532342021214,
      "loss": 4.9726,
      "mean_token_accuracy": 0.2253118634223938,
      "num_tokens": 85328779.0,
      "step": 37250
    },
    {
      "entropy": 5.723974180221558,
      "epoch": 3.682779754843812,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003695212103950321,
      "loss": 4.8827,
      "mean_token_accuracy": 0.2334430232644081,
      "num_tokens": 85339916.0,
      "step": 37255
    },
    {
      "entropy": 5.757748174667358,
      "epoch": 3.683274021352313,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00036948918426270646,
      "loss": 5.0362,
      "mean_token_accuracy": 0.2162937968969345,
      "num_tokens": 85351381.0,
      "step": 37260
    },
    {
      "entropy": 5.749101781845093,
      "epoch": 3.6837682878608144,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00036945715580593213,
      "loss": 4.984,
      "mean_token_accuracy": 0.21774595528841018,
      "num_tokens": 85362166.0,
      "step": 37265
    },
    {
      "entropy": 5.726912593841552,
      "epoch": 3.684262554369316,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003694251250254971,
      "loss": 4.8997,
      "mean_token_accuracy": 0.23004893958568573,
      "num_tokens": 85373626.0,
      "step": 37270
    },
    {
      "entropy": 5.781505012512207,
      "epoch": 3.6847568208778174,
      "grad_norm": 1.125,
      "learning_rate": 0.00036939309192218946,
      "loss": 4.9645,
      "mean_token_accuracy": 0.21542072147130967,
      "num_tokens": 85385217.0,
      "step": 37275
    },
    {
      "entropy": 5.7707232475280765,
      "epoch": 3.6852510873863187,
      "grad_norm": 1.375,
      "learning_rate": 0.000369361056496797,
      "loss": 4.9587,
      "mean_token_accuracy": 0.22747980803251266,
      "num_tokens": 85395944.0,
      "step": 37280
    },
    {
      "entropy": 5.7851881980896,
      "epoch": 3.68574535389482,
      "grad_norm": 1.296875,
      "learning_rate": 0.00036932901875010797,
      "loss": 4.9634,
      "mean_token_accuracy": 0.22452894002199172,
      "num_tokens": 85407614.0,
      "step": 37285
    },
    {
      "entropy": 5.737072992324829,
      "epoch": 3.686239620403321,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036929697868291045,
      "loss": 4.957,
      "mean_token_accuracy": 0.22738783657550812,
      "num_tokens": 85419006.0,
      "step": 37290
    },
    {
      "entropy": 5.7481636047363285,
      "epoch": 3.686733886911823,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00036926493629599257,
      "loss": 4.9246,
      "mean_token_accuracy": 0.23410102277994155,
      "num_tokens": 85430267.0,
      "step": 37295
    },
    {
      "entropy": 5.7388018608093265,
      "epoch": 3.687228153420324,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003692328915901427,
      "loss": 4.9552,
      "mean_token_accuracy": 0.22416764497756958,
      "num_tokens": 85441424.0,
      "step": 37300
    },
    {
      "entropy": 5.73291425704956,
      "epoch": 3.687722419928826,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003692008445661491,
      "loss": 4.8973,
      "mean_token_accuracy": 0.22834030240774156,
      "num_tokens": 85452606.0,
      "step": 37305
    },
    {
      "entropy": 5.732199954986572,
      "epoch": 3.688216686437327,
      "grad_norm": 1.203125,
      "learning_rate": 0.00036916879522480003,
      "loss": 4.9422,
      "mean_token_accuracy": 0.22722528576850892,
      "num_tokens": 85463532.0,
      "step": 37310
    },
    {
      "entropy": 5.793481922149658,
      "epoch": 3.6887109529458284,
      "grad_norm": 1.265625,
      "learning_rate": 0.000369136743566884,
      "loss": 4.9968,
      "mean_token_accuracy": 0.22339461296796798,
      "num_tokens": 85476014.0,
      "step": 37315
    },
    {
      "entropy": 5.74561619758606,
      "epoch": 3.6892052194543297,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003691046895931895,
      "loss": 4.9303,
      "mean_token_accuracy": 0.22855920493602752,
      "num_tokens": 85488296.0,
      "step": 37320
    },
    {
      "entropy": 5.806354331970215,
      "epoch": 3.689699485962831,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036907263330450505,
      "loss": 5.0219,
      "mean_token_accuracy": 0.2249247044324875,
      "num_tokens": 85498579.0,
      "step": 37325
    },
    {
      "entropy": 5.771112442016602,
      "epoch": 3.6901937524713326,
      "grad_norm": 1.234375,
      "learning_rate": 0.00036904057470161914,
      "loss": 4.9616,
      "mean_token_accuracy": 0.2243894636631012,
      "num_tokens": 85509978.0,
      "step": 37330
    },
    {
      "entropy": 5.785664653778076,
      "epoch": 3.690688018979834,
      "grad_norm": 1.28125,
      "learning_rate": 0.00036900851378532065,
      "loss": 4.9981,
      "mean_token_accuracy": 0.22123742550611497,
      "num_tokens": 85522277.0,
      "step": 37335
    },
    {
      "entropy": 5.715666723251343,
      "epoch": 3.691182285488335,
      "grad_norm": 1.203125,
      "learning_rate": 0.00036897645055639794,
      "loss": 4.9472,
      "mean_token_accuracy": 0.22633805721998215,
      "num_tokens": 85532880.0,
      "step": 37340
    },
    {
      "entropy": 5.6957906723022464,
      "epoch": 3.691676551996837,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003689443850156401,
      "loss": 4.8732,
      "mean_token_accuracy": 0.2336172491312027,
      "num_tokens": 85543714.0,
      "step": 37345
    },
    {
      "entropy": 5.703838682174682,
      "epoch": 3.692170818505338,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003689123171638358,
      "loss": 4.9103,
      "mean_token_accuracy": 0.23035956025123597,
      "num_tokens": 85555870.0,
      "step": 37350
    },
    {
      "entropy": 5.670854997634888,
      "epoch": 3.6926650850138394,
      "grad_norm": 1.171875,
      "learning_rate": 0.000368880247001774,
      "loss": 4.8497,
      "mean_token_accuracy": 0.23960686177015306,
      "num_tokens": 85567607.0,
      "step": 37355
    },
    {
      "entropy": 5.701727294921875,
      "epoch": 3.6931593515223406,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003688481745302435,
      "loss": 4.954,
      "mean_token_accuracy": 0.22836966216564178,
      "num_tokens": 85578448.0,
      "step": 37360
    },
    {
      "entropy": 5.794955587387085,
      "epoch": 3.6936536180308424,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003688160997500332,
      "loss": 5.0207,
      "mean_token_accuracy": 0.21871814280748367,
      "num_tokens": 85589220.0,
      "step": 37365
    },
    {
      "entropy": 5.816171312332154,
      "epoch": 3.6941478845393436,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00036878402266193243,
      "loss": 5.0986,
      "mean_token_accuracy": 0.21787215024232864,
      "num_tokens": 85600095.0,
      "step": 37370
    },
    {
      "entropy": 5.818463945388794,
      "epoch": 3.694642151047845,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036875194326673025,
      "loss": 4.9653,
      "mean_token_accuracy": 0.2266804203391075,
      "num_tokens": 85612188.0,
      "step": 37375
    },
    {
      "entropy": 5.719247579574585,
      "epoch": 3.6951364175563466,
      "grad_norm": 1.296875,
      "learning_rate": 0.00036871986156521554,
      "loss": 4.8576,
      "mean_token_accuracy": 0.23641509264707566,
      "num_tokens": 85623262.0,
      "step": 37380
    },
    {
      "entropy": 5.709928369522094,
      "epoch": 3.695630684064848,
      "grad_norm": 1.25,
      "learning_rate": 0.00036868777755817776,
      "loss": 4.9535,
      "mean_token_accuracy": 0.22405129224061965,
      "num_tokens": 85634811.0,
      "step": 37385
    },
    {
      "entropy": 5.8356575012207035,
      "epoch": 3.696124950573349,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003686556912464062,
      "loss": 5.0617,
      "mean_token_accuracy": 0.21919228434562682,
      "num_tokens": 85645739.0,
      "step": 37390
    },
    {
      "entropy": 5.785538005828857,
      "epoch": 3.6966192170818504,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00036862360263069,
      "loss": 4.9524,
      "mean_token_accuracy": 0.23166478425264359,
      "num_tokens": 85657124.0,
      "step": 37395
    },
    {
      "entropy": 5.829905939102173,
      "epoch": 3.697113483590352,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003685915117118187,
      "loss": 4.976,
      "mean_token_accuracy": 0.22730881720781326,
      "num_tokens": 85667750.0,
      "step": 37400
    },
    {
      "entropy": 5.701602792739868,
      "epoch": 3.6976077500988533,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003685594184905817,
      "loss": 4.9577,
      "mean_token_accuracy": 0.2274487942457199,
      "num_tokens": 85679009.0,
      "step": 37405
    },
    {
      "entropy": 5.757835435867309,
      "epoch": 3.6981020166073546,
      "grad_norm": 1.25,
      "learning_rate": 0.0003685273229677684,
      "loss": 4.9928,
      "mean_token_accuracy": 0.2267878010869026,
      "num_tokens": 85690925.0,
      "step": 37410
    },
    {
      "entropy": 5.872438430786133,
      "epoch": 3.6985962831158563,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00036849522514416854,
      "loss": 5.0809,
      "mean_token_accuracy": 0.21109531223773956,
      "num_tokens": 85702045.0,
      "step": 37415
    },
    {
      "entropy": 5.766263198852539,
      "epoch": 3.6990905496243576,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003684631250205716,
      "loss": 4.9627,
      "mean_token_accuracy": 0.2202484980225563,
      "num_tokens": 85713589.0,
      "step": 37420
    },
    {
      "entropy": 5.736242246627808,
      "epoch": 3.699584816132859,
      "grad_norm": 1.125,
      "learning_rate": 0.00036843102259776736,
      "loss": 4.9141,
      "mean_token_accuracy": 0.22963446080684663,
      "num_tokens": 85726060.0,
      "step": 37425
    },
    {
      "entropy": 5.89557147026062,
      "epoch": 3.70007908264136,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003683989178765454,
      "loss": 5.0684,
      "mean_token_accuracy": 0.21528128534555435,
      "num_tokens": 85737473.0,
      "step": 37430
    },
    {
      "entropy": 5.79092321395874,
      "epoch": 3.7005733491498614,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003683668108576955,
      "loss": 4.9794,
      "mean_token_accuracy": 0.22522441893815995,
      "num_tokens": 85749392.0,
      "step": 37435
    },
    {
      "entropy": 5.775051498413086,
      "epoch": 3.701067615658363,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036833470154200765,
      "loss": 4.9734,
      "mean_token_accuracy": 0.22117066085338594,
      "num_tokens": 85761180.0,
      "step": 37440
    },
    {
      "entropy": 5.689578104019165,
      "epoch": 3.7015618821668643,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003683025899302717,
      "loss": 4.9134,
      "mean_token_accuracy": 0.23618309944868088,
      "num_tokens": 85772230.0,
      "step": 37445
    },
    {
      "entropy": 5.766793251037598,
      "epoch": 3.702056148675366,
      "grad_norm": 1.296875,
      "learning_rate": 0.00036827047602327757,
      "loss": 5.0019,
      "mean_token_accuracy": 0.22699790745973586,
      "num_tokens": 85783413.0,
      "step": 37450
    },
    {
      "entropy": 5.808465003967285,
      "epoch": 3.7025504151838673,
      "grad_norm": 1.25,
      "learning_rate": 0.00036823835982181515,
      "loss": 4.9771,
      "mean_token_accuracy": 0.22568456083536148,
      "num_tokens": 85794057.0,
      "step": 37455
    },
    {
      "entropy": 5.775546598434448,
      "epoch": 3.7030446816923686,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036820624132667467,
      "loss": 4.8981,
      "mean_token_accuracy": 0.22219039499759674,
      "num_tokens": 85805379.0,
      "step": 37460
    },
    {
      "entropy": 5.796580457687378,
      "epoch": 3.70353894820087,
      "grad_norm": 1.25,
      "learning_rate": 0.00036817412053864626,
      "loss": 4.9577,
      "mean_token_accuracy": 0.22342815846204758,
      "num_tokens": 85817177.0,
      "step": 37465
    },
    {
      "entropy": 5.724299097061158,
      "epoch": 3.704033214709371,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036814199745851997,
      "loss": 4.939,
      "mean_token_accuracy": 0.23137965500354768,
      "num_tokens": 85830017.0,
      "step": 37470
    },
    {
      "entropy": 5.747087860107422,
      "epoch": 3.704527481217873,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00036810987208708607,
      "loss": 5.0443,
      "mean_token_accuracy": 0.21935572624206542,
      "num_tokens": 85840835.0,
      "step": 37475
    },
    {
      "entropy": 5.736013269424438,
      "epoch": 3.705021747726374,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036807774442513495,
      "loss": 4.9504,
      "mean_token_accuracy": 0.2279438003897667,
      "num_tokens": 85852492.0,
      "step": 37480
    },
    {
      "entropy": 5.753422451019287,
      "epoch": 3.7055160142348753,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003680456144734567,
      "loss": 4.9256,
      "mean_token_accuracy": 0.22800526171922683,
      "num_tokens": 85864570.0,
      "step": 37485
    },
    {
      "entropy": 5.745447826385498,
      "epoch": 3.706010280743377,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003680134822328421,
      "loss": 4.9378,
      "mean_token_accuracy": 0.22222978621721268,
      "num_tokens": 85875516.0,
      "step": 37490
    },
    {
      "entropy": 5.779710054397583,
      "epoch": 3.7065045472518783,
      "grad_norm": 1.125,
      "learning_rate": 0.0003679813477040814,
      "loss": 4.951,
      "mean_token_accuracy": 0.23086798340082168,
      "num_tokens": 85886051.0,
      "step": 37495
    },
    {
      "entropy": 5.797987794876098,
      "epoch": 3.7069988137603795,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000367949210887965,
      "loss": 5.0234,
      "mean_token_accuracy": 0.2208812028169632,
      "num_tokens": 85897996.0,
      "step": 37500
    },
    {
      "entropy": 5.7410407066345215,
      "epoch": 3.707493080268881,
      "grad_norm": 1.296875,
      "learning_rate": 0.00036791707178528355,
      "loss": 4.855,
      "mean_token_accuracy": 0.23410629630088806,
      "num_tokens": 85908562.0,
      "step": 37505
    },
    {
      "entropy": 5.802285385131836,
      "epoch": 3.7079873467773825,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003678849303968278,
      "loss": 4.9972,
      "mean_token_accuracy": 0.21652754098176957,
      "num_tokens": 85920859.0,
      "step": 37510
    },
    {
      "entropy": 5.6696741580963135,
      "epoch": 3.7084816132858838,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003678527867233884,
      "loss": 4.9356,
      "mean_token_accuracy": 0.22560035437345505,
      "num_tokens": 85932919.0,
      "step": 37515
    },
    {
      "entropy": 5.681406879425049,
      "epoch": 3.708975879794385,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036782064076575603,
      "loss": 4.8675,
      "mean_token_accuracy": 0.2346976190805435,
      "num_tokens": 85943399.0,
      "step": 37520
    },
    {
      "entropy": 5.776633596420288,
      "epoch": 3.7094701463028867,
      "grad_norm": 1.359375,
      "learning_rate": 0.00036778849252472144,
      "loss": 4.971,
      "mean_token_accuracy": 0.22241942584514618,
      "num_tokens": 85954110.0,
      "step": 37525
    },
    {
      "entropy": 5.804233169555664,
      "epoch": 3.709964412811388,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003677563420010755,
      "loss": 4.9651,
      "mean_token_accuracy": 0.22105648070573808,
      "num_tokens": 85965955.0,
      "step": 37530
    },
    {
      "entropy": 5.755979585647583,
      "epoch": 3.7104586793198893,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036772418919560937,
      "loss": 4.9104,
      "mean_token_accuracy": 0.2265198901295662,
      "num_tokens": 85977475.0,
      "step": 37535
    },
    {
      "entropy": 5.756081438064575,
      "epoch": 3.7109529458283905,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003676920341091136,
      "loss": 4.961,
      "mean_token_accuracy": 0.21792267709970475,
      "num_tokens": 85987995.0,
      "step": 37540
    },
    {
      "entropy": 5.789609479904175,
      "epoch": 3.711447212336892,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003676598767423795,
      "loss": 5.0449,
      "mean_token_accuracy": 0.21563535630702974,
      "num_tokens": 86000379.0,
      "step": 37545
    },
    {
      "entropy": 5.784220790863037,
      "epoch": 3.7119414788453935,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000367627717096198,
      "loss": 4.942,
      "mean_token_accuracy": 0.22381436675786973,
      "num_tokens": 86011959.0,
      "step": 37550
    },
    {
      "entropy": 5.792229747772216,
      "epoch": 3.7124357453538948,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003675955551713603,
      "loss": 4.9654,
      "mean_token_accuracy": 0.21988274455070494,
      "num_tokens": 86023015.0,
      "step": 37555
    },
    {
      "entropy": 5.703189134597778,
      "epoch": 3.7129300118623965,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003675633909686576,
      "loss": 4.9149,
      "mean_token_accuracy": 0.2233332008123398,
      "num_tokens": 86032802.0,
      "step": 37560
    },
    {
      "entropy": 5.772376632690429,
      "epoch": 3.7134242783708977,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036753122448888114,
      "loss": 4.9525,
      "mean_token_accuracy": 0.22448374480009078,
      "num_tokens": 86043499.0,
      "step": 37565
    },
    {
      "entropy": 5.813872766494751,
      "epoch": 3.713918544879399,
      "grad_norm": 1.203125,
      "learning_rate": 0.00036749905573282214,
      "loss": 4.9681,
      "mean_token_accuracy": 0.2173368901014328,
      "num_tokens": 86054327.0,
      "step": 37570
    },
    {
      "entropy": 5.835749864578247,
      "epoch": 3.7144128113879002,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036746688470127214,
      "loss": 5.0738,
      "mean_token_accuracy": 0.2144184723496437,
      "num_tokens": 86067023.0,
      "step": 37575
    },
    {
      "entropy": 5.753268623352051,
      "epoch": 3.7149070778964015,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036743471139502235,
      "loss": 4.944,
      "mean_token_accuracy": 0.2225894033908844,
      "num_tokens": 86079906.0,
      "step": 37580
    },
    {
      "entropy": 5.7769428253173825,
      "epoch": 3.715401344404903,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003674025358148643,
      "loss": 4.9809,
      "mean_token_accuracy": 0.22423101514577864,
      "num_tokens": 86090467.0,
      "step": 37585
    },
    {
      "entropy": 5.7302769184112545,
      "epoch": 3.7158956109134045,
      "grad_norm": 1.125,
      "learning_rate": 0.0003673703579615895,
      "loss": 4.9285,
      "mean_token_accuracy": 0.22742253243923188,
      "num_tokens": 86102908.0,
      "step": 37590
    },
    {
      "entropy": 5.667901563644409,
      "epoch": 3.7163898774219057,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00036733817783598973,
      "loss": 4.8432,
      "mean_token_accuracy": 0.23802122771739959,
      "num_tokens": 86114588.0,
      "step": 37595
    },
    {
      "entropy": 5.710298824310303,
      "epoch": 3.7168841439304074,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003673059954388563,
      "loss": 4.9711,
      "mean_token_accuracy": 0.22340784072875977,
      "num_tokens": 86126800.0,
      "step": 37600
    },
    {
      "entropy": 5.761493253707886,
      "epoch": 3.7173784104389087,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003672738107709811,
      "loss": 4.9836,
      "mean_token_accuracy": 0.22413140535354614,
      "num_tokens": 86138703.0,
      "step": 37605
    },
    {
      "entropy": 5.745538568496704,
      "epoch": 3.71787267694741,
      "grad_norm": 1.359375,
      "learning_rate": 0.00036724162383315584,
      "loss": 4.8832,
      "mean_token_accuracy": 0.22908671349287033,
      "num_tokens": 86148856.0,
      "step": 37610
    },
    {
      "entropy": 5.8843542575836185,
      "epoch": 3.7183669434559112,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00036720943462617236,
      "loss": 5.0125,
      "mean_token_accuracy": 0.22095763683319092,
      "num_tokens": 86160108.0,
      "step": 37615
    },
    {
      "entropy": 5.843805122375488,
      "epoch": 3.718861209964413,
      "grad_norm": 1.25,
      "learning_rate": 0.00036717724315082245,
      "loss": 4.9744,
      "mean_token_accuracy": 0.2241460680961609,
      "num_tokens": 86170065.0,
      "step": 37620
    },
    {
      "entropy": 5.768634939193726,
      "epoch": 3.719355476472914,
      "grad_norm": 1.234375,
      "learning_rate": 0.000367145049407898,
      "loss": 4.9579,
      "mean_token_accuracy": 0.22481756508350373,
      "num_tokens": 86180469.0,
      "step": 37625
    },
    {
      "entropy": 5.788803768157959,
      "epoch": 3.7198497429814155,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00036711285339819115,
      "loss": 5.05,
      "mean_token_accuracy": 0.2188575014472008,
      "num_tokens": 86191801.0,
      "step": 37630
    },
    {
      "entropy": 5.731822204589844,
      "epoch": 3.720344009489917,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003670806551224938,
      "loss": 4.9155,
      "mean_token_accuracy": 0.22321949005126954,
      "num_tokens": 86202967.0,
      "step": 37635
    },
    {
      "entropy": 5.811128568649292,
      "epoch": 3.7208382759984184,
      "grad_norm": 1.265625,
      "learning_rate": 0.00036704845458159803,
      "loss": 5.0458,
      "mean_token_accuracy": 0.21628097891807557,
      "num_tokens": 86215225.0,
      "step": 37640
    },
    {
      "entropy": 5.8112287521362305,
      "epoch": 3.7213325425069197,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00036701625177629595,
      "loss": 5.0336,
      "mean_token_accuracy": 0.21719338595867158,
      "num_tokens": 86227417.0,
      "step": 37645
    },
    {
      "entropy": 5.753141736984253,
      "epoch": 3.721826809015421,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003669840467073798,
      "loss": 4.9266,
      "mean_token_accuracy": 0.23020173460245133,
      "num_tokens": 86238749.0,
      "step": 37650
    },
    {
      "entropy": 5.728374290466308,
      "epoch": 3.7223210755239227,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003669518393756419,
      "loss": 5.0105,
      "mean_token_accuracy": 0.2173576608300209,
      "num_tokens": 86250146.0,
      "step": 37655
    },
    {
      "entropy": 5.766939640045166,
      "epoch": 3.722815342032424,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003669196297818745,
      "loss": 5.0126,
      "mean_token_accuracy": 0.2197690263390541,
      "num_tokens": 86262086.0,
      "step": 37660
    },
    {
      "entropy": 5.8283168315887455,
      "epoch": 3.723309608540925,
      "grad_norm": 1.375,
      "learning_rate": 0.00036688741792686985,
      "loss": 5.0147,
      "mean_token_accuracy": 0.22550718635320663,
      "num_tokens": 86273937.0,
      "step": 37665
    },
    {
      "entropy": 5.795480680465698,
      "epoch": 3.723803875049427,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003668552038114205,
      "loss": 4.9872,
      "mean_token_accuracy": 0.22517864406108856,
      "num_tokens": 86284451.0,
      "step": 37670
    },
    {
      "entropy": 5.779292345046997,
      "epoch": 3.724298141557928,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003668229874363189,
      "loss": 4.9447,
      "mean_token_accuracy": 0.22701106071472169,
      "num_tokens": 86296406.0,
      "step": 37675
    },
    {
      "entropy": 5.732065296173095,
      "epoch": 3.7247924080664294,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036679076880235764,
      "loss": 4.9052,
      "mean_token_accuracy": 0.23302721381187438,
      "num_tokens": 86306922.0,
      "step": 37680
    },
    {
      "entropy": 5.763846015930175,
      "epoch": 3.7252866745749307,
      "grad_norm": 1.265625,
      "learning_rate": 0.00036675854791032916,
      "loss": 4.9086,
      "mean_token_accuracy": 0.23191508948802947,
      "num_tokens": 86317785.0,
      "step": 37685
    },
    {
      "entropy": 5.851247930526734,
      "epoch": 3.725780941083432,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003667263247610261,
      "loss": 5.0655,
      "mean_token_accuracy": 0.21622992157936097,
      "num_tokens": 86329271.0,
      "step": 37690
    },
    {
      "entropy": 5.810261297225952,
      "epoch": 3.7262752075919336,
      "grad_norm": 1.28125,
      "learning_rate": 0.00036669409935524136,
      "loss": 5.0119,
      "mean_token_accuracy": 0.22189213335514069,
      "num_tokens": 86339739.0,
      "step": 37695
    },
    {
      "entropy": 5.699525785446167,
      "epoch": 3.726769474100435,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003666618716937674,
      "loss": 4.8984,
      "mean_token_accuracy": 0.23122749477624893,
      "num_tokens": 86351111.0,
      "step": 37700
    },
    {
      "entropy": 5.7642998695373535,
      "epoch": 3.7272637406089366,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003666296417773973,
      "loss": 5.0055,
      "mean_token_accuracy": 0.22127450853586197,
      "num_tokens": 86363320.0,
      "step": 37705
    },
    {
      "entropy": 5.798371458053589,
      "epoch": 3.727758007117438,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003665974096069237,
      "loss": 5.0125,
      "mean_token_accuracy": 0.22296567112207413,
      "num_tokens": 86374372.0,
      "step": 37710
    },
    {
      "entropy": 5.887238025665283,
      "epoch": 3.728252273625939,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003665651751831397,
      "loss": 5.0255,
      "mean_token_accuracy": 0.21387027949094772,
      "num_tokens": 86385545.0,
      "step": 37715
    },
    {
      "entropy": 5.750654602050782,
      "epoch": 3.7287465401344404,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036653293850683803,
      "loss": 4.9704,
      "mean_token_accuracy": 0.22632201313972472,
      "num_tokens": 86398240.0,
      "step": 37720
    },
    {
      "entropy": 5.733729219436645,
      "epoch": 3.7292408066429417,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003665006995788121,
      "loss": 4.9587,
      "mean_token_accuracy": 0.22277727723121643,
      "num_tokens": 86409024.0,
      "step": 37725
    },
    {
      "entropy": 5.835492563247681,
      "epoch": 3.7297350731514434,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036646845839985464,
      "loss": 4.9865,
      "mean_token_accuracy": 0.22047607600688934,
      "num_tokens": 86422114.0,
      "step": 37730
    },
    {
      "entropy": 5.817431545257568,
      "epoch": 3.7302293396599446,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036643621497075895,
      "loss": 5.0221,
      "mean_token_accuracy": 0.21398374140262605,
      "num_tokens": 86435230.0,
      "step": 37735
    },
    {
      "entropy": 5.754470586776733,
      "epoch": 3.730723606168446,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036640396929231813,
      "loss": 5.0548,
      "mean_token_accuracy": 0.219473434984684,
      "num_tokens": 86446449.0,
      "step": 37740
    },
    {
      "entropy": 5.789600801467896,
      "epoch": 3.7312178726769476,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003663717213653255,
      "loss": 4.975,
      "mean_token_accuracy": 0.22250833809375764,
      "num_tokens": 86456814.0,
      "step": 37745
    },
    {
      "entropy": 5.778157520294189,
      "epoch": 3.731712139185449,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003663394711905744,
      "loss": 4.9234,
      "mean_token_accuracy": 0.22234784960746765,
      "num_tokens": 86468135.0,
      "step": 37750
    },
    {
      "entropy": 5.81230206489563,
      "epoch": 3.73220640569395,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00036630721876885813,
      "loss": 5.0415,
      "mean_token_accuracy": 0.22250140458345413,
      "num_tokens": 86480268.0,
      "step": 37755
    },
    {
      "entropy": 5.73613224029541,
      "epoch": 3.7327006722024514,
      "grad_norm": 1.234375,
      "learning_rate": 0.00036627496410097005,
      "loss": 4.881,
      "mean_token_accuracy": 0.2334372490644455,
      "num_tokens": 86492660.0,
      "step": 37760
    },
    {
      "entropy": 5.755557250976563,
      "epoch": 3.733194938710953,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036624270718770376,
      "loss": 5.0016,
      "mean_token_accuracy": 0.2249850794672966,
      "num_tokens": 86504638.0,
      "step": 37765
    },
    {
      "entropy": 5.748024559020996,
      "epoch": 3.7336892052194544,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003662104480298527,
      "loss": 4.8957,
      "mean_token_accuracy": 0.2314486101269722,
      "num_tokens": 86514225.0,
      "step": 37770
    },
    {
      "entropy": 5.7848869323730465,
      "epoch": 3.7341834717279556,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003661781866282104,
      "loss": 4.981,
      "mean_token_accuracy": 0.22631637156009674,
      "num_tokens": 86525370.0,
      "step": 37775
    },
    {
      "entropy": 5.772403335571289,
      "epoch": 3.7346777382364573,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003661459229835707,
      "loss": 4.898,
      "mean_token_accuracy": 0.22576549053192138,
      "num_tokens": 86535137.0,
      "step": 37780
    },
    {
      "entropy": 5.67782039642334,
      "epoch": 3.7351720047449586,
      "grad_norm": 1.359375,
      "learning_rate": 0.00036611365709672707,
      "loss": 4.851,
      "mean_token_accuracy": 0.23626085072755815,
      "num_tokens": 86546922.0,
      "step": 37785
    },
    {
      "entropy": 5.802244758605957,
      "epoch": 3.73566627125346,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003660813889684733,
      "loss": 5.0252,
      "mean_token_accuracy": 0.21744414567947387,
      "num_tokens": 86559356.0,
      "step": 37790
    },
    {
      "entropy": 5.720613479614258,
      "epoch": 3.736160537761961,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003660491185996033,
      "loss": 4.8409,
      "mean_token_accuracy": 0.2380582332611084,
      "num_tokens": 86571515.0,
      "step": 37795
    },
    {
      "entropy": 5.796172380447388,
      "epoch": 3.7366548042704624,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003660168459909109,
      "loss": 4.9449,
      "mean_token_accuracy": 0.22473030984401704,
      "num_tokens": 86581641.0,
      "step": 37800
    },
    {
      "entropy": 5.73213529586792,
      "epoch": 3.737149070778964,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036598457114318984,
      "loss": 4.9443,
      "mean_token_accuracy": 0.22271333038806915,
      "num_tokens": 86594478.0,
      "step": 37805
    },
    {
      "entropy": 5.749959707260132,
      "epoch": 3.7376433372874653,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003659522940572343,
      "loss": 4.9176,
      "mean_token_accuracy": 0.22743833512067796,
      "num_tokens": 86605496.0,
      "step": 37810
    },
    {
      "entropy": 5.757935667037964,
      "epoch": 3.738137603795967,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036592001473383817,
      "loss": 4.9123,
      "mean_token_accuracy": 0.2287479966878891,
      "num_tokens": 86616745.0,
      "step": 37815
    },
    {
      "entropy": 5.793238306045533,
      "epoch": 3.7386318703044683,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003658877331737956,
      "loss": 4.996,
      "mean_token_accuracy": 0.21782800704240798,
      "num_tokens": 86628094.0,
      "step": 37820
    },
    {
      "entropy": 5.780686664581299,
      "epoch": 3.7391261368129696,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036585544937790067,
      "loss": 4.9489,
      "mean_token_accuracy": 0.23064619600772857,
      "num_tokens": 86639421.0,
      "step": 37825
    },
    {
      "entropy": 5.705244827270508,
      "epoch": 3.739620403321471,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00036582316334694766,
      "loss": 4.9208,
      "mean_token_accuracy": 0.22962794750928878,
      "num_tokens": 86650653.0,
      "step": 37830
    },
    {
      "entropy": 5.815390491485596,
      "epoch": 3.740114669829972,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003657908750817307,
      "loss": 4.937,
      "mean_token_accuracy": 0.22738177329301834,
      "num_tokens": 86660954.0,
      "step": 37835
    },
    {
      "entropy": 5.821546316146851,
      "epoch": 3.740608936338474,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036575858458304407,
      "loss": 5.0412,
      "mean_token_accuracy": 0.2155962720513344,
      "num_tokens": 86671983.0,
      "step": 37840
    },
    {
      "entropy": 5.72502179145813,
      "epoch": 3.741103202846975,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00036572629185168223,
      "loss": 4.872,
      "mean_token_accuracy": 0.23590594679117202,
      "num_tokens": 86682634.0,
      "step": 37845
    },
    {
      "entropy": 5.733956623077392,
      "epoch": 3.7415974693554763,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003656939968884396,
      "loss": 4.9969,
      "mean_token_accuracy": 0.22023146599531174,
      "num_tokens": 86693516.0,
      "step": 37850
    },
    {
      "entropy": 5.79281964302063,
      "epoch": 3.742091735863978,
      "grad_norm": 1.265625,
      "learning_rate": 0.00036566169969411047,
      "loss": 5.0152,
      "mean_token_accuracy": 0.21973765939474105,
      "num_tokens": 86705350.0,
      "step": 37855
    },
    {
      "entropy": 5.77914662361145,
      "epoch": 3.7425860023724793,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003656294002694895,
      "loss": 4.9348,
      "mean_token_accuracy": 0.22151776701211928,
      "num_tokens": 86715647.0,
      "step": 37860
    },
    {
      "entropy": 5.790775299072266,
      "epoch": 3.7430802688809806,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003655970986153712,
      "loss": 5.0592,
      "mean_token_accuracy": 0.21426394134759902,
      "num_tokens": 86726453.0,
      "step": 37865
    },
    {
      "entropy": 5.756238412857056,
      "epoch": 3.743574535389482,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003655647947325503,
      "loss": 4.9745,
      "mean_token_accuracy": 0.22652319073677063,
      "num_tokens": 86737400.0,
      "step": 37870
    },
    {
      "entropy": 5.865530109405517,
      "epoch": 3.7440688018979835,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003655324886218213,
      "loss": 5.0704,
      "mean_token_accuracy": 0.21921167224645616,
      "num_tokens": 86748667.0,
      "step": 37875
    },
    {
      "entropy": 5.748724794387817,
      "epoch": 3.744563068406485,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00036550018028397915,
      "loss": 4.9524,
      "mean_token_accuracy": 0.2274917781352997,
      "num_tokens": 86761081.0,
      "step": 37880
    },
    {
      "entropy": 5.811793279647827,
      "epoch": 3.745057334914986,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003654678697198184,
      "loss": 5.0224,
      "mean_token_accuracy": 0.2225654751062393,
      "num_tokens": 86770823.0,
      "step": 37885
    },
    {
      "entropy": 5.747380065917969,
      "epoch": 3.7455516014234878,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036543555693013406,
      "loss": 4.9805,
      "mean_token_accuracy": 0.22445109486579895,
      "num_tokens": 86784362.0,
      "step": 37890
    },
    {
      "entropy": 5.808714962005615,
      "epoch": 3.746045867931989,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003654032419157211,
      "loss": 5.0162,
      "mean_token_accuracy": 0.22368196696043013,
      "num_tokens": 86796679.0,
      "step": 37895
    },
    {
      "entropy": 5.780781984329224,
      "epoch": 3.7465401344404903,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003653709246773743,
      "loss": 5.0074,
      "mean_token_accuracy": 0.2254186525940895,
      "num_tokens": 86808272.0,
      "step": 37900
    },
    {
      "entropy": 5.795290946960449,
      "epoch": 3.7470344009489915,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003653386052158886,
      "loss": 4.9893,
      "mean_token_accuracy": 0.22253507375717163,
      "num_tokens": 86819428.0,
      "step": 37905
    },
    {
      "entropy": 5.7553528308868405,
      "epoch": 3.7475286674574932,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003653062835320593,
      "loss": 4.9781,
      "mean_token_accuracy": 0.21847597360610962,
      "num_tokens": 86831373.0,
      "step": 37910
    },
    {
      "entropy": 5.715044641494751,
      "epoch": 3.7480229339659945,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00036527395962668145,
      "loss": 4.8738,
      "mean_token_accuracy": 0.22778510451316833,
      "num_tokens": 86843196.0,
      "step": 37915
    },
    {
      "entropy": 5.7884166717529295,
      "epoch": 3.7485172004744958,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003652416335005502,
      "loss": 4.9839,
      "mean_token_accuracy": 0.2222740486264229,
      "num_tokens": 86855233.0,
      "step": 37920
    },
    {
      "entropy": 5.7970046043396,
      "epoch": 3.7490114669829975,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003652093051544607,
      "loss": 4.9249,
      "mean_token_accuracy": 0.2237451121211052,
      "num_tokens": 86865994.0,
      "step": 37925
    },
    {
      "entropy": 5.758708906173706,
      "epoch": 3.7495057334914987,
      "grad_norm": 1.265625,
      "learning_rate": 0.00036517697458920837,
      "loss": 4.9117,
      "mean_token_accuracy": 0.22558679729700087,
      "num_tokens": 86877451.0,
      "step": 37930
    },
    {
      "entropy": 5.694994068145752,
      "epoch": 3.75,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00036514464180558836,
      "loss": 4.9436,
      "mean_token_accuracy": 0.22306866496801375,
      "num_tokens": 86888390.0,
      "step": 37935
    },
    {
      "entropy": 5.7700052738189695,
      "epoch": 3.7504942665085013,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003651123068043963,
      "loss": 5.0286,
      "mean_token_accuracy": 0.22130186706781388,
      "num_tokens": 86899703.0,
      "step": 37940
    },
    {
      "entropy": 5.784519529342651,
      "epoch": 3.7509885330170025,
      "grad_norm": 1.125,
      "learning_rate": 0.00036507996958642746,
      "loss": 4.9668,
      "mean_token_accuracy": 0.22671309858560562,
      "num_tokens": 86911753.0,
      "step": 37945
    },
    {
      "entropy": 5.921967601776123,
      "epoch": 3.7514827995255042,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003650476301524773,
      "loss": 5.1182,
      "mean_token_accuracy": 0.21818505376577377,
      "num_tokens": 86924165.0,
      "step": 37950
    },
    {
      "entropy": 5.79480881690979,
      "epoch": 3.7519770660340055,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003650152885033416,
      "loss": 5.0231,
      "mean_token_accuracy": 0.21883623749017717,
      "num_tokens": 86935848.0,
      "step": 37955
    },
    {
      "entropy": 5.7117616653442385,
      "epoch": 3.7524713325425068,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036498294463981583,
      "loss": 4.9056,
      "mean_token_accuracy": 0.22507920414209365,
      "num_tokens": 86946816.0,
      "step": 37960
    },
    {
      "entropy": 5.78493103981018,
      "epoch": 3.7529655990510085,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036495059856269556,
      "loss": 4.999,
      "mean_token_accuracy": 0.22427811175584794,
      "num_tokens": 86958436.0,
      "step": 37965
    },
    {
      "entropy": 5.8226512432098385,
      "epoch": 3.7534598655595097,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003649182502727768,
      "loss": 4.9472,
      "mean_token_accuracy": 0.21884147077798843,
      "num_tokens": 86970338.0,
      "step": 37970
    },
    {
      "entropy": 5.773404979705811,
      "epoch": 3.753954132068011,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036488589977085495,
      "loss": 4.9471,
      "mean_token_accuracy": 0.2224264085292816,
      "num_tokens": 86981796.0,
      "step": 37975
    },
    {
      "entropy": 5.657529354095459,
      "epoch": 3.7544483985765122,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00036485354705772605,
      "loss": 4.9022,
      "mean_token_accuracy": 0.23338191360235214,
      "num_tokens": 86993681.0,
      "step": 37980
    },
    {
      "entropy": 5.798483562469483,
      "epoch": 3.754942665085014,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00036482119213418604,
      "loss": 5.098,
      "mean_token_accuracy": 0.21738347709178923,
      "num_tokens": 87005621.0,
      "step": 37985
    },
    {
      "entropy": 5.764780187606812,
      "epoch": 3.755436931593515,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036478883500103067,
      "loss": 4.9732,
      "mean_token_accuracy": 0.23007028102874755,
      "num_tokens": 87016054.0,
      "step": 37990
    },
    {
      "entropy": 5.831835222244263,
      "epoch": 3.7559311981020165,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003647564756590561,
      "loss": 4.964,
      "mean_token_accuracy": 0.22115644663572312,
      "num_tokens": 87026603.0,
      "step": 37995
    },
    {
      "entropy": 5.789140844345093,
      "epoch": 3.756425464610518,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003647241141090583,
      "loss": 4.9842,
      "mean_token_accuracy": 0.22459333389997482,
      "num_tokens": 87037036.0,
      "step": 38000
    },
    {
      "entropy": 5.695921468734741,
      "epoch": 3.7569197311190194,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003646917503518333,
      "loss": 4.9009,
      "mean_token_accuracy": 0.2345951572060585,
      "num_tokens": 87047360.0,
      "step": 38005
    },
    {
      "entropy": 5.675279569625855,
      "epoch": 3.7574139976275207,
      "grad_norm": 1.234375,
      "learning_rate": 0.00036465938438817735,
      "loss": 4.8977,
      "mean_token_accuracy": 0.23151865601539612,
      "num_tokens": 87058580.0,
      "step": 38010
    },
    {
      "entropy": 5.724679088592529,
      "epoch": 3.757908264136022,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003646270162188867,
      "loss": 5.001,
      "mean_token_accuracy": 0.22494527846574783,
      "num_tokens": 87070176.0,
      "step": 38015
    },
    {
      "entropy": 5.768527460098267,
      "epoch": 3.7584025306445237,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003645946458447575,
      "loss": 4.9804,
      "mean_token_accuracy": 0.22818768322467803,
      "num_tokens": 87082188.0,
      "step": 38020
    },
    {
      "entropy": 5.807416391372681,
      "epoch": 3.758896797153025,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003645622732665862,
      "loss": 4.9915,
      "mean_token_accuracy": 0.22055274695158006,
      "num_tokens": 87093784.0,
      "step": 38025
    },
    {
      "entropy": 5.748805332183838,
      "epoch": 3.759391063661526,
      "grad_norm": 1.328125,
      "learning_rate": 0.00036452989848516896,
      "loss": 4.9232,
      "mean_token_accuracy": 0.22302961945533753,
      "num_tokens": 87105102.0,
      "step": 38030
    },
    {
      "entropy": 5.703180599212646,
      "epoch": 3.759885330170028,
      "grad_norm": 1.1875,
      "learning_rate": 0.00036449752150130235,
      "loss": 4.901,
      "mean_token_accuracy": 0.23253726661205293,
      "num_tokens": 87115956.0,
      "step": 38035
    },
    {
      "entropy": 5.70514030456543,
      "epoch": 3.760379596678529,
      "grad_norm": 1.328125,
      "learning_rate": 0.00036446514231578287,
      "loss": 4.9139,
      "mean_token_accuracy": 0.22850992083549498,
      "num_tokens": 87126514.0,
      "step": 38040
    },
    {
      "entropy": 5.732977819442749,
      "epoch": 3.7608738631870304,
      "grad_norm": 1.203125,
      "learning_rate": 0.00036443276092940703,
      "loss": 4.9403,
      "mean_token_accuracy": 0.22415173649787903,
      "num_tokens": 87137249.0,
      "step": 38045
    },
    {
      "entropy": 5.780674314498901,
      "epoch": 3.7613681296955317,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036440037734297134,
      "loss": 5.0049,
      "mean_token_accuracy": 0.22554363161325455,
      "num_tokens": 87148593.0,
      "step": 38050
    },
    {
      "entropy": 5.813843727111816,
      "epoch": 3.761862396204033,
      "grad_norm": 1.234375,
      "learning_rate": 0.00036436799155727253,
      "loss": 4.993,
      "mean_token_accuracy": 0.2211841881275177,
      "num_tokens": 87160085.0,
      "step": 38055
    },
    {
      "entropy": 5.767389011383057,
      "epoch": 3.7623566627125347,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036433560357310727,
      "loss": 4.9323,
      "mean_token_accuracy": 0.22540901750326156,
      "num_tokens": 87171427.0,
      "step": 38060
    },
    {
      "entropy": 5.727397346496582,
      "epoch": 3.762850929221036,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003643032133912723,
      "loss": 4.9432,
      "mean_token_accuracy": 0.22716560661792756,
      "num_tokens": 87183073.0,
      "step": 38065
    },
    {
      "entropy": 5.77149019241333,
      "epoch": 3.7633451957295376,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003642708210125644,
      "loss": 4.9517,
      "mean_token_accuracy": 0.22665513008832933,
      "num_tokens": 87193199.0,
      "step": 38070
    },
    {
      "entropy": 5.829591274261475,
      "epoch": 3.763839462238039,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00036423842643778045,
      "loss": 4.9984,
      "mean_token_accuracy": 0.22349378019571303,
      "num_tokens": 87205052.0,
      "step": 38075
    },
    {
      "entropy": 5.781742095947266,
      "epoch": 3.76433372874654,
      "grad_norm": 1.296875,
      "learning_rate": 0.00036420602966771734,
      "loss": 5.044,
      "mean_token_accuracy": 0.21928689777851104,
      "num_tokens": 87217485.0,
      "step": 38080
    },
    {
      "entropy": 5.745099020004273,
      "epoch": 3.7648279952550414,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003641736307031721,
      "loss": 4.9294,
      "mean_token_accuracy": 0.2298099234700203,
      "num_tokens": 87228184.0,
      "step": 38085
    },
    {
      "entropy": 5.757774114608765,
      "epoch": 3.7653222617635427,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036414122954494177,
      "loss": 5.0109,
      "mean_token_accuracy": 0.2189857006072998,
      "num_tokens": 87240713.0,
      "step": 38090
    },
    {
      "entropy": 5.783568572998047,
      "epoch": 3.7658165282720444,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003641088261938233,
      "loss": 5.0223,
      "mean_token_accuracy": 0.21678822636604309,
      "num_tokens": 87252739.0,
      "step": 38095
    },
    {
      "entropy": 5.693881464004517,
      "epoch": 3.7663107947805456,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003640764206506138,
      "loss": 4.8399,
      "mean_token_accuracy": 0.24297914803028106,
      "num_tokens": 87263034.0,
      "step": 38100
    },
    {
      "entropy": 5.772025871276855,
      "epoch": 3.766805061289047,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003640440129161107,
      "loss": 4.9753,
      "mean_token_accuracy": 0.22777872234582902,
      "num_tokens": 87275384.0,
      "step": 38105
    },
    {
      "entropy": 5.781468772888184,
      "epoch": 3.7672993277975486,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00036401160299111095,
      "loss": 5.029,
      "mean_token_accuracy": 0.21649794578552245,
      "num_tokens": 87286568.0,
      "step": 38110
    },
    {
      "entropy": 5.744393968582154,
      "epoch": 3.76779359430605,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000363979190876412,
      "loss": 4.939,
      "mean_token_accuracy": 0.22868804782629013,
      "num_tokens": 87297972.0,
      "step": 38115
    },
    {
      "entropy": 5.837701463699341,
      "epoch": 3.768287860814551,
      "grad_norm": 1.3125,
      "learning_rate": 0.00036394677657281113,
      "loss": 4.9196,
      "mean_token_accuracy": 0.2304409921169281,
      "num_tokens": 87309137.0,
      "step": 38120
    },
    {
      "entropy": 5.75016975402832,
      "epoch": 3.7687821273230524,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036391436008110573,
      "loss": 4.9815,
      "mean_token_accuracy": 0.22741299569606782,
      "num_tokens": 87320097.0,
      "step": 38125
    },
    {
      "entropy": 5.710018873214722,
      "epoch": 3.769276393831554,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036388194140209336,
      "loss": 4.9501,
      "mean_token_accuracy": 0.23217149525880815,
      "num_tokens": 87332662.0,
      "step": 38130
    },
    {
      "entropy": 5.781949424743653,
      "epoch": 3.7697706603400554,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00036384952053657144,
      "loss": 4.9575,
      "mean_token_accuracy": 0.23605646342039108,
      "num_tokens": 87344118.0,
      "step": 38135
    },
    {
      "entropy": 5.705094909667968,
      "epoch": 3.7702649268485566,
      "grad_norm": 1.265625,
      "learning_rate": 0.00036381709748533746,
      "loss": 4.9084,
      "mean_token_accuracy": 0.2272249221801758,
      "num_tokens": 87355787.0,
      "step": 38140
    },
    {
      "entropy": 5.758694648742676,
      "epoch": 3.7707591933570583,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003637846722491892,
      "loss": 4.9122,
      "mean_token_accuracy": 0.23372262418270112,
      "num_tokens": 87367332.0,
      "step": 38145
    },
    {
      "entropy": 5.729998254776001,
      "epoch": 3.7712534598655596,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003637522448289241,
      "loss": 4.9295,
      "mean_token_accuracy": 0.2275598853826523,
      "num_tokens": 87378968.0,
      "step": 38150
    },
    {
      "entropy": 5.716853523254395,
      "epoch": 3.771747726374061,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003637198152253401,
      "loss": 4.9604,
      "mean_token_accuracy": 0.22874368727207184,
      "num_tokens": 87391573.0,
      "step": 38155
    },
    {
      "entropy": 5.7780735969543455,
      "epoch": 3.772241992882562,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003636873834392349,
      "loss": 4.9824,
      "mean_token_accuracy": 0.22612094581127168,
      "num_tokens": 87402431.0,
      "step": 38160
    },
    {
      "entropy": 5.752438974380493,
      "epoch": 3.7727362593910634,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003636549494714062,
      "loss": 4.9978,
      "mean_token_accuracy": 0.22652032226324081,
      "num_tokens": 87414156.0,
      "step": 38165
    },
    {
      "entropy": 5.728734302520752,
      "epoch": 3.773230525899565,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036362251332265204,
      "loss": 4.9489,
      "mean_token_accuracy": 0.23094669133424758,
      "num_tokens": 87425878.0,
      "step": 38170
    },
    {
      "entropy": 5.770079612731934,
      "epoch": 3.7737247924080664,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003635900749937703,
      "loss": 4.9723,
      "mean_token_accuracy": 0.22962706387043,
      "num_tokens": 87437171.0,
      "step": 38175
    },
    {
      "entropy": 5.750476408004761,
      "epoch": 3.774219058916568,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000363557634485559,
      "loss": 4.9188,
      "mean_token_accuracy": 0.22449419796466827,
      "num_tokens": 87448465.0,
      "step": 38180
    },
    {
      "entropy": 5.678539705276489,
      "epoch": 3.7747133254250693,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003635251917988162,
      "loss": 4.9099,
      "mean_token_accuracy": 0.2294145107269287,
      "num_tokens": 87458800.0,
      "step": 38185
    },
    {
      "entropy": 5.7286755561828615,
      "epoch": 3.7752075919335706,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003634927469343399,
      "loss": 4.8951,
      "mean_token_accuracy": 0.2320461317896843,
      "num_tokens": 87470427.0,
      "step": 38190
    },
    {
      "entropy": 5.757998752593994,
      "epoch": 3.775701858442072,
      "grad_norm": 1.25,
      "learning_rate": 0.0003634602998929282,
      "loss": 4.9212,
      "mean_token_accuracy": 0.22941216826438904,
      "num_tokens": 87481498.0,
      "step": 38195
    },
    {
      "entropy": 5.689374017715454,
      "epoch": 3.776196124950573,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003634278506753795,
      "loss": 4.9064,
      "mean_token_accuracy": 0.23252497613430023,
      "num_tokens": 87493578.0,
      "step": 38200
    },
    {
      "entropy": 5.775686645507813,
      "epoch": 3.776690391459075,
      "grad_norm": 1.390625,
      "learning_rate": 0.000363395399282492,
      "loss": 5.0054,
      "mean_token_accuracy": 0.2134772151708603,
      "num_tokens": 87504387.0,
      "step": 38205
    },
    {
      "entropy": 5.794479084014893,
      "epoch": 3.777184657967576,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00036336294571506387,
      "loss": 4.9309,
      "mean_token_accuracy": 0.22212982773780823,
      "num_tokens": 87514231.0,
      "step": 38210
    },
    {
      "entropy": 5.733604097366333,
      "epoch": 3.7776789244760773,
      "grad_norm": 1.1875,
      "learning_rate": 0.00036333048997389363,
      "loss": 4.9844,
      "mean_token_accuracy": 0.2264952078461647,
      "num_tokens": 87526483.0,
      "step": 38215
    },
    {
      "entropy": 5.763343286514282,
      "epoch": 3.778173190984579,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003632980320597797,
      "loss": 4.9629,
      "mean_token_accuracy": 0.2179630234837532,
      "num_tokens": 87537931.0,
      "step": 38220
    },
    {
      "entropy": 5.769645500183105,
      "epoch": 3.7786674574930803,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003632655719735203,
      "loss": 4.9481,
      "mean_token_accuracy": 0.22772042602300643,
      "num_tokens": 87549486.0,
      "step": 38225
    },
    {
      "entropy": 5.766765451431274,
      "epoch": 3.7791617240015816,
      "grad_norm": 1.125,
      "learning_rate": 0.0003632331097159143,
      "loss": 5.0256,
      "mean_token_accuracy": 0.22252532243728637,
      "num_tokens": 87562232.0,
      "step": 38230
    },
    {
      "entropy": 5.784458971023559,
      "epoch": 3.779655990510083,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036320064528776,
      "loss": 4.9106,
      "mean_token_accuracy": 0.22461774051189423,
      "num_tokens": 87574246.0,
      "step": 38235
    },
    {
      "entropy": 5.793606376647949,
      "epoch": 3.7801502570185845,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003631681786898562,
      "loss": 4.9309,
      "mean_token_accuracy": 0.22601214945316314,
      "num_tokens": 87586033.0,
      "step": 38240
    },
    {
      "entropy": 5.753064060211182,
      "epoch": 3.780644523527086,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003631357099230015,
      "loss": 4.9567,
      "mean_token_accuracy": 0.2259281322360039,
      "num_tokens": 87596468.0,
      "step": 38245
    },
    {
      "entropy": 5.767170810699463,
      "epoch": 3.781138790035587,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003631032389879947,
      "loss": 4.9951,
      "mean_token_accuracy": 0.2206801012158394,
      "num_tokens": 87608577.0,
      "step": 38250
    },
    {
      "entropy": 5.7897227764129635,
      "epoch": 3.7816330565440888,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003630707658856345,
      "loss": 5.0124,
      "mean_token_accuracy": 0.2210958331823349,
      "num_tokens": 87620773.0,
      "step": 38255
    },
    {
      "entropy": 5.835418081283569,
      "epoch": 3.78212732305259,
      "grad_norm": 1.234375,
      "learning_rate": 0.00036303829061671986,
      "loss": 5.0092,
      "mean_token_accuracy": 0.21755923926830292,
      "num_tokens": 87633248.0,
      "step": 38260
    },
    {
      "entropy": 5.782951831817627,
      "epoch": 3.7826215895610913,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003630058131820495,
      "loss": 4.9815,
      "mean_token_accuracy": 0.22020128220319748,
      "num_tokens": 87644710.0,
      "step": 38265
    },
    {
      "entropy": 5.73781909942627,
      "epoch": 3.7831158560695926,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003629733335824226,
      "loss": 4.9211,
      "mean_token_accuracy": 0.22974741756916045,
      "num_tokens": 87655297.0,
      "step": 38270
    },
    {
      "entropy": 5.74844069480896,
      "epoch": 3.7836101225780943,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000362940851818638,
      "loss": 4.9277,
      "mean_token_accuracy": 0.22986202389001847,
      "num_tokens": 87667595.0,
      "step": 38275
    },
    {
      "entropy": 5.748209142684937,
      "epoch": 3.7841043890865955,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036290836789149483,
      "loss": 4.9381,
      "mean_token_accuracy": 0.22753172218799592,
      "num_tokens": 87680041.0,
      "step": 38280
    },
    {
      "entropy": 5.7980231761932375,
      "epoch": 3.784598655595097,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00036287588180179215,
      "loss": 5.0026,
      "mean_token_accuracy": 0.22103807032108308,
      "num_tokens": 87691287.0,
      "step": 38285
    },
    {
      "entropy": 5.740351676940918,
      "epoch": 3.7850929221035985,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003628433935503291,
      "loss": 4.9872,
      "mean_token_accuracy": 0.22585272043943405,
      "num_tokens": 87704224.0,
      "step": 38290
    },
    {
      "entropy": 5.748596239089966,
      "epoch": 3.7855871886120998,
      "grad_norm": 1.265625,
      "learning_rate": 0.000362810903137905,
      "loss": 4.9601,
      "mean_token_accuracy": 0.22943371087312697,
      "num_tokens": 87714449.0,
      "step": 38295
    },
    {
      "entropy": 5.772917366027832,
      "epoch": 3.786081455120601,
      "grad_norm": 1.234375,
      "learning_rate": 0.000362778410565319,
      "loss": 4.9636,
      "mean_token_accuracy": 0.22030484229326247,
      "num_tokens": 87725348.0,
      "step": 38300
    },
    {
      "entropy": 5.754444026947022,
      "epoch": 3.7865757216291023,
      "grad_norm": 1.296875,
      "learning_rate": 0.00036274591583337055,
      "loss": 4.9565,
      "mean_token_accuracy": 0.22712165415287017,
      "num_tokens": 87737244.0,
      "step": 38305
    },
    {
      "entropy": 5.721923208236694,
      "epoch": 3.7870699881376035,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036271341894285894,
      "loss": 4.9572,
      "mean_token_accuracy": 0.22210686802864074,
      "num_tokens": 87748710.0,
      "step": 38310
    },
    {
      "entropy": 5.763736867904663,
      "epoch": 3.7875642546461052,
      "grad_norm": 1.203125,
      "learning_rate": 0.00036268091989458357,
      "loss": 5.0401,
      "mean_token_accuracy": 0.22098899781703948,
      "num_tokens": 87759944.0,
      "step": 38315
    },
    {
      "entropy": 5.8263712406158445,
      "epoch": 3.7880585211546065,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003626484186893441,
      "loss": 4.9724,
      "mean_token_accuracy": 0.22450798004865646,
      "num_tokens": 87771201.0,
      "step": 38320
    },
    {
      "entropy": 5.77383246421814,
      "epoch": 3.788552787663108,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003626159153279398,
      "loss": 4.9258,
      "mean_token_accuracy": 0.23120176643133164,
      "num_tokens": 87782129.0,
      "step": 38325
    },
    {
      "entropy": 5.724788331985474,
      "epoch": 3.7890470541716095,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036258340981117044,
      "loss": 4.9173,
      "mean_token_accuracy": 0.23080307841300965,
      "num_tokens": 87792915.0,
      "step": 38330
    },
    {
      "entropy": 5.767858219146729,
      "epoch": 3.7895413206801107,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00036255090213983555,
      "loss": 4.9226,
      "mean_token_accuracy": 0.22401067167520522,
      "num_tokens": 87803301.0,
      "step": 38335
    },
    {
      "entropy": 5.778872966766357,
      "epoch": 3.790035587188612,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036251839231473494,
      "loss": 5.0187,
      "mean_token_accuracy": 0.2227788582444191,
      "num_tokens": 87815167.0,
      "step": 38340
    },
    {
      "entropy": 5.753599786758423,
      "epoch": 3.7905298536971133,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003624858803366684,
      "loss": 4.9782,
      "mean_token_accuracy": 0.22499951273202895,
      "num_tokens": 87827348.0,
      "step": 38345
    },
    {
      "entropy": 5.729551124572754,
      "epoch": 3.791024120205615,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003624533662064355,
      "loss": 4.9247,
      "mean_token_accuracy": 0.23431379050016404,
      "num_tokens": 87837986.0,
      "step": 38350
    },
    {
      "entropy": 5.749391555786133,
      "epoch": 3.7915183867141162,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003624208499248362,
      "loss": 4.87,
      "mean_token_accuracy": 0.2357859343290329,
      "num_tokens": 87847992.0,
      "step": 38355
    },
    {
      "entropy": 5.801260662078858,
      "epoch": 3.7920126532226175,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003623883314926705,
      "loss": 5.0907,
      "mean_token_accuracy": 0.21952947825193406,
      "num_tokens": 87859768.0,
      "step": 38360
    },
    {
      "entropy": 5.760098600387574,
      "epoch": 3.792506919731119,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003623558109107383,
      "loss": 4.9696,
      "mean_token_accuracy": 0.21976616531610488,
      "num_tokens": 87871366.0,
      "step": 38365
    },
    {
      "entropy": 5.747028779983521,
      "epoch": 3.7930011862396205,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003623232881798395,
      "loss": 4.9494,
      "mean_token_accuracy": 0.2308356985449791,
      "num_tokens": 87883074.0,
      "step": 38370
    },
    {
      "entropy": 5.720154762268066,
      "epoch": 3.7934954527481217,
      "grad_norm": 1.1875,
      "learning_rate": 0.00036229076330077445,
      "loss": 4.9406,
      "mean_token_accuracy": 0.22300800830125808,
      "num_tokens": 87894906.0,
      "step": 38375
    },
    {
      "entropy": 5.738593196868896,
      "epoch": 3.793989719256623,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003622582362743429,
      "loss": 4.8787,
      "mean_token_accuracy": 0.2324230194091797,
      "num_tokens": 87906027.0,
      "step": 38380
    },
    {
      "entropy": 5.753184843063354,
      "epoch": 3.7944839857651247,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003622257071013453,
      "loss": 4.9649,
      "mean_token_accuracy": 0.21730988025665282,
      "num_tokens": 87916348.0,
      "step": 38385
    },
    {
      "entropy": 5.750005197525025,
      "epoch": 3.794978252273626,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003621931757825817,
      "loss": 4.9657,
      "mean_token_accuracy": 0.22124162316322327,
      "num_tokens": 87928397.0,
      "step": 38390
    },
    {
      "entropy": 5.79720983505249,
      "epoch": 3.795472518782127,
      "grad_norm": 1.203125,
      "learning_rate": 0.00036216064231885255,
      "loss": 4.9773,
      "mean_token_accuracy": 0.21621413081884383,
      "num_tokens": 87939925.0,
      "step": 38395
    },
    {
      "entropy": 5.734981346130371,
      "epoch": 3.795966785290629,
      "grad_norm": 1.265625,
      "learning_rate": 0.00036212810671095797,
      "loss": 4.9737,
      "mean_token_accuracy": 0.22742882668972014,
      "num_tokens": 87951166.0,
      "step": 38400
    },
    {
      "entropy": 5.7364801406860355,
      "epoch": 3.79646105179913,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036209556895969847,
      "loss": 4.9053,
      "mean_token_accuracy": 0.23710869252681732,
      "num_tokens": 87962861.0,
      "step": 38405
    },
    {
      "entropy": 5.68361611366272,
      "epoch": 3.7969553183076314,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003620630290658745,
      "loss": 4.8251,
      "mean_token_accuracy": 0.23367105126380922,
      "num_tokens": 87974362.0,
      "step": 38410
    },
    {
      "entropy": 5.745436143875122,
      "epoch": 3.7974495848161327,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003620304870302865,
      "loss": 4.9645,
      "mean_token_accuracy": 0.229224357008934,
      "num_tokens": 87984982.0,
      "step": 38415
    },
    {
      "entropy": 5.763537883758545,
      "epoch": 3.797943851324634,
      "grad_norm": 1.1875,
      "learning_rate": 0.00036199794285373505,
      "loss": 4.9161,
      "mean_token_accuracy": 0.23079071044921876,
      "num_tokens": 87996980.0,
      "step": 38420
    },
    {
      "entropy": 5.776569604873657,
      "epoch": 3.7984381178331357,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00036196539653702073,
      "loss": 4.9324,
      "mean_token_accuracy": 0.22842410057783127,
      "num_tokens": 88008585.0,
      "step": 38425
    },
    {
      "entropy": 5.730450010299682,
      "epoch": 3.798932384341637,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000361932848080944,
      "loss": 4.9199,
      "mean_token_accuracy": 0.2352994441986084,
      "num_tokens": 88020779.0,
      "step": 38430
    },
    {
      "entropy": 5.795297908782959,
      "epoch": 3.7994266508501386,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003619002974863059,
      "loss": 5.0268,
      "mean_token_accuracy": 0.21689335256814957,
      "num_tokens": 88033003.0,
      "step": 38435
    },
    {
      "entropy": 5.740248918533325,
      "epoch": 3.79992091735864,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036186774475390686,
      "loss": 4.9308,
      "mean_token_accuracy": 0.22629230618476867,
      "num_tokens": 88044095.0,
      "step": 38440
    },
    {
      "entropy": 5.802647495269776,
      "epoch": 3.800415183867141,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00036183518988454794,
      "loss": 5.0129,
      "mean_token_accuracy": 0.22315700501203536,
      "num_tokens": 88056487.0,
      "step": 38445
    },
    {
      "entropy": 5.753944826126099,
      "epoch": 3.8009094503756424,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003618026328790299,
      "loss": 4.976,
      "mean_token_accuracy": 0.2270748272538185,
      "num_tokens": 88067803.0,
      "step": 38450
    },
    {
      "entropy": 5.832457113265991,
      "epoch": 3.8014037168841437,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003617700737381535,
      "loss": 4.988,
      "mean_token_accuracy": 0.2283177763223648,
      "num_tokens": 88078914.0,
      "step": 38455
    },
    {
      "entropy": 5.796089220046997,
      "epoch": 3.8018979833926454,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003617375124627199,
      "loss": 4.9574,
      "mean_token_accuracy": 0.2189305081963539,
      "num_tokens": 88090567.0,
      "step": 38460
    },
    {
      "entropy": 5.752261209487915,
      "epoch": 3.8023922499011467,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00036170494905352996,
      "loss": 4.9626,
      "mean_token_accuracy": 0.21865417659282685,
      "num_tokens": 88101171.0,
      "step": 38465
    },
    {
      "entropy": 5.767463302612304,
      "epoch": 3.802886516409648,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036167238351138487,
      "loss": 4.9874,
      "mean_token_accuracy": 0.22491028010845185,
      "num_tokens": 88111629.0,
      "step": 38470
    },
    {
      "entropy": 5.83456597328186,
      "epoch": 3.8033807829181496,
      "grad_norm": 1.328125,
      "learning_rate": 0.00036163981583708574,
      "loss": 4.9812,
      "mean_token_accuracy": 0.21840023547410964,
      "num_tokens": 88122354.0,
      "step": 38475
    },
    {
      "entropy": 5.745080280303955,
      "epoch": 3.803875049426651,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003616072460314336,
      "loss": 4.9317,
      "mean_token_accuracy": 0.22894297242164613,
      "num_tokens": 88134830.0,
      "step": 38480
    },
    {
      "entropy": 5.703588581085205,
      "epoch": 3.804369315935152,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00036157467409522986,
      "loss": 4.8779,
      "mean_token_accuracy": 0.22895611822605133,
      "num_tokens": 88145995.0,
      "step": 38485
    },
    {
      "entropy": 5.796066761016846,
      "epoch": 3.8048635824436534,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003615421000292757,
      "loss": 5.0057,
      "mean_token_accuracy": 0.2224449634552002,
      "num_tokens": 88157777.0,
      "step": 38490
    },
    {
      "entropy": 5.819619464874267,
      "epoch": 3.805357848952155,
      "grad_norm": 1.265625,
      "learning_rate": 0.00036150952383437246,
      "loss": 4.9997,
      "mean_token_accuracy": 0.2208433046936989,
      "num_tokens": 88169971.0,
      "step": 38495
    },
    {
      "entropy": 5.732924938201904,
      "epoch": 3.8058521154606564,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003614769455113215,
      "loss": 4.8882,
      "mean_token_accuracy": 0.2283529594540596,
      "num_tokens": 88181317.0,
      "step": 38500
    },
    {
      "entropy": 5.817422389984131,
      "epoch": 3.8063463819691576,
      "grad_norm": 1.125,
      "learning_rate": 0.0003614443650609242,
      "loss": 5.0768,
      "mean_token_accuracy": 0.212154258787632,
      "num_tokens": 88194603.0,
      "step": 38505
    },
    {
      "entropy": 5.792971849441528,
      "epoch": 3.8068406484776594,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003614117824839822,
      "loss": 5.0323,
      "mean_token_accuracy": 0.22379914820194244,
      "num_tokens": 88205775.0,
      "step": 38510
    },
    {
      "entropy": 5.754872179031372,
      "epoch": 3.8073349149861606,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003613791977812969,
      "loss": 4.974,
      "mean_token_accuracy": 0.2208371192216873,
      "num_tokens": 88216784.0,
      "step": 38515
    },
    {
      "entropy": 5.796415519714356,
      "epoch": 3.807829181494662,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00036134661095367,
      "loss": 4.9373,
      "mean_token_accuracy": 0.22468699514865875,
      "num_tokens": 88228861.0,
      "step": 38520
    },
    {
      "entropy": 5.7819616317749025,
      "epoch": 3.808323448003163,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036131402200190296,
      "loss": 4.9524,
      "mean_token_accuracy": 0.22730890810489654,
      "num_tokens": 88241134.0,
      "step": 38525
    },
    {
      "entropy": 5.715646600723266,
      "epoch": 3.808817714511665,
      "grad_norm": 1.3125,
      "learning_rate": 0.00036128143092679764,
      "loss": 4.9452,
      "mean_token_accuracy": 0.2244659036397934,
      "num_tokens": 88253295.0,
      "step": 38530
    },
    {
      "entropy": 5.795124435424805,
      "epoch": 3.809311981020166,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036124883772915573,
      "loss": 4.9785,
      "mean_token_accuracy": 0.22221399694681168,
      "num_tokens": 88265906.0,
      "step": 38535
    },
    {
      "entropy": 5.742788076400757,
      "epoch": 3.8098062475286674,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003612162424097791,
      "loss": 4.8699,
      "mean_token_accuracy": 0.23514126539230346,
      "num_tokens": 88276928.0,
      "step": 38540
    },
    {
      "entropy": 5.716719150543213,
      "epoch": 3.810300514037169,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00036118364496946943,
      "loss": 4.9135,
      "mean_token_accuracy": 0.22747474014759064,
      "num_tokens": 88288125.0,
      "step": 38545
    },
    {
      "entropy": 5.765071630477905,
      "epoch": 3.8107947805456703,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003611510454090288,
      "loss": 4.9568,
      "mean_token_accuracy": 0.2240847334265709,
      "num_tokens": 88298778.0,
      "step": 38550
    },
    {
      "entropy": 5.796245813369751,
      "epoch": 3.8112890470541716,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000361118443729259,
      "loss": 4.9566,
      "mean_token_accuracy": 0.2243811532855034,
      "num_tokens": 88309086.0,
      "step": 38555
    },
    {
      "entropy": 5.761635398864746,
      "epoch": 3.811783313562673,
      "grad_norm": 1.125,
      "learning_rate": 0.0003610858399309622,
      "loss": 4.9834,
      "mean_token_accuracy": 0.22224853187799454,
      "num_tokens": 88320358.0,
      "step": 38560
    },
    {
      "entropy": 5.723652696609497,
      "epoch": 3.812277580071174,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003610532340149404,
      "loss": 4.8747,
      "mean_token_accuracy": 0.2352326676249504,
      "num_tokens": 88331767.0,
      "step": 38565
    },
    {
      "entropy": 5.722182750701904,
      "epoch": 3.812771846579676,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036102062598199566,
      "loss": 4.8738,
      "mean_token_accuracy": 0.22826676666736603,
      "num_tokens": 88342197.0,
      "step": 38570
    },
    {
      "entropy": 5.780563688278198,
      "epoch": 3.813266113088177,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003609880158329301,
      "loss": 4.9426,
      "mean_token_accuracy": 0.2273011952638626,
      "num_tokens": 88353378.0,
      "step": 38575
    },
    {
      "entropy": 5.6785595417022705,
      "epoch": 3.813760379596679,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003609554035685462,
      "loss": 4.8759,
      "mean_token_accuracy": 0.23317303210496904,
      "num_tokens": 88366006.0,
      "step": 38580
    },
    {
      "entropy": 5.821148347854614,
      "epoch": 3.81425464610518,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003609227891896459,
      "loss": 5.0539,
      "mean_token_accuracy": 0.22001373022794724,
      "num_tokens": 88378132.0,
      "step": 38585
    },
    {
      "entropy": 5.737678146362304,
      "epoch": 3.8147489126136813,
      "grad_norm": 1.28125,
      "learning_rate": 0.00036089017269703164,
      "loss": 4.9127,
      "mean_token_accuracy": 0.2250444307923317,
      "num_tokens": 88389760.0,
      "step": 38590
    },
    {
      "entropy": 5.703362846374512,
      "epoch": 3.8152431791221826,
      "grad_norm": 1.328125,
      "learning_rate": 0.00036085755409150577,
      "loss": 4.9425,
      "mean_token_accuracy": 0.2284134656190872,
      "num_tokens": 88402518.0,
      "step": 38595
    },
    {
      "entropy": 5.771936559677124,
      "epoch": 3.815737445630684,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003608249333738709,
      "loss": 5.0015,
      "mean_token_accuracy": 0.21708550304174423,
      "num_tokens": 88413359.0,
      "step": 38600
    },
    {
      "entropy": 5.813227319717408,
      "epoch": 3.8162317121391856,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003607923105449292,
      "loss": 4.9773,
      "mean_token_accuracy": 0.21970864832401277,
      "num_tokens": 88424966.0,
      "step": 38605
    },
    {
      "entropy": 5.718343353271484,
      "epoch": 3.816725978647687,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036075968560548346,
      "loss": 4.8407,
      "mean_token_accuracy": 0.23425922989845277,
      "num_tokens": 88435826.0,
      "step": 38610
    },
    {
      "entropy": 5.68151593208313,
      "epoch": 3.817220245156188,
      "grad_norm": 1.234375,
      "learning_rate": 0.00036072705855633613,
      "loss": 4.8962,
      "mean_token_accuracy": 0.23387911915779114,
      "num_tokens": 88447136.0,
      "step": 38615
    },
    {
      "entropy": 5.650468349456787,
      "epoch": 3.81771451166469,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003606944293982898,
      "loss": 4.8508,
      "mean_token_accuracy": 0.23945784121751784,
      "num_tokens": 88458394.0,
      "step": 38620
    },
    {
      "entropy": 5.805815696716309,
      "epoch": 3.818208778173191,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003606617981321473,
      "loss": 4.9495,
      "mean_token_accuracy": 0.2281596392393112,
      "num_tokens": 88469573.0,
      "step": 38625
    },
    {
      "entropy": 5.778375196456909,
      "epoch": 3.8187030446816923,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00036062916475871124,
      "loss": 4.8847,
      "mean_token_accuracy": 0.234144826233387,
      "num_tokens": 88479561.0,
      "step": 38630
    },
    {
      "entropy": 5.7266237258911135,
      "epoch": 3.8191973111901936,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036059652927878446,
      "loss": 4.9825,
      "mean_token_accuracy": 0.22871050089597703,
      "num_tokens": 88492296.0,
      "step": 38635
    },
    {
      "entropy": 5.782790946960449,
      "epoch": 3.8196915776986953,
      "grad_norm": 1.28125,
      "learning_rate": 0.00036056389169316977,
      "loss": 5.0642,
      "mean_token_accuracy": 0.22096779495477675,
      "num_tokens": 88504993.0,
      "step": 38640
    },
    {
      "entropy": 5.752505970001221,
      "epoch": 3.8201858442071965,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036053125200267,
      "loss": 4.9282,
      "mean_token_accuracy": 0.23099552989006042,
      "num_tokens": 88516126.0,
      "step": 38645
    },
    {
      "entropy": 5.7910315990448,
      "epoch": 3.820680110715698,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003604986102080882,
      "loss": 4.9952,
      "mean_token_accuracy": 0.21310860961675643,
      "num_tokens": 88527327.0,
      "step": 38650
    },
    {
      "entropy": 5.701392269134521,
      "epoch": 3.8211743772241995,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003604659663102274,
      "loss": 4.8475,
      "mean_token_accuracy": 0.23302048444747925,
      "num_tokens": 88539088.0,
      "step": 38655
    },
    {
      "entropy": 5.708590078353882,
      "epoch": 3.8216686437327008,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003604333203098905,
      "loss": 4.9111,
      "mean_token_accuracy": 0.22605914622545242,
      "num_tokens": 88551539.0,
      "step": 38660
    },
    {
      "entropy": 5.757379817962646,
      "epoch": 3.822162910241202,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036040067220788075,
      "loss": 4.9308,
      "mean_token_accuracy": 0.2261853575706482,
      "num_tokens": 88563111.0,
      "step": 38665
    },
    {
      "entropy": 5.754641342163086,
      "epoch": 3.8226571767497033,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003603680220050011,
      "loss": 4.9566,
      "mean_token_accuracy": 0.23046756088733672,
      "num_tokens": 88574181.0,
      "step": 38670
    },
    {
      "entropy": 5.782589101791382,
      "epoch": 3.8231514432582046,
      "grad_norm": 1.234375,
      "learning_rate": 0.00036033536970205487,
      "loss": 4.9981,
      "mean_token_accuracy": 0.2266621023416519,
      "num_tokens": 88584965.0,
      "step": 38675
    },
    {
      "entropy": 5.848534631729126,
      "epoch": 3.8236457097667063,
      "grad_norm": 1.40625,
      "learning_rate": 0.00036030271529984535,
      "loss": 5.0441,
      "mean_token_accuracy": 0.21533331423997878,
      "num_tokens": 88597390.0,
      "step": 38680
    },
    {
      "entropy": 5.809558963775634,
      "epoch": 3.8241399762752075,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003602700587991759,
      "loss": 4.9575,
      "mean_token_accuracy": 0.2284396171569824,
      "num_tokens": 88609470.0,
      "step": 38685
    },
    {
      "entropy": 5.7601337909698485,
      "epoch": 3.8246342427837092,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00036023740020084963,
      "loss": 4.9431,
      "mean_token_accuracy": 0.22481801360845566,
      "num_tokens": 88620858.0,
      "step": 38690
    },
    {
      "entropy": 5.779210090637207,
      "epoch": 3.8251285092922105,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003602047395056701,
      "loss": 4.9897,
      "mean_token_accuracy": 0.2229233890771866,
      "num_tokens": 88632191.0,
      "step": 38695
    },
    {
      "entropy": 5.78726601600647,
      "epoch": 3.8256227758007118,
      "grad_norm": 1.28125,
      "learning_rate": 0.00036017207671444086,
      "loss": 4.9828,
      "mean_token_accuracy": 0.22114255726337434,
      "num_tokens": 88643405.0,
      "step": 38700
    },
    {
      "entropy": 5.719884872436523,
      "epoch": 3.826117042309213,
      "grad_norm": 1.265625,
      "learning_rate": 0.00036013941182796516,
      "loss": 4.8149,
      "mean_token_accuracy": 0.23648451119661332,
      "num_tokens": 88653765.0,
      "step": 38705
    },
    {
      "entropy": 5.7795919418334964,
      "epoch": 3.8266113088177143,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036010674484704684,
      "loss": 4.9827,
      "mean_token_accuracy": 0.2190888747572899,
      "num_tokens": 88665038.0,
      "step": 38710
    },
    {
      "entropy": 5.726240968704223,
      "epoch": 3.827105575326216,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003600740757724894,
      "loss": 4.9119,
      "mean_token_accuracy": 0.23566055446863174,
      "num_tokens": 88675172.0,
      "step": 38715
    },
    {
      "entropy": 5.750473260879517,
      "epoch": 3.8275998418347172,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036004140460509646,
      "loss": 4.972,
      "mean_token_accuracy": 0.22573740780353546,
      "num_tokens": 88687393.0,
      "step": 38720
    },
    {
      "entropy": 5.853509521484375,
      "epoch": 3.8280941083432185,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003600087313456718,
      "loss": 5.0348,
      "mean_token_accuracy": 0.21617888510227204,
      "num_tokens": 88700117.0,
      "step": 38725
    },
    {
      "entropy": 5.762597131729126,
      "epoch": 3.82858837485172,
      "grad_norm": 1.4375,
      "learning_rate": 0.00035997605599501914,
      "loss": 4.9548,
      "mean_token_accuracy": 0.22774316370487213,
      "num_tokens": 88711329.0,
      "step": 38730
    },
    {
      "entropy": 5.81672945022583,
      "epoch": 3.8290826413602215,
      "grad_norm": 1.25,
      "learning_rate": 0.0003599433785539423,
      "loss": 5.0012,
      "mean_token_accuracy": 0.22332691103219987,
      "num_tokens": 88722017.0,
      "step": 38735
    },
    {
      "entropy": 5.722673416137695,
      "epoch": 3.8295769078687227,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00035991069902324517,
      "loss": 4.922,
      "mean_token_accuracy": 0.22938984334468843,
      "num_tokens": 88733749.0,
      "step": 38740
    },
    {
      "entropy": 5.773297119140625,
      "epoch": 3.830071174377224,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003598780174037318,
      "loss": 4.9374,
      "mean_token_accuracy": 0.2268606022000313,
      "num_tokens": 88745885.0,
      "step": 38745
    },
    {
      "entropy": 5.841207790374756,
      "epoch": 3.8305654408857257,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003598453336962059,
      "loss": 5.0833,
      "mean_token_accuracy": 0.21651489585638045,
      "num_tokens": 88759216.0,
      "step": 38750
    },
    {
      "entropy": 5.74736180305481,
      "epoch": 3.831059707394227,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00035981264790147167,
      "loss": 4.9892,
      "mean_token_accuracy": 0.2230536237359047,
      "num_tokens": 88770272.0,
      "step": 38755
    },
    {
      "entropy": 5.698786354064941,
      "epoch": 3.8315539739027282,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003597799600203331,
      "loss": 4.8689,
      "mean_token_accuracy": 0.2378442943096161,
      "num_tokens": 88781509.0,
      "step": 38760
    },
    {
      "entropy": 5.770243167877197,
      "epoch": 3.83204824041123,
      "grad_norm": 1.296875,
      "learning_rate": 0.00035974727005359434,
      "loss": 4.9788,
      "mean_token_accuracy": 0.22589323073625564,
      "num_tokens": 88791770.0,
      "step": 38765
    },
    {
      "entropy": 5.764696550369263,
      "epoch": 3.832542506919731,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003597145780020597,
      "loss": 4.9725,
      "mean_token_accuracy": 0.22748714089393615,
      "num_tokens": 88804344.0,
      "step": 38770
    },
    {
      "entropy": 5.753374528884888,
      "epoch": 3.8330367734282325,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003596818838665333,
      "loss": 4.8872,
      "mean_token_accuracy": 0.23273838460445403,
      "num_tokens": 88816251.0,
      "step": 38775
    },
    {
      "entropy": 5.693378686904907,
      "epoch": 3.8335310399367337,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003596491876478193,
      "loss": 4.9326,
      "mean_token_accuracy": 0.22621732652187349,
      "num_tokens": 88827805.0,
      "step": 38780
    },
    {
      "entropy": 5.763323593139648,
      "epoch": 3.8340253064452354,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003596164893467222,
      "loss": 4.9282,
      "mean_token_accuracy": 0.22901551723480223,
      "num_tokens": 88839237.0,
      "step": 38785
    },
    {
      "entropy": 5.882288789749145,
      "epoch": 3.8345195729537367,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00035958378896404635,
      "loss": 5.0416,
      "mean_token_accuracy": 0.21981064081192017,
      "num_tokens": 88850950.0,
      "step": 38790
    },
    {
      "entropy": 5.735190343856812,
      "epoch": 3.835013839462238,
      "grad_norm": 1.203125,
      "learning_rate": 0.00035955108650059613,
      "loss": 4.9189,
      "mean_token_accuracy": 0.23332116305828093,
      "num_tokens": 88863743.0,
      "step": 38795
    },
    {
      "entropy": 5.683094453811646,
      "epoch": 3.8355081059707397,
      "grad_norm": 1.109375,
      "learning_rate": 0.00035951838195717614,
      "loss": 4.8706,
      "mean_token_accuracy": 0.23135916441679,
      "num_tokens": 88875375.0,
      "step": 38800
    },
    {
      "entropy": 5.792744493484497,
      "epoch": 3.836002372479241,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00035948567533459076,
      "loss": 4.999,
      "mean_token_accuracy": 0.22894601672887802,
      "num_tokens": 88885890.0,
      "step": 38805
    },
    {
      "entropy": 5.807839727401733,
      "epoch": 3.836496638987742,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00035945296663364467,
      "loss": 4.9532,
      "mean_token_accuracy": 0.22057102620601654,
      "num_tokens": 88897783.0,
      "step": 38810
    },
    {
      "entropy": 5.7756999969482425,
      "epoch": 3.8369909054962434,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00035942025585514257,
      "loss": 4.9903,
      "mean_token_accuracy": 0.226309172809124,
      "num_tokens": 88908982.0,
      "step": 38815
    },
    {
      "entropy": 5.686149930953979,
      "epoch": 3.8374851720047447,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000359387542999889,
      "loss": 4.8565,
      "mean_token_accuracy": 0.24286628067493438,
      "num_tokens": 88921059.0,
      "step": 38820
    },
    {
      "entropy": 5.798223781585693,
      "epoch": 3.8379794385132464,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003593548280686889,
      "loss": 5.0786,
      "mean_token_accuracy": 0.20993549525737762,
      "num_tokens": 88932988.0,
      "step": 38825
    },
    {
      "entropy": 5.790928745269776,
      "epoch": 3.8384737050217477,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003593221110623468,
      "loss": 4.9542,
      "mean_token_accuracy": 0.22731160074472428,
      "num_tokens": 88944890.0,
      "step": 38830
    },
    {
      "entropy": 5.763184595108032,
      "epoch": 3.8389679715302494,
      "grad_norm": 1.1875,
      "learning_rate": 0.00035928939198166777,
      "loss": 4.9534,
      "mean_token_accuracy": 0.2276351422071457,
      "num_tokens": 88956936.0,
      "step": 38835
    },
    {
      "entropy": 5.776267957687378,
      "epoch": 3.8394622380387506,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003592566708274566,
      "loss": 4.9883,
      "mean_token_accuracy": 0.22827732414007187,
      "num_tokens": 88968705.0,
      "step": 38840
    },
    {
      "entropy": 5.74217209815979,
      "epoch": 3.839956504547252,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003592239476005183,
      "loss": 4.8825,
      "mean_token_accuracy": 0.23492239266633988,
      "num_tokens": 88979977.0,
      "step": 38845
    },
    {
      "entropy": 5.7370171546936035,
      "epoch": 3.840450771055753,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003591912223016578,
      "loss": 4.9955,
      "mean_token_accuracy": 0.21851376593112945,
      "num_tokens": 88991974.0,
      "step": 38850
    },
    {
      "entropy": 5.713030767440796,
      "epoch": 3.8409450375642544,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003591584949316801,
      "loss": 4.903,
      "mean_token_accuracy": 0.2345935359597206,
      "num_tokens": 89003733.0,
      "step": 38855
    },
    {
      "entropy": 5.847453927993774,
      "epoch": 3.841439304072756,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003591257654913904,
      "loss": 5.0432,
      "mean_token_accuracy": 0.20937418341636657,
      "num_tokens": 89015701.0,
      "step": 38860
    },
    {
      "entropy": 5.841371679306031,
      "epoch": 3.8419335705812574,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003590930339815938,
      "loss": 5.0209,
      "mean_token_accuracy": 0.22092653661966324,
      "num_tokens": 89027216.0,
      "step": 38865
    },
    {
      "entropy": 5.8067199230194095,
      "epoch": 3.8424278370897587,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003590603004030956,
      "loss": 4.9559,
      "mean_token_accuracy": 0.22416433542966843,
      "num_tokens": 89039305.0,
      "step": 38870
    },
    {
      "entropy": 5.7390532970428465,
      "epoch": 3.8429221035982604,
      "grad_norm": 1.21875,
      "learning_rate": 0.000359027564756701,
      "loss": 4.8797,
      "mean_token_accuracy": 0.23254089653491974,
      "num_tokens": 89049982.0,
      "step": 38875
    },
    {
      "entropy": 5.748529624938965,
      "epoch": 3.8434163701067616,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003589948270432151,
      "loss": 4.948,
      "mean_token_accuracy": 0.23012804090976716,
      "num_tokens": 89062039.0,
      "step": 38880
    },
    {
      "entropy": 5.755610513687134,
      "epoch": 3.843910636615263,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003589620872634436,
      "loss": 4.9483,
      "mean_token_accuracy": 0.22757059037685395,
      "num_tokens": 89072621.0,
      "step": 38885
    },
    {
      "entropy": 5.796428632736206,
      "epoch": 3.844404903123764,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003589293454181916,
      "loss": 4.9187,
      "mean_token_accuracy": 0.22796160131692886,
      "num_tokens": 89084019.0,
      "step": 38890
    },
    {
      "entropy": 5.777689504623413,
      "epoch": 3.844899169632266,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00035889660150826473,
      "loss": 4.9561,
      "mean_token_accuracy": 0.2256498709321022,
      "num_tokens": 89094305.0,
      "step": 38895
    },
    {
      "entropy": 5.707172632217407,
      "epoch": 3.845393436140767,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00035886385553446853,
      "loss": 4.9623,
      "mean_token_accuracy": 0.2203045204281807,
      "num_tokens": 89106261.0,
      "step": 38900
    },
    {
      "entropy": 5.69955096244812,
      "epoch": 3.8458877026492684,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00035883110749760836,
      "loss": 4.9013,
      "mean_token_accuracy": 0.22885473668575287,
      "num_tokens": 89118171.0,
      "step": 38905
    },
    {
      "entropy": 5.809811973571778,
      "epoch": 3.84638196915777,
      "grad_norm": 1.3125,
      "learning_rate": 0.00035879835739848994,
      "loss": 5.0386,
      "mean_token_accuracy": 0.22211336493492126,
      "num_tokens": 89129953.0,
      "step": 38910
    },
    {
      "entropy": 5.695519304275512,
      "epoch": 3.8468762356662713,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000358765605237919,
      "loss": 4.8589,
      "mean_token_accuracy": 0.23657006323337554,
      "num_tokens": 89143284.0,
      "step": 38915
    },
    {
      "entropy": 5.790096426010132,
      "epoch": 3.8473705021747726,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003587328510167012,
      "loss": 4.9947,
      "mean_token_accuracy": 0.22115783393383026,
      "num_tokens": 89154675.0,
      "step": 38920
    },
    {
      "entropy": 5.823217535018921,
      "epoch": 3.847864768683274,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003587000947356421,
      "loss": 5.0133,
      "mean_token_accuracy": 0.21787019371986388,
      "num_tokens": 89164576.0,
      "step": 38925
    },
    {
      "entropy": 5.771351003646851,
      "epoch": 3.848359035191775,
      "grad_norm": 1.1875,
      "learning_rate": 0.00035866733639554777,
      "loss": 5.0358,
      "mean_token_accuracy": 0.21166231334209443,
      "num_tokens": 89176106.0,
      "step": 38930
    },
    {
      "entropy": 5.773771953582764,
      "epoch": 3.848853301700277,
      "grad_norm": 1.171875,
      "learning_rate": 0.000358634575997224,
      "loss": 4.9562,
      "mean_token_accuracy": 0.223372446000576,
      "num_tokens": 89187214.0,
      "step": 38935
    },
    {
      "entropy": 5.695539712905884,
      "epoch": 3.849347568208778,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003586018135414767,
      "loss": 4.8395,
      "mean_token_accuracy": 0.2304440751671791,
      "num_tokens": 89198366.0,
      "step": 38940
    },
    {
      "entropy": 5.735632991790771,
      "epoch": 3.84984183471728,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003585690490291119,
      "loss": 5.0195,
      "mean_token_accuracy": 0.22199472039937973,
      "num_tokens": 89209122.0,
      "step": 38945
    },
    {
      "entropy": 5.809645557403565,
      "epoch": 3.850336101225781,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003585362824609354,
      "loss": 5.0177,
      "mean_token_accuracy": 0.21956222206354142,
      "num_tokens": 89221104.0,
      "step": 38950
    },
    {
      "entropy": 5.806887817382813,
      "epoch": 3.8508303677342823,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003585035138377534,
      "loss": 4.9837,
      "mean_token_accuracy": 0.22624120116233826,
      "num_tokens": 89231087.0,
      "step": 38955
    },
    {
      "entropy": 5.782963848114013,
      "epoch": 3.8513246342427836,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00035847074316037206,
      "loss": 5.0412,
      "mean_token_accuracy": 0.2204209163784981,
      "num_tokens": 89243198.0,
      "step": 38960
    },
    {
      "entropy": 5.713662385940552,
      "epoch": 3.851818900751285,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003584379704295975,
      "loss": 4.9431,
      "mean_token_accuracy": 0.22764402031898498,
      "num_tokens": 89254408.0,
      "step": 38965
    },
    {
      "entropy": 5.822534513473511,
      "epoch": 3.8523131672597866,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003584051956462359,
      "loss": 5.0057,
      "mean_token_accuracy": 0.2243173524737358,
      "num_tokens": 89266087.0,
      "step": 38970
    },
    {
      "entropy": 5.747350215911865,
      "epoch": 3.852807433768288,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003583724188110936,
      "loss": 4.8974,
      "mean_token_accuracy": 0.2257521629333496,
      "num_tokens": 89277039.0,
      "step": 38975
    },
    {
      "entropy": 5.809957265853882,
      "epoch": 3.853301700276789,
      "grad_norm": 1.3125,
      "learning_rate": 0.00035833963992497667,
      "loss": 5.0189,
      "mean_token_accuracy": 0.22204402834177017,
      "num_tokens": 89288940.0,
      "step": 38980
    },
    {
      "entropy": 5.805511426925659,
      "epoch": 3.853795966785291,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003583068589886919,
      "loss": 4.9196,
      "mean_token_accuracy": 0.22370391637086867,
      "num_tokens": 89301460.0,
      "step": 38985
    },
    {
      "entropy": 5.791597318649292,
      "epoch": 3.854290233293792,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00035827407600304533,
      "loss": 4.9899,
      "mean_token_accuracy": 0.22211606055498123,
      "num_tokens": 89312421.0,
      "step": 38990
    },
    {
      "entropy": 5.743489456176758,
      "epoch": 3.8547844998022933,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003582412909688436,
      "loss": 4.9546,
      "mean_token_accuracy": 0.2253192186355591,
      "num_tokens": 89323552.0,
      "step": 38995
    },
    {
      "entropy": 5.79524827003479,
      "epoch": 3.8552787663107946,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003582085038868932,
      "loss": 4.9981,
      "mean_token_accuracy": 0.21214788407087326,
      "num_tokens": 89333911.0,
      "step": 39000
    },
    {
      "epoch": 3.8552787663107946,
      "eval_entropy": 5.5270417131327765,
      "eval_loss": 5.015138149261475,
      "eval_mean_token_accuracy": 0.23101152317623663,
      "eval_num_tokens": 89333911.0,
      "eval_runtime": 20.4603,
      "eval_samples_per_second": 1589.076,
      "eval_steps_per_second": 198.677,
      "step": 39000
    },
    {
      "entropy": 5.843964576721191,
      "epoch": 3.8557730328192963,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003581757147580007,
      "loss": 5.0649,
      "mean_token_accuracy": 0.2166712090373039,
      "num_tokens": 89344632.0,
      "step": 39005
    },
    {
      "entropy": 5.776649713516235,
      "epoch": 3.8562672993277975,
      "grad_norm": 1.265625,
      "learning_rate": 0.00035814292358297273,
      "loss": 4.9481,
      "mean_token_accuracy": 0.23139783442020417,
      "num_tokens": 89354443.0,
      "step": 39010
    },
    {
      "entropy": 5.784475088119507,
      "epoch": 3.856761565836299,
      "grad_norm": 1.25,
      "learning_rate": 0.000358110130362616,
      "loss": 5.0145,
      "mean_token_accuracy": 0.22555150836706161,
      "num_tokens": 89365292.0,
      "step": 39015
    },
    {
      "entropy": 5.811525583267212,
      "epoch": 3.8572558323448005,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00035807733509773704,
      "loss": 5.0614,
      "mean_token_accuracy": 0.22084376066923142,
      "num_tokens": 89378280.0,
      "step": 39020
    },
    {
      "entropy": 5.8122330665588375,
      "epoch": 3.857750098853302,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003580445377891428,
      "loss": 4.9598,
      "mean_token_accuracy": 0.2288960337638855,
      "num_tokens": 89390360.0,
      "step": 39025
    },
    {
      "entropy": 5.784896373748779,
      "epoch": 3.858244365361803,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003580117384376401,
      "loss": 4.9859,
      "mean_token_accuracy": 0.22218020260334015,
      "num_tokens": 89402057.0,
      "step": 39030
    },
    {
      "entropy": 5.73190279006958,
      "epoch": 3.8587386318703043,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003579789370440358,
      "loss": 4.9314,
      "mean_token_accuracy": 0.22414183914661406,
      "num_tokens": 89413057.0,
      "step": 39035
    },
    {
      "entropy": 5.7765881538391115,
      "epoch": 3.859232898378806,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00035794613360913666,
      "loss": 5.027,
      "mean_token_accuracy": 0.2201632261276245,
      "num_tokens": 89425565.0,
      "step": 39040
    },
    {
      "entropy": 5.784101676940918,
      "epoch": 3.8597271648873073,
      "grad_norm": 1.296875,
      "learning_rate": 0.00035791332813374973,
      "loss": 4.9637,
      "mean_token_accuracy": 0.22456785887479783,
      "num_tokens": 89436806.0,
      "step": 39045
    },
    {
      "entropy": 5.799672174453735,
      "epoch": 3.8602214313958085,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00035788052061868215,
      "loss": 5.0198,
      "mean_token_accuracy": 0.2219857320189476,
      "num_tokens": 89448745.0,
      "step": 39050
    },
    {
      "entropy": 5.802977275848389,
      "epoch": 3.8607156979043102,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003578477110647409,
      "loss": 4.9753,
      "mean_token_accuracy": 0.22809406518936157,
      "num_tokens": 89460236.0,
      "step": 39055
    },
    {
      "entropy": 5.7846136569976805,
      "epoch": 3.8612099644128115,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003578148994727331,
      "loss": 4.9804,
      "mean_token_accuracy": 0.2210586816072464,
      "num_tokens": 89471534.0,
      "step": 39060
    },
    {
      "entropy": 5.822690391540528,
      "epoch": 3.8617042309213128,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003577820858434659,
      "loss": 5.0133,
      "mean_token_accuracy": 0.22286662012338637,
      "num_tokens": 89483803.0,
      "step": 39065
    },
    {
      "entropy": 5.763155078887939,
      "epoch": 3.862198497429814,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00035774927017774653,
      "loss": 4.9613,
      "mean_token_accuracy": 0.22060411423444748,
      "num_tokens": 89494236.0,
      "step": 39070
    },
    {
      "entropy": 5.808399677276611,
      "epoch": 3.8626927639383153,
      "grad_norm": 1.1875,
      "learning_rate": 0.00035771645247638243,
      "loss": 4.9682,
      "mean_token_accuracy": 0.22675992995500566,
      "num_tokens": 89506101.0,
      "step": 39075
    },
    {
      "entropy": 5.789398097991944,
      "epoch": 3.863187030446817,
      "grad_norm": 1.171875,
      "learning_rate": 0.00035768363274018064,
      "loss": 4.9351,
      "mean_token_accuracy": 0.23177928030490874,
      "num_tokens": 89518820.0,
      "step": 39080
    },
    {
      "entropy": 5.710754823684693,
      "epoch": 3.8636812969553183,
      "grad_norm": 1.265625,
      "learning_rate": 0.00035765081096994863,
      "loss": 4.8569,
      "mean_token_accuracy": 0.229751555621624,
      "num_tokens": 89530337.0,
      "step": 39085
    },
    {
      "entropy": 5.735813951492309,
      "epoch": 3.86417556346382,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00035761798716649385,
      "loss": 4.9064,
      "mean_token_accuracy": 0.22648853063583374,
      "num_tokens": 89541516.0,
      "step": 39090
    },
    {
      "entropy": 5.823968076705933,
      "epoch": 3.8646698299723212,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003575851613306238,
      "loss": 5.0128,
      "mean_token_accuracy": 0.2252624198794365,
      "num_tokens": 89553544.0,
      "step": 39095
    },
    {
      "entropy": 5.766701745986938,
      "epoch": 3.8651640964808225,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00035755233346314595,
      "loss": 4.9571,
      "mean_token_accuracy": 0.2299944654107094,
      "num_tokens": 89564472.0,
      "step": 39100
    },
    {
      "entropy": 5.777039337158203,
      "epoch": 3.8656583629893237,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00035751950356486796,
      "loss": 5.008,
      "mean_token_accuracy": 0.22157350480556487,
      "num_tokens": 89575708.0,
      "step": 39105
    },
    {
      "entropy": 5.809261274337769,
      "epoch": 3.866152629497825,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00035748667163659737,
      "loss": 4.9093,
      "mean_token_accuracy": 0.23332905173301696,
      "num_tokens": 89585643.0,
      "step": 39110
    },
    {
      "entropy": 5.696730041503907,
      "epoch": 3.8666468960063267,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003574538376791418,
      "loss": 4.8996,
      "mean_token_accuracy": 0.2408711552619934,
      "num_tokens": 89597069.0,
      "step": 39115
    },
    {
      "entropy": 5.864247179031372,
      "epoch": 3.867141162514828,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00035742100169330903,
      "loss": 5.1147,
      "mean_token_accuracy": 0.209864641726017,
      "num_tokens": 89608469.0,
      "step": 39120
    },
    {
      "entropy": 5.805208969116211,
      "epoch": 3.8676354290233292,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003573881636799068,
      "loss": 5.0326,
      "mean_token_accuracy": 0.2186299070715904,
      "num_tokens": 89620132.0,
      "step": 39125
    },
    {
      "entropy": 5.747498226165772,
      "epoch": 3.868129695531831,
      "grad_norm": 1.25,
      "learning_rate": 0.00035735532363974305,
      "loss": 4.9344,
      "mean_token_accuracy": 0.2288547471165657,
      "num_tokens": 89631565.0,
      "step": 39130
    },
    {
      "entropy": 5.754403495788575,
      "epoch": 3.868623962040332,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003573224815736256,
      "loss": 4.896,
      "mean_token_accuracy": 0.23093175739049912,
      "num_tokens": 89641530.0,
      "step": 39135
    },
    {
      "entropy": 5.772377061843872,
      "epoch": 3.8691182285488335,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003572896374823622,
      "loss": 5.0204,
      "mean_token_accuracy": 0.2199908599257469,
      "num_tokens": 89653588.0,
      "step": 39140
    },
    {
      "entropy": 5.8562065124511715,
      "epoch": 3.8696124950573347,
      "grad_norm": 1.234375,
      "learning_rate": 0.00035725679136676104,
      "loss": 5.062,
      "mean_token_accuracy": 0.22180762588977815,
      "num_tokens": 89665506.0,
      "step": 39145
    },
    {
      "entropy": 5.738216161727905,
      "epoch": 3.8701067615658364,
      "grad_norm": 1.234375,
      "learning_rate": 0.00035722394322763,
      "loss": 4.9335,
      "mean_token_accuracy": 0.22933071255683898,
      "num_tokens": 89678133.0,
      "step": 39150
    },
    {
      "entropy": 5.7754473209381105,
      "epoch": 3.8706010280743377,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003571910930657773,
      "loss": 4.8885,
      "mean_token_accuracy": 0.23716054111719131,
      "num_tokens": 89690102.0,
      "step": 39155
    },
    {
      "entropy": 5.709008598327637,
      "epoch": 3.871095294582839,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003571582408820109,
      "loss": 4.8603,
      "mean_token_accuracy": 0.23795785307884215,
      "num_tokens": 89701434.0,
      "step": 39160
    },
    {
      "entropy": 5.7675879955291744,
      "epoch": 3.8715895610913407,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00035712538667713895,
      "loss": 5.0563,
      "mean_token_accuracy": 0.2192534625530243,
      "num_tokens": 89711812.0,
      "step": 39165
    },
    {
      "entropy": 5.815571212768555,
      "epoch": 3.872083827599842,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00035709253045196977,
      "loss": 4.9905,
      "mean_token_accuracy": 0.22557516545057296,
      "num_tokens": 89723077.0,
      "step": 39170
    },
    {
      "entropy": 5.770314407348633,
      "epoch": 3.872578094108343,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003570596722073117,
      "loss": 4.9451,
      "mean_token_accuracy": 0.22723625004291534,
      "num_tokens": 89733964.0,
      "step": 39175
    },
    {
      "entropy": 5.776780796051026,
      "epoch": 3.8730723606168445,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000357026811943973,
      "loss": 4.9718,
      "mean_token_accuracy": 0.22622642070055007,
      "num_tokens": 89745086.0,
      "step": 39180
    },
    {
      "entropy": 5.771366834640503,
      "epoch": 3.8735666271253457,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00035699394966276184,
      "loss": 4.9699,
      "mean_token_accuracy": 0.21960407346487046,
      "num_tokens": 89756075.0,
      "step": 39185
    },
    {
      "entropy": 5.754063749313355,
      "epoch": 3.8740608936338474,
      "grad_norm": 1.25,
      "learning_rate": 0.00035696108536448684,
      "loss": 4.8912,
      "mean_token_accuracy": 0.232508347928524,
      "num_tokens": 89766271.0,
      "step": 39190
    },
    {
      "entropy": 5.7904266834259035,
      "epoch": 3.8745551601423487,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00035692821904995646,
      "loss": 5.0254,
      "mean_token_accuracy": 0.22071074694395065,
      "num_tokens": 89778408.0,
      "step": 39195
    },
    {
      "entropy": 5.778732776641846,
      "epoch": 3.8750494266508504,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00035689535071997913,
      "loss": 4.9626,
      "mean_token_accuracy": 0.22704945206642152,
      "num_tokens": 89789670.0,
      "step": 39200
    },
    {
      "entropy": 5.758708763122558,
      "epoch": 3.8755436931593517,
      "grad_norm": 1.171875,
      "learning_rate": 0.00035686248037536356,
      "loss": 4.9618,
      "mean_token_accuracy": 0.22741394639015197,
      "num_tokens": 89802269.0,
      "step": 39205
    },
    {
      "entropy": 5.744171953201294,
      "epoch": 3.876037959667853,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003568296080169182,
      "loss": 4.9393,
      "mean_token_accuracy": 0.22382222414016723,
      "num_tokens": 89814168.0,
      "step": 39210
    },
    {
      "entropy": 5.783205127716064,
      "epoch": 3.876532226176354,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003567967336454518,
      "loss": 4.973,
      "mean_token_accuracy": 0.22233276814222336,
      "num_tokens": 89826494.0,
      "step": 39215
    },
    {
      "entropy": 5.752947998046875,
      "epoch": 3.8770264926848554,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003567638572617731,
      "loss": 4.9105,
      "mean_token_accuracy": 0.22909229397773742,
      "num_tokens": 89838266.0,
      "step": 39220
    },
    {
      "entropy": 5.707412242889404,
      "epoch": 3.877520759193357,
      "grad_norm": 1.40625,
      "learning_rate": 0.00035673097886669086,
      "loss": 4.9074,
      "mean_token_accuracy": 0.2270902544260025,
      "num_tokens": 89848941.0,
      "step": 39225
    },
    {
      "entropy": 5.765740394592285,
      "epoch": 3.8780150257018584,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00035669809846101383,
      "loss": 4.9396,
      "mean_token_accuracy": 0.22830454856157303,
      "num_tokens": 89859878.0,
      "step": 39230
    },
    {
      "entropy": 5.765536165237426,
      "epoch": 3.8785092922103597,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003566652160455508,
      "loss": 4.9663,
      "mean_token_accuracy": 0.2269761949777603,
      "num_tokens": 89872044.0,
      "step": 39235
    },
    {
      "entropy": 5.786499977111816,
      "epoch": 3.8790035587188614,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003566323316211109,
      "loss": 4.9699,
      "mean_token_accuracy": 0.22595528066158294,
      "num_tokens": 89883950.0,
      "step": 39240
    },
    {
      "entropy": 5.794172143936157,
      "epoch": 3.8794978252273626,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003565994451885029,
      "loss": 5.0092,
      "mean_token_accuracy": 0.22047265768051147,
      "num_tokens": 89895944.0,
      "step": 39245
    },
    {
      "entropy": 5.78608341217041,
      "epoch": 3.879992091735864,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00035656655674853595,
      "loss": 5.0006,
      "mean_token_accuracy": 0.2194966718554497,
      "num_tokens": 89907413.0,
      "step": 39250
    },
    {
      "entropy": 5.749328851699829,
      "epoch": 3.880486358244365,
      "grad_norm": 1.265625,
      "learning_rate": 0.00035653366630201906,
      "loss": 4.8878,
      "mean_token_accuracy": 0.2271288201212883,
      "num_tokens": 89918910.0,
      "step": 39255
    },
    {
      "entropy": 5.808552408218384,
      "epoch": 3.880980624752867,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003565007738497612,
      "loss": 5.1022,
      "mean_token_accuracy": 0.2123834580183029,
      "num_tokens": 89931928.0,
      "step": 39260
    },
    {
      "entropy": 5.761010408401489,
      "epoch": 3.881474891261368,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003564678793925718,
      "loss": 4.954,
      "mean_token_accuracy": 0.2312987580895424,
      "num_tokens": 89942963.0,
      "step": 39265
    },
    {
      "entropy": 5.705652904510498,
      "epoch": 3.8819691577698694,
      "grad_norm": 1.265625,
      "learning_rate": 0.00035643498293125997,
      "loss": 4.898,
      "mean_token_accuracy": 0.22958698719739914,
      "num_tokens": 89953706.0,
      "step": 39270
    },
    {
      "entropy": 5.714903545379639,
      "epoch": 3.882463424278371,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003564020844666348,
      "loss": 4.9611,
      "mean_token_accuracy": 0.22449294626712799,
      "num_tokens": 89964243.0,
      "step": 39275
    },
    {
      "entropy": 5.846740865707398,
      "epoch": 3.8829576907868724,
      "grad_norm": 1.375,
      "learning_rate": 0.0003563691839995058,
      "loss": 5.0003,
      "mean_token_accuracy": 0.21891957372426987,
      "num_tokens": 89974708.0,
      "step": 39280
    },
    {
      "entropy": 5.742700481414795,
      "epoch": 3.8834519572953736,
      "grad_norm": 1.328125,
      "learning_rate": 0.00035633628153068213,
      "loss": 4.9023,
      "mean_token_accuracy": 0.24532349407672882,
      "num_tokens": 89984950.0,
      "step": 39285
    },
    {
      "entropy": 5.752230596542359,
      "epoch": 3.883946223803875,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003563033770609734,
      "loss": 4.8705,
      "mean_token_accuracy": 0.22767487317323684,
      "num_tokens": 89995605.0,
      "step": 39290
    },
    {
      "entropy": 5.691107702255249,
      "epoch": 3.8844404903123766,
      "grad_norm": 1.171875,
      "learning_rate": 0.00035627047059118907,
      "loss": 4.9116,
      "mean_token_accuracy": 0.23006444424390793,
      "num_tokens": 90007209.0,
      "step": 39295
    },
    {
      "entropy": 5.719086694717407,
      "epoch": 3.884934756820878,
      "grad_norm": 1.171875,
      "learning_rate": 0.00035623756212213847,
      "loss": 4.929,
      "mean_token_accuracy": 0.22537785917520523,
      "num_tokens": 90019042.0,
      "step": 39300
    },
    {
      "entropy": 5.776600408554077,
      "epoch": 3.885429023329379,
      "grad_norm": 1.25,
      "learning_rate": 0.00035620465165463124,
      "loss": 4.9565,
      "mean_token_accuracy": 0.23085272163152695,
      "num_tokens": 90030466.0,
      "step": 39305
    },
    {
      "entropy": 5.742365407943725,
      "epoch": 3.885923289837881,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00035617173918947695,
      "loss": 4.8892,
      "mean_token_accuracy": 0.22705599516630173,
      "num_tokens": 90041308.0,
      "step": 39310
    },
    {
      "entropy": 5.7418293952941895,
      "epoch": 3.886417556346382,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003561388247274853,
      "loss": 5.0175,
      "mean_token_accuracy": 0.22416626811027526,
      "num_tokens": 90052655.0,
      "step": 39315
    },
    {
      "entropy": 5.797838878631592,
      "epoch": 3.8869118228548833,
      "grad_norm": 1.25,
      "learning_rate": 0.000356105908269466,
      "loss": 5.0074,
      "mean_token_accuracy": 0.21475773304700851,
      "num_tokens": 90063643.0,
      "step": 39320
    },
    {
      "entropy": 5.723918342590332,
      "epoch": 3.8874060893633846,
      "grad_norm": 1.328125,
      "learning_rate": 0.00035607298981622885,
      "loss": 4.8963,
      "mean_token_accuracy": 0.23104674220085145,
      "num_tokens": 90076293.0,
      "step": 39325
    },
    {
      "entropy": 5.825060749053955,
      "epoch": 3.887900355871886,
      "grad_norm": 1.328125,
      "learning_rate": 0.00035604006936858343,
      "loss": 5.0224,
      "mean_token_accuracy": 0.2145558550953865,
      "num_tokens": 90086423.0,
      "step": 39330
    },
    {
      "entropy": 5.774041318893433,
      "epoch": 3.8883946223803876,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00035600714692733974,
      "loss": 4.8829,
      "mean_token_accuracy": 0.23435807526111602,
      "num_tokens": 90097330.0,
      "step": 39335
    },
    {
      "entropy": 5.796225118637085,
      "epoch": 3.888888888888889,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003559742224933078,
      "loss": 5.0374,
      "mean_token_accuracy": 0.22253907471895218,
      "num_tokens": 90109532.0,
      "step": 39340
    },
    {
      "entropy": 5.857407522201538,
      "epoch": 3.8893831553973905,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003559412960672972,
      "loss": 5.0724,
      "mean_token_accuracy": 0.21018495708703994,
