{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.04312910653114,
  "eval_steps": 3000,
  "global_step": 36000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 4.838842916488647,
      "epoch": 0.0005615769079575448,
      "grad_norm": 17.625,
      "learning_rate": 2e-06,
      "loss": 14.3444,
      "mean_token_accuracy": 0.0,
      "num_tokens": 10897.0,
      "step": 5
    },
    {
      "entropy": 4.85677433013916,
      "epoch": 0.0011231538159150896,
      "grad_norm": 27.375,
      "learning_rate": 4.5e-06,
      "loss": 14.2546,
      "mean_token_accuracy": 0.0,
      "num_tokens": 23084.0,
      "step": 10
    },
    {
      "entropy": 4.91091685295105,
      "epoch": 0.0016847307238726343,
      "grad_norm": 22.375,
      "learning_rate": 7e-06,
      "loss": 14.0991,
      "mean_token_accuracy": 0.0,
      "num_tokens": 35810.0,
      "step": 15
    },
    {
      "entropy": 5.158815002441406,
      "epoch": 0.0022463076318301792,
      "grad_norm": 30.625,
      "learning_rate": 9.5e-06,
      "loss": 13.5174,
      "mean_token_accuracy": 0.0,
      "num_tokens": 46019.0,
      "step": 20
    },
    {
      "entropy": 6.645208549499512,
      "epoch": 0.002807884539787724,
      "grad_norm": 26.75,
      "learning_rate": 1.2e-05,
      "loss": 12.2545,
      "mean_token_accuracy": 7.246377062983811e-05,
      "num_tokens": 57272.0,
      "step": 25
    },
    {
      "entropy": 9.382804870605469,
      "epoch": 0.0033694614477452686,
      "grad_norm": 5.15625,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 11.1662,
      "mean_token_accuracy": 0.00013372841058298946,
      "num_tokens": 68686.0,
      "step": 30
    },
    {
      "entropy": 10.640052032470702,
      "epoch": 0.003931038355702813,
      "grad_norm": 2.921875,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 10.7365,
      "mean_token_accuracy": 0.0,
      "num_tokens": 79187.0,
      "step": 35
    },
    {
      "entropy": 10.704295444488526,
      "epoch": 0.0044926152636603585,
      "grad_norm": 2.90625,
      "learning_rate": 1.95e-05,
      "loss": 10.4885,
      "mean_token_accuracy": 0.02202295488677919,
      "num_tokens": 90293.0,
      "step": 40
    },
    {
      "entropy": 10.693971633911133,
      "epoch": 0.005054192171617903,
      "grad_norm": 2.84375,
      "learning_rate": 2.2e-05,
      "loss": 10.1222,
      "mean_token_accuracy": 0.038746101595461366,
      "num_tokens": 100420.0,
      "step": 45
    },
    {
      "entropy": 10.644959449768066,
      "epoch": 0.005615769079575448,
      "grad_norm": 2.484375,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 9.8614,
      "mean_token_accuracy": 0.03684623762965202,
      "num_tokens": 111198.0,
      "step": 50
    },
    {
      "entropy": 10.567186164855958,
      "epoch": 0.006177345987532993,
      "grad_norm": 2.078125,
      "learning_rate": 2.7e-05,
      "loss": 9.6438,
      "mean_token_accuracy": 0.03930287417024374,
      "num_tokens": 124723.0,
      "step": 55
    },
    {
      "entropy": 10.53961763381958,
      "epoch": 0.006738922895490537,
      "grad_norm": 2.1875,
      "learning_rate": 2.95e-05,
      "loss": 9.4635,
      "mean_token_accuracy": 0.03558007832616568,
      "num_tokens": 136366.0,
      "step": 60
    },
    {
      "entropy": 10.547893810272218,
      "epoch": 0.007300499803448083,
      "grad_norm": 2.09375,
      "learning_rate": 3.2e-05,
      "loss": 9.3861,
      "mean_token_accuracy": 0.03799900617450476,
      "num_tokens": 146536.0,
      "step": 65
    },
    {
      "entropy": 10.548893260955811,
      "epoch": 0.007862076711405626,
      "grad_norm": 2.046875,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 9.3036,
      "mean_token_accuracy": 0.04170756805688143,
      "num_tokens": 157005.0,
      "step": 70
    },
    {
      "entropy": 10.522705841064454,
      "epoch": 0.008423653619363172,
      "grad_norm": 1.9765625,
      "learning_rate": 3.7e-05,
      "loss": 9.248,
      "mean_token_accuracy": 0.035757702216506,
      "num_tokens": 167864.0,
      "step": 75
    },
    {
      "entropy": 10.487411975860596,
      "epoch": 0.008985230527320717,
      "grad_norm": 1.984375,
      "learning_rate": 3.95e-05,
      "loss": 9.1701,
      "mean_token_accuracy": 0.0344331294298172,
      "num_tokens": 179635.0,
      "step": 80
    },
    {
      "entropy": 10.44954423904419,
      "epoch": 0.00954680743527826,
      "grad_norm": 1.9140625,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 9.0736,
      "mean_token_accuracy": 0.0393454410135746,
      "num_tokens": 190614.0,
      "step": 85
    },
    {
      "entropy": 10.411544609069825,
      "epoch": 0.010108384343235806,
      "grad_norm": 1.890625,
      "learning_rate": 4.45e-05,
      "loss": 8.9515,
      "mean_token_accuracy": 0.03861959706991911,
      "num_tokens": 201496.0,
      "step": 90
    },
    {
      "entropy": 10.388364028930663,
      "epoch": 0.010669961251193351,
      "grad_norm": 1.953125,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 8.8206,
      "mean_token_accuracy": 0.041578310914337634,
      "num_tokens": 211606.0,
      "step": 95
    },
    {
      "entropy": 10.31544532775879,
      "epoch": 0.011231538159150897,
      "grad_norm": 1.9375,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 8.6811,
      "mean_token_accuracy": 0.03809907417744398,
      "num_tokens": 223503.0,
      "step": 100
    },
    {
      "entropy": 10.235967540740967,
      "epoch": 0.01179311506710844,
      "grad_norm": 1.7421875,
      "learning_rate": 5.2e-05,
      "loss": 8.602,
      "mean_token_accuracy": 0.03533066865056753,
      "num_tokens": 234744.0,
      "step": 105
    },
    {
      "entropy": 10.127453708648682,
      "epoch": 0.012354691975065986,
      "grad_norm": 1.5,
      "learning_rate": 5.45e-05,
      "loss": 8.5013,
      "mean_token_accuracy": 0.04971061982214451,
      "num_tokens": 247663.0,
      "step": 110
    },
    {
      "entropy": 9.991131114959718,
      "epoch": 0.012916268883023531,
      "grad_norm": 1.5703125,
      "learning_rate": 5.7e-05,
      "loss": 8.3462,
      "mean_token_accuracy": 0.051536497101187705,
      "num_tokens": 260326.0,
      "step": 115
    },
    {
      "entropy": 9.806007671356202,
      "epoch": 0.013477845790981074,
      "grad_norm": 1.453125,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 8.186,
      "mean_token_accuracy": 0.04841248132288456,
      "num_tokens": 271568.0,
      "step": 120
    },
    {
      "entropy": 9.62636604309082,
      "epoch": 0.01403942269893862,
      "grad_norm": 1.3671875,
      "learning_rate": 6.2e-05,
      "loss": 8.0076,
      "mean_token_accuracy": 0.057324649021029475,
      "num_tokens": 282651.0,
      "step": 125
    },
    {
      "entropy": 9.388986778259277,
      "epoch": 0.014600999606896165,
      "grad_norm": 1.3359375,
      "learning_rate": 6.450000000000001e-05,
      "loss": 7.8519,
      "mean_token_accuracy": 0.0591864850372076,
      "num_tokens": 292753.0,
      "step": 130
    },
    {
      "entropy": 9.186084842681884,
      "epoch": 0.015162576514853709,
      "grad_norm": 1.0546875,
      "learning_rate": 6.7e-05,
      "loss": 7.8262,
      "mean_token_accuracy": 0.05541557185351849,
      "num_tokens": 302981.0,
      "step": 135
    },
    {
      "entropy": 8.970846366882324,
      "epoch": 0.015724153422811252,
      "grad_norm": 0.96875,
      "learning_rate": 6.950000000000001e-05,
      "loss": 7.7879,
      "mean_token_accuracy": 0.05500966273248196,
      "num_tokens": 314461.0,
      "step": 140
    },
    {
      "entropy": 8.709946346282958,
      "epoch": 0.0162857303307688,
      "grad_norm": 0.9453125,
      "learning_rate": 7.2e-05,
      "loss": 7.7088,
      "mean_token_accuracy": 0.0600316010415554,
      "num_tokens": 326877.0,
      "step": 145
    },
    {
      "entropy": 8.442815399169922,
      "epoch": 0.016847307238726343,
      "grad_norm": 1.0078125,
      "learning_rate": 7.45e-05,
      "loss": 7.5355,
      "mean_token_accuracy": 0.0628300279378891,
      "num_tokens": 337831.0,
      "step": 150
    },
    {
      "entropy": 8.346270942687989,
      "epoch": 0.017408884146683887,
      "grad_norm": 1.015625,
      "learning_rate": 7.7e-05,
      "loss": 7.4941,
      "mean_token_accuracy": 0.061401508003473285,
      "num_tokens": 348233.0,
      "step": 155
    },
    {
      "entropy": 8.139367294311523,
      "epoch": 0.017970461054641434,
      "grad_norm": 0.875,
      "learning_rate": 7.950000000000001e-05,
      "loss": 7.394,
      "mean_token_accuracy": 0.06936323903501034,
      "num_tokens": 359113.0,
      "step": 160
    },
    {
      "entropy": 8.002007293701173,
      "epoch": 0.018532037962598977,
      "grad_norm": 2.3125,
      "learning_rate": 8.2e-05,
      "loss": 7.4953,
      "mean_token_accuracy": 0.06876972652971744,
      "num_tokens": 370284.0,
      "step": 165
    },
    {
      "entropy": 7.985783815383911,
      "epoch": 0.01909361487055652,
      "grad_norm": 1.5390625,
      "learning_rate": 8.450000000000001e-05,
      "loss": 7.5029,
      "mean_token_accuracy": 0.06690707504749298,
      "num_tokens": 381415.0,
      "step": 170
    },
    {
      "entropy": 7.8451512336730955,
      "epoch": 0.019655191778514068,
      "grad_norm": 1.4765625,
      "learning_rate": 8.7e-05,
      "loss": 7.4341,
      "mean_token_accuracy": 0.07936849370598793,
      "num_tokens": 392080.0,
      "step": 175
    },
    {
      "entropy": 7.888399457931518,
      "epoch": 0.020216768686471612,
      "grad_norm": 0.91015625,
      "learning_rate": 8.95e-05,
      "loss": 7.4261,
      "mean_token_accuracy": 0.07295190393924714,
      "num_tokens": 402551.0,
      "step": 180
    },
    {
      "entropy": 7.864237499237061,
      "epoch": 0.020778345594429155,
      "grad_norm": 0.91796875,
      "learning_rate": 9.2e-05,
      "loss": 7.4271,
      "mean_token_accuracy": 0.07404583394527435,
      "num_tokens": 414670.0,
      "step": 185
    },
    {
      "entropy": 7.856368970870972,
      "epoch": 0.021339922502386702,
      "grad_norm": 0.8515625,
      "learning_rate": 9.45e-05,
      "loss": 7.3977,
      "mean_token_accuracy": 0.08026547133922576,
      "num_tokens": 427162.0,
      "step": 190
    },
    {
      "entropy": 7.7947780132293705,
      "epoch": 0.021901499410344246,
      "grad_norm": 0.94921875,
      "learning_rate": 9.7e-05,
      "loss": 7.2803,
      "mean_token_accuracy": 0.08303194791078568,
      "num_tokens": 438628.0,
      "step": 195
    },
    {
      "entropy": 7.755880117416382,
      "epoch": 0.022463076318301793,
      "grad_norm": 1.140625,
      "learning_rate": 9.95e-05,
      "loss": 7.2159,
      "mean_token_accuracy": 0.0891957201063633,
      "num_tokens": 449248.0,
      "step": 200
    },
    {
      "entropy": 7.714970827102661,
      "epoch": 0.023024653226259337,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000102,
      "loss": 7.1809,
      "mean_token_accuracy": 0.08612844422459602,
      "num_tokens": 460815.0,
      "step": 205
    },
    {
      "entropy": 7.662018966674805,
      "epoch": 0.02358623013421688,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00010449999999999999,
      "loss": 7.3416,
      "mean_token_accuracy": 0.0834441527724266,
      "num_tokens": 471972.0,
      "step": 210
    },
    {
      "entropy": 7.711150503158569,
      "epoch": 0.024147807042174427,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000107,
      "loss": 7.3092,
      "mean_token_accuracy": 0.0846812091767788,
      "num_tokens": 482775.0,
      "step": 215
    },
    {
      "entropy": 7.647646760940551,
      "epoch": 0.02470938395013197,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0001095,
      "loss": 7.2401,
      "mean_token_accuracy": 0.08639068603515625,
      "num_tokens": 495151.0,
      "step": 220
    },
    {
      "entropy": 7.704783630371094,
      "epoch": 0.025270960858089515,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000112,
      "loss": 7.2108,
      "mean_token_accuracy": 0.085585917532444,
      "num_tokens": 506947.0,
      "step": 225
    },
    {
      "entropy": 7.684891605377198,
      "epoch": 0.025832537766047062,
      "grad_norm": 1.140625,
      "learning_rate": 0.0001145,
      "loss": 7.2449,
      "mean_token_accuracy": 0.08590937703847885,
      "num_tokens": 519153.0,
      "step": 230
    },
    {
      "entropy": 7.573825359344482,
      "epoch": 0.026394114674004605,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.2067,
      "mean_token_accuracy": 0.08794588595628738,
      "num_tokens": 531050.0,
      "step": 235
    },
    {
      "entropy": 7.663270759582519,
      "epoch": 0.02695569158196215,
      "grad_norm": 1.140625,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.1534,
      "mean_token_accuracy": 0.09201830625534058,
      "num_tokens": 542744.0,
      "step": 240
    },
    {
      "entropy": 7.600920248031616,
      "epoch": 0.027517268489919696,
      "grad_norm": 1.375,
      "learning_rate": 0.000122,
      "loss": 7.228,
      "mean_token_accuracy": 0.08778905719518662,
      "num_tokens": 554748.0,
      "step": 245
    },
    {
      "entropy": 7.532622385025024,
      "epoch": 0.02807884539787724,
      "grad_norm": 1.390625,
      "learning_rate": 0.0001245,
      "loss": 7.1585,
      "mean_token_accuracy": 0.09387479797005653,
      "num_tokens": 564868.0,
      "step": 250
    },
    {
      "entropy": 7.697250652313232,
      "epoch": 0.028640422305834783,
      "grad_norm": 1.359375,
      "learning_rate": 0.000127,
      "loss": 7.1689,
      "mean_token_accuracy": 0.09540257602930069,
      "num_tokens": 575668.0,
      "step": 255
    },
    {
      "entropy": 7.548843145370483,
      "epoch": 0.02920199921379233,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0001295,
      "loss": 7.2777,
      "mean_token_accuracy": 0.08451436385512352,
      "num_tokens": 587956.0,
      "step": 260
    },
    {
      "entropy": 7.618549108505249,
      "epoch": 0.029763576121749874,
      "grad_norm": 1.1875,
      "learning_rate": 0.000132,
      "loss": 7.1221,
      "mean_token_accuracy": 0.09290947988629342,
      "num_tokens": 598246.0,
      "step": 265
    },
    {
      "entropy": 7.489711093902588,
      "epoch": 0.030325153029707418,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.2436,
      "mean_token_accuracy": 0.09298828765749931,
      "num_tokens": 610628.0,
      "step": 270
    },
    {
      "entropy": 7.488981866836548,
      "epoch": 0.030886729937664965,
      "grad_norm": 1.109375,
      "learning_rate": 0.00013700000000000002,
      "loss": 7.0009,
      "mean_token_accuracy": 0.09813047423958779,
      "num_tokens": 621214.0,
      "step": 275
    },
    {
      "entropy": 7.4872016429901125,
      "epoch": 0.031448306845622505,
      "grad_norm": 1.109375,
      "learning_rate": 0.0001395,
      "loss": 7.1012,
      "mean_token_accuracy": 0.10055555552244186,
      "num_tokens": 631123.0,
      "step": 280
    },
    {
      "entropy": 7.584370040893555,
      "epoch": 0.032009883753580055,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00014199999999999998,
      "loss": 7.085,
      "mean_token_accuracy": 0.09335025921463966,
      "num_tokens": 643434.0,
      "step": 285
    },
    {
      "entropy": 7.4562742710113525,
      "epoch": 0.0325714606615376,
      "grad_norm": 1.125,
      "learning_rate": 0.0001445,
      "loss": 7.0706,
      "mean_token_accuracy": 0.09464491978287697,
      "num_tokens": 655336.0,
      "step": 290
    },
    {
      "entropy": 7.467017602920532,
      "epoch": 0.03313303756949514,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000147,
      "loss": 7.0599,
      "mean_token_accuracy": 0.09872465133666992,
      "num_tokens": 666591.0,
      "step": 295
    },
    {
      "entropy": 7.420497798919678,
      "epoch": 0.033694614477452686,
      "grad_norm": 1.359375,
      "learning_rate": 0.0001495,
      "loss": 7.0739,
      "mean_token_accuracy": 0.09496475607156754,
      "num_tokens": 678288.0,
      "step": 300
    },
    {
      "entropy": 7.421144628524781,
      "epoch": 0.03425619138541023,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000152,
      "loss": 7.0016,
      "mean_token_accuracy": 0.10096974968910218,
      "num_tokens": 688526.0,
      "step": 305
    },
    {
      "entropy": 7.3301185131072994,
      "epoch": 0.03481776829336777,
      "grad_norm": 0.921875,
      "learning_rate": 0.00015450000000000001,
      "loss": 6.9177,
      "mean_token_accuracy": 0.110024693608284,
      "num_tokens": 698727.0,
      "step": 310
    },
    {
      "entropy": 7.411620664596557,
      "epoch": 0.035379345201325324,
      "grad_norm": 1.578125,
      "learning_rate": 0.000157,
      "loss": 6.9695,
      "mean_token_accuracy": 0.10299110189080238,
      "num_tokens": 710042.0,
      "step": 315
    },
    {
      "entropy": 7.385035133361816,
      "epoch": 0.03594092210928287,
      "grad_norm": 1.59375,
      "learning_rate": 0.0001595,
      "loss": 6.9146,
      "mean_token_accuracy": 0.10558308735489845,
      "num_tokens": 721390.0,
      "step": 320
    },
    {
      "entropy": 7.296188449859619,
      "epoch": 0.03650249901724041,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000162,
      "loss": 7.0823,
      "mean_token_accuracy": 0.09788208454847336,
      "num_tokens": 732949.0,
      "step": 325
    },
    {
      "entropy": 7.4511847496032715,
      "epoch": 0.037064075925197955,
      "grad_norm": 1.34375,
      "learning_rate": 0.00016450000000000001,
      "loss": 7.0935,
      "mean_token_accuracy": 0.09952903613448143,
      "num_tokens": 744943.0,
      "step": 330
    },
    {
      "entropy": 7.254492616653442,
      "epoch": 0.0376256528331555,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00016700000000000002,
      "loss": 6.8781,
      "mean_token_accuracy": 0.10532095208764077,
      "num_tokens": 756579.0,
      "step": 335
    },
    {
      "entropy": 7.256378412246704,
      "epoch": 0.03818722974111304,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00016950000000000003,
      "loss": 6.8154,
      "mean_token_accuracy": 0.11190035417675973,
      "num_tokens": 767260.0,
      "step": 340
    },
    {
      "entropy": 7.247639799118042,
      "epoch": 0.03874880664907059,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00017199999999999998,
      "loss": 6.9169,
      "mean_token_accuracy": 0.10428953543305397,
      "num_tokens": 778163.0,
      "step": 345
    },
    {
      "entropy": 7.356704139709473,
      "epoch": 0.039310383557028136,
      "grad_norm": 1.25,
      "learning_rate": 0.00017449999999999999,
      "loss": 6.9501,
      "mean_token_accuracy": 0.10315855517983437,
      "num_tokens": 788912.0,
      "step": 350
    },
    {
      "entropy": 7.303907775878907,
      "epoch": 0.03987196046498568,
      "grad_norm": 1.0625,
      "learning_rate": 0.000177,
      "loss": 7.0559,
      "mean_token_accuracy": 0.10429143980145454,
      "num_tokens": 801405.0,
      "step": 355
    },
    {
      "entropy": 7.173692321777343,
      "epoch": 0.040433537372943223,
      "grad_norm": 1.296875,
      "learning_rate": 0.0001795,
      "loss": 6.8638,
      "mean_token_accuracy": 0.10999086052179337,
      "num_tokens": 813023.0,
      "step": 360
    },
    {
      "entropy": 7.268525695800781,
      "epoch": 0.04099511428090077,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000182,
      "loss": 6.8755,
      "mean_token_accuracy": 0.1096720390021801,
      "num_tokens": 823993.0,
      "step": 365
    },
    {
      "entropy": 7.310594749450684,
      "epoch": 0.04155669118885831,
      "grad_norm": 2.15625,
      "learning_rate": 0.0001845,
      "loss": 6.9426,
      "mean_token_accuracy": 0.10535624772310256,
      "num_tokens": 836642.0,
      "step": 370
    },
    {
      "entropy": 7.09607253074646,
      "epoch": 0.04211826809681586,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000187,
      "loss": 6.7626,
      "mean_token_accuracy": 0.10821205154061317,
      "num_tokens": 847475.0,
      "step": 375
    },
    {
      "entropy": 7.293263673782349,
      "epoch": 0.042679845004773405,
      "grad_norm": 1.25,
      "learning_rate": 0.0001895,
      "loss": 6.8202,
      "mean_token_accuracy": 0.1073866032063961,
      "num_tokens": 857949.0,
      "step": 380
    },
    {
      "entropy": 7.130275249481201,
      "epoch": 0.04324142191273095,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000192,
      "loss": 6.8675,
      "mean_token_accuracy": 0.10518962070345879,
      "num_tokens": 868661.0,
      "step": 385
    },
    {
      "entropy": 7.215259170532226,
      "epoch": 0.04380299882068849,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0001945,
      "loss": 6.8914,
      "mean_token_accuracy": 0.10828885361552239,
      "num_tokens": 879651.0,
      "step": 390
    },
    {
      "entropy": 7.22428069114685,
      "epoch": 0.044364575728646036,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00019700000000000002,
      "loss": 6.8748,
      "mean_token_accuracy": 0.10419828295707703,
      "num_tokens": 891189.0,
      "step": 395
    },
    {
      "entropy": 7.181816673278808,
      "epoch": 0.044926152636603586,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00019950000000000002,
      "loss": 6.7881,
      "mean_token_accuracy": 0.10996332466602325,
      "num_tokens": 902915.0,
      "step": 400
    },
    {
      "entropy": 7.18764443397522,
      "epoch": 0.04548772954456113,
      "grad_norm": 1.15625,
      "learning_rate": 0.000202,
      "loss": 6.8843,
      "mean_token_accuracy": 0.10474307015538216,
      "num_tokens": 914108.0,
      "step": 405
    },
    {
      "entropy": 7.189007234573364,
      "epoch": 0.046049306452518673,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00020449999999999998,
      "loss": 6.8786,
      "mean_token_accuracy": 0.10835319235920907,
      "num_tokens": 925729.0,
      "step": 410
    },
    {
      "entropy": 7.182924604415893,
      "epoch": 0.04661088336047622,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000207,
      "loss": 6.8842,
      "mean_token_accuracy": 0.11298655942082406,
      "num_tokens": 936912.0,
      "step": 415
    },
    {
      "entropy": 7.123146343231201,
      "epoch": 0.04717246026843376,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002095,
      "loss": 6.8558,
      "mean_token_accuracy": 0.11014456525444985,
      "num_tokens": 948576.0,
      "step": 420
    },
    {
      "entropy": 7.098531770706177,
      "epoch": 0.047734037176391304,
      "grad_norm": 1.0625,
      "learning_rate": 0.000212,
      "loss": 6.7199,
      "mean_token_accuracy": 0.11547508239746093,
      "num_tokens": 959346.0,
      "step": 425
    },
    {
      "entropy": 7.038528633117676,
      "epoch": 0.048295614084348855,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002145,
      "loss": 6.7322,
      "mean_token_accuracy": 0.11771169304847717,
      "num_tokens": 970124.0,
      "step": 430
    },
    {
      "entropy": 7.150494194030761,
      "epoch": 0.0488571909923064,
      "grad_norm": 1.296875,
      "learning_rate": 0.00021700000000000002,
      "loss": 6.8081,
      "mean_token_accuracy": 0.11283196061849594,
      "num_tokens": 981628.0,
      "step": 435
    },
    {
      "entropy": 7.029330587387085,
      "epoch": 0.04941876790026394,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002195,
      "loss": 6.7668,
      "mean_token_accuracy": 0.12012017667293548,
      "num_tokens": 992423.0,
      "step": 440
    },
    {
      "entropy": 7.096112585067749,
      "epoch": 0.049980344808221486,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000222,
      "loss": 6.7728,
      "mean_token_accuracy": 0.1109731949865818,
      "num_tokens": 1002954.0,
      "step": 445
    },
    {
      "entropy": 7.020883989334107,
      "epoch": 0.05054192171617903,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002245,
      "loss": 6.6564,
      "mean_token_accuracy": 0.11653559058904647,
      "num_tokens": 1013351.0,
      "step": 450
    },
    {
      "entropy": 7.039500284194946,
      "epoch": 0.05110349862413657,
      "grad_norm": 1.484375,
      "learning_rate": 0.00022700000000000002,
      "loss": 6.7656,
      "mean_token_accuracy": 0.11227584257721901,
      "num_tokens": 1023781.0,
      "step": 455
    },
    {
      "entropy": 7.15218596458435,
      "epoch": 0.051665075532094124,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00022950000000000002,
      "loss": 6.8282,
      "mean_token_accuracy": 0.11297958865761756,
      "num_tokens": 1034565.0,
      "step": 460
    },
    {
      "entropy": 6.995783090591431,
      "epoch": 0.05222665244005167,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00023200000000000003,
      "loss": 6.6247,
      "mean_token_accuracy": 0.11796534955501556,
      "num_tokens": 1045595.0,
      "step": 465
    },
    {
      "entropy": 6.886814785003662,
      "epoch": 0.05278822934800921,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00023449999999999998,
      "loss": 6.6337,
      "mean_token_accuracy": 0.11582124829292298,
      "num_tokens": 1056084.0,
      "step": 470
    },
    {
      "entropy": 7.037261819839477,
      "epoch": 0.053349806255966754,
      "grad_norm": 1.421875,
      "learning_rate": 0.000237,
      "loss": 6.682,
      "mean_token_accuracy": 0.11749066337943077,
      "num_tokens": 1067014.0,
      "step": 475
    },
    {
      "entropy": 6.891424036026001,
      "epoch": 0.0539113831639243,
      "grad_norm": 1.375,
      "learning_rate": 0.0002395,
      "loss": 6.6497,
      "mean_token_accuracy": 0.12076789811253548,
      "num_tokens": 1078557.0,
      "step": 480
    },
    {
      "entropy": 6.8824389457702635,
      "epoch": 0.05447296007188184,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000242,
      "loss": 6.7126,
      "mean_token_accuracy": 0.12164518311619758,
      "num_tokens": 1090541.0,
      "step": 485
    },
    {
      "entropy": 6.996250343322754,
      "epoch": 0.05503453697983939,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002445,
      "loss": 6.7637,
      "mean_token_accuracy": 0.11996383219957352,
      "num_tokens": 1101851.0,
      "step": 490
    },
    {
      "entropy": 6.960945463180542,
      "epoch": 0.055596113887796936,
      "grad_norm": 1.359375,
      "learning_rate": 0.000247,
      "loss": 6.675,
      "mean_token_accuracy": 0.11683426722884178,
      "num_tokens": 1112728.0,
      "step": 495
    },
    {
      "entropy": 7.036937141418457,
      "epoch": 0.05615769079575448,
      "grad_norm": 1.65625,
      "learning_rate": 0.0002495,
      "loss": 6.7971,
      "mean_token_accuracy": 0.1172296643257141,
      "num_tokens": 1124537.0,
      "step": 500
    },
    {
      "entropy": 6.785452508926392,
      "epoch": 0.05671926770371202,
      "grad_norm": 1.28125,
      "learning_rate": 0.000252,
      "loss": 6.6958,
      "mean_token_accuracy": 0.11893488615751266,
      "num_tokens": 1135749.0,
      "step": 505
    },
    {
      "entropy": 6.959902667999268,
      "epoch": 0.05728084461166957,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002545,
      "loss": 6.6181,
      "mean_token_accuracy": 0.11765907853841781,
      "num_tokens": 1147597.0,
      "step": 510
    },
    {
      "entropy": 6.922120237350464,
      "epoch": 0.05784242151962711,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000257,
      "loss": 6.5849,
      "mean_token_accuracy": 0.1235293172299862,
      "num_tokens": 1158975.0,
      "step": 515
    },
    {
      "entropy": 6.856467056274414,
      "epoch": 0.05840399842758466,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002595,
      "loss": 6.6972,
      "mean_token_accuracy": 0.12084685266017914,
      "num_tokens": 1170556.0,
      "step": 520
    },
    {
      "entropy": 6.8622949600219725,
      "epoch": 0.058965575335542204,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000262,
      "loss": 6.5607,
      "mean_token_accuracy": 0.12300923243165016,
      "num_tokens": 1180757.0,
      "step": 525
    },
    {
      "entropy": 6.924650192260742,
      "epoch": 0.05952715224349975,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00026450000000000003,
      "loss": 6.596,
      "mean_token_accuracy": 0.12116657346487045,
      "num_tokens": 1192428.0,
      "step": 530
    },
    {
      "entropy": 6.890792942047119,
      "epoch": 0.06008872915145729,
      "grad_norm": 1.25,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.6508,
      "mean_token_accuracy": 0.11567564010620117,
      "num_tokens": 1203867.0,
      "step": 535
    },
    {
      "entropy": 6.883419179916382,
      "epoch": 0.060650306059414835,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00026950000000000005,
      "loss": 6.6555,
      "mean_token_accuracy": 0.12775546982884406,
      "num_tokens": 1214059.0,
      "step": 540
    },
    {
      "entropy": 6.8652650833129885,
      "epoch": 0.06121188296737238,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.7077,
      "mean_token_accuracy": 0.12304283306002617,
      "num_tokens": 1224750.0,
      "step": 545
    },
    {
      "entropy": 6.938561868667603,
      "epoch": 0.06177345987532993,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002745,
      "loss": 6.7628,
      "mean_token_accuracy": 0.11775858029723167,
      "num_tokens": 1235936.0,
      "step": 550
    },
    {
      "entropy": 6.777755689620972,
      "epoch": 0.06233503678328747,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000277,
      "loss": 6.5937,
      "mean_token_accuracy": 0.12519247382879256,
      "num_tokens": 1247619.0,
      "step": 555
    },
    {
      "entropy": 6.887727403640747,
      "epoch": 0.06289661369124501,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002795,
      "loss": 6.659,
      "mean_token_accuracy": 0.12322699278593063,
      "num_tokens": 1258089.0,
      "step": 560
    },
    {
      "entropy": 6.906085586547851,
      "epoch": 0.06345819059920256,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.7312,
      "mean_token_accuracy": 0.12112714126706123,
      "num_tokens": 1270727.0,
      "step": 565
    },
    {
      "entropy": 6.755902671813965,
      "epoch": 0.06401976750716011,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002845,
      "loss": 6.6007,
      "mean_token_accuracy": 0.12350654378533363,
      "num_tokens": 1280886.0,
      "step": 570
    },
    {
      "entropy": 6.883510446548462,
      "epoch": 0.06458134441511765,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000287,
      "loss": 6.6176,
      "mean_token_accuracy": 0.12934386506676673,
      "num_tokens": 1292205.0,
      "step": 575
    },
    {
      "entropy": 6.82674560546875,
      "epoch": 0.0651429213230752,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002895,
      "loss": 6.6112,
      "mean_token_accuracy": 0.11811202019453049,
      "num_tokens": 1304621.0,
      "step": 580
    },
    {
      "entropy": 6.911860084533691,
      "epoch": 0.06570449823103273,
      "grad_norm": 1.1875,
      "learning_rate": 0.000292,
      "loss": 6.687,
      "mean_token_accuracy": 0.12401566356420517,
      "num_tokens": 1315481.0,
      "step": 585
    },
    {
      "entropy": 6.72884316444397,
      "epoch": 0.06626607513899029,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002945,
      "loss": 6.672,
      "mean_token_accuracy": 0.1262218400835991,
      "num_tokens": 1328726.0,
      "step": 590
    },
    {
      "entropy": 6.845497131347656,
      "epoch": 0.06682765204694784,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000297,
      "loss": 6.7212,
      "mean_token_accuracy": 0.11815008223056793,
      "num_tokens": 1340151.0,
      "step": 595
    },
    {
      "entropy": 6.719886589050293,
      "epoch": 0.06738922895490537,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002995,
      "loss": 6.4967,
      "mean_token_accuracy": 0.12117979228496552,
      "num_tokens": 1350833.0,
      "step": 600
    },
    {
      "entropy": 6.805467367172241,
      "epoch": 0.06795080586286292,
      "grad_norm": 1.28125,
      "learning_rate": 0.000302,
      "loss": 6.6645,
      "mean_token_accuracy": 0.12367547824978828,
      "num_tokens": 1362332.0,
      "step": 605
    },
    {
      "entropy": 6.769433450698853,
      "epoch": 0.06851238277082046,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003045,
      "loss": 6.5527,
      "mean_token_accuracy": 0.12536538690328597,
      "num_tokens": 1373696.0,
      "step": 610
    },
    {
      "entropy": 6.676548099517822,
      "epoch": 0.06907395967877801,
      "grad_norm": 1.53125,
      "learning_rate": 0.000307,
      "loss": 6.4719,
      "mean_token_accuracy": 0.13480750992894172,
      "num_tokens": 1383553.0,
      "step": 615
    },
    {
      "entropy": 6.740024566650391,
      "epoch": 0.06963553658673555,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003095,
      "loss": 6.5834,
      "mean_token_accuracy": 0.12851510047912598,
      "num_tokens": 1394487.0,
      "step": 620
    },
    {
      "entropy": 6.825436305999756,
      "epoch": 0.0701971134946931,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000312,
      "loss": 6.6125,
      "mean_token_accuracy": 0.12515986561775208,
      "num_tokens": 1404928.0,
      "step": 625
    },
    {
      "entropy": 6.644348859786987,
      "epoch": 0.07075869040265065,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003145,
      "loss": 6.4801,
      "mean_token_accuracy": 0.13337204083800316,
      "num_tokens": 1415854.0,
      "step": 630
    },
    {
      "entropy": 6.6855494499206545,
      "epoch": 0.07132026731060818,
      "grad_norm": 1.140625,
      "learning_rate": 0.000317,
      "loss": 6.5309,
      "mean_token_accuracy": 0.12912384271621705,
      "num_tokens": 1427153.0,
      "step": 635
    },
    {
      "entropy": 6.709783935546875,
      "epoch": 0.07188184421856574,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003195,
      "loss": 6.6399,
      "mean_token_accuracy": 0.1278666652739048,
      "num_tokens": 1438907.0,
      "step": 640
    },
    {
      "entropy": 6.7197265625,
      "epoch": 0.07244342112652327,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000322,
      "loss": 6.5184,
      "mean_token_accuracy": 0.12605315521359445,
      "num_tokens": 1449975.0,
      "step": 645
    },
    {
      "entropy": 6.712170648574829,
      "epoch": 0.07300499803448082,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.4948,
      "mean_token_accuracy": 0.12966418787837028,
      "num_tokens": 1460815.0,
      "step": 650
    },
    {
      "entropy": 6.6579677104949955,
      "epoch": 0.07356657494243837,
      "grad_norm": 1.125,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.5567,
      "mean_token_accuracy": 0.12379679009318352,
      "num_tokens": 1472403.0,
      "step": 655
    },
    {
      "entropy": 6.710644960403442,
      "epoch": 0.07412815185039591,
      "grad_norm": 1.03125,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.5702,
      "mean_token_accuracy": 0.12100749611854553,
      "num_tokens": 1484015.0,
      "step": 660
    },
    {
      "entropy": 6.797943162918091,
      "epoch": 0.07468972875835346,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.5901,
      "mean_token_accuracy": 0.12561839520931245,
      "num_tokens": 1495383.0,
      "step": 665
    },
    {
      "entropy": 6.693433809280395,
      "epoch": 0.075251305666311,
      "grad_norm": 1.359375,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.5047,
      "mean_token_accuracy": 0.12705199867486955,
      "num_tokens": 1505893.0,
      "step": 670
    },
    {
      "entropy": 6.6826924800872805,
      "epoch": 0.07581288257426855,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000337,
      "loss": 6.6436,
      "mean_token_accuracy": 0.12257061004638672,
      "num_tokens": 1518933.0,
      "step": 675
    },
    {
      "entropy": 6.760889625549316,
      "epoch": 0.07637445948222608,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003395,
      "loss": 6.6673,
      "mean_token_accuracy": 0.12451218515634536,
      "num_tokens": 1530603.0,
      "step": 680
    },
    {
      "entropy": 6.696544408798218,
      "epoch": 0.07693603639018363,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000342,
      "loss": 6.5067,
      "mean_token_accuracy": 0.13363826498389245,
      "num_tokens": 1540977.0,
      "step": 685
    },
    {
      "entropy": 6.735008668899536,
      "epoch": 0.07749761329814119,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.5981,
      "mean_token_accuracy": 0.12193178161978721,
      "num_tokens": 1553004.0,
      "step": 690
    },
    {
      "entropy": 6.677850008010864,
      "epoch": 0.07805919020609872,
      "grad_norm": 1.109375,
      "learning_rate": 0.000347,
      "loss": 6.5178,
      "mean_token_accuracy": 0.13057287335395812,
      "num_tokens": 1563446.0,
      "step": 695
    },
    {
      "entropy": 6.671777582168579,
      "epoch": 0.07862076711405627,
      "grad_norm": 1.125,
      "learning_rate": 0.0003495,
      "loss": 6.5655,
      "mean_token_accuracy": 0.12624817118048667,
      "num_tokens": 1573884.0,
      "step": 700
    },
    {
      "entropy": 6.61401858329773,
      "epoch": 0.07918234402201381,
      "grad_norm": 1.09375,
      "learning_rate": 0.000352,
      "loss": 6.4291,
      "mean_token_accuracy": 0.13389793410897255,
      "num_tokens": 1585396.0,
      "step": 705
    },
    {
      "entropy": 6.501125383377075,
      "epoch": 0.07974392092997136,
      "grad_norm": 1.125,
      "learning_rate": 0.0003545,
      "loss": 6.4109,
      "mean_token_accuracy": 0.13774080127477645,
      "num_tokens": 1596934.0,
      "step": 710
    },
    {
      "entropy": 6.609948444366455,
      "epoch": 0.08030549783792891,
      "grad_norm": 1.078125,
      "learning_rate": 0.000357,
      "loss": 6.4683,
      "mean_token_accuracy": 0.1344927228987217,
      "num_tokens": 1608069.0,
      "step": 715
    },
    {
      "entropy": 6.599338579177856,
      "epoch": 0.08086707474588645,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003595,
      "loss": 6.3866,
      "mean_token_accuracy": 0.13470392003655435,
      "num_tokens": 1619532.0,
      "step": 720
    },
    {
      "entropy": 6.591270208358765,
      "epoch": 0.081428651653844,
      "grad_norm": 1.109375,
      "learning_rate": 0.000362,
      "loss": 6.5525,
      "mean_token_accuracy": 0.12656062543392183,
      "num_tokens": 1630697.0,
      "step": 725
    },
    {
      "entropy": 6.659748363494873,
      "epoch": 0.08199022856180153,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003645,
      "loss": 6.5249,
      "mean_token_accuracy": 0.1280876874923706,
      "num_tokens": 1641854.0,
      "step": 730
    },
    {
      "entropy": 6.703177356719971,
      "epoch": 0.08255180546975908,
      "grad_norm": 1.03125,
      "learning_rate": 0.000367,
      "loss": 6.6371,
      "mean_token_accuracy": 0.1277525916695595,
      "num_tokens": 1653036.0,
      "step": 735
    },
    {
      "entropy": 6.552266073226929,
      "epoch": 0.08311338237771662,
      "grad_norm": 1.125,
      "learning_rate": 0.0003695,
      "loss": 6.4117,
      "mean_token_accuracy": 0.12795400097966195,
      "num_tokens": 1664096.0,
      "step": 740
    },
    {
      "entropy": 6.613392782211304,
      "epoch": 0.08367495928567417,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000372,
      "loss": 6.5774,
      "mean_token_accuracy": 0.12923691421747208,
      "num_tokens": 1677890.0,
      "step": 745
    },
    {
      "entropy": 6.594800758361816,
      "epoch": 0.08423653619363172,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003745,
      "loss": 6.4451,
      "mean_token_accuracy": 0.12722701579332352,
      "num_tokens": 1687627.0,
      "step": 750
    },
    {
      "entropy": 6.683115100860595,
      "epoch": 0.08479811310158926,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000377,
      "loss": 6.6706,
      "mean_token_accuracy": 0.12636393308639526,
      "num_tokens": 1700843.0,
      "step": 755
    },
    {
      "entropy": 6.58741455078125,
      "epoch": 0.08535969000954681,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003795,
      "loss": 6.4928,
      "mean_token_accuracy": 0.12452271282672882,
      "num_tokens": 1712356.0,
      "step": 760
    },
    {
      "entropy": 6.643987846374512,
      "epoch": 0.08592126691750435,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000382,
      "loss": 6.4807,
      "mean_token_accuracy": 0.13168904408812523,
      "num_tokens": 1722041.0,
      "step": 765
    },
    {
      "entropy": 6.53457236289978,
      "epoch": 0.0864828438254619,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003845,
      "loss": 6.57,
      "mean_token_accuracy": 0.12712905704975128,
      "num_tokens": 1733628.0,
      "step": 770
    },
    {
      "entropy": 6.591041088104248,
      "epoch": 0.08704442073341945,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.4336,
      "mean_token_accuracy": 0.13341111317276955,
      "num_tokens": 1745174.0,
      "step": 775
    },
    {
      "entropy": 6.515659618377685,
      "epoch": 0.08760599764137698,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.38,
      "mean_token_accuracy": 0.13189554139971732,
      "num_tokens": 1757117.0,
      "step": 780
    },
    {
      "entropy": 6.525059461593628,
      "epoch": 0.08816757454933453,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.3682,
      "mean_token_accuracy": 0.1381472647190094,
      "num_tokens": 1767343.0,
      "step": 785
    },
    {
      "entropy": 6.593395042419433,
      "epoch": 0.08872915145729207,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.462,
      "mean_token_accuracy": 0.12963056042790413,
      "num_tokens": 1778500.0,
      "step": 790
    },
    {
      "entropy": 6.50290265083313,
      "epoch": 0.08929072836524962,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.4653,
      "mean_token_accuracy": 0.13221071287989616,
      "num_tokens": 1788474.0,
      "step": 795
    },
    {
      "entropy": 6.755932855606079,
      "epoch": 0.08985230527320717,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003995,
      "loss": 6.6218,
      "mean_token_accuracy": 0.12680470421910287,
      "num_tokens": 1800756.0,
      "step": 800
    },
    {
      "entropy": 6.268558931350708,
      "epoch": 0.09041388218116471,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000402,
      "loss": 6.3418,
      "mean_token_accuracy": 0.1349015899002552,
      "num_tokens": 1812172.0,
      "step": 805
    },
    {
      "entropy": 6.630968856811523,
      "epoch": 0.09097545908912226,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004045,
      "loss": 6.4633,
      "mean_token_accuracy": 0.1382911778986454,
      "num_tokens": 1822057.0,
      "step": 810
    },
    {
      "entropy": 6.488110589981079,
      "epoch": 0.0915370359970798,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.4049,
      "mean_token_accuracy": 0.13655738160014153,
      "num_tokens": 1833074.0,
      "step": 815
    },
    {
      "entropy": 6.4761574268341064,
      "epoch": 0.09209861290503735,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004095,
      "loss": 6.3869,
      "mean_token_accuracy": 0.14094622135162355,
      "num_tokens": 1844932.0,
      "step": 820
    },
    {
      "entropy": 6.494161558151245,
      "epoch": 0.09266018981299488,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000412,
      "loss": 6.4131,
      "mean_token_accuracy": 0.1332133084535599,
      "num_tokens": 1855591.0,
      "step": 825
    },
    {
      "entropy": 6.388966846466064,
      "epoch": 0.09322176672095243,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004145,
      "loss": 6.3704,
      "mean_token_accuracy": 0.13496313914656638,
      "num_tokens": 1866726.0,
      "step": 830
    },
    {
      "entropy": 6.572238254547119,
      "epoch": 0.09378334362890998,
      "grad_norm": 1.21875,
      "learning_rate": 0.000417,
      "loss": 6.4138,
      "mean_token_accuracy": 0.1391880229115486,
      "num_tokens": 1877385.0,
      "step": 835
    },
    {
      "entropy": 6.4263691902160645,
      "epoch": 0.09434492053686752,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004195,
      "loss": 6.3199,
      "mean_token_accuracy": 0.1338621348142624,
      "num_tokens": 1887905.0,
      "step": 840
    },
    {
      "entropy": 6.395117902755738,
      "epoch": 0.09490649744482507,
      "grad_norm": 1.046875,
      "learning_rate": 0.000422,
      "loss": 6.3642,
      "mean_token_accuracy": 0.13623479902744293,
      "num_tokens": 1898649.0,
      "step": 845
    },
    {
      "entropy": 6.444835424423218,
      "epoch": 0.09546807435278261,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004245,
      "loss": 6.3703,
      "mean_token_accuracy": 0.13008482083678247,
      "num_tokens": 1908808.0,
      "step": 850
    },
    {
      "entropy": 6.572300815582276,
      "epoch": 0.09602965126074016,
      "grad_norm": 1.125,
      "learning_rate": 0.000427,
      "loss": 6.4928,
      "mean_token_accuracy": 0.1304762378334999,
      "num_tokens": 1920360.0,
      "step": 855
    },
    {
      "entropy": 6.422976160049439,
      "epoch": 0.09659122816869771,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004295,
      "loss": 6.3844,
      "mean_token_accuracy": 0.13653280809521676,
      "num_tokens": 1931082.0,
      "step": 860
    },
    {
      "entropy": 6.534259939193726,
      "epoch": 0.09715280507665525,
      "grad_norm": 1.0625,
      "learning_rate": 0.000432,
      "loss": 6.5493,
      "mean_token_accuracy": 0.12917705476284028,
      "num_tokens": 1942712.0,
      "step": 865
    },
    {
      "entropy": 6.3769426345825195,
      "epoch": 0.0977143819846128,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004345,
      "loss": 6.2825,
      "mean_token_accuracy": 0.13908328264951705,
      "num_tokens": 1953316.0,
      "step": 870
    },
    {
      "entropy": 6.47237458229065,
      "epoch": 0.09827595889257033,
      "grad_norm": 1.28125,
      "learning_rate": 0.000437,
      "loss": 6.3548,
      "mean_token_accuracy": 0.1339756242930889,
      "num_tokens": 1963936.0,
      "step": 875
    },
    {
      "entropy": 6.405522155761719,
      "epoch": 0.09883753580052788,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004395,
      "loss": 6.3967,
      "mean_token_accuracy": 0.14097022041678428,
      "num_tokens": 1974502.0,
      "step": 880
    },
    {
      "entropy": 6.409914779663086,
      "epoch": 0.09939911270848542,
      "grad_norm": 0.96875,
      "learning_rate": 0.000442,
      "loss": 6.3222,
      "mean_token_accuracy": 0.1387803927063942,
      "num_tokens": 1985746.0,
      "step": 885
    },
    {
      "entropy": 6.426558923721314,
      "epoch": 0.09996068961644297,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004445,
      "loss": 6.4244,
      "mean_token_accuracy": 0.13661134615540504,
      "num_tokens": 1997696.0,
      "step": 890
    },
    {
      "entropy": 6.5053589820861815,
      "epoch": 0.10052226652440052,
      "grad_norm": 1.140625,
      "learning_rate": 0.000447,
      "loss": 6.3979,
      "mean_token_accuracy": 0.1326365053653717,
      "num_tokens": 2009462.0,
      "step": 895
    },
    {
      "entropy": 6.395185899734497,
      "epoch": 0.10108384343235806,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.3199,
      "mean_token_accuracy": 0.13828641325235366,
      "num_tokens": 2019547.0,
      "step": 900
    },
    {
      "entropy": 6.321804475784302,
      "epoch": 0.10164542034031561,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.3305,
      "mean_token_accuracy": 0.13299717009067535,
      "num_tokens": 2030990.0,
      "step": 905
    },
    {
      "entropy": 6.37341685295105,
      "epoch": 0.10220699724827315,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.4025,
      "mean_token_accuracy": 0.13234635218977928,
      "num_tokens": 2042488.0,
      "step": 910
    },
    {
      "entropy": 6.573526334762573,
      "epoch": 0.1027685741562307,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.4312,
      "mean_token_accuracy": 0.1312362991273403,
      "num_tokens": 2052822.0,
      "step": 915
    },
    {
      "entropy": 6.418108129501343,
      "epoch": 0.10333015106418825,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.3949,
      "mean_token_accuracy": 0.13608081564307212,
      "num_tokens": 2063274.0,
      "step": 920
    },
    {
      "entropy": 6.522128009796143,
      "epoch": 0.10389172797214578,
      "grad_norm": 1.0,
      "learning_rate": 0.000462,
      "loss": 6.4406,
      "mean_token_accuracy": 0.1320824518799782,
      "num_tokens": 2074424.0,
      "step": 925
    },
    {
      "entropy": 6.382816791534424,
      "epoch": 0.10445330488010333,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004645,
      "loss": 6.4306,
      "mean_token_accuracy": 0.13350846022367477,
      "num_tokens": 2085776.0,
      "step": 930
    },
    {
      "entropy": 6.57677001953125,
      "epoch": 0.10501488178806087,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000467,
      "loss": 6.4913,
      "mean_token_accuracy": 0.13423626497387886,
      "num_tokens": 2096789.0,
      "step": 935
    },
    {
      "entropy": 6.406237030029297,
      "epoch": 0.10557645869601842,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004695,
      "loss": 6.4001,
      "mean_token_accuracy": 0.1370982863008976,
      "num_tokens": 2108335.0,
      "step": 940
    },
    {
      "entropy": 6.491322803497314,
      "epoch": 0.10613803560397596,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000472,
      "loss": 6.4755,
      "mean_token_accuracy": 0.13377573490142822,
      "num_tokens": 2121063.0,
      "step": 945
    },
    {
      "entropy": 6.351012897491455,
      "epoch": 0.10669961251193351,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004745,
      "loss": 6.3532,
      "mean_token_accuracy": 0.14160069227218627,
      "num_tokens": 2132442.0,
      "step": 950
    },
    {
      "entropy": 6.3580584049224855,
      "epoch": 0.10726118941989106,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000477,
      "loss": 6.2661,
      "mean_token_accuracy": 0.14054919332265853,
      "num_tokens": 2142155.0,
      "step": 955
    },
    {
      "entropy": 6.399745845794678,
      "epoch": 0.1078227663278486,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004795,
      "loss": 6.336,
      "mean_token_accuracy": 0.14111177176237105,
      "num_tokens": 2152923.0,
      "step": 960
    },
    {
      "entropy": 6.359347009658814,
      "epoch": 0.10838434323580615,
      "grad_norm": 1.0,
      "learning_rate": 0.000482,
      "loss": 6.3205,
      "mean_token_accuracy": 0.13966229930520058,
      "num_tokens": 2164640.0,
      "step": 965
    },
    {
      "entropy": 6.258652973175049,
      "epoch": 0.10894592014376368,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004845,
      "loss": 6.2377,
      "mean_token_accuracy": 0.1407262295484543,
      "num_tokens": 2175403.0,
      "step": 970
    },
    {
      "entropy": 6.3965113162994385,
      "epoch": 0.10950749705172123,
      "grad_norm": 1.03125,
      "learning_rate": 0.000487,
      "loss": 6.3472,
      "mean_token_accuracy": 0.13764843419194223,
      "num_tokens": 2186031.0,
      "step": 975
    },
    {
      "entropy": 6.3890174388885494,
      "epoch": 0.11006907395967878,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004895,
      "loss": 6.3411,
      "mean_token_accuracy": 0.13227884098887444,
      "num_tokens": 2198152.0,
      "step": 980
    },
    {
      "entropy": 6.318311357498169,
      "epoch": 0.11063065086763632,
      "grad_norm": 1.046875,
      "learning_rate": 0.000492,
      "loss": 6.3484,
      "mean_token_accuracy": 0.13873713985085487,
      "num_tokens": 2209435.0,
      "step": 985
    },
    {
      "entropy": 6.323774909973144,
      "epoch": 0.11119222777559387,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004945,
      "loss": 6.3035,
      "mean_token_accuracy": 0.1427369087934494,
      "num_tokens": 2219476.0,
      "step": 990
    },
    {
      "entropy": 6.493531656265259,
      "epoch": 0.11175380468355141,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000497,
      "loss": 6.3993,
      "mean_token_accuracy": 0.1396132729947567,
      "num_tokens": 2231131.0,
      "step": 995
    },
    {
      "entropy": 6.319530439376831,
      "epoch": 0.11231538159150896,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004995,
      "loss": 6.3023,
      "mean_token_accuracy": 0.14031659737229346,
      "num_tokens": 2242309.0,
      "step": 1000
    },
    {
      "entropy": 6.262487936019897,
      "epoch": 0.1128769584994665,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499999997707492,
      "loss": 6.154,
      "mean_token_accuracy": 0.1367334969341755,
      "num_tokens": 2252236.0,
      "step": 1005
    },
    {
      "entropy": 6.368433952331543,
      "epoch": 0.11343853540742405,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999999883941787,
      "loss": 6.4644,
      "mean_token_accuracy": 0.12982962355017663,
      "num_tokens": 2263529.0,
      "step": 1010
    },
    {
      "entropy": 6.292843532562256,
      "epoch": 0.1140001123153816,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999999719167785,
      "loss": 6.3108,
      "mean_token_accuracy": 0.13819449320435523,
      "num_tokens": 2275318.0,
      "step": 1015
    },
    {
      "entropy": 6.369863128662109,
      "epoch": 0.11456168922333913,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999999482752917,
      "loss": 6.3412,
      "mean_token_accuracy": 0.136563753336668,
      "num_tokens": 2287434.0,
      "step": 1020
    },
    {
      "entropy": 6.253592109680175,
      "epoch": 0.11512326613129668,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004999999174697195,
      "loss": 6.2465,
      "mean_token_accuracy": 0.13917232006788255,
      "num_tokens": 2299145.0,
      "step": 1025
    },
    {
      "entropy": 6.300353574752807,
      "epoch": 0.11568484303925422,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999998795000627,
      "loss": 6.3371,
      "mean_token_accuracy": 0.1372515380382538,
      "num_tokens": 2309841.0,
      "step": 1030
    },
    {
      "entropy": 6.349047756195068,
      "epoch": 0.11624641994721177,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999998343663223,
      "loss": 6.3261,
      "mean_token_accuracy": 0.1395231880247593,
      "num_tokens": 2322143.0,
      "step": 1035
    },
    {
      "entropy": 6.3239373683929445,
      "epoch": 0.11680799685516932,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999997820685,
      "loss": 6.2641,
      "mean_token_accuracy": 0.13360065817832947,
      "num_tokens": 2333207.0,
      "step": 1040
    },
    {
      "entropy": 6.235982275009155,
      "epoch": 0.11736957376312686,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999997226065974,
      "loss": 6.3289,
      "mean_token_accuracy": 0.13780270218849183,
      "num_tokens": 2345542.0,
      "step": 1045
    },
    {
      "entropy": 6.4007457256317135,
      "epoch": 0.11793115067108441,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999996559806163,
      "loss": 6.314,
      "mean_token_accuracy": 0.14232611060142517,
      "num_tokens": 2356612.0,
      "step": 1050
    },
    {
      "entropy": 6.266652011871338,
      "epoch": 0.11849272757904195,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999995821905589,
      "loss": 6.2434,
      "mean_token_accuracy": 0.13702767938375474,
      "num_tokens": 2368204.0,
      "step": 1055
    },
    {
      "entropy": 6.304989004135132,
      "epoch": 0.1190543044869995,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999995012364276,
      "loss": 6.2753,
      "mean_token_accuracy": 0.14323293939232826,
      "num_tokens": 2379114.0,
      "step": 1060
    },
    {
      "entropy": 6.26793475151062,
      "epoch": 0.11961588139495703,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999994131182248,
      "loss": 6.3179,
      "mean_token_accuracy": 0.1384415589272976,
      "num_tokens": 2390562.0,
      "step": 1065
    },
    {
      "entropy": 6.350690174102783,
      "epoch": 0.12017745830291458,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999993178359535,
      "loss": 6.3826,
      "mean_token_accuracy": 0.13662955462932586,
      "num_tokens": 2402352.0,
      "step": 1070
    },
    {
      "entropy": 6.4283607006073,
      "epoch": 0.12073903521087213,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999992153896166,
      "loss": 6.4014,
      "mean_token_accuracy": 0.14080253392457961,
      "num_tokens": 2415140.0,
      "step": 1075
    },
    {
      "entropy": 6.249602460861206,
      "epoch": 0.12130061211882967,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999991057792175,
      "loss": 6.2969,
      "mean_token_accuracy": 0.13987230136990547,
      "num_tokens": 2426679.0,
      "step": 1080
    },
    {
      "entropy": 6.337273216247558,
      "epoch": 0.12186218902678722,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999989890047596,
      "loss": 6.3586,
      "mean_token_accuracy": 0.14236537516117095,
      "num_tokens": 2437735.0,
      "step": 1085
    },
    {
      "entropy": 6.343389844894409,
      "epoch": 0.12242376593474476,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999988650662467,
      "loss": 6.2561,
      "mean_token_accuracy": 0.14142627790570259,
      "num_tokens": 2448611.0,
      "step": 1090
    },
    {
      "entropy": 6.220987558364868,
      "epoch": 0.12298534284270231,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999987339636825,
      "loss": 6.1792,
      "mean_token_accuracy": 0.14699204117059708,
      "num_tokens": 2461101.0,
      "step": 1095
    },
    {
      "entropy": 6.273333358764648,
      "epoch": 0.12354691975065986,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999985956970713,
      "loss": 6.3255,
      "mean_token_accuracy": 0.1386694774031639,
      "num_tokens": 2473569.0,
      "step": 1100
    },
    {
      "entropy": 6.208774709701538,
      "epoch": 0.1241084966586174,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999984502664178,
      "loss": 6.2466,
      "mean_token_accuracy": 0.1429204486310482,
      "num_tokens": 2483970.0,
      "step": 1105
    },
    {
      "entropy": 6.335579538345337,
      "epoch": 0.12467007356657495,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999982976717261,
      "loss": 6.3077,
      "mean_token_accuracy": 0.14402595162391663,
      "num_tokens": 2494184.0,
      "step": 1110
    },
    {
      "entropy": 6.263895559310913,
      "epoch": 0.1252316504745325,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999981379130016,
      "loss": 6.3083,
      "mean_token_accuracy": 0.14719027727842332,
      "num_tokens": 2505715.0,
      "step": 1115
    },
    {
      "entropy": 6.438121223449707,
      "epoch": 0.12579322738249002,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499997970990249,
      "loss": 6.3564,
      "mean_token_accuracy": 0.1362408310174942,
      "num_tokens": 2517168.0,
      "step": 1120
    },
    {
      "entropy": 6.369924163818359,
      "epoch": 0.12635480429044757,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999977969034737,
      "loss": 6.3785,
      "mean_token_accuracy": 0.13768665492534637,
      "num_tokens": 2528296.0,
      "step": 1125
    },
    {
      "entropy": 6.212030839920044,
      "epoch": 0.12691638119840512,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999976156526814,
      "loss": 6.3186,
      "mean_token_accuracy": 0.14621145129203797,
      "num_tokens": 2540648.0,
      "step": 1130
    },
    {
      "entropy": 6.224785709381104,
      "epoch": 0.12747795810636267,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999974272378776,
      "loss": 6.196,
      "mean_token_accuracy": 0.14358407855033875,
      "num_tokens": 2551840.0,
      "step": 1135
    },
    {
      "entropy": 6.380413961410523,
      "epoch": 0.12803953501432022,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999972316590685,
      "loss": 6.263,
      "mean_token_accuracy": 0.14095489978790282,
      "num_tokens": 2561362.0,
      "step": 1140
    },
    {
      "entropy": 6.275799179077149,
      "epoch": 0.12860111192227774,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999970289162604,
      "loss": 6.3161,
      "mean_token_accuracy": 0.14759473651647567,
      "num_tokens": 2573584.0,
      "step": 1145
    },
    {
      "entropy": 6.472133827209473,
      "epoch": 0.1291626888302353,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999968190094595,
      "loss": 6.4268,
      "mean_token_accuracy": 0.13627540394663812,
      "num_tokens": 2585255.0,
      "step": 1150
    },
    {
      "entropy": 6.098404502868652,
      "epoch": 0.12972426573819285,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999966019386727,
      "loss": 6.2311,
      "mean_token_accuracy": 0.14041587859392166,
      "num_tokens": 2596849.0,
      "step": 1155
    },
    {
      "entropy": 6.301661539077759,
      "epoch": 0.1302858426461504,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999963777039067,
      "loss": 6.2686,
      "mean_token_accuracy": 0.1406569629907608,
      "num_tokens": 2606918.0,
      "step": 1160
    },
    {
      "entropy": 6.24212965965271,
