isl-latn-10mb-after-ppt-Dp-100mb-ckpt500_seed3407/checkpoint-6000/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.5931198102016607,
  "eval_steps": 3000,
  "global_step": 6000,
  "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
    }
  ],
  "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": 3184649031352320.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}