      "num_tokens": 90121655.0,
      "step": 39345
    },
    {
      "entropy": 5.734466981887818,
      "epoch": 3.889877421905892,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003559083676501184,
      "loss": 4.9445,
      "mean_token_accuracy": 0.23160445988178252,
      "num_tokens": 90133515.0,
      "step": 39350
    },
    {
      "entropy": 5.822032690048218,
      "epoch": 3.890371688414393,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003558754372425811,
      "loss": 4.9829,
      "mean_token_accuracy": 0.2198260620236397,
      "num_tokens": 90146364.0,
      "step": 39355
    },
    {
      "entropy": 5.807757711410522,
      "epoch": 3.8908659549228943,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00035584250484549547,
      "loss": 4.9658,
      "mean_token_accuracy": 0.22484630197286606,
      "num_tokens": 90158568.0,
      "step": 39360
    },
    {
      "entropy": 5.743658971786499,
      "epoch": 3.8913602214313956,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003558095704596717,
      "loss": 4.9395,
      "mean_token_accuracy": 0.22645913064479828,
      "num_tokens": 90169252.0,
      "step": 39365
    },
    {
      "entropy": 5.8253251075744625,
      "epoch": 3.8918544879398973,
      "grad_norm": 1.21875,
      "learning_rate": 0.00035577663408592003,
      "loss": 5.0521,
      "mean_token_accuracy": 0.2255424365401268,
      "num_tokens": 90180398.0,
      "step": 39370
    },
    {
      "entropy": 5.7431800365448,
      "epoch": 3.8923487544483986,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00035574369572505054,
      "loss": 4.899,
      "mean_token_accuracy": 0.23363993167877198,
      "num_tokens": 90191991.0,
      "step": 39375
    },
    {
      "entropy": 5.804829072952271,
      "epoch": 3.8928430209569,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003557107553778736,
      "loss": 4.9676,
      "mean_token_accuracy": 0.22032666504383086,
      "num_tokens": 90203812.0,
      "step": 39380
    },
    {
      "entropy": 5.762810182571411,
      "epoch": 3.8933372874654015,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003556778130451997,
      "loss": 4.9704,
      "mean_token_accuracy": 0.22746917754411697,
      "num_tokens": 90214793.0,
      "step": 39385
    },
    {
      "entropy": 5.740804862976074,
      "epoch": 3.893831553973903,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003556448687278389,
      "loss": 4.909,
      "mean_token_accuracy": 0.22290343195199966,
      "num_tokens": 90225145.0,
      "step": 39390
    },
    {
      "entropy": 5.708287858963013,
      "epoch": 3.894325820482404,
      "grad_norm": 1.25,
      "learning_rate": 0.00035561192242660193,
      "loss": 4.9453,
      "mean_token_accuracy": 0.225430528819561,
      "num_tokens": 90236394.0,
      "step": 39395
    },
    {
      "entropy": 5.793304920196533,
      "epoch": 3.8948200869909053,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003555789741422991,
      "loss": 4.963,
      "mean_token_accuracy": 0.22755814641714095,
      "num_tokens": 90246649.0,
      "step": 39400
    },
    {
      "entropy": 5.825546407699585,
      "epoch": 3.895314353499407,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00035554602387574105,
      "loss": 4.9811,
      "mean_token_accuracy": 0.22282395362854004,
      "num_tokens": 90258047.0,
      "step": 39405
    },
    {
      "entropy": 5.849278593063355,
      "epoch": 3.8958086200079083,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003555130716277383,
      "loss": 5.0662,
      "mean_token_accuracy": 0.21635986119508743,
      "num_tokens": 90269848.0,
      "step": 39410
    },
    {
      "entropy": 5.81041955947876,
      "epoch": 3.8963028865164095,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003554801173991016,
      "loss": 4.9991,
      "mean_token_accuracy": 0.22111031264066697,
      "num_tokens": 90281614.0,
      "step": 39415
    },
    {
      "entropy": 5.778650236129761,
      "epoch": 3.8967971530249113,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003554471611906414,
      "loss": 4.9607,
      "mean_token_accuracy": 0.22790725529193878,
      "num_tokens": 90292857.0,
      "step": 39420
    },
    {
      "entropy": 5.748271131515503,
      "epoch": 3.8972914195334125,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003554142030031685,
      "loss": 4.999,
      "mean_token_accuracy": 0.21897072046995164,
      "num_tokens": 90304558.0,
      "step": 39425
    },
    {
      "entropy": 5.764628934860229,
      "epoch": 3.8977856860419138,
      "grad_norm": 1.21875,
      "learning_rate": 0.00035538124283749387,
      "loss": 4.9448,
      "mean_token_accuracy": 0.2267860621213913,
      "num_tokens": 90315914.0,
      "step": 39430
    },
    {
      "entropy": 5.768094396591186,
      "epoch": 3.898279952550415,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003553482806944281,
      "loss": 4.8883,
      "mean_token_accuracy": 0.22569429874420166,
      "num_tokens": 90327165.0,
      "step": 39435
    },
    {
      "entropy": 5.7028614521026615,
      "epoch": 3.8987742190589163,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003553153165747822,
      "loss": 4.8732,
      "mean_token_accuracy": 0.23501939624547957,
      "num_tokens": 90338755.0,
      "step": 39440
    },
    {
      "entropy": 5.707259798049927,
      "epoch": 3.899268485567418,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000355282350479367,
      "loss": 4.9579,
      "mean_token_accuracy": 0.2260788857936859,
      "num_tokens": 90349694.0,
      "step": 39445
    },
    {
      "entropy": 5.800248384475708,
      "epoch": 3.8997627520759193,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003552493824089936,
      "loss": 4.9184,
      "mean_token_accuracy": 0.22780183106660842,
      "num_tokens": 90359604.0,
      "step": 39450
    },
    {
      "entropy": 5.6854407787323,
      "epoch": 3.900257018584421,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00035521641236447297,
      "loss": 4.8117,
      "mean_token_accuracy": 0.23606300801038743,
      "num_tokens": 90370607.0,
      "step": 39455
    },
    {
      "entropy": 5.712741947174072,
      "epoch": 3.9007512850929222,
      "grad_norm": 1.359375,
      "learning_rate": 0.00035518344034661607,
      "loss": 4.8712,
      "mean_token_accuracy": 0.2347013309597969,
      "num_tokens": 90381429.0,
      "step": 39460
    },
    {
      "entropy": 5.734273147583008,
      "epoch": 3.9012455516014235,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00035515046635623416,
      "loss": 4.9244,
      "mean_token_accuracy": 0.23089791983366012,
      "num_tokens": 90393550.0,
      "step": 39465
    },
    {
      "entropy": 5.738218116760254,
      "epoch": 3.9017398181099248,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003551174903941383,
      "loss": 4.9969,
      "mean_token_accuracy": 0.2252648264169693,
      "num_tokens": 90404380.0,
      "step": 39470
    },
    {
      "entropy": 5.807044887542725,
      "epoch": 3.902234084618426,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003550845124611397,
      "loss": 4.9492,
      "mean_token_accuracy": 0.2220213681459427,
      "num_tokens": 90414374.0,
      "step": 39475
    },
    {
      "entropy": 5.754154539108276,
      "epoch": 3.9027283511269277,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003550515325580498,
      "loss": 4.9113,
      "mean_token_accuracy": 0.23424049615859985,
      "num_tokens": 90426682.0,
      "step": 39480
    },
    {
      "entropy": 5.729799270629883,
      "epoch": 3.903222617635429,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003550185506856797,
      "loss": 4.8869,
      "mean_token_accuracy": 0.23171612322330476,
      "num_tokens": 90437992.0,
      "step": 39485
    },
    {
      "entropy": 5.781531620025635,
      "epoch": 3.9037168841439303,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00035498556684484086,
      "loss": 5.0343,
      "mean_token_accuracy": 0.22096819579601287,
      "num_tokens": 90450589.0,
      "step": 39490
    },
    {
      "entropy": 5.703736352920532,
      "epoch": 3.904211150652432,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003549525810363446,
      "loss": 4.8921,
      "mean_token_accuracy": 0.2307443305850029,
      "num_tokens": 90461723.0,
      "step": 39495
    },
    {
      "entropy": 5.791468906402588,
      "epoch": 3.9047054171609332,
      "grad_norm": 1.359375,
      "learning_rate": 0.00035491959326100253,
      "loss": 4.9495,
      "mean_token_accuracy": 0.22864301651716232,
      "num_tokens": 90472972.0,
      "step": 39500
    },
    {
      "entropy": 5.807104063034058,
      "epoch": 3.9051996836694345,
      "grad_norm": 1.25,
      "learning_rate": 0.000354886603519626,
      "loss": 5.029,
      "mean_token_accuracy": 0.21685766875743867,
      "num_tokens": 90483995.0,
      "step": 39505
    },
    {
      "entropy": 5.739621686935425,
      "epoch": 3.9056939501779357,
      "grad_norm": 1.25,
      "learning_rate": 0.0003548536118130267,
      "loss": 4.8817,
      "mean_token_accuracy": 0.2336791381239891,
      "num_tokens": 90495999.0,
      "step": 39510
    },
    {
      "entropy": 5.8262029647827145,
      "epoch": 3.9061882166864375,
      "grad_norm": 1.265625,
      "learning_rate": 0.00035482061814201614,
      "loss": 5.0161,
      "mean_token_accuracy": 0.22147536426782607,
      "num_tokens": 90507906.0,
      "step": 39515
    },
    {
      "entropy": 5.756185483932495,
      "epoch": 3.9066824831949387,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00035478762250740594,
      "loss": 4.9249,
      "mean_token_accuracy": 0.23043465316295625,
      "num_tokens": 90517938.0,
      "step": 39520
    },
    {
      "entropy": 5.768194627761841,
      "epoch": 3.90717674970344,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003547546249100079,
      "loss": 4.9686,
      "mean_token_accuracy": 0.22231417000293732,
      "num_tokens": 90528163.0,
      "step": 39525
    },
    {
      "entropy": 5.752123641967773,
      "epoch": 3.9076710162119417,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00035472162535063375,
      "loss": 4.9539,
      "mean_token_accuracy": 0.22275971621274948,
      "num_tokens": 90539846.0,
      "step": 39530
    },
    {
      "entropy": 5.7405637264251705,
      "epoch": 3.908165282720443,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00035468862383009514,
      "loss": 4.8615,
      "mean_token_accuracy": 0.23689231276512146,
      "num_tokens": 90550290.0,
      "step": 39535
    },
    {
      "entropy": 5.771781539916992,
      "epoch": 3.908659549228944,
      "grad_norm": 1.28125,
      "learning_rate": 0.00035465562034920413,
      "loss": 4.9277,
      "mean_token_accuracy": 0.23283457309007644,
      "num_tokens": 90561052.0,
      "step": 39540
    },
    {
      "entropy": 5.771683216094971,
      "epoch": 3.9091538157374455,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003546226149087724,
      "loss": 5.01,
      "mean_token_accuracy": 0.22289457321166992,
      "num_tokens": 90572007.0,
      "step": 39545
    },
    {
      "entropy": 5.772237825393677,
      "epoch": 3.909648082245947,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00035458960750961205,
      "loss": 4.9938,
      "mean_token_accuracy": 0.22165055125951766,
      "num_tokens": 90583538.0,
      "step": 39550
    },
    {
      "entropy": 5.77426700592041,
      "epoch": 3.9101423487544484,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000354556598152535,
      "loss": 4.9536,
      "mean_token_accuracy": 0.23117521405220032,
      "num_tokens": 90595464.0,
      "step": 39555
    },
    {
      "entropy": 5.8065886974334715,
      "epoch": 3.9106366152629497,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003545235868383533,
      "loss": 5.0243,
      "mean_token_accuracy": 0.22450612634420394,
      "num_tokens": 90607593.0,
      "step": 39560
    },
    {
      "entropy": 5.872856283187867,
      "epoch": 3.9111308817714514,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000354490573567879,
      "loss": 5.0475,
      "mean_token_accuracy": 0.21648077219724654,
      "num_tokens": 90618779.0,
      "step": 39565
    },
    {
      "entropy": 5.801626396179199,
      "epoch": 3.9116251482799527,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035445755834192426,
      "loss": 4.9726,
      "mean_token_accuracy": 0.22345806509256363,
      "num_tokens": 90630355.0,
      "step": 39570
    },
    {
      "entropy": 5.8214026927948,
      "epoch": 3.912119414788454,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003544245411613013,
      "loss": 4.9972,
      "mean_token_accuracy": 0.2175506293773651,
      "num_tokens": 90642042.0,
      "step": 39575
    },
    {
      "entropy": 5.78690767288208,
      "epoch": 3.912613681296955,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003543915220268223,
      "loss": 4.9395,
      "mean_token_accuracy": 0.2172304391860962,
      "num_tokens": 90653659.0,
      "step": 39580
    },
    {
      "entropy": 5.752219343185425,
      "epoch": 3.9131079478054565,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003543585009392996,
      "loss": 4.9002,
      "mean_token_accuracy": 0.23587049990892411,
      "num_tokens": 90664525.0,
      "step": 39585
    },
    {
      "entropy": 5.731579065322876,
      "epoch": 3.913602214313958,
      "grad_norm": 1.25,
      "learning_rate": 0.0003543254778995453,
      "loss": 4.9841,
      "mean_token_accuracy": 0.21812761723995208,
      "num_tokens": 90676231.0,
      "step": 39590
    },
    {
      "entropy": 5.821331787109375,
      "epoch": 3.9140964808224594,
      "grad_norm": 1.296875,
      "learning_rate": 0.00035429245290837197,
      "loss": 5.0274,
      "mean_token_accuracy": 0.20846595317125322,
      "num_tokens": 90686996.0,
      "step": 39595
    },
    {
      "entropy": 5.827587556838989,
      "epoch": 3.914590747330961,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00035425942596659213,
      "loss": 4.9961,
      "mean_token_accuracy": 0.22124552428722383,
      "num_tokens": 90699329.0,
      "step": 39600
    },
    {
      "entropy": 5.738008737564087,
      "epoch": 3.9150850138394624,
      "grad_norm": 1.234375,
      "learning_rate": 0.000354226397075018,
      "loss": 4.9556,
      "mean_token_accuracy": 0.23012987673282623,
      "num_tokens": 90710835.0,
      "step": 39605
    },
    {
      "entropy": 5.702221870422363,
      "epoch": 3.9155792803479637,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003541933662344623,
      "loss": 4.8781,
      "mean_token_accuracy": 0.23469143509864807,
      "num_tokens": 90722793.0,
      "step": 39610
    },
    {
      "entropy": 5.776153087615967,
      "epoch": 3.916073546856465,
      "grad_norm": 1.203125,
      "learning_rate": 0.00035416033344573745,
      "loss": 4.9882,
      "mean_token_accuracy": 0.22272575944662093,
      "num_tokens": 90733380.0,
      "step": 39615
    },
    {
      "entropy": 5.789487314224243,
      "epoch": 3.916567813364966,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00035412729870965606,
      "loss": 5.032,
      "mean_token_accuracy": 0.2201627939939499,
      "num_tokens": 90744801.0,
      "step": 39620
    },
    {
      "entropy": 5.783236503601074,
      "epoch": 3.917062079873468,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00035409426202703093,
      "loss": 4.9583,
      "mean_token_accuracy": 0.2277090683579445,
      "num_tokens": 90755826.0,
      "step": 39625
    },
    {
      "entropy": 5.7975671768188475,
      "epoch": 3.917556346381969,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003540612233986747,
      "loss": 4.9482,
      "mean_token_accuracy": 0.22587441354990007,
      "num_tokens": 90766943.0,
      "step": 39630
    },
    {
      "entropy": 5.848583984375,
      "epoch": 3.9180506128904704,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00035402818282540005,
      "loss": 5.1684,
      "mean_token_accuracy": 0.21577582359313965,
      "num_tokens": 90780413.0,
      "step": 39635
    },
    {
      "entropy": 5.80862455368042,
      "epoch": 3.918544879398972,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00035399514030801983,
      "loss": 4.9706,
      "mean_token_accuracy": 0.22604450285434724,
      "num_tokens": 90792923.0,
      "step": 39640
    },
    {
      "entropy": 5.727105951309204,
      "epoch": 3.9190391459074734,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00035396209584734697,
      "loss": 4.885,
      "mean_token_accuracy": 0.22646315395832062,
      "num_tokens": 90803457.0,
      "step": 39645
    },
    {
      "entropy": 5.75400218963623,
      "epoch": 3.9195334124159746,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00035392904944419424,
      "loss": 4.9205,
      "mean_token_accuracy": 0.22814578860998153,
      "num_tokens": 90815499.0,
      "step": 39650
    },
    {
      "entropy": 5.773372840881348,
      "epoch": 3.920027678924476,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003538960010993747,
      "loss": 4.9395,
      "mean_token_accuracy": 0.22199928760528564,
      "num_tokens": 90827263.0,
      "step": 39655
    },
    {
      "entropy": 5.810328340530395,
      "epoch": 3.9205219454329776,
      "grad_norm": 1.328125,
      "learning_rate": 0.00035386295081370125,
      "loss": 4.9386,
      "mean_token_accuracy": 0.23436135202646255,
      "num_tokens": 90837804.0,
      "step": 39660
    },
    {
      "entropy": 5.7375880718231205,
      "epoch": 3.921016211941479,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00035382989858798695,
      "loss": 4.9852,
      "mean_token_accuracy": 0.23022964149713515,
      "num_tokens": 90849331.0,
      "step": 39665
    },
    {
      "entropy": 5.7507225513458256,
      "epoch": 3.92151047844998,
      "grad_norm": 1.25,
      "learning_rate": 0.00035379684442304497,
      "loss": 4.9931,
      "mean_token_accuracy": 0.22130931913852692,
      "num_tokens": 90861164.0,
      "step": 39670
    },
    {
      "entropy": 5.817691373825073,
      "epoch": 3.922004744958482,
      "grad_norm": 1.25,
      "learning_rate": 0.00035376378831968844,
      "loss": 5.0939,
      "mean_token_accuracy": 0.20615292638540267,
      "num_tokens": 90872453.0,
      "step": 39675
    },
    {
      "entropy": 5.691192960739135,
      "epoch": 3.922499011466983,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003537307302787304,
      "loss": 4.8879,
      "mean_token_accuracy": 0.2311477079987526,
      "num_tokens": 90882111.0,
      "step": 39680
    },
    {
      "entropy": 5.737515544891357,
      "epoch": 3.9229932779754844,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00035369767030098424,
      "loss": 4.9697,
      "mean_token_accuracy": 0.22839185744524002,
      "num_tokens": 90894501.0,
      "step": 39685
    },
    {
      "entropy": 5.7266960620880125,
      "epoch": 3.9234875444839856,
      "grad_norm": 1.234375,
      "learning_rate": 0.00035366460838726316,
      "loss": 4.8439,
      "mean_token_accuracy": 0.2304357260465622,
      "num_tokens": 90905629.0,
      "step": 39690
    },
    {
      "entropy": 5.732442998886109,
      "epoch": 3.923981810992487,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003536315445383805,
      "loss": 4.957,
      "mean_token_accuracy": 0.2217630684375763,
      "num_tokens": 90916710.0,
      "step": 39695
    },
    {
      "entropy": 5.744244527816773,
      "epoch": 3.9244760775009886,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003535984787551497,
      "loss": 4.9475,
      "mean_token_accuracy": 0.2251925840973854,
      "num_tokens": 90927463.0,
      "step": 39700
    },
    {
      "entropy": 5.795710611343384,
      "epoch": 3.92497034400949,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003535654110383841,
      "loss": 4.9312,
      "mean_token_accuracy": 0.2223491832613945,
      "num_tokens": 90938387.0,
      "step": 39705
    },
    {
      "entropy": 5.753119325637817,
      "epoch": 3.9254646105179916,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003535323413888972,
      "loss": 4.9525,
      "mean_token_accuracy": 0.23341279625892639,
      "num_tokens": 90950326.0,
      "step": 39710
    },
    {
      "entropy": 5.796470975875854,
      "epoch": 3.925958877026493,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003534992698075025,
      "loss": 4.9816,
      "mean_token_accuracy": 0.22603694647550582,
      "num_tokens": 90962598.0,
      "step": 39715
    },
    {
      "entropy": 5.780059289932251,
      "epoch": 3.926453143534994,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003534661962950136,
      "loss": 4.9949,
      "mean_token_accuracy": 0.220823971927166,
      "num_tokens": 90975048.0,
      "step": 39720
    },
    {
      "entropy": 5.788070964813232,
      "epoch": 3.9269474100434953,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00035343312085224407,
      "loss": 4.9686,
      "mean_token_accuracy": 0.23019943684339522,
      "num_tokens": 90985768.0,
      "step": 39725
    },
    {
      "entropy": 5.807478666305542,
      "epoch": 3.9274416765519966,
      "grad_norm": 1.21875,
      "learning_rate": 0.00035340004348000765,
      "loss": 5.0031,
      "mean_token_accuracy": 0.21751173287630082,
      "num_tokens": 90998448.0,
      "step": 39730
    },
    {
      "entropy": 5.779744625091553,
      "epoch": 3.9279359430604983,
      "grad_norm": 1.3125,
      "learning_rate": 0.000353366964179118,
      "loss": 4.9919,
      "mean_token_accuracy": 0.2238590657711029,
      "num_tokens": 91010463.0,
      "step": 39735
    },
    {
      "entropy": 5.812977361679077,
      "epoch": 3.9284302095689996,
      "grad_norm": 1.25,
      "learning_rate": 0.0003533338829503888,
      "loss": 4.8949,
      "mean_token_accuracy": 0.2244887501001358,
      "num_tokens": 91022188.0,
      "step": 39740
    },
    {
      "entropy": 5.7933777332305905,
      "epoch": 3.928924476077501,
      "grad_norm": 1.28125,
      "learning_rate": 0.00035330079979463397,
      "loss": 4.9805,
      "mean_token_accuracy": 0.2278628706932068,
      "num_tokens": 91033272.0,
      "step": 39745
    },
    {
      "entropy": 5.713204097747803,
      "epoch": 3.9294187425860025,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00035326771471266735,
      "loss": 4.9702,
      "mean_token_accuracy": 0.22614417523145675,
      "num_tokens": 91044014.0,
      "step": 39750
    },
    {
      "entropy": 5.744575166702271,
      "epoch": 3.929913009094504,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003532346277053028,
      "loss": 4.9412,
      "mean_token_accuracy": 0.22631218582391738,
      "num_tokens": 91055601.0,
      "step": 39755
    },
    {
      "entropy": 5.702720594406128,
      "epoch": 3.930407275603005,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00035320153877335415,
      "loss": 4.9077,
      "mean_token_accuracy": 0.23022176027297975,
      "num_tokens": 91066384.0,
      "step": 39760
    },
    {
      "entropy": 5.732511520385742,
      "epoch": 3.9309015421115063,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003531684479176356,
      "loss": 4.8963,
      "mean_token_accuracy": 0.23080967962741852,
      "num_tokens": 91076675.0,
      "step": 39765
    },
    {
      "entropy": 5.755775308609008,
      "epoch": 3.931395808620008,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003531353551389611,
      "loss": 4.9685,
      "mean_token_accuracy": 0.2287505641579628,
      "num_tokens": 91088900.0,
      "step": 39770
    },
    {
      "entropy": 5.806204414367675,
      "epoch": 3.9318900751285093,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003531022604381448,
      "loss": 4.8769,
      "mean_token_accuracy": 0.23561695367097854,
      "num_tokens": 91100483.0,
      "step": 39775
    },
    {
      "entropy": 5.814331436157227,
      "epoch": 3.9323843416370106,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003530691638160007,
      "loss": 5.0635,
      "mean_token_accuracy": 0.2156870886683464,
      "num_tokens": 91112310.0,
      "step": 39780
    },
    {
      "entropy": 5.782642793655396,
      "epoch": 3.9328786081455123,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00035303606527334314,
      "loss": 4.9817,
      "mean_token_accuracy": 0.22016711980104448,
      "num_tokens": 91124086.0,
      "step": 39785
    },
    {
      "entropy": 5.69743013381958,
      "epoch": 3.9333728746540135,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003530029648109862,
      "loss": 4.9136,
      "mean_token_accuracy": 0.2258111596107483,
      "num_tokens": 91135044.0,
      "step": 39790
    },
    {
      "entropy": 5.782300710678101,
      "epoch": 3.933867141162515,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003529698624297442,
      "loss": 4.9747,
      "mean_token_accuracy": 0.22447331249713898,
      "num_tokens": 91145606.0,
      "step": 39795
    },
    {
      "entropy": 5.756932401657105,
      "epoch": 3.934361407671016,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003529367581304316,
      "loss": 4.9326,
      "mean_token_accuracy": 0.23109450191259384,
      "num_tokens": 91155709.0,
      "step": 39800
    },
    {
      "entropy": 5.734607934951782,
      "epoch": 3.9348556741795178,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003529036519138626,
      "loss": 4.9159,
      "mean_token_accuracy": 0.23251646608114243,
      "num_tokens": 91166511.0,
      "step": 39805
    },
    {
      "entropy": 5.75219693183899,
      "epoch": 3.935349940688019,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003528705437808516,
      "loss": 4.9118,
      "mean_token_accuracy": 0.22588907927274704,
      "num_tokens": 91178441.0,
      "step": 39810
    },
    {
      "entropy": 5.818978977203369,
      "epoch": 3.9358442071965203,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00035283743373221323,
      "loss": 5.042,
      "mean_token_accuracy": 0.2196223497390747,
      "num_tokens": 91189958.0,
      "step": 39815
    },
    {
      "entropy": 5.786458206176758,
      "epoch": 3.936338473705022,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000352804321768762,
      "loss": 4.9464,
      "mean_token_accuracy": 0.22719651013612746,
      "num_tokens": 91201461.0,
      "step": 39820
    },
    {
      "entropy": 5.736437273025513,
      "epoch": 3.9368327402135233,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00035277120789131217,
      "loss": 4.9009,
      "mean_token_accuracy": 0.22953283786773682,
      "num_tokens": 91212626.0,
      "step": 39825
    },
    {
      "entropy": 5.7448506355285645,
      "epoch": 3.9373270067220245,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003527380921006787,
      "loss": 4.938,
      "mean_token_accuracy": 0.22615545392036437,
      "num_tokens": 91224425.0,
      "step": 39830
    },
    {
      "entropy": 5.720529460906983,
      "epoch": 3.9378212732305258,
      "grad_norm": 1.21875,
      "learning_rate": 0.00035270497439767606,
      "loss": 4.8623,
      "mean_token_accuracy": 0.2317921668291092,
      "num_tokens": 91234989.0,
      "step": 39835
    },
    {
      "entropy": 5.736161422729492,
      "epoch": 3.938315539739027,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00035267185478311895,
      "loss": 4.8755,
      "mean_token_accuracy": 0.23517173230648042,
      "num_tokens": 91245713.0,
      "step": 39840
    },
    {
      "entropy": 5.742568635940552,
      "epoch": 3.9388098062475287,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003526387332578223,
      "loss": 4.918,
      "mean_token_accuracy": 0.2307419240474701,
      "num_tokens": 91256719.0,
      "step": 39845
    },
    {
      "entropy": 5.768957948684692,
      "epoch": 3.93930407275603,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003526056098226007,
      "loss": 5.0455,
      "mean_token_accuracy": 0.2200557678937912,
      "num_tokens": 91268167.0,
      "step": 39850
    },
    {
      "entropy": 5.793686962127685,
      "epoch": 3.9397983392645317,
      "grad_norm": 1.34375,
      "learning_rate": 0.000352572484478269,
      "loss": 5.0148,
      "mean_token_accuracy": 0.21721922755241393,
      "num_tokens": 91280182.0,
      "step": 39855
    },
    {
      "entropy": 5.791457605361939,
      "epoch": 3.940292605773033,
      "grad_norm": 1.234375,
      "learning_rate": 0.00035253935722564215,
      "loss": 5.0127,
      "mean_token_accuracy": 0.2197528213262558,
      "num_tokens": 91291755.0,
      "step": 39860
    },
    {
      "entropy": 5.769615745544433,
      "epoch": 3.9407868722815342,
      "grad_norm": 1.25,
      "learning_rate": 0.00035250622806553525,
      "loss": 4.886,
      "mean_token_accuracy": 0.23325519412755966,
      "num_tokens": 91302389.0,
      "step": 39865
    },
    {
      "entropy": 5.789990949630737,
      "epoch": 3.9412811387900355,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00035247309699876293,
      "loss": 4.9493,
      "mean_token_accuracy": 0.23177388608455657,
      "num_tokens": 91313354.0,
      "step": 39870
    },
    {
      "entropy": 5.7733869552612305,
      "epoch": 3.9417754052985368,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003524399640261404,
      "loss": 5.0637,
      "mean_token_accuracy": 0.21744525730609893,
      "num_tokens": 91325939.0,
      "step": 39875
    },
    {
      "entropy": 5.763187503814697,
      "epoch": 3.9422696718070385,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003524068291484829,
      "loss": 4.9646,
      "mean_token_accuracy": 0.22696375697851182,
      "num_tokens": 91336637.0,
      "step": 39880
    },
    {
      "entropy": 5.759377908706665,
      "epoch": 3.9427639383155397,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00035237369236660525,
      "loss": 4.9759,
      "mean_token_accuracy": 0.22297230511903762,
      "num_tokens": 91348058.0,
      "step": 39885
    },
    {
      "entropy": 5.751921463012695,
      "epoch": 3.943258204824041,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003523405536813229,
      "loss": 4.893,
      "mean_token_accuracy": 0.23175987154245375,
      "num_tokens": 91358630.0,
      "step": 39890
    },
    {
      "entropy": 5.792415904998779,
      "epoch": 3.9437524713325427,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003523074130934509,
      "loss": 4.9523,
      "mean_token_accuracy": 0.2214099258184433,
      "num_tokens": 91369989.0,
      "step": 39895
    },
    {
      "entropy": 5.773793363571167,
      "epoch": 3.944246737841044,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00035227427060380436,
      "loss": 5.0023,
      "mean_token_accuracy": 0.21726160049438475,
      "num_tokens": 91382145.0,
      "step": 39900
    },
    {
      "entropy": 5.790749359130859,
      "epoch": 3.944741004349545,
      "grad_norm": 1.3125,
      "learning_rate": 0.000352241126213199,
      "loss": 4.8955,
      "mean_token_accuracy": 0.23027364760637284,
      "num_tokens": 91392952.0,
      "step": 39905
    },
    {
      "entropy": 5.751145124435425,
      "epoch": 3.9452352708580465,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003522079799224499,
      "loss": 4.9776,
      "mean_token_accuracy": 0.22604178339242936,
      "num_tokens": 91404050.0,
      "step": 39910
    },
    {
      "entropy": 5.684496736526489,
      "epoch": 3.945729537366548,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003521748317323724,
      "loss": 4.8918,
      "mean_token_accuracy": 0.22975898236036302,
      "num_tokens": 91414326.0,
      "step": 39915
    },
    {
      "entropy": 5.834219026565552,
      "epoch": 3.9462238038750495,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003521416816437822,
      "loss": 5.0066,
      "mean_token_accuracy": 0.22006482183933257,
      "num_tokens": 91426774.0,
      "step": 39920
    },
    {
      "entropy": 5.790394401550293,
      "epoch": 3.9467180703835507,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003521085296574945,
      "loss": 4.9433,
      "mean_token_accuracy": 0.22705452889204025,
      "num_tokens": 91438303.0,
      "step": 39925
    },
    {
      "entropy": 5.804960918426514,
      "epoch": 3.9472123368920524,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00035207537577432516,
      "loss": 4.9646,
      "mean_token_accuracy": 0.2261754646897316,
      "num_tokens": 91449310.0,
      "step": 39930
    },
    {
      "entropy": 5.765598583221435,
      "epoch": 3.9477066034005537,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003520422199950896,
      "loss": 4.9401,
      "mean_token_accuracy": 0.22558145523071288,
      "num_tokens": 91460665.0,
      "step": 39935
    },
    {
      "entropy": 5.827822875976563,
      "epoch": 3.948200869909055,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003520090623206034,
      "loss": 4.9829,
      "mean_token_accuracy": 0.22303509563207627,
      "num_tokens": 91471111.0,
      "step": 39940
    },
    {
      "entropy": 5.680939960479736,
      "epoch": 3.948695136417556,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00035197590275168247,
      "loss": 4.8912,
      "mean_token_accuracy": 0.23884239345788955,
      "num_tokens": 91483525.0,
      "step": 39945
    },
    {
      "entropy": 5.72439923286438,
      "epoch": 3.9491894029260575,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003519427412891422,
      "loss": 4.92,
      "mean_token_accuracy": 0.22762472778558732,
      "num_tokens": 91494718.0,
      "step": 39950
    },
    {
      "entropy": 5.763533878326416,
      "epoch": 3.949683669434559,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003519095779337987,
      "loss": 4.913,
      "mean_token_accuracy": 0.23371028304100036,
      "num_tokens": 91507286.0,
      "step": 39955
    },
    {
      "entropy": 5.678683614730835,
      "epoch": 3.9501779359430604,
      "grad_norm": 1.484375,
      "learning_rate": 0.00035187641268646756,
      "loss": 4.812,
      "mean_token_accuracy": 0.24543562531471252,
      "num_tokens": 91517590.0,
      "step": 39960
    },
    {
      "entropy": 5.7644916534423825,
      "epoch": 3.950672202451562,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00035184324554796484,
      "loss": 4.8567,
      "mean_token_accuracy": 0.2372788518667221,
      "num_tokens": 91527995.0,
      "step": 39965
    },
    {
      "entropy": 5.855312156677246,
      "epoch": 3.9511664689600634,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00035181007651910624,
      "loss": 4.9909,
      "mean_token_accuracy": 0.21801034957170487,
      "num_tokens": 91539457.0,
      "step": 39970
    },
    {
      "entropy": 5.72074179649353,
      "epoch": 3.9516607354685647,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003517769056007079,
      "loss": 4.9221,
      "mean_token_accuracy": 0.23197518140077591,
      "num_tokens": 91551104.0,
      "step": 39975
    },
    {
      "entropy": 5.787520122528076,
      "epoch": 3.952155001977066,
      "grad_norm": 1.3125,
      "learning_rate": 0.00035174373279358586,
      "loss": 4.9351,
      "mean_token_accuracy": 0.22299662083387375,
      "num_tokens": 91561558.0,
      "step": 39980
    },
    {
      "entropy": 5.83480396270752,
      "epoch": 3.952649268485567,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003517105580985559,
      "loss": 5.0883,
      "mean_token_accuracy": 0.2192860037088394,
      "num_tokens": 91573179.0,
      "step": 39985
    },
    {
      "entropy": 5.784835147857666,
      "epoch": 3.953143534994069,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003516773815164345,
      "loss": 4.9925,
      "mean_token_accuracy": 0.22235371619462968,
      "num_tokens": 91584973.0,
      "step": 39990
    },
    {
      "entropy": 5.787422323226929,
      "epoch": 3.95363780150257,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003516442030480375,
      "loss": 4.9153,
      "mean_token_accuracy": 0.23015111982822417,
      "num_tokens": 91596681.0,
      "step": 39995
    },
    {
      "entropy": 5.720462465286255,
      "epoch": 3.9541320680110714,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003516110226941812,
      "loss": 4.894,
      "mean_token_accuracy": 0.2240665152668953,
      "num_tokens": 91608382.0,
      "step": 40000
    },
    {
      "entropy": 5.822912693023682,
      "epoch": 3.954626334519573,
      "grad_norm": 1.203125,
      "learning_rate": 0.00035157784045568194,
      "loss": 4.9771,
      "mean_token_accuracy": 0.22176067531108856,
      "num_tokens": 91619619.0,
      "step": 40005
    },
    {
      "entropy": 5.810512113571167,
      "epoch": 3.9551206010280744,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003515446563333559,
      "loss": 5.0111,
      "mean_token_accuracy": 0.21462950855493546,
      "num_tokens": 91631889.0,
      "step": 40010
    },
    {
      "entropy": 5.8214610576629635,
      "epoch": 3.9556148675365757,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00035151147032801947,
      "loss": 5.028,
      "mean_token_accuracy": 0.22076182216405868,
      "num_tokens": 91643982.0,
      "step": 40015
    },
    {
      "entropy": 5.774022340774536,
      "epoch": 3.956109134045077,
      "grad_norm": 1.1875,
      "learning_rate": 0.00035147828244048887,
      "loss": 5.0165,
      "mean_token_accuracy": 0.22320191860198973,
      "num_tokens": 91656322.0,
      "step": 40020
    },
    {
      "entropy": 5.764347219467163,
      "epoch": 3.9566034005535786,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003514450926715808,
      "loss": 4.9919,
      "mean_token_accuracy": 0.21665906012058259,
      "num_tokens": 91669419.0,
      "step": 40025
    },
    {
      "entropy": 5.772978973388672,
      "epoch": 3.95709766706208,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003514119010221115,
      "loss": 4.8776,
      "mean_token_accuracy": 0.23673643469810485,
      "num_tokens": 91679919.0,
      "step": 40030
    },
    {
      "entropy": 5.709691429138184,
      "epoch": 3.957591933570581,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003513787074928976,
      "loss": 4.8893,
      "mean_token_accuracy": 0.23461782336235046,
      "num_tokens": 91690985.0,
      "step": 40035
    },
    {
      "entropy": 5.7271809577941895,
      "epoch": 3.958086200079083,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003513455120847557,
      "loss": 4.9325,
      "mean_token_accuracy": 0.22589402496814728,
      "num_tokens": 91702944.0,
      "step": 40040
    },
    {
      "entropy": 5.771044397354126,
      "epoch": 3.958580466587584,
      "grad_norm": 1.28125,
      "learning_rate": 0.00035131231479850226,
      "loss": 4.9172,
      "mean_token_accuracy": 0.23129124641418458,
      "num_tokens": 91713608.0,
      "step": 40045
    },
    {
      "entropy": 5.8294273853302006,
      "epoch": 3.9590747330960854,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003512791156349541,
      "loss": 4.9581,
      "mean_token_accuracy": 0.22651818841695787,
      "num_tokens": 91725970.0,
      "step": 40050
    },
    {
      "entropy": 5.714255523681641,
      "epoch": 3.9595689996045866,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003512459145949279,
      "loss": 4.8957,
      "mean_token_accuracy": 0.22749619334936141,
      "num_tokens": 91738001.0,
      "step": 40055
    },
    {
      "entropy": 5.729784631729126,
      "epoch": 3.9600632661130883,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00035121271167924033,
      "loss": 4.9115,
      "mean_token_accuracy": 0.23294657766819,
      "num_tokens": 91747050.0,
      "step": 40060
    },
    {
      "entropy": 5.73633975982666,
      "epoch": 3.9605575326215896,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003511795068887082,
      "loss": 4.9224,
      "mean_token_accuracy": 0.2277794286608696,
      "num_tokens": 91757298.0,
      "step": 40065
    },
    {
      "entropy": 5.848687410354614,
      "epoch": 3.961051799130091,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003511463002241485,
      "loss": 5.0289,
      "mean_token_accuracy": 0.21902576982975006,
      "num_tokens": 91768276.0,
      "step": 40070
    },
    {
      "entropy": 5.777990531921387,
      "epoch": 3.9615460656385926,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003511130916863779,
      "loss": 4.923,
      "mean_token_accuracy": 0.23095069378614425,
      "num_tokens": 91779186.0,
      "step": 40075
    },
    {
      "entropy": 5.742792463302612,
      "epoch": 3.962040332147094,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003510798812762134,
      "loss": 4.9856,
      "mean_token_accuracy": 0.22668840289115905,
      "num_tokens": 91790503.0,
      "step": 40080
    },
    {
      "entropy": 5.762075471878052,
      "epoch": 3.962534598655595,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003510466689944721,
      "loss": 4.9492,
      "mean_token_accuracy": 0.22801593989133834,
      "num_tokens": 91802526.0,
      "step": 40085
    },
    {
      "entropy": 5.777751016616821,
      "epoch": 3.9630288651640964,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003510134548419709,
      "loss": 4.9477,
      "mean_token_accuracy": 0.2307625502347946,
      "num_tokens": 91814116.0,
      "step": 40090
    },
    {
      "entropy": 5.771556234359741,
      "epoch": 3.9635231316725976,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003509802388195269,
      "loss": 4.9749,
      "mean_token_accuracy": 0.2300652891397476,
      "num_tokens": 91825459.0,
      "step": 40095
    },
    {
      "entropy": 5.789162063598633,
      "epoch": 3.9640173981810993,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00035094702092795734,
      "loss": 5.0309,
      "mean_token_accuracy": 0.2215588092803955,
      "num_tokens": 91837656.0,
      "step": 40100
    },
    {
      "entropy": 5.751531887054443,
      "epoch": 3.9645116646896006,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003509138011680792,
      "loss": 4.9454,
      "mean_token_accuracy": 0.2293277069926262,
      "num_tokens": 91849384.0,
      "step": 40105
    },
    {
      "entropy": 5.717943572998047,
      "epoch": 3.965005931198102,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003508805795407097,
      "loss": 4.9062,
      "mean_token_accuracy": 0.23012191206216812,
      "num_tokens": 91861508.0,
      "step": 40110
    },
    {
      "entropy": 5.802012395858765,
      "epoch": 3.9655001977066036,
      "grad_norm": 1.203125,
      "learning_rate": 0.00035084735604666627,
      "loss": 5.0512,
      "mean_token_accuracy": 0.21989835500717164,
      "num_tokens": 91874439.0,
      "step": 40115
    },
    {
      "entropy": 5.821330547332764,
      "epoch": 3.965994464215105,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003508141306867662,
      "loss": 4.9527,
      "mean_token_accuracy": 0.22307491153478623,
      "num_tokens": 91884949.0,
      "step": 40120
    },
    {
      "entropy": 5.766239309310913,
      "epoch": 3.966488730723606,
      "grad_norm": 1.25,
      "learning_rate": 0.0003507809034618267,
      "loss": 4.9407,
      "mean_token_accuracy": 0.22629584074020387,
      "num_tokens": 91895636.0,
      "step": 40125
    },
    {
      "entropy": 5.794170808792114,
      "epoch": 3.9669829972321073,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00035074767437266507,
      "loss": 4.9556,
      "mean_token_accuracy": 0.2285407453775406,
      "num_tokens": 91908392.0,
      "step": 40130
    },
    {
      "entropy": 5.81916995048523,
      "epoch": 3.967477263740609,
      "grad_norm": 1.296875,
      "learning_rate": 0.000350714443420099,
      "loss": 4.9985,
      "mean_token_accuracy": 0.22149754762649537,
      "num_tokens": 91919976.0,
      "step": 40135
    },
    {
      "entropy": 5.7858967781066895,
      "epoch": 3.9679715302491103,
      "grad_norm": 1.296875,
      "learning_rate": 0.00035068121060494596,
      "loss": 4.9416,
      "mean_token_accuracy": 0.22283742129802703,
      "num_tokens": 91930770.0,
      "step": 40140
    },
    {
      "entropy": 5.863386011123657,
      "epoch": 3.9684657967576116,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003506479759280233,
      "loss": 5.0554,
      "mean_token_accuracy": 0.21945670545101165,
      "num_tokens": 91943317.0,
      "step": 40145
    },
    {
      "entropy": 5.766076850891113,
      "epoch": 3.9689600632661133,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003506147393901487,
      "loss": 4.9403,
      "mean_token_accuracy": 0.22656757086515428,
      "num_tokens": 91954949.0,
      "step": 40150
    },
    {
      "entropy": 5.763749885559082,
      "epoch": 3.9694543297746145,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003505815009921398,
      "loss": 4.9483,
      "mean_token_accuracy": 0.22425553500652312,
      "num_tokens": 91967392.0,
      "step": 40155
    },
    {
      "entropy": 5.7620312690734865,
      "epoch": 3.969948596283116,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003505482607348142,
      "loss": 4.8925,
      "mean_token_accuracy": 0.22799201756715776,
      "num_tokens": 91979116.0,
      "step": 40160
    },
    {
      "entropy": 5.822312211990356,
      "epoch": 3.970442862791617,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003505150186189898,
      "loss": 4.9634,
      "mean_token_accuracy": 0.2221357047557831,
      "num_tokens": 91989811.0,
      "step": 40165
    },
    {
      "entropy": 5.771495723724366,
      "epoch": 3.9709371293001188,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003504817746454841,
      "loss": 4.9497,
      "mean_token_accuracy": 0.2263193815946579,
      "num_tokens": 92000043.0,
      "step": 40170
    },
    {
      "entropy": 5.780642509460449,
      "epoch": 3.97143139580862,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00035044852881511507,
      "loss": 5.0498,
      "mean_token_accuracy": 0.21753509789705278,
      "num_tokens": 92013475.0,
      "step": 40175
    },
    {
      "entropy": 5.803360843658448,
      "epoch": 3.9719256623171213,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00035041528112870053,
      "loss": 4.9984,
      "mean_token_accuracy": 0.22429601550102235,
      "num_tokens": 92025369.0,
      "step": 40180
    },
    {
      "entropy": 5.808307266235351,
      "epoch": 3.972419928825623,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003503820315870583,
      "loss": 4.9739,
      "mean_token_accuracy": 0.22106163799762726,
      "num_tokens": 92035682.0,
      "step": 40185
    },
    {
      "entropy": 5.82649507522583,
      "epoch": 3.9729141953341243,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003503487801910065,
      "loss": 4.9731,
      "mean_token_accuracy": 0.22267328202724457,
      "num_tokens": 92046735.0,
      "step": 40190
    },
    {
      "entropy": 5.753697872161865,
      "epoch": 3.9734084618426255,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000350315526941363,
      "loss": 4.9508,
      "mean_token_accuracy": 0.22687720954418183,
      "num_tokens": 92057802.0,
      "step": 40195
    },
    {
      "entropy": 5.785489845275879,
      "epoch": 3.973902728351127,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003502822718389459,
      "loss": 5.0018,
      "mean_token_accuracy": 0.22397810220718384,
      "num_tokens": 92068656.0,
      "step": 40200
    },
    {
      "entropy": 5.721517753601074,
      "epoch": 3.974396994859628,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003502490148845732,
      "loss": 4.9125,
      "mean_token_accuracy": 0.23132238239049913,
      "num_tokens": 92079403.0,
      "step": 40205
    },
    {
      "entropy": 5.783994483947754,
      "epoch": 3.9748912613681298,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00035021575607906304,
      "loss": 4.9965,
      "mean_token_accuracy": 0.21775999516248704,
      "num_tokens": 92091482.0,
      "step": 40210
    },
    {
      "entropy": 5.8313556671142575,
      "epoch": 3.975385527876631,
      "grad_norm": 1.390625,
      "learning_rate": 0.00035018249542323357,
      "loss": 5.0096,
      "mean_token_accuracy": 0.2175132766366005,
      "num_tokens": 92101906.0,
      "step": 40215
    },
    {
      "entropy": 5.820845127105713,
      "epoch": 3.9758797943851327,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003501492329179031,
      "loss": 5.038,
      "mean_token_accuracy": 0.2138381227850914,
      "num_tokens": 92114683.0,
      "step": 40220
    },
    {
      "entropy": 5.873892879486084,
      "epoch": 3.976374060893634,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003501159685638898,
      "loss": 5.0269,
      "mean_token_accuracy": 0.2177908957004547,
      "num_tokens": 92126108.0,
      "step": 40225
    },
    {
      "entropy": 5.763959884643555,
      "epoch": 3.9768683274021353,
      "grad_norm": 1.1875,
      "learning_rate": 0.000350082702362012,
      "loss": 4.963,
      "mean_token_accuracy": 0.22574483156204223,
      "num_tokens": 92136970.0,
      "step": 40230
    },
    {
      "entropy": 5.758642911911011,
      "epoch": 3.9773625939106365,
      "grad_norm": 1.25,
      "learning_rate": 0.00035004943431308816,
      "loss": 5.0494,
      "mean_token_accuracy": 0.21973459869623185,
      "num_tokens": 92148919.0,
      "step": 40235
    },
    {
      "entropy": 5.785069131851197,
      "epoch": 3.9778568604191378,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003500161644179366,
      "loss": 4.8443,
      "mean_token_accuracy": 0.2346043646335602,
      "num_tokens": 92159492.0,
      "step": 40240
    },
    {
      "entropy": 5.833597803115845,
      "epoch": 3.9783511269276395,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003499828926773757,
      "loss": 4.9774,
      "mean_token_accuracy": 0.2258749157190323,
      "num_tokens": 92171324.0,
      "step": 40245
    },
    {
      "entropy": 5.705038022994995,
      "epoch": 3.9788453934361407,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00034994961909222405,
      "loss": 4.879,
      "mean_token_accuracy": 0.22789081186056137,
      "num_tokens": 92182336.0,
      "step": 40250
    },
    {
      "entropy": 5.757192611694336,
      "epoch": 3.979339659944642,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00034991634366330005,
      "loss": 4.8848,
      "mean_token_accuracy": 0.23378842324018478,
      "num_tokens": 92193965.0,
      "step": 40255
    },
    {
      "entropy": 5.753625679016113,
      "epoch": 3.9798339264531437,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003498830663914224,
      "loss": 4.8995,
      "mean_token_accuracy": 0.2297613501548767,
      "num_tokens": 92204921.0,
      "step": 40260
    },
    {
      "entropy": 5.734578418731689,
      "epoch": 3.980328192961645,
      "grad_norm": 1.265625,
      "learning_rate": 0.00034984978727740977,
      "loss": 4.9468,
      "mean_token_accuracy": 0.22469718158245086,
      "num_tokens": 92216557.0,
      "step": 40265
    },
    {
      "entropy": 5.762670040130615,
      "epoch": 3.9808224594701462,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003498165063220808,
      "loss": 4.9166,
      "mean_token_accuracy": 0.23029980808496475,
      "num_tokens": 92227346.0,
      "step": 40270
    },
    {
      "entropy": 5.733290815353394,
      "epoch": 3.9813167259786475,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003497832235262541,
      "loss": 4.8943,
      "mean_token_accuracy": 0.23587428033351898,
      "num_tokens": 92240099.0,
      "step": 40275
    },
    {
      "entropy": 5.771744298934936,
      "epoch": 3.981810992487149,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00034974993889074846,
      "loss": 4.9863,
      "mean_token_accuracy": 0.22145727127790452,
      "num_tokens": 92251194.0,
      "step": 40280
    },
    {
      "entropy": 5.821692180633545,
      "epoch": 3.9823052589956505,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003497166524163827,
      "loss": 4.9912,
      "mean_token_accuracy": 0.22347330152988434,
      "num_tokens": 92261926.0,
      "step": 40285
    },
    {
      "entropy": 5.792262983322144,
      "epoch": 3.9827995255041517,
      "grad_norm": 1.25,
      "learning_rate": 0.00034968336410397586,
      "loss": 4.9282,
      "mean_token_accuracy": 0.23435707539319992,
      "num_tokens": 92274292.0,
      "step": 40290
    },
    {
      "entropy": 5.734842824935913,
      "epoch": 3.9832937920126534,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00034965007395434657,
      "loss": 4.9587,
      "mean_token_accuracy": 0.22517624199390412,
      "num_tokens": 92285803.0,
      "step": 40295
    },
    {
      "entropy": 5.737663316726684,
      "epoch": 3.9837880585211547,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003496167819683138,
      "loss": 4.9404,
      "mean_token_accuracy": 0.2281440481543541,
      "num_tokens": 92297729.0,
      "step": 40300
    },
    {
      "entropy": 5.772334098815918,
      "epoch": 3.984282325029656,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003495834881466967,
      "loss": 4.9433,
      "mean_token_accuracy": 0.23039780408143998,
      "num_tokens": 92309215.0,
      "step": 40305
    },
    {
      "entropy": 5.7725506782531735,
      "epoch": 3.984776591538157,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003495501924903143,
      "loss": 4.9627,
      "mean_token_accuracy": 0.2204370990395546,
      "num_tokens": 92320320.0,
      "step": 40310
    },
    {
      "entropy": 5.781480836868286,
      "epoch": 3.9852708580466585,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003495168949999855,
      "loss": 4.9835,
      "mean_token_accuracy": 0.2208327829837799,
      "num_tokens": 92333054.0,
      "step": 40315
    },
    {
      "entropy": 5.699251174926758,
      "epoch": 3.98576512455516,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003494835956765295,
      "loss": 4.9092,
      "mean_token_accuracy": 0.22790580242872238,
      "num_tokens": 92345558.0,
      "step": 40320
    },
    {
      "entropy": 5.751389694213867,
      "epoch": 3.9862593910636615,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003494502945207656,
      "loss": 4.9092,
      "mean_token_accuracy": 0.22479078471660613,
      "num_tokens": 92356782.0,
      "step": 40325
    },
    {
      "entropy": 5.84130449295044,
      "epoch": 3.986753657572163,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00034941699153351286,
      "loss": 5.0467,
      "mean_token_accuracy": 0.22326316237449645,
      "num_tokens": 92368600.0,
      "step": 40330
    },
    {
      "entropy": 5.786207675933838,
      "epoch": 3.9872479240806644,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003493836867155905,
      "loss": 4.9192,
      "mean_token_accuracy": 0.22779793590307235,
      "num_tokens": 92380694.0,
      "step": 40335
    },
    {
      "entropy": 5.742025232315063,
      "epoch": 3.9877421905891657,
      "grad_norm": 1.125,
      "learning_rate": 0.00034935038006781805,
      "loss": 4.9451,
      "mean_token_accuracy": 0.22942596822977065,
      "num_tokens": 92392518.0,
      "step": 40340
    },
    {
      "entropy": 5.745583248138428,
      "epoch": 3.988236457097667,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00034931707159101467,
      "loss": 4.9769,
      "mean_token_accuracy": 0.22393186390399933,
      "num_tokens": 92403631.0,
      "step": 40345
    },
    {
      "entropy": 5.88181209564209,
      "epoch": 3.988730723606168,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003492837612859997,
      "loss": 5.1564,
      "mean_token_accuracy": 0.20926500707864762,
      "num_tokens": 92414606.0,
      "step": 40350
    },
    {
      "entropy": 5.7603480339050295,
      "epoch": 3.98922499011467,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003492504491535928,
      "loss": 4.905,
      "mean_token_accuracy": 0.2368936851620674,
      "num_tokens": 92425384.0,
      "step": 40355
    },
    {
      "entropy": 5.746561288833618,
      "epoch": 3.989719256623171,
      "grad_norm": 1.25,
      "learning_rate": 0.0003492171351946133,
      "loss": 4.8662,
      "mean_token_accuracy": 0.23603501319885253,
      "num_tokens": 92436179.0,
      "step": 40360
    },
    {
      "entropy": 5.672228717803955,
      "epoch": 3.9902135231316724,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00034918381940988075,
      "loss": 4.9061,
      "mean_token_accuracy": 0.2369284674525261,
      "num_tokens": 92447615.0,
      "step": 40365
    },
    {
      "entropy": 5.743786144256592,
      "epoch": 3.990707789640174,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003491505018002147,
      "loss": 4.9611,
      "mean_token_accuracy": 0.22974384874105452,
      "num_tokens": 92458900.0,
      "step": 40370
    },
    {
      "entropy": 5.747843027114868,
      "epoch": 3.9912020561486754,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003491171823664348,
      "loss": 4.8705,
      "mean_token_accuracy": 0.23919363170862198,
      "num_tokens": 92470065.0,
      "step": 40375
    },
    {
      "entropy": 5.866286516189575,
      "epoch": 3.9916963226571767,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00034908386110936077,
      "loss": 5.0603,
      "mean_token_accuracy": 0.2202727258205414,
      "num_tokens": 92482157.0,
      "step": 40380
    },
    {
      "entropy": 5.754137992858887,
      "epoch": 3.992190589165678,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003490505380298122,
      "loss": 5.0023,
      "mean_token_accuracy": 0.21203076839447021,
      "num_tokens": 92496438.0,
      "step": 40385
    },
    {
      "entropy": 5.72451605796814,
      "epoch": 3.9926848556741796,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000349017213128609,
      "loss": 4.8854,
      "mean_token_accuracy": 0.23598654121160506,
      "num_tokens": 92507832.0,
      "step": 40390
    },
    {
      "entropy": 5.75435094833374,
      "epoch": 3.993179122182681,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003489838864065708,
      "loss": 4.9257,
      "mean_token_accuracy": 0.2287950873374939,
      "num_tokens": 92519465.0,
      "step": 40395
    },
    {
      "entropy": 5.743924808502197,
      "epoch": 3.993673388691182,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003489505578645175,
      "loss": 4.9734,
      "mean_token_accuracy": 0.21963551193475722,
      "num_tokens": 92531051.0,
      "step": 40400
    },
    {
      "entropy": 5.740170431137085,
      "epoch": 3.994167655199684,
      "grad_norm": 1.203125,
      "learning_rate": 0.00034891722750326907,
      "loss": 4.9826,
      "mean_token_accuracy": 0.22334738373756408,
      "num_tokens": 92543654.0,
      "step": 40405
    },
    {
      "entropy": 5.8207605361938475,
      "epoch": 3.994661921708185,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00034888389532364533,
      "loss": 4.9824,
      "mean_token_accuracy": 0.225418059527874,
      "num_tokens": 92554126.0,
      "step": 40410
    },
    {
      "entropy": 5.807209396362305,
      "epoch": 3.9951561882166864,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00034885056132646637,
      "loss": 5.0786,
      "mean_token_accuracy": 0.21850505322217942,
      "num_tokens": 92567205.0,
      "step": 40415
    },
    {
      "entropy": 5.778727054595947,
      "epoch": 3.9956504547251877,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00034881722551255206,
      "loss": 4.9405,
      "mean_token_accuracy": 0.22961635142564774,
      "num_tokens": 92579402.0,
      "step": 40420
    },
    {
      "entropy": 5.82212266921997,
      "epoch": 3.9961447212336894,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003487838878827225,
      "loss": 4.9814,
      "mean_token_accuracy": 0.2220032572746277,
      "num_tokens": 92592514.0,
      "step": 40425
    },
    {
      "entropy": 5.777477264404297,
      "epoch": 3.9966389877421906,
      "grad_norm": 1.265625,
      "learning_rate": 0.00034875054843779804,
      "loss": 4.9444,
      "mean_token_accuracy": 0.22618235051631927,
      "num_tokens": 92604571.0,
      "step": 40430
    },
    {
      "entropy": 5.724583482742309,
      "epoch": 3.997133254250692,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003487172071785985,
      "loss": 4.9916,
      "mean_token_accuracy": 0.2290568083524704,
      "num_tokens": 92616867.0,
      "step": 40435
    },
    {
      "entropy": 5.758102321624756,
      "epoch": 3.9976275207591936,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003486838641059443,
      "loss": 4.9523,
      "mean_token_accuracy": 0.2272259071469307,
      "num_tokens": 92628958.0,
      "step": 40440
    },
    {
      "entropy": 5.806486463546753,
      "epoch": 3.998121787267695,
      "grad_norm": 1.25,
      "learning_rate": 0.00034865051922065556,
      "loss": 4.9863,
      "mean_token_accuracy": 0.22635818719863893,
      "num_tokens": 92639076.0,
      "step": 40445
    },
    {
      "entropy": 5.850391244888305,
      "epoch": 3.998616053776196,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00034861717252355264,
      "loss": 5.0205,
      "mean_token_accuracy": 0.22004858255386353,
      "num_tokens": 92650824.0,
      "step": 40450
    },
    {
      "entropy": 5.801622819900513,
      "epoch": 3.9991103202846974,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003485838240154559,
      "loss": 5.0758,
      "mean_token_accuracy": 0.2143637418746948,
      "num_tokens": 92662371.0,
      "step": 40455
    },
    {
      "entropy": 5.839067077636718,
      "epoch": 3.9996045867931986,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00034855047369718563,
      "loss": 5.0721,
      "mean_token_accuracy": 0.21548788249492645,
      "num_tokens": 92673604.0,
      "step": 40460
    },
    {
      "entropy": 5.785769891738892,
      "epoch": 4.0000988533017,
      "grad_norm": 1.125,
      "learning_rate": 0.00034851712156956226,
      "loss": 4.9656,
      "mean_token_accuracy": 0.218556709587574,
      "num_tokens": 92684130.0,
      "step": 40465
    },
    {
      "entropy": 5.762605953216553,
      "epoch": 4.000593119810202,
      "grad_norm": 1.25,
      "learning_rate": 0.00034848376763340635,
      "loss": 4.881,
      "mean_token_accuracy": 0.22869669944047927,
      "num_tokens": 92694625.0,
      "step": 40470
    },
    {
      "entropy": 5.761461544036865,
      "epoch": 4.001087386318703,
      "grad_norm": 1.296875,
      "learning_rate": 0.00034845041188953836,
      "loss": 4.9901,
      "mean_token_accuracy": 0.2163677215576172,
      "num_tokens": 92705243.0,
      "step": 40475
    },
    {
      "entropy": 5.712785053253174,
      "epoch": 4.001581652827205,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003484170543387788,
      "loss": 4.8023,
      "mean_token_accuracy": 0.24021390229463577,
      "num_tokens": 92716064.0,
      "step": 40480
    },
    {
      "entropy": 5.824042749404907,
      "epoch": 4.002075919335706,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003483836949819483,
      "loss": 4.9341,
      "mean_token_accuracy": 0.22667334824800492,
      "num_tokens": 92726910.0,
      "step": 40485
    },
    {
      "entropy": 5.803349924087525,
      "epoch": 4.002570185844207,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00034835033381986757,
      "loss": 4.9713,
      "mean_token_accuracy": 0.224851331114769,
      "num_tokens": 92738847.0,
      "step": 40490
    },
    {
      "entropy": 5.7535951137542725,
      "epoch": 4.003064452352708,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003483169708533572,
      "loss": 4.9799,
      "mean_token_accuracy": 0.22277723848819733,
      "num_tokens": 92750471.0,
      "step": 40495
    },
    {
      "entropy": 5.709720087051392,
      "epoch": 4.00355871886121,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003482836060832381,
      "loss": 4.8768,
      "mean_token_accuracy": 0.23464414477348328,
      "num_tokens": 92762619.0,
      "step": 40500
    },
    {
      "entropy": 5.702446556091308,
      "epoch": 4.004052985369712,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003482502395103309,
      "loss": 4.8284,
      "mean_token_accuracy": 0.23986173719167708,
      "num_tokens": 92774308.0,
      "step": 40505
    },
    {
      "entropy": 5.783125019073486,
      "epoch": 4.004547251878213,
      "grad_norm": 1.234375,
      "learning_rate": 0.00034821687113545624,
      "loss": 5.0138,
      "mean_token_accuracy": 0.21683288663625716,
      "num_tokens": 92784408.0,
      "step": 40510
    },
    {
      "entropy": 5.805502843856812,
      "epoch": 4.005041518386714,
      "grad_norm": 1.296875,
      "learning_rate": 0.00034818350095943535,
      "loss": 4.9917,
      "mean_token_accuracy": 0.2325211927294731,
      "num_tokens": 92795615.0,
      "step": 40515
    },
    {
      "entropy": 5.814350032806397,
      "epoch": 4.005535784895216,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00034815012898308894,
      "loss": 4.9934,
      "mean_token_accuracy": 0.2276730492711067,
      "num_tokens": 92807850.0,
      "step": 40520
    },
    {
      "entropy": 5.881468486785889,
      "epoch": 4.006030051403717,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000348116755207238,
      "loss": 5.0107,
      "mean_token_accuracy": 0.2245783507823944,
      "num_tokens": 92818462.0,
      "step": 40525
    },
    {
      "entropy": 5.776654672622681,
      "epoch": 4.006524317912218,
      "grad_norm": 1.15625,
      "learning_rate": 0.00034808337963270364,
      "loss": 4.8652,
      "mean_token_accuracy": 0.23501743525266647,
      "num_tokens": 92830716.0,
      "step": 40530
    },
    {
      "entropy": 5.815789651870728,
      "epoch": 4.007018584420719,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00034805000226030677,
      "loss": 4.9927,
      "mean_token_accuracy": 0.2213546723127365,
      "num_tokens": 92842576.0,
      "step": 40535
    },
    {
      "entropy": 5.794633245468139,
      "epoch": 4.007512850929221,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003480166230908683,
      "loss": 4.9798,
      "mean_token_accuracy": 0.222476963698864,
      "num_tokens": 92853981.0,
      "step": 40540
    },
    {
      "entropy": 5.797709703445435,
      "epoch": 4.008007117437723,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003479832421252098,
      "loss": 4.971,
      "mean_token_accuracy": 0.21914765536785125,
      "num_tokens": 92866059.0,
      "step": 40545
    },
    {
      "entropy": 5.7646660804748535,
      "epoch": 4.008501383946224,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003479498593641522,
      "loss": 4.9291,
      "mean_token_accuracy": 0.22424253821372986,
      "num_tokens": 92877362.0,
      "step": 40550
    },
    {
      "entropy": 5.807667016983032,
      "epoch": 4.008995650454725,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003479164748085167,
      "loss": 4.9909,
      "mean_token_accuracy": 0.2231794148683548,
      "num_tokens": 92890910.0,
      "step": 40555
    },
    {
      "entropy": 5.786786651611328,
      "epoch": 4.0094899169632265,
      "grad_norm": 1.390625,
      "learning_rate": 0.00034788308845912463,
      "loss": 4.8903,
      "mean_token_accuracy": 0.23297762274742126,
      "num_tokens": 92901463.0,
      "step": 40560
    },
    {
      "entropy": 5.7811933040618895,
      "epoch": 4.009984183471728,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003478497003167972,
      "loss": 4.9014,
      "mean_token_accuracy": 0.23313555717468262,
      "num_tokens": 92912964.0,
      "step": 40565
    },
    {
      "entropy": 5.732276153564453,
      "epoch": 4.010478449980229,
      "grad_norm": 1.421875,
      "learning_rate": 0.00034781631038235597,
      "loss": 4.882,
      "mean_token_accuracy": 0.2361373260617256,
      "num_tokens": 92923725.0,
      "step": 40570
    },
    {
      "entropy": 5.696112251281738,
      "epoch": 4.01097271648873,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003477829186566221,
      "loss": 4.9464,
      "mean_token_accuracy": 0.23318856358528137,
      "num_tokens": 92934982.0,
      "step": 40575
    },
    {
      "entropy": 5.828345251083374,
      "epoch": 4.0114669829972325,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003477495251404172,
      "loss": 4.9474,
      "mean_token_accuracy": 0.22557390630245208,
      "num_tokens": 92945525.0,
      "step": 40580
    },
    {
      "entropy": 5.735734272003174,
      "epoch": 4.011961249505734,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003477161298345627,
      "loss": 4.8962,
      "mean_token_accuracy": 0.23384736478328705,
      "num_tokens": 92957802.0,
      "step": 40585
    },
    {
      "entropy": 5.83856692314148,
      "epoch": 4.012455516014235,
      "grad_norm": 1.34375,
      "learning_rate": 0.00034768273273988004,
      "loss": 4.9609,
      "mean_token_accuracy": 0.22502784430980682,
      "num_tokens": 92969010.0,
      "step": 40590
    },
    {
      "entropy": 5.811841201782227,
      "epoch": 4.012949782522736,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003476493338571909,
      "loss": 5.0109,
      "mean_token_accuracy": 0.22335454672574998,
      "num_tokens": 92979342.0,
      "step": 40595
    },
    {
      "entropy": 5.81382737159729,
      "epoch": 4.0134440490312375,
      "grad_norm": 1.328125,
      "learning_rate": 0.00034761593318731704,
      "loss": 4.9488,
      "mean_token_accuracy": 0.22952488213777542,
      "num_tokens": 92990092.0,
      "step": 40600
    },
    {
      "entropy": 5.7955405712127686,
      "epoch": 4.013938315539739,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00034758253073107983,
      "loss": 4.8631,
      "mean_token_accuracy": 0.23247399628162385,
      "num_tokens": 92999658.0,
      "step": 40605
    },
    {
      "entropy": 5.737318325042724,
      "epoch": 4.01443258204824,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003475491264893011,
      "loss": 4.9727,
      "mean_token_accuracy": 0.22823771685361863,
      "num_tokens": 93010771.0,
      "step": 40610
    },
    {
      "entropy": 5.788852262496948,
      "epoch": 4.014926848556742,
      "grad_norm": 1.265625,
      "learning_rate": 0.00034751572046280254,
      "loss": 4.9029,
      "mean_token_accuracy": 0.22613521069288253,
      "num_tokens": 93020993.0,
      "step": 40615
    },
    {
      "entropy": 5.820985412597656,
      "epoch": 4.0154211150652435,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00034748231265240617,
      "loss": 4.9961,
      "mean_token_accuracy": 0.22618257254362106,
      "num_tokens": 93032072.0,
      "step": 40620
    },
    {
      "entropy": 5.798431825637818,
      "epoch": 4.015915381573745,
      "grad_norm": 1.3125,
      "learning_rate": 0.00034744890305893355,
      "loss": 5.0358,
      "mean_token_accuracy": 0.22091520428657532,
      "num_tokens": 93042768.0,
      "step": 40625
    },
    {
      "entropy": 5.822737407684326,
      "epoch": 4.016409648082246,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003474154916832067,
      "loss": 4.9771,
      "mean_token_accuracy": 0.23114742189645768,
      "num_tokens": 93054162.0,
      "step": 40630
    },
    {
      "entropy": 5.739299964904785,
      "epoch": 4.016903914590747,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00034738207852604746,
      "loss": 4.7729,
      "mean_token_accuracy": 0.24725307673215866,
      "num_tokens": 93065124.0,
      "step": 40635
    },
    {
      "entropy": 5.769763088226318,
      "epoch": 4.0173981810992485,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00034734866358827795,
      "loss": 4.9424,
      "mean_token_accuracy": 0.23085163086652755,
      "num_tokens": 93076439.0,
      "step": 40640
    },
    {
      "entropy": 5.744402122497559,
      "epoch": 4.01789244760775,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00034731524687072005,
      "loss": 4.9154,
      "mean_token_accuracy": 0.22195419520139695,
      "num_tokens": 93087441.0,
      "step": 40645
    },
    {
      "entropy": 5.873094511032105,
      "epoch": 4.018386714116252,
      "grad_norm": 1.375,
      "learning_rate": 0.0003472818283741959,
      "loss": 5.0439,
      "mean_token_accuracy": 0.20976882576942443,
      "num_tokens": 93099098.0,
      "step": 40650
    },
    {
      "entropy": 5.762462902069092,
      "epoch": 4.018880980624753,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003472484080995275,
      "loss": 4.8994,
      "mean_token_accuracy": 0.2354673907160759,
      "num_tokens": 93109927.0,
      "step": 40655
    },
    {
      "entropy": 5.741521215438842,
      "epoch": 4.0193752471332544,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00034721498604753706,
      "loss": 4.8254,
      "mean_token_accuracy": 0.2353705108165741,
      "num_tokens": 93121910.0,
      "step": 40660
    },
    {
      "entropy": 5.711540842056275,
      "epoch": 4.019869513641756,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00034718156221904677,
      "loss": 4.9215,
      "mean_token_accuracy": 0.23290601670742034,
      "num_tokens": 93133598.0,
      "step": 40665
    },
    {
      "entropy": 5.7512321949005125,
      "epoch": 4.020363780150257,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003471481366148789,
      "loss": 4.9255,
      "mean_token_accuracy": 0.2258656218647957,
      "num_tokens": 93145189.0,
      "step": 40670
    },
    {
      "entropy": 5.786343765258789,
      "epoch": 4.020858046658758,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00034711470923585563,
      "loss": 4.9327,
      "mean_token_accuracy": 0.2321312204003334,
      "num_tokens": 93157630.0,
      "step": 40675
    },
    {
      "entropy": 5.731643772125244,
      "epoch": 4.0213523131672595,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003470812800827993,
      "loss": 4.8698,
      "mean_token_accuracy": 0.23693819493055343,
      "num_tokens": 93167640.0,
      "step": 40680
    },
    {
      "entropy": 5.722732257843018,
      "epoch": 4.021846579675761,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003470478491565323,
      "loss": 4.8762,
      "mean_token_accuracy": 0.23768391609191894,
      "num_tokens": 93179088.0,
      "step": 40685
    },
    {
      "entropy": 5.785084962844849,
      "epoch": 4.022340846184263,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00034701441645787706,
      "loss": 4.9433,
      "mean_token_accuracy": 0.23518577665090562,
      "num_tokens": 93190345.0,
      "step": 40690
    },
    {
      "entropy": 5.815078830718994,
      "epoch": 4.022835112692764,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003469809819876561,
      "loss": 4.9557,
      "mean_token_accuracy": 0.22196510881185533,
      "num_tokens": 93201206.0,
      "step": 40695
    },
    {
      "entropy": 5.763471603393555,
      "epoch": 4.023329379201265,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00034694754574669166,
      "loss": 4.8781,
      "mean_token_accuracy": 0.2307511568069458,
      "num_tokens": 93212204.0,
      "step": 40700
    },
    {
      "entropy": 5.7550232887268065,
      "epoch": 4.023823645709767,
      "grad_norm": 1.203125,
      "learning_rate": 0.00034691410773580653,
      "loss": 4.8637,
      "mean_token_accuracy": 0.24097003936767578,
      "num_tokens": 93224530.0,
      "step": 40705
    },
    {
      "entropy": 5.756622266769409,
      "epoch": 4.024317912218268,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003468806679558231,
      "loss": 4.9272,
      "mean_token_accuracy": 0.2391957238316536,
      "num_tokens": 93237403.0,
      "step": 40710
    },
    {
      "entropy": 5.745161056518555,
      "epoch": 4.024812178726769,
      "grad_norm": 1.234375,
      "learning_rate": 0.00034684722640756415,
      "loss": 4.9322,
      "mean_token_accuracy": 0.23341405987739564,
      "num_tokens": 93249168.0,
      "step": 40715
    },
    {
      "entropy": 5.78914065361023,
      "epoch": 4.0253064452352705,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003468137830918523,
      "loss": 4.933,
      "mean_token_accuracy": 0.22577230483293534,
      "num_tokens": 93260050.0,
      "step": 40720
    },
    {
      "entropy": 5.767939805984497,
      "epoch": 4.025800711743773,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003467803380095102,
      "loss": 4.9352,
      "mean_token_accuracy": 0.22619079798460007,
      "num_tokens": 93272514.0,
      "step": 40725
    },
    {
      "entropy": 5.71905746459961,
      "epoch": 4.026294978252274,
      "grad_norm": 1.140625,
      "learning_rate": 0.00034674689116136073,
      "loss": 4.9405,
      "mean_token_accuracy": 0.23253145962953567,
      "num_tokens": 93283824.0,
      "step": 40730
    },
    {
      "entropy": 5.732578086853027,
      "epoch": 4.026789244760775,
      "grad_norm": 1.234375,
      "learning_rate": 0.00034671344254822654,
      "loss": 4.8791,
      "mean_token_accuracy": 0.23702708780765533,
      "num_tokens": 93294864.0,
      "step": 40735
    },
    {
      "entropy": 5.765891647338867,
      "epoch": 4.027283511269276,
      "grad_norm": 1.265625,
      "learning_rate": 0.00034667999217093056,
      "loss": 4.9978,
      "mean_token_accuracy": 0.22929890751838683,
      "num_tokens": 93306166.0,
      "step": 40740
    },
    {
      "entropy": 5.804338550567627,
      "epoch": 4.027777777777778,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003466465400302956,
      "loss": 4.998,
      "mean_token_accuracy": 0.22286831736564636,
      "num_tokens": 93318622.0,
      "step": 40745
    },
    {
      "entropy": 5.741446828842163,
      "epoch": 4.028272044286279,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003466130861271447,
      "loss": 4.8809,
      "mean_token_accuracy": 0.2385230302810669,
      "num_tokens": 93330487.0,
      "step": 40750
    },
    {
      "entropy": 5.797368812561035,
      "epoch": 4.02876631079478,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003465796304623007,
      "loss": 4.9931,
      "mean_token_accuracy": 0.22030794322490693,
      "num_tokens": 93342639.0,
      "step": 40755
    },
    {
      "entropy": 5.733156490325928,
      "epoch": 4.029260577303282,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003465461730365868,
      "loss": 4.8791,
      "mean_token_accuracy": 0.23482966125011445,
      "num_tokens": 93355696.0,
      "step": 40760
    },
    {
      "entropy": 5.786558294296265,
      "epoch": 4.029754843811784,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003465127138508259,
      "loss": 4.9096,
      "mean_token_accuracy": 0.22907466441392899,
      "num_tokens": 93367913.0,
      "step": 40765
    },
    {
      "entropy": 5.756281185150146,
      "epoch": 4.030249110320285,
      "grad_norm": 1.328125,
      "learning_rate": 0.00034647925290584114,
      "loss": 4.9418,
      "mean_token_accuracy": 0.22875956445932388,
      "num_tokens": 93378872.0,
      "step": 40770
    },
    {
      "entropy": 5.75783052444458,
      "epoch": 4.030743376828786,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003464457902024556,
      "loss": 4.9522,
      "mean_token_accuracy": 0.22891180515289306,
      "num_tokens": 93390029.0,
      "step": 40775
    },
    {
      "entropy": 5.817436361312867,
      "epoch": 4.031237643337287,
      "grad_norm": 1.25,
      "learning_rate": 0.00034641232574149263,
      "loss": 4.9621,
      "mean_token_accuracy": 0.22530021965503694,
      "num_tokens": 93401518.0,
      "step": 40780
    },
    {
      "entropy": 5.7952308177948,
      "epoch": 4.031731909845789,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003463788595237753,
      "loss": 4.8551,
      "mean_token_accuracy": 0.2315203756093979,
      "num_tokens": 93412399.0,
      "step": 40785
    },
    {
      "entropy": 5.796579647064209,
      "epoch": 4.03222617635429,
      "grad_norm": 1.28125,
      "learning_rate": 0.00034634539155012703,
      "loss": 4.9637,
      "mean_token_accuracy": 0.22577570527791976,
      "num_tokens": 93423815.0,
      "step": 40790
    },
    {
      "entropy": 5.79377179145813,
      "epoch": 4.032720442862791,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000346311921821371,
      "loss": 4.9242,
      "mean_token_accuracy": 0.23173152804374694,
      "num_tokens": 93434635.0,
      "step": 40795
    },
    {
      "entropy": 5.7649143695831295,
      "epoch": 4.033214709371293,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00034627845033833053,
      "loss": 4.9407,
      "mean_token_accuracy": 0.23091771006584166,
      "num_tokens": 93445138.0,
      "step": 40800
    },
    {
      "entropy": 5.762605762481689,
      "epoch": 4.033708975879795,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003462449771018293,
      "loss": 4.9324,
      "mean_token_accuracy": 0.23476919680833816,
      "num_tokens": 93456249.0,
      "step": 40805
    },
    {
      "entropy": 5.8352518558502195,
      "epoch": 4.034203242388296,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003462115021126906,
      "loss": 4.9729,
      "mean_token_accuracy": 0.22602119892835618,
      "num_tokens": 93468576.0,
      "step": 40810
    },
    {
      "entropy": 5.813147830963135,
      "epoch": 4.034697508896797,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003461780253717378,
      "loss": 4.9036,
      "mean_token_accuracy": 0.22892917990684508,
      "num_tokens": 93479643.0,
      "step": 40815
    },
    {
      "entropy": 5.786829233169556,
      "epoch": 4.035191775405298,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003461445468797946,
      "loss": 4.8802,
      "mean_token_accuracy": 0.2335035979747772,
      "num_tokens": 93490619.0,
      "step": 40820
    },
    {
      "entropy": 5.688183116912842,
      "epoch": 4.0356860419138,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00034611106663768443,
      "loss": 4.8481,
      "mean_token_accuracy": 0.22841022461652755,
      "num_tokens": 93501763.0,
      "step": 40825
    },
    {
      "entropy": 5.73259015083313,
      "epoch": 4.036180308422301,
      "grad_norm": 1.3125,
      "learning_rate": 0.000346077584646231,
      "loss": 4.8408,
      "mean_token_accuracy": 0.23093935698270798,
      "num_tokens": 93513851.0,
      "step": 40830
    },
    {
      "entropy": 5.71949577331543,
      "epoch": 4.036674574930803,
      "grad_norm": 1.328125,
      "learning_rate": 0.00034604410090625795,
      "loss": 4.9332,
      "mean_token_accuracy": 0.22738130986690522,
      "num_tokens": 93525049.0,
      "step": 40835
    },
    {
      "entropy": 5.723616600036621,
      "epoch": 4.037168841439304,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003460106154185891,
      "loss": 4.926,
      "mean_token_accuracy": 0.23043525218963623,
      "num_tokens": 93537068.0,
      "step": 40840
    },
    {
      "entropy": 5.797301578521728,
      "epoch": 4.037663107947806,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00034597712818404793,
      "loss": 4.9098,
      "mean_token_accuracy": 0.2364737495779991,
      "num_tokens": 93548787.0,
      "step": 40845
    },
    {
      "entropy": 5.7573511600494385,
      "epoch": 4.038157374456307,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003459436392034585,
      "loss": 4.9187,
      "mean_token_accuracy": 0.23165173381567,
      "num_tokens": 93559781.0,
      "step": 40850
    },
    {
      "entropy": 5.769538259506225,
      "epoch": 4.038651640964808,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00034591014847764457,
      "loss": 4.897,
      "mean_token_accuracy": 0.22607481479644775,
      "num_tokens": 93570775.0,
      "step": 40855
    },
    {
      "entropy": 5.738147020339966,
      "epoch": 4.039145907473309,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003458766560074299,
      "loss": 4.8908,
      "mean_token_accuracy": 0.23456590324640275,
      "num_tokens": 93581222.0,
      "step": 40860
    },
    {
      "entropy": 5.775625705718994,
      "epoch": 4.039640173981811,
      "grad_norm": 1.25,
      "learning_rate": 0.0003458431617936386,
      "loss": 4.9712,
      "mean_token_accuracy": 0.22165210992097856,
      "num_tokens": 93592071.0,
      "step": 40865
    },
    {
      "entropy": 5.796543884277344,
      "epoch": 4.040134440490313,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003458096658370944,
      "loss": 4.9556,
      "mean_token_accuracy": 0.22880680561065675,
      "num_tokens": 93603570.0,
      "step": 40870
    },
    {
      "entropy": 5.840706920623779,
      "epoch": 4.040628706998814,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003457761681386215,
      "loss": 4.9349,
      "mean_token_accuracy": 0.2240452766418457,
      "num_tokens": 93614552.0,
      "step": 40875
    },
    {
      "entropy": 5.765056705474853,
      "epoch": 4.041122973507315,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003457426686990439,
      "loss": 4.9383,
      "mean_token_accuracy": 0.23604277819395064,
      "num_tokens": 93625593.0,
      "step": 40880
    },
    {
      "entropy": 5.757392263412475,
      "epoch": 4.041617240015817,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003457091675191857,
      "loss": 4.936,
      "mean_token_accuracy": 0.23400370180606841,
      "num_tokens": 93635420.0,
      "step": 40885
    },
    {
      "entropy": 5.742556428909301,
      "epoch": 4.042111506524318,
      "grad_norm": 1.328125,
      "learning_rate": 0.000345675664599871,
      "loss": 4.869,
      "mean_token_accuracy": 0.23507955372333528,
      "num_tokens": 93647458.0,
      "step": 40890
    },
    {
      "entropy": 5.796082639694214,
      "epoch": 4.042605773032819,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003456421599419239,
      "loss": 4.9119,
      "mean_token_accuracy": 0.2297603502869606,
      "num_tokens": 93658867.0,
      "step": 40895
    },
    {
      "entropy": 5.834683990478515,
      "epoch": 4.04310003954132,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00034560865354616875,
      "loss": 4.97,
      "mean_token_accuracy": 0.22013348788022996,
      "num_tokens": 93670391.0,
      "step": 40900
    },
    {
      "entropy": 5.739375066757202,
      "epoch": 4.0435943060498225,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003455751454134298,
      "loss": 4.9012,
      "mean_token_accuracy": 0.23456461131572723,
      "num_tokens": 93681670.0,
      "step": 40905
    },
    {
      "entropy": 5.757953786849976,
      "epoch": 4.044088572558324,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003455416355445313,
      "loss": 4.8941,
      "mean_token_accuracy": 0.22779715657234192,
      "num_tokens": 93692529.0,
      "step": 40910
    },
    {
      "entropy": 5.815210199356079,
      "epoch": 4.044582839066825,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00034550812394029765,
      "loss": 4.9258,
      "mean_token_accuracy": 0.22130864560604097,
      "num_tokens": 93704259.0,
      "step": 40915
    },
    {
      "entropy": 5.716447591781616,
      "epoch": 4.045077105575326,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00034547461060155317,
      "loss": 4.9079,
      "mean_token_accuracy": 0.22683641761541368,
      "num_tokens": 93715832.0,
      "step": 40920
    },
    {
      "entropy": 5.776839113235473,
      "epoch": 4.0455713720838276,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003454410955291224,
      "loss": 4.9515,
      "mean_token_accuracy": 0.22794792652130128,
      "num_tokens": 93728022.0,
      "step": 40925
    },
    {
      "entropy": 5.769331359863282,
      "epoch": 4.046065638592329,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00034540757872382965,
      "loss": 4.9093,
      "mean_token_accuracy": 0.23582160025835036,
      "num_tokens": 93739445.0,
      "step": 40930
    },
    {
      "entropy": 5.663434267044067,
      "epoch": 4.04655990510083,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003453740601864997,
      "loss": 4.8767,
      "mean_token_accuracy": 0.24382007867097855,
      "num_tokens": 93752132.0,
      "step": 40935
    },
    {
      "entropy": 5.783597755432129,
      "epoch": 4.047054171609331,
      "grad_norm": 1.109375,
      "learning_rate": 0.00034534053991795683,
      "loss": 4.9491,
      "mean_token_accuracy": 0.22663235664367676,
      "num_tokens": 93763346.0,
      "step": 40940
    },
    {
      "entropy": 5.8040412902832035,
      "epoch": 4.0475484381178335,
      "grad_norm": 1.328125,
      "learning_rate": 0.00034530701791902574,
      "loss": 4.9652,
      "mean_token_accuracy": 0.22557296901941298,
      "num_tokens": 93775877.0,
      "step": 40945
    },
    {
      "entropy": 5.766940450668335,
      "epoch": 4.048042704626335,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00034527349419053116,
      "loss": 4.9599,
      "mean_token_accuracy": 0.2233266532421112,
      "num_tokens": 93788741.0,
      "step": 40950
    },
    {
      "entropy": 5.732238006591797,
      "epoch": 4.048536971134836,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003452399687332978,
      "loss": 4.9262,
      "mean_token_accuracy": 0.23133978396654128,
      "num_tokens": 93798865.0,
      "step": 40955
    },
    {
      "entropy": 5.794459104537964,
      "epoch": 4.049031237643337,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003452064415481502,
      "loss": 4.9413,
      "mean_token_accuracy": 0.23377889692783355,
      "num_tokens": 93810131.0,
      "step": 40960
    },
    {
      "entropy": 5.776198530197144,
      "epoch": 4.0495255041518385,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003451729126359132,
      "loss": 4.9106,
      "mean_token_accuracy": 0.23329171538352966,
      "num_tokens": 93821696.0,
      "step": 40965
    },
    {
      "entropy": 5.81839919090271,
      "epoch": 4.05001977066034,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003451393819974118,
      "loss": 4.9593,
      "mean_token_accuracy": 0.22210249155759812,
      "num_tokens": 93833297.0,
      "step": 40970
    },
    {
      "entropy": 5.757199001312256,
      "epoch": 4.050514037168841,
      "grad_norm": 1.234375,
      "learning_rate": 0.00034510584963347063,
      "loss": 4.9656,
      "mean_token_accuracy": 0.22226360738277434,
      "num_tokens": 93844573.0,
      "step": 40975
    },
    {
      "entropy": 5.81689419746399,
      "epoch": 4.051008303677343,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003450723155449148,
      "loss": 4.9323,
      "mean_token_accuracy": 0.22625923305749893,
      "num_tokens": 93856542.0,
      "step": 40980
    },
    {
      "entropy": 5.7723846435546875,
      "epoch": 4.0515025701858445,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000345038779732569,
      "loss": 4.8945,
      "mean_token_accuracy": 0.22036599069833757,
      "num_tokens": 93868750.0,
      "step": 40985
    },
    {
      "entropy": 5.77267107963562,
      "epoch": 4.051996836694346,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003450052421972583,
      "loss": 4.8809,
      "mean_token_accuracy": 0.23302657455205916,
      "num_tokens": 93879506.0,
      "step": 40990
    },
    {
      "entropy": 5.734283685684204,
      "epoch": 4.052491103202847,
      "grad_norm": 1.265625,
      "learning_rate": 0.00034497170293980794,
      "loss": 4.8758,
      "mean_token_accuracy": 0.23572975695133208,
      "num_tokens": 93890068.0,
      "step": 40995
    },
    {
      "entropy": 5.766786241531372,
      "epoch": 4.052985369711348,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003449381619610428,
      "loss": 5.0101,
      "mean_token_accuracy": 0.22521729469299318,
      "num_tokens": 93902869.0,
      "step": 41000
    },
    {
      "entropy": 5.7843513011932375,
      "epoch": 4.0534796362198495,
      "grad_norm": 1.125,
      "learning_rate": 0.00034490461926178794,
      "loss": 4.9603,
      "mean_token_accuracy": 0.2256326735019684,
      "num_tokens": 93915403.0,
      "step": 41005
    },
    {
      "entropy": 5.778074169158936,
      "epoch": 4.053973902728351,
      "grad_norm": 1.34375,
      "learning_rate": 0.00034487107484286873,
      "loss": 4.8964,
      "mean_token_accuracy": 0.2346753165125847,
      "num_tokens": 93926616.0,
      "step": 41010
    },
    {
      "entropy": 5.783435869216919,
      "epoch": 4.054468169236853,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00034483752870511015,
      "loss": 4.8924,
      "mean_token_accuracy": 0.2339733734726906,
      "num_tokens": 93938840.0,
      "step": 41015
    },
    {
      "entropy": 5.7416908740997314,
      "epoch": 4.054962435745354,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003448039808493375,
      "loss": 4.9461,
      "mean_token_accuracy": 0.2251197636127472,
      "num_tokens": 93951294.0,
      "step": 41020
    },
    {
      "entropy": 5.756411695480347,
      "epoch": 4.0554567022538555,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003447704312763762,
      "loss": 4.9183,
      "mean_token_accuracy": 0.23253450244665147,
      "num_tokens": 93962596.0,
      "step": 41025
    },
    {
      "entropy": 5.850187873840332,
      "epoch": 4.055950968762357,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00034473687998705144,
      "loss": 4.9916,
      "mean_token_accuracy": 0.22448416352272033,
      "num_tokens": 93973699.0,
      "step": 41030
    },
    {
      "entropy": 5.757361125946045,
      "epoch": 4.056445235270858,
      "grad_norm": 1.234375,
      "learning_rate": 0.00034470332698218853,
      "loss": 4.8888,
      "mean_token_accuracy": 0.22993982285261155,
      "num_tokens": 93984711.0,
      "step": 41035
    },
    {
      "entropy": 5.798279857635498,
      "epoch": 4.056939501779359,
      "grad_norm": 1.1875,
      "learning_rate": 0.00034466977226261305,
      "loss": 5.0305,
      "mean_token_accuracy": 0.22364769577980043,
      "num_tokens": 93996217.0,
      "step": 41040
    },
    {
      "entropy": 5.732394647598267,
      "epoch": 4.0574337682878605,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003446362158291503,
      "loss": 4.9302,
      "mean_token_accuracy": 0.22893664091825486,
      "num_tokens": 94007738.0,
      "step": 41045
    },
    {
      "entropy": 5.780115079879761,
      "epoch": 4.057928034796362,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003446026576826258,
      "loss": 4.9278,
      "mean_token_accuracy": 0.23068011552095413,
      "num_tokens": 94018473.0,
      "step": 41050
    },
    {
      "entropy": 5.818477725982666,
      "epoch": 4.058422301304864,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003445690978238653,
      "loss": 4.9817,
      "mean_token_accuracy": 0.22401828318834305,
      "num_tokens": 94029461.0,
      "step": 41055
    },
    {
      "entropy": 5.734799480438232,
      "epoch": 4.058916567813365,
      "grad_norm": 1.28125,
      "learning_rate": 0.00034453553625369405,
      "loss": 4.9754,
      "mean_token_accuracy": 0.22892947793006896,
      "num_tokens": 94042399.0,
      "step": 41060
    },
    {
      "entropy": 5.80694727897644,
      "epoch": 4.0594108343218664,
      "grad_norm": 1.1875,
      "learning_rate": 0.00034450197297293783,
      "loss": 4.8984,
      "mean_token_accuracy": 0.23060431331396103,
      "num_tokens": 94054694.0,
      "step": 41065
    },
    {
      "entropy": 5.811329936981201,
      "epoch": 4.059905100830368,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003444684079824223,
      "loss": 4.9664,
      "mean_token_accuracy": 0.2272731676697731,
      "num_tokens": 94065426.0,
      "step": 41070
    },
    {
      "entropy": 5.736282777786255,
      "epoch": 4.060399367338869,
      "grad_norm": 1.28125,
      "learning_rate": 0.00034443484128297326,
      "loss": 4.8465,
      "mean_token_accuracy": 0.2341087654232979,
      "num_tokens": 94078119.0,
      "step": 41075
    },
    {
      "entropy": 5.719613838195801,
      "epoch": 4.06089363384737,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0003444012728754162,
      "loss": 4.9002,
      "mean_token_accuracy": 0.23489640206098555,
      "num_tokens": 94089157.0,
      "step": 41080
    },
    {
      "entropy": 5.748463153839111,
      "epoch": 4.0613879003558715,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00034436770276057715,
      "loss": 4.9151,
      "mean_token_accuracy": 0.22995193153619767,
      "num_tokens": 94101534.0,
      "step": 41085
    },
    {
      "entropy": 5.768482398986817,
      "epoch": 4.061882166864374,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003443341309392818,
      "loss": 4.9585,
      "mean_token_accuracy": 0.22236736714839936,
      "num_tokens": 94113310.0,
      "step": 41090
    },
    {
      "entropy": 5.780543422698974,
      "epoch": 4.062376433372875,
      "grad_norm": 1.234375,
      "learning_rate": 0.00034430055741235615,
      "loss": 4.9008,
      "mean_token_accuracy": 0.2296526089310646,
      "num_tokens": 94124189.0,
      "step": 41095
    },
    {
      "entropy": 5.745877408981324,
      "epoch": 4.062870699881376,
      "grad_norm": 1.28125,
      "learning_rate": 0.00034426698218062603,
      "loss": 4.8898,
      "mean_token_accuracy": 0.23208441138267516,
      "num_tokens": 94135458.0,
      "step": 41100
    },
    {
      "entropy": 5.69871335029602,
      "epoch": 4.063364966389877,
      "grad_norm": 1.34375,
      "learning_rate": 0.00034423340524491737,
      "loss": 4.903,
      "mean_token_accuracy": 0.2315752238035202,
      "num_tokens": 94146270.0,
      "step": 41105
    },
    {
      "entropy": 5.778707981109619,
      "epoch": 4.063859232898379,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003441998266060561,
      "loss": 4.9022,
      "mean_token_accuracy": 0.23032476156949996,
      "num_tokens": 94157814.0,
      "step": 41110
    },
    {
      "entropy": 5.83778223991394,
      "epoch": 4.06435349940688,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00034416624626486855,
      "loss": 4.9689,
      "mean_token_accuracy": 0.23451650887727737,
      "num_tokens": 94170424.0,
      "step": 41115
    },
    {
      "entropy": 5.783183288574219,
      "epoch": 4.064847765915381,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003441326642221805,
      "loss": 4.9329,
      "mean_token_accuracy": 0.2251179501414299,
      "num_tokens": 94181913.0,
      "step": 41120
    },
    {
      "entropy": 5.688137483596802,
      "epoch": 4.065342032423883,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003440990804788182,
      "loss": 4.9108,
      "mean_token_accuracy": 0.22789467573165895,
      "num_tokens": 94193370.0,
      "step": 41125
    },
    {
      "entropy": 5.734695625305176,
      "epoch": 4.065836298932385,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003440654950356078,
      "loss": 4.8451,
      "mean_token_accuracy": 0.2342666804790497,
      "num_tokens": 94204176.0,
      "step": 41130
    },
    {
      "entropy": 5.829955577850342,
      "epoch": 4.066330565440886,
      "grad_norm": 1.296875,
      "learning_rate": 0.00034403190789337546,
      "loss": 4.949,
      "mean_token_accuracy": 0.22796981483697892,
      "num_tokens": 94214355.0,
      "step": 41135
    },
    {
      "entropy": 5.800871992111206,
      "epoch": 4.066824831949387,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003439983190529476,
      "loss": 4.9679,
      "mean_token_accuracy": 0.22392816990613937,
      "num_tokens": 94226248.0,
      "step": 41140
    },
    {
      "entropy": 5.758503389358521,
      "epoch": 4.067319098457888,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003439647285151503,
      "loss": 4.873,
      "mean_token_accuracy": 0.23600295782089234,
      "num_tokens": 94236847.0,
      "step": 41145
    },
    {
      "entropy": 5.783296966552735,
      "epoch": 4.06781336496639,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00034393113628080994,
      "loss": 4.9682,
      "mean_token_accuracy": 0.22489985525608064,
      "num_tokens": 94248387.0,
      "step": 41150
    },
    {
      "entropy": 5.766246223449707,
      "epoch": 4.068307631474891,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00034389754235075303,
      "loss": 4.8917,
      "mean_token_accuracy": 0.2326929062604904,
      "num_tokens": 94259069.0,
      "step": 41155
    },
    {
      "entropy": 5.713758754730224,
      "epoch": 4.068801897983393,
      "grad_norm": 1.265625,
      "learning_rate": 0.00034386394672580585,
      "loss": 4.8865,
      "mean_token_accuracy": 0.23263249546289444,
      "num_tokens": 94270390.0,
      "step": 41160
    },
    {
      "entropy": 5.77420597076416,
      "epoch": 4.069296164491894,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003438303494067949,
      "loss": 4.9757,
      "mean_token_accuracy": 0.2271587073802948,
      "num_tokens": 94281712.0,
      "step": 41165
    },
    {
      "entropy": 5.7641973972320555,
      "epoch": 4.069790431000396,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003437967503945467,
      "loss": 4.892,
      "mean_token_accuracy": 0.2344158485531807,
      "num_tokens": 94292277.0,
      "step": 41170
    },
    {
      "entropy": 5.7991039752960205,
      "epoch": 4.070284697508897,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00034376314968988773,
      "loss": 5.0016,
      "mean_token_accuracy": 0.22252244949340821,
      "num_tokens": 94303276.0,
      "step": 41175
    },
    {
      "entropy": 5.751678419113159,
      "epoch": 4.070778964017398,
      "grad_norm": 1.40625,
      "learning_rate": 0.00034372954729364457,
      "loss": 4.8597,
      "mean_token_accuracy": 0.24168787002563477,
      "num_tokens": 94313692.0,
      "step": 41180
    },
    {
      "entropy": 5.802755832672119,
      "epoch": 4.071273230525899,
      "grad_norm": 1.21875,
      "learning_rate": 0.000343695943206644,
      "loss": 4.9339,
      "mean_token_accuracy": 0.2270935520529747,
      "num_tokens": 94326859.0,
      "step": 41185
    },
    {
      "entropy": 5.79927020072937,
      "epoch": 4.071767497034401,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003436623374297125,
      "loss": 4.9852,
      "mean_token_accuracy": 0.2233715310692787,
      "num_tokens": 94339417.0,
      "step": 41190
    },
    {
      "entropy": 5.820031929016113,
      "epoch": 4.072261763542902,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003436287299636768,
      "loss": 5.0127,
      "mean_token_accuracy": 0.22170572131872177,
      "num_tokens": 94351462.0,
      "step": 41195
    },
    {
      "entropy": 5.775681352615356,
      "epoch": 4.072756030051404,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003435951208093638,
      "loss": 4.8849,
      "mean_token_accuracy": 0.2265557661652565,
      "num_tokens": 94362455.0,
      "step": 41200
    },
    {
      "entropy": 5.760687303543091,
      "epoch": 4.073250296559905,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003435615099676001,
      "loss": 4.9154,
      "mean_token_accuracy": 0.22505036890506744,
      "num_tokens": 94373005.0,
      "step": 41205
    },
    {
      "entropy": 5.7610198020935055,
      "epoch": 4.073744563068407,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003435278974392126,
      "loss": 4.9173,
      "mean_token_accuracy": 0.2382674530148506,
      "num_tokens": 94385619.0,
      "step": 41210
    },
    {
      "entropy": 5.808843326568604,
      "epoch": 4.074238829576908,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003434942832250283,
      "loss": 4.928,
      "mean_token_accuracy": 0.22959907054901124,
      "num_tokens": 94397281.0,
      "step": 41215
    },
    {
      "entropy": 5.798768138885498,
      "epoch": 4.074733096085409,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00034346066732587394,
      "loss": 4.9541,
      "mean_token_accuracy": 0.22725386023521424,
      "num_tokens": 94408222.0,
      "step": 41220
    },
    {
      "entropy": 5.812325429916382,
      "epoch": 4.07522736259391,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00034342704974257653,
      "loss": 4.9574,
      "mean_token_accuracy": 0.228243188560009,
      "num_tokens": 94419540.0,
      "step": 41225
    },
    {
      "entropy": 5.731334590911866,
      "epoch": 4.075721629102412,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003433934304759631,
      "loss": 4.9271,
      "mean_token_accuracy": 0.22931711822748185,
      "num_tokens": 94431372.0,
      "step": 41230
    },
    {
      "entropy": 5.769289112091064,
      "epoch": 4.076215895610914,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00034335980952686064,
      "loss": 4.8539,
      "mean_token_accuracy": 0.2335490494966507,
      "num_tokens": 94441922.0,
      "step": 41235
    },
    {
      "entropy": 5.734683847427368,
      "epoch": 4.076710162119415,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003433261868960964,
      "loss": 4.82,
      "mean_token_accuracy": 0.23569392114877702,
      "num_tokens": 94453031.0,
      "step": 41240
    },
    {
      "entropy": 5.8063396453857425,
      "epoch": 4.077204428627916,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00034329256258449725,
      "loss": 4.9513,
      "mean_token_accuracy": 0.22520948052406312,
      "num_tokens": 94463952.0,
      "step": 41245
    },
    {
      "entropy": 5.80389256477356,
      "epoch": 4.077698695136418,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003432589365928904,
      "loss": 4.9184,
      "mean_token_accuracy": 0.23384554088115692,
      "num_tokens": 94474811.0,
      "step": 41250
    },
    {
      "entropy": 5.71890115737915,
      "epoch": 4.078192961644919,
      "grad_norm": 1.234375,
      "learning_rate": 0.00034322530892210324,
      "loss": 4.9015,
      "mean_token_accuracy": 0.22836243659257888,
      "num_tokens": 94485901.0,
      "step": 41255
    },
    {
      "entropy": 5.735397243499756,
      "epoch": 4.07868722815342,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003431916795729628,
      "loss": 4.9323,
      "mean_token_accuracy": 0.22868915647268295,
      "num_tokens": 94498147.0,
      "step": 41260
    },
    {
      "entropy": 5.740125417709351,
      "epoch": 4.079181494661921,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003431580485462966,
      "loss": 4.8752,
      "mean_token_accuracy": 0.2339969828724861,
      "num_tokens": 94510787.0,
      "step": 41265
    },
    {
      "entropy": 5.775384092330933,
      "epoch": 4.0796757611704235,
      "grad_norm": 1.25,
      "learning_rate": 0.0003431244158429317,
      "loss": 5.0165,
      "mean_token_accuracy": 0.22178604155778886,
      "num_tokens": 94523459.0,
      "step": 41270
    },
    {
      "entropy": 5.766784620285034,
      "epoch": 4.080170027678925,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00034309078146369567,
      "loss": 4.9288,
      "mean_token_accuracy": 0.22403345853090287,
      "num_tokens": 94534579.0,
      "step": 41275
    },
    {
      "entropy": 5.8723752975463865,
      "epoch": 4.080664294187426,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003430571454094158,
      "loss": 5.0086,
      "mean_token_accuracy": 0.22319798320531845,
      "num_tokens": 94545233.0,
      "step": 41280
    },
    {
      "entropy": 5.8177896499633786,
      "epoch": 4.081158560695927,
      "grad_norm": 1.390625,
      "learning_rate": 0.00034302350768091963,
      "loss": 4.9582,
      "mean_token_accuracy": 0.2297793671488762,
      "num_tokens": 94555189.0,
      "step": 41285
    },
    {
      "entropy": 5.854527282714844,
      "epoch": 4.081652827204429,
      "grad_norm": 1.34375,
      "learning_rate": 0.00034298986827903464,
      "loss": 4.985,
      "mean_token_accuracy": 0.2198912352323532,
      "num_tokens": 94566001.0,
      "step": 41290
    },
    {
      "entropy": 5.710570478439331,
      "epoch": 4.08214709371293,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003429562272045883,
      "loss": 4.8493,
      "mean_token_accuracy": 0.23932705223560333,
      "num_tokens": 94577578.0,
      "step": 41295
    },
    {
      "entropy": 5.767728328704834,
      "epoch": 4.082641360221431,
      "grad_norm": 1.296875,
      "learning_rate": 0.00034292258445840816,
      "loss": 4.854,
      "mean_token_accuracy": 0.23665076196193696,
      "num_tokens": 94589162.0,
      "step": 41300
    },
    {
      "entropy": 5.776793336868286,
      "epoch": 4.083135626729932,
      "grad_norm": 1.296875,
      "learning_rate": 0.00034288894004132203,
      "loss": 5.0262,
      "mean_token_accuracy": 0.21865688711404802,
      "num_tokens": 94601883.0,
      "step": 41305
    },
    {
      "entropy": 5.767431211471558,
      "epoch": 4.0836298932384345,
      "grad_norm": 1.3125,
      "learning_rate": 0.00034285529395415744,
      "loss": 4.9682,
      "mean_token_accuracy": 0.22047097980976105,
      "num_tokens": 94612944.0,
      "step": 41310
    },
    {
      "entropy": 5.77349739074707,
      "epoch": 4.084124159746936,
      "grad_norm": 1.421875,
      "learning_rate": 0.000342821646197742,
      "loss": 4.9159,
      "mean_token_accuracy": 0.22595989853143691,
      "num_tokens": 94624042.0,
      "step": 41315
    },
    {
      "entropy": 5.72882719039917,
      "epoch": 4.084618426255437,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00034278799677290363,
      "loss": 4.8783,
      "mean_token_accuracy": 0.23347859382629393,
      "num_tokens": 94635149.0,
      "step": 41320
    },
    {
      "entropy": 5.7230040550231935,
      "epoch": 4.085112692763938,
      "grad_norm": 1.328125,
      "learning_rate": 0.00034275434568047,
      "loss": 4.9281,
      "mean_token_accuracy": 0.2301914244890213,
      "num_tokens": 94646586.0,
      "step": 41325
    },
    {
      "entropy": 5.725247192382812,
      "epoch": 4.0856069592724396,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003427206929212689,
      "loss": 4.9851,
      "mean_token_accuracy": 0.22446895092725755,
      "num_tokens": 94658602.0,
      "step": 41330
    },
    {
      "entropy": 5.785997343063355,
      "epoch": 4.086101225780941,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003426870384961283,
      "loss": 4.9171,
      "mean_token_accuracy": 0.2248311385512352,
      "num_tokens": 94668472.0,
      "step": 41335
    },
    {
      "entropy": 5.806724882125854,
      "epoch": 4.086595492289442,
      "grad_norm": 1.203125,
      "learning_rate": 0.000342653382405876,
      "loss": 4.9173,
      "mean_token_accuracy": 0.22882626205682755,
      "num_tokens": 94679397.0,
      "step": 41340
    },
    {
      "entropy": 5.851347637176514,
      "epoch": 4.087089758797944,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003426197246513401,
      "loss": 4.9839,
      "mean_token_accuracy": 0.22656087577342987,
      "num_tokens": 94690770.0,
      "step": 41345
    },
    {
      "entropy": 5.758445405960083,
      "epoch": 4.0875840253064455,
      "grad_norm": 1.21875,
      "learning_rate": 0.00034258606523334843,
      "loss": 4.9234,
      "mean_token_accuracy": 0.22523144632577896,
      "num_tokens": 94702426.0,
      "step": 41350
    },
    {
      "entropy": 5.672008848190307,
      "epoch": 4.088078291814947,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00034255240415272906,
      "loss": 4.8211,
      "mean_token_accuracy": 0.23724103569984437,
      "num_tokens": 94713403.0,
      "step": 41355
    },
    {
      "entropy": 5.797766160964966,
      "epoch": 4.088572558323448,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003425187414103101,
      "loss": 4.9518,
      "mean_token_accuracy": 0.23036193251609802,
      "num_tokens": 94724441.0,
      "step": 41360
    },
    {
      "entropy": 5.754850578308106,
      "epoch": 4.089066824831949,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00034248507700691957,
      "loss": 4.9526,
      "mean_token_accuracy": 0.23229407668113708,
      "num_tokens": 94735048.0,
      "step": 41365
    },
    {
      "entropy": 5.827209520339966,
      "epoch": 4.0895610913404505,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003424514109433857,
      "loss": 4.9731,
      "mean_token_accuracy": 0.22553397417068483,
      "num_tokens": 94745435.0,
      "step": 41370
    },
    {
      "entropy": 5.766744136810303,
      "epoch": 4.090055357848952,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003424177432205366,
      "loss": 4.8938,
      "mean_token_accuracy": 0.231175796687603,
      "num_tokens": 94756495.0,
      "step": 41375
    },
    {
      "entropy": 5.7455644607543945,
      "epoch": 4.090549624357454,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003423840738392007,
      "loss": 4.9044,
      "mean_token_accuracy": 0.2333471104502678,
      "num_tokens": 94767876.0,
      "step": 41380
    },
    {
      "entropy": 5.782841157913208,
      "epoch": 4.091043890865955,
      "grad_norm": 1.40625,
      "learning_rate": 0.000342350402800206,
      "loss": 4.9388,
      "mean_token_accuracy": 0.22504641264677047,
      "num_tokens": 94778855.0,
      "step": 41385
    },
    {
      "entropy": 5.827413558959961,
      "epoch": 4.0915381573744565,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00034231673010438104,
      "loss": 4.9025,
      "mean_token_accuracy": 0.2313551649451256,
      "num_tokens": 94788433.0,
      "step": 41390
    },
    {
      "entropy": 5.672556018829345,
      "epoch": 4.092032423882958,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00034228305575255397,
      "loss": 4.8024,
      "mean_token_accuracy": 0.24233658909797667,
      "num_tokens": 94799630.0,
      "step": 41395
    },
    {
      "entropy": 5.709322261810303,
      "epoch": 4.092526690391459,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00034224937974555334,
      "loss": 4.8542,
      "mean_token_accuracy": 0.23055158108472823,
      "num_tokens": 94811696.0,
      "step": 41400
    },
    {
      "entropy": 5.823106288909912,
      "epoch": 4.09302095689996,
      "grad_norm": 1.28125,
      "learning_rate": 0.00034221570208420755,
      "loss": 4.9787,
      "mean_token_accuracy": 0.2263631045818329,
      "num_tokens": 94822648.0,
      "step": 41405
    },
    {
      "entropy": 5.760322618484497,
      "epoch": 4.0935152234084615,
      "grad_norm": 1.265625,
      "learning_rate": 0.00034218202276934507,
      "loss": 4.8895,
      "mean_token_accuracy": 0.2305459722876549,
      "num_tokens": 94833690.0,
      "step": 41410
    },
    {
      "entropy": 5.728191661834717,
      "epoch": 4.094009489916964,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003421483418017944,
      "loss": 4.8494,
      "mean_token_accuracy": 0.23938871771097184,
      "num_tokens": 94843859.0,
      "step": 41415
    },
    {
      "entropy": 5.767635250091553,
      "epoch": 4.094503756425465,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003421146591823841,
      "loss": 4.9008,
      "mean_token_accuracy": 0.23263185620307922,
      "num_tokens": 94855156.0,
      "step": 41420
    },
    {
      "entropy": 5.8042725086212155,
      "epoch": 4.094998022933966,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003420809749119427,
      "loss": 5.0027,
      "mean_token_accuracy": 0.2186235785484314,
      "num_tokens": 94867082.0,
      "step": 41425
    },
    {
      "entropy": 5.733512687683105,
      "epoch": 4.0954922894424675,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003420472889912991,
      "loss": 4.8498,
      "mean_token_accuracy": 0.23714292496442796,
      "num_tokens": 94877606.0,
      "step": 41430
    },
    {
      "entropy": 5.778481340408325,
      "epoch": 4.095986555950969,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00034201360142128165,
      "loss": 4.8537,
      "mean_token_accuracy": 0.2366952121257782,
      "num_tokens": 94888114.0,
      "step": 41435
    },
    {
      "entropy": 5.7050361156463625,
      "epoch": 4.09648082245947,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003419799122027192,
      "loss": 4.8414,
      "mean_token_accuracy": 0.2384738028049469,
      "num_tokens": 94900135.0,
      "step": 41440
    },
    {
      "entropy": 5.768857192993164,
      "epoch": 4.096975088967971,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003419462213364405,
      "loss": 4.8938,
      "mean_token_accuracy": 0.2314760833978653,
      "num_tokens": 94911444.0,
      "step": 41445
    },
    {
      "entropy": 5.688500595092774,