      "epoch": 0.13084741955410795,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499996146305169,
      "loss": 6.2067,
      "mean_token_accuracy": 0.1421182483434677,
      "num_tokens": 2617779.0,
      "step": 1165
    },
    {
      "entropy": 6.263619422912598,
      "epoch": 0.13140899646206547,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999959077424665,
      "loss": 6.2691,
      "mean_token_accuracy": 0.14162113517522812,
      "num_tokens": 2630294.0,
      "step": 1170
    },
    {
      "entropy": 6.185452175140381,
      "epoch": 0.13197057337002302,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999956620158072,
      "loss": 6.1681,
      "mean_token_accuracy": 0.1465056374669075,
      "num_tokens": 2641002.0,
      "step": 1175
    },
    {
      "entropy": 6.21528000831604,
      "epoch": 0.13253215027798057,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999954091251987,
      "loss": 6.1524,
      "mean_token_accuracy": 0.14625044465065,
      "num_tokens": 2652188.0,
      "step": 1180
    },
    {
      "entropy": 6.194118452072144,
      "epoch": 0.13309372718593812,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999951490706491,
      "loss": 6.2151,
      "mean_token_accuracy": 0.14109446853399277,
      "num_tokens": 2662542.0,
      "step": 1185
    },
    {
      "entropy": 6.315721750259399,
      "epoch": 0.13365530409389567,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999948818521667,
      "loss": 6.2176,
      "mean_token_accuracy": 0.14606809467077256,
      "num_tokens": 2673292.0,
      "step": 1190
    },
    {
      "entropy": 6.192793464660644,
      "epoch": 0.1342168810018532,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00049999460746976,
      "loss": 6.2188,
      "mean_token_accuracy": 0.14586541950702667,
      "num_tokens": 2684042.0,
      "step": 1195
    },
    {
      "entropy": 6.236338186264038,
      "epoch": 0.13477845790981074,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999943259234377,
      "loss": 6.1383,
      "mean_token_accuracy": 0.14659893661737441,
      "num_tokens": 2694716.0,
      "step": 1200
    },
    {
      "entropy": 6.314930438995361,
      "epoch": 0.1353400348177683,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999940372132089,
      "loss": 6.2874,
      "mean_token_accuracy": 0.14430573657155038,
      "num_tokens": 2706200.0,
      "step": 1205
    },
    {
      "entropy": 6.087372541427612,
      "epoch": 0.13590161172572585,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999937413390826,
      "loss": 6.1129,
      "mean_token_accuracy": 0.1477084070444107,
      "num_tokens": 2716814.0,
      "step": 1210
    },
    {
      "entropy": 6.265769004821777,
      "epoch": 0.1364631886336834,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999934383010683,
      "loss": 6.2639,
      "mean_token_accuracy": 0.1386260524392128,
      "num_tokens": 2728042.0,
      "step": 1215
    },
    {
      "entropy": 6.220071077346802,
      "epoch": 0.13702476554164092,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999931280991757,
      "loss": 6.2398,
      "mean_token_accuracy": 0.13685657903552056,
      "num_tokens": 2739093.0,
      "step": 1220
    },
    {
      "entropy": 6.315275382995606,
      "epoch": 0.13758634244959847,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999928107334146,
      "loss": 6.273,
      "mean_token_accuracy": 0.13907489106059073,
      "num_tokens": 2750307.0,
      "step": 1225
    },
    {
      "entropy": 6.200074577331543,
      "epoch": 0.13814791935755602,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999924862037952,
      "loss": 6.1898,
      "mean_token_accuracy": 0.14381182566285133,
      "num_tokens": 2760534.0,
      "step": 1230
    },
    {
      "entropy": 6.189889335632325,
      "epoch": 0.13870949626551357,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999921545103278,
      "loss": 6.217,
      "mean_token_accuracy": 0.1399001397192478,
      "num_tokens": 2772459.0,
      "step": 1235
    },
    {
      "entropy": 6.180458927154541,
      "epoch": 0.1392710731734711,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999918156530229,
      "loss": 6.1231,
      "mean_token_accuracy": 0.1533089265227318,
      "num_tokens": 2784038.0,
      "step": 1240
    },
    {
      "entropy": 6.275652027130127,
      "epoch": 0.13983265008142864,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999914696318913,
      "loss": 6.2431,
      "mean_token_accuracy": 0.15094640329480172,
      "num_tokens": 2795756.0,
      "step": 1245
    },
    {
      "entropy": 6.123087310791016,
      "epoch": 0.1403942269893862,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999911164469441,
      "loss": 6.1844,
      "mean_token_accuracy": 0.14661128968000411,
      "num_tokens": 2806242.0,
      "step": 1250
    },
    {
      "entropy": 6.29753270149231,
      "epoch": 0.14095580389734375,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999907560981925,
      "loss": 6.2712,
      "mean_token_accuracy": 0.14754103049635886,
      "num_tokens": 2817017.0,
      "step": 1255
    },
    {
      "entropy": 6.216703319549561,
      "epoch": 0.1415173808053013,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999903885856479,
      "loss": 6.1715,
      "mean_token_accuracy": 0.15246719866991043,
      "num_tokens": 2827415.0,
      "step": 1260
    },
    {
      "entropy": 6.185984897613525,
      "epoch": 0.14207895771325882,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000499990013909322,
      "loss": 6.1905,
      "mean_token_accuracy": 0.1513909749686718,
      "num_tokens": 2837595.0,
      "step": 1265
    },
    {
      "entropy": 6.154061126708984,
      "epoch": 0.14264053462121637,
      "grad_norm": 1.0,
      "learning_rate": 0.000499989632069227,
      "loss": 6.1333,
      "mean_token_accuracy": 0.14768583253026007,
      "num_tokens": 2847906.0,
      "step": 1270
    },
    {
      "entropy": 6.224718379974365,
      "epoch": 0.14320211152917392,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999892430653747,
      "loss": 6.2518,
      "mean_token_accuracy": 0.13458375483751298,
      "num_tokens": 2859145.0,
      "step": 1275
    },
    {
      "entropy": 6.155771541595459,
      "epoch": 0.14376368843713147,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999888468977776,
      "loss": 6.165,
      "mean_token_accuracy": 0.14523654133081437,
      "num_tokens": 2870704.0,
      "step": 1280
    },
    {
      "entropy": 6.266523933410644,
      "epoch": 0.14432526534508902,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999884435664485,
      "loss": 6.2767,
      "mean_token_accuracy": 0.1407052122056484,
      "num_tokens": 2881861.0,
      "step": 1285
    },
    {
      "entropy": 6.25180025100708,
      "epoch": 0.14488684225304654,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004999880330714,
      "loss": 6.2025,
      "mean_token_accuracy": 0.1466933399438858,
      "num_tokens": 2893247.0,
      "step": 1290
    },
    {
      "entropy": 6.105476760864258,
      "epoch": 0.1454484191610041,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999876154126453,
      "loss": 6.0418,
      "mean_token_accuracy": 0.14936177730560302,
      "num_tokens": 2905351.0,
      "step": 1295
    },
    {
      "entropy": 6.1718381404876705,
      "epoch": 0.14600999606896164,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999871905901977,
      "loss": 6.2019,
      "mean_token_accuracy": 0.1479304015636444,
      "num_tokens": 2915153.0,
      "step": 1300
    },
    {
      "entropy": 6.215558004379273,
      "epoch": 0.1465715729769192,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999867586040706,
      "loss": 6.158,
      "mean_token_accuracy": 0.13994157537817956,
      "num_tokens": 2926234.0,
      "step": 1305
    },
    {
      "entropy": 6.141939926147461,
      "epoch": 0.14713314988487675,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999863194542779,
      "loss": 6.1102,
      "mean_token_accuracy": 0.14791369438171387,
      "num_tokens": 2937033.0,
      "step": 1310
    },
    {
      "entropy": 6.09499454498291,
      "epoch": 0.14769472679283427,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999858731408336,
      "loss": 6.0657,
      "mean_token_accuracy": 0.1484006866812706,
      "num_tokens": 2945904.0,
      "step": 1315
    },
    {
      "entropy": 6.192239475250244,
      "epoch": 0.14825630370079182,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999854196637518,
      "loss": 6.2645,
      "mean_token_accuracy": 0.13989078402519226,
      "num_tokens": 2956298.0,
      "step": 1320
    },
    {
      "entropy": 6.196642541885376,
      "epoch": 0.14881788060874937,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999849590230469,
      "loss": 6.1577,
      "mean_token_accuracy": 0.14732514470815658,
      "num_tokens": 2967660.0,
      "step": 1325
    },
    {
      "entropy": 6.234364652633667,
      "epoch": 0.14937945751670692,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999844912187336,
      "loss": 6.2437,
      "mean_token_accuracy": 0.14329968988895417,
      "num_tokens": 2978616.0,
      "step": 1330
    },
    {
      "entropy": 6.08893051147461,
      "epoch": 0.14994103442466447,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499984016250827,
      "loss": 6.0911,
      "mean_token_accuracy": 0.14890712201595308,
      "num_tokens": 2989731.0,
      "step": 1335
    },
    {
      "entropy": 6.246871900558472,
      "epoch": 0.150502611332622,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499983534119342,
      "loss": 6.1816,
      "mean_token_accuracy": 0.14426462948322297,
      "num_tokens": 3002110.0,
      "step": 1340
    },
    {
      "entropy": 5.933366966247559,
      "epoch": 0.15106418824057954,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999830448242941,
      "loss": 6.0031,
      "mean_token_accuracy": 0.1531351014971733,
      "num_tokens": 3012709.0,
      "step": 1345
    },
    {
      "entropy": 6.321645641326905,
      "epoch": 0.1516257651485371,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999825483656986,
      "loss": 6.3087,
      "mean_token_accuracy": 0.13611152917146682,
      "num_tokens": 3025738.0,
      "step": 1350
    },
    {
      "entropy": 6.149252319335938,
      "epoch": 0.15218734205649465,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999820447435716,
      "loss": 6.1525,
      "mean_token_accuracy": 0.14370962455868722,
      "num_tokens": 3037711.0,
      "step": 1355
    },
    {
      "entropy": 6.265703773498535,
      "epoch": 0.15274891896445217,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999815339579291,
      "loss": 6.3363,
      "mean_token_accuracy": 0.1384710282087326,
      "num_tokens": 3048929.0,
      "step": 1360
    },
    {
      "entropy": 6.190625715255737,
      "epoch": 0.15331049587240972,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004999810160087872,
      "loss": 6.0812,
      "mean_token_accuracy": 0.1597306936979294,
      "num_tokens": 3059717.0,
      "step": 1365
    },
    {
      "entropy": 6.122515726089477,
      "epoch": 0.15387207278036727,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999804908961626,
      "loss": 6.0872,
      "mean_token_accuracy": 0.14993196427822114,
      "num_tokens": 3069530.0,
      "step": 1370
    },
    {
      "entropy": 6.177134799957275,
      "epoch": 0.15443364968832482,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999799586200719,
      "loss": 6.2762,
      "mean_token_accuracy": 0.14048854932188987,
      "num_tokens": 3080705.0,
      "step": 1375
    },
    {
      "entropy": 6.211107969284058,
      "epoch": 0.15499522659628237,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499979419180532,
      "loss": 6.1237,
      "mean_token_accuracy": 0.1432718813419342,
      "num_tokens": 3092553.0,
      "step": 1380
    },
    {
      "entropy": 6.260249090194702,
      "epoch": 0.1555568035042399,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999788725775601,
      "loss": 6.2522,
      "mean_token_accuracy": 0.14531850293278695,
      "num_tokens": 3102947.0,
      "step": 1385
    },
    {
      "entropy": 6.119967412948609,
      "epoch": 0.15611838041219744,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999783188111737,
      "loss": 6.0963,
      "mean_token_accuracy": 0.14920552223920822,
      "num_tokens": 3114157.0,
      "step": 1390
    },
    {
      "entropy": 6.029822540283203,
      "epoch": 0.156679957320155,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999777578813904,
      "loss": 5.9551,
      "mean_token_accuracy": 0.1501602292060852,
      "num_tokens": 3125739.0,
      "step": 1395
    },
    {
      "entropy": 6.111610746383667,
      "epoch": 0.15724153422811254,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499977189788228,
      "loss": 6.1285,
      "mean_token_accuracy": 0.1495620235800743,
      "num_tokens": 3136970.0,
      "step": 1400
    },
    {
      "entropy": 6.170126581192017,
      "epoch": 0.1578031111360701,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999766145317046,
      "loss": 6.0804,
      "mean_token_accuracy": 0.15332209914922715,
      "num_tokens": 3148512.0,
      "step": 1405
    },
    {
      "entropy": 6.112670230865478,
      "epoch": 0.15836468804402762,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999760321118385,
      "loss": 6.1281,
      "mean_token_accuracy": 0.15459683984518052,
      "num_tokens": 3159477.0,
      "step": 1410
    },
    {
      "entropy": 6.104544401168823,
      "epoch": 0.15892626495198517,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999754425286483,
      "loss": 6.15,
      "mean_token_accuracy": 0.14965095221996308,
      "num_tokens": 3171361.0,
      "step": 1415
    },
    {
      "entropy": 6.267684602737427,
      "epoch": 0.15948784185994272,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999748457821529,
      "loss": 6.2915,
      "mean_token_accuracy": 0.1418513722717762,
      "num_tokens": 3184652.0,
      "step": 1420
    },
    {
      "entropy": 6.195340013504028,
      "epoch": 0.16004941876790027,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499974241872371,
      "loss": 6.1786,
      "mean_token_accuracy": 0.14732735827565194,
      "num_tokens": 3195557.0,
      "step": 1425
    },
    {
      "entropy": 6.219116926193237,
      "epoch": 0.16061099567585782,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999736307993221,
      "loss": 6.2129,
      "mean_token_accuracy": 0.1357981599867344,
      "num_tokens": 3207139.0,
      "step": 1430
    },
    {
      "entropy": 6.232678031921386,
      "epoch": 0.16117257258381534,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999730125630255,
      "loss": 6.2839,
      "mean_token_accuracy": 0.14023263975977898,
      "num_tokens": 3218967.0,
      "step": 1435
    },
    {
      "entropy": 6.271196174621582,
      "epoch": 0.1617341494917729,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999723871635008,
      "loss": 6.109,
      "mean_token_accuracy": 0.14901139065623284,
      "num_tokens": 3230168.0,
      "step": 1440
    },
    {
      "entropy": 6.015722990036011,
      "epoch": 0.16229572639973044,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999717546007681,
      "loss": 6.1545,
      "mean_token_accuracy": 0.14376315474510193,
      "num_tokens": 3241829.0,
      "step": 1445
    },
    {
      "entropy": 6.264339351654053,
      "epoch": 0.162857303307688,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999711148748476,
      "loss": 6.18,
      "mean_token_accuracy": 0.1445568896830082,
      "num_tokens": 3253195.0,
      "step": 1450
    },
    {
      "entropy": 6.123285531997681,
      "epoch": 0.16341888021564555,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999704679857594,
      "loss": 6.2042,
      "mean_token_accuracy": 0.13930154219269753,
      "num_tokens": 3264951.0,
      "step": 1455
    },
    {
      "entropy": 6.061741495132447,
      "epoch": 0.16398045712360307,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999698139335243,
      "loss": 6.0429,
      "mean_token_accuracy": 0.15055785626173018,
      "num_tokens": 3276054.0,
      "step": 1460
    },
    {
      "entropy": 6.117865514755249,
      "epoch": 0.16454203403156062,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999691527181632,
      "loss": 6.0213,
      "mean_token_accuracy": 0.15988054275512695,
      "num_tokens": 3285540.0,
      "step": 1465
    },
    {
      "entropy": 6.181283855438233,
      "epoch": 0.16510361093951817,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999684843396968,
      "loss": 6.1457,
      "mean_token_accuracy": 0.15024212151765823,
      "num_tokens": 3296869.0,
      "step": 1470
    },
    {
      "entropy": 6.154390048980713,
      "epoch": 0.16566518784747572,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999678087981467,
      "loss": 6.0919,
      "mean_token_accuracy": 0.15537651777267455,
      "num_tokens": 3307554.0,
      "step": 1475
    },
    {
      "entropy": 6.024777555465699,
      "epoch": 0.16622676475543324,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999671260935343,
      "loss": 6.0384,
      "mean_token_accuracy": 0.1448281340301037,
      "num_tokens": 3319685.0,
      "step": 1480
    },
    {
      "entropy": 6.090826416015625,
      "epoch": 0.1667883416633908,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999664362258814,
      "loss": 6.1633,
      "mean_token_accuracy": 0.1465357907116413,
      "num_tokens": 3330742.0,
      "step": 1485
    },
    {
      "entropy": 6.111087989807129,
      "epoch": 0.16734991857134834,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999657391952099,
      "loss": 6.169,
      "mean_token_accuracy": 0.1444841891527176,
      "num_tokens": 3343613.0,
      "step": 1490
    },
    {
      "entropy": 6.308396863937378,
      "epoch": 0.1679114954793059,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000499965035001542,
      "loss": 6.2148,
      "mean_token_accuracy": 0.15076154470443726,
      "num_tokens": 3355162.0,
      "step": 1495
    },
    {
      "entropy": 6.160010528564453,
      "epoch": 0.16847307238726345,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999643236449001,
      "loss": 6.1347,
      "mean_token_accuracy": 0.15359269082546234,
      "num_tokens": 3366324.0,
      "step": 1500
    },
    {
      "entropy": 6.283501052856446,
      "epoch": 0.16903464929522097,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999636051253069,
      "loss": 6.2882,
      "mean_token_accuracy": 0.14009708762168885,
      "num_tokens": 3378571.0,
      "step": 1505
    },
    {
      "entropy": 6.109706830978394,
      "epoch": 0.16959622620317852,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999628794427854,
      "loss": 6.168,
      "mean_token_accuracy": 0.14259269535541536,
      "num_tokens": 3390605.0,
      "step": 1510
    },
    {
      "entropy": 6.130779981613159,
      "epoch": 0.17015780311113607,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999621465973583,
      "loss": 6.0931,
      "mean_token_accuracy": 0.15212171971797944,
      "num_tokens": 3402735.0,
      "step": 1515
    },
    {
      "entropy": 6.165618658065796,
      "epoch": 0.17071938001909362,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999614065890495,
      "loss": 6.1082,
      "mean_token_accuracy": 0.15175931602716447,
      "num_tokens": 3413079.0,
      "step": 1520
    },
    {
      "entropy": 6.059429597854614,
      "epoch": 0.17128095692705117,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999606594178821,
      "loss": 6.037,
      "mean_token_accuracy": 0.15060681253671646,
      "num_tokens": 3423098.0,
      "step": 1525
    },
    {
      "entropy": 6.137203025817871,
      "epoch": 0.1718425338350087,
      "grad_norm": 0.90625,
      "learning_rate": 0.00049995990508388,
      "loss": 6.081,
      "mean_token_accuracy": 0.14952545762062072,
      "num_tokens": 3435794.0,
      "step": 1530
    },
    {
      "entropy": 6.122977447509766,
      "epoch": 0.17240411074296624,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999591435870674,
      "loss": 6.2098,
      "mean_token_accuracy": 0.14280164167284964,
      "num_tokens": 3448125.0,
      "step": 1535
    },
    {
      "entropy": 6.121371603012085,
      "epoch": 0.1729656876509238,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999583749274684,
      "loss": 6.0418,
      "mean_token_accuracy": 0.151399427652359,
      "num_tokens": 3458217.0,
      "step": 1540
    },
    {
      "entropy": 6.071384716033935,
      "epoch": 0.17352726455888134,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999575991051074,
      "loss": 6.0979,
      "mean_token_accuracy": 0.15133253037929534,
      "num_tokens": 3469553.0,
      "step": 1545
    },
    {
      "entropy": 6.166380977630615,
      "epoch": 0.1740888414668389,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999568161200093,
      "loss": 6.0439,
      "mean_token_accuracy": 0.1554693400859833,
      "num_tokens": 3480928.0,
      "step": 1550
    },
    {
      "entropy": 6.097128963470459,
      "epoch": 0.17465041837479642,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999560259721989,
      "loss": 6.1325,
      "mean_token_accuracy": 0.1535270929336548,
      "num_tokens": 3492074.0,
      "step": 1555
    },
    {
      "entropy": 6.1337090015411375,
      "epoch": 0.17521199528275397,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999552286617014,
      "loss": 6.1384,
      "mean_token_accuracy": 0.14494589790701867,
      "num_tokens": 3503320.0,
      "step": 1560
    },
    {
      "entropy": 6.057374572753906,
      "epoch": 0.17577357219071152,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004999544241885421,
      "loss": 6.1406,
      "mean_token_accuracy": 0.15341189801692962,
      "num_tokens": 3515466.0,
      "step": 1565
    },
    {
      "entropy": 6.160209703445434,
      "epoch": 0.17633514909866907,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999536125527467,
      "loss": 6.0176,
      "mean_token_accuracy": 0.1521401584148407,
      "num_tokens": 3525584.0,
      "step": 1570
    },
    {
      "entropy": 6.019997692108154,
      "epoch": 0.17689672600662662,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499952793754341,
      "loss": 6.1022,
      "mean_token_accuracy": 0.14772207736968995,
      "num_tokens": 3536973.0,
      "step": 1575
    },
    {
      "entropy": 6.149301910400391,
      "epoch": 0.17745830291458414,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999519677933512,
      "loss": 6.1208,
      "mean_token_accuracy": 0.14433789029717445,
      "num_tokens": 3547799.0,
      "step": 1580
    },
    {
      "entropy": 6.182972002029419,
      "epoch": 0.1780198798225417,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999511346698035,
      "loss": 6.1857,
      "mean_token_accuracy": 0.14982210397720336,
      "num_tokens": 3558436.0,
      "step": 1585
    },
    {
      "entropy": 6.146703910827637,
      "epoch": 0.17858145673049924,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999502943837244,
      "loss": 6.1605,
      "mean_token_accuracy": 0.14957652762532234,
      "num_tokens": 3569851.0,
      "step": 1590
    },
    {
      "entropy": 6.185033321380615,
      "epoch": 0.1791430336384568,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999494469351406,
      "loss": 6.0591,
      "mean_token_accuracy": 0.15206927061080933,
      "num_tokens": 3580797.0,
      "step": 1595
    },
    {
      "entropy": 6.020595216751099,
      "epoch": 0.17970461054641435,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999485923240793,
      "loss": 6.0549,
      "mean_token_accuracy": 0.15080841779708862,
      "num_tokens": 3592552.0,
      "step": 1600
    },
    {
      "entropy": 6.074697542190552,
      "epoch": 0.18026618745437187,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999477305505675,
      "loss": 6.0382,
      "mean_token_accuracy": 0.15332526564598084,
      "num_tokens": 3603932.0,
      "step": 1605
    },
    {
      "entropy": 6.0842041015625,
      "epoch": 0.18082776436232942,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999468616146328,
      "loss": 6.0442,
      "mean_token_accuracy": 0.16077973172068596,
      "num_tokens": 3614865.0,
      "step": 1610
    },
    {
      "entropy": 5.924830675125122,
      "epoch": 0.18138934127028697,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999459855163026,
      "loss": 5.9775,
      "mean_token_accuracy": 0.15816942751407623,
      "num_tokens": 3626275.0,
      "step": 1615
    },
    {
      "entropy": 6.170416784286499,
      "epoch": 0.18195091817824452,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499945102255605,
      "loss": 6.1048,
      "mean_token_accuracy": 0.14498920738697052,
      "num_tokens": 3638416.0,
      "step": 1620
    },
    {
      "entropy": 5.96520586013794,
      "epoch": 0.18251249508620204,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999442118325682,
      "loss": 5.9654,
      "mean_token_accuracy": 0.1536343052983284,
      "num_tokens": 3649651.0,
      "step": 1625
    },
    {
      "entropy": 6.211764001846314,
      "epoch": 0.1830740719941596,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999433142472205,
      "loss": 6.1963,
      "mean_token_accuracy": 0.15030549317598343,
      "num_tokens": 3661172.0,
      "step": 1630
    },
    {
      "entropy": 6.097804975509644,
      "epoch": 0.18363564890211714,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999424094995902,
      "loss": 6.0371,
      "mean_token_accuracy": 0.15204803496599198,
      "num_tokens": 3672086.0,
      "step": 1635
    },
    {
      "entropy": 6.043741655349732,
      "epoch": 0.1841972258100747,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999414975897066,
      "loss": 6.0105,
      "mean_token_accuracy": 0.1487807512283325,
      "num_tokens": 3682579.0,
      "step": 1640
    },
    {
      "entropy": 6.1238837242126465,
      "epoch": 0.18475880271803224,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999405785175983,
      "loss": 6.1187,
      "mean_token_accuracy": 0.1473942831158638,
      "num_tokens": 3694616.0,
      "step": 1645
    },
    {
      "entropy": 5.982372426986695,
      "epoch": 0.18532037962598977,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999396522832949,
      "loss": 5.919,
      "mean_token_accuracy": 0.15665805488824844,
      "num_tokens": 3706821.0,
      "step": 1650
    },
    {
      "entropy": 6.029320240020752,
      "epoch": 0.18588195653394732,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999387188868256,
      "loss": 5.9732,
      "mean_token_accuracy": 0.1571487694978714,
      "num_tokens": 3717552.0,
      "step": 1655
    },
    {
      "entropy": 6.036300897598267,
      "epoch": 0.18644353344190487,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999377783282203,
      "loss": 6.0474,
      "mean_token_accuracy": 0.14972800016403198,
      "num_tokens": 3728300.0,
      "step": 1660
    },
    {
      "entropy": 6.082576656341553,
      "epoch": 0.18700511034986242,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999368306075089,
      "loss": 6.0455,
      "mean_token_accuracy": 0.16109108030796052,
      "num_tokens": 3738545.0,
      "step": 1665
    },
    {
      "entropy": 6.077401065826416,
      "epoch": 0.18756668725781997,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999358757247215,
      "loss": 6.022,
      "mean_token_accuracy": 0.15040296018123628,
      "num_tokens": 3749190.0,
      "step": 1670
    },
    {
      "entropy": 6.0228570938110355,
      "epoch": 0.1881282641657775,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999349136798886,
      "loss": 6.0348,
      "mean_token_accuracy": 0.15334153771400452,
      "num_tokens": 3760978.0,
      "step": 1675
    },
    {
      "entropy": 5.990633153915406,
      "epoch": 0.18868984107373504,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999339444730409,
      "loss": 5.9467,
      "mean_token_accuracy": 0.1572801098227501,
      "num_tokens": 3771339.0,
      "step": 1680
    },
    {
      "entropy": 6.005237627029419,
      "epoch": 0.1892514179816926,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999329681042092,
      "loss": 5.9206,
      "mean_token_accuracy": 0.15659749656915664,
      "num_tokens": 3781550.0,
      "step": 1685
    },
    {
      "entropy": 6.061115407943726,
      "epoch": 0.18981299488965014,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999319845734243,
      "loss": 6.1359,
      "mean_token_accuracy": 0.14560328647494317,
      "num_tokens": 3793700.0,
      "step": 1690
    },
    {
      "entropy": 6.180884456634521,
      "epoch": 0.1903745717976077,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999309938807181,
      "loss": 6.1379,
      "mean_token_accuracy": 0.14918680638074874,
      "num_tokens": 3806114.0,
      "step": 1695
    },
    {
      "entropy": 6.095144414901734,
      "epoch": 0.19093614870556522,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999299960261216,
      "loss": 6.0594,
      "mean_token_accuracy": 0.1524960294365883,
      "num_tokens": 3817251.0,
      "step": 1700
    },
    {
      "entropy": 6.046061992645264,
      "epoch": 0.19149772561352277,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999289910096669,
      "loss": 6.055,
      "mean_token_accuracy": 0.15111273676156997,
      "num_tokens": 3828544.0,
      "step": 1705
    },
    {
      "entropy": 6.141667985916138,
      "epoch": 0.19205930252148032,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999279788313858,
      "loss": 6.1126,
      "mean_token_accuracy": 0.14861491471529006,
      "num_tokens": 3840423.0,
      "step": 1710
    },
    {
      "entropy": 6.044348049163818,
      "epoch": 0.19262087942943787,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999269594913107,
      "loss": 6.0097,
      "mean_token_accuracy": 0.15253075659275056,
      "num_tokens": 3852751.0,
      "step": 1715
    },
    {
      "entropy": 5.987351751327514,
      "epoch": 0.19318245633739542,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999259329894739,
      "loss": 6.0073,
      "mean_token_accuracy": 0.15642097666859628,
      "num_tokens": 3863382.0,
      "step": 1720
    },
    {
      "entropy": 6.021144962310791,
      "epoch": 0.19374403324535294,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999248993259081,
      "loss": 6.0064,
      "mean_token_accuracy": 0.15451697632670403,
      "num_tokens": 3874294.0,
      "step": 1725
    },
    {
      "entropy": 6.07633786201477,
      "epoch": 0.1943056101533105,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999238585006463,
      "loss": 6.0426,
      "mean_token_accuracy": 0.14845188707113266,
      "num_tokens": 3886693.0,
      "step": 1730
    },
    {
      "entropy": 6.009305143356324,
      "epoch": 0.19486718706126804,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999228105137218,
      "loss": 5.9498,
      "mean_token_accuracy": 0.15883688777685165,
      "num_tokens": 3897755.0,
      "step": 1735
    },
    {
      "entropy": 6.021089506149292,
      "epoch": 0.1954287639692256,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999217553651676,
      "loss": 5.9524,
      "mean_token_accuracy": 0.15532039254903793,
      "num_tokens": 3909418.0,
      "step": 1740
    },
    {
      "entropy": 6.1032227039337155,
      "epoch": 0.19599034087718312,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999206930550175,
      "loss": 6.1072,
      "mean_token_accuracy": 0.1495990239083767,
      "num_tokens": 3919992.0,
      "step": 1745
    },
    {
      "entropy": 6.063982629776001,
      "epoch": 0.19655191778514067,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999196235833054,
      "loss": 5.9716,
      "mean_token_accuracy": 0.1587469384074211,
      "num_tokens": 3930648.0,
      "step": 1750
    },
    {
      "entropy": 5.948443031311035,
      "epoch": 0.19711349469309822,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999185469500651,
      "loss": 5.9572,
      "mean_token_accuracy": 0.1576173797249794,
      "num_tokens": 3941883.0,
      "step": 1755
    },
    {
      "entropy": 5.984550046920776,
      "epoch": 0.19767507160105577,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499917463155331,
      "loss": 6.004,
      "mean_token_accuracy": 0.15634143352508545,
      "num_tokens": 3952005.0,
      "step": 1760
    },
    {
      "entropy": 6.034853076934814,
      "epoch": 0.19823664850901332,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999163721991378,
      "loss": 5.9225,
      "mean_token_accuracy": 0.15769157111644744,
      "num_tokens": 3963720.0,
      "step": 1765
    },
    {
      "entropy": 5.936740970611572,
      "epoch": 0.19879822541697084,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00049991527408152,
      "loss": 6.0118,
      "mean_token_accuracy": 0.15719937086105346,
      "num_tokens": 3974998.0,
      "step": 1770
    },
    {
      "entropy": 6.06788649559021,
      "epoch": 0.1993598023249284,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999141688025127,
      "loss": 6.0574,
      "mean_token_accuracy": 0.14897572845220566,
      "num_tokens": 3986469.0,
      "step": 1775
    },
    {
      "entropy": 6.033199405670166,
      "epoch": 0.19992137923288594,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000499913056362151,
      "loss": 5.876,
      "mean_token_accuracy": 0.15733214020729064,
      "num_tokens": 3997553.0,
      "step": 1780
    },
    {
      "entropy": 5.917162466049194,
      "epoch": 0.2004829561408435,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999119367604703,
      "loss": 5.9558,
      "mean_token_accuracy": 0.15673445910215378,
      "num_tokens": 4008613.0,
      "step": 1785
    },
    {
      "entropy": 6.153360414505005,
      "epoch": 0.20104453304880104,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999108099975064,
      "loss": 6.1283,
      "mean_token_accuracy": 0.1536238044500351,
      "num_tokens": 4019860.0,
      "step": 1790
    },
    {
      "entropy": 6.023882675170898,
      "epoch": 0.20160610995675857,
      "grad_norm": 1.0,
      "learning_rate": 0.000499909676073295,
      "loss": 6.017,
      "mean_token_accuracy": 0.15675632730126382,
      "num_tokens": 4030602.0,
      "step": 1795
    },
    {
      "entropy": 5.985197401046753,
      "epoch": 0.20216768686471612,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999085349878723,
      "loss": 5.8936,
      "mean_token_accuracy": 0.15248199701309204,
      "num_tokens": 4041878.0,
      "step": 1800
    },
    {
      "entropy": 5.950754308700562,
      "epoch": 0.20272926377267367,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999073867412746,
      "loss": 5.9882,
      "mean_token_accuracy": 0.1526031345129013,
      "num_tokens": 4052799.0,
      "step": 1805
    },
    {
      "entropy": 6.037546443939209,
      "epoch": 0.20329084068063122,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999062313335385,
      "loss": 5.9902,
      "mean_token_accuracy": 0.1529676452279091,
      "num_tokens": 4064339.0,
      "step": 1810
    },
    {
      "entropy": 5.8854522705078125,
      "epoch": 0.20385241758858877,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999050687647008,
      "loss": 5.9338,
      "mean_token_accuracy": 0.15455373972654343,
      "num_tokens": 4076246.0,
      "step": 1815
    },
    {
      "entropy": 5.971135139465332,
      "epoch": 0.2044139944965463,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999038990347984,
      "loss": 6.019,
      "mean_token_accuracy": 0.1492105543613434,
      "num_tokens": 4088006.0,
      "step": 1820
    },
    {
      "entropy": 6.113638019561767,
      "epoch": 0.20497557140450384,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999027221438688,
      "loss": 6.0213,
      "mean_token_accuracy": 0.1529981389641762,
      "num_tokens": 4098122.0,
      "step": 1825
    },
    {
      "entropy": 6.144082927703858,
      "epoch": 0.2055371483124614,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499901538091949,
      "loss": 6.2063,
      "mean_token_accuracy": 0.1428961783647537,
      "num_tokens": 4109958.0,
      "step": 1830
    },
    {
      "entropy": 6.17197060585022,
      "epoch": 0.20609872522041894,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999003468790773,
      "loss": 6.0981,
      "mean_token_accuracy": 0.14817476719617845,
      "num_tokens": 4121262.0,
      "step": 1835
    },
    {
      "entropy": 5.936773729324341,
      "epoch": 0.2066603021283765,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998991485052911,
      "loss": 5.8681,
      "mean_token_accuracy": 0.16058467477560043,
      "num_tokens": 4130958.0,
      "step": 1840
    },
    {
      "entropy": 6.0173663139343265,
      "epoch": 0.20722187903633402,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499897942970629,
      "loss": 6.0781,
      "mean_token_accuracy": 0.14642804861068726,
      "num_tokens": 4143779.0,
      "step": 1845
    },
    {
      "entropy": 6.1406467914581295,
      "epoch": 0.20778345594429157,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499896730275129,
      "loss": 6.1001,
      "mean_token_accuracy": 0.14446378648281097,
      "num_tokens": 4155367.0,
      "step": 1850
    },
    {
      "entropy": 5.987906074523925,
      "epoch": 0.20834503285224912,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00049989551041883,
      "loss": 5.912,
      "mean_token_accuracy": 0.15731868371367455,
      "num_tokens": 4166657.0,
      "step": 1855
    },
    {
      "entropy": 6.03128080368042,
      "epoch": 0.20890660976020667,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998942834017707,
      "loss": 6.0393,
      "mean_token_accuracy": 0.14800393357872962,
      "num_tokens": 4177517.0,
      "step": 1860
    },
    {
      "entropy": 5.982046890258789,
      "epoch": 0.2094681866681642,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998930492239901,
      "loss": 5.894,
      "mean_token_accuracy": 0.15527052879333497,
      "num_tokens": 4188182.0,
      "step": 1865
    },
    {
      "entropy": 6.043006801605225,
      "epoch": 0.21002976357612174,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998918078855277,
      "loss": 6.0166,
      "mean_token_accuracy": 0.15780879855155944,
      "num_tokens": 4199415.0,
      "step": 1870
    },
    {
      "entropy": 5.954761743545532,
      "epoch": 0.2105913404840793,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998905593864227,
      "loss": 5.9667,
      "mean_token_accuracy": 0.15191469490528106,
      "num_tokens": 4210025.0,
      "step": 1875
    },
    {
      "entropy": 6.013121223449707,
      "epoch": 0.21115291739203684,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998893037267153,
      "loss": 5.8829,
      "mean_token_accuracy": 0.15989519357681276,
      "num_tokens": 4219920.0,
      "step": 1880
    },
    {
      "entropy": 5.9762413024902346,
      "epoch": 0.2117144942999944,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998880409064452,
      "loss": 6.0603,
      "mean_token_accuracy": 0.15269239693880082,
      "num_tokens": 4232647.0,
      "step": 1885
    },
    {
      "entropy": 6.130297899246216,
      "epoch": 0.21227607120795192,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004998867709256527,
      "loss": 5.9489,
      "mean_token_accuracy": 0.1568835824728012,
      "num_tokens": 4242637.0,
      "step": 1890
    },
    {
      "entropy": 5.8553040504455565,
      "epoch": 0.21283764811590947,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499885493784378,
      "loss": 5.9229,
      "mean_token_accuracy": 0.15738577097654344,
      "num_tokens": 4253633.0,
      "step": 1895
    },
    {
      "entropy": 6.022933912277222,
      "epoch": 0.21339922502386702,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998842094826621,
      "loss": 5.8827,
      "mean_token_accuracy": 0.15891942679882048,
      "num_tokens": 4264472.0,
      "step": 1900
    },
    {
      "entropy": 5.982009840011597,
      "epoch": 0.21396080193182457,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998829180205458,
      "loss": 6.0074,
      "mean_token_accuracy": 0.15923700332641602,
      "num_tokens": 4275233.0,
      "step": 1905
    },
    {
      "entropy": 6.03095121383667,
      "epoch": 0.21452237883978212,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998816193980701,
      "loss": 5.9611,
      "mean_token_accuracy": 0.15337118357419968,
      "num_tokens": 4287063.0,
      "step": 1910
    },
    {
      "entropy": 5.989390468597412,
      "epoch": 0.21508395574773964,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998803136152764,
      "loss": 5.9785,
      "mean_token_accuracy": 0.14854297190904617,
      "num_tokens": 4298761.0,
      "step": 1915
    },
    {
      "entropy": 5.828275489807129,
      "epoch": 0.2156455326556972,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998790006722063,
      "loss": 5.8355,
      "mean_token_accuracy": 0.16531981974840165,
      "num_tokens": 4309650.0,
      "step": 1920
    },
    {
      "entropy": 6.0078376770019535,
      "epoch": 0.21620710956365474,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998776805689016,
      "loss": 5.9047,
      "mean_token_accuracy": 0.16718189269304276,
      "num_tokens": 4320403.0,
      "step": 1925
    },
    {
      "entropy": 5.839940357208252,
      "epoch": 0.2167686864716123,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998763533054045,
      "loss": 5.9284,
      "mean_token_accuracy": 0.1600561797618866,
      "num_tokens": 4332049.0,
      "step": 1930
    },
    {
      "entropy": 5.9344322681427,
      "epoch": 0.21733026337956984,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998750188817568,
      "loss": 5.8379,
      "mean_token_accuracy": 0.1621323823928833,
      "num_tokens": 4342551.0,
      "step": 1935
    },
    {
      "entropy": 5.97367901802063,
      "epoch": 0.21789184028752737,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998736772980015,
      "loss": 5.9608,
      "mean_token_accuracy": 0.16004323810338975,
      "num_tokens": 4355156.0,
      "step": 1940
    },
    {
      "entropy": 5.941838502883911,
      "epoch": 0.21845341719548492,
      "grad_norm": 1.0,
      "learning_rate": 0.000499872328554181,
      "loss": 6.0201,
      "mean_token_accuracy": 0.15124871134757994,
      "num_tokens": 4367402.0,
      "step": 1945
    },
    {
      "entropy": 6.041738891601563,
      "epoch": 0.21901499410344247,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998709726503383,
      "loss": 5.9018,
      "mean_token_accuracy": 0.16085198670625686,
      "num_tokens": 4378687.0,
      "step": 1950
    },
    {
      "entropy": 6.0037284851074215,
      "epoch": 0.21957657101140002,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998696095865168,
      "loss": 6.0324,
      "mean_token_accuracy": 0.16538831740617752,
      "num_tokens": 4388871.0,
      "step": 1955
    },
    {
      "entropy": 5.935435390472412,
      "epoch": 0.22013814791935757,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998682393627595,
      "loss": 5.861,
      "mean_token_accuracy": 0.1636255994439125,
      "num_tokens": 4399440.0,
      "step": 1960
    },
    {
      "entropy": 5.838962459564209,
      "epoch": 0.2206997248273151,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998668619791104,
      "loss": 5.8651,
      "mean_token_accuracy": 0.16348468512296677,
      "num_tokens": 4409989.0,
      "step": 1965
    },
    {
      "entropy": 6.090083599090576,
      "epoch": 0.22126130173527264,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998654774356131,
      "loss": 6.0124,
      "mean_token_accuracy": 0.1552437886595726,
      "num_tokens": 4422149.0,
      "step": 1970
    },
    {
      "entropy": 6.077429008483887,
      "epoch": 0.2218228786432302,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998640857323118,
      "loss": 6.0456,
      "mean_token_accuracy": 0.15136015340685843,
      "num_tokens": 4433157.0,
      "step": 1975
    },
    {
      "entropy": 6.002451324462891,
      "epoch": 0.22238445555118774,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004998626868692509,
      "loss": 5.9592,
      "mean_token_accuracy": 0.15625503659248352,
      "num_tokens": 4444595.0,
      "step": 1980
    },
    {
      "entropy": 6.024257373809815,
      "epoch": 0.2229460324591453,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998612808464748,
      "loss": 6.1028,
      "mean_token_accuracy": 0.15059059411287307,
      "num_tokens": 4456592.0,
      "step": 1985
    },
    {
      "entropy": 6.03296799659729,
      "epoch": 0.22350760936710282,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998598676640283,
      "loss": 5.867,
      "mean_token_accuracy": 0.17013804167509078,
      "num_tokens": 4467624.0,
      "step": 1990
    },
    {
      "entropy": 5.895969343185425,
      "epoch": 0.22406918627506037,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998584473219563,
      "loss": 6.052,
      "mean_token_accuracy": 0.15197935700416565,
      "num_tokens": 4479766.0,
      "step": 1995
    },
    {
      "entropy": 6.083270072937012,
      "epoch": 0.22463076318301792,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998570198203043,
      "loss": 5.9056,
      "mean_token_accuracy": 0.1571629002690315,
      "num_tokens": 4490023.0,
      "step": 2000
    },
    {
      "entropy": 5.897563028335571,
      "epoch": 0.22519234009097547,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998555851591175,
      "loss": 5.9511,
      "mean_token_accuracy": 0.15516175776720048,
      "num_tokens": 4500651.0,
      "step": 2005
    },
    {
      "entropy": 6.082723474502563,
      "epoch": 0.225753916998933,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998541433384416,
      "loss": 6.0535,
      "mean_token_accuracy": 0.15102613866329193,
      "num_tokens": 4512934.0,
      "step": 2010
    },
    {
      "entropy": 5.9272867202758786,
      "epoch": 0.22631549390689054,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998526943583226,
      "loss": 5.9453,
      "mean_token_accuracy": 0.15935770124197007,
      "num_tokens": 4524251.0,
      "step": 2015
    },
    {
      "entropy": 6.003200340270996,
      "epoch": 0.2268770708148481,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998512382188066,
      "loss": 6.0203,
      "mean_token_accuracy": 0.1512477733194828,
      "num_tokens": 4536025.0,
      "step": 2020
    },
    {
      "entropy": 6.054232501983643,
      "epoch": 0.22743864772280564,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998497749199401,
      "loss": 5.9695,
      "mean_token_accuracy": 0.15859290361404418,
      "num_tokens": 4546717.0,
      "step": 2025
    },
    {
      "entropy": 5.903731346130371,
      "epoch": 0.2280002246307632,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998483044617694,
      "loss": 5.8621,
      "mean_token_accuracy": 0.16061438173055648,
      "num_tokens": 4557206.0,
      "step": 2030
    },
    {
      "entropy": 5.900349473953247,
      "epoch": 0.22856180153872072,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998468268443416,
      "loss": 5.8563,
      "mean_token_accuracy": 0.16150960475206375,
      "num_tokens": 4568123.0,
      "step": 2035
    },
    {
      "entropy": 5.986484718322754,
      "epoch": 0.22912337844667827,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998453420677034,
      "loss": 5.9679,
      "mean_token_accuracy": 0.15943958312273027,
      "num_tokens": 4579385.0,
      "step": 2040
    },
    {
      "entropy": 6.026423215866089,
      "epoch": 0.22968495535463582,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998438501319025,
      "loss": 6.0272,
      "mean_token_accuracy": 0.15108214914798737,
      "num_tokens": 4590514.0,
      "step": 2045
    },
    {
      "entropy": 5.957902145385742,
      "epoch": 0.23024653226259337,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998423510369862,
      "loss": 5.9695,
      "mean_token_accuracy": 0.1499602884054184,
      "num_tokens": 4601624.0,
      "step": 2050
    },
    {
      "entropy": 6.06238203048706,
      "epoch": 0.23080810917055092,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998408447830022,
      "loss": 5.9435,
      "mean_token_accuracy": 0.1589498057961464,
      "num_tokens": 4612822.0,
      "step": 2055
    },
    {
      "entropy": 5.938096332550049,
      "epoch": 0.23136968607850844,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998393313699985,
      "loss": 5.9516,
      "mean_token_accuracy": 0.16269320100545884,
      "num_tokens": 4623443.0,
      "step": 2060
    },
    {
      "entropy": 5.958241176605225,
      "epoch": 0.231931262986466,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998378107980233,
      "loss": 5.8527,
      "mean_token_accuracy": 0.16247376054525375,
      "num_tokens": 4633600.0,
      "step": 2065
    },
    {
      "entropy": 5.959673881530762,
      "epoch": 0.23249283989442354,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998362830671249,
      "loss": 6.0057,
      "mean_token_accuracy": 0.15157989487051965,
      "num_tokens": 4644217.0,
      "step": 2070
    },
    {
      "entropy": 6.015304946899414,
      "epoch": 0.2330544168023811,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998347481773522,
      "loss": 6.0174,
      "mean_token_accuracy": 0.15409978926181794,
      "num_tokens": 4654671.0,
      "step": 2075
    },
    {
      "entropy": 6.001746320724488,
      "epoch": 0.23361599371033864,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998332061287538,
      "loss": 6.0231,
      "mean_token_accuracy": 0.15379573553800582,
      "num_tokens": 4665920.0,
      "step": 2080
    },
    {
      "entropy": 5.931762742996216,
      "epoch": 0.23417757061829617,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000499831656921379,
      "loss": 5.9091,
      "mean_token_accuracy": 0.15680453628301622,
      "num_tokens": 4675406.0,
      "step": 2085
    },
    {
      "entropy": 6.051575613021851,
      "epoch": 0.23473914752625372,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499830100555277,
      "loss": 5.9274,
      "mean_token_accuracy": 0.16170108020305635,
      "num_tokens": 4686870.0,
      "step": 2090
    },
    {
      "entropy": 5.986575174331665,
      "epoch": 0.23530072443421127,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998285370304974,
      "loss": 6.0236,
      "mean_token_accuracy": 0.15555590838193895,
      "num_tokens": 4698926.0,
      "step": 2095
    },
    {
      "entropy": 6.053481721878052,
      "epoch": 0.23586230134216882,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00049982696634709,
      "loss": 6.0906,
      "mean_token_accuracy": 0.1449894279241562,
      "num_tokens": 4709318.0,
      "step": 2100
    },
    {
      "entropy": 5.874515628814697,
      "epoch": 0.23642387825012637,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998253885051048,
      "loss": 5.7865,
      "mean_token_accuracy": 0.17088475227355956,
      "num_tokens": 4718819.0,
      "step": 2105
    },
    {
      "entropy": 5.872879409790039,
      "epoch": 0.2369854551580839,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499823803504592,
      "loss": 5.8319,
      "mean_token_accuracy": 0.165112866461277,
      "num_tokens": 4729513.0,
      "step": 2110
    },
    {
      "entropy": 5.943530511856079,
      "epoch": 0.23754703206604144,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998222113456022,
      "loss": 5.935,
      "mean_token_accuracy": 0.15231743901968003,
      "num_tokens": 4739875.0,
      "step": 2115
    },
    {
      "entropy": 6.088411569595337,
      "epoch": 0.238108608973999,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998206120281858,
      "loss": 6.0438,
      "mean_token_accuracy": 0.15248556584119796,
      "num_tokens": 4752933.0,
      "step": 2120
    },
    {
      "entropy": 5.917099142074585,
      "epoch": 0.23867018588195654,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499819005552394,
      "loss": 5.84,
      "mean_token_accuracy": 0.16093057990074158,
      "num_tokens": 4762969.0,
      "step": 2125
    },
    {
      "entropy": 5.932834434509277,
      "epoch": 0.23923176278991407,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998173919182779,
      "loss": 5.9482,
      "mean_token_accuracy": 0.15672307163476945,
      "num_tokens": 4773495.0,
      "step": 2130
    },
    {
      "entropy": 5.969760227203369,
      "epoch": 0.23979333969787162,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499815771125889,
      "loss": 5.9697,
      "mean_token_accuracy": 0.1630027860403061,
      "num_tokens": 4785335.0,
      "step": 2135
    },
    {
      "entropy": 6.086810684204101,
      "epoch": 0.24035491660582917,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998141431752785,
      "loss": 6.118,
      "mean_token_accuracy": 0.15253835916519165,
      "num_tokens": 4798068.0,
      "step": 2140
    },
    {
      "entropy": 6.0190082550048825,
      "epoch": 0.24091649351378672,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998125080664987,
      "loss": 5.9132,
      "mean_token_accuracy": 0.16539788246154785,
      "num_tokens": 4809338.0,
      "step": 2145
    },
    {
      "entropy": 5.937499141693115,
      "epoch": 0.24147807042174427,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998108657996012,
      "loss": 5.9263,
      "mean_token_accuracy": 0.15961964651942254,
      "num_tokens": 4819636.0,
      "step": 2150
    },
    {
      "entropy": 5.9000465869903564,
      "epoch": 0.2420396473297018,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998092163746387,
      "loss": 5.8958,
      "mean_token_accuracy": 0.15810615718364715,
      "num_tokens": 4830975.0,
      "step": 2155
    },
    {
      "entropy": 5.922979879379272,
      "epoch": 0.24260122423765934,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998075597916636,
      "loss": 5.9436,
      "mean_token_accuracy": 0.15569430440664292,
      "num_tokens": 4841951.0,
      "step": 2160
    },
    {
      "entropy": 6.041098976135254,
      "epoch": 0.2431628011456169,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998058960507286,
      "loss": 5.9101,
      "mean_token_accuracy": 0.16229601576924324,
      "num_tokens": 4852919.0,
      "step": 2165
    },
    {
      "entropy": 5.922900772094726,
      "epoch": 0.24372437805357444,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998042251518866,
      "loss": 5.9418,
      "mean_token_accuracy": 0.15963752567768097,
      "num_tokens": 4863433.0,
      "step": 2170
    },
    {
      "entropy": 5.997221231460571,
      "epoch": 0.244285954961532,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998025470951908,
      "loss": 6.0079,
      "mean_token_accuracy": 0.15224221497774124,
      "num_tokens": 4875685.0,
      "step": 2175
    },
    {
      "entropy": 5.946529865264893,
      "epoch": 0.24484753186948952,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998008618806949,
      "loss": 5.9148,
      "mean_token_accuracy": 0.1580193430185318,
      "num_tokens": 4886347.0,
      "step": 2180
    },
    {
      "entropy": 5.936647653579712,
      "epoch": 0.24540910877744707,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997991695084523,
      "loss": 5.9116,
      "mean_token_accuracy": 0.16283671855926513,
      "num_tokens": 4896898.0,
      "step": 2185
    },
    {
      "entropy": 5.9814536571502686,
      "epoch": 0.24597068568540462,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997974699785169,
      "loss": 5.9969,
      "mean_token_accuracy": 0.1565718963742256,
      "num_tokens": 4908892.0,
      "step": 2190
    },
    {
      "entropy": 5.995170927047729,
      "epoch": 0.24653226259336217,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997957632909429,
      "loss": 5.953,
      "mean_token_accuracy": 0.15219104290008545,
      "num_tokens": 4920855.0,
      "step": 2195
    },
    {
      "entropy": 5.939112997055053,
      "epoch": 0.24709383950131972,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997940494457846,
      "loss": 5.9266,
      "mean_token_accuracy": 0.1611930876970291,
      "num_tokens": 4932857.0,
      "step": 2200
    },
    {
      "entropy": 5.920130109786987,
      "epoch": 0.24765541640927724,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997923284430967,
      "loss": 5.9635,
      "mean_token_accuracy": 0.157567498087883,
      "num_tokens": 4944063.0,
      "step": 2205
    },
    {
      "entropy": 5.916885423660278,
      "epoch": 0.2482169933172348,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997906002829338,
      "loss": 5.8422,
      "mean_token_accuracy": 0.16268620938062667,
      "num_tokens": 4955126.0,
      "step": 2210
    },
    {
      "entropy": 6.011972093582154,
      "epoch": 0.24877857022519234,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499788864965351,
      "loss": 6.0297,
      "mean_token_accuracy": 0.1572894498705864,
      "num_tokens": 4965738.0,
      "step": 2215
    },
    {
      "entropy": 6.104279661178589,
      "epoch": 0.2493401471331499,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997871224904036,
      "loss": 6.0573,
      "mean_token_accuracy": 0.1530898168683052,
      "num_tokens": 4976963.0,
      "step": 2220
    },
    {
      "entropy": 5.95416898727417,
      "epoch": 0.24990172404110744,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997853728581468,
      "loss": 5.9342,
      "mean_token_accuracy": 0.15562186539173126,
      "num_tokens": 4987577.0,
      "step": 2225
    },
    {
      "entropy": 5.972849702835083,
      "epoch": 0.250463300949065,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004997836160686368,
      "loss": 6.0588,
      "mean_token_accuracy": 0.15612355023622512,
      "num_tokens": 5000234.0,
      "step": 2230
    },
    {
      "entropy": 6.0028489112854,
      "epoch": 0.25102487785702254,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997818521219291,
      "loss": 5.8785,
      "mean_token_accuracy": 0.1534307286143303,
      "num_tokens": 5011652.0,
      "step": 2235
    },
    {
      "entropy": 5.960933828353882,
      "epoch": 0.25158645476498004,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997800810180803,
      "loss": 5.924,
      "mean_token_accuracy": 0.15806746035814284,
      "num_tokens": 5023603.0,
      "step": 2240
    },
    {
      "entropy": 5.890470409393311,
      "epoch": 0.2521480316729376,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997783027571464,
      "loss": 5.9027,
      "mean_token_accuracy": 0.16501567810773848,
      "num_tokens": 5034427.0,
      "step": 2245
    },
    {
      "entropy": 5.9391638278961185,
      "epoch": 0.25270960858089514,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997765173391841,
      "loss": 5.8942,
      "mean_token_accuracy": 0.16304780095815657,
      "num_tokens": 5045738.0,
      "step": 2250
    },
    {
      "entropy": 5.828226089477539,
      "epoch": 0.2532711854888527,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997747247642505,
      "loss": 5.8315,
      "mean_token_accuracy": 0.15862570255994796,
      "num_tokens": 5057021.0,
      "step": 2255
    },
    {
      "entropy": 5.92571177482605,
      "epoch": 0.25383276239681024,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997729250324023,
      "loss": 5.9125,
      "mean_token_accuracy": 0.1542244702577591,
      "num_tokens": 5067676.0,
      "step": 2260
    },
    {
      "entropy": 5.909444570541382,
      "epoch": 0.2543943393047678,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997711181436971,
      "loss": 5.8705,
      "mean_token_accuracy": 0.16087970584630967,
      "num_tokens": 5079422.0,
      "step": 2265
    },
    {
      "entropy": 5.894988298416138,
      "epoch": 0.25495591621272534,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997693040981922,
      "loss": 5.8818,
      "mean_token_accuracy": 0.160662043094635,
      "num_tokens": 5091242.0,
      "step": 2270
    },
    {
      "entropy": 5.955714464187622,
      "epoch": 0.2555174931206829,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997674828959456,
      "loss": 5.8787,
      "mean_token_accuracy": 0.15537350177764891,
      "num_tokens": 5102251.0,
      "step": 2275
    },
    {
      "entropy": 5.890719270706176,
      "epoch": 0.25607907002864044,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004997656545370151,
      "loss": 5.9183,
      "mean_token_accuracy": 0.16142252534627916,
      "num_tokens": 5113276.0,
      "step": 2280
    },
    {
      "entropy": 5.97783522605896,
      "epoch": 0.256640646936598,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499763819021459,
      "loss": 5.9129,
      "mean_token_accuracy": 0.15958033353090287,
      "num_tokens": 5124472.0,
      "step": 2285
    },
    {
      "entropy": 5.908177518844605,
      "epoch": 0.2572022238445555,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997619763493357,
      "loss": 5.86,