      "epoch": 4.0974693554764725,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003419125288232745,
      "loss": 4.8984,
      "mean_token_accuracy": 0.23025558441877364,
      "num_tokens": 94923378.0,
      "step": 41450
    },
    {
      "entropy": 5.709119510650635,
      "epoch": 4.097963621984975,
      "grad_norm": 1.25,
      "learning_rate": 0.0003418788346640498,
      "loss": 4.8494,
      "mean_token_accuracy": 0.23043256849050522,
      "num_tokens": 94933784.0,
      "step": 41455
    },
    {
      "entropy": 5.7248087406158445,
      "epoch": 4.098457888493476,
      "grad_norm": 1.1875,
      "learning_rate": 0.00034184513885959544,
      "loss": 4.8828,
      "mean_token_accuracy": 0.23019013553857803,
      "num_tokens": 94945814.0,
      "step": 41460
    },
    {
      "entropy": 5.715527486801148,
      "epoch": 4.098952155001977,
      "grad_norm": 1.140625,
      "learning_rate": 0.00034181144141074023,
      "loss": 4.9327,
      "mean_token_accuracy": 0.22753919363021852,
      "num_tokens": 94957792.0,
      "step": 41465
    },
    {
      "entropy": 5.732485961914063,
      "epoch": 4.0994464215104784,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00034177774231831324,
      "loss": 4.9009,
      "mean_token_accuracy": 0.23239139318466187,
      "num_tokens": 94968524.0,
      "step": 41470
    },
    {
      "entropy": 5.777075529098511,
      "epoch": 4.09994068801898,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00034174404158314345,
      "loss": 4.9398,
      "mean_token_accuracy": 0.22249210774898528,
      "num_tokens": 94979213.0,
      "step": 41475
    },
    {
      "entropy": 5.693767738342285,
      "epoch": 4.100434954527481,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003417103392060599,
      "loss": 4.8408,
      "mean_token_accuracy": 0.2336836040019989,
      "num_tokens": 94991068.0,
      "step": 41480
    },
    {
      "entropy": 5.752703857421875,
      "epoch": 4.100929221035982,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00034167663518789156,
      "loss": 4.939,
      "mean_token_accuracy": 0.22881190478801727,
      "num_tokens": 95001780.0,
      "step": 41485
    },
    {
      "entropy": 5.822382068634033,
      "epoch": 4.101423487544484,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003416429295294677,
      "loss": 4.9949,
      "mean_token_accuracy": 0.22056188732385634,
      "num_tokens": 95013244.0,
      "step": 41490
    },
    {
      "entropy": 5.749713277816772,
      "epoch": 4.101917754052986,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00034160922223161743,
      "loss": 4.8634,
      "mean_token_accuracy": 0.23021193593740463,
      "num_tokens": 95024103.0,
      "step": 41495
    },
    {
      "entropy": 5.725566577911377,
      "epoch": 4.102412020561487,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003415755132951699,
      "loss": 4.8711,
      "mean_token_accuracy": 0.23483748286962508,
      "num_tokens": 95034132.0,
      "step": 41500
    },
    {
      "entropy": 5.73406457901001,
      "epoch": 4.102906287069988,
      "grad_norm": 1.25,
      "learning_rate": 0.0003415418027209546,
      "loss": 4.8841,
      "mean_token_accuracy": 0.22875937521457673,
      "num_tokens": 95046735.0,
      "step": 41505
    },
    {
      "entropy": 5.75829586982727,
      "epoch": 4.103400553578489,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003415080905098005,
      "loss": 4.8693,
      "mean_token_accuracy": 0.23368439227342605,
      "num_tokens": 95058538.0,
      "step": 41510
    },
    {
      "entropy": 5.758951187133789,
      "epoch": 4.103894820086991,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003414743766625371,
      "loss": 4.869,
      "mean_token_accuracy": 0.23296344578266143,
      "num_tokens": 95070049.0,
      "step": 41515
    },
    {
      "entropy": 5.741114854812622,
      "epoch": 4.104389086595492,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00034144066117999366,
      "loss": 4.9227,
      "mean_token_accuracy": 0.23145071417093277,
      "num_tokens": 95081702.0,
      "step": 41520
    },
    {
      "entropy": 5.7907140254974365,
      "epoch": 4.104883353103994,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003414069440629997,
      "loss": 4.9953,
      "mean_token_accuracy": 0.21967278122901918,
      "num_tokens": 95093831.0,
      "step": 41525
    },
    {
      "entropy": 5.7364953517913815,
      "epoch": 4.105377619612495,
      "grad_norm": 1.25,
      "learning_rate": 0.0003413732253123846,
      "loss": 4.8617,
      "mean_token_accuracy": 0.23206555396318435,
      "num_tokens": 95104960.0,
      "step": 41530
    },
    {
      "entropy": 5.791954851150512,
      "epoch": 4.105871886120997,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003413395049289778,
      "loss": 5.0021,
      "mean_token_accuracy": 0.2272172152996063,
      "num_tokens": 95115578.0,
      "step": 41535
    },
    {
      "entropy": 5.776717042922973,
      "epoch": 4.106366152629498,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003413057829136089,
      "loss": 4.9354,
      "mean_token_accuracy": 0.2249817132949829,
      "num_tokens": 95129237.0,
      "step": 41540
    },
    {
      "entropy": 5.808293724060059,
      "epoch": 4.106860419137999,
      "grad_norm": 1.421875,
      "learning_rate": 0.00034127205926710755,
      "loss": 5.0049,
      "mean_token_accuracy": 0.21929070651531218,
      "num_tokens": 95141500.0,
      "step": 41545
    },
    {
      "entropy": 5.737236499786377,
      "epoch": 4.1073546856465,
      "grad_norm": 1.328125,
      "learning_rate": 0.00034123833399030324,
      "loss": 4.8628,
      "mean_token_accuracy": 0.237789523601532,
      "num_tokens": 95152669.0,
      "step": 41550
    },
    {
      "entropy": 5.822025108337402,
      "epoch": 4.107848952155002,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003412046070840256,
      "loss": 4.9046,
      "mean_token_accuracy": 0.2299567922949791,
      "num_tokens": 95163181.0,
      "step": 41555
    },
    {
      "entropy": 5.8268636703491214,
      "epoch": 4.108343218663503,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00034117087854910436,
      "loss": 4.9984,
      "mean_token_accuracy": 0.21720368564128875,
      "num_tokens": 95176321.0,
      "step": 41560
    },
    {
      "entropy": 5.7686468124389645,
      "epoch": 4.108837485172005,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003411371483863692,
      "loss": 4.929,
      "mean_token_accuracy": 0.22899484187364577,
      "num_tokens": 95188097.0,
      "step": 41565
    },
    {
      "entropy": 5.679139041900635,
      "epoch": 4.109331751680506,
      "grad_norm": 1.171875,
      "learning_rate": 0.00034110341659665,
      "loss": 4.8546,
      "mean_token_accuracy": 0.24654210060834886,
      "num_tokens": 95199629.0,
      "step": 41570
    },
    {
      "entropy": 5.8065441131591795,
      "epoch": 4.109826018189008,
      "grad_norm": 1.125,
      "learning_rate": 0.0003410696831807765,
      "loss": 4.9852,
      "mean_token_accuracy": 0.21947593092918397,
      "num_tokens": 95212035.0,
      "step": 41575
    },
    {
      "entropy": 5.858055353164673,
      "epoch": 4.110320284697509,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00034103594813957846,
      "loss": 4.9899,
      "mean_token_accuracy": 0.22308092415332795,
      "num_tokens": 95222705.0,
      "step": 41580
    },
    {
      "entropy": 5.740157175064087,
      "epoch": 4.11081455120601,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003410022114738859,
      "loss": 4.9599,
      "mean_token_accuracy": 0.2264355167746544,
      "num_tokens": 95234806.0,
      "step": 41585
    },
    {
      "entropy": 5.743243074417114,
      "epoch": 4.111308817714511,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003409684731845287,
      "loss": 4.872,
      "mean_token_accuracy": 0.23121480643749237,
      "num_tokens": 95245789.0,
      "step": 41590
    },
    {
      "entropy": 5.806638145446778,
      "epoch": 4.111803084223013,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003409347332723369,
      "loss": 4.9604,
      "mean_token_accuracy": 0.22663187831640244,
      "num_tokens": 95258295.0,
      "step": 41595
    },
    {
      "entropy": 5.827193927764893,
      "epoch": 4.112297350731515,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003409009917381403,
      "loss": 4.938,
      "mean_token_accuracy": 0.2247425600886345,
      "num_tokens": 95270704.0,
      "step": 41600
    },
    {
      "entropy": 5.762149477005005,
      "epoch": 4.112791617240016,
      "grad_norm": 1.140625,
      "learning_rate": 0.00034086724858276913,
      "loss": 4.9273,
      "mean_token_accuracy": 0.2296552389860153,
      "num_tokens": 95283321.0,
      "step": 41605
    },
    {
      "entropy": 5.767711448669433,
      "epoch": 4.113285883748517,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00034083350380705344,
      "loss": 4.946,
      "mean_token_accuracy": 0.22966258674860002,
      "num_tokens": 95294592.0,
      "step": 41610
    },
    {
      "entropy": 5.726210451126098,
      "epoch": 4.113780150257019,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00034079975741182334,
      "loss": 4.9003,
      "mean_token_accuracy": 0.2387520655989647,
      "num_tokens": 95306619.0,
      "step": 41615
    },
    {
      "entropy": 5.712404870986939,
      "epoch": 4.11427441676552,
      "grad_norm": 1.203125,
      "learning_rate": 0.00034076600939790904,
      "loss": 4.8921,
      "mean_token_accuracy": 0.2362593322992325,
      "num_tokens": 95318660.0,
      "step": 41620
    },
    {
      "entropy": 5.755060625076294,
      "epoch": 4.114768683274021,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003407322597661407,
      "loss": 4.9723,
      "mean_token_accuracy": 0.2230877161026001,
      "num_tokens": 95330355.0,
      "step": 41625
    },
    {
      "entropy": 5.768701362609863,
      "epoch": 4.115262949782522,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003406985085173484,
      "loss": 4.924,
      "mean_token_accuracy": 0.23125091195106506,
      "num_tokens": 95343390.0,
      "step": 41630
    },
    {
      "entropy": 5.73344578742981,
      "epoch": 4.1157572162910245,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003406647556523628,
      "loss": 4.9221,
      "mean_token_accuracy": 0.22505954504013062,
      "num_tokens": 95355379.0,
      "step": 41635
    },
    {
      "entropy": 5.714823532104492,
      "epoch": 4.116251482799526,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000340631001172014,
      "loss": 4.9917,
      "mean_token_accuracy": 0.22489867061376573,
      "num_tokens": 95367166.0,
      "step": 41640
    },
    {
      "entropy": 5.802769517898559,
      "epoch": 4.116745749308027,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003405972450771323,
      "loss": 4.8362,
      "mean_token_accuracy": 0.23994607627391815,
      "num_tokens": 95377692.0,
      "step": 41645
    },
    {
      "entropy": 5.833525085449219,
      "epoch": 4.117240015816528,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003405634873685483,
      "loss": 4.9531,
      "mean_token_accuracy": 0.22516766041517258,
      "num_tokens": 95390057.0,
      "step": 41650
    },
    {
      "entropy": 5.7912674903869625,
      "epoch": 4.11773428232503,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00034052972804709224,
      "loss": 4.9227,
      "mean_token_accuracy": 0.22753215730190277,
      "num_tokens": 95401525.0,
      "step": 41655
    },
    {
      "entropy": 5.7854469299316404,
      "epoch": 4.118228548833531,
      "grad_norm": 1.296875,
      "learning_rate": 0.00034049596711359473,
      "loss": 4.9483,
      "mean_token_accuracy": 0.21727627664804458,
      "num_tokens": 95413659.0,
      "step": 41660
    },
    {
      "entropy": 5.800493669509888,
      "epoch": 4.118722815342032,
      "grad_norm": 1.390625,
      "learning_rate": 0.00034046220456888633,
      "loss": 4.9865,
      "mean_token_accuracy": 0.22885150015354155,
      "num_tokens": 95424269.0,
      "step": 41665
    },
    {
      "entropy": 5.771414947509766,
      "epoch": 4.119217081850534,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003404284404137975,
      "loss": 4.8838,
      "mean_token_accuracy": 0.2324856013059616,
      "num_tokens": 95434683.0,
      "step": 41670
    },
    {
      "entropy": 5.731257820129395,
      "epoch": 4.1197113483590355,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00034039467464915885,
      "loss": 5.0134,
      "mean_token_accuracy": 0.22297765463590621,
      "num_tokens": 95446561.0,
      "step": 41675
    },
    {
      "entropy": 5.777362203598022,
      "epoch": 4.120205614867537,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00034036090727580104,
      "loss": 4.9459,
      "mean_token_accuracy": 0.2209606260061264,
      "num_tokens": 95458716.0,
      "step": 41680
    },
    {
      "entropy": 5.831929492950439,
      "epoch": 4.120699881376038,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00034032713829455487,
      "loss": 4.9821,
      "mean_token_accuracy": 0.22794946283102036,
      "num_tokens": 95470628.0,
      "step": 41685
    },
    {
      "entropy": 5.786280822753906,
      "epoch": 4.121194147884539,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00034029336770625095,
      "loss": 4.8781,
      "mean_token_accuracy": 0.23393614590168,
      "num_tokens": 95481995.0,
      "step": 41690
    },
    {
      "entropy": 5.714264726638794,
      "epoch": 4.121688414393041,
      "grad_norm": 1.28125,
      "learning_rate": 0.00034025959551172005,
      "loss": 4.8229,
      "mean_token_accuracy": 0.23844441324472426,
      "num_tokens": 95493518.0,
      "step": 41695
    },
    {
      "entropy": 5.764245510101318,
      "epoch": 4.122182680901542,
      "grad_norm": 1.5625,
      "learning_rate": 0.00034022582171179284,
      "loss": 4.9767,
      "mean_token_accuracy": 0.22553161680698394,
      "num_tokens": 95504089.0,
      "step": 41700
    },
    {
      "entropy": 5.842084693908691,
      "epoch": 4.122676947410043,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00034019204630730045,
      "loss": 4.9682,
      "mean_token_accuracy": 0.21809196770191192,
      "num_tokens": 95513658.0,
      "step": 41705
    },
    {
      "entropy": 5.801896381378174,
      "epoch": 4.123171213918545,
      "grad_norm": 1.296875,
      "learning_rate": 0.00034015826929907357,
      "loss": 4.9524,
      "mean_token_accuracy": 0.22594311982393264,
      "num_tokens": 95524969.0,
      "step": 41710
    },
    {
      "entropy": 5.840292024612427,
      "epoch": 4.1236654804270465,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003401244906879431,
      "loss": 5.0183,
      "mean_token_accuracy": 0.2240653708577156,
      "num_tokens": 95535432.0,
      "step": 41715
    },
    {
      "entropy": 5.809587717056274,
      "epoch": 4.124159746935548,
      "grad_norm": 1.171875,
      "learning_rate": 0.00034009071047474016,
      "loss": 4.9514,
      "mean_token_accuracy": 0.23011977672576905,
      "num_tokens": 95546389.0,
      "step": 41720
    },
    {
      "entropy": 5.737720489501953,
      "epoch": 4.124654013444049,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003400569286602955,
      "loss": 4.9031,
      "mean_token_accuracy": 0.23201536983251572,
      "num_tokens": 95558398.0,
      "step": 41725
    },
    {
      "entropy": 5.767891359329224,
      "epoch": 4.12514827995255,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003400231452454404,
      "loss": 4.9855,
      "mean_token_accuracy": 0.22396766841411592,
      "num_tokens": 95570707.0,
      "step": 41730
    },
    {
      "entropy": 5.75676155090332,
      "epoch": 4.1256425464610516,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00033998936023100585,
      "loss": 4.9429,
      "mean_token_accuracy": 0.22870891839265822,
      "num_tokens": 95583187.0,
      "step": 41735
    },
    {
      "entropy": 5.848313474655152,
      "epoch": 4.126136812969553,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000339955573617823,
      "loss": 4.9706,
      "mean_token_accuracy": 0.2235844373703003,
      "num_tokens": 95595121.0,
      "step": 41740
    },
    {
      "entropy": 5.845588636398316,
      "epoch": 4.126631079478055,
      "grad_norm": 1.15625,
      "learning_rate": 0.00033992178540672287,
      "loss": 5.006,
      "mean_token_accuracy": 0.22153065651655196,
      "num_tokens": 95606992.0,
      "step": 41745
    },
    {
      "entropy": 5.7620946884155275,
      "epoch": 4.127125345986556,
      "grad_norm": 1.390625,
      "learning_rate": 0.00033988799559853675,
      "loss": 4.916,
      "mean_token_accuracy": 0.23088352382183075,
      "num_tokens": 95618460.0,
      "step": 41750
    },
    {
      "entropy": 5.824718856811524,
      "epoch": 4.1276196124950575,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000339854204194096,
      "loss": 4.9726,
      "mean_token_accuracy": 0.227671879529953,
      "num_tokens": 95629690.0,
      "step": 41755
    },
    {
      "entropy": 5.678577136993408,
      "epoch": 4.128113879003559,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00033982041119423156,
      "loss": 4.7584,
      "mean_token_accuracy": 0.24933973103761672,
      "num_tokens": 95640732.0,
      "step": 41760
    },
    {
      "entropy": 5.712448167800903,
      "epoch": 4.12860814551206,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003397866165997751,
      "loss": 4.9168,
      "mean_token_accuracy": 0.23318539410829545,
      "num_tokens": 95651292.0,
      "step": 41765
    },
    {
      "entropy": 5.735893821716308,
      "epoch": 4.129102412020561,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003397528204115578,
      "loss": 4.9115,
      "mean_token_accuracy": 0.23449735045433046,
      "num_tokens": 95662499.0,
      "step": 41770
    },
    {
      "entropy": 5.776894998550415,
      "epoch": 4.1295966785290625,
      "grad_norm": 1.203125,
      "learning_rate": 0.000339719022630411,
      "loss": 4.949,
      "mean_token_accuracy": 0.2284708723425865,
      "num_tokens": 95674097.0,
      "step": 41775
    },
    {
      "entropy": 5.823663854598999,
      "epoch": 4.130090945037564,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003396852232571663,
      "loss": 4.9111,
      "mean_token_accuracy": 0.22634780108928682,
      "num_tokens": 95684917.0,
      "step": 41780
    },
    {
      "entropy": 5.834529685974121,
      "epoch": 4.130585211546066,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000339651422292655,
      "loss": 5.0003,
      "mean_token_accuracy": 0.2175817906856537,
      "num_tokens": 95696588.0,
      "step": 41785
    },
    {
      "entropy": 5.775815725326538,
      "epoch": 4.131079478054567,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00033961761973770865,
      "loss": 4.922,
      "mean_token_accuracy": 0.23107039630413057,
      "num_tokens": 95707246.0,
      "step": 41790
    },
    {
      "entropy": 5.7839210510253904,
      "epoch": 4.1315737445630685,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003395838155931589,
      "loss": 4.935,
      "mean_token_accuracy": 0.22675368338823318,
      "num_tokens": 95717432.0,
      "step": 41795
    },
    {
      "entropy": 5.765988731384278,
      "epoch": 4.13206801107157,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00033955000985983727,
      "loss": 4.9127,
      "mean_token_accuracy": 0.22939050197601318,
      "num_tokens": 95728102.0,
      "step": 41800
    },
    {
      "entropy": 5.783460807800293,
      "epoch": 4.132562277580071,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003395162025385753,
      "loss": 4.9032,
      "mean_token_accuracy": 0.22884804010391235,
      "num_tokens": 95738691.0,
      "step": 41805
    },
    {
      "entropy": 5.728170585632324,
      "epoch": 4.133056544088572,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003394823936302048,
      "loss": 4.9318,
      "mean_token_accuracy": 0.22652331590652466,
      "num_tokens": 95750716.0,
      "step": 41810
    },
    {
      "entropy": 5.772882747650146,
      "epoch": 4.1335508105970735,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003394485831355574,
      "loss": 4.9012,
      "mean_token_accuracy": 0.22775209844112396,
      "num_tokens": 95764130.0,
      "step": 41815
    },
    {
      "entropy": 5.7816931247711185,
      "epoch": 4.134045077105576,
      "grad_norm": 1.25,
      "learning_rate": 0.00033941477105546493,
      "loss": 4.9648,
      "mean_token_accuracy": 0.22361060231924057,
      "num_tokens": 95776222.0,
      "step": 41820
    },
    {
      "entropy": 5.805946731567383,
      "epoch": 4.134539343614077,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00033938095739075903,
      "loss": 4.914,
      "mean_token_accuracy": 0.22808899134397506,
      "num_tokens": 95786986.0,
      "step": 41825
    },
    {
      "entropy": 5.7656896114349365,
      "epoch": 4.135033610122578,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00033934714214227163,
      "loss": 4.8685,
      "mean_token_accuracy": 0.23751085698604585,
      "num_tokens": 95797206.0,
      "step": 41830
    },
    {
      "entropy": 5.797953987121582,
      "epoch": 4.1355278766310795,
      "grad_norm": 1.25,
      "learning_rate": 0.00033931332531083447,
      "loss": 4.9914,
      "mean_token_accuracy": 0.23031650930643083,
      "num_tokens": 95809277.0,
      "step": 41835
    },
    {
      "entropy": 5.83391695022583,
      "epoch": 4.136022143139581,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003392795068972796,
      "loss": 4.9849,
      "mean_token_accuracy": 0.218040369451046,
      "num_tokens": 95820981.0,
      "step": 41840
    },
    {
      "entropy": 5.803598546981812,
      "epoch": 4.136516409648082,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00033924568690243885,
      "loss": 4.9312,
      "mean_token_accuracy": 0.22516164779663086,
      "num_tokens": 95832287.0,
      "step": 41845
    },
    {
      "entropy": 5.7404989242553714,
      "epoch": 4.137010676156583,
      "grad_norm": 1.25,
      "learning_rate": 0.00033921186532714426,
      "loss": 4.902,
      "mean_token_accuracy": 0.23195397704839707,
      "num_tokens": 95842650.0,
      "step": 41850
    },
    {
      "entropy": 5.789996004104614,
      "epoch": 4.137504942665085,
      "grad_norm": 1.5,
      "learning_rate": 0.0003391780421722279,
      "loss": 4.9765,
      "mean_token_accuracy": 0.2248440146446228,
      "num_tokens": 95852946.0,
      "step": 41855
    },
    {
      "entropy": 5.764745569229126,
      "epoch": 4.137999209173587,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003391442174385216,
      "loss": 4.9749,
      "mean_token_accuracy": 0.22567938417196273,
      "num_tokens": 95864373.0,
      "step": 41860
    },
    {
      "entropy": 5.78674578666687,
      "epoch": 4.138493475682088,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00033911039112685773,
      "loss": 4.9359,
      "mean_token_accuracy": 0.2320415958762169,
      "num_tokens": 95875932.0,
      "step": 41865
    },
    {
      "entropy": 5.78922963142395,
      "epoch": 4.138987742190589,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003390765632380681,
      "loss": 4.916,
      "mean_token_accuracy": 0.23035893142223357,
      "num_tokens": 95887677.0,
      "step": 41870
    },
    {
      "entropy": 5.7822576522827145,
      "epoch": 4.13948200869909,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003390427337729852,
      "loss": 4.9467,
      "mean_token_accuracy": 0.22918060272932053,
      "num_tokens": 95898966.0,
      "step": 41875
    },
    {
      "entropy": 5.766458892822266,
      "epoch": 4.139976275207592,
      "grad_norm": 1.25,
      "learning_rate": 0.00033900890273244117,
      "loss": 4.9692,
      "mean_token_accuracy": 0.2234041139483452,
      "num_tokens": 95909645.0,
      "step": 41880
    },
    {
      "entropy": 5.781626844406128,
      "epoch": 4.140470541716093,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00033897507011726815,
      "loss": 4.9202,
      "mean_token_accuracy": 0.2280095174908638,
      "num_tokens": 95920365.0,
      "step": 41885
    },
    {
      "entropy": 5.784172391891479,
      "epoch": 4.140964808224595,
      "grad_norm": 1.21875,
      "learning_rate": 0.00033894123592829843,
      "loss": 4.967,
      "mean_token_accuracy": 0.2250792443752289,
      "num_tokens": 95932770.0,
      "step": 41890
    },
    {
      "entropy": 5.793095493316651,
      "epoch": 4.141459074733096,
      "grad_norm": 1.3125,
      "learning_rate": 0.00033890740016636445,
      "loss": 4.9064,
      "mean_token_accuracy": 0.22764774262905121,
      "num_tokens": 95942792.0,
      "step": 41895
    },
    {
      "entropy": 5.752836132049561,
      "epoch": 4.141953341241598,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003388735628322985,
      "loss": 4.8426,
      "mean_token_accuracy": 0.23235284835100173,
      "num_tokens": 95953286.0,
      "step": 41900
    },
    {
      "entropy": 5.781254291534424,
      "epoch": 4.142447607750099,
      "grad_norm": 1.234375,
      "learning_rate": 0.00033883972392693295,
      "loss": 4.9655,
      "mean_token_accuracy": 0.22055097818374633,
      "num_tokens": 95964419.0,
      "step": 41905
    },
    {
      "entropy": 5.847490739822388,
      "epoch": 4.1429418742586,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00033880588345110044,
      "loss": 4.9896,
      "mean_token_accuracy": 0.22141663134098052,
      "num_tokens": 95975558.0,
      "step": 41910
    },
    {
      "entropy": 5.875650024414062,
      "epoch": 4.143436140767101,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003387720414056331,
      "loss": 5.0509,
      "mean_token_accuracy": 0.22070700079202651,
      "num_tokens": 95987421.0,
      "step": 41915
    },
    {
      "entropy": 5.847297859191895,
      "epoch": 4.143930407275603,
      "grad_norm": 1.25,
      "learning_rate": 0.00033873819779136374,
      "loss": 4.9499,
      "mean_token_accuracy": 0.22520407289266586,
      "num_tokens": 95999765.0,
      "step": 41920
    },
    {
      "entropy": 5.74351487159729,
      "epoch": 4.144424673784105,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003387043526091248,
      "loss": 4.9498,
      "mean_token_accuracy": 0.2274530276656151,
      "num_tokens": 96010795.0,
      "step": 41925
    },
    {
      "entropy": 5.79376974105835,
      "epoch": 4.144918940292606,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000338670505859749,
      "loss": 4.9707,
      "mean_token_accuracy": 0.22821931838989257,
      "num_tokens": 96022885.0,
      "step": 41930
    },
    {
      "entropy": 5.786641263961792,
      "epoch": 4.145413206801107,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00033863665754406874,
      "loss": 4.9339,
      "mean_token_accuracy": 0.2335208147764206,
      "num_tokens": 96034636.0,
      "step": 41935
    },
    {
      "entropy": 5.806053113937378,
      "epoch": 4.145907473309609,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003386028076629169,
      "loss": 4.9398,
      "mean_token_accuracy": 0.2244931563735008,
      "num_tokens": 96044880.0,
      "step": 41940
    },
    {
      "entropy": 5.7830219745635985,
      "epoch": 4.14640173981811,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003385689562171261,
      "loss": 4.9953,
      "mean_token_accuracy": 0.22309200018644332,
      "num_tokens": 96056856.0,
      "step": 41945
    },
    {
      "entropy": 5.7712890148162845,
      "epoch": 4.146896006326611,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003385351032075291,
      "loss": 4.8746,
      "mean_token_accuracy": 0.2320969060063362,
      "num_tokens": 96068843.0,
      "step": 41950
    },
    {
      "entropy": 5.754102611541748,
      "epoch": 4.147390272835112,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003385012486349588,
      "loss": 4.9188,
      "mean_token_accuracy": 0.23111398816108703,
      "num_tokens": 96079780.0,
      "step": 41955
    },
    {
      "entropy": 5.722984266281128,
      "epoch": 4.147884539343614,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003384673925002478,
      "loss": 4.8778,
      "mean_token_accuracy": 0.23693632185459138,
      "num_tokens": 96090747.0,
      "step": 41960
    },
    {
      "entropy": 5.753922939300537,
      "epoch": 4.148378805852116,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00033843353480422924,
      "loss": 4.8568,
      "mean_token_accuracy": 0.2386754870414734,
      "num_tokens": 96102895.0,
      "step": 41965
    },
    {
      "entropy": 5.7776401996612545,
      "epoch": 4.148873072360617,
      "grad_norm": 1.25,
      "learning_rate": 0.0003383996755477358,
      "loss": 4.906,
      "mean_token_accuracy": 0.22976725548505783,
      "num_tokens": 96113860.0,
      "step": 41970
    },
    {
      "entropy": 5.733097934722901,
      "epoch": 4.149367338869118,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003383658147316006,
      "loss": 4.8939,
      "mean_token_accuracy": 0.23128117024898528,
      "num_tokens": 96124154.0,
      "step": 41975
    },
    {
      "entropy": 5.770008373260498,
      "epoch": 4.14986160537762,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00033833195235665636,
      "loss": 4.9653,
      "mean_token_accuracy": 0.22047383338212967,
      "num_tokens": 96137137.0,
      "step": 41980
    },
    {
      "entropy": 5.752041149139404,
      "epoch": 4.150355871886121,
      "grad_norm": 1.1875,
      "learning_rate": 0.00033829808842373633,
      "loss": 4.9188,
      "mean_token_accuracy": 0.23087651133537293,
      "num_tokens": 96148683.0,
      "step": 41985
    },
    {
      "entropy": 5.690748167037964,
      "epoch": 4.150850138394622,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00033826422293367354,
      "loss": 4.8403,
      "mean_token_accuracy": 0.23442936837673187,
      "num_tokens": 96159240.0,
      "step": 41990
    },
    {
      "entropy": 5.797614336013794,
      "epoch": 4.151344404903123,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000338230355887301,
      "loss": 4.9834,
      "mean_token_accuracy": 0.2292370840907097,
      "num_tokens": 96170479.0,
      "step": 41995
    },
    {
      "entropy": 5.831102752685547,
      "epoch": 4.1518386714116255,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000338196487285452,
      "loss": 4.9615,
      "mean_token_accuracy": 0.22549984455108643,
      "num_tokens": 96181242.0,
      "step": 42000
    },
    {
      "epoch": 4.1518386714116255,
      "eval_entropy": 5.533678781502362,
      "eval_loss": 5.009167194366455,
      "eval_mean_token_accuracy": 0.23254616406790884,
      "eval_num_tokens": 96181242.0,
      "eval_runtime": 20.5425,
      "eval_samples_per_second": 1582.72,
      "eval_steps_per_second": 197.883,
      "step": 42000
    },
    {
      "entropy": 5.85185227394104,
      "epoch": 4.152332937920127,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003381626171289595,
      "loss": 5.0112,
      "mean_token_accuracy": 0.22486206591129304,
      "num_tokens": 96192049.0,
      "step": 42005
    },
    {
      "entropy": 5.745854711532592,
      "epoch": 4.152827204428628,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00033812874541865685,
      "loss": 4.8998,
      "mean_token_accuracy": 0.23356309086084365,
      "num_tokens": 96203334.0,
      "step": 42010
    },
    {
      "entropy": 5.745328617095947,
      "epoch": 4.153321470937129,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003380948721553773,
      "loss": 4.9403,
      "mean_token_accuracy": 0.22854797244071962,
      "num_tokens": 96214790.0,
      "step": 42015
    },
    {
      "entropy": 5.809774446487427,
      "epoch": 4.153815737445631,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003380609973399541,
      "loss": 5.0114,
      "mean_token_accuracy": 0.22639363259077072,
      "num_tokens": 96227832.0,
      "step": 42020
    },
    {
      "entropy": 5.922045946121216,
      "epoch": 4.154310003954132,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003380271209732205,
      "loss": 5.0891,
      "mean_token_accuracy": 0.21870527863502504,
      "num_tokens": 96238501.0,
      "step": 42025
    },
    {
      "entropy": 5.7928547859191895,
      "epoch": 4.154804270462633,
      "grad_norm": 1.375,
      "learning_rate": 0.00033799324305601,
      "loss": 4.9373,
      "mean_token_accuracy": 0.2300085484981537,
      "num_tokens": 96249605.0,
      "step": 42030
    },
    {
      "entropy": 5.790035820007324,
      "epoch": 4.155298536971134,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00033795936358915594,
      "loss": 4.9913,
      "mean_token_accuracy": 0.22123833745718002,
      "num_tokens": 96259754.0,
      "step": 42035
    },
    {
      "entropy": 5.748666477203369,
      "epoch": 4.1557928034796365,
      "grad_norm": 1.25,
      "learning_rate": 0.00033792548257349184,
      "loss": 4.9344,
      "mean_token_accuracy": 0.23001475036144256,
      "num_tokens": 96272566.0,
      "step": 42040
    },
    {
      "entropy": 5.810056924819946,
      "epoch": 4.156287069988138,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003378916000098511,
      "loss": 4.955,
      "mean_token_accuracy": 0.2234086126089096,
      "num_tokens": 96283560.0,
      "step": 42045
    },
    {
      "entropy": 5.833791494369507,
      "epoch": 4.156781336496639,
      "grad_norm": 1.28125,
      "learning_rate": 0.00033785771589906724,
      "loss": 4.926,
      "mean_token_accuracy": 0.22924749851226806,
      "num_tokens": 96295892.0,
      "step": 42050
    },
    {
      "entropy": 5.806897783279419,
      "epoch": 4.15727560300514,
      "grad_norm": 1.28125,
      "learning_rate": 0.00033782383024197365,
      "loss": 4.9655,
      "mean_token_accuracy": 0.2264800950884819,
      "num_tokens": 96307561.0,
      "step": 42055
    },
    {
      "entropy": 5.666009616851807,
      "epoch": 4.157769869513642,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003377899430394043,
      "loss": 4.8547,
      "mean_token_accuracy": 0.24008260369300843,
      "num_tokens": 96318782.0,
      "step": 42060
    },
    {
      "entropy": 5.777905750274658,
      "epoch": 4.158264136022143,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003377560542921925,
      "loss": 4.9237,
      "mean_token_accuracy": 0.23734949976205827,
      "num_tokens": 96329090.0,
      "step": 42065
    },
    {
      "entropy": 5.878195810317993,
      "epoch": 4.158758402530644,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00033772216400117206,
      "loss": 5.0285,
      "mean_token_accuracy": 0.2202135592699051,
      "num_tokens": 96338681.0,
      "step": 42070
    },
    {
      "entropy": 5.738410711288452,
      "epoch": 4.159252669039146,
      "grad_norm": 1.265625,
      "learning_rate": 0.00033768827216717676,
      "loss": 4.883,
      "mean_token_accuracy": 0.23235542923212052,
      "num_tokens": 96348772.0,
      "step": 42075
    },
    {
      "entropy": 5.797555637359619,
      "epoch": 4.1597469355476475,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00033765437879104004,
      "loss": 4.9756,
      "mean_token_accuracy": 0.22796508818864822,
      "num_tokens": 96360772.0,
      "step": 42080
    },
    {
      "entropy": 5.834882020950317,
      "epoch": 4.160241202056149,
      "grad_norm": 1.296875,
      "learning_rate": 0.00033762048387359593,
      "loss": 4.9892,
      "mean_token_accuracy": 0.22521932870149614,
      "num_tokens": 96371729.0,
      "step": 42085
    },
    {
      "entropy": 5.808661460876465,
      "epoch": 4.16073546856465,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003375865874156783,
      "loss": 4.9443,
      "mean_token_accuracy": 0.22904234677553176,
      "num_tokens": 96382422.0,
      "step": 42090
    },
    {
      "entropy": 5.809221076965332,
      "epoch": 4.161229735073151,
      "grad_norm": 1.25,
      "learning_rate": 0.0003375526894181208,
      "loss": 4.9748,
      "mean_token_accuracy": 0.22535169422626494,
      "num_tokens": 96393411.0,
      "step": 42095
    },
    {
      "entropy": 5.713190650939941,
      "epoch": 4.161724001581653,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003375187898817575,
      "loss": 4.8651,
      "mean_token_accuracy": 0.23432136625051497,
      "num_tokens": 96404133.0,
      "step": 42100
    },
    {
      "entropy": 5.831609487533569,
      "epoch": 4.162218268090154,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003374848888074223,
      "loss": 4.9837,
      "mean_token_accuracy": 0.22206680476665497,
      "num_tokens": 96416651.0,
      "step": 42105
    },
    {
      "entropy": 5.8268908023834225,
      "epoch": 4.162712534598656,
      "grad_norm": 1.265625,
      "learning_rate": 0.00033745098619594906,
      "loss": 4.9889,
      "mean_token_accuracy": 0.22366662472486495,
      "num_tokens": 96429831.0,
      "step": 42110
    },
    {
      "entropy": 5.72743673324585,
      "epoch": 4.163206801107157,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000337417082048172,
      "loss": 4.8792,
      "mean_token_accuracy": 0.2341558203101158,
      "num_tokens": 96440227.0,
      "step": 42115
    },
    {
      "entropy": 5.789227867126465,
      "epoch": 4.1637010676156585,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00033738317636492496,
      "loss": 4.965,
      "mean_token_accuracy": 0.22414795011281968,
      "num_tokens": 96451975.0,
      "step": 42120
    },
    {
      "entropy": 5.724147891998291,
      "epoch": 4.16419533412416,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003373492691470421,
      "loss": 4.8604,
      "mean_token_accuracy": 0.23121273815631865,
      "num_tokens": 96462416.0,
      "step": 42125
    },
    {
      "entropy": 5.783310747146606,
      "epoch": 4.164689600632661,
      "grad_norm": 1.390625,
      "learning_rate": 0.00033731536039535747,
      "loss": 4.9266,
      "mean_token_accuracy": 0.228180493414402,
      "num_tokens": 96475450.0,
      "step": 42130
    },
    {
      "entropy": 5.852667999267578,
      "epoch": 4.165183867141162,
      "grad_norm": 1.25,
      "learning_rate": 0.00033728145011070537,
      "loss": 5.0241,
      "mean_token_accuracy": 0.21669515818357468,
      "num_tokens": 96486885.0,
      "step": 42135
    },
    {
      "entropy": 5.752664232254029,
      "epoch": 4.1656781336496636,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00033724753829391996,
      "loss": 4.9065,
      "mean_token_accuracy": 0.22948591858148576,
      "num_tokens": 96498401.0,
      "step": 42140
    },
    {
      "entropy": 5.728801679611206,
      "epoch": 4.166172400158166,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003372136249458355,
      "loss": 4.9219,
      "mean_token_accuracy": 0.23119271397590638,
      "num_tokens": 96509824.0,
      "step": 42145
    },
    {
      "entropy": 5.755523443222046,
      "epoch": 4.166666666666667,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003371797100672861,
      "loss": 4.9261,
      "mean_token_accuracy": 0.2321494162082672,
      "num_tokens": 96521261.0,
      "step": 42150
    },
    {
      "entropy": 5.830453634262085,
      "epoch": 4.167160933175168,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00033714579365910617,
      "loss": 4.9643,
      "mean_token_accuracy": 0.22389705926179887,
      "num_tokens": 96533015.0,
      "step": 42155
    },
    {
      "entropy": 5.8386155128479,
      "epoch": 4.1676551996836695,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003371118757221302,
      "loss": 4.9272,
      "mean_token_accuracy": 0.2348856583237648,
      "num_tokens": 96543076.0,
      "step": 42160
    },
    {
      "entropy": 5.746624708175659,
      "epoch": 4.168149466192171,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00033707795625719243,
      "loss": 4.874,
      "mean_token_accuracy": 0.2364471897482872,
      "num_tokens": 96554003.0,
      "step": 42165
    },
    {
      "entropy": 5.710572195053101,
      "epoch": 4.168643732700672,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003370440352651273,
      "loss": 4.8629,
      "mean_token_accuracy": 0.23399485349655152,
      "num_tokens": 96567005.0,
      "step": 42170
    },
    {
      "entropy": 5.7476881980896,
      "epoch": 4.169137999209173,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00033701011274676917,
      "loss": 4.9295,
      "mean_token_accuracy": 0.22620429545640947,
      "num_tokens": 96577403.0,
      "step": 42175
    },
    {
      "entropy": 5.811063432693482,
      "epoch": 4.169632265717675,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00033697618870295274,
      "loss": 4.9218,
      "mean_token_accuracy": 0.23040825873613358,
      "num_tokens": 96588511.0,
      "step": 42180
    },
    {
      "entropy": 5.84308876991272,
      "epoch": 4.170126532226177,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003369422631345124,
      "loss": 4.9864,
      "mean_token_accuracy": 0.22484734058380126,
      "num_tokens": 96599422.0,
      "step": 42185
    },
    {
      "entropy": 5.795851421356201,
      "epoch": 4.170620798734678,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00033690833604228283,
      "loss": 4.9697,
      "mean_token_accuracy": 0.22498807162046433,
      "num_tokens": 96611230.0,
      "step": 42190
    },
    {
      "entropy": 5.702176713943482,
      "epoch": 4.171115065243179,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003368744074270986,
      "loss": 4.8636,
      "mean_token_accuracy": 0.24526637494564058,
      "num_tokens": 96622341.0,
      "step": 42195
    },
    {
      "entropy": 5.7744073390960695,
      "epoch": 4.1716093317516805,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00033684047728979426,
      "loss": 4.892,
      "mean_token_accuracy": 0.22548425048589707,
      "num_tokens": 96632987.0,
      "step": 42200
    },
    {
      "entropy": 5.740166521072387,
      "epoch": 4.172103598260182,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003368065456312047,
      "loss": 4.9358,
      "mean_token_accuracy": 0.23454935550689698,
      "num_tokens": 96645262.0,
      "step": 42205
    },
    {
      "entropy": 5.737353181838989,
      "epoch": 4.172597864768683,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00033677261245216444,
      "loss": 4.9058,
      "mean_token_accuracy": 0.23785075545310974,
      "num_tokens": 96657873.0,
      "step": 42210
    },
    {
      "entropy": 5.773782730102539,
      "epoch": 4.173092131277184,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003367386777535084,
      "loss": 4.9743,
      "mean_token_accuracy": 0.2265040621161461,
      "num_tokens": 96668845.0,
      "step": 42215
    },
    {
      "entropy": 5.7452627658844,
      "epoch": 4.173586397785686,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00033670474153607127,
      "loss": 4.8884,
      "mean_token_accuracy": 0.23731583803892137,
      "num_tokens": 96680787.0,
      "step": 42220
    },
    {
      "entropy": 5.805747175216675,
      "epoch": 4.174080664294188,
      "grad_norm": 1.328125,
      "learning_rate": 0.00033667080380068786,
      "loss": 4.9334,
      "mean_token_accuracy": 0.23413556069135666,
      "num_tokens": 96692783.0,
      "step": 42225
    },
    {
      "entropy": 5.899657773971557,
      "epoch": 4.174574930802689,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00033663686454819326,
      "loss": 5.086,
      "mean_token_accuracy": 0.21467451006174088,
      "num_tokens": 96704000.0,
      "step": 42230
    },
    {
      "entropy": 5.792047119140625,
      "epoch": 4.17506919731119,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003366029237794221,
      "loss": 4.9794,
      "mean_token_accuracy": 0.22851030379533768,
      "num_tokens": 96715791.0,
      "step": 42235
    },
    {
      "entropy": 5.772071790695191,
      "epoch": 4.1755634638196915,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003365689814952095,
      "loss": 4.8715,
      "mean_token_accuracy": 0.23212676495313644,
      "num_tokens": 96726510.0,
      "step": 42240
    },
    {
      "entropy": 5.771815061569214,
      "epoch": 4.176057730328193,
      "grad_norm": 1.265625,
      "learning_rate": 0.00033653503769639045,
      "loss": 4.9258,
      "mean_token_accuracy": 0.22770150750875473,
      "num_tokens": 96738147.0,
      "step": 42245
    },
    {
      "entropy": 5.749001169204712,
      "epoch": 4.176551996836694,
      "grad_norm": 1.546875,
      "learning_rate": 0.0003365010923837999,
      "loss": 4.9401,
      "mean_token_accuracy": 0.23551376461982726,
      "num_tokens": 96749553.0,
      "step": 42250
    },
    {
      "entropy": 5.819431114196777,
      "epoch": 4.177046263345196,
      "grad_norm": 1.25,
      "learning_rate": 0.0003364671455582729,
      "loss": 4.9453,
      "mean_token_accuracy": 0.22581138014793395,
      "num_tokens": 96761728.0,
      "step": 42255
    },
    {
      "entropy": 5.770155763626098,
      "epoch": 4.177540529853697,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00033643319722064457,
      "loss": 4.8949,
      "mean_token_accuracy": 0.23484534919261932,
      "num_tokens": 96772382.0,
      "step": 42260
    },
    {
      "entropy": 5.636398077011108,
      "epoch": 4.178034796362199,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003363992473717502,
      "loss": 4.7689,
      "mean_token_accuracy": 0.2548342704772949,
      "num_tokens": 96783844.0,
      "step": 42265
    },
    {
      "entropy": 5.755936145782471,
      "epoch": 4.1785290628707,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003363652960124246,
      "loss": 4.9963,
      "mean_token_accuracy": 0.22712917178869246,
      "num_tokens": 96795256.0,
      "step": 42270
    },
    {
      "entropy": 5.79455075263977,
      "epoch": 4.179023329379201,
      "grad_norm": 1.34375,
      "learning_rate": 0.00033633134314350335,
      "loss": 4.9887,
      "mean_token_accuracy": 0.22693213522434236,
      "num_tokens": 96807085.0,
      "step": 42275
    },
    {
      "entropy": 5.790366363525391,
      "epoch": 4.179517595887702,
      "grad_norm": 1.3125,
      "learning_rate": 0.00033629738876582156,
      "loss": 4.8891,
      "mean_token_accuracy": 0.2325239360332489,
      "num_tokens": 96818351.0,
      "step": 42280
    },
    {
      "entropy": 5.762495660781861,
      "epoch": 4.180011862396204,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00033626343288021437,
      "loss": 4.8851,
      "mean_token_accuracy": 0.23101907968521118,
      "num_tokens": 96830093.0,
      "step": 42285
    },
    {
      "entropy": 5.785472869873047,
      "epoch": 4.180506128904705,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00033622947548751737,
      "loss": 4.9323,
      "mean_token_accuracy": 0.22993303686380387,
      "num_tokens": 96842124.0,
      "step": 42290
    },
    {
      "entropy": 5.757175493240356,
      "epoch": 4.181000395413207,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00033619551658856567,
      "loss": 4.9744,
      "mean_token_accuracy": 0.21841660737991334,
      "num_tokens": 96853764.0,
      "step": 42295
    },
    {
      "entropy": 5.764651155471801,
      "epoch": 4.181494661921708,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003361615561841948,
      "loss": 4.925,
      "mean_token_accuracy": 0.2266773283481598,
      "num_tokens": 96865356.0,
      "step": 42300
    },
    {
      "entropy": 5.773111391067505,
      "epoch": 4.18198892843021,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00033612759427524016,
      "loss": 4.9253,
      "mean_token_accuracy": 0.23118874728679656,
      "num_tokens": 96876747.0,
      "step": 42305
    },
    {
      "entropy": 5.756291246414184,
      "epoch": 4.182483194938711,
      "grad_norm": 1.28125,
      "learning_rate": 0.00033609363086253727,
      "loss": 4.8756,
      "mean_token_accuracy": 0.23611579239368438,
      "num_tokens": 96888723.0,
      "step": 42310
    },
    {
      "entropy": 5.798770999908447,
      "epoch": 4.182977461447212,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003360596659469215,
      "loss": 4.9507,
      "mean_token_accuracy": 0.2297533854842186,
      "num_tokens": 96900772.0,
      "step": 42315
    },
    {
      "entropy": 5.828733348846436,
      "epoch": 4.183471727955713,
      "grad_norm": 1.25,
      "learning_rate": 0.0003360256995292285,
      "loss": 4.9663,
      "mean_token_accuracy": 0.22560759484767914,
      "num_tokens": 96911739.0,
      "step": 42320
    },
    {
      "entropy": 5.76828236579895,
      "epoch": 4.183965994464215,
      "grad_norm": 1.25,
      "learning_rate": 0.00033599173161029383,
      "loss": 4.8803,
      "mean_token_accuracy": 0.2311836913228035,
      "num_tokens": 96923135.0,
      "step": 42325
    },
    {
      "entropy": 5.741417407989502,
      "epoch": 4.184460260972717,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003359577621909531,
      "loss": 4.8929,
      "mean_token_accuracy": 0.23270026445388795,
      "num_tokens": 96934835.0,
      "step": 42330
    },
    {
      "entropy": 5.801205253601074,
      "epoch": 4.184954527481218,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000335923791272042,
      "loss": 5.0142,
      "mean_token_accuracy": 0.2273244485259056,
      "num_tokens": 96944870.0,
      "step": 42335
    },
    {
      "entropy": 5.881455326080323,
      "epoch": 4.185448793989719,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003358898188543961,
      "loss": 5.028,
      "mean_token_accuracy": 0.22067079544067383,
      "num_tokens": 96955562.0,
      "step": 42340
    },
    {
      "entropy": 5.77050371170044,
      "epoch": 4.185943060498221,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00033585584493885123,
      "loss": 4.9238,
      "mean_token_accuracy": 0.23313048779964446,
      "num_tokens": 96966445.0,
      "step": 42345
    },
    {
      "entropy": 5.837284517288208,
      "epoch": 4.186437327006722,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00033582186952624314,
      "loss": 5.0434,
      "mean_token_accuracy": 0.2157254621386528,
      "num_tokens": 96978162.0,
      "step": 42350
    },
    {
      "entropy": 5.7980328559875485,
      "epoch": 4.186931593515223,
      "grad_norm": 1.171875,
      "learning_rate": 0.00033578789261740766,
      "loss": 4.9486,
      "mean_token_accuracy": 0.22475740313529968,
      "num_tokens": 96989348.0,
      "step": 42355
    },
    {
      "entropy": 5.850470447540284,
      "epoch": 4.187425860023724,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00033575391421318063,
      "loss": 4.9671,
      "mean_token_accuracy": 0.2292979434132576,
      "num_tokens": 97001291.0,
      "step": 42360
    },
    {
      "entropy": 5.764464282989502,
      "epoch": 4.187920126532227,
      "grad_norm": 1.1875,
      "learning_rate": 0.00033571993431439777,
      "loss": 5.0261,
      "mean_token_accuracy": 0.2221033275127411,
      "num_tokens": 97013568.0,
      "step": 42365
    },
    {
      "entropy": 5.783184385299682,
      "epoch": 4.188414393040728,
      "grad_norm": 1.375,
      "learning_rate": 0.0003356859529218952,
      "loss": 4.8943,
      "mean_token_accuracy": 0.22824627310037612,
      "num_tokens": 97024009.0,
      "step": 42370
    },
    {
      "entropy": 5.785395574569702,
      "epoch": 4.188908659549229,
      "grad_norm": 1.328125,
      "learning_rate": 0.00033565197003650876,
      "loss": 4.9495,
      "mean_token_accuracy": 0.2278834268450737,
      "num_tokens": 97035582.0,
      "step": 42375
    },
    {
      "entropy": 5.758455896377564,
      "epoch": 4.18940292605773,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003356179856590744,
      "loss": 4.9837,
      "mean_token_accuracy": 0.22433990985155106,
      "num_tokens": 97048193.0,
      "step": 42380
    },
    {
      "entropy": 5.777410268783569,
      "epoch": 4.189897192566232,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00033558399979042826,
      "loss": 4.9649,
      "mean_token_accuracy": 0.23234901279211045,
      "num_tokens": 97058977.0,
      "step": 42385
    },
    {
      "entropy": 5.823492240905762,
      "epoch": 4.190391459074733,
      "grad_norm": 1.296875,
      "learning_rate": 0.00033555001243140627,
      "loss": 4.9663,
      "mean_token_accuracy": 0.2248974099755287,
      "num_tokens": 97070757.0,
      "step": 42390
    },
    {
      "entropy": 5.857982444763183,
      "epoch": 4.190885725583234,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003355160235828446,
      "loss": 4.98,
      "mean_token_accuracy": 0.22358244210481643,
      "num_tokens": 97081062.0,
      "step": 42395
    },
    {
      "entropy": 5.793693017959595,
      "epoch": 4.191379992091736,
      "grad_norm": 1.3125,
      "learning_rate": 0.00033548203324557937,
      "loss": 4.9494,
      "mean_token_accuracy": 0.22338467985391616,
      "num_tokens": 97093090.0,
      "step": 42400
    },
    {
      "entropy": 5.773223066329956,
      "epoch": 4.1918742586002375,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00033544804142044673,
      "loss": 4.9568,
      "mean_token_accuracy": 0.22342406809329987,
      "num_tokens": 97103936.0,
      "step": 42405
    },
    {
      "entropy": 5.7943885803222654,
      "epoch": 4.192368525108739,
      "grad_norm": 1.34375,
      "learning_rate": 0.00033541404810828284,
      "loss": 4.9703,
      "mean_token_accuracy": 0.21995019912719727,
      "num_tokens": 97115797.0,
      "step": 42410
    },
    {
      "entropy": 5.840319585800171,
      "epoch": 4.19286279161724,
      "grad_norm": 1.375,
      "learning_rate": 0.00033538005330992405,
      "loss": 4.9905,
      "mean_token_accuracy": 0.23146339058876036,
      "num_tokens": 97126920.0,
      "step": 42415
    },
    {
      "entropy": 5.759835577011108,
      "epoch": 4.193357058125741,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003353460570262066,
      "loss": 4.8082,
      "mean_token_accuracy": 0.23873633295297622,
      "num_tokens": 97138236.0,
      "step": 42420
    },
    {
      "entropy": 5.810373210906983,
      "epoch": 4.193851324634243,
      "grad_norm": 1.21875,
      "learning_rate": 0.00033531205925796674,
      "loss": 4.9728,
      "mean_token_accuracy": 0.2259466528892517,
      "num_tokens": 97150666.0,
      "step": 42425
    },
    {
      "entropy": 5.790635585784912,
      "epoch": 4.194345591142744,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003352780600060409,
      "loss": 4.913,
      "mean_token_accuracy": 0.2279958412051201,
      "num_tokens": 97161747.0,
      "step": 42430
    },
    {
      "entropy": 5.770146656036377,
      "epoch": 4.194839857651246,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00033524405927126536,
      "loss": 4.9333,
      "mean_token_accuracy": 0.22841710597276688,
      "num_tokens": 97173582.0,
      "step": 42435
    },
    {
      "entropy": 5.787171936035156,
      "epoch": 4.195334124159747,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003352100570544767,
      "loss": 4.8791,
      "mean_token_accuracy": 0.23148779273033143,
      "num_tokens": 97184707.0,
      "step": 42440
    },
    {
      "entropy": 5.743264198303223,
      "epoch": 4.1958283906682485,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003351760533565113,
      "loss": 4.9201,
      "mean_token_accuracy": 0.2276039019227028,
      "num_tokens": 97196496.0,
      "step": 42445
    },
    {
      "entropy": 5.739260911941528,
      "epoch": 4.19632265717675,
      "grad_norm": 1.375,
      "learning_rate": 0.00033514204817820564,
      "loss": 4.9156,
      "mean_token_accuracy": 0.2269454151391983,
      "num_tokens": 97207291.0,
      "step": 42450
    },
    {
      "entropy": 5.811512184143067,
      "epoch": 4.196816923685251,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003351080415203962,
      "loss": 5.0051,
      "mean_token_accuracy": 0.22769604474306107,
      "num_tokens": 97219001.0,
      "step": 42455
    },
    {
      "entropy": 5.821756744384766,
      "epoch": 4.197311190193752,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003350740333839197,
      "loss": 4.9325,
      "mean_token_accuracy": 0.23457292467355728,
      "num_tokens": 97230451.0,
      "step": 42460
    },
    {
      "entropy": 5.8236688613891605,
      "epoch": 4.197805456702254,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003350400237696126,
      "loss": 4.9494,
      "mean_token_accuracy": 0.22913777083158493,
      "num_tokens": 97241752.0,
      "step": 42465
    },
    {
      "entropy": 5.724374580383301,
      "epoch": 4.198299723210755,
      "grad_norm": 1.25,
      "learning_rate": 0.00033500601267831163,
      "loss": 4.9359,
      "mean_token_accuracy": 0.22753970474004745,
      "num_tokens": 97253453.0,
      "step": 42470
    },
    {
      "entropy": 5.7651282787323,
      "epoch": 4.198793989719257,
      "grad_norm": 1.171875,
      "learning_rate": 0.00033497200011085344,
      "loss": 4.8894,
      "mean_token_accuracy": 0.23417162597179414,
      "num_tokens": 97265963.0,
      "step": 42475
    },
    {
      "entropy": 5.699686050415039,
      "epoch": 4.199288256227758,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003349379860680747,
      "loss": 4.8875,
      "mean_token_accuracy": 0.23095064014196395,
      "num_tokens": 97278311.0,
      "step": 42480
    },
    {
      "entropy": 5.817659854888916,
      "epoch": 4.1997825227362595,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003349039705508123,
      "loss": 4.9071,
      "mean_token_accuracy": 0.22736934721469879,
      "num_tokens": 97289437.0,
      "step": 42485
    },
    {
      "entropy": 5.745040798187256,
      "epoch": 4.200276789244761,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003348699535599028,
      "loss": 4.8753,
      "mean_token_accuracy": 0.23689419031143188,
      "num_tokens": 97300623.0,
      "step": 42490
    },
    {
      "entropy": 5.777965068817139,
      "epoch": 4.200771055753262,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003348359350961833,
      "loss": 4.953,
      "mean_token_accuracy": 0.23104121685028076,
      "num_tokens": 97311584.0,
      "step": 42495
    },
    {
      "entropy": 5.777185773849487,
      "epoch": 4.201265322261763,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003348019151604904,
      "loss": 4.9249,
      "mean_token_accuracy": 0.22905660420656204,
      "num_tokens": 97323536.0,
      "step": 42500
    },
    {
      "entropy": 5.750197219848633,
      "epoch": 4.201759588770265,
      "grad_norm": 1.296875,
      "learning_rate": 0.00033476789375366106,
      "loss": 4.9347,
      "mean_token_accuracy": 0.23070674389600754,
      "num_tokens": 97333203.0,
      "step": 42505
    },
    {
      "entropy": 5.749302673339844,
      "epoch": 4.202253855278767,
      "grad_norm": 1.296875,
      "learning_rate": 0.00033473387087653244,
      "loss": 4.9233,
      "mean_token_accuracy": 0.2303317129611969,
      "num_tokens": 97343970.0,
      "step": 42510
    },
    {
      "entropy": 5.791309595108032,
      "epoch": 4.202748121787268,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003346998465299412,
      "loss": 4.9467,
      "mean_token_accuracy": 0.22809857428073882,
      "num_tokens": 97355636.0,
      "step": 42515
    },
    {
      "entropy": 5.837331485748291,
      "epoch": 4.203242388295769,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00033466582071472453,
      "loss": 4.965,
      "mean_token_accuracy": 0.2257887005805969,
      "num_tokens": 97366756.0,
      "step": 42520
    },
    {
      "entropy": 5.798159790039063,
      "epoch": 4.2037366548042705,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003346317934317194,
      "loss": 5.008,
      "mean_token_accuracy": 0.22483496963977814,
      "num_tokens": 97379304.0,
      "step": 42525
    },
    {
      "entropy": 5.768119955062867,
      "epoch": 4.204230921312772,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003345977646817629,
      "loss": 4.8859,
      "mean_token_accuracy": 0.23493802100419997,
      "num_tokens": 97390162.0,
      "step": 42530
    },
    {
      "entropy": 5.78220386505127,
      "epoch": 4.204725187821273,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003345637344656921,
      "loss": 4.9649,
      "mean_token_accuracy": 0.22246380895376205,
      "num_tokens": 97401380.0,
      "step": 42535
    },
    {
      "entropy": 5.766062450408936,
      "epoch": 4.205219454329774,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00033452970278434425,
      "loss": 4.9988,
      "mean_token_accuracy": 0.22351522892713546,
      "num_tokens": 97413056.0,
      "step": 42540
    },
    {
      "entropy": 5.745073556900024,
      "epoch": 4.2057137208382755,
      "grad_norm": 1.140625,
      "learning_rate": 0.00033449566963855644,
      "loss": 4.9071,
      "mean_token_accuracy": 0.23116191923618318,
      "num_tokens": 97424143.0,
      "step": 42545
    },
    {
      "entropy": 5.769159412384033,
      "epoch": 4.206207987346778,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000334461635029166,
      "loss": 4.9198,
      "mean_token_accuracy": 0.2286946713924408,
      "num_tokens": 97435417.0,
      "step": 42550
    },
    {
      "entropy": 5.822003126144409,
      "epoch": 4.206702253855279,
      "grad_norm": 1.25,
      "learning_rate": 0.00033442759895701,
      "loss": 5.052,
      "mean_token_accuracy": 0.21964351534843446,
      "num_tokens": 97446769.0,
      "step": 42555
    },
    {
      "entropy": 5.822015619277954,
      "epoch": 4.20719652036378,
      "grad_norm": 1.328125,
      "learning_rate": 0.00033439356142292594,
      "loss": 5.0,
      "mean_token_accuracy": 0.22792186588048935,
      "num_tokens": 97456878.0,
      "step": 42560
    },
    {
      "entropy": 5.769874429702758,
      "epoch": 4.2076907868722815,
      "grad_norm": 1.34375,
      "learning_rate": 0.00033435952242775096,
      "loss": 4.9282,
      "mean_token_accuracy": 0.23150235712528228,
      "num_tokens": 97467593.0,
      "step": 42565
    },
    {
      "entropy": 5.773221158981324,
      "epoch": 4.208185053380783,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003343254819723226,
      "loss": 4.9098,
      "mean_token_accuracy": 0.23176957666873932,
      "num_tokens": 97478024.0,
      "step": 42570
    },
    {
      "entropy": 5.696961307525635,
      "epoch": 4.208679319889284,
      "grad_norm": 1.25,
      "learning_rate": 0.0003342914400574782,
      "loss": 4.8314,
      "mean_token_accuracy": 0.23367846012115479,
      "num_tokens": 97489028.0,
      "step": 42575
    },
    {
      "entropy": 5.719100666046143,
      "epoch": 4.209173586397785,
      "grad_norm": 1.390625,
      "learning_rate": 0.00033425739668405515,
      "loss": 4.8887,
      "mean_token_accuracy": 0.2301653653383255,
      "num_tokens": 97500521.0,
      "step": 42580
    },
    {
      "entropy": 5.879099607467651,
      "epoch": 4.209667852906287,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00033422335185289093,
      "loss": 5.0855,
      "mean_token_accuracy": 0.21434663534164428,
      "num_tokens": 97511781.0,
      "step": 42585
    },
    {
      "entropy": 5.788574647903443,
      "epoch": 4.210162119414789,
      "grad_norm": 1.328125,
      "learning_rate": 0.00033418930556482305,
      "loss": 4.9209,
      "mean_token_accuracy": 0.22930293083190917,
      "num_tokens": 97523232.0,
      "step": 42590
    },
    {
      "entropy": 5.729521656036377,
      "epoch": 4.21065638592329,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003341552578206891,
      "loss": 4.8879,
      "mean_token_accuracy": 0.23320727348327636,
      "num_tokens": 97534072.0,
      "step": 42595
    },
    {
      "entropy": 5.765132236480713,
      "epoch": 4.211150652431791,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003341212086213266,
      "loss": 4.9128,
      "mean_token_accuracy": 0.2346050411462784,
      "num_tokens": 97546480.0,
      "step": 42600
    },
    {
      "entropy": 5.790809583663941,
      "epoch": 4.2116449189402925,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00033408715796757316,
      "loss": 4.9126,
      "mean_token_accuracy": 0.2208746165037155,
      "num_tokens": 97557840.0,
      "step": 42605
    },
    {
      "entropy": 5.835300016403198,
      "epoch": 4.212139185448794,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003340531058602666,
      "loss": 4.9745,
      "mean_token_accuracy": 0.22754185050725936,
      "num_tokens": 97570244.0,
      "step": 42610
    },
    {
      "entropy": 5.68229112625122,
      "epoch": 4.212633451957295,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003340190523002444,
      "loss": 4.7454,
      "mean_token_accuracy": 0.2451862722635269,
      "num_tokens": 97580093.0,
      "step": 42615
    },
    {
      "entropy": 5.764474248886108,
      "epoch": 4.213127718465797,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00033398499728834436,
      "loss": 4.9688,
      "mean_token_accuracy": 0.22876218259334563,
      "num_tokens": 97592134.0,
      "step": 42620
    },
    {
      "entropy": 5.775645542144775,
      "epoch": 4.213621984974298,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003339509408254042,
      "loss": 4.9668,
      "mean_token_accuracy": 0.22776615470647812,
      "num_tokens": 97604881.0,
      "step": 42625
    },
    {
      "entropy": 5.756598949432373,
      "epoch": 4.2141162514828,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003339168829122618,
      "loss": 4.9378,
      "mean_token_accuracy": 0.22826602756977082,
      "num_tokens": 97615916.0,
      "step": 42630
    },
    {
      "entropy": 5.8059889793396,
      "epoch": 4.214610517991301,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003338828235497549,
      "loss": 4.9581,
      "mean_token_accuracy": 0.23205505758523942,
      "num_tokens": 97627184.0,
      "step": 42635
    },
    {
      "entropy": 5.827702569961548,
      "epoch": 4.215104784499802,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00033384876273872145,
      "loss": 4.9709,
      "mean_token_accuracy": 0.2198186472058296,
      "num_tokens": 97637981.0,
      "step": 42640
    },
    {
      "entropy": 5.81188850402832,
      "epoch": 4.2155990510083035,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003338147004799993,
      "loss": 4.9967,
      "mean_token_accuracy": 0.22466234862804413,
      "num_tokens": 97649103.0,
      "step": 42645
    },
    {
      "entropy": 5.745229721069336,
      "epoch": 4.216093317516805,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003337806367744263,
      "loss": 4.8695,
      "mean_token_accuracy": 0.23384003192186356,
      "num_tokens": 97661380.0,
      "step": 42650
    },
    {
      "entropy": 5.779924011230468,
      "epoch": 4.216587584025307,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003337465716228406,
      "loss": 4.9465,
      "mean_token_accuracy": 0.22884174585342407,
      "num_tokens": 97672283.0,
      "step": 42655
    },
    {
      "entropy": 5.760346412658691,
      "epoch": 4.217081850533808,
      "grad_norm": 1.390625,
      "learning_rate": 0.00033371250502608015,
      "loss": 4.957,
      "mean_token_accuracy": 0.22185326218605042,
      "num_tokens": 97683918.0,
      "step": 42660
    },
    {
      "entropy": 5.82781572341919,
      "epoch": 4.217576117042309,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003336784369849829,
      "loss": 4.9104,
      "mean_token_accuracy": 0.23541373908519744,
      "num_tokens": 97695391.0,
      "step": 42665
    },
    {
      "entropy": 5.79263916015625,
      "epoch": 4.218070383550811,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003336443675003871,
      "loss": 4.9378,
      "mean_token_accuracy": 0.23558879196643828,
      "num_tokens": 97705932.0,
      "step": 42670
    },
    {
      "entropy": 5.772899007797241,
      "epoch": 4.218564650059312,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003336102965731306,
      "loss": 4.9245,
      "mean_token_accuracy": 0.23542657345533372,
      "num_tokens": 97717298.0,
      "step": 42675
    },
    {
      "entropy": 5.742224597930909,
      "epoch": 4.219058916567813,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003335762242040519,
      "loss": 4.9096,
      "mean_token_accuracy": 0.22689489424228668,
      "num_tokens": 97727631.0,
      "step": 42680
    },
    {
      "entropy": 5.718751859664917,
      "epoch": 4.219553183076314,
      "grad_norm": 1.4375,
      "learning_rate": 0.00033354215039398886,
      "loss": 4.8922,
      "mean_token_accuracy": 0.23641736060380936,
      "num_tokens": 97738363.0,
      "step": 42685
    },
    {
      "entropy": 5.793566942214966,
      "epoch": 4.220047449584817,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003335080751437799,
      "loss": 4.9351,
      "mean_token_accuracy": 0.23322536796331406,
      "num_tokens": 97750413.0,
      "step": 42690
    },
    {
      "entropy": 5.853152894973755,
      "epoch": 4.220541716093318,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003334739984542632,
      "loss": 5.0274,
      "mean_token_accuracy": 0.21706680655479432,
      "num_tokens": 97762284.0,
      "step": 42695
    },
    {
      "entropy": 5.702807760238647,
      "epoch": 4.221035982601819,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003334399203262771,
      "loss": 4.903,
      "mean_token_accuracy": 0.22967411577701569,
      "num_tokens": 97774006.0,
      "step": 42700
    },
    {
      "entropy": 5.825601863861084,
      "epoch": 4.22153024911032,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003334058407606598,
      "loss": 5.0081,
      "mean_token_accuracy": 0.222252556681633,
      "num_tokens": 97785373.0,
      "step": 42705
    },
    {
      "entropy": 5.8319169044494625,
      "epoch": 4.222024515618822,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00033337175975824986,
      "loss": 4.9621,
      "mean_token_accuracy": 0.22701905518770218,
      "num_tokens": 97796455.0,
      "step": 42710
    },
    {
      "entropy": 5.7989696025848385,
      "epoch": 4.222518782127323,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00033333767731988557,
      "loss": 4.9224,
      "mean_token_accuracy": 0.22690904289484023,
      "num_tokens": 97808248.0,
      "step": 42715
    },
    {
      "entropy": 5.797176647186279,
      "epoch": 4.223013048635824,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003333035934464052,
      "loss": 4.9684,
      "mean_token_accuracy": 0.22452993988990783,
      "num_tokens": 97821318.0,
      "step": 42720
    },
    {
      "entropy": 5.673232555389404,
      "epoch": 4.223507315144325,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003332695081386476,
      "loss": 4.9054,
      "mean_token_accuracy": 0.2260928452014923,
      "num_tokens": 97832213.0,
      "step": 42725
    },
    {
      "entropy": 5.866296482086182,
      "epoch": 4.224001581652828,
      "grad_norm": 1.234375,
      "learning_rate": 0.00033323542139745093,
      "loss": 5.1157,
      "mean_token_accuracy": 0.2082962140440941,
      "num_tokens": 97844577.0,
      "step": 42730
    },
    {
      "entropy": 5.844697523117065,
      "epoch": 4.224495848161329,
      "grad_norm": 1.25,
      "learning_rate": 0.00033320133322365395,
      "loss": 4.963,
      "mean_token_accuracy": 0.2305160492658615,
      "num_tokens": 97857366.0,
      "step": 42735
    },
    {
      "entropy": 5.790691280364991,
      "epoch": 4.22499011466983,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003331672436180951,
      "loss": 4.9793,
      "mean_token_accuracy": 0.22226182073354722,
      "num_tokens": 97869107.0,
      "step": 42740
    },
    {
      "entropy": 5.742409896850586,
      "epoch": 4.225484381178331,
      "grad_norm": 1.359375,
      "learning_rate": 0.000333133152581613,
      "loss": 4.8984,
      "mean_token_accuracy": 0.2426530107855797,
      "num_tokens": 97879741.0,
      "step": 42745
    },
    {
      "entropy": 5.798948192596436,
      "epoch": 4.225978647686833,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00033309906011504635,
      "loss": 4.9951,
      "mean_token_accuracy": 0.22139669358730316,
      "num_tokens": 97890025.0,
      "step": 42750
    },
    {
      "entropy": 5.7380938053131105,
      "epoch": 4.226472914195334,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00033306496621923385,
      "loss": 4.8855,
      "mean_token_accuracy": 0.23137569129467012,
      "num_tokens": 97901584.0,
      "step": 42755
    },
    {
      "entropy": 5.783602094650268,
      "epoch": 4.226967180703835,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003330308708950141,
      "loss": 4.8868,
      "mean_token_accuracy": 0.2366688385605812,
      "num_tokens": 97912264.0,
      "step": 42760
    },
    {
      "entropy": 5.81136736869812,
      "epoch": 4.227461447212337,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000332996774143226,
      "loss": 4.9289,
      "mean_token_accuracy": 0.22855636477470398,
      "num_tokens": 97922971.0,
      "step": 42765
    },
    {
      "entropy": 5.775972032546997,
      "epoch": 4.227955713720839,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003329626759647082,
      "loss": 4.9137,
      "mean_token_accuracy": 0.23125038146972657,
      "num_tokens": 97934528.0,
      "step": 42770
    },
    {
      "entropy": 5.74899525642395,
      "epoch": 4.22844998022934,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003329285763602996,
      "loss": 4.9645,
      "mean_token_accuracy": 0.23216992020606994,
      "num_tokens": 97947426.0,
      "step": 42775
    },
    {
      "entropy": 5.789149761199951,
      "epoch": 4.228944246737841,
      "grad_norm": 1.234375,
      "learning_rate": 0.000332894475330839,
      "loss": 4.9157,
      "mean_token_accuracy": 0.22987958043813705,
      "num_tokens": 97958675.0,
      "step": 42780
    },
    {
      "entropy": 5.756489324569702,
      "epoch": 4.229438513246342,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00033286037287716543,
      "loss": 4.815,
      "mean_token_accuracy": 0.24156129360198975,
      "num_tokens": 97970100.0,
      "step": 42785
    },
    {
      "entropy": 5.812063837051392,
      "epoch": 4.229932779754844,
      "grad_norm": 1.328125,
      "learning_rate": 0.00033282626900011763,
      "loss": 4.9986,
      "mean_token_accuracy": 0.21871910840272904,
      "num_tokens": 97982265.0,
      "step": 42790
    },
    {
      "entropy": 5.743394136428833,
      "epoch": 4.230427046263345,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003327921637005346,
      "loss": 4.8615,
      "mean_token_accuracy": 0.23582167625427247,
      "num_tokens": 97993804.0,
      "step": 42795
    },
    {
      "entropy": 5.771084022521973,
      "epoch": 4.230921312771846,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003327580569792554,
      "loss": 4.9483,
      "mean_token_accuracy": 0.2285108119249344,
      "num_tokens": 98005377.0,
      "step": 42800
    },
    {
      "entropy": 5.77469391822815,
      "epoch": 4.231415579280348,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00033272394883711916,
      "loss": 4.9059,
      "mean_token_accuracy": 0.23201585114002227,
      "num_tokens": 98016492.0,
      "step": 42805
    },
    {
      "entropy": 5.746594762802124,
      "epoch": 4.2319098457888495,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003326898392749647,
      "loss": 4.8984,
      "mean_token_accuracy": 0.2318544402718544,
      "num_tokens": 98027924.0,
      "step": 42810
    },
    {
      "entropy": 5.776256942749024,
      "epoch": 4.232404112297351,
      "grad_norm": 1.421875,
      "learning_rate": 0.00033265572829363125,
      "loss": 4.8876,
      "mean_token_accuracy": 0.23608448058366777,
      "num_tokens": 98039993.0,
      "step": 42815
    },
    {
      "entropy": 5.745669221878051,
      "epoch": 4.232898378805852,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00033262161589395797,
      "loss": 4.8844,
      "mean_token_accuracy": 0.23217971324920655,
      "num_tokens": 98051292.0,
      "step": 42820
    },
    {
      "entropy": 5.783245038986206,
      "epoch": 4.233392645314353,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00033258750207678403,
      "loss": 4.9699,
      "mean_token_accuracy": 0.22795554101467133,
      "num_tokens": 98063462.0,
      "step": 42825
    },
    {
      "entropy": 5.7545631408691404,
      "epoch": 4.233886911822855,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00033255338684294854,
      "loss": 4.9163,
      "mean_token_accuracy": 0.23314686566591264,
      "num_tokens": 98074960.0,
      "step": 42830
    },
    {
      "entropy": 5.813656663894653,
      "epoch": 4.234381178331356,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003325192701932908,
      "loss": 4.9111,
      "mean_token_accuracy": 0.23063104450702668,
      "num_tokens": 98086266.0,
      "step": 42835
    },
    {
      "entropy": 5.841728353500367,
      "epoch": 4.234875444839858,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003324851521286501,
      "loss": 4.9848,
      "mean_token_accuracy": 0.22389378100633622,
      "num_tokens": 98097276.0,
      "step": 42840
    },
    {
      "entropy": 5.839176750183105,
      "epoch": 4.235369711348359,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003324510326498658,
      "loss": 4.9983,
      "mean_token_accuracy": 0.22117218971252442,
      "num_tokens": 98108462.0,
      "step": 42845
    },
    {
      "entropy": 5.751401424407959,
      "epoch": 4.2358639778568605,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003324169117577771,
      "loss": 4.9058,
      "mean_token_accuracy": 0.23278176188468933,
      "num_tokens": 98120384.0,
      "step": 42850
    },
    {
      "entropy": 5.788185119628906,
      "epoch": 4.236358244365362,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003323827894532236,
      "loss": 4.9395,
      "mean_token_accuracy": 0.23105270862579347,
      "num_tokens": 98132281.0,
      "step": 42855
    },
    {
      "entropy": 5.803438138961792,
      "epoch": 4.236852510873863,
      "grad_norm": 1.34375,
      "learning_rate": 0.00033234866573704443,
      "loss": 4.9636,
      "mean_token_accuracy": 0.2211210161447525,
      "num_tokens": 98142027.0,
      "step": 42860
    },
    {
      "entropy": 5.78488335609436,
      "epoch": 4.237346777382364,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003323145406100792,
      "loss": 4.9248,
      "mean_token_accuracy": 0.22945363372564315,
      "num_tokens": 98152452.0,
      "step": 42865
    },
    {
      "entropy": 5.777389812469482,
      "epoch": 4.237841043890866,
      "grad_norm": 1.296875,
      "learning_rate": 0.00033228041407316753,
      "loss": 4.9808,
      "mean_token_accuracy": 0.2221358373761177,
      "num_tokens": 98162968.0,
      "step": 42870
    },
    {
      "entropy": 5.761802005767822,
      "epoch": 4.238335310399368,
      "grad_norm": 1.296875,
      "learning_rate": 0.00033224628612714863,
      "loss": 4.8894,
      "mean_token_accuracy": 0.2369239702820778,
      "num_tokens": 98174768.0,
      "step": 42875
    },
    {
      "entropy": 5.850575256347656,
      "epoch": 4.238829576907869,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003322121567728623,
      "loss": 5.0211,
      "mean_token_accuracy": 0.22095994800329208,
      "num_tokens": 98185602.0,
      "step": 42880
    },
    {
      "entropy": 5.750828886032105,
      "epoch": 4.23932384341637,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000332178026011148,
      "loss": 4.8437,
      "mean_token_accuracy": 0.24256231188774108,
      "num_tokens": 98197142.0,
      "step": 42885
    },
    {
      "entropy": 5.785069274902344,
      "epoch": 4.2398181099248715,
      "grad_norm": 1.328125,
      "learning_rate": 0.00033214389384284536,
      "loss": 5.0271,
      "mean_token_accuracy": 0.2202734872698784,
      "num_tokens": 98209396.0,
      "step": 42890
    },
    {
      "entropy": 5.820632886886597,
      "epoch": 4.240312376433373,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003321097602687941,
      "loss": 4.9619,
      "mean_token_accuracy": 0.2264051541686058,
      "num_tokens": 98221099.0,
      "step": 42895
    },
    {
      "entropy": 5.7308117866516115,
      "epoch": 4.240806642941874,
      "grad_norm": 1.25,
      "learning_rate": 0.0003320756252898339,
      "loss": 4.8704,
      "mean_token_accuracy": 0.2408662647008896,
      "num_tokens": 98232475.0,
      "step": 42900
    },
    {
      "entropy": 5.7716302394866945,
      "epoch": 4.241300909450375,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003320414889068044,
      "loss": 4.9587,
      "mean_token_accuracy": 0.22904403507709503,
      "num_tokens": 98243819.0,
      "step": 42905
    },
    {
      "entropy": 5.845891618728638,
      "epoch": 4.2417951759588775,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003320073511205454,
      "loss": 4.9841,
      "mean_token_accuracy": 0.22430334240198135,
      "num_tokens": 98254695.0,
      "step": 42910
    },
    {
      "entropy": 5.735282754898071,
      "epoch": 4.242289442467379,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00033197321193189684,
      "loss": 4.8939,
      "mean_token_accuracy": 0.2293570414185524,
      "num_tokens": 98265882.0,
      "step": 42915
    },
    {
      "entropy": 5.726566505432129,
      "epoch": 4.24278370897588,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003319390713416984,
      "loss": 4.8842,
      "mean_token_accuracy": 0.23846326768398285,
      "num_tokens": 98277273.0,
      "step": 42920
    },
    {
      "entropy": 5.6971132278442385,
      "epoch": 4.243277975484381,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003319049293507899,
      "loss": 4.8091,
      "mean_token_accuracy": 0.23734006583690642,
      "num_tokens": 98287197.0,
      "step": 42925
    },
    {
      "entropy": 5.799550628662109,
      "epoch": 4.2437722419928825,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00033187078596001143,
      "loss": 4.9896,
      "mean_token_accuracy": 0.2240047961473465,
      "num_tokens": 98299264.0,
      "step": 42930
    },
    {
      "entropy": 5.767678451538086,
      "epoch": 4.244266508501384,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00033183664117020266,
      "loss": 4.9042,
      "mean_token_accuracy": 0.22502769231796266,
      "num_tokens": 98310521.0,
      "step": 42935
    },
    {
      "entropy": 5.772043895721436,
      "epoch": 4.244760775009885,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003318024949822038,
      "loss": 4.9917,
      "mean_token_accuracy": 0.227236770093441,
      "num_tokens": 98322806.0,
      "step": 42940
    },
    {
      "entropy": 5.810443687438965,
      "epoch": 4.245255041518386,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003317683473968547,
      "loss": 4.9357,
      "mean_token_accuracy": 0.22973594069480896,
      "num_tokens": 98334825.0,
      "step": 42945
    },
    {
      "entropy": 5.726431179046631,
      "epoch": 4.245749308026888,
      "grad_norm": 1.28125,
      "learning_rate": 0.00033173419841499546,
      "loss": 4.9873,
      "mean_token_accuracy": 0.22865829467773438,
      "num_tokens": 98346331.0,
      "step": 42950
    },
    {
      "entropy": 5.761148405075073,
      "epoch": 4.24624357453539,
      "grad_norm": 1.171875,
      "learning_rate": 0.00033170004803746613,
      "loss": 4.9299,
      "mean_token_accuracy": 0.23412587195634843,
      "num_tokens": 98357523.0,
      "step": 42955
    },
    {
      "entropy": 5.793221187591553,
      "epoch": 4.246737841043891,
      "grad_norm": 1.1875,
      "learning_rate": 0.00033166589626510676,
      "loss": 4.9303,
      "mean_token_accuracy": 0.2237307012081146,
      "num_tokens": 98368552.0,
      "step": 42960
    },
    {
      "entropy": 5.761098051071167,
      "epoch": 4.247232107552392,
      "grad_norm": 1.25,
      "learning_rate": 0.00033163174309875764,
      "loss": 4.9256,
      "mean_token_accuracy": 0.23658025711774827,
      "num_tokens": 98380473.0,
      "step": 42965
    },
    {
      "entropy": 5.735262060165406,
      "epoch": 4.2477263740608935,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003315975885392589,
      "loss": 4.864,
      "mean_token_accuracy": 0.2377377927303314,
      "num_tokens": 98391018.0,
      "step": 42970
    },
    {
      "entropy": 5.798591232299804,
      "epoch": 4.248220640569395,
      "grad_norm": 1.390625,
      "learning_rate": 0.00033156343258745067,
      "loss": 5.0,
      "mean_token_accuracy": 0.22954386472702026,
      "num_tokens": 98403177.0,
      "step": 42975
    },
    {
      "entropy": 5.826274394989014,
      "epoch": 4.248714907077896,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003315292752441731,
      "loss": 5.0078,
      "mean_token_accuracy": 0.23046029657125472,
      "num_tokens": 98415118.0,
      "step": 42980
    },
    {
      "entropy": 5.708406066894531,
      "epoch": 4.249209173586398,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00033149511651026667,
      "loss": 4.8437,
      "mean_token_accuracy": 0.23663249462842942,
      "num_tokens": 98427951.0,
      "step": 42985
    },
    {
      "entropy": 5.794697618484497,
      "epoch": 4.249703440094899,
      "grad_norm": 1.3125,
      "learning_rate": 0.00033146095638657166,
      "loss": 4.9484,
      "mean_token_accuracy": 0.22799890041351317,
      "num_tokens": 98440935.0,
      "step": 42990
    },
    {
      "entropy": 5.732081270217895,
      "epoch": 4.250197706603401,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003314267948739283,
      "loss": 4.8723,
      "mean_token_accuracy": 0.23388630151748657,
      "num_tokens": 98452950.0,
      "step": 42995
    },
    {
      "entropy": 5.818053197860718,
      "epoch": 4.250691973111902,
      "grad_norm": 1.390625,
      "learning_rate": 0.00033139263197317704,
      "loss": 4.9618,
      "mean_token_accuracy": 0.22811376452445983,
      "num_tokens": 98463725.0,
      "step": 43000
    },
    {
      "entropy": 5.806340837478638,
      "epoch": 4.251186239620403,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003313584676851583,
      "loss": 4.9512,
      "mean_token_accuracy": 0.22498218715190887,
      "num_tokens": 98475244.0,
      "step": 43005
    },
    {
      "entropy": 5.765746593475342,
      "epoch": 4.2516805061289045,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003313243020107125,
      "loss": 4.917,
      "mean_token_accuracy": 0.22897647470235824,
      "num_tokens": 98485625.0,
      "step": 43010
    },
    {
      "entropy": 5.850155639648437,
      "epoch": 4.252174772637406,
      "grad_norm": 1.375,
      "learning_rate": 0.00033129013495068017,
      "loss": 5.0224,
      "mean_token_accuracy": 0.2202941969037056,
      "num_tokens": 98496572.0,
      "step": 43015
    },
    {
      "entropy": 5.816200828552246,
      "epoch": 4.252669039145908,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003312559665059017,
      "loss": 5.0636,
      "mean_token_accuracy": 0.21866766065359117,
      "num_tokens": 98509460.0,
      "step": 43020
    },
    {
      "entropy": 5.778518295288086,
      "epoch": 4.253163305654409,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003312217966772178,
      "loss": 4.9937,
      "mean_token_accuracy": 0.2309087783098221,
      "num_tokens": 98521415.0,
      "step": 43025
    },
    {
      "entropy": 5.783109045028686,
      "epoch": 4.25365757216291,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00033118762546546895,
      "loss": 4.8591,
      "mean_token_accuracy": 0.2348336935043335,
      "num_tokens": 98532188.0,
      "step": 43030
    },
    {
      "entropy": 5.744780206680298,
      "epoch": 4.254151838671412,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003311534528714959,
      "loss": 4.9282,
      "mean_token_accuracy": 0.2324213445186615,
      "num_tokens": 98544112.0,
      "step": 43035
    },
    {
      "entropy": 5.783785820007324,
      "epoch": 4.254646105179913,
      "grad_norm": 1.328125,
      "learning_rate": 0.00033111927889613904,
      "loss": 4.8796,
      "mean_token_accuracy": 0.22901627719402312,
      "num_tokens": 98555054.0,
      "step": 43040
    },
    {
      "entropy": 5.7991087436676025,
      "epoch": 4.255140371688414,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003310851035402393,
      "loss": 4.9364,
      "mean_token_accuracy": 0.2239627331495285,
      "num_tokens": 98565956.0,
      "step": 43045
    },
    {
      "entropy": 5.799546718597412,
      "epoch": 4.2556346381969155,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00033105092680463726,
      "loss": 4.9949,
      "mean_token_accuracy": 0.2288043513894081,
      "num_tokens": 98577976.0,
      "step": 43050
    },
    {
      "entropy": 5.787723207473755,
      "epoch": 4.256128904705417,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00033101674869017377,
      "loss": 4.9573,
      "mean_token_accuracy": 0.22977009564638137,
      "num_tokens": 98589143.0,
      "step": 43055
    },
    {
      "entropy": 5.758676958084107,
      "epoch": 4.256623171213919,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003309825691976896,
      "loss": 4.9056,
      "mean_token_accuracy": 0.2288531944155693,
      "num_tokens": 98599323.0,
      "step": 43060
    },
    {
      "entropy": 5.771621370315552,
      "epoch": 4.25711743772242,
      "grad_norm": 1.265625,
      "learning_rate": 0.00033094838832802546,
      "loss": 5.0568,
      "mean_token_accuracy": 0.22085707485675812,
      "num_tokens": 98610869.0,
      "step": 43065
    },
    {
      "entropy": 5.811565256118774,
      "epoch": 4.257611704230921,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00033091420608202235,
      "loss": 4.8738,
      "mean_token_accuracy": 0.23123203665018083,
      "num_tokens": 98622022.0,
      "step": 43070
    },
    {
      "entropy": 5.812743330001831,
      "epoch": 4.258105970739423,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00033088002246052105,
      "loss": 4.9355,
      "mean_token_accuracy": 0.23160478323698044,
      "num_tokens": 98634110.0,
      "step": 43075
    },
    {
      "entropy": 5.7493304252624515,
      "epoch": 4.258600237247924,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003308458374643626,
      "loss": 4.9542,
      "mean_token_accuracy": 0.23056052774190902,
      "num_tokens": 98645182.0,
      "step": 43080
    },
    {
      "entropy": 5.809789371490479,
      "epoch": 4.259094503756425,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003308116510943878,
      "loss": 4.9846,
      "mean_token_accuracy": 0.22502518743276595,
      "num_tokens": 98655651.0,
      "step": 43085
    },
    {
      "entropy": 5.804376268386841,
      "epoch": 4.259588770264926,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003307774633514378,
      "loss": 4.9025,
      "mean_token_accuracy": 0.2311157241463661,
      "num_tokens": 98668191.0,
      "step": 43090
    },
    {
      "entropy": 5.741907215118408,
      "epoch": 4.260083036773429,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00033074327423635356,
      "loss": 4.905,
      "mean_token_accuracy": 0.2314024716615677,
      "num_tokens": 98678479.0,
      "step": 43095
    },
    {
      "entropy": 5.728987789154052,
      "epoch": 4.26057730328193,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00033070908374997597,
      "loss": 4.793,
      "mean_token_accuracy": 0.24201630204916,
      "num_tokens": 98688702.0,
      "step": 43100
    },
    {
      "entropy": 5.774787378311157,
      "epoch": 4.261071569790431,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00033067489189314645,
      "loss": 4.9243,
      "mean_token_accuracy": 0.22774240970611573,
      "num_tokens": 98700131.0,
      "step": 43105
    },
    {
      "entropy": 5.763353157043457,
      "epoch": 4.261565836298932,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003306406986667059,
      "loss": 4.8973,
      "mean_token_accuracy": 0.22950106710195542,
      "num_tokens": 98711134.0,
      "step": 43110
    },
    {
      "entropy": 5.810732698440551,
      "epoch": 4.262060102807434,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003306065040714955,
      "loss": 4.9547,
      "mean_token_accuracy": 0.24042146205902098,
      "num_tokens": 98721892.0,
      "step": 43115
    },
    {
      "entropy": 5.698682832717895,
      "epoch": 4.262554369315935,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003305723081083565,
      "loss": 4.9539,
      "mean_token_accuracy": 0.2260197088122368,
      "num_tokens": 98734663.0,
      "step": 43120
    },
    {
      "entropy": 5.829581022262573,
      "epoch": 4.263048635824436,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003305381107781301,
      "loss": 4.9285,
      "mean_token_accuracy": 0.22908418476581574,
      "num_tokens": 98746216.0,
      "step": 43125
    },
    {
      "entropy": 5.769526481628418,
      "epoch": 4.263542902332938,
      "grad_norm": 1.453125,
      "learning_rate": 0.00033050391208165755,
      "loss": 4.8267,
      "mean_token_accuracy": 0.22889414429664612,
      "num_tokens": 98756322.0,
      "step": 43130
    },
    {
      "entropy": 5.789292097091675,
      "epoch": 4.26403716884144,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00033046971201978017,
      "loss": 5.0151,
      "mean_token_accuracy": 0.22542168945074081,
      "num_tokens": 98768256.0,
      "step": 43135
    },
    {
      "entropy": 5.804489612579346,
      "epoch": 4.264531435349941,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003304355105933393,
      "loss": 4.9662,
      "mean_token_accuracy": 0.2250652477145195,
      "num_tokens": 98780787.0,
      "step": 43140
    },
    {
      "entropy": 5.827536869049072,
      "epoch": 4.265025701858442,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00033040130780317615,
      "loss": 4.9671,
      "mean_token_accuracy": 0.22966360598802565,
      "num_tokens": 98791965.0,
      "step": 43145
    },
    {
      "entropy": 5.790524244308472,
      "epoch": 4.265519968366943,
      "grad_norm": 1.34375,
      "learning_rate": 0.00033036710365013234,
      "loss": 4.9165,
      "mean_token_accuracy": 0.22805018872022628,
      "num_tokens": 98803087.0,
      "step": 43150
    },
    {
      "entropy": 5.822976922988891,
      "epoch": 4.266014234875445,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00033033289813504916,
      "loss": 5.0361,
      "mean_token_accuracy": 0.22536057233810425,
      "num_tokens": 98814829.0,
      "step": 43155
    },
    {
      "entropy": 5.751730585098267,
      "epoch": 4.266508501383946,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003302986912587681,
      "loss": 4.9275,
      "mean_token_accuracy": 0.22343873083591462,
      "num_tokens": 98826144.0,
      "step": 43160
    },
    {
      "entropy": 5.790082836151123,
      "epoch": 4.267002767892448,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003302644830221307,
      "loss": 4.9702,
      "mean_token_accuracy": 0.22530618757009507,
      "num_tokens": 98836233.0,
      "step": 43165
    },
    {
      "entropy": 5.767870569229126,
      "epoch": 4.267497034400949,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003302302734259784,
      "loss": 4.9602,
      "mean_token_accuracy": 0.22929328829050064,
      "num_tokens": 98849011.0,
      "step": 43170
    },
    {
      "entropy": 5.781072092056275,
      "epoch": 4.267991300909451,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00033019606247115276,
      "loss": 4.8902,
      "mean_token_accuracy": 0.23799516260623932,
      "num_tokens": 98859843.0,
      "step": 43175
    },
    {
      "entropy": 5.777816390991211,
      "epoch": 4.268485567417952,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00033016185015849545,
      "loss": 4.9463,
      "mean_token_accuracy": 0.22494549602270125,
      "num_tokens": 98871261.0,
      "step": 43180
    },
    {
      "entropy": 5.746015787124634,
      "epoch": 4.268979833926453,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00033012763648884806,
      "loss": 4.9529,
      "mean_token_accuracy": 0.232180655002594,
      "num_tokens": 98882046.0,
      "step": 43185
    },
    {
      "entropy": 5.761539268493652,
      "epoch": 4.269474100434954,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00033009342146305234,
      "loss": 4.9591,
      "mean_token_accuracy": 0.2292468100786209,
      "num_tokens": 98894337.0,
      "step": 43190
    },
    {
      "entropy": 5.79749059677124,
      "epoch": 4.269968366943456,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003300592050819498,
      "loss": 4.9364,
      "mean_token_accuracy": 0.23068920522928238,
      "num_tokens": 98904766.0,
      "step": 43195
    },
    {
      "entropy": 5.74762806892395,
      "epoch": 4.270462633451958,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00033002498734638233,
      "loss": 5.0142,
      "mean_token_accuracy": 0.22265869081020356,
      "num_tokens": 98917278.0,
      "step": 43200
    },
    {
      "entropy": 5.861811780929566,
      "epoch": 4.270956899960459,
      "grad_norm": 1.140625,
      "learning_rate": 0.00032999076825719165,
      "loss": 5.0422,
      "mean_token_accuracy": 0.2192737877368927,
      "num_tokens": 98930417.0,
      "step": 43205
    },
    {
      "entropy": 5.8522233963012695,
      "epoch": 4.27145116646896,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003299565478152196,
      "loss": 4.9628,
      "mean_token_accuracy": 0.21862891018390657,
      "num_tokens": 98941105.0,
      "step": 43210
    },
    {
      "entropy": 5.7517982006073,
      "epoch": 4.2719454329774615,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003299223260213078,
      "loss": 4.905,
      "mean_token_accuracy": 0.23683905452489853,
      "num_tokens": 98953178.0,
      "step": 43215
    },
    {
      "entropy": 5.774961376190186,
      "epoch": 4.272439699485963,
      "grad_norm": 1.34375,
      "learning_rate": 0.00032988810287629827,
      "loss": 4.9052,
      "mean_token_accuracy": 0.22653292268514633,
      "num_tokens": 98964394.0,
      "step": 43220
    },
    {
      "entropy": 5.854064702987671,
      "epoch": 4.272933965994464,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00032985387838103296,
      "loss": 5.0056,
      "mean_token_accuracy": 0.22290830314159393,
      "num_tokens": 98975522.0,
      "step": 43225
    },
    {
      "entropy": 5.7996526718139645,
      "epoch": 4.273428232502965,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003298196525363537,
      "loss": 4.8958,
      "mean_token_accuracy": 0.2281111881136894,
      "num_tokens": 98987965.0,
      "step": 43230
    },
    {
      "entropy": 5.7047285556793215,
      "epoch": 4.273922499011467,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003297854253431025,
      "loss": 4.8676,
      "mean_token_accuracy": 0.2389364719390869,
      "num_tokens": 98999773.0,
      "step": 43235
    },
    {
      "entropy": 5.778940677642822,
      "epoch": 4.274416765519969,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003297511968021214,
      "loss": 4.9155,
      "mean_token_accuracy": 0.22905998677015305,
      "num_tokens": 99010487.0,
      "step": 43240
    },
    {
      "entropy": 5.730114459991455,
      "epoch": 4.27491103202847,
      "grad_norm": 1.28125,
      "learning_rate": 0.00032971696691425225,
      "loss": 4.9464,
      "mean_token_accuracy": 0.2264096647500992,
      "num_tokens": 99021876.0,
      "step": 43245
    },
    {
      "entropy": 5.76324691772461,
      "epoch": 4.275405298536971,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003296827356803373,
      "loss": 4.903,
      "mean_token_accuracy": 0.2262290194630623,
      "num_tokens": 99032507.0,
      "step": 43250
    },
    {
      "entropy": 5.846307229995728,
      "epoch": 4.2758995650454725,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003296485031012186,
      "loss": 4.9781,
      "mean_token_accuracy": 0.22036830931901932,
      "num_tokens": 99043139.0,
      "step": 43255
    },
    {
      "entropy": 5.895349550247192,
      "epoch": 4.276393831553974,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003296142691777382,
      "loss": 5.0101,
      "mean_token_accuracy": 0.2225436344742775,
      "num_tokens": 99053966.0,
      "step": 43260
    },
    {
      "entropy": 5.8293062210083,
      "epoch": 4.276888098062475,
      "grad_norm": 1.265625,
      "learning_rate": 0.00032958003391073826,
      "loss": 5.0248,
      "mean_token_accuracy": 0.22121129631996156,
      "num_tokens": 99066118.0,
      "step": 43265
    },
    {
      "entropy": 5.710164356231689,
      "epoch": 4.277382364570976,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003295457973010611,
      "loss": 4.8534,
      "mean_token_accuracy": 0.23426484018564225,
      "num_tokens": 99078898.0,
      "step": 43270
    },
    {
      "entropy": 5.792031192779541,
      "epoch": 4.2778766310794785,
      "grad_norm": 1.40625,
      "learning_rate": 0.00032951155934954877,
      "loss": 4.944,
      "mean_token_accuracy": 0.22990844398736954,
      "num_tokens": 99089816.0,
      "step": 43275
    },
    {
      "entropy": 5.811977434158325,
      "epoch": 4.27837089758798,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003294773200570437,
      "loss": 4.9355,
      "mean_token_accuracy": 0.23657287508249283,
      "num_tokens": 99101033.0,
      "step": 43280
    },
    {
      "entropy": 5.692740345001221,
      "epoch": 4.278865164096481,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003294430794243881,
      "loss": 4.8411,
      "mean_token_accuracy": 0.2332753837108612,
      "num_tokens": 99111706.0,
      "step": 43285
    },
    {
      "entropy": 5.717005014419556,
      "epoch": 4.279359430604982,
      "grad_norm": 1.15625,
      "learning_rate": 0.00032940883745242425,
      "loss": 4.8442,
      "mean_token_accuracy": 0.23728483468294143,
      "num_tokens": 99123461.0,
      "step": 43290
    },
    {
      "entropy": 5.765208148956299,
      "epoch": 4.2798536971134835,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003293745941419946,
      "loss": 4.871,
      "mean_token_accuracy": 0.23692589700222016,
      "num_tokens": 99134309.0,
      "step": 43295
    },
    {
      "entropy": 5.7740672588348385,
      "epoch": 4.280347963621985,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003293403494939415,
      "loss": 4.9485,
      "mean_token_accuracy": 0.2295371860265732,
      "num_tokens": 99145416.0,
      "step": 43300
    },
    {
      "entropy": 5.748534917831421,
      "epoch": 4.280842230130486,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00032930610350910734,
      "loss": 4.9294,
      "mean_token_accuracy": 0.23051405549049378,
      "num_tokens": 99157004.0,
      "step": 43305
    },
    {
      "entropy": 5.820102214813232,
      "epoch": 4.281336496638987,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003292718561883346,
      "loss": 4.9008,
      "mean_token_accuracy": 0.23587457686662675,
      "num_tokens": 99168563.0,
      "step": 43310
    },
    {
      "entropy": 5.821598196029663,
      "epoch": 4.2818307631474894,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003292376075324658,
      "loss": 4.914,
      "mean_token_accuracy": 0.2315291255712509,
      "num_tokens": 99180522.0,
      "step": 43315
    },
    {
      "entropy": 5.738863325119018,
      "epoch": 4.282325029655991,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00032920335754234347,
      "loss": 4.9269,
      "mean_token_accuracy": 0.22914780378341676,
      "num_tokens": 99192487.0,
      "step": 43320
    },
    {
      "entropy": 5.71196665763855,
      "epoch": 4.282819296164492,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00032916910621881015,
      "loss": 4.8969,
      "mean_token_accuracy": 0.23040175139904023,
      "num_tokens": 99203808.0,
      "step": 43325
    },
    {
      "entropy": 5.773679208755493,
      "epoch": 4.283313562672993,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00032913485356270833,
      "loss": 4.9558,
      "mean_token_accuracy": 0.22506097406148912,
      "num_tokens": 99216459.0,
      "step": 43330
    },
    {
      "entropy": 5.750085973739624,
      "epoch": 4.2838078291814945,
      "grad_norm": 1.28125,
      "learning_rate": 0.00032910059957488077,
      "loss": 4.8843,
      "mean_token_accuracy": 0.23416091352701188,
      "num_tokens": 99227244.0,
      "step": 43335
    },
    {
      "entropy": 5.813915681838989,
      "epoch": 4.284302095689996,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003290663442561701,
      "loss": 4.9569,
      "mean_token_accuracy": 0.22644461691379547,
      "num_tokens": 99239034.0,
      "step": 43340
    },
    {
      "entropy": 5.737202739715576,
      "epoch": 4.284796362198497,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00032903208760741887,
      "loss": 4.904,
      "mean_token_accuracy": 0.23592038899660112,
      "num_tokens": 99250466.0,
      "step": 43345
    },
    {
      "entropy": 5.775089406967163,
      "epoch": 4.285290628706999,
      "grad_norm": 1.265625,
      "learning_rate": 0.00032899782962947,
      "loss": 4.9546,
      "mean_token_accuracy": 0.22616446167230606,
      "num_tokens": 99261720.0,
      "step": 43350
    },
    {
      "entropy": 5.838986730575561,
      "epoch": 4.2857848952155,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003289635703231662,
      "loss": 4.9506,
      "mean_token_accuracy": 0.2297082930803299,
      "num_tokens": 99272597.0,
      "step": 43355
    },
    {
      "entropy": 5.7523571968078615,
      "epoch": 4.286279161724002,
      "grad_norm": 1.3125,
      "learning_rate": 0.00032892930968934995,
      "loss": 4.9109,
      "mean_token_accuracy": 0.2346217229962349,
      "num_tokens": 99284941.0,
      "step": 43360
    },
    {
      "entropy": 5.80360050201416,
      "epoch": 4.286773428232503,
      "grad_norm": 1.34375,
      "learning_rate": 0.00032889504772886454,
      "loss": 4.964,
      "mean_token_accuracy": 0.22827168405056,
      "num_tokens": 99296096.0,
      "step": 43365
    },
    {
      "entropy": 5.736228561401367,
      "epoch": 4.287267694741004,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003288607844425526,
      "loss": 4.8913,
      "mean_token_accuracy": 0.23487668484449387,
      "num_tokens": 99307779.0,
      "step": 43370
    },
    {
      "entropy": 5.818888425827026,
      "epoch": 4.2877619612495055,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003288265198312568,
      "loss": 5.0094,
      "mean_token_accuracy": 0.22803349643945695,
      "num_tokens": 99319179.0,
      "step": 43375
    },
    {
      "entropy": 5.8140768051147464,
      "epoch": 4.288256227758007,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003287922538958205,
      "loss": 5.0082,
      "mean_token_accuracy": 0.21821558922529222,
      "num_tokens": 99331706.0,
      "step": 43380
    },
    {
      "entropy": 5.822761917114258,
      "epoch": 4.288750494266509,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00032875798663708626,
      "loss": 5.0124,
      "mean_token_accuracy": 0.22471252530813218,
      "num_tokens": 99343527.0,
      "step": 43385
    },
    {
      "entropy": 5.760383033752442,
      "epoch": 4.28924476077501,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00032872371805589724,
      "loss": 4.9695,
      "mean_token_accuracy": 0.2258553609251976,
      "num_tokens": 99354593.0,
      "step": 43390
    },
    {
      "entropy": 5.788660764694214,
      "epoch": 4.289739027283511,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003286894481530964,
      "loss": 4.9963,
      "mean_token_accuracy": 0.22517401576042176,
      "num_tokens": 99367204.0,
      "step": 43395
    },
    {
      "entropy": 5.79248080253601,
      "epoch": 4.290233293792013,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003286551769295269,
      "loss": 4.9139,
      "mean_token_accuracy": 0.23274416476488113,
      "num_tokens": 99377846.0,
      "step": 43400
    },
    {
      "entropy": 5.795986175537109,
      "epoch": 4.290727560300514,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003286209043860317,
      "loss": 5.0161,
      "mean_token_accuracy": 0.21947432905435563,
      "num_tokens": 99388894.0,
      "step": 43405
    },
    {
      "entropy": 5.756523704528808,
      "epoch": 4.291221826809015,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003285866305234539,
      "loss": 4.9194,
      "mean_token_accuracy": 0.23076125234365463,
      "num_tokens": 99399666.0,
      "step": 43410
    },
    {
      "entropy": 5.799577617645264,
      "epoch": 4.2917160933175165,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003285523553426367,
      "loss": 4.8756,
      "mean_token_accuracy": 0.22879491299390792,
      "num_tokens": 99412390.0,
      "step": 43415
    },
    {
      "entropy": 5.785652446746826,
      "epoch": 4.292210359826019,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003285180788444232,
      "loss": 4.9478,
      "mean_token_accuracy": 0.2291867330670357,
      "num_tokens": 99424572.0,
      "step": 43420
    },
    {
      "entropy": 5.804775905609131,
      "epoch": 4.29270462633452,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003284838010296567,
      "loss": 5.0151,
      "mean_token_accuracy": 0.22339471131563188,
      "num_tokens": 99436911.0,
      "step": 43425
    },
    {
      "entropy": 5.8251588344573975,
      "epoch": 4.293198892843021,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003284495218991803,
      "loss": 4.9582,
      "mean_token_accuracy": 0.22570926696062088,
      "num_tokens": 99448259.0,
      "step": 43430
    },
    {
      "entropy": 5.819349527359009,
      "epoch": 4.293693159351522,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003284152414538374,
      "loss": 4.9485,
      "mean_token_accuracy": 0.232858307659626,
      "num_tokens": 99459807.0,
      "step": 43435
    },
    {
      "entropy": 5.841078901290894,
      "epoch": 4.294187425860024,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00032838095969447143,
      "loss": 5.0083,
      "mean_token_accuracy": 0.2254486247897148,
      "num_tokens": 99471204.0,
      "step": 43440
    },
    {
      "entropy": 5.790413665771484,
      "epoch": 4.294681692368525,
      "grad_norm": 1.25,
      "learning_rate": 0.00032834667662192546,
      "loss": 4.9805,
      "mean_token_accuracy": 0.23049046844244003,
      "num_tokens": 99483906.0,
      "step": 43445
    },
    {
      "entropy": 5.750399208068847,
      "epoch": 4.295175958877026,
      "grad_norm": 1.21875,
      "learning_rate": 0.000328312392237043,
      "loss": 4.8427,
      "mean_token_accuracy": 0.23797794431447983,
      "num_tokens": 99494088.0,
      "step": 43450
    },
    {
      "entropy": 5.751242780685425,
      "epoch": 4.295670225385528,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003282781065406674,
      "loss": 4.9267,
      "mean_token_accuracy": 0.2288760870695114,
      "num_tokens": 99504398.0,
      "step": 43455
    },
    {
      "entropy": 5.7648131370544435,
      "epoch": 4.29616449189403,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00032824381953364205,
      "loss": 4.9458,
      "mean_token_accuracy": 0.22770990580320358,
      "num_tokens": 99515574.0,
      "step": 43460
    },
    {
      "entropy": 5.742452716827392,
      "epoch": 4.296658758402531,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003282095312168106,
      "loss": 4.8772,
      "mean_token_accuracy": 0.23198122382164002,
      "num_tokens": 99528322.0,
      "step": 43465
    },
    {
      "entropy": 5.754184436798096,
      "epoch": 4.297153024911032,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00032817524159101634,
      "loss": 4.8988,
      "mean_token_accuracy": 0.23320045620203017,
      "num_tokens": 99539202.0,
      "step": 43470
    },
    {
      "entropy": 5.770015811920166,
      "epoch": 4.297647291419533,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003281409506571029,
      "loss": 4.9794,
      "mean_token_accuracy": 0.22583311051130295,
      "num_tokens": 99552190.0,
      "step": 43475
    },
    {
      "entropy": 5.797441911697388,
      "epoch": 4.298141557928035,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003281066584159138,
      "loss": 4.9969,
      "mean_token_accuracy": 0.22186241596937178,
      "num_tokens": 99564125.0,
      "step": 43480
    },
    {
      "entropy": 5.815305423736572,
      "epoch": 4.298635824436536,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00032807236486829275,
      "loss": 4.9229,
      "mean_token_accuracy": 0.2317344442009926,
      "num_tokens": 99575607.0,
      "step": 43485
    },
    {
      "entropy": 5.8023299217224125,
      "epoch": 4.299130090945037,
      "grad_norm": 1.34375,
      "learning_rate": 0.00032803807001508326,
      "loss": 4.9974,
      "mean_token_accuracy": 0.22914122343063353,
      "num_tokens": 99587054.0,
      "step": 43490
    },
    {
      "entropy": 5.81861138343811,
      "epoch": 4.299624357453539,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000328003773857129,
      "loss": 4.9609,
      "mean_token_accuracy": 0.22524659782648088,
      "num_tokens": 99598839.0,
      "step": 43495
    },
    {
      "entropy": 5.73095383644104,
      "epoch": 4.300118623962041,
      "grad_norm": 1.515625,
      "learning_rate": 0.00032796947639527366,
      "loss": 4.875,
      "mean_token_accuracy": 0.24024249464273453,
      "num_tokens": 99609953.0,
      "step": 43500
    },
    {
      "entropy": 5.679924631118775,
      "epoch": 4.300612890470542,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000327935177630361,
      "loss": 4.8477,
      "mean_token_accuracy": 0.23327869325876235,
      "num_tokens": 99620916.0,
      "step": 43505
    },
    {
      "entropy": 5.761402654647827,
      "epoch": 4.301107156979043,
      "grad_norm": 1.34375,
      "learning_rate": 0.00032790087756323476,
      "loss": 4.9493,
      "mean_token_accuracy": 0.22225290089845656,
      "num_tokens": 99632615.0,
      "step": 43510
    },
    {
      "entropy": 5.858375167846679,
      "epoch": 4.301601423487544,
      "grad_norm": 1.40625,
      "learning_rate": 0.00032786657619473874,
      "loss": 4.9329,
      "mean_token_accuracy": 0.23221989274024962,
      "num_tokens": 99643140.0,
      "step": 43515
    },
    {
      "entropy": 5.761956405639649,
      "epoch": 4.302095689996046,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003278322735257167,
      "loss": 4.9369,
      "mean_token_accuracy": 0.22405422180891038,
      "num_tokens": 99654514.0,
      "step": 43520
    },
    {
      "entropy": 5.807722663879394,
      "epoch": 4.302589956504547,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003277979695570126,
      "loss": 4.9919,
      "mean_token_accuracy": 0.22425485998392106,
      "num_tokens": 99666168.0,
      "step": 43525
    },
    {
      "entropy": 5.789457225799561,
      "epoch": 4.303084223013049,
      "grad_norm": 1.390625,
      "learning_rate": 0.00032776366428947023,
      "loss": 4.9288,
      "mean_token_accuracy": 0.22506711930036544,
      "num_tokens": 99677911.0,
      "step": 43530
    },
    {
      "entropy": 5.778418684005738,
      "epoch": 4.30357848952155,
      "grad_norm": 1.3125,
      "learning_rate": 0.00032772935772393356,
      "loss": 4.906,
      "mean_token_accuracy": 0.2303909331560135,
      "num_tokens": 99688805.0,
      "step": 43535
    },
    {
      "entropy": 5.787128353118897,
      "epoch": 4.304072756030052,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003276950498612465,
      "loss": 4.9629,
      "mean_token_accuracy": 0.22370685487985612,
      "num_tokens": 99699613.0,
      "step": 43540
    },
    {
      "entropy": 5.7627091884613035,
      "epoch": 4.304567022538553,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00032766074070225297,
      "loss": 4.9225,
      "mean_token_accuracy": 0.2237701639533043,
      "num_tokens": 99709687.0,
      "step": 43545
    },
    {