      "mean_token_accuracy": 0.1621762201189995,
      "num_tokens": 5134896.0,
      "step": 2290
    },
    {
      "entropy": 5.960345888137818,
      "epoch": 0.25776380075251304,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499760126520704,
      "loss": 5.9422,
      "mean_token_accuracy": 0.16326863765716554,
      "num_tokens": 5145115.0,
      "step": 2295
    },
    {
      "entropy": 5.935664415359497,
      "epoch": 0.2583253776604706,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997582695356225,
      "loss": 5.8934,
      "mean_token_accuracy": 0.16419054567813873,
      "num_tokens": 5155620.0,
      "step": 2300
    },
    {
      "entropy": 6.026849174499512,
      "epoch": 0.25888695456842814,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997564053941507,
      "loss": 5.9954,
      "mean_token_accuracy": 0.15510585457086562,
      "num_tokens": 5166260.0,
      "step": 2305
    },
    {
      "entropy": 5.943055200576782,
      "epoch": 0.2594485314763857,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997545340963478,
      "loss": 5.8667,
      "mean_token_accuracy": 0.15928099900484086,
      "num_tokens": 5177143.0,
      "step": 2310
    },
    {
      "entropy": 5.971165609359741,
      "epoch": 0.26001010838434324,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997526556422733,
      "loss": 6.0204,
      "mean_token_accuracy": 0.15262550860643387,
      "num_tokens": 5187795.0,
      "step": 2315
    },
    {
      "entropy": 5.947878408432007,
      "epoch": 0.2605716852923008,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004997507700319872,
      "loss": 5.8785,
      "mean_token_accuracy": 0.16017002314329148,
      "num_tokens": 5199365.0,
      "step": 2320
    },
    {
      "entropy": 5.992268705368042,
      "epoch": 0.26113326220025834,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004997488772655492,
      "loss": 6.097,
      "mean_token_accuracy": 0.1522466480731964,
      "num_tokens": 5212264.0,
      "step": 2325
    },
    {
      "entropy": 6.11021203994751,
      "epoch": 0.2616948391082159,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997469773430199,
      "loss": 5.924,
      "mean_token_accuracy": 0.1561689630150795,
      "num_tokens": 5223544.0,
      "step": 2330
    },
    {
      "entropy": 5.8078991889953615,
      "epoch": 0.2622564160161734,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997450702644596,
      "loss": 5.8672,
      "mean_token_accuracy": 0.16755220293998718,
      "num_tokens": 5234713.0,
      "step": 2335
    },
    {
      "entropy": 6.004143762588501,
      "epoch": 0.26281799292413094,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997431560299292,
      "loss": 5.9525,
      "mean_token_accuracy": 0.15500690788030624,
      "num_tokens": 5246103.0,
      "step": 2340
    },
    {
      "entropy": 5.804888391494751,
      "epoch": 0.2633795698320885,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997412346394894,
      "loss": 5.7883,
      "mean_token_accuracy": 0.16899138689041138,
      "num_tokens": 5257372.0,
      "step": 2345
    },
    {
      "entropy": 5.825739240646362,
      "epoch": 0.26394114674004604,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997393060932017,
      "loss": 5.8548,
      "mean_token_accuracy": 0.16804155260324477,
      "num_tokens": 5268677.0,
      "step": 2350
    },
    {
      "entropy": 5.981194734573364,
      "epoch": 0.2645027236480036,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997373703911273,
      "loss": 5.9211,
      "mean_token_accuracy": 0.16036693155765533,
      "num_tokens": 5279056.0,
      "step": 2355
    },
    {
      "entropy": 5.791534566879273,
      "epoch": 0.26506430055596114,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997354275333277,
      "loss": 5.8086,
      "mean_token_accuracy": 0.16567573547363282,
      "num_tokens": 5289310.0,
      "step": 2360
    },
    {
      "entropy": 6.013248252868652,
      "epoch": 0.2656258774639187,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499733477519865,
      "loss": 5.8963,
      "mean_token_accuracy": 0.16800552755594253,
      "num_tokens": 5300619.0,
      "step": 2365
    },
    {
      "entropy": 6.035416316986084,
      "epoch": 0.26618745437187624,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997315203508013,
      "loss": 6.0486,
      "mean_token_accuracy": 0.15519964396953584,
      "num_tokens": 5311554.0,
      "step": 2370
    },
    {
      "entropy": 5.981985998153687,
      "epoch": 0.2667490312798338,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004997295560261988,
      "loss": 5.8789,
      "mean_token_accuracy": 0.16367268562316895,
      "num_tokens": 5323592.0,
      "step": 2375
    },
    {
      "entropy": 5.978114938735962,
      "epoch": 0.26731060818779134,
      "grad_norm": 1.0,
      "learning_rate": 0.00049972758454612,
      "loss": 5.9728,
      "mean_token_accuracy": 0.15567181259393692,
      "num_tokens": 5334978.0,
      "step": 2380
    },
    {
      "entropy": 5.869251680374146,
      "epoch": 0.26787218509574884,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997256059106279,
      "loss": 5.8687,
      "mean_token_accuracy": 0.15930577069520951,
      "num_tokens": 5345724.0,
      "step": 2385
    },
    {
      "entropy": 5.983371686935425,
      "epoch": 0.2684337620037064,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997236201197853,
      "loss": 5.9829,
      "mean_token_accuracy": 0.15415222197771072,
      "num_tokens": 5357141.0,
      "step": 2390
    },
    {
      "entropy": 5.8907865524292,
      "epoch": 0.26899533891166394,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997216271736554,
      "loss": 5.9781,
      "mean_token_accuracy": 0.1598132699728012,
      "num_tokens": 5367889.0,
      "step": 2395
    },
    {
      "entropy": 5.8703772068023685,
      "epoch": 0.2695569158196215,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997196270723018,
      "loss": 5.7156,
      "mean_token_accuracy": 0.16515835821628572,
      "num_tokens": 5379329.0,
      "step": 2400
    },
    {
      "entropy": 5.878870630264283,
      "epoch": 0.27011849272757904,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997176198157881,
      "loss": 5.892,
      "mean_token_accuracy": 0.16479273587465287,
      "num_tokens": 5389525.0,
      "step": 2405
    },
    {
      "entropy": 5.872372484207153,
      "epoch": 0.2706800696355366,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004997156054041781,
      "loss": 5.791,
      "mean_token_accuracy": 0.16410698741674423,
      "num_tokens": 5401061.0,
      "step": 2410
    },
    {
      "entropy": 5.8996857643127445,
      "epoch": 0.27124164654349414,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997135838375362,
      "loss": 5.8973,
      "mean_token_accuracy": 0.16323280036449433,
      "num_tokens": 5411791.0,
      "step": 2415
    },
    {
      "entropy": 5.924712800979615,
      "epoch": 0.2718032234514517,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997115551159267,
      "loss": 5.8386,
      "mean_token_accuracy": 0.16478586345911025,
      "num_tokens": 5422790.0,
      "step": 2420
    },
    {
      "entropy": 5.759177160263062,
      "epoch": 0.27236480035940924,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499709519239414,
      "loss": 5.7834,
      "mean_token_accuracy": 0.16483882069587708,
      "num_tokens": 5434170.0,
      "step": 2425
    },
    {
      "entropy": 6.053406143188477,
      "epoch": 0.2729263772673668,
      "grad_norm": 1.875,
      "learning_rate": 0.0004997074762080632,
      "loss": 5.9184,
      "mean_token_accuracy": 0.1545036256313324,
      "num_tokens": 5445948.0,
      "step": 2430
    },
    {
      "entropy": 5.891904878616333,
      "epoch": 0.2734879541753243,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499705426021939,
      "loss": 5.8834,
      "mean_token_accuracy": 0.15728461295366286,
      "num_tokens": 5457786.0,
      "step": 2435
    },
    {
      "entropy": 5.781130790710449,
      "epoch": 0.27404953108328184,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499703368681107,
      "loss": 5.8408,
      "mean_token_accuracy": 0.1613198235630989,
      "num_tokens": 5469434.0,
      "step": 2440
    },
    {
      "entropy": 5.9204620838165285,
      "epoch": 0.2746111079912394,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997013041856326,
      "loss": 5.8404,
      "mean_token_accuracy": 0.1561868131160736,
      "num_tokens": 5480393.0,
      "step": 2445
    },
    {
      "entropy": 5.874335432052613,
      "epoch": 0.27517268489919694,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996992325355815,
      "loss": 5.8282,
      "mean_token_accuracy": 0.16535558551549911,
      "num_tokens": 5490452.0,
      "step": 2450
    },
    {
      "entropy": 5.791786336898804,
      "epoch": 0.2757342618071545,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996971537310198,
      "loss": 5.7938,
      "mean_token_accuracy": 0.1647621661424637,
      "num_tokens": 5501404.0,
      "step": 2455
    },
    {
      "entropy": 5.994299411773682,
      "epoch": 0.27629583871511204,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004996950677720133,
      "loss": 5.9414,
      "mean_token_accuracy": 0.15404624938964845,
      "num_tokens": 5512938.0,
      "step": 2460
    },
    {
      "entropy": 5.8537561893463135,
      "epoch": 0.2768574156230696,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499692974658629,
      "loss": 5.8529,
      "mean_token_accuracy": 0.1536116473376751,
      "num_tokens": 5524009.0,
      "step": 2465
    },
    {
      "entropy": 5.983808088302612,
      "epoch": 0.27741899253102714,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996908743909331,
      "loss": 5.9422,
      "mean_token_accuracy": 0.15818442404270172,
      "num_tokens": 5535930.0,
      "step": 2470
    },
    {
      "entropy": 5.950139951705933,
      "epoch": 0.2779805694389847,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996887669689926,
      "loss": 5.9345,
      "mean_token_accuracy": 0.15633012354373932,
      "num_tokens": 5547295.0,
      "step": 2475
    },
    {
      "entropy": 5.969197559356689,
      "epoch": 0.2785421463469422,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004996866523928747,
      "loss": 5.8346,
      "mean_token_accuracy": 0.16864263713359834,
      "num_tokens": 5558447.0,
      "step": 2480
    },
    {
      "entropy": 5.92009596824646,
      "epoch": 0.27910372325489974,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004996845306626467,
      "loss": 5.932,
      "mean_token_accuracy": 0.1618420287966728,
      "num_tokens": 5569822.0,
      "step": 2485
    },
    {
      "entropy": 5.958059930801392,
      "epoch": 0.2796653001628573,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996824017783759,
      "loss": 5.9856,
      "mean_token_accuracy": 0.15708132237195968,
      "num_tokens": 5582284.0,
      "step": 2490
    },
    {
      "entropy": 5.984194135665893,
      "epoch": 0.28022687707081484,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004996802657401304,
      "loss": 5.869,
      "mean_token_accuracy": 0.163690784573555,
      "num_tokens": 5593765.0,
      "step": 2495
    },
    {
      "entropy": 5.780626010894776,
      "epoch": 0.2807884539787724,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996781225479781,
      "loss": 5.8481,
      "mean_token_accuracy": 0.1617444708943367,
      "num_tokens": 5605080.0,
      "step": 2500
    },
    {
      "entropy": 5.99798469543457,
      "epoch": 0.28135003088672994,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004996759722019872,
      "loss": 6.0017,
      "mean_token_accuracy": 0.15487387031316757,
      "num_tokens": 5616739.0,
      "step": 2505
    },
    {
      "entropy": 5.904121208190918,
      "epoch": 0.2819116077946875,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996738147022262,
      "loss": 5.8621,
      "mean_token_accuracy": 0.15306745395064353,
      "num_tokens": 5628779.0,
      "step": 2510
    },
    {
      "entropy": 5.885381984710693,
      "epoch": 0.28247318470264504,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004996716500487638,
      "loss": 5.8984,
      "mean_token_accuracy": 0.1586427852511406,
      "num_tokens": 5639168.0,
      "step": 2515
    },
    {
      "entropy": 5.909733247756958,
      "epoch": 0.2830347616106026,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499669478241669,
      "loss": 5.8667,
      "mean_token_accuracy": 0.15645023584365844,
      "num_tokens": 5650739.0,
      "step": 2520
    },
    {
      "entropy": 5.885955667495727,
      "epoch": 0.28359633851856014,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996672992810108,
      "loss": 5.8945,
      "mean_token_accuracy": 0.15876920521259308,
      "num_tokens": 5662630.0,
      "step": 2525
    },
    {
      "entropy": 5.927337265014648,
      "epoch": 0.28415791542651764,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004996651131668587,
      "loss": 5.91,
      "mean_token_accuracy": 0.16869498640298844,
      "num_tokens": 5673473.0,
      "step": 2530
    },
    {
      "entropy": 5.940307760238648,
      "epoch": 0.2847194923344752,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004996629198992823,
      "loss": 6.0388,
      "mean_token_accuracy": 0.150181782245636,
      "num_tokens": 5686577.0,
      "step": 2535
    },
    {
      "entropy": 5.947848510742188,
      "epoch": 0.28528106924243274,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004996607194783513,
      "loss": 5.7909,
      "mean_token_accuracy": 0.16303919181227683,
      "num_tokens": 5697570.0,
      "step": 2540
    },
    {
      "entropy": 5.871793365478515,
      "epoch": 0.2858426461503903,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996585119041359,
      "loss": 5.8567,
      "mean_token_accuracy": 0.15702579021453858,
      "num_tokens": 5708601.0,
      "step": 2545
    },
    {
      "entropy": 5.792822456359863,
      "epoch": 0.28640422305834784,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004996562971767063,
      "loss": 5.8127,
      "mean_token_accuracy": 0.16065748929977416,
      "num_tokens": 5719619.0,
      "step": 2550
    },
    {
      "entropy": 5.899547290802002,
      "epoch": 0.2869657999663054,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499654075296133,
      "loss": 5.8429,
      "mean_token_accuracy": 0.16420732587575912,
      "num_tokens": 5730061.0,
      "step": 2555
    },
    {
      "entropy": 5.912689304351806,
      "epoch": 0.28752737687426294,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499651846262487,
      "loss": 5.8523,
      "mean_token_accuracy": 0.1673080876469612,
      "num_tokens": 5741579.0,
      "step": 2560
    },
    {
      "entropy": 5.8733302593231205,
      "epoch": 0.2880889537822205,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996496100758389,
      "loss": 5.8671,
      "mean_token_accuracy": 0.16228052973747253,
      "num_tokens": 5754001.0,
      "step": 2565
    },
    {
      "entropy": 5.9434535026550295,
      "epoch": 0.28865053069017804,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996473667362604,
      "loss": 5.9618,
      "mean_token_accuracy": 0.1558829851448536,
      "num_tokens": 5766224.0,
      "step": 2570
    },
    {
      "entropy": 5.918426132202148,
      "epoch": 0.28921210759813554,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004996451162438224,
      "loss": 5.9264,
      "mean_token_accuracy": 0.15533761382102967,
      "num_tokens": 5777475.0,
      "step": 2575
    },
    {
      "entropy": 6.021241569519043,
      "epoch": 0.2897736845060931,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996428585985967,
      "loss": 5.9082,
      "mean_token_accuracy": 0.15379685014486313,
      "num_tokens": 5788359.0,
      "step": 2580
    },
    {
      "entropy": 5.839427900314331,
      "epoch": 0.29033526141405064,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004996405938006553,
      "loss": 5.8604,
      "mean_token_accuracy": 0.16365527361631393,
      "num_tokens": 5799116.0,
      "step": 2585
    },
    {
      "entropy": 5.8858864307403564,
      "epoch": 0.2908968383220082,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996383218500703,
      "loss": 5.8009,
      "mean_token_accuracy": 0.16506671160459518,
      "num_tokens": 5811221.0,
      "step": 2590
    },
    {
      "entropy": 5.835213947296142,
      "epoch": 0.29145841522996574,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499636042746914,
      "loss": 5.8785,
      "mean_token_accuracy": 0.16208941340446473,
      "num_tokens": 5822099.0,
      "step": 2595
    },
    {
      "entropy": 5.913222026824951,
      "epoch": 0.2920199921379233,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499633756491259,
      "loss": 5.8475,
      "mean_token_accuracy": 0.15887113511562348,
      "num_tokens": 5832836.0,
      "step": 2600
    },
    {
      "entropy": 5.820470571517944,
      "epoch": 0.29258156904588084,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004996314630831781,
      "loss": 5.7915,
      "mean_token_accuracy": 0.15489693582057953,
      "num_tokens": 5842892.0,
      "step": 2605
    },
    {
      "entropy": 5.879719972610474,
      "epoch": 0.2931431459538384,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996291625227443,
      "loss": 5.775,
      "mean_token_accuracy": 0.17257018089294435,
      "num_tokens": 5852909.0,
      "step": 2610
    },
    {
      "entropy": 5.919102716445923,
      "epoch": 0.29370472286179594,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004996268548100307,
      "loss": 5.9454,
      "mean_token_accuracy": 0.15507653206586838,
      "num_tokens": 5864676.0,
      "step": 2615
    },
    {
      "entropy": 5.755217504501343,
      "epoch": 0.2942662997697535,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499624539945111,
      "loss": 5.6732,
      "mean_token_accuracy": 0.17032847702503204,
      "num_tokens": 5875258.0,
      "step": 2620
    },
    {
      "entropy": 5.909181356430054,
      "epoch": 0.294827876677711,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996222179280589,
      "loss": 5.8435,
      "mean_token_accuracy": 0.16484241038560868,
      "num_tokens": 5885611.0,
      "step": 2625
    },
    {
      "entropy": 5.845112323760986,
      "epoch": 0.29538945358566854,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996198887589481,
      "loss": 5.804,
      "mean_token_accuracy": 0.16185440346598626,
      "num_tokens": 5896726.0,
      "step": 2630
    },
    {
      "entropy": 5.90363187789917,
      "epoch": 0.2959510304936261,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996175524378531,
      "loss": 5.836,
      "mean_token_accuracy": 0.1556556850671768,
      "num_tokens": 5907860.0,
      "step": 2635
    },
    {
      "entropy": 5.824535226821899,
      "epoch": 0.29651260740158364,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499615208964848,
      "loss": 5.7633,
      "mean_token_accuracy": 0.16664294600486756,
      "num_tokens": 5918676.0,
      "step": 2640
    },
    {
      "entropy": 5.854987764358521,
      "epoch": 0.2970741843095412,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996128583400077,
      "loss": 5.8401,
      "mean_token_accuracy": 0.1673136442899704,
      "num_tokens": 5928646.0,
      "step": 2645
    },
    {
      "entropy": 5.853833866119385,
      "epoch": 0.29763576121749874,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996105005634066,
      "loss": 5.8502,
      "mean_token_accuracy": 0.16196250915527344,
      "num_tokens": 5938990.0,
      "step": 2650
    },
    {
      "entropy": 5.856926727294922,
      "epoch": 0.2981973381254563,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996081356351202,
      "loss": 5.7999,
      "mean_token_accuracy": 0.16858235150575637,
      "num_tokens": 5948811.0,
      "step": 2655
    },
    {
      "entropy": 5.798100328445434,
      "epoch": 0.29875891503341384,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004996057635552237,
      "loss": 5.8268,
      "mean_token_accuracy": 0.16800669431686402,
      "num_tokens": 5959227.0,
      "step": 2660
    },
    {
      "entropy": 5.866962957382202,
      "epoch": 0.2993204919413714,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004996033843237925,
      "loss": 5.748,
      "mean_token_accuracy": 0.1633942097425461,
      "num_tokens": 5970546.0,
      "step": 2665
    },
    {
      "entropy": 5.948568105697632,
      "epoch": 0.29988206884932894,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996009979409025,
      "loss": 6.0184,
      "mean_token_accuracy": 0.15058949068188668,
      "num_tokens": 5983936.0,
      "step": 2670
    },
    {
      "entropy": 5.829497051239014,
      "epoch": 0.30044364575728644,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995986044066295,
      "loss": 5.8335,
      "mean_token_accuracy": 0.15705229640007018,
      "num_tokens": 5994963.0,
      "step": 2675
    },
    {
      "entropy": 5.724363040924072,
      "epoch": 0.301005222665244,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00049959620372105,
      "loss": 5.7848,
      "mean_token_accuracy": 0.1620802253484726,
      "num_tokens": 6005836.0,
      "step": 2680
    },
    {
      "entropy": 5.882124423980713,
      "epoch": 0.30156679957320154,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995937958842401,
      "loss": 5.7842,
      "mean_token_accuracy": 0.16714142858982087,
      "num_tokens": 6017408.0,
      "step": 2685
    },
    {
      "entropy": 5.831092786788941,
      "epoch": 0.3021283764811591,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004995913808962767,
      "loss": 5.7877,
      "mean_token_accuracy": 0.162654872238636,
      "num_tokens": 6029281.0,
      "step": 2690
    },
    {
      "entropy": 5.836739444732666,
      "epoch": 0.30268995338911664,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995889587572366,
      "loss": 5.7974,
      "mean_token_accuracy": 0.171444633603096,
      "num_tokens": 6039354.0,
      "step": 2695
    },
    {
      "entropy": 5.8621213912963865,
      "epoch": 0.3032515302970742,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499586529467197,
      "loss": 5.775,
      "mean_token_accuracy": 0.16184911429882048,
      "num_tokens": 6050053.0,
      "step": 2700
    },
    {
      "entropy": 5.8485065460205075,
      "epoch": 0.30381310720503174,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995840930262351,
      "loss": 5.8236,
      "mean_token_accuracy": 0.1607857584953308,
      "num_tokens": 6060221.0,
      "step": 2705
    },
    {
      "entropy": 5.851527500152588,
      "epoch": 0.3043746841129893,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004995816494344287,
      "loss": 5.8527,
      "mean_token_accuracy": 0.16010224372148513,
      "num_tokens": 6070963.0,
      "step": 2710
    },
    {
      "entropy": 5.9478880882263185,
      "epoch": 0.30493626102094684,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995791986918554,
      "loss": 5.9101,
      "mean_token_accuracy": 0.1615760624408722,
      "num_tokens": 6082028.0,
      "step": 2715
    },
    {
      "entropy": 5.897542142868042,
      "epoch": 0.30549783792890434,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995767407985933,
      "loss": 5.8573,
      "mean_token_accuracy": 0.16264641284942627,
      "num_tokens": 6092778.0,
      "step": 2720
    },
    {
      "entropy": 5.689213371276855,
      "epoch": 0.3060594148368619,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995742757547207,
      "loss": 5.6296,
      "mean_token_accuracy": 0.17164971232414244,
      "num_tokens": 6103093.0,
      "step": 2725
    },
    {
      "entropy": 5.817043876647949,
      "epoch": 0.30662099174481944,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499571803560316,
      "loss": 5.7244,
      "mean_token_accuracy": 0.16921960562467575,
      "num_tokens": 6113516.0,
      "step": 2730
    },
    {
      "entropy": 5.806196832656861,
      "epoch": 0.307182568652777,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995693242154581,
      "loss": 5.7965,
      "mean_token_accuracy": 0.16070803254842758,
      "num_tokens": 6125723.0,
      "step": 2735
    },
    {
      "entropy": 5.849681663513183,
      "epoch": 0.30774414556073454,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995668377202258,
      "loss": 5.7988,
      "mean_token_accuracy": 0.16376324146986007,
      "num_tokens": 6137091.0,
      "step": 2740
    },
    {
      "entropy": 5.76956524848938,
      "epoch": 0.3083057224686921,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004995643440746982,
      "loss": 5.7289,
      "mean_token_accuracy": 0.17301594614982604,
      "num_tokens": 6147941.0,
      "step": 2745
    },
    {
      "entropy": 5.743298673629761,
      "epoch": 0.30886729937664964,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004995618432789548,
      "loss": 5.7378,
      "mean_token_accuracy": 0.1669941335916519,
      "num_tokens": 6158493.0,
      "step": 2750
    },
    {
      "entropy": 5.804452991485595,
      "epoch": 0.3094288762846072,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004995593353330753,
      "loss": 5.8589,
      "mean_token_accuracy": 0.16288769990205765,
      "num_tokens": 6170759.0,
      "step": 2755
    },
    {
      "entropy": 5.855029296875,
      "epoch": 0.30999045319256474,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004995568202371393,
      "loss": 5.7168,
      "mean_token_accuracy": 0.16882307976484298,
      "num_tokens": 6182488.0,
      "step": 2760
    },
    {
      "entropy": 5.854329395294189,
      "epoch": 0.3105520301005223,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004995542979912272,
      "loss": 5.891,
      "mean_token_accuracy": 0.15524321794509888,
      "num_tokens": 6193849.0,
      "step": 2765
    },
    {
      "entropy": 5.789872646331787,
      "epoch": 0.3111136070084798,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995517685954192,
      "loss": 5.7515,
      "mean_token_accuracy": 0.16951252967119218,
      "num_tokens": 6204132.0,
      "step": 2770
    },
    {
      "entropy": 5.9421745300292965,
      "epoch": 0.31167518391643734,
      "grad_norm": 0.875,
      "learning_rate": 0.0004995492320497958,
      "loss": 5.8861,
      "mean_token_accuracy": 0.1572794422507286,
      "num_tokens": 6215547.0,
      "step": 2775
    },
    {
      "entropy": 5.9671179294586185,
      "epoch": 0.3122367608243949,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004995466883544376,
      "loss": 5.9168,
      "mean_token_accuracy": 0.15148334205150604,
      "num_tokens": 6227150.0,
      "step": 2780
    },
    {
      "entropy": 5.834510707855225,
      "epoch": 0.31279833773235244,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004995441375094258,
      "loss": 5.911,
      "mean_token_accuracy": 0.16446434408426286,
      "num_tokens": 6237387.0,
      "step": 2785
    },
    {
      "entropy": 5.939399671554566,
      "epoch": 0.31335991464031,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995415795148417,
      "loss": 5.8125,
      "mean_token_accuracy": 0.16512976735830306,
      "num_tokens": 6248623.0,
      "step": 2790
    },
    {
      "entropy": 5.766932392120362,
      "epoch": 0.31392149154826754,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995390143707665,
      "loss": 5.7239,
      "mean_token_accuracy": 0.1762712210416794,
      "num_tokens": 6258596.0,
      "step": 2795
    },
    {
      "entropy": 5.81198902130127,
      "epoch": 0.3144830684562251,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995364420772821,
      "loss": 5.935,
      "mean_token_accuracy": 0.15414962992072107,
      "num_tokens": 6270529.0,
      "step": 2800
    },
    {
      "entropy": 6.066451454162598,
      "epoch": 0.31504464536418264,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995338626344702,
      "loss": 5.8459,
      "mean_token_accuracy": 0.16282169967889787,
      "num_tokens": 6280891.0,
      "step": 2805
    },
    {
      "entropy": 5.736438417434693,
      "epoch": 0.3156062222721402,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499531276042413,
      "loss": 5.7929,
      "mean_token_accuracy": 0.15952381491661072,
      "num_tokens": 6292247.0,
      "step": 2810
    },
    {
      "entropy": 5.746332216262817,
      "epoch": 0.31616779918009774,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004995286823011929,
      "loss": 5.7661,
      "mean_token_accuracy": 0.16169815212488176,
      "num_tokens": 6302403.0,
      "step": 2815
    },
    {
      "entropy": 5.739615869522095,
      "epoch": 0.31672937608805524,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995260814108926,
      "loss": 5.6451,
      "mean_token_accuracy": 0.17269758135080338,
      "num_tokens": 6313901.0,
      "step": 2820
    },
    {
      "entropy": 5.937485456466675,
      "epoch": 0.3172909529960128,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995234733715947,
      "loss": 5.8759,
      "mean_token_accuracy": 0.16334352642297745,
      "num_tokens": 6323711.0,
      "step": 2825
    },
    {
      "entropy": 5.813136100769043,
      "epoch": 0.31785252990397034,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004995208581833824,
      "loss": 5.7666,
      "mean_token_accuracy": 0.16345760077238083,
      "num_tokens": 6334937.0,
      "step": 2830
    },
    {
      "entropy": 5.919223499298096,
      "epoch": 0.3184141068119279,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995182358463388,
      "loss": 5.8575,
      "mean_token_accuracy": 0.16153930276632308,
      "num_tokens": 6345723.0,
      "step": 2835
    },
    {
      "entropy": 5.8121337890625,
      "epoch": 0.31897568371988544,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004995156063605475,
      "loss": 5.8015,
      "mean_token_accuracy": 0.1603420630097389,
      "num_tokens": 6357573.0,
      "step": 2840
    },
    {
      "entropy": 5.871078062057495,
      "epoch": 0.319537260627843,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995129697260922,
      "loss": 5.8159,
      "mean_token_accuracy": 0.16848725825548172,
      "num_tokens": 6367417.0,
      "step": 2845
    },
    {
      "entropy": 5.8781421184539795,
      "epoch": 0.32009883753580054,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004995103259430569,
      "loss": 5.8276,
      "mean_token_accuracy": 0.16959178149700166,
      "num_tokens": 6377753.0,
      "step": 2850
    },
    {
      "entropy": 5.828861999511719,
      "epoch": 0.3206604144437581,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995076750115258,
      "loss": 5.8189,
      "mean_token_accuracy": 0.16387695223093032,
      "num_tokens": 6388613.0,
      "step": 2855
    },
    {
      "entropy": 5.8316224098205565,
      "epoch": 0.32122199135171564,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995050169315832,
      "loss": 5.7769,
      "mean_token_accuracy": 0.1644709199666977,
      "num_tokens": 6399415.0,
      "step": 2860
    },
    {
      "entropy": 5.866077947616577,
      "epoch": 0.32178356825967314,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004995023517033138,
      "loss": 5.8072,
      "mean_token_accuracy": 0.16304630041122437,
      "num_tokens": 6411435.0,
      "step": 2865
    },
    {
      "entropy": 5.856478118896485,
      "epoch": 0.3223451451676307,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994996793268024,
      "loss": 5.9623,
      "mean_token_accuracy": 0.16070206016302108,
      "num_tokens": 6423030.0,
      "step": 2870
    },
    {
      "entropy": 5.8936666488647464,
      "epoch": 0.32290672207558824,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004994969998021342,
      "loss": 5.7486,
      "mean_token_accuracy": 0.1634832188487053,
      "num_tokens": 6434330.0,
      "step": 2875
    },
    {
      "entropy": 5.910007047653198,
      "epoch": 0.3234682989835458,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994943131293944,
      "loss": 5.8438,
      "mean_token_accuracy": 0.1616750717163086,
      "num_tokens": 6446452.0,
      "step": 2880
    },
    {
      "entropy": 5.813801622390747,
      "epoch": 0.32402987589150334,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004994916193086687,
      "loss": 5.8075,
      "mean_token_accuracy": 0.16977857053279877,
      "num_tokens": 6456723.0,
      "step": 2885
    },
    {
      "entropy": 5.856975555419922,
      "epoch": 0.3245914527994609,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004994889183400428,
      "loss": 5.9158,
      "mean_token_accuracy": 0.15541210025548935,
      "num_tokens": 6468117.0,
      "step": 2890
    },
    {
      "entropy": 5.875374364852905,
      "epoch": 0.32515302970741844,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004994862102236025,
      "loss": 5.6984,
      "mean_token_accuracy": 0.16900594830513,
      "num_tokens": 6479465.0,
      "step": 2895
    },
    {
      "entropy": 5.878074026107788,
      "epoch": 0.325714606615376,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994834949594342,
      "loss": 5.9302,
      "mean_token_accuracy": 0.16639080941677092,
      "num_tokens": 6490868.0,
      "step": 2900
    },
    {
      "entropy": 5.9626843452453615,
      "epoch": 0.32627618352333354,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994807725476246,
      "loss": 5.81,
      "mean_token_accuracy": 0.16306480169296264,
      "num_tokens": 6502126.0,
      "step": 2905
    },
    {
      "entropy": 5.729597043991089,
      "epoch": 0.3268377604312911,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994780429882599,
      "loss": 5.892,
      "mean_token_accuracy": 0.15703100711107254,
      "num_tokens": 6514382.0,
      "step": 2910
    },
    {
      "entropy": 5.896344518661499,
      "epoch": 0.3273993373392486,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004994753062814274,
      "loss": 5.7714,
      "mean_token_accuracy": 0.16577536463737488,
      "num_tokens": 6526272.0,
      "step": 2915
    },
    {
      "entropy": 5.8439172267913815,
      "epoch": 0.32796091424720614,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994725624272139,
      "loss": 5.7837,
      "mean_token_accuracy": 0.1664314329624176,
      "num_tokens": 6537511.0,
      "step": 2920
    },
    {
      "entropy": 5.755551195144653,
      "epoch": 0.3285224911551637,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004994698114257071,
      "loss": 5.7261,
      "mean_token_accuracy": 0.16514882296323777,
      "num_tokens": 6549737.0,
      "step": 2925
    },
    {
      "entropy": 5.780179738998413,
      "epoch": 0.32908406806312124,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004994670532769944,
      "loss": 5.8268,
      "mean_token_accuracy": 0.1637934371829033,
      "num_tokens": 6560318.0,
      "step": 2930
    },
    {
      "entropy": 5.810793733596801,
      "epoch": 0.3296456449710788,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004994642879811637,
      "loss": 5.7381,
      "mean_token_accuracy": 0.16960348188877106,
      "num_tokens": 6570575.0,
      "step": 2935
    },
    {
      "entropy": 5.7971124172210695,
      "epoch": 0.33020722187903634,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499461515538303,
      "loss": 5.7894,
      "mean_token_accuracy": 0.15714770257472993,
      "num_tokens": 6582083.0,
      "step": 2940
    },
    {
      "entropy": 5.988094854354858,
      "epoch": 0.3307687987869939,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994587359485006,
      "loss": 5.9175,
      "mean_token_accuracy": 0.16404277086257935,
      "num_tokens": 6592240.0,
      "step": 2945
    },
    {
      "entropy": 5.789616870880127,
      "epoch": 0.33133037569495144,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499455949211845,
      "loss": 5.7105,
      "mean_token_accuracy": 0.16835165172815322,
      "num_tokens": 6602680.0,
      "step": 2950
    },
    {
      "entropy": 5.819018793106079,
      "epoch": 0.331891952602909,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004994531553284248,
      "loss": 5.8154,
      "mean_token_accuracy": 0.1665824145078659,
      "num_tokens": 6613399.0,
      "step": 2955
    },
    {
      "entropy": 5.913612318038941,
      "epoch": 0.3324535295108665,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004994503542983293,
      "loss": 5.8909,
      "mean_token_accuracy": 0.160952527821064,
      "num_tokens": 6624410.0,
      "step": 2960
    },
    {
      "entropy": 5.85199613571167,
      "epoch": 0.33301510641882404,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004994475461216472,
      "loss": 5.7373,
      "mean_token_accuracy": 0.16876164823770523,
      "num_tokens": 6635523.0,
      "step": 2965
    },
    {
      "entropy": 5.898876953125,
      "epoch": 0.3335766833267816,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004994447307984684,
      "loss": 5.8893,
      "mean_token_accuracy": 0.1576322928071022,
      "num_tokens": 6648416.0,
      "step": 2970
    },
    {
      "entropy": 5.7013694763183596,
      "epoch": 0.33413826023473914,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004994419083288823,
      "loss": 5.7018,
      "mean_token_accuracy": 0.16272213459014892,
      "num_tokens": 6658907.0,
      "step": 2975
    },
    {
      "entropy": 5.858709526062012,
      "epoch": 0.3346998371426967,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994390787129787,
      "loss": 5.8528,
      "mean_token_accuracy": 0.16023978292942048,
      "num_tokens": 6669682.0,
      "step": 2980
    },
    {
      "entropy": 5.921603679656982,
      "epoch": 0.33526141405065424,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004994362419508478,
      "loss": 5.9159,
      "mean_token_accuracy": 0.15644512325525284,
      "num_tokens": 6681919.0,
      "step": 2985
    },
    {
      "entropy": 5.877898931503296,
      "epoch": 0.3358229909586118,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00049943339804258,
      "loss": 5.8463,
      "mean_token_accuracy": 0.164728282392025,
      "num_tokens": 6693988.0,
      "step": 2990
    },
    {
      "entropy": 5.7946343421936035,
      "epoch": 0.33638456786656934,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004994305469882657,
      "loss": 5.7797,
      "mean_token_accuracy": 0.16247627437114714,
      "num_tokens": 6705413.0,
      "step": 2995
    },
    {
      "entropy": 5.833203840255737,
      "epoch": 0.3369461447745269,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004994276887879957,
      "loss": 5.7304,
      "mean_token_accuracy": 0.1643373355269432,
      "num_tokens": 6716217.0,
      "step": 3000
    },
    {
      "epoch": 0.3369461447745269,
      "eval_entropy": 5.671144631587263,
      "eval_loss": 5.824180603027344,
      "eval_mean_token_accuracy": 0.1703288601276599,
      "eval_num_tokens": 6716217.0,
      "eval_runtime": 24.7059,
      "eval_samples_per_second": 1255.289,
      "eval_steps_per_second": 156.926,
      "step": 3000
    },
    {
      "entropy": 5.783461570739746,
      "epoch": 0.33750772168248444,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004994248234418611,
      "loss": 5.798,
      "mean_token_accuracy": 0.17027225643396376,
      "num_tokens": 6728059.0,
      "step": 3005
    },
    {
      "entropy": 5.841067123413086,
      "epoch": 0.33806929859044194,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499421950949953,
      "loss": 5.7348,
      "mean_token_accuracy": 0.1664648473262787,
      "num_tokens": 6740650.0,
      "step": 3010
    },
    {
      "entropy": 5.7244490623474125,
      "epoch": 0.3386308754983995,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499419071312363,
      "loss": 5.6661,
      "mean_token_accuracy": 0.16809621155261995,
      "num_tokens": 6751467.0,
      "step": 3015
    },
    {
      "entropy": 5.7594749450683596,
      "epoch": 0.33919245240635704,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004994161845291828,
      "loss": 5.7664,
      "mean_token_accuracy": 0.16994708329439162,
      "num_tokens": 6761377.0,
      "step": 3020
    },
    {
      "entropy": 5.758836936950684,
      "epoch": 0.3397540293143146,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004994132906005041,
      "loss": 5.718,
      "mean_token_accuracy": 0.164249050617218,
      "num_tokens": 6773296.0,
      "step": 3025
    },
    {
      "entropy": 5.842982196807862,
      "epoch": 0.34031560622227214,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994103895264193,
      "loss": 5.7694,
      "mean_token_accuracy": 0.16293616145849227,
      "num_tokens": 6784122.0,
      "step": 3030
    },
    {
      "entropy": 5.708171463012695,
      "epoch": 0.3408771831302297,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004994074813070207,
      "loss": 5.6114,
      "mean_token_accuracy": 0.16830061823129655,
      "num_tokens": 6795604.0,
      "step": 3035
    },
    {
      "entropy": 5.7460723400115965,
      "epoch": 0.34143876003818724,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004994045659424007,
      "loss": 5.7745,
      "mean_token_accuracy": 0.16574450433254242,
      "num_tokens": 6806671.0,
      "step": 3040
    },
    {
      "entropy": 5.839440774917603,
      "epoch": 0.3420003369461448,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004994016434326524,
      "loss": 5.7955,
      "mean_token_accuracy": 0.16577693223953247,
      "num_tokens": 6818271.0,
      "step": 3045
    },
    {
      "entropy": 5.7390542984008786,
      "epoch": 0.34256191385410234,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004993987137778687,
      "loss": 5.8033,
      "mean_token_accuracy": 0.16315534114837646,
      "num_tokens": 6830049.0,
      "step": 3050
    },
    {
      "entropy": 5.958957481384277,
      "epoch": 0.3431234907620599,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499395776978143,
      "loss": 5.9135,
      "mean_token_accuracy": 0.15180308371782303,
      "num_tokens": 6842275.0,
      "step": 3055
    },
    {
      "entropy": 5.943307113647461,
      "epoch": 0.3436850676700174,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004993928330335686,
      "loss": 5.8598,
      "mean_token_accuracy": 0.16282831132411957,
      "num_tokens": 6854910.0,
      "step": 3060
    },
    {
      "entropy": 5.793823719024658,
      "epoch": 0.34424664457797494,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993898819442395,
      "loss": 5.807,
      "mean_token_accuracy": 0.16762974634766578,
      "num_tokens": 6866031.0,
      "step": 3065
    },
    {
      "entropy": 5.871393156051636,
      "epoch": 0.3448082214859325,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004993869237102494,
      "loss": 5.8228,
      "mean_token_accuracy": 0.15970202833414077,
      "num_tokens": 6877672.0,
      "step": 3070
    },
    {
      "entropy": 5.761143684387207,
      "epoch": 0.34536979839389004,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004993839583316927,
      "loss": 5.7247,
      "mean_token_accuracy": 0.17055800259113313,
      "num_tokens": 6889089.0,
      "step": 3075
    },
    {
      "entropy": 5.664313745498657,
      "epoch": 0.3459313753018476,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004993809858086638,
      "loss": 5.6823,
      "mean_token_accuracy": 0.17068515717983246,
      "num_tokens": 6899877.0,
      "step": 3080
    },
    {
      "entropy": 5.812488031387329,
      "epoch": 0.34649295220980514,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004993780061412571,
      "loss": 5.6589,
      "mean_token_accuracy": 0.17749509513378142,
      "num_tokens": 6910686.0,
      "step": 3085
    },
    {
      "entropy": 5.706931829452515,
      "epoch": 0.3470545291177627,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004993750193295678,
      "loss": 5.6692,
      "mean_token_accuracy": 0.17425832152366638,
      "num_tokens": 6920867.0,
      "step": 3090
    },
    {
      "entropy": 5.850731134414673,
      "epoch": 0.34761610602572024,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004993720253736909,
      "loss": 5.8867,
      "mean_token_accuracy": 0.16133658438920975,
      "num_tokens": 6932779.0,
      "step": 3095
    },
    {
      "entropy": 5.8591066837310795,
      "epoch": 0.3481776829336778,
      "grad_norm": 1.125,
      "learning_rate": 0.0004993690242737216,
      "loss": 5.7881,
      "mean_token_accuracy": 0.16611640602350236,
      "num_tokens": 6943378.0,
      "step": 3100
    },
    {
      "entropy": 5.807201147079468,
      "epoch": 0.3487392598416353,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004993660160297555,
      "loss": 5.7406,
      "mean_token_accuracy": 0.17002221643924714,
      "num_tokens": 6955469.0,
      "step": 3105
    },
    {
      "entropy": 5.86589674949646,
      "epoch": 0.34930083674959284,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004993630006418885,
      "loss": 5.8558,
      "mean_token_accuracy": 0.16300770342350007,
      "num_tokens": 6966750.0,
      "step": 3110
    },
    {
      "entropy": 5.798392200469971,
      "epoch": 0.3498624136575504,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004993599781102166,
      "loss": 5.7582,
      "mean_token_accuracy": 0.1637483909726143,
      "num_tokens": 6979010.0,
      "step": 3115
    },
    {
      "entropy": 5.802719974517823,
      "epoch": 0.35042399056550794,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004993569484348359,
      "loss": 5.7499,
      "mean_token_accuracy": 0.16707883924245834,
      "num_tokens": 6990566.0,
      "step": 3120
    },
    {
      "entropy": 5.834181118011474,
      "epoch": 0.3509855674734655,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004993539116158429,
      "loss": 5.7652,
      "mean_token_accuracy": 0.1657142534852028,
      "num_tokens": 7000288.0,
      "step": 3125
    },
    {
      "entropy": 5.645972061157226,
      "epoch": 0.35154714438142304,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004993508676533346,
      "loss": 5.7088,
      "mean_token_accuracy": 0.16696739345788955,
      "num_tokens": 7011143.0,
      "step": 3130
    },
    {
      "entropy": 5.985333442687988,
      "epoch": 0.3521087212893806,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004993478165474074,
      "loss": 5.8602,
      "mean_token_accuracy": 0.15459210872650148,
      "num_tokens": 7022797.0,
      "step": 3135
    },
    {
      "entropy": 5.85102014541626,
      "epoch": 0.35267029819733814,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004993447582981588,
      "loss": 5.8302,
      "mean_token_accuracy": 0.16041662842035292,
      "num_tokens": 7033477.0,
      "step": 3140
    },
    {
      "entropy": 5.82341685295105,
      "epoch": 0.3532318751052957,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499341692905686,
      "loss": 5.8544,
      "mean_token_accuracy": 0.16658125072717667,
      "num_tokens": 7045249.0,
      "step": 3145
    },
    {
      "entropy": 5.9159962177276615,
      "epoch": 0.35379345201325324,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004993386203700868,
      "loss": 5.873,
      "mean_token_accuracy": 0.158836430311203,
      "num_tokens": 7056890.0,
      "step": 3150
    },
    {
      "entropy": 5.8627392768859865,
      "epoch": 0.35435502892121074,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993355406914589,
      "loss": 5.7546,
      "mean_token_accuracy": 0.16837516576051711,
      "num_tokens": 7067181.0,
      "step": 3155
    },
    {
      "entropy": 5.777601909637451,
      "epoch": 0.3549166058291683,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004993324538699003,
      "loss": 5.7169,
      "mean_token_accuracy": 0.17076447010040283,
      "num_tokens": 7078220.0,
      "step": 3160
    },
    {
      "entropy": 5.781362104415893,
      "epoch": 0.35547818273712584,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993293599055094,
      "loss": 5.7824,
      "mean_token_accuracy": 0.16769789457321166,
      "num_tokens": 7089536.0,
      "step": 3165
    },
    {
      "entropy": 5.889367294311524,
      "epoch": 0.3560397596450834,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004993262587983848,
      "loss": 5.8619,
      "mean_token_accuracy": 0.16396937817335128,
      "num_tokens": 7101535.0,
      "step": 3170
    },
    {
      "entropy": 5.833146476745606,
      "epoch": 0.35660133655304094,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499323150548625,
      "loss": 5.7677,
      "mean_token_accuracy": 0.16556764245033265,
      "num_tokens": 7112468.0,
      "step": 3175
    },
    {
      "entropy": 5.795244979858398,
      "epoch": 0.3571629134609985,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004993200351563291,
      "loss": 5.7518,
      "mean_token_accuracy": 0.1701154515147209,
      "num_tokens": 7123192.0,
      "step": 3180
    },
    {
      "entropy": 5.722583484649658,
      "epoch": 0.35772449036895604,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004993169126215962,
      "loss": 5.6744,
      "mean_token_accuracy": 0.17081747353076934,
      "num_tokens": 7133063.0,
      "step": 3185
    },
    {
      "entropy": 5.796593475341797,
      "epoch": 0.3582860672769136,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004993137829445259,
      "loss": 5.7686,
      "mean_token_accuracy": 0.16542071104049683,
      "num_tokens": 7144190.0,
      "step": 3190
    },
    {
      "entropy": 5.819895172119141,
      "epoch": 0.35884764418487114,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004993106461252177,
      "loss": 5.7482,
      "mean_token_accuracy": 0.166579270362854,
      "num_tokens": 7155433.0,
      "step": 3195
    },
    {
      "entropy": 5.85274076461792,
      "epoch": 0.3594092210928287,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004993075021637716,
      "loss": 5.824,
      "mean_token_accuracy": 0.16794340312480927,
      "num_tokens": 7166084.0,
      "step": 3200
    },
    {
      "entropy": 5.842438650131226,
      "epoch": 0.3599707980007862,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993043510602875,
      "loss": 5.8464,
      "mean_token_accuracy": 0.16460557132959366,
      "num_tokens": 7176948.0,
      "step": 3205
    },
    {
      "entropy": 5.817403078079224,
      "epoch": 0.36053237490874374,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499301192814866,
      "loss": 5.6936,
      "mean_token_accuracy": 0.17103520631790162,
      "num_tokens": 7188601.0,
      "step": 3210
    },
    {
      "entropy": 5.786190128326416,
      "epoch": 0.3610939518167013,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004992980274276074,
      "loss": 5.7437,
      "mean_token_accuracy": 0.17011172771453859,
      "num_tokens": 7198495.0,
      "step": 3215
    },
    {
      "entropy": 5.820584297180176,
      "epoch": 0.36165552872465884,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004992948548986128,
      "loss": 5.7303,
      "mean_token_accuracy": 0.16564846634864808,
      "num_tokens": 7209848.0,
      "step": 3220
    },
    {
      "entropy": 5.795607185363769,
      "epoch": 0.3622171056326164,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004992916752279829,
      "loss": 5.8558,
      "mean_token_accuracy": 0.16129396557807923,
      "num_tokens": 7220366.0,
      "step": 3225
    },
    {
      "entropy": 5.897544002532959,
      "epoch": 0.36277868254057394,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499288488415819,
      "loss": 5.7705,
      "mean_token_accuracy": 0.1602724403142929,
      "num_tokens": 7231442.0,
      "step": 3230
    },
    {
      "entropy": 5.750069427490234,
      "epoch": 0.3633402594485315,
      "grad_norm": 1.125,
      "learning_rate": 0.0004992852944622229,
      "loss": 5.8141,
      "mean_token_accuracy": 0.1618262603878975,
      "num_tokens": 7242391.0,
      "step": 3235
    },
    {
      "entropy": 5.774437808990479,
      "epoch": 0.36390183635648904,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004992820933672958,
      "loss": 5.7006,
      "mean_token_accuracy": 0.1609469935297966,
      "num_tokens": 7252061.0,
      "step": 3240
    },
    {
      "entropy": 5.821221542358399,
      "epoch": 0.3644634132644466,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004992788851311399,
      "loss": 5.7641,
      "mean_token_accuracy": 0.16593098938465117,
      "num_tokens": 7264035.0,
      "step": 3245
    },
    {
      "entropy": 5.917165231704712,
      "epoch": 0.3650249901724041,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004992756697538574,
      "loss": 5.8315,
      "mean_token_accuracy": 0.16129237413406372,
      "num_tokens": 7275988.0,
      "step": 3250
    },
    {
      "entropy": 5.73224196434021,
      "epoch": 0.36558656708036164,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004992724472355505,
      "loss": 5.7022,
      "mean_token_accuracy": 0.16566504389047623,
      "num_tokens": 7286509.0,
      "step": 3255
    },
    {
      "entropy": 5.863594579696655,
      "epoch": 0.3661481439883192,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004992692175763219,
      "loss": 5.7925,
      "mean_token_accuracy": 0.1689484089612961,
      "num_tokens": 7296916.0,
      "step": 3260
    },
    {
      "entropy": 5.829982566833496,
      "epoch": 0.36670972089627674,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004992659807762745,
      "loss": 5.8744,
      "mean_token_accuracy": 0.15968306809663774,
      "num_tokens": 7309563.0,
      "step": 3265
    },
    {
      "entropy": 5.838032150268555,
      "epoch": 0.3672712978042343,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004992627368355112,
      "loss": 5.7243,
      "mean_token_accuracy": 0.16306615769863128,
      "num_tokens": 7320264.0,
      "step": 3270
    },
    {
      "entropy": 5.724431037902832,
      "epoch": 0.36783287471219184,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004992594857541355,
      "loss": 5.7052,
      "mean_token_accuracy": 0.16838785856962205,
      "num_tokens": 7330442.0,
      "step": 3275
    },
    {
      "entropy": 5.888395214080811,
      "epoch": 0.3683944516201494,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004992562275322506,
      "loss": 5.895,
      "mean_token_accuracy": 0.15699345022439956,
      "num_tokens": 7342292.0,
      "step": 3280
    },
    {
      "entropy": 5.8640961170196535,
      "epoch": 0.36895602852810694,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004992529621699606,
      "loss": 5.7705,
      "mean_token_accuracy": 0.16837097555398942,
      "num_tokens": 7353857.0,
      "step": 3285
    },
    {
      "entropy": 5.775230503082275,
      "epoch": 0.3695176054360645,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004992496896673692,
      "loss": 5.7152,
      "mean_token_accuracy": 0.16670513898134232,
      "num_tokens": 7365653.0,
      "step": 3290
    },
    {
      "entropy": 5.740857458114624,
      "epoch": 0.37007918234402204,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004992464100245809,
      "loss": 5.7305,
      "mean_token_accuracy": 0.16270318627357483,
      "num_tokens": 7376522.0,
      "step": 3295
    },
    {
      "entropy": 5.747436046600342,
      "epoch": 0.37064075925197953,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004992431232416997,
      "loss": 5.7154,
      "mean_token_accuracy": 0.17133116871118545,
      "num_tokens": 7388562.0,
      "step": 3300
    },
    {
      "entropy": 5.800361824035645,
      "epoch": 0.3712023361599371,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004992398293188307,
      "loss": 5.7742,
      "mean_token_accuracy": 0.16429976969957352,
      "num_tokens": 7398597.0,
      "step": 3305
    },
    {
      "entropy": 5.7546545505523685,
      "epoch": 0.37176391306789464,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004992365282560784,
      "loss": 5.7275,
      "mean_token_accuracy": 0.16918642073869705,
      "num_tokens": 7409656.0,
      "step": 3310
    },
    {
      "entropy": 5.805946731567383,
      "epoch": 0.3723254899758522,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004992332200535483,
      "loss": 5.775,
      "mean_token_accuracy": 0.16006744652986526,
      "num_tokens": 7420621.0,
      "step": 3315
    },
    {
      "entropy": 5.888463115692138,
      "epoch": 0.37288706688380974,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004992299047113452,
      "loss": 5.805,
      "mean_token_accuracy": 0.1692685827612877,
      "num_tokens": 7432951.0,
      "step": 3320
    },
    {
      "entropy": 5.7518034934997555,
      "epoch": 0.3734486437917673,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004992265822295752,
      "loss": 5.7799,
      "mean_token_accuracy": 0.1667136251926422,
      "num_tokens": 7443630.0,
      "step": 3325
    },
    {
      "entropy": 5.860118579864502,
      "epoch": 0.37401022069972484,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004992232526083438,
      "loss": 5.7719,
      "mean_token_accuracy": 0.16329481303691865,
      "num_tokens": 7454535.0,
      "step": 3330
    },
    {
      "entropy": 5.820288276672363,
      "epoch": 0.3745717976076824,
      "grad_norm": 1.0,
      "learning_rate": 0.000499219915847757,
      "loss": 5.8174,
      "mean_token_accuracy": 0.16804846227169037,
      "num_tokens": 7467751.0,
      "step": 3335
    },
    {
      "entropy": 5.762420225143432,
      "epoch": 0.37513337451563994,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004992165719479211,
      "loss": 5.6913,
      "mean_token_accuracy": 0.17562256008386612,
      "num_tokens": 7479347.0,
      "step": 3340
    },
    {
      "entropy": 5.752630519866943,
      "epoch": 0.37569495142359743,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004992132209089427,
      "loss": 5.7231,
      "mean_token_accuracy": 0.16887853890657425,
      "num_tokens": 7490916.0,
      "step": 3345
    },
    {
      "entropy": 5.8285187721252445,
      "epoch": 0.376256528331555,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004992098627309283,
      "loss": 5.7154,
      "mean_token_accuracy": 0.17069905400276184,
      "num_tokens": 7500824.0,
      "step": 3350
    },
    {
      "entropy": 5.852393579483032,
      "epoch": 0.37681810523951254,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499206497413985,
      "loss": 5.8538,
      "mean_token_accuracy": 0.1615035891532898,
      "num_tokens": 7511778.0,
      "step": 3355
    },
    {
      "entropy": 5.893685340881348,
      "epoch": 0.3773796821474701,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004992031249582197,
      "loss": 5.9243,
      "mean_token_accuracy": 0.14992212802171706,
      "num_tokens": 7522616.0,
      "step": 3360
    },
    {
      "entropy": 5.873114585876465,
      "epoch": 0.37794125905542764,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004991997453637401,
      "loss": 5.7852,
      "mean_token_accuracy": 0.1683845803141594,
      "num_tokens": 7533019.0,
      "step": 3365
    },
    {
      "entropy": 5.765808391571045,
      "epoch": 0.3785028359633852,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004991963586306536,
      "loss": 5.7314,
      "mean_token_accuracy": 0.16272798478603362,
      "num_tokens": 7544139.0,
      "step": 3370
    },
    {
      "entropy": 5.830184459686279,
      "epoch": 0.37906441287134274,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004991929647590682,
      "loss": 5.7091,
      "mean_token_accuracy": 0.16443547755479812,
      "num_tokens": 7554638.0,
      "step": 3375
    },
    {
      "entropy": 5.897355794906616,
      "epoch": 0.3796259897793003,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991895637490917,
      "loss": 5.7721,
      "mean_token_accuracy": 0.16925213634967803,
      "num_tokens": 7565158.0,
      "step": 3380
    },
    {
      "entropy": 5.794806432723999,
      "epoch": 0.38018756668725784,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991861556008326,
      "loss": 5.7784,
      "mean_token_accuracy": 0.16760217994451523,
      "num_tokens": 7576381.0,
      "step": 3385
    },
    {
      "entropy": 5.717622900009156,
      "epoch": 0.3807491435952154,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004991827403143992,
      "loss": 5.6744,
      "mean_token_accuracy": 0.1701246201992035,
      "num_tokens": 7586631.0,
      "step": 3390
    },
    {
      "entropy": 5.785594701766968,
      "epoch": 0.3813107205031729,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004991793178899006,
      "loss": 5.7166,
      "mean_token_accuracy": 0.16686752289533616,
      "num_tokens": 7598321.0,
      "step": 3395
    },
    {
      "entropy": 5.751121282577515,
      "epoch": 0.38187229741113043,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004991758883274455,
      "loss": 5.7492,
      "mean_token_accuracy": 0.1697208911180496,
      "num_tokens": 7609336.0,
      "step": 3400
    },
    {
      "entropy": 5.673071670532226,
      "epoch": 0.382433874319088,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004991724516271432,
      "loss": 5.6488,
      "mean_token_accuracy": 0.17070717364549637,
      "num_tokens": 7619116.0,