      "entropy": 5.75416898727417,
      "epoch": 4.305061289047054,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003276264302477971,
      "loss": 4.9235,
      "mean_token_accuracy": 0.23285234570503235,
      "num_tokens": 99721105.0,
      "step": 43550
    },
    {
      "entropy": 5.74481315612793,
      "epoch": 4.305555555555555,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003275921184987229,
      "loss": 4.9252,
      "mean_token_accuracy": 0.2334502950310707,
      "num_tokens": 99732979.0,
      "step": 43555
    },
    {
      "entropy": 5.7675505638122555,
      "epoch": 4.306049822064057,
      "grad_norm": 1.234375,
      "learning_rate": 0.00032755780545587444,
      "loss": 4.8915,
      "mean_token_accuracy": 0.22913219034671783,
      "num_tokens": 99743964.0,
      "step": 43560
    },
    {
      "entropy": 5.77692437171936,
      "epoch": 4.306544088572558,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003275234911200958,
      "loss": 4.9183,
      "mean_token_accuracy": 0.23403078615665435,
      "num_tokens": 99755313.0,
      "step": 43565
    },
    {
      "entropy": 5.817244052886963,
      "epoch": 4.30703835508106,
      "grad_norm": 1.40625,
      "learning_rate": 0.00032748917549223113,
      "loss": 4.8889,
      "mean_token_accuracy": 0.23346937596797943,
      "num_tokens": 99765807.0,
      "step": 43570
    },
    {
      "entropy": 5.876787328720093,
      "epoch": 4.307532621589561,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003274548585731245,
      "loss": 5.0171,
      "mean_token_accuracy": 0.21789332032203673,
      "num_tokens": 99777167.0,
      "step": 43575
    },
    {
      "entropy": 5.763022136688233,
      "epoch": 4.308026888098063,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003274205403636203,
      "loss": 4.9218,
      "mean_token_accuracy": 0.2234768182039261,
      "num_tokens": 99787022.0,
      "step": 43580
    },
    {
      "entropy": 5.838502597808838,
      "epoch": 4.308521154606564,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0003273862208645627,
      "loss": 5.0444,
      "mean_token_accuracy": 0.21600334644317626,
      "num_tokens": 99798498.0,
      "step": 43585
    },
    {
      "entropy": 5.760817813873291,
      "epoch": 4.309015421115065,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003273519000767959,
      "loss": 4.9615,
      "mean_token_accuracy": 0.23053132891654968,
      "num_tokens": 99808462.0,
      "step": 43590
    },
    {
      "entropy": 5.831525278091431,
      "epoch": 4.309509687623566,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00032731757800116417,
      "loss": 5.0252,
      "mean_token_accuracy": 0.21893076598644257,
      "num_tokens": 99820403.0,
      "step": 43595
    },
    {
      "entropy": 5.656565856933594,
      "epoch": 4.310003954132068,
      "grad_norm": 1.203125,
      "learning_rate": 0.000327283254638512,
      "loss": 4.8039,
      "mean_token_accuracy": 0.24927282631397246,
      "num_tokens": 99831802.0,
      "step": 43600
    },
    {
      "entropy": 5.759474897384644,
      "epoch": 4.31049822064057,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003272489299896837,
      "loss": 4.8698,
      "mean_token_accuracy": 0.23000855296850203,
      "num_tokens": 99844433.0,
      "step": 43605
    },
    {
      "entropy": 5.809208154678345,
      "epoch": 4.310992487149071,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003272146040555235,
      "loss": 4.9667,
      "mean_token_accuracy": 0.2286195531487465,
      "num_tokens": 99856425.0,
      "step": 43610
    },
    {
      "entropy": 5.792343950271606,
      "epoch": 4.311486753657572,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000327180276836876,
      "loss": 5.0061,
      "mean_token_accuracy": 0.22227270603179933,
      "num_tokens": 99868091.0,
      "step": 43615
    },
    {
      "entropy": 5.686750316619873,
      "epoch": 4.3119810201660735,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00032714594833458554,
      "loss": 4.8799,
      "mean_token_accuracy": 0.23801347017288207,
      "num_tokens": 99880118.0,
      "step": 43620
    },
    {
      "entropy": 5.817808675765991,
      "epoch": 4.312475286674575,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00032711161854949666,
      "loss": 4.9964,
      "mean_token_accuracy": 0.21977361142635346,
      "num_tokens": 99891397.0,
      "step": 43625
    },
    {
      "entropy": 5.803943538665772,
      "epoch": 4.312969553183076,
      "grad_norm": 1.1875,
      "learning_rate": 0.00032707728748245397,
      "loss": 4.9097,
      "mean_token_accuracy": 0.23356318473815918,
      "num_tokens": 99903184.0,
      "step": 43630
    },
    {
      "entropy": 5.813151693344116,
      "epoch": 4.313463819691577,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003270429551343019,
      "loss": 4.9793,
      "mean_token_accuracy": 0.22186863273382187,
      "num_tokens": 99913946.0,
      "step": 43635
    },
    {
      "entropy": 5.736412000656128,
      "epoch": 4.3139580862000795,
      "grad_norm": 1.390625,
      "learning_rate": 0.00032700862150588506,
      "loss": 4.8827,
      "mean_token_accuracy": 0.23413027226924896,
      "num_tokens": 99925367.0,
      "step": 43640
    },
    {
      "entropy": 5.740288305282593,
      "epoch": 4.314452352708581,
      "grad_norm": 1.421875,
      "learning_rate": 0.00032697428659804803,
      "loss": 4.9177,
      "mean_token_accuracy": 0.22883841395378113,
      "num_tokens": 99937605.0,
      "step": 43645
    },
    {
      "entropy": 5.802655982971191,
      "epoch": 4.314946619217082,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003269399504116355,
      "loss": 4.9881,
      "mean_token_accuracy": 0.22271736562252045,
      "num_tokens": 99948508.0,
      "step": 43650
    },
    {
      "entropy": 5.82040581703186,
      "epoch": 4.315440885725583,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003269056129474921,
      "loss": 4.9625,
      "mean_token_accuracy": 0.22900949716567992,
      "num_tokens": 99959436.0,
      "step": 43655
    },
    {
      "entropy": 5.795877313613891,
      "epoch": 4.3159351522340845,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003268712742064626,
      "loss": 4.9415,
      "mean_token_accuracy": 0.22824318706989288,
      "num_tokens": 99970164.0,
      "step": 43660
    },
    {
      "entropy": 5.783881330490113,
      "epoch": 4.316429418742586,
      "grad_norm": 1.296875,
      "learning_rate": 0.00032683693418939175,
      "loss": 4.9454,
      "mean_token_accuracy": 0.2307812362909317,
      "num_tokens": 99982571.0,
      "step": 43665
    },
    {
      "entropy": 5.6900654315948485,
      "epoch": 4.316923685251087,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00032680259289712414,
      "loss": 4.8306,
      "mean_token_accuracy": 0.2340555027127266,
      "num_tokens": 99994482.0,
      "step": 43670
    },
    {
      "entropy": 5.8299415588378904,
      "epoch": 4.317417951759589,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003267682503305049,
      "loss": 4.9755,
      "mean_token_accuracy": 0.22461063265800477,
      "num_tokens": 100006962.0,
      "step": 43675
    },
    {
      "entropy": 5.842651653289795,
      "epoch": 4.3179122182680905,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00032673390649037853,
      "loss": 5.0373,
      "mean_token_accuracy": 0.22794072926044465,
      "num_tokens": 100018804.0,
      "step": 43680
    },
    {
      "entropy": 5.7152337551116945,
      "epoch": 4.318406484776592,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003266995613775901,
      "loss": 4.8864,
      "mean_token_accuracy": 0.23669615387916565,
      "num_tokens": 100029876.0,
      "step": 43685
    },
    {
      "entropy": 5.820630073547363,
      "epoch": 4.318900751285093,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003266652149929844,
      "loss": 4.9643,
      "mean_token_accuracy": 0.22326465398073198,
      "num_tokens": 100042368.0,
      "step": 43690
    },
    {
      "entropy": 5.846385860443116,
      "epoch": 4.319395017793594,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00032663086733740644,
      "loss": 5.0422,
      "mean_token_accuracy": 0.22350607812404633,
      "num_tokens": 100054800.0,
      "step": 43695
    },
    {
      "entropy": 5.720472288131714,
      "epoch": 4.3198892843020955,
      "grad_norm": 1.25,
      "learning_rate": 0.0003265965184117011,
      "loss": 4.9138,
      "mean_token_accuracy": 0.2326551765203476,
      "num_tokens": 100067480.0,
      "step": 43700
    },
    {
      "entropy": 5.7658435821533205,
      "epoch": 4.320383550810597,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003265621682167135,
      "loss": 4.9783,
      "mean_token_accuracy": 0.23047412633895875,
      "num_tokens": 100080414.0,
      "step": 43705
    },
    {
      "entropy": 5.892251300811767,
      "epoch": 4.320877817319099,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003265278167532884,
      "loss": 4.9955,
      "mean_token_accuracy": 0.22642307579517365,
      "num_tokens": 100090909.0,
      "step": 43710
    },
    {
      "entropy": 5.76643648147583,
      "epoch": 4.3213720838276,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003264934640222711,
      "loss": 4.9051,
      "mean_token_accuracy": 0.22523427605628968,
      "num_tokens": 100101892.0,
      "step": 43715
    },
    {
      "entropy": 5.801924562454223,
      "epoch": 4.3218663503361014,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003264591100245066,
      "loss": 4.9646,
      "mean_token_accuracy": 0.22455812394618987,
      "num_tokens": 100115158.0,
      "step": 43720
    },
    {
      "entropy": 5.7676451206207275,
      "epoch": 4.322360616844603,
      "grad_norm": 1.328125,
      "learning_rate": 0.00032642475476084,
      "loss": 4.8877,
      "mean_token_accuracy": 0.23035601824522017,
      "num_tokens": 100125732.0,
      "step": 43725
    },
    {
      "entropy": 5.742232990264893,
      "epoch": 4.322854883353104,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003263903982321164,
      "loss": 4.8897,
      "mean_token_accuracy": 0.23454792499542237,
      "num_tokens": 100138500.0,
      "step": 43730
    },
    {
      "entropy": 5.774654197692871,
      "epoch": 4.323349149861605,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003263560404391811,
      "loss": 4.9508,
      "mean_token_accuracy": 0.23044732809066773,
      "num_tokens": 100151320.0,
      "step": 43735
    },
    {
      "entropy": 5.753339147567749,
      "epoch": 4.3238434163701065,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00032632168138287923,
      "loss": 4.914,
      "mean_token_accuracy": 0.2280087113380432,
      "num_tokens": 100161839.0,
      "step": 43740
    },
    {
      "entropy": 5.792183065414429,
      "epoch": 4.324337682878608,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000326287321064056,
      "loss": 4.9712,
      "mean_token_accuracy": 0.22993913441896438,
      "num_tokens": 100174077.0,
      "step": 43745
    },
    {
      "entropy": 5.78464412689209,
      "epoch": 4.32483194938711,
      "grad_norm": 1.34375,
      "learning_rate": 0.00032625295948355675,
      "loss": 4.9604,
      "mean_token_accuracy": 0.23379905223846437,
      "num_tokens": 100186971.0,
      "step": 43750
    },
    {
      "entropy": 5.803415250778198,
      "epoch": 4.325326215895611,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00032621859664222675,
      "loss": 4.8946,
      "mean_token_accuracy": 0.2362338125705719,
      "num_tokens": 100198976.0,
      "step": 43755
    },
    {
      "entropy": 5.79419355392456,
      "epoch": 4.325820482404112,
      "grad_norm": 1.28125,
      "learning_rate": 0.00032618423254091136,
      "loss": 4.9206,
      "mean_token_accuracy": 0.2339049234986305,
      "num_tokens": 100210052.0,
      "step": 43760
    },
    {
      "entropy": 5.803069257736206,
      "epoch": 4.326314748912614,
      "grad_norm": 1.359375,
      "learning_rate": 0.00032614986718045583,
      "loss": 4.992,
      "mean_token_accuracy": 0.22524020075798035,
      "num_tokens": 100221360.0,
      "step": 43765
    },
    {
      "entropy": 5.709546613693237,
      "epoch": 4.326809015421115,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003261155005617056,
      "loss": 4.8772,
      "mean_token_accuracy": 0.23937513530254365,
      "num_tokens": 100232746.0,
      "step": 43770
    },
    {
      "entropy": 5.77418007850647,
      "epoch": 4.327303281929616,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003260811326855063,
      "loss": 4.8729,
      "mean_token_accuracy": 0.22601770758628845,
      "num_tokens": 100243604.0,
      "step": 43775
    },
    {
      "entropy": 5.732131624221802,
      "epoch": 4.3277975484381175,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003260467635527031,
      "loss": 4.8521,
      "mean_token_accuracy": 0.24189838171005248,
      "num_tokens": 100255139.0,
      "step": 43780
    },
    {
      "entropy": 5.829389476776123,
      "epoch": 4.32829181494662,
      "grad_norm": 1.140625,
      "learning_rate": 0.00032601239316414147,
      "loss": 4.998,
      "mean_token_accuracy": 0.2263636514544487,
      "num_tokens": 100266925.0,
      "step": 43785
    },
    {
      "entropy": 5.852699184417725,
      "epoch": 4.328786081455121,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00032597802152066724,
      "loss": 4.9719,
      "mean_token_accuracy": 0.22343516796827317,
      "num_tokens": 100277237.0,
      "step": 43790
    },
    {
      "entropy": 5.730988311767578,
      "epoch": 4.329280347963622,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003259436486231257,
      "loss": 4.9022,
      "mean_token_accuracy": 0.24145944118499757,
      "num_tokens": 100289212.0,
      "step": 43795
    },
    {
      "entropy": 5.769581937789917,
      "epoch": 4.329774614472123,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003259092744723625,
      "loss": 4.9579,
      "mean_token_accuracy": 0.22664784640073776,
      "num_tokens": 100300749.0,
      "step": 43800
    },
    {
      "entropy": 5.813533639907837,
      "epoch": 4.330268880980625,
      "grad_norm": 1.359375,
      "learning_rate": 0.00032587489906922334,
      "loss": 4.8888,
      "mean_token_accuracy": 0.23205453306436538,
      "num_tokens": 100309895.0,
      "step": 43805
    },
    {
      "entropy": 5.784304761886597,
      "epoch": 4.330763147489126,
      "grad_norm": 1.328125,
      "learning_rate": 0.00032584052241455366,
      "loss": 4.9018,
      "mean_token_accuracy": 0.22921976447105408,
      "num_tokens": 100321639.0,
      "step": 43810
    },
    {
      "entropy": 5.75391321182251,
      "epoch": 4.331257413997627,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00032580614450919925,
      "loss": 4.9558,
      "mean_token_accuracy": 0.2304617941379547,
      "num_tokens": 100334461.0,
      "step": 43815
    },
    {
      "entropy": 5.787412214279175,
      "epoch": 4.3317516805061285,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003257717653540059,
      "loss": 4.9698,
      "mean_token_accuracy": 0.22787330597639083,
      "num_tokens": 100345404.0,
      "step": 43820
    },
    {
      "entropy": 5.7872556209564205,
      "epoch": 4.332245947014631,
      "grad_norm": 1.359375,
      "learning_rate": 0.00032573738494981913,
      "loss": 4.9293,
      "mean_token_accuracy": 0.22544123679399491,
      "num_tokens": 100356229.0,
      "step": 43825
    },
    {
      "entropy": 5.803913974761963,
      "epoch": 4.332740213523132,
      "grad_norm": 1.203125,
      "learning_rate": 0.00032570300329748495,
      "loss": 4.9618,
      "mean_token_accuracy": 0.231622114777565,
      "num_tokens": 100367965.0,
      "step": 43830
    },
    {
      "entropy": 5.815386772155762,
      "epoch": 4.333234480031633,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00032566862039784886,
      "loss": 4.9644,
      "mean_token_accuracy": 0.2266029089689255,
      "num_tokens": 100380435.0,
      "step": 43835
    },
    {
      "entropy": 5.71528058052063,
      "epoch": 4.333728746540134,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003256342362517569,
      "loss": 4.8555,
      "mean_token_accuracy": 0.23829855024814606,
      "num_tokens": 100391811.0,
      "step": 43840
    },
    {
      "entropy": 5.743194532394409,
      "epoch": 4.334223013048636,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003255998508600549,
      "loss": 4.9916,
      "mean_token_accuracy": 0.2225760981440544,
      "num_tokens": 100403124.0,
      "step": 43845
    },
    {
      "entropy": 5.8112959384918215,
      "epoch": 4.334717279557137,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003255654642235888,
      "loss": 4.9625,
      "mean_token_accuracy": 0.22332628071308136,
      "num_tokens": 100416047.0,
      "step": 43850
    },
    {
      "entropy": 5.815351390838623,
      "epoch": 4.335211546065638,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003255310763432043,
      "loss": 4.9264,
      "mean_token_accuracy": 0.22594129741191865,
      "num_tokens": 100427730.0,
      "step": 43855
    },
    {
      "entropy": 5.805792903900146,
      "epoch": 4.33570581257414,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003254966872197475,
      "loss": 4.9689,
      "mean_token_accuracy": 0.2294496178627014,
      "num_tokens": 100439108.0,
      "step": 43860
    },
    {
      "entropy": 5.819001340866089,
      "epoch": 4.336200079082642,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003254622968540645,
      "loss": 4.9485,
      "mean_token_accuracy": 0.23153590112924577,
      "num_tokens": 100451240.0,
      "step": 43865
    },
    {
      "entropy": 5.820013570785522,
      "epoch": 4.336694345591143,
      "grad_norm": 1.328125,
      "learning_rate": 0.000325427905247001,
      "loss": 4.9538,
      "mean_token_accuracy": 0.23107928931713104,
      "num_tokens": 100462875.0,
      "step": 43870
    },
    {
      "entropy": 5.77145562171936,
      "epoch": 4.337188612099644,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003253935123994033,
      "loss": 4.872,
      "mean_token_accuracy": 0.24394679814577103,
      "num_tokens": 100473206.0,
      "step": 43875
    },
    {
      "entropy": 5.7274768352508545,
      "epoch": 4.337682878608145,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003253591183121173,
      "loss": 4.8703,
      "mean_token_accuracy": 0.23346533328294755,
      "num_tokens": 100483905.0,
      "step": 43880
    },
    {
      "entropy": 5.768246269226074,
      "epoch": 4.338177145116647,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00032532472298598926,
      "loss": 4.839,
      "mean_token_accuracy": 0.23923424929380416,
      "num_tokens": 100495432.0,
      "step": 43885
    },
    {
      "entropy": 5.77139139175415,
      "epoch": 4.338671411625148,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00032529032642186523,
      "loss": 4.9425,
      "mean_token_accuracy": 0.2307194098830223,
      "num_tokens": 100507300.0,
      "step": 43890
    },
    {
      "entropy": 5.761772155761719,
      "epoch": 4.33916567813365,
      "grad_norm": 1.265625,
      "learning_rate": 0.00032525592862059133,
      "loss": 4.9388,
      "mean_token_accuracy": 0.2232879489660263,
      "num_tokens": 100518797.0,
      "step": 43895
    },
    {
      "entropy": 5.799756717681885,
      "epoch": 4.339659944642151,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003252215295830138,
      "loss": 4.9904,
      "mean_token_accuracy": 0.22461570501327516,
      "num_tokens": 100530250.0,
      "step": 43900
    },
    {
      "entropy": 5.739356184005738,
      "epoch": 4.340154211150653,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003251871293099788,
      "loss": 4.9059,
      "mean_token_accuracy": 0.23649802803993225,
      "num_tokens": 100541711.0,
      "step": 43905
    },
    {
      "entropy": 5.76520528793335,
      "epoch": 4.340648477659154,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003251527278023327,
      "loss": 4.9113,
      "mean_token_accuracy": 0.23077207803726196,
      "num_tokens": 100553322.0,
      "step": 43910
    },
    {
      "entropy": 5.808765268325805,
      "epoch": 4.341142744167655,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003251183250609217,
      "loss": 4.9401,
      "mean_token_accuracy": 0.22486316710710524,
      "num_tokens": 100564891.0,
      "step": 43915
    },
    {
      "entropy": 5.8058404445648195,
      "epoch": 4.341637010676156,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00032508392108659213,
      "loss": 4.9399,
      "mean_token_accuracy": 0.22869712859392166,
      "num_tokens": 100575211.0,
      "step": 43920
    },
    {
      "entropy": 5.744402122497559,
      "epoch": 4.342131277184658,
      "grad_norm": 1.3125,
      "learning_rate": 0.00032504951588019034,
      "loss": 4.9048,
      "mean_token_accuracy": 0.22673576474189758,
      "num_tokens": 100586785.0,
      "step": 43925
    },
    {
      "entropy": 5.826343631744384,
      "epoch": 4.34262554369316,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00032501510944256266,
      "loss": 4.9685,
      "mean_token_accuracy": 0.2246192827820778,
      "num_tokens": 100596834.0,
      "step": 43930
    },
    {
      "entropy": 5.750299119949341,
      "epoch": 4.343119810201661,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003249807017745555,
      "loss": 4.9351,
      "mean_token_accuracy": 0.22991953343153,
      "num_tokens": 100608059.0,
      "step": 43935
    },
    {
      "entropy": 5.788659381866455,
      "epoch": 4.343614076710162,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00032494629287701545,
      "loss": 4.9795,
      "mean_token_accuracy": 0.22982373386621474,
      "num_tokens": 100619134.0,
      "step": 43940
    },
    {
      "entropy": 5.780305242538452,
      "epoch": 4.344108343218664,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0003249118827507887,
      "loss": 4.8952,
      "mean_token_accuracy": 0.23748289197683334,
      "num_tokens": 100629963.0,
      "step": 43945
    },
    {
      "entropy": 5.894381999969482,
      "epoch": 4.344602609727165,
      "grad_norm": 1.375,
      "learning_rate": 0.0003248774713967219,
      "loss": 5.0818,
      "mean_token_accuracy": 0.21382415741682054,
      "num_tokens": 100641267.0,
      "step": 43950
    },
    {
      "entropy": 5.806288146972657,
      "epoch": 4.345096876235666,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003248430588156616,
      "loss": 4.95,
      "mean_token_accuracy": 0.22947254478931428,
      "num_tokens": 100652664.0,
      "step": 43955
    },
    {
      "entropy": 5.799053716659546,
      "epoch": 4.345591142744167,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003248086450084544,
      "loss": 4.9062,
      "mean_token_accuracy": 0.23578690588474274,
      "num_tokens": 100664792.0,
      "step": 43960
    },
    {
      "entropy": 5.856228256225586,
      "epoch": 4.3460854092526695,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003247742299759468,
      "loss": 5.017,
      "mean_token_accuracy": 0.22255590111017226,
      "num_tokens": 100679028.0,
      "step": 43965
    },
    {
      "entropy": 5.69987006187439,
      "epoch": 4.346579675761171,
      "grad_norm": 1.21875,
      "learning_rate": 0.00032473981371898536,
      "loss": 4.8813,
      "mean_token_accuracy": 0.2368882805109024,
      "num_tokens": 100691119.0,
      "step": 43970
    },
    {
      "entropy": 5.708284997940064,
      "epoch": 4.347073942269672,
      "grad_norm": 1.3125,
      "learning_rate": 0.00032470539623841675,
      "loss": 4.9007,
      "mean_token_accuracy": 0.2335027977824211,
      "num_tokens": 100702046.0,
      "step": 43975
    },
    {
      "entropy": 5.766903829574585,
      "epoch": 4.347568208778173,
      "grad_norm": 1.25,
      "learning_rate": 0.00032467097753508775,
      "loss": 4.942,
      "mean_token_accuracy": 0.23029682785272598,
      "num_tokens": 100714345.0,
      "step": 43980
    },
    {
      "entropy": 5.75723557472229,
      "epoch": 4.3480624752866746,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00032463655760984496,
      "loss": 4.9131,
      "mean_token_accuracy": 0.2349873885512352,
      "num_tokens": 100724526.0,
      "step": 43985
    },
    {
      "entropy": 5.814954090118408,
      "epoch": 4.348556741795176,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003246021364635352,
      "loss": 4.9496,
      "mean_token_accuracy": 0.22874481678009034,
      "num_tokens": 100736413.0,
      "step": 43990
    },
    {
      "entropy": 5.821660900115967,
      "epoch": 4.349051008303677,
      "grad_norm": 1.296875,
      "learning_rate": 0.00032456771409700516,
      "loss": 5.0345,
      "mean_token_accuracy": 0.22705846279859543,
      "num_tokens": 100747431.0,
      "step": 43995
    },
    {
      "entropy": 5.763138055801392,
      "epoch": 4.349545274812178,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00032453329051110156,
      "loss": 4.9464,
      "mean_token_accuracy": 0.2330540046095848,
      "num_tokens": 100758475.0,
      "step": 44000
    },
    {
      "entropy": 5.7330591678619385,
      "epoch": 4.3500395413206805,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003244988657066715,
      "loss": 4.8511,
      "mean_token_accuracy": 0.23591730892658233,
      "num_tokens": 100769145.0,
      "step": 44005
    },
    {
      "entropy": 5.731269216537475,
      "epoch": 4.350533807829182,
      "grad_norm": 1.125,
      "learning_rate": 0.00032446443968456166,
      "loss": 4.8973,
      "mean_token_accuracy": 0.24055728912353516,
      "num_tokens": 100780074.0,
      "step": 44010
    },
    {
      "entropy": 5.769398260116577,
      "epoch": 4.351028074337683,
      "grad_norm": 1.28125,
      "learning_rate": 0.00032443001244561876,
      "loss": 4.8985,
      "mean_token_accuracy": 0.2310859367251396,
      "num_tokens": 100790752.0,
      "step": 44015
    },
    {
      "entropy": 5.7649352073669435,
      "epoch": 4.351522340846184,
      "grad_norm": 1.125,
      "learning_rate": 0.00032439558399069006,
      "loss": 5.0108,
      "mean_token_accuracy": 0.22790098935365677,
      "num_tokens": 100801803.0,
      "step": 44020
    },
    {
      "entropy": 5.773670816421509,
      "epoch": 4.3520166073546855,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00032436115432062226,
      "loss": 4.8883,
      "mean_token_accuracy": 0.2344165936112404,
      "num_tokens": 100813221.0,
      "step": 44025
    },
    {
      "entropy": 5.837436389923096,
      "epoch": 4.352510873863187,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00032432672343626234,
      "loss": 4.8986,
      "mean_token_accuracy": 0.23521172106266022,
      "num_tokens": 100825159.0,
      "step": 44030
    },
    {
      "entropy": 5.813451051712036,
      "epoch": 4.353005140371688,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00032429229133845743,
      "loss": 4.9553,
      "mean_token_accuracy": 0.22492115050554276,
      "num_tokens": 100836282.0,
      "step": 44035
    },
    {
      "entropy": 5.783771610260009,
      "epoch": 4.35349940688019,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003242578580280546,
      "loss": 4.8935,
      "mean_token_accuracy": 0.22847584635019302,
      "num_tokens": 100847543.0,
      "step": 44040
    },
    {
      "entropy": 5.702975702285767,
      "epoch": 4.3539936733886915,
      "grad_norm": 1.421875,
      "learning_rate": 0.00032422342350590065,
      "loss": 4.8674,
      "mean_token_accuracy": 0.2368446409702301,
      "num_tokens": 100859038.0,
      "step": 44045
    },
    {
      "entropy": 5.7889069557189945,
      "epoch": 4.354487939897193,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00032418898777284295,
      "loss": 4.9464,
      "mean_token_accuracy": 0.22795826941728592,
      "num_tokens": 100869678.0,
      "step": 44050
    },
    {
      "entropy": 5.780865097045899,
      "epoch": 4.354982206405694,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003241545508297285,
      "loss": 4.9324,
      "mean_token_accuracy": 0.22249923646450043,
      "num_tokens": 100881305.0,
      "step": 44055
    },
    {
      "entropy": 5.784142684936524,
      "epoch": 4.355476472914195,
      "grad_norm": 1.4375,
      "learning_rate": 0.00032412011267740447,
      "loss": 4.9837,
      "mean_token_accuracy": 0.22497359961271285,
      "num_tokens": 100892085.0,
      "step": 44060
    },
    {
      "entropy": 5.875810861587524,
      "epoch": 4.3559707394226965,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003240856733167181,
      "loss": 5.0956,
      "mean_token_accuracy": 0.21268861293792723,
      "num_tokens": 100903558.0,
      "step": 44065
    },
    {
      "entropy": 5.800432252883911,
      "epoch": 4.356465005931198,
      "grad_norm": 1.421875,
      "learning_rate": 0.00032405123274851643,
      "loss": 4.9038,
      "mean_token_accuracy": 0.23196739107370376,
      "num_tokens": 100914219.0,
      "step": 44070
    },
    {
      "entropy": 5.830919981002808,
      "epoch": 4.356959272439699,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003240167909736469,
      "loss": 4.9902,
      "mean_token_accuracy": 0.22438732534646988,
      "num_tokens": 100926846.0,
      "step": 44075
    },
    {
      "entropy": 5.785229063034057,
      "epoch": 4.357453538948201,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003239823479929567,
      "loss": 4.9523,
      "mean_token_accuracy": 0.22845074832439421,
      "num_tokens": 100937502.0,
      "step": 44080
    },
    {
      "entropy": 5.76610312461853,
      "epoch": 4.3579478054567025,
      "grad_norm": 1.265625,
      "learning_rate": 0.00032394790380729316,
      "loss": 4.9504,
      "mean_token_accuracy": 0.2245262771844864,
      "num_tokens": 100949797.0,
      "step": 44085
    },
    {
      "entropy": 5.745938491821289,
      "epoch": 4.358442071965204,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00032391345841750356,
      "loss": 4.8859,
      "mean_token_accuracy": 0.23861430287361146,
      "num_tokens": 100960952.0,
      "step": 44090
    },
    {
      "entropy": 5.782887268066406,
      "epoch": 4.358936338473705,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003238790118244353,
      "loss": 4.9602,
      "mean_token_accuracy": 0.23222958892583848,
      "num_tokens": 100971691.0,
      "step": 44095
    },
    {
      "entropy": 5.744265985488892,
      "epoch": 4.359430604982206,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0003238445640289358,
      "loss": 4.8584,
      "mean_token_accuracy": 0.23746822476387025,
      "num_tokens": 100982872.0,
      "step": 44100
    },
    {
      "entropy": 5.743630266189575,
      "epoch": 4.3599248714907075,
      "grad_norm": 1.265625,
      "learning_rate": 0.00032381011503185243,
      "loss": 4.8877,
      "mean_token_accuracy": 0.23484389632940292,
      "num_tokens": 100994435.0,
      "step": 44105
    },
    {
      "entropy": 5.817675256729126,
      "epoch": 4.360419137999209,
      "grad_norm": 1.296875,
      "learning_rate": 0.00032377566483403266,
      "loss": 4.9654,
      "mean_token_accuracy": 0.22597405314445496,
      "num_tokens": 101006647.0,
      "step": 44110
    },
    {
      "entropy": 5.733247232437134,
      "epoch": 4.360913404507711,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00032374121343632396,
      "loss": 4.8817,
      "mean_token_accuracy": 0.23568636775016785,
      "num_tokens": 101018156.0,
      "step": 44115
    },
    {
      "entropy": 5.7707600593566895,
      "epoch": 4.361407671016212,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003237067608395738,
      "loss": 4.9617,
      "mean_token_accuracy": 0.22437798380851745,
      "num_tokens": 101028691.0,
      "step": 44120
    },
    {
      "entropy": 5.801129150390625,
      "epoch": 4.3619019375247134,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00032367230704462985,
      "loss": 4.9567,
      "mean_token_accuracy": 0.22752400636672973,
      "num_tokens": 101040658.0,
      "step": 44125
    },
    {
      "entropy": 5.848366546630859,
      "epoch": 4.362396204033215,
      "grad_norm": 1.25,
      "learning_rate": 0.00032363785205233954,
      "loss": 4.9654,
      "mean_token_accuracy": 0.2246871203184128,
      "num_tokens": 101052879.0,
      "step": 44130
    },
    {
      "entropy": 5.78539776802063,
      "epoch": 4.362890470541716,
      "grad_norm": 1.28125,
      "learning_rate": 0.00032360339586355054,
      "loss": 4.9289,
      "mean_token_accuracy": 0.2264582708477974,
      "num_tokens": 101064349.0,
      "step": 44135
    },
    {
      "entropy": 5.730290317535401,
      "epoch": 4.363384737050217,
      "grad_norm": 1.234375,
      "learning_rate": 0.00032356893847911044,
      "loss": 4.8542,
      "mean_token_accuracy": 0.23592747747898102,
      "num_tokens": 101078624.0,
      "step": 44140
    },
    {
      "entropy": 5.745179224014282,
      "epoch": 4.3638790035587185,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00032353447989986697,
      "loss": 4.8376,
      "mean_token_accuracy": 0.23632629662752153,
      "num_tokens": 101089822.0,
      "step": 44145
    },
    {
      "entropy": 5.7355554580688475,
      "epoch": 4.364373270067221,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003235000201266678,
      "loss": 4.8418,
      "mean_token_accuracy": 0.23961308598518372,
      "num_tokens": 101100499.0,
      "step": 44150
    },
    {
      "entropy": 5.619287061691284,
      "epoch": 4.364867536575722,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00032346555916036053,
      "loss": 4.7916,
      "mean_token_accuracy": 0.2498767614364624,
      "num_tokens": 101112331.0,
      "step": 44155
    },
    {
      "entropy": 5.852122974395752,
      "epoch": 4.365361803084223,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00032343109700179294,
      "loss": 5.0324,
      "mean_token_accuracy": 0.22152429670095444,
      "num_tokens": 101123491.0,
      "step": 44160
    },
    {
      "entropy": 5.775107288360596,
      "epoch": 4.365856069592724,
      "grad_norm": 1.375,
      "learning_rate": 0.0003233966336518129,
      "loss": 4.9086,
      "mean_token_accuracy": 0.23637375086545945,
      "num_tokens": 101134039.0,
      "step": 44165
    },
    {
      "entropy": 5.8704845905303955,
      "epoch": 4.366350336101226,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003233621691112683,
      "loss": 4.9936,
      "mean_token_accuracy": 0.22862096428871154,
      "num_tokens": 101145910.0,
      "step": 44170
    },
    {
      "entropy": 5.81949553489685,
      "epoch": 4.366844602609727,
      "grad_norm": 1.375,
      "learning_rate": 0.00032332770338100664,
      "loss": 4.9977,
      "mean_token_accuracy": 0.22273940294981004,
      "num_tokens": 101155924.0,
      "step": 44175
    },
    {
      "entropy": 5.731286525726318,
      "epoch": 4.367338869118228,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003232932364618761,
      "loss": 4.9118,
      "mean_token_accuracy": 0.22744144648313522,
      "num_tokens": 101166656.0,
      "step": 44180
    },
    {
      "entropy": 5.819930124282837,
      "epoch": 4.36783313562673,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00032325876835472445,
      "loss": 5.0248,
      "mean_token_accuracy": 0.22852956652641296,
      "num_tokens": 101177868.0,
      "step": 44185
    },
    {
      "entropy": 5.799646282196045,
      "epoch": 4.368327402135232,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003232242990603995,
      "loss": 4.9401,
      "mean_token_accuracy": 0.23055540919303893,
      "num_tokens": 101190114.0,
      "step": 44190
    },
    {
      "entropy": 5.787968873977661,
      "epoch": 4.368821668643733,
      "grad_norm": 1.25,
      "learning_rate": 0.0003231898285797494,
      "loss": 4.9692,
      "mean_token_accuracy": 0.22656389325857162,
      "num_tokens": 101202417.0,
      "step": 44195
    },
    {
      "entropy": 5.819163131713867,
      "epoch": 4.369315935152234,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000323155356913622,
      "loss": 4.96,
      "mean_token_accuracy": 0.22407441288232804,
      "num_tokens": 101214416.0,
      "step": 44200
    },
    {
      "entropy": 5.769860553741455,
      "epoch": 4.369810201660735,
      "grad_norm": 1.265625,
      "learning_rate": 0.00032312088406286543,
      "loss": 4.8766,
      "mean_token_accuracy": 0.22877698689699172,
      "num_tokens": 101225304.0,
      "step": 44205
    },
    {
      "entropy": 5.7694511890411375,
      "epoch": 4.370304468169237,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003230864100283276,
      "loss": 4.9847,
      "mean_token_accuracy": 0.2290133625268936,
      "num_tokens": 101236988.0,
      "step": 44210
    },
    {
      "entropy": 5.6366394519805905,
      "epoch": 4.370798734677738,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003230519348108566,
      "loss": 4.7387,
      "mean_token_accuracy": 0.25698686838150026,
      "num_tokens": 101248105.0,
      "step": 44215
    },
    {
      "entropy": 5.789619493484497,
      "epoch": 4.37129300118624,
      "grad_norm": 1.25,
      "learning_rate": 0.0003230174584113005,
      "loss": 4.996,
      "mean_token_accuracy": 0.21850239783525466,
      "num_tokens": 101260180.0,
      "step": 44220
    },
    {
      "entropy": 5.775313091278076,
      "epoch": 4.371787267694741,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00032298298083050756,
      "loss": 4.9368,
      "mean_token_accuracy": 0.22804800420999527,
      "num_tokens": 101271799.0,
      "step": 44225
    },
    {
      "entropy": 5.696804618835449,
      "epoch": 4.372281534203243,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003229485020693259,
      "loss": 4.8212,
      "mean_token_accuracy": 0.24237262457609177,
      "num_tokens": 101283111.0,
      "step": 44230
    },
    {
      "entropy": 5.7484715461730955,
      "epoch": 4.372775800711744,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00032291402212860347,
      "loss": 4.9471,
      "mean_token_accuracy": 0.22551867067813874,
      "num_tokens": 101295419.0,
      "step": 44235
    },
    {
      "entropy": 5.84717173576355,
      "epoch": 4.373270067220245,
      "grad_norm": 1.234375,
      "learning_rate": 0.00032287954100918877,
      "loss": 5.039,
      "mean_token_accuracy": 0.21360981911420823,
      "num_tokens": 101307431.0,
      "step": 44240
    },
    {
      "entropy": 5.817013263702393,
      "epoch": 4.373764333728746,
      "grad_norm": 1.359375,
      "learning_rate": 0.00032284505871193,
      "loss": 4.9054,
      "mean_token_accuracy": 0.22775947004556657,
      "num_tokens": 101318080.0,
      "step": 44245
    },
    {
      "entropy": 5.785833406448364,
      "epoch": 4.374258600237248,
      "grad_norm": 1.234375,
      "learning_rate": 0.00032281057523767524,
      "loss": 4.9248,
      "mean_token_accuracy": 0.2356586217880249,
      "num_tokens": 101330628.0,
      "step": 44250
    },
    {
      "entropy": 5.763400840759277,
      "epoch": 4.374752866745749,
      "grad_norm": 1.296875,
      "learning_rate": 0.00032277609058727297,
      "loss": 4.9275,
      "mean_token_accuracy": 0.23164038956165314,
      "num_tokens": 101340823.0,
      "step": 44255
    },
    {
      "entropy": 5.754184007644653,
      "epoch": 4.375247133254251,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003227416047615714,
      "loss": 4.8925,
      "mean_token_accuracy": 0.22541674524545668,
      "num_tokens": 101352329.0,
      "step": 44260
    },
    {
      "entropy": 5.821122455596924,
      "epoch": 4.375741399762752,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00032270711776141904,
      "loss": 5.0091,
      "mean_token_accuracy": 0.21585677862167357,
      "num_tokens": 101364223.0,
      "step": 44265
    },
    {
      "entropy": 5.790074586868286,
      "epoch": 4.376235666271254,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003226726295876642,
      "loss": 4.9387,
      "mean_token_accuracy": 0.2250300258398056,
      "num_tokens": 101375935.0,
      "step": 44270
    },
    {
      "entropy": 5.7145483016967775,
      "epoch": 4.376729932779755,
      "grad_norm": 1.203125,
      "learning_rate": 0.00032263814024115515,
      "loss": 4.8539,
      "mean_token_accuracy": 0.23135704398155213,
      "num_tokens": 101386751.0,
      "step": 44275
    },
    {
      "entropy": 5.797204542160034,
      "epoch": 4.377224199288256,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003226036497227406,
      "loss": 4.9771,
      "mean_token_accuracy": 0.2224976345896721,
      "num_tokens": 101397863.0,
      "step": 44280
    },
    {
      "entropy": 5.741680669784546,
      "epoch": 4.377718465796757,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00032256915803326874,
      "loss": 4.9209,
      "mean_token_accuracy": 0.23197393119335175,
      "num_tokens": 101408912.0,
      "step": 44285
    },
    {
      "entropy": 5.735201120376587,
      "epoch": 4.378212732305259,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003225346651735883,
      "loss": 4.8474,
      "mean_token_accuracy": 0.24280088543891906,
      "num_tokens": 101419784.0,
      "step": 44290
    },
    {
      "entropy": 5.76398458480835,
      "epoch": 4.37870699881376,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00032250017114454773,
      "loss": 4.8758,
      "mean_token_accuracy": 0.23235227912664413,
      "num_tokens": 101431148.0,
      "step": 44295
    },
    {
      "entropy": 5.7205362796783445,
      "epoch": 4.379201265322262,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003224656759469956,
      "loss": 4.9408,
      "mean_token_accuracy": 0.2321418270468712,
      "num_tokens": 101443034.0,
      "step": 44300
    },
    {
      "entropy": 5.775005912780761,
      "epoch": 4.379695531830763,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00032243117958178036,
      "loss": 4.9422,
      "mean_token_accuracy": 0.2310222268104553,
      "num_tokens": 101455242.0,
      "step": 44305
    },
    {
      "entropy": 5.787845468521118,
      "epoch": 4.380189798339265,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003223966820497508,
      "loss": 5.0084,
      "mean_token_accuracy": 0.2278506875038147,
      "num_tokens": 101466838.0,
      "step": 44310
    },
    {
      "entropy": 5.860180997848511,
      "epoch": 4.380684064847766,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003223621833517555,
      "loss": 4.9993,
      "mean_token_accuracy": 0.23098972737789153,
      "num_tokens": 101479333.0,
      "step": 44315
    },
    {
      "entropy": 5.7658881664276125,
      "epoch": 4.381178331356267,
      "grad_norm": 1.234375,
      "learning_rate": 0.00032232768348864314,
      "loss": 4.9276,
      "mean_token_accuracy": 0.2284233182668686,
      "num_tokens": 101490891.0,
      "step": 44320
    },
    {
      "entropy": 5.858300256729126,
      "epoch": 4.381672597864768,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003222931824612626,
      "loss": 5.0803,
      "mean_token_accuracy": 0.21535296142101287,
      "num_tokens": 101502146.0,
      "step": 44325
    },
    {
      "entropy": 5.7702418804168705,
      "epoch": 4.38216686437327,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003222586802704622,
      "loss": 4.936,
      "mean_token_accuracy": 0.22951242476701736,
      "num_tokens": 101514619.0,
      "step": 44330
    },
    {
      "entropy": 5.836402511596679,
      "epoch": 4.382661130881772,
      "grad_norm": 1.4375,
      "learning_rate": 0.00032222417691709097,
      "loss": 4.9897,
      "mean_token_accuracy": 0.22398712784051894,
      "num_tokens": 101526103.0,
      "step": 44335
    },
    {
      "entropy": 5.749256610870361,
      "epoch": 4.383155397390273,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00032218967240199776,
      "loss": 4.872,
      "mean_token_accuracy": 0.2353816568851471,
      "num_tokens": 101538503.0,
      "step": 44340
    },
    {
      "entropy": 5.750564002990723,
      "epoch": 4.383649663898774,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003221551667260313,
      "loss": 5.0058,
      "mean_token_accuracy": 0.22326190173625945,
      "num_tokens": 101549854.0,
      "step": 44345
    },
    {
      "entropy": 5.813726854324341,
      "epoch": 4.384143930407276,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00032212065989004036,
      "loss": 5.0005,
      "mean_token_accuracy": 0.2255925416946411,
      "num_tokens": 101561272.0,
      "step": 44350
    },
    {
      "entropy": 5.800186681747436,
      "epoch": 4.384638196915777,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003220861518948738,
      "loss": 4.8628,
      "mean_token_accuracy": 0.23794747442007064,
      "num_tokens": 101570941.0,
      "step": 44355
    },
    {
      "entropy": 5.673511171340943,
      "epoch": 4.385132463424278,
      "grad_norm": 1.25,
      "learning_rate": 0.00032205164274138075,
      "loss": 4.8616,
      "mean_token_accuracy": 0.23980370312929153,
      "num_tokens": 101582530.0,
      "step": 44360
    },
    {
      "entropy": 5.7948277473449705,
      "epoch": 4.385626729932779,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003220171324304099,
      "loss": 5.0374,
      "mean_token_accuracy": 0.2218106359243393,
      "num_tokens": 101594402.0,
      "step": 44365
    },
    {
      "entropy": 5.760267448425293,
      "epoch": 4.3861209964412815,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003219826209628103,
      "loss": 4.8925,
      "mean_token_accuracy": 0.2339683637022972,
      "num_tokens": 101606399.0,
      "step": 44370
    },
    {
      "entropy": 5.774322986602783,
      "epoch": 4.386615262949783,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003219481083394309,
      "loss": 4.9276,
      "mean_token_accuracy": 0.22042375504970552,
      "num_tokens": 101618403.0,
      "step": 44375
    },
    {
      "entropy": 5.8222596645355225,
      "epoch": 4.387109529458284,
      "grad_norm": 1.46875,
      "learning_rate": 0.00032191359456112083,
      "loss": 4.9801,
      "mean_token_accuracy": 0.22815112471580506,
      "num_tokens": 101630553.0,
      "step": 44380
    },
    {
      "entropy": 5.803790521621704,
      "epoch": 4.387603795966785,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003218790796287291,
      "loss": 4.939,
      "mean_token_accuracy": 0.22676888406276702,
      "num_tokens": 101642188.0,
      "step": 44385
    },
    {
      "entropy": 5.7963869094848635,
      "epoch": 4.3880980624752866,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00032184456354310465,
      "loss": 4.9138,
      "mean_token_accuracy": 0.23316336721181868,
      "num_tokens": 101654202.0,
      "step": 44390
    },
    {
      "entropy": 5.803341484069824,
      "epoch": 4.388592328983788,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003218100463050967,
      "loss": 4.9247,
      "mean_token_accuracy": 0.226852910220623,
      "num_tokens": 101665017.0,
      "step": 44395
    },
    {
      "entropy": 5.784501838684082,
      "epoch": 4.389086595492289,
      "grad_norm": 1.25,
      "learning_rate": 0.0003217755279155543,
      "loss": 4.9223,
      "mean_token_accuracy": 0.22675565034151077,
      "num_tokens": 101675662.0,
      "step": 44400
    },
    {
      "entropy": 5.766637325286865,
      "epoch": 4.389580862000791,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003217410083753267,
      "loss": 4.9191,
      "mean_token_accuracy": 0.22631852626800536,
      "num_tokens": 101687707.0,
      "step": 44405
    },
    {
      "entropy": 5.702825450897217,
      "epoch": 4.3900751285092925,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000321706487685263,
      "loss": 4.8991,
      "mean_token_accuracy": 0.22825476974248887,
      "num_tokens": 101701557.0,
      "step": 44410
    },
    {
      "entropy": 5.7099542140960695,
      "epoch": 4.390569395017794,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00032167196584621257,
      "loss": 4.8175,
      "mean_token_accuracy": 0.2400932714343071,
      "num_tokens": 101713305.0,
      "step": 44415
    },
    {
      "entropy": 5.7351367473602295,
      "epoch": 4.391063661526295,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00032163744285902446,
      "loss": 4.8962,
      "mean_token_accuracy": 0.2351185992360115,
      "num_tokens": 101725390.0,
      "step": 44420
    },
    {
      "entropy": 5.800309324264527,
      "epoch": 4.391557928034796,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000321602918724548,
      "loss": 4.9999,
      "mean_token_accuracy": 0.22799553871154785,
      "num_tokens": 101737191.0,
      "step": 44425
    },
    {
      "entropy": 5.800535535812378,
      "epoch": 4.3920521945432975,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003215683934436325,
      "loss": 4.9401,
      "mean_token_accuracy": 0.22713883817195893,
      "num_tokens": 101747326.0,
      "step": 44430
    },
    {
      "entropy": 5.7912821769714355,
      "epoch": 4.392546461051799,
      "grad_norm": 1.46875,
      "learning_rate": 0.00032153386701712735,
      "loss": 4.955,
      "mean_token_accuracy": 0.23055837005376817,
      "num_tokens": 101758214.0,
      "step": 44435
    },
    {
      "entropy": 5.724551105499268,
      "epoch": 4.393040727560301,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003214993394458818,
      "loss": 4.892,
      "mean_token_accuracy": 0.23377511352300645,
      "num_tokens": 101769045.0,
      "step": 44440
    },
    {
      "entropy": 5.79332218170166,
      "epoch": 4.393534994068802,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003214648107307454,
      "loss": 4.954,
      "mean_token_accuracy": 0.22634820491075516,
      "num_tokens": 101780358.0,
      "step": 44445
    },
    {
      "entropy": 5.764217138290405,
      "epoch": 4.3940292605773035,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003214302808725673,
      "loss": 4.9749,
      "mean_token_accuracy": 0.225906802713871,
      "num_tokens": 101792370.0,
      "step": 44450
    },
    {
      "entropy": 5.781581735610962,
      "epoch": 4.394523527085805,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003213957498721971,
      "loss": 4.904,
      "mean_token_accuracy": 0.23213106244802476,
      "num_tokens": 101803105.0,
      "step": 44455
    },
    {
      "entropy": 5.728999328613281,
      "epoch": 4.395017793594306,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00032136121773048445,
      "loss": 4.8944,
      "mean_token_accuracy": 0.24072905927896499,
      "num_tokens": 101814237.0,
      "step": 44460
    },
    {
      "entropy": 5.817507219314575,
      "epoch": 4.395512060102807,
      "grad_norm": 1.203125,
      "learning_rate": 0.00032132668444827847,
      "loss": 4.959,
      "mean_token_accuracy": 0.22507347613573075,
      "num_tokens": 101825190.0,
      "step": 44465
    },
    {
      "entropy": 5.783156299591065,
      "epoch": 4.3960063266113085,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00032129215002642893,
      "loss": 4.9132,
      "mean_token_accuracy": 0.22980836033821106,
      "num_tokens": 101835512.0,
      "step": 44470
    },
    {
      "entropy": 5.812867593765259,
      "epoch": 4.39650059311981,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003212576144657852,
      "loss": 5.0065,
      "mean_token_accuracy": 0.22016580849885942,
      "num_tokens": 101847329.0,
      "step": 44475
    },
    {
      "entropy": 5.861368846893311,
      "epoch": 4.396994859628312,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003212230777671971,
      "loss": 5.0394,
      "mean_token_accuracy": 0.22342875003814697,
      "num_tokens": 101858371.0,
      "step": 44480
    },
    {
      "entropy": 5.831535053253174,
      "epoch": 4.397489126136813,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003211885399315141,
      "loss": 4.9874,
      "mean_token_accuracy": 0.2299030140042305,
      "num_tokens": 101869011.0,
      "step": 44485
    },
    {
      "entropy": 5.774883556365967,
      "epoch": 4.3979833926453145,
      "grad_norm": 1.3125,
      "learning_rate": 0.00032115400095958585,
      "loss": 4.9289,
      "mean_token_accuracy": 0.2337183192372322,
      "num_tokens": 101880147.0,
      "step": 44490
    },
    {
      "entropy": 5.774054479598999,
      "epoch": 4.398477659153816,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000321119460852262,
      "loss": 4.8976,
      "mean_token_accuracy": 0.2280077025294304,
      "num_tokens": 101891519.0,
      "step": 44495
    },
    {
      "entropy": 5.712663888931274,
      "epoch": 4.398971925662317,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003210849196103922,
      "loss": 4.8371,
      "mean_token_accuracy": 0.23728777170181276,
      "num_tokens": 101902575.0,
      "step": 44500
    },
    {
      "entropy": 5.758746194839477,
      "epoch": 4.399466192170818,
      "grad_norm": 1.28125,
      "learning_rate": 0.00032105037723482635,
      "loss": 4.9415,
      "mean_token_accuracy": 0.22427077293395997,
      "num_tokens": 101913717.0,
      "step": 44505
    },
    {
      "entropy": 5.783820962905883,
      "epoch": 4.3999604586793195,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00032101583372641404,
      "loss": 4.8967,
      "mean_token_accuracy": 0.23310937732458115,
      "num_tokens": 101925196.0,
      "step": 44510
    },
    {
      "entropy": 5.814217901229858,
      "epoch": 4.400454725187822,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00032098128908600505,
      "loss": 4.897,
      "mean_token_accuracy": 0.2318163186311722,
      "num_tokens": 101936430.0,
      "step": 44515
    },
    {
      "entropy": 5.785718393325806,
      "epoch": 4.400948991696323,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003209467433144492,
      "loss": 4.9347,
      "mean_token_accuracy": 0.22434916347265244,
      "num_tokens": 101947368.0,
      "step": 44520
    },
    {
      "entropy": 5.7272134780883786,
      "epoch": 4.401443258204824,
      "grad_norm": 1.359375,
      "learning_rate": 0.00032091219641259633,
      "loss": 4.8713,
      "mean_token_accuracy": 0.23744321316480638,
      "num_tokens": 101957474.0,
      "step": 44525
    },
    {
      "entropy": 5.846186113357544,
      "epoch": 4.4019375247133254,
      "grad_norm": 1.1875,
      "learning_rate": 0.00032087764838129636,
      "loss": 5.0503,
      "mean_token_accuracy": 0.22099510878324508,
      "num_tokens": 101970141.0,
      "step": 44530
    },
    {
      "entropy": 5.825059986114502,
      "epoch": 4.402431791221827,
      "grad_norm": 1.34375,
      "learning_rate": 0.00032084309922139907,
      "loss": 5.0037,
      "mean_token_accuracy": 0.21945223063230515,
      "num_tokens": 101980851.0,
      "step": 44535
    },
    {
      "entropy": 5.779472732543946,
      "epoch": 4.402926057730328,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00032080854893375447,
      "loss": 4.9409,
      "mean_token_accuracy": 0.22958521246910096,
      "num_tokens": 101992716.0,
      "step": 44540
    },
    {
      "entropy": 5.752793169021606,
      "epoch": 4.403420324238829,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003207739975192124,
      "loss": 4.9206,
      "mean_token_accuracy": 0.22481743693351747,
      "num_tokens": 102003419.0,
      "step": 44545
    },
    {
      "entropy": 5.796084451675415,
      "epoch": 4.4039145907473305,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003207394449786231,
      "loss": 4.938,
      "mean_token_accuracy": 0.23021966218948364,
      "num_tokens": 102014188.0,
      "step": 44550
    },
    {
      "entropy": 5.804183769226074,
      "epoch": 4.404408857255833,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003207048913128361,
      "loss": 4.9279,
      "mean_token_accuracy": 0.22860852330923082,
      "num_tokens": 102024925.0,
      "step": 44555
    },
    {
      "entropy": 5.709659481048584,
      "epoch": 4.404903123764334,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003206703365227018,
      "loss": 4.8879,
      "mean_token_accuracy": 0.2354836270213127,
      "num_tokens": 102036846.0,
      "step": 44560
    },
    {
      "entropy": 5.6926289081573485,
      "epoch": 4.405397390272835,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003206357806090701,
      "loss": 4.9004,
      "mean_token_accuracy": 0.23062173128128052,
      "num_tokens": 102050187.0,
      "step": 44565
    },
    {
      "entropy": 5.774506998062134,
      "epoch": 4.405891656781336,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003206012235727911,
      "loss": 4.97,
      "mean_token_accuracy": 0.22743042409420014,
      "num_tokens": 102062071.0,
      "step": 44570
    },
    {
      "entropy": 5.805512809753418,
      "epoch": 4.406385923289838,
      "grad_norm": 1.34375,
      "learning_rate": 0.00032056666541471497,
      "loss": 4.9176,
      "mean_token_accuracy": 0.23443736732006074,
      "num_tokens": 102073641.0,
      "step": 44575
    },
    {
      "entropy": 5.815981101989746,
      "epoch": 4.406880189798339,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003205321061356918,
      "loss": 4.9042,
      "mean_token_accuracy": 0.2320135623216629,
      "num_tokens": 102085328.0,
      "step": 44580
    },
    {
      "entropy": 5.764554023742676,
      "epoch": 4.40737445630684,
      "grad_norm": 1.421875,
      "learning_rate": 0.00032049754573657176,
      "loss": 4.9506,
      "mean_token_accuracy": 0.2369006544351578,
      "num_tokens": 102096866.0,
      "step": 44585
    },
    {
      "entropy": 5.724488067626953,
      "epoch": 4.407868722815342,
      "grad_norm": 1.421875,
      "learning_rate": 0.000320462984218205,
      "loss": 4.9233,
      "mean_token_accuracy": 0.22877844125032426,
      "num_tokens": 102107823.0,
      "step": 44590
    },
    {
      "entropy": 5.759817838668823,
      "epoch": 4.408362989323844,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00032042842158144173,
      "loss": 4.863,
      "mean_token_accuracy": 0.236982823908329,
      "num_tokens": 102118443.0,
      "step": 44595
    },
    {
      "entropy": 5.796397686004639,
      "epoch": 4.408857255832345,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003203938578271324,
      "loss": 4.9228,
      "mean_token_accuracy": 0.23113987296819688,
      "num_tokens": 102131745.0,
      "step": 44600
    },
    {
      "entropy": 5.762889051437378,
      "epoch": 4.409351522340846,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00032035929295612695,
      "loss": 4.9622,
      "mean_token_accuracy": 0.22129473984241485,
      "num_tokens": 102143275.0,
      "step": 44605
    },
    {
      "entropy": 5.762961959838867,
      "epoch": 4.409845788849347,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000320324726969276,
      "loss": 4.9041,
      "mean_token_accuracy": 0.2342536687850952,
      "num_tokens": 102156125.0,
      "step": 44610
    },
    {
      "entropy": 5.762206745147705,
      "epoch": 4.410340055357849,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00032029015986742966,
      "loss": 4.8654,
      "mean_token_accuracy": 0.23709594011306762,
      "num_tokens": 102166599.0,
      "step": 44615
    },
    {
      "entropy": 5.649644422531128,
      "epoch": 4.41083432186635,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003202555916514384,
      "loss": 4.8285,
      "mean_token_accuracy": 0.24055459797382356,
      "num_tokens": 102179320.0,
      "step": 44620
    },
    {
      "entropy": 5.785014581680298,
      "epoch": 4.411328588374852,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003202210223221525,
      "loss": 5.0023,
      "mean_token_accuracy": 0.2238845556974411,
      "num_tokens": 102191786.0,
      "step": 44625
    },
    {
      "entropy": 5.799521446228027,
      "epoch": 4.411822854883353,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003201864518804225,
      "loss": 4.8851,
      "mean_token_accuracy": 0.23715116530656816,
      "num_tokens": 102202955.0,
      "step": 44630
    },
    {
      "entropy": 5.789065170288086,
      "epoch": 4.412317121391855,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003201518803270987,
      "loss": 4.8469,
      "mean_token_accuracy": 0.24145787954330444,
      "num_tokens": 102214930.0,
      "step": 44635
    },
    {
      "entropy": 5.793876457214355,
      "epoch": 4.412811387900356,
      "grad_norm": 1.46875,
      "learning_rate": 0.00032011730766303163,
      "loss": 4.9112,
      "mean_token_accuracy": 0.22901122719049455,
      "num_tokens": 102226009.0,
      "step": 44640
    },
    {
      "entropy": 5.7417479991912845,
      "epoch": 4.413305654408857,
      "grad_norm": 1.25,
      "learning_rate": 0.0003200827338890719,
      "loss": 4.8675,
      "mean_token_accuracy": 0.23515848368406295,
      "num_tokens": 102237486.0,
      "step": 44645
    },
    {
      "entropy": 5.718871831893921,
      "epoch": 4.413799920917358,
      "grad_norm": 1.390625,
      "learning_rate": 0.00032004815900606994,
      "loss": 4.9275,
      "mean_token_accuracy": 0.22771875113248824,
      "num_tokens": 102249094.0,
      "step": 44650
    },
    {
      "entropy": 5.753797340393066,
      "epoch": 4.41429418742586,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003200135830148762,
      "loss": 4.9953,
      "mean_token_accuracy": 0.22277871668338775,
      "num_tokens": 102260388.0,
      "step": 44655
    },
    {
      "entropy": 5.794828462600708,
      "epoch": 4.414788453934362,
      "grad_norm": 1.25,
      "learning_rate": 0.0003199790059163414,
      "loss": 4.9439,
      "mean_token_accuracy": 0.22106139063835145,
      "num_tokens": 102272745.0,
      "step": 44660
    },
    {
      "entropy": 5.754795837402344,
      "epoch": 4.415282720442863,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003199444277113159,
      "loss": 4.8701,
      "mean_token_accuracy": 0.23313587605953218,
      "num_tokens": 102283566.0,
      "step": 44665
    },
    {
      "entropy": 5.810240173339844,
      "epoch": 4.415776986951364,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003199098484006507,
      "loss": 4.9849,
      "mean_token_accuracy": 0.22759221643209457,
      "num_tokens": 102296125.0,
      "step": 44670
    },
    {
      "entropy": 5.8383808612823485,
      "epoch": 4.416271253459866,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00031987526798519616,
      "loss": 4.9189,
      "mean_token_accuracy": 0.2265412464737892,
      "num_tokens": 102307925.0,
      "step": 44675
    },
    {
      "entropy": 5.850876426696777,
      "epoch": 4.416765519968367,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00031984068646580315,
      "loss": 4.9779,
      "mean_token_accuracy": 0.22305086106061936,
      "num_tokens": 102318913.0,
      "step": 44680
    },
    {
      "entropy": 5.705397129058838,
      "epoch": 4.417259786476868,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003198061038433222,
      "loss": 4.887,
      "mean_token_accuracy": 0.23596783429384233,
      "num_tokens": 102330659.0,
      "step": 44685
    },
    {
      "entropy": 5.785893678665161,
      "epoch": 4.417754052985369,
      "grad_norm": 1.21875,
      "learning_rate": 0.00031977152011860416,
      "loss": 4.9318,
      "mean_token_accuracy": 0.22578260600566863,
      "num_tokens": 102342389.0,
      "step": 44690
    },
    {
      "entropy": 5.871178913116455,
      "epoch": 4.4182483194938715,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00031973693529249985,
      "loss": 4.9935,
      "mean_token_accuracy": 0.22863285094499589,
      "num_tokens": 102354810.0,
      "step": 44695
    },
    {
      "entropy": 5.7864807605743405,
      "epoch": 4.418742586002373,
      "grad_norm": 1.390625,
      "learning_rate": 0.00031970234936585997,
      "loss": 4.936,
      "mean_token_accuracy": 0.22711482793092727,
      "num_tokens": 102366078.0,
      "step": 44700
    },
    {
      "entropy": 5.758639717102051,
      "epoch": 4.419236852510874,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00031966776233953534,
      "loss": 4.95,
      "mean_token_accuracy": 0.2346431568264961,
      "num_tokens": 102378480.0,
      "step": 44705
    },
    {
      "entropy": 5.747470331192017,
      "epoch": 4.419731119019375,
      "grad_norm": 1.28125,
      "learning_rate": 0.00031963317421437675,
      "loss": 4.8619,
      "mean_token_accuracy": 0.2429773062467575,
      "num_tokens": 102389493.0,
      "step": 44710
    },
    {
      "entropy": 5.887057447433472,
      "epoch": 4.420225385527877,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003195985849912353,
      "loss": 5.035,
      "mean_token_accuracy": 0.21522504985332488,
      "num_tokens": 102401543.0,
      "step": 44715
    },
    {
      "entropy": 5.7858387470245365,
      "epoch": 4.420719652036378,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003195639946709616,
      "loss": 5.0028,
      "mean_token_accuracy": 0.22263889759778976,
      "num_tokens": 102412758.0,
      "step": 44720
    },
    {
      "entropy": 5.794162178039551,
      "epoch": 4.421213918544879,
      "grad_norm": 1.34375,
      "learning_rate": 0.00031952940325440685,
      "loss": 4.9044,
      "mean_token_accuracy": 0.229052072763443,
      "num_tokens": 102422297.0,
      "step": 44725
    },
    {
      "entropy": 5.730296945571899,
      "epoch": 4.42170818505338,
      "grad_norm": 1.25,
      "learning_rate": 0.00031949481074242174,
      "loss": 4.8925,
      "mean_token_accuracy": 0.2313983842730522,
      "num_tokens": 102433838.0,
      "step": 44730
    },
    {
      "entropy": 5.816375350952148,
      "epoch": 4.4222024515618825,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00031946021713585745,
      "loss": 4.9512,
      "mean_token_accuracy": 0.22502039819955827,
      "num_tokens": 102445138.0,
      "step": 44735
    },
    {
      "entropy": 5.749163389205933,
      "epoch": 4.422696718070384,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003194256224355649,
      "loss": 4.8828,
      "mean_token_accuracy": 0.23647888898849487,
      "num_tokens": 102457346.0,
      "step": 44740
    },
    {
      "entropy": 5.732340860366821,
      "epoch": 4.423190984578885,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003193910266423952,
      "loss": 4.9289,
      "mean_token_accuracy": 0.23004593849182128,
      "num_tokens": 102469463.0,
      "step": 44745
    },
    {
      "entropy": 5.727120399475098,
      "epoch": 4.423685251087386,
      "grad_norm": 1.3125,
      "learning_rate": 0.00031935642975719936,
      "loss": 4.9015,
      "mean_token_accuracy": 0.23419105410575866,
      "num_tokens": 102480623.0,
      "step": 44750
    },
    {
      "entropy": 5.756617021560669,
      "epoch": 4.424179517595888,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003193218317808284,
      "loss": 4.9276,
      "mean_token_accuracy": 0.23384655714035035,
      "num_tokens": 102492768.0,
      "step": 44755
    },
    {
      "entropy": 5.801549863815308,
      "epoch": 4.424673784104389,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003192872327141335,
      "loss": 4.8924,
      "mean_token_accuracy": 0.23001836240291595,
      "num_tokens": 102503592.0,
      "step": 44760
    },
    {
      "entropy": 5.766797828674316,
      "epoch": 4.42516805061289,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00031925263255796587,
      "loss": 4.9237,
      "mean_token_accuracy": 0.2280314087867737,
      "num_tokens": 102515056.0,
      "step": 44765
    },
    {
      "entropy": 5.78840708732605,
      "epoch": 4.425662317121392,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00031921803131317657,
      "loss": 4.9155,
      "mean_token_accuracy": 0.22566602677106856,
      "num_tokens": 102525749.0,
      "step": 44770
    },
    {
      "entropy": 5.7522176742553714,
      "epoch": 4.4261565836298935,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003191834289806169,
      "loss": 4.9238,
      "mean_token_accuracy": 0.23341690003871918,
      "num_tokens": 102537760.0,
      "step": 44775
    },
    {
      "entropy": 5.816337060928345,
      "epoch": 4.426650850138395,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003191488255611379,
      "loss": 5.0436,
      "mean_token_accuracy": 0.2216949224472046,
      "num_tokens": 102548997.0,
      "step": 44780
    },
    {
      "entropy": 5.750407409667969,
      "epoch": 4.427145116646896,
      "grad_norm": 1.171875,
      "learning_rate": 0.000319114221055591,
      "loss": 4.9493,
      "mean_token_accuracy": 0.2295815259218216,
      "num_tokens": 102562056.0,
      "step": 44785
    },
    {
      "entropy": 5.703689670562744,
      "epoch": 4.427639383155397,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003190796154648275,
      "loss": 4.7663,
      "mean_token_accuracy": 0.24367156326770784,
      "num_tokens": 102573336.0,
      "step": 44790
    },
    {
      "entropy": 5.773833847045898,
      "epoch": 4.4281336496638986,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003190450087896986,
      "loss": 4.9161,
      "mean_token_accuracy": 0.2274806395173073,
      "num_tokens": 102583929.0,
      "step": 44795
    },
    {
      "entropy": 5.750168943405152,
      "epoch": 4.4286279161724,
      "grad_norm": 1.40625,
      "learning_rate": 0.00031901040103105553,
      "loss": 4.8336,
      "mean_token_accuracy": 0.2391642227768898,
      "num_tokens": 102594444.0,
      "step": 44800
    },
    {
      "entropy": 5.757972240447998,
      "epoch": 4.429122182680901,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003189757921897498,
      "loss": 4.9625,
      "mean_token_accuracy": 0.22552814185619355,
      "num_tokens": 102606357.0,
      "step": 44805
    },
    {
      "entropy": 5.780731248855591,
      "epoch": 4.429616449189403,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003189411822666329,
      "loss": 4.9497,
      "mean_token_accuracy": 0.23164233416318894,
      "num_tokens": 102617798.0,
      "step": 44810
    },
    {
      "entropy": 5.769657850265503,
      "epoch": 4.4301107156979045,
      "grad_norm": 1.265625,
      "learning_rate": 0.00031890657126255596,
      "loss": 4.8855,
      "mean_token_accuracy": 0.237362901866436,
      "num_tokens": 102630104.0,
      "step": 44815
    },
    {
      "entropy": 5.76247501373291,
      "epoch": 4.430604982206406,
      "grad_norm": 1.34375,
      "learning_rate": 0.00031887195917837066,
      "loss": 4.8496,
      "mean_token_accuracy": 0.23561314195394517,
      "num_tokens": 102641469.0,
      "step": 44820
    },
    {
      "entropy": 5.733041667938233,
      "epoch": 4.431099248714907,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00031883734601492823,
      "loss": 4.9591,
      "mean_token_accuracy": 0.2274942323565483,
      "num_tokens": 102653023.0,
      "step": 44825
    },
    {
      "entropy": 5.804011535644531,
      "epoch": 4.431593515223408,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003188027317730803,
      "loss": 4.9387,
      "mean_token_accuracy": 0.22652135491371156,
      "num_tokens": 102665074.0,
      "step": 44830
    },
    {
      "entropy": 5.7715263843536375,
      "epoch": 4.4320877817319095,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003187681164536785,
      "loss": 4.897,
      "mean_token_accuracy": 0.23389161229133607,
      "num_tokens": 102677277.0,
      "step": 44835
    },
    {
      "entropy": 5.786307430267334,
      "epoch": 4.432582048240411,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00031873350005757415,
      "loss": 4.9277,
      "mean_token_accuracy": 0.23112163543701172,
      "num_tokens": 102687375.0,
      "step": 44840
    },
    {
      "entropy": 5.778110694885254,
      "epoch": 4.433076314748913,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00031869888258561897,
      "loss": 4.899,
      "mean_token_accuracy": 0.23145822137594224,
      "num_tokens": 102698305.0,
      "step": 44845
    },
    {
      "entropy": 5.753381633758545,
      "epoch": 4.433570581257414,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00031866426403866444,
      "loss": 4.9309,
      "mean_token_accuracy": 0.23085810244083405,
      "num_tokens": 102711887.0,
      "step": 44850
    },
    {
      "entropy": 5.829583692550659,
      "epoch": 4.4340648477659155,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00031862964441756224,
      "loss": 4.8807,
      "mean_token_accuracy": 0.22851230055093766,
      "num_tokens": 102723187.0,
      "step": 44855
    },
    {
      "entropy": 5.697841024398803,
      "epoch": 4.434559114274417,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003185950237231641,
      "loss": 4.8611,
      "mean_token_accuracy": 0.24009753465652467,
      "num_tokens": 102733380.0,
      "step": 44860
    },
    {
      "entropy": 5.705118036270141,
      "epoch": 4.435053380782918,
      "grad_norm": 1.296875,
      "learning_rate": 0.00031856040195632154,
      "loss": 4.9342,
      "mean_token_accuracy": 0.22359001338481904,
      "num_tokens": 102746371.0,
      "step": 44865
    },
    {
      "entropy": 5.772717046737671,
      "epoch": 4.435547647291419,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00031852577911788637,
      "loss": 4.8881,
      "mean_token_accuracy": 0.24111796915531158,
      "num_tokens": 102757069.0,
      "step": 44870
    },
    {
      "entropy": 5.756084537506103,
      "epoch": 4.4360419137999205,
      "grad_norm": 1.296875,
      "learning_rate": 0.00031849115520871026,
      "loss": 4.9206,
      "mean_token_accuracy": 0.23426035046577454,
      "num_tokens": 102768323.0,
      "step": 44875
    },
    {
      "entropy": 5.829871797561646,
      "epoch": 4.436536180308423,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003184565302296449,
      "loss": 4.9544,
      "mean_token_accuracy": 0.22645423263311387,
      "num_tokens": 102779330.0,
      "step": 44880
    },
    {
      "entropy": 5.762402009963989,
      "epoch": 4.437030446816924,
      "grad_norm": 1.40625,
      "learning_rate": 0.00031842190418154235,
      "loss": 4.8828,
      "mean_token_accuracy": 0.23722873777151107,
      "num_tokens": 102790288.0,
      "step": 44885
    },
    {
      "entropy": 5.724287700653076,
      "epoch": 4.437524713325425,
      "grad_norm": 1.375,
      "learning_rate": 0.00031838727706525404,
      "loss": 4.8157,
      "mean_token_accuracy": 0.24305551946163179,
      "num_tokens": 102800799.0,
      "step": 44890
    },
    {
      "entropy": 5.772220468521118,
      "epoch": 4.4380189798339265,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003183526488816321,
      "loss": 4.9205,
      "mean_token_accuracy": 0.2321889728307724,
      "num_tokens": 102811811.0,
      "step": 44895
    },
    {
      "entropy": 5.786501741409301,
      "epoch": 4.438513246342428,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003183180196315282,
      "loss": 4.9479,
      "mean_token_accuracy": 0.22081732153892517,
      "num_tokens": 102822851.0,
      "step": 44900
    },
    {
      "entropy": 5.7800191879272464,
      "epoch": 4.439007512850929,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003182833893157943,
      "loss": 4.8946,
      "mean_token_accuracy": 0.225760518014431,
      "num_tokens": 102833333.0,
      "step": 44905
    },
    {
      "entropy": 5.782329273223877,
      "epoch": 4.43950177935943,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003182487579352824,
      "loss": 4.8911,
      "mean_token_accuracy": 0.2317438989877701,
      "num_tokens": 102844667.0,
      "step": 44910
    },
    {
      "entropy": 5.8237776279449465,
      "epoch": 4.439996045867932,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00031821412549084426,
      "loss": 4.9808,
      "mean_token_accuracy": 0.22645284980535507,
      "num_tokens": 102856701.0,
      "step": 44915
    },
    {
      "entropy": 5.8112269878387455,
      "epoch": 4.440490312376434,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003181794919833319,
      "loss": 4.9259,
      "mean_token_accuracy": 0.22987623810768126,
      "num_tokens": 102869518.0,
      "step": 44920
    },
    {
      "entropy": 5.759421634674072,
      "epoch": 4.440984578884935,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003181448574135974,
      "loss": 4.9414,
      "mean_token_accuracy": 0.2303444415330887,
      "num_tokens": 102879839.0,
      "step": 44925
    },
    {
      "entropy": 5.781966018676758,
      "epoch": 4.441478845393436,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003181102217824927,
      "loss": 4.8911,
      "mean_token_accuracy": 0.2298296019434929,
      "num_tokens": 102892034.0,
      "step": 44930
    },
    {
      "entropy": 5.803639316558838,
      "epoch": 4.441973111901937,
      "grad_norm": 1.359375,
      "learning_rate": 0.00031807558509086987,
      "loss": 4.9738,
      "mean_token_accuracy": 0.22538159787654877,
      "num_tokens": 102903683.0,
      "step": 44935
    },
    {
      "entropy": 5.827106857299805,
      "epoch": 4.442467378410439,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000318040947339581,
      "loss": 4.9507,
      "mean_token_accuracy": 0.2275276467204094,
      "num_tokens": 102915115.0,
      "step": 44940
    },
    {
      "entropy": 5.796533107757568,
      "epoch": 4.44296164491894,
      "grad_norm": 1.40625,
      "learning_rate": 0.000318006308529478,
      "loss": 4.9146,
      "mean_token_accuracy": 0.22475071251392365,
      "num_tokens": 102925765.0,
      "step": 44945
    },
    {
      "entropy": 5.7181822776794435,
      "epoch": 4.443455911427442,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003179716686614133,
      "loss": 4.9081,
      "mean_token_accuracy": 0.227689728140831,
      "num_tokens": 102937611.0,
      "step": 44950
    },
    {
      "entropy": 5.732023859024048,
      "epoch": 4.443950177935943,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00031793702773623883,
      "loss": 4.8888,
      "mean_token_accuracy": 0.2343665897846222,
      "num_tokens": 102947975.0,
      "step": 44955
    },
    {
      "entropy": 5.730938196182251,
      "epoch": 4.444444444444445,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003179023857548068,
      "loss": 4.9394,
      "mean_token_accuracy": 0.23144738078117372,
      "num_tokens": 102960040.0,
      "step": 44960
    },
    {
      "entropy": 5.852973175048828,
      "epoch": 4.444938710952946,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003178677427179694,
      "loss": 5.055,
      "mean_token_accuracy": 0.22154628336429597,
      "num_tokens": 102971723.0,
      "step": 44965
    },
    {
      "entropy": 5.781847095489502,
      "epoch": 4.445432977461447,
      "grad_norm": 1.234375,
      "learning_rate": 0.00031783309862657897,
      "loss": 4.9187,
      "mean_token_accuracy": 0.2291146770119667,
      "num_tokens": 102982776.0,
      "step": 44970
    },
    {
      "entropy": 5.849120140075684,
      "epoch": 4.445927243969948,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003177984534814875,
      "loss": 5.0446,
      "mean_token_accuracy": 0.2212783470749855,
      "num_tokens": 102994043.0,
      "step": 44975
    },
    {
      "entropy": 5.808528757095337,
      "epoch": 4.44642151047845,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003177638072835476,
      "loss": 4.9233,
      "mean_token_accuracy": 0.23067558109760283,
      "num_tokens": 103005863.0,
      "step": 44980
    },
    {
      "entropy": 5.80425329208374,
      "epoch": 4.446915776986951,
      "grad_norm": 1.28125,
      "learning_rate": 0.00031772916003361136,
      "loss": 4.9637,
      "mean_token_accuracy": 0.2263253942131996,
      "num_tokens": 103017584.0,
      "step": 44985
    },
    {
      "entropy": 5.7601556301116945,
      "epoch": 4.447410043495453,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00031769451173253117,
      "loss": 4.9273,
      "mean_token_accuracy": 0.228986257314682,
      "num_tokens": 103029505.0,
      "step": 44990
    },
    {
      "entropy": 5.824105501174927,
      "epoch": 4.447904310003954,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003176598623811593,
      "loss": 4.8684,
      "mean_token_accuracy": 0.23501897156238555,
      "num_tokens": 103039897.0,
      "step": 44995
    },
    {
      "entropy": 5.819701910018921,
      "epoch": 4.448398576512456,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003176252119803483,
      "loss": 4.9786,
      "mean_token_accuracy": 0.22547285556793212,
      "num_tokens": 103052674.0,
      "step": 45000
    },
    {
      "epoch": 4.448398576512456,
      "eval_entropy": 5.51085833189379,
      "eval_loss": 5.003839015960693,
      "eval_mean_token_accuracy": 0.23369953760838302,
      "eval_num_tokens": 103052674.0,
      "eval_runtime": 20.5812,
      "eval_samples_per_second": 1579.742,
      "eval_steps_per_second": 197.51,
      "step": 45000
    },
    {
      "entropy": 5.7594677925109865,
      "epoch": 4.448892843020957,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00031759056053095035,
      "loss": 4.9525,
      "mean_token_accuracy": 0.2280846953392029,
      "num_tokens": 103063772.0,
      "step": 45005
    },
    {
      "entropy": 5.784536504745484,
      "epoch": 4.449387109529458,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00031755590803381813,
      "loss": 4.9093,
      "mean_token_accuracy": 0.23169808983802795,
      "num_tokens": 103074784.0,
      "step": 45010
    },
    {
      "entropy": 5.770496559143067,
      "epoch": 4.449881376037959,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003175212544898038,
      "loss": 4.8371,
      "mean_token_accuracy": 0.23846487253904342,
      "num_tokens": 103085552.0,
      "step": 45015
    },
    {
      "entropy": 5.764013624191284,
      "epoch": 4.450375642546461,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003174865998997602,
      "loss": 4.9187,
      "mean_token_accuracy": 0.23576008081436156,
      "num_tokens": 103096499.0,
      "step": 45020
    },
    {
      "entropy": 5.693491744995117,
      "epoch": 4.450869909054963,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00031745194426453956,
      "loss": 4.9192,
      "mean_token_accuracy": 0.23411534130573272,
      "num_tokens": 103107669.0,
      "step": 45025
    },
    {
      "entropy": 5.816270923614502,
      "epoch": 4.451364175563464,
      "grad_norm": 1.375,
      "learning_rate": 0.00031741728758499457,
      "loss": 4.9011,
      "mean_token_accuracy": 0.23865312933921815,
      "num_tokens": 103118661.0,
      "step": 45030
    },
    {
      "entropy": 5.892057085037232,
      "epoch": 4.451858442071965,
      "grad_norm": 1.46875,
      "learning_rate": 0.0003173826298619776,
      "loss": 5.0419,
      "mean_token_accuracy": 0.22481542378664016,
      "num_tokens": 103130131.0,
      "step": 45035
    },
    {
      "entropy": 5.81878695487976,
      "epoch": 4.452352708580467,
      "grad_norm": 1.234375,
      "learning_rate": 0.00031734797109634147,
      "loss": 4.9833,
      "mean_token_accuracy": 0.22596470713615419,
      "num_tokens": 103141985.0,
      "step": 45040
    },
    {
      "entropy": 5.844881391525268,
      "epoch": 4.452846975088968,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00031731331128893864,
      "loss": 5.0043,
      "mean_token_accuracy": 0.22548459470272064,
      "num_tokens": 103152442.0,
      "step": 45045
    },
    {
      "entropy": 5.781526565551758,
      "epoch": 4.453341241597469,
      "grad_norm": 1.359375,
      "learning_rate": 0.00031727865044062177,
      "loss": 4.8951,
      "mean_token_accuracy": 0.23354577124118805,
      "num_tokens": 103163196.0,
      "step": 45050
    },
    {
      "entropy": 5.819967317581177,
      "epoch": 4.45383550810597,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003172439885522436,
      "loss": 4.9286,
      "mean_token_accuracy": 0.23250508606433867,
      "num_tokens": 103173933.0,
      "step": 45055
    },
    {
      "entropy": 5.783967447280884,
      "epoch": 4.454329774614472,
      "grad_norm": 1.265625,
      "learning_rate": 0.00031720932562465673,
      "loss": 4.9552,
      "mean_token_accuracy": 0.22930959910154342,
      "num_tokens": 103186445.0,
      "step": 45060
    },
    {
      "entropy": 5.771194314956665,
      "epoch": 4.454824041122974,
      "grad_norm": 1.390625,
      "learning_rate": 0.00031717466165871386,
      "loss": 4.9158,
      "mean_token_accuracy": 0.23100268244743347,
      "num_tokens": 103198381.0,
      "step": 45065
    },
    {
      "entropy": 5.730855512619018,
      "epoch": 4.455318307631475,
      "grad_norm": 1.265625,
      "learning_rate": 0.00031713999665526784,
      "loss": 4.8717,
      "mean_token_accuracy": 0.23673943877220155,
      "num_tokens": 103209668.0,
      "step": 45070
    },
    {
      "entropy": 5.804381036758423,
      "epoch": 4.455812574139976,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003171053306151713,
      "loss": 4.9639,
      "mean_token_accuracy": 0.21901979744434358,
      "num_tokens": 103222601.0,
      "step": 45075
    },
    {
      "entropy": 5.759120845794678,
      "epoch": 4.456306840648478,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003170706635392771,
      "loss": 4.8942,
      "mean_token_accuracy": 0.22490824908018112,
      "num_tokens": 103233224.0,
      "step": 45080
    },
    {
      "entropy": 5.741366910934448,
      "epoch": 4.456801107156979,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003170359954284381,
      "loss": 4.8667,
      "mean_token_accuracy": 0.2370065540075302,
      "num_tokens": 103244276.0,
      "step": 45085
    },
    {
      "entropy": 5.809133291244507,
      "epoch": 4.45729537366548,
      "grad_norm": 1.578125,
      "learning_rate": 0.00031700132628350696,
      "loss": 4.9522,
      "mean_token_accuracy": 0.23240016996860505,
      "num_tokens": 103254063.0,
      "step": 45090
    },
    {
      "entropy": 5.810419225692749,
      "epoch": 4.457789640173981,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003169666561053368,
      "loss": 4.9716,
      "mean_token_accuracy": 0.22476155310869217,
      "num_tokens": 103266634.0,
      "step": 45095
    },
    {
      "entropy": 5.79051513671875,
      "epoch": 4.4582839066824835,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003169319848947805,
      "loss": 4.8627,
      "mean_token_accuracy": 0.23638947755098344,
      "num_tokens": 103276858.0,
      "step": 45100
    },
    {
      "entropy": 5.796458435058594,
      "epoch": 4.458778173190985,
      "grad_norm": 1.1875,
      "learning_rate": 0.00031689731265269064,
      "loss": 4.9791,
      "mean_token_accuracy": 0.2236119583249092,
      "num_tokens": 103287582.0,
      "step": 45105
    },
    {
      "entropy": 5.78113956451416,
      "epoch": 4.459272439699486,
      "grad_norm": 1.5,
      "learning_rate": 0.00031686263937992054,
      "loss": 4.9012,
      "mean_token_accuracy": 0.23414964824914933,
      "num_tokens": 103298628.0,
      "step": 45110
    },
    {
      "entropy": 5.813367080688477,
      "epoch": 4.459766706207987,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003168279650773229,
      "loss": 4.9236,
      "mean_token_accuracy": 0.23301488757133484,
      "num_tokens": 103310055.0,
      "step": 45115
    },
    {
      "entropy": 5.768561267852784,
      "epoch": 4.460260972716489,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003167932897457509,
      "loss": 4.9862,
      "mean_token_accuracy": 0.2228391408920288,
      "num_tokens": 103321337.0,
      "step": 45120
    },
    {
      "entropy": 5.804376888275146,
      "epoch": 4.46075523922499,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003167586133860575,
      "loss": 4.9828,
      "mean_token_accuracy": 0.2259013384580612,
      "num_tokens": 103333753.0,
      "step": 45125
    },
    {
      "entropy": 5.762835931777954,
      "epoch": 4.461249505733491,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003167239359990957,
      "loss": 4.8127,
      "mean_token_accuracy": 0.23836423307657242,
      "num_tokens": 103345106.0,
      "step": 45130
    },
    {
      "entropy": 5.828259754180908,
      "epoch": 4.461743772241993,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00031668925758571864,
      "loss": 4.9527,
      "mean_token_accuracy": 0.22639204412698746,
      "num_tokens": 103355470.0,
      "step": 45135
    },
    {
      "entropy": 5.799319076538086,
      "epoch": 4.4622380387504945,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003166545781467793,
      "loss": 4.9397,
      "mean_token_accuracy": 0.22289031445980073,
      "num_tokens": 103366422.0,
      "step": 45140
    },
    {
      "entropy": 5.715981197357178,
      "epoch": 4.462732305258996,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003166198976831309,
      "loss": 4.7817,
      "mean_token_accuracy": 0.24110711812973024,
      "num_tokens": 103377721.0,
      "step": 45145
    },
    {
      "entropy": 5.8008959770202635,
      "epoch": 4.463226571767497,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00031658521619562665,
      "loss": 4.982,
      "mean_token_accuracy": 0.22364811599254608,
      "num_tokens": 103389399.0,
      "step": 45150
    },
    {
      "entropy": 5.702796506881714,
      "epoch": 4.463720838275998,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00031655053368511956,
      "loss": 4.93,
      "mean_token_accuracy": 0.23522112518548965,
      "num_tokens": 103400664.0,
      "step": 45155
    },
    {
      "entropy": 5.7671815872192385,
      "epoch": 4.4642151047845,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003165158501524629,
      "loss": 4.9009,
      "mean_token_accuracy": 0.2301260784268379,
      "num_tokens": 103411220.0,
      "step": 45160
    },
    {
      "entropy": 5.746367454528809,
      "epoch": 4.464709371293001,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00031648116559850986,
      "loss": 4.805,
      "mean_token_accuracy": 0.24461638331413268,
      "num_tokens": 103421987.0,
      "step": 45165
    },
    {
      "entropy": 5.690417146682739,
      "epoch": 4.465203637801503,
      "grad_norm": 1.34375,
      "learning_rate": 0.00031644648002411373,
      "loss": 4.835,
      "mean_token_accuracy": 0.23668057322502137,
      "num_tokens": 103433397.0,
      "step": 45170
    },
    {
      "entropy": 5.851757335662842,
      "epoch": 4.465697904310004,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003164117934301278,
      "loss": 5.0187,
      "mean_token_accuracy": 0.22407246530056,
      "num_tokens": 103445332.0,
      "step": 45175
    },
    {
      "entropy": 5.810300636291504,
      "epoch": 4.4661921708185055,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003163771058174053,
      "loss": 4.9308,
      "mean_token_accuracy": 0.22863934487104415,
      "num_tokens": 103456052.0,
      "step": 45180
    },
    {
      "entropy": 5.812199831008911,
      "epoch": 4.466686437327007,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003163424171867996,
      "loss": 4.9676,
      "mean_token_accuracy": 0.22579726874828338,
      "num_tokens": 103467341.0,
      "step": 45185
    },
    {
      "entropy": 5.716512012481689,
      "epoch": 4.467180703835508,
      "grad_norm": 1.296875,
      "learning_rate": 0.000316307727539164,
      "loss": 4.8285,
      "mean_token_accuracy": 0.2429486021399498,
      "num_tokens": 103479383.0,
      "step": 45190
    },
    {
      "entropy": 5.779064750671386,
      "epoch": 4.467674970344009,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003162730368753518,
      "loss": 4.9392,
      "mean_token_accuracy": 0.22418877482414246,
      "num_tokens": 103491650.0,
      "step": 45195
    },
    {
      "entropy": 5.847826766967773,
      "epoch": 4.4681692368525106,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00031623834519621665,
      "loss": 5.0074,
      "mean_token_accuracy": 0.22580087780952454,
      "num_tokens": 103503276.0,
      "step": 45200
    },
    {
      "entropy": 5.8123682022094725,
      "epoch": 4.468663503361013,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00031620365250261176,
      "loss": 4.9401,
      "mean_token_accuracy": 0.23036534190177918,
      "num_tokens": 103515425.0,
      "step": 45205
    },
    {
      "entropy": 5.706355810165405,
      "epoch": 4.469157769869514,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00031616895879539055,
      "loss": 4.9065,
      "mean_token_accuracy": 0.23564016222953796,
      "num_tokens": 103526166.0,
      "step": 45210
    },
    {
      "entropy": 5.8396265506744385,
      "epoch": 4.469652036378015,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00031613426407540665,
      "loss": 5.0278,
      "mean_token_accuracy": 0.22693254947662353,
      "num_tokens": 103537679.0,
      "step": 45215
    },
    {
      "entropy": 5.795495080947876,
      "epoch": 4.4701463028865165,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00031609956834351344,
      "loss": 4.8677,
      "mean_token_accuracy": 0.23599029034376146,
      "num_tokens": 103548792.0,
      "step": 45220
    },
    {
      "entropy": 5.816911363601685,
      "epoch": 4.470640569395018,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003160648716005645,
      "loss": 4.9626,
      "mean_token_accuracy": 0.22733725607395172,
      "num_tokens": 103559900.0,
      "step": 45225
    },
    {
      "entropy": 5.788603830337524,
      "epoch": 4.471134835903519,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003160301738474133,
      "loss": 4.9519,
      "mean_token_accuracy": 0.23107523918151857,
      "num_tokens": 103571820.0,
      "step": 45230
    },
    {
      "entropy": 5.7808185577392575,
      "epoch": 4.47162910241202,
      "grad_norm": 1.375,
      "learning_rate": 0.0003159954750849135,
      "loss": 4.9083,
      "mean_token_accuracy": 0.23181209862232208,
      "num_tokens": 103583246.0,
      "step": 45235
    },
    {
      "entropy": 5.850015926361084,
      "epoch": 4.4721233689205215,
      "grad_norm": 1.109375,
      "learning_rate": 0.00031596077531391865,
      "loss": 4.975,
      "mean_token_accuracy": 0.22433124631643295,
      "num_tokens": 103594951.0,
      "step": 45240
    },
    {
      "entropy": 5.827924823760986,
      "epoch": 4.472617635429024,
      "grad_norm": 1.375,
      "learning_rate": 0.00031592607453528236,
      "loss": 4.982,
      "mean_token_accuracy": 0.2310337319970131,
      "num_tokens": 103608304.0,
      "step": 45245
    },
    {
      "entropy": 5.854845142364502,
      "epoch": 4.473111901937525,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00031589137274985827,
      "loss": 4.9621,
      "mean_token_accuracy": 0.22753060460090638,
      "num_tokens": 103619051.0,
      "step": 45250
    },
    {
      "entropy": 5.8117451667785645,
      "epoch": 4.473606168446026,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003158566699585001,
      "loss": 4.9618,
      "mean_token_accuracy": 0.23229935616254807,
      "num_tokens": 103630330.0,
      "step": 45255
    },
    {
      "entropy": 5.848821592330933,
      "epoch": 4.4741004349545275,
      "grad_norm": 1.21875,
      "learning_rate": 0.00031582196616206147,
      "loss": 5.0284,
      "mean_token_accuracy": 0.22590140849351883,
      "num_tokens": 103642450.0,
      "step": 45260
    },
    {
      "entropy": 5.817962217330932,
      "epoch": 4.474594701463029,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00031578726136139617,
      "loss": 4.966,
      "mean_token_accuracy": 0.2260540872812271,
      "num_tokens": 103653389.0,
      "step": 45265
    },
    {
      "entropy": 5.775904130935669,
      "epoch": 4.47508896797153,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003157525555573579,
      "loss": 4.9264,
      "mean_token_accuracy": 0.2303704097867012,
      "num_tokens": 103664586.0,
      "step": 45270
    },
    {
      "entropy": 5.7523026943206785,
      "epoch": 4.475583234480031,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003157178487508005,
      "loss": 4.9426,
      "mean_token_accuracy": 0.23125007450580598,
      "num_tokens": 103675987.0,
      "step": 45275
    },
    {
      "entropy": 5.761595344543457,
      "epoch": 4.476077500988533,
      "grad_norm": 1.375,
      "learning_rate": 0.00031568314094257763,
      "loss": 4.9427,
      "mean_token_accuracy": 0.22179912477731706,
      "num_tokens": 103688649.0,
      "step": 45280
    },
    {
      "entropy": 5.799959373474121,
      "epoch": 4.476571767497035,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003156484321335432,
      "loss": 4.9848,
      "mean_token_accuracy": 0.23477091640233994,
      "num_tokens": 103700582.0,
      "step": 45285
    },
    {
      "entropy": 5.882960033416748,
      "epoch": 4.477066034005536,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00031561372232455105,
      "loss": 5.0937,
      "mean_token_accuracy": 0.218164823949337,
      "num_tokens": 103712705.0,
      "step": 45290
    },
    {
      "entropy": 5.76187915802002,
      "epoch": 4.477560300514037,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003155790115164551,
      "loss": 4.9139,
      "mean_token_accuracy": 0.23022040575742722,
      "num_tokens": 103723128.0,
      "step": 45295
    },
    {
      "entropy": 5.819494581222534,
      "epoch": 4.4780545670225385,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00031554429971010907,
      "loss": 4.9181,
      "mean_token_accuracy": 0.22745622992515563,
      "num_tokens": 103734443.0,
      "step": 45300
    },
    {
      "entropy": 5.732424211502075,
      "epoch": 4.47854883353104,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00031550958690636693,
      "loss": 4.8758,
      "mean_token_accuracy": 0.2363263949751854,
      "num_tokens": 103746838.0,
      "step": 45305
    },
    {
      "entropy": 5.768497657775879,
      "epoch": 4.479043100039541,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00031547487310608276,
      "loss": 4.942,
      "mean_token_accuracy": 0.23609296828508378,
      "num_tokens": 103756924.0,
      "step": 45310
    },
    {
      "entropy": 5.803349876403809,
      "epoch": 4.479537366548042,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003154401583101104,
      "loss": 4.9651,
      "mean_token_accuracy": 0.2180570513010025,
      "num_tokens": 103769332.0,
      "step": 45315
    },
    {
      "entropy": 5.789024734497071,
      "epoch": 4.480031633056544,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003154054425193039,
      "loss": 4.9036,
      "mean_token_accuracy": 0.2341454282402992,
      "num_tokens": 103781149.0,
      "step": 45320
    },
    {
      "entropy": 5.85286545753479,
      "epoch": 4.480525899565046,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00031537072573451717,
      "loss": 5.0133,
      "mean_token_accuracy": 0.22293965071439742,
      "num_tokens": 103792217.0,
      "step": 45325
    },
    {
      "entropy": 5.742894458770752,
      "epoch": 4.481020166073547,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003153360079566044,
      "loss": 4.8979,
      "mean_token_accuracy": 0.23371906727552413,
      "num_tokens": 103803628.0,
      "step": 45330
    },
    {
      "entropy": 5.748565149307251,
      "epoch": 4.481514432582048,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003153012891864196,
      "loss": 4.9089,
      "mean_token_accuracy": 0.22793481796979903,
      "num_tokens": 103814735.0,
      "step": 45335
    },
    {
      "entropy": 5.792451572418213,
      "epoch": 4.482008699090549,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00031526656942481673,
      "loss": 4.9336,
      "mean_token_accuracy": 0.23010524362325668,
      "num_tokens": 103825952.0,
      "step": 45340
    },
    {
      "entropy": 5.71683259010315,
      "epoch": 4.482502965599051,
      "grad_norm": 1.28125,
      "learning_rate": 0.00031523184867265013,
      "loss": 4.9292,
      "mean_token_accuracy": 0.2290325328707695,
      "num_tokens": 103836558.0,
      "step": 45345
    },
    {
      "entropy": 5.789231777191162,
      "epoch": 4.482997232107552,
      "grad_norm": 1.296875,
      "learning_rate": 0.00031519712693077375,
      "loss": 4.9554,
      "mean_token_accuracy": 0.22570794969797134,
      "num_tokens": 103848453.0,
      "step": 45350
    },
    {
      "entropy": 5.769028854370117,
      "epoch": 4.483491498616054,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003151624042000418,
      "loss": 4.8409,
      "mean_token_accuracy": 0.23806987851858138,
      "num_tokens": 103860603.0,
      "step": 45355
    },
    {
      "entropy": 5.764129686355591,
      "epoch": 4.483985765124555,
      "grad_norm": 1.34375,
      "learning_rate": 0.00031512768048130846,
      "loss": 4.9581,
      "mean_token_accuracy": 0.23115486800670623,
      "num_tokens": 103870678.0,
      "step": 45360
    },
    {
      "entropy": 5.759457969665528,
      "epoch": 4.484480031633057,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00031509295577542804,
      "loss": 4.9178,
      "mean_token_accuracy": 0.22873513251543046,
      "num_tokens": 103881139.0,
      "step": 45365
    },
    {
      "entropy": 5.766952800750732,
      "epoch": 4.484974298141558,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003150582300832546,
      "loss": 4.8924,
      "mean_token_accuracy": 0.2364226222038269,
      "num_tokens": 103892623.0,
      "step": 45370
    },
    {
      "entropy": 5.85888442993164,
      "epoch": 4.485468564650059,
      "grad_norm": 1.25,
      "learning_rate": 0.0003150235034056425,
      "loss": 4.9464,
      "mean_token_accuracy": 0.22653366774320602,
      "num_tokens": 103905278.0,
      "step": 45375
    },
    {
      "entropy": 5.813764762878418,
      "epoch": 4.48596283115856,
      "grad_norm": 1.265625,
      "learning_rate": 0.00031498877574344603,
      "loss": 4.9142,
      "mean_token_accuracy": 0.22859687060117723,
      "num_tokens": 103916363.0,
      "step": 45380
    },
    {
      "entropy": 5.725140190124511,
      "epoch": 4.486457097667062,
      "grad_norm": 1.140625,
      "learning_rate": 0.00031495404709751945,
      "loss": 4.9219,
      "mean_token_accuracy": 0.2254478380084038,
      "num_tokens": 103927764.0,
      "step": 45385
    },
    {
      "entropy": 5.737674713134766,
      "epoch": 4.486951364175564,
      "grad_norm": 1.453125,
      "learning_rate": 0.00031491931746871715,
      "loss": 4.9028,
      "mean_token_accuracy": 0.2328400582075119,
      "num_tokens": 103938415.0,
      "step": 45390
    },
    {
      "entropy": 5.7559370517730715,
      "epoch": 4.487445630684065,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003148845868578934,
      "loss": 4.8451,
      "mean_token_accuracy": 0.23926866203546523,
      "num_tokens": 103949282.0,
      "step": 45395
    },
    {
      "entropy": 5.72900800704956,
      "epoch": 4.487939897192566,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003148498552659026,
      "loss": 4.9005,
      "mean_token_accuracy": 0.23042674511671066,
      "num_tokens": 103960896.0,
      "step": 45400
    },
    {
      "entropy": 5.753850936889648,
      "epoch": 4.488434163701068,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003148151226935993,
      "loss": 4.8691,
      "mean_token_accuracy": 0.24018344581127166,
      "num_tokens": 103970696.0,
      "step": 45405
    },
    {
      "entropy": 5.900739955902099,
      "epoch": 4.488928430209569,
      "grad_norm": 1.265625,
      "learning_rate": 0.00031478038914183764,
      "loss": 5.1279,
      "mean_token_accuracy": 0.210573410987854,
      "num_tokens": 103984218.0,
      "step": 45410
    },
    {
      "entropy": 5.766367197036743,
      "epoch": 4.48942269671807,
      "grad_norm": 1.390625,
      "learning_rate": 0.00031474565461147235,
      "loss": 4.9062,
      "mean_token_accuracy": 0.23101238161325455,
      "num_tokens": 103994871.0,
      "step": 45415
    },
    {
      "entropy": 5.656657123565674,
      "epoch": 4.489916963226571,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003147109191033578,
      "loss": 4.7248,
      "mean_token_accuracy": 0.24869392067193985,
      "num_tokens": 104005629.0,
      "step": 45420
    },
    {
      "entropy": 5.748292541503906,
      "epoch": 4.490411229735074,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003146761826183484,
      "loss": 4.8395,
      "mean_token_accuracy": 0.2364925593137741,
      "num_tokens": 104016257.0,
      "step": 45425
    },
    {
      "entropy": 5.741846513748169,
      "epoch": 4.490905496243575,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003146414451572988,
      "loss": 4.8516,
      "mean_token_accuracy": 0.23508195877075194,
      "num_tokens": 104028322.0,
      "step": 45430
    },
    {
      "entropy": 5.849905490875244,
      "epoch": 4.491399762752076,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00031460670672106355,
      "loss": 4.8885,
      "mean_token_accuracy": 0.23126174956560136,
      "num_tokens": 104039349.0,
      "step": 45435
    },
    {
      "entropy": 5.797709989547729,
      "epoch": 4.491894029260577,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003145719673104971,
      "loss": 4.9685,
      "mean_token_accuracy": 0.21894969940185546,
      "num_tokens": 104050585.0,
      "step": 45440
    },
    {
      "entropy": 5.7235959529876705,
      "epoch": 4.492388295769079,
      "grad_norm": 1.359375,
      "learning_rate": 0.00031453722692645407,
      "loss": 4.831,
      "mean_token_accuracy": 0.23474867045879363,
      "num_tokens": 104060623.0,
      "step": 45445
    },
    {
      "entropy": 5.712519216537475,
      "epoch": 4.49288256227758,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00031450248556978927,
      "loss": 4.8772,
      "mean_token_accuracy": 0.23690192252397538,
      "num_tokens": 104070880.0,
      "step": 45450
    },
    {
      "entropy": 5.808254051208496,
      "epoch": 4.493376828786081,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003144677432413571,
      "loss": 4.9881,
      "mean_token_accuracy": 0.22447519600391388,
      "num_tokens": 104082686.0,
      "step": 45455
    },
    {
      "entropy": 5.764213514328003,
      "epoch": 4.493871095294583,
      "grad_norm": 1.421875,
      "learning_rate": 0.00031443299994201234,
      "loss": 4.9617,
      "mean_token_accuracy": 0.2254466861486435,
      "num_tokens": 104093746.0,
      "step": 45460
    },
    {
      "entropy": 5.806276178359985,
      "epoch": 4.4943653618030845,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003143982556726097,
      "loss": 4.9247,
      "mean_token_accuracy": 0.2323767676949501,
      "num_tokens": 104104843.0,
      "step": 45465
    },
    {
      "entropy": 5.718568229675293,
      "epoch": 4.494859628311586,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003143635104340037,
      "loss": 4.8062,
      "mean_token_accuracy": 0.24710734635591508,
      "num_tokens": 104115317.0,
      "step": 45470
    },
    {
      "entropy": 5.7181707382202145,
      "epoch": 4.495353894820087,
      "grad_norm": 1.15625,
      "learning_rate": 0.00031432876422704943,
      "loss": 4.8836,
      "mean_token_accuracy": 0.22716282606124877,
      "num_tokens": 104126775.0,
      "step": 45475
    },
    {
      "entropy": 5.7147480010986325,
      "epoch": 4.495848161328588,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003142940170526014,
      "loss": 4.873,
      "mean_token_accuracy": 0.23321637362241746,
      "num_tokens": 104138343.0,
      "step": 45480
    },
    {
      "entropy": 5.861667728424072,
      "epoch": 4.49634242783709,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003142592689115145,
      "loss": 4.9535,
      "mean_token_accuracy": 0.22753961831331254,
      "num_tokens": 104149406.0,
      "step": 45485
    },
    {
      "entropy": 5.820281553268432,
      "epoch": 4.496836694345591,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003142245198046434,
      "loss": 4.9479,
      "mean_token_accuracy": 0.22834640890359878,
      "num_tokens": 104159665.0,
      "step": 45490
    },
    {
      "entropy": 5.78868670463562,
      "epoch": 4.497330960854092,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003141897697328431,
      "loss": 4.9912,
      "mean_token_accuracy": 0.2187335729598999,
      "num_tokens": 104171195.0,
      "step": 45495
    },
    {
      "entropy": 5.774057769775391,
      "epoch": 4.497825227362594,
      "grad_norm": 1.5,
      "learning_rate": 0.0003141550186969683,
      "loss": 4.9262,
      "mean_token_accuracy": 0.23265533596277238,
      "num_tokens": 104182035.0,
      "step": 45500
    },
    {
      "entropy": 5.798000621795654,
      "epoch": 4.4983194938710955,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000314120266697874,
      "loss": 4.9079,
      "mean_token_accuracy": 0.22686358839273452,
      "num_tokens": 104192643.0,
      "step": 45505
    },
    {
      "entropy": 5.732927227020264,
      "epoch": 4.498813760379597,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00031408551373641514,
      "loss": 4.8567,
      "mean_token_accuracy": 0.23118343949317932,
      "num_tokens": 104205079.0,
      "step": 45510
    },
    {
      "entropy": 5.758198213577271,
      "epoch": 4.499308026888098,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003140507598134465,
      "loss": 4.9619,
      "mean_token_accuracy": 0.22676891833543777,
      "num_tokens": 104216495.0,
      "step": 45515
    },
    {
      "entropy": 5.81792140007019,
      "epoch": 4.499802293396599,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003140160049298231,
      "loss": 5.0228,
      "mean_token_accuracy": 0.22965183854103088,
      "num_tokens": 104228115.0,
      "step": 45520
    },
    {
      "entropy": 5.784226560592652,
      "epoch": 4.500296559905101,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00031398124908639996,
      "loss": 4.9113,
      "mean_token_accuracy": 0.231858529150486,
      "num_tokens": 104241184.0,
      "step": 45525
    },
    {
      "entropy": 5.786034202575683,
      "epoch": 4.500790826413602,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00031394649228403207,
      "loss": 4.9453,
      "mean_token_accuracy": 0.22771105915308,
      "num_tokens": 104253039.0,
      "step": 45530
    },
    {
      "entropy": 5.731417846679688,
      "epoch": 4.501285092922103,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0003139117345235744,
      "loss": 4.86,
      "mean_token_accuracy": 0.23112557083368301,
      "num_tokens": 104263536.0,
      "step": 45535
    },
    {
      "entropy": 5.697473049163818,
      "epoch": 4.501779359430605,
      "grad_norm": 1.375,
      "learning_rate": 0.000313876975805882,
      "loss": 4.8434,
      "mean_token_accuracy": 0.24049308001995087,
      "num_tokens": 104275744.0,
      "step": 45540
    },
    {
      "entropy": 5.7183098793029785,
      "epoch": 4.5022736259391065,
      "grad_norm": 1.25,
      "learning_rate": 0.00031384221613181,
      "loss": 4.8633,
      "mean_token_accuracy": 0.23792820274829865,
      "num_tokens": 104286338.0,
      "step": 45545
    },
    {
      "entropy": 5.7764688491821286,
      "epoch": 4.502767892447608,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00031380745550221346,
      "loss": 4.899,
      "mean_token_accuracy": 0.2315663665533066,
      "num_tokens": 104298641.0,
      "step": 45550
    },
    {
      "entropy": 5.7993558883667,
      "epoch": 4.503262158956109,
      "grad_norm": 1.34375,
      "learning_rate": 0.00031377269391794744,
      "loss": 4.9367,
      "mean_token_accuracy": 0.22540636956691742,
      "num_tokens": 104308973.0,
      "step": 45555
    },
    {
      "entropy": 5.724284648895264,
      "epoch": 4.50375642546461,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00031373793137986724,
      "loss": 4.8633,
      "mean_token_accuracy": 0.24258564561605453,
      "num_tokens": 104320949.0,
      "step": 45560
    },
    {
      "entropy": 5.799285650253296,
      "epoch": 4.504250691973112,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003137031678888278,
      "loss": 4.9345,
      "mean_token_accuracy": 0.23465042263269426,
      "num_tokens": 104332190.0,
      "step": 45565
    },
    {
      "entropy": 5.818480157852173,
      "epoch": 4.504744958481613,
      "grad_norm": 1.265625,
      "learning_rate": 0.00031366840344568443,
      "loss": 4.9429,
      "mean_token_accuracy": 0.22992178946733474,
      "num_tokens": 104342936.0,
      "step": 45570
    },
    {
      "entropy": 5.789445304870606,
      "epoch": 4.505239224990115,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003136336380512925,
      "loss": 4.9511,
      "mean_token_accuracy": 0.22872791588306426,
      "num_tokens": 104353556.0,
      "step": 45575
    },
    {
      "entropy": 5.818964195251465,
      "epoch": 4.505733491498616,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00031359887170650695,
      "loss": 4.9685,
      "mean_token_accuracy": 0.22081490308046342,
      "num_tokens": 104364960.0,
      "step": 45580
    },
    {
      "entropy": 5.8182331085205075,
      "epoch": 4.5062277580071175,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003135641044121831,
      "loss": 4.9641,
      "mean_token_accuracy": 0.2231404036283493,
      "num_tokens": 104376101.0,
      "step": 45585
    },
    {
      "entropy": 5.7885191440582275,
      "epoch": 4.506722024515619,
      "grad_norm": 1.359375,
      "learning_rate": 0.00031352933616917637,
      "loss": 4.9624,
      "mean_token_accuracy": 0.22957054078578948,
      "num_tokens": 104386981.0,
      "step": 45590
    },
    {
      "entropy": 5.755187892913819,
      "epoch": 4.50721629102412,
      "grad_norm": 1.328125,
      "learning_rate": 0.00031349456697834203,
      "loss": 4.9001,
      "mean_token_accuracy": 0.22705262899398804,
      "num_tokens": 104398275.0,
      "step": 45595
    },
    {
      "entropy": 5.805354499816895,
      "epoch": 4.507710557532621,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003134597968405354,
      "loss": 4.9888,
      "mean_token_accuracy": 0.22353918552398683,
      "num_tokens": 104409337.0,
      "step": 45600
    },
    {
      "entropy": 5.807986831665039,
      "epoch": 4.5082048240411225,
      "grad_norm": 1.34375,
      "learning_rate": 0.00031342502575661173,
      "loss": 4.9553,
      "mean_token_accuracy": 0.23028065413236617,
      "num_tokens": 104420880.0,
      "step": 45605
    },
    {
      "entropy": 5.744704484939575,
      "epoch": 4.508699090549625,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00031339025372742647,
      "loss": 4.8511,
      "mean_token_accuracy": 0.23837293088436126,
      "num_tokens": 104432398.0,
      "step": 45610
    },
    {
      "entropy": 5.760972690582276,
      "epoch": 4.509193357058126,
      "grad_norm": 1.296875,
      "learning_rate": 0.00031335548075383504,
      "loss": 4.939,
      "mean_token_accuracy": 0.22678867727518082,
      "num_tokens": 104444008.0,
      "step": 45615
    },
    {
      "entropy": 5.8074994564056395,
      "epoch": 4.509687623566627,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003133207068366928,
      "loss": 4.9785,
      "mean_token_accuracy": 0.22550204545259475,
      "num_tokens": 104456059.0,
      "step": 45620
    },
    {
      "entropy": 5.758225345611573,
      "epoch": 4.5101818900751285,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003132859319768553,
      "loss": 4.9449,
      "mean_token_accuracy": 0.23146769553422927,
      "num_tokens": 104467757.0,
      "step": 45625
    },
    {
      "entropy": 5.782956886291504,
      "epoch": 4.51067615658363,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00031325115617517784,
      "loss": 4.9561,
      "mean_token_accuracy": 0.22635255008935928,
      "num_tokens": 104479102.0,
      "step": 45630
    },
    {
      "entropy": 5.733843851089477,
      "epoch": 4.511170423092131,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000313216379432516,
      "loss": 4.8669,
      "mean_token_accuracy": 0.23859147876501083,
      "num_tokens": 104490515.0,
      "step": 45635