      "step": 3405
    },
    {
      "entropy": 5.8170229434967045,
      "epoch": 0.38299545122704554,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004991690077891032,
      "loss": 5.782,
      "mean_token_accuracy": 0.1625625655055046,
      "num_tokens": 7630567.0,
      "step": 3410
    },
    {
      "entropy": 5.899796676635742,
      "epoch": 0.3835570281350031,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991655568134348,
      "loss": 5.8416,
      "mean_token_accuracy": 0.1650968074798584,
      "num_tokens": 7642564.0,
      "step": 3415
    },
    {
      "entropy": 5.928177785873413,
      "epoch": 0.38411860504296064,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004991620987002483,
      "loss": 5.8619,
      "mean_token_accuracy": 0.15847996026277542,
      "num_tokens": 7654588.0,
      "step": 3420
    },
    {
      "entropy": 5.73178539276123,
      "epoch": 0.3846801819509182,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004991586334496536,
      "loss": 5.6525,
      "mean_token_accuracy": 0.16577123254537582,
      "num_tokens": 7666120.0,
      "step": 3425
    },
    {
      "entropy": 5.76026554107666,
      "epoch": 0.38524175885887574,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004991551610617611,
      "loss": 5.8045,
      "mean_token_accuracy": 0.16211747676134108,
      "num_tokens": 7677132.0,
      "step": 3430
    },
    {
      "entropy": 5.945665550231934,
      "epoch": 0.3858033357668333,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004991516815366812,
      "loss": 5.871,
      "mean_token_accuracy": 0.16256911009550096,
      "num_tokens": 7687515.0,
      "step": 3435
    },
    {
      "entropy": 5.772604322433471,
      "epoch": 0.38636491267479084,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499148194874525,
      "loss": 5.6858,
      "mean_token_accuracy": 0.16287670731544496,
      "num_tokens": 7698454.0,
      "step": 3440
    },
    {
      "entropy": 5.6769756317138675,
      "epoch": 0.38692648958274833,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004991447010754032,
      "loss": 5.6686,
      "mean_token_accuracy": 0.17033511698246,
      "num_tokens": 7708940.0,
      "step": 3445
    },
    {
      "entropy": 5.819909334182739,
      "epoch": 0.3874880664907059,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004991412001394273,
      "loss": 5.7115,
      "mean_token_accuracy": 0.16921086460351945,
      "num_tokens": 7719705.0,
      "step": 3450
    },
    {
      "entropy": 5.867383575439453,
      "epoch": 0.38804964339866344,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004991376920667086,
      "loss": 5.8453,
      "mean_token_accuracy": 0.1619514159858227,
      "num_tokens": 7730649.0,
      "step": 3455
    },
    {
      "entropy": 5.810446166992188,
      "epoch": 0.388611220306621,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004991341768573589,
      "loss": 5.8605,
      "mean_token_accuracy": 0.16137317419052125,
      "num_tokens": 7742014.0,
      "step": 3460
    },
    {
      "entropy": 5.896458959579467,
      "epoch": 0.38917279721457854,
      "grad_norm": 0.96875,
      "learning_rate": 0.00049913065451149,
      "loss": 5.784,
      "mean_token_accuracy": 0.16017531454563141,
      "num_tokens": 7752892.0,
      "step": 3465
    },
    {
      "entropy": 5.821871614456176,
      "epoch": 0.3897343741225361,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004991271250292141,
      "loss": 5.7861,
      "mean_token_accuracy": 0.16442208290100097,
      "num_tokens": 7762959.0,
      "step": 3470
    },
    {
      "entropy": 5.79083251953125,
      "epoch": 0.39029595103049364,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004991235884106437,
      "loss": 5.6881,
      "mean_token_accuracy": 0.16737891733646393,
      "num_tokens": 7773566.0,
      "step": 3475
    },
    {
      "entropy": 5.765025424957275,
      "epoch": 0.3908575279384512,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004991200446558913,
      "loss": 5.7114,
      "mean_token_accuracy": 0.16073740422725677,
      "num_tokens": 7785782.0,
      "step": 3480
    },
    {
      "entropy": 5.862476444244384,
      "epoch": 0.39141910484640874,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004991164937650698,
      "loss": 5.8762,
      "mean_token_accuracy": 0.16760316342115403,
      "num_tokens": 7797551.0,
      "step": 3485
    },
    {
      "entropy": 5.73495044708252,
      "epoch": 0.39198068175436623,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004991129357382922,
      "loss": 5.6645,
      "mean_token_accuracy": 0.17506183534860612,
      "num_tokens": 7808819.0,
      "step": 3490
    },
    {
      "entropy": 5.616777992248535,
      "epoch": 0.3925422586623238,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004991093705756718,
      "loss": 5.6024,
      "mean_token_accuracy": 0.17233946174383163,
      "num_tokens": 7818940.0,
      "step": 3495
    },
    {
      "entropy": 5.686883592605591,
      "epoch": 0.39310383557028133,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004991057982773221,
      "loss": 5.6274,
      "mean_token_accuracy": 0.17493174970149994,
      "num_tokens": 7829733.0,
      "step": 3500
    },
    {
      "entropy": 5.808896207809449,
      "epoch": 0.3936654124782389,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004991022188433569,
      "loss": 5.6848,
      "mean_token_accuracy": 0.17338285148143767,
      "num_tokens": 7840427.0,
      "step": 3505
    },
    {
      "entropy": 5.790255498886109,
      "epoch": 0.39422698938619644,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00049909863227389,
      "loss": 5.7601,
      "mean_token_accuracy": 0.1668383851647377,
      "num_tokens": 7852834.0,
      "step": 3510
    },
    {
      "entropy": 5.78264799118042,
      "epoch": 0.394788566294154,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004990950385690358,
      "loss": 5.7846,
      "mean_token_accuracy": 0.17013094276189805,
      "num_tokens": 7864993.0,
      "step": 3515
    },
    {
      "entropy": 5.773603677749634,
      "epoch": 0.39535014320211154,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004990914377289088,
      "loss": 5.7661,
      "mean_token_accuracy": 0.16848256886005403,
      "num_tokens": 7876578.0,
      "step": 3520
    },
    {
      "entropy": 5.680042886734009,
      "epoch": 0.3959117201100691,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004990878297536233,
      "loss": 5.614,
      "mean_token_accuracy": 0.17408079355955125,
      "num_tokens": 7886770.0,
      "step": 3525
    },
    {
      "entropy": 5.811942529678345,
      "epoch": 0.39647329701802664,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004990842146432946,
      "loss": 5.7683,
      "mean_token_accuracy": 0.1661121889948845,
      "num_tokens": 7897106.0,
      "step": 3530
    },
    {
      "entropy": 5.751025629043579,
      "epoch": 0.3970348739259842,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004990805923980375,
      "loss": 5.6216,
      "mean_token_accuracy": 0.17755733132362367,
      "num_tokens": 7907321.0,
      "step": 3535
    },
    {
      "entropy": 5.655942153930664,
      "epoch": 0.3975964508339417,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004990769630179673,
      "loss": 5.6403,
      "mean_token_accuracy": 0.16999269723892213,
      "num_tokens": 7918628.0,
      "step": 3540
    },
    {
      "entropy": 5.836121892929077,
      "epoch": 0.39815802774189923,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004990733265032,
      "loss": 5.8027,
      "mean_token_accuracy": 0.1559404879808426,
      "num_tokens": 7929387.0,
      "step": 3545
    },
    {
      "entropy": 5.927747631072998,
      "epoch": 0.3987196046498568,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004990696828538508,
      "loss": 5.8298,
      "mean_token_accuracy": 0.16399966925382614,
      "num_tokens": 7940840.0,
      "step": 3550
    },
    {
      "entropy": 5.6598316669464115,
      "epoch": 0.39928118155781434,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499066032070036,
      "loss": 5.6769,
      "mean_token_accuracy": 0.16840001195669174,
      "num_tokens": 7952243.0,
      "step": 3555
    },
    {
      "entropy": 5.821429252624512,
      "epoch": 0.3998427584657719,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004990623741518719,
      "loss": 5.7262,
      "mean_token_accuracy": 0.16655853092670442,
      "num_tokens": 7963150.0,
      "step": 3560
    },
    {
      "entropy": 5.814081430435181,
      "epoch": 0.40040433537372944,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499058709099475,
      "loss": 5.8311,
      "mean_token_accuracy": 0.15919842272996904,
      "num_tokens": 7974862.0,
      "step": 3565
    },
    {
      "entropy": 5.867576789855957,
      "epoch": 0.400965912281687,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004990550369129618,
      "loss": 5.8455,
      "mean_token_accuracy": 0.16491233110427855,
      "num_tokens": 7986438.0,
      "step": 3570
    },
    {
      "entropy": 5.817252349853516,
      "epoch": 0.40152748918964454,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990513575924494,
      "loss": 5.6855,
      "mean_token_accuracy": 0.17492170333862306,
      "num_tokens": 7996958.0,
      "step": 3575
    },
    {
      "entropy": 5.792129468917847,
      "epoch": 0.4020890660976021,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004990476711380548,
      "loss": 5.7799,
      "mean_token_accuracy": 0.16117825210094452,
      "num_tokens": 8008862.0,
      "step": 3580
    },
    {
      "entropy": 5.815643358230591,
      "epoch": 0.40265064300555964,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004990439775498954,
      "loss": 5.7644,
      "mean_token_accuracy": 0.16977428197860717,
      "num_tokens": 8018940.0,
      "step": 3585
    },
    {
      "entropy": 5.828982782363892,
      "epoch": 0.40321221991351713,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499040276828089,
      "loss": 5.763,
      "mean_token_accuracy": 0.1631990298628807,
      "num_tokens": 8029148.0,
      "step": 3590
    },
    {
      "entropy": 5.729877281188965,
      "epoch": 0.4037737968214747,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004990365689727533,
      "loss": 5.6978,
      "mean_token_accuracy": 0.1689377024769783,
      "num_tokens": 8039618.0,
      "step": 3595
    },
    {
      "entropy": 5.757413005828857,
      "epoch": 0.40433537372943223,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004990328539840062,
      "loss": 5.6564,
      "mean_token_accuracy": 0.16500634849071502,
      "num_tokens": 8051145.0,
      "step": 3600
    },
    {
      "entropy": 5.8003099918365475,
      "epoch": 0.4048969506373898,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004990291318619663,
      "loss": 5.7344,
      "mean_token_accuracy": 0.16073449105024337,
      "num_tokens": 8061520.0,
      "step": 3605
    },
    {
      "entropy": 5.667002534866333,
      "epoch": 0.40545852754534734,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990254026067518,
      "loss": 5.6818,
      "mean_token_accuracy": 0.1623705357313156,
      "num_tokens": 8072876.0,
      "step": 3610
    },
    {
      "entropy": 5.872017192840576,
      "epoch": 0.4060201044533049,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004990216662184817,
      "loss": 5.7732,
      "mean_token_accuracy": 0.16956481486558914,
      "num_tokens": 8084240.0,
      "step": 3615
    },
    {
      "entropy": 5.7483558177948,
      "epoch": 0.40658168136126244,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004990179226972749,
      "loss": 5.6591,
      "mean_token_accuracy": 0.1743858516216278,
      "num_tokens": 8095773.0,
      "step": 3620
    },
    {
      "entropy": 5.592015504837036,
      "epoch": 0.40714325826922,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004990141720432504,
      "loss": 5.6567,
      "mean_token_accuracy": 0.16978682577610016,
      "num_tokens": 8107003.0,
      "step": 3625
    },
    {
      "entropy": 5.759105825424195,
      "epoch": 0.40770483517717754,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004990104142565279,
      "loss": 5.6366,
      "mean_token_accuracy": 0.17539675235748292,
      "num_tokens": 8117864.0,
      "step": 3630
    },
    {
      "entropy": 5.809414434432983,
      "epoch": 0.40826641208513503,
      "grad_norm": 1.0,
      "learning_rate": 0.000499006649337227,
      "loss": 5.8345,
      "mean_token_accuracy": 0.16757829636335372,
      "num_tokens": 8129363.0,
      "step": 3635
    },
    {
      "entropy": 5.834819698333741,
      "epoch": 0.4088279889930926,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004990028772854674,
      "loss": 5.744,
      "mean_token_accuracy": 0.16895247399806976,
      "num_tokens": 8141319.0,
      "step": 3640
    },
    {
      "entropy": 5.747754859924316,
      "epoch": 0.40938956590105013,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004989990981013693,
      "loss": 5.6388,
      "mean_token_accuracy": 0.1729058414697647,
      "num_tokens": 8153344.0,
      "step": 3645
    },
    {
      "entropy": 5.738060092926025,
      "epoch": 0.4099511428090077,
      "grad_norm": 1.125,
      "learning_rate": 0.000498995311785053,
      "loss": 5.8254,
      "mean_token_accuracy": 0.16335516422986984,
      "num_tokens": 8164196.0,
      "step": 3650
    },
    {
      "entropy": 5.7848913192749025,
      "epoch": 0.41051271971696524,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004989915183366392,
      "loss": 5.6866,
      "mean_token_accuracy": 0.1649090826511383,
      "num_tokens": 8174663.0,
      "step": 3655
    },
    {
      "entropy": 5.721457862854004,
      "epoch": 0.4110742966249228,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004989877177562485,
      "loss": 5.6467,
      "mean_token_accuracy": 0.17456351667642594,
      "num_tokens": 8185215.0,
      "step": 3660
    },
    {
      "entropy": 5.749952888488769,
      "epoch": 0.41163587353288034,
      "grad_norm": 1.09375,
      "learning_rate": 0.000498983910044002,
      "loss": 5.7424,
      "mean_token_accuracy": 0.17318557500839232,
      "num_tokens": 8195694.0,
      "step": 3665
    },
    {
      "entropy": 5.635189437866211,
      "epoch": 0.4121974504408379,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004989800952000209,
      "loss": 5.6268,
      "mean_token_accuracy": 0.1810888469219208,
      "num_tokens": 8206112.0,
      "step": 3670
    },
    {
      "entropy": 5.759236907958984,
      "epoch": 0.41275902734879544,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004989762732244268,
      "loss": 5.7231,
      "mean_token_accuracy": 0.16745658367872238,
      "num_tokens": 8216548.0,
      "step": 3675
    },
    {
      "entropy": 5.85249171257019,
      "epoch": 0.413320604256753,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004989724441173413,
      "loss": 5.7553,
      "mean_token_accuracy": 0.16827837824821473,
      "num_tokens": 8226643.0,
      "step": 3680
    },
    {
      "entropy": 5.826248264312744,
      "epoch": 0.4138821811647105,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004989686078788862,
      "loss": 5.7753,
      "mean_token_accuracy": 0.16633398830890656,
      "num_tokens": 8236755.0,
      "step": 3685
    },
    {
      "entropy": 5.724143409729004,
      "epoch": 0.41444375807266803,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004989647645091839,
      "loss": 5.673,
      "mean_token_accuracy": 0.1689850628376007,
      "num_tokens": 8247526.0,
      "step": 3690
    },
    {
      "entropy": 5.706833410263061,
      "epoch": 0.4150053349806256,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004989609140083565,
      "loss": 5.7571,
      "mean_token_accuracy": 0.1614609733223915,
      "num_tokens": 8258649.0,
      "step": 3695
    },
    {
      "entropy": 5.779714632034302,
      "epoch": 0.41556691188858313,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004989570563765269,
      "loss": 5.6419,
      "mean_token_accuracy": 0.17595813125371934,
      "num_tokens": 8268260.0,
      "step": 3700
    },
    {
      "entropy": 5.8270526885986325,
      "epoch": 0.4161284887965407,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004989531916138176,
      "loss": 5.8555,
      "mean_token_accuracy": 0.1657438427209854,
      "num_tokens": 8280090.0,
      "step": 3705
    },
    {
      "entropy": 5.791995334625244,
      "epoch": 0.41669006570449824,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004989493197203519,
      "loss": 5.723,
      "mean_token_accuracy": 0.16886358708143234,
      "num_tokens": 8291217.0,
      "step": 3710
    },
    {
      "entropy": 5.726297903060913,
      "epoch": 0.4172516426124558,
      "grad_norm": 0.90625,
      "learning_rate": 0.000498945440696253,
      "loss": 5.6492,
      "mean_token_accuracy": 0.16829246431589126,
      "num_tokens": 8303541.0,
      "step": 3715
    },
    {
      "entropy": 5.762456846237183,
      "epoch": 0.41781321952041334,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004989415545416444,
      "loss": 5.6985,
      "mean_token_accuracy": 0.1673440992832184,
      "num_tokens": 8316104.0,
      "step": 3720
    },
    {
      "entropy": 5.744006252288818,
      "epoch": 0.4183747964283709,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004989376612566498,
      "loss": 5.6514,
      "mean_token_accuracy": 0.16804916858673097,
      "num_tokens": 8327158.0,
      "step": 3725
    },
    {
      "entropy": 5.8159033298492435,
      "epoch": 0.4189363733363284,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004989337608413933,
      "loss": 5.808,
      "mean_token_accuracy": 0.16719546616077424,
      "num_tokens": 8337685.0,
      "step": 3730
    },
    {
      "entropy": 5.878854513168335,
      "epoch": 0.41949795024428593,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004989298532959989,
      "loss": 5.8328,
      "mean_token_accuracy": 0.1629587471485138,
      "num_tokens": 8349778.0,
      "step": 3735
    },
    {
      "entropy": 5.78654465675354,
      "epoch": 0.4200595271522435,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004989259386205912,
      "loss": 5.6681,
      "mean_token_accuracy": 0.17906660735607147,
      "num_tokens": 8361041.0,
      "step": 3740
    },
    {
      "entropy": 5.802013111114502,
      "epoch": 0.42062110406020103,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004989220168152948,
      "loss": 5.8528,
      "mean_token_accuracy": 0.16143392473459245,
      "num_tokens": 8373227.0,
      "step": 3745
    },
    {
      "entropy": 5.715717411041259,
      "epoch": 0.4211826809681586,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004989180878802346,
      "loss": 5.6921,
      "mean_token_accuracy": 0.1718795359134674,
      "num_tokens": 8384223.0,
      "step": 3750
    },
    {
      "entropy": 5.748624086380005,
      "epoch": 0.42174425787611614,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004989141518155354,
      "loss": 5.7212,
      "mean_token_accuracy": 0.16813842505216597,
      "num_tokens": 8395014.0,
      "step": 3755
    },
    {
      "entropy": 5.7476129055023195,
      "epoch": 0.4223058347840737,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004989102086213229,
      "loss": 5.6997,
      "mean_token_accuracy": 0.16538315266370773,
      "num_tokens": 8406445.0,
      "step": 3760
    },
    {
      "entropy": 5.689257097244263,
      "epoch": 0.42286741169203124,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004989062582977225,
      "loss": 5.6201,
      "mean_token_accuracy": 0.1783268302679062,
      "num_tokens": 8417488.0,
      "step": 3765
    },
    {
      "entropy": 5.838154745101929,
      "epoch": 0.4234289885999888,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004989023008448599,
      "loss": 5.8424,
      "mean_token_accuracy": 0.15957439690828323,
      "num_tokens": 8430582.0,
      "step": 3770
    },
    {
      "entropy": 5.741817569732666,
      "epoch": 0.42399056550794634,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004988983362628614,
      "loss": 5.6359,
      "mean_token_accuracy": 0.16756061762571334,
      "num_tokens": 8442033.0,
      "step": 3775
    },
    {
      "entropy": 5.67078595161438,
      "epoch": 0.42455214241590383,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004988943645518528,
      "loss": 5.6296,
      "mean_token_accuracy": 0.1648266240954399,
      "num_tokens": 8451722.0,
      "step": 3780
    },
    {
      "entropy": 5.661726951599121,
      "epoch": 0.4251137193238614,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004988903857119611,
      "loss": 5.7103,
      "mean_token_accuracy": 0.17043628990650178,
      "num_tokens": 8462753.0,
      "step": 3785
    },
    {
      "entropy": 5.858794832229615,
      "epoch": 0.42567529623181893,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004988863997433125,
      "loss": 5.7323,
      "mean_token_accuracy": 0.1649400696158409,
      "num_tokens": 8472755.0,
      "step": 3790
    },
    {
      "entropy": 5.6707133769989015,
      "epoch": 0.4262368731397765,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004988824066460341,
      "loss": 5.7121,
      "mean_token_accuracy": 0.17042610347270964,
      "num_tokens": 8483296.0,
      "step": 3795
    },
    {
      "entropy": 5.7635962009429935,
      "epoch": 0.42679845004773403,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000498878406420253,
      "loss": 5.5839,
      "mean_token_accuracy": 0.17600233256816863,
      "num_tokens": 8493739.0,
      "step": 3800
    },
    {
      "entropy": 5.72014741897583,
      "epoch": 0.4273600269556916,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004988743990660967,
      "loss": 5.7239,
      "mean_token_accuracy": 0.16405366212129593,
      "num_tokens": 8504788.0,
      "step": 3805
    },
    {
      "entropy": 5.748248672485351,
      "epoch": 0.42792160386364914,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004988703845836927,
      "loss": 5.7044,
      "mean_token_accuracy": 0.1693815343081951,
      "num_tokens": 8514708.0,
      "step": 3810
    },
    {
      "entropy": 5.679748487472534,
      "epoch": 0.4284831807716067,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004988663629731687,
      "loss": 5.588,
      "mean_token_accuracy": 0.17928462773561477,
      "num_tokens": 8525227.0,
      "step": 3815
    },
    {
      "entropy": 5.6420917987823485,
      "epoch": 0.42904475767956424,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000498862334234653,
      "loss": 5.6362,
      "mean_token_accuracy": 0.17540821731090545,
      "num_tokens": 8536502.0,
      "step": 3820
    },
    {
      "entropy": 5.781670999526978,
      "epoch": 0.4296063345875218,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004988582983682737,
      "loss": 5.7711,
      "mean_token_accuracy": 0.16840056627988814,
      "num_tokens": 8548087.0,
      "step": 3825
    },
    {
      "entropy": 5.795170736312866,
      "epoch": 0.4301679114954793,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004988542553741594,
      "loss": 5.761,
      "mean_token_accuracy": 0.16793481409549713,
      "num_tokens": 8560856.0,
      "step": 3830
    },
    {
      "entropy": 5.855878400802612,
      "epoch": 0.43072948840343683,
      "grad_norm": 1.125,
      "learning_rate": 0.0004988502052524388,
      "loss": 5.8761,
      "mean_token_accuracy": 0.16418124437332154,
      "num_tokens": 8572521.0,
      "step": 3835
    },
    {
      "entropy": 5.816007995605469,
      "epoch": 0.4312910653113944,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004988461480032408,
      "loss": 5.7061,
      "mean_token_accuracy": 0.17180412262678146,
      "num_tokens": 8583977.0,
      "step": 3840
    },
    {
      "entropy": 5.716819429397583,
      "epoch": 0.43185264221935193,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004988420836266946,
      "loss": 5.6652,
      "mean_token_accuracy": 0.16814473569393157,
      "num_tokens": 8594441.0,
      "step": 3845
    },
    {
      "entropy": 5.816855192184448,
      "epoch": 0.4324142191273095,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004988380121229298,
      "loss": 5.7001,
      "mean_token_accuracy": 0.16635711044073104,
      "num_tokens": 8604843.0,
      "step": 3850
    },
    {
      "entropy": 5.7296374320983885,
      "epoch": 0.43297579603526704,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004988339334920757,
      "loss": 5.7164,
      "mean_token_accuracy": 0.16814279556274414,
      "num_tokens": 8616059.0,
      "step": 3855
    },
    {
      "entropy": 5.692005634307861,
      "epoch": 0.4335373729432246,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004988298477342625,
      "loss": 5.6582,
      "mean_token_accuracy": 0.1732110321521759,
      "num_tokens": 8626203.0,
      "step": 3860
    },
    {
      "entropy": 5.70359992980957,
      "epoch": 0.43409894985118214,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00049882575484962,
      "loss": 5.6599,
      "mean_token_accuracy": 0.1660165533423424,
      "num_tokens": 8636921.0,
      "step": 3865
    },
    {
      "entropy": 5.611678266525269,
      "epoch": 0.4346605267591397,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004988216548382787,
      "loss": 5.6259,
      "mean_token_accuracy": 0.1729949086904526,
      "num_tokens": 8648526.0,
      "step": 3870
    },
    {
      "entropy": 5.830500364303589,
      "epoch": 0.4352221036670972,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004988175477003691,
      "loss": 5.7701,
      "mean_token_accuracy": 0.16089984625577927,
      "num_tokens": 8660339.0,
      "step": 3875
    },
    {
      "entropy": 5.931962203979492,
      "epoch": 0.43578368057505473,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000498813433436022,
      "loss": 5.8243,
      "mean_token_accuracy": 0.16056531369686128,
      "num_tokens": 8671835.0,
      "step": 3880
    },
    {
      "entropy": 5.566000127792359,
      "epoch": 0.4363452574830123,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004988093120453684,
      "loss": 5.5178,
      "mean_token_accuracy": 0.17349252998828887,
      "num_tokens": 8681146.0,
      "step": 3885
    },
    {
      "entropy": 5.632588624954224,
      "epoch": 0.43690683439096983,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004988051835285394,
      "loss": 5.5873,
      "mean_token_accuracy": 0.1765430301427841,
      "num_tokens": 8691623.0,
      "step": 3890
    },
    {
      "entropy": 5.7326610565185545,
      "epoch": 0.4374684112989274,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004988010478856667,
      "loss": 5.6815,
      "mean_token_accuracy": 0.16309479176998137,
      "num_tokens": 8703702.0,
      "step": 3895
    },
    {
      "entropy": 5.947009515762329,
      "epoch": 0.43802998820688493,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004987969051168816,
      "loss": 5.833,
      "mean_token_accuracy": 0.17143723219633103,
      "num_tokens": 8715434.0,
      "step": 3900
    },
    {
      "entropy": 5.65341625213623,
      "epoch": 0.4385915651148425,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004987927552223165,
      "loss": 5.6245,
      "mean_token_accuracy": 0.18007058501243592,
      "num_tokens": 8726135.0,
      "step": 3905
    },
    {
      "entropy": 5.631536531448364,
      "epoch": 0.43915314202280004,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004987885982021032,
      "loss": 5.7136,
      "mean_token_accuracy": 0.17419808506965637,
      "num_tokens": 8736538.0,
      "step": 3910
    },
    {
      "entropy": 5.8478046417236325,
      "epoch": 0.4397147189307576,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004987844340563741,
      "loss": 5.6466,
      "mean_token_accuracy": 0.1773888498544693,
      "num_tokens": 8746456.0,
      "step": 3915
    },
    {
      "entropy": 5.69115891456604,
      "epoch": 0.44027629583871514,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004987802627852619,
      "loss": 5.7475,
      "mean_token_accuracy": 0.17206081598997117,
      "num_tokens": 8757143.0,
      "step": 3920
    },
    {
      "entropy": 5.810414505004883,
      "epoch": 0.44083787274667263,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004987760843888992,
      "loss": 5.7378,
      "mean_token_accuracy": 0.17338018864393234,
      "num_tokens": 8769084.0,
      "step": 3925
    },
    {
      "entropy": 5.753427219390869,
      "epoch": 0.4413994496546302,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004987718988674193,
      "loss": 5.743,
      "mean_token_accuracy": 0.1651982232928276,
      "num_tokens": 8780707.0,
      "step": 3930
    },
    {
      "entropy": 5.72292742729187,
      "epoch": 0.44196102656258773,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004987677062209552,
      "loss": 5.7107,
      "mean_token_accuracy": 0.16147968918085098,
      "num_tokens": 8792157.0,
      "step": 3935
    },
    {
      "entropy": 5.810204982757568,
      "epoch": 0.4425226034705453,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004987635064496407,
      "loss": 5.7168,
      "mean_token_accuracy": 0.1728065565228462,
      "num_tokens": 8802691.0,
      "step": 3940
    },
    {
      "entropy": 5.810116243362427,
      "epoch": 0.44308418037850283,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004987592995536093,
      "loss": 5.7195,
      "mean_token_accuracy": 0.16920821368694305,
      "num_tokens": 8814869.0,
      "step": 3945
    },
    {
      "entropy": 5.681972312927246,
      "epoch": 0.4436457572864604,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004987550855329951,
      "loss": 5.6692,
      "mean_token_accuracy": 0.17656219750642776,
      "num_tokens": 8825264.0,
      "step": 3950
    },
    {
      "entropy": 5.748923397064209,
      "epoch": 0.44420733419441794,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004987508643879322,
      "loss": 5.7473,
      "mean_token_accuracy": 0.1682357296347618,
      "num_tokens": 8837128.0,
      "step": 3955
    },
    {
      "entropy": 5.801842880249024,
      "epoch": 0.4447689111023755,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004987466361185549,
      "loss": 5.7386,
      "mean_token_accuracy": 0.16442462354898452,
      "num_tokens": 8848587.0,
      "step": 3960
    },
    {
      "entropy": 5.819173049926758,
      "epoch": 0.44533048801033304,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004987424007249979,
      "loss": 5.8059,
      "mean_token_accuracy": 0.16234633326530457,
      "num_tokens": 8859219.0,
      "step": 3965
    },
    {
      "entropy": 5.819671678543091,
      "epoch": 0.4458920649182906,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004987381582073962,
      "loss": 5.7066,
      "mean_token_accuracy": 0.17399350851774215,
      "num_tokens": 8868780.0,
      "step": 3970
    },
    {
      "entropy": 5.786139106750488,
      "epoch": 0.4464536418262481,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004987339085658848,
      "loss": 5.6722,
      "mean_token_accuracy": 0.16677722632884978,
      "num_tokens": 8880903.0,
      "step": 3975
    },
    {
      "entropy": 5.6065737247467045,
      "epoch": 0.44701521873420563,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004987296518005988,
      "loss": 5.5296,
      "mean_token_accuracy": 0.17554202377796174,
      "num_tokens": 8890335.0,
      "step": 3980
    },
    {
      "entropy": 5.659189176559448,
      "epoch": 0.4475767956421632,
      "grad_norm": 1.015625,
      "learning_rate": 0.000498725387911674,
      "loss": 5.6548,
      "mean_token_accuracy": 0.1714469239115715,
      "num_tokens": 8904104.0,
      "step": 3985
    },
    {
      "entropy": 5.81782169342041,
      "epoch": 0.44813837255012073,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004987211168992461,
      "loss": 5.753,
      "mean_token_accuracy": 0.16336047351360322,
      "num_tokens": 8915118.0,
      "step": 3990
    },
    {
      "entropy": 5.765303611755371,
      "epoch": 0.4486999494580783,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000498716838763451,
      "loss": 5.7033,
      "mean_token_accuracy": 0.16566676199436187,
      "num_tokens": 8927450.0,
      "step": 3995
    },
    {
      "entropy": 5.841678524017334,
      "epoch": 0.44926152636603583,
      "grad_norm": 1.03125,
      "learning_rate": 0.000498712553504425,
      "loss": 5.7872,
      "mean_token_accuracy": 0.16800279319286346,
      "num_tokens": 8939902.0,
      "step": 4000
    },
    {
      "entropy": 5.718601369857788,
      "epoch": 0.4498231032739934,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004987082611223046,
      "loss": 5.6537,
      "mean_token_accuracy": 0.16262869536876678,
      "num_tokens": 8951736.0,
      "step": 4005
    },
    {
      "entropy": 5.787780618667602,
      "epoch": 0.45038468018195094,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004987039616172265,
      "loss": 5.7581,
      "mean_token_accuracy": 0.1674249663949013,
      "num_tokens": 8963283.0,
      "step": 4010
    },
    {
      "entropy": 5.740637540817261,
      "epoch": 0.4509462570899085,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004986996549893273,
      "loss": 5.6376,
      "mean_token_accuracy": 0.1759241685271263,
      "num_tokens": 8974323.0,
      "step": 4015
    },
    {
      "entropy": 5.709171104431152,
      "epoch": 0.451507833997866,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004986953412387445,
      "loss": 5.7253,
      "mean_token_accuracy": 0.16495283246040343,
      "num_tokens": 8986687.0,
      "step": 4020
    },
    {
      "entropy": 5.728184700012207,
      "epoch": 0.45206941090582353,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004986910203656151,
      "loss": 5.6307,
      "mean_token_accuracy": 0.17774972021579744,
      "num_tokens": 8997734.0,
      "step": 4025
    },
    {
      "entropy": 5.695507574081421,
      "epoch": 0.4526309878137811,
      "grad_norm": 1.0,
      "learning_rate": 0.0004986866923700769,
      "loss": 5.7386,
      "mean_token_accuracy": 0.16319615244865418,
      "num_tokens": 9008743.0,
      "step": 4030
    },
    {
      "entropy": 5.8246533393859865,
      "epoch": 0.45319256472173863,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004986823572522677,
      "loss": 5.6201,
      "mean_token_accuracy": 0.17015985697507857,
      "num_tokens": 9019495.0,
      "step": 4035
    },
    {
      "entropy": 5.715194749832153,
      "epoch": 0.4537541416296962,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004986780150123254,
      "loss": 5.7272,
      "mean_token_accuracy": 0.1648936837911606,
      "num_tokens": 9029546.0,
      "step": 4040
    },
    {
      "entropy": 5.575207471847534,
      "epoch": 0.45431571853765373,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004986736656503885,
      "loss": 5.4721,
      "mean_token_accuracy": 0.17512325048446656,
      "num_tokens": 9040393.0,
      "step": 4045
    },
    {
      "entropy": 5.707759714126587,
      "epoch": 0.4548772954456113,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004986693091665952,
      "loss": 5.6945,
      "mean_token_accuracy": 0.16999636441469193,
      "num_tokens": 9050840.0,
      "step": 4050
    },
    {
      "entropy": 5.72128119468689,
      "epoch": 0.45543887235356884,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004986649455610845,
      "loss": 5.6179,
      "mean_token_accuracy": 0.171335069835186,
      "num_tokens": 9061615.0,
      "step": 4055
    },
    {
      "entropy": 5.659548997879028,
      "epoch": 0.4560004492615264,
      "grad_norm": 1.109375,
      "learning_rate": 0.000498660574833995,
      "loss": 5.5899,
      "mean_token_accuracy": 0.1790798544883728,
      "num_tokens": 9071902.0,
      "step": 4060
    },
    {
      "entropy": 5.709374570846558,
      "epoch": 0.45656202616948394,
      "grad_norm": 1.0,
      "learning_rate": 0.0004986561969854661,
      "loss": 5.7615,
      "mean_token_accuracy": 0.1679736390709877,
      "num_tokens": 9083613.0,
      "step": 4065
    },
    {
      "entropy": 5.707869672775269,
      "epoch": 0.45712360307744143,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004986518120156373,
      "loss": 5.651,
      "mean_token_accuracy": 0.17007273733615874,
      "num_tokens": 9095678.0,
      "step": 4070
    },
    {
      "entropy": 5.757339191436768,
      "epoch": 0.457685179985399,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000498647419924648,
      "loss": 5.6644,
      "mean_token_accuracy": 0.17293123602867128,
      "num_tokens": 9106637.0,
      "step": 4075
    },
    {
      "entropy": 5.621027708053589,
      "epoch": 0.45824675689335653,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000498643020712638,
      "loss": 5.6203,
      "mean_token_accuracy": 0.17639168947935105,
      "num_tokens": 9117343.0,
      "step": 4080
    },
    {
      "entropy": 5.746330738067627,
      "epoch": 0.4588083338013141,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004986386143797475,
      "loss": 5.7277,
      "mean_token_accuracy": 0.16545779407024383,
      "num_tokens": 9129315.0,
      "step": 4085
    },
    {
      "entropy": 5.73464002609253,
      "epoch": 0.45936991070927163,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004986342009261169,
      "loss": 5.6359,
      "mean_token_accuracy": 0.167977724224329,
      "num_tokens": 9139701.0,
      "step": 4090
    },
    {
      "entropy": 5.74483060836792,
      "epoch": 0.4599314876172292,
      "grad_norm": 1.0,
      "learning_rate": 0.0004986297803518865,
      "loss": 5.759,
      "mean_token_accuracy": 0.16751545965671538,
      "num_tokens": 9152334.0,
      "step": 4095
    },
    {
      "entropy": 5.765616512298584,
      "epoch": 0.46049306452518673,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004986253526571971,
      "loss": 5.576,
      "mean_token_accuracy": 0.1734239786863327,
      "num_tokens": 9163277.0,
      "step": 4100
    },
    {
      "entropy": 5.768707799911499,
      "epoch": 0.4610546414331443,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004986209178421899,
      "loss": 5.7614,
      "mean_token_accuracy": 0.1655284732580185,
      "num_tokens": 9174909.0,
      "step": 4105
    },
    {
      "entropy": 5.6475273132324215,
      "epoch": 0.46161621834110184,
      "grad_norm": 0.984375,
      "learning_rate": 0.000498616475907006,
      "loss": 5.5571,
      "mean_token_accuracy": 0.1686522975564003,
      "num_tokens": 9186058.0,
      "step": 4110
    },
    {
      "entropy": 5.674168920516967,
      "epoch": 0.46217779524905933,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004986120268517865,
      "loss": 5.7231,
      "mean_token_accuracy": 0.1727948561310768,
      "num_tokens": 9196919.0,
      "step": 4115
    },
    {
      "entropy": 5.765881633758545,
      "epoch": 0.4627393721570169,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004986075706766736,
      "loss": 5.6947,
      "mean_token_accuracy": 0.16965437233448027,
      "num_tokens": 9208452.0,
      "step": 4120
    },
    {
      "entropy": 5.759398698806763,
      "epoch": 0.46330094906497443,
      "grad_norm": 1.0,
      "learning_rate": 0.0004986031073818088,
      "loss": 5.7586,
      "mean_token_accuracy": 0.16571177393198014,
      "num_tokens": 9219605.0,
      "step": 4125
    },
    {
      "entropy": 5.737103414535523,
      "epoch": 0.463862525972932,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004985986369673343,
      "loss": 5.5798,
      "mean_token_accuracy": 0.17850485891103746,
      "num_tokens": 9230491.0,
      "step": 4130
    },
    {
      "entropy": 5.662662029266357,
      "epoch": 0.46442410288088953,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004985941594333926,
      "loss": 5.725,
      "mean_token_accuracy": 0.1656407356262207,
      "num_tokens": 9241768.0,
      "step": 4135
    },
    {
      "entropy": 5.733066082000732,
      "epoch": 0.4649856797888471,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000498589674780126,
      "loss": 5.7548,
      "mean_token_accuracy": 0.16551038324832917,
      "num_tokens": 9253779.0,
      "step": 4140
    },
    {
      "entropy": 5.88292088508606,
      "epoch": 0.46554725669680463,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004985851830076775,
      "loss": 5.6474,
      "mean_token_accuracy": 0.17941390872001647,
      "num_tokens": 9264567.0,
      "step": 4145
    },
    {
      "entropy": 5.555645227432251,
      "epoch": 0.4661088336047622,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004985806841161901,
      "loss": 5.6184,
      "mean_token_accuracy": 0.17442478984594345,
      "num_tokens": 9275960.0,
      "step": 4150
    },
    {
      "entropy": 5.769675016403198,
      "epoch": 0.46667041051271974,
      "grad_norm": 1.015625,
      "learning_rate": 0.000498576178105807,
      "loss": 5.7443,
      "mean_token_accuracy": 0.17543013840913774,
      "num_tokens": 9288305.0,
      "step": 4155
    },
    {
      "entropy": 5.7806731224060055,
      "epoch": 0.4672319874206773,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004985716649766717,
      "loss": 5.5678,
      "mean_token_accuracy": 0.17707031667232515,
      "num_tokens": 9299579.0,
      "step": 4160
    },
    {
      "entropy": 5.547669410705566,
      "epoch": 0.4677935643286348,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004985671447289277,
      "loss": 5.612,
      "mean_token_accuracy": 0.17254957109689711,
      "num_tokens": 9311221.0,
      "step": 4165
    },
    {
      "entropy": 5.721201992034912,
      "epoch": 0.46835514123659233,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004985626173627194,
      "loss": 5.7159,
      "mean_token_accuracy": 0.16610774844884874,
      "num_tokens": 9322488.0,
      "step": 4170
    },
    {
      "entropy": 5.68211784362793,
      "epoch": 0.4689167181445499,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004985580828781904,
      "loss": 5.6654,
      "mean_token_accuracy": 0.17230002731084823,
      "num_tokens": 9333672.0,
      "step": 4175
    },
    {
      "entropy": 5.837147235870361,
      "epoch": 0.46947829505250743,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004985535412754855,
      "loss": 5.6984,
      "mean_token_accuracy": 0.17192493826150895,
      "num_tokens": 9344449.0,
      "step": 4180
    },
    {
      "entropy": 5.7961784362792965,
      "epoch": 0.470039871960465,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000498548992554749,
      "loss": 5.7641,
      "mean_token_accuracy": 0.16886739283800126,
      "num_tokens": 9355279.0,
      "step": 4185
    },
    {
      "entropy": 5.712259769439697,
      "epoch": 0.47060144886842253,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004985444367161259,
      "loss": 5.6734,
      "mean_token_accuracy": 0.1700341910123825,
      "num_tokens": 9366942.0,
      "step": 4190
    },
    {
      "entropy": 5.659228897094726,
      "epoch": 0.4711630257763801,
      "grad_norm": 1.0,
      "learning_rate": 0.0004985398737597614,
      "loss": 5.5857,
      "mean_token_accuracy": 0.17830372601747513,
      "num_tokens": 9378129.0,
      "step": 4195
    },
    {
      "entropy": 5.772541093826294,
      "epoch": 0.47172460268433763,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004985353036858004,
      "loss": 5.6896,
      "mean_token_accuracy": 0.17231922447681428,
      "num_tokens": 9388757.0,
      "step": 4200
    },
    {
      "entropy": 5.684602355957031,
      "epoch": 0.4722861795922952,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004985307264943887,
      "loss": 5.6498,
      "mean_token_accuracy": 0.17196862101554872,
      "num_tokens": 9399112.0,
      "step": 4205
    },
    {
      "entropy": 5.770534038543701,
      "epoch": 0.47284775650025274,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004985261421856721,
      "loss": 5.7325,
      "mean_token_accuracy": 0.17345496267080307,
      "num_tokens": 9410156.0,
      "step": 4210
    },
    {
      "entropy": 5.819583177566528,
      "epoch": 0.47340933340821023,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004985215507597963,
      "loss": 5.7198,
      "mean_token_accuracy": 0.1737724334001541,
      "num_tokens": 9420402.0,
      "step": 4215
    },
    {
      "entropy": 5.735470008850098,
      "epoch": 0.4739709103161678,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004985169522169076,
      "loss": 5.7144,
      "mean_token_accuracy": 0.16731610596179963,
      "num_tokens": 9431159.0,
      "step": 4220
    },
    {
      "entropy": 5.678283309936523,
      "epoch": 0.47453248722412533,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004985123465571526,
      "loss": 5.632,
      "mean_token_accuracy": 0.17130510807037352,
      "num_tokens": 9441737.0,
      "step": 4225
    },
    {
      "entropy": 5.7234447479248045,
      "epoch": 0.4750940641320829,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004985077337806777,
      "loss": 5.5744,
      "mean_token_accuracy": 0.17791515290737153,
      "num_tokens": 9452756.0,
      "step": 4230
    },
    {
      "entropy": 5.703112936019897,
      "epoch": 0.47565564104004043,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004985031138876299,
      "loss": 5.6551,
      "mean_token_accuracy": 0.17064587622880936,
      "num_tokens": 9464231.0,
      "step": 4235
    },
    {
      "entropy": 5.7311310291290285,
      "epoch": 0.476217217947998,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004984984868781563,
      "loss": 5.6729,
      "mean_token_accuracy": 0.1737397164106369,
      "num_tokens": 9475648.0,
      "step": 4240
    },
    {
      "entropy": 5.733070755004883,
      "epoch": 0.47677879485595553,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004984938527524041,
      "loss": 5.7043,
      "mean_token_accuracy": 0.17226987034082414,
      "num_tokens": 9486594.0,
      "step": 4245
    },
    {
      "entropy": 5.652930355072021,
      "epoch": 0.4773403717639131,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000498489211510521,
      "loss": 5.6218,
      "mean_token_accuracy": 0.17511968761682511,
      "num_tokens": 9497085.0,
      "step": 4250
    },
    {
      "entropy": 5.772219705581665,
      "epoch": 0.47790194867187064,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004984845631526547,
      "loss": 5.7044,
      "mean_token_accuracy": 0.16952602714300155,
      "num_tokens": 9509765.0,
      "step": 4255
    },
    {
      "entropy": 5.670475578308105,
      "epoch": 0.47846352557982813,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004984799076789534,
      "loss": 5.5285,
      "mean_token_accuracy": 0.1708536058664322,
      "num_tokens": 9521341.0,
      "step": 4260
    },
    {
      "entropy": 5.631149101257324,
      "epoch": 0.4790251024877857,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004984752450895649,
      "loss": 5.6455,
      "mean_token_accuracy": 0.16713927537202836,
      "num_tokens": 9533196.0,
      "step": 4265
    },
    {
      "entropy": 5.850339698791504,
      "epoch": 0.47958667939574323,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000498470575384638,
      "loss": 5.7875,
      "mean_token_accuracy": 0.1595136046409607,
      "num_tokens": 9543687.0,
      "step": 4270
    },
    {
      "entropy": 5.761205673217773,
      "epoch": 0.4801482563037008,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004984658985643214,
      "loss": 5.6716,
      "mean_token_accuracy": 0.172142693400383,
      "num_tokens": 9555320.0,
      "step": 4275
    },
    {
      "entropy": 5.67142014503479,
      "epoch": 0.48070983321165833,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004984612146287638,
      "loss": 5.6228,
      "mean_token_accuracy": 0.1735290065407753,
      "num_tokens": 9566347.0,
      "step": 4280
    },
    {
      "entropy": 5.66849308013916,
      "epoch": 0.4812714101196159,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004984565235781147,
      "loss": 5.6213,
      "mean_token_accuracy": 0.1703104555606842,
      "num_tokens": 9577111.0,
      "step": 4285
    },
    {
      "entropy": 5.701872062683106,
      "epoch": 0.48183298702757343,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000498451825412523,
      "loss": 5.6347,
      "mean_token_accuracy": 0.16810290664434432,
      "num_tokens": 9586612.0,
      "step": 4290
    },
    {
      "entropy": 5.757062673568726,
      "epoch": 0.482394563935531,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004984471201321386,
      "loss": 5.6413,
      "mean_token_accuracy": 0.1697199061512947,
      "num_tokens": 9597347.0,
      "step": 4295
    },
    {
      "entropy": 5.687151956558227,
      "epoch": 0.48295614084348854,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004984424077371112,
      "loss": 5.6178,
      "mean_token_accuracy": 0.17745056003332138,
      "num_tokens": 9608507.0,
      "step": 4300
    },
    {
      "entropy": 5.673877429962158,
      "epoch": 0.4835177177514461,
      "grad_norm": 1.0,
      "learning_rate": 0.0004984376882275909,
      "loss": 5.6808,
      "mean_token_accuracy": 0.17530004531145096,
      "num_tokens": 9619449.0,
      "step": 4305
    },
    {
      "entropy": 5.65759506225586,
      "epoch": 0.4840792946594036,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004984329616037279,
      "loss": 5.5893,
      "mean_token_accuracy": 0.17190126776695253,
      "num_tokens": 9630613.0,
      "step": 4310
    },
    {
      "entropy": 5.736999607086181,
      "epoch": 0.48464087156736113,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004984282278656727,
      "loss": 5.6791,
      "mean_token_accuracy": 0.1694868117570877,
      "num_tokens": 9642149.0,
      "step": 4315
    },
    {
      "entropy": 5.757570791244507,
      "epoch": 0.4852024484753187,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004984234870135762,
      "loss": 5.6028,
      "mean_token_accuracy": 0.17882782816886902,
      "num_tokens": 9652960.0,
      "step": 4320
    },
    {
      "entropy": 5.621033573150635,
      "epoch": 0.48576402538327623,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004984187390475891,
      "loss": 5.6171,
      "mean_token_accuracy": 0.17553129345178603,
      "num_tokens": 9662676.0,
      "step": 4325
    },
    {
      "entropy": 5.819079494476318,
      "epoch": 0.4863256022912338,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004984139839678628,
      "loss": 5.7111,
      "mean_token_accuracy": 0.16809459924697875,
      "num_tokens": 9674138.0,
      "step": 4330
    },
    {
      "entropy": 5.703151273727417,
      "epoch": 0.48688717919919133,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004984092217745486,
      "loss": 5.6657,
      "mean_token_accuracy": 0.17142535001039505,
      "num_tokens": 9685397.0,
      "step": 4335
    },
    {
      "entropy": 5.774184131622315,
      "epoch": 0.4874487561071489,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004984044524677981,
      "loss": 5.7848,
      "mean_token_accuracy": 0.1624930515885353,
      "num_tokens": 9697266.0,
      "step": 4340
    },
    {
      "entropy": 5.826199245452881,
      "epoch": 0.48801033301510643,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004983996760477632,
      "loss": 5.6162,
      "mean_token_accuracy": 0.16484991014003753,
      "num_tokens": 9707967.0,
      "step": 4345
    },
    {
      "entropy": 5.560380268096924,
      "epoch": 0.488571909923064,
      "grad_norm": 1.046875,
      "learning_rate": 0.000498394892514596,
      "loss": 5.5933,
      "mean_token_accuracy": 0.17131706923246384,
      "num_tokens": 9718162.0,
      "step": 4350
    },
    {
      "entropy": 5.7999358654022215,
      "epoch": 0.48913348683102154,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004983901018684486,
      "loss": 5.6509,
      "mean_token_accuracy": 0.17103128731250763,
      "num_tokens": 9729439.0,
      "step": 4355
    },
    {
      "entropy": 5.683565473556518,
      "epoch": 0.48969506373897903,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004983853041094739,
      "loss": 5.6757,
      "mean_token_accuracy": 0.17455968558788298,
      "num_tokens": 9741383.0,
      "step": 4360
    },
    {
      "entropy": 5.714941501617432,
      "epoch": 0.4902566406469366,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004983804992378243,
      "loss": 5.6927,
      "mean_token_accuracy": 0.17366560846567153,
      "num_tokens": 9753050.0,
      "step": 4365
    },
    {
      "entropy": 5.620019578933716,
      "epoch": 0.49081821755489413,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004983756872536532,
      "loss": 5.5683,
      "mean_token_accuracy": 0.18012312352657317,
      "num_tokens": 9763715.0,
      "step": 4370
    },
    {
      "entropy": 5.695003032684326,
      "epoch": 0.4913797944628517,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004983708681571134,
      "loss": 5.5528,
      "mean_token_accuracy": 0.18305423706769944,
      "num_tokens": 9774145.0,
      "step": 4375
    },
    {
      "entropy": 5.680963850021362,
      "epoch": 0.49194137137080923,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004983660419483586,
      "loss": 5.6078,
      "mean_token_accuracy": 0.17759615331888198,
      "num_tokens": 9784703.0,
      "step": 4380
    },
    {
      "entropy": 5.679491186141968,
      "epoch": 0.4925029482787668,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004983612086275424,
      "loss": 5.6362,
      "mean_token_accuracy": 0.17147394567728041,
      "num_tokens": 9795426.0,
      "step": 4385
    },
    {
      "entropy": 5.696656703948975,
      "epoch": 0.49306452518672433,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004983563681948187,
      "loss": 5.7071,
      "mean_token_accuracy": 0.1721432939171791,
      "num_tokens": 9806539.0,
      "step": 4390
    },
    {
      "entropy": 5.826297426223755,
      "epoch": 0.4936261020946819,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004983515206503416,
      "loss": 5.7002,
      "mean_token_accuracy": 0.16714772433042527,
      "num_tokens": 9816806.0,
      "step": 4395
    },
    {
      "entropy": 5.648491716384887,
      "epoch": 0.49418767900263944,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004983466659942655,
      "loss": 5.5917,
      "mean_token_accuracy": 0.1751645863056183,
      "num_tokens": 9828001.0,
      "step": 4400
    },
    {
      "entropy": 5.5533051013946535,
      "epoch": 0.49474925591059693,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004983418042267449,
      "loss": 5.639,
      "mean_token_accuracy": 0.1717042788863182,
      "num_tokens": 9838773.0,
      "step": 4405
    },
    {
      "entropy": 5.716422510147095,
      "epoch": 0.4953108328185545,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004983369353479345,
      "loss": 5.6217,
      "mean_token_accuracy": 0.17908157855272294,
      "num_tokens": 9848680.0,
      "step": 4410
    },
    {
      "entropy": 5.658833932876587,
      "epoch": 0.49587240972651203,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004983320593579895,
      "loss": 5.6063,
      "mean_token_accuracy": 0.16746314913034438,
      "num_tokens": 9859168.0,
      "step": 4415
    },
    {
      "entropy": 5.6701929569244385,
      "epoch": 0.4964339866344696,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004983271762570653,
      "loss": 5.6299,
      "mean_token_accuracy": 0.17470429837703705,
      "num_tokens": 9869071.0,
      "step": 4420
    },
    {
      "entropy": 5.723834419250489,
      "epoch": 0.49699556354242713,
      "grad_norm": 1.015625,
      "learning_rate": 0.000498322286045317,
      "loss": 5.6382,
      "mean_token_accuracy": 0.17299242913722992,
      "num_tokens": 9880208.0,
      "step": 4425
    },
    {
      "entropy": 5.748816013336182,
      "epoch": 0.4975571404503847,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004983173887229006,
      "loss": 5.6825,
      "mean_token_accuracy": 0.17058278769254684,
      "num_tokens": 9891192.0,
      "step": 4430
    },
    {
      "entropy": 5.800527906417846,
      "epoch": 0.49811871735834223,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004983124842899719,
      "loss": 5.6943,
      "mean_token_accuracy": 0.1687977910041809,
      "num_tokens": 9903038.0,
      "step": 4435
    },
    {
      "entropy": 5.710976600646973,
      "epoch": 0.4986802942662998,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004983075727466871,
      "loss": 5.7407,
      "mean_token_accuracy": 0.16949210166931153,
      "num_tokens": 9916104.0,
      "step": 4440
    },
    {
      "entropy": 5.665004396438599,
      "epoch": 0.49924187117425733,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004983026540932026,
      "loss": 5.5764,
      "mean_token_accuracy": 0.17195432782173156,
      "num_tokens": 9927094.0,
      "step": 4445
    },
    {
      "entropy": 5.798759841918946,
      "epoch": 0.4998034480822149,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004982977283296749,
      "loss": 5.7795,
      "mean_token_accuracy": 0.16310768872499465,
      "num_tokens": 9937964.0,
      "step": 4450
    },
    {
      "entropy": 5.817939710617066,
      "epoch": 0.5003650249901724,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000498292795456261,
      "loss": 5.7309,
      "mean_token_accuracy": 0.1622211217880249,
      "num_tokens": 9949238.0,
      "step": 4455
    },
    {
      "entropy": 5.688595342636108,
      "epoch": 0.50092660189813,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004982878554731178,
      "loss": 5.5736,
      "mean_token_accuracy": 0.17607400566339493,
      "num_tokens": 9960125.0,
      "step": 4460
    },
    {
      "entropy": 5.533368349075317,
      "epoch": 0.5014881788060875,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004982829083804028,
      "loss": 5.5922,
      "mean_token_accuracy": 0.1714958816766739,
      "num_tokens": 9971130.0,
      "step": 4465
    },
    {
      "entropy": 5.670532083511352,
      "epoch": 0.5020497557140451,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004982779541782733,
      "loss": 5.504,
      "mean_token_accuracy": 0.18260910958051682,
      "num_tokens": 9981516.0,
      "step": 4470
    },
    {
      "entropy": 5.606881427764892,
      "epoch": 0.5026113326220026,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004982729928668872,
      "loss": 5.7221,
      "mean_token_accuracy": 0.16584337502717972,
      "num_tokens": 9992360.0,
      "step": 4475
    },
    {
      "entropy": 5.906071424484253,
      "epoch": 0.5031729095299601,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004982680244464024,
      "loss": 5.8343,
      "mean_token_accuracy": 0.16307681500911714,
      "num_tokens": 10003410.0,
      "step": 4480
    },
    {
      "entropy": 5.768603515625,
      "epoch": 0.5037344864379176,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004982630489169771,
      "loss": 5.6211,
      "mean_token_accuracy": 0.17453329265117645,
      "num_tokens": 10014974.0,
      "step": 4485
    },
    {
      "entropy": 5.625658369064331,
      "epoch": 0.5042960633458752,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004982580662787697,
      "loss": 5.6752,
      "mean_token_accuracy": 0.17374566048383713,
      "num_tokens": 10026702.0,
      "step": 4490
    },
    {
      "entropy": 5.722785615921021,
      "epoch": 0.5048576402538327,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004982530765319389,
      "loss": 5.6928,
      "mean_token_accuracy": 0.16920074373483657,
      "num_tokens": 10039318.0,
      "step": 4495
    },
    {
      "entropy": 5.749802827835083,
      "epoch": 0.5054192171617903,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004982480796766436,
      "loss": 5.6242,
      "mean_token_accuracy": 0.16932324469089508,
      "num_tokens": 10049726.0,
      "step": 4500
    },
    {
      "entropy": 5.612262010574341,
      "epoch": 0.5059807940697478,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004982430757130429,
      "loss": 5.5047,
      "mean_token_accuracy": 0.18139116764068602,
      "num_tokens": 10059663.0,
      "step": 4505
    },
    {
      "entropy": 5.5416709899902346,
      "epoch": 0.5065423709777054,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004982380646412959,
      "loss": 5.5581,
      "mean_token_accuracy": 0.17046761363744736,
      "num_tokens": 10071195.0,
      "step": 4510
    },
    {
      "entropy": 5.844079303741455,
      "epoch": 0.5071039478856629,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004982330464615625,
      "loss": 5.7765,
      "mean_token_accuracy": 0.1632475271821022,
      "num_tokens": 10082014.0,
      "step": 4515
    },
    {
      "entropy": 5.751795768737793,
      "epoch": 0.5076655247936205,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004982280211740023,
      "loss": 5.5744,
      "mean_token_accuracy": 0.17851648032665252,
      "num_tokens": 10093156.0,
      "step": 4520
    },
    {
      "entropy": 5.613743448257447,
      "epoch": 0.508227101701578,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004982229887787753,
      "loss": 5.6992,
      "mean_token_accuracy": 0.15935521125793456,
      "num_tokens": 10104993.0,
      "step": 4525
    },
    {
      "entropy": 5.648493957519531,
      "epoch": 0.5087886786095356,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004982179492760418,
      "loss": 5.5339,
      "mean_token_accuracy": 0.18565938919782637,
      "num_tokens": 10114590.0,
      "step": 4530
    },
    {
      "entropy": 5.615663385391235,
      "epoch": 0.5093502555174931,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004982129026659621,
      "loss": 5.5248,
      "mean_token_accuracy": 0.1792694315314293,
      "num_tokens": 10125849.0,
      "step": 4535
    },
    {
      "entropy": 5.648001050949096,
      "epoch": 0.5099118324254507,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004982078489486972,
      "loss": 5.5531,
      "mean_token_accuracy": 0.17856017500162125,
      "num_tokens": 10135493.0,
      "step": 4540
    },
    {
      "entropy": 5.713650226593018,
      "epoch": 0.5104734093334082,
      "grad_norm": 1.0,
      "learning_rate": 0.0004982027881244076,
      "loss": 5.7338,
      "mean_token_accuracy": 0.16947841942310332,
      "num_tokens": 10146987.0,
      "step": 4545
    },
    {
      "entropy": 5.71418023109436,
      "epoch": 0.5110349862413658,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004981977201932549,
      "loss": 5.6533,
      "mean_token_accuracy": 0.18045909255743026,
      "num_tokens": 10157823.0,
      "step": 4550
    },
    {
      "entropy": 5.619814920425415,
      "epoch": 0.5115965631493233,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004981926451554001,
      "loss": 5.6355,
      "mean_token_accuracy": 0.1748201385140419,
      "num_tokens": 10169305.0,
      "step": 4555
    },
    {
      "entropy": 5.792585229873657,
      "epoch": 0.5121581400572809,
      "grad_norm": 1.0,
      "learning_rate": 0.0004981875630110048,
      "loss": 5.7278,
      "mean_token_accuracy": 0.16290395855903625,
      "num_tokens": 10180346.0,
      "step": 4560
    },
    {
      "entropy": 5.696116018295288,
      "epoch": 0.5127197169652384,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004981824737602311,
      "loss": 5.606,
      "mean_token_accuracy": 0.17742037624120713,
      "num_tokens": 10190870.0,
      "step": 4565
    },
    {
      "entropy": 5.708626985549927,
      "epoch": 0.513281293873196,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004981773774032409,
      "loss": 5.6616,
      "mean_token_accuracy": 0.1595981314778328,
      "num_tokens": 10202408.0,
      "step": 4570
    },
    {
      "entropy": 5.574900913238525,
      "epoch": 0.5138428707811534,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004981722739401962,
      "loss": 5.4893,
      "mean_token_accuracy": 0.18211452811956405,
      "num_tokens": 10213367.0,
      "step": 4575
    },
    {
      "entropy": 5.738127088546753,
      "epoch": 0.514404447689111,
      "grad_norm": 1.09375,
      "learning_rate": 0.00049816716337126,
      "loss": 5.64,
      "mean_token_accuracy": 0.17113033086061477,
      "num_tokens": 10223865.0,
      "step": 4580
    },
    {
      "entropy": 5.5362961769104,
      "epoch": 0.5149660245970685,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004981620456965946,
      "loss": 5.4073,
      "mean_token_accuracy": 0.1773922085762024,
      "num_tokens": 10235268.0,
      "step": 4585
    },
    {
      "entropy": 5.609420347213745,
      "epoch": 0.5155276015050261,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004981569209163633,
      "loss": 5.6565,
      "mean_token_accuracy": 0.17316273003816604,
      "num_tokens": 10247700.0,
      "step": 4590
    },
    {
      "entropy": 5.723442077636719,
      "epoch": 0.5160891784129836,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004981517890307288,
      "loss": 5.7071,
      "mean_token_accuracy": 0.1703304648399353,
      "num_tokens": 10258553.0,
      "step": 4595
    },
    {
      "entropy": 5.744851779937744,
      "epoch": 0.5166507553209412,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000498146650039855,
      "loss": 5.5959,
      "mean_token_accuracy": 0.17326550483703612,
      "num_tokens": 10268496.0,
      "step": 4600
    },
    {
      "entropy": 5.71726393699646,
      "epoch": 0.5172123322288987,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004981415039439054,
      "loss": 5.668,
      "mean_token_accuracy": 0.169342340528965,
      "num_tokens": 10278803.0,
      "step": 4605
    },
    {
      "entropy": 5.572047185897827,
      "epoch": 0.5177739091368563,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004981363507430436,
      "loss": 5.5351,
      "mean_token_accuracy": 0.17788060009479523,
      "num_tokens": 10289805.0,
      "step": 4610
    },
    {
      "entropy": 5.656609630584716,
      "epoch": 0.5183354860448138,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004981311904374339,
      "loss": 5.6699,
      "mean_token_accuracy": 0.17499706149101257,
      "num_tokens": 10300295.0,
      "step": 4615
    },
    {
      "entropy": 5.859008026123047,
      "epoch": 0.5188970629527714,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004981260230272405,
      "loss": 5.7341,
      "mean_token_accuracy": 0.1685648515820503,
      "num_tokens": 10312175.0,
      "step": 4620
    },
    {
      "entropy": 5.597790479660034,
      "epoch": 0.5194586398607289,
      "grad_norm": 1.0,
      "learning_rate": 0.0004981208485126281,
      "loss": 5.6148,
      "mean_token_accuracy": 0.17157396376132966,
      "num_tokens": 10322888.0,
      "step": 4625
    },
    {
      "entropy": 5.718346166610718,
      "epoch": 0.5200202167686865,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004981156668937613,
      "loss": 5.608,
      "mean_token_accuracy": 0.18176369220018387,
      "num_tokens": 10333305.0,
      "step": 4630
    },
    {
      "entropy": 5.677349710464478,
      "epoch": 0.520581793676644,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004981104781708051,
      "loss": 5.6577,
      "mean_token_accuracy": 0.17384644746780395,
      "num_tokens": 10344190.0,
      "step": 4635
    },
    {
      "entropy": 5.698106813430786,
      "epoch": 0.5211433705846016,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004981052823439247,
      "loss": 5.7185,
      "mean_token_accuracy": 0.17180264741182327,
      "num_tokens": 10356271.0,
      "step": 4640
    },
    {
      "entropy": 5.837706089019775,
      "epoch": 0.5217049474925591,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004981000794132858,
      "loss": 5.7849,
      "mean_token_accuracy": 0.15830366164445878,
      "num_tokens": 10368756.0,
      "step": 4645
    },
    {
      "entropy": 5.713723230361938,
      "epoch": 0.5222665244005167,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004980948693790536,
      "loss": 5.6552,
      "mean_token_accuracy": 0.17654035687446595,
      "num_tokens": 10380029.0,
      "step": 4650
    },
    {
      "entropy": 5.648865699768066,
      "epoch": 0.5228281013084742,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004980896522413943,
      "loss": 5.6426,
      "mean_token_accuracy": 0.17456530779600143,
      "num_tokens": 10390358.0,
      "step": 4655
    },
    {
      "entropy": 5.631305456161499,
      "epoch": 0.5233896782164318,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004980844280004739,
      "loss": 5.5963,
      "mean_token_accuracy": 0.17203012406826018,
      "num_tokens": 10401331.0,
      "step": 4660
    },
    {
      "entropy": 5.744878959655762,
      "epoch": 0.5239512551243893,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004980791966564589,
      "loss": 5.6369,
      "mean_token_accuracy": 0.17498970329761504,
      "num_tokens": 10412755.0,
      "step": 4665
    },
    {
      "entropy": 5.655697345733643,
      "epoch": 0.5245128320323468,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004980739582095158,
      "loss": 5.5698,
      "mean_token_accuracy": 0.1805332988500595,
      "num_tokens": 10422698.0,
      "step": 4670
    },
    {
      "entropy": 5.684418535232544,
      "epoch": 0.5250744089403043,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004980687126598111,
      "loss": 5.7125,
      "mean_token_accuracy": 0.16871879249811172,
      "num_tokens": 10434467.0,
      "step": 4675
    },
    {
      "entropy": 5.668355417251587,
      "epoch": 0.5256359858482619,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004980634600075123,
      "loss": 5.6176,
      "mean_token_accuracy": 0.1768874704837799,
      "num_tokens": 10446007.0,
      "step": 4680
    },
    {
      "entropy": 5.7380529880523685,
      "epoch": 0.5261975627562194,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004980582002527863,
      "loss": 5.6297,
      "mean_token_accuracy": 0.17300729006528853,
      "num_tokens": 10457538.0,
      "step": 4685
    },
    {
      "entropy": 5.654391670227051,
      "epoch": 0.526759139664177,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004980529333958008,
      "loss": 5.5928,
      "mean_token_accuracy": 0.17372933626174927,
      "num_tokens": 10468389.0,
      "step": 4690
    },
    {
      "entropy": 5.6444511890411375,
      "epoch": 0.5273207165721345,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004980476594367233,
      "loss": 5.5816,
      "mean_token_accuracy": 0.17045760601758958,
      "num_tokens": 10479032.0,
      "step": 4695
    },
    {
      "entropy": 5.6966166496276855,
      "epoch": 0.5278822934800921,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004980423783757219,
      "loss": 5.578,
      "mean_token_accuracy": 0.1791645720601082,
      "num_tokens": 10490505.0,
      "step": 4700
    },
    {
      "entropy": 5.682454395294189,
      "epoch": 0.5284438703880496,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004980370902129646,
      "loss": 5.6246,
      "mean_token_accuracy": 0.17562898099422455,
      "num_tokens": 10501382.0,
      "step": 4705
    },
    {
      "entropy": 5.717043256759643,
      "epoch": 0.5290054472960072,
      "grad_norm": 0.953125,
      "learning_rate": 0.00049803179494862,
      "loss": 5.5992,
      "mean_token_accuracy": 0.17553628236055374,
      "num_tokens": 10512727.0,
      "step": 4710
    },
    {
      "entropy": 5.556482124328613,
      "epoch": 0.5295670242039647,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004980264925828563,
      "loss": 5.5775,
      "mean_token_accuracy": 0.17804460376501083,
      "num_tokens": 10523383.0,
      "step": 4715
    },
    {
      "entropy": 5.709814786911011,
      "epoch": 0.5301286011119223,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004980211831158427,
      "loss": 5.5989,
      "mean_token_accuracy": 0.1751595303416252,
      "num_tokens": 10534416.0,
      "step": 4720
    },
    {
      "entropy": 5.625168704986573,
      "epoch": 0.5306901780198798,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004980158665477481,
      "loss": 5.5719,
      "mean_token_accuracy": 0.17959113717079161,
      "num_tokens": 10544291.0,
      "step": 4725
    },
    {
      "entropy": 5.584088706970215,
      "epoch": 0.5312517549278374,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004980105428787419,
      "loss": 5.5392,
      "mean_token_accuracy": 0.17734111696481705,
      "num_tokens": 10555291.0,
      "step": 4730
    },
    {
      "entropy": 5.70352520942688,
      "epoch": 0.5318133318357949,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004980052121089934,
      "loss": 5.621,
      "mean_token_accuracy": 0.1686886802315712,
      "num_tokens": 10565812.0,
      "step": 4735
    },
    {
      "entropy": 5.655381917953491,
      "epoch": 0.5323749087437525,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004979998742386725,
      "loss": 5.5652,
      "mean_token_accuracy": 0.17929922342300414,
      "num_tokens": 10577351.0,
      "step": 4740
    },
    {
      "entropy": 5.672246265411377,
      "epoch": 0.53293648565171,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004979945292679492,
      "loss": 5.6352,
      "mean_token_accuracy": 0.17564911842346193,
      "num_tokens": 10589236.0,
      "step": 4745
    },
    {
      "entropy": 5.578131532669067,
      "epoch": 0.5334980625596676,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004979891771969935,
      "loss": 5.5861,
      "mean_token_accuracy": 0.1787547141313553,
      "num_tokens": 10599926.0,
      "step": 4750
    },
    {
      "entropy": 5.7819067478179935,
      "epoch": 0.5340596394676251,
      "grad_norm": 1.015625,
      "learning_rate": 0.000497983818025976,
      "loss": 5.6961,
      "mean_token_accuracy": 0.17263125479221345,
      "num_tokens": 10611495.0,
      "step": 4755
    },
    {
      "entropy": 5.685853099822998,
      "epoch": 0.5346212163755827,
      "grad_norm": 1.125,
      "learning_rate": 0.0004979784517550671,
      "loss": 5.7153,
      "mean_token_accuracy": 0.17006094455718995,
      "num_tokens": 10622455.0,
      "step": 4760
    },
    {
      "entropy": 5.7067320346832275,
      "epoch": 0.5351827932835401,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004979730783844379,
      "loss": 5.5581,
      "mean_token_accuracy": 0.17890345454216003,
      "num_tokens": 10634255.0,
      "step": 4765
    },
    {
      "entropy": 5.698330736160278,
      "epoch": 0.5357443701914977,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004979676979142595,
      "loss": 5.6687,
      "mean_token_accuracy": 0.17115973979234694,
      "num_tokens": 10644816.0,
      "step": 4770
    },
    {
      "entropy": 5.727415084838867,
      "epoch": 0.5363059470994552,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000497962310344703,
      "loss": 5.5968,
      "mean_token_accuracy": 0.17313093245029448,
      "num_tokens": 10655449.0,
      "step": 4775
    },
    {
      "entropy": 5.684287357330322,
      "epoch": 0.5368675240074128,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00049795691567594,
      "loss": 5.5985,
      "mean_token_accuracy": 0.1717308908700943,
      "num_tokens": 10666343.0,
      "step": 4780
    },
    {
      "entropy": 5.582318305969238,
      "epoch": 0.5374291009153703,
      "grad_norm": 1.125,
      "learning_rate": 0.0004979515139081426,
      "loss": 5.6555,
      "mean_token_accuracy": 0.17162540256977082,
      "num_tokens": 10678511.0,
      "step": 4785
    },
    {
      "entropy": 5.626242160797119,
      "epoch": 0.5379906778233279,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004979461050414822,
      "loss": 5.6289,
      "mean_token_accuracy": 0.18092743307352066,
      "num_tokens": 10690005.0,
      "step": 4790
    },
    {
      "entropy": 5.68348240852356,
      "epoch": 0.5385522547312854,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004979406890761315,
      "loss": 5.6349,
      "mean_token_accuracy": 0.16977559328079223,
      "num_tokens": 10701062.0,
      "step": 4795
    },
    {
      "entropy": 5.75759072303772,
      "epoch": 0.539113831639243,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004979352660122629,
      "loss": 5.55,
      "mean_token_accuracy": 0.1803233802318573,
      "num_tokens": 10711986.0,
      "step": 4800
    },
    {
      "entropy": 5.666845035552979,
      "epoch": 0.5396754085472005,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004979298358500488,
      "loss": 5.6675,
      "mean_token_accuracy": 0.17071848660707473,
      "num_tokens": 10722509.0,
      "step": 4805
    },
    {
      "entropy": 5.741609716415406,
      "epoch": 0.5402369854551581,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004979243985896624,
      "loss": 5.6079,
      "mean_token_accuracy": 0.17072016447782518,
      "num_tokens": 10733755.0,
      "step": 4810
    },
    {
      "entropy": 5.615660715103149,
      "epoch": 0.5407985623631156,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004979189542312765,
      "loss": 5.6384,
      "mean_token_accuracy": 0.1729228302836418,
      "num_tokens": 10744356.0,
      "step": 4815
    },
    {
      "entropy": 5.677237510681152,
      "epoch": 0.5413601392710732,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004979135027750649,
      "loss": 5.5567,
      "mean_token_accuracy": 0.17308810353279114,
      "num_tokens": 10756474.0,
      "step": 4820
    },
    {
      "entropy": 5.656378126144409,
      "epoch": 0.5419217161790307,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004979080442212007,
      "loss": 5.6192,
      "mean_token_accuracy": 0.1746382936835289,
      "num_tokens": 10767551.0,
      "step": 4825
    },
    {
      "entropy": 5.680651378631592,
      "epoch": 0.5424832930869883,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004979025785698581,
      "loss": 5.712,
      "mean_token_accuracy": 0.16779966354370118,
      "num_tokens": 10778524.0,
      "step": 4830
    },
    {
      "entropy": 5.664962291717529,
      "epoch": 0.5430448699949458,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004978971058212108,
      "loss": 5.6191,
      "mean_token_accuracy": 0.17625763714313508,
      "num_tokens": 10789798.0,
      "step": 4835
    },
    {
      "entropy": 5.704985332489014,
      "epoch": 0.5436064469029034,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004978916259754331,
      "loss": 5.567,
      "mean_token_accuracy": 0.1782965049147606,
      "num_tokens": 10800500.0,
      "step": 4840
    },
    {
      "entropy": 5.676136541366577,
      "epoch": 0.5441680238108609,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004978861390326998,
      "loss": 5.5904,
      "mean_token_accuracy": 0.17779603600502014,
      "num_tokens": 10812047.0,
      "step": 4845
    },
    {
      "entropy": 5.631122589111328,
      "epoch": 0.5447296007188185,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004978806449931854,
      "loss": 5.5945,
      "mean_token_accuracy": 0.17298661023378373,
      "num_tokens": 10823272.0,
      "step": 4850
    },
    {
      "entropy": 5.60619797706604,
      "epoch": 0.545291177626776,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004978751438570647,
      "loss": 5.4942,
      "mean_token_accuracy": 0.17209938615560533,
      "num_tokens": 10834485.0,
      "step": 4855
    },
    {
      "entropy": 5.587714338302613,
      "epoch": 0.5458527545347336,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004978696356245129,
      "loss": 5.5583,
      "mean_token_accuracy": 0.17484517991542817,
      "num_tokens": 10843547.0,
      "step": 4860
    },
    {
      "entropy": 5.656753206253052,
      "epoch": 0.546414331442691,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004978641202957057,
      "loss": 5.6862,
      "mean_token_accuracy": 0.16544173657894135,
      "num_tokens": 10855890.0,
      "step": 4865
    },
    {
      "entropy": 5.86716594696045,
      "epoch": 0.5469759083506486,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004978585978708183,
      "loss": 5.7401,
      "mean_token_accuracy": 0.16506602764129638,
      "num_tokens": 10868060.0,
      "step": 4870
    },
    {
      "entropy": 5.70542688369751,
      "epoch": 0.5475374852586061,
      "grad_norm": 1.0,
      "learning_rate": 0.0004978530683500267,
      "loss": 5.6527,
      "mean_token_accuracy": 0.1707741618156433,
      "num_tokens": 10879323.0,
      "step": 4875
    },
    {
      "entropy": 5.627936220169067,
      "epoch": 0.5480990621665637,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004978475317335071,
      "loss": 5.5041,
      "mean_token_accuracy": 0.1808686375617981,
      "num_tokens": 10890545.0,
      "step": 4880
    },
    {
      "entropy": 5.6057947158813475,
      "epoch": 0.5486606390745212,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004978419880214356,
      "loss": 5.6164,
      "mean_token_accuracy": 0.17315196841955185,
      "num_tokens": 10900761.0,
      "step": 4885
    },
    {
      "entropy": 5.692386198043823,
      "epoch": 0.5492222159824788,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004978364372139887,
      "loss": 5.6013,
      "mean_token_accuracy": 0.17580977231264114,
      "num_tokens": 10911874.0,
      "step": 4890
    },
    {
      "entropy": 5.600283193588257,
      "epoch": 0.5497837928904363,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004978308793113434,
      "loss": 5.5124,
      "mean_token_accuracy": 0.18189069181680678,
      "num_tokens": 10922949.0,
      "step": 4895
    },
    {
      "entropy": 5.653483057022095,
      "epoch": 0.5503453697983939,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004978253143136762,
      "loss": 5.7108,
      "mean_token_accuracy": 0.16567208468914033,
      "num_tokens": 10935415.0,
      "step": 4900
    },
    {
      "entropy": 5.613330411911011,
      "epoch": 0.5509069467063514,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004978197422211647,
      "loss": 5.538,
      "mean_token_accuracy": 0.17912916243076324,
      "num_tokens": 10946983.0,
      "step": 4905
    },
    {
      "entropy": 5.621643495559693,
      "epoch": 0.551468523614309,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004978141630339863,
      "loss": 5.608,
      "mean_token_accuracy": 0.16836465448141097,
      "num_tokens": 10957884.0,
      "step": 4910
    },
    {
      "entropy": 5.72468318939209,
      "epoch": 0.5520301005222665,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004978085767523184,
      "loss": 5.6406,
      "mean_token_accuracy": 0.17867001444101333,
      "num_tokens": 10968518.0,
      "step": 4915
    },
    {
      "entropy": 5.660504865646362,
      "epoch": 0.5525916774302241,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000497802983376339,
      "loss": 5.6311,
      "mean_token_accuracy": 0.16997804641723632,
      "num_tokens": 10979621.0,
      "step": 4920
    },
    {
      "entropy": 5.66439299583435,
      "epoch": 0.5531532543381816,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004977973829062263,
      "loss": 5.565,
      "mean_token_accuracy": 0.17659663856029512,
      "num_tokens": 10990158.0,
      "step": 4925
    },
    {
      "entropy": 5.6029417514801025,
      "epoch": 0.5537148312461392,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004977917753421585,
      "loss": 5.6683,
      "mean_token_accuracy": 0.17195623815059663,
      "num_tokens": 11001732.0,
      "step": 4930
    },
    {
      "entropy": 5.768234539031982,
      "epoch": 0.5542764081540967,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000497786160684314,
      "loss": 5.6845,
      "mean_token_accuracy": 0.17094241380691527,
      "num_tokens": 11013483.0,
      "step": 4935
    },
    {
      "entropy": 5.720606136322021,
      "epoch": 0.5548379850620543,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004977805389328719,
      "loss": 5.6785,
      "mean_token_accuracy": 0.17289019227027894,
      "num_tokens": 11025423.0,
      "step": 4940
    },
    {
      "entropy": 5.6656688213348385,
      "epoch": 0.5553995619700118,
      "grad_norm": 1.03125,
      "learning_rate": 0.000497774910088011,
      "loss": 5.5109,
      "mean_token_accuracy": 0.17695456445217134,
      "num_tokens": 11037041.0,
      "step": 4945
    },
    {
      "entropy": 5.643969488143921,
      "epoch": 0.5559611388779694,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004977692741499106,
      "loss": 5.6696,
      "mean_token_accuracy": 0.17007917761802674,
      "num_tokens": 11049206.0,
      "step": 4950
    },
    {
      "entropy": 5.696540069580078,
      "epoch": 0.5565227157859269,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00049776363111875,
      "loss": 5.6304,
      "mean_token_accuracy": 0.1747553288936615,
      "num_tokens": 11060407.0,
      "step": 4955
    },
    {
      "entropy": 5.631777429580689,
      "epoch": 0.5570842926938844,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004977579809947091,
      "loss": 5.571,
      "mean_token_accuracy": 0.17024786919355392,
      "num_tokens": 11071269.0,
      "step": 4960
    },
    {
      "entropy": 5.616387462615966,
      "epoch": 0.5576458696018419,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004977523237779676,
      "loss": 5.5456,
      "mean_token_accuracy": 0.18107982277870177,
      "num_tokens": 11082078.0,
      "step": 4965
    },
    {
      "entropy": 5.849941396713257,
      "epoch": 0.5582074465097995,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004977466594687057,
      "loss": 5.79,
      "mean_token_accuracy": 0.16486690640449525,
      "num_tokens": 11094596.0,
      "step": 4970
    },
    {
      "entropy": 5.670880126953125,
      "epoch": 0.558769023417757,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004977409880671039,
      "loss": 5.6464,
      "mean_token_accuracy": 0.17041069865226746,
      "num_tokens": 11107705.0,
      "step": 4975
    },
    {
      "entropy": 5.656254816055298,
      "epoch": 0.5593306003257146,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004977353095733427,
      "loss": 5.5672,
      "mean_token_accuracy": 0.17746357023715972,
      "num_tokens": 11118152.0,
      "step": 4980
    },
    {
      "entropy": 5.619640588760376,
      "epoch": 0.5598921772336721,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004977296239876026,
      "loss": 5.5918,
      "mean_token_accuracy": 0.17338587194681168,
      "num_tokens": 11129261.0,
      "step": 4985
    },
    {
      "entropy": 5.706889677047729,
      "epoch": 0.5604537541416297,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000497723931310065,
      "loss": 5.7086,
      "mean_token_accuracy": 0.16488064974546432,
      "num_tokens": 11142227.0,
      "step": 4990
    },
    {
      "entropy": 5.786289644241333,
      "epoch": 0.5610153310495872,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004977182315409111,
      "loss": 5.6409,
      "mean_token_accuracy": 0.17092662453651428,
      "num_tokens": 11154277.0,
      "step": 4995
    },
    {
      "entropy": 5.608156490325928,
      "epoch": 0.5615769079575448,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004977125246803224,
      "loss": 5.6366,
      "mean_token_accuracy": 0.1718232586979866,
      "num_tokens": 11164902.0,
      "step": 5000
    },
    {
      "entropy": 5.611583423614502,
      "epoch": 0.5621384848655023,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004977068107284803,
      "loss": 5.498,
      "mean_token_accuracy": 0.16827967017889023,
      "num_tokens": 11176010.0,
      "step": 5005
    },
    {
      "entropy": 5.711164474487305,
      "epoch": 0.5627000617734599,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004977010896855672,
      "loss": 5.5542,
      "mean_token_accuracy": 0.1738850563764572,
      "num_tokens": 11187773.0,
      "step": 5010
    },
    {
      "entropy": 5.5657257556915285,
      "epoch": 0.5632616386814174,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004976953615517649,
      "loss": 5.6249,
      "mean_token_accuracy": 0.17674653679132463,
      "num_tokens": 11198420.0,
      "step": 5015
    },
    {
      "entropy": 5.560083484649658,
      "epoch": 0.563823215589375,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000497689626327256,
      "loss": 5.431,
      "mean_token_accuracy": 0.18361275792121887,
      "num_tokens": 11208553.0,
      "step": 5020
    },
    {
      "entropy": 5.655757188796997,
      "epoch": 0.5643847924973325,
      "grad_norm": 1.078125,
      "learning_rate": 0.000497683884012223,
      "loss": 5.6071,
      "mean_token_accuracy": 0.17579771876335143,
      "num_tokens": 11219737.0,
      "step": 5025
    },
    {
      "entropy": 5.575431442260742,
      "epoch": 0.5649463694052901,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004976781346068487,
      "loss": 5.5922,
      "mean_token_accuracy": 0.1733000248670578,
      "num_tokens": 11230997.0,
      "step": 5030
    },
    {
      "entropy": 5.735964012145996,
      "epoch": 0.5655079463132476,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004976723781113163,
      "loss": 5.6008,
      "mean_token_accuracy": 0.17403126358985901,
      "num_tokens": 11242338.0,
      "step": 5035
    },
    {
      "entropy": 5.759188175201416,
      "epoch": 0.5660695232212052,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004976666145258089,
      "loss": 5.6422,
      "mean_token_accuracy": 0.1733418211340904,
      "num_tokens": 11252671.0,
      "step": 5040
    },
    {
      "entropy": 5.484661960601807,
      "epoch": 0.5666311001291627,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004976608438505103,
      "loss": 5.4814,
      "mean_token_accuracy": 0.1841064766049385,
      "num_tokens": 11262721.0,
      "step": 5045
    },
    {
      "entropy": 5.6478406429290775,
      "epoch": 0.5671926770371203,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004976550660856041,
      "loss": 5.6504,
      "mean_token_accuracy": 0.17501976042985917,
      "num_tokens": 11274074.0,
      "step": 5050
    },
    {
      "entropy": 5.772860145568847,
      "epoch": 0.5677542539450777,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000497649281231274,
      "loss": 5.5935,
      "mean_token_accuracy": 0.1755838707089424,
      "num_tokens": 11286051.0,
      "step": 5055
    },
    {
      "entropy": 5.708038234710694,
      "epoch": 0.5683158308530353,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004976434892877046,
      "loss": 5.717,
      "mean_token_accuracy": 0.1660558745265007,
      "num_tokens": 11298841.0,
      "step": 5060
    },
    {
      "entropy": 5.728675174713135,
      "epoch": 0.5688774077609928,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004976376902550802,
      "loss": 5.7089,
      "mean_token_accuracy": 0.1739212304353714,
      "num_tokens": 11309898.0,
      "step": 5065
    },
    {
      "entropy": 5.7357440948486325,
      "epoch": 0.5694389846689504,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004976318841335854,
      "loss": 5.7239,
      "mean_token_accuracy": 0.17025599181652068,
      "num_tokens": 11322163.0,
      "step": 5070
    },
    {
      "entropy": 5.702219438552857,
      "epoch": 0.5700005615769079,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004976260709234049,
      "loss": 5.6723,
      "mean_token_accuracy": 0.17023052424192428,
      "num_tokens": 11333405.0,
      "step": 5075
    },
    {
      "entropy": 5.8639789581298825,
      "epoch": 0.5705621384848655,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000497620250624724,
      "loss": 5.732,
      "mean_token_accuracy": 0.1701212376356125,
      "num_tokens": 11345037.0,
      "step": 5080
    },
    {
      "entropy": 5.7345935821533205,
      "epoch": 0.571123715392823,
      "grad_norm": 1.109375,
      "learning_rate": 0.000497614423237728,
      "loss": 5.6561,
      "mean_token_accuracy": 0.17391154319047927,
      "num_tokens": 11356403.0,
      "step": 5085
    },
    {
      "entropy": 5.550114631652832,
      "epoch": 0.5716852923007806,
      "grad_norm": 1.125,
      "learning_rate": 0.0004976085887626026,
      "loss": 5.5886,
      "mean_token_accuracy": 0.17732116878032683,
      "num_tokens": 11367913.0,
      "step": 5090
    },
    {
      "entropy": 5.6702498435974125,
      "epoch": 0.5722468692087381,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004976027471995331,
      "loss": 5.5927,
      "mean_token_accuracy": 0.18221637308597566,
      "num_tokens": 11378519.0,
      "step": 5095
    },
    {
      "entropy": 5.70367956161499,
      "epoch": 0.5728084461166957,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000497596898548706,
      "loss": 5.5413,
      "mean_token_accuracy": 0.18071655333042144,
      "num_tokens": 11389530.0,
      "step": 5100
    },
    {
      "entropy": 5.489148712158203,
      "epoch": 0.5733700230246532,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004975910428103072,
      "loss": 5.4969,
      "mean_token_accuracy": 0.1796567976474762,
      "num_tokens": 11400468.0,
      "step": 5105
    },
    {
      "entropy": 5.462692499160767,
      "epoch": 0.5739315999326108,
      "grad_norm": 1.0,
      "learning_rate": 0.0004975851799845233,
      "loss": 5.488,
      "mean_token_accuracy": 0.17597452253103257,
      "num_tokens": 11411242.0,
      "step": 5110
    },
    {
      "entropy": 5.778983449935913,
      "epoch": 0.5744931768405683,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000497579310071541,
      "loss": 5.5757,
      "mean_token_accuracy": 0.1733377143740654,
      "num_tokens": 11420909.0,
      "step": 5115
    },
    {
      "entropy": 5.539997625350952,
      "epoch": 0.5750547537485259,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004975734330715471,
      "loss": 5.4521,
      "mean_token_accuracy": 0.1738223224878311,
      "num_tokens": 11431791.0,
      "step": 5120
    },
    {
      "entropy": 5.716046142578125,
      "epoch": 0.5756163306564834,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004975675489847287,
      "loss": 5.6822,
      "mean_token_accuracy": 0.17156284302473068,
      "num_tokens": 11443056.0,
      "step": 5125
    },
    {
      "entropy": 5.680481433868408,
      "epoch": 0.576177907564441,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004975616578112734,
      "loss": 5.6469,
      "mean_token_accuracy": 0.1763196349143982,
      "num_tokens": 11453413.0,
      "step": 5130
    },
    {
      "entropy": 5.808458614349365,
      "epoch": 0.5767394844723985,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004975557595513684,
      "loss": 5.8541,
      "mean_token_accuracy": 0.1605907917022705,
      "num_tokens": 11465457.0,
      "step": 5135
    },
    {
      "entropy": 5.891483497619629,
      "epoch": 0.5773010613803561,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004975498542052018,
      "loss": 5.6332,
      "mean_token_accuracy": 0.17838063091039658,
      "num_tokens": 11476899.0,
      "step": 5140
    },
    {
      "entropy": 5.546449661254883,
      "epoch": 0.5778626382883136,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004975439417729615,
      "loss": 5.5953,
      "mean_token_accuracy": 0.1756868079304695,
      "num_tokens": 11488652.0,
      "step": 5145
    },
    {
      "entropy": 5.7409721374511715,
      "epoch": 0.5784242151962711,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004975380222548358,
      "loss": 5.6501,
      "mean_token_accuracy": 0.1696472465991974,
      "num_tokens": 11499540.0,
      "step": 5150
    },
    {
      "entropy": 5.78712158203125,
      "epoch": 0.5789857921042286,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004975320956510131,
      "loss": 5.6785,
      "mean_token_accuracy": 0.17673855423927307,
      "num_tokens": 11511156.0,
      "step": 5155
    },
    {
      "entropy": 5.741092824935913,
      "epoch": 0.5795473690121862,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004975261619616822,
      "loss": 5.7344,
      "mean_token_accuracy": 0.16098379641771315,
      "num_tokens": 11523489.0,
      "step": 5160
    },
    {
      "entropy": 5.663554048538208,
      "epoch": 0.5801089459201437,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004975202211870321,
      "loss": 5.5607,
      "mean_token_accuracy": 0.17262763679027557,
      "num_tokens": 11534497.0,
      "step": 5165
    },
    {
      "entropy": 5.777981901168824,
      "epoch": 0.5806705228281013,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004975142733272518,
      "loss": 5.6513,
      "mean_token_accuracy": 0.16769843697547912,
      "num_tokens": 11545585.0,
      "step": 5170
    },
    {
      "entropy": 5.563515138626099,
      "epoch": 0.5812320997360588,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004975083183825306,
      "loss": 5.4882,
      "mean_token_accuracy": 0.17500709146261215,
      "num_tokens": 11556278.0,
      "step": 5175
    },
    {
      "entropy": 5.626966762542724,
      "epoch": 0.5817936766440164,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004975023563530584,
      "loss": 5.5748,
      "mean_token_accuracy": 0.18051354438066483,
      "num_tokens": 11568358.0,
      "step": 5180
    },
    {
      "entropy": 5.565542268753052,
      "epoch": 0.5823552535519739,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004974963872390249,
      "loss": 5.5384,
      "mean_token_accuracy": 0.18231007307767869,
      "num_tokens": 11579777.0,
      "step": 5185
    },
    {
      "entropy": 5.641102981567383,
      "epoch": 0.5829168304599315,
      "grad_norm": 1.078125,
      "learning_rate": 0.00049749041104062,
      "loss": 5.6098,
      "mean_token_accuracy": 0.17667368203401565,
      "num_tokens": 11591479.0,
      "step": 5190
    },
    {
      "entropy": 5.7074541568756105,
      "epoch": 0.583478407367889,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004974844277580343,
      "loss": 5.5524,
      "mean_token_accuracy": 0.17349864393472672,
      "num_tokens": 11601452.0,
      "step": 5195
    },
    {
      "entropy": 5.577162837982177,
      "epoch": 0.5840399842758466,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004974784373914579,
      "loss": 5.4333,
      "mean_token_accuracy": 0.18361603319644929,
      "num_tokens": 11611661.0,
      "step": 5200
    },
    {
      "entropy": 5.599547243118286,
      "epoch": 0.5846015611838041,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004974724399410818,
      "loss": 5.6047,
      "mean_token_accuracy": 0.16931488811969758,
      "num_tokens": 11622940.0,
      "step": 5205
    },
    {
      "entropy": 5.690115451812744,
      "epoch": 0.5851631380917617,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004974664354070971,
      "loss": 5.6227,
      "mean_token_accuracy": 0.1699385091662407,
      "num_tokens": 11634290.0,
      "step": 5210
    },
    {
      "entropy": 5.728141355514526,
      "epoch": 0.5857247149997192,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004974604237896947,
      "loss": 5.6574,
      "mean_token_accuracy": 0.1699557989835739,
      "num_tokens": 11645812.0,
      "step": 5215
    },
    {
      "entropy": 5.6514918327331545,
      "epoch": 0.5862862919076768,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004974544050890662,
      "loss": 5.64,
      "mean_token_accuracy": 0.16767231971025467,
      "num_tokens": 11657520.0,
      "step": 5220
    },
    {
      "entropy": 5.598641109466553,
      "epoch": 0.5868478688156343,
      "grad_norm": 1.03125,
      "learning_rate": 0.000497448379305403,
      "loss": 5.5069,
      "mean_token_accuracy": 0.18663187325000763,
      "num_tokens": 11666887.0,
      "step": 5225
    },
    {
      "entropy": 5.767738676071167,
      "epoch": 0.5874094457235919,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004974423464388973,
      "loss": 5.6995,
      "mean_token_accuracy": 0.169911178201437,
      "num_tokens": 11677337.0,
      "step": 5230
    },
    {
      "entropy": 5.663180017471314,
      "epoch": 0.5879710226315494,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000497436306489741,
      "loss": 5.6141,
      "mean_token_accuracy": 0.17291259318590163,
      "num_tokens": 11688203.0,
      "step": 5235
    },
    {
      "entropy": 5.619011163711548,
      "epoch": 0.588532599539507,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004974302594581264,
      "loss": 5.6417,
      "mean_token_accuracy": 0.16967344433069229,
      "num_tokens": 11699727.0,
      "step": 5240
    },
    {
      "entropy": 5.7239940643310545,
      "epoch": 0.5890941764474645,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004974242053442461,
      "loss": 5.6512,
      "mean_token_accuracy": 0.17492523789405823,
      "num_tokens": 11710956.0,
      "step": 5245
    },
    {
      "entropy": 5.641733169555664,
      "epoch": 0.589655753355422,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004974181441482927,
      "loss": 5.4931,
      "mean_token_accuracy": 0.18178997188806534,
      "num_tokens": 11721681.0,
      "step": 5250
    },
    {
      "entropy": 5.628269052505493,
      "epoch": 0.5902173302633795,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004974120758704595,
      "loss": 5.6147,
      "mean_token_accuracy": 0.16814662367105485,
      "num_tokens": 11731679.0,
      "step": 5255
    },
    {
      "entropy": 5.767731237411499,
      "epoch": 0.5907789071713371,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004974060005109394,
      "loss": 5.7352,
      "mean_token_accuracy": 0.1669013872742653,
      "num_tokens": 11743953.0,
      "step": 5260
    },
    {
      "entropy": 5.543233251571655,
      "epoch": 0.5913404840792946,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000497399918069926,
      "loss": 5.5312,
      "mean_token_accuracy": 0.18252587020397187,
      "num_tokens": 11754091.0,
      "step": 5265
    },
    {
      "entropy": 5.627281427383423,
      "epoch": 0.5919020609872522,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004973938285476131,
      "loss": 5.5562,
      "mean_token_accuracy": 0.1774235710501671,
      "num_tokens": 11764771.0,
      "step": 5270
    },
    {
      "entropy": 5.690001058578491,
      "epoch": 0.5924636378952097,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004973877319441943,
      "loss": 5.6003,
      "mean_token_accuracy": 0.17352219372987748,
      "num_tokens": 11776287.0,
      "step": 5275
    },
    {
      "entropy": 5.614457988739014,
      "epoch": 0.5930252148031673,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004973816282598639,
      "loss": 5.6158,
      "mean_token_accuracy": 0.17027940452098847,
      "num_tokens": 11787142.0,
      "step": 5280
    },
    {
      "entropy": 5.660101699829101,
      "epoch": 0.5935867917111248,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004973755174948162,
      "loss": 5.6097,
      "mean_token_accuracy": 0.1729769602417946,
      "num_tokens": 11798966.0,
      "step": 5285
    },
    {
      "entropy": 5.738249206542969,
      "epoch": 0.5941483686190824,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004973693996492458,
      "loss": 5.6907,
      "mean_token_accuracy": 0.1670216962695122,
      "num_tokens": 11810529.0,
      "step": 5290
    },
    {
      "entropy": 5.606731319427491,
      "epoch": 0.5947099455270399,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004973632747233475,
      "loss": 5.5176,
      "mean_token_accuracy": 0.18147186636924745,
      "num_tokens": 11821965.0,
      "step": 5295
    },
    {
      "entropy": 5.690023565292359,
      "epoch": 0.5952715224349975,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004973571427173163,
      "loss": 5.6884,
      "mean_token_accuracy": 0.1674957126379013,
      "num_tokens": 11833936.0,
      "step": 5300
    },
    {
      "entropy": 5.707571601867675,
      "epoch": 0.595833099342955,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004973510036313473,
      "loss": 5.6043,
      "mean_token_accuracy": 0.18068814724683763,
      "num_tokens": 11845245.0,
      "step": 5305
    },
    {
      "entropy": 5.6551947593688965,
      "epoch": 0.5963946762509126,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004973448574656362,
      "loss": 5.5744,
      "mean_token_accuracy": 0.1755778968334198,
      "num_tokens": 11856010.0,
      "step": 5310
    },
    {
      "entropy": 5.729648447036743,
      "epoch": 0.5969562531588701,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004973387042203786,
      "loss": 5.7334,
      "mean_token_accuracy": 0.1702040694653988,
      "num_tokens": 11866194.0,
      "step": 5315
    },
    {
      "entropy": 5.672245216369629,
      "epoch": 0.5975178300668277,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004973325438957704,
      "loss": 5.5636,
      "mean_token_accuracy": 0.17379278242588042,
      "num_tokens": 11877224.0,
      "step": 5320
    },
    {
      "entropy": 5.586542654037475,
      "epoch": 0.5980794069747852,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004973263764920079,
      "loss": 5.4943,
      "mean_token_accuracy": 0.17458819299936296,
      "num_tokens": 11887410.0,
      "step": 5325
    },
    {
      "entropy": 5.5934830665588375,
      "epoch": 0.5986409838827428,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004973202020092871,
      "loss": 5.6495,
      "mean_token_accuracy": 0.1742006853222847,
      "num_tokens": 11898816.0,
      "step": 5330
    },
    {
      "entropy": 5.703639984130859,
      "epoch": 0.5992025607907003,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000497314020447805,
      "loss": 5.6084,
      "mean_token_accuracy": 0.17020292431116105,
      "num_tokens": 11909220.0,
      "step": 5335
    },
    {
      "entropy": 5.71856722831726,
      "epoch": 0.5997641376986579,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004973078318077583,
      "loss": 5.6072,
      "mean_token_accuracy": 0.17614903450012206,
      "num_tokens": 11919584.0,
      "step": 5340
    },
    {
      "entropy": 5.5959783554077145,
      "epoch": 0.6003257146066153,
      "grad_norm": 1.140625,
      "learning_rate": 0.000497301636089344,
      "loss": 5.6118,
      "mean_token_accuracy": 0.17212496548891068,
      "num_tokens": 11929298.0,
      "step": 5345
    },
    {
      "entropy": 5.660267543792725,
      "epoch": 0.6008872915145729,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004972954332927592,
      "loss": 5.6009,
      "mean_token_accuracy": 0.1719886213541031,
      "num_tokens": 11939809.0,
      "step": 5350
    },
    {
      "entropy": 5.671508884429931,
      "epoch": 0.6014488684225304,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004972892234182018,
      "loss": 5.4893,
      "mean_token_accuracy": 0.18083265721797942,
      "num_tokens": 11950292.0,
      "step": 5355
    },
    {
      "entropy": 5.556065845489502,
      "epoch": 0.602010445330488,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004972830064658692,
      "loss": 5.5023,
      "mean_token_accuracy": 0.17756835669279097,
      "num_tokens": 11961774.0,
      "step": 5360
    },
    {
      "entropy": 5.66449556350708,
      "epoch": 0.6025720222384455,
      "grad_norm": 1.0,
      "learning_rate": 0.0004972767824359596,
      "loss": 5.6029,
      "mean_token_accuracy": 0.1695825606584549,
      "num_tokens": 11973351.0,
      "step": 5365
    },
    {
      "entropy": 5.555609703063965,
      "epoch": 0.6031335991464031,
      "grad_norm": 1.0,
      "learning_rate": 0.0004972705513286709,
      "loss": 5.4777,
      "mean_token_accuracy": 0.17793078422546388,
      "num_tokens": 11984472.0,
      "step": 5370
    },
    {
      "entropy": 5.632693433761597,
      "epoch": 0.6036951760543606,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004972643131442017,
      "loss": 5.5798,
      "mean_token_accuracy": 0.17433574795722961,
      "num_tokens": 11994271.0,
      "step": 5375
    },
    {
      "entropy": 5.790684080123901,
      "epoch": 0.6042567529623182,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004972580678827506,
      "loss": 5.6862,
      "mean_token_accuracy": 0.16546384394168853,
      "num_tokens": 12007647.0,
      "step": 5380
    },
    {
      "entropy": 5.731524085998535,
      "epoch": 0.6048183298702757,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004972518155445163,
      "loss": 5.6494,
      "mean_token_accuracy": 0.17192945629358292,
      "num_tokens": 12020331.0,
      "step": 5385
    },
    {
      "entropy": 5.588660812377929,
      "epoch": 0.6053799067782333,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004972455561296982,
      "loss": 5.6414,
      "mean_token_accuracy": 0.16670804917812349,
      "num_tokens": 12031616.0,
      "step": 5390
    },
    {
      "entropy": 5.76715316772461,
      "epoch": 0.6059414836861908,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004972392896384952,
      "loss": 5.5916,
      "mean_token_accuracy": 0.1751255512237549,
      "num_tokens": 12044107.0,
      "step": 5395
    },
    {
      "entropy": 5.657217121124267,
      "epoch": 0.6065030605941484,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004972330160711071,
      "loss": 5.5721,
      "mean_token_accuracy": 0.17411974668502808,
      "num_tokens": 12055747.0,
      "step": 5400
    },
    {
      "entropy": 5.6032792091369625,
      "epoch": 0.6070646375021059,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004972267354277336,
      "loss": 5.5794,
      "mean_token_accuracy": 0.1721959114074707,
      "num_tokens": 12067849.0,
      "step": 5405
    },
    {
      "entropy": 5.683909177780151,
      "epoch": 0.6076262144100635,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004972204477085746,
      "loss": 5.6133,
      "mean_token_accuracy": 0.17528761625289918,
      "num_tokens": 12078534.0,
      "step": 5410
    },
    {
      "entropy": 5.559306383132935,
      "epoch": 0.608187791318021,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004972141529138303,
      "loss": 5.5438,
      "mean_token_accuracy": 0.1765257254242897,
      "num_tokens": 12089534.0,
      "step": 5415
    },
    {
      "entropy": 5.594197463989258,
      "epoch": 0.6087493682259786,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004972078510437013,
      "loss": 5.4506,
      "mean_token_accuracy": 0.19235396534204482,
      "num_tokens": 12099513.0,
      "step": 5420
    },
    {
      "entropy": 5.698856830596924,
      "epoch": 0.6093109451339361,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004972015420983882,
      "loss": 5.6569,
      "mean_token_accuracy": 0.17706738412380219,
      "num_tokens": 12110106.0,
      "step": 5425
    },
    {
      "entropy": 5.603834581375122,
      "epoch": 0.6098725220418937,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004971952260780915,
      "loss": 5.5443,
      "mean_token_accuracy": 0.1780934900045395,
      "num_tokens": 12121302.0,
      "step": 5430
    },
    {
      "entropy": 5.620524740219116,
      "epoch": 0.6104340989498512,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000497188902983013,
      "loss": 5.5821,
      "mean_token_accuracy": 0.1753602981567383,
      "num_tokens": 12132879.0,
      "step": 5435
    },
    {
      "entropy": 5.651505422592163,
      "epoch": 0.6109956758578087,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004971825728133535,
      "loss": 5.5151,
      "mean_token_accuracy": 0.1849575787782669,
      "num_tokens": 12143597.0,
      "step": 5440
    },
    {
      "entropy": 5.654022741317749,
      "epoch": 0.6115572527657662,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004971762355693146,
      "loss": 5.5902,
      "mean_token_accuracy": 0.1785195827484131,
      "num_tokens": 12156036.0,
      "step": 5445
    },
    {
      "entropy": 5.616942453384399,
      "epoch": 0.6121188296737238,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004971698912510983,
      "loss": 5.6135,
      "mean_token_accuracy": 0.17735974937677385,
      "num_tokens": 12167305.0,
      "step": 5450
    },
    {
      "entropy": 5.668393564224243,
      "epoch": 0.6126804065816813,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004971635398589066,
      "loss": 5.5772,
      "mean_token_accuracy": 0.17645054757595063,
      "num_tokens": 12176772.0,
      "step": 5455
    },
    {
      "entropy": 5.651277637481689,
      "epoch": 0.6132419834896389,
      "grad_norm": 1.125,
      "learning_rate": 0.0004971571813929415,
      "loss": 5.5844,
      "mean_token_accuracy": 0.17520664632320404,
      "num_tokens": 12187852.0,
      "step": 5460
    },
    {
      "entropy": 5.670183801651001,
      "epoch": 0.6138035603975964,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004971508158534056,
      "loss": 5.6387,
      "mean_token_accuracy": 0.17312638610601425,
      "num_tokens": 12198574.0,
      "step": 5465
    },
    {
      "entropy": 5.5702825546264645,
      "epoch": 0.614365137305554,
      "grad_norm": 1.0,
      "learning_rate": 0.0004971444432405015,
      "loss": 5.418,
      "mean_token_accuracy": 0.18247187584638597,
      "num_tokens": 12209463.0,
      "step": 5470
    },
    {
      "entropy": 5.700192022323608,
      "epoch": 0.6149267142135115,
      "grad_norm": 1.0,
      "learning_rate": 0.0004971380635544323,
      "loss": 5.6809,
      "mean_token_accuracy": 0.17231896817684172,
      "num_tokens": 12220993.0,
      "step": 5475
    },
    {
      "entropy": 5.669808197021484,
      "epoch": 0.6154882911214691,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004971316767954009,
      "loss": 5.5519,
      "mean_token_accuracy": 0.18184880763292313,
      "num_tokens": 12232648.0,
      "step": 5480
    },
    {
      "entropy": 5.6592018604278564,
      "epoch": 0.6160498680294266,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004971252829636108,
      "loss": 5.5922,
      "mean_token_accuracy": 0.17842979580163956,
      "num_tokens": 12243931.0,
      "step": 5485
    },
    {
      "entropy": 5.681622648239136,
      "epoch": 0.6166114449373842,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004971188820592656,
      "loss": 5.593,
      "mean_token_accuracy": 0.1769217938184738,
      "num_tokens": 12254821.0,
      "step": 5490
    },
    {
      "entropy": 5.565288066864014,
      "epoch": 0.6171730218453417,
      "grad_norm": 1.078125,
      "learning_rate": 0.000497112474082569,
      "loss": 5.5907,
      "mean_token_accuracy": 0.17548164278268813,
      "num_tokens": 12265440.0,
      "step": 5495
    },
    {
      "entropy": 5.718973827362061,
      "epoch": 0.6177345987532993,
      "grad_norm": 1.09375,
      "learning_rate": 0.000497106059033725,
      "loss": 5.5731,
      "mean_token_accuracy": 0.17688217163085937,
      "num_tokens": 12276636.0,
      "step": 5500
    },
    {
      "entropy": 5.622245597839355,
      "epoch": 0.6182961756612568,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004970996369129381,
      "loss": 5.4534,
      "mean_token_accuracy": 0.18576505780220032,
      "num_tokens": 12287404.0,
      "step": 5505
    },
    {
      "entropy": 5.563075256347656,
      "epoch": 0.6188577525692144,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004970932077204125,
      "loss": 5.6683,
      "mean_token_accuracy": 0.1664884641766548,
      "num_tokens": 12299283.0,
      "step": 5510
    },
    {
      "entropy": 5.727330064773559,
      "epoch": 0.6194193294771719,
      "grad_norm": 1.109375,
      "learning_rate": 0.000497086771456353,
      "loss": 5.581,
      "mean_token_accuracy": 0.16977208703756333,
      "num_tokens": 12310281.0,
      "step": 5515
    },
    {
      "entropy": 5.663081502914428,
      "epoch": 0.6199809063851295,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004970803281209646,
      "loss": 5.5978,
      "mean_token_accuracy": 0.17746423780918122,
      "num_tokens": 12320839.0,
      "step": 5520
    },
    {
      "entropy": 5.533584928512573,
      "epoch": 0.620542483293087,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004970738777144524,
      "loss": 5.5409,
      "mean_token_accuracy": 0.17699830383062362,
      "num_tokens": 12332845.0,
      "step": 5525
    },
    {
      "entropy": 5.586899328231811,
      "epoch": 0.6211040602010446,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004970674202370218,
      "loss": 5.4627,
      "mean_token_accuracy": 0.18716193437576295,
      "num_tokens": 12343327.0,
      "step": 5530
    },
    {
      "entropy": 5.609989213943481,
      "epoch": 0.621665637109002,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004970609556888785,
      "loss": 5.5949,
      "mean_token_accuracy": 0.17614777088165284,
      "num_tokens": 12353224.0,
      "step": 5535
    },
    {
      "entropy": 5.603342199325562,
      "epoch": 0.6222272140169596,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004970544840702282,
      "loss": 5.495,
      "mean_token_accuracy": 0.1759045422077179,
      "num_tokens": 12363981.0,
      "step": 5540
    },
    {
      "entropy": 5.5095680236816404,
      "epoch": 0.6227887909249171,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004970480053812771,
      "loss": 5.5194,
      "mean_token_accuracy": 0.18178894519805908,
      "num_tokens": 12374769.0,
      "step": 5545
    },
    {
      "entropy": 5.761261701583862,
      "epoch": 0.6233503678328747,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004970415196222312,
      "loss": 5.7276,
      "mean_token_accuracy": 0.17217909097671508,
      "num_tokens": 12387247.0,
      "step": 5550
    },
    {
      "entropy": 5.685073900222778,
      "epoch": 0.6239119447408322,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004970350267932973,
      "loss": 5.5811,
      "mean_token_accuracy": 0.18184039294719695,
      "num_tokens": 12398347.0,
      "step": 5555
    },
    {
      "entropy": 5.448342895507812,
      "epoch": 0.6244735216487898,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000497028526894682,
      "loss": 5.483,
      "mean_token_accuracy": 0.1800021141767502,
      "num_tokens": 12409763.0,
      "step": 5560
    },
    {
      "entropy": 5.711596441268921,
      "epoch": 0.6250350985567473,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004970220199265923,
      "loss": 5.5847,
      "mean_token_accuracy": 0.17855364829301834,
      "num_tokens": 12420658.0,
      "step": 5565
    },
    {
      "entropy": 5.741230106353759,
      "epoch": 0.6255966754647049,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004970155058892354,
      "loss": 5.6642,
      "mean_token_accuracy": 0.1671621546149254,
      "num_tokens": 12433959.0,
      "step": 5570
    },
    {
      "entropy": 5.622898769378662,
      "epoch": 0.6261582523726624,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004970089847828187,
      "loss": 5.658,
      "mean_token_accuracy": 0.16862257421016694,
      "num_tokens": 12444756.0,
      "step": 5575
    },
    {
      "entropy": 5.6203995704650875,
      "epoch": 0.62671982928062,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004970024566075497,
      "loss": 5.5605,
      "mean_token_accuracy": 0.18181279450654983,
      "num_tokens": 12456114.0,
      "step": 5580
    },
    {
      "entropy": 5.70815896987915,
      "epoch": 0.6272814061885775,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004969959213636365,
      "loss": 5.6191,
      "mean_token_accuracy": 0.17511922717094422,
      "num_tokens": 12466927.0,
      "step": 5585
    },
    {
      "entropy": 5.722125434875489,
      "epoch": 0.6278429830965351,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004969893790512871,
      "loss": 5.6545,
      "mean_token_accuracy": 0.16113144904375076,
      "num_tokens": 12477834.0,
      "step": 5590
    },
    {
      "entropy": 5.6834611892700195,
      "epoch": 0.6284045600044926,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004969828296707097,
      "loss": 5.5586,
      "mean_token_accuracy": 0.1791980355978012,
      "num_tokens": 12488219.0,
      "step": 5595
    },
    {
      "entropy": 5.713558530807495,
      "epoch": 0.6289661369124502,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004969762732221129,
      "loss": 5.6641,
      "mean_token_accuracy": 0.18148195445537568,
      "num_tokens": 12498947.0,
      "step": 5600
    },
    {
      "entropy": 5.559450149536133,
      "epoch": 0.6295277138204077,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004969697097057054,
      "loss": 5.5685,
      "mean_token_accuracy": 0.17765460163354874,
      "num_tokens": 12510142.0,
      "step": 5605
    },
    {
      "entropy": 5.717630434036255,
      "epoch": 0.6300892907283653,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004969631391216964,
      "loss": 5.5866,
      "mean_token_accuracy": 0.17643727213144303,
      "num_tokens": 12521350.0,
      "step": 5610
    },
    {
      "entropy": 5.418944931030273,
      "epoch": 0.6306508676363228,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004969565614702948,
      "loss": 5.4369,
      "mean_token_accuracy": 0.17893686443567275,
      "num_tokens": 12532579.0,
      "step": 5615
    },
    {
      "entropy": 5.705151700973511,
      "epoch": 0.6312124445442804,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004969499767517104,
      "loss": 5.5417,
      "mean_token_accuracy": 0.17601590752601623,
      "num_tokens": 12542885.0,
      "step": 5620
    },
    {
      "entropy": 5.754283237457275,
      "epoch": 0.6317740214522379,
      "grad_norm": 1.0,
      "learning_rate": 0.0004969433849661525,
      "loss": 5.7003,
      "mean_token_accuracy": 0.16661942452192308,
      "num_tokens": 12554881.0,
      "step": 5625
    },
    {
      "entropy": 5.677257251739502,
      "epoch": 0.6323355983601955,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004969367861138313,
      "loss": 5.6471,
      "mean_token_accuracy": 0.1723466694355011,
      "num_tokens": 12564626.0,
      "step": 5630
    },
    {
      "entropy": 5.643770456314087,
      "epoch": 0.6328971752681529,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004969301801949565,
      "loss": 5.5037,
      "mean_token_accuracy": 0.17835188060998916,
      "num_tokens": 12576180.0,
      "step": 5635
    },
    {
      "entropy": 5.563134002685547,
      "epoch": 0.6334587521761105,
      "grad_norm": 1.0625,