    },
    {
      "entropy": 5.817013931274414,
      "epoch": 4.511664689600632,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003131816017497254,
      "loss": 4.9485,
      "mean_token_accuracy": 0.22880192399024962,
      "num_tokens": 104502539.0,
      "step": 45640
    },
    {
      "entropy": 5.784474802017212,
      "epoch": 4.512158956109134,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003131468231276614,
      "loss": 4.9145,
      "mean_token_accuracy": 0.22501565963029863,
      "num_tokens": 104514763.0,
      "step": 45645
    },
    {
      "entropy": 5.838056564331055,
      "epoch": 4.512653222617636,
      "grad_norm": 1.171875,
      "learning_rate": 0.00031311204356717964,
      "loss": 4.9709,
      "mean_token_accuracy": 0.22233615964651107,
      "num_tokens": 104527192.0,
      "step": 45650
    },
    {
      "entropy": 5.766119289398193,
      "epoch": 4.513147489126137,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003130772630691357,
      "loss": 4.9044,
      "mean_token_accuracy": 0.240135994553566,
      "num_tokens": 104539482.0,
      "step": 45655
    },
    {
      "entropy": 5.793220186233521,
      "epoch": 4.513641755634638,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00031304248163438513,
      "loss": 4.9257,
      "mean_token_accuracy": 0.2305889382958412,
      "num_tokens": 104551168.0,
      "step": 45660
    },
    {
      "entropy": 5.826167106628418,
      "epoch": 4.5141360221431395,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00031300769926378367,
      "loss": 4.9876,
      "mean_token_accuracy": 0.22297198474407195,
      "num_tokens": 104562516.0,
      "step": 45665
    },
    {
      "entropy": 5.810970640182495,
      "epoch": 4.514630288651641,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00031297291595818697,
      "loss": 4.9268,
      "mean_token_accuracy": 0.23705515265464783,
      "num_tokens": 104573665.0,
      "step": 45670
    },
    {
      "entropy": 5.7266350269317625,
      "epoch": 4.515124555160142,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003129381317184506,
      "loss": 4.8754,
      "mean_token_accuracy": 0.23252833336591722,
      "num_tokens": 104584596.0,
      "step": 45675
    },
    {
      "entropy": 5.692915964126587,
      "epoch": 4.515618821668644,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003129033465454303,
      "loss": 4.8809,
      "mean_token_accuracy": 0.23240439593791962,
      "num_tokens": 104596961.0,
      "step": 45680
    },
    {
      "entropy": 5.813670492172241,
      "epoch": 4.516113088177145,
      "grad_norm": 1.234375,
      "learning_rate": 0.00031286856043998175,
      "loss": 4.9821,
      "mean_token_accuracy": 0.23345856368541718,
      "num_tokens": 104609479.0,
      "step": 45685
    },
    {
      "entropy": 5.839987182617188,
      "epoch": 4.516607354685647,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003128337734029607,
      "loss": 4.9974,
      "mean_token_accuracy": 0.22323267608880998,
      "num_tokens": 104622690.0,
      "step": 45690
    },
    {
      "entropy": 5.807674551010132,
      "epoch": 4.517101621194148,
      "grad_norm": 1.375,
      "learning_rate": 0.0003127989854352231,
      "loss": 4.8806,
      "mean_token_accuracy": 0.23564257323741913,
      "num_tokens": 104633005.0,
      "step": 45695
    },
    {
      "entropy": 5.799054098129273,
      "epoch": 4.517595887702649,
      "grad_norm": 1.25,
      "learning_rate": 0.0003127641965376245,
      "loss": 4.9637,
      "mean_token_accuracy": 0.22782188057899475,
      "num_tokens": 104646209.0,
      "step": 45700
    },
    {
      "entropy": 5.773272943496704,
      "epoch": 4.5180901542111505,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003127294067110208,
      "loss": 4.9308,
      "mean_token_accuracy": 0.2275946170091629,
      "num_tokens": 104657833.0,
      "step": 45705
    },
    {
      "entropy": 5.786072015762329,
      "epoch": 4.518584420719652,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00031269461595626784,
      "loss": 4.9649,
      "mean_token_accuracy": 0.2236754685640335,
      "num_tokens": 104668858.0,
      "step": 45710
    },
    {
      "entropy": 5.78107590675354,
      "epoch": 4.519078687228154,
      "grad_norm": 1.3125,
      "learning_rate": 0.00031265982427422154,
      "loss": 4.908,
      "mean_token_accuracy": 0.22265292406082154,
      "num_tokens": 104681231.0,
      "step": 45715
    },
    {
      "entropy": 5.852201128005982,
      "epoch": 4.519572953736655,
      "grad_norm": 1.234375,
      "learning_rate": 0.00031262503166573776,
      "loss": 5.0271,
      "mean_token_accuracy": 0.22966112047433854,
      "num_tokens": 104693543.0,
      "step": 45720
    },
    {
      "entropy": 5.773343944549561,
      "epoch": 4.520067220245156,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00031259023813167227,
      "loss": 4.8934,
      "mean_token_accuracy": 0.2293206572532654,
      "num_tokens": 104704642.0,
      "step": 45725
    },
    {
      "entropy": 5.7721751689910885,
      "epoch": 4.520561486753658,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003125554436728811,
      "loss": 4.9291,
      "mean_token_accuracy": 0.2256588250398636,
      "num_tokens": 104715931.0,
      "step": 45730
    },
    {
      "entropy": 5.772691535949707,
      "epoch": 4.521055753262159,
      "grad_norm": 1.3125,
      "learning_rate": 0.00031252064829022017,
      "loss": 4.9025,
      "mean_token_accuracy": 0.2356062114238739,
      "num_tokens": 104727561.0,
      "step": 45735
    },
    {
      "entropy": 5.802690696716309,
      "epoch": 4.52155001977066,
      "grad_norm": 1.265625,
      "learning_rate": 0.00031248585198454556,
      "loss": 4.9542,
      "mean_token_accuracy": 0.22885042130947114,
      "num_tokens": 104740127.0,
      "step": 45740
    },
    {
      "entropy": 5.8079160213470455,
      "epoch": 4.522044286279161,
      "grad_norm": 1.296875,
      "learning_rate": 0.00031245105475671305,
      "loss": 4.9222,
      "mean_token_accuracy": 0.23039889186620713,
      "num_tokens": 104752255.0,
      "step": 45745
    },
    {
      "entropy": 5.810000658035278,
      "epoch": 4.522538552787664,
      "grad_norm": 1.390625,
      "learning_rate": 0.00031241625660757887,
      "loss": 4.9906,
      "mean_token_accuracy": 0.21968691498041154,
      "num_tokens": 104763956.0,
      "step": 45750
    },
    {
      "entropy": 5.761193752288818,
      "epoch": 4.523032819296165,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00031238145753799895,
      "loss": 4.9083,
      "mean_token_accuracy": 0.22877920120954515,
      "num_tokens": 104775608.0,
      "step": 45755
    },
    {
      "entropy": 5.797948503494263,
      "epoch": 4.523527085804666,
      "grad_norm": 1.15625,
      "learning_rate": 0.00031234665754882937,
      "loss": 4.8653,
      "mean_token_accuracy": 0.2278096243739128,
      "num_tokens": 104786958.0,
      "step": 45760
    },
    {
      "entropy": 5.829562187194824,
      "epoch": 4.524021352313167,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00031231185664092623,
      "loss": 4.9983,
      "mean_token_accuracy": 0.21904897093772888,
      "num_tokens": 104798155.0,
      "step": 45765
    },
    {
      "entropy": 5.781052780151367,
      "epoch": 4.524515618821669,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00031227705481514566,
      "loss": 4.9041,
      "mean_token_accuracy": 0.22937185317277908,
      "num_tokens": 104809378.0,
      "step": 45770
    },
    {
      "entropy": 5.740417003631592,
      "epoch": 4.52500988533017,
      "grad_norm": 1.265625,
      "learning_rate": 0.00031224225207234375,
      "loss": 4.8765,
      "mean_token_accuracy": 0.2324560448527336,
      "num_tokens": 104820223.0,
      "step": 45775
    },
    {
      "entropy": 5.826227712631225,
      "epoch": 4.525504151838671,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003122074484133766,
      "loss": 4.9391,
      "mean_token_accuracy": 0.23290713280439376,
      "num_tokens": 104833665.0,
      "step": 45780
    },
    {
      "entropy": 5.781940126419068,
      "epoch": 4.525998418347172,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003121726438391005,
      "loss": 4.9134,
      "mean_token_accuracy": 0.23552767634391786,
      "num_tokens": 104845405.0,
      "step": 45785
    },
    {
      "entropy": 5.8112023830413815,
      "epoch": 4.526492684855674,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003121378383503716,
      "loss": 5.017,
      "mean_token_accuracy": 0.2193110242486,
      "num_tokens": 104856388.0,
      "step": 45790
    },
    {
      "entropy": 5.776507425308227,
      "epoch": 4.526986951364176,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003121030319480461,
      "loss": 4.9327,
      "mean_token_accuracy": 0.2277138724923134,
      "num_tokens": 104868214.0,
      "step": 45795
    },
    {
      "entropy": 5.772354698181152,
      "epoch": 4.527481217872677,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00031206822463298017,
      "loss": 4.8776,
      "mean_token_accuracy": 0.2362642616033554,
      "num_tokens": 104880894.0,
      "step": 45800
    },
    {
      "entropy": 5.808596515655518,
      "epoch": 4.527975484381178,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0003120334164060303,
      "loss": 4.9552,
      "mean_token_accuracy": 0.22429129034280776,
      "num_tokens": 104894244.0,
      "step": 45805
    },
    {
      "entropy": 5.79736180305481,
      "epoch": 4.52846975088968,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003119986072680526,
      "loss": 4.8902,
      "mean_token_accuracy": 0.2254765287041664,
      "num_tokens": 104904960.0,
      "step": 45810
    },
    {
      "entropy": 5.778193140029908,
      "epoch": 4.528964017398181,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003119637972199034,
      "loss": 4.9993,
      "mean_token_accuracy": 0.22231073528528214,
      "num_tokens": 104917574.0,
      "step": 45815
    },
    {
      "entropy": 5.716688108444214,
      "epoch": 4.529458283906682,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00031192898626243907,
      "loss": 4.8244,
      "mean_token_accuracy": 0.23645983934402465,
      "num_tokens": 104928131.0,
      "step": 45820
    },
    {
      "entropy": 5.824159383773804,
      "epoch": 4.529952550415183,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00031189417439651584,
      "loss": 4.9424,
      "mean_token_accuracy": 0.22612955868244172,
      "num_tokens": 104939166.0,
      "step": 45825
    },
    {
      "entropy": 5.758881902694702,
      "epoch": 4.530446816923686,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003118593616229904,
      "loss": 4.9334,
      "mean_token_accuracy": 0.23210021555423738,
      "num_tokens": 104949595.0,
      "step": 45830
    },
    {
      "entropy": 5.810052919387817,
      "epoch": 4.530941083432187,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003118245479427188,
      "loss": 4.9275,
      "mean_token_accuracy": 0.22450329959392548,
      "num_tokens": 104961089.0,
      "step": 45835
    },
    {
      "entropy": 5.780639266967773,
      "epoch": 4.531435349940688,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003117897333565577,
      "loss": 4.9542,
      "mean_token_accuracy": 0.22893834710121155,
      "num_tokens": 104972199.0,
      "step": 45840
    },
    {
      "entropy": 5.761220932006836,
      "epoch": 4.531929616449189,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003117549178653634,
      "loss": 4.8822,
      "mean_token_accuracy": 0.24257535338401795,
      "num_tokens": 104984286.0,
      "step": 45845
    },
    {
      "entropy": 5.80683708190918,
      "epoch": 4.532423882957691,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003117201014699924,
      "loss": 4.8793,
      "mean_token_accuracy": 0.23434792459011078,
      "num_tokens": 104995339.0,
      "step": 45850
    },
    {
      "entropy": 5.837441444396973,
      "epoch": 4.532918149466192,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003116852841713012,
      "loss": 5.0316,
      "mean_token_accuracy": 0.216287998855114,
      "num_tokens": 105006199.0,
      "step": 45855
    },
    {
      "entropy": 5.794378137588501,
      "epoch": 4.533412415974693,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003116504659701463,
      "loss": 4.9235,
      "mean_token_accuracy": 0.2343118518590927,
      "num_tokens": 105017222.0,
      "step": 45860
    },
    {
      "entropy": 5.753310632705689,
      "epoch": 4.533906682483195,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00031161564686738437,
      "loss": 4.8946,
      "mean_token_accuracy": 0.22981908917427063,
      "num_tokens": 105028260.0,
      "step": 45865
    },
    {
      "entropy": 5.770794534683228,
      "epoch": 4.5344009489916965,
      "grad_norm": 1.390625,
      "learning_rate": 0.00031158082686387175,
      "loss": 4.9073,
      "mean_token_accuracy": 0.23393017202615737,
      "num_tokens": 105040397.0,
      "step": 45870
    },
    {
      "entropy": 5.7791064262390135,
      "epoch": 4.534895215500198,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003115460059604652,
      "loss": 4.9727,
      "mean_token_accuracy": 0.23050335496664048,
      "num_tokens": 105051291.0,
      "step": 45875
    },
    {
      "entropy": 5.86417965888977,
      "epoch": 4.535389482008699,
      "grad_norm": 1.375,
      "learning_rate": 0.0003115111841580211,
      "loss": 5.022,
      "mean_token_accuracy": 0.22395497411489487,
      "num_tokens": 105063224.0,
      "step": 45880
    },
    {
      "entropy": 5.839791774749756,
      "epoch": 4.5358837485172,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00031147636145739633,
      "loss": 4.9596,
      "mean_token_accuracy": 0.22914834022521974,
      "num_tokens": 105074750.0,
      "step": 45885
    },
    {
      "entropy": 5.760359907150269,
      "epoch": 4.536378015025702,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003114415378594473,
      "loss": 4.8946,
      "mean_token_accuracy": 0.23531904369592666,
      "num_tokens": 105086782.0,
      "step": 45890
    },
    {
      "entropy": 5.8248707294464115,
      "epoch": 4.536872281534203,
      "grad_norm": 1.203125,
      "learning_rate": 0.00031140671336503086,
      "loss": 5.0713,
      "mean_token_accuracy": 0.22133518904447555,
      "num_tokens": 105099540.0,
      "step": 45895
    },
    {
      "entropy": 5.826449537277222,
      "epoch": 4.537366548042705,
      "grad_norm": 1.296875,
      "learning_rate": 0.00031137188797500365,
      "loss": 5.0316,
      "mean_token_accuracy": 0.22160473167896272,
      "num_tokens": 105112212.0,
      "step": 45900
    },
    {
      "entropy": 5.823426818847656,
      "epoch": 4.537860814551206,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003113370616902223,
      "loss": 4.9833,
      "mean_token_accuracy": 0.22607220560312272,
      "num_tokens": 105124397.0,
      "step": 45905
    },
    {
      "entropy": 5.879618072509766,
      "epoch": 4.5383550810597075,
      "grad_norm": 1.25,
      "learning_rate": 0.0003113022345115436,
      "loss": 5.0173,
      "mean_token_accuracy": 0.2288053959608078,
      "num_tokens": 105138114.0,
      "step": 45910
    },
    {
      "entropy": 5.85983247756958,
      "epoch": 4.538849347568209,
      "grad_norm": 1.296875,
      "learning_rate": 0.00031126740643982434,
      "loss": 4.9456,
      "mean_token_accuracy": 0.23041774481534957,
      "num_tokens": 105148967.0,
      "step": 45915
    },
    {
      "entropy": 5.724219799041748,
      "epoch": 4.53934361407671,
      "grad_norm": 1.28125,
      "learning_rate": 0.00031123257747592113,
      "loss": 4.8607,
      "mean_token_accuracy": 0.23958863615989684,
      "num_tokens": 105161472.0,
      "step": 45920
    },
    {
      "entropy": 5.7243719577789305,
      "epoch": 4.539837880585211,
      "grad_norm": 1.25,
      "learning_rate": 0.00031119774762069103,
      "loss": 4.9277,
      "mean_token_accuracy": 0.2267378032207489,
      "num_tokens": 105174213.0,
      "step": 45925
    },
    {
      "entropy": 5.865203762054444,
      "epoch": 4.540332147093713,
      "grad_norm": 1.375,
      "learning_rate": 0.00031116291687499067,
      "loss": 5.0321,
      "mean_token_accuracy": 0.22186678797006607,
      "num_tokens": 105186011.0,
      "step": 45930
    },
    {
      "entropy": 5.791933488845825,
      "epoch": 4.540826413602215,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00031112808523967695,
      "loss": 4.9214,
      "mean_token_accuracy": 0.23054639250040054,
      "num_tokens": 105197605.0,
      "step": 45935
    },
    {
      "entropy": 5.833133840560913,
      "epoch": 4.541320680110716,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003110932527156067,
      "loss": 5.0409,
      "mean_token_accuracy": 0.22330638766288757,
      "num_tokens": 105209030.0,
      "step": 45940
    },
    {
      "entropy": 5.771899557113647,
      "epoch": 4.541814946619217,
      "grad_norm": 1.296875,
      "learning_rate": 0.00031105841930363686,
      "loss": 4.8418,
      "mean_token_accuracy": 0.24034301191568375,
      "num_tokens": 105220167.0,
      "step": 45945
    },
    {
      "entropy": 5.745329999923706,
      "epoch": 4.5423092131277185,
      "grad_norm": 1.359375,
      "learning_rate": 0.00031102358500462434,
      "loss": 4.8841,
      "mean_token_accuracy": 0.23243952691555023,
      "num_tokens": 105231646.0,
      "step": 45950
    },
    {
      "entropy": 5.788830900192261,
      "epoch": 4.54280347963622,
      "grad_norm": 1.40625,
      "learning_rate": 0.000310988749819426,
      "loss": 4.9592,
      "mean_token_accuracy": 0.22363257259130478,
      "num_tokens": 105243822.0,
      "step": 45955
    },
    {
      "entropy": 5.832404041290284,
      "epoch": 4.543297746144721,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00031095391374889885,
      "loss": 4.974,
      "mean_token_accuracy": 0.22787881642580032,
      "num_tokens": 105256267.0,
      "step": 45960
    },
    {
      "entropy": 5.805338191986084,
      "epoch": 4.543792012653222,
      "grad_norm": 1.3125,
      "learning_rate": 0.00031091907679389985,
      "loss": 4.9642,
      "mean_token_accuracy": 0.23044847697019577,
      "num_tokens": 105267748.0,
      "step": 45965
    },
    {
      "entropy": 5.815966796875,
      "epoch": 4.5442862791617245,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000310884238955286,
      "loss": 4.9786,
      "mean_token_accuracy": 0.22754546105861664,
      "num_tokens": 105278696.0,
      "step": 45970
    },
    {
      "entropy": 5.755666780471802,
      "epoch": 4.544780545670226,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003108494002339143,
      "loss": 4.8618,
      "mean_token_accuracy": 0.2380600929260254,
      "num_tokens": 105290053.0,
      "step": 45975
    },
    {
      "entropy": 5.828767013549805,
      "epoch": 4.545274812178727,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003108145606306417,
      "loss": 4.9531,
      "mean_token_accuracy": 0.23300205767154694,
      "num_tokens": 105301196.0,
      "step": 45980
    },
    {
      "entropy": 5.7889210224151615,
      "epoch": 4.545769078687228,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00031077972014632545,
      "loss": 4.9834,
      "mean_token_accuracy": 0.22678053826093675,
      "num_tokens": 105312799.0,
      "step": 45985
    },
    {
      "entropy": 5.866717147827148,
      "epoch": 4.5462633451957295,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003107448787818225,
      "loss": 5.0084,
      "mean_token_accuracy": 0.21666594892740249,
      "num_tokens": 105324818.0,
      "step": 45990
    },
    {
      "entropy": 5.7314784049987795,
      "epoch": 4.546757611704231,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00031071003653799,
      "loss": 4.8994,
      "mean_token_accuracy": 0.23617913872003554,
      "num_tokens": 105335916.0,
      "step": 45995
    },
    {
      "entropy": 5.774842357635498,
      "epoch": 4.547251878212732,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00031067519341568506,
      "loss": 4.8853,
      "mean_token_accuracy": 0.23889122307300567,
      "num_tokens": 105347257.0,
      "step": 46000
    },
    {
      "entropy": 5.843125534057617,
      "epoch": 4.547746144721234,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003106403494157649,
      "loss": 5.034,
      "mean_token_accuracy": 0.21733746230602263,
      "num_tokens": 105359285.0,
      "step": 46005
    },
    {
      "entropy": 5.830749273300171,
      "epoch": 4.548240411229735,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003106055045390865,
      "loss": 5.0512,
      "mean_token_accuracy": 0.22584494203329086,
      "num_tokens": 105372809.0,
      "step": 46010
    },
    {
      "entropy": 5.758221387863159,
      "epoch": 4.548734677738237,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00031057065878650715,
      "loss": 4.8403,
      "mean_token_accuracy": 0.23986903876066207,
      "num_tokens": 105384341.0,
      "step": 46015
    },
    {
      "entropy": 5.792311716079712,
      "epoch": 4.549228944246738,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00031053581215888417,
      "loss": 4.9135,
      "mean_token_accuracy": 0.23019323945045472,
      "num_tokens": 105397222.0,
      "step": 46020
    },
    {
      "entropy": 5.771563529968262,
      "epoch": 4.549723210755239,
      "grad_norm": 1.421875,
      "learning_rate": 0.00031050096465707466,
      "loss": 4.8649,
      "mean_token_accuracy": 0.2273395150899887,
      "num_tokens": 105407924.0,
      "step": 46025
    },
    {
      "entropy": 5.741381216049194,
      "epoch": 4.5502174772637405,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003104661162819359,
      "loss": 4.9629,
      "mean_token_accuracy": 0.23127864599227904,
      "num_tokens": 105419390.0,
      "step": 46030
    },
    {
      "entropy": 5.802460527420044,
      "epoch": 4.550711743772242,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003104312670343252,
      "loss": 4.9171,
      "mean_token_accuracy": 0.23314529061317443,
      "num_tokens": 105430963.0,
      "step": 46035
    },
    {
      "entropy": 5.742901039123535,
      "epoch": 4.551206010280743,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00031039641691509983,
      "loss": 4.8804,
      "mean_token_accuracy": 0.23332030475139617,
      "num_tokens": 105443176.0,
      "step": 46040
    },
    {
      "entropy": 5.81176643371582,
      "epoch": 4.551700276789244,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003103615659251171,
      "loss": 4.9025,
      "mean_token_accuracy": 0.23529370427131652,
      "num_tokens": 105454523.0,
      "step": 46045
    },
    {
      "entropy": 5.772581434249878,
      "epoch": 4.552194543297746,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003103267140652344,
      "loss": 4.9296,
      "mean_token_accuracy": 0.23143180757761,
      "num_tokens": 105466523.0,
      "step": 46050
    },
    {
      "entropy": 5.722124624252319,
      "epoch": 4.552688809806248,
      "grad_norm": 1.390625,
      "learning_rate": 0.000310291861336309,
      "loss": 4.8932,
      "mean_token_accuracy": 0.23618594706058502,
      "num_tokens": 105478799.0,
      "step": 46055
    },
    {
      "entropy": 5.798241043090821,
      "epoch": 4.553183076314749,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00031025700773919837,
      "loss": 4.9618,
      "mean_token_accuracy": 0.22804735749959945,
      "num_tokens": 105491088.0,
      "step": 46060
    },
    {
      "entropy": 5.833500146865845,
      "epoch": 4.55367734282325,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003102221532747598,
      "loss": 4.9212,
      "mean_token_accuracy": 0.2349862903356552,
      "num_tokens": 105502258.0,
      "step": 46065
    },
    {
      "entropy": 5.782088708877564,
      "epoch": 4.5541716093317515,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003101872979438509,
      "loss": 4.9453,
      "mean_token_accuracy": 0.22302112430334092,
      "num_tokens": 105514414.0,
      "step": 46070
    },
    {
      "entropy": 5.768762111663818,
      "epoch": 4.554665875840253,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000310152441747329,
      "loss": 4.8752,
      "mean_token_accuracy": 0.23776285648345946,
      "num_tokens": 105526023.0,
      "step": 46075
    },
    {
      "entropy": 5.761282539367675,
      "epoch": 4.555160142348754,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003101175846860515,
      "loss": 4.9404,
      "mean_token_accuracy": 0.22767429649829865,
      "num_tokens": 105538593.0,
      "step": 46080
    },
    {
      "entropy": 5.781795835494995,
      "epoch": 4.555654408857256,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003100827267608761,
      "loss": 4.8882,
      "mean_token_accuracy": 0.23408602625131608,
      "num_tokens": 105549854.0,
      "step": 46085
    },
    {
      "entropy": 5.773876428604126,
      "epoch": 4.556148675365757,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00031004786797266004,
      "loss": 4.9651,
      "mean_token_accuracy": 0.22132352739572525,
      "num_tokens": 105561778.0,
      "step": 46090
    },
    {
      "entropy": 5.793002891540527,
      "epoch": 4.556642941874259,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00031001300832226105,
      "loss": 4.9957,
      "mean_token_accuracy": 0.2241692900657654,
      "num_tokens": 105573829.0,
      "step": 46095
    },
    {
      "entropy": 5.742995309829712,
      "epoch": 4.55713720838276,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00030997814781053663,
      "loss": 4.8984,
      "mean_token_accuracy": 0.23615624755620956,
      "num_tokens": 105585456.0,
      "step": 46100
    },
    {
      "entropy": 5.834888553619384,
      "epoch": 4.557631474891261,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00030994328643834437,
      "loss": 4.9534,
      "mean_token_accuracy": 0.2263108894228935,
      "num_tokens": 105596926.0,
      "step": 46105
    },
    {
      "entropy": 5.874442672729492,
      "epoch": 4.5581257413997625,
      "grad_norm": 1.4375,
      "learning_rate": 0.00030990842420654176,
      "loss": 5.0391,
      "mean_token_accuracy": 0.21975065171718597,
      "num_tokens": 105608316.0,
      "step": 46110
    },
    {
      "entropy": 5.758781671524048,
      "epoch": 4.558620007908264,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003098735611159866,
      "loss": 4.9183,
      "mean_token_accuracy": 0.23203592300415038,
      "num_tokens": 105618921.0,
      "step": 46115
    },
    {
      "entropy": 5.783056592941284,
      "epoch": 4.559114274416766,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003098386971675364,
      "loss": 5.0214,
      "mean_token_accuracy": 0.21841016858816148,
      "num_tokens": 105629981.0,
      "step": 46120
    },
    {
      "entropy": 5.8702027797698975,
      "epoch": 4.559608540925267,
      "grad_norm": 1.359375,
      "learning_rate": 0.00030980383236204885,
      "loss": 4.9979,
      "mean_token_accuracy": 0.22250534296035768,
      "num_tokens": 105640975.0,
      "step": 46125
    },
    {
      "entropy": 5.7482696056365965,
      "epoch": 4.560102807433768,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00030976896670038156,
      "loss": 4.8916,
      "mean_token_accuracy": 0.23418245762586593,
      "num_tokens": 105652075.0,
      "step": 46130
    },
    {
      "entropy": 5.846658182144165,
      "epoch": 4.56059707394227,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00030973410018339233,
      "loss": 5.0241,
      "mean_token_accuracy": 0.21871226131916047,
      "num_tokens": 105663016.0,
      "step": 46135
    },
    {
      "entropy": 5.738672733306885,
      "epoch": 4.561091340450771,
      "grad_norm": 1.328125,
      "learning_rate": 0.00030969923281193887,
      "loss": 4.8968,
      "mean_token_accuracy": 0.22797385156154631,
      "num_tokens": 105674032.0,
      "step": 46140
    },
    {
      "entropy": 5.8253308773040775,
      "epoch": 4.561585606959272,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00030966436458687883,
      "loss": 4.9525,
      "mean_token_accuracy": 0.2279127135872841,
      "num_tokens": 105685411.0,
      "step": 46145
    },
    {
      "entropy": 5.764629936218261,
      "epoch": 4.562079873467773,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00030962949550907015,
      "loss": 4.8691,
      "mean_token_accuracy": 0.23304575383663179,
      "num_tokens": 105695898.0,
      "step": 46150
    },
    {
      "entropy": 5.774698877334595,
      "epoch": 4.562574139976276,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003095946255793705,
      "loss": 4.9636,
      "mean_token_accuracy": 0.22783783227205276,
      "num_tokens": 105706889.0,
      "step": 46155
    },
    {
      "entropy": 5.777381038665771,
      "epoch": 4.563068406484777,
      "grad_norm": 1.25,
      "learning_rate": 0.0003095597547986376,
      "loss": 4.9408,
      "mean_token_accuracy": 0.22896846383810043,
      "num_tokens": 105718700.0,
      "step": 46160
    },
    {
      "entropy": 5.7424046993255615,
      "epoch": 4.563562672993278,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003095248831677294,
      "loss": 4.82,
      "mean_token_accuracy": 0.2347561925649643,
      "num_tokens": 105729790.0,
      "step": 46165
    },
    {
      "entropy": 5.7797459125518795,
      "epoch": 4.564056939501779,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003094900106875037,
      "loss": 4.9192,
      "mean_token_accuracy": 0.2315300539135933,
      "num_tokens": 105740370.0,
      "step": 46170
    },
    {
      "entropy": 5.8432698249816895,
      "epoch": 4.564551206010281,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003094551373588185,
      "loss": 4.8982,
      "mean_token_accuracy": 0.22627813667058944,
      "num_tokens": 105749992.0,
      "step": 46175
    },
    {
      "entropy": 5.760067987442016,
      "epoch": 4.565045472518782,
      "grad_norm": 1.359375,
      "learning_rate": 0.00030942026318253146,
      "loss": 4.8802,
      "mean_token_accuracy": 0.23362171947956084,
      "num_tokens": 105760491.0,
      "step": 46180
    },
    {
      "entropy": 5.667845344543457,
      "epoch": 4.565539739027283,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003093853881595006,
      "loss": 4.8756,
      "mean_token_accuracy": 0.24643746465444566,
      "num_tokens": 105772307.0,
      "step": 46185
    },
    {
      "entropy": 5.778562831878662,
      "epoch": 4.566034005535785,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000309350512290584,
      "loss": 4.9558,
      "mean_token_accuracy": 0.22749521285295488,
      "num_tokens": 105784108.0,
      "step": 46190
    },
    {
      "entropy": 5.7500465393066404,
      "epoch": 4.566528272044287,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00030931563557663934,
      "loss": 4.9404,
      "mean_token_accuracy": 0.2314898043870926,
      "num_tokens": 105795833.0,
      "step": 46195
    },
    {
      "entropy": 5.848878335952759,
      "epoch": 4.567022538552788,
      "grad_norm": 1.25,
      "learning_rate": 0.00030928075801852473,
      "loss": 4.963,
      "mean_token_accuracy": 0.23106114715337753,
      "num_tokens": 105807170.0,
      "step": 46200
    },
    {
      "entropy": 5.86488184928894,
      "epoch": 4.567516805061289,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003092458796170982,
      "loss": 4.9657,
      "mean_token_accuracy": 0.22627424448728561,
      "num_tokens": 105818620.0,
      "step": 46205
    },
    {
      "entropy": 5.7761985778808596,
      "epoch": 4.56801107156979,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00030921100037321767,
      "loss": 4.8719,
      "mean_token_accuracy": 0.23799406588077546,
      "num_tokens": 105830872.0,
      "step": 46210
    },
    {
      "entropy": 5.694473457336426,
      "epoch": 4.568505338078292,
      "grad_norm": 1.390625,
      "learning_rate": 0.00030917612028774123,
      "loss": 4.7964,
      "mean_token_accuracy": 0.24616227596998214,
      "num_tokens": 105840132.0,
      "step": 46215
    },
    {
      "entropy": 5.751914358139038,
      "epoch": 4.568999604586793,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000309141239361527,
      "loss": 4.9238,
      "mean_token_accuracy": 0.23768649101257325,
      "num_tokens": 105851949.0,
      "step": 46220
    },
    {
      "entropy": 5.7951353073120115,
      "epoch": 4.569493871095295,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003091063575954329,
      "loss": 4.903,
      "mean_token_accuracy": 0.23163083344697952,
      "num_tokens": 105863340.0,
      "step": 46225
    },
    {
      "entropy": 5.82391324043274,
      "epoch": 4.569988137603796,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003090714749903171,
      "loss": 5.0361,
      "mean_token_accuracy": 0.2240614339709282,
      "num_tokens": 105875158.0,
      "step": 46230
    },
    {
      "entropy": 5.74729905128479,
      "epoch": 4.570482404112298,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00030903659154703786,
      "loss": 4.9588,
      "mean_token_accuracy": 0.23066673129796983,
      "num_tokens": 105886369.0,
      "step": 46235
    },
    {
      "entropy": 5.799321079254151,
      "epoch": 4.570976670620799,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000309001707266453,
      "loss": 4.9152,
      "mean_token_accuracy": 0.23584138751029968,
      "num_tokens": 105897341.0,
      "step": 46240
    },
    {
      "entropy": 5.762801313400269,
      "epoch": 4.5714709371293,
      "grad_norm": 1.5,
      "learning_rate": 0.00030896682214942106,
      "loss": 4.9257,
      "mean_token_accuracy": 0.22668698728084563,
      "num_tokens": 105909132.0,
      "step": 46245
    },
    {
      "entropy": 5.853288555145264,
      "epoch": 4.571965203637801,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00030893193619679985,
      "loss": 4.9597,
      "mean_token_accuracy": 0.22201996445655822,
      "num_tokens": 105919916.0,
      "step": 46250
    },
    {
      "entropy": 5.810869121551514,
      "epoch": 4.572459470146303,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003088970494094479,
      "loss": 4.9534,
      "mean_token_accuracy": 0.22364890575408936,
      "num_tokens": 105930104.0,
      "step": 46255
    },
    {
      "entropy": 5.785264110565185,
      "epoch": 4.572953736654805,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003088621617882232,
      "loss": 4.8927,
      "mean_token_accuracy": 0.22587590366601945,
      "num_tokens": 105942382.0,
      "step": 46260
    },
    {
      "entropy": 5.763106775283814,
      "epoch": 4.573448003163306,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003088272733339841,
      "loss": 4.9324,
      "mean_token_accuracy": 0.22870266437530518,
      "num_tokens": 105954878.0,
      "step": 46265
    },
    {
      "entropy": 5.728130435943603,
      "epoch": 4.573942269671807,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003087923840475888,
      "loss": 4.8523,
      "mean_token_accuracy": 0.23797095268964769,
      "num_tokens": 105965277.0,
      "step": 46270
    },
    {
      "entropy": 5.822075843811035,
      "epoch": 4.5744365361803085,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00030875749392989557,
      "loss": 4.985,
      "mean_token_accuracy": 0.22939426004886626,
      "num_tokens": 105977660.0,
      "step": 46275
    },
    {
      "entropy": 5.730556631088257,
      "epoch": 4.57493080268881,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003087226029817628,
      "loss": 4.9276,
      "mean_token_accuracy": 0.2355106756091118,
      "num_tokens": 105989401.0,
      "step": 46280
    },
    {
      "entropy": 5.810043239593506,
      "epoch": 4.575425069197311,
      "grad_norm": 1.25,
      "learning_rate": 0.0003086877112040487,
      "loss": 4.9331,
      "mean_token_accuracy": 0.22445629686117172,
      "num_tokens": 105999801.0,
      "step": 46285
    },
    {
      "entropy": 5.763574075698853,
      "epoch": 4.575919335705812,
      "grad_norm": 1.125,
      "learning_rate": 0.0003086528185976118,
      "loss": 4.9301,
      "mean_token_accuracy": 0.22380468398332595,
      "num_tokens": 106011263.0,
      "step": 46290
    },
    {
      "entropy": 5.779327058792115,
      "epoch": 4.576413602214314,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003086179251633103,
      "loss": 4.953,
      "mean_token_accuracy": 0.2292763113975525,
      "num_tokens": 106023166.0,
      "step": 46295
    },
    {
      "entropy": 5.785010528564453,
      "epoch": 4.576907868722815,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00030858303090200255,
      "loss": 4.9852,
      "mean_token_accuracy": 0.22796581089496612,
      "num_tokens": 106036143.0,
      "step": 46300
    },
    {
      "entropy": 5.743170499801636,
      "epoch": 4.577402135231317,
      "grad_norm": 1.4375,
      "learning_rate": 0.00030854813581454707,
      "loss": 4.8557,
      "mean_token_accuracy": 0.2360699236392975,
      "num_tokens": 106047160.0,
      "step": 46305
    },
    {
      "entropy": 5.8159160137176515,
      "epoch": 4.577896401739818,
      "grad_norm": 1.125,
      "learning_rate": 0.0003085132399018023,
      "loss": 4.9303,
      "mean_token_accuracy": 0.23296049237251282,
      "num_tokens": 106059384.0,
      "step": 46310
    },
    {
      "entropy": 5.794270706176758,
      "epoch": 4.5783906682483195,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003084783431646266,
      "loss": 4.9491,
      "mean_token_accuracy": 0.22866792380809783,
      "num_tokens": 106069859.0,
      "step": 46315
    },
    {
      "entropy": 5.749280166625977,
      "epoch": 4.578884934756821,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00030844344560387844,
      "loss": 4.9157,
      "mean_token_accuracy": 0.22616037279367446,
      "num_tokens": 106081292.0,
      "step": 46320
    },
    {
      "entropy": 5.782117176055908,
      "epoch": 4.579379201265322,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003084085472204163,
      "loss": 4.9008,
      "mean_token_accuracy": 0.230620676279068,
      "num_tokens": 106093236.0,
      "step": 46325
    },
    {
      "entropy": 5.788677453994751,
      "epoch": 4.579873467773823,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00030837364801509874,
      "loss": 4.9301,
      "mean_token_accuracy": 0.23032985031604766,
      "num_tokens": 106104386.0,
      "step": 46330
    },
    {
      "entropy": 5.789829254150391,
      "epoch": 4.580367734282325,
      "grad_norm": 1.421875,
      "learning_rate": 0.00030833874798878437,
      "loss": 4.8943,
      "mean_token_accuracy": 0.23631369471549987,
      "num_tokens": 106114812.0,
      "step": 46335
    },
    {
      "entropy": 5.770914554595947,
      "epoch": 4.580862000790827,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00030830384714233155,
      "loss": 4.8823,
      "mean_token_accuracy": 0.23000660240650178,
      "num_tokens": 106125434.0,
      "step": 46340
    },
    {
      "entropy": 5.782076740264893,
      "epoch": 4.581356267299328,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00030826894547659876,
      "loss": 4.9645,
      "mean_token_accuracy": 0.22369841635227203,
      "num_tokens": 106138218.0,
      "step": 46345
    },
    {
      "entropy": 5.764967632293701,
      "epoch": 4.581850533807829,
      "grad_norm": 1.359375,
      "learning_rate": 0.00030823404299244495,
      "loss": 4.8582,
      "mean_token_accuracy": 0.23870978951454164,
      "num_tokens": 106149408.0,
      "step": 46350
    },
    {
      "entropy": 5.861570167541504,
      "epoch": 4.5823448003163305,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003081991396907285,
      "loss": 5.0841,
      "mean_token_accuracy": 0.21974919885396957,
      "num_tokens": 106162589.0,
      "step": 46355
    },
    {
      "entropy": 5.789119863510132,
      "epoch": 4.582839066824832,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00030816423557230807,
      "loss": 4.9921,
      "mean_token_accuracy": 0.23710990697145462,
      "num_tokens": 106175539.0,
      "step": 46360
    },
    {
      "entropy": 5.829701662063599,
      "epoch": 4.583333333333333,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00030812933063804226,
      "loss": 4.9771,
      "mean_token_accuracy": 0.2270043596625328,
      "num_tokens": 106187545.0,
      "step": 46365
    },
    {
      "entropy": 5.8006213188171385,
      "epoch": 4.583827599841834,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00030809442488878976,
      "loss": 4.9125,
      "mean_token_accuracy": 0.22982307225465776,
      "num_tokens": 106199512.0,
      "step": 46370
    },
    {
      "entropy": 5.78091721534729,
      "epoch": 4.5843218663503364,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003080595183254093,
      "loss": 4.9568,
      "mean_token_accuracy": 0.22119287997484208,
      "num_tokens": 106211185.0,
      "step": 46375
    },
    {
      "entropy": 5.8040971755981445,
      "epoch": 4.584816132858838,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003080246109487596,
      "loss": 4.9548,
      "mean_token_accuracy": 0.22680919021368026,
      "num_tokens": 106222604.0,
      "step": 46380
    },
    {
      "entropy": 5.8519572734832765,
      "epoch": 4.585310399367339,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003079897027596992,
      "loss": 4.9919,
      "mean_token_accuracy": 0.224563167989254,
      "num_tokens": 106236134.0,
      "step": 46385
    },
    {
      "entropy": 5.795613384246826,
      "epoch": 4.58580466587584,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003079547937590872,
      "loss": 4.9356,
      "mean_token_accuracy": 0.2309411346912384,
      "num_tokens": 106247553.0,
      "step": 46390
    },
    {
      "entropy": 5.745602607727051,
      "epoch": 4.5862989323843415,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000307919883947782,
      "loss": 4.8643,
      "mean_token_accuracy": 0.24145733416080475,
      "num_tokens": 106258356.0,
      "step": 46395
    },
    {
      "entropy": 5.831272745132447,
      "epoch": 4.586793198892843,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003078849733266426,
      "loss": 4.9914,
      "mean_token_accuracy": 0.22067010253667832,
      "num_tokens": 106269428.0,
      "step": 46400
    },
    {
      "entropy": 5.742138338088989,
      "epoch": 4.587287465401344,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003078500618965278,
      "loss": 4.8299,
      "mean_token_accuracy": 0.2401998907327652,
      "num_tokens": 106280600.0,
      "step": 46405
    },
    {
      "entropy": 5.820037937164306,
      "epoch": 4.587781731909846,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00030781514965829636,
      "loss": 4.9104,
      "mean_token_accuracy": 0.22838593125343323,
      "num_tokens": 106291720.0,
      "step": 46410
    },
    {
      "entropy": 5.746521997451782,
      "epoch": 4.588275998418347,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003077802366128071,
      "loss": 4.9372,
      "mean_token_accuracy": 0.2251282438635826,
      "num_tokens": 106303156.0,
      "step": 46415
    },
    {
      "entropy": 5.755290126800537,
      "epoch": 4.588770264926849,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000307745322760919,
      "loss": 4.9195,
      "mean_token_accuracy": 0.23144638389348984,
      "num_tokens": 106314671.0,
      "step": 46420
    },
    {
      "entropy": 5.774840068817139,
      "epoch": 4.58926453143535,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003077104081034909,
      "loss": 4.9047,
      "mean_token_accuracy": 0.23309523165225982,
      "num_tokens": 106327246.0,
      "step": 46425
    },
    {
      "entropy": 5.8008284091949465,
      "epoch": 4.589758797943851,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00030767549264138156,
      "loss": 4.8775,
      "mean_token_accuracy": 0.24055134057998656,
      "num_tokens": 106337632.0,
      "step": 46430
    },
    {
      "entropy": 5.806554985046387,
      "epoch": 4.5902530644523525,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003076405763754501,
      "loss": 4.9799,
      "mean_token_accuracy": 0.22656079083681108,
      "num_tokens": 106348885.0,
      "step": 46435
    },
    {
      "entropy": 5.802888584136963,
      "epoch": 4.590747330960854,
      "grad_norm": 1.375,
      "learning_rate": 0.00030760565930655544,
      "loss": 4.9178,
      "mean_token_accuracy": 0.23548796772956848,
      "num_tokens": 106359369.0,
      "step": 46440
    },
    {
      "entropy": 5.723750925064087,
      "epoch": 4.591241597469356,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003075707414355564,
      "loss": 4.8823,
      "mean_token_accuracy": 0.23327623903751374,
      "num_tokens": 106371198.0,
      "step": 46445
    },
    {
      "entropy": 5.836214065551758,
      "epoch": 4.591735863977857,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003075358227633122,
      "loss": 4.9871,
      "mean_token_accuracy": 0.22521165758371353,
      "num_tokens": 106382874.0,
      "step": 46450
    },
    {
      "entropy": 5.871688270568848,
      "epoch": 4.592230130486358,
      "grad_norm": 1.234375,
      "learning_rate": 0.00030750090329068167,
      "loss": 5.0364,
      "mean_token_accuracy": 0.2219303399324417,
      "num_tokens": 106394329.0,
      "step": 46455
    },
    {
      "entropy": 5.748979473114014,
      "epoch": 4.59272439699486,
      "grad_norm": 1.125,
      "learning_rate": 0.0003074659830185238,
      "loss": 4.919,
      "mean_token_accuracy": 0.2393765166401863,
      "num_tokens": 106407223.0,
      "step": 46460
    },
    {
      "entropy": 5.781836223602295,
      "epoch": 4.593218663503361,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003074310619476978,
      "loss": 4.9704,
      "mean_token_accuracy": 0.23156144618988037,
      "num_tokens": 106418622.0,
      "step": 46465
    },
    {
      "entropy": 5.821830177307129,
      "epoch": 4.593712930011862,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003073961400790626,
      "loss": 4.955,
      "mean_token_accuracy": 0.2286625936627388,
      "num_tokens": 106430575.0,
      "step": 46470
    },
    {
      "entropy": 5.762187337875366,
      "epoch": 4.5942071965203635,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003073612174134773,
      "loss": 4.894,
      "mean_token_accuracy": 0.22554826289415358,
      "num_tokens": 106441707.0,
      "step": 46475
    },
    {
      "entropy": 5.8026206493377686,
      "epoch": 4.594701463028866,
      "grad_norm": 1.25,
      "learning_rate": 0.0003073262939518011,
      "loss": 4.9582,
      "mean_token_accuracy": 0.22558055818080902,
      "num_tokens": 106453470.0,
      "step": 46480
    },
    {
      "entropy": 5.744352245330811,
      "epoch": 4.595195729537367,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000307291369694893,
      "loss": 4.9067,
      "mean_token_accuracy": 0.23441220372915267,
      "num_tokens": 106464140.0,
      "step": 46485
    },
    {
      "entropy": 5.783983373641968,
      "epoch": 4.595689996045868,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003072564446436122,
      "loss": 4.8943,
      "mean_token_accuracy": 0.2320398911833763,
      "num_tokens": 106476052.0,
      "step": 46490
    },
    {
      "entropy": 5.713870573043823,
      "epoch": 4.596184262554369,
      "grad_norm": 1.359375,
      "learning_rate": 0.00030722151879881786,
      "loss": 4.8435,
      "mean_token_accuracy": 0.2404267519712448,
      "num_tokens": 106486568.0,
      "step": 46495
    },
    {
      "entropy": 5.702104806900024,
      "epoch": 4.596678529062871,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00030718659216136924,
      "loss": 4.9011,
      "mean_token_accuracy": 0.23873864710330964,
      "num_tokens": 106498770.0,
      "step": 46500
    },
    {
      "entropy": 5.797584676742554,
      "epoch": 4.597172795571372,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003071516647321254,
      "loss": 4.9424,
      "mean_token_accuracy": 0.22179621011018752,
      "num_tokens": 106510384.0,
      "step": 46505
    },
    {
      "entropy": 5.729933166503907,
      "epoch": 4.597667062079873,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003071167365119455,
      "loss": 4.8791,
      "mean_token_accuracy": 0.23266719430685043,
      "num_tokens": 106522197.0,
      "step": 46510
    },
    {
      "entropy": 5.785855388641357,
      "epoch": 4.598161328588375,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003070818075016891,
      "loss": 4.9219,
      "mean_token_accuracy": 0.22925770431756973,
      "num_tokens": 106533218.0,
      "step": 46515
    },
    {
      "entropy": 5.780378246307373,
      "epoch": 4.598655595096877,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003070468777022151,
      "loss": 4.8978,
      "mean_token_accuracy": 0.23363303989171982,
      "num_tokens": 106545391.0,
      "step": 46520
    },
    {
      "entropy": 5.826485300064087,
      "epoch": 4.599149861605378,
      "grad_norm": 1.296875,
      "learning_rate": 0.00030701194711438306,
      "loss": 4.9556,
      "mean_token_accuracy": 0.22816691994667054,
      "num_tokens": 106556122.0,
      "step": 46525
    },
    {
      "entropy": 5.7969639778137205,
      "epoch": 4.599644128113879,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00030697701573905204,
      "loss": 4.9147,
      "mean_token_accuracy": 0.2263759881258011,
      "num_tokens": 106567876.0,
      "step": 46530
    },
    {
      "entropy": 5.704545974731445,
      "epoch": 4.60013839462238,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003069420835770816,
      "loss": 4.8612,
      "mean_token_accuracy": 0.2347429007291794,
      "num_tokens": 106578083.0,
      "step": 46535
    },
    {
      "entropy": 5.764178466796875,
      "epoch": 4.600632661130882,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003069071506293308,
      "loss": 4.9773,
      "mean_token_accuracy": 0.2270976036787033,
      "num_tokens": 106589986.0,
      "step": 46540
    },
    {
      "entropy": 5.843314170837402,
      "epoch": 4.601126927639383,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003068722168966593,
      "loss": 4.9726,
      "mean_token_accuracy": 0.22991180568933486,
      "num_tokens": 106602055.0,
      "step": 46545
    },
    {
      "entropy": 5.869006538391114,
      "epoch": 4.601621194147884,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00030683728237992626,
      "loss": 4.8934,
      "mean_token_accuracy": 0.2325399711728096,
      "num_tokens": 106613394.0,
      "step": 46550
    },
    {
      "entropy": 5.806365966796875,
      "epoch": 4.602115460656385,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00030680234707999117,
      "loss": 4.8175,
      "mean_token_accuracy": 0.24214028269052507,
      "num_tokens": 106624190.0,
      "step": 46555
    },
    {
      "entropy": 5.742762613296509,
      "epoch": 4.602609727164888,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003067674109977133,
      "loss": 4.9174,
      "mean_token_accuracy": 0.22881741672754288,
      "num_tokens": 106636030.0,
      "step": 46560
    },
    {
      "entropy": 5.828702926635742,
      "epoch": 4.603103993673389,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003067324741339523,
      "loss": 4.9943,
      "mean_token_accuracy": 0.22057678401470185,
      "num_tokens": 106648648.0,
      "step": 46565
    },
    {
      "entropy": 5.726020860671997,
      "epoch": 4.60359826018189,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00030669753648956743,
      "loss": 4.8756,
      "mean_token_accuracy": 0.23147964477539062,
      "num_tokens": 106659902.0,
      "step": 46570
    },
    {
      "entropy": 5.821587562561035,
      "epoch": 4.604092526690391,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00030666259806541823,
      "loss": 4.9719,
      "mean_token_accuracy": 0.22490687817335128,
      "num_tokens": 106670768.0,
      "step": 46575
    },
    {
      "entropy": 5.763034534454346,
      "epoch": 4.604586793198893,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003066276588623643,
      "loss": 4.9066,
      "mean_token_accuracy": 0.22534407526254654,
      "num_tokens": 106682331.0,
      "step": 46580
    },
    {
      "entropy": 5.814328193664551,
      "epoch": 4.605081059707394,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000306592718881265,
      "loss": 5.0036,
      "mean_token_accuracy": 0.22627761960029602,
      "num_tokens": 106692562.0,
      "step": 46585
    },
    {
      "entropy": 5.741906356811524,
      "epoch": 4.605575326215895,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003065577781229799,
      "loss": 4.881,
      "mean_token_accuracy": 0.23673069179058076,
      "num_tokens": 106705208.0,
      "step": 46590
    },
    {
      "entropy": 5.890418243408203,
      "epoch": 4.606069592724397,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003065228365883686,
      "loss": 5.0577,
      "mean_token_accuracy": 0.22078993171453476,
      "num_tokens": 106717875.0,
      "step": 46595
    },
    {
      "entropy": 5.835211658477784,
      "epoch": 4.606563859232899,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00030648789427829067,
      "loss": 4.9417,
      "mean_token_accuracy": 0.22563556730747222,
      "num_tokens": 106729237.0,
      "step": 46600
    },
    {
      "entropy": 5.794555425643921,
      "epoch": 4.6070581257414,
      "grad_norm": 1.390625,
      "learning_rate": 0.00030645295119360557,
      "loss": 5.0049,
      "mean_token_accuracy": 0.2262166440486908,
      "num_tokens": 106740805.0,
      "step": 46605
    },
    {
      "entropy": 5.729637384414673,
      "epoch": 4.607552392249901,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000306418007335173,
      "loss": 4.907,
      "mean_token_accuracy": 0.23758375346660615,
      "num_tokens": 106752069.0,
      "step": 46610
    },
    {
      "entropy": 5.923226022720337,
      "epoch": 4.608046658758402,
      "grad_norm": 1.25,
      "learning_rate": 0.00030638306270385266,
      "loss": 5.0759,
      "mean_token_accuracy": 0.21541512459516526,
      "num_tokens": 106764885.0,
      "step": 46615
    },
    {
      "entropy": 5.797978734970092,
      "epoch": 4.608540925266904,
      "grad_norm": 1.390625,
      "learning_rate": 0.000306348117300504,
      "loss": 4.9559,
      "mean_token_accuracy": 0.22912544012069702,
      "num_tokens": 106775712.0,
      "step": 46620
    },
    {
      "entropy": 5.796900510787964,
      "epoch": 4.609035191775405,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003063131711259868,
      "loss": 4.9206,
      "mean_token_accuracy": 0.23948630094528198,
      "num_tokens": 106788068.0,
      "step": 46625
    },
    {
      "entropy": 5.759616470336914,
      "epoch": 4.609529458283907,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00030627822418116083,
      "loss": 4.9147,
      "mean_token_accuracy": 0.2280917078256607,
      "num_tokens": 106798984.0,
      "step": 46630
    },
    {
      "entropy": 5.835804414749146,
      "epoch": 4.610023724792408,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003062432764668856,
      "loss": 4.954,
      "mean_token_accuracy": 0.22887501269578933,
      "num_tokens": 106811447.0,
      "step": 46635
    },
    {
      "entropy": 5.748229551315307,
      "epoch": 4.61051799130091,
      "grad_norm": 1.296875,
      "learning_rate": 0.000306208327984021,
      "loss": 4.9854,
      "mean_token_accuracy": 0.22281210571527482,
      "num_tokens": 106823622.0,
      "step": 46640
    },
    {
      "entropy": 5.79205322265625,
      "epoch": 4.611012257809411,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00030617337873342664,
      "loss": 5.0006,
      "mean_token_accuracy": 0.22198145240545272,
      "num_tokens": 106836767.0,
      "step": 46645
    },
    {
      "entropy": 5.862561655044556,
      "epoch": 4.611506524317912,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003061384287159623,
      "loss": 5.0463,
      "mean_token_accuracy": 0.22319204062223436,
      "num_tokens": 106849027.0,
      "step": 46650
    },
    {
      "entropy": 5.76710090637207,
      "epoch": 4.612000790826413,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003061034779324878,
      "loss": 4.8934,
      "mean_token_accuracy": 0.23429109901189804,
      "num_tokens": 106859543.0,
      "step": 46655
    },
    {
      "entropy": 5.783667707443238,
      "epoch": 4.612495057334915,
      "grad_norm": 1.421875,
      "learning_rate": 0.00030606852638386304,
      "loss": 4.8737,
      "mean_token_accuracy": 0.23549743592739106,
      "num_tokens": 106870300.0,
      "step": 46660
    },
    {
      "entropy": 5.7949793338775635,
      "epoch": 4.612989323843417,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0003060335740709475,
      "loss": 4.9021,
      "mean_token_accuracy": 0.23392112404108048,
      "num_tokens": 106880706.0,
      "step": 46665
    },
    {
      "entropy": 5.781295442581177,
      "epoch": 4.613483590351918,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003059986209946014,
      "loss": 5.0002,
      "mean_token_accuracy": 0.22184660136699677,
      "num_tokens": 106890906.0,
      "step": 46670
    },
    {
      "entropy": 5.79620213508606,
      "epoch": 4.613977856860419,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003059636671556843,
      "loss": 4.9573,
      "mean_token_accuracy": 0.22817874997854232,
      "num_tokens": 106902175.0,
      "step": 46675
    },
    {
      "entropy": 5.815825796127319,
      "epoch": 4.6144721233689205,
      "grad_norm": 1.359375,
      "learning_rate": 0.00030592871255505615,
      "loss": 5.0141,
      "mean_token_accuracy": 0.21588873863220215,
      "num_tokens": 106913267.0,
      "step": 46680
    },
    {
      "entropy": 5.76472840309143,
      "epoch": 4.614966389877422,
      "grad_norm": 1.140625,
      "learning_rate": 0.00030589375719357705,
      "loss": 4.942,
      "mean_token_accuracy": 0.2238152578473091,
      "num_tokens": 106925603.0,
      "step": 46685
    },
    {
      "entropy": 5.798580503463745,
      "epoch": 4.615460656385923,
      "grad_norm": 1.484375,
      "learning_rate": 0.00030585880107210654,
      "loss": 4.9537,
      "mean_token_accuracy": 0.2326124280691147,
      "num_tokens": 106937038.0,
      "step": 46690
    },
    {
      "entropy": 5.757341527938843,
      "epoch": 4.615954922894424,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003058238441915049,
      "loss": 4.8391,
      "mean_token_accuracy": 0.231806880235672,
      "num_tokens": 106948302.0,
      "step": 46695
    },
    {
      "entropy": 5.753542566299439,
      "epoch": 4.6164491894029265,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003057888865526318,
      "loss": 4.8282,
      "mean_token_accuracy": 0.2423550859093666,
      "num_tokens": 106958347.0,
      "step": 46700
    },
    {
      "entropy": 5.805516386032105,
      "epoch": 4.616943455911428,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003057539281563473,
      "loss": 4.9869,
      "mean_token_accuracy": 0.22660465240478517,
      "num_tokens": 106969816.0,
      "step": 46705
    },
    {
      "entropy": 5.791575384140015,
      "epoch": 4.617437722419929,
      "grad_norm": 1.375,
      "learning_rate": 0.00030571896900351156,
      "loss": 4.9669,
      "mean_token_accuracy": 0.22817349880933763,
      "num_tokens": 106981100.0,
      "step": 46710
    },
    {
      "entropy": 5.705214929580689,
      "epoch": 4.61793198892843,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003056840090949843,
      "loss": 4.9013,
      "mean_token_accuracy": 0.23403552174568176,
      "num_tokens": 106993135.0,
      "step": 46715
    },
    {
      "entropy": 5.827020359039307,
      "epoch": 4.6184262554369315,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003056490484316256,
      "loss": 4.967,
      "mean_token_accuracy": 0.22484555691480637,
      "num_tokens": 107005375.0,
      "step": 46720
    },
    {
      "entropy": 5.856458234786987,
      "epoch": 4.618920521945433,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003056140870142957,
      "loss": 4.9575,
      "mean_token_accuracy": 0.2268575206398964,
      "num_tokens": 107017932.0,
      "step": 46725
    },
    {
      "entropy": 5.796388006210327,
      "epoch": 4.619414788453934,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003055791248438544,
      "loss": 4.8776,
      "mean_token_accuracy": 0.23005934059619904,
      "num_tokens": 107029122.0,
      "step": 46730
    },
    {
      "entropy": 5.718320274353028,
      "epoch": 4.619909054962436,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00030554416192116185,
      "loss": 4.9566,
      "mean_token_accuracy": 0.23064990043640138,
      "num_tokens": 107041501.0,
      "step": 46735
    },
    {
      "entropy": 5.77813868522644,
      "epoch": 4.6204033214709375,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003055091982470783,
      "loss": 4.9406,
      "mean_token_accuracy": 0.23280662000179292,
      "num_tokens": 107051948.0,
      "step": 46740
    },
    {
      "entropy": 5.810642433166504,
      "epoch": 4.620897587979439,
      "grad_norm": 1.375,
      "learning_rate": 0.00030547423382246364,
      "loss": 4.9292,
      "mean_token_accuracy": 0.23355851769447328,
      "num_tokens": 107063186.0,
      "step": 46745
    },
    {
      "entropy": 5.722560119628906,
      "epoch": 4.62139185448794,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003054392686481781,
      "loss": 4.8857,
      "mean_token_accuracy": 0.2381330594420433,
      "num_tokens": 107075617.0,
      "step": 46750
    },
    {
      "entropy": 5.782483863830566,
      "epoch": 4.621886120996441,
      "grad_norm": 1.40625,
      "learning_rate": 0.00030540430272508193,
      "loss": 4.9748,
      "mean_token_accuracy": 0.22900398969650268,
      "num_tokens": 107087505.0,
      "step": 46755
    },
    {
      "entropy": 5.825350570678711,
      "epoch": 4.6223803875049425,
      "grad_norm": 1.3125,
      "learning_rate": 0.00030536933605403513,
      "loss": 4.9214,
      "mean_token_accuracy": 0.23312136828899382,
      "num_tokens": 107099062.0,
      "step": 46760
    },
    {
      "entropy": 5.740026521682739,
      "epoch": 4.622874654013444,
      "grad_norm": 1.234375,
      "learning_rate": 0.00030533436863589794,
      "loss": 4.8709,
      "mean_token_accuracy": 0.23595583885908128,
      "num_tokens": 107112014.0,
      "step": 46765
    },
    {
      "entropy": 5.729031610488891,
      "epoch": 4.623368920521946,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003052994004715306,
      "loss": 4.8819,
      "mean_token_accuracy": 0.23987618535757066,
      "num_tokens": 107124458.0,
      "step": 46770
    },
    {
      "entropy": 5.833514547348022,
      "epoch": 4.623863187030447,
      "grad_norm": 1.390625,
      "learning_rate": 0.00030526443156179327,
      "loss": 5.0123,
      "mean_token_accuracy": 0.22328999489545823,
      "num_tokens": 107135701.0,
      "step": 46775
    },
    {
      "entropy": 5.793702745437622,
      "epoch": 4.6243574535389484,
      "grad_norm": 1.203125,
      "learning_rate": 0.00030522946190754624,
      "loss": 4.9065,
      "mean_token_accuracy": 0.2326708570122719,
      "num_tokens": 107147347.0,
      "step": 46780
    },
    {
      "entropy": 5.81126561164856,
      "epoch": 4.62485172004745,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003051944915096498,
      "loss": 4.9735,
      "mean_token_accuracy": 0.2295999437570572,
      "num_tokens": 107159287.0,
      "step": 46785
    },
    {
      "entropy": 5.770407867431641,
      "epoch": 4.625345986555951,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003051595203689642,
      "loss": 4.9808,
      "mean_token_accuracy": 0.22392118126153945,
      "num_tokens": 107170445.0,
      "step": 46790
    },
    {
      "entropy": 5.837595891952515,
      "epoch": 4.625840253064452,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00030512454848634964,
      "loss": 4.9958,
      "mean_token_accuracy": 0.22195052355527878,
      "num_tokens": 107182676.0,
      "step": 46795
    },
    {
      "entropy": 5.804289770126343,
      "epoch": 4.6263345195729535,
      "grad_norm": 1.328125,
      "learning_rate": 0.00030508957586266656,
      "loss": 4.9829,
      "mean_token_accuracy": 0.2260310471057892,
      "num_tokens": 107192929.0,
      "step": 46800
    },
    {
      "entropy": 5.747317361831665,
      "epoch": 4.626828786081455,
      "grad_norm": 1.296875,
      "learning_rate": 0.00030505460249877523,
      "loss": 4.8671,
      "mean_token_accuracy": 0.237581966817379,
      "num_tokens": 107204649.0,
      "step": 46805
    },
    {
      "entropy": 5.771765613555909,
      "epoch": 4.627323052589956,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00030501962839553603,
      "loss": 4.9419,
      "mean_token_accuracy": 0.23457082360982895,
      "num_tokens": 107216784.0,
      "step": 46810
    },
    {
      "entropy": 5.743585157394409,
      "epoch": 4.627817319098458,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003049846535538094,
      "loss": 4.8987,
      "mean_token_accuracy": 0.23351547867059708,
      "num_tokens": 107228265.0,
      "step": 46815
    },
    {
      "entropy": 5.847634267807007,
      "epoch": 4.628311585606959,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003049496779744555,
      "loss": 4.9971,
      "mean_token_accuracy": 0.22111781537532807,
      "num_tokens": 107239433.0,
      "step": 46820
    },
    {
      "entropy": 5.803410005569458,
      "epoch": 4.628805852115461,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00030491470165833483,
      "loss": 4.8886,
      "mean_token_accuracy": 0.23405026495456696,
      "num_tokens": 107250780.0,
      "step": 46825
    },
    {
      "entropy": 5.722461748123169,
      "epoch": 4.629300118623962,
      "grad_norm": 1.421875,
      "learning_rate": 0.00030487972460630804,
      "loss": 4.8818,
      "mean_token_accuracy": 0.2355014681816101,
      "num_tokens": 107261914.0,
      "step": 46830
    },
    {
      "entropy": 5.761139249801635,
      "epoch": 4.629794385132463,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00030484474681923533,
      "loss": 4.9282,
      "mean_token_accuracy": 0.23302968442440034,
      "num_tokens": 107272697.0,
      "step": 46835
    },
    {
      "entropy": 5.845745086669922,
      "epoch": 4.6302886516409645,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003048097682979772,
      "loss": 5.0165,
      "mean_token_accuracy": 0.23046574145555496,
      "num_tokens": 107283722.0,
      "step": 46840
    },
    {
      "entropy": 5.823620843887329,
      "epoch": 4.630782918149466,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00030477478904339405,
      "loss": 4.9982,
      "mean_token_accuracy": 0.22307006418704986,
      "num_tokens": 107296605.0,
      "step": 46845
    },
    {
      "entropy": 5.8146570205688475,
      "epoch": 4.631277184657968,
      "grad_norm": 1.171875,
      "learning_rate": 0.00030473980905634655,
      "loss": 4.9811,
      "mean_token_accuracy": 0.22446367591619493,
      "num_tokens": 107309564.0,
      "step": 46850
    },
    {
      "entropy": 5.792740392684936,
      "epoch": 4.631771451166469,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003047048283376952,
      "loss": 4.8964,
      "mean_token_accuracy": 0.23018056601285936,
      "num_tokens": 107321400.0,
      "step": 46855
    },
    {
      "entropy": 5.818086576461792,
      "epoch": 4.63226571767497,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003046698468883004,
      "loss": 5.0012,
      "mean_token_accuracy": 0.22375721782445906,
      "num_tokens": 107333242.0,
      "step": 46860
    },
    {
      "entropy": 5.657722091674804,
      "epoch": 4.632759984183472,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003046348647090227,
      "loss": 4.8171,
      "mean_token_accuracy": 0.24118132442235946,
      "num_tokens": 107346641.0,
      "step": 46865
    },
    {
      "entropy": 5.7459029197692875,
      "epoch": 4.633254250691973,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003045998818007228,
      "loss": 4.8513,
      "mean_token_accuracy": 0.23637654185295104,
      "num_tokens": 107358274.0,
      "step": 46870
    },
    {
      "entropy": 5.838780164718628,
      "epoch": 4.633748517200474,
      "grad_norm": 1.34375,
      "learning_rate": 0.00030456489816426123,
      "loss": 4.9924,
      "mean_token_accuracy": 0.22585046887397767,
      "num_tokens": 107370984.0,
      "step": 46875
    },
    {
      "entropy": 5.805145740509033,
      "epoch": 4.6342427837089755,
      "grad_norm": 1.25,
      "learning_rate": 0.0003045299138004986,
      "loss": 4.9704,
      "mean_token_accuracy": 0.23411972522735597,
      "num_tokens": 107383416.0,
      "step": 46880
    },
    {
      "entropy": 5.749883604049683,
      "epoch": 4.634737050217478,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003044949287102955,
      "loss": 4.8764,
      "mean_token_accuracy": 0.2316595658659935,
      "num_tokens": 107394949.0,
      "step": 46885
    },
    {
      "entropy": 5.7576038360595705,
      "epoch": 4.635231316725979,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003044599428945125,
      "loss": 4.9278,
      "mean_token_accuracy": 0.23068660646677017,
      "num_tokens": 107406310.0,
      "step": 46890
    },
    {
      "entropy": 5.809793853759766,
      "epoch": 4.63572558323448,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003044249563540104,
      "loss": 4.9693,
      "mean_token_accuracy": 0.2268818974494934,
      "num_tokens": 107417988.0,
      "step": 46895
    },
    {
      "entropy": 5.752955722808838,
      "epoch": 4.636219849742981,
      "grad_norm": 1.25,
      "learning_rate": 0.00030438996908964985,
      "loss": 4.8777,
      "mean_token_accuracy": 0.23615337163209915,
      "num_tokens": 107430254.0,
      "step": 46900
    },
    {
      "entropy": 5.7924481391906735,
      "epoch": 4.636714116251483,
      "grad_norm": 1.328125,
      "learning_rate": 0.00030435498110229143,
      "loss": 4.9305,
      "mean_token_accuracy": 0.23742320984601975,
      "num_tokens": 107441674.0,
      "step": 46905
    },
    {
      "entropy": 5.842913198471069,
      "epoch": 4.637208382759984,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003043199923927959,
      "loss": 4.9677,
      "mean_token_accuracy": 0.228306382894516,
      "num_tokens": 107452019.0,
      "step": 46910
    },
    {
      "entropy": 5.854248142242431,
      "epoch": 4.637702649268485,
      "grad_norm": 1.453125,
      "learning_rate": 0.000304285002962024,
      "loss": 5.0642,
      "mean_token_accuracy": 0.21787166297435762,
      "num_tokens": 107463254.0,
      "step": 46915
    },
    {
      "entropy": 5.764175462722778,
      "epoch": 4.638196915776987,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003042500128108366,
      "loss": 4.9216,
      "mean_token_accuracy": 0.23076693713665009,
      "num_tokens": 107474674.0,
      "step": 46920
    },
    {
      "entropy": 5.801751518249512,
      "epoch": 4.638691182285489,
      "grad_norm": 1.234375,
      "learning_rate": 0.00030421502194009426,
      "loss": 4.9709,
      "mean_token_accuracy": 0.22785632759332658,
      "num_tokens": 107486374.0,
      "step": 46925
    },
    {
      "entropy": 5.795235633850098,
      "epoch": 4.63918544879399,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00030418003035065784,
      "loss": 4.937,
      "mean_token_accuracy": 0.22354020923376083,
      "num_tokens": 107497759.0,
      "step": 46930
    },
    {
      "entropy": 5.777407789230347,
      "epoch": 4.639679715302491,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003041450380433882,
      "loss": 4.9102,
      "mean_token_accuracy": 0.23000673204660416,
      "num_tokens": 107509693.0,
      "step": 46935
    },
    {
      "entropy": 5.696861124038696,
      "epoch": 4.640173981810992,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00030411004501914603,
      "loss": 4.8585,
      "mean_token_accuracy": 0.23283033967018127,
      "num_tokens": 107521453.0,
      "step": 46940
    },
    {
      "entropy": 5.7536849021911625,
      "epoch": 4.640668248319494,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0003040750512787923,
      "loss": 4.8734,
      "mean_token_accuracy": 0.24750007539987565,
      "num_tokens": 107533401.0,
      "step": 46945
    },
    {
      "entropy": 5.815132761001587,
      "epoch": 4.641162514827995,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003040400568231878,
      "loss": 4.89,
      "mean_token_accuracy": 0.2341687023639679,
      "num_tokens": 107545126.0,
      "step": 46950
    },
    {
      "entropy": 5.893264818191528,
      "epoch": 4.641656781336497,
      "grad_norm": 1.265625,
      "learning_rate": 0.00030400506165319333,
      "loss": 5.0632,
      "mean_token_accuracy": 0.21849648803472518,
      "num_tokens": 107555774.0,
      "step": 46955
    },
    {
      "entropy": 5.847916126251221,
      "epoch": 4.642151047844998,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003039700657696699,
      "loss": 5.0371,
      "mean_token_accuracy": 0.21739352196455003,
      "num_tokens": 107567522.0,
      "step": 46960
    },
    {
      "entropy": 5.724497699737549,
      "epoch": 4.6426453143535,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00030393506917347833,
      "loss": 4.8422,
      "mean_token_accuracy": 0.24883191883563996,
      "num_tokens": 107578637.0,
      "step": 46965
    },
    {
      "entropy": 5.713301801681519,
      "epoch": 4.643139580862001,
      "grad_norm": 1.328125,
      "learning_rate": 0.00030390007186547956,
      "loss": 4.9028,
      "mean_token_accuracy": 0.22882129698991777,
      "num_tokens": 107590750.0,
      "step": 46970
    },
    {
      "entropy": 5.775608205795288,
      "epoch": 4.643633847370502,
      "grad_norm": 1.375,
      "learning_rate": 0.0003038650738465347,
      "loss": 4.8998,
      "mean_token_accuracy": 0.23181559294462203,
      "num_tokens": 107600660.0,
      "step": 46975
    },
    {
      "entropy": 5.803049373626709,
      "epoch": 4.644128113879003,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003038300751175044,
      "loss": 4.9801,
      "mean_token_accuracy": 0.22562056928873062,
      "num_tokens": 107612880.0,
      "step": 46980
    },
    {
      "entropy": 5.728144025802612,
      "epoch": 4.644622380387505,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003037950756792497,
      "loss": 4.9074,
      "mean_token_accuracy": 0.2424024224281311,
      "num_tokens": 107625651.0,
      "step": 46985
    },
    {
      "entropy": 5.806266117095947,
      "epoch": 4.645116646896007,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003037600755326317,
      "loss": 4.9098,
      "mean_token_accuracy": 0.23096064627170562,
      "num_tokens": 107636418.0,
      "step": 46990
    },
    {
      "entropy": 5.811736583709717,
      "epoch": 4.645610913404508,
      "grad_norm": 1.265625,
      "learning_rate": 0.00030372507467851155,
      "loss": 4.9868,
      "mean_token_accuracy": 0.22928935885429383,
      "num_tokens": 107648643.0,
      "step": 46995
    },
    {
      "entropy": 5.779860019683838,
      "epoch": 4.646105179913009,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00030369007311775,
      "loss": 4.9521,
      "mean_token_accuracy": 0.2262760728597641,
      "num_tokens": 107660183.0,
      "step": 47000
    },
    {
      "entropy": 5.77252049446106,
      "epoch": 4.646599446421511,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003036550708512082,
      "loss": 4.9462,
      "mean_token_accuracy": 0.23115115314722062,
      "num_tokens": 107670868.0,
      "step": 47005
    },
    {
      "entropy": 5.7599915027618405,
      "epoch": 4.647093712930012,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003036200678797471,
      "loss": 4.9047,
      "mean_token_accuracy": 0.23555329591035842,
      "num_tokens": 107682502.0,
      "step": 47010
    },
    {
      "entropy": 5.819391965866089,
      "epoch": 4.647587979438513,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000303585064204228,
      "loss": 4.9538,
      "mean_token_accuracy": 0.22348420917987824,
      "num_tokens": 107694490.0,
      "step": 47015
    },
    {
      "entropy": 5.848630809783936,
      "epoch": 4.648082245947014,
      "grad_norm": 1.25,
      "learning_rate": 0.000303550059825512,
      "loss": 4.9652,
      "mean_token_accuracy": 0.2221877709031105,
      "num_tokens": 107707172.0,
      "step": 47020
    },
    {
      "entropy": 5.8045569896698,
      "epoch": 4.6485765124555165,
      "grad_norm": 1.484375,
      "learning_rate": 0.00030351505474445997,
      "loss": 4.8944,
      "mean_token_accuracy": 0.23476910442113877,
      "num_tokens": 107718919.0,
      "step": 47025
    },
    {
      "entropy": 5.798918342590332,
      "epoch": 4.649070778964018,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003034800489619332,
      "loss": 4.9583,
      "mean_token_accuracy": 0.22852526903152465,
      "num_tokens": 107729839.0,
      "step": 47030
    },
    {
      "entropy": 5.736470365524292,
      "epoch": 4.649565045472519,
      "grad_norm": 1.265625,
      "learning_rate": 0.00030344504247879273,
      "loss": 4.8666,
      "mean_token_accuracy": 0.23711312115192412,
      "num_tokens": 107740551.0,
      "step": 47035
    },
    {
      "entropy": 5.834483671188354,
      "epoch": 4.65005931198102,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003034100352958999,
      "loss": 5.0346,
      "mean_token_accuracy": 0.21883748918771745,
      "num_tokens": 107754594.0,
      "step": 47040
    },
    {
      "entropy": 5.788775205612183,
      "epoch": 4.6505535784895216,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003033750274141157,
      "loss": 4.9434,
      "mean_token_accuracy": 0.22954196631908416,
      "num_tokens": 107765957.0,
      "step": 47045
    },
    {
      "entropy": 5.81618914604187,
      "epoch": 4.651047844998023,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003033400188343015,
      "loss": 4.9178,
      "mean_token_accuracy": 0.23099416494369507,
      "num_tokens": 107776336.0,
      "step": 47050
    },
    {
      "entropy": 5.705657148361206,
      "epoch": 4.651542111506524,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003033050095573185,
      "loss": 4.8482,
      "mean_token_accuracy": 0.24015369564294814,
      "num_tokens": 107788394.0,
      "step": 47055
    },
    {
      "entropy": 5.830919408798218,
      "epoch": 4.652036378015025,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003032699995840277,
      "loss": 4.9709,
      "mean_token_accuracy": 0.2256106972694397,
      "num_tokens": 107800398.0,
      "step": 47060
    },
    {
      "entropy": 5.815746927261353,
      "epoch": 4.652530644523527,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003032349889152907,
      "loss": 4.9686,
      "mean_token_accuracy": 0.22416827976703643,
      "num_tokens": 107810906.0,
      "step": 47065
    },
    {
      "entropy": 5.747481489181519,
      "epoch": 4.653024911032029,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003031999775519685,
      "loss": 4.8307,
      "mean_token_accuracy": 0.2402476578950882,
      "num_tokens": 107821864.0,
      "step": 47070
    },
    {
      "entropy": 5.768503618240357,
      "epoch": 4.65351917754053,
      "grad_norm": 1.28125,
      "learning_rate": 0.00030316496549492243,
      "loss": 4.8854,
      "mean_token_accuracy": 0.23162124156951905,
      "num_tokens": 107832898.0,
      "step": 47075
    },
    {
      "entropy": 5.7859814167022705,
      "epoch": 4.654013444049031,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003031299527450139,
      "loss": 4.9559,
      "mean_token_accuracy": 0.2311214253306389,
      "num_tokens": 107845075.0,
      "step": 47080
    },
    {
      "entropy": 5.895735883712769,
      "epoch": 4.6545077105575325,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00030309493930310416,
      "loss": 5.0282,
      "mean_token_accuracy": 0.21744897216558456,
      "num_tokens": 107856735.0,
      "step": 47085
    },
    {
      "entropy": 5.771990633010864,
      "epoch": 4.655001977066034,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003030599251700546,
      "loss": 4.8759,
      "mean_token_accuracy": 0.24318237900733947,
      "num_tokens": 107868224.0,
      "step": 47090
    },
    {
      "entropy": 5.7607152462005615,
      "epoch": 4.655496243574535,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003030249103467265,
      "loss": 4.9575,
      "mean_token_accuracy": 0.22793270349502565,
      "num_tokens": 107879332.0,
      "step": 47095
    },
    {
      "entropy": 5.778506803512573,
      "epoch": 4.655990510083036,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003029898948339812,
      "loss": 4.9154,
      "mean_token_accuracy": 0.2332310274243355,
      "num_tokens": 107891537.0,
      "step": 47100
    },
    {
      "entropy": 5.736543750762939,
      "epoch": 4.6564847765915385,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00030295487863268015,
      "loss": 4.8442,
      "mean_token_accuracy": 0.2425287976861,
      "num_tokens": 107904127.0,
      "step": 47105
    },
    {
      "entropy": 5.822204828262329,
      "epoch": 4.65697904310004,
      "grad_norm": 1.28125,
      "learning_rate": 0.00030291986174368484,
      "loss": 4.9488,
      "mean_token_accuracy": 0.22527184933423997,
      "num_tokens": 107915436.0,
      "step": 47110
    },
    {
      "entropy": 5.673159790039063,
      "epoch": 4.657473309608541,
      "grad_norm": 1.515625,
      "learning_rate": 0.00030288484416785656,
      "loss": 4.7931,
      "mean_token_accuracy": 0.23887789994478226,
      "num_tokens": 107927335.0,
      "step": 47115
    },
    {
      "entropy": 5.771707677841187,
      "epoch": 4.657967576117042,
      "grad_norm": 1.234375,
      "learning_rate": 0.00030284982590605687,
      "loss": 4.9419,
      "mean_token_accuracy": 0.22586944550275803,
      "num_tokens": 107938941.0,
      "step": 47120
    },
    {
      "entropy": 5.810255908966065,
      "epoch": 4.6584618426255435,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00030281480695914696,
      "loss": 4.9732,
      "mean_token_accuracy": 0.22516824901103974,
      "num_tokens": 107950333.0,
      "step": 47125
    },
    {
      "entropy": 5.821358728408813,
      "epoch": 4.658956109134045,
      "grad_norm": 1.34375,
      "learning_rate": 0.00030277978732798854,
      "loss": 4.9632,
      "mean_token_accuracy": 0.23040594905614853,
      "num_tokens": 107960951.0,
      "step": 47130
    },
    {
      "entropy": 5.80390419960022,
      "epoch": 4.659450375642546,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00030274476701344313,
      "loss": 4.9638,
      "mean_token_accuracy": 0.23027291446924208,
      "num_tokens": 107972383.0,
      "step": 47135
    },
    {
      "entropy": 5.751413106918335,
      "epoch": 4.659944642151048,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00030270974601637207,
      "loss": 4.8784,
      "mean_token_accuracy": 0.23280862271785735,
      "num_tokens": 107984562.0,
      "step": 47140
    },
    {
      "entropy": 5.732869243621826,
      "epoch": 4.6604389086595495,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000302674724337637,
      "loss": 4.8765,
      "mean_token_accuracy": 0.23776807636022568,
      "num_tokens": 107995445.0,
      "step": 47145
    },
    {
      "entropy": 5.824235534667968,
      "epoch": 4.660933175168051,
      "grad_norm": 1.3125,
      "learning_rate": 0.00030263970197809935,
      "loss": 5.0088,
      "mean_token_accuracy": 0.22183357328176498,
      "num_tokens": 108007444.0,
      "step": 47150
    },
    {
      "entropy": 5.808983230590821,
      "epoch": 4.661427441676552,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00030260467893862086,
      "loss": 4.9533,
      "mean_token_accuracy": 0.23162726014852525,
      "num_tokens": 108019813.0,
      "step": 47155
    },
    {
      "entropy": 5.756083583831787,
      "epoch": 4.661921708185053,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003025696552200629,
      "loss": 4.8481,
      "mean_token_accuracy": 0.24116761684417726,
      "num_tokens": 108031309.0,
      "step": 47160
    },
    {
      "entropy": 5.720457220077515,
      "epoch": 4.6624159746935545,
      "grad_norm": 1.5,
      "learning_rate": 0.0003025346308232872,
      "loss": 4.8871,
      "mean_token_accuracy": 0.2379217267036438,
      "num_tokens": 108042480.0,
      "step": 47165
    },
    {
      "entropy": 5.807332468032837,
      "epoch": 4.662910241202056,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00030249960574915523,
      "loss": 4.9219,
      "mean_token_accuracy": 0.22399173974990844,
      "num_tokens": 108054297.0,
      "step": 47170
    },
    {
      "entropy": 5.816686916351318,
      "epoch": 4.663404507710558,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003024645799985287,
      "loss": 4.9969,
      "mean_token_accuracy": 0.2252377226948738,
      "num_tokens": 108066065.0,
      "step": 47175
    },
    {
      "entropy": 5.868224334716797,
      "epoch": 4.663898774219059,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003024295535722693,
      "loss": 5.0065,
      "mean_token_accuracy": 0.2236034169793129,
      "num_tokens": 108078421.0,
      "step": 47180
    },
    {
      "entropy": 5.76887698173523,
      "epoch": 4.6643930407275604,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003023945264712386,
      "loss": 4.9721,
      "mean_token_accuracy": 0.2306017830967903,
      "num_tokens": 108090022.0,
      "step": 47185
    },
    {
      "entropy": 5.7657371997833256,
      "epoch": 4.664887307236062,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003023594986962984,
      "loss": 4.9484,
      "mean_token_accuracy": 0.22866140455007553,
      "num_tokens": 108102425.0,
      "step": 47190
    },
    {
      "entropy": 5.796149492263794,
      "epoch": 4.665381573744563,
      "grad_norm": 1.3125,
      "learning_rate": 0.00030232447024831015,
      "loss": 4.9653,
      "mean_token_accuracy": 0.22979059666395188,
      "num_tokens": 108114801.0,
      "step": 47195
    },
    {
      "entropy": 5.7963744640350345,
      "epoch": 4.665875840253064,
      "grad_norm": 1.359375,
      "learning_rate": 0.00030228944112813567,
      "loss": 4.9023,
      "mean_token_accuracy": 0.2323967769742012,
      "num_tokens": 108126089.0,
      "step": 47200
    },
    {
      "entropy": 5.747124576568604,
      "epoch": 4.6663701067615655,
      "grad_norm": 1.375,
      "learning_rate": 0.0003022544113366368,
      "loss": 4.8385,
      "mean_token_accuracy": 0.2410220116376877,
      "num_tokens": 108137500.0,
      "step": 47205
    },
    {
      "entropy": 5.780154418945313,
      "epoch": 4.666864373270068,
      "grad_norm": 1.3125,
      "learning_rate": 0.00030221938087467517,
      "loss": 5.0089,
      "mean_token_accuracy": 0.216172057390213,
      "num_tokens": 108148645.0,
      "step": 47210
    },
    {
      "entropy": 5.800406265258789,
      "epoch": 4.667358639778569,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003021843497431125,
      "loss": 4.9626,
      "mean_token_accuracy": 0.2223236531019211,
      "num_tokens": 108159227.0,
      "step": 47215
    },
    {
      "entropy": 5.767984294891358,
      "epoch": 4.66785290628707,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003021493179428107,
      "loss": 4.9131,
      "mean_token_accuracy": 0.23571093380451202,
      "num_tokens": 108170686.0,
      "step": 47220
    },
    {
      "entropy": 5.845607566833496,
      "epoch": 4.668347172795571,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003021142854746314,
      "loss": 4.9711,
      "mean_token_accuracy": 0.2292397752404213,
      "num_tokens": 108182170.0,
      "step": 47225
    },
    {
      "entropy": 5.730560541152954,
      "epoch": 4.668841439304073,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003020792523394364,
      "loss": 4.8647,
      "mean_token_accuracy": 0.23827758580446243,
      "num_tokens": 108193638.0,
      "step": 47230
    },
    {
      "entropy": 5.744837665557862,
      "epoch": 4.669335705812574,
      "grad_norm": 1.390625,
      "learning_rate": 0.00030204421853808765,
      "loss": 4.9482,
      "mean_token_accuracy": 0.22724126428365707,
      "num_tokens": 108205455.0,
      "step": 47235
    },
    {
      "entropy": 5.80811915397644,
      "epoch": 4.669829972321075,
      "grad_norm": 1.265625,
      "learning_rate": 0.000302009184071447,
      "loss": 4.937,
      "mean_token_accuracy": 0.22525297701358796,
      "num_tokens": 108216949.0,
      "step": 47240
    },
    {
      "entropy": 5.847018527984619,
      "epoch": 4.670324238829577,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003019741489403762,
      "loss": 4.9467,
      "mean_token_accuracy": 0.2233959749341011,
      "num_tokens": 108227792.0,
      "step": 47245
    },
    {
      "entropy": 5.788420820236206,
      "epoch": 4.670818505338079,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003019391131457371,
      "loss": 4.9056,
      "mean_token_accuracy": 0.2358655020594597,
      "num_tokens": 108239167.0,
      "step": 47250
    },
    {
      "entropy": 5.8632443904876705,
      "epoch": 4.67131277184658,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003019040766883917,
      "loss": 5.0346,
      "mean_token_accuracy": 0.22102455347776412,
      "num_tokens": 108251241.0,
      "step": 47255
    },
    {
      "entropy": 5.755634927749634,
      "epoch": 4.671807038355081,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00030186903956920183,
      "loss": 4.8762,
      "mean_token_accuracy": 0.23423669934272767,
      "num_tokens": 108262287.0,
      "step": 47260
    },
    {
      "entropy": 5.7509729862213135,
      "epoch": 4.672301304863582,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003018340017890294,
      "loss": 4.847,
      "mean_token_accuracy": 0.23831482529640197,
      "num_tokens": 108275126.0,
      "step": 47265
    },
    {
      "entropy": 5.726219034194946,
      "epoch": 4.672795571372084,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003017989633487363,
      "loss": 4.8827,
      "mean_token_accuracy": 0.2269839733839035,
      "num_tokens": 108287021.0,
      "step": 47270
    },
    {
      "entropy": 5.774707841873169,
      "epoch": 4.673289837880585,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003017639242491846,
      "loss": 4.8505,
      "mean_token_accuracy": 0.23849994391202928,
      "num_tokens": 108298512.0,
      "step": 47275
    },
    {
      "entropy": 5.727691411972046,
      "epoch": 4.673784104389087,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003017288844912362,
      "loss": 4.8511,
      "mean_token_accuracy": 0.24037442356348038,
      "num_tokens": 108309505.0,
      "step": 47280
    },
    {
      "entropy": 5.793265295028687,
      "epoch": 4.674278370897588,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00030169384407575315,
      "loss": 4.9246,
      "mean_token_accuracy": 0.22959048599004744,
      "num_tokens": 108320404.0,
      "step": 47285
    },
    {
      "entropy": 5.778944587707519,
      "epoch": 4.67477263740609,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003016588030035973,
      "loss": 4.9897,
      "mean_token_accuracy": 0.22556092590093613,
      "num_tokens": 108332221.0,
      "step": 47290
    },
    {
      "entropy": 5.7904242992401125,
      "epoch": 4.675266903914591,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00030162376127563085,
      "loss": 4.9523,
      "mean_token_accuracy": 0.22357306629419327,
      "num_tokens": 108344962.0,
      "step": 47295
    },
    {
      "entropy": 5.727663230895996,
      "epoch": 4.675761170423092,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00030158871889271574,
      "loss": 4.8374,
      "mean_token_accuracy": 0.2410629078745842,
      "num_tokens": 108356515.0,
      "step": 47300
    },
    {
      "entropy": 5.773973178863526,
      "epoch": 4.676255436931593,
      "grad_norm": 1.328125,
      "learning_rate": 0.00030155367585571393,
      "loss": 4.9319,
      "mean_token_accuracy": 0.2320661038160324,
      "num_tokens": 108367807.0,
      "step": 47305
    },
    {
      "entropy": 5.828750944137573,
      "epoch": 4.676749703440095,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003015186321654877,
      "loss": 5.0066,
      "mean_token_accuracy": 0.22485716491937638,
      "num_tokens": 108381296.0,
      "step": 47310
    },
    {
      "entropy": 5.814452123641968,
      "epoch": 4.677243969948596,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00030148358782289893,
      "loss": 5.0048,
      "mean_token_accuracy": 0.2319187968969345,
      "num_tokens": 108392331.0,
      "step": 47315
    },
    {
      "entropy": 5.802259826660157,
      "epoch": 4.677738236457097,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00030144854282880974,
      "loss": 4.8828,
      "mean_token_accuracy": 0.24043433964252472,
      "num_tokens": 108402523.0,
      "step": 47320
    },
    {
      "entropy": 5.8346014499664305,
      "epoch": 4.678232502965599,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00030141349718408237,
      "loss": 4.9929,
      "mean_token_accuracy": 0.22858950942754747,
      "num_tokens": 108413828.0,
      "step": 47325
    },
    {
      "entropy": 5.754173517227173,
      "epoch": 4.678726769474101,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003013784508895788,
      "loss": 4.861,
      "mean_token_accuracy": 0.23203733265399934,
      "num_tokens": 108424273.0,
      "step": 47330
    },
    {
      "entropy": 5.789046001434326,
      "epoch": 4.679221035982602,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00030134340394616133,
      "loss": 4.9718,
      "mean_token_accuracy": 0.2277196690440178,
      "num_tokens": 108435620.0,
      "step": 47335
    },
    {
      "entropy": 5.679608774185181,
      "epoch": 4.679715302491103,
      "grad_norm": 1.484375,
      "learning_rate": 0.000301308356354692,
      "loss": 4.7852,
      "mean_token_accuracy": 0.23625802099704743,
      "num_tokens": 108445947.0,
      "step": 47340
    },
    {
      "entropy": 5.779399061203003,
      "epoch": 4.680209568999604,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00030127330811603297,
      "loss": 4.9038,
      "mean_token_accuracy": 0.23002164661884308,
      "num_tokens": 108457343.0,
      "step": 47345
    },
    {
      "entropy": 5.684987592697143,
      "epoch": 4.680703835508106,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003012382592310466,
      "loss": 4.803,
      "mean_token_accuracy": 0.24076834917068482,
      "num_tokens": 108468393.0,
      "step": 47350
    },
    {
      "entropy": 5.833012247085572,
      "epoch": 4.681198102016607,
      "grad_norm": 1.390625,
      "learning_rate": 0.00030120320970059494,
      "loss": 5.0168,
      "mean_token_accuracy": 0.22196914106607438,
      "num_tokens": 108479170.0,
      "step": 47355
    },
    {
      "entropy": 5.835642623901367,
      "epoch": 4.681692368525109,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003011681595255402,
      "loss": 4.9376,
      "mean_token_accuracy": 0.23087151497602462,
      "num_tokens": 108490807.0,
      "step": 47360
    },
    {
      "entropy": 5.911527395248413,
      "epoch": 4.68218663503361,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003011331087067447,
      "loss": 5.0513,
      "mean_token_accuracy": 0.21888690441846848,
      "num_tokens": 108501582.0,
      "step": 47365
    },
    {
      "entropy": 5.795808029174805,
      "epoch": 4.682680901542112,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00030109805724507074,
      "loss": 4.9256,
      "mean_token_accuracy": 0.23486603796482086,
      "num_tokens": 108513805.0,
      "step": 47370
    },
    {
      "entropy": 5.870370197296142,
      "epoch": 4.683175168050613,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003010630051413804,
      "loss": 4.9745,
      "mean_token_accuracy": 0.21895952075719832,
      "num_tokens": 108525486.0,
      "step": 47375
    },
    {
      "entropy": 5.815187931060791,
      "epoch": 4.683669434559114,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003010279523965362,
      "loss": 4.9698,
      "mean_token_accuracy": 0.23192799985408782,
      "num_tokens": 108537194.0,
      "step": 47380
    },
    {
      "entropy": 5.8313206195831295,
      "epoch": 4.684163701067615,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003009928990114003,
      "loss": 4.9922,
      "mean_token_accuracy": 0.22354118824005126,
      "num_tokens": 108548422.0,
      "step": 47385
    },
    {
      "entropy": 5.844069862365723,
      "epoch": 4.684657967576117,
      "grad_norm": 1.359375,
      "learning_rate": 0.000300957844986835,
      "loss": 4.9823,
      "mean_token_accuracy": 0.22397423535585403,
      "num_tokens": 108559604.0,
      "step": 47390
    },
    {
      "entropy": 5.818086242675781,
      "epoch": 4.685152234084619,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003009227903237027,
      "loss": 5.0051,
      "mean_token_accuracy": 0.22341058552265167,
      "num_tokens": 108570878.0,
      "step": 47395
    },
    {
      "entropy": 5.768146419525147,
      "epoch": 4.68564650059312,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003008877350228659,
      "loss": 4.8693,
      "mean_token_accuracy": 0.23457152843475343,
      "num_tokens": 108582687.0,
      "step": 47400
    },
    {
      "entropy": 5.762358236312866,
      "epoch": 4.686140767101621,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003008526790851866,
      "loss": 4.9127,
      "mean_token_accuracy": 0.23555808812379836,
      "num_tokens": 108593505.0,
      "step": 47405
    },
    {
      "entropy": 5.740628242492676,
      "epoch": 4.686635033610123,
      "grad_norm": 1.234375,
      "learning_rate": 0.00030081762251152745,
      "loss": 4.8659,
      "mean_token_accuracy": 0.2364867866039276,
      "num_tokens": 108604809.0,
      "step": 47410
    },
    {
      "entropy": 5.732062959671021,
      "epoch": 4.687129300118624,
      "grad_norm": 1.265625,
      "learning_rate": 0.00030078256530275077,
      "loss": 4.9121,
      "mean_token_accuracy": 0.23210962861776352,
      "num_tokens": 108616601.0,
      "step": 47415
    },
    {
      "entropy": 5.80835280418396,
      "epoch": 4.687623566627125,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000300747507459719,
      "loss": 4.893,
      "mean_token_accuracy": 0.23503482192754746,
      "num_tokens": 108627521.0,
      "step": 47420
    },
    {
      "entropy": 5.818966627120972,
      "epoch": 4.688117833135626,
      "grad_norm": 1.203125,
      "learning_rate": 0.00030071244898329457,
      "loss": 4.9856,
      "mean_token_accuracy": 0.2256098970770836,
      "num_tokens": 108639602.0,
      "step": 47425
    },
    {
      "entropy": 5.891282367706299,
      "epoch": 4.6886120996441285,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003006773898743399,
      "loss": 5.03,
      "mean_token_accuracy": 0.21880697160959245,
      "num_tokens": 108652546.0,
      "step": 47430
    },
    {
      "entropy": 5.750076484680176,
      "epoch": 4.68910636615263,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003006423301337174,
      "loss": 4.8464,
      "mean_token_accuracy": 0.23219281286001206,
      "num_tokens": 108663772.0,
      "step": 47435
    },
    {
      "entropy": 5.814196157455444,
      "epoch": 4.689600632661131,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00030060726976228964,
      "loss": 4.9296,
      "mean_token_accuracy": 0.22501252889633178,
      "num_tokens": 108674865.0,
      "step": 47440
    },
    {
      "entropy": 5.835595655441284,
      "epoch": 4.690094899169632,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003005722087609191,
      "loss": 5.0479,
      "mean_token_accuracy": 0.22176049053668975,
      "num_tokens": 108687127.0,
      "step": 47445
    },
    {
      "entropy": 5.825901174545288,
      "epoch": 4.6905891656781336,
      "grad_norm": 1.421875,
      "learning_rate": 0.00030053714713046826,
      "loss": 4.9602,
      "mean_token_accuracy": 0.22832885831594468,
      "num_tokens": 108698119.0,
      "step": 47450
    },
    {
      "entropy": 5.757703685760498,
      "epoch": 4.691083432186635,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003005020848717996,
      "loss": 4.9173,
      "mean_token_accuracy": 0.23430751711130143,
      "num_tokens": 108709104.0,
      "step": 47455
    },
    {
      "entropy": 5.710097885131836,
      "epoch": 4.691577698695136,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003004670219857757,
      "loss": 4.8887,
      "mean_token_accuracy": 0.24015873968601226,
      "num_tokens": 108720111.0,
      "step": 47460
    },
    {
      "entropy": 5.796454668045044,
      "epoch": 4.692071965203638,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003004319584732591,
      "loss": 4.9776,
      "mean_token_accuracy": 0.22292574048042296,
      "num_tokens": 108731496.0,
      "step": 47465
    },
    {
      "entropy": 5.746397352218628,
      "epoch": 4.6925662317121395,
      "grad_norm": 1.328125,
      "learning_rate": 0.00030039689433511234,
      "loss": 4.9263,
      "mean_token_accuracy": 0.2338717073202133,
      "num_tokens": 108742760.0,
      "step": 47470
    },
    {
      "entropy": 5.722792768478394,
      "epoch": 4.693060498220641,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003003618295721982,
      "loss": 4.8493,
      "mean_token_accuracy": 0.24326559007167817,
      "num_tokens": 108754554.0,
      "step": 47475
    },
    {
      "entropy": 5.82582802772522,
      "epoch": 4.693554764729142,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003003267641853789,
      "loss": 4.9277,
      "mean_token_accuracy": 0.232988241314888,
      "num_tokens": 108765832.0,
      "step": 47480
    },
    {
      "entropy": 5.793525505065918,
      "epoch": 4.694049031237643,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00030029169817551746,
      "loss": 4.9176,
      "mean_token_accuracy": 0.22953367084264756,
      "num_tokens": 108776898.0,
      "step": 47485
    },
    {
      "entropy": 5.723574876785278,
      "epoch": 4.6945432977461445,
      "grad_norm": 1.390625,
      "learning_rate": 0.00030025663154347615,
      "loss": 4.802,
      "mean_token_accuracy": 0.24124278277158737,
      "num_tokens": 108786724.0,
      "step": 47490
    },
    {
      "entropy": 5.774144887924194,
      "epoch": 4.695037564254646,
      "grad_norm": 1.390625,
      "learning_rate": 0.00030022156429011787,
      "loss": 4.9493,
      "mean_token_accuracy": 0.22150613218545914,
      "num_tokens": 108797950.0,
      "step": 47495
    },
    {
      "entropy": 5.728539037704468,
      "epoch": 4.695531830763148,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003001864964163053,
      "loss": 4.8585,
      "mean_token_accuracy": 0.236504989862442,
      "num_tokens": 108809931.0,
      "step": 47500
    },
    {
      "entropy": 5.792474794387817,
      "epoch": 4.696026097271649,
      "grad_norm": 1.359375,
      "learning_rate": 0.00030015142792290086,
      "loss": 4.9638,
      "mean_token_accuracy": 0.2231461688876152,
      "num_tokens": 108820893.0,
      "step": 47505
    },
    {
      "entropy": 5.821643543243408,
      "epoch": 4.6965203637801505,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003001163588107674,
      "loss": 4.9628,
      "mean_token_accuracy": 0.22859786003828048,
      "num_tokens": 108832208.0,
      "step": 47510
    },
    {
      "entropy": 5.840558242797852,
      "epoch": 4.697014630288652,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003000812890807677,
      "loss": 5.0164,
      "mean_token_accuracy": 0.21479549258947372,
      "num_tokens": 108843297.0,
      "step": 47515
    },
    {
      "entropy": 5.82894082069397,
      "epoch": 4.697508896797153,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003000462187337644,
      "loss": 4.91,
      "mean_token_accuracy": 0.23035038709640504,
      "num_tokens": 108853503.0,
      "step": 47520
    },
    {
      "entropy": 5.8051676750183105,
      "epoch": 4.698003163305654,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00030001114777062016,
      "loss": 4.9426,
      "mean_token_accuracy": 0.23255185335874556,
      "num_tokens": 108864305.0,
      "step": 47525
    },
    {
      "entropy": 5.670496797561645,
      "epoch": 4.6984974298141555,
      "grad_norm": 1.375,
      "learning_rate": 0.0002999760761921978,
      "loss": 4.837,
      "mean_token_accuracy": 0.24469426572322844,
      "num_tokens": 108875096.0,
      "step": 47530
    },
    {
      "entropy": 5.73917965888977,
      "epoch": 4.698991696322658,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00029994100399936016,
      "loss": 4.8892,
      "mean_token_accuracy": 0.2391519844532013,
      "num_tokens": 108886708.0,
      "step": 47535
    },
    {
      "entropy": 5.826587677001953,
      "epoch": 4.699485962831159,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00029990593119296995,
      "loss": 4.9843,
      "mean_token_accuracy": 0.21747240126132966,
      "num_tokens": 108897706.0,
      "step": 47540
    },
    {
      "entropy": 5.777768087387085,
      "epoch": 4.69998022933966,
      "grad_norm": 1.3125,
      "learning_rate": 0.00029987085777388994,
      "loss": 4.9189,
      "mean_token_accuracy": 0.2314580038189888,
      "num_tokens": 108908356.0,
      "step": 47545
    },
    {
      "entropy": 5.830490159988403,
      "epoch": 4.7004744958481615,
      "grad_norm": 1.34375,
      "learning_rate": 0.000299835783742983,
      "loss": 4.9415,
      "mean_token_accuracy": 0.2284214034676552,
      "num_tokens": 108920090.0,
      "step": 47550
    },
    {
      "entropy": 5.741714334487915,
      "epoch": 4.700968762356663,
      "grad_norm": 1.375,
      "learning_rate": 0.00029980070910111184,
      "loss": 4.8407,
      "mean_token_accuracy": 0.24353937804698944,
      "num_tokens": 108931171.0,
      "step": 47555
    },
    {
      "entropy": 5.782672357559204,
      "epoch": 4.701463028865164,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002997656338491395,
      "loss": 4.941,
      "mean_token_accuracy": 0.23217135965824126,
      "num_tokens": 108941915.0,
      "step": 47560
    },
    {
      "entropy": 5.78099684715271,
      "epoch": 4.701957295373665,
      "grad_norm": 1.296875,
      "learning_rate": 0.00029973055798792865,
      "loss": 4.9203,
      "mean_token_accuracy": 0.2261556178331375,
      "num_tokens": 108953983.0,
      "step": 47565
    },
    {
      "entropy": 5.709943675994873,
      "epoch": 4.7024515618821665,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00029969548151834223,
      "loss": 4.8448,
      "mean_token_accuracy": 0.24059588015079497,
      "num_tokens": 108965372.0,
      "step": 47570
    },
    {
      "entropy": 5.783597373962403,
      "epoch": 4.702945828390668,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002996604044412431,
      "loss": 4.9309,
      "mean_token_accuracy": 0.22744090110063553,
      "num_tokens": 108975739.0,
      "step": 47575
    },
    {
      "entropy": 5.849443531036377,
      "epoch": 4.70344009489917,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00029962532675749425,
      "loss": 5.0114,
      "mean_token_accuracy": 0.22142528891563415,
      "num_tokens": 108987005.0,
      "step": 47580
    },
    {
      "entropy": 5.789889287948609,
      "epoch": 4.703934361407671,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002995902484679586,
      "loss": 4.9208,
      "mean_token_accuracy": 0.22667604982852935,
      "num_tokens": 108998353.0,
      "step": 47585
    },
    {
      "entropy": 5.878598880767822,
      "epoch": 4.7044286279161724,
      "grad_norm": 1.328125,
      "learning_rate": 0.000299555169573499,
      "loss": 5.1309,
      "mean_token_accuracy": 0.22114192694425583,
      "num_tokens": 109010991.0,
      "step": 47590
    },
    {
      "entropy": 5.776914167404175,
      "epoch": 4.704922894424674,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002995200900749784,
      "loss": 4.8383,
      "mean_token_accuracy": 0.2362937346100807,
      "num_tokens": 109021850.0,
      "step": 47595
    },
    {
      "entropy": 5.8423261642456055,
      "epoch": 4.705417160933175,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00029948500997325974,
      "loss": 4.9823,
      "mean_token_accuracy": 0.2264585167169571,
      "num_tokens": 109032900.0,
      "step": 47600
    },
    {
      "entropy": 5.815677547454834,
      "epoch": 4.705911427441676,
      "grad_norm": 1.40625,
      "learning_rate": 0.00029944992926920604,
      "loss": 5.0111,
      "mean_token_accuracy": 0.22767894715070724,
      "num_tokens": 109043861.0,
      "step": 47605
    },
    {
      "entropy": 5.7531952381134035,
      "epoch": 4.7064056939501775,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002994148479636803,
      "loss": 4.8749,
      "mean_token_accuracy": 0.22969327569007875,
      "num_tokens": 109054662.0,
      "step": 47610
    },
    {
      "entropy": 5.745825719833374,
      "epoch": 4.70689996045868,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002993797660575456,
      "loss": 4.8384,
      "mean_token_accuracy": 0.24698904007673264,
      "num_tokens": 109065767.0,
      "step": 47615
    },
    {
      "entropy": 5.760192346572876,
      "epoch": 4.707394226967181,
      "grad_norm": 1.25,
      "learning_rate": 0.00029934468355166476,
      "loss": 4.9693,
      "mean_token_accuracy": 0.2245113432407379,
      "num_tokens": 109076984.0,
      "step": 47620
    },
    {
      "entropy": 5.807067060470581,
      "epoch": 4.707888493475682,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002993096004469009,
      "loss": 4.9363,
      "mean_token_accuracy": 0.2207971140742302,
      "num_tokens": 109089105.0,
      "step": 47625
    },
    {
      "entropy": 5.834498739242553,
      "epoch": 4.708382759984183,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002992745167441172,
      "loss": 4.9532,
      "mean_token_accuracy": 0.22714738845825194,
      "num_tokens": 109100918.0,
      "step": 47630
    },
    {
      "entropy": 5.806296682357788,
      "epoch": 4.708877026492685,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00029923943244417663,
      "loss": 4.9078,
      "mean_token_accuracy": 0.23116590231657028,
      "num_tokens": 109112304.0,
      "step": 47635
    },
    {
      "entropy": 5.782422256469727,
      "epoch": 4.709371293001186,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00029920434754794216,
      "loss": 4.9864,
      "mean_token_accuracy": 0.23077698200941085,
      "num_tokens": 109122651.0,
      "step": 47640
    },
    {
      "entropy": 5.730442476272583,
      "epoch": 4.709865559509687,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002991692620562771,
      "loss": 4.8151,
      "mean_token_accuracy": 0.23456472307443618,
      "num_tokens": 109135064.0,
      "step": 47645
    },
    {
      "entropy": 5.789420890808105,
      "epoch": 4.710359826018189,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002991341759700443,
      "loss": 4.9548,
      "mean_token_accuracy": 0.23002425134181975,
      "num_tokens": 109148202.0,
      "step": 47650
    },
    {
      "entropy": 5.791964721679688,
      "epoch": 4.710854092526691,
      "grad_norm": 1.21875,
      "learning_rate": 0.00029909908929010715,
      "loss": 4.9165,
      "mean_token_accuracy": 0.23380051404237748,
      "num_tokens": 109160867.0,
      "step": 47655
    },
    {
      "entropy": 5.7236556053161625,
      "epoch": 4.711348359035192,
      "grad_norm": 1.25,
      "learning_rate": 0.0002990640020173286,
      "loss": 4.9109,
      "mean_token_accuracy": 0.22859922498464585,
      "num_tokens": 109173107.0,
      "step": 47660
    },
    {
      "entropy": 5.750639581680298,
      "epoch": 4.711842625543693,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00029902891415257196,
      "loss": 4.9391,
      "mean_token_accuracy": 0.22540190815925598,
      "num_tokens": 109184148.0,
      "step": 47665
    },
    {
      "entropy": 5.779976320266724,
      "epoch": 4.712336892052194,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00029899382569670025,
      "loss": 4.8291,
      "mean_token_accuracy": 0.2380516678094864,
      "num_tokens": 109194647.0,
      "step": 47670
    },
    {
      "entropy": 5.8154045104980465,
      "epoch": 4.712831158560696,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00029895873665057663,
      "loss": 4.953,
      "mean_token_accuracy": 0.2364491879940033,
      "num_tokens": 109206495.0,
      "step": 47675
    },
    {
      "entropy": 5.778707790374756,
      "epoch": 4.713325425069197,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002989236470150644,
      "loss": 4.9636,
      "mean_token_accuracy": 0.2275765359401703,
      "num_tokens": 109217131.0,
      "step": 47680
    },
    {
      "entropy": 5.718992137908936,
      "epoch": 4.713819691577699,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002988885567910269,
      "loss": 4.9023,
      "mean_token_accuracy": 0.23317519724369049,
      "num_tokens": 109228785.0,
      "step": 47685
    },
    {
      "entropy": 5.816817855834961,
      "epoch": 4.7143139580862,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00029885346597932703,
      "loss": 4.9616,
      "mean_token_accuracy": 0.22616491764783858,
      "num_tokens": 109240821.0,
      "step": 47690
    },
    {
      "entropy": 5.802253532409668,
      "epoch": 4.714808224594702,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002988183745808282,
      "loss": 4.9147,
      "mean_token_accuracy": 0.23629493564367293,
      "num_tokens": 109251982.0,
      "step": 47695
    },
    {
      "entropy": 5.75986762046814,
      "epoch": 4.715302491103203,
      "grad_norm": 1.40625,
      "learning_rate": 0.00029878328259639377,
      "loss": 4.9598,
      "mean_token_accuracy": 0.22361217439174652,
      "num_tokens": 109263176.0,
      "step": 47700
    },
    {
      "entropy": 5.74373950958252,
      "epoch": 4.715796757611704,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002987481900268868,
      "loss": 4.9484,
      "mean_token_accuracy": 0.2228730872273445,
      "num_tokens": 109275076.0,
      "step": 47705
    },
    {
      "entropy": 5.829935598373413,
      "epoch": 4.716291024120205,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00029871309687317066,
      "loss": 4.9589,
      "mean_token_accuracy": 0.2250740945339203,
      "num_tokens": 109285872.0,
      "step": 47710
    },
    {
      "entropy": 5.818625593185425,
      "epoch": 4.716785290628707,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002986780031361087,
      "loss": 4.9686,
      "mean_token_accuracy": 0.22753921151161194,
      "num_tokens": 109298268.0,
      "step": 47715
    },
    {
      "entropy": 5.8017435550689695,
      "epoch": 4.717279557137209,
      "grad_norm": 1.296875,
      "learning_rate": 0.00029864290881656427,
      "loss": 4.9348,
      "mean_token_accuracy": 0.22436753064393997,
      "num_tokens": 109309739.0,
      "step": 47720
    },
    {
      "entropy": 5.726105833053589,
      "epoch": 4.71777382364571,
      "grad_norm": 1.3125,
      "learning_rate": 0.00029860781391540046,
      "loss": 4.8924,
      "mean_token_accuracy": 0.23759746849536895,
      "num_tokens": 109321876.0,
      "step": 47725
    },
    {
      "entropy": 5.84837121963501,
      "epoch": 4.718268090154211,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002985727184334809,
      "loss": 4.9115,
      "mean_token_accuracy": 0.2332331657409668,
      "num_tokens": 109333112.0,
      "step": 47730
    },
    {
      "entropy": 5.761255359649658,
      "epoch": 4.718762356662713,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002985376223716688,
      "loss": 4.9378,
      "mean_token_accuracy": 0.2280514121055603,
      "num_tokens": 109344590.0,
      "step": 47735
    },
    {
      "entropy": 5.732264661788941,
      "epoch": 4.719256623171214,
      "grad_norm": 1.328125,
      "learning_rate": 0.00029850252573082744,
      "loss": 4.9082,
      "mean_token_accuracy": 0.2339516818523407,
      "num_tokens": 109355824.0,
      "step": 47740
    },
    {
      "entropy": 5.753519487380982,
      "epoch": 4.719750889679715,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002984674285118203,
      "loss": 4.8412,
      "mean_token_accuracy": 0.23846886157989503,
      "num_tokens": 109366753.0,
      "step": 47745
    },
    {
      "entropy": 5.760152673721313,
      "epoch": 4.720245156188216,
      "grad_norm": 1.265625,
      "learning_rate": 0.00029843233071551086,
      "loss": 4.9244,
      "mean_token_accuracy": 0.22944518625736238,
      "num_tokens": 109378279.0,
      "step": 47750
    },
    {
      "entropy": 5.7465921401977536,
      "epoch": 4.7207394226967185,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002983972323427624,
      "loss": 4.957,
      "mean_token_accuracy": 0.22748399078845977,
      "num_tokens": 109390485.0,
      "step": 47755
    },
    {
      "entropy": 5.837819576263428,
      "epoch": 4.72123368920522,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002983621333944384,
      "loss": 4.9837,
      "mean_token_accuracy": 0.2216181293129921,
      "num_tokens": 109402245.0,
      "step": 47760
    },
    {
      "entropy": 5.8557665824890135,
      "epoch": 4.721727955713721,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002983270338714023,
      "loss": 5.0042,
      "mean_token_accuracy": 0.22224336117506027,
      "num_tokens": 109413531.0,
      "step": 47765
    },
    {
      "entropy": 5.797545003890991,
      "epoch": 4.722222222222222,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002982919337745175,
      "loss": 4.9778,
      "mean_token_accuracy": 0.23266494572162627,
      "num_tokens": 109424959.0,
      "step": 47770
    },
    {
      "entropy": 5.833150959014892,
      "epoch": 4.722716488730724,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002982568331046476,
      "loss": 4.9601,
      "mean_token_accuracy": 0.2265777826309204,
      "num_tokens": 109436981.0,
      "step": 47775
    },
    {
      "entropy": 5.7705484390258786,
      "epoch": 4.723210755239225,
      "grad_norm": 1.328125,
      "learning_rate": 0.00029822173186265597,
      "loss": 4.8825,
      "mean_token_accuracy": 0.23940133303403854,
      "num_tokens": 109448948.0,
      "step": 47780
    },
    {
      "entropy": 5.7298290729522705,
      "epoch": 4.723705021747726,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00029818663004940613,
      "loss": 4.8985,
      "mean_token_accuracy": 0.23255247473716736,
      "num_tokens": 109459729.0,
      "step": 47785
    },
    {
      "entropy": 5.780205297470093,
      "epoch": 4.724199288256228,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002981515276657616,
      "loss": 4.8932,
      "mean_token_accuracy": 0.23434288650751114,
      "num_tokens": 109471283.0,
      "step": 47790
    },
    {
      "entropy": 5.726873111724854,
      "epoch": 4.7246935547647295,
      "grad_norm": 1.5,
      "learning_rate": 0.0002981164247125859,
      "loss": 4.7543,
      "mean_token_accuracy": 0.24842661768198013,
      "num_tokens": 109482626.0,
      "step": 47795
    },
    {
      "entropy": 5.814235591888428,
      "epoch": 4.725187821273231,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002980813211907426,
      "loss": 4.931,
      "mean_token_accuracy": 0.230115170776844,
      "num_tokens": 109495072.0,
      "step": 47800
    },
    {
      "entropy": 5.769908094406128,
      "epoch": 4.725682087781732,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00029804621710109513,
      "loss": 4.9744,
      "mean_token_accuracy": 0.233517324924469,
      "num_tokens": 109506104.0,
      "step": 47805
    },
    {
      "entropy": 5.739403009414673,
      "epoch": 4.726176354290233,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002980111124445072,
      "loss": 4.8629,
      "mean_token_accuracy": 0.24353299736976625,
      "num_tokens": 109517001.0,
      "step": 47810
    },
    {
      "entropy": 5.764806127548217,
      "epoch": 4.726670620798735,
      "grad_norm": 1.34375,
      "learning_rate": 0.00029797600722184225,
      "loss": 4.8738,
      "mean_token_accuracy": 0.24242452532052994,
      "num_tokens": 109528432.0,
      "step": 47815
    },
    {
      "entropy": 5.779729795455933,
      "epoch": 4.727164887307236,
      "grad_norm": 1.25,
      "learning_rate": 0.00029794090143396405,
      "loss": 4.9572,
      "mean_token_accuracy": 0.23355347365140916,
      "num_tokens": 109539584.0,
      "step": 47820
    },
    {
      "entropy": 5.720501470565796,
      "epoch": 4.727659153815737,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002979057950817361,
      "loss": 4.7928,
      "mean_token_accuracy": 0.2456271231174469,
      "num_tokens": 109550795.0,
      "step": 47825
    },
    {
      "entropy": 5.737788343429566,
      "epoch": 4.728153420324238,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002978706881660221,
      "loss": 4.9155,
      "mean_token_accuracy": 0.23036651015281678,
      "num_tokens": 109561741.0,
      "step": 47830
    },
    {
      "entropy": 5.758519458770752,
      "epoch": 4.7286476868327405,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002978355806876856,
      "loss": 4.9322,
      "mean_token_accuracy": 0.22955139130353927,
      "num_tokens": 109574246.0,
      "step": 47835
    },
    {
      "entropy": 5.852679634094239,
      "epoch": 4.729141953341242,
      "grad_norm": 1.375,
      "learning_rate": 0.00029780047264759015,
      "loss": 4.9474,
      "mean_token_accuracy": 0.2323753535747528,
      "num_tokens": 109584133.0,
      "step": 47840
    },
    {
      "entropy": 5.7726890563964846,
      "epoch": 4.729636219849743,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002977653640465996,
      "loss": 4.8921,
      "mean_token_accuracy": 0.23043569922447205,
      "num_tokens": 109596128.0,
      "step": 47845
    },
    {
      "entropy": 5.714256286621094,
      "epoch": 4.730130486358244,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002977302548855776,
      "loss": 4.8963,
      "mean_token_accuracy": 0.23256790935993193,
      "num_tokens": 109608309.0,
      "step": 47850
    },
    {
      "entropy": 5.761639738082886,
      "epoch": 4.7306247528667456,
      "grad_norm": 1.46875,
      "learning_rate": 0.00029769514516538774,
      "loss": 4.929,
      "mean_token_accuracy": 0.23441541343927383,
      "num_tokens": 109620153.0,
      "step": 47855
    },
    {
      "entropy": 5.7831024646759035,
      "epoch": 4.731119019375247,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002976600348868939,
      "loss": 4.8653,
      "mean_token_accuracy": 0.2412390500307083,
      "num_tokens": 109630775.0,
      "step": 47860
    },
    {
      "entropy": 5.8199068069458,
      "epoch": 4.731613285883748,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002976249240509595,
      "loss": 4.8938,
      "mean_token_accuracy": 0.23597402721643448,
      "num_tokens": 109641623.0,
      "step": 47865
    },
    {
      "entropy": 5.731820821762085,
      "epoch": 4.73210755239225,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002975898126584486,
      "loss": 4.9027,
      "mean_token_accuracy": 0.2401573270559311,
      "num_tokens": 109652071.0,
      "step": 47870
    },
    {
      "entropy": 5.800443887710571,
      "epoch": 4.7326018189007515,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002975547007102248,
      "loss": 4.9075,
      "mean_token_accuracy": 0.22722297310829162,
      "num_tokens": 109662734.0,
      "step": 47875
    },
    {
      "entropy": 5.79953293800354,
      "epoch": 4.733096085409253,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002975195882071519,
      "loss": 5.0085,
      "mean_token_accuracy": 0.22036950886249543,
      "num_tokens": 109674419.0,
      "step": 47880
    },
    {
      "entropy": 5.800833082199096,
      "epoch": 4.733590351917754,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002974844751500935,
      "loss": 4.8919,
      "mean_token_accuracy": 0.23005912750959395,
      "num_tokens": 109686442.0,
      "step": 47885
    },
    {
      "entropy": 5.843943071365357,
      "epoch": 4.734084618426255,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002974493615399136,
      "loss": 4.9378,
      "mean_token_accuracy": 0.23649304807186128,
      "num_tokens": 109698191.0,
      "step": 47890
    },
    {
      "entropy": 5.781586599349976,
      "epoch": 4.7345788849347565,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00029741424737747595,
      "loss": 4.9634,
      "mean_token_accuracy": 0.22625308632850646,
      "num_tokens": 109709822.0,
      "step": 47895
    },
    {
      "entropy": 5.837997913360596,
      "epoch": 4.735073151443258,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002973791326636443,
      "loss": 4.9948,
      "mean_token_accuracy": 0.22127400189638138,
      "num_tokens": 109720638.0,
      "step": 47900
    },
    {
      "entropy": 5.812130355834961,
      "epoch": 4.73556741795176,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002973440173992826,
      "loss": 4.9357,
      "mean_token_accuracy": 0.2291085436940193,
      "num_tokens": 109731678.0,
      "step": 47905
    },
    {
      "entropy": 5.845504188537598,
      "epoch": 4.736061684460261,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00029730890158525446,
      "loss": 4.9668,
      "mean_token_accuracy": 0.22876352369785308,
      "num_tokens": 109742089.0,
      "step": 47910
    },
    {
      "entropy": 5.837890720367431,
      "epoch": 4.7365559509687625,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00029727378522242395,
      "loss": 4.9429,
      "mean_token_accuracy": 0.2281051680445671,
      "num_tokens": 109754392.0,
      "step": 47915
    },
    {
      "entropy": 5.7649506568908695,
      "epoch": 4.737050217477264,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00029723866831165484,
      "loss": 4.9906,
      "mean_token_accuracy": 0.22172244191169738,
      "num_tokens": 109766019.0,
      "step": 47920
    },
    {
      "entropy": 5.736168003082275,
      "epoch": 4.737544483985765,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002972035508538112,
      "loss": 4.8867,
      "mean_token_accuracy": 0.2382505491375923,
      "num_tokens": 109776545.0,
      "step": 47925
    },
    {
      "entropy": 5.86067533493042,
      "epoch": 4.738038750494266,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002971684328497565,
      "loss": 5.0261,
      "mean_token_accuracy": 0.22011249810457229,
      "num_tokens": 109789356.0,
      "step": 47930
    },
    {
      "entropy": 5.741080379486084,
      "epoch": 4.7385330170027675,
      "grad_norm": 1.296875,
      "learning_rate": 0.00029713331430035505,
      "loss": 4.882,
      "mean_token_accuracy": 0.23431582152843475,
      "num_tokens": 109800823.0,
      "step": 47935
    },
    {
      "entropy": 5.775539255142212,
      "epoch": 4.73902728351127,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002970981952064707,
      "loss": 4.9352,
      "mean_token_accuracy": 0.2327557235956192,
      "num_tokens": 109812512.0,
      "step": 47940
    },
    {
      "entropy": 5.764524030685425,
      "epoch": 4.739521550019771,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00029706307556896723,
      "loss": 4.8705,
      "mean_token_accuracy": 0.23699752539396285,
      "num_tokens": 109824026.0,
      "step": 47945
    },
    {
      "entropy": 5.812251424789428,
      "epoch": 4.740015816528272,
      "grad_norm": 1.359375,
      "learning_rate": 0.00029702795538870874,
      "loss": 4.9772,
      "mean_token_accuracy": 0.22571152597665786,
      "num_tokens": 109834792.0,
      "step": 47950
    },
    {
      "entropy": 5.778350305557251,
      "epoch": 4.7405100830367735,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000296992834666559,
      "loss": 4.8854,
      "mean_token_accuracy": 0.23837599009275437,
      "num_tokens": 109846435.0,
      "step": 47955
    },
    {
      "entropy": 5.735034847259522,
      "epoch": 4.741004349545275,
      "grad_norm": 1.359375,
      "learning_rate": 0.00029695771340338225,
      "loss": 4.834,
      "mean_token_accuracy": 0.23706765621900558,
      "num_tokens": 109857399.0,
      "step": 47960
    },
    {
      "entropy": 5.794822549819946,
      "epoch": 4.741498616053776,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002969225916000423,
      "loss": 4.9143,
      "mean_token_accuracy": 0.23088134080171585,
      "num_tokens": 109868872.0,
      "step": 47965
    },
    {
      "entropy": 5.756534051895142,
      "epoch": 4.741992882562277,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002968874692574032,
      "loss": 4.9757,
      "mean_token_accuracy": 0.22160680443048478,
      "num_tokens": 109880724.0,
      "step": 47970
    },
    {
      "entropy": 5.868090486526489,
      "epoch": 4.742487149070779,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00029685234637632894,
      "loss": 4.9689,
      "mean_token_accuracy": 0.22669397443532943,
      "num_tokens": 109891513.0,
      "step": 47975
    },
    {
      "entropy": 5.763829898834229,
      "epoch": 4.742981415579281,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002968172229576836,
      "loss": 4.9078,
      "mean_token_accuracy": 0.23452693670988084,
      "num_tokens": 109904148.0,
      "step": 47980
    },
    {
      "entropy": 5.848297786712647,
      "epoch": 4.743475682087782,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002967820990023311,
      "loss": 4.9206,
      "mean_token_accuracy": 0.23044856488704682,
      "num_tokens": 109915835.0,
      "step": 47985
    },
    {
      "entropy": 5.781224918365479,
      "epoch": 4.743969948596283,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00029674697451113576,
      "loss": 4.9838,
      "mean_token_accuracy": 0.22415730655193328,
      "num_tokens": 109927538.0,
      "step": 47990
    },
    {
      "entropy": 5.831448411941528,
      "epoch": 4.744464215104784,
      "grad_norm": 1.390625,
      "learning_rate": 0.00029671184948496134,
      "loss": 4.9777,
      "mean_token_accuracy": 0.222627255320549,
      "num_tokens": 109938379.0,
      "step": 47995
    },
    {
      "entropy": 5.74734206199646,
      "epoch": 4.744958481613286,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002966767239246721,
      "loss": 4.8522,
      "mean_token_accuracy": 0.23739527612924577,
      "num_tokens": 109949155.0,
      "step": 48000
    },
    {
      "epoch": 4.744958481613286,
      "eval_entropy": 5.495092670152109,
      "eval_loss": 4.997913360595703,
      "eval_mean_token_accuracy": 0.2346619784337128,
      "eval_num_tokens": 109949155.0,
      "eval_runtime": 20.3323,
      "eval_samples_per_second": 1599.078,
      "eval_steps_per_second": 199.928,
      "step": 48000
    },
    {
      "entropy": 5.815223264694214,
      "epoch": 4.745452748121787,
      "grad_norm": 1.390625,
      "learning_rate": 0.00029664159783113203,
      "loss": 4.9439,
      "mean_token_accuracy": 0.22734562158584595,
      "num_tokens": 109959968.0,
      "step": 48005
    },
    {
      "entropy": 5.860291862487793,
      "epoch": 4.745947014630289,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002966064712052054,
      "loss": 5.0425,
      "mean_token_accuracy": 0.2229422450065613,
      "num_tokens": 109972247.0,
      "step": 48010
    },
    {
      "entropy": 5.822835874557495,
      "epoch": 4.74644128113879,
      "grad_norm": 1.421875,
      "learning_rate": 0.00029657134404775614,
      "loss": 4.894,
      "mean_token_accuracy": 0.22919864505529403,
      "num_tokens": 109982905.0,
      "step": 48015
    },
    {
      "entropy": 5.839215660095215,
      "epoch": 4.746935547647292,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002965362163596485,
      "loss": 5.0102,
      "mean_token_accuracy": 0.22506968230009078,
      "num_tokens": 109994155.0,
      "step": 48020
    },
    {
      "entropy": 5.788618850708008,
      "epoch": 4.747429814155793,
      "grad_norm": 1.359375,
      "learning_rate": 0.00029650108814174657,
      "loss": 4.8963,
      "mean_token_accuracy": 0.23855823576450347,
      "num_tokens": 110005642.0,
      "step": 48025
    },
    {
      "entropy": 5.75698184967041,
      "epoch": 4.747924080664294,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002964659593949146,
      "loss": 4.8804,
      "mean_token_accuracy": 0.23293833881616594,
      "num_tokens": 110016627.0,
      "step": 48030
    },
    {
      "entropy": 5.767724800109863,
      "epoch": 4.748418347172795,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00029643083012001665,
      "loss": 4.8652,
      "mean_token_accuracy": 0.23284413516521454,
      "num_tokens": 110026758.0,
      "step": 48035
    },
    {
      "entropy": 5.847751474380493,
      "epoch": 4.748912613681297,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00029639570031791696,
      "loss": 5.0288,
      "mean_token_accuracy": 0.22261364609003068,
      "num_tokens": 110038652.0,
      "step": 48040
    },
    {
      "entropy": 5.805406808853149,
      "epoch": 4.749406880189798,
      "grad_norm": 1.28125,
      "learning_rate": 0.00029636056998947973,
      "loss": 4.938,
      "mean_token_accuracy": 0.23308006525039673,
      "num_tokens": 110049483.0,
      "step": 48045
    },
    {
      "entropy": 5.7864031314849855,
      "epoch": 4.7499011466983,
      "grad_norm": 1.375,
      "learning_rate": 0.0002963254391355691,
      "loss": 4.95,
      "mean_token_accuracy": 0.2258678987622261,
      "num_tokens": 110059925.0,
      "step": 48050
    },
    {
      "entropy": 5.8283538818359375,
      "epoch": 4.750395413206801,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00029629030775704944,
      "loss": 5.0001,
      "mean_token_accuracy": 0.21990932077169417,
      "num_tokens": 110071953.0,
      "step": 48055
    },
    {
      "entropy": 5.758293485641479,
      "epoch": 4.750889679715303,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002962551758547849,
      "loss": 4.8956,
      "mean_token_accuracy": 0.23699200451374053,
      "num_tokens": 110083742.0,
      "step": 48060
    },
    {
      "entropy": 5.8650599956512455,
      "epoch": 4.751383946223804,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002962200434296397,
      "loss": 4.9866,
      "mean_token_accuracy": 0.22963346689939498,
      "num_tokens": 110094920.0,
      "step": 48065
    },
    {
      "entropy": 5.849640274047852,
      "epoch": 4.751878212732305,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002961849104824782,
      "loss": 5.0062,
      "mean_token_accuracy": 0.22333280444145204,
      "num_tokens": 110107109.0,
      "step": 48070
    },
    {
      "entropy": 5.88192138671875,
      "epoch": 4.752372479240806,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002961497770141645,
      "loss": 4.9998,
      "mean_token_accuracy": 0.22173140347003936,
      "num_tokens": 110117529.0,
      "step": 48075
    },
    {
      "entropy": 5.755938720703125,
      "epoch": 4.752866745749308,
      "grad_norm": 1.375,
      "learning_rate": 0.0002961146430255631,
      "loss": 4.8945,
      "mean_token_accuracy": 0.23653606027364732,
      "num_tokens": 110130356.0,
      "step": 48080
    },
    {
      "entropy": 5.74698314666748,
      "epoch": 4.753361012257809,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002960795085175382,
      "loss": 4.8572,
      "mean_token_accuracy": 0.2323732390999794,
      "num_tokens": 110142176.0,
      "step": 48085
    },
    {
      "entropy": 5.842442178726197,
      "epoch": 4.753855278766311,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002960443734909541,
      "loss": 5.021,
      "mean_token_accuracy": 0.22443126291036605,
      "num_tokens": 110153777.0,
      "step": 48090
    },
    {
      "entropy": 5.8104053974151615,
      "epoch": 4.754349545274812,
      "grad_norm": 1.21875,
      "learning_rate": 0.00029600923794667524,
      "loss": 4.9568,
      "mean_token_accuracy": 0.2292995572090149,
      "num_tokens": 110165346.0,
      "step": 48095
    },
    {
      "entropy": 5.832502841949463,
      "epoch": 4.754843811783314,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002959741018855658,
      "loss": 4.9026,
      "mean_token_accuracy": 0.23744392842054368,
      "num_tokens": 110175714.0,
      "step": 48100
    },
    {
      "entropy": 5.7123020648956295,
      "epoch": 4.755338078291815,
      "grad_norm": 1.40625,
      "learning_rate": 0.00029593896530849014,
      "loss": 4.9019,
      "mean_token_accuracy": 0.23093169182538986,
      "num_tokens": 110187451.0,
      "step": 48105
    },
    {
      "entropy": 5.762855768203735,
      "epoch": 4.755832344800316,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002959038282163127,
      "loss": 4.9158,
      "mean_token_accuracy": 0.2250763162970543,
      "num_tokens": 110199199.0,
      "step": 48110
    },
    {
      "entropy": 5.8097158908844,
      "epoch": 4.756326611308817,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002958686906098979,
      "loss": 4.9576,
      "mean_token_accuracy": 0.22751531451940538,
      "num_tokens": 110209936.0,
      "step": 48115
    },
    {
      "entropy": 5.745332860946656,
      "epoch": 4.756820877817319,
      "grad_norm": 1.375,
      "learning_rate": 0.00029583355249011004,
      "loss": 4.905,
      "mean_token_accuracy": 0.23148689419031143,
      "num_tokens": 110220646.0,
      "step": 48120
    },
    {
      "entropy": 5.741123151779175,
      "epoch": 4.757315144325821,
      "grad_norm": 1.28125,
      "learning_rate": 0.00029579841385781353,
      "loss": 4.8301,
      "mean_token_accuracy": 0.24511925578117372,
      "num_tokens": 110232314.0,
      "step": 48125
    },
    {
      "entropy": 5.807451152801514,
      "epoch": 4.757809410834322,
      "grad_norm": 1.125,
      "learning_rate": 0.00029576327471387285,
      "loss": 4.9192,
      "mean_token_accuracy": 0.22710827738046646,
      "num_tokens": 110245566.0,
      "step": 48130
    },
    {
      "entropy": 5.7495331287384035,
      "epoch": 4.758303677342823,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002957281350591524,
      "loss": 4.8615,
      "mean_token_accuracy": 0.24180029928684235,
      "num_tokens": 110256133.0,
      "step": 48135
    },
    {
      "entropy": 5.812336111068726,
      "epoch": 4.758797943851325,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002956929948945166,
      "loss": 4.9506,
      "mean_token_accuracy": 0.23354557752609253,
      "num_tokens": 110268080.0,
      "step": 48140
    },
    {
      "entropy": 5.766601610183716,
      "epoch": 4.759292210359826,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00029565785422082993,
      "loss": 4.9396,
      "mean_token_accuracy": 0.22591589987277985,
      "num_tokens": 110278940.0,
      "step": 48145
    },
    {
      "entropy": 5.7701996803283695,
      "epoch": 4.759786476868327,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002956227130389568,
      "loss": 4.9317,
      "mean_token_accuracy": 0.22912800759077073,
      "num_tokens": 110289610.0,
      "step": 48150
    },
    {
      "entropy": 5.780746793746948,
      "epoch": 4.760280743376828,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002955875713497617,
      "loss": 4.904,
      "mean_token_accuracy": 0.23255961239337922,
      "num_tokens": 110300563.0,
      "step": 48155
    },
    {
      "entropy": 5.829084491729736,
      "epoch": 4.7607750098853305,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00029555242915410925,
      "loss": 4.9784,
      "mean_token_accuracy": 0.23344335407018663,
      "num_tokens": 110312352.0,
      "step": 48160
    },
    {
      "entropy": 5.784983921051025,
      "epoch": 4.761269276393832,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002955172864528638,
      "loss": 4.9463,
      "mean_token_accuracy": 0.23354115337133408,
      "num_tokens": 110323964.0,
      "step": 48165
    },
    {
      "entropy": 5.766483402252197,
      "epoch": 4.761763542902333,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00029548214324688995,
      "loss": 4.9613,
      "mean_token_accuracy": 0.2292727544903755,
      "num_tokens": 110337369.0,
      "step": 48170
    },
    {
      "entropy": 5.7960272312164305,
      "epoch": 4.762257809410834,
      "grad_norm": 1.359375,
      "learning_rate": 0.00029544699953705213,
      "loss": 4.9452,
      "mean_token_accuracy": 0.2351111114025116,
      "num_tokens": 110349554.0,
      "step": 48175
    },
    {
      "entropy": 5.778850507736206,
      "epoch": 4.762752075919336,
      "grad_norm": 1.421875,
      "learning_rate": 0.000295411855324215,
      "loss": 4.9069,
      "mean_token_accuracy": 0.22932820916175842,
      "num_tokens": 110360254.0,
      "step": 48180
    },
    {
      "entropy": 5.793631744384766,
      "epoch": 4.763246342427837,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002953767106092431,
      "loss": 4.8869,
      "mean_token_accuracy": 0.22682023793458939,
      "num_tokens": 110371765.0,
      "step": 48185
    },
    {
      "entropy": 5.754798555374146,
      "epoch": 4.763740608936338,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002953415653930009,
      "loss": 4.8855,
      "mean_token_accuracy": 0.23610733449459076,
      "num_tokens": 110383061.0,
      "step": 48190
    },
    {
      "entropy": 5.835553455352783,
      "epoch": 4.76423487544484,
      "grad_norm": 1.171875,
      "learning_rate": 0.00029530641967635303,
      "loss": 4.9996,
      "mean_token_accuracy": 0.22661035507917404,
      "num_tokens": 110395851.0,
      "step": 48195
    },
    {
      "entropy": 5.851271915435791,
      "epoch": 4.7647291419533415,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002952712734601641,
      "loss": 5.0046,
      "mean_token_accuracy": 0.22413257360458375,
      "num_tokens": 110408221.0,
      "step": 48200
    },
    {
      "entropy": 5.754383754730225,
      "epoch": 4.765223408461843,
      "grad_norm": 1.21875,
      "learning_rate": 0.00029523612674529876,
      "loss": 4.9157,
      "mean_token_accuracy": 0.23201589584350585,
      "num_tokens": 110420390.0,
      "step": 48205
    },
    {
      "entropy": 5.790556573867798,
      "epoch": 4.765717674970344,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00029520097953262154,
      "loss": 4.9483,
      "mean_token_accuracy": 0.22475413531064986,
      "num_tokens": 110431365.0,
      "step": 48210
    },
    {
      "entropy": 5.7308553695678714,
      "epoch": 4.766211941478845,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000295165831822997,
      "loss": 4.8642,
      "mean_token_accuracy": 0.23766261488199233,
      "num_tokens": 110442116.0,
      "step": 48215
    },
    {
      "entropy": 5.691875219345093,
      "epoch": 4.766706207987347,
      "grad_norm": 1.25,
      "learning_rate": 0.0002951306836172899,
      "loss": 4.7975,
      "mean_token_accuracy": 0.24256475567817687,
      "num_tokens": 110453015.0,
      "step": 48220
    },
    {
      "entropy": 5.75885591506958,
      "epoch": 4.767200474495848,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00029509553491636494,
      "loss": 4.9363,
      "mean_token_accuracy": 0.23123945593833922,
      "num_tokens": 110464957.0,
      "step": 48225
    },
    {
      "entropy": 5.8461285591125485,
      "epoch": 4.76769474100435,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002950603857210867,
      "loss": 5.0141,
      "mean_token_accuracy": 0.22275579422712327,
      "num_tokens": 110477179.0,
      "step": 48230
    },
    {
      "entropy": 5.819657802581787,
      "epoch": 4.768189007512851,
      "grad_norm": 1.296875,
      "learning_rate": 0.00029502523603231975,
      "loss": 4.8995,
      "mean_token_accuracy": 0.2339450240135193,
      "num_tokens": 110488235.0,
      "step": 48235
    },
    {
      "entropy": 5.739092826843262,
      "epoch": 4.7686832740213525,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000294990085850929,
      "loss": 4.8591,
      "mean_token_accuracy": 0.2429361879825592,
      "num_tokens": 110501083.0,
      "step": 48240
    },
    {
      "entropy": 5.8375,
      "epoch": 4.769177540529854,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002949549351777789,
      "loss": 4.9498,
      "mean_token_accuracy": 0.22401072531938554,
      "num_tokens": 110511819.0,
      "step": 48245
    },
    {
      "entropy": 5.828218126296997,
      "epoch": 4.769671807038355,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00029491978401373435,
      "loss": 4.9394,
      "mean_token_accuracy": 0.2392090827226639,
      "num_tokens": 110522243.0,
      "step": 48250
    },
    {
      "entropy": 5.786455678939819,
      "epoch": 4.770166073546856,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002948846323596601,
      "loss": 4.9621,
      "mean_token_accuracy": 0.22299021035432814,
      "num_tokens": 110533069.0,
      "step": 48255
    },
    {
      "entropy": 5.769337606430054,
      "epoch": 4.7706603400553576,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00029484948021642085,
      "loss": 4.9494,
      "mean_token_accuracy": 0.2294630229473114,
      "num_tokens": 110543740.0,
      "step": 48260
    },
    {
      "entropy": 5.844870615005493,
      "epoch": 4.77115460656386,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00029481432758488106,
      "loss": 4.9747,
      "mean_token_accuracy": 0.21788849383592607,
      "num_tokens": 110555335.0,
      "step": 48265
    },
    {
      "entropy": 5.807349348068238,
      "epoch": 4.771648873072361,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000294779174465906,
      "loss": 4.9251,
      "mean_token_accuracy": 0.2316594660282135,
      "num_tokens": 110566645.0,
      "step": 48270
    },
    {
      "entropy": 5.795274877548218,
      "epoch": 4.772143139580862,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002947440208603601,
      "loss": 4.9156,
      "mean_token_accuracy": 0.23267343938350676,
      "num_tokens": 110578001.0,
      "step": 48275
    },
    {
      "entropy": 5.804475259780884,
      "epoch": 4.7726374060893635,
      "grad_norm": 1.28125,
      "learning_rate": 0.00029470886676910817,
      "loss": 4.9601,
      "mean_token_accuracy": 0.23293033838272095,
      "num_tokens": 110589432.0,
      "step": 48280
    },
    {
      "entropy": 5.767077875137329,
      "epoch": 4.773131672597865,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002946737121930151,
      "loss": 4.865,
      "mean_token_accuracy": 0.23637205362319946,
      "num_tokens": 110601213.0,
      "step": 48285
    },
    {
      "entropy": 5.744436502456665,
      "epoch": 4.773625939106366,
      "grad_norm": 1.328125,
      "learning_rate": 0.00029463855713294563,
      "loss": 4.912,
      "mean_token_accuracy": 0.2326609954237938,
      "num_tokens": 110612898.0,
      "step": 48290
    },
    {
      "entropy": 5.724813032150268,
      "epoch": 4.774120205614867,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00029460340158976464,
      "loss": 4.9013,
      "mean_token_accuracy": 0.23118477314710617,
      "num_tokens": 110624345.0,
      "step": 48295
    },
    {
      "entropy": 5.804712295532227,
      "epoch": 4.7746144721233685,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002945682455643369,
      "loss": 4.9098,
      "mean_token_accuracy": 0.23064847737550737,
      "num_tokens": 110635758.0,
      "step": 48300
    },
    {
      "entropy": 5.800288963317871,
      "epoch": 4.775108738631871,
      "grad_norm": 1.5,
      "learning_rate": 0.0002945330890575273,
      "loss": 4.947,
      "mean_token_accuracy": 0.23002690821886063,
      "num_tokens": 110647641.0,
      "step": 48305
    },
    {
      "entropy": 5.787571716308594,
      "epoch": 4.775603005140372,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002944979320702007,
      "loss": 4.9468,
      "mean_token_accuracy": 0.23017821758985518,
      "num_tokens": 110659498.0,
      "step": 48310
    },
    {
      "entropy": 5.784953927993774,
      "epoch": 4.776097271648873,
      "grad_norm": 1.25,
      "learning_rate": 0.0002944627746032219,
      "loss": 4.9201,
      "mean_token_accuracy": 0.22933128476142883,
      "num_tokens": 110671303.0,
      "step": 48315
    },
    {
      "entropy": 5.836751556396484,
      "epoch": 4.7765915381573745,
      "grad_norm": 1.1875,
      "learning_rate": 0.00029442761665745583,
      "loss": 5.0158,
      "mean_token_accuracy": 0.22049447447061538,
      "num_tokens": 110684377.0,
      "step": 48320
    },
    {
      "entropy": 5.754684972763061,
      "epoch": 4.777085804665876,
      "grad_norm": 2.140625,
      "learning_rate": 0.00029439245823376743,
      "loss": 4.8962,
      "mean_token_accuracy": 0.2303997904062271,
      "num_tokens": 110697433.0,
      "step": 48325
    },
    {
      "entropy": 5.864385414123535,
      "epoch": 4.777580071174377,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002943572993330215,
      "loss": 5.0253,
      "mean_token_accuracy": 0.22800776809453965,
      "num_tokens": 110709178.0,
      "step": 48330
    },
    {
      "entropy": 5.814106798171997,
      "epoch": 4.778074337682878,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002943221399560831,
      "loss": 4.9305,
      "mean_token_accuracy": 0.22009014785289766,
      "num_tokens": 110719512.0,
      "step": 48335
    },
    {
      "entropy": 5.780506896972656,
      "epoch": 4.7785686041913795,
      "grad_norm": 1.25,
      "learning_rate": 0.00029428698010381695,
      "loss": 4.8522,
      "mean_token_accuracy": 0.23838422149419786,
      "num_tokens": 110730726.0,
      "step": 48340
    },
    {
      "entropy": 5.7862035751342775,
      "epoch": 4.779062870699882,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00029425181977708816,
      "loss": 5.0049,
      "mean_token_accuracy": 0.22065076380968093,
      "num_tokens": 110743767.0,
      "step": 48345
    },
    {
      "entropy": 5.716965866088867,
      "epoch": 4.779557137208383,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00029421665897676154,
      "loss": 4.813,
      "mean_token_accuracy": 0.24135451763868332,
      "num_tokens": 110753744.0,
      "step": 48350
    },
    {
      "entropy": 5.821961402893066,
      "epoch": 4.780051403716884,
      "grad_norm": 1.28125,
      "learning_rate": 0.00029418149770370217,
      "loss": 5.0581,
      "mean_token_accuracy": 0.22011152803897857,
      "num_tokens": 110766036.0,
      "step": 48355
    },
    {
      "entropy": 5.771739435195923,
      "epoch": 4.7805456702253855,
      "grad_norm": 1.3125,
      "learning_rate": 0.00029414633595877497,
      "loss": 4.8801,
      "mean_token_accuracy": 0.23368316888809204,
      "num_tokens": 110778739.0,
      "step": 48360
    },
    {
      "entropy": 5.768290042877197,
      "epoch": 4.781039936733887,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00029411117374284493,
      "loss": 4.8913,
      "mean_token_accuracy": 0.24399167597293853,
      "num_tokens": 110790675.0,
      "step": 48365
    },
    {
      "entropy": 5.719346570968628,
      "epoch": 4.781534203242388,
      "grad_norm": 1.21875,
      "learning_rate": 0.00029407601105677707,
      "loss": 4.9572,
      "mean_token_accuracy": 0.22749950736761093,
      "num_tokens": 110803807.0,
      "step": 48370
    },
    {
      "entropy": 5.697475862503052,
      "epoch": 4.782028469750889,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002940408479014364,
      "loss": 4.7803,
      "mean_token_accuracy": 0.25386126935482023,
      "num_tokens": 110815543.0,
      "step": 48375
    },
    {
      "entropy": 5.841625738143921,
      "epoch": 4.782522736259391,
      "grad_norm": 1.421875,
      "learning_rate": 0.00029400568427768786,
      "loss": 4.9112,
      "mean_token_accuracy": 0.229024076461792,
      "num_tokens": 110826919.0,
      "step": 48380
    },
    {
      "entropy": 5.772989749908447,
      "epoch": 4.783017002767893,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00029397052018639657,
      "loss": 4.9061,
      "mean_token_accuracy": 0.22511757910251617,
      "num_tokens": 110839314.0,
      "step": 48385
    },
    {
      "entropy": 5.819831418991089,
      "epoch": 4.783511269276394,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002939353556284276,
      "loss": 4.9702,
      "mean_token_accuracy": 0.22306977063417435,
      "num_tokens": 110850762.0,
      "step": 48390
    },
    {
      "entropy": 5.709316778182983,
      "epoch": 4.784005535784895,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002939001906046458,
      "loss": 4.84,
      "mean_token_accuracy": 0.23457081168889998,
      "num_tokens": 110862177.0,
      "step": 48395
    },
    {
      "entropy": 5.845559692382812,
      "epoch": 4.784499802293396,
      "grad_norm": 1.484375,
      "learning_rate": 0.00029386502511591654,
      "loss": 5.006,
      "mean_token_accuracy": 0.22857406437397004,
      "num_tokens": 110873647.0,
      "step": 48400
    },
    {
      "entropy": 5.820308065414428,
      "epoch": 4.784994068801898,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00029382985916310466,
      "loss": 4.9888,
      "mean_token_accuracy": 0.21962386816740037,
      "num_tokens": 110885996.0,
      "step": 48405
    },
    {
      "entropy": 5.703316593170166,
      "epoch": 4.785488335310399,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00029379469274707525,
      "loss": 4.8289,
      "mean_token_accuracy": 0.23729819804430008,
      "num_tokens": 110897600.0,
      "step": 48410
    },
    {
      "entropy": 5.811141633987427,
      "epoch": 4.785982601818901,
      "grad_norm": 1.3125,
      "learning_rate": 0.00029375952586869363,
      "loss": 4.9427,
      "mean_token_accuracy": 0.2299438863992691,
      "num_tokens": 110909853.0,
      "step": 48415
    },
    {
      "entropy": 5.78523998260498,
      "epoch": 4.786476868327402,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002937243585288247,
      "loss": 4.9566,
      "mean_token_accuracy": 0.22216263115406037,
      "num_tokens": 110921938.0,
      "step": 48420
    },
    {
      "entropy": 5.8564962387084964,
      "epoch": 4.786971134835904,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002936891907283336,
      "loss": 5.0192,
      "mean_token_accuracy": 0.22907274812459946,
      "num_tokens": 110932856.0,
      "step": 48425
    },
    {
      "entropy": 5.780996227264405,
      "epoch": 4.787465401344405,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002936540224680855,
      "loss": 4.873,
      "mean_token_accuracy": 0.24152058213949204,
      "num_tokens": 110944601.0,
      "step": 48430
    },
    {
      "entropy": 5.758243799209595,
      "epoch": 4.787959667852906,
      "grad_norm": 1.3125,
      "learning_rate": 0.00029361885374894566,
      "loss": 4.8789,
      "mean_token_accuracy": 0.24083559066057206,
      "num_tokens": 110955916.0,
      "step": 48435
    },
    {
      "entropy": 5.790015983581543,
      "epoch": 4.788453934361407,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002935836845717791,
      "loss": 4.9446,
      "mean_token_accuracy": 0.22843215316534043,
      "num_tokens": 110968068.0,
      "step": 48440
    },
    {
      "entropy": 5.725407218933105,
      "epoch": 4.788948200869909,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000293548514937451,
      "loss": 4.8935,
      "mean_token_accuracy": 0.23900855034589769,
      "num_tokens": 110979251.0,
      "step": 48445
    },
    {
      "entropy": 5.799242067337036,
      "epoch": 4.789442467378411,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002935133448468265,
      "loss": 4.9468,
      "mean_token_accuracy": 0.23031775206327437,
      "num_tokens": 110991029.0,
      "step": 48450
    },
    {
      "entropy": 5.848375511169434,
      "epoch": 4.789936733886912,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0002934781743007709,
      "loss": 4.9987,
      "mean_token_accuracy": 0.2227770984172821,
      "num_tokens": 111001500.0,
      "step": 48455
    },
    {
      "entropy": 5.833744668960572,
      "epoch": 4.790431000395413,
      "grad_norm": 1.34375,
      "learning_rate": 0.00029344300330014933,
      "loss": 4.9052,
      "mean_token_accuracy": 0.23683041781187059,
      "num_tokens": 111012838.0,
      "step": 48460
    },
    {
      "entropy": 5.7820264339447025,
      "epoch": 4.790925266903915,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000293407831845827,
      "loss": 4.9918,
      "mean_token_accuracy": 0.2264212980866432,
      "num_tokens": 111024937.0,
      "step": 48465
    },
    {
      "entropy": 5.820897626876831,
      "epoch": 4.791419533412416,
      "grad_norm": 1.328125,
      "learning_rate": 0.00029337265993866916,
      "loss": 5.0035,
      "mean_token_accuracy": 0.22224596440792083,
      "num_tokens": 111035700.0,
      "step": 48470
    },
    {
      "entropy": 5.751635026931763,
      "epoch": 4.791913799920917,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002933374875795411,
      "loss": 4.818,
      "mean_token_accuracy": 0.24409843534231185,
      "num_tokens": 111047605.0,
      "step": 48475
    },
    {
      "entropy": 5.818263816833496,
      "epoch": 4.792408066429418,
      "grad_norm": 1.34375,
      "learning_rate": 0.00029330231476930785,
      "loss": 4.94,
      "mean_token_accuracy": 0.23142216205596924,
      "num_tokens": 111059209.0,
      "step": 48480
    },
    {
      "entropy": 5.81302695274353,
      "epoch": 4.792902332937921,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000293267141508835,
      "loss": 4.9639,
      "mean_token_accuracy": 0.2209448531270027,
      "num_tokens": 111069862.0,
      "step": 48485
    },
    {
      "entropy": 5.784808206558227,
      "epoch": 4.793396599446422,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002932319677989875,
      "loss": 4.9339,
      "mean_token_accuracy": 0.22385257333517075,
      "num_tokens": 111081646.0,
      "step": 48490
    },
    {
      "entropy": 5.77811017036438,
      "epoch": 4.793890865954923,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002931967936406308,
      "loss": 4.8895,
      "mean_token_accuracy": 0.23336291313171387,
      "num_tokens": 111091763.0,
      "step": 48495
    },
    {
      "entropy": 5.749838495254517,
      "epoch": 4.794385132463424,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00029316161903463026,
      "loss": 4.895,
      "mean_token_accuracy": 0.23352765589952468,
      "num_tokens": 111103777.0,
      "step": 48500
    },
    {
      "entropy": 5.7744375705719,
      "epoch": 4.794879398971926,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00029312644398185096,
      "loss": 4.9442,
      "mean_token_accuracy": 0.23032571375370026,
      "num_tokens": 111115100.0,
      "step": 48505
    },
    {
      "entropy": 5.750958824157715,
      "epoch": 4.795373665480427,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002930912684831583,
      "loss": 4.8923,
      "mean_token_accuracy": 0.22976758033037187,
      "num_tokens": 111125563.0,
      "step": 48510
    },
    {
      "entropy": 5.761693334579467,
      "epoch": 4.795867931988928,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002930560925394177,
      "loss": 4.9321,
      "mean_token_accuracy": 0.2354413613677025,
      "num_tokens": 111136306.0,
      "step": 48515
    },
    {
      "entropy": 5.820460224151612,
      "epoch": 4.79636219849743,
      "grad_norm": 1.28125,
      "learning_rate": 0.00029302091615149447,
      "loss": 4.9638,
      "mean_token_accuracy": 0.21639739871025085,
      "num_tokens": 111148533.0,
      "step": 48520
    },
    {
      "entropy": 5.734290838241577,
      "epoch": 4.7968564650059315,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00029298573932025373,
      "loss": 4.9142,
      "mean_token_accuracy": 0.23580976873636245,
      "num_tokens": 111160285.0,
      "step": 48525
    },
    {
      "entropy": 5.729951190948486,
      "epoch": 4.797350731514433,
      "grad_norm": 1.328125,
      "learning_rate": 0.00029295056204656114,
      "loss": 4.8722,
      "mean_token_accuracy": 0.23296225517988206,
      "num_tokens": 111171803.0,
      "step": 48530
    },
    {
      "entropy": 5.816208171844482,
      "epoch": 4.797844998022934,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002929153843312819,
      "loss": 4.932,
      "mean_token_accuracy": 0.22507246136665343,
      "num_tokens": 111184819.0,
      "step": 48535
    },
    {
      "entropy": 5.751821899414063,
      "epoch": 4.798339264531435,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002928802061752814,
      "loss": 4.8715,
      "mean_token_accuracy": 0.23102896958589553,
      "num_tokens": 111195592.0,
      "step": 48540
    },
    {
      "entropy": 5.757092618942261,
      "epoch": 4.798833531039937,
      "grad_norm": 1.265625,
      "learning_rate": 0.00029284502757942515,
      "loss": 4.9312,
      "mean_token_accuracy": 0.2373213067650795,
      "num_tokens": 111207300.0,
      "step": 48545
    },
    {
      "entropy": 5.762848138809204,
      "epoch": 4.799327797548438,
      "grad_norm": 1.421875,
      "learning_rate": 0.00029280984854457835,
      "loss": 4.9164,
      "mean_token_accuracy": 0.22977296113967896,
      "num_tokens": 111217542.0,
      "step": 48550
    },
    {
      "entropy": 5.740937566757202,
      "epoch": 4.799822064056939,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00029277466907160653,
      "loss": 4.8462,
      "mean_token_accuracy": 0.23858896493911744,
      "num_tokens": 111229422.0,
      "step": 48555
    },
    {
      "entropy": 5.8081145763397215,
      "epoch": 4.800316330565441,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002927394891613752,
      "loss": 4.9984,
      "mean_token_accuracy": 0.22640254497528076,
      "num_tokens": 111241547.0,
      "step": 48560
    },
    {
      "entropy": 5.772748947143555,
      "epoch": 4.8008105970739425,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002927043088147496,
      "loss": 4.9083,
      "mean_token_accuracy": 0.2349537044763565,
      "num_tokens": 111252123.0,
      "step": 48565
    },
    {
      "entropy": 5.765936660766601,
      "epoch": 4.801304863582444,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002926691280325953,
      "loss": 4.9163,
      "mean_token_accuracy": 0.2331869900226593,
      "num_tokens": 111263862.0,
      "step": 48570
    },
    {
      "entropy": 5.778378105163574,
      "epoch": 4.801799130090945,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00029263394681577757,
      "loss": 4.9468,
      "mean_token_accuracy": 0.23231848627328872,
      "num_tokens": 111274525.0,
      "step": 48575
    },
    {
      "entropy": 5.84856104850769,
      "epoch": 4.802293396599446,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00029259876516516217,
      "loss": 4.9655,
      "mean_token_accuracy": 0.2263644114136696,
      "num_tokens": 111284972.0,
      "step": 48580
    },
    {
      "entropy": 5.807890844345093,
      "epoch": 4.802787663107948,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002925635830816144,
      "loss": 4.9957,
      "mean_token_accuracy": 0.23001225888729096,
      "num_tokens": 111297214.0,
      "step": 48585
    },
    {
      "entropy": 5.767961692810059,
      "epoch": 4.803281929616449,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002925284005659997,
      "loss": 4.8524,
      "mean_token_accuracy": 0.23326777815818786,
      "num_tokens": 111307731.0,
      "step": 48590
    },
    {
      "entropy": 5.860457563400269,
      "epoch": 4.80377619612495,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00029249321761918364,
      "loss": 4.9794,
      "mean_token_accuracy": 0.2264727845788002,
      "num_tokens": 111319474.0,
      "step": 48595
    },
    {
      "entropy": 5.797132778167724,
      "epoch": 4.804270462633452,
      "grad_norm": 1.359375,
      "learning_rate": 0.00029245803424203173,
      "loss": 4.8552,
      "mean_token_accuracy": 0.240923647582531,
      "num_tokens": 111329821.0,
      "step": 48600
    },
    {
      "entropy": 5.781694984436035,
      "epoch": 4.8047647291419535,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002924228504354095,
      "loss": 4.9481,
      "mean_token_accuracy": 0.22714199274778366,
      "num_tokens": 111340761.0,
      "step": 48605
    },
    {
      "entropy": 5.7036017894744875,
      "epoch": 4.805258995650455,
      "grad_norm": 1.34375,
      "learning_rate": 0.00029238766620018244,
      "loss": 4.8156,
      "mean_token_accuracy": 0.2418240413069725,
      "num_tokens": 111352208.0,
      "step": 48610
    },
    {
      "entropy": 5.849536037445068,
      "epoch": 4.805753262158956,
      "grad_norm": 1.421875,
      "learning_rate": 0.000292352481537216,
      "loss": 5.0061,
      "mean_token_accuracy": 0.22162528336048126,
      "num_tokens": 111363489.0,
      "step": 48615
    },
    {
      "entropy": 5.824751949310302,
      "epoch": 4.806247528667457,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00029231729644737587,
      "loss": 4.9938,
      "mean_token_accuracy": 0.22761696726083755,
      "num_tokens": 111375791.0,
      "step": 48620
    },
    {
      "entropy": 5.818091154098511,
      "epoch": 4.806741795175959,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002922821109315275,
      "loss": 4.9133,
      "mean_token_accuracy": 0.23071672916412353,
      "num_tokens": 111387349.0,
      "step": 48625
    },
    {
      "entropy": 5.750053024291992,
      "epoch": 4.80723606168446,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00029224692499053666,
      "loss": 4.9078,
      "mean_token_accuracy": 0.23563915938138963,
      "num_tokens": 111398342.0,
      "step": 48630
    },
    {
      "entropy": 5.777785301208496,
      "epoch": 4.807730328192962,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002922117386252687,
      "loss": 4.9048,
      "mean_token_accuracy": 0.23507018387317657,
      "num_tokens": 111410208.0,
      "step": 48635
    },
    {
      "entropy": 5.806124925613403,
      "epoch": 4.808224594701463,
      "grad_norm": 1.3125,
      "learning_rate": 0.00029217655183658945,
      "loss": 4.8935,
      "mean_token_accuracy": 0.23420380502939225,
      "num_tokens": 111421273.0,
      "step": 48640
    },
    {
      "entropy": 5.798521232604981,
      "epoch": 4.8087188612099645,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002921413646253642,
      "loss": 4.9389,
      "mean_token_accuracy": 0.23665851056575776,
      "num_tokens": 111432496.0,
      "step": 48645
    },
    {
      "entropy": 5.857860708236695,
      "epoch": 4.809213127718466,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002921061769924587,
      "loss": 5.0156,
      "mean_token_accuracy": 0.22493495047092438,
      "num_tokens": 111443830.0,
      "step": 48650
    },
    {
      "entropy": 5.787855625152588,
      "epoch": 4.809707394226967,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00029207098893873864,
      "loss": 4.9359,
      "mean_token_accuracy": 0.2311509519815445,
      "num_tokens": 111455056.0,
      "step": 48655
    },
    {
      "entropy": 5.819750452041626,
      "epoch": 4.810201660735468,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00029203580046506967,
      "loss": 4.9837,
      "mean_token_accuracy": 0.22390341013669968,
      "num_tokens": 111466829.0,
      "step": 48660
    },
    {
      "entropy": 5.7724113941192625,
      "epoch": 4.8106959272439695,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00029200061157231725,
      "loss": 4.8824,
      "mean_token_accuracy": 0.23262255340814592,
      "num_tokens": 111477967.0,
      "step": 48665
    },
    {
      "entropy": 5.804455041885376,
      "epoch": 4.811190193752472,
      "grad_norm": 1.453125,
      "learning_rate": 0.00029196542226134724,
      "loss": 4.9368,
      "mean_token_accuracy": 0.23116700649261473,
      "num_tokens": 111488376.0,
      "step": 48670
    },
    {
      "entropy": 5.78844985961914,
      "epoch": 4.811684460260973,
      "grad_norm": 1.28125,
      "learning_rate": 0.00029193023253302515,
      "loss": 4.9489,
      "mean_token_accuracy": 0.23101856261491777,
      "num_tokens": 111500460.0,
      "step": 48675
    },
    {
      "entropy": 5.830862617492675,
      "epoch": 4.812178726769474,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00029189504238821677,
      "loss": 4.9645,
      "mean_token_accuracy": 0.2189592733979225,
      "num_tokens": 111511229.0,
      "step": 48680
    },
    {
      "entropy": 5.811045932769775,
      "epoch": 4.8126729932779755,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00029185985182778767,
      "loss": 4.895,
      "mean_token_accuracy": 0.2434671476483345,
      "num_tokens": 111521829.0,
      "step": 48685
    },
    {
      "entropy": 5.812684726715088,
      "epoch": 4.813167259786477,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00029182466085260364,
      "loss": 4.9121,
      "mean_token_accuracy": 0.2291320487856865,
      "num_tokens": 111533674.0,
      "step": 48690
    },
    {
      "entropy": 5.836369752883911,
      "epoch": 4.813661526294978,
      "grad_norm": 1.328125,
      "learning_rate": 0.00029178946946353037,
      "loss": 4.9889,
      "mean_token_accuracy": 0.230241996049881,
      "num_tokens": 111545588.0,
      "step": 48695
    },
    {
      "entropy": 5.8090471744537355,
      "epoch": 4.814155792803479,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002917542776614335,
      "loss": 4.8779,
      "mean_token_accuracy": 0.24001488089561462,
      "num_tokens": 111557316.0,
      "step": 48700
    },
    {
      "entropy": 5.759321784973144,
      "epoch": 4.814650059311981,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002917190854471788,
      "loss": 4.9495,
      "mean_token_accuracy": 0.23272371888160706,
      "num_tokens": 111569402.0,
      "step": 48705
    },
    {
      "entropy": 5.749652338027954,
      "epoch": 4.815144325820483,
      "grad_norm": 1.3125,
      "learning_rate": 0.00029168389282163205,
      "loss": 4.9348,
      "mean_token_accuracy": 0.23140799403190612,
      "num_tokens": 111581155.0,
      "step": 48710
    },
    {
      "entropy": 5.729934358596802,
      "epoch": 4.815638592328984,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002916486997856588,
      "loss": 4.8529,
      "mean_token_accuracy": 0.2374156266450882,
      "num_tokens": 111592566.0,
      "step": 48715
    },
    {
      "entropy": 5.810738039016724,
      "epoch": 4.816132858837485,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00029161350634012513,
      "loss": 4.955,
      "mean_token_accuracy": 0.23040008693933486,
      "num_tokens": 111604039.0,
      "step": 48720
    },
    {
      "entropy": 5.871062326431274,
      "epoch": 4.8166271253459865,
      "grad_norm": 1.328125,
      "learning_rate": 0.00029157831248589655,
      "loss": 5.0336,
      "mean_token_accuracy": 0.21991561055183412,
      "num_tokens": 111614943.0,
      "step": 48725
    },
    {
      "entropy": 5.741412019729614,
      "epoch": 4.817121391854488,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002915431182238389,
      "loss": 4.875,
      "mean_token_accuracy": 0.23893264383077623,
      "num_tokens": 111626011.0,
      "step": 48730
    },
    {
      "entropy": 5.764105319976807,
      "epoch": 4.817615658362989,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00029150792355481806,
      "loss": 4.8899,
      "mean_token_accuracy": 0.23161435425281524,
      "num_tokens": 111636980.0,
      "step": 48735
    },
    {
      "entropy": 5.790144252777099,
      "epoch": 4.818109924871491,
      "grad_norm": 1.8125,
      "learning_rate": 0.0002914727284796996,
      "loss": 4.9802,
      "mean_token_accuracy": 0.22901953905820846,
      "num_tokens": 111649603.0,
      "step": 48740
    },
    {
      "entropy": 5.806594276428223,
      "epoch": 4.818604191379992,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00029143753299934954,
      "loss": 4.9974,
      "mean_token_accuracy": 0.2225810781121254,
      "num_tokens": 111661554.0,
      "step": 48745
    },
    {
      "entropy": 5.848444700241089,
      "epoch": 4.819098457888494,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00029140233711463364,
      "loss": 4.9605,
      "mean_token_accuracy": 0.22503461241722106,
      "num_tokens": 111673938.0,
      "step": 48750
    },
    {
      "entropy": 5.807308530807495,
      "epoch": 4.819592724396995,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002913671408264177,
      "loss": 4.9169,
      "mean_token_accuracy": 0.22926753014326096,
      "num_tokens": 111685526.0,
      "step": 48755
    },
    {
      "entropy": 5.809872817993164,
      "epoch": 4.820086990905496,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002913319441355675,
      "loss": 5.0441,
      "mean_token_accuracy": 0.22487984150648116,
      "num_tokens": 111698195.0,
      "step": 48760
    },
    {
      "entropy": 5.817394065856933,
      "epoch": 4.8205812574139975,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00029129674704294896,
      "loss": 4.9714,
      "mean_token_accuracy": 0.2356329083442688,
      "num_tokens": 111708793.0,
      "step": 48765
    },
    {
      "entropy": 5.797488451004028,
      "epoch": 4.821075523922499,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000291261549549428,
      "loss": 4.9079,
      "mean_token_accuracy": 0.23432808220386506,
      "num_tokens": 111720388.0,
      "step": 48770
    },
    {
      "entropy": 5.7515308380126955,
      "epoch": 4.821569790431001,
      "grad_norm": 1.375,
      "learning_rate": 0.00029122635165587023,
      "loss": 4.9001,
      "mean_token_accuracy": 0.2344530001282692,
      "num_tokens": 111733178.0,
      "step": 48775
    },
    {
      "entropy": 5.762867164611817,
      "epoch": 4.822064056939502,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00029119115336314185,
      "loss": 4.8647,
      "mean_token_accuracy": 0.23720869421958923,
      "num_tokens": 111744131.0,
      "step": 48780
    },
    {
      "entropy": 5.804542350769043,
      "epoch": 4.822558323448003,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002911559546721085,
      "loss": 4.9472,
      "mean_token_accuracy": 0.2304277941584587,
      "num_tokens": 111755926.0,
      "step": 48785
    },
    {
      "entropy": 5.772536659240723,
      "epoch": 4.823052589956505,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00029112075558363615,
      "loss": 4.8663,
      "mean_token_accuracy": 0.23646577596664428,
      "num_tokens": 111766252.0,
      "step": 48790
    },
    {
      "entropy": 5.762499189376831,
      "epoch": 4.823546856465006,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002910855560985908,
      "loss": 4.8314,
      "mean_token_accuracy": 0.2392893150448799,
      "num_tokens": 111776703.0,
      "step": 48795
    },
    {
      "entropy": 5.712212085723877,
      "epoch": 4.824041122973507,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002910503562178382,
      "loss": 4.837,
      "mean_token_accuracy": 0.24385461658239366,
      "num_tokens": 111788026.0,
      "step": 48800
    },
    {
      "entropy": 5.7266027450561525,
      "epoch": 4.824535389482008,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00029101515594224435,
      "loss": 4.8196,
      "mean_token_accuracy": 0.24569595158100127,
      "num_tokens": 111798673.0,
      "step": 48805
    },
    {
      "entropy": 5.7184511661529545,
      "epoch": 4.82502965599051,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002909799552726752,
      "loss": 4.9206,
      "mean_token_accuracy": 0.23068345040082933,
      "num_tokens": 111810502.0,
      "step": 48810
    },
    {
      "entropy": 5.829906940460205,
      "epoch": 4.825523922499012,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00029094475420999666,
      "loss": 4.9866,
      "mean_token_accuracy": 0.2190413221716881,
      "num_tokens": 111822597.0,
      "step": 48815
    },
    {
      "entropy": 5.817552995681763,
      "epoch": 4.826018189007513,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00029090955275507473,
      "loss": 4.9587,
      "mean_token_accuracy": 0.2273958370089531,
      "num_tokens": 111833372.0,
      "step": 48820
    },
    {
      "entropy": 5.821482801437378,
      "epoch": 4.826512455516014,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00029087435090877534,
      "loss": 4.9747,
      "mean_token_accuracy": 0.2231215789914131,
      "num_tokens": 111845323.0,
      "step": 48825
    },
    {
      "entropy": 5.792683887481689,
      "epoch": 4.827006722024516,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002908391486719644,
      "loss": 4.8908,
      "mean_token_accuracy": 0.2401478484272957,
      "num_tokens": 111855953.0,
      "step": 48830
    },
    {
      "entropy": 5.793558120727539,
      "epoch": 4.827500988533017,
      "grad_norm": 1.421875,
      "learning_rate": 0.00029080394604550795,
      "loss": 4.9729,
      "mean_token_accuracy": 0.22883465886116028,
      "num_tokens": 111866433.0,
      "step": 48835
    },
    {
      "entropy": 5.811889266967773,
      "epoch": 4.827995255041518,
      "grad_norm": 1.28125,
      "learning_rate": 0.000290768743030272,
      "loss": 5.0015,
      "mean_token_accuracy": 0.2313559353351593,
      "num_tokens": 111877774.0,
      "step": 48840
    },
    {
      "entropy": 5.810166263580323,
      "epoch": 4.828489521550019,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002907335396271225,
      "loss": 4.9672,
      "mean_token_accuracy": 0.22693869322538376,
      "num_tokens": 111888966.0,
      "step": 48845
    },
    {
      "entropy": 5.81334958076477,
      "epoch": 4.828983788058521,
      "grad_norm": 1.53125,
      "learning_rate": 0.00029069833583692553,
      "loss": 4.9628,
      "mean_token_accuracy": 0.22860878854990005,
      "num_tokens": 111899571.0,
      "step": 48850
    },
    {
      "entropy": 5.758677053451538,
      "epoch": 4.829478054567023,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000290663131660547,
      "loss": 4.9622,
      "mean_token_accuracy": 0.2224608764052391,
      "num_tokens": 111911280.0,
      "step": 48855
    },
    {
      "entropy": 5.880365800857544,
      "epoch": 4.829972321075524,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000290627927098853,
      "loss": 5.0463,
      "mean_token_accuracy": 0.21710314154624938,
      "num_tokens": 111922692.0,
      "step": 48860
    },
    {
      "entropy": 5.772500896453858,
      "epoch": 4.830466587584025,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00029059272215270965,
      "loss": 4.906,
      "mean_token_accuracy": 0.23387258648872375,
      "num_tokens": 111934910.0,
      "step": 48865
    },
    {
      "entropy": 5.787146329879761,
      "epoch": 4.830960854092527,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00029055751682298295,
      "loss": 4.8561,
      "mean_token_accuracy": 0.23517969846725464,
      "num_tokens": 111945539.0,
      "step": 48870
    },
    {
      "entropy": 5.74071159362793,
      "epoch": 4.831455120601028,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002905223111105388,
      "loss": 4.8961,
      "mean_token_accuracy": 0.23512951135635377,
      "num_tokens": 111956456.0,
      "step": 48875
    },
    {
      "entropy": 5.714801073074341,
      "epoch": 4.831949387109529,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002904871050162434,
      "loss": 4.849,
      "mean_token_accuracy": 0.23729708641767502,
      "num_tokens": 111967054.0,
      "step": 48880
    },
    {
      "entropy": 5.719619083404541,
      "epoch": 4.83244365361803,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002904518985409629,
      "loss": 4.8962,
      "mean_token_accuracy": 0.2348434656858444,
      "num_tokens": 111978774.0,
      "step": 48885
    },
    {
      "entropy": 5.798677158355713,
      "epoch": 4.832937920126533,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00029041669168556315,
      "loss": 4.8781,
      "mean_token_accuracy": 0.2366304099559784,
      "num_tokens": 111989459.0,
      "step": 48890
    },
    {
      "entropy": 5.850676774978638,
      "epoch": 4.833432186635034,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002903814844509106,
      "loss": 4.9716,
      "mean_token_accuracy": 0.22540567368268966,
      "num_tokens": 112002016.0,
      "step": 48895
    },
    {
      "entropy": 5.840531730651856,
      "epoch": 4.833926453143535,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000290346276837871,
      "loss": 4.9496,
      "mean_token_accuracy": 0.2277243033051491,
      "num_tokens": 112014697.0,
      "step": 48900
    },
    {
      "entropy": 5.821752691268921,
      "epoch": 4.834420719652036,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002903110688473106,
      "loss": 4.9555,
      "mean_token_accuracy": 0.23304442018270494,
      "num_tokens": 112026425.0,
      "step": 48905
    },
    {
      "entropy": 5.773401546478271,
      "epoch": 4.834914986160538,
      "grad_norm": 1.265625,
      "learning_rate": 0.00029027586048009554,
      "loss": 4.8961,
      "mean_token_accuracy": 0.2319907397031784,
      "num_tokens": 112036980.0,
      "step": 48910
    },
    {
      "entropy": 5.7356030464172365,
      "epoch": 4.835409252669039,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000290240651737092,
      "loss": 4.9678,
      "mean_token_accuracy": 0.22903912216424943,
      "num_tokens": 112048746.0,
      "step": 48915
    },
    {
      "entropy": 5.836233901977539,
      "epoch": 4.83590351917754,
      "grad_norm": 1.265625,
      "learning_rate": 0.000290205442619166,
      "loss": 4.9349,
      "mean_token_accuracy": 0.2324625074863434,
      "num_tokens": 112059635.0,
      "step": 48920
    },
    {
      "entropy": 5.895700216293335,
      "epoch": 4.836397785686042,
      "grad_norm": 1.3125,
      "learning_rate": 0.00029017023312718387,
      "loss": 5.0301,
      "mean_token_accuracy": 0.22018679976463318,
      "num_tokens": 112070994.0,
      "step": 48925
    },
    {
      "entropy": 5.769571733474732,
      "epoch": 4.8368920521945435,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002901350232620116,
      "loss": 4.9048,
      "mean_token_accuracy": 0.2372240036725998,
      "num_tokens": 112080897.0,
      "step": 48930
    },
    {
      "entropy": 5.842018890380859,
      "epoch": 4.837386318703045,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002900998130245153,
      "loss": 4.9622,
      "mean_token_accuracy": 0.23688526600599288,
      "num_tokens": 112091658.0,
      "step": 48935
    },
    {
      "entropy": 5.847526884078979,
      "epoch": 4.837880585211546,
      "grad_norm": 1.28125,
      "learning_rate": 0.00029006460241556137,
      "loss": 4.9904,
      "mean_token_accuracy": 0.22483591586351395,
      "num_tokens": 112103490.0,
      "step": 48940
    },
    {
      "entropy": 5.817656230926514,
      "epoch": 4.838374851720047,
      "grad_norm": 1.390625,
      "learning_rate": 0.00029002939143601584,
      "loss": 4.9286,
      "mean_token_accuracy": 0.23140643388032914,
      "num_tokens": 112115255.0,
      "step": 48945
    },
    {
      "entropy": 5.7761311531066895,
      "epoch": 4.838869118228549,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000289994180086745,
      "loss": 4.9144,
      "mean_token_accuracy": 0.22883250415325165,
      "num_tokens": 112126188.0,
      "step": 48950
    },
    {
      "entropy": 5.793883037567139,
      "epoch": 4.83936338473705,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002899589683686149,
      "loss": 4.9628,
      "mean_token_accuracy": 0.22414219230413437,
      "num_tokens": 112138006.0,
      "step": 48955
    },
    {
      "entropy": 5.801441192626953,
      "epoch": 4.839857651245552,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00028992375628249196,
      "loss": 4.9681,
      "mean_token_accuracy": 0.22325728833675385,
      "num_tokens": 112149828.0,
      "step": 48960
    },
    {
      "entropy": 5.755842018127441,
      "epoch": 4.840351917754053,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002898885438292423,
      "loss": 4.9574,
      "mean_token_accuracy": 0.2333178073167801,
      "num_tokens": 112161615.0,
      "step": 48965
    },
    {
      "entropy": 5.739009046554566,
      "epoch": 4.8408461842625545,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00028985333100973207,
      "loss": 4.8255,
      "mean_token_accuracy": 0.24662744402885436,
      "num_tokens": 112172116.0,
      "step": 48970
    },
    {
      "entropy": 5.790988779067993,
      "epoch": 4.841340450771056,
      "grad_norm": 1.34375,
      "learning_rate": 0.00028981811782482765,
      "loss": 4.872,
      "mean_token_accuracy": 0.23302360028028488,
      "num_tokens": 112184469.0,
      "step": 48975
    },
    {
      "entropy": 5.793516302108765,
      "epoch": 4.841834717279557,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002897829042753952,
      "loss": 4.9208,
      "mean_token_accuracy": 0.2327763095498085,
      "num_tokens": 112195473.0,
      "step": 48980
    },
    {
      "entropy": 5.851512813568116,
      "epoch": 4.842328983788058,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00028974769036230105,
      "loss": 5.0018,
      "mean_token_accuracy": 0.22331580519676208,
      "num_tokens": 112207339.0,
      "step": 48985
    },
    {
      "entropy": 5.774642562866211,
      "epoch": 4.84282325029656,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00028971247608641136,
      "loss": 4.8942,
      "mean_token_accuracy": 0.23702143430709838,
      "num_tokens": 112218444.0,
      "step": 48990
    },
    {
      "entropy": 5.868919038772583,
      "epoch": 4.843317516805062,
      "grad_norm": 1.375,
      "learning_rate": 0.00028967726144859255,
      "loss": 5.0071,
      "mean_token_accuracy": 0.22231556326150895,
      "num_tokens": 112230108.0,
      "step": 48995
    },
    {
      "entropy": 5.7322290420532225,
      "epoch": 4.843811783313563,
      "grad_norm": 1.265625,
      "learning_rate": 0.00028964204644971083,
      "loss": 4.8853,
      "mean_token_accuracy": 0.24014401584863662,
      "num_tokens": 112241731.0,
      "step": 49000
    },
    {
      "entropy": 5.753686714172363,
      "epoch": 4.844306049822064,
      "grad_norm": 1.3125,
      "learning_rate": 0.00028960683109063244,
      "loss": 4.8892,
      "mean_token_accuracy": 0.23182919025421142,
      "num_tokens": 112253501.0,
      "step": 49005
    },
    {
      "entropy": 5.750853586196899,
      "epoch": 4.8448003163305655,
      "grad_norm": 1.453125,
      "learning_rate": 0.00028957161537222374,
      "loss": 4.8545,
      "mean_token_accuracy": 0.24411375373601912,
      "num_tokens": 112264128.0,
      "step": 49010
    },
    {
      "entropy": 5.826304817199707,
      "epoch": 4.845294582839067,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00028953639929535114,
      "loss": 5.0,
      "mean_token_accuracy": 0.21745189279317856,
      "num_tokens": 112275602.0,
      "step": 49015
    },
    {
      "entropy": 5.815337753295898,
      "epoch": 4.845788849347568,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002895011828608808,
      "loss": 4.967,
      "mean_token_accuracy": 0.2273756667971611,
      "num_tokens": 112287158.0,
      "step": 49020
    },
    {
      "entropy": 5.702516984939575,
      "epoch": 4.846283115856069,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002894659660696791,
      "loss": 4.7557,
      "mean_token_accuracy": 0.2473243370652199,
      "num_tokens": 112298764.0,
      "step": 49025
    },
    {
      "entropy": 5.7469700336456295,
      "epoch": 4.8467773823645715,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002894307489226124,
      "loss": 4.8252,
      "mean_token_accuracy": 0.24225077778100967,
      "num_tokens": 112309976.0,
      "step": 49030
    },
    {
      "entropy": 5.828213787078857,
      "epoch": 4.847271648873073,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000289395531420547,
      "loss": 4.9363,
      "mean_token_accuracy": 0.22556197196245192,
      "num_tokens": 112321950.0,
      "step": 49035
    },
    {
      "entropy": 5.714060211181641,
      "epoch": 4.847765915381574,
      "grad_norm": 1.25,
      "learning_rate": 0.0002893603135643494,
      "loss": 4.8093,
      "mean_token_accuracy": 0.23687225431203843,
      "num_tokens": 112332823.0,
      "step": 49040
    },
    {
      "entropy": 5.734771680831909,
      "epoch": 4.848260181890075,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002893250953548857,
      "loss": 4.9261,
      "mean_token_accuracy": 0.2350659266114235,
      "num_tokens": 112345374.0,
      "step": 49045
    },
    {
      "entropy": 5.83812313079834,
      "epoch": 4.8487544483985765,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00028928987679302255,
      "loss": 5.0093,
      "mean_token_accuracy": 0.22093958854675294,
      "num_tokens": 112357963.0,
      "step": 49050
    },
    {
      "entropy": 5.782468175888061,
      "epoch": 4.849248714907078,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002892546578796262,
      "loss": 4.9514,
      "mean_token_accuracy": 0.22908113300800323,
      "num_tokens": 112368410.0,
      "step": 49055
    },
    {
      "entropy": 5.8211737155914305,
      "epoch": 4.849742981415579,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00028921943861556304,
      "loss": 4.9288,
      "mean_token_accuracy": 0.23487457633018494,
      "num_tokens": 112379202.0,
      "step": 49060
    },
    {
      "entropy": 5.695417404174805,
      "epoch": 4.85023724792408,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002891842190016995,
      "loss": 4.8308,
      "mean_token_accuracy": 0.23586246073246003,
      "num_tokens": 112389908.0,
      "step": 49065
    },
    {
      "entropy": 5.752830171585083,
      "epoch": 4.850731514432582,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00028914899903890193,
      "loss": 4.989,
      "mean_token_accuracy": 0.22820772230625153,
      "num_tokens": 112400917.0,
      "step": 49070
    },
    {
      "entropy": 5.718752145767212,
      "epoch": 4.851225780941084,
      "grad_norm": 1.390625,
      "learning_rate": 0.00028911377872803683,
      "loss": 4.859,
      "mean_token_accuracy": 0.23485609591007234,
      "num_tokens": 112413540.0,
      "step": 49075
    },
    {
      "entropy": 5.784301662445069,
      "epoch": 4.851720047449585,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00028907855806997057,
      "loss": 4.8963,
      "mean_token_accuracy": 0.2355680987238884,
      "num_tokens": 112423779.0,
      "step": 49080
    },
    {
      "entropy": 5.826023101806641,
      "epoch": 4.852214313958086,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002890433370655696,
      "loss": 4.9603,
      "mean_token_accuracy": 0.2272229865193367,
      "num_tokens": 112435705.0,
      "step": 49085
    },
    {
      "entropy": 5.780259704589843,
      "epoch": 4.8527085804665875,
      "grad_norm": 1.296875,
      "learning_rate": 0.00028900811571570036,
      "loss": 4.8864,
      "mean_token_accuracy": 0.24538090080022812,
      "num_tokens": 112446138.0,
      "step": 49090
    },
    {
      "entropy": 5.751820850372314,
      "epoch": 4.853202846975089,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002889728940212293,
      "loss": 4.9237,
      "mean_token_accuracy": 0.23289150297641753,
      "num_tokens": 112457616.0,
      "step": 49095
    },
    {
      "entropy": 5.7912389755249025,
      "epoch": 4.85369711348359,
      "grad_norm": 1.28125,
      "learning_rate": 0.00028893767198302284,
      "loss": 4.9601,
      "mean_token_accuracy": 0.23024620711803437,
      "num_tokens": 112470367.0,
      "step": 49100
    },
    {
      "entropy": 5.786993408203125,
      "epoch": 4.854191379992091,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002889024496019476,
      "loss": 4.9246,
      "mean_token_accuracy": 0.23474255651235582,
      "num_tokens": 112481075.0,
      "step": 49105
    },
    {
      "entropy": 5.781751441955566,
      "epoch": 4.854685646500593,
      "grad_norm": 1.203125,
      "learning_rate": 0.00028886722687886986,
      "loss": 4.9271,
      "mean_token_accuracy": 0.23021235466003417,
      "num_tokens": 112493846.0,
      "step": 49110
    },
    {
      "entropy": 5.727640628814697,
      "epoch": 4.855179913009095,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002888320038146562,
      "loss": 4.9143,
      "mean_token_accuracy": 0.2323154479265213,
      "num_tokens": 112506464.0,
      "step": 49115
    },
    {
      "entropy": 5.779295825958252,
      "epoch": 4.855674179517596,
      "grad_norm": 1.265625,
      "learning_rate": 0.00028879678041017297,
      "loss": 4.8463,
      "mean_token_accuracy": 0.23435731679201127,
      "num_tokens": 112517603.0,
      "step": 49120
    },
    {
      "entropy": 5.792895126342773,
      "epoch": 4.856168446026097,
      "grad_norm": 1.296875,
      "learning_rate": 0.000288761556666287,
      "loss": 4.8955,
      "mean_token_accuracy": 0.23019715994596482,
      "num_tokens": 112528736.0,
      "step": 49125
    },
    {
      "entropy": 5.830591297149658,
      "epoch": 4.8566627125345985,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00028872633258386453,
      "loss": 4.9937,
      "mean_token_accuracy": 0.22849510759115219,
      "num_tokens": 112540864.0,
      "step": 49130
    },
    {
      "entropy": 5.803523826599121,
      "epoch": 4.8571569790431,
      "grad_norm": 1.4375,
      "learning_rate": 0.00028869110816377215,
      "loss": 4.8823,
      "mean_token_accuracy": 0.23310040682554245,
      "num_tokens": 112551563.0,
      "step": 49135
    },
    {
      "entropy": 5.799018430709839,
      "epoch": 4.857651245551601,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00028865588340687637,
      "loss": 5.0196,
      "mean_token_accuracy": 0.22205927073955536,
      "num_tokens": 112563445.0,
      "step": 49140
    },
    {
      "entropy": 5.824177980422974,
      "epoch": 4.858145512060103,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002886206583140437,
      "loss": 4.9826,
      "mean_token_accuracy": 0.22782248705625535,
      "num_tokens": 112573601.0,
      "step": 49145
    },
    {
      "entropy": 5.804582309722901,
      "epoch": 4.858639778568604,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002885854328861407,
      "loss": 4.9275,
      "mean_token_accuracy": 0.23240430057048797,
      "num_tokens": 112584251.0,
      "step": 49150
    },
    {
      "entropy": 5.720830869674683,
      "epoch": 4.859134045077106,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002885502071240341,
      "loss": 4.8328,
      "mean_token_accuracy": 0.24566471725702285,
      "num_tokens": 112595580.0,
      "step": 49155
    },
    {
      "entropy": 5.774464702606201,
      "epoch": 4.859628311585607,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002885149810285903,
      "loss": 4.8949,
      "mean_token_accuracy": 0.23415710031986237,
      "num_tokens": 112606806.0,
      "step": 49160
    },
    {
      "entropy": 5.776809644699097,
      "epoch": 4.860122578094108,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002884797546006757,
      "loss": 4.912,
      "mean_token_accuracy": 0.23020249754190444,
      "num_tokens": 112618508.0,
      "step": 49165
    },
    {
      "entropy": 5.767145109176636,
      "epoch": 4.8606168446026095,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002884445278411572,
      "loss": 4.873,
      "mean_token_accuracy": 0.23561283200979233,
      "num_tokens": 112630830.0,
      "step": 49170
    },
    {
      "entropy": 5.783767461776733,
      "epoch": 4.861111111111111,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002884093007509012,
      "loss": 4.9272,
      "mean_token_accuracy": 0.23199794292449952,
      "num_tokens": 112642432.0,
      "step": 49175
    },
    {
      "entropy": 5.816671514511109,
      "epoch": 4.861605377619613,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002883740733307742,
      "loss": 4.9322,
      "mean_token_accuracy": 0.23084331601858138,
      "num_tokens": 112653756.0,
      "step": 49180
    },
    {
      "entropy": 5.809394073486328,
      "epoch": 4.862099644128114,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000288338845581643,
      "loss": 4.8939,
      "mean_token_accuracy": 0.23427701592445374,
      "num_tokens": 112664495.0,
      "step": 49185
    },
    {
      "entropy": 5.847043037414551,
      "epoch": 4.862593910636615,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002883036175043742,
      "loss": 4.9377,
      "mean_token_accuracy": 0.22541873455047606,
      "num_tokens": 112676232.0,
      "step": 49190
    },
    {
      "entropy": 5.819697570800781,
      "epoch": 4.863088177145117,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002882683890998342,
      "loss": 4.9255,
      "mean_token_accuracy": 0.21987585127353668,
      "num_tokens": 112687145.0,
      "step": 49195
    },
    {
      "entropy": 5.784909963607788,
      "epoch": 4.863582443653618,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002882331603688899,
      "loss": 4.941,
      "mean_token_accuracy": 0.22892287820577623,
      "num_tokens": 112697231.0,
      "step": 49200
    },
    {
      "entropy": 5.726517963409424,
      "epoch": 4.864076710162119,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002881979313124078,
      "loss": 4.8345,
      "mean_token_accuracy": 0.23340236991643906,
      "num_tokens": 112708478.0,
      "step": 49205
    },
    {
      "entropy": 5.768404769897461,
      "epoch": 4.86457097667062,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00028816270193125445,
      "loss": 4.8578,
      "mean_token_accuracy": 0.23843661993741988,
      "num_tokens": 112718564.0,
      "step": 49210
    },
    {
      "entropy": 5.78004412651062,
      "epoch": 4.865065243179123,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002881274722262966,
      "loss": 4.9074,
      "mean_token_accuracy": 0.23452680855989455,
      "num_tokens": 112728913.0,
      "step": 49215
    },
    {
      "entropy": 5.7678955078125,
      "epoch": 4.865559509687624,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002880922421984009,
      "loss": 4.8433,
      "mean_token_accuracy": 0.23443082720041275,
      "num_tokens": 112739136.0,
      "step": 49220
    },
    {
      "entropy": 5.755817556381226,
      "epoch": 4.866053776196125,
      "grad_norm": 1.46875,
      "learning_rate": 0.00028805701184843406,
      "loss": 4.89,
      "mean_token_accuracy": 0.227299202978611,
      "num_tokens": 112750279.0,
      "step": 49225
    },
    {
      "entropy": 5.751411294937133,
      "epoch": 4.866548042704626,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002880217811772627,
      "loss": 4.9179,
      "mean_token_accuracy": 0.233957003057003,
      "num_tokens": 112761814.0,
      "step": 49230
    },
    {
      "entropy": 5.743126487731933,
      "epoch": 4.867042309213128,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002879865501857535,
      "loss": 4.8618,
      "mean_token_accuracy": 0.22636641263961793,
      "num_tokens": 112771806.0,
      "step": 49235
    },
    {
      "entropy": 5.715855646133423,
      "epoch": 4.867536575721629,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002879513188747731,
      "loss": 4.8543,
      "mean_token_accuracy": 0.23425724655389785,
      "num_tokens": 112782071.0,
      "step": 49240
    },
    {
      "entropy": 5.696483135223389,
      "epoch": 4.86803084223013,
      "grad_norm": 1.421875,
      "learning_rate": 0.00028791608724518833,
      "loss": 4.9395,
      "mean_token_accuracy": 0.2379143238067627,
      "num_tokens": 112793312.0,
      "step": 49245
    },
    {
      "entropy": 5.742300796508789,
      "epoch": 4.868525108738632,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002878808552978658,
      "loss": 4.932,
      "mean_token_accuracy": 0.2354622781276703,
      "num_tokens": 112804925.0,
      "step": 49250
    },
    {
      "entropy": 5.812800073623658,
      "epoch": 4.869019375247134,
      "grad_norm": 1.28125,
      "learning_rate": 0.00028784562303367213,
      "loss": 4.9597,
      "mean_token_accuracy": 0.22812983244657517,
      "num_tokens": 112817947.0,
      "step": 49255
    },
    {
      "entropy": 5.854327297210693,
      "epoch": 4.869513641755635,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002878103904534742,
      "loss": 4.9815,
      "mean_token_accuracy": 0.22626199573278427,
      "num_tokens": 112829182.0,
      "step": 49260
    },
    {
      "entropy": 5.843795442581177,
      "epoch": 4.870007908264136,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00028777515755813867,
      "loss": 4.9784,
      "mean_token_accuracy": 0.22238934338092803,
      "num_tokens": 112841024.0,
      "step": 49265
    },
    {
      "entropy": 5.8153845310211185,
      "epoch": 4.870502174772637,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00028773992434853236,
      "loss": 4.8998,
      "mean_token_accuracy": 0.23224059790372847,
      "num_tokens": 112852175.0,
      "step": 49270
    },
    {
      "entropy": 5.699468564987183,
      "epoch": 4.870996441281139,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002877046908255219,
      "loss": 4.8467,
      "mean_token_accuracy": 0.2413761019706726,
      "num_tokens": 112863392.0,
      "step": 49275
    },
    {
      "entropy": 5.812985563278199,
      "epoch": 4.87149070778964,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00028766945698997397,
      "loss": 5.0034,
      "mean_token_accuracy": 0.21540404111146927,
      "num_tokens": 112876393.0,
      "step": 49280
    },
    {
      "entropy": 5.806315755844116,
      "epoch": 4.871984974298142,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00028763422284275547,
      "loss": 4.9588,
      "mean_token_accuracy": 0.23059856593608857,
      "num_tokens": 112886540.0,
      "step": 49285
    },
    {
      "entropy": 5.785651969909668,
      "epoch": 4.872479240806643,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002875989883847332,
      "loss": 4.8964,
      "mean_token_accuracy": 0.2378912091255188,
      "num_tokens": 112898057.0,
      "step": 49290
    },
    {
      "entropy": 5.74857006072998,
      "epoch": 4.872973507315145,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002875637536167738,
      "loss": 4.8772,
      "mean_token_accuracy": 0.23588797003030776,
      "num_tokens": 112910135.0,
      "step": 49295
    },
    {
      "entropy": 5.844922399520874,
      "epoch": 4.873467773823646,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002875285185397442,
      "loss": 5.0153,
      "mean_token_accuracy": 0.22356097549200057,
      "num_tokens": 112923374.0,
      "step": 49300
    },
    {
      "entropy": 5.725541353225708,
      "epoch": 4.873962040332147,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000287493283154511,
      "loss": 4.8864,
      "mean_token_accuracy": 0.23751910775899887,
      "num_tokens": 112934332.0,
      "step": 49305
    },
    {
      "entropy": 5.663888549804687,
      "epoch": 4.874456306840648,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002874580474619411,
      "loss": 4.7899,
      "mean_token_accuracy": 0.24756349474191666,
      "num_tokens": 112944630.0,
      "step": 49310
    },
    {
      "entropy": 5.810069847106933,
      "epoch": 4.87495057334915,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002874228114629014,
      "loss": 4.9777,
      "mean_token_accuracy": 0.22409878373146058,
      "num_tokens": 112957589.0,
      "step": 49315
    },
    {
      "entropy": 5.759831523895263,
      "epoch": 4.875444839857651,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002873875751582586,
      "loss": 4.8533,
      "mean_token_accuracy": 0.23885130137205124,
      "num_tokens": 112969052.0,
      "step": 49320
    },
    {
      "entropy": 5.812015390396118,
      "epoch": 4.875939106366152,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002873523385488794,
      "loss": 4.8968,
      "mean_token_accuracy": 0.2364566758275032,
      "num_tokens": 112979578.0,
      "step": 49325
    },
    {
      "entropy": 5.74316372871399,
      "epoch": 4.876433372874654,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00028731710163563083,
      "loss": 4.8596,
      "mean_token_accuracy": 0.23841979950666428,
      "num_tokens": 112991353.0,
      "step": 49330
    },
    {
      "entropy": 5.829902076721192,
      "epoch": 4.8769276393831555,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002872818644193797,
      "loss": 5.0518,
      "mean_token_accuracy": 0.21806159168481826,
      "num_tokens": 113003918.0,
      "step": 49335
    },
    {
      "entropy": 5.770462703704834,
      "epoch": 4.877421905891657,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002872466269009927,
      "loss": 4.9765,
      "mean_token_accuracy": 0.23322107195854186,
      "num_tokens": 113015251.0,
      "step": 49340
    },
    {
      "entropy": 5.772735643386841,
      "epoch": 4.877916172400158,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00028721138908133686,
      "loss": 4.8542,
      "mean_token_accuracy": 0.23681345731019973,
      "num_tokens": 113027076.0,
      "step": 49345
    },
    {
      "entropy": 5.856437063217163,
      "epoch": 4.878410438908659,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00028717615096127895,
      "loss": 5.0068,
      "mean_token_accuracy": 0.22299409359693528,
      "num_tokens": 113037622.0,
      "step": 49350
    },
    {
      "entropy": 5.812705945968628,
      "epoch": 4.878904705417161,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00028714091254168574,
      "loss": 4.9186,
      "mean_token_accuracy": 0.23489080965518952,
      "num_tokens": 113048687.0,
      "step": 49355
    },
    {
      "entropy": 5.847864055633545,
      "epoch": 4.879398971925662,
      "grad_norm": 1.453125,
      "learning_rate": 0.00028710567382342426,
      "loss": 5.0568,
      "mean_token_accuracy": 0.2227376565337181,
      "num_tokens": 113060185.0,
      "step": 49360
    },
    {
      "entropy": 5.871094036102295,
      "epoch": 4.879893238434164,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002870704348073614,
      "loss": 5.0525,
      "mean_token_accuracy": 0.21605550348758698,
      "num_tokens": 113072743.0,
      "step": 49365
    },
    {
      "entropy": 5.8603894233703615,
      "epoch": 4.880387504942665,
      "grad_norm": 1.453125,
      "learning_rate": 0.00028703519549436385,
      "loss": 5.0096,
      "mean_token_accuracy": 0.2306547373533249,
      "num_tokens": 113082990.0,
      "step": 49370
    },
    {
      "entropy": 5.863398027420044,
      "epoch": 4.8808817714511665,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002869999558852988,
      "loss": 4.963,
      "mean_token_accuracy": 0.21795407235622405,
      "num_tokens": 113094316.0,
      "step": 49375
    },
    {
      "entropy": 5.802748298645019,
      "epoch": 4.881376037959668,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002869647159810327,
      "loss": 4.9457,
      "mean_token_accuracy": 0.22640439867973328,
      "num_tokens": 113105297.0,
      "step": 49380
    },
    {
      "entropy": 5.762135076522827,
      "epoch": 4.881870304468169,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00028692947578243293,
      "loss": 4.8535,
      "mean_token_accuracy": 0.24241251349449158,
      "num_tokens": 113116851.0,
      "step": 49385
    },
    {
      "entropy": 5.750718212127685,
      "epoch": 4.88236457097667,
      "grad_norm": 1.375,
      "learning_rate": 0.0002868942352903662,
      "loss": 4.8731,
      "mean_token_accuracy": 0.23560429066419603,
      "num_tokens": 113127429.0,
      "step": 49390
    },
    {
      "entropy": 5.783828973770142,
      "epoch": 4.882858837485172,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002868589945056994,
      "loss": 4.9684,
      "mean_token_accuracy": 0.22898115813732148,
      "num_tokens": 113140173.0,
      "step": 49395
    },
    {
      "entropy": 5.8312843322753904,
      "epoch": 4.883353103993674,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002868237534292995,
      "loss": 4.9201,
      "mean_token_accuracy": 0.23126276284456254,
      "num_tokens": 113151147.0,
      "step": 49400
    },
    {
      "entropy": 5.801044845581055,
      "epoch": 4.883847370502175,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002867885120620334,
      "loss": 4.9338,
      "mean_token_accuracy": 0.2244262397289276,
      "num_tokens": 113163308.0,
      "step": 49405
    },
    {
      "entropy": 5.839687776565552,
      "epoch": 4.884341637010676,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002867532704047681,
      "loss": 4.9784,
      "mean_token_accuracy": 0.22870284020900727,
      "num_tokens": 113174405.0,
      "step": 49410
    },
    {
      "entropy": 5.792166423797608,
      "epoch": 4.8848359035191775,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002867180284583705,
      "loss": 4.9526,
      "mean_token_accuracy": 0.23022861033678055,
      "num_tokens": 113186662.0,
      "step": 49415
    },
    {
      "entropy": 5.756495189666748,
      "epoch": 4.885330170027679,
      "grad_norm": 1.265625,
      "learning_rate": 0.00028668278622370763,
      "loss": 4.8723,
      "mean_token_accuracy": 0.2418500378727913,
      "num_tokens": 113198489.0,
      "step": 49420
    },
    {
      "entropy": 5.788332462310791,
      "epoch": 4.88582443653618,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002866475437016464,
      "loss": 4.9504,
      "mean_token_accuracy": 0.2251896381378174,
      "num_tokens": 113210238.0,
      "step": 49425
    },
    {
      "entropy": 5.70060362815857,
      "epoch": 4.886318703044681,
      "grad_norm": 1.34375,
      "learning_rate": 0.00028661230089305375,
      "loss": 4.886,
      "mean_token_accuracy": 0.2377672776579857,
      "num_tokens": 113222008.0,
      "step": 49430
    },
    {
      "entropy": 5.752803707122803,
      "epoch": 4.8868129695531834,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002865770577987967,
      "loss": 4.9072,
      "mean_token_accuracy": 0.2328586310148239,
      "num_tokens": 113234187.0,
      "step": 49435
    },
    {
      "entropy": 5.761597537994385,
      "epoch": 4.887307236061685,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002865418144197423,
      "loss": 4.8911,
      "mean_token_accuracy": 0.23951933085918425,
      "num_tokens": 113246418.0,
      "step": 49440
    },
    {
      "entropy": 5.742188405990601,
      "epoch": 4.887801502570186,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00028650657075675735,
      "loss": 4.894,
      "mean_token_accuracy": 0.2345789149403572,
      "num_tokens": 113256561.0,
      "step": 49445
    },
    {
      "entropy": 5.7131763935089115,
      "epoch": 4.888295769078687,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002864713268107091,
      "loss": 4.9019,
      "mean_token_accuracy": 0.23587365448474884,
      "num_tokens": 113269145.0,
      "step": 49450
    },
    {
      "entropy": 5.758551073074341,
      "epoch": 4.8887900355871885,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002864360825824643,
      "loss": 4.8734,
      "mean_token_accuracy": 0.23137819468975068,
      "num_tokens": 113280545.0,
      "step": 49455
    },
    {
      "entropy": 5.845311450958252,
      "epoch": 4.88928430209569,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002864008380728902,
      "loss": 4.9316,
      "mean_token_accuracy": 0.23153275102376938,
      "num_tokens": 113292276.0,
      "step": 49460
    },
    {
      "entropy": 5.820145845413208,
      "epoch": 4.889778568604191,
      "grad_norm": 1.25,
      "learning_rate": 0.00028636559328285374,
      "loss": 4.9682,
      "mean_token_accuracy": 0.23152431845664978,
      "num_tokens": 113304486.0,
      "step": 49465
    },
    {
      "entropy": 5.757752323150635,
      "epoch": 4.890272835112693,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00028633034821322177,
      "loss": 4.9553,
      "mean_token_accuracy": 0.22723328173160554,
      "num_tokens": 113316374.0,
      "step": 49470
    },
    {
      "entropy": 5.816243886947632,
      "epoch": 4.890767101621194,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002862951028648616,
      "loss": 4.8702,
      "mean_token_accuracy": 0.23911852687597274,
      "num_tokens": 113327151.0,
      "step": 49475
    },
    {
      "entropy": 5.7520115852355955,
      "epoch": 4.891261368129696,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0002862598572386402,
      "loss": 4.8899,
      "mean_token_accuracy": 0.23756948858499527,
      "num_tokens": 113336314.0,
      "step": 49480
    },
    {
      "entropy": 5.7892567157745365,
      "epoch": 4.891755634638197,
      "grad_norm": 1.390625,
      "learning_rate": 0.00028622461133542436,
      "loss": 4.9411,
      "mean_token_accuracy": 0.2231715813279152,
      "num_tokens": 113346671.0,
      "step": 49485
    },
    {
      "entropy": 5.83081374168396,
      "epoch": 4.892249901146698,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002861893651560816,
      "loss": 4.9655,
      "mean_token_accuracy": 0.22435983121395112,
      "num_tokens": 113357781.0,
      "step": 49490
    },
    {
      "entropy": 5.809107971191406,
      "epoch": 4.8927441676551995,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002861541187014785,
      "loss": 4.9487,
      "mean_token_accuracy": 0.2285016655921936,
      "num_tokens": 113369691.0,
      "step": 49495
    },
    {
      "entropy": 5.782532119750977,
      "epoch": 4.893238434163701,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002861188719724824,
      "loss": 4.9617,
      "mean_token_accuracy": 0.22591428309679032,
      "num_tokens": 113381586.0,
      "step": 49500
    },
    {
      "entropy": 5.7742259979248045,
      "epoch": 4.893732700672203,
      "grad_norm": 1.3125,
      "learning_rate": 0.00028608362496996046,
      "loss": 4.9505,
      "mean_token_accuracy": 0.23338802456855773,
      "num_tokens": 113393551.0,
      "step": 49505
    },
    {
      "entropy": 5.841367483139038,
      "epoch": 4.894226967180704,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002860483776947795,
      "loss": 4.891,
      "mean_token_accuracy": 0.23188011050224305,
      "num_tokens": 113404615.0,
      "step": 49510
    },
    {
      "entropy": 5.752622985839844,
      "epoch": 4.894721233689205,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00028601313014780676,
      "loss": 4.9234,
      "mean_token_accuracy": 0.23329124748706817,
      "num_tokens": 113416186.0,
      "step": 49515
    },
    {
      "entropy": 5.795762062072754,
      "epoch": 4.895215500197707,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002859778823299093,
      "loss": 4.9408,
      "mean_token_accuracy": 0.2286386325955391,
      "num_tokens": 113428037.0,
      "step": 49520
    },
    {
      "entropy": 5.783979511260986,
      "epoch": 4.895709766706208,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0002859426342419543,
      "loss": 4.8809,
      "mean_token_accuracy": 0.23774271011352538,
      "num_tokens": 113439020.0,
      "step": 49525
    },
    {
      "entropy": 5.845564603805542,
      "epoch": 4.896204033214709,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002859073858848087,
      "loss": 4.9994,
      "mean_token_accuracy": 0.22532339990139008,
      "num_tokens": 113451060.0,
      "step": 49530
    },
    {
      "entropy": 5.834315061569214,
      "epoch": 4.8966982997232105,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00028587213725933975,
      "loss": 4.9626,
      "mean_token_accuracy": 0.22629932463169097,
      "num_tokens": 113463485.0,
      "step": 49535
    },
    {
      "entropy": 5.751065444946289,
      "epoch": 4.897192566231713,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002858368883664146,
      "loss": 4.9011,
      "mean_token_accuracy": 0.23899421691894532,
      "num_tokens": 113476259.0,
      "step": 49540
    },
    {
      "entropy": 5.837675619125366,
      "epoch": 4.897686832740214,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00028580163920690016,
      "loss": 5.0094,
      "mean_token_accuracy": 0.2246207684278488,
      "num_tokens": 113487957.0,
      "step": 49545
    },
    {
      "entropy": 5.84792914390564,
      "epoch": 4.898181099248715,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002857663897816639,
      "loss": 5.0165,
      "mean_token_accuracy": 0.21944080293178558,
      "num_tokens": 113500066.0,
      "step": 49550
    },
    {
      "entropy": 5.749091625213623,
      "epoch": 4.898675365757216,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002857311400915726,
      "loss": 4.806,
      "mean_token_accuracy": 0.2427174910902977,
      "num_tokens": 113511032.0,
      "step": 49555
    },
    {
      "entropy": 5.729857110977173,
      "epoch": 4.899169632265718,
      "grad_norm": 1.3125,
      "learning_rate": 0.00028569589013749375,
      "loss": 4.8534,
      "mean_token_accuracy": 0.2420921131968498,
      "num_tokens": 113522472.0,
      "step": 49560
    },
    {
      "entropy": 5.750196695327759,
      "epoch": 4.899663898774219,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002856606399202943,
      "loss": 4.9583,
      "mean_token_accuracy": 0.2297402262687683,
      "num_tokens": 113534841.0,
      "step": 49565
    },
    {
      "entropy": 5.855951929092408,
      "epoch": 4.90015816528272,
      "grad_norm": 1.28125,
      "learning_rate": 0.00028562538944084135,
      "loss": 4.9642,
      "mean_token_accuracy": 0.2299104690551758,
      "num_tokens": 113545659.0,
      "step": 49570
    },
    {
      "entropy": 5.788914060592651,
      "epoch": 4.9006524317912215,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002855901387000023,
      "loss": 4.9505,
      "mean_token_accuracy": 0.23686325848102568,
      "num_tokens": 113557293.0,
      "step": 49575
    },
    {
      "entropy": 5.813724088668823,
      "epoch": 4.901146698299723,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002855548876986442,
      "loss": 4.9616,
      "mean_token_accuracy": 0.22703695595264434,
      "num_tokens": 113568270.0,
      "step": 49580
    },
    {
      "entropy": 5.779737901687622,
      "epoch": 4.901640964808225,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002855196364376341,
      "loss": 4.9126,
      "mean_token_accuracy": 0.22952039688825607,
      "num_tokens": 113579838.0,
      "step": 49585
    },
    {
      "entropy": 5.815392303466797,
      "epoch": 4.902135231316726,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002854843849178395,
      "loss": 4.9733,
      "mean_token_accuracy": 0.23557699471712112,
      "num_tokens": 113590817.0,
      "step": 49590
    },
    {
      "entropy": 5.768790197372437,
      "epoch": 4.902629497825227,
      "grad_norm": 1.53125,
      "learning_rate": 0.0002854491331401273,
      "loss": 4.9101,
      "mean_token_accuracy": 0.234692944586277,
      "num_tokens": 113602912.0,
      "step": 49595
    },
    {
      "entropy": 5.773423910140991,
      "epoch": 4.903123764333729,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002854138811053647,
      "loss": 4.9112,
      "mean_token_accuracy": 0.23110030293464662,
      "num_tokens": 113613508.0,
      "step": 49600
    },
    {
      "entropy": 5.809531879425049,
      "epoch": 4.90361803084223,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00028537862881441913,
      "loss": 4.9847,
      "mean_token_accuracy": 0.22846920192241668,
      "num_tokens": 113625221.0,
      "step": 49605
    },
    {
      "entropy": 5.763628339767456,
      "epoch": 4.904112297350731,
      "grad_norm": 1.375,
      "learning_rate": 0.0002853433762681577,
      "loss": 4.8548,
      "mean_token_accuracy": 0.23607393205165864,
      "num_tokens": 113636367.0,
      "step": 49610
    },
    {
      "entropy": 5.774369621276856,
      "epoch": 4.904606563859232,
      "grad_norm": 1.25,
      "learning_rate": 0.00028530812346744754,
      "loss": 4.9599,
      "mean_token_accuracy": 0.22789814472198486,
      "num_tokens": 113648667.0,
      "step": 49615
    },
    {
      "entropy": 5.786903238296508,
      "epoch": 4.905100830367735,
      "grad_norm": 1.375,
      "learning_rate": 0.000285272870413156,
      "loss": 4.951,
      "mean_token_accuracy": 0.23696109056472778,
      "num_tokens": 113659331.0,
      "step": 49620
    },
    {
      "entropy": 5.74313006401062,
      "epoch": 4.905595096876236,
      "grad_norm": 1.328125,
      "learning_rate": 0.00028523761710615023,
      "loss": 4.8871,
      "mean_token_accuracy": 0.2343490019440651,
      "num_tokens": 113669874.0,
      "step": 49625
    },
    {
      "entropy": 5.849160671234131,
      "epoch": 4.906089363384737,
      "grad_norm": 1.671875,
      "learning_rate": 0.0002852023635472974,
      "loss": 4.9733,
      "mean_token_accuracy": 0.22268903702497483,
      "num_tokens": 113681626.0,
      "step": 49630
    },
    {
      "entropy": 5.826789712905883,
      "epoch": 4.906583629893238,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000285167109737465,
      "loss": 4.9867,
      "mean_token_accuracy": 0.228760027885437,
      "num_tokens": 113692778.0,
      "step": 49635
    },
    {
      "entropy": 5.7404790878295895,
      "epoch": 4.90707789640174,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00028513185567752004,
      "loss": 4.9258,
      "mean_token_accuracy": 0.23725828975439073,
      "num_tokens": 113704061.0,
      "step": 49640
    },
    {
      "entropy": 5.720127725601197,
      "epoch": 4.907572162910241,
      "grad_norm": 1.4375,
      "learning_rate": 0.00028509660136832984,
      "loss": 4.85,
      "mean_token_accuracy": 0.242891925573349,
      "num_tokens": 113715184.0,
      "step": 49645
    },
    {
      "entropy": 5.773281002044678,
      "epoch": 4.908066429418742,
      "grad_norm": 1.25,
      "learning_rate": 0.0002850613468107617,
      "loss": 4.963,
      "mean_token_accuracy": 0.22806744873523713,
      "num_tokens": 113726987.0,
      "step": 49650
    },
    {
      "entropy": 5.809217071533203,
      "epoch": 4.908560695927244,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000285026092005683,
      "loss": 4.9935,
      "mean_token_accuracy": 0.22644792199134828,
      "num_tokens": 113738516.0,
      "step": 49655
    },
    {
      "entropy": 5.909839200973511,
      "epoch": 4.909054962435746,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002849908369539607,
      "loss": 5.1007,
      "mean_token_accuracy": 0.21748928874731063,
      "num_tokens": 113750087.0,
      "step": 49660
    },
    {
      "entropy": 5.717170667648316,
      "epoch": 4.909549228944247,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002849555816564623,
      "loss": 4.7555,
      "mean_token_accuracy": 0.25533356964588166,
      "num_tokens": 113760837.0,
      "step": 49665
    },
    {
      "entropy": 5.825530576705932,
      "epoch": 4.910043495452748,
      "grad_norm": 1.53125,
      "learning_rate": 0.00028492032611405506,
      "loss": 4.9159,
      "mean_token_accuracy": 0.23628135323524474,
      "num_tokens": 113770496.0,
      "step": 49670
    },
    {
      "entropy": 5.821273994445801,
      "epoch": 4.910537761961249,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002848850703276062,
      "loss": 5.0018,
      "mean_token_accuracy": 0.2215157702565193,
      "num_tokens": 113781720.0,
      "step": 49675
    },
    {
      "entropy": 5.744265460968018,
      "epoch": 4.911032028469751,
      "grad_norm": 1.296875,
      "learning_rate": 0.00028484981429798326,
      "loss": 4.8594,
      "mean_token_accuracy": 0.23632450699806212,
      "num_tokens": 113793140.0,
      "step": 49680
    },
    {
      "entropy": 5.827163887023926,
      "epoch": 4.911526294978252,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002848145580260532,
      "loss": 4.9677,
      "mean_token_accuracy": 0.219768488407135,
      "num_tokens": 113804584.0,
      "step": 49685
    },
    {
      "entropy": 5.71978611946106,
      "epoch": 4.912020561486754,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002847793015126835,
      "loss": 4.8745,
      "mean_token_accuracy": 0.2271732896566391,
      "num_tokens": 113815701.0,
      "step": 49690
    },
    {
      "entropy": 5.752813482284546,
      "epoch": 4.912514827995255,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00028474404475874155,
      "loss": 4.853,
      "mean_token_accuracy": 0.23758791238069535,
      "num_tokens": 113826878.0,
      "step": 49695
    },
    {
      "entropy": 5.7857531070709225,
      "epoch": 4.913009094503757,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002847087877650946,
      "loss": 4.901,
      "mean_token_accuracy": 0.23775359094142914,
      "num_tokens": 113837320.0,
      "step": 49700
    },
    {
      "entropy": 5.874799537658691,
      "epoch": 4.913503361012258,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002846735305326098,
      "loss": 5.1032,
      "mean_token_accuracy": 0.21485828310251237,
      "num_tokens": 113849460.0,
      "step": 49705
    },
    {
      "entropy": 5.745315313339233,
      "epoch": 4.913997627520759,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002846382730621548,
      "loss": 4.8557,
      "mean_token_accuracy": 0.24056546986103058,
      "num_tokens": 113860998.0,
      "step": 49710
    },
    {
      "entropy": 5.794930124282837,
      "epoch": 4.91449189402926,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002846030153545967,
      "loss": 4.9141,
      "mean_token_accuracy": 0.2289827585220337,
      "num_tokens": 113873100.0,
      "step": 49715
    },
    {
      "entropy": 5.799171400070191,
      "epoch": 4.914986160537762,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00028456775741080297,
      "loss": 4.9979,
      "mean_token_accuracy": 0.22023083865642548,
      "num_tokens": 113885444.0,
      "step": 49720
    },
    {
      "entropy": 5.807810354232788,
      "epoch": 4.915480427046264,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00028453249923164094,
      "loss": 4.9349,
      "mean_token_accuracy": 0.23151204884052276,
      "num_tokens": 113897198.0,
      "step": 49725
    },
    {
      "entropy": 5.822122287750244,
      "epoch": 4.915974693554765,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00028449724081797796,
      "loss": 4.9538,
      "mean_token_accuracy": 0.22858942449092864,
      "num_tokens": 113908436.0,
      "step": 49730
    },
    {
      "entropy": 5.800542879104614,
      "epoch": 4.916468960063266,
      "grad_norm": 1.328125,
      "learning_rate": 0.00028446198217068133,
      "loss": 4.9914,
      "mean_token_accuracy": 0.23040324449539185,
      "num_tokens": 113918955.0,
      "step": 49735
    },
    {
      "entropy": 5.744678020477295,
      "epoch": 4.9169632265717675,
      "grad_norm": 1.40625,
      "learning_rate": 0.00028442672329061854,
      "loss": 4.9525,
      "mean_token_accuracy": 0.2347753643989563,
      "num_tokens": 113931299.0,
      "step": 49740
    },
    {
      "entropy": 5.855995416641235,
      "epoch": 4.917457493080269,
      "grad_norm": 1.25,
      "learning_rate": 0.0002843914641786568,
      "loss": 4.961,
      "mean_token_accuracy": 0.22710224986076355,
      "num_tokens": 113942396.0,
      "step": 49745
    },
    {
      "entropy": 5.842866897583008,
      "epoch": 4.91795175958877,
      "grad_norm": 1.390625,
      "learning_rate": 0.00028435620483566366,
      "loss": 4.9478,
      "mean_token_accuracy": 0.2239452451467514,
      "num_tokens": 113953815.0,
      "step": 49750
    },
    {
      "entropy": 5.851448488235474,
      "epoch": 4.918446026097271,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002843209452625064,
      "loss": 4.99,
      "mean_token_accuracy": 0.23054806143045425,
      "num_tokens": 113965567.0,
      "step": 49755
    },
    {
      "entropy": 5.763729429244995,
      "epoch": 4.9189402926057735,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002842856854600524,
      "loss": 4.9664,
      "mean_token_accuracy": 0.2292360171675682,
      "num_tokens": 113977350.0,
      "step": 49760
    },
    {
      "entropy": 5.862820100784302,
      "epoch": 4.919434559114275,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00028425042542916916,
      "loss": 5.0441,
      "mean_token_accuracy": 0.2248494103550911,
      "num_tokens": 113987997.0,
      "step": 49765
    },
    {
      "entropy": 5.774439096450806,
      "epoch": 4.919928825622776,
      "grad_norm": 1.375,
      "learning_rate": 0.00028421516517072404,
      "loss": 4.8376,
      "mean_token_accuracy": 0.23773832470178605,
      "num_tokens": 113998754.0,
      "step": 49770
    },
    {
      "entropy": 5.782770776748658,
      "epoch": 4.920423092131277,
      "grad_norm": 1.3125,
      "learning_rate": 0.00028417990468558433,
      "loss": 4.9206,
      "mean_token_accuracy": 0.2305078998208046,
      "num_tokens": 114009619.0,
      "step": 49775
    },
    {
      "entropy": 5.742311000823975,
      "epoch": 4.9209173586397785,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002841446439746176,
      "loss": 4.8635,
      "mean_token_accuracy": 0.23687660247087478,
      "num_tokens": 114020581.0,
      "step": 49780
    },
    {
      "entropy": 5.795451498031616,
      "epoch": 4.92141162514828,
      "grad_norm": 1.3125,
      "learning_rate": 0.00028410938303869126,
      "loss": 4.9184,
      "mean_token_accuracy": 0.2348613977432251,
      "num_tokens": 114031796.0,
      "step": 49785
    },
    {
      "entropy": 5.806154537200928,
      "epoch": 4.921905891656781,
      "grad_norm": 1.375,
      "learning_rate": 0.0002840741218786726,
      "loss": 4.9424,
      "mean_token_accuracy": 0.2264641135931015,
      "num_tokens": 114042125.0,
      "step": 49790
    },
    {
      "entropy": 5.790507650375366,
      "epoch": 4.922400158165283,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002840388604954292,
      "loss": 4.9393,
      "mean_token_accuracy": 0.22884512394666673,
      "num_tokens": 114054546.0,
      "step": 49795
    },
    {
      "entropy": 5.826856231689453,
      "epoch": 4.9228944246737845,
      "grad_norm": 1.578125,
      "learning_rate": 0.0002840035988898284,
      "loss": 4.9468,
      "mean_token_accuracy": 0.23112366646528243,
      "num_tokens": 114065290.0,
      "step": 49800
    },
    {
      "entropy": 5.863804817199707,
      "epoch": 4.923388691182286,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002839683370627377,
      "loss": 4.9316,
      "mean_token_accuracy": 0.2260959655046463,
      "num_tokens": 114075560.0,
      "step": 49805
    },
    {
      "entropy": 5.730663824081421,
      "epoch": 4.923882957690787,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002839330750150244,
      "loss": 4.7917,
      "mean_token_accuracy": 0.2465095192193985,
      "num_tokens": 114086806.0,
      "step": 49810
    },
    {
      "entropy": 5.742004776000977,
      "epoch": 4.924377224199288,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00028389781274755624,
      "loss": 4.9371,
      "mean_token_accuracy": 0.23309370130300522,
      "num_tokens": 114098204.0,
      "step": 49815
    },
    {
      "entropy": 5.76456503868103,
      "epoch": 4.9248714907077895,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002838625502612004,
      "loss": 4.9408,
      "mean_token_accuracy": 0.23431558907032013,
      "num_tokens": 114111331.0,
      "step": 49820
    },
    {
      "entropy": 5.813323926925659,