      "learning_rate": 0.000496923567209739,
      "loss": 5.5774,
      "mean_token_accuracy": 0.174107725918293,
      "num_tokens": 12587487.0,
      "step": 5640
    },
    {
      "entropy": 5.745814180374145,
      "epoch": 0.634020329084068,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004969169471583889,
      "loss": 5.7229,
      "mean_token_accuracy": 0.16757968813180923,
      "num_tokens": 12598268.0,
      "step": 5645
    },
    {
      "entropy": 5.715684080123902,
      "epoch": 0.6345819059920256,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004969103200411171,
      "loss": 5.6033,
      "mean_token_accuracy": 0.17257933020591737,
      "num_tokens": 12609287.0,
      "step": 5650
    },
    {
      "entropy": 5.6591414451599125,
      "epoch": 0.6351434828999831,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004969036858581346,
      "loss": 5.5018,
      "mean_token_accuracy": 0.17563670724630356,
      "num_tokens": 12620317.0,
      "step": 5655
    },
    {
      "entropy": 5.607477045059204,
      "epoch": 0.6357050598079407,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004968970446096527,
      "loss": 5.5794,
      "mean_token_accuracy": 0.18153921514749527,
      "num_tokens": 12631420.0,
      "step": 5660
    },
    {
      "entropy": 5.591618061065674,
      "epoch": 0.6362666367158982,
      "grad_norm": 1.28125,
      "learning_rate": 0.000496890396295883,
      "loss": 5.5336,
      "mean_token_accuracy": 0.18361421972513198,
      "num_tokens": 12641319.0,
      "step": 5665
    },
    {
      "entropy": 5.704915857315063,
      "epoch": 0.6368282136238558,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004968837409170368,
      "loss": 5.6281,
      "mean_token_accuracy": 0.1668785482645035,
      "num_tokens": 12651662.0,
      "step": 5670
    },
    {
      "entropy": 5.696206760406494,
      "epoch": 0.6373897905318133,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004968770784733263,
      "loss": 5.657,
      "mean_token_accuracy": 0.1661613643169403,
      "num_tokens": 12663488.0,
      "step": 5675
    },
    {
      "entropy": 5.607398462295532,
      "epoch": 0.6379513674397709,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004968704089649636,
      "loss": 5.561,
      "mean_token_accuracy": 0.17629160284996032,
      "num_tokens": 12675848.0,
      "step": 5680
    },
    {
      "entropy": 5.668233633041382,
      "epoch": 0.6385129443477284,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004968637323921611,
      "loss": 5.5559,
      "mean_token_accuracy": 0.17808516770601274,
      "num_tokens": 12686541.0,
      "step": 5685
    },
    {
      "entropy": 5.621880674362183,
      "epoch": 0.639074521255686,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004968570487551312,
      "loss": 5.5768,
      "mean_token_accuracy": 0.1741996854543686,
      "num_tokens": 12698001.0,
      "step": 5690
    },
    {
      "entropy": 5.6410235404968265,
      "epoch": 0.6396360981636435,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004968503580540869,
      "loss": 5.5873,
      "mean_token_accuracy": 0.17105681151151658,
      "num_tokens": 12709069.0,
      "step": 5695
    },
    {
      "entropy": 5.700166130065918,
      "epoch": 0.6401976750716011,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004968436602892411,
      "loss": 5.5645,
      "mean_token_accuracy": 0.1719864621758461,
      "num_tokens": 12719916.0,
      "step": 5700
    },
    {
      "entropy": 5.637115240097046,
      "epoch": 0.6407592519795586,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004968369554608071,
      "loss": 5.7004,
      "mean_token_accuracy": 0.16285409182310104,
      "num_tokens": 12732079.0,
      "step": 5705
    },
    {
      "entropy": 5.82158350944519,
      "epoch": 0.6413208288875162,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004968302435689985,
      "loss": 5.665,
      "mean_token_accuracy": 0.17319654673337936,
      "num_tokens": 12742708.0,
      "step": 5710
    },
    {
      "entropy": 5.7179646492004395,
      "epoch": 0.6418824057954737,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004968235246140288,
      "loss": 5.6111,
      "mean_token_accuracy": 0.1778605744242668,
      "num_tokens": 12753952.0,
      "step": 5715
    },
    {
      "entropy": 5.531432247161865,
      "epoch": 0.6424439827034313,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004968167985961121,
      "loss": 5.5362,
      "mean_token_accuracy": 0.17749983221292495,
      "num_tokens": 12765226.0,
      "step": 5720
    },
    {
      "entropy": 5.591689348220825,
      "epoch": 0.6430055596113888,
      "grad_norm": 1.0,
      "learning_rate": 0.0004968100655154625,
      "loss": 5.466,
      "mean_token_accuracy": 0.18534375578165055,
      "num_tokens": 12776455.0,
      "step": 5725
    },
    {
      "entropy": 5.59759750366211,
      "epoch": 0.6435671365193463,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004968033253722943,
      "loss": 5.4942,
      "mean_token_accuracy": 0.17470683455467223,
      "num_tokens": 12787054.0,
      "step": 5730
    },
    {
      "entropy": 5.544394207000733,
      "epoch": 0.6441287134273038,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004967965781668223,
      "loss": 5.6279,
      "mean_token_accuracy": 0.17692292779684066,
      "num_tokens": 12798223.0,
      "step": 5735
    },
    {
      "entropy": 5.735273742675782,
      "epoch": 0.6446902903352614,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004967898238992613,
      "loss": 5.4996,
      "mean_token_accuracy": 0.17831794917583466,
      "num_tokens": 12809388.0,
      "step": 5740
    },
    {
      "entropy": 5.609306335449219,
      "epoch": 0.6452518672432189,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004967830625698262,
      "loss": 5.611,
      "mean_token_accuracy": 0.1793173745274544,
      "num_tokens": 12820116.0,
      "step": 5745
    },
    {
      "entropy": 5.588194179534912,
      "epoch": 0.6458134441511765,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004967762941787323,
      "loss": 5.542,
      "mean_token_accuracy": 0.18204563707113267,
      "num_tokens": 12831590.0,
      "step": 5750
    },
    {
      "entropy": 5.747331142425537,
      "epoch": 0.646375021059134,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004967695187261952,
      "loss": 5.6858,
      "mean_token_accuracy": 0.1728877142071724,
      "num_tokens": 12843510.0,
      "step": 5755
    },
    {
      "entropy": 5.608614253997803,
      "epoch": 0.6469365979670916,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004967627362124306,
      "loss": 5.5609,
      "mean_token_accuracy": 0.17334567308425902,
      "num_tokens": 12854907.0,
      "step": 5760
    },
    {
      "entropy": 5.587041234970092,
      "epoch": 0.6474981748750491,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004967559466376546,
      "loss": 5.5234,
      "mean_token_accuracy": 0.17716541588306428,
      "num_tokens": 12865753.0,
      "step": 5765
    },
    {
      "entropy": 5.554595422744751,
      "epoch": 0.6480597517830067,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004967491500020831,
      "loss": 5.4336,
      "mean_token_accuracy": 0.18294782489538192,
      "num_tokens": 12877116.0,
      "step": 5770
    },
    {
      "entropy": 5.483822631835937,
      "epoch": 0.6486213286909642,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004967423463059328,
      "loss": 5.4352,
      "mean_token_accuracy": 0.17747284173965455,
      "num_tokens": 12887685.0,
      "step": 5775
    },
    {
      "entropy": 5.638487148284912,
      "epoch": 0.6491829055989218,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004967355355494201,
      "loss": 5.5896,
      "mean_token_accuracy": 0.17525444626808168,
      "num_tokens": 12898519.0,
      "step": 5780
    },
    {
      "entropy": 5.663454532623291,
      "epoch": 0.6497444825068793,
      "grad_norm": 1.125,
      "learning_rate": 0.000496728717732762,
      "loss": 5.5181,
      "mean_token_accuracy": 0.18272124230861664,
      "num_tokens": 12908400.0,
      "step": 5785
    },
    {
      "entropy": 5.527053546905518,
      "epoch": 0.6503060594148369,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004967218928561755,
      "loss": 5.4906,
      "mean_token_accuracy": 0.179681658744812,
      "num_tokens": 12919159.0,
      "step": 5790
    },
    {
      "entropy": 5.5719407081604,
      "epoch": 0.6508676363227944,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004967150609198779,
      "loss": 5.4453,
      "mean_token_accuracy": 0.1843953549861908,
      "num_tokens": 12929838.0,
      "step": 5795
    },
    {
      "entropy": 5.621962022781372,
      "epoch": 0.651429213230752,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004967082219240869,
      "loss": 5.5878,
      "mean_token_accuracy": 0.17661426216363907,
      "num_tokens": 12939976.0,
      "step": 5800
    },
    {
      "entropy": 5.627374505996704,
      "epoch": 0.6519907901387095,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004967013758690199,
      "loss": 5.5756,
      "mean_token_accuracy": 0.18177225887775422,
      "num_tokens": 12950493.0,
      "step": 5805
    },
    {
      "entropy": 5.550268268585205,
      "epoch": 0.6525523670466671,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004966945227548953,
      "loss": 5.5494,
      "mean_token_accuracy": 0.17592897564172744,
      "num_tokens": 12961267.0,
      "step": 5810
    },
    {
      "entropy": 5.636107158660889,
      "epoch": 0.6531139439546246,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004966876625819309,
      "loss": 5.5945,
      "mean_token_accuracy": 0.17396701127290726,
      "num_tokens": 12972720.0,
      "step": 5815
    },
    {
      "entropy": 5.736133337020874,
      "epoch": 0.6536755208625822,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004966807953503456,
      "loss": 5.6785,
      "mean_token_accuracy": 0.17026133388280867,
      "num_tokens": 12984025.0,
      "step": 5820
    },
    {
      "entropy": 5.66474461555481,
      "epoch": 0.6542370977705396,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004966739210603578,
      "loss": 5.5474,
      "mean_token_accuracy": 0.18265554457902908,
      "num_tokens": 12994422.0,
      "step": 5825
    },
    {
      "entropy": 5.683717918395996,
      "epoch": 0.6547986746784972,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004966670397121862,
      "loss": 5.7054,
      "mean_token_accuracy": 0.16815723925828935,
      "num_tokens": 13005040.0,
      "step": 5830
    },
    {
      "entropy": 5.588709115982056,
      "epoch": 0.6553602515864547,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004966601513060502,
      "loss": 5.4606,
      "mean_token_accuracy": 0.17782158851623536,
      "num_tokens": 13015720.0,
      "step": 5835
    },
    {
      "entropy": 5.605904150009155,
      "epoch": 0.6559218284944123,
      "grad_norm": 1.171875,
      "learning_rate": 0.000496653255842169,
      "loss": 5.4243,
      "mean_token_accuracy": 0.18126147985458374,
      "num_tokens": 13026269.0,
      "step": 5840
    },
    {
      "entropy": 5.58271598815918,
      "epoch": 0.6564834054023698,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004966463533207621,
      "loss": 5.6139,
      "mean_token_accuracy": 0.1731554314494133,
      "num_tokens": 13037102.0,
      "step": 5845
    },
    {
      "entropy": 5.605773305892944,
      "epoch": 0.6570449823103274,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004966394437420493,
      "loss": 5.4847,
      "mean_token_accuracy": 0.18165120780467986,
      "num_tokens": 13047149.0,
      "step": 5850
    },
    {
      "entropy": 5.592484378814698,
      "epoch": 0.6576065592182849,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004966325271062508,
      "loss": 5.4902,
      "mean_token_accuracy": 0.18418198078870773,
      "num_tokens": 13058196.0,
      "step": 5855
    },
    {
      "entropy": 5.549489974975586,
      "epoch": 0.6581681361262425,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004966256034135866,
      "loss": 5.5066,
      "mean_token_accuracy": 0.17483980655670167,
      "num_tokens": 13067897.0,
      "step": 5860
    },
    {
      "entropy": 5.6510625839233395,
      "epoch": 0.6587297130342,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004966186726642772,
      "loss": 5.601,
      "mean_token_accuracy": 0.17503603547811508,
      "num_tokens": 13079720.0,
      "step": 5865
    },
    {
      "entropy": 5.553473663330078,
      "epoch": 0.6592912899421576,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004966117348585434,
      "loss": 5.4526,
      "mean_token_accuracy": 0.17671313732862473,
      "num_tokens": 13090995.0,
      "step": 5870
    },
    {
      "entropy": 5.555782318115234,
      "epoch": 0.6598528668501151,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004966047899966059,
      "loss": 5.4995,
      "mean_token_accuracy": 0.18041776716709138,
      "num_tokens": 13101945.0,
      "step": 5875
    },
    {
      "entropy": 5.6111548900604244,
      "epoch": 0.6604144437580727,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000496597838078686,
      "loss": 5.4583,
      "mean_token_accuracy": 0.1832595482468605,
      "num_tokens": 13112869.0,
      "step": 5880
    },
    {
      "entropy": 5.653467321395874,
      "epoch": 0.6609760206660302,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000496590879105005,
      "loss": 5.5377,
      "mean_token_accuracy": 0.1737200826406479,
      "num_tokens": 13123242.0,
      "step": 5885
    },
    {
      "entropy": 5.692622661590576,
      "epoch": 0.6615375975739878,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004965839130757845,
      "loss": 5.6504,
      "mean_token_accuracy": 0.16568809598684311,
      "num_tokens": 13133981.0,
      "step": 5890
    },
    {
      "entropy": 5.600550270080566,
      "epoch": 0.6620991744819453,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004965769399912462,
      "loss": 5.6044,
      "mean_token_accuracy": 0.17274861112236978,
      "num_tokens": 13144696.0,
      "step": 5895
    },
    {
      "entropy": 5.752182531356811,
      "epoch": 0.6626607513899029,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004965699598516122,
      "loss": 5.6717,
      "mean_token_accuracy": 0.17395811080932616,
      "num_tokens": 13156674.0,
      "step": 5900
    },
    {
      "entropy": 5.6548622131347654,
      "epoch": 0.6632223282978604,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004965629726571047,
      "loss": 5.5044,
      "mean_token_accuracy": 0.1797347277402878,
      "num_tokens": 13167757.0,
      "step": 5905
    },
    {
      "entropy": 5.641356420516968,
      "epoch": 0.663783905205818,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004965559784079463,
      "loss": 5.5814,
      "mean_token_accuracy": 0.17386945337057114,
      "num_tokens": 13177701.0,
      "step": 5910
    },
    {
      "entropy": 5.6706444263458256,
      "epoch": 0.6643454821137755,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004965489771043596,
      "loss": 5.6056,
      "mean_token_accuracy": 0.1673995077610016,
      "num_tokens": 13189840.0,
      "step": 5915
    },
    {
      "entropy": 5.465957355499268,
      "epoch": 0.664907059021733,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004965419687465676,
      "loss": 5.4098,
      "mean_token_accuracy": 0.18124558925628662,
      "num_tokens": 13200206.0,
      "step": 5920
    },
    {
      "entropy": 5.577658987045288,
      "epoch": 0.6654686359296905,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004965349533347934,
      "loss": 5.5136,
      "mean_token_accuracy": 0.18476506173610688,
      "num_tokens": 13211474.0,
      "step": 5925
    },
    {
      "entropy": 5.664955425262451,
      "epoch": 0.6660302128376481,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004965279308692602,
      "loss": 5.5777,
      "mean_token_accuracy": 0.1884651154279709,
      "num_tokens": 13223337.0,
      "step": 5930
    },
    {
      "entropy": 5.680084991455078,
      "epoch": 0.6665917897456056,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004965209013501919,
      "loss": 5.6529,
      "mean_token_accuracy": 0.17240284383296967,
      "num_tokens": 13234977.0,
      "step": 5935
    },
    {
      "entropy": 5.669816732406616,
      "epoch": 0.6671533666535632,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004965138647778122,
      "loss": 5.6202,
      "mean_token_accuracy": 0.18018104135990143,
      "num_tokens": 13245315.0,
      "step": 5940
    },
    {
      "entropy": 5.650509548187256,
      "epoch": 0.6677149435615207,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000496506821152345,
      "loss": 5.4993,
      "mean_token_accuracy": 0.18011848330497743,
      "num_tokens": 13255829.0,
      "step": 5945
    },
    {
      "entropy": 5.581755065917969,
      "epoch": 0.6682765204694783,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004964997704740147,
      "loss": 5.5492,
      "mean_token_accuracy": 0.17475591748952865,
      "num_tokens": 13266964.0,
      "step": 5950
    },
    {
      "entropy": 5.686089038848877,
      "epoch": 0.6688380973774358,
      "grad_norm": 1.171875,
      "learning_rate": 0.000496492712743046,
      "loss": 5.5911,
      "mean_token_accuracy": 0.174809929728508,
      "num_tokens": 13279416.0,
      "step": 5955
    },
    {
      "entropy": 5.6538951873779295,
      "epoch": 0.6693996742853934,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004964856479596632,
      "loss": 5.6308,
      "mean_token_accuracy": 0.17145811766386032,
      "num_tokens": 13290633.0,
      "step": 5960
    },
    {
      "entropy": 5.632832479476929,
      "epoch": 0.6699612511933509,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004964785761240915,
      "loss": 5.506,
      "mean_token_accuracy": 0.17835093289613724,
      "num_tokens": 13302236.0,
      "step": 5965
    },
    {
      "entropy": 5.652864074707031,
      "epoch": 0.6705228281013085,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004964714972365562,
      "loss": 5.5896,
      "mean_token_accuracy": 0.17846613228321076,
      "num_tokens": 13315356.0,
      "step": 5970
    },
    {
      "entropy": 5.429797649383545,
      "epoch": 0.671084405009266,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004964644112972824,
      "loss": 5.471,
      "mean_token_accuracy": 0.1815232589840889,
      "num_tokens": 13326037.0,
      "step": 5975
    },
    {
      "entropy": 5.597888708114624,
      "epoch": 0.6716459819172236,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004964573183064959,
      "loss": 5.5461,
      "mean_token_accuracy": 0.1743439257144928,
      "num_tokens": 13336278.0,
      "step": 5980
    },
    {
      "entropy": 5.718096590042114,
      "epoch": 0.6722075588251811,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004964502182644225,
      "loss": 5.6411,
      "mean_token_accuracy": 0.1770189583301544,
      "num_tokens": 13348147.0,
      "step": 5985
    },
    {
      "entropy": 5.6348189353942875,
      "epoch": 0.6727691357331387,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004964431111712883,
      "loss": 5.5684,
      "mean_token_accuracy": 0.17162202000617982,
      "num_tokens": 13360181.0,
      "step": 5990
    },
    {
      "entropy": 5.72844705581665,
      "epoch": 0.6733307126410962,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004964359970273195,
      "loss": 5.6547,
      "mean_token_accuracy": 0.17094038128852845,
      "num_tokens": 13372258.0,
      "step": 5995
    },
    {
      "entropy": 5.685292148590088,
      "epoch": 0.6738922895490538,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004964288758327426,
      "loss": 5.5639,
      "mean_token_accuracy": 0.17274709343910216,
      "num_tokens": 13384332.0,
      "step": 6000
    },
    {
      "epoch": 0.6738922895490538,
      "eval_entropy": 5.501720400547779,
      "eval_loss": 5.592982292175293,
      "eval_mean_token_accuracy": 0.181518482865402,
      "eval_num_tokens": 13384332.0,
      "eval_runtime": 24.2081,
      "eval_samples_per_second": 1281.101,
      "eval_steps_per_second": 160.153,
      "step": 6000
    },
    {
      "entropy": 5.618107748031616,
      "epoch": 0.6744538664570113,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004964217475877846,
      "loss": 5.5533,
      "mean_token_accuracy": 0.17436386197805404,
      "num_tokens": 13396345.0,
      "step": 6005
    },
    {
      "entropy": 5.5280087947845455,
      "epoch": 0.6750154433649689,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000496414612292672,
      "loss": 5.4029,
      "mean_token_accuracy": 0.18110353648662567,
      "num_tokens": 13407325.0,
      "step": 6010
    },
    {
      "entropy": 5.646691608428955,
      "epoch": 0.6755770202729264,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004964074699476325,
      "loss": 5.5251,
      "mean_token_accuracy": 0.18078237920999526,
      "num_tokens": 13417476.0,
      "step": 6015
    },
    {
      "entropy": 5.538549613952637,
      "epoch": 0.6761385971808839,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004964003205528933,
      "loss": 5.5597,
      "mean_token_accuracy": 0.17804979383945466,
      "num_tokens": 13428792.0,
      "step": 6020
    },
    {
      "entropy": 5.663299608230591,
      "epoch": 0.6767001740888414,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004963931641086819,
      "loss": 5.6011,
      "mean_token_accuracy": 0.17148593813180923,
      "num_tokens": 13440043.0,
      "step": 6025
    },
    {
      "entropy": 5.714299488067627,
      "epoch": 0.677261750996799,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004963860006152262,
      "loss": 5.5503,
      "mean_token_accuracy": 0.18272247016429902,
      "num_tokens": 13451637.0,
      "step": 6030
    },
    {
      "entropy": 5.672242164611816,
      "epoch": 0.6778233279047565,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004963788300727546,
      "loss": 5.5296,
      "mean_token_accuracy": 0.17724456489086152,
      "num_tokens": 13463205.0,
      "step": 6035
    },
    {
      "entropy": 5.61183385848999,
      "epoch": 0.6783849048127141,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004963716524814951,
      "loss": 5.5812,
      "mean_token_accuracy": 0.17874963730573654,
      "num_tokens": 13474201.0,
      "step": 6040
    },
    {
      "entropy": 5.553326892852783,
      "epoch": 0.6789464817206716,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004963644678416763,
      "loss": 5.4711,
      "mean_token_accuracy": 0.18081430047750474,
      "num_tokens": 13485439.0,
      "step": 6045
    },
    {
      "entropy": 5.609672117233276,
      "epoch": 0.6795080586286292,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000496357276153527,
      "loss": 5.677,
      "mean_token_accuracy": 0.1681491807103157,
      "num_tokens": 13497257.0,
      "step": 6050
    },
    {
      "entropy": 5.670541715621948,
      "epoch": 0.6800696355365867,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004963500774172763,
      "loss": 5.5205,
      "mean_token_accuracy": 0.17992660403251648,
      "num_tokens": 13508689.0,
      "step": 6055
    },
    {
      "entropy": 5.634726619720459,
      "epoch": 0.6806312124445443,
      "grad_norm": 1.140625,
      "learning_rate": 0.000496342871633153,
      "loss": 5.4843,
      "mean_token_accuracy": 0.1851092979311943,
      "num_tokens": 13519392.0,
      "step": 6060
    },
    {
      "entropy": 5.561260080337524,
      "epoch": 0.6811927893525018,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004963356588013871,
      "loss": 5.5287,
      "mean_token_accuracy": 0.18115695416927338,
      "num_tokens": 13530503.0,
      "step": 6065
    },
    {
      "entropy": 5.682749891281128,
      "epoch": 0.6817543662604594,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004963284389222078,
      "loss": 5.61,
      "mean_token_accuracy": 0.17627547383308412,
      "num_tokens": 13541668.0,
      "step": 6070
    },
    {
      "entropy": 5.56047830581665,
      "epoch": 0.6823159431684169,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004963212119958453,
      "loss": 5.4538,
      "mean_token_accuracy": 0.18298290818929672,
      "num_tokens": 13552692.0,
      "step": 6075
    },
    {
      "entropy": 5.580601358413697,
      "epoch": 0.6828775200763745,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004963139780225295,
      "loss": 5.5376,
      "mean_token_accuracy": 0.17632260024547577,
      "num_tokens": 13563813.0,
      "step": 6080
    },
    {
      "entropy": 5.668052339553833,
      "epoch": 0.683439096984332,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004963067370024909,
      "loss": 5.486,
      "mean_token_accuracy": 0.1808595821261406,
      "num_tokens": 13573818.0,
      "step": 6085
    },
    {
      "entropy": 5.603163623809815,
      "epoch": 0.6840006738922896,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004962994889359598,
      "loss": 5.5982,
      "mean_token_accuracy": 0.17281472235918044,
      "num_tokens": 13585416.0,
      "step": 6090
    },
    {
      "entropy": 5.5317394733428955,
      "epoch": 0.6845622508002471,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004962922338231673,
      "loss": 5.5021,
      "mean_token_accuracy": 0.18008041083812715,
      "num_tokens": 13596645.0,
      "step": 6095
    },
    {
      "entropy": 5.722293949127197,
      "epoch": 0.6851238277082047,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004962849716643441,
      "loss": 5.4517,
      "mean_token_accuracy": 0.18034529089927673,
      "num_tokens": 13607206.0,
      "step": 6100
    },
    {
      "entropy": 5.639233922958374,
      "epoch": 0.6856854046161622,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004962777024597216,
      "loss": 5.6667,
      "mean_token_accuracy": 0.1742106184363365,
      "num_tokens": 13619418.0,
      "step": 6105
    },
    {
      "entropy": 5.578310108184814,
      "epoch": 0.6862469815241198,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004962704262095314,
      "loss": 5.4494,
      "mean_token_accuracy": 0.18483205139636993,
      "num_tokens": 13630098.0,
      "step": 6110
    },
    {
      "entropy": 5.578888559341431,
      "epoch": 0.6868085584320772,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004962631429140048,
      "loss": 5.5845,
      "mean_token_accuracy": 0.17956829071044922,
      "num_tokens": 13641401.0,
      "step": 6115
    },
    {
      "entropy": 5.6445728778839115,
      "epoch": 0.6873701353400348,
      "grad_norm": 1.09375,
      "learning_rate": 0.000496255852573374,
      "loss": 5.5259,
      "mean_token_accuracy": 0.17465951293706894,
      "num_tokens": 13653099.0,
      "step": 6120
    },
    {
      "entropy": 5.729621410369873,
      "epoch": 0.6879317122479923,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004962485551878711,
      "loss": 5.5622,
      "mean_token_accuracy": 0.17419738918542862,
      "num_tokens": 13665381.0,
      "step": 6125
    },
    {
      "entropy": 5.626136779785156,
      "epoch": 0.6884932891559499,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004962412507577283,
      "loss": 5.561,
      "mean_token_accuracy": 0.1779366299510002,
      "num_tokens": 13676767.0,
      "step": 6130
    },
    {
      "entropy": 5.560002660751342,
      "epoch": 0.6890548660639074,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004962339392831783,
      "loss": 5.5566,
      "mean_token_accuracy": 0.17694362103939057,
      "num_tokens": 13688552.0,
      "step": 6135
    },
    {
      "entropy": 5.619175672531128,
      "epoch": 0.689616442971865,
      "grad_norm": 1.09375,
      "learning_rate": 0.000496226620764454,
      "loss": 5.5219,
      "mean_token_accuracy": 0.17612551301717758,
      "num_tokens": 13700160.0,
      "step": 6140
    },
    {
      "entropy": 5.625892305374146,
      "epoch": 0.6901780198798225,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000496219295201788,
      "loss": 5.529,
      "mean_token_accuracy": 0.17607801854610444,
      "num_tokens": 13710828.0,
      "step": 6145
    },
    {
      "entropy": 5.641735219955445,
      "epoch": 0.6907395967877801,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004962119625954141,
      "loss": 5.6466,
      "mean_token_accuracy": 0.1706322029232979,
      "num_tokens": 13721912.0,
      "step": 6150
    },
    {
      "entropy": 5.740659761428833,
      "epoch": 0.6913011736957376,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004962046229455653,
      "loss": 5.6702,
      "mean_token_accuracy": 0.1701244443655014,
      "num_tokens": 13733662.0,
      "step": 6155
    },
    {
      "entropy": 5.599560832977295,
      "epoch": 0.6918627506036952,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004961972762524757,
      "loss": 5.5011,
      "mean_token_accuracy": 0.1789718046784401,
      "num_tokens": 13743699.0,
      "step": 6160
    },
    {
      "entropy": 5.517910480499268,
      "epoch": 0.6924243275116527,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000496189922516379,
      "loss": 5.4846,
      "mean_token_accuracy": 0.1773657575249672,
      "num_tokens": 13754918.0,
      "step": 6165
    },
    {
      "entropy": 5.643311166763306,
      "epoch": 0.6929859044196103,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004961825617375093,
      "loss": 5.5331,
      "mean_token_accuracy": 0.18123019933700563,
      "num_tokens": 13764432.0,
      "step": 6170
    },
    {
      "entropy": 5.542857933044433,
      "epoch": 0.6935474813275678,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004961751939161011,
      "loss": 5.4477,
      "mean_token_accuracy": 0.1756064176559448,
      "num_tokens": 13777049.0,
      "step": 6175
    },
    {
      "entropy": 5.603497886657715,
      "epoch": 0.6941090582355254,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000496167819052389,
      "loss": 5.5517,
      "mean_token_accuracy": 0.17919934391975403,
      "num_tokens": 13788324.0,
      "step": 6180
    },
    {
      "entropy": 5.691399669647216,
      "epoch": 0.6946706351434829,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004961604371466077,
      "loss": 5.6361,
      "mean_token_accuracy": 0.17262129187583924,
      "num_tokens": 13800443.0,
      "step": 6185
    },
    {
      "entropy": 5.755781364440918,
      "epoch": 0.6952322120514405,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004961530481989923,
      "loss": 5.6756,
      "mean_token_accuracy": 0.17884768843650817,
      "num_tokens": 13811982.0,
      "step": 6190
    },
    {
      "entropy": 5.588156127929688,
      "epoch": 0.695793788959398,
      "grad_norm": 1.125,
      "learning_rate": 0.0004961456522097782,
      "loss": 5.4438,
      "mean_token_accuracy": 0.1794885739684105,
      "num_tokens": 13822005.0,
      "step": 6195
    },
    {
      "entropy": 5.532123804092407,
      "epoch": 0.6963553658673556,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004961382491792006,
      "loss": 5.4862,
      "mean_token_accuracy": 0.17643697708845138,
      "num_tokens": 13831943.0,
      "step": 6200
    },
    {
      "entropy": 5.711555671691895,
      "epoch": 0.6969169427753131,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004961308391074953,
      "loss": 5.6511,
      "mean_token_accuracy": 0.1729170188307762,
      "num_tokens": 13842517.0,
      "step": 6205
    },
    {
      "entropy": 5.636730909347534,
      "epoch": 0.6974785196832706,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004961234219948986,
      "loss": 5.5414,
      "mean_token_accuracy": 0.1776733160018921,
      "num_tokens": 13853636.0,
      "step": 6210
    },
    {
      "entropy": 5.567195415496826,
      "epoch": 0.6980400965912281,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004961159978416461,
      "loss": 5.4802,
      "mean_token_accuracy": 0.1734500765800476,
      "num_tokens": 13864378.0,
      "step": 6215
    },
    {
      "entropy": 5.625928211212158,
      "epoch": 0.6986016734991857,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004961085666479746,
      "loss": 5.5175,
      "mean_token_accuracy": 0.1804937392473221,
      "num_tokens": 13874964.0,
      "step": 6220
    },
    {
      "entropy": 5.54618010520935,
      "epoch": 0.6991632504071432,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004961011284141206,
      "loss": 5.4804,
      "mean_token_accuracy": 0.17753927409648895,
      "num_tokens": 13885852.0,
      "step": 6225
    },
    {
      "entropy": 5.561345338821411,
      "epoch": 0.6997248273151008,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004960936831403209,
      "loss": 5.5519,
      "mean_token_accuracy": 0.17466597408056259,
      "num_tokens": 13896101.0,
      "step": 6230
    },
    {
      "entropy": 5.604340124130249,
      "epoch": 0.7002864042230583,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004960862308268126,
      "loss": 5.4392,
      "mean_token_accuracy": 0.18222669661045074,
      "num_tokens": 13907129.0,
      "step": 6235
    },
    {
      "entropy": 5.617907619476318,
      "epoch": 0.7008479811310159,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004960787714738329,
      "loss": 5.6198,
      "mean_token_accuracy": 0.17847588360309602,
      "num_tokens": 13918662.0,
      "step": 6240
    },
    {
      "entropy": 5.60418586730957,
      "epoch": 0.7014095580389734,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004960713050816194,
      "loss": 5.5826,
      "mean_token_accuracy": 0.17841585576534272,
      "num_tokens": 13930717.0,
      "step": 6245
    },
    {
      "entropy": 5.584712982177734,
      "epoch": 0.701971134946931,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004960638316504098,
      "loss": 5.4722,
      "mean_token_accuracy": 0.17472756803035736,
      "num_tokens": 13941299.0,
      "step": 6250
    },
    {
      "entropy": 5.735639905929565,
      "epoch": 0.7025327118548885,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004960563511804421,
      "loss": 5.5502,
      "mean_token_accuracy": 0.1777105763554573,
      "num_tokens": 13951807.0,
      "step": 6255
    },
    {
      "entropy": 5.556670236587524,
      "epoch": 0.7030942887628461,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004960488636719545,
      "loss": 5.4491,
      "mean_token_accuracy": 0.18050920814275742,
      "num_tokens": 13962994.0,
      "step": 6260
    },
    {
      "entropy": 5.487055253982544,
      "epoch": 0.7036558656708036,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004960413691251852,
      "loss": 5.4873,
      "mean_token_accuracy": 0.17508239299058914,
      "num_tokens": 13973731.0,
      "step": 6265
    },
    {
      "entropy": 5.6760049819946286,
      "epoch": 0.7042174425787612,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004960338675403731,
      "loss": 5.5542,
      "mean_token_accuracy": 0.1789546400308609,
      "num_tokens": 13983772.0,
      "step": 6270
    },
    {
      "entropy": 5.576485872268677,
      "epoch": 0.7047790194867187,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004960263589177568,
      "loss": 5.5639,
      "mean_token_accuracy": 0.18367017954587936,
      "num_tokens": 13994935.0,
      "step": 6275
    },
    {
      "entropy": 5.599701929092407,
      "epoch": 0.7053405963946763,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004960188432575757,
      "loss": 5.5262,
      "mean_token_accuracy": 0.18418762385845183,
      "num_tokens": 14006225.0,
      "step": 6280
    },
    {
      "entropy": 5.621382999420166,
      "epoch": 0.7059021733026338,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004960113205600688,
      "loss": 5.5526,
      "mean_token_accuracy": 0.18518788665533065,
      "num_tokens": 14017693.0,
      "step": 6285
    },
    {
      "entropy": 5.610233116149902,
      "epoch": 0.7064637502105914,
      "grad_norm": 1.125,
      "learning_rate": 0.0004960037908254757,
      "loss": 5.5083,
      "mean_token_accuracy": 0.1801917716860771,
      "num_tokens": 14027264.0,
      "step": 6290
    },
    {
      "entropy": 5.615220403671264,
      "epoch": 0.7070253271185489,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004959962540540361,
      "loss": 5.6,
      "mean_token_accuracy": 0.1790604904294014,
      "num_tokens": 14039422.0,
      "step": 6295
    },
    {
      "entropy": 5.603928327560425,
      "epoch": 0.7075869040265065,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004959887102459903,
      "loss": 5.5191,
      "mean_token_accuracy": 0.17634472846984864,
      "num_tokens": 14051619.0,
      "step": 6300
    },
    {
      "entropy": 5.4765581607818605,
      "epoch": 0.7081484809344639,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004959811594015781,
      "loss": 5.519,
      "mean_token_accuracy": 0.17275591790676117,
      "num_tokens": 14061752.0,
      "step": 6305
    },
    {
      "entropy": 5.643550777435303,
      "epoch": 0.7087100578424215,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004959736015210401,
      "loss": 5.475,
      "mean_token_accuracy": 0.1777690604329109,
      "num_tokens": 14072931.0,
      "step": 6310
    },
    {
      "entropy": 5.6225982189178465,
      "epoch": 0.709271634750379,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004959660366046169,
      "loss": 5.5643,
      "mean_token_accuracy": 0.18406163901090622,
      "num_tokens": 14084137.0,
      "step": 6315
    },
    {
      "entropy": 5.635232162475586,
      "epoch": 0.7098332116583366,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004959584646525494,
      "loss": 5.5652,
      "mean_token_accuracy": 0.17078970968723298,
      "num_tokens": 14096251.0,
      "step": 6320
    },
    {
      "entropy": 5.765122604370117,
      "epoch": 0.7103947885662941,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004959508856650788,
      "loss": 5.6863,
      "mean_token_accuracy": 0.16424456089735032,
      "num_tokens": 14107844.0,
      "step": 6325
    },
    {
      "entropy": 5.630089235305786,
      "epoch": 0.7109563654742517,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004959432996424463,
      "loss": 5.4599,
      "mean_token_accuracy": 0.17952507734298706,
      "num_tokens": 14118344.0,
      "step": 6330
    },
    {
      "entropy": 5.553254890441894,
      "epoch": 0.7115179423822092,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004959357065848932,
      "loss": 5.5036,
      "mean_token_accuracy": 0.17871073633432388,
      "num_tokens": 14128767.0,
      "step": 6335
    },
    {
      "entropy": 5.619081354141235,
      "epoch": 0.7120795192901668,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004959281064926618,
      "loss": 5.681,
      "mean_token_accuracy": 0.16622405201196672,
      "num_tokens": 14141593.0,
      "step": 6340
    },
    {
      "entropy": 5.728466653823853,
      "epoch": 0.7126410961981243,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004959204993659936,
      "loss": 5.5785,
      "mean_token_accuracy": 0.1735784649848938,
      "num_tokens": 14151488.0,
      "step": 6345
    },
    {
      "entropy": 5.602538156509399,
      "epoch": 0.7132026731060819,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004959128852051311,
      "loss": 5.5771,
      "mean_token_accuracy": 0.1768488958477974,
      "num_tokens": 14162399.0,
      "step": 6350
    },
    {
      "entropy": 5.593472957611084,
      "epoch": 0.7137642500140394,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004959052640103167,
      "loss": 5.5416,
      "mean_token_accuracy": 0.16909119188785554,
      "num_tokens": 14173528.0,
      "step": 6355
    },
    {
      "entropy": 5.682034206390381,
      "epoch": 0.714325826921997,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000495897635781793,
      "loss": 5.5338,
      "mean_token_accuracy": 0.1783952072262764,
      "num_tokens": 14184119.0,
      "step": 6360
    },
    {
      "entropy": 5.697862195968628,
      "epoch": 0.7148874038299545,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004958900005198029,
      "loss": 5.607,
      "mean_token_accuracy": 0.17433831989765167,
      "num_tokens": 14196225.0,
      "step": 6365
    },
    {
      "entropy": 5.524587726593017,
      "epoch": 0.7154489807379121,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004958823582245894,
      "loss": 5.4343,
      "mean_token_accuracy": 0.18652019649744034,
      "num_tokens": 14207541.0,
      "step": 6370
    },
    {
      "entropy": 5.632974243164062,
      "epoch": 0.7160105576458696,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004958747088963961,
      "loss": 5.5327,
      "mean_token_accuracy": 0.17961990386247634,
      "num_tokens": 14218713.0,
      "step": 6375
    },
    {
      "entropy": 5.494053936004638,
      "epoch": 0.7165721345538272,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004958670525354663,
      "loss": 5.4032,
      "mean_token_accuracy": 0.18424461632966996,
      "num_tokens": 14229626.0,
      "step": 6380
    },
    {
      "entropy": 5.609897804260254,
      "epoch": 0.7171337114617847,
      "grad_norm": 1.09375,
      "learning_rate": 0.000495859389142044,
      "loss": 5.5803,
      "mean_token_accuracy": 0.1705845594406128,
      "num_tokens": 14240320.0,
      "step": 6385
    },
    {
      "entropy": 5.729384040832519,
      "epoch": 0.7176952883697423,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000495851718716373,
      "loss": 5.6394,
      "mean_token_accuracy": 0.17609725147485733,
      "num_tokens": 14252233.0,
      "step": 6390
    },
    {
      "entropy": 5.705629062652588,
      "epoch": 0.7182568652776998,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004958440412586977,
      "loss": 5.6605,
      "mean_token_accuracy": 0.16531089097261428,
      "num_tokens": 14264421.0,
      "step": 6395
    },
    {
      "entropy": 5.728456354141235,
      "epoch": 0.7188184421856574,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004958363567692624,
      "loss": 5.5369,
      "mean_token_accuracy": 0.18062908798456193,
      "num_tokens": 14275320.0,
      "step": 6400
    },
    {
      "entropy": 5.538830661773682,
      "epoch": 0.7193800190936148,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004958286652483118,
      "loss": 5.5013,
      "mean_token_accuracy": 0.17964332848787307,
      "num_tokens": 14286099.0,
      "step": 6405
    },
    {
      "entropy": 5.5996184825897215,
      "epoch": 0.7199415960015724,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004958209666960909,
      "loss": 5.6219,
      "mean_token_accuracy": 0.17469322234392165,
      "num_tokens": 14297795.0,
      "step": 6410
    },
    {
      "entropy": 5.59598126411438,
      "epoch": 0.7205031729095299,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004958132611128448,
      "loss": 5.5165,
      "mean_token_accuracy": 0.17891257256269455,
      "num_tokens": 14309424.0,
      "step": 6415
    },
    {
      "entropy": 5.7096476554870605,
      "epoch": 0.7210647498174875,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004958055484988188,
      "loss": 5.5964,
      "mean_token_accuracy": 0.17317477017641067,
      "num_tokens": 14320763.0,
      "step": 6420
    },
    {
      "entropy": 5.604258584976196,
      "epoch": 0.721626326725445,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004957978288542585,
      "loss": 5.555,
      "mean_token_accuracy": 0.17598283737897874,
      "num_tokens": 14331704.0,
      "step": 6425
    },
    {
      "entropy": 5.654898071289063,
      "epoch": 0.7221879036334026,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004957901021794098,
      "loss": 5.6018,
      "mean_token_accuracy": 0.1679856985807419,
      "num_tokens": 14343694.0,
      "step": 6430
    },
    {
      "entropy": 5.635381412506104,
      "epoch": 0.7227494805413601,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004957823684745184,
      "loss": 5.492,
      "mean_token_accuracy": 0.1778458148241043,
      "num_tokens": 14355408.0,
      "step": 6435
    },
    {
      "entropy": 5.595786666870117,
      "epoch": 0.7233110574493177,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004957746277398309,
      "loss": 5.6563,
      "mean_token_accuracy": 0.16623536497354507,
      "num_tokens": 14366778.0,
      "step": 6440
    },
    {
      "entropy": 5.620596981048584,
      "epoch": 0.7238726343572752,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004957668799755937,
      "loss": 5.5183,
      "mean_token_accuracy": 0.17913369536399842,
      "num_tokens": 14377947.0,
      "step": 6445
    },
    {
      "entropy": 5.6885699272155765,
      "epoch": 0.7244342112652328,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004957591251820532,
      "loss": 5.6014,
      "mean_token_accuracy": 0.17538815289735793,
      "num_tokens": 14389571.0,
      "step": 6450
    },
    {
      "entropy": 5.772772932052613,
      "epoch": 0.7249957881731903,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004957513633594567,
      "loss": 5.6556,
      "mean_token_accuracy": 0.1718176931142807,
      "num_tokens": 14401865.0,
      "step": 6455
    },
    {
      "entropy": 5.521078777313233,
      "epoch": 0.7255573650811479,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004957435945080512,
      "loss": 5.4903,
      "mean_token_accuracy": 0.1813999369740486,
      "num_tokens": 14413676.0,
      "step": 6460
    },
    {
      "entropy": 5.5760499954223635,
      "epoch": 0.7261189419891054,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004957358186280839,
      "loss": 5.5221,
      "mean_token_accuracy": 0.18442534506320954,
      "num_tokens": 14424675.0,
      "step": 6465
    },
    {
      "entropy": 5.569950723648072,
      "epoch": 0.726680518897063,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004957280357198026,
      "loss": 5.4312,
      "mean_token_accuracy": 0.18597756177186966,
      "num_tokens": 14437238.0,
      "step": 6470
    },
    {
      "entropy": 5.519033432006836,
      "epoch": 0.7272420958050205,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004957202457834551,
      "loss": 5.4075,
      "mean_token_accuracy": 0.18195293843746185,
      "num_tokens": 14448003.0,
      "step": 6475
    },
    {
      "entropy": 5.437840843200684,
      "epoch": 0.7278036727129781,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004957124488192893,
      "loss": 5.3511,
      "mean_token_accuracy": 0.18920646607875824,
      "num_tokens": 14458440.0,
      "step": 6480
    },
    {
      "entropy": 5.590056228637695,
      "epoch": 0.7283652496209356,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004957046448275535,
      "loss": 5.5352,
      "mean_token_accuracy": 0.17582074999809266,
      "num_tokens": 14470127.0,
      "step": 6485
    },
    {
      "entropy": 5.604477024078369,
      "epoch": 0.7289268265288932,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004956968338084961,
      "loss": 5.4478,
      "mean_token_accuracy": 0.18338292390108107,
      "num_tokens": 14480993.0,
      "step": 6490
    },
    {
      "entropy": 5.588405609130859,
      "epoch": 0.7294884034368507,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004956890157623661,
      "loss": 5.557,
      "mean_token_accuracy": 0.17237964421510696,
      "num_tokens": 14492789.0,
      "step": 6495
    },
    {
      "entropy": 5.644979190826416,
      "epoch": 0.7300499803448082,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004956811906894122,
      "loss": 5.6378,
      "mean_token_accuracy": 0.17463567852973938,
      "num_tokens": 14504528.0,
      "step": 6500
    },
    {
      "entropy": 5.753073358535767,
      "epoch": 0.7306115572527657,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004956733585898836,
      "loss": 5.5334,
      "mean_token_accuracy": 0.1780219629406929,
      "num_tokens": 14515739.0,
      "step": 6505
    },
    {
      "entropy": 5.698412227630615,
      "epoch": 0.7311731341607233,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004956655194640297,
      "loss": 5.648,
      "mean_token_accuracy": 0.1695479065179825,
      "num_tokens": 14527120.0,
      "step": 6510
    },
    {
      "entropy": 5.605332660675049,
      "epoch": 0.7317347110686808,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004956576733121,
      "loss": 5.5983,
      "mean_token_accuracy": 0.1782694473862648,
      "num_tokens": 14539390.0,
      "step": 6515
    },
    {
      "entropy": 5.639449262619019,
      "epoch": 0.7322962879766384,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004956498201343444,
      "loss": 5.5758,
      "mean_token_accuracy": 0.17402366697788238,
      "num_tokens": 14549905.0,
      "step": 6520
    },
    {
      "entropy": 5.661479854583741,
      "epoch": 0.7328578648845959,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000495641959931013,
      "loss": 5.6073,
      "mean_token_accuracy": 0.17633203715085982,
      "num_tokens": 14561519.0,
      "step": 6525
    },
    {
      "entropy": 5.56331353187561,
      "epoch": 0.7334194417925535,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004956340927023561,
      "loss": 5.3983,
      "mean_token_accuracy": 0.1783517822623253,
      "num_tokens": 14572704.0,
      "step": 6530
    },
    {
      "entropy": 5.63326587677002,
      "epoch": 0.733981018700511,
      "grad_norm": 1.078125,
      "learning_rate": 0.000495626218448624,
      "loss": 5.5781,
      "mean_token_accuracy": 0.1766029641032219,
      "num_tokens": 14584055.0,
      "step": 6535
    },
    {
      "entropy": 5.484734582901001,
      "epoch": 0.7345425956084686,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004956183371700676,
      "loss": 5.3838,
      "mean_token_accuracy": 0.18476226031780243,
      "num_tokens": 14595271.0,
      "step": 6540
    },
    {
      "entropy": 5.622118520736694,
      "epoch": 0.7351041725164261,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004956104488669377,
      "loss": 5.5382,
      "mean_token_accuracy": 0.17977642714977266,
      "num_tokens": 14605748.0,
      "step": 6545
    },
    {
      "entropy": 5.598040819168091,
      "epoch": 0.7356657494243837,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004956025535394856,
      "loss": 5.4699,
      "mean_token_accuracy": 0.18558113425970077,
      "num_tokens": 14616295.0,
      "step": 6550
    },
    {
      "entropy": 5.562770795822144,
      "epoch": 0.7362273263323412,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004955946511879626,
      "loss": 5.5733,
      "mean_token_accuracy": 0.172361159324646,
      "num_tokens": 14626690.0,
      "step": 6555
    },
    {
      "entropy": 5.627026748657227,
      "epoch": 0.7367889032402988,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004955867418126205,
      "loss": 5.5523,
      "mean_token_accuracy": 0.17980725616216658,
      "num_tokens": 14637685.0,
      "step": 6560
    },
    {
      "entropy": 5.65916953086853,
      "epoch": 0.7373504801482563,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004955788254137108,
      "loss": 5.4638,
      "mean_token_accuracy": 0.1704305499792099,
      "num_tokens": 14648098.0,
      "step": 6565
    },
    {
      "entropy": 5.682143974304199,
      "epoch": 0.7379120570562139,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004955709019914857,
      "loss": 5.6049,
      "mean_token_accuracy": 0.1708393543958664,
      "num_tokens": 14661134.0,
      "step": 6570
    },
    {
      "entropy": 5.6985547065734865,
      "epoch": 0.7384736339641714,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004955629715461977,
      "loss": 5.5701,
      "mean_token_accuracy": 0.17628692984580993,
      "num_tokens": 14672699.0,
      "step": 6575
    },
    {
      "entropy": 5.600019931793213,
      "epoch": 0.739035210872129,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004955550340780991,
      "loss": 5.5509,
      "mean_token_accuracy": 0.17778549641370772,
      "num_tokens": 14682994.0,
      "step": 6580
    },
    {
      "entropy": 5.531461000442505,
      "epoch": 0.7395967877800865,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004955470895874427,
      "loss": 5.5462,
      "mean_token_accuracy": 0.17700110077857972,
      "num_tokens": 14694828.0,
      "step": 6585
    },
    {
      "entropy": 5.6243321895599365,
      "epoch": 0.7401583646880441,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004955391380744813,
      "loss": 5.5241,
      "mean_token_accuracy": 0.18042442947626114,
      "num_tokens": 14705975.0,
      "step": 6590
    },
    {
      "entropy": 5.681596088409424,
      "epoch": 0.7407199415960015,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004955311795394684,
      "loss": 5.5246,
      "mean_token_accuracy": 0.17880332916975022,
      "num_tokens": 14717717.0,
      "step": 6595
    },
    {
      "entropy": 5.607036161422729,
      "epoch": 0.7412815185039591,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000495523213982657,
      "loss": 5.484,
      "mean_token_accuracy": 0.1803207665681839,
      "num_tokens": 14729254.0,
      "step": 6600
    },
    {
      "entropy": 5.637073707580567,
      "epoch": 0.7418430954119166,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004955152414043011,
      "loss": 5.6554,
      "mean_token_accuracy": 0.17397006899118422,
      "num_tokens": 14739733.0,
      "step": 6605
    },
    {
      "entropy": 5.668559455871582,
      "epoch": 0.7424046723198742,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004955072618046543,
      "loss": 5.5754,
      "mean_token_accuracy": 0.1776353806257248,
      "num_tokens": 14750855.0,
      "step": 6610
    },
    {
      "entropy": 5.623273038864136,
      "epoch": 0.7429662492278317,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004954992751839708,
      "loss": 5.5917,
      "mean_token_accuracy": 0.1725652515888214,
      "num_tokens": 14762762.0,
      "step": 6615
    },
    {
      "entropy": 5.632060384750366,
      "epoch": 0.7435278261357893,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004954912815425048,
      "loss": 5.4507,
      "mean_token_accuracy": 0.17733130604028702,
      "num_tokens": 14773690.0,
      "step": 6620
    },
    {
      "entropy": 5.599242305755615,
      "epoch": 0.7440894030437468,
      "grad_norm": 1.1875,
      "learning_rate": 0.000495483280880511,
      "loss": 5.5628,
      "mean_token_accuracy": 0.1783632293343544,
      "num_tokens": 14785019.0,
      "step": 6625
    },
    {
      "entropy": 5.529038906097412,
      "epoch": 0.7446509799517044,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000495475273198244,
      "loss": 5.527,
      "mean_token_accuracy": 0.17582466453313828,
      "num_tokens": 14795638.0,
      "step": 6630
    },
    {
      "entropy": 5.626880168914795,
      "epoch": 0.7452125568596619,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004954672584959588,
      "loss": 5.4774,
      "mean_token_accuracy": 0.18109178692102432,
      "num_tokens": 14806546.0,
      "step": 6635
    },
    {
      "entropy": 5.6049220085144045,
      "epoch": 0.7457741337676195,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004954592367739104,
      "loss": 5.4968,
      "mean_token_accuracy": 0.18364625722169875,
      "num_tokens": 14816918.0,
      "step": 6640
    },
    {
      "entropy": 5.579300022125244,
      "epoch": 0.746335710675577,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004954512080323546,
      "loss": 5.4855,
      "mean_token_accuracy": 0.18444073349237441,
      "num_tokens": 14827420.0,
      "step": 6645
    },
    {
      "entropy": 5.677469062805176,
      "epoch": 0.7468972875835346,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004954431722715468,
      "loss": 5.7127,
      "mean_token_accuracy": 0.1737464800477028,
      "num_tokens": 14839209.0,
      "step": 6650
    },
    {
      "entropy": 5.652341699600219,
      "epoch": 0.7474588644914921,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004954351294917428,
      "loss": 5.6023,
      "mean_token_accuracy": 0.17281786054372789,
      "num_tokens": 14849916.0,
      "step": 6655
    },
    {
      "entropy": 5.570694541931152,
      "epoch": 0.7480204413994497,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004954270796931989,
      "loss": 5.4618,
      "mean_token_accuracy": 0.182657852768898,
      "num_tokens": 14860946.0,
      "step": 6660
    },
    {
      "entropy": 5.72287883758545,
      "epoch": 0.7485820183074072,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004954190228761712,
      "loss": 5.5855,
      "mean_token_accuracy": 0.17253236323595048,
      "num_tokens": 14872750.0,
      "step": 6665
    },
    {
      "entropy": 5.6262044429779055,
      "epoch": 0.7491435952153648,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004954109590409164,
      "loss": 5.539,
      "mean_token_accuracy": 0.17555639892816544,
      "num_tokens": 14883714.0,
      "step": 6670
    },
    {
      "entropy": 5.62942214012146,
      "epoch": 0.7497051721233223,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004954028881876911,
      "loss": 5.5595,
      "mean_token_accuracy": 0.17766272276639938,
      "num_tokens": 14895147.0,
      "step": 6675
    },
    {
      "entropy": 5.5647547245025635,
      "epoch": 0.7502667490312799,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004953948103167524,
      "loss": 5.4565,
      "mean_token_accuracy": 0.18300872892141343,
      "num_tokens": 14905914.0,
      "step": 6680
    },
    {
      "entropy": 5.510549879074096,
      "epoch": 0.7508283259392374,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004953867254283575,
      "loss": 5.47,
      "mean_token_accuracy": 0.18229593336582184,
      "num_tokens": 14916988.0,
      "step": 6685
    },
    {
      "entropy": 5.701801776885986,
      "epoch": 0.7513899028471949,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004953786335227636,
      "loss": 5.5688,
      "mean_token_accuracy": 0.17668356597423554,
      "num_tokens": 14928675.0,
      "step": 6690
    },
    {
      "entropy": 5.712063598632812,
      "epoch": 0.7519514797551524,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004953705346002287,
      "loss": 5.5868,
      "mean_token_accuracy": 0.18081481009721756,
      "num_tokens": 14939422.0,
      "step": 6695
    },
    {
      "entropy": 5.584119415283203,
      "epoch": 0.75251305666311,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004953624286610106,
      "loss": 5.5314,
      "mean_token_accuracy": 0.18003119081258773,
      "num_tokens": 14950156.0,
      "step": 6700
    },
    {
      "entropy": 5.632153129577636,
      "epoch": 0.7530746335710675,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004953543157053672,
      "loss": 5.5357,
      "mean_token_accuracy": 0.18005113303661346,
      "num_tokens": 14960535.0,
      "step": 6705
    },
    {
      "entropy": 5.616408395767212,
      "epoch": 0.7536362104790251,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004953461957335569,
      "loss": 5.5452,
      "mean_token_accuracy": 0.17996135503053665,
      "num_tokens": 14970975.0,
      "step": 6710
    },
    {
      "entropy": 5.51405782699585,
      "epoch": 0.7541977873869826,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004953380687458382,
      "loss": 5.4198,
      "mean_token_accuracy": 0.18629451245069503,
      "num_tokens": 14981204.0,
      "step": 6715
    },
    {
      "entropy": 5.639434194564819,
      "epoch": 0.7547593642949402,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004953299347424701,
      "loss": 5.5734,
      "mean_token_accuracy": 0.1737573802471161,
      "num_tokens": 14991631.0,
      "step": 6720
    },
    {
      "entropy": 5.578642225265503,
      "epoch": 0.7553209412028977,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004953217937237114,
      "loss": 5.4879,
      "mean_token_accuracy": 0.18210435658693314,
      "num_tokens": 15003873.0,
      "step": 6725
    },
    {
      "entropy": 5.598605060577393,
      "epoch": 0.7558825181108553,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004953136456898212,
      "loss": 5.5749,
      "mean_token_accuracy": 0.17554473280906677,
      "num_tokens": 15014890.0,
      "step": 6730
    },
    {
      "entropy": 5.601700782775879,
      "epoch": 0.7564440950188128,
      "grad_norm": 1.125,
      "learning_rate": 0.0004953054906410593,
      "loss": 5.4654,
      "mean_token_accuracy": 0.183247472345829,
      "num_tokens": 15025037.0,
      "step": 6735
    },
    {
      "entropy": 5.638119792938232,
      "epoch": 0.7570056719267704,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004952973285776848,
      "loss": 5.5595,
      "mean_token_accuracy": 0.17968008071184158,
      "num_tokens": 15035722.0,
      "step": 6740
    },
    {
      "entropy": 5.621606111526489,
      "epoch": 0.7575672488347279,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004952891594999582,
      "loss": 5.5468,
      "mean_token_accuracy": 0.17412951290607454,
      "num_tokens": 15047138.0,
      "step": 6745
    },
    {
      "entropy": 5.561537742614746,
      "epoch": 0.7581288257426855,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004952809834081392,
      "loss": 5.5078,
      "mean_token_accuracy": 0.18185122311115265,
      "num_tokens": 15058559.0,
      "step": 6750
    },
    {
      "entropy": 5.609898042678833,
      "epoch": 0.758690402650643,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004952728003024883,
      "loss": 5.5402,
      "mean_token_accuracy": 0.1776107594370842,
      "num_tokens": 15070481.0,
      "step": 6755
    },
    {
      "entropy": 5.661940670013427,
      "epoch": 0.7592519795586006,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004952646101832659,
      "loss": 5.4662,
      "mean_token_accuracy": 0.1806693732738495,
      "num_tokens": 15079844.0,
      "step": 6760
    },
    {
      "entropy": 5.4710187911987305,
      "epoch": 0.7598135564665581,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004952564130507331,
      "loss": 5.4847,
      "mean_token_accuracy": 0.18411026149988174,
      "num_tokens": 15091197.0,
      "step": 6765
    },
    {
      "entropy": 5.572347736358642,
      "epoch": 0.7603751333745157,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004952482089051506,
      "loss": 5.5489,
      "mean_token_accuracy": 0.17732695043087005,
      "num_tokens": 15102130.0,
      "step": 6770
    },
    {
      "entropy": 5.751537132263183,
      "epoch": 0.7609367102824732,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004952399977467796,
      "loss": 5.6385,
      "mean_token_accuracy": 0.17081636041402817,
      "num_tokens": 15113886.0,
      "step": 6775
    },
    {
      "entropy": 5.564804983139038,
      "epoch": 0.7614982871904308,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004952317795758817,
      "loss": 5.4532,
      "mean_token_accuracy": 0.17811087965965272,
      "num_tokens": 15125284.0,
      "step": 6780
    },
    {
      "entropy": 5.617397785186768,
      "epoch": 0.7620598640983883,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004952235543927185,
      "loss": 5.5343,
      "mean_token_accuracy": 0.17885010689496994,
      "num_tokens": 15136486.0,
      "step": 6785
    },
    {
      "entropy": 5.525661993026733,
      "epoch": 0.7626214410063458,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000495215322197552,
      "loss": 5.4205,
      "mean_token_accuracy": 0.18539561927318574,
      "num_tokens": 15145937.0,
      "step": 6790
    },
    {
      "entropy": 5.495564413070679,
      "epoch": 0.7631830179143033,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004952070829906442,
      "loss": 5.4968,
      "mean_token_accuracy": 0.17593696862459182,
      "num_tokens": 15156923.0,
      "step": 6795
    },
    {
      "entropy": 5.737236595153808,
      "epoch": 0.7637445948222609,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004951988367722574,
      "loss": 5.6229,
      "mean_token_accuracy": 0.1781747668981552,
      "num_tokens": 15169793.0,
      "step": 6800
    },
    {
      "entropy": 5.547433233261108,
      "epoch": 0.7643061717302184,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004951905835426544,
      "loss": 5.3874,
      "mean_token_accuracy": 0.18833620548248292,
      "num_tokens": 15180679.0,
      "step": 6805
    },
    {
      "entropy": 5.567129516601563,
      "epoch": 0.764867748638176,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004951823233020976,
      "loss": 5.5818,
      "mean_token_accuracy": 0.17742714583873748,
      "num_tokens": 15192298.0,
      "step": 6810
    },
    {
      "entropy": 5.671421337127685,
      "epoch": 0.7654293255461335,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004951740560508504,
      "loss": 5.6584,
      "mean_token_accuracy": 0.17175425589084625,
      "num_tokens": 15204759.0,
      "step": 6815
    },
    {
      "entropy": 5.702127742767334,
      "epoch": 0.7659909024540911,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004951657817891758,
      "loss": 5.5318,
      "mean_token_accuracy": 0.18169393688440322,
      "num_tokens": 15217166.0,
      "step": 6820
    },
    {
      "entropy": 5.550262022018432,
      "epoch": 0.7665524793620486,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004951575005173373,
      "loss": 5.5179,
      "mean_token_accuracy": 0.18110382705926895,
      "num_tokens": 15227022.0,
      "step": 6825
    },
    {
      "entropy": 5.631211233139038,
      "epoch": 0.7671140562700062,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004951492122355986,
      "loss": 5.5561,
      "mean_token_accuracy": 0.17720902115106582,
      "num_tokens": 15239088.0,
      "step": 6830
    },
    {
      "entropy": 5.676913547515869,
      "epoch": 0.7676756331779637,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004951409169442236,
      "loss": 5.5942,
      "mean_token_accuracy": 0.18003331422805785,
      "num_tokens": 15250263.0,
      "step": 6835
    },
    {
      "entropy": 5.541953372955322,
      "epoch": 0.7682372100859213,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004951326146434765,
      "loss": 5.4774,
      "mean_token_accuracy": 0.18118782788515092,
      "num_tokens": 15263014.0,
      "step": 6840
    },
    {
      "entropy": 5.647920036315918,
      "epoch": 0.7687987869938788,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004951243053336214,
      "loss": 5.4753,
      "mean_token_accuracy": 0.18037050515413283,
      "num_tokens": 15274261.0,
      "step": 6845
    },
    {
      "entropy": 5.55603928565979,
      "epoch": 0.7693603639018364,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004951159890149232,
      "loss": 5.4189,
      "mean_token_accuracy": 0.18435574620962142,
      "num_tokens": 15284929.0,
      "step": 6850
    },
    {
      "entropy": 5.4724915504455565,
      "epoch": 0.7699219408097939,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004951076656876466,
      "loss": 5.5106,
      "mean_token_accuracy": 0.1790434405207634,
      "num_tokens": 15296622.0,
      "step": 6855
    },
    {
      "entropy": 5.589769315719605,
      "epoch": 0.7704835177177515,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004950993353520564,
      "loss": 5.567,
      "mean_token_accuracy": 0.17600751519203187,
      "num_tokens": 15308373.0,
      "step": 6860
    },
    {
      "entropy": 5.599905920028687,
      "epoch": 0.771045094625709,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000495090998008418,
      "loss": 5.443,
      "mean_token_accuracy": 0.1868746668100357,
      "num_tokens": 15319705.0,
      "step": 6865
    },
    {
      "entropy": 5.6429096221923825,
      "epoch": 0.7716066715336666,
      "grad_norm": 1.203125,
      "learning_rate": 0.000495082653656997,
      "loss": 5.6041,
      "mean_token_accuracy": 0.17266785949468613,
      "num_tokens": 15330883.0,
      "step": 6870
    },
    {
      "entropy": 5.605991697311401,
      "epoch": 0.7721682484416241,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004950743022980587,
      "loss": 5.5143,
      "mean_token_accuracy": 0.17568910270929336,
      "num_tokens": 15342294.0,
      "step": 6875
    },
    {
      "entropy": 5.627771997451783,
      "epoch": 0.7727298253495817,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004950659439318696,
      "loss": 5.5091,
      "mean_token_accuracy": 0.17947840988636016,
      "num_tokens": 15352615.0,
      "step": 6880
    },
    {
      "entropy": 5.564583730697632,
      "epoch": 0.7732914022575391,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004950575785586952,
      "loss": 5.564,
      "mean_token_accuracy": 0.17500838935375213,
      "num_tokens": 15364516.0,
      "step": 6885
    },
    {
      "entropy": 5.607831716537476,
      "epoch": 0.7738529791654967,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004950492061788023,
      "loss": 5.4594,
      "mean_token_accuracy": 0.17648799419403077,
      "num_tokens": 15374877.0,
      "step": 6890
    },
    {
      "entropy": 5.542214727401733,
      "epoch": 0.7744145560734542,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004950408267924573,
      "loss": 5.3786,
      "mean_token_accuracy": 0.19236088395118714,
      "num_tokens": 15385250.0,
      "step": 6895
    },
    {
      "entropy": 5.529676914215088,
      "epoch": 0.7749761329814118,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004950324403999271,
      "loss": 5.5327,
      "mean_token_accuracy": 0.18624390959739684,
      "num_tokens": 15397074.0,
      "step": 6900
    },
    {
      "entropy": 5.651473903656006,
      "epoch": 0.7755377098893693,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004950240470014785,
      "loss": 5.588,
      "mean_token_accuracy": 0.1724044218659401,
      "num_tokens": 15408575.0,
      "step": 6905
    },
    {
      "entropy": 5.647315263748169,
      "epoch": 0.7760992867973269,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004950156465973791,
      "loss": 5.5141,
      "mean_token_accuracy": 0.18157450556755067,
      "num_tokens": 15419071.0,
      "step": 6910
    },
    {
      "entropy": 5.509777021408081,
      "epoch": 0.7766608637052844,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000495007239187896,
      "loss": 5.384,
      "mean_token_accuracy": 0.18187790364027023,
      "num_tokens": 15430137.0,
      "step": 6915
    },
    {
      "entropy": 5.524733924865723,
      "epoch": 0.777222440613242,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004949988247732972,
      "loss": 5.465,
      "mean_token_accuracy": 0.1827964335680008,
      "num_tokens": 15440661.0,
      "step": 6920
    },
    {
      "entropy": 5.533017110824585,
      "epoch": 0.7777840175211995,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004949904033538504,
      "loss": 5.4651,
      "mean_token_accuracy": 0.18287097066640853,
      "num_tokens": 15451909.0,
      "step": 6925
    },
    {
      "entropy": 5.639970874786377,
      "epoch": 0.7783455944291571,
      "grad_norm": 1.078125,
      "learning_rate": 0.000494981974929824,
      "loss": 5.5471,
      "mean_token_accuracy": 0.17527093440294267,
      "num_tokens": 15462244.0,
      "step": 6930
    },
    {
      "entropy": 5.60972228050232,
      "epoch": 0.7789071713371146,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000494973539501486,
      "loss": 5.5867,
      "mean_token_accuracy": 0.1783525377511978,
      "num_tokens": 15473284.0,
      "step": 6935
    },
    {
      "entropy": 5.580111360549926,
      "epoch": 0.7794687482450722,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004949650970691052,
      "loss": 5.4725,
      "mean_token_accuracy": 0.1878479540348053,
      "num_tokens": 15484964.0,
      "step": 6940
    },
    {
      "entropy": 5.594936656951904,
      "epoch": 0.7800303251530297,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004949566476329505,
      "loss": 5.5062,
      "mean_token_accuracy": 0.1817149266600609,
      "num_tokens": 15495922.0,
      "step": 6945
    },
    {
      "entropy": 5.674933338165284,
      "epoch": 0.7805919020609873,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004949481911932908,
      "loss": 5.5752,
      "mean_token_accuracy": 0.1781456872820854,
      "num_tokens": 15507173.0,
      "step": 6950
    },
    {
      "entropy": 5.590296363830566,
      "epoch": 0.7811534789689448,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004949397277503954,
      "loss": 5.5302,
      "mean_token_accuracy": 0.18023878186941147,
      "num_tokens": 15517720.0,
      "step": 6955
    },
    {
      "entropy": 5.69305305480957,
      "epoch": 0.7817150558769024,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004949312573045338,
      "loss": 5.5468,
      "mean_token_accuracy": 0.17897407561540604,
      "num_tokens": 15529482.0,
      "step": 6960
    },
    {
      "entropy": 5.694714498519898,
      "epoch": 0.7822766327848599,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004949227798559757,
      "loss": 5.5829,
      "mean_token_accuracy": 0.1729534775018692,
      "num_tokens": 15540522.0,
      "step": 6965
    },
    {
      "entropy": 5.541285848617553,
      "epoch": 0.7828382096928175,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004949142954049909,
      "loss": 5.5517,
      "mean_token_accuracy": 0.17297329306602477,
      "num_tokens": 15552443.0,
      "step": 6970
    },
    {
      "entropy": 5.68100209236145,
      "epoch": 0.783399786600775,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004949058039518496,
      "loss": 5.5764,
      "mean_token_accuracy": 0.18182898163795472,
      "num_tokens": 15563418.0,
      "step": 6975
    },
    {
      "entropy": 5.671764802932739,
      "epoch": 0.7839613635087325,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004948973054968222,
      "loss": 5.5919,
      "mean_token_accuracy": 0.17396205365657808,
      "num_tokens": 15574301.0,
      "step": 6980
    },
    {
      "entropy": 5.6782971858978275,
      "epoch": 0.78452294041669,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004948888000401794,
      "loss": 5.5227,
      "mean_token_accuracy": 0.1690296858549118,
      "num_tokens": 15585125.0,
      "step": 6985
    },
    {
      "entropy": 5.66305456161499,
      "epoch": 0.7850845173246476,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004948802875821918,
      "loss": 5.6155,
      "mean_token_accuracy": 0.1790311560034752,
      "num_tokens": 15595822.0,
      "step": 6990
    },
    {
      "entropy": 5.543122959136963,
      "epoch": 0.7856460942326051,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004948717681231308,
      "loss": 5.4496,
      "mean_token_accuracy": 0.18060397654771804,
      "num_tokens": 15606835.0,
      "step": 6995
    },
    {
      "entropy": 5.634728717803955,
      "epoch": 0.7862076711405627,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004948632416632672,
      "loss": 5.5385,
      "mean_token_accuracy": 0.17847790718078613,
      "num_tokens": 15617287.0,
      "step": 7000
    },
    {
      "entropy": 5.682579708099365,
      "epoch": 0.7867692480485202,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004948547082028728,
      "loss": 5.5701,
      "mean_token_accuracy": 0.17271330952644348,
      "num_tokens": 15628440.0,
      "step": 7005
    },
    {
      "entropy": 5.624586391448974,
      "epoch": 0.7873308249564778,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004948461677422193,
      "loss": 5.6636,
      "mean_token_accuracy": 0.1696137636899948,
      "num_tokens": 15640677.0,
      "step": 7010
    },
    {
      "entropy": 5.656479740142823,
      "epoch": 0.7878924018644353,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004948376202815785,
      "loss": 5.5121,
      "mean_token_accuracy": 0.18355311155319215,
      "num_tokens": 15651941.0,
      "step": 7015
    },
    {
      "entropy": 5.569307041168213,
      "epoch": 0.7884539787723929,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004948290658212225,
      "loss": 5.4934,
      "mean_token_accuracy": 0.18243621587753295,
      "num_tokens": 15663513.0,
      "step": 7020
    },
    {
      "entropy": 5.544464063644409,
      "epoch": 0.7890155556803504,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004948205043614239,
      "loss": 5.4909,
      "mean_token_accuracy": 0.17685482650995255,
      "num_tokens": 15675548.0,
      "step": 7025
    },
    {
      "entropy": 5.688881158828735,
      "epoch": 0.789577132588308,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004948119359024551,
      "loss": 5.533,
      "mean_token_accuracy": 0.18039870858192444,
      "num_tokens": 15685365.0,
      "step": 7030
    },
    {
      "entropy": 5.542220163345337,
      "epoch": 0.7901387094962655,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004948033604445891,
      "loss": 5.5407,
      "mean_token_accuracy": 0.1748532086610794,
      "num_tokens": 15696112.0,
      "step": 7035
    },
    {
      "entropy": 5.55469856262207,
      "epoch": 0.7907002864042231,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004947947779880988,
      "loss": 5.4823,
      "mean_token_accuracy": 0.18138503134250641,
      "num_tokens": 15707414.0,
      "step": 7040
    },
    {
      "entropy": 5.654241371154785,
      "epoch": 0.7912618633121806,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004947861885332576,
      "loss": 5.5664,
      "mean_token_accuracy": 0.17848818451166154,
      "num_tokens": 15719077.0,
      "step": 7045
    },
    {
      "entropy": 5.618268728256226,
      "epoch": 0.7918234402201382,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004947775920803388,
      "loss": 5.5662,
      "mean_token_accuracy": 0.17513773441314698,
      "num_tokens": 15730497.0,
      "step": 7050
    },
    {
      "entropy": 5.595116710662841,
      "epoch": 0.7923850171280957,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004947689886296162,
      "loss": 5.4589,
      "mean_token_accuracy": 0.17772922664880753,
      "num_tokens": 15741724.0,
      "step": 7055
    },
    {
      "entropy": 5.623131275177002,
      "epoch": 0.7929465940360533,
      "grad_norm": 1.1875,
      "learning_rate": 0.000494760378181364,
      "loss": 5.5574,
      "mean_token_accuracy": 0.17355152368545532,
      "num_tokens": 15751868.0,
      "step": 7060
    },
    {
      "entropy": 5.5758336067199705,
      "epoch": 0.7935081709440108,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004947517607358559,
      "loss": 5.5692,
      "mean_token_accuracy": 0.1746085524559021,
      "num_tokens": 15762783.0,
      "step": 7065
    },
    {
      "entropy": 5.649895334243775,
      "epoch": 0.7940697478519684,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004947431362933666,
      "loss": 5.5423,
      "mean_token_accuracy": 0.17847929894924164,
      "num_tokens": 15774155.0,
      "step": 7070
    },
    {
      "entropy": 5.75625376701355,
      "epoch": 0.7946313247599258,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004947345048541706,
      "loss": 5.5787,
      "mean_token_accuracy": 0.17354002594947815,
      "num_tokens": 15785034.0,
      "step": 7075
    },
    {
      "entropy": 5.489012670516968,
      "epoch": 0.7951929016678834,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004947258664185427,
      "loss": 5.4894,
      "mean_token_accuracy": 0.18017226457595825,
      "num_tokens": 15795959.0,
      "step": 7080
    },
    {
      "entropy": 5.543977546691894,
      "epoch": 0.7957544785758409,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004947172209867581,
      "loss": 5.3857,
      "mean_token_accuracy": 0.1887056425213814,
      "num_tokens": 15805561.0,
      "step": 7085
    },
    {
      "entropy": 5.635234260559082,
      "epoch": 0.7963160554837985,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004947085685590919,
      "loss": 5.4964,
      "mean_token_accuracy": 0.1761572927236557,
      "num_tokens": 15816883.0,
      "step": 7090
    },
    {
      "entropy": 5.4980443000793455,
      "epoch": 0.796877632391756,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004946999091358196,
      "loss": 5.4681,
      "mean_token_accuracy": 0.17913821786642076,
      "num_tokens": 15827753.0,
      "step": 7095
    },
    {
      "entropy": 5.591536998748779,
      "epoch": 0.7974392092997136,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000494691242717217,
      "loss": 5.5482,
      "mean_token_accuracy": 0.17684003859758377,
      "num_tokens": 15840346.0,
      "step": 7100
    },
    {
      "entropy": 5.6175682067871096,
      "epoch": 0.7980007862076711,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004946825693035601,
      "loss": 5.5118,
      "mean_token_accuracy": 0.17747934460639953,
      "num_tokens": 15851036.0,
      "step": 7105
    },
    {
      "entropy": 5.573828268051147,
      "epoch": 0.7985623631156287,
      "grad_norm": 1.125,
      "learning_rate": 0.000494673888895125,
      "loss": 5.4767,
      "mean_token_accuracy": 0.17412922978401185,
      "num_tokens": 15863154.0,
      "step": 7110
    },
    {
      "entropy": 5.6976847648620605,
      "epoch": 0.7991239400235862,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004946652014921881,
      "loss": 5.5819,
      "mean_token_accuracy": 0.17622600942850114,
      "num_tokens": 15874543.0,
      "step": 7115
    },
    {
      "entropy": 5.609316015243531,
      "epoch": 0.7996855169315438,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004946565070950259,
      "loss": 5.5146,
      "mean_token_accuracy": 0.183436518907547,
      "num_tokens": 15886221.0,
      "step": 7120
    },
    {
      "entropy": 5.594043350219726,
      "epoch": 0.8002470938395013,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004946478057039153,
      "loss": 5.5211,
      "mean_token_accuracy": 0.17511670291423798,
      "num_tokens": 15897602.0,
      "step": 7125
    },
    {
      "entropy": 5.625231838226318,
      "epoch": 0.8008086707474589,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004946390973191335,
      "loss": 5.5227,
      "mean_token_accuracy": 0.17719862908124923,
      "num_tokens": 15908030.0,
      "step": 7130
    },
    {
      "entropy": 5.691114854812622,
      "epoch": 0.8013702476554164,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004946303819409575,
      "loss": 5.576,
      "mean_token_accuracy": 0.17731900811195372,
      "num_tokens": 15918341.0,
      "step": 7135
    },
    {
      "entropy": 5.664419794082642,
      "epoch": 0.801931824563374,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004946216595696652,
      "loss": 5.5549,
      "mean_token_accuracy": 0.181121726334095,
      "num_tokens": 15929144.0,
      "step": 7140
    },
    {
      "entropy": 5.643019914627075,
      "epoch": 0.8024934014713315,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004946129302055338,
      "loss": 5.4811,
      "mean_token_accuracy": 0.17576255947351455,
      "num_tokens": 15940013.0,
      "step": 7145
    },
    {
      "entropy": 5.581735897064209,
      "epoch": 0.8030549783792891,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004946041938488417,
      "loss": 5.5277,
      "mean_token_accuracy": 0.1781127780675888,
      "num_tokens": 15950680.0,
      "step": 7150
    },
    {
      "entropy": 5.582114028930664,
      "epoch": 0.8036165552872466,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004945954504998668,
      "loss": 5.5268,
      "mean_token_accuracy": 0.18031199276447296,
      "num_tokens": 15961462.0,
      "step": 7155
    },
    {
      "entropy": 5.602687406539917,
      "epoch": 0.8041781321952042,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004945867001588877,
      "loss": 5.4325,
      "mean_token_accuracy": 0.18474943786859513,
      "num_tokens": 15972788.0,
      "step": 7160
    },
    {
      "entropy": 5.557822370529175,
      "epoch": 0.8047397091031617,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004945779428261827,
      "loss": 5.499,
      "mean_token_accuracy": 0.18163815885782242,
      "num_tokens": 15983244.0,
      "step": 7165
    },
    {
      "entropy": 5.503819370269776,
      "epoch": 0.8053012860111193,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000494569178502031,
      "loss": 5.5909,
      "mean_token_accuracy": 0.175319604575634,
      "num_tokens": 15994325.0,
      "step": 7170
    },
    {
      "entropy": 5.739853048324585,
      "epoch": 0.8058628629190767,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004945604071867114,
      "loss": 5.5865,
      "mean_token_accuracy": 0.17490192204713823,
      "num_tokens": 16005237.0,
      "step": 7175
    },
    {
      "entropy": 5.613080358505249,
      "epoch": 0.8064244398270343,
      "grad_norm": 1.25,
      "learning_rate": 0.0004945516288805035,
      "loss": 5.4179,
      "mean_token_accuracy": 0.18339292854070663,
      "num_tokens": 16015943.0,
      "step": 7180
    },
    {
      "entropy": 5.535682630538941,
      "epoch": 0.8069860167349918,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004945428435836864,
      "loss": 5.4284,
      "mean_token_accuracy": 0.18102275878190993,
      "num_tokens": 16026727.0,
      "step": 7185
    },
    {
      "entropy": 5.566647338867187,
      "epoch": 0.8075475936429494,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004945340512965401,
      "loss": 5.4629,
      "mean_token_accuracy": 0.17425420731306077,
      "num_tokens": 16039021.0,
      "step": 7190
    },
    {
      "entropy": 5.561880207061767,
      "epoch": 0.8081091705509069,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004945252520193443,
      "loss": 5.4167,
      "mean_token_accuracy": 0.1859311878681183,
      "num_tokens": 16050312.0,
      "step": 7195
    },
    {
      "entropy": 5.574615049362182,
      "epoch": 0.8086707474588645,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004945164457523795,
      "loss": 5.3728,
      "mean_token_accuracy": 0.18789263665676117,
      "num_tokens": 16060009.0,
      "step": 7200
    },
    {
      "entropy": 5.4924399852752686,
      "epoch": 0.809232324366822,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004945076324959258,
      "loss": 5.5195,
      "mean_token_accuracy": 0.17271893173456193,
      "num_tokens": 16072025.0,
      "step": 7205
    },
    {
      "entropy": 5.686260652542114,
      "epoch": 0.8097939012747796,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004944988122502641,
      "loss": 5.5172,
      "mean_token_accuracy": 0.17981114685535432,
      "num_tokens": 16083229.0,
      "step": 7210
    },
    {
      "entropy": 5.4941551208496096,
      "epoch": 0.8103554781827371,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000494489985015675,
      "loss": 5.3871,
      "mean_token_accuracy": 0.1825522854924202,
      "num_tokens": 16094936.0,
      "step": 7215
    },
    {
      "entropy": 5.526007795333863,
      "epoch": 0.8109170550906947,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004944811507924396,
      "loss": 5.5244,
      "mean_token_accuracy": 0.1825345501303673,
      "num_tokens": 16106182.0,
      "step": 7220
    },
    {
      "entropy": 5.537650918960571,
      "epoch": 0.8114786319986522,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004944723095808393,
      "loss": 5.424,
      "mean_token_accuracy": 0.18856724202632905,
      "num_tokens": 16117006.0,
      "step": 7225
    },
    {
      "entropy": 5.521580028533935,
      "epoch": 0.8120402089066098,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004944634613811555,
      "loss": 5.4534,
      "mean_token_accuracy": 0.1759404256939888,
      "num_tokens": 16128892.0,
      "step": 7230
    },
    {
      "entropy": 5.6342175006866455,
      "epoch": 0.8126017858145673,
      "grad_norm": 1.09375,
      "learning_rate": 0.00049445460619367,
      "loss": 5.5773,
      "mean_token_accuracy": 0.18207518011331558,
      "num_tokens": 16139953.0,
      "step": 7235
    },
    {
      "entropy": 5.708789825439453,
      "epoch": 0.8131633627225249,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004944457440186648,
      "loss": 5.5375,
      "mean_token_accuracy": 0.17561133503913878,
      "num_tokens": 16150103.0,
      "step": 7240
    },
    {
      "entropy": 5.5955424308776855,
      "epoch": 0.8137249396304824,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004944368748564219,
      "loss": 5.6237,
      "mean_token_accuracy": 0.169679594039917,
      "num_tokens": 16161666.0,
      "step": 7245
    },
    {
      "entropy": 5.619999742507934,
      "epoch": 0.81428651653844,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004944279987072238,
      "loss": 5.5675,
      "mean_token_accuracy": 0.1718843996524811,
      "num_tokens": 16172524.0,
      "step": 7250
    },
    {
      "entropy": 5.672878122329712,
      "epoch": 0.8148480934463975,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004944191155713531,
      "loss": 5.4897,
      "mean_token_accuracy": 0.18235614597797395,
      "num_tokens": 16183023.0,
      "step": 7255
    },
    {
      "entropy": 5.606191349029541,
      "epoch": 0.8154096703543551,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004944102254490928,
      "loss": 5.4984,
      "mean_token_accuracy": 0.18199875503778457,
      "num_tokens": 16193504.0,
      "step": 7260
    },
    {
      "entropy": 5.550720834732056,
      "epoch": 0.8159712472623126,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004944013283407257,
      "loss": 5.4585,
      "mean_token_accuracy": 0.18619175106287003,
      "num_tokens": 16204730.0,
      "step": 7265
    },
    {
      "entropy": 5.565019130706787,
      "epoch": 0.8165328241702701,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004943924242465352,
      "loss": 5.4753,
      "mean_token_accuracy": 0.18423691689968108,
      "num_tokens": 16214496.0,
      "step": 7270
    },
    {
      "entropy": 5.524086284637451,
      "epoch": 0.8170944010782276,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004943835131668049,
      "loss": 5.426,
      "mean_token_accuracy": 0.18748211562633516,
      "num_tokens": 16225494.0,
      "step": 7275
    },
    {
      "entropy": 5.4620363235473635,
      "epoch": 0.8176559779861852,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004943745951018184,
      "loss": 5.3817,
      "mean_token_accuracy": 0.1807102605700493,
      "num_tokens": 16236385.0,
      "step": 7280
    },
    {
      "entropy": 5.583876895904541,
      "epoch": 0.8182175548941427,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004943656700518597,
      "loss": 5.5446,
      "mean_token_accuracy": 0.1831841751933098,
      "num_tokens": 16247300.0,
      "step": 7285
    },
    {
      "entropy": 5.631110048294067,
      "epoch": 0.8187791318021003,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004943567380172129,
      "loss": 5.5699,
      "mean_token_accuracy": 0.1760045111179352,
      "num_tokens": 16259326.0,
      "step": 7290
    },
    {
      "entropy": 5.593363904953003,
      "epoch": 0.8193407087100578,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004943477989981627,
      "loss": 5.533,
      "mean_token_accuracy": 0.18199775218963624,
      "num_tokens": 16270559.0,
      "step": 7295
    },
    {
      "entropy": 5.6280618667602536,
      "epoch": 0.8199022856180154,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004943388529949933,
      "loss": 5.5132,
      "mean_token_accuracy": 0.17606371641159058,
      "num_tokens": 16283120.0,
      "step": 7300
    },
    {
      "entropy": 5.693280553817749,
      "epoch": 0.8204638625259729,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004943299000079899,
      "loss": 5.5307,
      "mean_token_accuracy": 0.18374147713184358,
      "num_tokens": 16293449.0,
      "step": 7305
    },
    {
      "entropy": 5.583550024032593,
      "epoch": 0.8210254394339305,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004943209400374373,
      "loss": 5.5774,
      "mean_token_accuracy": 0.1706283748149872,
      "num_tokens": 16305144.0,
      "step": 7310
    },
    {
      "entropy": 5.533866262435913,
      "epoch": 0.821587016341888,
      "grad_norm": 1.109375,
      "learning_rate": 0.000494311973083621,
      "loss": 5.373,
      "mean_token_accuracy": 0.186729696393013,
      "num_tokens": 16316728.0,
      "step": 7315
    },
    {
      "entropy": 5.641025495529175,
      "epoch": 0.8221485932498456,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004943029991468264,
      "loss": 5.6503,
      "mean_token_accuracy": 0.17274267375469207,
      "num_tokens": 16329117.0,
      "step": 7320
    },
    {
      "entropy": 5.735594606399536,
      "epoch": 0.8227101701578031,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004942940182273393,
      "loss": 5.5968,
      "mean_token_accuracy": 0.1784449487924576,
      "num_tokens": 16341651.0,
      "step": 7325
    },
    {
      "entropy": 5.681063222885132,
      "epoch": 0.8232717470657607,
      "grad_norm": 1.125,
      "learning_rate": 0.0004942850303254455,
      "loss": 5.5806,
      "mean_token_accuracy": 0.1788855344057083,
      "num_tokens": 16353220.0,
      "step": 7330
    },
    {
      "entropy": 5.68979229927063,
      "epoch": 0.8238333239737182,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004942760354414313,
      "loss": 5.612,
      "mean_token_accuracy": 0.17333096116781235,
      "num_tokens": 16363748.0,
      "step": 7335
    },
    {
      "entropy": 5.439294481277466,
      "epoch": 0.8243949008816758,
      "grad_norm": 1.125,
      "learning_rate": 0.0004942670335755831,
      "loss": 5.2984,
      "mean_token_accuracy": 0.19079074263572693,
      "num_tokens": 16374086.0,
      "step": 7340
    },
    {
      "entropy": 5.546598100662232,
      "epoch": 0.8249564777896333,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004942580247281876,
      "loss": 5.4665,
      "mean_token_accuracy": 0.18552974313497544,
      "num_tokens": 16384140.0,
      "step": 7345
    },
    {
      "entropy": 5.602595853805542,
      "epoch": 0.8255180546975909,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004942490088995315,
      "loss": 5.5712,
      "mean_token_accuracy": 0.1743219807744026,
      "num_tokens": 16395940.0,
      "step": 7350
    },
    {
      "entropy": 5.570687770843506,
      "epoch": 0.8260796316055484,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004942399860899018,
      "loss": 5.4248,
      "mean_token_accuracy": 0.18304139822721482,
      "num_tokens": 16406667.0,
      "step": 7355
    },
    {
      "entropy": 5.623033571243286,
      "epoch": 0.826641208513506,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004942309562995862,
      "loss": 5.5829,
      "mean_token_accuracy": 0.1722528338432312,
      "num_tokens": 16418621.0,
      "step": 7360
    },
    {
      "entropy": 5.543011903762817,
      "epoch": 0.8272027854214634,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004942219195288716,
      "loss": 5.4797,
      "mean_token_accuracy": 0.18262684643268584,
      "num_tokens": 16429576.0,
      "step": 7365
    },
    {
      "entropy": 5.677867650985718,
      "epoch": 0.827764362329421,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004942128757780463,
      "loss": 5.5986,
      "mean_token_accuracy": 0.1752135455608368,
      "num_tokens": 16441004.0,
      "step": 7370
    },
    {
      "entropy": 5.616818618774414,
      "epoch": 0.8283259392373785,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004942038250473979,
      "loss": 5.5489,
      "mean_token_accuracy": 0.1737070068717003,
      "num_tokens": 16452183.0,
      "step": 7375
    },
    {
      "entropy": 5.549670743942261,
      "epoch": 0.8288875161453361,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004941947673372147,
      "loss": 5.49,
      "mean_token_accuracy": 0.18317970633506775,
      "num_tokens": 16463552.0,
      "step": 7380
    },
    {
      "entropy": 5.66449465751648,
      "epoch": 0.8294490930532936,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004941857026477851,
      "loss": 5.5333,
      "mean_token_accuracy": 0.1872756540775299,
      "num_tokens": 16474062.0,
      "step": 7385
    },
    {
      "entropy": 5.69481635093689,
      "epoch": 0.8300106699612512,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004941766309793978,
      "loss": 5.6122,
      "mean_token_accuracy": 0.17872422188520432,
      "num_tokens": 16487377.0,
      "step": 7390
    },
    {
      "entropy": 5.664995956420898,
      "epoch": 0.8305722468692087,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004941675523323414,
      "loss": 5.5449,
      "mean_token_accuracy": 0.1841708779335022,
      "num_tokens": 16499704.0,
      "step": 7395
    },
    {
      "entropy": 5.569231557846069,
      "epoch": 0.8311338237771663,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004941584667069052,
      "loss": 5.574,
      "mean_token_accuracy": 0.17898223400115967,
      "num_tokens": 16511954.0,
      "step": 7400
    },
    {
      "entropy": 5.558076286315918,
      "epoch": 0.8316954006851238,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004941493741033784,
      "loss": 5.4097,
      "mean_token_accuracy": 0.1776068016886711,
      "num_tokens": 16522602.0,
      "step": 7405
    },
    {
      "entropy": 5.603202533721924,
      "epoch": 0.8322569775930814,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004941402745220507,
      "loss": 5.5625,
      "mean_token_accuracy": 0.17015616297721864,
      "num_tokens": 16534233.0,
      "step": 7410
    },
    {
      "entropy": 5.684613752365112,
      "epoch": 0.8328185545010389,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004941311679632115,
      "loss": 5.6202,
      "mean_token_accuracy": 0.17935654819011687,
      "num_tokens": 16546120.0,
      "step": 7415
    },
    {
      "entropy": 5.670873737335205,
      "epoch": 0.8333801314089965,
      "grad_norm": 1.171875,
      "learning_rate": 0.000494122054427151,
      "loss": 5.5101,
      "mean_token_accuracy": 0.17859695702791215,
      "num_tokens": 16556830.0,
      "step": 7420
    },
    {
      "entropy": 5.539680290222168,
      "epoch": 0.833941708316954,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004941129339141592,
      "loss": 5.4875,
      "mean_token_accuracy": 0.17416851669549943,
      "num_tokens": 16567436.0,
      "step": 7425
    },
    {
      "entropy": 5.602557945251465,
      "epoch": 0.8345032852249116,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004941038064245267,
      "loss": 5.4443,
      "mean_token_accuracy": 0.17892910838127135,
      "num_tokens": 16577807.0,
      "step": 7430
    },
    {
      "entropy": 5.6734942436218265,
      "epoch": 0.8350648621328691,
      "grad_norm": 1.109375,
      "learning_rate": 0.000494094671958544,
      "loss": 5.5403,
      "mean_token_accuracy": 0.17929859906435014,
      "num_tokens": 16589745.0,
      "step": 7435
    },
    {
      "entropy": 5.70731782913208,
      "epoch": 0.8356264390408267,
      "grad_norm": 1.125,
      "learning_rate": 0.0004940855305165019,
      "loss": 5.5778,
      "mean_token_accuracy": 0.17803435623645783,
      "num_tokens": 16600727.0,
      "step": 7440
    },
    {
      "entropy": 5.565154123306274,
      "epoch": 0.8361880159487842,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004940763820986916,
      "loss": 5.4729,
      "mean_token_accuracy": 0.17871894389390947,
      "num_tokens": 16611091.0,
      "step": 7445
    },
    {
      "entropy": 5.48442234992981,
      "epoch": 0.8367495928567418,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004940672267054044,
      "loss": 5.4498,
      "mean_token_accuracy": 0.18054181188344956,
      "num_tokens": 16622550.0,
      "step": 7450
    },
    {
      "entropy": 5.578350639343261,
      "epoch": 0.8373111697646993,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004940580643369316,
      "loss": 5.4573,
      "mean_token_accuracy": 0.17644257396459578,
      "num_tokens": 16634243.0,
      "step": 7455
    },
    {
      "entropy": 5.567024040222168,
      "epoch": 0.8378727466726568,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004940488949935651,
      "loss": 5.4766,
      "mean_token_accuracy": 0.1778653308749199,
      "num_tokens": 16645266.0,
      "step": 7460
    },
    {
      "entropy": 5.622676944732666,
      "epoch": 0.8384343235806143,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004940397186755968,
      "loss": 5.5031,
      "mean_token_accuracy": 0.17560533732175826,
      "num_tokens": 16654365.0,
      "step": 7465
    },
    {
      "entropy": 5.563875770568847,
      "epoch": 0.8389959004885719,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004940305353833191,
      "loss": 5.4517,
      "mean_token_accuracy": 0.1842338338494301,
      "num_tokens": 16665251.0,
      "step": 7470
    },
    {
      "entropy": 5.555642890930176,
      "epoch": 0.8395574773965294,
      "grad_norm": 1.1875,
      "learning_rate": 0.000494021345117024,
      "loss": 5.5093,
      "mean_token_accuracy": 0.18142585456371307,
      "num_tokens": 16676209.0,
      "step": 7475
    },
    {
      "entropy": 5.60004448890686,
      "epoch": 0.840119054304487,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004940121478770044,
      "loss": 5.4704,
      "mean_token_accuracy": 0.17672317773103713,
      "num_tokens": 16687213.0,
      "step": 7480
    },
    {
      "entropy": 5.560383701324463,
      "epoch": 0.8406806312124445,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004940029436635531,
      "loss": 5.4949,
      "mean_token_accuracy": 0.1826560840010643,
      "num_tokens": 16697920.0,
      "step": 7485
    },
    {
      "entropy": 5.5280232429504395,
      "epoch": 0.8412422081204021,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004939937324769631,
      "loss": 5.484,
      "mean_token_accuracy": 0.17732658386230468,
      "num_tokens": 16707815.0,
      "step": 7490
    },
    {
      "entropy": 5.55229697227478,
      "epoch": 0.8418037850283596,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004939845143175276,
      "loss": 5.3819,
      "mean_token_accuracy": 0.18761899918317795,
      "num_tokens": 16719221.0,
      "step": 7495
    },
    {
      "entropy": 5.50514783859253,
      "epoch": 0.8423653619363172,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004939752891855404,
      "loss": 5.432,
      "mean_token_accuracy": 0.17894004434347152,
      "num_tokens": 16731741.0,
      "step": 7500
    },
    {
      "entropy": 5.553002643585205,
      "epoch": 0.8429269388442747,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004939660570812951,
      "loss": 5.4408,
      "mean_token_accuracy": 0.18835001289844513,
      "num_tokens": 16742795.0,
      "step": 7505
    },
    {
      "entropy": 5.560106325149536,
      "epoch": 0.8434885157522323,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004939568180050855,
      "loss": 5.4416,
      "mean_token_accuracy": 0.18138732612133027,
      "num_tokens": 16753139.0,
      "step": 7510
    },
    {
      "entropy": 5.549212551116943,
      "epoch": 0.8440500926601898,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004939475719572059,
      "loss": 5.4572,
      "mean_token_accuracy": 0.1833619624376297,
      "num_tokens": 16764246.0,
      "step": 7515
    },
    {
      "entropy": 5.6717616558074955,
      "epoch": 0.8446116695681474,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004939383189379507,
      "loss": 5.5755,
      "mean_token_accuracy": 0.17061865776777269,
      "num_tokens": 16776039.0,
      "step": 7520
    },
    {
      "entropy": 5.670359802246094,
      "epoch": 0.8451732464761049,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004939290589476145,
      "loss": 5.5513,
      "mean_token_accuracy": 0.18261810392141342,
      "num_tokens": 16786979.0,
      "step": 7525
    },
    {
      "entropy": 5.582773733139038,
      "epoch": 0.8457348233840625,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004939197919864922,
      "loss": 5.5748,
      "mean_token_accuracy": 0.1797413408756256,
      "num_tokens": 16798113.0,
      "step": 7530
    },
    {
      "entropy": 5.622071743011475,
      "epoch": 0.84629640029202,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004939105180548788,
      "loss": 5.533,
      "mean_token_accuracy": 0.17459736466407777,
      "num_tokens": 16809072.0,
      "step": 7535
    },
    {
      "entropy": 5.516169548034668,
      "epoch": 0.8468579771999776,
      "grad_norm": 1.125,
      "learning_rate": 0.0004939012371530696,
      "loss": 5.3069,
      "mean_token_accuracy": 0.1848357766866684,
      "num_tokens": 16819548.0,
      "step": 7540
    },
    {
      "entropy": 5.602190446853638,
      "epoch": 0.8474195541079351,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004938919492813601,
      "loss": 5.6197,
      "mean_token_accuracy": 0.16935227066278458,
      "num_tokens": 16831935.0,
      "step": 7545
    },
    {
      "entropy": 5.618028211593628,
      "epoch": 0.8479811310158927,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004938826544400461,
      "loss": 5.4925,
      "mean_token_accuracy": 0.18564322143793105,
      "num_tokens": 16842961.0,
      "step": 7550
    },
    {
      "entropy": 5.56212477684021,
      "epoch": 0.8485427079238502,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004938733526294234,
      "loss": 5.4252,
      "mean_token_accuracy": 0.1770094096660614,
      "num_tokens": 16854392.0,
      "step": 7555
    },
    {
      "entropy": 5.560294818878174,
      "epoch": 0.8491042848318077,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004938640438497883,
      "loss": 5.5217,
      "mean_token_accuracy": 0.17702339887619017,
      "num_tokens": 16865443.0,
      "step": 7560
    },
    {
      "entropy": 5.608917284011841,
      "epoch": 0.8496658617397652,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004938547281014372,
      "loss": 5.4715,
      "mean_token_accuracy": 0.18097776025533677,
      "num_tokens": 16876854.0,
      "step": 7565
    },
    {
      "entropy": 5.650098085403442,
      "epoch": 0.8502274386477228,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004938454053846666,
      "loss": 5.5407,
      "mean_token_accuracy": 0.18218712508678436,
      "num_tokens": 16888823.0,
      "step": 7570
    },
    {
      "entropy": 5.551910924911499,
      "epoch": 0.8507890155556803,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004938360756997735,
      "loss": 5.489,
      "mean_token_accuracy": 0.17805470675230026,
      "num_tokens": 16898807.0,
      "step": 7575
    },
    {
      "entropy": 5.697138738632202,
      "epoch": 0.8513505924636379,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004938267390470547,
      "loss": 5.5876,
      "mean_token_accuracy": 0.17045462429523467,
      "num_tokens": 16911258.0,
      "step": 7580
    },
    {
      "entropy": 5.6649847507476805,
      "epoch": 0.8519121693715954,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004938173954268079,
      "loss": 5.5573,
      "mean_token_accuracy": 0.1735591858625412,
      "num_tokens": 16922557.0,
      "step": 7585
    },
    {
      "entropy": 5.616597032546997,
      "epoch": 0.852473746279553,
      "grad_norm": 1.125,
      "learning_rate": 0.0004938080448393302,
      "loss": 5.5519,
      "mean_token_accuracy": 0.1746073842048645,
      "num_tokens": 16935330.0,
      "step": 7590
    },
    {
      "entropy": 5.635535860061646,
      "epoch": 0.8530353231875105,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004937986872849195,
      "loss": 5.5035,
      "mean_token_accuracy": 0.18408897668123245,
      "num_tokens": 16945361.0,
      "step": 7595
    },
    {
      "entropy": 5.463597202301026,
      "epoch": 0.8535969000954681,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004937893227638738,
      "loss": 5.3727,
      "mean_token_accuracy": 0.1869575098156929,
      "num_tokens": 16955363.0,
      "step": 7600
    },
    {
      "entropy": 5.4785126686096195,
      "epoch": 0.8541584770034256,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004937799512764912,
      "loss": 5.4248,
      "mean_token_accuracy": 0.17820223271846772,
      "num_tokens": 16965267.0,
      "step": 7605
    },
    {
      "entropy": 5.631610918045044,
      "epoch": 0.8547200539113832,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00049377057282307,
      "loss": 5.6411,
      "mean_token_accuracy": 0.17408629655838012,
      "num_tokens": 16977528.0,
      "step": 7610
    },
    {
      "entropy": 5.610168933868408,
      "epoch": 0.8552816308193407,
      "grad_norm": 1.921875,
      "learning_rate": 0.000493761187403909,
      "loss": 5.503,
      "mean_token_accuracy": 0.17779579609632493,
      "num_tokens": 16987326.0,
      "step": 7615
    },
    {
      "entropy": 5.672342205047608,
      "epoch": 0.8558432077272983,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000493751795019307,
      "loss": 5.5333,
      "mean_token_accuracy": 0.17939226925373078,
      "num_tokens": 16999127.0,
      "step": 7620
    },
    {
      "entropy": 5.512845087051391,
      "epoch": 0.8564047846352558,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004937423956695629,
      "loss": 5.4426,
      "mean_token_accuracy": 0.18400541990995406,
      "num_tokens": 17009929.0,
      "step": 7625
    },
    {
      "entropy": 5.712212991714478,
      "epoch": 0.8569663615432134,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004937329893549762,
      "loss": 5.5686,
      "mean_token_accuracy": 0.1770023763179779,
      "num_tokens": 17020909.0,
      "step": 7630
    },
    {
      "entropy": 5.595722961425781,
      "epoch": 0.8575279384511709,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004937235760758462,
      "loss": 5.4991,
      "mean_token_accuracy": 0.18141598999500275,
      "num_tokens": 17031358.0,
      "step": 7635
    },
    {
      "entropy": 5.563752126693726,
      "epoch": 0.8580895153591285,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004937141558324726,
      "loss": 5.5481,
      "mean_token_accuracy": 0.18018485158681868,
      "num_tokens": 17042566.0,
      "step": 7640
    },
    {
      "entropy": 5.719388771057129,
      "epoch": 0.858651092267086,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004937047286251556,
      "loss": 5.5641,
      "mean_token_accuracy": 0.16669271886348724,
      "num_tokens": 17054998.0,
      "step": 7645
    },
    {
      "entropy": 5.528896808624268,
      "epoch": 0.8592126691750436,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004936952944541952,
      "loss": 5.4363,
      "mean_token_accuracy": 0.17745328843593597,
      "num_tokens": 17064876.0,
      "step": 7650
    },
    {
      "entropy": 5.52032618522644,
      "epoch": 0.859774246083001,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004936858533198917,
      "loss": 5.4865,
      "mean_token_accuracy": 0.1823222517967224,
      "num_tokens": 17075310.0,
      "step": 7655
    },
    {
      "entropy": 5.468863916397095,
      "epoch": 0.8603358229909586,
      "grad_norm": 1.09375,
      "learning_rate": 0.000493676405222546,
      "loss": 5.3901,
      "mean_token_accuracy": 0.18763498812913895,
      "num_tokens": 17086023.0,
      "step": 7660
    },
    {
      "entropy": 5.528159093856812,
      "epoch": 0.8608973998989161,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004936669501624586,
      "loss": 5.4737,
      "mean_token_accuracy": 0.18357129395008087,
      "num_tokens": 17097554.0,
      "step": 7665
    },
    {
      "entropy": 5.824870252609253,
      "epoch": 0.8614589768068737,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004936574881399308,
      "loss": 5.5917,
      "mean_token_accuracy": 0.17834053784608841,
      "num_tokens": 17108890.0,
      "step": 7670
    },
    {
      "entropy": 5.517649364471436,
      "epoch": 0.8620205537148312,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004936480191552637,
      "loss": 5.4205,
      "mean_token_accuracy": 0.17925918102264404,
      "num_tokens": 17120651.0,
      "step": 7675
    },
    {
      "entropy": 5.573526811599732,
      "epoch": 0.8625821306227888,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004936385432087589,
      "loss": 5.5555,
      "mean_token_accuracy": 0.18306990563869477,
      "num_tokens": 17131306.0,
      "step": 7680
    },
    {
      "entropy": 5.764016628265381,
      "epoch": 0.8631437075307463,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004936290603007181,
      "loss": 5.6931,
      "mean_token_accuracy": 0.1679967001080513,
      "num_tokens": 17142779.0,
      "step": 7685
    },
    {
      "entropy": 5.726499080657959,
      "epoch": 0.8637052844387039,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004936195704314432,
      "loss": 5.5719,
      "mean_token_accuracy": 0.18899858593940735,
      "num_tokens": 17154992.0,
      "step": 7690
    },
    {
      "entropy": 5.6286399364471436,
      "epoch": 0.8642668613466614,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004936100736012363,
      "loss": 5.5506,
      "mean_token_accuracy": 0.18483092933893203,
      "num_tokens": 17166089.0,
      "step": 7695
    },
    {
      "entropy": 5.566044807434082,
      "epoch": 0.864828438254619,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004936005698104,
      "loss": 5.4691,
      "mean_token_accuracy": 0.17727130353450776,
      "num_tokens": 17175691.0,
      "step": 7700
    },
    {
      "entropy": 5.576226806640625,
      "epoch": 0.8653900151625765,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004935910590592367,
      "loss": 5.5128,
      "mean_token_accuracy": 0.17505736351013185,
      "num_tokens": 17186466.0,
      "step": 7705
    },
    {
      "entropy": 5.455333232879639,
      "epoch": 0.8659515920705341,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004935815413480494,
      "loss": 5.2936,
      "mean_token_accuracy": 0.19221956878900529,
      "num_tokens": 17198520.0,
      "step": 7710
    },
    {
      "entropy": 5.580836582183838,
      "epoch": 0.8665131689784916,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000493572016677141,
      "loss": 5.5266,
      "mean_token_accuracy": 0.1748068481683731,
      "num_tokens": 17210496.0,
      "step": 7715
    },
    {
      "entropy": 5.599395322799682,
      "epoch": 0.8670747458864492,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004935624850468148,
      "loss": 5.5113,
      "mean_token_accuracy": 0.1853922963142395,
      "num_tokens": 17221631.0,
      "step": 7720
    },
    {
      "entropy": 5.594560241699218,
      "epoch": 0.8676363227944067,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004935529464573742,
      "loss": 5.5698,
      "mean_token_accuracy": 0.17323410958051683,
      "num_tokens": 17233316.0,
      "step": 7725
    },
    {
      "entropy": 5.598330116271972,
      "epoch": 0.8681978997023643,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004935434009091232,
      "loss": 5.4645,
      "mean_token_accuracy": 0.18653941303491592,
      "num_tokens": 17245761.0,
      "step": 7730
    },
    {
      "entropy": 5.604849529266358,
      "epoch": 0.8687594766103218,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004935338484023655,
      "loss": 5.4821,
      "mean_token_accuracy": 0.18092142641544343,
      "num_tokens": 17255007.0,
      "step": 7735
    },
    {
      "entropy": 5.585720872879028,
      "epoch": 0.8693210535182794,
      "grad_norm": 1.0,
      "learning_rate": 0.0004935242889374052,
      "loss": 5.4433,
      "mean_token_accuracy": 0.1823727682232857,
      "num_tokens": 17265378.0,
      "step": 7740
    },
    {
      "entropy": 5.581008720397949,
      "epoch": 0.8698826304262369,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004935147225145469,
      "loss": 5.5197,
      "mean_token_accuracy": 0.17471259832382202,
      "num_tokens": 17277202.0,
      "step": 7745
    },
    {
      "entropy": 5.531655645370483,
      "epoch": 0.8704442073341944,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004935051491340951,
      "loss": 5.4736,
      "mean_token_accuracy": 0.18472496271133423,
      "num_tokens": 17287006.0,
      "step": 7750
    },
    {
      "entropy": 5.573832416534424,
      "epoch": 0.8710057842421519,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004934955687963546,
      "loss": 5.446,
      "mean_token_accuracy": 0.187906976044178,
      "num_tokens": 17297387.0,
      "step": 7755
    },
    {
      "entropy": 5.730034589767456,
      "epoch": 0.8715673611501095,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004934859815016304,
      "loss": 5.649,
      "mean_token_accuracy": 0.17365161925554276,
      "num_tokens": 17310364.0,
      "step": 7760
    },
    {
      "entropy": 5.543053722381591,
      "epoch": 0.872128938058067,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004934763872502278,
      "loss": 5.4994,
      "mean_token_accuracy": 0.18727059960365294,
      "num_tokens": 17321282.0,
      "step": 7765
    },
    {
      "entropy": 5.610386800765991,
      "epoch": 0.8726905149660246,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004934667860424523,
      "loss": 5.5458,
      "mean_token_accuracy": 0.17405443638563156,
      "num_tokens": 17331755.0,
      "step": 7770
    },
    {
      "entropy": 5.604792165756225,
      "epoch": 0.8732520918739821,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004934571778786096,
      "loss": 5.4798,
      "mean_token_accuracy": 0.17688754498958587,
      "num_tokens": 17342542.0,
      "step": 7775
    },
    {
      "entropy": 5.558994340896606,
      "epoch": 0.8738136687819397,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004934475627590057,
      "loss": 5.4511,
      "mean_token_accuracy": 0.1801168590784073,
      "num_tokens": 17353150.0,
      "step": 7780
    },
    {
      "entropy": 5.560382652282715,
      "epoch": 0.8743752456898972,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004934379406839466,
      "loss": 5.4911,
      "mean_token_accuracy": 0.17974613159894942,
      "num_tokens": 17363901.0,
      "step": 7785
    },
    {
      "entropy": 5.590231657028198,
      "epoch": 0.8749368225978548,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004934283116537387,
      "loss": 5.4052,
      "mean_token_accuracy": 0.18424939215183259,
      "num_tokens": 17374429.0,
      "step": 7790
    },
    {
      "entropy": 5.598772621154785,
      "epoch": 0.8754983995058123,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004934186756686888,
      "loss": 5.5535,
      "mean_token_accuracy": 0.17777433544397353,
      "num_tokens": 17384952.0,
      "step": 7795
    },
    {
      "entropy": 5.53681640625,
      "epoch": 0.8760599764137699,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004934090327291034,
      "loss": 5.5025,
      "mean_token_accuracy": 0.17633518129587172,
      "num_tokens": 17397771.0,
      "step": 7800
    },
    {
      "entropy": 5.5828369140625,
      "epoch": 0.8766215533217274,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004933993828352898,
      "loss": 5.4045,
      "mean_token_accuracy": 0.18528952151536943,
      "num_tokens": 17407748.0,
      "step": 7805
    },
    {
      "entropy": 5.5529238224029545,
      "epoch": 0.877183130229685,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004933897259875551,
      "loss": 5.403,
      "mean_token_accuracy": 0.18062180578708648,
      "num_tokens": 17418572.0,
      "step": 7810
    },
    {
      "entropy": 5.4584471702575685,
      "epoch": 0.8777447071376425,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004933800621862069,
      "loss": 5.4421,
      "mean_token_accuracy": 0.18314612954854964,
      "num_tokens": 17429909.0,
      "step": 7815
    },
    {
      "entropy": 5.532392072677612,
      "epoch": 0.8783062840456001,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004933703914315527,
      "loss": 5.466,
      "mean_token_accuracy": 0.17753292471170426,
      "num_tokens": 17441143.0,
      "step": 7820
    },
    {
      "entropy": 5.5883691787719725,
      "epoch": 0.8788678609535576,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004933607137239006,
      "loss": 5.5835,
      "mean_token_accuracy": 0.16914670914411545,
      "num_tokens": 17452156.0,
      "step": 7825
    },
    {
      "entropy": 5.54946985244751,
      "epoch": 0.8794294378615152,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004933510290635587,
      "loss": 5.4842,
      "mean_token_accuracy": 0.17690777480602266,
      "num_tokens": 17463873.0,
      "step": 7830
    },
    {
      "entropy": 5.577990484237671,
      "epoch": 0.8799910147694727,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004933413374508353,
      "loss": 5.4141,
      "mean_token_accuracy": 0.18533463925123214,
      "num_tokens": 17474180.0,
      "step": 7835
    },
    {
      "entropy": 5.557853364944458,
      "epoch": 0.8805525916774303,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004933316388860391,
      "loss": 5.4376,
      "mean_token_accuracy": 0.18558824211359023,
      "num_tokens": 17486903.0,
      "step": 7840
    },
    {
      "entropy": 5.465214347839355,
      "epoch": 0.8811141685853877,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004933219333694787,
      "loss": 5.4814,
      "mean_token_accuracy": 0.17692553550004958,
      "num_tokens": 17498790.0,
      "step": 7845
    },
    {
      "entropy": 5.6459153175354,
      "epoch": 0.8816757454933453,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004933122209014634,
      "loss": 5.5261,
      "mean_token_accuracy": 0.17037224024534225,
      "num_tokens": 17510061.0,
      "step": 7850
    },
    {
      "entropy": 5.571832084655762,
      "epoch": 0.8822373224013028,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004933025014823022,
      "loss": 5.5205,
      "mean_token_accuracy": 0.18104423731565475,
      "num_tokens": 17522414.0,
      "step": 7855
    },
    {
      "entropy": 5.824170780181885,
      "epoch": 0.8827988993092604,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004932927751123048,
      "loss": 5.6447,
      "mean_token_accuracy": 0.16792079359292983,
      "num_tokens": 17532745.0,
      "step": 7860
    },
    {
      "entropy": 5.551614761352539,