      "epoch": 4.925365757216291,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00028382728755682444,
      "loss": 4.973,
      "mean_token_accuracy": 0.23562365770339966,
      "num_tokens": 114122878.0,
      "step": 49825
    },
    {
      "entropy": 5.866981029510498,
      "epoch": 4.925860023724792,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00028379202463529586,
      "loss": 4.8962,
      "mean_token_accuracy": 0.2335548982024193,
      "num_tokens": 114133320.0,
      "step": 49830
    },
    {
      "entropy": 5.760624980926513,
      "epoch": 4.926354290233293,
      "grad_norm": 1.3125,
      "learning_rate": 0.00028375676149748207,
      "loss": 4.8903,
      "mean_token_accuracy": 0.23015797585248948,
      "num_tokens": 114145975.0,
      "step": 49835
    },
    {
      "entropy": 5.76899824142456,
      "epoch": 4.9268485567417954,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002837214981442507,
      "loss": 4.9441,
      "mean_token_accuracy": 0.22269562780857086,
      "num_tokens": 114156812.0,
      "step": 49840
    },
    {
      "entropy": 5.806515836715699,
      "epoch": 4.927342823250297,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002836862345764691,
      "loss": 4.998,
      "mean_token_accuracy": 0.2285042390227318,
      "num_tokens": 114169009.0,
      "step": 49845
    },
    {
      "entropy": 5.811793565750122,
      "epoch": 4.927837089758798,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00028365097079500475,
      "loss": 4.879,
      "mean_token_accuracy": 0.2363865539431572,
      "num_tokens": 114181427.0,
      "step": 49850
    },
    {
      "entropy": 5.795496034622192,
      "epoch": 4.928331356267299,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00028361570680072507,
      "loss": 4.9727,
      "mean_token_accuracy": 0.22633173167705536,
      "num_tokens": 114192427.0,
      "step": 49855
    },
    {
      "entropy": 5.7613029956817625,
      "epoch": 4.9288256227758005,
      "grad_norm": 1.390625,
      "learning_rate": 0.00028358044259449785,
      "loss": 4.8681,
      "mean_token_accuracy": 0.24114388972520828,
      "num_tokens": 114202914.0,
      "step": 49860
    },
    {
      "entropy": 5.731849336624146,
      "epoch": 4.929319889284302,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00028354517817719024,
      "loss": 4.899,
      "mean_token_accuracy": 0.23744490295648574,
      "num_tokens": 114214359.0,
      "step": 49865
    },
    {
      "entropy": 5.834128952026367,
      "epoch": 4.929814155792803,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00028350991354967003,
      "loss": 5.0506,
      "mean_token_accuracy": 0.23135192096233367,
      "num_tokens": 114225766.0,
      "step": 49870
    },
    {
      "entropy": 5.826089715957641,
      "epoch": 4.930308422301305,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002834746487128046,
      "loss": 4.9655,
      "mean_token_accuracy": 0.232780584692955,
      "num_tokens": 114236488.0,
      "step": 49875
    },
    {
      "entropy": 5.85629620552063,
      "epoch": 4.930802688809806,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002834393836674615,
      "loss": 4.9476,
      "mean_token_accuracy": 0.23046936392784118,
      "num_tokens": 114248524.0,
      "step": 49880
    },
    {
      "entropy": 5.767047452926636,
      "epoch": 4.931296955318308,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002834041184145082,
      "loss": 4.8612,
      "mean_token_accuracy": 0.2411500871181488,
      "num_tokens": 114260523.0,
      "step": 49885
    },
    {
      "entropy": 5.76344256401062,
      "epoch": 4.931791221826809,
      "grad_norm": 1.25,
      "learning_rate": 0.0002833688529548124,
      "loss": 4.8648,
      "mean_token_accuracy": 0.2361275389790535,
      "num_tokens": 114272150.0,
      "step": 49890
    },
    {
      "entropy": 5.789968538284302,
      "epoch": 4.93228548833531,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002833335872892414,
      "loss": 5.001,
      "mean_token_accuracy": 0.22087265253067018,
      "num_tokens": 114284127.0,
      "step": 49895
    },
    {
      "entropy": 5.841102409362793,
      "epoch": 4.9327797548438115,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002832983214186629,
      "loss": 5.0396,
      "mean_token_accuracy": 0.22227319329977036,
      "num_tokens": 114295995.0,
      "step": 49900
    },
    {
      "entropy": 5.817886257171631,
      "epoch": 4.933274021352313,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002832630553439444,
      "loss": 4.9849,
      "mean_token_accuracy": 0.22462240159511565,
      "num_tokens": 114307148.0,
      "step": 49905
    },
    {
      "entropy": 5.774989795684815,
      "epoch": 4.933768287860815,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002832277890659535,
      "loss": 4.952,
      "mean_token_accuracy": 0.2342192381620407,
      "num_tokens": 114319062.0,
      "step": 49910
    },
    {
      "entropy": 5.810780096054077,
      "epoch": 4.934262554369316,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002831925225855577,
      "loss": 4.9351,
      "mean_token_accuracy": 0.23211680203676224,
      "num_tokens": 114330958.0,
      "step": 49915
    },
    {
      "entropy": 5.867742538452148,
      "epoch": 4.934756820877817,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002831572559036245,
      "loss": 4.9718,
      "mean_token_accuracy": 0.22906851023435593,
      "num_tokens": 114342249.0,
      "step": 49920
    },
    {
      "entropy": 5.84346833229065,
      "epoch": 4.935251087386319,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00028312198902102166,
      "loss": 4.9386,
      "mean_token_accuracy": 0.2293518677353859,
      "num_tokens": 114352834.0,
      "step": 49925
    },
    {
      "entropy": 5.880119180679321,
      "epoch": 4.93574535389482,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002830867219386166,
      "loss": 4.9961,
      "mean_token_accuracy": 0.22626329362392425,
      "num_tokens": 114364397.0,
      "step": 49930
    },
    {
      "entropy": 5.715747547149658,
      "epoch": 4.936239620403321,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002830514546572768,
      "loss": 4.8404,
      "mean_token_accuracy": 0.24229886084795,
      "num_tokens": 114375588.0,
      "step": 49935
    },
    {
      "entropy": 5.850084352493286,
      "epoch": 4.9367338869118225,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00028301618717787013,
      "loss": 5.0128,
      "mean_token_accuracy": 0.2190149173140526,
      "num_tokens": 114387467.0,
      "step": 49940
    },
    {
      "entropy": 5.834776735305786,
      "epoch": 4.937228153420325,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002829809195012639,
      "loss": 4.9504,
      "mean_token_accuracy": 0.23169461041688919,
      "num_tokens": 114399024.0,
      "step": 49945
    },
    {
      "entropy": 5.777243947982788,
      "epoch": 4.937722419928826,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00028294565162832584,
      "loss": 4.9013,
      "mean_token_accuracy": 0.23407716006040574,
      "num_tokens": 114409441.0,
      "step": 49950
    },
    {
      "entropy": 5.782475519180298,
      "epoch": 4.938216686437327,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00028291038355992356,
      "loss": 4.8911,
      "mean_token_accuracy": 0.23453335464000702,
      "num_tokens": 114422033.0,
      "step": 49955
    },
    {
      "entropy": 5.813448095321656,
      "epoch": 4.938710952945828,
      "grad_norm": 1.265625,
      "learning_rate": 0.00028287511529692465,
      "loss": 4.8757,
      "mean_token_accuracy": 0.23964359760284423,
      "num_tokens": 114432850.0,
      "step": 49960
    },
    {
      "entropy": 5.833718013763428,
      "epoch": 4.93920521945433,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002828398468401966,
      "loss": 4.9923,
      "mean_token_accuracy": 0.22636514157056808,
      "num_tokens": 114444381.0,
      "step": 49965
    },
    {
      "entropy": 5.823450422286987,
      "epoch": 4.939699485962831,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00028280457819060704,
      "loss": 4.978,
      "mean_token_accuracy": 0.2295143187046051,
      "num_tokens": 114456052.0,
      "step": 49970
    },
    {
      "entropy": 5.762185716629029,
      "epoch": 4.940193752471332,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00028276930934902376,
      "loss": 4.865,
      "mean_token_accuracy": 0.24041374325752257,
      "num_tokens": 114466933.0,
      "step": 49975
    },
    {
      "entropy": 5.772572183609009,
      "epoch": 4.940688018979834,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002827340403163143,
      "loss": 4.9292,
      "mean_token_accuracy": 0.2306359201669693,
      "num_tokens": 114478688.0,
      "step": 49980
    },
    {
      "entropy": 5.821403312683105,
      "epoch": 4.941182285488336,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002826987710933462,
      "loss": 4.976,
      "mean_token_accuracy": 0.22346433997154236,
      "num_tokens": 114489655.0,
      "step": 49985
    },
    {
      "entropy": 5.792050361633301,
      "epoch": 4.941676551996837,
      "grad_norm": 1.21875,
      "learning_rate": 0.00028266350168098706,
      "loss": 4.8529,
      "mean_token_accuracy": 0.2348995640873909,
      "num_tokens": 114500791.0,
      "step": 49990
    },
    {
      "entropy": 5.859633302688598,
      "epoch": 4.942170818505338,
      "grad_norm": 1.265625,
      "learning_rate": 0.00028262823208010464,
      "loss": 5.0045,
      "mean_token_accuracy": 0.2238997608423233,
      "num_tokens": 114513562.0,
      "step": 49995
    },
    {
      "entropy": 5.817596673965454,
      "epoch": 4.942665085013839,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002825929622915665,
      "loss": 4.9875,
      "mean_token_accuracy": 0.2229781374335289,
      "num_tokens": 114524973.0,
      "step": 50000
    },
    {
      "entropy": 5.861956024169922,
      "epoch": 4.943159351522341,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002825576923162404,
      "loss": 4.9679,
      "mean_token_accuracy": 0.22606209218502044,
      "num_tokens": 114536511.0,
      "step": 50005
    },
    {
      "entropy": 5.873575115203858,
      "epoch": 4.943653618030842,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00028252242215499385,
      "loss": 5.016,
      "mean_token_accuracy": 0.22243181616067886,
      "num_tokens": 114548544.0,
      "step": 50010
    },
    {
      "entropy": 5.819559574127197,
      "epoch": 4.944147884539344,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00028248715180869457,
      "loss": 4.9019,
      "mean_token_accuracy": 0.23738448619842528,
      "num_tokens": 114559759.0,
      "step": 50015
    },
    {
      "entropy": 5.7790083408355715,
      "epoch": 4.944642151047845,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00028245188127821015,
      "loss": 4.9006,
      "mean_token_accuracy": 0.2331358566880226,
      "num_tokens": 114570746.0,
      "step": 50020
    },
    {
      "entropy": 5.801171922683716,
      "epoch": 4.945136417556347,
      "grad_norm": 1.375,
      "learning_rate": 0.00028241661056440824,
      "loss": 4.9646,
      "mean_token_accuracy": 0.23100357353687287,
      "num_tokens": 114582411.0,
      "step": 50025
    },
    {
      "entropy": 5.735310363769531,
      "epoch": 4.945630684064848,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00028238133966815667,
      "loss": 4.9028,
      "mean_token_accuracy": 0.24055004715919495,
      "num_tokens": 114594692.0,
      "step": 50030
    },
    {
      "entropy": 5.773415374755859,
      "epoch": 4.946124950573349,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00028234606859032303,
      "loss": 4.8689,
      "mean_token_accuracy": 0.23279104232788086,
      "num_tokens": 114606346.0,
      "step": 50035
    },
    {
      "entropy": 5.8222709655761715,
      "epoch": 4.94661921708185,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002823107973317748,
      "loss": 4.9734,
      "mean_token_accuracy": 0.23279672116041183,
      "num_tokens": 114617657.0,
      "step": 50040
    },
    {
      "entropy": 5.798703145980835,
      "epoch": 4.947113483590352,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002822755258933799,
      "loss": 4.9108,
      "mean_token_accuracy": 0.23610049337148667,
      "num_tokens": 114629203.0,
      "step": 50045
    },
    {
      "entropy": 5.787212467193603,
      "epoch": 4.947607750098854,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00028224025427600597,
      "loss": 5.0014,
      "mean_token_accuracy": 0.22070663124322892,
      "num_tokens": 114640441.0,
      "step": 50050
    },
    {
      "entropy": 5.800938844680786,
      "epoch": 4.948102016607355,
      "grad_norm": 1.359375,
      "learning_rate": 0.00028220498248052064,
      "loss": 4.9329,
      "mean_token_accuracy": 0.22891813665628433,
      "num_tokens": 114651095.0,
      "step": 50055
    },
    {
      "entropy": 5.778642702102661,
      "epoch": 4.948596283115856,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00028216971050779163,
      "loss": 4.9129,
      "mean_token_accuracy": 0.22912734597921372,
      "num_tokens": 114662691.0,
      "step": 50060
    },
    {
      "entropy": 5.797781658172608,
      "epoch": 4.949090549624358,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00028213443835868655,
      "loss": 4.9096,
      "mean_token_accuracy": 0.23262368589639665,
      "num_tokens": 114674618.0,
      "step": 50065
    },
    {
      "entropy": 5.816585731506348,
      "epoch": 4.949584816132859,
      "grad_norm": 1.5625,
      "learning_rate": 0.00028209916603407333,
      "loss": 4.9616,
      "mean_token_accuracy": 0.2245672821998596,
      "num_tokens": 114685832.0,
      "step": 50070
    },
    {
      "entropy": 5.777929639816284,
      "epoch": 4.95007908264136,
      "grad_norm": 1.375,
      "learning_rate": 0.0002820638935348194,
      "loss": 4.8835,
      "mean_token_accuracy": 0.23668137937784195,
      "num_tokens": 114696723.0,
      "step": 50075
    },
    {
      "entropy": 5.817441034317016,
      "epoch": 4.950573349149861,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002820286208617926,
      "loss": 4.9448,
      "mean_token_accuracy": 0.2272737056016922,
      "num_tokens": 114707064.0,
      "step": 50080
    },
    {
      "entropy": 5.783890104293823,
      "epoch": 4.951067615658363,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002819933480158607,
      "loss": 4.9033,
      "mean_token_accuracy": 0.23640350252389908,
      "num_tokens": 114717121.0,
      "step": 50085
    },
    {
      "entropy": 5.778128242492675,
      "epoch": 4.951561882166864,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002819580749978913,
      "loss": 4.8801,
      "mean_token_accuracy": 0.23893888145685196,
      "num_tokens": 114728539.0,
      "step": 50090
    },
    {
      "entropy": 5.749608325958252,
      "epoch": 4.952056148675366,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002819228018087522,
      "loss": 4.8648,
      "mean_token_accuracy": 0.23393135964870454,
      "num_tokens": 114739318.0,
      "step": 50095
    },
    {
      "entropy": 5.864294385910034,
      "epoch": 4.952550415183867,
      "grad_norm": 1.296875,
      "learning_rate": 0.00028188752844931115,
      "loss": 4.9744,
      "mean_token_accuracy": 0.22932562083005906,
      "num_tokens": 114751118.0,
      "step": 50100
    },
    {
      "entropy": 5.810138130187989,
      "epoch": 4.953044681692369,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002818522549204358,
      "loss": 4.9082,
      "mean_token_accuracy": 0.22569647282361985,
      "num_tokens": 114762309.0,
      "step": 50105
    },
    {
      "entropy": 5.847204303741455,
      "epoch": 4.95353894820087,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002818169812229939,
      "loss": 4.9518,
      "mean_token_accuracy": 0.22461484372615814,
      "num_tokens": 114774279.0,
      "step": 50110
    },
    {
      "entropy": 5.812509822845459,
      "epoch": 4.954033214709371,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00028178170735785316,
      "loss": 5.0113,
      "mean_token_accuracy": 0.22614845037460327,
      "num_tokens": 114786188.0,
      "step": 50115
    },
    {
      "entropy": 5.743883228302002,
      "epoch": 4.954527481217872,
      "grad_norm": 1.359375,
      "learning_rate": 0.00028174643332588145,
      "loss": 4.8474,
      "mean_token_accuracy": 0.23588586002588272,
      "num_tokens": 114797720.0,
      "step": 50120
    },
    {
      "entropy": 5.777899312973022,
      "epoch": 4.955021747726374,
      "grad_norm": 1.3125,
      "learning_rate": 0.00028171115912794636,
      "loss": 4.9468,
      "mean_token_accuracy": 0.23775314390659333,
      "num_tokens": 114809717.0,
      "step": 50125
    },
    {
      "entropy": 5.81242995262146,
      "epoch": 4.955516014234876,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002816758847649158,
      "loss": 4.918,
      "mean_token_accuracy": 0.2350105121731758,
      "num_tokens": 114819900.0,
      "step": 50130
    },
    {
      "entropy": 5.83761887550354,
      "epoch": 4.956010280743377,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002816406102376574,
      "loss": 4.9627,
      "mean_token_accuracy": 0.2272122621536255,
      "num_tokens": 114832089.0,
      "step": 50135
    },
    {
      "entropy": 5.833520841598511,
      "epoch": 4.956504547251878,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00028160533554703887,
      "loss": 5.0526,
      "mean_token_accuracy": 0.22129807770252227,
      "num_tokens": 114844261.0,
      "step": 50140
    },
    {
      "entropy": 5.774409484863281,
      "epoch": 4.9569988137603795,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00028157006069392816,
      "loss": 4.8963,
      "mean_token_accuracy": 0.23243194669485093,
      "num_tokens": 114854903.0,
      "step": 50145
    },
    {
      "entropy": 5.814947509765625,
      "epoch": 4.957493080268881,
      "grad_norm": 1.265625,
      "learning_rate": 0.00028153478567919293,
      "loss": 4.8819,
      "mean_token_accuracy": 0.23764284402132035,
      "num_tokens": 114865997.0,
      "step": 50150
    },
    {
      "entropy": 5.829001760482788,
      "epoch": 4.957987346777382,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00028149951050370094,
      "loss": 5.0398,
      "mean_token_accuracy": 0.21498369425535202,
      "num_tokens": 114878522.0,
      "step": 50155
    },
    {
      "entropy": 5.83411054611206,
      "epoch": 4.958481613285883,
      "grad_norm": 1.4375,
      "learning_rate": 0.00028146423516831997,
      "loss": 4.953,
      "mean_token_accuracy": 0.23039473593235016,
      "num_tokens": 114890277.0,
      "step": 50160
    },
    {
      "entropy": 5.736070346832276,
      "epoch": 4.9589758797943855,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00028142895967391793,
      "loss": 4.8513,
      "mean_token_accuracy": 0.23783244490623473,
      "num_tokens": 114902019.0,
      "step": 50165
    },
    {
      "entropy": 5.775155544281006,
      "epoch": 4.959470146302887,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002813936840213623,
      "loss": 4.9182,
      "mean_token_accuracy": 0.23343444615602493,
      "num_tokens": 114913375.0,
      "step": 50170
    },
    {
      "entropy": 5.8057445049285885,
      "epoch": 4.959964412811388,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002813584082115212,
      "loss": 4.932,
      "mean_token_accuracy": 0.23340822160243987,
      "num_tokens": 114924793.0,
      "step": 50175
    },
    {
      "entropy": 5.80465874671936,
      "epoch": 4.960458679319889,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00028132313224526226,
      "loss": 4.9206,
      "mean_token_accuracy": 0.23438309282064437,
      "num_tokens": 114936255.0,
      "step": 50180
    },
    {
      "entropy": 5.810911035537719,
      "epoch": 4.9609529458283905,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002812878561234531,
      "loss": 4.9359,
      "mean_token_accuracy": 0.22913830429315568,
      "num_tokens": 114947832.0,
      "step": 50185
    },
    {
      "entropy": 5.747545385360718,
      "epoch": 4.961447212336892,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00028125257984696185,
      "loss": 4.9271,
      "mean_token_accuracy": 0.23852724730968475,
      "num_tokens": 114960245.0,
      "step": 50190
    },
    {
      "entropy": 5.784721755981446,
      "epoch": 4.961941478845393,
      "grad_norm": 1.28125,
      "learning_rate": 0.00028121730341665617,
      "loss": 4.9296,
      "mean_token_accuracy": 0.22984112203121185,
      "num_tokens": 114970931.0,
      "step": 50195
    },
    {
      "entropy": 5.795871829986572,
      "epoch": 4.962435745353895,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002811820268334038,
      "loss": 4.9267,
      "mean_token_accuracy": 0.23059766739606857,
      "num_tokens": 114982106.0,
      "step": 50200
    },
    {
      "entropy": 5.859975576400757,
      "epoch": 4.9629300118623965,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002811467500980727,
      "loss": 4.9456,
      "mean_token_accuracy": 0.23010871559381485,
      "num_tokens": 114994194.0,
      "step": 50205
    },
    {
      "entropy": 5.7653028011322025,
      "epoch": 4.963424278370898,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002811114732115305,
      "loss": 4.936,
      "mean_token_accuracy": 0.23017161041498185,
      "num_tokens": 115006526.0,
      "step": 50210
    },
    {
      "entropy": 5.770809602737427,
      "epoch": 4.963918544879399,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002810761961746452,
      "loss": 4.9183,
      "mean_token_accuracy": 0.23486812561750411,
      "num_tokens": 115017863.0,
      "step": 50215
    },
    {
      "entropy": 5.801350116729736,
      "epoch": 4.9644128113879,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00028104091898828454,
      "loss": 4.9338,
      "mean_token_accuracy": 0.2330716520547867,
      "num_tokens": 115029597.0,
      "step": 50220
    },
    {
      "entropy": 5.7715836524963375,
      "epoch": 4.9649070778964015,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002810056416533162,
      "loss": 4.9222,
      "mean_token_accuracy": 0.23138584047555924,
      "num_tokens": 115041027.0,
      "step": 50225
    },
    {
      "entropy": 5.793311882019043,
      "epoch": 4.965401344404903,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002809703641706083,
      "loss": 4.9315,
      "mean_token_accuracy": 0.2287633612751961,
      "num_tokens": 115051508.0,
      "step": 50230
    },
    {
      "entropy": 5.706295824050903,
      "epoch": 4.965895610913405,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002809350865410284,
      "loss": 4.8438,
      "mean_token_accuracy": 0.23879338651895524,
      "num_tokens": 115061766.0,
      "step": 50235
    },
    {
      "entropy": 5.808468914031982,
      "epoch": 4.966389877421906,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002808998087654445,
      "loss": 4.9523,
      "mean_token_accuracy": 0.22999130189418793,
      "num_tokens": 115072858.0,
      "step": 50240
    },
    {
      "entropy": 5.874891996383667,
      "epoch": 4.9668841439304074,
      "grad_norm": 1.34375,
      "learning_rate": 0.00028086453084472443,
      "loss": 5.0388,
      "mean_token_accuracy": 0.21961769610643386,
      "num_tokens": 115084714.0,
      "step": 50245
    },
    {
      "entropy": 5.749356937408447,
      "epoch": 4.967378410438909,
      "grad_norm": 1.34375,
      "learning_rate": 0.00028082925277973596,
      "loss": 4.8929,
      "mean_token_accuracy": 0.23632720857858658,
      "num_tokens": 115095647.0,
      "step": 50250
    },
    {
      "entropy": 5.790010404586792,
      "epoch": 4.96787267694741,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002807939745713469,
      "loss": 4.924,
      "mean_token_accuracy": 0.22470076829195024,
      "num_tokens": 115106030.0,
      "step": 50255
    },
    {
      "entropy": 5.89292426109314,
      "epoch": 4.968366943455911,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002807586962204252,
      "loss": 5.0839,
      "mean_token_accuracy": 0.2159691721200943,
      "num_tokens": 115117941.0,
      "step": 50260
    },
    {
      "entropy": 5.792506361007691,
      "epoch": 4.9688612099644125,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002807234177278388,
      "loss": 4.8367,
      "mean_token_accuracy": 0.2352254331111908,
      "num_tokens": 115130230.0,
      "step": 50265
    },
    {
      "entropy": 5.779163408279419,
      "epoch": 4.969355476472915,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00028068813909445527,
      "loss": 4.9265,
      "mean_token_accuracy": 0.23493935614824296,
      "num_tokens": 115141925.0,
      "step": 50270
    },
    {
      "entropy": 5.744682455062867,
      "epoch": 4.969849742981416,
      "grad_norm": 1.359375,
      "learning_rate": 0.00028065286032114266,
      "loss": 4.9493,
      "mean_token_accuracy": 0.23053625524044036,
      "num_tokens": 115154295.0,
      "step": 50275
    },
    {
      "entropy": 5.730276679992675,
      "epoch": 4.970344009489917,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002806175814087689,
      "loss": 4.827,
      "mean_token_accuracy": 0.24380260407924653,
      "num_tokens": 115165133.0,
      "step": 50280
    },
    {
      "entropy": 5.806663894653321,
      "epoch": 4.970838275998418,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002805823023582017,
      "loss": 4.903,
      "mean_token_accuracy": 0.23219995349645614,
      "num_tokens": 115176570.0,
      "step": 50285
    },
    {
      "entropy": 5.793891382217407,
      "epoch": 4.97133254250692,
      "grad_norm": 1.40625,
      "learning_rate": 0.000280547023170309,
      "loss": 4.9861,
      "mean_token_accuracy": 0.2218492478132248,
      "num_tokens": 115188464.0,
      "step": 50290
    },
    {
      "entropy": 5.8527429580688475,
      "epoch": 4.971826809015421,
      "grad_norm": 1.3125,
      "learning_rate": 0.00028051174384595873,
      "loss": 5.0055,
      "mean_token_accuracy": 0.22439261823892592,
      "num_tokens": 115199422.0,
      "step": 50295
    },
    {
      "entropy": 5.793202495574951,
      "epoch": 4.972321075523922,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00028047646438601867,
      "loss": 4.9546,
      "mean_token_accuracy": 0.23006391227245332,
      "num_tokens": 115211281.0,
      "step": 50300
    },
    {
      "entropy": 5.883242273330689,
      "epoch": 4.972815342032424,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002804411847913567,
      "loss": 5.026,
      "mean_token_accuracy": 0.22225152850151061,
      "num_tokens": 115223194.0,
      "step": 50305
    },
    {
      "entropy": 5.80102949142456,
      "epoch": 4.973309608540926,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00028040590506284076,
      "loss": 4.9049,
      "mean_token_accuracy": 0.22802323549985887,
      "num_tokens": 115233919.0,
      "step": 50310
    },
    {
      "entropy": 5.882370281219482,
      "epoch": 4.973803875049427,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002803706252013387,
      "loss": 5.0482,
      "mean_token_accuracy": 0.223785500228405,
      "num_tokens": 115246245.0,
      "step": 50315
    },
    {
      "entropy": 5.753377103805542,
      "epoch": 4.974298141557928,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002803353452077185,
      "loss": 4.8445,
      "mean_token_accuracy": 0.24095330238342286,
      "num_tokens": 115257247.0,
      "step": 50320
    },
    {
      "entropy": 5.8253552436828615,
      "epoch": 4.974792408066429,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002803000650828479,
      "loss": 4.9763,
      "mean_token_accuracy": 0.23062942624092103,
      "num_tokens": 115268962.0,
      "step": 50325
    },
    {
      "entropy": 5.7928262710571286,
      "epoch": 4.975286674574931,
      "grad_norm": 1.3125,
      "learning_rate": 0.00028026478482759485,
      "loss": 4.8955,
      "mean_token_accuracy": 0.23038111478090287,
      "num_tokens": 115281034.0,
      "step": 50330
    },
    {
      "entropy": 5.811640977859497,
      "epoch": 4.975780941083432,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002802295044428274,
      "loss": 4.989,
      "mean_token_accuracy": 0.2286304324865341,
      "num_tokens": 115292904.0,
      "step": 50335
    },
    {
      "entropy": 5.863275480270386,
      "epoch": 4.976275207591933,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002801942239294133,
      "loss": 4.9547,
      "mean_token_accuracy": 0.22226293236017228,
      "num_tokens": 115304723.0,
      "step": 50340
    },
    {
      "entropy": 5.795854282379151,
      "epoch": 4.9767694741004345,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002801589432882204,
      "loss": 4.9765,
      "mean_token_accuracy": 0.22963757514953614,
      "num_tokens": 115316106.0,
      "step": 50345
    },
    {
      "entropy": 5.79200553894043,
      "epoch": 4.977263740608937,
      "grad_norm": 1.265625,
      "learning_rate": 0.00028012366252011674,
      "loss": 4.9415,
      "mean_token_accuracy": 0.2274552971124649,
      "num_tokens": 115328981.0,
      "step": 50350
    },
    {
      "entropy": 5.87086033821106,
      "epoch": 4.977758007117438,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002800883816259702,
      "loss": 4.9905,
      "mean_token_accuracy": 0.22318440824747085,
      "num_tokens": 115340655.0,
      "step": 50355
    },
    {
      "entropy": 5.707351016998291,
      "epoch": 4.978252273625939,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00028005310060664866,
      "loss": 4.8143,
      "mean_token_accuracy": 0.24641534239053725,
      "num_tokens": 115352130.0,
      "step": 50360
    },
    {
      "entropy": 5.759663867950439,
      "epoch": 4.97874654013444,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002800178194630201,
      "loss": 4.9147,
      "mean_token_accuracy": 0.22972965091466904,
      "num_tokens": 115363419.0,
      "step": 50365
    },
    {
      "entropy": 5.797353982925415,
      "epoch": 4.979240806642942,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00027998253819595234,
      "loss": 4.8559,
      "mean_token_accuracy": 0.23890891522169114,
      "num_tokens": 115374008.0,
      "step": 50370
    },
    {
      "entropy": 5.840556192398071,
      "epoch": 4.979735073151443,
      "grad_norm": 1.328125,
      "learning_rate": 0.00027994725680631336,
      "loss": 4.969,
      "mean_token_accuracy": 0.2232537493109703,
      "num_tokens": 115385829.0,
      "step": 50375
    },
    {
      "entropy": 5.776134586334228,
      "epoch": 4.980229339659944,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0002799119752949712,
      "loss": 4.9171,
      "mean_token_accuracy": 0.22979263812303544,
      "num_tokens": 115396747.0,
      "step": 50380
    },
    {
      "entropy": 5.770176029205322,
      "epoch": 4.980723606168446,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00027987669366279363,
      "loss": 4.8846,
      "mean_token_accuracy": 0.23460711985826493,
      "num_tokens": 115407436.0,
      "step": 50385
    },
    {
      "entropy": 5.820124626159668,
      "epoch": 4.981217872676948,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002798414119106487,
      "loss": 4.9435,
      "mean_token_accuracy": 0.23083589524030684,
      "num_tokens": 115418395.0,
      "step": 50390
    },
    {
      "entropy": 5.8134846687316895,
      "epoch": 4.981712139185449,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002798061300394041,
      "loss": 4.899,
      "mean_token_accuracy": 0.23319198191165924,
      "num_tokens": 115428501.0,
      "step": 50395
    },
    {
      "entropy": 5.745656681060791,
      "epoch": 4.98220640569395,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00027977084804992807,
      "loss": 4.8882,
      "mean_token_accuracy": 0.23176613003015517,
      "num_tokens": 115440557.0,
      "step": 50400
    },
    {
      "entropy": 5.834184360504151,
      "epoch": 4.982700672202451,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002797355659430885,
      "loss": 4.9913,
      "mean_token_accuracy": 0.22125499099493026,
      "num_tokens": 115451349.0,
      "step": 50405
    },
    {
      "entropy": 5.841407775878906,
      "epoch": 4.983194938710953,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002797002837197532,
      "loss": 4.932,
      "mean_token_accuracy": 0.22379605025053023,
      "num_tokens": 115461816.0,
      "step": 50410
    },
    {
      "entropy": 5.709989547729492,
      "epoch": 4.983689205219454,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002796650013807902,
      "loss": 4.8132,
      "mean_token_accuracy": 0.24233080297708512,
      "num_tokens": 115473670.0,
      "step": 50415
    },
    {
      "entropy": 5.769084835052491,
      "epoch": 4.984183471727956,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002796297189270675,
      "loss": 4.9353,
      "mean_token_accuracy": 0.23125441670417785,
      "num_tokens": 115485533.0,
      "step": 50420
    },
    {
      "entropy": 5.753753662109375,
      "epoch": 4.984677738236457,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00027959443635945295,
      "loss": 4.8606,
      "mean_token_accuracy": 0.23900100141763686,
      "num_tokens": 115497226.0,
      "step": 50425
    },
    {
      "entropy": 5.811904191970825,
      "epoch": 4.985172004744959,
      "grad_norm": 1.3125,
      "learning_rate": 0.00027955915367881444,
      "loss": 4.9319,
      "mean_token_accuracy": 0.23003616333007812,
      "num_tokens": 115509202.0,
      "step": 50430
    },
    {
      "entropy": 5.8007896900177,
      "epoch": 4.98566627125346,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002795238708860202,
      "loss": 4.9408,
      "mean_token_accuracy": 0.23013546466827392,
      "num_tokens": 115520548.0,
      "step": 50435
    },
    {
      "entropy": 5.73655686378479,
      "epoch": 4.986160537761961,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00027948858798193803,
      "loss": 4.8741,
      "mean_token_accuracy": 0.23425007313489915,
      "num_tokens": 115532016.0,
      "step": 50440
    },
    {
      "entropy": 5.814308404922485,
      "epoch": 4.986654804270462,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00027945330496743584,
      "loss": 4.9567,
      "mean_token_accuracy": 0.234225831925869,
      "num_tokens": 115542420.0,
      "step": 50445
    },
    {
      "entropy": 5.7929274559021,
      "epoch": 4.987149070778964,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002794180218433816,
      "loss": 4.8819,
      "mean_token_accuracy": 0.22814449220895766,
      "num_tokens": 115552947.0,
      "step": 50450
    },
    {
      "entropy": 5.776124095916748,
      "epoch": 4.987643337287466,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002793827386106435,
      "loss": 4.9331,
      "mean_token_accuracy": 0.23522318452596663,
      "num_tokens": 115564564.0,
      "step": 50455
    },
    {
      "entropy": 5.80662088394165,
      "epoch": 4.988137603795967,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002793474552700892,
      "loss": 4.9229,
      "mean_token_accuracy": 0.2334411084651947,
      "num_tokens": 115575235.0,
      "step": 50460
    },
    {
      "entropy": 5.843799114227295,
      "epoch": 4.988631870304468,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00027931217182258696,
      "loss": 4.989,
      "mean_token_accuracy": 0.22882246226072311,
      "num_tokens": 115587142.0,
      "step": 50465
    },
    {
      "entropy": 5.758025169372559,
      "epoch": 4.98912613681297,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002792768882690045,
      "loss": 4.9044,
      "mean_token_accuracy": 0.22891243696212768,
      "num_tokens": 115597895.0,
      "step": 50470
    },
    {
      "entropy": 5.857814979553223,
      "epoch": 4.989620403321471,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00027924160461020994,
      "loss": 4.9571,
      "mean_token_accuracy": 0.2320897176861763,
      "num_tokens": 115609606.0,
      "step": 50475
    },
    {
      "entropy": 5.747151708602905,
      "epoch": 4.990114669829972,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002792063208470713,
      "loss": 4.8583,
      "mean_token_accuracy": 0.24255991876125335,
      "num_tokens": 115620280.0,
      "step": 50480
    },
    {
      "entropy": 5.823554754257202,
      "epoch": 4.990608936338473,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002791710369804565,
      "loss": 4.94,
      "mean_token_accuracy": 0.23338314294815063,
      "num_tokens": 115631970.0,
      "step": 50485
    },
    {
      "entropy": 5.791849899291992,
      "epoch": 4.9911032028469755,
      "grad_norm": 1.3125,
      "learning_rate": 0.00027913575301123357,
      "loss": 4.8523,
      "mean_token_accuracy": 0.22632969170808792,
      "num_tokens": 115642186.0,
      "step": 50490
    },
    {
      "entropy": 5.754032802581787,
      "epoch": 4.991597469355477,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002791004689402704,
      "loss": 4.8641,
      "mean_token_accuracy": 0.2316341444849968,
      "num_tokens": 115653658.0,
      "step": 50495
    },
    {
      "entropy": 5.777803993225097,
      "epoch": 4.992091735863978,
      "grad_norm": 1.375,
      "learning_rate": 0.00027906518476843514,
      "loss": 4.8821,
      "mean_token_accuracy": 0.23468457013368607,
      "num_tokens": 115665517.0,
      "step": 50500
    },
    {
      "entropy": 5.784101486206055,
      "epoch": 4.992586002372479,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002790299004965956,
      "loss": 4.9131,
      "mean_token_accuracy": 0.233921679854393,
      "num_tokens": 115677317.0,
      "step": 50505
    },
    {
      "entropy": 5.848688459396362,
      "epoch": 4.9930802688809806,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00027899461612562005,
      "loss": 5.0241,
      "mean_token_accuracy": 0.2204493299126625,
      "num_tokens": 115689732.0,
      "step": 50510
    },
    {
      "entropy": 5.775991296768188,
      "epoch": 4.993574535389482,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00027895933165637615,
      "loss": 4.8777,
      "mean_token_accuracy": 0.23720050603151321,
      "num_tokens": 115701660.0,
      "step": 50515
    },
    {
      "entropy": 5.862777900695801,
      "epoch": 4.994068801897983,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002789240470897322,
      "loss": 4.9335,
      "mean_token_accuracy": 0.22486172318458558,
      "num_tokens": 115713113.0,
      "step": 50520
    },
    {
      "entropy": 5.7619537830352785,
      "epoch": 4.994563068406485,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000278888762426556,
      "loss": 4.9028,
      "mean_token_accuracy": 0.23057316541671752,
      "num_tokens": 115725613.0,
      "step": 50525
    },
    {
      "entropy": 5.7636620044708256,
      "epoch": 4.9950573349149865,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002788534776677157,
      "loss": 4.9216,
      "mean_token_accuracy": 0.2327973112463951,
      "num_tokens": 115737680.0,
      "step": 50530
    },
    {
      "entropy": 5.77488899230957,
      "epoch": 4.995551601423488,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002788181928140792,
      "loss": 4.9283,
      "mean_token_accuracy": 0.23227963000535964,
      "num_tokens": 115747847.0,
      "step": 50535
    },
    {
      "entropy": 5.801711511611939,
      "epoch": 4.996045867931989,
      "grad_norm": 1.359375,
      "learning_rate": 0.00027878290786651454,
      "loss": 4.9864,
      "mean_token_accuracy": 0.2283879041671753,
      "num_tokens": 115758679.0,
      "step": 50540
    },
    {
      "entropy": 5.896848678588867,
      "epoch": 4.99654013444049,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002787476228258898,
      "loss": 5.004,
      "mean_token_accuracy": 0.2216818302869797,
      "num_tokens": 115769496.0,
      "step": 50545
    },
    {
      "entropy": 5.799694347381592,
      "epoch": 4.9970344009489915,
      "grad_norm": 1.4375,
      "learning_rate": 0.000278712337693073,
      "loss": 4.8992,
      "mean_token_accuracy": 0.23330094665288925,
      "num_tokens": 115781422.0,
      "step": 50550
    },
    {
      "entropy": 5.833764219284058,
      "epoch": 4.997528667457493,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002786770524689321,
      "loss": 5.0305,
      "mean_token_accuracy": 0.22634673863649368,
      "num_tokens": 115792612.0,
      "step": 50555
    },
    {
      "entropy": 5.779271984100342,
      "epoch": 4.998022933965995,
      "grad_norm": 1.265625,
      "learning_rate": 0.00027864176715433517,
      "loss": 4.9542,
      "mean_token_accuracy": 0.2259409934282303,
      "num_tokens": 115805479.0,
      "step": 50560
    },
    {
      "entropy": 5.7756028175354,
      "epoch": 4.998517200474496,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00027860648175015006,
      "loss": 4.8415,
      "mean_token_accuracy": 0.24140482693910598,
      "num_tokens": 115817582.0,
      "step": 50565
    },
    {
      "entropy": 5.804584980010986,
      "epoch": 4.9990114669829975,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002785711962572451,
      "loss": 4.9351,
      "mean_token_accuracy": 0.23388731181621553,
      "num_tokens": 115829413.0,
      "step": 50570
    },
    {
      "entropy": 5.735230922698975,
      "epoch": 4.999505733491499,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00027853591067648805,
      "loss": 4.8583,
      "mean_token_accuracy": 0.23544179797172546,
      "num_tokens": 115840247.0,
      "step": 50575
    },
    {
      "entropy": 5.680647850036621,
      "epoch": 5.0,
      "grad_norm": 1.625,
      "learning_rate": 0.0002785006250087471,
      "loss": 4.7671,
      "mean_token_accuracy": 0.24508559405803682,
      "num_tokens": 115851460.0,
      "step": 50580
    },
    {
      "entropy": 5.790252256393432,
      "epoch": 5.000494266508501,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002784653392548902,
      "loss": 4.8662,
      "mean_token_accuracy": 0.24249252080917358,
      "num_tokens": 115863069.0,
      "step": 50585
    },
    {
      "entropy": 5.764139223098755,
      "epoch": 5.0009885330170025,
      "grad_norm": 1.34375,
      "learning_rate": 0.00027843005341578544,
      "loss": 4.8154,
      "mean_token_accuracy": 0.242041651904583,
      "num_tokens": 115875628.0,
      "step": 50590
    },
    {
      "entropy": 5.809588527679443,
      "epoch": 5.001482799525504,
      "grad_norm": 1.421875,
      "learning_rate": 0.00027839476749230084,
      "loss": 4.9477,
      "mean_token_accuracy": 0.22751238495111464,
      "num_tokens": 115886404.0,
      "step": 50595
    },
    {
      "entropy": 5.752597618103027,
      "epoch": 5.001977066034006,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002783594814853045,
      "loss": 4.8199,
      "mean_token_accuracy": 0.2419012278318405,
      "num_tokens": 115897927.0,
      "step": 50600
    },
    {
      "entropy": 5.781824684143066,
      "epoch": 5.002471332542507,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00027832419539566426,
      "loss": 4.9222,
      "mean_token_accuracy": 0.23271087408065796,
      "num_tokens": 115908908.0,
      "step": 50605
    },
    {
      "entropy": 5.7605384349822994,
      "epoch": 5.0029655990510085,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00027828890922424837,
      "loss": 4.8484,
      "mean_token_accuracy": 0.23954013139009475,
      "num_tokens": 115919783.0,
      "step": 50610
    },
    {
      "entropy": 5.846383857727051,
      "epoch": 5.00345986555951,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002782536229719248,
      "loss": 4.9161,
      "mean_token_accuracy": 0.22566650062799454,
      "num_tokens": 115931939.0,
      "step": 50615
    },
    {
      "entropy": 5.812274026870727,
      "epoch": 5.003954132068011,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002782183366395616,
      "loss": 4.9171,
      "mean_token_accuracy": 0.23325273692607879,
      "num_tokens": 115943606.0,
      "step": 50620
    },
    {
      "entropy": 5.636936521530151,
      "epoch": 5.004448398576512,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002781830502280268,
      "loss": 4.7947,
      "mean_token_accuracy": 0.24648771584033966,
      "num_tokens": 115955121.0,
      "step": 50625
    },
    {
      "entropy": 5.773365068435669,
      "epoch": 5.0049426650850135,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00027814776373818854,
      "loss": 4.8971,
      "mean_token_accuracy": 0.2371148183941841,
      "num_tokens": 115965720.0,
      "step": 50630
    },
    {
      "entropy": 5.81085057258606,
      "epoch": 5.005436931593516,
      "grad_norm": 1.28125,
      "learning_rate": 0.00027811247717091466,
      "loss": 4.9044,
      "mean_token_accuracy": 0.23276356011629104,
      "num_tokens": 115977318.0,
      "step": 50635
    },
    {
      "entropy": 5.706590890884399,
      "epoch": 5.005931198102017,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002780771905270735,
      "loss": 4.816,
      "mean_token_accuracy": 0.24623806774616241,
      "num_tokens": 115989541.0,
      "step": 50640
    },
    {
      "entropy": 5.705309677124023,
      "epoch": 5.006425464610518,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002780419038075329,
      "loss": 4.7189,
      "mean_token_accuracy": 0.25526981949806216,
      "num_tokens": 116000227.0,
      "step": 50645
    },
    {
      "entropy": 5.821117162704468,
      "epoch": 5.0069197311190194,
      "grad_norm": 1.34375,
      "learning_rate": 0.00027800661701316096,
      "loss": 4.9536,
      "mean_token_accuracy": 0.2256518065929413,
      "num_tokens": 116012822.0,
      "step": 50650
    },
    {
      "entropy": 5.886148977279663,
      "epoch": 5.007413997627521,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002779713301448259,
      "loss": 4.9913,
      "mean_token_accuracy": 0.2290161594748497,
      "num_tokens": 116024113.0,
      "step": 50655
    },
    {
      "entropy": 5.764233112335205,
      "epoch": 5.007908264136022,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00027793604320339556,
      "loss": 4.8759,
      "mean_token_accuracy": 0.24002205282449723,
      "num_tokens": 116035381.0,
      "step": 50660
    },
    {
      "entropy": 5.756098222732544,
      "epoch": 5.008402530644523,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002779007561897381,
      "loss": 4.8114,
      "mean_token_accuracy": 0.2432310089468956,
      "num_tokens": 116045673.0,
      "step": 50665
    },
    {
      "entropy": 5.8016825199127195,
      "epoch": 5.0088967971530245,
      "grad_norm": 1.328125,
      "learning_rate": 0.00027786546910472167,
      "loss": 4.9509,
      "mean_token_accuracy": 0.23153091967105865,
      "num_tokens": 116057010.0,
      "step": 50670
    },
    {
      "entropy": 5.804245376586914,
      "epoch": 5.009391063661527,
      "grad_norm": 1.328125,
      "learning_rate": 0.00027783018194921423,
      "loss": 4.9009,
      "mean_token_accuracy": 0.22945568859577178,
      "num_tokens": 116067197.0,
      "step": 50675
    },
    {
      "entropy": 5.785934638977051,
      "epoch": 5.009885330170028,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002777948947240838,
      "loss": 4.9018,
      "mean_token_accuracy": 0.23685529381036757,
      "num_tokens": 116080409.0,
      "step": 50680
    },
    {
      "entropy": 5.813837146759033,
      "epoch": 5.010379596678529,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00027775960743019857,
      "loss": 4.9412,
      "mean_token_accuracy": 0.22483682483434678,
      "num_tokens": 116091935.0,
      "step": 50685
    },
    {
      "entropy": 5.828999471664429,
      "epoch": 5.01087386318703,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002777243200684266,
      "loss": 4.9161,
      "mean_token_accuracy": 0.2295263960957527,
      "num_tokens": 116103326.0,
      "step": 50690
    },
    {
      "entropy": 5.735544109344483,
      "epoch": 5.011368129695532,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002776890326396359,
      "loss": 4.7934,
      "mean_token_accuracy": 0.24206490367650985,
      "num_tokens": 116114711.0,
      "step": 50695
    },
    {
      "entropy": 5.717087125778198,
      "epoch": 5.011862396204033,
      "grad_norm": 1.296875,
      "learning_rate": 0.00027765374514469453,
      "loss": 4.8552,
      "mean_token_accuracy": 0.24208480417728423,
      "num_tokens": 116126023.0,
      "step": 50700
    },
    {
      "entropy": 5.760784864425659,
      "epoch": 5.012356662712534,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002776184575844707,
      "loss": 4.8811,
      "mean_token_accuracy": 0.23942002654075623,
      "num_tokens": 116137847.0,
      "step": 50705
    },
    {
      "entropy": 5.7322986125946045,
      "epoch": 5.012850929221036,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00027758316995983237,
      "loss": 4.8455,
      "mean_token_accuracy": 0.2383717492222786,
      "num_tokens": 116148635.0,
      "step": 50710
    },
    {
      "entropy": 5.7208997249603275,
      "epoch": 5.013345195729538,
      "grad_norm": 1.359375,
      "learning_rate": 0.00027754788227164767,
      "loss": 4.8511,
      "mean_token_accuracy": 0.24311988055706024,
      "num_tokens": 116159312.0,
      "step": 50715
    },
    {
      "entropy": 5.858943319320678,
      "epoch": 5.013839462238039,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00027751259452078465,
      "loss": 4.9907,
      "mean_token_accuracy": 0.22615025341510772,
      "num_tokens": 116171754.0,
      "step": 50720
    },
    {
      "entropy": 5.857482004165649,
      "epoch": 5.01433372874654,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002774773067081114,
      "loss": 4.9496,
      "mean_token_accuracy": 0.22050189226865768,
      "num_tokens": 116182691.0,
      "step": 50725
    },
    {
      "entropy": 5.823702573776245,
      "epoch": 5.014827995255041,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002774420188344961,
      "loss": 4.9983,
      "mean_token_accuracy": 0.2181121051311493,
      "num_tokens": 116194580.0,
      "step": 50730
    },
    {
      "entropy": 5.809158611297607,
      "epoch": 5.015322261763543,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002774067309008067,
      "loss": 4.9455,
      "mean_token_accuracy": 0.22156823724508284,
      "num_tokens": 116205634.0,
      "step": 50735
    },
    {
      "entropy": 5.783737230300903,
      "epoch": 5.015816528272044,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00027737144290791135,
      "loss": 4.8842,
      "mean_token_accuracy": 0.23593875169754028,
      "num_tokens": 116217246.0,
      "step": 50740
    },
    {
      "entropy": 5.842929887771606,
      "epoch": 5.016310794780546,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00027733615485667825,
      "loss": 4.9873,
      "mean_token_accuracy": 0.2224092662334442,
      "num_tokens": 116229725.0,
      "step": 50745
    },
    {
      "entropy": 5.747458219528198,
      "epoch": 5.016805061289047,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00027730086674797517,
      "loss": 4.8616,
      "mean_token_accuracy": 0.24245306700468064,
      "num_tokens": 116240706.0,
      "step": 50750
    },
    {
      "entropy": 5.791763114929199,
      "epoch": 5.017299327797549,
      "grad_norm": 1.28125,
      "learning_rate": 0.00027726557858267055,
      "loss": 4.8758,
      "mean_token_accuracy": 0.23667625337839127,
      "num_tokens": 116251750.0,
      "step": 50755
    },
    {
      "entropy": 5.853065919876099,
      "epoch": 5.01779359430605,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002772302903616323,
      "loss": 4.9191,
      "mean_token_accuracy": 0.22910383939743043,
      "num_tokens": 116263116.0,
      "step": 50760
    },
    {
      "entropy": 5.87770676612854,
      "epoch": 5.018287860814551,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00027719500208572855,
      "loss": 4.9871,
      "mean_token_accuracy": 0.22318989038467407,
      "num_tokens": 116274425.0,
      "step": 50765
    },
    {
      "entropy": 5.781600522994995,
      "epoch": 5.018782127323052,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002771597137558274,
      "loss": 4.8569,
      "mean_token_accuracy": 0.24118724912405015,
      "num_tokens": 116287535.0,
      "step": 50770
    },
    {
      "entropy": 5.768781423568726,
      "epoch": 5.019276393831554,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000277124425372797,
      "loss": 4.9089,
      "mean_token_accuracy": 0.2295445665717125,
      "num_tokens": 116299530.0,
      "step": 50775
    },
    {
      "entropy": 5.761230325698852,
      "epoch": 5.019770660340055,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0002770891369375054,
      "loss": 4.9352,
      "mean_token_accuracy": 0.2340790331363678,
      "num_tokens": 116311114.0,
      "step": 50780
    },
    {
      "entropy": 5.790247249603271,
      "epoch": 5.020264926848557,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002770538484508207,
      "loss": 4.9176,
      "mean_token_accuracy": 0.23376774787902832,
      "num_tokens": 116322149.0,
      "step": 50785
    },
    {
      "entropy": 5.69867730140686,
      "epoch": 5.020759193357058,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002770185599136111,
      "loss": 4.8104,
      "mean_token_accuracy": 0.23930883705615996,
      "num_tokens": 116333905.0,
      "step": 50790
    },
    {
      "entropy": 5.842459201812744,
      "epoch": 5.02125345986556,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002769832713267445,
      "loss": 4.9055,
      "mean_token_accuracy": 0.23321451842784882,
      "num_tokens": 116344415.0,
      "step": 50795
    },
    {
      "entropy": 5.847038221359253,
      "epoch": 5.021747726374061,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002769479826910891,
      "loss": 5.0301,
      "mean_token_accuracy": 0.22335854768753052,
      "num_tokens": 116358272.0,
      "step": 50800
    },
    {
      "entropy": 5.836552143096924,
      "epoch": 5.022241992882562,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00027691269400751306,
      "loss": 4.9463,
      "mean_token_accuracy": 0.22245769649744035,
      "num_tokens": 116370880.0,
      "step": 50805
    },
    {
      "entropy": 5.799551248550415,
      "epoch": 5.022736259391063,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00027687740527688445,
      "loss": 4.9103,
      "mean_token_accuracy": 0.23957173824310302,
      "num_tokens": 116382525.0,
      "step": 50810
    },
    {
      "entropy": 5.763939380645752,
      "epoch": 5.023230525899565,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002768421165000714,
      "loss": 4.8467,
      "mean_token_accuracy": 0.23824166208505632,
      "num_tokens": 116393810.0,
      "step": 50815
    },
    {
      "entropy": 5.7464484691619875,
      "epoch": 5.023724792408067,
      "grad_norm": 1.375,
      "learning_rate": 0.000276806827677942,
      "loss": 4.8797,
      "mean_token_accuracy": 0.2380839541554451,
      "num_tokens": 116405935.0,
      "step": 50820
    },
    {
      "entropy": 5.814163541793823,
      "epoch": 5.024219058916568,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002767715388113643,
      "loss": 4.9537,
      "mean_token_accuracy": 0.23389680832624435,
      "num_tokens": 116416026.0,
      "step": 50825
    },
    {
      "entropy": 5.820384693145752,
      "epoch": 5.024713325425069,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00027673624990120653,
      "loss": 4.9567,
      "mean_token_accuracy": 0.22330318242311478,
      "num_tokens": 116426387.0,
      "step": 50830
    },
    {
      "entropy": 5.851880598068237,
      "epoch": 5.025207591933571,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00027670096094833675,
      "loss": 4.9937,
      "mean_token_accuracy": 0.22774713933467866,
      "num_tokens": 116438793.0,
      "step": 50835
    },
    {
      "entropy": 5.834858417510986,
      "epoch": 5.025701858442072,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000276665671953623,
      "loss": 4.8818,
      "mean_token_accuracy": 0.23656534552574157,
      "num_tokens": 116451524.0,
      "step": 50840
    },
    {
      "entropy": 5.768569231033325,
      "epoch": 5.026196124950573,
      "grad_norm": 1.3125,
      "learning_rate": 0.00027663038291793357,
      "loss": 4.895,
      "mean_token_accuracy": 0.2359624296426773,
      "num_tokens": 116462856.0,
      "step": 50845
    },
    {
      "entropy": 5.782940530776978,
      "epoch": 5.026690391459074,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00027659509384213635,
      "loss": 4.9199,
      "mean_token_accuracy": 0.2310938686132431,
      "num_tokens": 116473355.0,
      "step": 50850
    },
    {
      "entropy": 5.8317756175994875,
      "epoch": 5.0271846579675765,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002765598047270996,
      "loss": 4.9158,
      "mean_token_accuracy": 0.23415391147136688,
      "num_tokens": 116485286.0,
      "step": 50855
    },
    {
      "entropy": 5.82098879814148,
      "epoch": 5.027678924476078,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002765245155736915,
      "loss": 4.9454,
      "mean_token_accuracy": 0.23094826936721802,
      "num_tokens": 116497962.0,
      "step": 50860
    },
    {
      "entropy": 5.7679750442504885,
      "epoch": 5.028173190984579,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00027648922638278,
      "loss": 4.8679,
      "mean_token_accuracy": 0.2334793269634247,
      "num_tokens": 116510085.0,
      "step": 50865
    },
    {
      "entropy": 5.685627031326294,
      "epoch": 5.02866745749308,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002764539371552333,
      "loss": 4.7505,
      "mean_token_accuracy": 0.25049273371696473,
      "num_tokens": 116520547.0,
      "step": 50870
    },
    {
      "entropy": 5.750125360488892,
      "epoch": 5.029161724001582,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002764186478919195,
      "loss": 4.8902,
      "mean_token_accuracy": 0.23526969701051711,
      "num_tokens": 116531181.0,
      "step": 50875
    },
    {
      "entropy": 5.7176519393920895,
      "epoch": 5.029655990510083,
      "grad_norm": 1.328125,
      "learning_rate": 0.00027638335859370676,
      "loss": 4.8572,
      "mean_token_accuracy": 0.24460895955562592,
      "num_tokens": 116543262.0,
      "step": 50880
    },
    {
      "entropy": 5.780966091156006,
      "epoch": 5.030150257018584,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002763480692614632,
      "loss": 4.9134,
      "mean_token_accuracy": 0.23494441211223602,
      "num_tokens": 116555942.0,
      "step": 50885
    },
    {
      "entropy": 5.7937249660491945,
      "epoch": 5.030644523527086,
      "grad_norm": 1.34375,
      "learning_rate": 0.00027631277989605704,
      "loss": 4.9094,
      "mean_token_accuracy": 0.2349216341972351,
      "num_tokens": 116566964.0,
      "step": 50890
    },
    {
      "entropy": 5.763413000106811,
      "epoch": 5.0311387900355875,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002762774904983561,
      "loss": 4.8207,
      "mean_token_accuracy": 0.23895571380853653,
      "num_tokens": 116578731.0,
      "step": 50895
    },
    {
      "entropy": 5.775255441665649,
      "epoch": 5.031633056544089,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00027624220106922876,
      "loss": 4.8862,
      "mean_token_accuracy": 0.2385069340467453,
      "num_tokens": 116590010.0,
      "step": 50900
    },
    {
      "entropy": 5.800702953338623,
      "epoch": 5.03212732305259,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00027620691160954315,
      "loss": 4.8779,
      "mean_token_accuracy": 0.23042651414871215,
      "num_tokens": 116601477.0,
      "step": 50905
    },
    {
      "entropy": 5.838814210891724,
      "epoch": 5.032621589561091,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002761716221201672,
      "loss": 4.9078,
      "mean_token_accuracy": 0.2336578533053398,
      "num_tokens": 116612543.0,
      "step": 50910
    },
    {
      "entropy": 5.763749599456787,
      "epoch": 5.0331158560695926,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002761363326019693,
      "loss": 5.0075,
      "mean_token_accuracy": 0.23233336061239243,
      "num_tokens": 116626754.0,
      "step": 50915
    },
    {
      "entropy": 5.7157008171081545,
      "epoch": 5.033610122578094,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002761010430558173,
      "loss": 4.8178,
      "mean_token_accuracy": 0.24412799179553984,
      "num_tokens": 116636486.0,
      "step": 50920
    },
    {
      "entropy": 5.796712589263916,
      "epoch": 5.034104389086595,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002760657534825796,
      "loss": 4.9187,
      "mean_token_accuracy": 0.23088960349559784,
      "num_tokens": 116647455.0,
      "step": 50925
    },
    {
      "entropy": 5.8264930725097654,
      "epoch": 5.034598655595097,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00027603046388312416,
      "loss": 4.9446,
      "mean_token_accuracy": 0.22876252979040146,
      "num_tokens": 116659505.0,
      "step": 50930
    },
    {
      "entropy": 5.787563133239746,
      "epoch": 5.0350929221035985,
      "grad_norm": 1.3125,
      "learning_rate": 0.00027599517425831916,
      "loss": 4.8807,
      "mean_token_accuracy": 0.23339801281690598,
      "num_tokens": 116672275.0,
      "step": 50935
    },
    {
      "entropy": 5.749746179580688,
      "epoch": 5.0355871886121,
      "grad_norm": 1.390625,
      "learning_rate": 0.00027595988460903266,
      "loss": 4.8473,
      "mean_token_accuracy": 0.234682796895504,
      "num_tokens": 116684834.0,
      "step": 50940
    },
    {
      "entropy": 5.845160961151123,
      "epoch": 5.036081455120601,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002759245949361329,
      "loss": 4.9744,
      "mean_token_accuracy": 0.23578402847051622,
      "num_tokens": 116696651.0,
      "step": 50945
    },
    {
      "entropy": 5.764265871047973,
      "epoch": 5.036575721629102,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002758893052404879,
      "loss": 4.8225,
      "mean_token_accuracy": 0.2406596377491951,
      "num_tokens": 116707726.0,
      "step": 50950
    },
    {
      "entropy": 5.843689775466919,
      "epoch": 5.0370699881376035,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002758540155229659,
      "loss": 4.9886,
      "mean_token_accuracy": 0.22216079384088516,
      "num_tokens": 116720014.0,
      "step": 50955
    },
    {
      "entropy": 5.750387573242188,
      "epoch": 5.037564254646105,
      "grad_norm": 1.359375,
      "learning_rate": 0.000275818725784435,
      "loss": 4.8278,
      "mean_token_accuracy": 0.2400175452232361,
      "num_tokens": 116731361.0,
      "step": 50960
    },
    {
      "entropy": 5.775577068328857,
      "epoch": 5.038058521154607,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002757834360257633,
      "loss": 4.9043,
      "mean_token_accuracy": 0.2347245082259178,
      "num_tokens": 116742962.0,
      "step": 50965
    },
    {
      "entropy": 5.757082319259643,
      "epoch": 5.038552787663108,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002757481462478189,
      "loss": 4.8924,
      "mean_token_accuracy": 0.23865432292222977,
      "num_tokens": 116756177.0,
      "step": 50970
    },
    {
      "entropy": 5.783158016204834,
      "epoch": 5.0390470541716095,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00027571285645147013,
      "loss": 4.8865,
      "mean_token_accuracy": 0.2359623685479164,
      "num_tokens": 116766671.0,
      "step": 50975
    },
    {
      "entropy": 5.774899101257324,
      "epoch": 5.039541320680111,
      "grad_norm": 1.53125,
      "learning_rate": 0.00027567756663758494,
      "loss": 4.8961,
      "mean_token_accuracy": 0.22883131057024003,
      "num_tokens": 116777815.0,
      "step": 50980
    },
    {
      "entropy": 5.762152051925659,
      "epoch": 5.040035587188612,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002756422768070314,
      "loss": 4.8475,
      "mean_token_accuracy": 0.23789254873991011,
      "num_tokens": 116788125.0,
      "step": 50985
    },
    {
      "entropy": 5.815701723098755,
      "epoch": 5.040529853697113,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00027560698696067775,
      "loss": 4.9605,
      "mean_token_accuracy": 0.22900315672159194,
      "num_tokens": 116800548.0,
      "step": 50990
    },
    {
      "entropy": 5.791468000411987,
      "epoch": 5.0410241202056145,
      "grad_norm": 1.265625,
      "learning_rate": 0.00027557169709939227,
      "loss": 4.8125,
      "mean_token_accuracy": 0.24385957270860673,
      "num_tokens": 116812807.0,
      "step": 50995
    },
    {
      "entropy": 5.829205083847046,
      "epoch": 5.041518386714117,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00027553640722404284,
      "loss": 4.9503,
      "mean_token_accuracy": 0.22996021211147308,
      "num_tokens": 116823751.0,
      "step": 51000
    },
    {
      "epoch": 5.041518386714117,
      "eval_entropy": 5.506667784776488,
      "eval_loss": 4.990108489990234,
      "eval_mean_token_accuracy": 0.2360359296037702,
      "eval_num_tokens": 116823751.0,
      "eval_runtime": 20.5791,
      "eval_samples_per_second": 1579.901,
      "eval_steps_per_second": 197.53,
      "step": 51000
    },
    {
      "entropy": 5.755015087127686,
      "epoch": 5.042012653222618,
      "grad_norm": 1.328125,
      "learning_rate": 0.00027550111733549775,
      "loss": 4.8912,
      "mean_token_accuracy": 0.2353619173169136,
      "num_tokens": 116834903.0,
      "step": 51005
    },
    {
      "entropy": 5.756914901733398,
      "epoch": 5.042506919731119,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00027546582743462507,
      "loss": 4.8477,
      "mean_token_accuracy": 0.23741211891174316,
      "num_tokens": 116846399.0,
      "step": 51010
    },
    {
      "entropy": 5.820778894424438,
      "epoch": 5.0430011862396205,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00027543053752229296,
      "loss": 4.8768,
      "mean_token_accuracy": 0.23498898446559907,
      "num_tokens": 116858114.0,
      "step": 51015
    },
    {
      "entropy": 5.784293079376221,
      "epoch": 5.043495452748122,
      "grad_norm": 1.25,
      "learning_rate": 0.0002753952475993696,
      "loss": 4.9113,
      "mean_token_accuracy": 0.2370942622423172,
      "num_tokens": 116869447.0,
      "step": 51020
    },
    {
      "entropy": 5.754377174377441,
      "epoch": 5.043989719256623,
      "grad_norm": 1.359375,
      "learning_rate": 0.00027535995766672313,
      "loss": 4.8334,
      "mean_token_accuracy": 0.23920430094003678,
      "num_tokens": 116880527.0,
      "step": 51025
    },
    {
      "entropy": 5.71414852142334,
      "epoch": 5.044483985765124,
      "grad_norm": 1.4375,
      "learning_rate": 0.00027532466772522155,
      "loss": 4.8431,
      "mean_token_accuracy": 0.23739852756261826,
      "num_tokens": 116892521.0,
      "step": 51030
    },
    {
      "entropy": 5.785099267959595,
      "epoch": 5.0449782522736255,
      "grad_norm": 1.28125,
      "learning_rate": 0.000275289377775733,
      "loss": 4.9662,
      "mean_token_accuracy": 0.23111690431833268,
      "num_tokens": 116905032.0,
      "step": 51035
    },
    {
      "entropy": 5.769398593902588,
      "epoch": 5.045472518782128,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002752540878191259,
      "loss": 4.9089,
      "mean_token_accuracy": 0.23312829285860062,
      "num_tokens": 116916388.0,
      "step": 51040
    },
    {
      "entropy": 5.801865148544311,
      "epoch": 5.045966785290629,
      "grad_norm": 1.5,
      "learning_rate": 0.0002752187978562681,
      "loss": 4.8578,
      "mean_token_accuracy": 0.23967996686697007,
      "num_tokens": 116928825.0,
      "step": 51045
    },
    {
      "entropy": 5.883091020584106,
      "epoch": 5.04646105179913,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00027518350788802794,
      "loss": 4.9819,
      "mean_token_accuracy": 0.22765196561813356,
      "num_tokens": 116940230.0,
      "step": 51050
    },
    {
      "entropy": 5.810816860198974,
      "epoch": 5.046955318307631,
      "grad_norm": 1.578125,
      "learning_rate": 0.00027514821791527335,
      "loss": 4.8819,
      "mean_token_accuracy": 0.2356828659772873,
      "num_tokens": 116951967.0,
      "step": 51055
    },
    {
      "entropy": 5.884723377227783,
      "epoch": 5.047449584816133,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002751129279388726,
      "loss": 5.009,
      "mean_token_accuracy": 0.2252824619412422,
      "num_tokens": 116962727.0,
      "step": 51060
    },
    {
      "entropy": 5.790974569320679,
      "epoch": 5.047943851324634,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002750776379596938,
      "loss": 4.8414,
      "mean_token_accuracy": 0.24004556834697724,
      "num_tokens": 116974429.0,
      "step": 51065
    },
    {
      "entropy": 5.7647583961486815,
      "epoch": 5.048438117833135,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00027504234797860516,
      "loss": 4.9206,
      "mean_token_accuracy": 0.23701458275318146,
      "num_tokens": 116986095.0,
      "step": 51070
    },
    {
      "entropy": 5.764050531387329,
      "epoch": 5.048932384341637,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002750070579964747,
      "loss": 4.8604,
      "mean_token_accuracy": 0.24363618493080139,
      "num_tokens": 116997725.0,
      "step": 51075
    },
    {
      "entropy": 5.747545289993286,
      "epoch": 5.049426650850139,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002749717680141707,
      "loss": 4.8313,
      "mean_token_accuracy": 0.2449190154671669,
      "num_tokens": 117007764.0,
      "step": 51080
    },
    {
      "entropy": 5.750375843048095,
      "epoch": 5.04992091735864,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002749364780325611,
      "loss": 4.8838,
      "mean_token_accuracy": 0.24095782339572908,
      "num_tokens": 117019289.0,
      "step": 51085
    },
    {
      "entropy": 5.766412782669067,
      "epoch": 5.050415183867141,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002749011880525142,
      "loss": 4.8932,
      "mean_token_accuracy": 0.23451274037361144,
      "num_tokens": 117030397.0,
      "step": 51090
    },
    {
      "entropy": 5.763336563110352,
      "epoch": 5.050909450375642,
      "grad_norm": 1.1875,
      "learning_rate": 0.000274865898074898,
      "loss": 4.8867,
      "mean_token_accuracy": 0.23976435214281083,
      "num_tokens": 117042224.0,
      "step": 51095
    },
    {
      "entropy": 5.825167512893676,
      "epoch": 5.051403716884144,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00027483060810058085,
      "loss": 4.9442,
      "mean_token_accuracy": 0.23388280868530273,
      "num_tokens": 117053172.0,
      "step": 51100
    },
    {
      "entropy": 5.808302688598633,
      "epoch": 5.051897983392645,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00027479531813043073,
      "loss": 4.9268,
      "mean_token_accuracy": 0.23059697598218917,
      "num_tokens": 117065171.0,
      "step": 51105
    },
    {
      "entropy": 5.763390588760376,
      "epoch": 5.052392249901147,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002747600281653158,
      "loss": 4.9038,
      "mean_token_accuracy": 0.24061799347400664,
      "num_tokens": 117076718.0,
      "step": 51110
    },
    {
      "entropy": 5.788515520095825,
      "epoch": 5.052886516409648,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002747247382061042,
      "loss": 4.9157,
      "mean_token_accuracy": 0.2341613695025444,
      "num_tokens": 117089439.0,
      "step": 51115
    },
    {
      "entropy": 5.8271935939788815,
      "epoch": 5.05338078291815,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00027468944825366406,
      "loss": 4.9079,
      "mean_token_accuracy": 0.22733000665903091,
      "num_tokens": 117102022.0,
      "step": 51120
    },
    {
      "entropy": 5.786767768859863,
      "epoch": 5.053875049426651,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00027465415830886357,
      "loss": 4.9048,
      "mean_token_accuracy": 0.23422137200832366,
      "num_tokens": 117113379.0,
      "step": 51125
    },
    {
      "entropy": 5.7811966896057125,
      "epoch": 5.054369315935152,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0002746188683725709,
      "loss": 4.851,
      "mean_token_accuracy": 0.2377065345644951,
      "num_tokens": 117123859.0,
      "step": 51130
    },
    {
      "entropy": 5.867577028274536,
      "epoch": 5.054863582443653,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00027458357844565407,
      "loss": 4.9621,
      "mean_token_accuracy": 0.22797959446907043,
      "num_tokens": 117134556.0,
      "step": 51135
    },
    {
      "entropy": 5.768825483322144,
      "epoch": 5.055357848952155,
      "grad_norm": 1.375,
      "learning_rate": 0.00027454828852898136,
      "loss": 4.839,
      "mean_token_accuracy": 0.2465412899851799,
      "num_tokens": 117146210.0,
      "step": 51140
    },
    {
      "entropy": 5.748956871032715,
      "epoch": 5.055852115460657,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002745129986234207,
      "loss": 4.8994,
      "mean_token_accuracy": 0.2402205228805542,
      "num_tokens": 117158932.0,
      "step": 51145
    },
    {
      "entropy": 5.7720640182495115,
      "epoch": 5.056346381969158,
      "grad_norm": 1.546875,
      "learning_rate": 0.00027447770872984035,
      "loss": 4.9323,
      "mean_token_accuracy": 0.22607089728116989,
      "num_tokens": 117170591.0,
      "step": 51150
    },
    {
      "entropy": 5.770310401916504,
      "epoch": 5.056840648477659,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002744424188491085,
      "loss": 4.9409,
      "mean_token_accuracy": 0.22987632006406783,
      "num_tokens": 117182513.0,
      "step": 51155
    },
    {
      "entropy": 5.788483905792236,
      "epoch": 5.057334914986161,
      "grad_norm": 1.390625,
      "learning_rate": 0.00027440712898209324,
      "loss": 4.8643,
      "mean_token_accuracy": 0.23142679929733276,
      "num_tokens": 117193924.0,
      "step": 51160
    },
    {
      "entropy": 5.752580213546753,
      "epoch": 5.057829181494662,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002743718391296627,
      "loss": 4.8662,
      "mean_token_accuracy": 0.23860539942979814,
      "num_tokens": 117204309.0,
      "step": 51165
    },
    {
      "entropy": 5.837085962295532,
      "epoch": 5.058323448003163,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00027433654929268504,
      "loss": 4.9532,
      "mean_token_accuracy": 0.233220411837101,
      "num_tokens": 117215936.0,
      "step": 51170
    },
    {
      "entropy": 5.735549640655518,
      "epoch": 5.058817714511664,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002743012594720283,
      "loss": 4.7975,
      "mean_token_accuracy": 0.24481100142002105,
      "num_tokens": 117226392.0,
      "step": 51175
    },
    {
      "entropy": 5.703374242782592,
      "epoch": 5.059311981020166,
      "grad_norm": 1.375,
      "learning_rate": 0.0002742659696685607,
      "loss": 4.8227,
      "mean_token_accuracy": 0.24700641483068467,
      "num_tokens": 117238149.0,
      "step": 51180
    },
    {
      "entropy": 5.750212240219116,
      "epoch": 5.059806247528668,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00027423067988315036,
      "loss": 4.8946,
      "mean_token_accuracy": 0.23320370465517043,
      "num_tokens": 117248686.0,
      "step": 51185
    },
    {
      "entropy": 5.827206230163574,
      "epoch": 5.060300514037169,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00027419539011666545,
      "loss": 4.9388,
      "mean_token_accuracy": 0.22561368644237517,
      "num_tokens": 117260894.0,
      "step": 51190
    },
    {
      "entropy": 5.773357725143432,
      "epoch": 5.06079478054567,
      "grad_norm": 1.34375,
      "learning_rate": 0.00027416010036997407,
      "loss": 4.8991,
      "mean_token_accuracy": 0.24014170616865158,
      "num_tokens": 117272860.0,
      "step": 51195
    },
    {
      "entropy": 5.7768370628356935,
      "epoch": 5.061289047054172,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00027412481064394427,
      "loss": 4.9092,
      "mean_token_accuracy": 0.2359883576631546,
      "num_tokens": 117285068.0,
      "step": 51200
    },
    {
      "entropy": 5.826574420928955,
      "epoch": 5.061783313562673,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002740895209394444,
      "loss": 4.974,
      "mean_token_accuracy": 0.21939457803964615,
      "num_tokens": 117297451.0,
      "step": 51205
    },
    {
      "entropy": 5.783771896362305,
      "epoch": 5.062277580071174,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002740542312573423,
      "loss": 4.8556,
      "mean_token_accuracy": 0.23927167057991028,
      "num_tokens": 117307257.0,
      "step": 51210
    },
    {
      "entropy": 5.775034666061401,
      "epoch": 5.062771846579675,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002740189415985063,
      "loss": 4.8754,
      "mean_token_accuracy": 0.23333913683891297,
      "num_tokens": 117317920.0,
      "step": 51215
    },
    {
      "entropy": 5.783985137939453,
      "epoch": 5.0632661130881775,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002739836519638046,
      "loss": 4.927,
      "mean_token_accuracy": 0.23144375085830687,
      "num_tokens": 117328858.0,
      "step": 51220
    },
    {
      "entropy": 5.7549608707427975,
      "epoch": 5.063760379596679,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002739483623541051,
      "loss": 4.8941,
      "mean_token_accuracy": 0.23410273641347884,
      "num_tokens": 117340155.0,
      "step": 51225
    },
    {
      "entropy": 5.842346143722534,
      "epoch": 5.06425464610518,
      "grad_norm": 1.21875,
      "learning_rate": 0.00027391307277027616,
      "loss": 4.9146,
      "mean_token_accuracy": 0.23020991086959838,
      "num_tokens": 117351905.0,
      "step": 51230
    },
    {
      "entropy": 5.863545560836792,
      "epoch": 5.064748912613681,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002738777832131857,
      "loss": 4.9519,
      "mean_token_accuracy": 0.22844569981098176,
      "num_tokens": 117363415.0,
      "step": 51235
    },
    {
      "entropy": 5.737932777404785,
      "epoch": 5.065243179122183,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00027384249368370194,
      "loss": 4.8948,
      "mean_token_accuracy": 0.23542826473712922,
      "num_tokens": 117374073.0,
      "step": 51240
    },
    {
      "entropy": 5.813953018188476,
      "epoch": 5.065737445630684,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00027380720418269304,
      "loss": 4.9298,
      "mean_token_accuracy": 0.22165605723857879,
      "num_tokens": 117385208.0,
      "step": 51245
    },
    {
      "entropy": 5.817016553878784,
      "epoch": 5.066231712139185,
      "grad_norm": 1.390625,
      "learning_rate": 0.00027377191471102716,
      "loss": 4.9219,
      "mean_token_accuracy": 0.22853710800409316,
      "num_tokens": 117396218.0,
      "step": 51250
    },
    {
      "entropy": 5.802625465393066,
      "epoch": 5.066725978647687,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00027373662526957235,
      "loss": 4.8359,
      "mean_token_accuracy": 0.24194348007440566,
      "num_tokens": 117407453.0,
      "step": 51255
    },
    {
      "entropy": 5.756240367889404,
      "epoch": 5.0672202451561885,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002737013358591967,
      "loss": 4.8848,
      "mean_token_accuracy": 0.23029125481843948,
      "num_tokens": 117419223.0,
      "step": 51260
    },
    {
      "entropy": 5.722092151641846,
      "epoch": 5.06771451166469,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002736660464807685,
      "loss": 4.8164,
      "mean_token_accuracy": 0.2363016724586487,
      "num_tokens": 117430182.0,
      "step": 51265
    },
    {
      "entropy": 5.81238431930542,
      "epoch": 5.068208778173191,
      "grad_norm": 1.328125,
      "learning_rate": 0.00027363075713515554,
      "loss": 4.9231,
      "mean_token_accuracy": 0.2234948009252548,
      "num_tokens": 117441006.0,
      "step": 51270
    },
    {
      "entropy": 5.797903823852539,
      "epoch": 5.068703044681692,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002735954678232263,
      "loss": 4.9196,
      "mean_token_accuracy": 0.2344592899084091,
      "num_tokens": 117452320.0,
      "step": 51275
    },
    {
      "entropy": 5.756849956512451,
      "epoch": 5.069197311190194,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002735601785458488,
      "loss": 4.8655,
      "mean_token_accuracy": 0.23825550973415374,
      "num_tokens": 117462783.0,
      "step": 51280
    },
    {
      "entropy": 5.897156858444214,
      "epoch": 5.069691577698695,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002735248893038912,
      "loss": 4.9538,
      "mean_token_accuracy": 0.22331099659204484,
      "num_tokens": 117473078.0,
      "step": 51285
    },
    {
      "entropy": 5.719188117980957,
      "epoch": 5.070185844207196,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002734896000982215,
      "loss": 4.7797,
      "mean_token_accuracy": 0.24512575417757035,
      "num_tokens": 117484592.0,
      "step": 51290
    },
    {
      "entropy": 5.726627016067505,
      "epoch": 5.070680110715698,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002734543109297077,
      "loss": 4.9229,
      "mean_token_accuracy": 0.23180221021175385,
      "num_tokens": 117496039.0,
      "step": 51295
    },
    {
      "entropy": 5.788932132720947,
      "epoch": 5.0711743772241995,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00027341902179921825,
      "loss": 4.9265,
      "mean_token_accuracy": 0.23673444837331772,
      "num_tokens": 117507284.0,
      "step": 51300
    },
    {
      "entropy": 5.849740219116211,
      "epoch": 5.071668643732701,
      "grad_norm": 1.359375,
      "learning_rate": 0.00027338373270762107,
      "loss": 4.9192,
      "mean_token_accuracy": 0.22814241498708726,
      "num_tokens": 117519389.0,
      "step": 51305
    },
    {
      "entropy": 5.803204584121704,
      "epoch": 5.072162910241202,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002733484436557844,
      "loss": 4.8891,
      "mean_token_accuracy": 0.2370512753725052,
      "num_tokens": 117530727.0,
      "step": 51310
    },
    {
      "entropy": 5.84442458152771,
      "epoch": 5.072657176749703,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002733131546445762,
      "loss": 4.9468,
      "mean_token_accuracy": 0.22916903346776962,
      "num_tokens": 117543079.0,
      "step": 51315
    },
    {
      "entropy": 5.835252857208252,
      "epoch": 5.0731514432582046,
      "grad_norm": 1.34375,
      "learning_rate": 0.00027327786567486463,
      "loss": 4.9305,
      "mean_token_accuracy": 0.2286427929997444,
      "num_tokens": 117555128.0,
      "step": 51320
    },
    {
      "entropy": 5.768091106414795,
      "epoch": 5.073645709766706,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002732425767475178,
      "loss": 4.8696,
      "mean_token_accuracy": 0.24293849170207976,
      "num_tokens": 117566394.0,
      "step": 51325
    },
    {
      "entropy": 5.808863592147827,
      "epoch": 5.074139976275208,
      "grad_norm": 1.375,
      "learning_rate": 0.00027320728786340393,
      "loss": 4.9059,
      "mean_token_accuracy": 0.23259452730417252,
      "num_tokens": 117577247.0,
      "step": 51330
    },
    {
      "entropy": 5.770870971679687,
      "epoch": 5.074634242783709,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002731719990233911,
      "loss": 4.9608,
      "mean_token_accuracy": 0.23561477959156035,
      "num_tokens": 117589976.0,
      "step": 51335
    },
    {
      "entropy": 5.791405820846558,
      "epoch": 5.0751285092922105,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002731367102283474,
      "loss": 4.9252,
      "mean_token_accuracy": 0.23132380694150925,
      "num_tokens": 117599963.0,
      "step": 51340
    },
    {
      "entropy": 5.723532676696777,
      "epoch": 5.075622775800712,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00027310142147914075,
      "loss": 4.8256,
      "mean_token_accuracy": 0.239030359685421,
      "num_tokens": 117612396.0,
      "step": 51345
    },
    {
      "entropy": 5.740130376815796,
      "epoch": 5.076117042309213,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002730661327766395,
      "loss": 4.8778,
      "mean_token_accuracy": 0.23406336903572084,
      "num_tokens": 117624815.0,
      "step": 51350
    },
    {
      "entropy": 5.821886348724365,
      "epoch": 5.076611308817714,
      "grad_norm": 1.375,
      "learning_rate": 0.00027303084412171166,
      "loss": 4.9271,
      "mean_token_accuracy": 0.23350560367107392,
      "num_tokens": 117636962.0,
      "step": 51355
    },
    {
      "entropy": 5.8264405727386475,
      "epoch": 5.0771055753262155,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002729955555152254,
      "loss": 4.9324,
      "mean_token_accuracy": 0.23494893163442612,
      "num_tokens": 117648484.0,
      "step": 51360
    },
    {
      "entropy": 5.814236068725586,
      "epoch": 5.077599841834718,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002729602669580488,
      "loss": 4.9294,
      "mean_token_accuracy": 0.22854012101888657,
      "num_tokens": 117661590.0,
      "step": 51365
    },
    {
      "entropy": 5.78513126373291,
      "epoch": 5.078094108343219,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002729249784510499,
      "loss": 4.8793,
      "mean_token_accuracy": 0.2338658258318901,
      "num_tokens": 117673860.0,
      "step": 51370
    },
    {
      "entropy": 5.790787696838379,
      "epoch": 5.07858837485172,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00027288968999509693,
      "loss": 4.9694,
      "mean_token_accuracy": 0.22769856899976731,
      "num_tokens": 117686602.0,
      "step": 51375
    },
    {
      "entropy": 5.749928617477417,
      "epoch": 5.0790826413602215,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002728544015910579,
      "loss": 4.9093,
      "mean_token_accuracy": 0.23692586719989778,
      "num_tokens": 117698267.0,
      "step": 51380
    },
    {
      "entropy": 5.755554580688477,
      "epoch": 5.079576907868723,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002728191132398008,
      "loss": 4.8461,
      "mean_token_accuracy": 0.23987818956375123,
      "num_tokens": 117709542.0,
      "step": 51385
    },
    {
      "entropy": 5.756363582611084,
      "epoch": 5.080071174377224,
      "grad_norm": 1.515625,
      "learning_rate": 0.000272783824942194,
      "loss": 4.8142,
      "mean_token_accuracy": 0.2448135420680046,
      "num_tokens": 117720413.0,
      "step": 51390
    },
    {
      "entropy": 5.819203948974609,
      "epoch": 5.080565440885725,
      "grad_norm": 1.5,
      "learning_rate": 0.00027274853669910536,
      "loss": 4.9663,
      "mean_token_accuracy": 0.23029715865850447,
      "num_tokens": 117732115.0,
      "step": 51395
    },
    {
      "entropy": 5.7345868110656735,
      "epoch": 5.0810597073942265,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00027271324851140314,
      "loss": 4.8789,
      "mean_token_accuracy": 0.2378255009651184,
      "num_tokens": 117742571.0,
      "step": 51400
    },
    {
      "entropy": 5.825289011001587,
      "epoch": 5.081553973902729,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00027267796037995534,
      "loss": 4.9736,
      "mean_token_accuracy": 0.2241715356707573,
      "num_tokens": 117755159.0,
      "step": 51405
    },
    {
      "entropy": 5.819095659255981,
      "epoch": 5.08204824041123,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00027264267230563014,
      "loss": 4.9511,
      "mean_token_accuracy": 0.22886353433132173,
      "num_tokens": 117765482.0,
      "step": 51410
    },
    {
      "entropy": 5.8019555568695065,
      "epoch": 5.082542506919731,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002726073842892954,
      "loss": 4.928,
      "mean_token_accuracy": 0.23271461129188536,
      "num_tokens": 117776878.0,
      "step": 51415
    },
    {
      "entropy": 5.742416143417358,
      "epoch": 5.0830367734282325,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00027257209633181937,
      "loss": 4.8418,
      "mean_token_accuracy": 0.23938788920640947,
      "num_tokens": 117788873.0,
      "step": 51420
    },
    {
      "entropy": 5.7196937084198,
      "epoch": 5.083531039936734,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002725368084340703,
      "loss": 4.7872,
      "mean_token_accuracy": 0.24716376960277558,
      "num_tokens": 117799497.0,
      "step": 51425
    },
    {
      "entropy": 5.789262390136718,
      "epoch": 5.084025306445235,
      "grad_norm": 1.375,
      "learning_rate": 0.0002725015205969161,
      "loss": 4.9387,
      "mean_token_accuracy": 0.22567253857851027,
      "num_tokens": 117810203.0,
      "step": 51430
    },
    {
      "entropy": 5.74295711517334,
      "epoch": 5.084519572953736,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002724662328212248,
      "loss": 4.8672,
      "mean_token_accuracy": 0.24587636888027192,
      "num_tokens": 117821057.0,
      "step": 51435
    },
    {
      "entropy": 5.747556257247925,
      "epoch": 5.085013839462238,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002724309451078646,
      "loss": 4.8521,
      "mean_token_accuracy": 0.23266409039497377,
      "num_tokens": 117832294.0,
      "step": 51440
    },
    {
      "entropy": 5.81097731590271,
      "epoch": 5.08550810597074,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002723956574577035,
      "loss": 4.8777,
      "mean_token_accuracy": 0.23060377836227416,
      "num_tokens": 117842909.0,
      "step": 51445
    },
    {
      "entropy": 5.859164428710938,
      "epoch": 5.086002372479241,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002723603698716097,
      "loss": 5.0007,
      "mean_token_accuracy": 0.22439118176698686,
      "num_tokens": 117855881.0,
      "step": 51450
    },
    {
      "entropy": 5.7852214813232425,
      "epoch": 5.086496638987742,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002723250823504513,
      "loss": 4.9614,
      "mean_token_accuracy": 0.23355901688337327,
      "num_tokens": 117867716.0,
      "step": 51455
    },
    {
      "entropy": 5.755856990814209,
      "epoch": 5.086990905496243,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002722897948950962,
      "loss": 4.8849,
      "mean_token_accuracy": 0.24029562175273894,
      "num_tokens": 117878049.0,
      "step": 51460
    },
    {
      "entropy": 5.883109283447266,
      "epoch": 5.087485172004745,
      "grad_norm": 1.265625,
      "learning_rate": 0.00027225450750641253,
      "loss": 5.0212,
      "mean_token_accuracy": 0.21799014806747435,
      "num_tokens": 117889593.0,
      "step": 51465
    },
    {
      "entropy": 5.745349454879761,
      "epoch": 5.087979438513246,
      "grad_norm": 1.625,
      "learning_rate": 0.0002722192201852683,
      "loss": 4.8763,
      "mean_token_accuracy": 0.23730818033218384,
      "num_tokens": 117900012.0,
      "step": 51470
    },
    {
      "entropy": 5.772338914871216,
      "epoch": 5.088473705021748,
      "grad_norm": 1.28125,
      "learning_rate": 0.00027218393293253185,
      "loss": 4.9318,
      "mean_token_accuracy": 0.2317038208246231,
      "num_tokens": 117911483.0,
      "step": 51475
    },
    {
      "entropy": 5.769381523132324,
      "epoch": 5.088967971530249,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00027214864574907113,
      "loss": 4.9621,
      "mean_token_accuracy": 0.2316024348139763,
      "num_tokens": 117923718.0,
      "step": 51480
    },
    {
      "entropy": 5.778090524673462,
      "epoch": 5.089462238038751,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002721133586357541,
      "loss": 4.8687,
      "mean_token_accuracy": 0.23859770596027374,
      "num_tokens": 117936014.0,
      "step": 51485
    },
    {
      "entropy": 5.776889514923096,
      "epoch": 5.089956504547252,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002720780715934488,
      "loss": 4.8484,
      "mean_token_accuracy": 0.2326500818133354,
      "num_tokens": 117947593.0,
      "step": 51490
    },
    {
      "entropy": 5.854087591171265,
      "epoch": 5.090450771055753,
      "grad_norm": 1.375,
      "learning_rate": 0.0002720427846230235,
      "loss": 4.94,
      "mean_token_accuracy": 0.22942851781845092,
      "num_tokens": 117959155.0,
      "step": 51495
    },
    {
      "entropy": 5.834285306930542,
      "epoch": 5.090945037564254,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00027200749772534607,
      "loss": 4.9229,
      "mean_token_accuracy": 0.2299712911248207,
      "num_tokens": 117970580.0,
      "step": 51500
    },
    {
      "entropy": 5.777625989913941,
      "epoch": 5.091439304072756,
      "grad_norm": 1.40625,
      "learning_rate": 0.00027197221090128476,
      "loss": 4.8765,
      "mean_token_accuracy": 0.23451332598924637,
      "num_tokens": 117983026.0,
      "step": 51505
    },
    {
      "entropy": 5.791086483001709,
      "epoch": 5.091933570581258,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002719369241517075,
      "loss": 4.9301,
      "mean_token_accuracy": 0.2303038716316223,
      "num_tokens": 117994317.0,
      "step": 51510
    },
    {
      "entropy": 5.7648838520050045,
      "epoch": 5.092427837089759,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002719016374774824,
      "loss": 4.8985,
      "mean_token_accuracy": 0.23311750143766402,
      "num_tokens": 118005825.0,
      "step": 51515
    },
    {
      "entropy": 5.831339883804321,
      "epoch": 5.09292210359826,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002718663508794775,
      "loss": 4.9314,
      "mean_token_accuracy": 0.23071902990341187,
      "num_tokens": 118017437.0,
      "step": 51520
    },
    {
      "entropy": 5.711911869049072,
      "epoch": 5.093416370106762,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002718310643585608,
      "loss": 4.8511,
      "mean_token_accuracy": 0.24188246279954911,
      "num_tokens": 118029714.0,
      "step": 51525
    },
    {
      "entropy": 5.8974226951599125,
      "epoch": 5.093910636615263,
      "grad_norm": 1.25,
      "learning_rate": 0.0002717957779156004,
      "loss": 5.0794,
      "mean_token_accuracy": 0.21733969151973725,
      "num_tokens": 118042102.0,
      "step": 51530
    },
    {
      "entropy": 5.870179605484009,
      "epoch": 5.094404903123764,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002717604915514644,
      "loss": 4.8881,
      "mean_token_accuracy": 0.23331099152565002,
      "num_tokens": 118052985.0,
      "step": 51535
    },
    {
      "entropy": 5.812343692779541,
      "epoch": 5.094899169632265,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002717252052670208,
      "loss": 4.9761,
      "mean_token_accuracy": 0.2258494660258293,
      "num_tokens": 118064969.0,
      "step": 51540
    },
    {
      "entropy": 5.683795928955078,
      "epoch": 5.095393436140767,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002716899190631377,
      "loss": 4.7496,
      "mean_token_accuracy": 0.24493298381567002,
      "num_tokens": 118076489.0,
      "step": 51545
    },
    {
      "entropy": 5.752049779891967,
      "epoch": 5.095887702649269,
      "grad_norm": 1.390625,
      "learning_rate": 0.000271654632940683,
      "loss": 4.8971,
      "mean_token_accuracy": 0.2394574910402298,
      "num_tokens": 118087512.0,
      "step": 51550
    },
    {
      "entropy": 5.848593044281006,
      "epoch": 5.09638196915777,
      "grad_norm": 1.3125,
      "learning_rate": 0.000271619346900525,
      "loss": 4.9627,
      "mean_token_accuracy": 0.23201024681329727,
      "num_tokens": 118099797.0,
      "step": 51555
    },
    {
      "entropy": 5.736326408386231,
      "epoch": 5.096876235666271,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002715840609435314,
      "loss": 4.8083,
      "mean_token_accuracy": 0.24461707323789597,
      "num_tokens": 118111854.0,
      "step": 51560
    },
    {
      "entropy": 5.802618217468262,
      "epoch": 5.097370502174773,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002715487750705706,
      "loss": 4.9575,
      "mean_token_accuracy": 0.22569347620010377,
      "num_tokens": 118123901.0,
      "step": 51565
    },
    {
      "entropy": 5.756728410720825,
      "epoch": 5.097864768683274,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002715134892825104,
      "loss": 4.797,
      "mean_token_accuracy": 0.2479052349925041,
      "num_tokens": 118133844.0,
      "step": 51570
    },
    {
      "entropy": 5.769483137130737,
      "epoch": 5.098359035191775,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002714782035802189,
      "loss": 4.8848,
      "mean_token_accuracy": 0.23666471391916274,
      "num_tokens": 118145978.0,
      "step": 51575
    },
    {
      "entropy": 5.799629831314087,
      "epoch": 5.098853301700276,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00027144291796456417,
      "loss": 4.9581,
      "mean_token_accuracy": 0.22932883352041245,
      "num_tokens": 118157532.0,
      "step": 51580
    },
    {
      "entropy": 5.803337717056275,
      "epoch": 5.0993475682087785,
      "grad_norm": 1.375,
      "learning_rate": 0.0002714076324364142,
      "loss": 4.931,
      "mean_token_accuracy": 0.22640581876039506,
      "num_tokens": 118168700.0,
      "step": 51585
    },
    {
      "entropy": 5.789723443984985,
      "epoch": 5.09984183471728,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002713723469966369,
      "loss": 4.8873,
      "mean_token_accuracy": 0.23550939410924912,
      "num_tokens": 118178961.0,
      "step": 51590
    },
    {
      "entropy": 5.783200359344482,
      "epoch": 5.100336101225781,
      "grad_norm": 1.59375,
      "learning_rate": 0.0002713370616461006,
      "loss": 4.8798,
      "mean_token_accuracy": 0.23656709790229796,
      "num_tokens": 118189074.0,
      "step": 51595
    },
    {
      "entropy": 5.817804288864136,
      "epoch": 5.100830367734282,
      "grad_norm": 1.375,
      "learning_rate": 0.00027130177638567314,
      "loss": 4.9036,
      "mean_token_accuracy": 0.23428381830453873,
      "num_tokens": 118201606.0,
      "step": 51600
    },
    {
      "entropy": 5.795020484924317,
      "epoch": 5.101324634242784,
      "grad_norm": 1.390625,
      "learning_rate": 0.00027126649121622247,
      "loss": 4.9384,
      "mean_token_accuracy": 0.2288828283548355,
      "num_tokens": 118212645.0,
      "step": 51605
    },
    {
      "entropy": 5.762093782424927,
      "epoch": 5.101818900751285,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002712312061386168,
      "loss": 4.8606,
      "mean_token_accuracy": 0.23501413464546203,
      "num_tokens": 118223870.0,
      "step": 51610
    },
    {
      "entropy": 5.836634683609009,
      "epoch": 5.102313167259786,
      "grad_norm": 1.375,
      "learning_rate": 0.0002711959211537239,
      "loss": 4.9459,
      "mean_token_accuracy": 0.2303989514708519,
      "num_tokens": 118235144.0,
      "step": 51615
    },
    {
      "entropy": 5.752189826965332,
      "epoch": 5.102807433768288,
      "grad_norm": 1.359375,
      "learning_rate": 0.000271160636262412,
      "loss": 4.8579,
      "mean_token_accuracy": 0.23388785719871522,
      "num_tokens": 118246006.0,
      "step": 51620
    },
    {
      "entropy": 5.796044731140137,
      "epoch": 5.1033017002767895,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00027112535146554915,
      "loss": 4.9118,
      "mean_token_accuracy": 0.23701255023479462,
      "num_tokens": 118258486.0,
      "step": 51625
    },
    {
      "entropy": 5.826635837554932,
      "epoch": 5.103795966785291,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00027109006676400305,
      "loss": 4.9252,
      "mean_token_accuracy": 0.22975379079580308,
      "num_tokens": 118269266.0,
      "step": 51630
    },
    {
      "entropy": 5.796469068527221,
      "epoch": 5.104290233293792,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002710547821586421,
      "loss": 4.9448,
      "mean_token_accuracy": 0.23124364167451858,
      "num_tokens": 118280586.0,
      "step": 51635
    },
    {
      "entropy": 5.737385559082031,
      "epoch": 5.104784499802293,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002710194976503341,
      "loss": 4.8198,
      "mean_token_accuracy": 0.2418235421180725,
      "num_tokens": 118292463.0,
      "step": 51640
    },
    {
      "entropy": 5.71797661781311,
      "epoch": 5.105278766310795,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002709842132399469,
      "loss": 4.8412,
      "mean_token_accuracy": 0.24015462547540664,
      "num_tokens": 118304777.0,
      "step": 51645
    },
    {
      "entropy": 5.79847092628479,
      "epoch": 5.105773032819296,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002709489289283488,
      "loss": 4.9012,
      "mean_token_accuracy": 0.2336844816803932,
      "num_tokens": 118315402.0,
      "step": 51650
    },
    {
      "entropy": 5.786246109008789,
      "epoch": 5.106267299327797,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00027091364471640766,
      "loss": 4.9075,
      "mean_token_accuracy": 0.23111010044813157,
      "num_tokens": 118328933.0,
      "step": 51655
    },
    {
      "entropy": 5.74150595664978,
      "epoch": 5.106761565836299,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00027087836060499156,
      "loss": 4.8258,
      "mean_token_accuracy": 0.2404320701956749,
      "num_tokens": 118340724.0,
      "step": 51660
    },
    {
      "entropy": 5.78893404006958,
      "epoch": 5.1072558323448005,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002708430765949684,
      "loss": 4.9983,
      "mean_token_accuracy": 0.22240907549858094,
      "num_tokens": 118353524.0,
      "step": 51665
    },
    {
      "entropy": 5.793966484069824,
      "epoch": 5.107750098853302,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00027080779268720616,
      "loss": 4.8628,
      "mean_token_accuracy": 0.23624544590711594,
      "num_tokens": 118365038.0,
      "step": 51670
    },
    {
      "entropy": 5.766090393066406,
      "epoch": 5.108244365361803,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002707725088825728,
      "loss": 4.8929,
      "mean_token_accuracy": 0.23490609377622604,
      "num_tokens": 118377230.0,
      "step": 51675
    },
    {
      "entropy": 5.811159896850586,
      "epoch": 5.108738631870304,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00027073722518193647,
      "loss": 4.867,
      "mean_token_accuracy": 0.23770486265420915,
      "num_tokens": 118387407.0,
      "step": 51680
    },
    {
      "entropy": 5.775157165527344,
      "epoch": 5.109232898378806,
      "grad_norm": 1.34375,
      "learning_rate": 0.00027070194158616514,
      "loss": 4.8966,
      "mean_token_accuracy": 0.23446733951568605,
      "num_tokens": 118398116.0,
      "step": 51685
    },
    {
      "entropy": 5.737257814407348,
      "epoch": 5.109727164887307,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002706666580961266,
      "loss": 4.8518,
      "mean_token_accuracy": 0.2353003814816475,
      "num_tokens": 118409486.0,
      "step": 51690
    },
    {
      "entropy": 5.69752779006958,
      "epoch": 5.110221431395809,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000270631374712689,
      "loss": 4.8149,
      "mean_token_accuracy": 0.24391796737909316,
      "num_tokens": 118420469.0,
      "step": 51695
    },
    {
      "entropy": 5.77099289894104,
      "epoch": 5.11071569790431,
      "grad_norm": 1.3125,
      "learning_rate": 0.00027059609143672024,
      "loss": 4.9176,
      "mean_token_accuracy": 0.2394423559308052,
      "num_tokens": 118431506.0,
      "step": 51700
    },
    {
      "entropy": 5.73484845161438,
      "epoch": 5.1112099644128115,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002705608082690882,
      "loss": 4.904,
      "mean_token_accuracy": 0.24268007427453994,
      "num_tokens": 118443311.0,
      "step": 51705
    },
    {
      "entropy": 5.773115873336792,
      "epoch": 5.111704230921313,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00027052552521066107,
      "loss": 4.8789,
      "mean_token_accuracy": 0.23386637717485428,
      "num_tokens": 118454997.0,
      "step": 51710
    },
    {
      "entropy": 5.759063100814819,
      "epoch": 5.112198497429814,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002704902422623068,
      "loss": 4.835,
      "mean_token_accuracy": 0.24104640483856202,
      "num_tokens": 118468145.0,
      "step": 51715
    },
    {
      "entropy": 5.798162841796875,
      "epoch": 5.112692763938315,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002704549594248932,
      "loss": 4.9001,
      "mean_token_accuracy": 0.2305876299738884,
      "num_tokens": 118479904.0,
      "step": 51720
    },
    {
      "entropy": 5.670382070541382,
      "epoch": 5.1131870304468166,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002704196766992883,
      "loss": 4.7857,
      "mean_token_accuracy": 0.2452153041958809,
      "num_tokens": 118491227.0,
      "step": 51725
    },
    {
      "entropy": 5.761623573303223,
      "epoch": 5.113681296955319,
      "grad_norm": 1.453125,
      "learning_rate": 0.00027038439408636004,
      "loss": 4.8667,
      "mean_token_accuracy": 0.24027989208698272,
      "num_tokens": 118501910.0,
      "step": 51730
    },
    {
      "entropy": 5.858758640289307,
      "epoch": 5.11417556346382,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00027034911158697634,
      "loss": 4.9791,
      "mean_token_accuracy": 0.2316924065351486,
      "num_tokens": 118512972.0,
      "step": 51735
    },
    {
      "entropy": 5.702692890167237,
      "epoch": 5.114669829972321,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002703138292020053,
      "loss": 4.8811,
      "mean_token_accuracy": 0.23641757071018218,
      "num_tokens": 118523105.0,
      "step": 51740
    },
    {
      "entropy": 5.816646766662598,
      "epoch": 5.1151640964808225,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00027027854693231486,
      "loss": 4.9871,
      "mean_token_accuracy": 0.22500278502702714,
      "num_tokens": 118532938.0,
      "step": 51745
    },
    {
      "entropy": 5.816492080688477,
      "epoch": 5.115658362989324,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002702432647787728,
      "loss": 4.9096,
      "mean_token_accuracy": 0.22594673335552215,
      "num_tokens": 118543863.0,
      "step": 51750
    },
    {
      "entropy": 5.788086795806885,
      "epoch": 5.116152629497825,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002702079827422472,
      "loss": 4.9469,
      "mean_token_accuracy": 0.2373531311750412,
      "num_tokens": 118554406.0,
      "step": 51755
    },
    {
      "entropy": 5.782899284362793,
      "epoch": 5.116646896006326,
      "grad_norm": 1.1875,
      "learning_rate": 0.00027017270082360585,
      "loss": 4.8917,
      "mean_token_accuracy": 0.23370036631822586,
      "num_tokens": 118566881.0,
      "step": 51760
    },
    {
      "entropy": 5.861611318588257,
      "epoch": 5.117141162514828,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002701374190237169,
      "loss": 4.9982,
      "mean_token_accuracy": 0.22707634866237641,
      "num_tokens": 118578591.0,
      "step": 51765
    },
    {
      "entropy": 5.839370203018189,
      "epoch": 5.11763542902333,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00027010213734344823,
      "loss": 4.9389,
      "mean_token_accuracy": 0.23113846480846406,
      "num_tokens": 118590618.0,
      "step": 51770
    },
    {
      "entropy": 5.884372425079346,
      "epoch": 5.118129695531831,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00027006685578366774,
      "loss": 4.9946,
      "mean_token_accuracy": 0.2217216521501541,
      "num_tokens": 118603064.0,
      "step": 51775
    },
    {
      "entropy": 5.740394353866577,
      "epoch": 5.118623962040332,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00027003157434524335,
      "loss": 4.8216,
      "mean_token_accuracy": 0.2463473603129387,
      "num_tokens": 118613604.0,
      "step": 51780
    },
    {
      "entropy": 5.795015525817871,
      "epoch": 5.1191182285488335,
      "grad_norm": 1.375,
      "learning_rate": 0.00026999629302904303,
      "loss": 4.8838,
      "mean_token_accuracy": 0.2323044091463089,
      "num_tokens": 118625430.0,
      "step": 51785
    },
    {
      "entropy": 5.704747295379638,
      "epoch": 5.119612495057335,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00026996101183593456,
      "loss": 4.8746,
      "mean_token_accuracy": 0.2386867344379425,
      "num_tokens": 118636844.0,
      "step": 51790
    },
    {
      "entropy": 5.74528226852417,
      "epoch": 5.120106761565836,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002699257307667861,
      "loss": 4.8794,
      "mean_token_accuracy": 0.23548623621463777,
      "num_tokens": 118648143.0,
      "step": 51795
    },
    {
      "entropy": 5.82378363609314,
      "epoch": 5.120601028074337,
      "grad_norm": 1.40625,
      "learning_rate": 0.00026989044982246536,
      "loss": 4.9603,
      "mean_token_accuracy": 0.22603877931833266,
      "num_tokens": 118659110.0,
      "step": 51800
    },
    {
      "entropy": 5.819825649261475,
      "epoch": 5.121095294582839,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00026985516900384043,
      "loss": 4.8788,
      "mean_token_accuracy": 0.23614778965711594,
      "num_tokens": 118670641.0,
      "step": 51805
    },
    {
      "entropy": 5.816943502426147,
      "epoch": 5.121589561091341,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00026981988831177914,
      "loss": 4.9942,
      "mean_token_accuracy": 0.2260827124118805,
      "num_tokens": 118682308.0,
      "step": 51810
    },
    {
      "entropy": 5.840769577026367,
      "epoch": 5.122083827599842,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00026978460774714934,
      "loss": 4.9802,
      "mean_token_accuracy": 0.22392056584358216,
      "num_tokens": 118695631.0,
      "step": 51815
    },
    {
      "entropy": 5.757374334335327,
      "epoch": 5.122578094108343,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00026974932731081903,
      "loss": 4.8796,
      "mean_token_accuracy": 0.23899724334478378,
      "num_tokens": 118706961.0,
      "step": 51820
    },
    {
      "entropy": 5.805899238586425,
      "epoch": 5.1230723606168445,
      "grad_norm": 1.375,
      "learning_rate": 0.000269714047003656,
      "loss": 4.8831,
      "mean_token_accuracy": 0.23812064230442048,
      "num_tokens": 118718809.0,
      "step": 51825
    },
    {
      "entropy": 5.7707830429077145,
      "epoch": 5.123566627125346,
      "grad_norm": 1.375,
      "learning_rate": 0.00026967876682652836,
      "loss": 4.8765,
      "mean_token_accuracy": 0.23799988925457,
      "num_tokens": 118732278.0,
      "step": 51830
    },
    {
      "entropy": 5.786280965805053,
      "epoch": 5.124060893633847,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00026964348678030386,
      "loss": 4.9199,
      "mean_token_accuracy": 0.22972101271152495,
      "num_tokens": 118743446.0,
      "step": 51835
    },
    {
      "entropy": 5.8399045944213865,
      "epoch": 5.124555160142349,
      "grad_norm": 1.5,
      "learning_rate": 0.00026960820686585036,
      "loss": 4.9332,
      "mean_token_accuracy": 0.22787681967020035,
      "num_tokens": 118754220.0,
      "step": 51840
    },
    {
      "entropy": 5.84801139831543,
      "epoch": 5.12504942665085,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00026957292708403594,
      "loss": 4.9699,
      "mean_token_accuracy": 0.2322448745369911,
      "num_tokens": 118765422.0,
      "step": 51845
    },
    {
      "entropy": 5.731307888031006,
      "epoch": 5.125543693159352,
      "grad_norm": 1.375,
      "learning_rate": 0.0002695376474357281,
      "loss": 4.8227,
      "mean_token_accuracy": 0.23956434279680253,
      "num_tokens": 118777244.0,
      "step": 51850
    },
    {
      "entropy": 5.797716236114502,
      "epoch": 5.126037959667853,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002695023679217952,
      "loss": 4.8993,
      "mean_token_accuracy": 0.2291640043258667,
      "num_tokens": 118789142.0,
      "step": 51855
    },
    {
      "entropy": 5.843733787536621,
      "epoch": 5.126532226176354,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002694670885431049,
      "loss": 4.9094,
      "mean_token_accuracy": 0.23270364999771118,
      "num_tokens": 118800697.0,
      "step": 51860
    },
    {
      "entropy": 5.809617948532105,
      "epoch": 5.127026492684855,
      "grad_norm": 1.375,
      "learning_rate": 0.00026943180930052497,
      "loss": 4.944,
      "mean_token_accuracy": 0.23590609878301622,
      "num_tokens": 118813200.0,
      "step": 51865
    },
    {
      "entropy": 5.865290737152099,
      "epoch": 5.127520759193357,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002693965301949234,
      "loss": 5.0053,
      "mean_token_accuracy": 0.22642103284597398,
      "num_tokens": 118827037.0,
      "step": 51870
    },
    {
      "entropy": 5.753869533538818,
      "epoch": 5.128015025701859,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00026936125122716816,
      "loss": 4.8657,
      "mean_token_accuracy": 0.2366800546646118,
      "num_tokens": 118837579.0,
      "step": 51875
    },
    {
      "entropy": 5.812687826156616,
      "epoch": 5.12850929221036,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002693259723981269,
      "loss": 4.9071,
      "mean_token_accuracy": 0.23242078721523285,
      "num_tokens": 118849270.0,
      "step": 51880
    },
    {
      "entropy": 5.697208881378174,
      "epoch": 5.129003558718861,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00026929069370866765,
      "loss": 4.8501,
      "mean_token_accuracy": 0.23358366638422012,
      "num_tokens": 118861521.0,
      "step": 51885
    },
    {
      "entropy": 5.808628511428833,
      "epoch": 5.129497825227363,
      "grad_norm": 1.4375,
      "learning_rate": 0.00026925541515965826,
      "loss": 4.8755,
      "mean_token_accuracy": 0.2357863187789917,
      "num_tokens": 118872617.0,
      "step": 51890
    },
    {
      "entropy": 5.833275508880615,
      "epoch": 5.129992091735864,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002692201367519665,
      "loss": 4.9145,
      "mean_token_accuracy": 0.23210688382387162,
      "num_tokens": 118884831.0,
      "step": 51895
    },
    {
      "entropy": 5.759375381469726,
      "epoch": 5.130486358244365,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002691848584864603,
      "loss": 4.857,
      "mean_token_accuracy": 0.2415716230869293,
      "num_tokens": 118896468.0,
      "step": 51900
    },
    {
      "entropy": 5.719766759872437,
      "epoch": 5.130980624752866,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002691495803640075,
      "loss": 4.8038,
      "mean_token_accuracy": 0.2399125576019287,
      "num_tokens": 118907605.0,
      "step": 51905
    },
    {
      "entropy": 5.701514387130738,
      "epoch": 5.131474891261368,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00026911430238547586,
      "loss": 4.8651,
      "mean_token_accuracy": 0.24075173884630202,
      "num_tokens": 118919767.0,
      "step": 51910
    },
    {
      "entropy": 5.83512396812439,
      "epoch": 5.13196915776987,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002690790245517333,
      "loss": 4.9528,
      "mean_token_accuracy": 0.22484129369258882,
      "num_tokens": 118931964.0,
      "step": 51915
    },
    {
      "entropy": 5.769226455688477,
      "epoch": 5.132463424278371,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002690437468636477,
      "loss": 4.8429,
      "mean_token_accuracy": 0.23760697543621062,
      "num_tokens": 118943738.0,
      "step": 51920
    },
    {
      "entropy": 5.800049543380737,
      "epoch": 5.132957690786872,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00026900846932208683,
      "loss": 4.9542,
      "mean_token_accuracy": 0.22608914971351624,
      "num_tokens": 118954700.0,
      "step": 51925
    },
    {
      "entropy": 5.774595594406128,
      "epoch": 5.133451957295374,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002689731919279186,
      "loss": 4.9142,
      "mean_token_accuracy": 0.23789456188678743,
      "num_tokens": 118966422.0,
      "step": 51930
    },
    {
      "entropy": 5.794870805740357,
      "epoch": 5.133946223803875,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002689379146820106,
      "loss": 4.8912,
      "mean_token_accuracy": 0.23002667725086212,
      "num_tokens": 118977166.0,
      "step": 51935
    },
    {
      "entropy": 5.731141233444214,
      "epoch": 5.134440490312376,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00026890263758523104,
      "loss": 4.7967,
      "mean_token_accuracy": 0.23931180834770202,
      "num_tokens": 118987611.0,
      "step": 51940
    },
    {
      "entropy": 5.776971769332886,
      "epoch": 5.134934756820877,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002688673606384475,
      "loss": 4.9227,
      "mean_token_accuracy": 0.23252964913845062,
      "num_tokens": 118998565.0,
      "step": 51945
    },
    {
      "entropy": 5.839393854141235,
      "epoch": 5.13542902332938,
      "grad_norm": 1.40625,
      "learning_rate": 0.00026883208384252785,
      "loss": 4.961,
      "mean_token_accuracy": 0.2233155131340027,
      "num_tokens": 119009528.0,
      "step": 51950
    },
    {
      "entropy": 5.853628492355346,
      "epoch": 5.135923289837881,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002687968071983399,
      "loss": 4.9676,
      "mean_token_accuracy": 0.22640328705310822,
      "num_tokens": 119021663.0,
      "step": 51955
    },
    {
      "entropy": 5.858878421783447,
      "epoch": 5.136417556346382,
      "grad_norm": 1.265625,
      "learning_rate": 0.00026876153070675134,
      "loss": 4.9881,
      "mean_token_accuracy": 0.22769499570131302,
      "num_tokens": 119034357.0,
      "step": 51960
    },
    {
      "entropy": 5.820227527618409,
      "epoch": 5.136911822854883,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00026872625436863006,
      "loss": 4.8937,
      "mean_token_accuracy": 0.23418477028608323,
      "num_tokens": 119044049.0,
      "step": 51965
    },
    {
      "entropy": 5.77783236503601,
      "epoch": 5.137406089363385,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002686909781848441,
      "loss": 4.8793,
      "mean_token_accuracy": 0.2370429053902626,
      "num_tokens": 119055359.0,
      "step": 51970
    },
    {
      "entropy": 5.755026054382324,
      "epoch": 5.137900355871886,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00026865570215626095,
      "loss": 4.7959,
      "mean_token_accuracy": 0.2403356522321701,
      "num_tokens": 119065984.0,
      "step": 51975
    },
    {
      "entropy": 5.7757243633270265,
      "epoch": 5.138394622380387,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00026862042628374855,
      "loss": 4.8661,
      "mean_token_accuracy": 0.23437810689210892,
      "num_tokens": 119077260.0,
      "step": 51980
    },
    {
      "entropy": 5.733433675765991,
      "epoch": 5.138888888888889,
      "grad_norm": 1.53125,
      "learning_rate": 0.00026858515056817463,
      "loss": 4.914,
      "mean_token_accuracy": 0.23037851452827454,
      "num_tokens": 119088370.0,
      "step": 51985
    },
    {
      "entropy": 5.78386082649231,
      "epoch": 5.1393831553973905,
      "grad_norm": 1.328125,
      "learning_rate": 0.000268549875010407,
      "loss": 4.8959,
      "mean_token_accuracy": 0.23075580447912217,
      "num_tokens": 119099566.0,
      "step": 51990
    },
    {
      "entropy": 5.834256792068482,
      "epoch": 5.139877421905892,
      "grad_norm": 1.421875,
      "learning_rate": 0.00026851459961131337,
      "loss": 4.9096,
      "mean_token_accuracy": 0.23344081044197082,
      "num_tokens": 119111144.0,
      "step": 51995
    },
    {
      "entropy": 5.778498125076294,
      "epoch": 5.140371688414393,
      "grad_norm": 1.5,
      "learning_rate": 0.0002684793243717617,
      "loss": 4.8875,
      "mean_token_accuracy": 0.22877756804227828,
      "num_tokens": 119122936.0,
      "step": 52000
    },
    {
      "entropy": 5.776762819290161,
      "epoch": 5.140865954922894,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00026844404929261963,
      "loss": 4.9339,
      "mean_token_accuracy": 0.23260470032691954,
      "num_tokens": 119134794.0,
      "step": 52005
    },
    {
      "entropy": 5.75736174583435,
      "epoch": 5.141360221431396,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002684087743747549,
      "loss": 4.8785,
      "mean_token_accuracy": 0.23193226456642152,
      "num_tokens": 119146114.0,
      "step": 52010
    },
    {
      "entropy": 5.824868154525757,
      "epoch": 5.141854487939897,
      "grad_norm": 1.171875,
      "learning_rate": 0.00026837349961903546,
      "loss": 4.9654,
      "mean_token_accuracy": 0.22668658047914506,
      "num_tokens": 119158137.0,
      "step": 52015
    },
    {
      "entropy": 5.84308385848999,
      "epoch": 5.142348754448399,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002683382250263289,
      "loss": 4.9441,
      "mean_token_accuracy": 0.2307843506336212,
      "num_tokens": 119170422.0,
      "step": 52020
    },
    {
      "entropy": 5.69630880355835,
      "epoch": 5.1428430209569,
      "grad_norm": 1.359375,
      "learning_rate": 0.00026830295059750305,
      "loss": 4.7966,
      "mean_token_accuracy": 0.2432638570666313,
      "num_tokens": 119182250.0,
      "step": 52025
    },
    {
      "entropy": 5.830500793457031,
      "epoch": 5.1433372874654015,
      "grad_norm": 1.53125,
      "learning_rate": 0.0002682676763334256,
      "loss": 4.9619,
      "mean_token_accuracy": 0.23052581697702407,
      "num_tokens": 119192810.0,
      "step": 52030
    },
    {
      "entropy": 5.7857825756073,
      "epoch": 5.143831553973903,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00026823240223496444,
      "loss": 4.8529,
      "mean_token_accuracy": 0.2390858858823776,
      "num_tokens": 119203754.0,
      "step": 52035
    },
    {
      "entropy": 5.76120548248291,
      "epoch": 5.144325820482404,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002681971283029872,
      "loss": 4.902,
      "mean_token_accuracy": 0.2340383991599083,
      "num_tokens": 119214956.0,
      "step": 52040
    },
    {
      "entropy": 5.812171792984008,
      "epoch": 5.144820086990905,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00026816185453836167,
      "loss": 4.8952,
      "mean_token_accuracy": 0.23928548991680146,
      "num_tokens": 119227608.0,
      "step": 52045
    },
    {
      "entropy": 5.837193250656128,
      "epoch": 5.145314353499407,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002681265809419555,
      "loss": 4.9939,
      "mean_token_accuracy": 0.2248193770647049,
      "num_tokens": 119239700.0,
      "step": 52050
    },
    {
      "entropy": 5.822922706604004,
      "epoch": 5.145808620007908,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002680913075146366,
      "loss": 4.9259,
      "mean_token_accuracy": 0.23404424637556076,
      "num_tokens": 119250803.0,
      "step": 52055
    },
    {
      "entropy": 5.775328969955444,
      "epoch": 5.14630288651641,
      "grad_norm": 1.328125,
      "learning_rate": 0.00026805603425727255,
      "loss": 4.8786,
      "mean_token_accuracy": 0.23798750191926957,
      "num_tokens": 119261574.0,
      "step": 52060
    },
    {
      "entropy": 5.7581535339355465,
      "epoch": 5.146797153024911,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002680207611707312,
      "loss": 4.8844,
      "mean_token_accuracy": 0.23909699469804763,
      "num_tokens": 119274558.0,
      "step": 52065
    },
    {
      "entropy": 5.8164139747619625,
      "epoch": 5.1472914195334125,
      "grad_norm": 1.359375,
      "learning_rate": 0.00026798548825588016,
      "loss": 4.9086,
      "mean_token_accuracy": 0.23062155693769454,
      "num_tokens": 119286262.0,
      "step": 52070
    },
    {
      "entropy": 5.825852108001709,
      "epoch": 5.147785686041914,
      "grad_norm": 1.5,
      "learning_rate": 0.0002679502155135872,
      "loss": 4.9988,
      "mean_token_accuracy": 0.23776006549596787,
      "num_tokens": 119297345.0,
      "step": 52075
    },
    {
      "entropy": 5.773509836196899,
      "epoch": 5.148279952550415,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00026791494294472004,
      "loss": 4.896,
      "mean_token_accuracy": 0.23074910193681716,
      "num_tokens": 119309356.0,
      "step": 52080
    },
    {
      "entropy": 5.781809854507446,
      "epoch": 5.148774219058916,
      "grad_norm": 1.234375,
      "learning_rate": 0.00026787967055014633,
      "loss": 4.9063,
      "mean_token_accuracy": 0.23062957376241683,
      "num_tokens": 119321614.0,
      "step": 52085
    },
    {
      "entropy": 5.727697134017944,
      "epoch": 5.149268485567418,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00026784439833073395,
      "loss": 4.8001,
      "mean_token_accuracy": 0.24458896964788437,
      "num_tokens": 119332568.0,
      "step": 52090
    },
    {
      "entropy": 5.670125150680542,
      "epoch": 5.14976275207592,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002678091262873504,
      "loss": 4.7541,
      "mean_token_accuracy": 0.2525578305125237,
      "num_tokens": 119343424.0,
      "step": 52095
    },
    {
      "entropy": 5.754666519165039,
      "epoch": 5.150257018584421,
      "grad_norm": 1.5,
      "learning_rate": 0.0002677738544208635,
      "loss": 4.8999,
      "mean_token_accuracy": 0.2323012113571167,
      "num_tokens": 119354630.0,
      "step": 52100
    },
    {
      "entropy": 5.7501060485839846,
      "epoch": 5.150751285092922,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00026773858273214087,
      "loss": 4.9166,
      "mean_token_accuracy": 0.23349232524633406,
      "num_tokens": 119365222.0,
      "step": 52105
    },
    {
      "entropy": 5.752210330963135,
      "epoch": 5.1512455516014235,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002677033112220501,
      "loss": 4.8277,
      "mean_token_accuracy": 0.24399359822273253,
      "num_tokens": 119377510.0,
      "step": 52110
    },
    {
      "entropy": 5.846501159667969,
      "epoch": 5.151739818109925,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002676680398914592,
      "loss": 4.9423,
      "mean_token_accuracy": 0.23093689233064651,
      "num_tokens": 119389168.0,
      "step": 52115
    },
    {
      "entropy": 5.791972064971924,
      "epoch": 5.152234084618426,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002676327687412356,
      "loss": 4.8536,
      "mean_token_accuracy": 0.23433460891246796,
      "num_tokens": 119399739.0,
      "step": 52120
    },
    {
      "entropy": 5.851007604598999,
      "epoch": 5.152728351126927,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002675974977722471,
      "loss": 4.97,
      "mean_token_accuracy": 0.22799548953771592,
      "num_tokens": 119411654.0,
      "step": 52125
    },
    {
      "entropy": 5.8112188339233395,
      "epoch": 5.153222617635429,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002675622269853612,
      "loss": 4.979,
      "mean_token_accuracy": 0.2324247255921364,
      "num_tokens": 119423003.0,
      "step": 52130
    },
    {
      "entropy": 5.806789922714233,
      "epoch": 5.153716884143931,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002675269563814457,
      "loss": 4.8934,
      "mean_token_accuracy": 0.2416821911931038,
      "num_tokens": 119435359.0,
      "step": 52135
    },
    {
      "entropy": 5.7659525871276855,
      "epoch": 5.154211150652432,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00026749168596136813,
      "loss": 4.9244,
      "mean_token_accuracy": 0.2316432073712349,
      "num_tokens": 119445740.0,
      "step": 52140
    },
    {
      "entropy": 5.778765916824341,
      "epoch": 5.154705417160933,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00026745641572599644,
      "loss": 4.9409,
      "mean_token_accuracy": 0.23024870008230208,
      "num_tokens": 119456862.0,
      "step": 52145
    },
    {
      "entropy": 5.820015239715576,
      "epoch": 5.1551996836694345,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000267421145676198,
      "loss": 4.953,
      "mean_token_accuracy": 0.22794984579086303,
      "num_tokens": 119469405.0,
      "step": 52150
    },
    {
      "entropy": 5.719579076766967,
      "epoch": 5.155693950177936,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002673858758128405,
      "loss": 4.8077,
      "mean_token_accuracy": 0.24238479435443877,
      "num_tokens": 119481739.0,
      "step": 52155
    },
    {
      "entropy": 5.7426918029785154,
      "epoch": 5.156188216686437,
      "grad_norm": 1.296875,
      "learning_rate": 0.00026735060613679177,
      "loss": 4.8735,
      "mean_token_accuracy": 0.23772998601198198,
      "num_tokens": 119494542.0,
      "step": 52160
    },
    {
      "entropy": 5.814143943786621,
      "epoch": 5.156682483194938,
      "grad_norm": 1.25,
      "learning_rate": 0.00026731533664891926,
      "loss": 4.9214,
      "mean_token_accuracy": 0.23200863152742385,
      "num_tokens": 119506360.0,
      "step": 52165
    },
    {
      "entropy": 5.885232830047608,
      "epoch": 5.15717674970344,
      "grad_norm": 1.359375,
      "learning_rate": 0.00026728006735009057,
      "loss": 5.0034,
      "mean_token_accuracy": 0.2238052934408188,
      "num_tokens": 119517401.0,
      "step": 52170
    },
    {
      "entropy": 5.761955118179321,
      "epoch": 5.157671016211942,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00026724479824117355,
      "loss": 4.8385,
      "mean_token_accuracy": 0.23778898566961287,
      "num_tokens": 119528486.0,
      "step": 52175
    },
    {
      "entropy": 5.823266553878784,
      "epoch": 5.158165282720443,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002672095293230356,
      "loss": 4.9784,
      "mean_token_accuracy": 0.2273479089140892,
      "num_tokens": 119539549.0,
      "step": 52180
    },
    {
      "entropy": 5.778042364120483,
      "epoch": 5.158659549228944,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00026717426059654447,
      "loss": 4.9017,
      "mean_token_accuracy": 0.23521946370601654,
      "num_tokens": 119550895.0,
      "step": 52185
    },
    {
      "entropy": 5.808334302902222,
      "epoch": 5.1591538157374455,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002671389920625678,
      "loss": 4.8802,
      "mean_token_accuracy": 0.2393190011382103,
      "num_tokens": 119561950.0,
      "step": 52190
    },
    {
      "entropy": 5.766501426696777,
      "epoch": 5.159648082245947,
      "grad_norm": 1.4375,
      "learning_rate": 0.00026710372372197306,
      "loss": 4.938,
      "mean_token_accuracy": 0.2287602096796036,
      "num_tokens": 119573002.0,
      "step": 52195
    },
    {
      "entropy": 5.791385746002197,
      "epoch": 5.160142348754448,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000267068455575628,
      "loss": 4.8289,
      "mean_token_accuracy": 0.23940352499485015,
      "num_tokens": 119585056.0,
      "step": 52200
    },
    {
      "entropy": 5.759680509567261,
      "epoch": 5.16063661526295,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00026703318762440003,
      "loss": 4.8864,
      "mean_token_accuracy": 0.23162442594766616,
      "num_tokens": 119597048.0,
      "step": 52205
    },
    {
      "entropy": 5.8073410987854,
      "epoch": 5.161130881771451,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002669979198691571,
      "loss": 4.935,
      "mean_token_accuracy": 0.23327739387750626,
      "num_tokens": 119609094.0,
      "step": 52210
    },
    {
      "entropy": 5.752017068862915,
      "epoch": 5.161625148279953,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002669626523107664,
      "loss": 4.8932,
      "mean_token_accuracy": 0.23307142406702042,
      "num_tokens": 119619598.0,
      "step": 52215
    },
    {
      "entropy": 5.820453548431397,
      "epoch": 5.162119414788454,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002669273849500958,
      "loss": 4.8642,
      "mean_token_accuracy": 0.24119096696376802,
      "num_tokens": 119632024.0,
      "step": 52220
    },
    {
      "entropy": 5.832971668243408,
      "epoch": 5.162613681296955,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002668921177880128,
      "loss": 4.9537,
      "mean_token_accuracy": 0.2316473349928856,
      "num_tokens": 119644609.0,
      "step": 52225
    },
    {
      "entropy": 5.8075333595275875,
      "epoch": 5.1631079478054565,
      "grad_norm": 1.234375,
      "learning_rate": 0.00026685685082538486,
      "loss": 4.9169,
      "mean_token_accuracy": 0.23027351945638658,
      "num_tokens": 119657029.0,
      "step": 52230
    },
    {
      "entropy": 5.778679752349854,
      "epoch": 5.163602214313958,
      "grad_norm": 1.484375,
      "learning_rate": 0.00026682158406307973,
      "loss": 4.9172,
      "mean_token_accuracy": 0.23953217417001724,
      "num_tokens": 119669104.0,
      "step": 52235
    },
    {
      "entropy": 5.752894687652588,
      "epoch": 5.16409648082246,
      "grad_norm": 1.375,
      "learning_rate": 0.00026678631750196485,
      "loss": 4.9114,
      "mean_token_accuracy": 0.23704142421483992,
      "num_tokens": 119679772.0,
      "step": 52240
    },
    {
      "entropy": 5.80510106086731,
      "epoch": 5.164590747330961,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00026675105114290786,
      "loss": 4.8676,
      "mean_token_accuracy": 0.23960766941308975,
      "num_tokens": 119691481.0,
      "step": 52245
    },
    {
      "entropy": 5.741370820999146,
      "epoch": 5.165085013839462,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00026671578498677637,
      "loss": 4.8577,
      "mean_token_accuracy": 0.2386803776025772,
      "num_tokens": 119703593.0,
      "step": 52250
    },
    {
      "entropy": 5.777454471588134,
      "epoch": 5.165579280347964,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002666805190344377,
      "loss": 4.8753,
      "mean_token_accuracy": 0.2390339970588684,
      "num_tokens": 119715362.0,
      "step": 52255
    },
    {
      "entropy": 5.783180093765258,
      "epoch": 5.166073546856465,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00026664525328675965,
      "loss": 4.9289,
      "mean_token_accuracy": 0.2284673497080803,
      "num_tokens": 119726975.0,
      "step": 52260
    },
    {
      "entropy": 5.726257848739624,
      "epoch": 5.166567813364966,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00026660998774460964,
      "loss": 4.8011,
      "mean_token_accuracy": 0.2488056406378746,
      "num_tokens": 119738093.0,
      "step": 52265
    },
    {
      "entropy": 5.810727262496949,
      "epoch": 5.167062079873467,
      "grad_norm": 1.328125,
      "learning_rate": 0.00026657472240885523,
      "loss": 4.947,
      "mean_token_accuracy": 0.23611362725496293,
      "num_tokens": 119750136.0,
      "step": 52270
    },
    {
      "entropy": 5.748880910873413,
      "epoch": 5.16755634638197,
      "grad_norm": 1.28125,
      "learning_rate": 0.000266539457280364,
      "loss": 4.8313,
      "mean_token_accuracy": 0.2422332912683487,
      "num_tokens": 119760954.0,
      "step": 52275
    },
    {
      "entropy": 5.7187803268432615,
      "epoch": 5.168050612890471,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002665041923600034,
      "loss": 4.8228,
      "mean_token_accuracy": 0.23759566992521286,
      "num_tokens": 119771330.0,
      "step": 52280
    },
    {
      "entropy": 5.695650196075439,
      "epoch": 5.168544879398972,
      "grad_norm": 1.34375,
      "learning_rate": 0.00026646892764864096,
      "loss": 4.8654,
      "mean_token_accuracy": 0.24107057154178618,
      "num_tokens": 119782269.0,
      "step": 52285
    },
    {
      "entropy": 5.755441379547119,
      "epoch": 5.169039145907473,
      "grad_norm": 1.296875,
      "learning_rate": 0.00026643366314714424,
      "loss": 4.9433,
      "mean_token_accuracy": 0.22619068175554274,
      "num_tokens": 119793671.0,
      "step": 52290
    },
    {
      "entropy": 5.770061874389649,
      "epoch": 5.169533412415975,
      "grad_norm": 1.3125,
      "learning_rate": 0.00026639839885638075,
      "loss": 4.8589,
      "mean_token_accuracy": 0.23388723731040956,
      "num_tokens": 119804416.0,
      "step": 52295
    },
    {
      "entropy": 5.749242353439331,
      "epoch": 5.170027678924476,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00026636313477721796,
      "loss": 4.8245,
      "mean_token_accuracy": 0.2407286658883095,
      "num_tokens": 119816269.0,
      "step": 52300
    },
    {
      "entropy": 5.766209459304809,
      "epoch": 5.170521945432977,
      "grad_norm": 1.546875,
      "learning_rate": 0.00026632787091052345,
      "loss": 4.8892,
      "mean_token_accuracy": 0.23497200161218643,
      "num_tokens": 119827899.0,
      "step": 52305
    },
    {
      "entropy": 5.841371822357178,
      "epoch": 5.171016211941478,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00026629260725716464,
      "loss": 4.8971,
      "mean_token_accuracy": 0.23108644038438797,
      "num_tokens": 119840300.0,
      "step": 52310
    },
    {
      "entropy": 5.8218026638031,
      "epoch": 5.171510478449981,
      "grad_norm": 1.34375,
      "learning_rate": 0.000266257343818009,
      "loss": 4.9399,
      "mean_token_accuracy": 0.23020581752061844,
      "num_tokens": 119852260.0,
      "step": 52315
    },
    {
      "entropy": 5.760693311691284,
      "epoch": 5.172004744958482,
      "grad_norm": 1.34375,
      "learning_rate": 0.000266222080593924,
      "loss": 4.8804,
      "mean_token_accuracy": 0.23899160027503968,
      "num_tokens": 119864348.0,
      "step": 52320
    },
    {
      "entropy": 5.754494476318359,
      "epoch": 5.172499011466983,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002661868175857773,
      "loss": 4.8424,
      "mean_token_accuracy": 0.2402345985174179,
      "num_tokens": 119875667.0,
      "step": 52325
    },
    {
      "entropy": 5.759509515762329,
      "epoch": 5.172993277975484,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00026615155479443626,
      "loss": 4.8565,
      "mean_token_accuracy": 0.24012009501457215,
      "num_tokens": 119886944.0,
      "step": 52330
    },
    {
      "entropy": 5.772789144515992,
      "epoch": 5.173487544483986,
      "grad_norm": 1.25,
      "learning_rate": 0.00026611629222076825,
      "loss": 4.9107,
      "mean_token_accuracy": 0.23416221439838408,
      "num_tokens": 119898477.0,
      "step": 52335
    },
    {
      "entropy": 5.807863855361939,
      "epoch": 5.173981810992487,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00026608102986564093,
      "loss": 4.9488,
      "mean_token_accuracy": 0.23154127448797227,
      "num_tokens": 119910500.0,
      "step": 52340
    },
    {
      "entropy": 5.736991739273071,
      "epoch": 5.174476077500988,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002660457677299215,
      "loss": 4.8589,
      "mean_token_accuracy": 0.24088558405637742,
      "num_tokens": 119922493.0,
      "step": 52345
    },
    {
      "entropy": 5.8319010734558105,
      "epoch": 5.17497034400949,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002660105058144777,
      "loss": 4.8905,
      "mean_token_accuracy": 0.23249326050281524,
      "num_tokens": 119933215.0,
      "step": 52350
    },
    {
      "entropy": 5.811773061752319,
      "epoch": 5.175464610517992,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002659752441201769,
      "loss": 4.9945,
      "mean_token_accuracy": 0.227525731921196,
      "num_tokens": 119945337.0,
      "step": 52355
    },
    {
      "entropy": 5.780213356018066,
      "epoch": 5.175958877026493,
      "grad_norm": 1.375,
      "learning_rate": 0.00026593998264788643,
      "loss": 4.8358,
      "mean_token_accuracy": 0.24163734316825866,
      "num_tokens": 119955027.0,
      "step": 52360
    },
    {
      "entropy": 5.826695442199707,
      "epoch": 5.176453143534994,
      "grad_norm": 1.375,
      "learning_rate": 0.00026590472139847375,
      "loss": 5.0171,
      "mean_token_accuracy": 0.22727564722299576,
      "num_tokens": 119966994.0,
      "step": 52365
    },
    {
      "entropy": 5.768813991546631,
      "epoch": 5.176947410043495,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002658694603728064,
      "loss": 4.85,
      "mean_token_accuracy": 0.2390779286623001,
      "num_tokens": 119977412.0,
      "step": 52370
    },
    {
      "entropy": 5.813506269454956,
      "epoch": 5.177441676551997,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00026583419957175165,
      "loss": 4.9556,
      "mean_token_accuracy": 0.22832420468330383,
      "num_tokens": 119990088.0,
      "step": 52375
    },
    {
      "entropy": 5.814873838424683,
      "epoch": 5.177935943060498,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002657989389961771,
      "loss": 4.9127,
      "mean_token_accuracy": 0.2346550017595291,
      "num_tokens": 120000284.0,
      "step": 52380
    },
    {
      "entropy": 5.744707345962524,
      "epoch": 5.178430209569,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00026576367864695013,
      "loss": 4.8556,
      "mean_token_accuracy": 0.2412243440747261,
      "num_tokens": 120011805.0,
      "step": 52385
    },
    {
      "entropy": 5.81412992477417,
      "epoch": 5.178924476077501,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000265728418524938,
      "loss": 4.9265,
      "mean_token_accuracy": 0.2295692890882492,
      "num_tokens": 120023391.0,
      "step": 52390
    },
    {
      "entropy": 5.785387802124023,
      "epoch": 5.1794187425860025,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00026569315863100824,
      "loss": 4.8602,
      "mean_token_accuracy": 0.23535608649253845,
      "num_tokens": 120034613.0,
      "step": 52395
    },
    {
      "entropy": 5.81029782295227,
      "epoch": 5.179913009094504,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00026565789896602817,
      "loss": 4.9922,
      "mean_token_accuracy": 0.22745047509670258,
      "num_tokens": 120046843.0,
      "step": 52400
    },
    {
      "entropy": 5.829337215423584,
      "epoch": 5.180407275603005,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002656226395308653,
      "loss": 4.9619,
      "mean_token_accuracy": 0.22829216420650483,
      "num_tokens": 120058823.0,
      "step": 52405
    },
    {
      "entropy": 5.810755825042724,
      "epoch": 5.180901542111506,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00026558738032638694,
      "loss": 4.9291,
      "mean_token_accuracy": 0.2341179832816124,
      "num_tokens": 120071082.0,
      "step": 52410
    },
    {
      "entropy": 5.854770994186401,
      "epoch": 5.181395808620008,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00026555212135346044,
      "loss": 5.031,
      "mean_token_accuracy": 0.22222233265638353,
      "num_tokens": 120084009.0,
      "step": 52415
    },
    {
      "entropy": 5.769263982772827,
      "epoch": 5.181890075128509,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00026551686261295333,
      "loss": 4.8919,
      "mean_token_accuracy": 0.23295036256313323,
      "num_tokens": 120094927.0,
      "step": 52420
    },
    {
      "entropy": 5.797933673858642,
      "epoch": 5.182384341637011,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002654816041057328,
      "loss": 4.9269,
      "mean_token_accuracy": 0.23214564025402068,
      "num_tokens": 120108343.0,
      "step": 52425
    },
    {
      "entropy": 5.803945589065552,
      "epoch": 5.182878608145512,
      "grad_norm": 1.421875,
      "learning_rate": 0.00026544634583266623,
      "loss": 4.9127,
      "mean_token_accuracy": 0.23175592720508575,
      "num_tokens": 120119660.0,
      "step": 52430
    },
    {
      "entropy": 5.795047473907471,
      "epoch": 5.1833728746540135,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00026541108779462114,
      "loss": 4.8973,
      "mean_token_accuracy": 0.2254657492041588,
      "num_tokens": 120130827.0,
      "step": 52435
    },
    {
      "entropy": 5.8015471458435055,
      "epoch": 5.183867141162515,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00026537582999246466,
      "loss": 4.8663,
      "mean_token_accuracy": 0.23336290717124938,
      "num_tokens": 120142243.0,
      "step": 52440
    },
    {
      "entropy": 5.697288084030151,
      "epoch": 5.184361407671016,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002653405724270644,
      "loss": 4.8074,
      "mean_token_accuracy": 0.2458158776164055,
      "num_tokens": 120153419.0,
      "step": 52445
    },
    {
      "entropy": 5.769218444824219,
      "epoch": 5.184855674179517,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00026530531509928745,
      "loss": 4.8782,
      "mean_token_accuracy": 0.23883734047412872,
      "num_tokens": 120165622.0,
      "step": 52450
    },
    {
      "entropy": 5.807198715209961,
      "epoch": 5.185349940688019,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00026527005801000133,
      "loss": 4.9155,
      "mean_token_accuracy": 0.2306995674967766,
      "num_tokens": 120176333.0,
      "step": 52455
    },
    {
      "entropy": 5.816861534118653,
      "epoch": 5.185844207196521,
      "grad_norm": 1.25,
      "learning_rate": 0.0002652348011600732,
      "loss": 4.8985,
      "mean_token_accuracy": 0.2344875916838646,
      "num_tokens": 120187910.0,
      "step": 52460
    },
    {
      "entropy": 5.770047664642334,
      "epoch": 5.186338473705022,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002651995445503705,
      "loss": 4.8852,
      "mean_token_accuracy": 0.23669556230306626,
      "num_tokens": 120199169.0,
      "step": 52465
    },
    {
      "entropy": 5.790364694595337,
      "epoch": 5.186832740213523,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00026516428818176054,
      "loss": 4.8445,
      "mean_token_accuracy": 0.23834069818258286,
      "num_tokens": 120209230.0,
      "step": 52470
    },
    {
      "entropy": 5.791821432113648,
      "epoch": 5.1873270067220245,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002651290320551107,
      "loss": 4.9556,
      "mean_token_accuracy": 0.22904090881347655,
      "num_tokens": 120220681.0,
      "step": 52475
    },
    {
      "entropy": 5.779607820510864,
      "epoch": 5.187821273230526,
      "grad_norm": 1.265625,
      "learning_rate": 0.00026509377617128804,
      "loss": 4.9211,
      "mean_token_accuracy": 0.2310744047164917,
      "num_tokens": 120232823.0,
      "step": 52480
    },
    {
      "entropy": 5.810683679580689,
      "epoch": 5.188315539739027,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002650585205311602,
      "loss": 4.9093,
      "mean_token_accuracy": 0.2327719360589981,
      "num_tokens": 120244319.0,
      "step": 52485
    },
    {
      "entropy": 5.7898133277893065,
      "epoch": 5.188809806247528,
      "grad_norm": 1.25,
      "learning_rate": 0.0002650232651355941,
      "loss": 4.8733,
      "mean_token_accuracy": 0.24025457948446274,
      "num_tokens": 120255415.0,
      "step": 52490
    },
    {
      "entropy": 5.772481250762939,
      "epoch": 5.1893040727560305,
      "grad_norm": 1.328125,
      "learning_rate": 0.00026498800998545736,
      "loss": 4.9021,
      "mean_token_accuracy": 0.23460862338542937,
      "num_tokens": 120266074.0,
      "step": 52495
    },
    {
      "entropy": 5.728386783599854,
      "epoch": 5.189798339264532,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002649527550816171,
      "loss": 4.89,
      "mean_token_accuracy": 0.23353151232004166,
      "num_tokens": 120277491.0,
      "step": 52500
    },
    {
      "entropy": 5.760611057281494,
      "epoch": 5.190292605773033,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002649175004249406,
      "loss": 4.8192,
      "mean_token_accuracy": 0.2410721018910408,
      "num_tokens": 120288190.0,
      "step": 52505
    },
    {
      "entropy": 5.7306712627410885,
      "epoch": 5.190786872281534,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00026488224601629526,
      "loss": 4.833,
      "mean_token_accuracy": 0.23720658272504808,
      "num_tokens": 120299001.0,
      "step": 52510
    },
    {
      "entropy": 5.767369985580444,
      "epoch": 5.1912811387900355,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002648469918565481,
      "loss": 4.8187,
      "mean_token_accuracy": 0.24812449067831038,
      "num_tokens": 120310965.0,
      "step": 52515
    },
    {
      "entropy": 5.753214597702026,
      "epoch": 5.191775405298537,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00026481173794656653,
      "loss": 4.8619,
      "mean_token_accuracy": 0.23485870212316512,
      "num_tokens": 120322954.0,
      "step": 52520
    },
    {
      "entropy": 5.654851675033569,
      "epoch": 5.192269671807038,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002647764842872179,
      "loss": 4.7815,
      "mean_token_accuracy": 0.2452276885509491,
      "num_tokens": 120335131.0,
      "step": 52525
    },
    {
      "entropy": 5.708177280426026,
      "epoch": 5.19276393831554,
      "grad_norm": 1.203125,
      "learning_rate": 0.00026474123087936926,
      "loss": 4.8008,
      "mean_token_accuracy": 0.24617787599563598,
      "num_tokens": 120346068.0,
      "step": 52530
    },
    {
      "entropy": 5.7385354995727536,
      "epoch": 5.193258204824041,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002647059777238879,
      "loss": 4.842,
      "mean_token_accuracy": 0.2427709922194481,
      "num_tokens": 120357781.0,
      "step": 52535
    },
    {
      "entropy": 5.70217456817627,
      "epoch": 5.193752471332543,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002646707248216412,
      "loss": 4.7563,
      "mean_token_accuracy": 0.2522758573293686,
      "num_tokens": 120367910.0,
      "step": 52540
    },
    {
      "entropy": 5.754609107971191,
      "epoch": 5.194246737841044,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002646354721734962,
      "loss": 4.8785,
      "mean_token_accuracy": 0.23436597287654876,
      "num_tokens": 120381215.0,
      "step": 52545
    },
    {
      "entropy": 5.826053333282471,
      "epoch": 5.194741004349545,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002646002197803201,
      "loss": 4.9672,
      "mean_token_accuracy": 0.23190339356660844,
      "num_tokens": 120392733.0,
      "step": 52550
    },
    {
      "entropy": 5.752792263031006,
      "epoch": 5.1952352708580465,
      "grad_norm": 1.4375,
      "learning_rate": 0.00026456496764298035,
      "loss": 4.9242,
      "mean_token_accuracy": 0.22858668863773346,
      "num_tokens": 120403668.0,
      "step": 52555
    },
    {
      "entropy": 5.8375904083251955,
      "epoch": 5.195729537366548,
      "grad_norm": 1.421875,
      "learning_rate": 0.000264529715762344,
      "loss": 5.0054,
      "mean_token_accuracy": 0.2244014024734497,
      "num_tokens": 120416104.0,
      "step": 52560
    },
    {
      "entropy": 5.7775591850280765,
      "epoch": 5.196223803875049,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00026449446413927816,
      "loss": 4.8315,
      "mean_token_accuracy": 0.23757384568452836,
      "num_tokens": 120426460.0,
      "step": 52565
    },
    {
      "entropy": 5.795022916793823,
      "epoch": 5.196718070383551,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002644592127746503,
      "loss": 4.9293,
      "mean_token_accuracy": 0.22576959133148194,
      "num_tokens": 120437236.0,
      "step": 52570
    },
    {
      "entropy": 5.785792493820191,
      "epoch": 5.197212336892052,
      "grad_norm": 1.390625,
      "learning_rate": 0.00026442396166932725,
      "loss": 4.8824,
      "mean_token_accuracy": 0.23254936784505845,
      "num_tokens": 120448054.0,
      "step": 52575
    },
    {
      "entropy": 5.775089359283447,
      "epoch": 5.197706603400554,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002643887108241765,
      "loss": 4.9329,
      "mean_token_accuracy": 0.22964668422937393,
      "num_tokens": 120460284.0,
      "step": 52580
    },
    {
      "entropy": 5.795315408706665,
      "epoch": 5.198200869909055,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002643534602400651,
      "loss": 4.9252,
      "mean_token_accuracy": 0.22973159104585647,
      "num_tokens": 120472268.0,
      "step": 52585
    },
    {
      "entropy": 5.830442142486572,
      "epoch": 5.198695136417556,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002643182099178602,
      "loss": 4.9559,
      "mean_token_accuracy": 0.23254806250333787,
      "num_tokens": 120483117.0,
      "step": 52590
    },
    {
      "entropy": 5.809105443954468,
      "epoch": 5.1991894029260575,
      "grad_norm": 1.265625,
      "learning_rate": 0.00026428295985842903,
      "loss": 4.8763,
      "mean_token_accuracy": 0.23150597512722015,
      "num_tokens": 120494493.0,
      "step": 52595
    },
    {
      "entropy": 5.799513101577759,
      "epoch": 5.199683669434559,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00026424771006263863,
      "loss": 4.8718,
      "mean_token_accuracy": 0.2364787846803665,
      "num_tokens": 120504737.0,
      "step": 52600
    },
    {
      "entropy": 5.686538124084473,
      "epoch": 5.200177935943061,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00026421246053135627,
      "loss": 4.8208,
      "mean_token_accuracy": 0.24905122816562653,
      "num_tokens": 120516516.0,
      "step": 52605
    },
    {
      "entropy": 5.712759733200073,
      "epoch": 5.200672202451562,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002641772112654491,
      "loss": 4.8169,
      "mean_token_accuracy": 0.2359617218375206,
      "num_tokens": 120527327.0,
      "step": 52610
    },
    {
      "entropy": 5.799140977859497,
      "epoch": 5.201166468960063,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00026414196226578416,
      "loss": 4.8824,
      "mean_token_accuracy": 0.23932477533817292,
      "num_tokens": 120538578.0,
      "step": 52615
    },
    {
      "entropy": 5.7796721935272215,
      "epoch": 5.201660735468565,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002641067135332287,
      "loss": 4.8994,
      "mean_token_accuracy": 0.24005429148674012,
      "num_tokens": 120550682.0,
      "step": 52620
    },
    {
      "entropy": 5.68453950881958,
      "epoch": 5.202155001977066,
      "grad_norm": 1.390625,
      "learning_rate": 0.00026407146506864977,
      "loss": 4.7472,
      "mean_token_accuracy": 0.2489650994539261,
      "num_tokens": 120561110.0,
      "step": 52625
    },
    {
      "entropy": 5.745421361923218,
      "epoch": 5.202649268485567,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002640362168729145,
      "loss": 4.8587,
      "mean_token_accuracy": 0.24351892173290252,
      "num_tokens": 120572104.0,
      "step": 52630
    },
    {
      "entropy": 5.75730299949646,
      "epoch": 5.2031435349940685,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00026400096894688987,
      "loss": 4.8862,
      "mean_token_accuracy": 0.22695060670375825,
      "num_tokens": 120582639.0,
      "step": 52635
    },
    {
      "entropy": 5.773979330062867,
      "epoch": 5.203637801502571,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002639657212914432,
      "loss": 4.8756,
      "mean_token_accuracy": 0.23909149467945098,
      "num_tokens": 120594615.0,
      "step": 52640
    },
    {
      "entropy": 5.785218715667725,
      "epoch": 5.204132068011072,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002639304739074415,
      "loss": 4.9096,
      "mean_token_accuracy": 0.23432474732398986,
      "num_tokens": 120604970.0,
      "step": 52645
    },
    {
      "entropy": 5.750272464752197,
      "epoch": 5.204626334519573,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002638952267957518,
      "loss": 4.8702,
      "mean_token_accuracy": 0.23436136543750763,
      "num_tokens": 120616280.0,
      "step": 52650
    },
    {
      "entropy": 5.7475183486938475,
      "epoch": 5.205120601028074,
      "grad_norm": 1.375,
      "learning_rate": 0.00026385997995724135,
      "loss": 4.9071,
      "mean_token_accuracy": 0.2341076448559761,
      "num_tokens": 120628790.0,
      "step": 52655
    },
    {
      "entropy": 5.797174739837646,
      "epoch": 5.205614867536576,
      "grad_norm": 1.328125,
      "learning_rate": 0.00026382473339277713,
      "loss": 4.9405,
      "mean_token_accuracy": 0.22537988126277925,
      "num_tokens": 120639906.0,
      "step": 52660
    },
    {
      "entropy": 5.740252876281739,
      "epoch": 5.206109134045077,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002637894871032261,
      "loss": 4.8585,
      "mean_token_accuracy": 0.23370161205530166,
      "num_tokens": 120651429.0,
      "step": 52665
    },
    {
      "entropy": 5.696439218521118,
      "epoch": 5.206603400553578,
      "grad_norm": 1.421875,
      "learning_rate": 0.00026375424108945547,
      "loss": 4.6946,
      "mean_token_accuracy": 0.24924720525741578,
      "num_tokens": 120662119.0,
      "step": 52670
    },
    {
      "entropy": 5.7125077724456785,
      "epoch": 5.207097667062079,
      "grad_norm": 1.375,
      "learning_rate": 0.00026371899535233216,
      "loss": 4.8322,
      "mean_token_accuracy": 0.24521962851285933,
      "num_tokens": 120672719.0,
      "step": 52675
    },
    {
      "entropy": 5.720586585998535,
      "epoch": 5.207591933570582,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00026368374989272345,
      "loss": 4.9323,
      "mean_token_accuracy": 0.23373921364545822,
      "num_tokens": 120683287.0,
      "step": 52680
    },
    {
      "entropy": 5.755691766738892,
      "epoch": 5.208086200079083,
      "grad_norm": 1.390625,
      "learning_rate": 0.00026364850471149624,
      "loss": 4.9555,
      "mean_token_accuracy": 0.2326302111148834,
      "num_tokens": 120693867.0,
      "step": 52685
    },
    {
      "entropy": 5.803948974609375,
      "epoch": 5.208580466587584,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002636132598095175,
      "loss": 4.9,
      "mean_token_accuracy": 0.23293645828962325,
      "num_tokens": 120706068.0,
      "step": 52690
    },
    {
      "entropy": 5.7815336227417,
      "epoch": 5.209074733096085,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002635780151876544,
      "loss": 4.8891,
      "mean_token_accuracy": 0.24202115833759308,
      "num_tokens": 120717111.0,
      "step": 52695
    },
    {
      "entropy": 5.8339159965515135,
      "epoch": 5.209568999604587,
      "grad_norm": 1.34375,
      "learning_rate": 0.00026354277084677386,
      "loss": 4.9059,
      "mean_token_accuracy": 0.24094317406415938,
      "num_tokens": 120729468.0,
      "step": 52700
    },
    {
      "entropy": 5.829070329666138,
      "epoch": 5.210063266113088,
      "grad_norm": 1.375,
      "learning_rate": 0.000263507526787743,
      "loss": 4.9657,
      "mean_token_accuracy": 0.22731245160102845,
      "num_tokens": 120740101.0,
      "step": 52705
    },
    {
      "entropy": 5.803519058227539,
      "epoch": 5.210557532621589,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002634722830114287,
      "loss": 4.9429,
      "mean_token_accuracy": 0.23000066727399826,
      "num_tokens": 120750250.0,
      "step": 52710
    },
    {
      "entropy": 5.72346076965332,
      "epoch": 5.211051799130091,
      "grad_norm": 1.40625,
      "learning_rate": 0.00026343703951869804,
      "loss": 4.8339,
      "mean_token_accuracy": 0.24094051420688628,
      "num_tokens": 120761955.0,
      "step": 52715
    },
    {
      "entropy": 5.711988306045532,
      "epoch": 5.211546065638593,
      "grad_norm": 1.40625,
      "learning_rate": 0.00026340179631041796,
      "loss": 4.8072,
      "mean_token_accuracy": 0.24470532387495042,
      "num_tokens": 120773660.0,
      "step": 52720
    },
    {
      "entropy": 5.805677032470703,
      "epoch": 5.212040332147094,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002633665533874555,
      "loss": 4.9201,
      "mean_token_accuracy": 0.2288750573992729,
      "num_tokens": 120785571.0,
      "step": 52725
    },
    {
      "entropy": 5.803398323059082,
      "epoch": 5.212534598655595,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002633313107506777,
      "loss": 4.9718,
      "mean_token_accuracy": 0.23289743661880494,
      "num_tokens": 120799292.0,
      "step": 52730
    },
    {
      "entropy": 5.785973358154297,
      "epoch": 5.213028865164096,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00026329606840095146,
      "loss": 4.9188,
      "mean_token_accuracy": 0.23506598323583602,
      "num_tokens": 120811839.0,
      "step": 52735
    },
    {
      "entropy": 5.819769191741943,
      "epoch": 5.213523131672598,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00026326082633914373,
      "loss": 4.8808,
      "mean_token_accuracy": 0.24286891371011735,
      "num_tokens": 120822859.0,
      "step": 52740
    },
    {
      "entropy": 5.818783712387085,
      "epoch": 5.214017398181099,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00026322558456612144,
      "loss": 4.9483,
      "mean_token_accuracy": 0.21918484419584275,
      "num_tokens": 120833146.0,
      "step": 52745
    },
    {
      "entropy": 5.8454365730285645,
      "epoch": 5.214511664689601,
      "grad_norm": 1.515625,
      "learning_rate": 0.00026319034308275156,
      "loss": 4.9414,
      "mean_token_accuracy": 0.23100204914808273,
      "num_tokens": 120843229.0,
      "step": 52750
    },
    {
      "entropy": 5.733302402496338,
      "epoch": 5.215005931198102,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002631551018899012,
      "loss": 4.8743,
      "mean_token_accuracy": 0.2432454526424408,
      "num_tokens": 120855374.0,
      "step": 52755
    },
    {
      "entropy": 5.763633298873901,
      "epoch": 5.215500197706604,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00026311986098843715,
      "loss": 4.897,
      "mean_token_accuracy": 0.2358341783285141,
      "num_tokens": 120867762.0,
      "step": 52760
    },
    {
      "entropy": 5.798912525177002,
      "epoch": 5.215994464215105,
      "grad_norm": 1.4375,
      "learning_rate": 0.00026308462037922636,
      "loss": 4.8722,
      "mean_token_accuracy": 0.23197089582681657,
      "num_tokens": 120879235.0,
      "step": 52765
    },
    {
      "entropy": 5.811575031280517,
      "epoch": 5.216488730723606,
      "grad_norm": 1.359375,
      "learning_rate": 0.00026304938006313565,
      "loss": 4.908,
      "mean_token_accuracy": 0.23741862326860427,
      "num_tokens": 120890500.0,
      "step": 52770
    },
    {
      "entropy": 5.752939081192016,
      "epoch": 5.216982997232107,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002630141400410322,
      "loss": 5.0079,
      "mean_token_accuracy": 0.22500530034303665,
      "num_tokens": 120901812.0,
      "step": 52775
    },
    {
      "entropy": 5.7808174133300785,
      "epoch": 5.217477263740609,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00026297890031378265,
      "loss": 4.905,
      "mean_token_accuracy": 0.2354174107313156,
      "num_tokens": 120913934.0,
      "step": 52780
    },
    {
      "entropy": 5.79930624961853,
      "epoch": 5.217971530249111,
      "grad_norm": 1.4375,
      "learning_rate": 0.00026294366088225405,
      "loss": 4.9072,
      "mean_token_accuracy": 0.22858721911907195,
      "num_tokens": 120924851.0,
      "step": 52785
    },
    {
      "entropy": 5.749112939834594,
      "epoch": 5.218465796757612,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00026290842174731335,
      "loss": 4.8067,
      "mean_token_accuracy": 0.24057770520448685,
      "num_tokens": 120935986.0,
      "step": 52790
    },
    {
      "entropy": 5.868067789077759,
      "epoch": 5.218960063266113,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002628731829098272,
      "loss": 4.9976,
      "mean_token_accuracy": 0.22129117101430892,
      "num_tokens": 120948153.0,
      "step": 52795
    },
    {
      "entropy": 5.847903108596801,
      "epoch": 5.2194543297746145,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002628379443706627,
      "loss": 4.9229,
      "mean_token_accuracy": 0.2340480923652649,
      "num_tokens": 120958524.0,
      "step": 52800
    },
    {
      "entropy": 5.783440017700196,
      "epoch": 5.219948596283116,
      "grad_norm": 1.40625,
      "learning_rate": 0.00026280270613068666,
      "loss": 4.8904,
      "mean_token_accuracy": 0.23051270097494125,
      "num_tokens": 120970332.0,
      "step": 52805
    },
    {
      "entropy": 5.796902799606324,
      "epoch": 5.220442862791617,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002627674681907659,
      "loss": 4.9586,
      "mean_token_accuracy": 0.22956990897655488,
      "num_tokens": 120982315.0,
      "step": 52810
    },
    {
      "entropy": 5.670535612106323,
      "epoch": 5.220937129300118,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002627322305517674,
      "loss": 4.7897,
      "mean_token_accuracy": 0.24570701867341996,
      "num_tokens": 120993631.0,
      "step": 52815
    },
    {
      "entropy": 5.743099641799927,
      "epoch": 5.22143139580862,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002626969932145579,
      "loss": 4.8457,
      "mean_token_accuracy": 0.24059172421693803,
      "num_tokens": 121005392.0,
      "step": 52820
    },
    {
      "entropy": 5.703053188323975,
      "epoch": 5.221925662317122,
      "grad_norm": 1.46875,
      "learning_rate": 0.00026266175618000427,
      "loss": 4.7339,
      "mean_token_accuracy": 0.24860429465770723,
      "num_tokens": 121015775.0,
      "step": 52825
    },
    {
      "entropy": 5.811665964126587,
      "epoch": 5.222419928825623,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002626265194489734,
      "loss": 4.9037,
      "mean_token_accuracy": 0.23336833566427231,
      "num_tokens": 121026635.0,
      "step": 52830
    },
    {
      "entropy": 5.756366729736328,
      "epoch": 5.222914195334124,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000262591283022332,
      "loss": 4.8625,
      "mean_token_accuracy": 0.24737353920936583,
      "num_tokens": 121037643.0,
      "step": 52835
    },
    {
      "entropy": 5.853870534896851,
      "epoch": 5.2234084618426255,
      "grad_norm": 1.328125,
      "learning_rate": 0.00026255604690094694,
      "loss": 5.0181,
      "mean_token_accuracy": 0.22399319261312484,
      "num_tokens": 121049936.0,
      "step": 52840
    },
    {
      "entropy": 5.717159366607666,
      "epoch": 5.223902728351127,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002625208110856851,
      "loss": 4.771,
      "mean_token_accuracy": 0.25098422914743423,
      "num_tokens": 121062594.0,
      "step": 52845
    },
    {
      "entropy": 5.7594257354736325,
      "epoch": 5.224396994859628,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002624855755774133,
      "loss": 4.8816,
      "mean_token_accuracy": 0.23950533270835878,
      "num_tokens": 121074129.0,
      "step": 52850
    },
    {
      "entropy": 5.885507011413575,
      "epoch": 5.224891261368129,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002624503403769982,
      "loss": 4.9987,
      "mean_token_accuracy": 0.22132306694984435,
      "num_tokens": 121085829.0,
      "step": 52855
    },
    {
      "entropy": 5.8054876804351805,
      "epoch": 5.2253855278766315,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002624151054853067,
      "loss": 4.9541,
      "mean_token_accuracy": 0.2266192153096199,
      "num_tokens": 121096645.0,
      "step": 52860
    },
    {
      "entropy": 5.770385885238648,
      "epoch": 5.225879794385133,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002623798709032056,
      "loss": 4.8762,
      "mean_token_accuracy": 0.2361864998936653,
      "num_tokens": 121109236.0,
      "step": 52865
    },
    {
      "entropy": 5.722325420379638,
      "epoch": 5.226374060893634,
      "grad_norm": 1.375,
      "learning_rate": 0.0002623446366315615,
      "loss": 4.8687,
      "mean_token_accuracy": 0.2388315826654434,
      "num_tokens": 121121139.0,
      "step": 52870
    },
    {
      "entropy": 5.794505929946899,
      "epoch": 5.226868327402135,
      "grad_norm": 1.734375,
      "learning_rate": 0.00026230940267124136,
      "loss": 4.892,
      "mean_token_accuracy": 0.23590610921382904,
      "num_tokens": 121133112.0,
      "step": 52875
    },
    {
      "entropy": 5.788093185424804,
      "epoch": 5.2273625939106365,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00026227416902311197,
      "loss": 4.8811,
      "mean_token_accuracy": 0.23319605737924576,
      "num_tokens": 121144414.0,
      "step": 52880
    },
    {
      "entropy": 5.742389488220215,
      "epoch": 5.227856860419138,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00026223893568803994,
      "loss": 4.8539,
      "mean_token_accuracy": 0.236684150993824,
      "num_tokens": 121156742.0,
      "step": 52885
    },
    {
      "entropy": 5.729253911972046,
      "epoch": 5.228351126927639,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002622037026668921,
      "loss": 4.8961,
      "mean_token_accuracy": 0.24452172219753265,
      "num_tokens": 121169494.0,
      "step": 52890
    },
    {
      "entropy": 5.838353538513184,
      "epoch": 5.228845393436141,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002621684699605351,
      "loss": 5.0321,
      "mean_token_accuracy": 0.22803181111812593,
      "num_tokens": 121181700.0,
      "step": 52895
    },
    {
      "entropy": 5.78857011795044,
      "epoch": 5.2293396599446424,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002621332375698357,
      "loss": 4.8444,
      "mean_token_accuracy": 0.2422074556350708,
      "num_tokens": 121193428.0,
      "step": 52900
    },
    {
      "entropy": 5.773274087905884,
      "epoch": 5.229833926453144,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002620980054956608,
      "loss": 4.8954,
      "mean_token_accuracy": 0.2380651742219925,
      "num_tokens": 121205103.0,
      "step": 52905
    },
    {
      "entropy": 5.79285831451416,
      "epoch": 5.230328192961645,
      "grad_norm": 1.375,
      "learning_rate": 0.0002620627737388768,
      "loss": 4.9043,
      "mean_token_accuracy": 0.23574793338775635,
      "num_tokens": 121217149.0,
      "step": 52910
    },
    {
      "entropy": 5.8601312160491945,
      "epoch": 5.230822459470146,
      "grad_norm": 1.375,
      "learning_rate": 0.0002620275423003507,
      "loss": 4.9804,
      "mean_token_accuracy": 0.23017164915800095,
      "num_tokens": 121228604.0,
      "step": 52915
    },
    {
      "entropy": 5.790221929550171,
      "epoch": 5.2313167259786475,
      "grad_norm": 1.40625,
      "learning_rate": 0.00026199231118094907,
      "loss": 4.8897,
      "mean_token_accuracy": 0.2373717397451401,
      "num_tokens": 121240290.0,
      "step": 52920
    },
    {
      "entropy": 5.779712915420532,
      "epoch": 5.231810992487149,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002619570803815385,
      "loss": 4.8973,
      "mean_token_accuracy": 0.23402367830276488,
      "num_tokens": 121252294.0,
      "step": 52925
    },
    {
      "entropy": 5.822255849838257,
      "epoch": 5.23230525899565,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00026192184990298585,
      "loss": 4.9429,
      "mean_token_accuracy": 0.23206371515989305,
      "num_tokens": 121263709.0,
      "step": 52930
    },
    {
      "entropy": 5.735537004470825,
      "epoch": 5.232799525504152,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002618866197461577,
      "loss": 4.7626,
      "mean_token_accuracy": 0.24475300759077073,
      "num_tokens": 121275218.0,
      "step": 52935
    },
    {
      "entropy": 5.769990348815918,
      "epoch": 5.233293792012653,
      "grad_norm": 1.234375,
      "learning_rate": 0.00026185138991192074,
      "loss": 4.8284,
      "mean_token_accuracy": 0.2373397782444954,
      "num_tokens": 121287236.0,
      "step": 52940
    },
    {
      "entropy": 5.790617370605469,
      "epoch": 5.233788058521155,
      "grad_norm": 1.375,
      "learning_rate": 0.00026181616040114164,
      "loss": 4.8862,
      "mean_token_accuracy": 0.23425393104553222,
      "num_tokens": 121299510.0,
      "step": 52945
    },
    {
      "entropy": 5.762645578384399,
      "epoch": 5.234282325029656,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002617809312146871,
      "loss": 4.8813,
      "mean_token_accuracy": 0.235842165350914,
      "num_tokens": 121311222.0,
      "step": 52950
    },
    {
      "entropy": 5.777142524719238,
      "epoch": 5.234776591538157,
      "grad_norm": 1.25,
      "learning_rate": 0.00026174570235342354,
      "loss": 4.876,
      "mean_token_accuracy": 0.24115837812423707,
      "num_tokens": 121322217.0,
      "step": 52955
    },
    {
      "entropy": 5.795700931549073,
      "epoch": 5.2352708580466585,
      "grad_norm": 1.421875,
      "learning_rate": 0.00026171047381821786,
      "loss": 4.9108,
      "mean_token_accuracy": 0.23465074002742767,
      "num_tokens": 121333386.0,
      "step": 52960
    },
    {
      "entropy": 5.744344186782837,
      "epoch": 5.23576512455516,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002616752456099366,
      "loss": 4.8731,
      "mean_token_accuracy": 0.235241162776947,
      "num_tokens": 121345492.0,
      "step": 52965
    },
    {
      "entropy": 5.773640489578247,
      "epoch": 5.236259391063662,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002616400177294463,
      "loss": 4.8829,
      "mean_token_accuracy": 0.23715141415596008,
      "num_tokens": 121356953.0,
      "step": 52970
    },
    {
      "entropy": 5.796369123458862,
      "epoch": 5.236753657572163,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002616047901776138,
      "loss": 4.9193,
      "mean_token_accuracy": 0.23460109084844588,
      "num_tokens": 121368366.0,
      "step": 52975
    },
    {
      "entropy": 5.869020557403564,
      "epoch": 5.237247924080664,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002615695629553054,
      "loss": 4.9925,
      "mean_token_accuracy": 0.22700978815555573,
      "num_tokens": 121379541.0,
      "step": 52980
    },
    {
      "entropy": 5.828376388549804,
      "epoch": 5.237742190589166,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00026153433606338774,
      "loss": 4.9733,
      "mean_token_accuracy": 0.22856339812278748,
      "num_tokens": 121391745.0,
      "step": 52985
    },
    {
      "entropy": 5.776850461959839,
      "epoch": 5.238236457097667,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00026149910950272763,
      "loss": 4.8976,
      "mean_token_accuracy": 0.23703269213438033,
      "num_tokens": 121402880.0,
      "step": 52990
    },
    {
      "entropy": 5.834416198730469,
      "epoch": 5.238730723606168,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00026146388327419155,
      "loss": 5.0187,
      "mean_token_accuracy": 0.22637251317501067,
      "num_tokens": 121414061.0,
      "step": 52995
    },
    {
      "entropy": 5.790109777450562,
      "epoch": 5.2392249901146695,
      "grad_norm": 1.34375,
      "learning_rate": 0.000261428657378646,
      "loss": 4.9185,
      "mean_token_accuracy": 0.22891247272491455,
      "num_tokens": 121425016.0,
      "step": 53000
    },
    {
      "entropy": 5.843472146987915,
      "epoch": 5.239719256623172,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002613934318169576,
      "loss": 4.9375,
      "mean_token_accuracy": 0.23408869802951812,
      "num_tokens": 121436955.0,
      "step": 53005
    },
    {
      "entropy": 5.789486026763916,
      "epoch": 5.240213523131673,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002613582065899929,
      "loss": 4.8945,
      "mean_token_accuracy": 0.23274163603782655,
      "num_tokens": 121449690.0,
      "step": 53010
    },
    {
      "entropy": 5.826712942123413,
      "epoch": 5.240707789640174,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00026132298169861837,
      "loss": 4.9575,
      "mean_token_accuracy": 0.22878662943840028,
      "num_tokens": 121462009.0,
      "step": 53015
    },
    {
      "entropy": 5.760797786712646,
      "epoch": 5.241202056148675,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002612877571437006,
      "loss": 4.9027,
      "mean_token_accuracy": 0.23074429035186766,
      "num_tokens": 121474416.0,
      "step": 53020
    },
    {
      "entropy": 5.819698143005371,
      "epoch": 5.241696322657177,
      "grad_norm": 1.328125,
      "learning_rate": 0.00026125253292610625,
      "loss": 4.9637,
      "mean_token_accuracy": 0.2304188132286072,
      "num_tokens": 121485990.0,
      "step": 53025
    },
    {
      "entropy": 5.853346252441407,
      "epoch": 5.242190589165678,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002612173090467016,
      "loss": 4.8806,
      "mean_token_accuracy": 0.2361729383468628,
      "num_tokens": 121496924.0,
      "step": 53030
    },
    {
      "entropy": 5.813127851486206,
      "epoch": 5.242684855674179,
      "grad_norm": 1.34375,
      "learning_rate": 0.00026118208550635335,
      "loss": 4.9147,
      "mean_token_accuracy": 0.2351555734872818,
      "num_tokens": 121508601.0,
      "step": 53035
    },
    {
      "entropy": 5.831164073944092,
      "epoch": 5.243179122182681,
      "grad_norm": 1.296875,
      "learning_rate": 0.00026114686230592797,
      "loss": 4.9857,
      "mean_token_accuracy": 0.2275465816259384,
      "num_tokens": 121521004.0,
      "step": 53040
    },
    {
      "entropy": 5.786175918579102,
      "epoch": 5.243673388691183,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002611116394462918,
      "loss": 4.8971,
      "mean_token_accuracy": 0.23269540816545486,
      "num_tokens": 121531660.0,
      "step": 53045
    },
    {
      "entropy": 5.757289838790894,
      "epoch": 5.244167655199684,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002610764169283116,
      "loss": 4.8634,
      "mean_token_accuracy": 0.23809027522802353,
      "num_tokens": 121542614.0,
      "step": 53050
    },
    {
      "entropy": 5.827285194396973,
      "epoch": 5.244661921708185,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002610411947528536,
      "loss": 4.9741,
      "mean_token_accuracy": 0.2335599899291992,
      "num_tokens": 121553428.0,
      "step": 53055
    },
    {
      "entropy": 5.8337095260620115,
      "epoch": 5.245156188216686,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00026100597292078444,
      "loss": 4.982,
      "mean_token_accuracy": 0.2228473022580147,
      "num_tokens": 121565957.0,
      "step": 53060
    },
    {
      "entropy": 5.755543851852417,
      "epoch": 5.245650454725188,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002609707514329705,
      "loss": 4.8157,
      "mean_token_accuracy": 0.238124817609787,
      "num_tokens": 121576229.0,
      "step": 53065
    },
    {
      "entropy": 5.732319593429565,
      "epoch": 5.246144721233689,
      "grad_norm": 1.375,
      "learning_rate": 0.00026093553029027817,
      "loss": 4.9057,
      "mean_token_accuracy": 0.22830894589424133,
      "num_tokens": 121587938.0,
      "step": 53070
    },
    {
      "entropy": 5.719219493865967,
      "epoch": 5.24663898774219,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002609003094935741,
      "loss": 4.8698,
      "mean_token_accuracy": 0.23851091861724855,
      "num_tokens": 121598605.0,
      "step": 53075
    },
    {
      "entropy": 5.762585639953613,
      "epoch": 5.247133254250692,
      "grad_norm": 1.328125,
      "learning_rate": 0.00026086508904372453,
      "loss": 4.8946,
      "mean_token_accuracy": 0.2397138223052025,
      "num_tokens": 121611963.0,
      "step": 53080
    },
    {
      "entropy": 5.8020587921142575,
      "epoch": 5.247627520759194,
      "grad_norm": 1.359375,
      "learning_rate": 0.00026082986894159605,
      "loss": 4.9213,
      "mean_token_accuracy": 0.22964800149202347,
      "num_tokens": 121623510.0,
      "step": 53085
    },
    {
      "entropy": 5.831209802627564,
      "epoch": 5.248121787267695,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00026079464918805494,
      "loss": 4.9406,
      "mean_token_accuracy": 0.22570960372686386,
      "num_tokens": 121635337.0,
      "step": 53090
    },
    {
      "entropy": 5.782417869567871,
      "epoch": 5.248616053776196,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002607594297839676,
      "loss": 4.8709,
      "mean_token_accuracy": 0.23625948131084443,
      "num_tokens": 121645682.0,
      "step": 53095
    },
    {
      "entropy": 5.863605117797851,
      "epoch": 5.249110320284697,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00026072421073020053,
      "loss": 5.0281,
      "mean_token_accuracy": 0.2218097984790802,
      "num_tokens": 121658531.0,
      "step": 53100
    },
    {
      "entropy": 5.761544418334961,
      "epoch": 5.249604586793199,
      "grad_norm": 1.328125,
      "learning_rate": 0.00026068899202762007,
      "loss": 4.9153,
      "mean_token_accuracy": 0.23979781121015548,
      "num_tokens": 121669721.0,
      "step": 53105
    },
    {
      "entropy": 5.830628681182861,
      "epoch": 5.2500988533017,
      "grad_norm": 1.265625,
      "learning_rate": 0.00026065377367709273,
      "loss": 4.9563,
      "mean_token_accuracy": 0.22691141664981843,
      "num_tokens": 121680438.0,
      "step": 53110
    },
    {
      "entropy": 5.867693996429443,
      "epoch": 5.250593119810202,
      "grad_norm": 1.359375,
      "learning_rate": 0.00026061855567948474,
      "loss": 4.9664,
      "mean_token_accuracy": 0.2308885633945465,
      "num_tokens": 121691342.0,
      "step": 53115
    },
    {
      "entropy": 5.760453796386718,
      "epoch": 5.251087386318703,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002605833380356625,
      "loss": 4.8727,
      "mean_token_accuracy": 0.2322571337223053,
      "num_tokens": 121703022.0,
      "step": 53120
    },
    {
      "entropy": 5.781603193283081,
      "epoch": 5.251581652827205,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0002605481207464924,
      "loss": 4.911,
      "mean_token_accuracy": 0.23846331387758254,
      "num_tokens": 121713738.0,
      "step": 53125
    },
    {
      "entropy": 5.758379554748535,
      "epoch": 5.252075919335706,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00026051290381284067,
      "loss": 4.8633,
      "mean_token_accuracy": 0.23346432149410248,
      "num_tokens": 121724640.0,
      "step": 53130
    },
    {
      "entropy": 5.829484653472901,
      "epoch": 5.252570185844207,
      "grad_norm": 1.25,
      "learning_rate": 0.00026047768723557385,
      "loss": 4.9838,
      "mean_token_accuracy": 0.2274206295609474,
      "num_tokens": 121737469.0,
      "step": 53135
    },
    {
      "entropy": 5.839113521575928,
      "epoch": 5.253064452352708,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002604424710155582,
      "loss": 4.9588,
      "mean_token_accuracy": 0.23049239814281464,
      "num_tokens": 121749483.0,
      "step": 53140
    },
    {
      "entropy": 5.859843730926514,
      "epoch": 5.25355871886121,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002604072551536599,
      "loss": 4.9951,
      "mean_token_accuracy": 0.22963541895151138,
      "num_tokens": 121761065.0,
      "step": 53145
    },
    {
      "entropy": 5.82386040687561,
      "epoch": 5.254052985369711,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002603720396507455,
      "loss": 4.9464,
      "mean_token_accuracy": 0.22915802747011185,
      "num_tokens": 121772382.0,
      "step": 53150
    },
    {
      "entropy": 5.841601037979126,
      "epoch": 5.254547251878213,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002603368245076812,
      "loss": 4.9567,
      "mean_token_accuracy": 0.23309761434793472,
      "num_tokens": 121783565.0,
      "step": 53155
    },
    {
      "entropy": 5.80554165840149,
      "epoch": 5.255041518386714,
      "grad_norm": 1.3125,
      "learning_rate": 0.00026030160972533316,
      "loss": 4.949,
      "mean_token_accuracy": 0.22750390321016312,
      "num_tokens": 121795560.0,
      "step": 53160
    },
    {
      "entropy": 5.797992181777954,
      "epoch": 5.255535784895216,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00026026639530456793,
      "loss": 4.9693,
      "mean_token_accuracy": 0.2280057191848755,
      "num_tokens": 121807199.0,
      "step": 53165
    },
    {
      "entropy": 5.785141658782959,
      "epoch": 5.256030051403717,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002602311812462516,
      "loss": 4.9023,
      "mean_token_accuracy": 0.23679499477148055,
      "num_tokens": 121818923.0,
      "step": 53170
    },
    {
      "entropy": 5.797716951370239,
      "epoch": 5.256524317912218,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002601959675512505,
      "loss": 4.9167,
      "mean_token_accuracy": 0.2287073776125908,
      "num_tokens": 121830544.0,
      "step": 53175
    },
    {
      "entropy": 5.821096134185791,
      "epoch": 5.257018584420719,
      "grad_norm": 1.34375,
      "learning_rate": 0.000260160754220431,
      "loss": 4.9215,
      "mean_token_accuracy": 0.22517766505479814,
      "num_tokens": 121842557.0,
      "step": 53180
    },
    {
      "entropy": 5.781222295761109,
      "epoch": 5.257512850929221,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002601255412546592,
      "loss": 4.8883,
      "mean_token_accuracy": 0.23780398964881896,
      "num_tokens": 121855207.0,
      "step": 53185
    },
    {
      "entropy": 5.798046636581421,
      "epoch": 5.258007117437723,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002600903286548013,
      "loss": 4.998,
      "mean_token_accuracy": 0.2235134482383728,
      "num_tokens": 121867516.0,
      "step": 53190
    },
    {
      "entropy": 5.799411725997925,
      "epoch": 5.258501383946224,
      "grad_norm": 1.375,
      "learning_rate": 0.00026005511642172364,
      "loss": 4.887,
      "mean_token_accuracy": 0.2334684267640114,
      "num_tokens": 121879036.0,
      "step": 53195
    },
    {
      "entropy": 5.827278423309326,
      "epoch": 5.258995650454725,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002600199045562925,
      "loss": 4.9692,
      "mean_token_accuracy": 0.22800715267658234,
      "num_tokens": 121891575.0,
      "step": 53200
    },
    {
      "entropy": 5.860282135009766,
      "epoch": 5.2594899169632265,
      "grad_norm": 1.3125,
      "learning_rate": 0.00025998469305937405,
      "loss": 4.9424,
      "mean_token_accuracy": 0.22373806834220886,
      "num_tokens": 121902665.0,
      "step": 53205
    },
    {
      "entropy": 5.757001352310181,
      "epoch": 5.259984183471728,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002599494819318345,
      "loss": 4.8451,
      "mean_token_accuracy": 0.24240019917488098,
      "num_tokens": 121914045.0,
      "step": 53210
    },
    {
      "entropy": 5.690659809112549,
      "epoch": 5.260478449980229,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002599142711745399,
      "loss": 4.7675,
      "mean_token_accuracy": 0.24499010145664216,
      "num_tokens": 121925183.0,
      "step": 53215
    },
    {
      "entropy": 5.790027523040772,
      "epoch": 5.26097271648873,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002598790607883566,
      "loss": 4.9531,
      "mean_token_accuracy": 0.2322533279657364,
      "num_tokens": 121936742.0,
      "step": 53220
    },
    {
      "entropy": 5.768565988540649,
      "epoch": 5.2614669829972325,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002598438507741508,
      "loss": 4.8616,
      "mean_token_accuracy": 0.2418904572725296,
      "num_tokens": 121948832.0,
      "step": 53225
    },
    {
      "entropy": 5.8038280487060545,
      "epoch": 5.261961249505734,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00025980864113278867,
      "loss": 4.9412,
      "mean_token_accuracy": 0.23113111555576324,
      "num_tokens": 121959603.0,
      "step": 53230
    },
    {
      "entropy": 5.766725635528564,
      "epoch": 5.262455516014235,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002597734318651363,
      "loss": 4.9051,
      "mean_token_accuracy": 0.2330106481909752,
      "num_tokens": 121970646.0,
      "step": 53235
    },
    {
      "entropy": 5.746005964279175,
      "epoch": 5.262949782522736,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002597382229720598,
      "loss": 4.8098,
      "mean_token_accuracy": 0.24316336959600449,
      "num_tokens": 121981704.0,
      "step": 53240
    },
    {
      "entropy": 5.790102195739746,
      "epoch": 5.2634440490312375,
      "grad_norm": 1.328125,
      "learning_rate": 0.00025970301445442536,
      "loss": 4.9281,
      "mean_token_accuracy": 0.23448754996061325,
      "num_tokens": 121993328.0,
      "step": 53245
    },
    {
      "entropy": 5.8121058464050295,
      "epoch": 5.263938315539739,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00025966780631309923,
      "loss": 4.9099,
      "mean_token_accuracy": 0.23548202961683273,
      "num_tokens": 122003672.0,
      "step": 53250
    },
    {
      "entropy": 5.808974027633667,
      "epoch": 5.26443258204824,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002596325985489474,
      "loss": 4.9318,
      "mean_token_accuracy": 0.23892140686511992,
      "num_tokens": 122013690.0,
      "step": 53255
    },
    {
      "entropy": 5.785220623016357,
      "epoch": 5.264926848556742,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002595973911628361,
      "loss": 4.8634,
      "mean_token_accuracy": 0.24277035444974898,
      "num_tokens": 122026636.0,
      "step": 53260
    },
    {
      "entropy": 5.826932096481324,
      "epoch": 5.2654211150652435,
      "grad_norm": 1.390625,
      "learning_rate": 0.00025956218415563125,
      "loss": 4.9433,
      "mean_token_accuracy": 0.22814250439405442,
      "num_tokens": 122037543.0,
      "step": 53265
    },
    {
      "entropy": 5.845264530181884,
      "epoch": 5.265915381573745,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002595269775281992,
      "loss": 4.9378,
      "mean_token_accuracy": 0.22860383838415146,
      "num_tokens": 122048833.0,
      "step": 53270
    },
    {
      "entropy": 5.8074116706848145,
      "epoch": 5.266409648082246,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002594917712814057,
      "loss": 4.9373,
      "mean_token_accuracy": 0.23358654975891113,
      "num_tokens": 122060294.0,
      "step": 53275
    },
    {
      "entropy": 5.715169143676758,
      "epoch": 5.266903914590747,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002594565654161172,
      "loss": 4.8348,
      "mean_token_accuracy": 0.23334084451198578,
      "num_tokens": 122071551.0,
      "step": 53280
    },
    {
      "entropy": 5.75650429725647,
      "epoch": 5.2673981810992485,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002594213599331996,
      "loss": 4.8564,
      "mean_token_accuracy": 0.2303932264447212,
      "num_tokens": 122082331.0,
      "step": 53285
    },
    {
      "entropy": 5.801055288314819,
      "epoch": 5.26789244760775,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002593861548335189,
      "loss": 4.9263,
      "mean_token_accuracy": 0.23169513344764708,
      "num_tokens": 122094303.0,
      "step": 53290
    },
    {
      "entropy": 5.772427415847778,
      "epoch": 5.268386714116252,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002593509501179413,
      "loss": 4.8573,
      "mean_token_accuracy": 0.2410147726535797,
      "num_tokens": 122104435.0,
      "step": 53295
    },
    {
      "entropy": 5.853130292892456,
      "epoch": 5.268880980624753,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002593157457873327,
      "loss": 4.9439,
      "mean_token_accuracy": 0.22760724425315856,
      "num_tokens": 122115623.0,
      "step": 53300
    },
    {
      "entropy": 5.847374820709229,
      "epoch": 5.2693752471332544,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002592805418425591,
      "loss": 4.9652,
      "mean_token_accuracy": 0.22753052562475204,
      "num_tokens": 122127128.0,
      "step": 53305
    },
    {
      "entropy": 5.831954145431519,
      "epoch": 5.269869513641756,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002592453382844867,
      "loss": 4.9669,
      "mean_token_accuracy": 0.2299848198890686,
      "num_tokens": 122138305.0,
      "step": 53310
    },
    {
      "entropy": 5.775985240936279,
      "epoch": 5.270363780150257,
      "grad_norm": 1.421875,
      "learning_rate": 0.00025921013511398137,
      "loss": 4.8831,
      "mean_token_accuracy": 0.23105882853269577,
      "num_tokens": 122149092.0,
      "step": 53315
    },
    {
      "entropy": 5.763343620300293,
      "epoch": 5.270858046658758,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00025917493233190925,
      "loss": 4.865,
      "mean_token_accuracy": 0.2370038241147995,
      "num_tokens": 122160152.0,
      "step": 53320
    },
    {
      "entropy": 5.794892883300781,
      "epoch": 5.2713523131672595,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00025913972993913623,
      "loss": 4.8944,
      "mean_token_accuracy": 0.23668826669454573,
      "num_tokens": 122170889.0,
      "step": 53325
    },
    {
      "entropy": 5.861569118499756,
      "epoch": 5.271846579675761,
      "grad_norm": 1.578125,
      "learning_rate": 0.0002591045279365283,
      "loss": 5.0081,
      "mean_token_accuracy": 0.22176806330680848,
      "num_tokens": 122182881.0,
      "step": 53330
    },
    {
      "entropy": 5.7466450214385985,
      "epoch": 5.272340846184263,
      "grad_norm": 1.40625,
      "learning_rate": 0.00025906932632495145,
      "loss": 4.8254,
      "mean_token_accuracy": 0.24544476419687272,
      "num_tokens": 122194258.0,
      "step": 53335
    },
    {
      "entropy": 5.748769855499267,
      "epoch": 5.272835112692764,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002590341251052716,
      "loss": 4.8769,
      "mean_token_accuracy": 0.2347778007388115,
      "num_tokens": 122205678.0,
      "step": 53340
    },
    {
      "entropy": 5.753863763809204,
      "epoch": 5.273329379201265,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00025899892427835473,
      "loss": 4.9126,
      "mean_token_accuracy": 0.2355031579732895,
      "num_tokens": 122217446.0,
      "step": 53345
    },
    {
      "entropy": 5.7574296474456785,
      "epoch": 5.273823645709767,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00025896372384506697,
      "loss": 4.9009,
      "mean_token_accuracy": 0.23715901076793672,
      "num_tokens": 122228466.0,
      "step": 53350
    },
    {
      "entropy": 5.775121212005615,
      "epoch": 5.274317912218268,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002589285238062739,
      "loss": 4.9433,
      "mean_token_accuracy": 0.23403626978397368,
      "num_tokens": 122240528.0,
      "step": 53355
    },
    {
      "entropy": 5.884656620025635,
      "epoch": 5.274812178726769,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00025889332416284175,
      "loss": 4.9678,
      "mean_token_accuracy": 0.23180302530527114,
      "num_tokens": 122251267.0,
      "step": 53360
    },
    {
      "entropy": 5.90588526725769,
      "epoch": 5.2753064452352705,
      "grad_norm": 1.28125,
      "learning_rate": 0.00025885812491563626,
      "loss": 4.986,
      "mean_token_accuracy": 0.22455986589193344,
      "num_tokens": 122262999.0,
      "step": 53365
    },
    {
      "entropy": 5.867618465423584,
      "epoch": 5.275800711743773,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00025882292606552344,
      "loss": 4.9825,
      "mean_token_accuracy": 0.22695672810077666,
      "num_tokens": 122275216.0,
      "step": 53370
    },
    {
      "entropy": 5.721565675735474,
      "epoch": 5.276294978252274,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00025878772761336915,
      "loss": 4.7812,
      "mean_token_accuracy": 0.24817583411931993,
      "num_tokens": 122285385.0,
      "step": 53375
    },
    {
      "entropy": 5.7855224609375,
      "epoch": 5.276789244760775,
      "grad_norm": 1.421875,
      "learning_rate": 0.00025875252956003927,
      "loss": 4.8958,
      "mean_token_accuracy": 0.23777510076761246,
      "num_tokens": 122296386.0,
      "step": 53380
    },
    {
      "entropy": 5.767304944992065,
      "epoch": 5.277283511269276,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00025871733190639966,
      "loss": 4.8624,
      "mean_token_accuracy": 0.23948730677366256,
      "num_tokens": 122307339.0,
      "step": 53385
    },
    {
      "entropy": 5.7609882831573485,
      "epoch": 5.277777777777778,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0002586821346533162,
      "loss": 4.8475,
      "mean_token_accuracy": 0.24567288160324097,
      "num_tokens": 122318712.0,
      "step": 53390
    },
    {
      "entropy": 5.754209756851196,
      "epoch": 5.278272044286279,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00025864693780165476,
      "loss": 4.8606,
      "mean_token_accuracy": 0.23581949323415757,
      "num_tokens": 122329890.0,
      "step": 53395
    },
    {
      "entropy": 5.760194206237793,
      "epoch": 5.27876631079478,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002586117413522813,
      "loss": 4.8721,
      "mean_token_accuracy": 0.2352184236049652,
      "num_tokens": 122340582.0,
      "step": 53400
    },
    {
      "entropy": 5.82879490852356,
      "epoch": 5.2792605773032815,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002585765453060615,
      "loss": 4.9769,
      "mean_token_accuracy": 0.2244429975748062,
      "num_tokens": 122353661.0,
      "step": 53405
    },
    {
      "entropy": 5.869246006011963,
      "epoch": 5.279754843811784,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002585413496638611,
      "loss": 5.0376,
      "mean_token_accuracy": 0.2292223945260048,
      "num_tokens": 122365223.0,
      "step": 53410
    },
    {
      "entropy": 5.822266530990601,
      "epoch": 5.280249110320285,
      "grad_norm": 1.328125,
      "learning_rate": 0.00025850615442654624,
      "loss": 4.8944,
      "mean_token_accuracy": 0.24017884880304335,
      "num_tokens": 122376148.0,
      "step": 53415
    },
    {
      "entropy": 5.757019710540772,
      "epoch": 5.280743376828786,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00025847095959498237,
      "loss": 4.9525,
      "mean_token_accuracy": 0.2271917462348938,
      "num_tokens": 122388108.0,
      "step": 53420
    },
    {
      "entropy": 5.781629943847657,
      "epoch": 5.281237643337287,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002584357651700355,
      "loss": 4.8961,
      "mean_token_accuracy": 0.23520078510046005,
      "num_tokens": 122400999.0,
      "step": 53425
    },
    {
      "entropy": 5.706395101547241,
      "epoch": 5.281731909845789,
      "grad_norm": 1.25,
      "learning_rate": 0.00025840057115257147,
      "loss": 4.8346,
      "mean_token_accuracy": 0.24182889461517335,
      "num_tokens": 122412398.0,
      "step": 53430
    },
    {
      "entropy": 5.844756841659546,
      "epoch": 5.28222617635429,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002583653775434559,
      "loss": 4.9489,
      "mean_token_accuracy": 0.23105632960796357,
      "num_tokens": 122423877.0,
      "step": 53435
    },
    {
      "entropy": 5.770450496673584,
      "epoch": 5.282720442862791,
      "grad_norm": 1.34375,
      "learning_rate": 0.00025833018434355466,
      "loss": 4.8644,
      "mean_token_accuracy": 0.2328769788146019,
      "num_tokens": 122435030.0,
      "step": 53440
    },
    {
      "entropy": 5.825913047790527,
      "epoch": 5.283214709371293,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00025829499155373344,
      "loss": 4.9869,
      "mean_token_accuracy": 0.22283226549625396,
      "num_tokens": 122446550.0,
      "step": 53445
    },
    {
      "entropy": 5.825781917572021,
      "epoch": 5.283708975879795,
      "grad_norm": 1.328125,
      "learning_rate": 0.00025825979917485796,
      "loss": 4.9395,
      "mean_token_accuracy": 0.22569505572319032,
      "num_tokens": 122457662.0,
      "step": 53450
    },
    {
      "entropy": 5.760116815567017,
      "epoch": 5.284203242388296,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002582246072077941,
      "loss": 4.9363,
      "mean_token_accuracy": 0.23199324309825897,
      "num_tokens": 122469514.0,
      "step": 53455
    },
    {
      "entropy": 5.796312379837036,
      "epoch": 5.284697508896797,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002581894156534075,
      "loss": 4.948,
      "mean_token_accuracy": 0.22823249846696853,
      "num_tokens": 122480795.0,
      "step": 53460
    },
    {
      "entropy": 5.761532735824585,
      "epoch": 5.285191775405298,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002581542245125639,
      "loss": 4.8367,
      "mean_token_accuracy": 0.24709578305482865,
      "num_tokens": 122492018.0,
      "step": 53465
    },
    {
      "entropy": 5.714620685577392,
      "epoch": 5.2856860419138,
      "grad_norm": 1.4375,
      "learning_rate": 0.00025811903378612896,
      "loss": 4.8335,
      "mean_token_accuracy": 0.23537764996290206,
      "num_tokens": 122503551.0,
      "step": 53470
    },
    {
      "entropy": 5.782076072692871,
      "epoch": 5.286180308422301,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002580838434749684,
      "loss": 4.8626,
      "mean_token_accuracy": 0.23754838854074478,
      "num_tokens": 122515145.0,
      "step": 53475
    },
    {
      "entropy": 5.814262056350708,
      "epoch": 5.286674574930803,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00025804865357994785,
      "loss": 4.9514,
      "mean_token_accuracy": 0.23070691972970964,
      "num_tokens": 122527399.0,
      "step": 53480
    },
    {
      "entropy": 5.7863842964172365,
      "epoch": 5.287168841439304,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002580134641019331,
      "loss": 4.8143,
      "mean_token_accuracy": 0.2445555180311203,
      "num_tokens": 122537608.0,
      "step": 53485
    },
    {
      "entropy": 5.796369457244873,
      "epoch": 5.287663107947806,
      "grad_norm": 1.375,
      "learning_rate": 0.0002579782750417898,
      "loss": 4.9158,
      "mean_token_accuracy": 0.23440824449062347,
      "num_tokens": 122548289.0,
      "step": 53490
    },
    {
      "entropy": 5.8116865158081055,
      "epoch": 5.288157374456307,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002579430864003836,
      "loss": 4.9958,
      "mean_token_accuracy": 0.23201971352100373,
      "num_tokens": 122560787.0,
      "step": 53495
    },
    {
      "entropy": 5.82585654258728,
      "epoch": 5.288651640964808,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00025790789817858005,
      "loss": 4.9306,
      "mean_token_accuracy": 0.2301764592528343,
      "num_tokens": 122572794.0,
      "step": 53500
    },
    {
      "entropy": 5.798297595977783,
      "epoch": 5.289145907473309,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00025787271037724496,
      "loss": 4.9091,
      "mean_token_accuracy": 0.24008553773164748,
      "num_tokens": 122586200.0,
      "step": 53505
    },
    {
      "entropy": 5.862014532089233,
      "epoch": 5.289640173981811,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00025783752299724373,
      "loss": 5.0593,
      "mean_token_accuracy": 0.22071299105882644,
      "num_tokens": 122597392.0,
      "step": 53510
    },
    {
      "entropy": 5.764507865905761,
      "epoch": 5.290134440490313,
      "grad_norm": 1.46875,
      "learning_rate": 0.00025780233603944216,
      "loss": 4.8932,
      "mean_token_accuracy": 0.2311556190252304,
      "num_tokens": 122608907.0,
      "step": 53515
    },
    {
      "entropy": 5.778825139999389,
      "epoch": 5.290628706998814,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002577671495047058,
      "loss": 4.8015,
      "mean_token_accuracy": 0.25134068429470063,
      "num_tokens": 122619792.0,
      "step": 53520
    },
    {
      "entropy": 5.800314950942993,
      "epoch": 5.291122973507315,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00025773196339390017,
      "loss": 4.8668,
      "mean_token_accuracy": 0.23419223129749298,
      "num_tokens": 122630113.0,
      "step": 53525
    },
    {
      "entropy": 5.775997591018677,
      "epoch": 5.291617240015817,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000257696777707891,
      "loss": 4.9743,
      "mean_token_accuracy": 0.22858325093984605,
      "num_tokens": 122642045.0,
      "step": 53530
    },
    {
      "entropy": 5.696418190002442,
      "epoch": 5.292111506524318,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00025766159244754367,
      "loss": 4.8024,
      "mean_token_accuracy": 0.2495665580034256,
      "num_tokens": 122654148.0,
      "step": 53535
    },
    {
      "entropy": 5.8170921325683596,
      "epoch": 5.292605773032819,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00025762640761372397,
      "loss": 4.9169,
      "mean_token_accuracy": 0.23545741885900498,
      "num_tokens": 122665012.0,
      "step": 53540
    },
    {
      "entropy": 5.778345823287964,
      "epoch": 5.29310003954132,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002575912232072973,
      "loss": 4.8748,
      "mean_token_accuracy": 0.23926792442798614,
      "num_tokens": 122677009.0,
      "step": 53545
    },
    {
      "entropy": 5.793969631195068,
      "epoch": 5.2935943060498225,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00025755603922912926,
      "loss": 4.8931,
      "mean_token_accuracy": 0.23192045390605925,
      "num_tokens": 122688828.0,
      "step": 53550
    },
    {
      "entropy": 5.754349279403686,
      "epoch": 5.294088572558324,
      "grad_norm": 1.5,
      "learning_rate": 0.00025752085568008534,
      "loss": 4.9133,
      "mean_token_accuracy": 0.23478081524372102,
      "num_tokens": 122699408.0,
      "step": 53555
    },
    {
      "entropy": 5.834153890609741,
      "epoch": 5.294582839066825,
      "grad_norm": 1.25,
      "learning_rate": 0.0002574856725610311,
      "loss": 4.9163,
      "mean_token_accuracy": 0.23136269450187683,
      "num_tokens": 122711885.0,
      "step": 53560
    },
    {
      "entropy": 5.761081171035767,
      "epoch": 5.295077105575326,
      "grad_norm": 1.296875,
      "learning_rate": 0.00025745048987283197,
      "loss": 4.7701,
      "mean_token_accuracy": 0.24466915875673295,
      "num_tokens": 122722805.0,
      "step": 53565
    },
    {
      "entropy": 5.7405171394348145,
      "epoch": 5.2955713720838276,
      "grad_norm": 1.40625,
      "learning_rate": 0.00025741530761635355,
      "loss": 4.8005,
      "mean_token_accuracy": 0.24088837057352067,
      "num_tokens": 122734595.0,
      "step": 53570
    },
    {
      "entropy": 5.726275396347046,
      "epoch": 5.296065638592329,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002573801257924613,
      "loss": 4.8843,
      "mean_token_accuracy": 0.23586251586675644,
      "num_tokens": 122747299.0,
      "step": 53575
    },
    {
      "entropy": 5.787537717819214,
      "epoch": 5.29655990510083,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00025734494440202073,
      "loss": 4.9464,
      "mean_token_accuracy": 0.22741032987833024,
      "num_tokens": 122759504.0,
      "step": 53580
    },
    {
      "entropy": 5.798646688461304,
      "epoch": 5.297054171609331,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002573097634458972,
      "loss": 4.9107,
      "mean_token_accuracy": 0.23337824791669845,
      "num_tokens": 122772220.0,
      "step": 53585
    },
    {
      "entropy": 5.744379281997681,
      "epoch": 5.2975484381178335,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002572745829249563,
      "loss": 4.8376,
      "mean_token_accuracy": 0.24275604635477066,
      "num_tokens": 122784585.0,
      "step": 53590
    },
    {
      "entropy": 5.854198837280274,
      "epoch": 5.298042704626335,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002572394028400633,
      "loss": 4.9364,
      "mean_token_accuracy": 0.2304521232843399,
      "num_tokens": 122795465.0,
      "step": 53595
    },
    {
      "entropy": 5.6753538131713865,
      "epoch": 5.298536971134836,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002572042231920838,
      "loss": 4.8524,
      "mean_token_accuracy": 0.2389932841062546,
      "num_tokens": 122807129.0,
      "step": 53600
    },
    {
      "entropy": 5.768017053604126,
      "epoch": 5.299031237643337,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00025716904398188326,
      "loss": 4.8854,
      "mean_token_accuracy": 0.22844571471214295,
      "num_tokens": 122818201.0,
      "step": 53605
    },
    {
      "entropy": 5.831211948394776,
      "epoch": 5.2995255041518385,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00025713386521032693,
      "loss": 4.912,
      "mean_token_accuracy": 0.23061102777719497,
      "num_tokens": 122828540.0,
      "step": 53610
    },
    {
      "entropy": 5.805179119110107,
      "epoch": 5.30001977066034,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002570986868782803,
      "loss": 4.9248,
      "mean_token_accuracy": 0.235867702960968,
      "num_tokens": 122840829.0,
      "step": 53615
    },
    {
      "entropy": 5.758443784713745,
      "epoch": 5.300514037168841,
      "grad_norm": 1.296875,
      "learning_rate": 0.00025706350898660875,
      "loss": 4.909,
      "mean_token_accuracy": 0.23470396250486375,
      "num_tokens": 122853131.0,
      "step": 53620
    },
    {
      "entropy": 5.809024667739868,
      "epoch": 5.301008303677343,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00025702833153617755,
      "loss": 4.9268,
      "mean_token_accuracy": 0.22747260481119155,
      "num_tokens": 122865062.0,
      "step": 53625
    },
    {
      "entropy": 5.77620038986206,
      "epoch": 5.3015025701858445,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0002569931545278523,
      "loss": 4.9044,
      "mean_token_accuracy": 0.2313554272055626,
      "num_tokens": 122875696.0,
      "step": 53630
    },
    {
      "entropy": 5.800540018081665,
      "epoch": 5.301996836694346,
      "grad_norm": 1.234375,
      "learning_rate": 0.00025695797796249824,
      "loss": 4.933,
      "mean_token_accuracy": 0.22957686483860015,
      "num_tokens": 122888383.0,
      "step": 53635
    },
    {
      "entropy": 5.803252220153809,
      "epoch": 5.302491103202847,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0002569228018409807,
      "loss": 4.9011,
      "mean_token_accuracy": 0.2318322092294693,
      "num_tokens": 122899345.0,
      "step": 53640
    },
    {
      "entropy": 5.856547021865845,
      "epoch": 5.302985369711348,
      "grad_norm": 1.328125,
      "learning_rate": 0.00025688762616416503,
      "loss": 5.0513,
      "mean_token_accuracy": 0.22793136537075043,
      "num_tokens": 122911104.0,
      "step": 53645
    },
    {
      "entropy": 5.808743715286255,
      "epoch": 5.3034796362198495,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00025685245093291667,
      "loss": 4.878,
      "mean_token_accuracy": 0.23417021483182907,
      "num_tokens": 122921844.0,
      "step": 53650
    },
    {
      "entropy": 5.779083108901977,
      "epoch": 5.303973902728351,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00025681727614810065,
      "loss": 4.8968,
      "mean_token_accuracy": 0.23549798727035523,
      "num_tokens": 122932774.0,
      "step": 53655
    },
    {
      "entropy": 5.762442970275879,
      "epoch": 5.304468169236852,
      "grad_norm": 1.421875,
      "learning_rate": 0.00025678210181058253,
      "loss": 4.9152,
      "mean_token_accuracy": 0.2330689698457718,
      "num_tokens": 122944565.0,
      "step": 53660
    },
    {
      "entropy": 5.723546838760376,
      "epoch": 5.304962435745354,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002567469279212276,
      "loss": 4.7803,
      "mean_token_accuracy": 0.24965546727180482,
      "num_tokens": 122955989.0,
      "step": 53665
    },
    {
      "entropy": 5.807387447357177,
      "epoch": 5.3054567022538555,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000256711754480901,
      "loss": 4.8836,
      "mean_token_accuracy": 0.23474696427583694,
      "num_tokens": 122967862.0,
      "step": 53670
    },
    {
      "entropy": 5.815908718109131,
      "epoch": 5.305950968762357,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00025667658149046814,
      "loss": 4.9174,
      "mean_token_accuracy": 0.22924703657627105,
      "num_tokens": 122978832.0,
      "step": 53675
    },
    {
      "entropy": 5.7551641941070555,
      "epoch": 5.306445235270858,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002566414089507942,
      "loss": 4.8151,
      "mean_token_accuracy": 0.24581925868988036,
      "num_tokens": 122991342.0,
      "step": 53680
    },
    {
      "entropy": 5.717195081710815,
      "epoch": 5.306939501779359,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002566062368627443,
      "loss": 4.8566,
      "mean_token_accuracy": 0.23960224837064742,
      "num_tokens": 123003377.0,
      "step": 53685
    },
    {
      "entropy": 5.743018436431885,
      "epoch": 5.3074337682878605,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000256571065227184,
      "loss": 4.7898,
      "mean_token_accuracy": 0.24273349791765214,
      "num_tokens": 123013797.0,
      "step": 53690
    },
    {
      "entropy": 5.791771793365479,
      "epoch": 5.307928034796362,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002565358940449783,
      "loss": 4.8808,
      "mean_token_accuracy": 0.23681527972221375,
      "num_tokens": 123023858.0,
      "step": 53695
    },
    {
      "entropy": 5.780860710144043,
      "epoch": 5.308422301304864,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002565007233169925,
      "loss": 4.9415,
      "mean_token_accuracy": 0.23055520206689833,
      "num_tokens": 123034410.0,
      "step": 53700
    },
    {
      "entropy": 5.7003649234771725,
      "epoch": 5.308916567813365,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002564655530440918,
      "loss": 4.8356,
      "mean_token_accuracy": 0.2374408334493637,
      "num_tokens": 123045422.0,
      "step": 53705
    },
    {
      "entropy": 5.832486820220947,
      "epoch": 5.3094108343218664,
      "grad_norm": 1.453125,
      "learning_rate": 0.00025643038322714125,
      "loss": 4.9442,
      "mean_token_accuracy": 0.23420112431049347,
      "num_tokens": 123056549.0,
      "step": 53710
    },
    {
      "entropy": 5.838386726379395,
      "epoch": 5.309905100830368,
      "grad_norm": 1.34375,
      "learning_rate": 0.00025639521386700627,
      "loss": 4.9536,
      "mean_token_accuracy": 0.23329526782035828,
      "num_tokens": 123067983.0,
      "step": 53715
    },
    {
      "entropy": 5.812074899673462,
      "epoch": 5.310399367338869,
      "grad_norm": 1.25,
      "learning_rate": 0.00025636004496455185,
      "loss": 4.9561,
      "mean_token_accuracy": 0.23123320192098618,
      "num_tokens": 123080332.0,
      "step": 53720
    },
    {
      "entropy": 5.860667943954468,
      "epoch": 5.31089363384737,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002563248765206432,
      "loss": 4.984,
      "mean_token_accuracy": 0.23098001927137374,
      "num_tokens": 123092905.0,
      "step": 53725
    },
    {
      "entropy": 5.841025161743164,
      "epoch": 5.3113879003558715,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00025628970853614565,
      "loss": 4.9124,
      "mean_token_accuracy": 0.22909849435091018,
      "num_tokens": 123105042.0,
      "step": 53730
    },
    {
      "entropy": 5.777275705337525,
      "epoch": 5.311882166864374,
      "grad_norm": 1.375,
      "learning_rate": 0.000256254541011924,
      "loss": 4.8971,
      "mean_token_accuracy": 0.2328811302781105,
      "num_tokens": 123115986.0,
      "step": 53735
    },
    {
      "entropy": 5.878675889968872,
      "epoch": 5.312376433372875,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002562193739488436,
      "loss": 5.0443,
      "mean_token_accuracy": 0.22539665848016738,
      "num_tokens": 123127453.0,
      "step": 53740
    },
    {
      "entropy": 5.849966335296631,
      "epoch": 5.312870699881376,
      "grad_norm": 1.453125,
      "learning_rate": 0.00025618420734776945,
      "loss": 5.0009,
      "mean_token_accuracy": 0.23008051216602327,
      "num_tokens": 123138806.0,
      "step": 53745
    },
    {
      "entropy": 5.828011798858642,
      "epoch": 5.313364966389877,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002561490412095668,
      "loss": 4.9009,
      "mean_token_accuracy": 0.24043914079666137,
      "num_tokens": 123149498.0,
      "step": 53750
    },
    {
      "entropy": 5.780609226226806,
      "epoch": 5.313859232898379,
      "grad_norm": 1.421875,
      "learning_rate": 0.00025611387553510067,
      "loss": 4.8998,
      "mean_token_accuracy": 0.2309871271252632,
      "num_tokens": 123160669.0,
      "step": 53755
    },
    {
      "entropy": 5.754515504837036,
      "epoch": 5.31435349940688,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002560787103252361,
      "loss": 4.8475,
      "mean_token_accuracy": 0.23811665624380113,
      "num_tokens": 123171330.0,
      "step": 53760
    },
    {
      "entropy": 5.784481143951416,
      "epoch": 5.314847765915381,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002560435455808382,
      "loss": 4.9315,
      "mean_token_accuracy": 0.2334736093878746,
      "num_tokens": 123183449.0,
      "step": 53765
    },
    {
      "entropy": 5.767261171340943,
      "epoch": 5.315342032423883,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002560083813027719,
      "loss": 4.9035,
      "mean_token_accuracy": 0.24158943593502044,
      "num_tokens": 123194904.0,
      "step": 53770
    },
    {
      "entropy": 5.806327486038208,
      "epoch": 5.315836298932385,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002559732174919026,
      "loss": 4.8445,
      "mean_token_accuracy": 0.23293237388134003,
      "num_tokens": 123206240.0,
      "step": 53775
    },
    {
      "entropy": 5.868523406982422,
      "epoch": 5.316330565440886,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002559380541490949,
      "loss": 4.8723,
      "mean_token_accuracy": 0.23655816316604614,
      "num_tokens": 123217277.0,
      "step": 53780
    },
    {
      "entropy": 5.712953376770019,
      "epoch": 5.316824831949387,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00025590289127521413,
      "loss": 4.8075,
      "mean_token_accuracy": 0.24541327208280564,
      "num_tokens": 123228074.0,
      "step": 53785
    },
    {
      "entropy": 5.720205068588257,
      "epoch": 5.317319098457888,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002558677288711252,
      "loss": 4.8537,
      "mean_token_accuracy": 0.23487648218870164,
      "num_tokens": 123238030.0,
      "step": 53790
    },
    {
      "entropy": 5.757998657226563,
      "epoch": 5.31781336496639,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002558325669376931,
      "loss": 4.915,
      "mean_token_accuracy": 0.234829343855381,
      "num_tokens": 123248716.0,
      "step": 53795
    },
    {
      "entropy": 5.739530372619629,
      "epoch": 5.318307631474891,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002557974054757827,
      "loss": 4.8314,
      "mean_token_accuracy": 0.24374352991580964,
      "num_tokens": 123260235.0,
      "step": 53800
    },
    {
      "entropy": 5.797208547592163,
      "epoch": 5.318801897983393,
      "grad_norm": 1.5,
      "learning_rate": 0.0002557622444862593,
      "loss": 4.8867,
      "mean_token_accuracy": 0.2449814960360527,
      "num_tokens": 123272214.0,
      "step": 53805
    },
    {
      "entropy": 5.790957403182984,
      "epoch": 5.319296164491894,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00025572708396998757,
      "loss": 4.9089,
      "mean_token_accuracy": 0.2318798616528511,
      "num_tokens": 123283644.0,
      "step": 53810
    },
    {
      "entropy": 5.726669406890869,
      "epoch": 5.319790431000396,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00025569192392783257,
      "loss": 4.8332,
      "mean_token_accuracy": 0.23835319727659227,
      "num_tokens": 123295487.0,
      "step": 53815
    },
    {
      "entropy": 5.783889675140381,
      "epoch": 5.320284697508897,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002556567643606593,
      "loss": 4.8998,
      "mean_token_accuracy": 0.23681072890758514,
      "num_tokens": 123306912.0,
      "step": 53820
    },
    {
      "entropy": 5.824093341827393,
      "epoch": 5.320778964017398,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002556216052693325,
      "loss": 4.8826,
      "mean_token_accuracy": 0.2317727565765381,
      "num_tokens": 123318875.0,
      "step": 53825
    },
    {
      "entropy": 5.796582412719727,
      "epoch": 5.321273230525899,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002555864466547173,
      "loss": 4.9275,
      "mean_token_accuracy": 0.23717094212770462,
      "num_tokens": 123330877.0,
      "step": 53830
    },
    {
      "entropy": 5.703095388412476,
      "epoch": 5.321767497034401,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00025555128851767846,
      "loss": 4.7724,
      "mean_token_accuracy": 0.2533256858587265,
      "num_tokens": 123341720.0,
      "step": 53835
    },
    {
      "entropy": 5.771019411087036,
      "epoch": 5.322261763542902,
      "grad_norm": 1.265625,
      "learning_rate": 0.000255516130859081,
      "loss": 4.8932,
      "mean_token_accuracy": 0.2376276135444641,
      "num_tokens": 123353160.0,
      "step": 53840
    },
    {
      "entropy": 5.759971714019775,
      "epoch": 5.322756030051404,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002554809736797898,
      "loss": 4.8859,
      "mean_token_accuracy": 0.2334063395857811,
      "num_tokens": 123363726.0,
      "step": 53845
    },
    {
      "entropy": 5.832391595840454,
      "epoch": 5.323250296559905,
      "grad_norm": 1.375,
      "learning_rate": 0.00025544581698066967,
      "loss": 4.9205,
      "mean_token_accuracy": 0.23166232258081437,
      "num_tokens": 123375152.0,
      "step": 53850
    },
    {
      "entropy": 5.780570983886719,
      "epoch": 5.323744563068407,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00025541066076258535,
      "loss": 4.8816,
      "mean_token_accuracy": 0.23595014810562134,
      "num_tokens": 123386266.0,
      "step": 53855
    },
    {
      "entropy": 5.778565835952759,
      "epoch": 5.324238829576908,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00025537550502640184,
      "loss": 4.8379,
      "mean_token_accuracy": 0.2453214555978775,
      "num_tokens": 123396423.0,
      "step": 53860
    },
    {
      "entropy": 5.813871002197265,
      "epoch": 5.324733096085409,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00025534034977298396,
      "loss": 4.9422,
      "mean_token_accuracy": 0.23119716495275497,
      "num_tokens": 123407907.0,
      "step": 53865
    },
    {
      "entropy": 5.8000068187713625,
      "epoch": 5.32522736259391,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002553051950031966,
      "loss": 4.948,
      "mean_token_accuracy": 0.22675623297691344,
      "num_tokens": 123419968.0,
      "step": 53870
    },
    {
      "entropy": 5.827806091308593,
      "epoch": 5.325721629102412,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002552700407179045,
      "loss": 4.9759,
      "mean_token_accuracy": 0.22901190519332887,
      "num_tokens": 123432089.0,
      "step": 53875
    },
    {
      "entropy": 5.759852886199951,
      "epoch": 5.326215895610914,
      "grad_norm": 1.3125,
      "learning_rate": 0.00025523488691797235,
      "loss": 4.8898,
      "mean_token_accuracy": 0.23924679607152938,
      "num_tokens": 123443623.0,
      "step": 53880
    },
    {
      "entropy": 5.825639629364014,
      "epoch": 5.326710162119415,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00025519973360426507,
      "loss": 4.9283,
      "mean_token_accuracy": 0.23681171387434005,
      "num_tokens": 123454200.0,
      "step": 53885
    },
    {
      "entropy": 5.885631847381592,
      "epoch": 5.327204428627916,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002551645807776475,
      "loss": 4.992,
      "mean_token_accuracy": 0.23027701377868653,
      "num_tokens": 123466361.0,
      "step": 53890
    },
    {
      "entropy": 5.787861299514771,
      "epoch": 5.327698695136418,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002551294284389843,
      "loss": 4.9379,
      "mean_token_accuracy": 0.2290056824684143,
      "num_tokens": 123478355.0,
      "step": 53895
    },
    {
      "entropy": 5.736255025863647,
      "epoch": 5.328192961644919,
      "grad_norm": 1.28125,
      "learning_rate": 0.00025509427658914024,
      "loss": 4.8883,
      "mean_token_accuracy": 0.23788779973983765,
      "num_tokens": 123489463.0,
      "step": 53900
    },
    {
      "entropy": 5.791329050064087,
      "epoch": 5.32868722815342,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00025505912522898,
      "loss": 4.8826,
      "mean_token_accuracy": 0.23348919302225113,
      "num_tokens": 123500924.0,
      "step": 53905
    },
    {
      "entropy": 5.759895467758179,
      "epoch": 5.329181494661921,
      "grad_norm": 1.390625,
      "learning_rate": 0.00025502397435936847,
      "loss": 4.7633,
      "mean_token_accuracy": 0.24542152136564255,
      "num_tokens": 123511414.0,
      "step": 53910
    },
    {
      "entropy": 5.738644933700561,
      "epoch": 5.329675761170423,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002549888239811701,
      "loss": 4.8044,
      "mean_token_accuracy": 0.24394217729568482,
      "num_tokens": 123521741.0,
      "step": 53915
    },
    {
      "entropy": 5.690075063705445,
      "epoch": 5.330170027678925,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002549536740952499,
      "loss": 4.8444,
      "mean_token_accuracy": 0.2390352576971054,
      "num_tokens": 123532469.0,
      "step": 53920
    },
    {
      "entropy": 5.714728689193725,
      "epoch": 5.330664294187426,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002549185247024725,
      "loss": 4.8293,
      "mean_token_accuracy": 0.24825237840414047,
      "num_tokens": 123544931.0,
      "step": 53925
    },
    {
      "entropy": 5.82331953048706,
      "epoch": 5.331158560695927,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00025488337580370236,
      "loss": 4.9349,
      "mean_token_accuracy": 0.22830101549625398,
      "num_tokens": 123557042.0,
      "step": 53930
    },
    {
      "entropy": 5.795929765701294,
      "epoch": 5.331652827204429,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002548482273998044,
      "loss": 4.8645,
      "mean_token_accuracy": 0.23205238282680513,
      "num_tokens": 123567740.0,
      "step": 53935
    },
    {
      "entropy": 5.808051061630249,
      "epoch": 5.33214709371293,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0002548130794916432,
      "loss": 4.9107,
      "mean_token_accuracy": 0.22945208102464676,
      "num_tokens": 123577756.0,
      "step": 53940
    },
    {
      "entropy": 5.780467844009399,
      "epoch": 5.332641360221431,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00025477793208008323,
      "loss": 4.9381,
      "mean_token_accuracy": 0.23680469542741775,
      "num_tokens": 123590160.0,
      "step": 53945
    },
    {
      "entropy": 5.708120965957642,
      "epoch": 5.333135626729932,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002547427851659893,
      "loss": 4.8526,
      "mean_token_accuracy": 0.23673400580883025,
      "num_tokens": 123602240.0,
      "step": 53950
    },
    {
      "entropy": 5.836886930465698,
      "epoch": 5.3336298932384345,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000254707638750226,
      "loss": 4.9871,
      "mean_token_accuracy": 0.22607833743095399,
      "num_tokens": 123613634.0,
      "step": 53955
    },
    {
      "entropy": 5.818803501129151,
      "epoch": 5.334124159746936,
      "grad_norm": 1.359375,
      "learning_rate": 0.000254672492833658,
      "loss": 4.91,
      "mean_token_accuracy": 0.24193129688501358,
      "num_tokens": 123625094.0,
      "step": 53960
    },
    {
      "entropy": 5.803480434417724,
      "epoch": 5.334618426255437,
      "grad_norm": 1.46875,
      "learning_rate": 0.00025463734741714984,
      "loss": 4.9333,
      "mean_token_accuracy": 0.23162195086479187,
      "num_tokens": 123636481.0,
      "step": 53965
    },
    {
      "entropy": 5.709811210632324,
      "epoch": 5.335112692763938,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000254602202501566,
      "loss": 4.7513,
      "mean_token_accuracy": 0.252083446085453,
      "num_tokens": 123645674.0,
      "step": 53970
    },
    {
      "entropy": 5.757914018630982,
      "epoch": 5.3356069592724396,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002545670580877712,
      "loss": 4.7922,
      "mean_token_accuracy": 0.244442018866539,
      "num_tokens": 123656592.0,
      "step": 53975
    },
    {
      "entropy": 5.768587207794189,
      "epoch": 5.336101225780941,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002545319141766299,
      "loss": 4.8711,
      "mean_token_accuracy": 0.23446458876132964,
      "num_tokens": 123668084.0,
      "step": 53980
    },
    {
      "entropy": 5.811733341217041,
      "epoch": 5.336595492289442,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002544967707690068,
      "loss": 5.0349,
      "mean_token_accuracy": 0.22667082995176316,
      "num_tokens": 123679416.0,
      "step": 53985
    },
    {
      "entropy": 5.794829654693603,
      "epoch": 5.337089758797944,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00025446162786576626,
      "loss": 4.8732,
      "mean_token_accuracy": 0.2356559678912163,
      "num_tokens": 123690165.0,
      "step": 53990
    },
    {
      "entropy": 5.854612588882446,
      "epoch": 5.3375840253064455,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002544264854677728,
      "loss": 4.9644,
      "mean_token_accuracy": 0.22763216942548753,
      "num_tokens": 123702294.0,
      "step": 53995
    },
    {
      "entropy": 5.854082918167114,
      "epoch": 5.338078291814947,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00025439134357589107,
      "loss": 4.9205,
      "mean_token_accuracy": 0.23585116118192673,
      "num_tokens": 123712872.0,
      "step": 54000
    },
    {
      "epoch": 5.338078291814947,
      "eval_entropy": 5.5482322355038125,
      "eval_loss": 4.986919403076172,
      "eval_mean_token_accuracy": 0.2368713241313363,
      "eval_num_tokens": 123712872.0,
      "eval_runtime": 20.2791,
      "eval_samples_per_second": 1603.279,
      "eval_steps_per_second": 200.453,
      "step": 54000
    }
  ],
  "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": 2.872000832246784e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}