      "epoch": 0.8833604762172179,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004932830417917808,
      "loss": 5.4429,
      "mean_token_accuracy": 0.18071972578763962,
      "num_tokens": 17545329.0,
      "step": 7865
    },
    {
      "entropy": 5.5655975341796875,
      "epoch": 0.8839220531251755,
      "grad_norm": 1.171875,
      "learning_rate": 0.00049327330152104,
      "loss": 5.5412,
      "mean_token_accuracy": 0.17881713509559632,
      "num_tokens": 17557789.0,
      "step": 7870
    },
    {
      "entropy": 5.54974331855774,
      "epoch": 0.884483630033133,
      "grad_norm": 1.375,
      "learning_rate": 0.0004932635543003927,
      "loss": 5.3834,
      "mean_token_accuracy": 0.1882578507065773,
      "num_tokens": 17568338.0,
      "step": 7875
    },
    {
      "entropy": 5.558633947372437,
      "epoch": 0.8850452069410906,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004932538001301491,
      "loss": 5.4791,
      "mean_token_accuracy": 0.18262417763471603,
      "num_tokens": 17581199.0,
      "step": 7880
    },
    {
      "entropy": 5.4577343463897705,
      "epoch": 0.8856067838490481,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004932440390106199,
      "loss": 5.4295,
      "mean_token_accuracy": 0.17994212359189987,
      "num_tokens": 17592937.0,
      "step": 7885
    },
    {
      "entropy": 5.573095846176147,
      "epoch": 0.8861683607570057,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004932342709421159,
      "loss": 5.4975,
      "mean_token_accuracy": 0.18612425178289413,
      "num_tokens": 17604003.0,
      "step": 7890
    },
    {
      "entropy": 5.786309099197387,
      "epoch": 0.8867299376649632,
      "grad_norm": 1.03125,
      "learning_rate": 0.000493224495924948,
      "loss": 5.613,
      "mean_token_accuracy": 0.17902036011219025,
      "num_tokens": 17617283.0,
      "step": 7895
    },
    {
      "entropy": 5.56037163734436,
      "epoch": 0.8872915145729208,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004932147139594276,
      "loss": 5.4764,
      "mean_token_accuracy": 0.18633088618516921,
      "num_tokens": 17628161.0,
      "step": 7900
    },
    {
      "entropy": 5.680006313323974,
      "epoch": 0.8878530914808783,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004932049250458659,
      "loss": 5.558,
      "mean_token_accuracy": 0.17434368431568145,
      "num_tokens": 17639226.0,
      "step": 7905
    },
    {
      "entropy": 5.508235073089599,
      "epoch": 0.8884146683888359,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004931951291845749,
      "loss": 5.3868,
      "mean_token_accuracy": 0.18843825310468673,
      "num_tokens": 17652114.0,
      "step": 7910
    },
    {
      "entropy": 5.502807331085205,
      "epoch": 0.8889762452967934,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004931853263758663,
      "loss": 5.4405,
      "mean_token_accuracy": 0.1851072296500206,
      "num_tokens": 17663280.0,
      "step": 7915
    },
    {
      "entropy": 5.576106357574463,
      "epoch": 0.889537822204751,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004931755166200523,
      "loss": 5.5052,
      "mean_token_accuracy": 0.17715532183647156,
      "num_tokens": 17674445.0,
      "step": 7920
    },
    {
      "entropy": 5.598239469528198,
      "epoch": 0.8900993991127085,
      "grad_norm": 1.125,
      "learning_rate": 0.0004931656999174452,
      "loss": 5.4209,
      "mean_token_accuracy": 0.1800720989704132,
      "num_tokens": 17685046.0,
      "step": 7925
    },
    {
      "entropy": 5.494156217575073,
      "epoch": 0.8906609760206661,
      "grad_norm": 1.125,
      "learning_rate": 0.0004931558762683577,
      "loss": 5.4098,
      "mean_token_accuracy": 0.18820843249559402,
      "num_tokens": 17695269.0,
      "step": 7930
    },
    {
      "entropy": 5.513292264938355,
      "epoch": 0.8912225529286236,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004931460456731024,
      "loss": 5.4406,
      "mean_token_accuracy": 0.18482877612113952,
      "num_tokens": 17706296.0,
      "step": 7935
    },
    {
      "entropy": 5.594622373580933,
      "epoch": 0.8917841298365812,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004931362081319925,
      "loss": 5.3944,
      "mean_token_accuracy": 0.1828088492155075,
      "num_tokens": 17716262.0,
      "step": 7940
    },
    {
      "entropy": 5.540141201019287,
      "epoch": 0.8923457067445386,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004931263636453409,
      "loss": 5.4172,
      "mean_token_accuracy": 0.18624761551618577,
      "num_tokens": 17726818.0,
      "step": 7945
    },
    {
      "entropy": 5.486764144897461,
      "epoch": 0.8929072836524962,
      "grad_norm": 1.125,
      "learning_rate": 0.0004931165122134616,
      "loss": 5.4625,
      "mean_token_accuracy": 0.18201241940259932,
      "num_tokens": 17737112.0,
      "step": 7950
    },
    {
      "entropy": 5.620662879943848,
      "epoch": 0.8934688605604537,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004931066538366677,
      "loss": 5.5159,
      "mean_token_accuracy": 0.17521448135375978,
      "num_tokens": 17747558.0,
      "step": 7955
    },
    {
      "entropy": 5.6320953369140625,
      "epoch": 0.8940304374684113,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004930967885152734,
      "loss": 5.4411,
      "mean_token_accuracy": 0.1846146136522293,
      "num_tokens": 17757495.0,
      "step": 7960
    },
    {
      "entropy": 5.51068263053894,
      "epoch": 0.8945920143763688,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004930869162495929,
      "loss": 5.4398,
      "mean_token_accuracy": 0.1829340025782585,
      "num_tokens": 17767562.0,
      "step": 7965
    },
    {
      "entropy": 5.5893453598022464,
      "epoch": 0.8951535912843264,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004930770370399403,
      "loss": 5.5465,
      "mean_token_accuracy": 0.17439816892147064,
      "num_tokens": 17779800.0,
      "step": 7970
    },
    {
      "entropy": 5.52977933883667,
      "epoch": 0.8957151681922839,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004930671508866303,
      "loss": 5.3452,
      "mean_token_accuracy": 0.18539544492959975,
      "num_tokens": 17791183.0,
      "step": 7975
    },
    {
      "entropy": 5.571494626998901,
      "epoch": 0.8962767451002415,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004930572577899776,
      "loss": 5.5048,
      "mean_token_accuracy": 0.17850027680397035,
      "num_tokens": 17802547.0,
      "step": 7980
    },
    {
      "entropy": 5.443140363693237,
      "epoch": 0.896838322008199,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004930473577502972,
      "loss": 5.3999,
      "mean_token_accuracy": 0.18462604582309722,
      "num_tokens": 17815406.0,
      "step": 7985
    },
    {
      "entropy": 5.503264999389648,
      "epoch": 0.8973998989161566,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004930374507679044,
      "loss": 5.4144,
      "mean_token_accuracy": 0.18130321204662322,
      "num_tokens": 17826516.0,
      "step": 7990
    },
    {
      "entropy": 5.5420280456542965,
      "epoch": 0.8979614758241141,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004930275368431146,
      "loss": 5.4107,
      "mean_token_accuracy": 0.18517792373895645,
      "num_tokens": 17838193.0,
      "step": 7995
    },
    {
      "entropy": 5.520239400863647,
      "epoch": 0.8985230527320717,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004930176159762435,
      "loss": 5.4442,
      "mean_token_accuracy": 0.1820875659584999,
      "num_tokens": 17849564.0,
      "step": 8000
    },
    {
      "entropy": 5.6418212890625,
      "epoch": 0.8990846296400292,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004930076881676069,
      "loss": 5.5538,
      "mean_token_accuracy": 0.17510098665952684,
      "num_tokens": 17861874.0,
      "step": 8005
    },
    {
      "entropy": 5.491885805130005,
      "epoch": 0.8996462065479868,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004929977534175211,
      "loss": 5.4211,
      "mean_token_accuracy": 0.182296022772789,
      "num_tokens": 17873139.0,
      "step": 8010
    },
    {
      "entropy": 5.610055541992187,
      "epoch": 0.9002077834559443,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004929878117263021,
      "loss": 5.4701,
      "mean_token_accuracy": 0.1862593933939934,
      "num_tokens": 17883262.0,
      "step": 8015
    },
    {
      "entropy": 5.469749927520752,
      "epoch": 0.9007693603639019,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004929778630942668,
      "loss": 5.3675,
      "mean_token_accuracy": 0.1919856533408165,
      "num_tokens": 17895041.0,
      "step": 8020
    },
    {
      "entropy": 5.52387170791626,
      "epoch": 0.9013309372718594,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004929679075217317,
      "loss": 5.419,
      "mean_token_accuracy": 0.17987386584281922,
      "num_tokens": 17905154.0,
      "step": 8025
    },
    {
      "entropy": 5.524198818206787,
      "epoch": 0.901892514179817,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004929579450090139,
      "loss": 5.4599,
      "mean_token_accuracy": 0.17928175032138824,
      "num_tokens": 17916287.0,
      "step": 8030
    },
    {
      "entropy": 5.610744285583496,
      "epoch": 0.9024540910877745,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004929479755564305,
      "loss": 5.4789,
      "mean_token_accuracy": 0.18802051544189452,
      "num_tokens": 17927269.0,
      "step": 8035
    },
    {
      "entropy": 5.6165242195129395,
      "epoch": 0.903015667995732,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004929379991642991,
      "loss": 5.4954,
      "mean_token_accuracy": 0.17794259488582612,
      "num_tokens": 17938142.0,
      "step": 8040
    },
    {
      "entropy": 5.593934726715088,
      "epoch": 0.9035772449036895,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004929280158329373,
      "loss": 5.4683,
      "mean_token_accuracy": 0.18169382214546204,
      "num_tokens": 17950229.0,
      "step": 8045
    },
    {
      "entropy": 5.498715353012085,
      "epoch": 0.9041388218116471,
      "grad_norm": 1.21875,
      "learning_rate": 0.000492918025562663,
      "loss": 5.4533,
      "mean_token_accuracy": 0.1903412565588951,
      "num_tokens": 17961757.0,
      "step": 8050
    },
    {
      "entropy": 5.630081605911255,
      "epoch": 0.9047003987196046,
      "grad_norm": 1.25,
      "learning_rate": 0.0004929080283537942,
      "loss": 5.5378,
      "mean_token_accuracy": 0.17213768661022186,
      "num_tokens": 17972564.0,
      "step": 8055
    },
    {
      "entropy": 5.668456745147705,
      "epoch": 0.9052619756275622,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004928980242066492,
      "loss": 5.4462,
      "mean_token_accuracy": 0.18704118579626083,
      "num_tokens": 17982776.0,
      "step": 8060
    },
    {
      "entropy": 5.584461402893067,
      "epoch": 0.9058235525355197,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004928880131215467,
      "loss": 5.5478,
      "mean_token_accuracy": 0.18279054015874863,
      "num_tokens": 17994460.0,
      "step": 8065
    },
    {
      "entropy": 5.529378986358642,
      "epoch": 0.9063851294434773,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004928779950988053,
      "loss": 5.5203,
      "mean_token_accuracy": 0.17869157195091248,
      "num_tokens": 18005744.0,
      "step": 8070
    },
    {
      "entropy": 5.538164520263672,
      "epoch": 0.9069467063514348,
      "grad_norm": 1.15625,
      "learning_rate": 0.000492867970138744,
      "loss": 5.3855,
      "mean_token_accuracy": 0.1909807249903679,
      "num_tokens": 18016517.0,
      "step": 8075
    },
    {
      "entropy": 5.538620519638061,
      "epoch": 0.9075082832593924,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004928579382416821,
      "loss": 5.5059,
      "mean_token_accuracy": 0.1774222433567047,
      "num_tokens": 18027653.0,
      "step": 8080
    },
    {
      "entropy": 5.573276138305664,
      "epoch": 0.9080698601673499,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000492847899407939,
      "loss": 5.5373,
      "mean_token_accuracy": 0.17884869873523712,
      "num_tokens": 18038741.0,
      "step": 8085
    },
    {
      "entropy": 5.662741804122925,
      "epoch": 0.9086314370753075,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004928378536378341,
      "loss": 5.4833,
      "mean_token_accuracy": 0.18136989325284958,
      "num_tokens": 18049188.0,
      "step": 8090
    },
    {
      "entropy": 5.520866203308105,
      "epoch": 0.909193013983265,
      "grad_norm": 1.125,
      "learning_rate": 0.0004928278009316877,
      "loss": 5.4015,
      "mean_token_accuracy": 0.1818578362464905,
      "num_tokens": 18059979.0,
      "step": 8095
    },
    {
      "entropy": 5.5410027503967285,
      "epoch": 0.9097545908912226,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004928177412898195,
      "loss": 5.3964,
      "mean_token_accuracy": 0.18654996901750565,
      "num_tokens": 18070712.0,
      "step": 8100
    },
    {
      "entropy": 5.612816429138183,
      "epoch": 0.9103161677991801,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004928076747125499,
      "loss": 5.4865,
      "mean_token_accuracy": 0.18552515655755997,
      "num_tokens": 18081372.0,
      "step": 8105
    },
    {
      "entropy": 5.508888244628906,
      "epoch": 0.9108777447071377,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004927976012001995,
      "loss": 5.3912,
      "mean_token_accuracy": 0.18342946767807006,
      "num_tokens": 18093033.0,
      "step": 8110
    },
    {
      "entropy": 5.474383354187012,
      "epoch": 0.9114393216150952,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004927875207530889,
      "loss": 5.3563,
      "mean_token_accuracy": 0.18753510266542434,
      "num_tokens": 18104808.0,
      "step": 8115
    },
    {
      "entropy": 5.600753831863403,
      "epoch": 0.9120008985230528,
      "grad_norm": 1.125,
      "learning_rate": 0.0004927774333715394,
      "loss": 5.4994,
      "mean_token_accuracy": 0.17992793023586273,
      "num_tokens": 18115450.0,
      "step": 8120
    },
    {
      "entropy": 5.613992691040039,
      "epoch": 0.9125624754310103,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004927673390558718,
      "loss": 5.5808,
      "mean_token_accuracy": 0.1797471523284912,
      "num_tokens": 18127437.0,
      "step": 8125
    },
    {
      "entropy": 5.586420583724975,
      "epoch": 0.9131240523389679,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004927572378064077,
      "loss": 5.4287,
      "mean_token_accuracy": 0.18430067002773284,
      "num_tokens": 18137371.0,
      "step": 8130
    },
    {
      "entropy": 5.53119969367981,
      "epoch": 0.9136856292469253,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004927471296234687,
      "loss": 5.4894,
      "mean_token_accuracy": 0.17454011738300323,
      "num_tokens": 18148332.0,
      "step": 8135
    },
    {
      "entropy": 5.514609003067017,
      "epoch": 0.9142472061548829,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004927370145073767,
      "loss": 5.4643,
      "mean_token_accuracy": 0.18311170786619185,
      "num_tokens": 18159918.0,
      "step": 8140
    },
    {
      "entropy": 5.561512231826782,
      "epoch": 0.9148087830628404,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004927268924584536,
      "loss": 5.4438,
      "mean_token_accuracy": 0.18352374136447908,
      "num_tokens": 18170911.0,
      "step": 8145
    },
    {
      "entropy": 5.5542713642120365,
      "epoch": 0.915370359970798,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004927167634770219,
      "loss": 5.4494,
      "mean_token_accuracy": 0.18428521752357482,
      "num_tokens": 18183089.0,
      "step": 8150
    },
    {
      "entropy": 5.719759750366211,
      "epoch": 0.9159319368787555,
      "grad_norm": 1.046875,
      "learning_rate": 0.000492706627563404,
      "loss": 5.7225,
      "mean_token_accuracy": 0.16299810111522675,
      "num_tokens": 18195467.0,
      "step": 8155
    },
    {
      "entropy": 5.704501008987426,
      "epoch": 0.9164935137867131,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004926964847179225,
      "loss": 5.4899,
      "mean_token_accuracy": 0.18084044754505157,
      "num_tokens": 18206766.0,
      "step": 8160
    },
    {
      "entropy": 5.549582529067993,
      "epoch": 0.9170550906946706,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004926863349409006,
      "loss": 5.4687,
      "mean_token_accuracy": 0.1784779578447342,
      "num_tokens": 18217840.0,
      "step": 8165
    },
    {
      "entropy": 5.535226583480835,
      "epoch": 0.9176166676026282,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004926761782326613,
      "loss": 5.5037,
      "mean_token_accuracy": 0.17805845588445662,
      "num_tokens": 18229514.0,
      "step": 8170
    },
    {
      "entropy": 5.5417369365692135,
      "epoch": 0.9181782445105857,
      "grad_norm": 1.125,
      "learning_rate": 0.0004926660145935282,
      "loss": 5.4574,
      "mean_token_accuracy": 0.17741518318653107,
      "num_tokens": 18241949.0,
      "step": 8175
    },
    {
      "entropy": 5.605534982681275,
      "epoch": 0.9187398214185433,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004926558440238247,
      "loss": 5.4503,
      "mean_token_accuracy": 0.18233083337545394,
      "num_tokens": 18252938.0,
      "step": 8180
    },
    {
      "entropy": 5.542209005355835,
      "epoch": 0.9193013983265008,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004926456665238746,
      "loss": 5.3997,
      "mean_token_accuracy": 0.1899299904704094,
      "num_tokens": 18263417.0,
      "step": 8185
    },
    {
      "entropy": 5.550479698181152,
      "epoch": 0.9198629752344584,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004926354820940023,
      "loss": 5.5537,
      "mean_token_accuracy": 0.17874219715595246,
      "num_tokens": 18275595.0,
      "step": 8190
    },
    {
      "entropy": 5.530245018005371,
      "epoch": 0.9204245521424159,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004926252907345317,
      "loss": 5.4031,
      "mean_token_accuracy": 0.18436510264873504,
      "num_tokens": 18287510.0,
      "step": 8195
    },
    {
      "entropy": 5.610671949386597,
      "epoch": 0.9209861290503735,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004926150924457874,
      "loss": 5.4884,
      "mean_token_accuracy": 0.17641472071409225,
      "num_tokens": 18298764.0,
      "step": 8200
    },
    {
      "entropy": 5.425869131088257,
      "epoch": 0.921547705958331,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004926048872280943,
      "loss": 5.3101,
      "mean_token_accuracy": 0.18456410765647888,
      "num_tokens": 18310427.0,
      "step": 8205
    },
    {
      "entropy": 5.565805959701538,
      "epoch": 0.9221092828662886,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000492594675081777,
      "loss": 5.4549,
      "mean_token_accuracy": 0.18020558953285218,
      "num_tokens": 18322174.0,
      "step": 8210
    },
    {
      "entropy": 5.540886783599854,
      "epoch": 0.9226708597742461,
      "grad_norm": 1.125,
      "learning_rate": 0.000492584456007161,
      "loss": 5.3931,
      "mean_token_accuracy": 0.18410781323909758,
      "num_tokens": 18332372.0,
      "step": 8215
    },
    {
      "entropy": 5.57743878364563,
      "epoch": 0.9232324366822037,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004925742300045716,
      "loss": 5.4518,
      "mean_token_accuracy": 0.17820675373077394,
      "num_tokens": 18343201.0,
      "step": 8220
    },
    {
      "entropy": 5.502658653259277,
      "epoch": 0.9237940135901612,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004925639970743341,
      "loss": 5.4724,
      "mean_token_accuracy": 0.1798643797636032,
      "num_tokens": 18354124.0,
      "step": 8225
    },
    {
      "entropy": 5.536202430725098,
      "epoch": 0.9243555904981187,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004925537572167746,
      "loss": 5.4168,
      "mean_token_accuracy": 0.1862925335764885,
      "num_tokens": 18365148.0,
      "step": 8230
    },
    {
      "entropy": 5.406712245941162,
      "epoch": 0.9249171674060762,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004925435104322192,
      "loss": 5.3001,
      "mean_token_accuracy": 0.18789792954921722,
      "num_tokens": 18376244.0,
      "step": 8235
    },
    {
      "entropy": 5.575233888626099,
      "epoch": 0.9254787443140338,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000492533256720994,
      "loss": 5.473,
      "mean_token_accuracy": 0.18039861917495728,
      "num_tokens": 18386728.0,
      "step": 8240
    },
    {
      "entropy": 5.565883779525757,
      "epoch": 0.9260403212219913,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004925229960834256,
      "loss": 5.4469,
      "mean_token_accuracy": 0.18403992354869841,
      "num_tokens": 18397008.0,
      "step": 8245
    },
    {
      "entropy": 5.544535350799561,
      "epoch": 0.9266018981299489,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004925127285198407,
      "loss": 5.4256,
      "mean_token_accuracy": 0.1897626891732216,
      "num_tokens": 18408348.0,
      "step": 8250
    },
    {
      "entropy": 5.567032814025879,
      "epoch": 0.9271634750379064,
      "grad_norm": 1.5,
      "learning_rate": 0.000492502454030566,
      "loss": 5.4668,
      "mean_token_accuracy": 0.1817547619342804,
      "num_tokens": 18419239.0,
      "step": 8255
    },
    {
      "entropy": 5.582904529571533,
      "epoch": 0.927725051945864,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004924921726159288,
      "loss": 5.4637,
      "mean_token_accuracy": 0.17974608689546584,
      "num_tokens": 18428918.0,
      "step": 8260
    },
    {
      "entropy": 5.597472333908081,
      "epoch": 0.9282866288538215,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004924818842762565,
      "loss": 5.4501,
      "mean_token_accuracy": 0.18555693775415422,
      "num_tokens": 18440361.0,
      "step": 8265
    },
    {
      "entropy": 5.450218343734742,
      "epoch": 0.9288482057617791,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004924715890118768,
      "loss": 5.3301,
      "mean_token_accuracy": 0.19059054106473922,
      "num_tokens": 18450678.0,
      "step": 8270
    },
    {
      "entropy": 5.541691637039184,
      "epoch": 0.9294097826697366,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004924612868231173,
      "loss": 5.5538,
      "mean_token_accuracy": 0.1785343185067177,
      "num_tokens": 18462364.0,
      "step": 8275
    },
    {
      "entropy": 5.574919891357422,
      "epoch": 0.9299713595776942,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004924509777103059,
      "loss": 5.4373,
      "mean_token_accuracy": 0.18601747900247573,
      "num_tokens": 18472896.0,
      "step": 8280
    },
    {
      "entropy": 5.639607858657837,
      "epoch": 0.9305329364856517,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004924406616737712,
      "loss": 5.5577,
      "mean_token_accuracy": 0.17171333730220795,
      "num_tokens": 18484951.0,
      "step": 8285
    },
    {
      "entropy": 5.534964847564697,
      "epoch": 0.9310945133936093,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004924303387138415,
      "loss": 5.5212,
      "mean_token_accuracy": 0.17847742289304733,
      "num_tokens": 18496006.0,
      "step": 8290
    },
    {
      "entropy": 5.641701507568359,
      "epoch": 0.9316560903015668,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004924200088308455,
      "loss": 5.4895,
      "mean_token_accuracy": 0.18168483823537826,
      "num_tokens": 18508069.0,
      "step": 8295
    },
    {
      "entropy": 5.655606365203857,
      "epoch": 0.9322176672095244,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000492409672025112,
      "loss": 5.6028,
      "mean_token_accuracy": 0.17496457695960999,
      "num_tokens": 18519746.0,
      "step": 8300
    },
    {
      "entropy": 5.6863316059112545,
      "epoch": 0.9327792441174819,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004923993282969703,
      "loss": 5.5126,
      "mean_token_accuracy": 0.18416012525558473,
      "num_tokens": 18530249.0,
      "step": 8305
    },
    {
      "entropy": 5.570022344589233,
      "epoch": 0.9333408210254395,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004923889776467495,
      "loss": 5.5203,
      "mean_token_accuracy": 0.18545539677143097,
      "num_tokens": 18543306.0,
      "step": 8310
    },
    {
      "entropy": 5.588349056243897,
      "epoch": 0.933902397933397,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004923786200747795,
      "loss": 5.5655,
      "mean_token_accuracy": 0.17473100423812865,
      "num_tokens": 18555250.0,
      "step": 8315
    },
    {
      "entropy": 5.561530494689942,
      "epoch": 0.9344639748413546,
      "grad_norm": 1.109375,
      "learning_rate": 0.00049236825558139,
      "loss": 5.3202,
      "mean_token_accuracy": 0.18172765970230104,
      "num_tokens": 18566378.0,
      "step": 8320
    },
    {
      "entropy": 5.56834921836853,
      "epoch": 0.9350255517493121,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004923578841669108,
      "loss": 5.5203,
      "mean_token_accuracy": 0.17748099714517593,
      "num_tokens": 18576827.0,
      "step": 8325
    },
    {
      "entropy": 5.554826164245606,
      "epoch": 0.9355871286572696,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004923475058316723,
      "loss": 5.4668,
      "mean_token_accuracy": 0.18439360558986664,
      "num_tokens": 18589209.0,
      "step": 8330
    },
    {
      "entropy": 5.54688777923584,
      "epoch": 0.9361487055652271,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004923371205760049,
      "loss": 5.4766,
      "mean_token_accuracy": 0.18281819969415664,
      "num_tokens": 18599367.0,
      "step": 8335
    },
    {
      "entropy": 5.582248783111572,
      "epoch": 0.9367102824731847,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004923267284002394,
      "loss": 5.4859,
      "mean_token_accuracy": 0.1811878576874733,
      "num_tokens": 18610505.0,
      "step": 8340
    },
    {
      "entropy": 5.636788845062256,
      "epoch": 0.9372718593811422,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004923163293047065,
      "loss": 5.4784,
      "mean_token_accuracy": 0.18135701417922973,
      "num_tokens": 18622616.0,
      "step": 8345
    },
    {
      "entropy": 5.502398300170898,
      "epoch": 0.9378334362890998,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004923059232897374,
      "loss": 5.3479,
      "mean_token_accuracy": 0.18603970855474472,
      "num_tokens": 18634127.0,
      "step": 8350
    },
    {
      "entropy": 5.54591999053955,
      "epoch": 0.9383950131970573,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004922955103556633,
      "loss": 5.5454,
      "mean_token_accuracy": 0.18128058910369874,
      "num_tokens": 18646196.0,
      "step": 8355
    },
    {
      "entropy": 5.601352548599243,
      "epoch": 0.9389565901050149,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000492285090502816,
      "loss": 5.4723,
      "mean_token_accuracy": 0.181783090531826,
      "num_tokens": 18656911.0,
      "step": 8360
    },
    {
      "entropy": 5.590421867370606,
      "epoch": 0.9395181670129724,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004922746637315272,
      "loss": 5.4329,
      "mean_token_accuracy": 0.18029942363500595,
      "num_tokens": 18667553.0,
      "step": 8365
    },
    {
      "entropy": 5.486310529708862,
      "epoch": 0.94007974392093,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004922642300421288,
      "loss": 5.4341,
      "mean_token_accuracy": 0.18416116237640381,
      "num_tokens": 18678087.0,
      "step": 8370
    },
    {
      "entropy": 5.63065357208252,
      "epoch": 0.9406413208288875,
      "grad_norm": 1.171875,
      "learning_rate": 0.000492253789434953,
      "loss": 5.5169,
      "mean_token_accuracy": 0.1799348324537277,
      "num_tokens": 18689946.0,
      "step": 8375
    },
    {
      "entropy": 5.566393184661865,
      "epoch": 0.9412028977368451,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004922433419103323,
      "loss": 5.4456,
      "mean_token_accuracy": 0.1850202798843384,
      "num_tokens": 18700944.0,
      "step": 8380
    },
    {
      "entropy": 5.562990045547485,
      "epoch": 0.9417644746448026,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004922328874685993,
      "loss": 5.5103,
      "mean_token_accuracy": 0.18239544481039047,
      "num_tokens": 18712901.0,
      "step": 8385
    },
    {
      "entropy": 5.490732955932617,
      "epoch": 0.9423260515527602,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004922224261100869,
      "loss": 5.3666,
      "mean_token_accuracy": 0.18870144635438918,
      "num_tokens": 18724322.0,
      "step": 8390
    },
    {
      "entropy": 5.496328067779541,
      "epoch": 0.9428876284607177,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004922119578351283,
      "loss": 5.4325,
      "mean_token_accuracy": 0.18145502358675003,
      "num_tokens": 18735315.0,
      "step": 8395
    },
    {
      "entropy": 5.634676313400268,
      "epoch": 0.9434492053686753,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004922014826440566,
      "loss": 5.4601,
      "mean_token_accuracy": 0.18332717567682266,
      "num_tokens": 18747497.0,
      "step": 8400
    },
    {
      "entropy": 5.546881532669067,
      "epoch": 0.9440107822766328,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004921910005372056,
      "loss": 5.499,
      "mean_token_accuracy": 0.17813228368759154,
      "num_tokens": 18758835.0,
      "step": 8405
    },
    {
      "entropy": 5.485214424133301,
      "epoch": 0.9445723591845904,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004921805115149088,
      "loss": 5.3901,
      "mean_token_accuracy": 0.19434336572885513,
      "num_tokens": 18770325.0,
      "step": 8410
    },
    {
      "entropy": 5.483918523788452,
      "epoch": 0.9451339360925479,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004921700155775002,
      "loss": 5.2918,
      "mean_token_accuracy": 0.19442717581987382,
      "num_tokens": 18781204.0,
      "step": 8415
    },
    {
      "entropy": 5.497776222229004,
      "epoch": 0.9456955130005055,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004921595127253142,
      "loss": 5.3796,
      "mean_token_accuracy": 0.18635170608758928,
      "num_tokens": 18791899.0,
      "step": 8420
    },
    {
      "entropy": 5.445171213150024,
      "epoch": 0.9462570899084629,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004921490029586852,
      "loss": 5.3901,
      "mean_token_accuracy": 0.18424216657876968,
      "num_tokens": 18803043.0,
      "step": 8425
    },
    {
      "entropy": 5.542738533020019,
      "epoch": 0.9468186668164205,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004921384862779475,
      "loss": 5.4018,
      "mean_token_accuracy": 0.18350836932659148,
      "num_tokens": 18814316.0,
      "step": 8430
    },
    {
      "entropy": 5.552679395675659,
      "epoch": 0.947380243724378,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004921279626834363,
      "loss": 5.4642,
      "mean_token_accuracy": 0.18264802992343904,
      "num_tokens": 18825490.0,
      "step": 8435
    },
    {
      "entropy": 5.680787849426269,
      "epoch": 0.9479418206323356,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004921174321754865,
      "loss": 5.5629,
      "mean_token_accuracy": 0.17903024852275848,
      "num_tokens": 18836479.0,
      "step": 8440
    },
    {
      "entropy": 5.574717426300049,
      "epoch": 0.9485033975402931,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004921068947544335,
      "loss": 5.4451,
      "mean_token_accuracy": 0.18462834060192107,
      "num_tokens": 18848634.0,
      "step": 8445
    },
    {
      "entropy": 5.517153215408325,
      "epoch": 0.9490649744482507,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004920963504206128,
      "loss": 5.416,
      "mean_token_accuracy": 0.18347776234149932,
      "num_tokens": 18858451.0,
      "step": 8450
    },
    {
      "entropy": 5.613441610336304,
      "epoch": 0.9496265513562082,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00049208579917436,
      "loss": 5.5598,
      "mean_token_accuracy": 0.17390970885753632,
      "num_tokens": 18869561.0,
      "step": 8455
    },
    {
      "entropy": 5.689114570617676,
      "epoch": 0.9501881282641658,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004920752410160113,
      "loss": 5.5584,
      "mean_token_accuracy": 0.1736804276704788,
      "num_tokens": 18880232.0,
      "step": 8460
    },
    {
      "entropy": 5.6100428104400635,
      "epoch": 0.9507497051721233,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004920646759459026,
      "loss": 5.5541,
      "mean_token_accuracy": 0.18129395842552185,
      "num_tokens": 18892562.0,
      "step": 8465
    },
    {
      "entropy": 5.481253433227539,
      "epoch": 0.9513112820800809,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004920541039643705,
      "loss": 5.4733,
      "mean_token_accuracy": 0.18009088784456254,
      "num_tokens": 18903566.0,
      "step": 8470
    },
    {
      "entropy": 5.557294750213623,
      "epoch": 0.9518728589880384,
      "grad_norm": 1.125,
      "learning_rate": 0.0004920435250717515,
      "loss": 5.4893,
      "mean_token_accuracy": 0.18462950140237808,
      "num_tokens": 18914619.0,
      "step": 8475
    },
    {
      "entropy": 5.662855911254883,
      "epoch": 0.952434435895996,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004920329392683826,
      "loss": 5.5469,
      "mean_token_accuracy": 0.1810678094625473,
      "num_tokens": 18927039.0,
      "step": 8480
    },
    {
      "entropy": 5.6880840301513675,
      "epoch": 0.9529960128039535,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004920223465546006,
      "loss": 5.6112,
      "mean_token_accuracy": 0.17829274982213975,
      "num_tokens": 18940449.0,
      "step": 8485
    },
    {
      "entropy": 5.6132728099823,
      "epoch": 0.9535575897119111,
      "grad_norm": 1.125,
      "learning_rate": 0.0004920117469307429,
      "loss": 5.5171,
      "mean_token_accuracy": 0.18039372563362122,
      "num_tokens": 18950798.0,
      "step": 8490
    },
    {
      "entropy": 5.526393604278565,
      "epoch": 0.9541191666198686,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004920011403971471,
      "loss": 5.3583,
      "mean_token_accuracy": 0.18431814908981323,
      "num_tokens": 18960958.0,
      "step": 8495
    },
    {
      "entropy": 5.441552066802979,
      "epoch": 0.9546807435278262,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004919905269541508,
      "loss": 5.3945,
      "mean_token_accuracy": 0.18452154695987702,
      "num_tokens": 18971890.0,
      "step": 8500
    },
    {
      "entropy": 5.492977571487427,
      "epoch": 0.9552423204357837,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000491979906602092,
      "loss": 5.3395,
      "mean_token_accuracy": 0.19218335300683975,
      "num_tokens": 18982309.0,
      "step": 8505
    },
    {
      "entropy": 5.621786642074585,
      "epoch": 0.9558038973437413,
      "grad_norm": 1.125,
      "learning_rate": 0.0004919692793413089,
      "loss": 5.5579,
      "mean_token_accuracy": 0.17565633207559586,
      "num_tokens": 18993648.0,
      "step": 8510
    },
    {
      "entropy": 5.556702375411987,
      "epoch": 0.9563654742516988,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004919586451721396,
      "loss": 5.32,
      "mean_token_accuracy": 0.19345040619373322,
      "num_tokens": 19003613.0,
      "step": 8515
    },
    {
      "entropy": 5.503692007064819,
      "epoch": 0.9569270511596563,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004919480040949229,
      "loss": 5.4276,
      "mean_token_accuracy": 0.180625756084919,
      "num_tokens": 19014720.0,
      "step": 8520
    },
    {
      "entropy": 5.572077751159668,
      "epoch": 0.9574886280676138,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004919373561099977,
      "loss": 5.4694,
      "mean_token_accuracy": 0.18018868565559387,
      "num_tokens": 19026635.0,
      "step": 8525
    },
    {
      "entropy": 5.5622203826904295,
      "epoch": 0.9580502049755714,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000491926701217703,
      "loss": 5.4798,
      "mean_token_accuracy": 0.17838658392429352,
      "num_tokens": 19037433.0,
      "step": 8530
    },
    {
      "entropy": 5.5688458442687985,
      "epoch": 0.9586117818835289,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004919160394183779,
      "loss": 5.3506,
      "mean_token_accuracy": 0.18753570169210435,
      "num_tokens": 19047351.0,
      "step": 8535
    },
    {
      "entropy": 5.416022920608521,
      "epoch": 0.9591733587914865,
      "grad_norm": 1.25,
      "learning_rate": 0.000491905370712362,
      "loss": 5.3887,
      "mean_token_accuracy": 0.1824868232011795,
      "num_tokens": 19058561.0,
      "step": 8540
    },
    {
      "entropy": 5.553125047683716,
      "epoch": 0.959734935699444,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004918946950999949,
      "loss": 5.4784,
      "mean_token_accuracy": 0.1771282434463501,
      "num_tokens": 19069877.0,
      "step": 8545
    },
    {
      "entropy": 5.649783658981323,
      "epoch": 0.9602965126074016,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004918840125816167,
      "loss": 5.4817,
      "mean_token_accuracy": 0.18284020572900772,
      "num_tokens": 19081962.0,
      "step": 8550
    },
    {
      "entropy": 5.610421895980835,
      "epoch": 0.9608580895153591,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004918733231575673,
      "loss": 5.5596,
      "mean_token_accuracy": 0.17947977781295776,
      "num_tokens": 19092955.0,
      "step": 8555
    },
    {
      "entropy": 5.531349754333496,
      "epoch": 0.9614196664233167,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004918626268281874,
      "loss": 5.3564,
      "mean_token_accuracy": 0.1943444862961769,
      "num_tokens": 19103521.0,
      "step": 8560
    },
    {
      "entropy": 5.4311846733093265,
      "epoch": 0.9619812433312742,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004918519235938171,
      "loss": 5.3767,
      "mean_token_accuracy": 0.19120045006275177,
      "num_tokens": 19114190.0,
      "step": 8565
    },
    {
      "entropy": 5.568183946609497,
      "epoch": 0.9625428202392318,
      "grad_norm": 1.25,
      "learning_rate": 0.0004918412134547976,
      "loss": 5.4324,
      "mean_token_accuracy": 0.18152185678482055,
      "num_tokens": 19123815.0,
      "step": 8570
    },
    {
      "entropy": 5.57272777557373,
      "epoch": 0.9631043971471893,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004918304964114696,
      "loss": 5.573,
      "mean_token_accuracy": 0.17766182273626327,
      "num_tokens": 19135805.0,
      "step": 8575
    },
    {
      "entropy": 5.59317536354065,
      "epoch": 0.9636659740551469,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004918197724641747,
      "loss": 5.4026,
      "mean_token_accuracy": 0.18198202103376387,
      "num_tokens": 19146656.0,
      "step": 8580
    },
    {
      "entropy": 5.461778593063355,
      "epoch": 0.9642275509631044,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000491809041613254,
      "loss": 5.4873,
      "mean_token_accuracy": 0.1770074263215065,
      "num_tokens": 19158271.0,
      "step": 8585
    },
    {
      "entropy": 5.533248376846314,
      "epoch": 0.964789127871062,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004917983038590495,
      "loss": 5.3945,
      "mean_token_accuracy": 0.17924244850873947,
      "num_tokens": 19168439.0,
      "step": 8590
    },
    {
      "entropy": 5.586949396133423,
      "epoch": 0.9653507047790195,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004917875592019028,
      "loss": 5.4771,
      "mean_token_accuracy": 0.18402522504329683,
      "num_tokens": 19179411.0,
      "step": 8595
    },
    {
      "entropy": 5.548010921478271,
      "epoch": 0.9659122816869771,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004917768076421561,
      "loss": 5.5001,
      "mean_token_accuracy": 0.18540108799934388,
      "num_tokens": 19189377.0,
      "step": 8600
    },
    {
      "entropy": 5.5190184116363525,
      "epoch": 0.9664738585949346,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004917660491801518,
      "loss": 5.4844,
      "mean_token_accuracy": 0.1811209052801132,
      "num_tokens": 19200662.0,
      "step": 8605
    },
    {
      "entropy": 5.557572555541992,
      "epoch": 0.9670354355028922,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004917552838162324,
      "loss": 5.3632,
      "mean_token_accuracy": 0.18630160838365556,
      "num_tokens": 19211535.0,
      "step": 8610
    },
    {
      "entropy": 5.505265951156616,
      "epoch": 0.9675970124108496,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004917445115507408,
      "loss": 5.3482,
      "mean_token_accuracy": 0.1947336509823799,
      "num_tokens": 19223402.0,
      "step": 8615
    },
    {
      "entropy": 5.387311220169067,
      "epoch": 0.9681585893188072,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004917337323840198,
      "loss": 5.4373,
      "mean_token_accuracy": 0.18396042436361312,
      "num_tokens": 19234976.0,
      "step": 8620
    },
    {
      "entropy": 5.429571056365967,
      "epoch": 0.9687201662267647,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004917229463164128,
      "loss": 5.2941,
      "mean_token_accuracy": 0.19184065461158753,
      "num_tokens": 19247328.0,
      "step": 8625
    },
    {
      "entropy": 5.5710437297821045,
      "epoch": 0.9692817431347223,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004917121533482631,
      "loss": 5.4292,
      "mean_token_accuracy": 0.17496708333492278,
      "num_tokens": 19259021.0,
      "step": 8630
    },
    {
      "entropy": 5.499776697158813,
      "epoch": 0.9698433200426798,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004917013534799144,
      "loss": 5.3959,
      "mean_token_accuracy": 0.18414133936166763,
      "num_tokens": 19270021.0,
      "step": 8635
    },
    {
      "entropy": 5.5824216365814205,
      "epoch": 0.9704048969506374,
      "grad_norm": 1.125,
      "learning_rate": 0.0004916905467117104,
      "loss": 5.4521,
      "mean_token_accuracy": 0.18951265811920165,
      "num_tokens": 19281520.0,
      "step": 8640
    },
    {
      "entropy": 5.491623163223267,
      "epoch": 0.9709664738585949,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004916797330439955,
      "loss": 5.3854,
      "mean_token_accuracy": 0.1859089031815529,
      "num_tokens": 19292761.0,
      "step": 8645
    },
    {
      "entropy": 5.491825962066651,
      "epoch": 0.9715280507665525,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004916689124771138,
      "loss": 5.4875,
      "mean_token_accuracy": 0.18376732021570205,
      "num_tokens": 19305022.0,
      "step": 8650
    },
    {
      "entropy": 5.675540494918823,
      "epoch": 0.97208962767451,
      "grad_norm": 1.125,
      "learning_rate": 0.0004916580850114099,
      "loss": 5.5921,
      "mean_token_accuracy": 0.17631538808345795,
      "num_tokens": 19316162.0,
      "step": 8655
    },
    {
      "entropy": 5.752667856216431,
      "epoch": 0.9726512045824676,
      "grad_norm": 1.25,
      "learning_rate": 0.0004916472506472286,
      "loss": 5.5995,
      "mean_token_accuracy": 0.16986350268125533,
      "num_tokens": 19327015.0,
      "step": 8660
    },
    {
      "entropy": 5.523492336273193,
      "epoch": 0.9732127814904251,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004916364093849147,
      "loss": 5.4718,
      "mean_token_accuracy": 0.18123141974210738,
      "num_tokens": 19338384.0,
      "step": 8665
    },
    {
      "entropy": 5.556578207015991,
      "epoch": 0.9737743583983827,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004916255612248136,
      "loss": 5.3981,
      "mean_token_accuracy": 0.18360741436481476,
      "num_tokens": 19349846.0,
      "step": 8670
    },
    {
      "entropy": 5.575875282287598,
      "epoch": 0.9743359353063402,
      "grad_norm": 1.375,
      "learning_rate": 0.0004916147061672706,
      "loss": 5.4363,
      "mean_token_accuracy": 0.18419525325298308,
      "num_tokens": 19361399.0,
      "step": 8675
    },
    {
      "entropy": 5.49245400428772,
      "epoch": 0.9748975122142978,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004916038442126314,
      "loss": 5.3987,
      "mean_token_accuracy": 0.18022444248199462,
      "num_tokens": 19371990.0,
      "step": 8680
    },
    {
      "entropy": 5.509934949874878,
      "epoch": 0.9754590891222553,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004915929753612416,
      "loss": 5.4086,
      "mean_token_accuracy": 0.1809844970703125,
      "num_tokens": 19383135.0,
      "step": 8685
    },
    {
      "entropy": 5.560032320022583,
      "epoch": 0.9760206660302129,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004915820996134477,
      "loss": 5.481,
      "mean_token_accuracy": 0.18229870796203612,
      "num_tokens": 19393942.0,
      "step": 8690
    },
    {
      "entropy": 5.51714015007019,
      "epoch": 0.9765822429381704,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004915712169695956,
      "loss": 5.314,
      "mean_token_accuracy": 0.18732747733592986,
      "num_tokens": 19404255.0,
      "step": 8695
    },
    {
      "entropy": 5.550613260269165,
      "epoch": 0.977143819846128,
      "grad_norm": 1.15625,
      "learning_rate": 0.000491560327430032,
      "loss": 5.4194,
      "mean_token_accuracy": 0.18320263624191285,
      "num_tokens": 19415425.0,
      "step": 8700
    },
    {
      "entropy": 5.4144432067871096,
      "epoch": 0.9777053967540855,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004915494309951035,
      "loss": 5.364,
      "mean_token_accuracy": 0.18927911818027496,
      "num_tokens": 19426050.0,
      "step": 8705
    },
    {
      "entropy": 5.538457584381104,
      "epoch": 0.9782669736620431,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004915385276651573,
      "loss": 5.4625,
      "mean_token_accuracy": 0.17654711306095122,
      "num_tokens": 19437471.0,
      "step": 8710
    },
    {
      "entropy": 5.6011223793029785,
      "epoch": 0.9788285505700005,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004915276174405403,
      "loss": 5.4986,
      "mean_token_accuracy": 0.18278201967477797,
      "num_tokens": 19449569.0,
      "step": 8715
    },
    {
      "entropy": 5.57077202796936,
      "epoch": 0.9793901274779581,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004915167003215999,
      "loss": 5.4881,
      "mean_token_accuracy": 0.1808352753520012,
      "num_tokens": 19461630.0,
      "step": 8720
    },
    {
      "entropy": 5.581459617614746,
      "epoch": 0.9799517043859156,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004915057763086838,
      "loss": 5.4395,
      "mean_token_accuracy": 0.18539508879184724,
      "num_tokens": 19472626.0,
      "step": 8725
    },
    {
      "entropy": 5.479933977127075,
      "epoch": 0.9805132812938732,
      "grad_norm": 1.34375,
      "learning_rate": 0.00049149484540214,
      "loss": 5.3999,
      "mean_token_accuracy": 0.18088460862636566,
      "num_tokens": 19483141.0,
      "step": 8730
    },
    {
      "entropy": 5.559927225112915,
      "epoch": 0.9810748582018307,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004914839076023161,
      "loss": 5.3922,
      "mean_token_accuracy": 0.1912113159894943,
      "num_tokens": 19495052.0,
      "step": 8735
    },
    {
      "entropy": 5.599201822280884,
      "epoch": 0.9816364351097883,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004914729629095609,
      "loss": 5.4547,
      "mean_token_accuracy": 0.1782187670469284,
      "num_tokens": 19506911.0,
      "step": 8740
    },
    {
      "entropy": 5.458628797531128,
      "epoch": 0.9821980120177458,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004914620113242224,
      "loss": 5.345,
      "mean_token_accuracy": 0.18721815943717957,
      "num_tokens": 19517812.0,
      "step": 8745
    },
    {
      "entropy": 5.538042879104614,
      "epoch": 0.9827595889257034,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004914510528466495,
      "loss": 5.427,
      "mean_token_accuracy": 0.18227364867925644,
      "num_tokens": 19529857.0,
      "step": 8750
    },
    {
      "entropy": 5.519110822677613,
      "epoch": 0.9833211658336609,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004914400874771913,
      "loss": 5.4617,
      "mean_token_accuracy": 0.18361574560403823,
      "num_tokens": 19541082.0,
      "step": 8755
    },
    {
      "entropy": 5.509701299667358,
      "epoch": 0.9838827427416185,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004914291152161967,
      "loss": 5.3975,
      "mean_token_accuracy": 0.19055985063314437,
      "num_tokens": 19551266.0,
      "step": 8760
    },
    {
      "entropy": 5.617368459701538,
      "epoch": 0.984444319649576,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000491418136064015,
      "loss": 5.5709,
      "mean_token_accuracy": 0.18123659491539001,
      "num_tokens": 19563381.0,
      "step": 8765
    },
    {
      "entropy": 5.672165203094482,
      "epoch": 0.9850058965575336,
      "grad_norm": 1.203125,
      "learning_rate": 0.000491407150020996,
      "loss": 5.6015,
      "mean_token_accuracy": 0.1808273136615753,
      "num_tokens": 19575078.0,
      "step": 8770
    },
    {
      "entropy": 5.589415597915649,
      "epoch": 0.9855674734654911,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004913961570874895,
      "loss": 5.3965,
      "mean_token_accuracy": 0.1857842966914177,
      "num_tokens": 19585273.0,
      "step": 8775
    },
    {
      "entropy": 5.478651762008667,
      "epoch": 0.9861290503734487,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004913851572638453,
      "loss": 5.4127,
      "mean_token_accuracy": 0.18559372425079346,
      "num_tokens": 19594245.0,
      "step": 8780
    },
    {
      "entropy": 5.60260124206543,
      "epoch": 0.9866906272814062,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004913741505504138,
      "loss": 5.4401,
      "mean_token_accuracy": 0.1803918659687042,
      "num_tokens": 19604920.0,
      "step": 8785
    },
    {
      "entropy": 5.6152997493743895,
      "epoch": 0.9872522041893638,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004913631369475454,
      "loss": 5.5372,
      "mean_token_accuracy": 0.17321525067090987,
      "num_tokens": 19616592.0,
      "step": 8790
    },
    {
      "entropy": 5.610131072998047,
      "epoch": 0.9878137810973213,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004913521164555909,
      "loss": 5.5009,
      "mean_token_accuracy": 0.17741193771362304,
      "num_tokens": 19627359.0,
      "step": 8795
    },
    {
      "entropy": 5.610148572921753,
      "epoch": 0.9883753580052789,
      "grad_norm": 1.0625,
      "learning_rate": 0.000491341089074901,
      "loss": 5.4751,
      "mean_token_accuracy": 0.1841354176402092,
      "num_tokens": 19639685.0,
      "step": 8800
    },
    {
      "entropy": 5.444202184677124,
      "epoch": 0.9889369349132364,
      "grad_norm": 1.0625,
      "learning_rate": 0.000491330054805827,
      "loss": 5.3076,
      "mean_token_accuracy": 0.1948076456785202,
      "num_tokens": 19650525.0,
      "step": 8805
    },
    {
      "entropy": 5.430496454238892,
      "epoch": 0.9894985118211939,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004913190136487201,
      "loss": 5.359,
      "mean_token_accuracy": 0.18796226233243943,
      "num_tokens": 19662241.0,
      "step": 8810
    },
    {
      "entropy": 5.588249969482422,
      "epoch": 0.9900600887291514,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004913079656039318,
      "loss": 5.4442,
      "mean_token_accuracy": 0.1836868479847908,
      "num_tokens": 19673752.0,
      "step": 8815
    },
    {
      "entropy": 5.49324746131897,
      "epoch": 0.990621665637109,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004912969106718142,
      "loss": 5.4307,
      "mean_token_accuracy": 0.18198232501745223,
      "num_tokens": 19685309.0,
      "step": 8820
    },
    {
      "entropy": 5.509182024002075,
      "epoch": 0.9911832425450665,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000491285848852719,
      "loss": 5.5403,
      "mean_token_accuracy": 0.18304151743650438,
      "num_tokens": 19697391.0,
      "step": 8825
    },
    {
      "entropy": 5.67175817489624,
      "epoch": 0.9917448194530241,
      "grad_norm": 1.25,
      "learning_rate": 0.0004912747801469984,
      "loss": 5.4896,
      "mean_token_accuracy": 0.1824740380048752,
      "num_tokens": 19707426.0,
      "step": 8830
    },
    {
      "entropy": 5.588723850250244,
      "epoch": 0.9923063963609816,
      "grad_norm": 1.09375,
      "learning_rate": 0.000491263704555005,
      "loss": 5.497,
      "mean_token_accuracy": 0.18016745448112487,
      "num_tokens": 19719800.0,
      "step": 8835
    },
    {
      "entropy": 5.467302989959717,
      "epoch": 0.9928679732689392,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004912526220770913,
      "loss": 5.3954,
      "mean_token_accuracy": 0.17956112772226335,
      "num_tokens": 19731600.0,
      "step": 8840
    },
    {
      "entropy": 5.556245040893555,
      "epoch": 0.9934295501768967,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004912415327136103,
      "loss": 5.4059,
      "mean_token_accuracy": 0.18448400646448135,
      "num_tokens": 19742610.0,
      "step": 8845
    },
    {
      "entropy": 5.621222734451294,
      "epoch": 0.9939911270848543,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000491230436464915,
      "loss": 5.6071,
      "mean_token_accuracy": 0.17388029992580414,
      "num_tokens": 19754207.0,
      "step": 8850
    },
    {
      "entropy": 5.535386848449707,
      "epoch": 0.9945527039928118,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004912193333313589,
      "loss": 5.3261,
      "mean_token_accuracy": 0.18944538980722428,
      "num_tokens": 19765894.0,
      "step": 8855
    },
    {
      "entropy": 5.497449779510498,
      "epoch": 0.9951142809007694,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004912082233132951,
      "loss": 5.4882,
      "mean_token_accuracy": 0.18101087361574172,
      "num_tokens": 19777867.0,
      "step": 8860
    },
    {
      "entropy": 5.4919740676879885,
      "epoch": 0.9956758578087269,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004911971064110778,
      "loss": 5.3688,
      "mean_token_accuracy": 0.1865628778934479,
      "num_tokens": 19788346.0,
      "step": 8865
    },
    {
      "entropy": 5.630201530456543,
      "epoch": 0.9962374347166845,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004911859826250607,
      "loss": 5.513,
      "mean_token_accuracy": 0.18572864085435867,
      "num_tokens": 19799752.0,
      "step": 8870
    },
    {
      "entropy": 5.560143041610718,
      "epoch": 0.996799011624642,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004911748519555981,
      "loss": 5.4694,
      "mean_token_accuracy": 0.18455371409654617,
      "num_tokens": 19810387.0,
      "step": 8875
    },
    {
      "entropy": 5.510699081420898,
      "epoch": 0.9973605885325996,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004911637144030444,
      "loss": 5.4324,
      "mean_token_accuracy": 0.18388308882713317,
      "num_tokens": 19822092.0,
      "step": 8880
    },
    {
      "entropy": 5.564812278747558,
      "epoch": 0.9979221654405571,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004911525699677541,
      "loss": 5.5362,
      "mean_token_accuracy": 0.1704825609922409,
      "num_tokens": 19833731.0,
      "step": 8885
    },
    {
      "entropy": 5.545547819137573,
      "epoch": 0.9984837423485147,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004911414186500822,
      "loss": 5.4738,
      "mean_token_accuracy": 0.17951803505420685,
      "num_tokens": 19845101.0,
      "step": 8890
    },
    {
      "entropy": 5.5828978538513185,
      "epoch": 0.9990453192564722,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004911302604503837,
      "loss": 5.5271,
      "mean_token_accuracy": 0.18034427911043166,
      "num_tokens": 19857006.0,
      "step": 8895
    },
    {
      "entropy": 5.5666149139404295,
      "epoch": 0.9996068961644298,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004911190953690139,
      "loss": 5.4083,
      "mean_token_accuracy": 0.1853315055370331,
      "num_tokens": 19868007.0,
      "step": 8900
    },
    {
      "entropy": 5.6724240514967175,
      "epoch": 1.0001123153815914,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004911079234063284,
      "loss": 5.3248,
      "mean_token_accuracy": 0.18326379689905378,
      "num_tokens": 19877897.0,
      "step": 8905
    },
    {
      "entropy": 5.52108244895935,
      "epoch": 1.000673892289549,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004910967445626826,
      "loss": 5.3879,
      "mean_token_accuracy": 0.1833508253097534,
      "num_tokens": 19888449.0,
      "step": 8910
    },
    {
      "entropy": 5.477229404449463,
      "epoch": 1.0012354691975065,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004910855588384327,
      "loss": 5.3625,
      "mean_token_accuracy": 0.18187325596809387,
      "num_tokens": 19898998.0,
      "step": 8915
    },
    {
      "entropy": 5.5847920894622805,
      "epoch": 1.0017970461054642,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004910743662339348,
      "loss": 5.3721,
      "mean_token_accuracy": 0.18803175240755082,
      "num_tokens": 19910369.0,
      "step": 8920
    },
    {
      "entropy": 5.587127685546875,
      "epoch": 1.0023586230134216,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004910631667495452,
      "loss": 5.4154,
      "mean_token_accuracy": 0.19084014296531676,
      "num_tokens": 19921936.0,
      "step": 8925
    },
    {
      "entropy": 5.478022575378418,
      "epoch": 1.0029201999213793,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004910519603856207,
      "loss": 5.396,
      "mean_token_accuracy": 0.18707295060157775,
      "num_tokens": 19933943.0,
      "step": 8930
    },
    {
      "entropy": 5.398440265655518,
      "epoch": 1.0034817768293367,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004910407471425179,
      "loss": 5.1986,
      "mean_token_accuracy": 0.19720520228147506,
      "num_tokens": 19944377.0,
      "step": 8935
    },
    {
      "entropy": 5.436029863357544,
      "epoch": 1.0040433537372944,
      "grad_norm": 1.328125,
      "learning_rate": 0.000491029527020594,
      "loss": 5.309,
      "mean_token_accuracy": 0.19129389971494676,
      "num_tokens": 19954865.0,
      "step": 8940
    },
    {
      "entropy": 5.403644323348999,
      "epoch": 1.0046049306452518,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004910183000202061,
      "loss": 5.231,
      "mean_token_accuracy": 0.19731839001178741,
      "num_tokens": 19965021.0,
      "step": 8945
    },
    {
      "entropy": 5.4806358337402346,
      "epoch": 1.0051665075532095,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004910070661417119,
      "loss": 5.3227,
      "mean_token_accuracy": 0.18687240928411483,
      "num_tokens": 19977068.0,
      "step": 8950
    },
    {
      "entropy": 5.43753137588501,
      "epoch": 1.005728084461167,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004909958253854687,
      "loss": 5.3716,
      "mean_token_accuracy": 0.18202826976776124,
      "num_tokens": 19987994.0,
      "step": 8955
    },
    {
      "entropy": 5.562107849121094,
      "epoch": 1.0062896613691246,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004909845777518347,
      "loss": 5.3708,
      "mean_token_accuracy": 0.18176992237567902,
      "num_tokens": 19998598.0,
      "step": 8960
    },
    {
      "entropy": 5.489080953598022,
      "epoch": 1.006851238277082,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004909733232411681,
      "loss": 5.2657,
      "mean_token_accuracy": 0.1959861010313034,
      "num_tokens": 20008881.0,
      "step": 8965
    },
    {
      "entropy": 5.4734149932861325,
      "epoch": 1.0074128151850397,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004909620618538271,
      "loss": 5.3175,
      "mean_token_accuracy": 0.18793402910232543,
      "num_tokens": 20019105.0,
      "step": 8970
    },
    {
      "entropy": 5.479021835327148,
      "epoch": 1.007974392092997,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004909507935901702,
      "loss": 5.3599,
      "mean_token_accuracy": 0.18327762335538864,
      "num_tokens": 20031805.0,
      "step": 8975
    },
    {
      "entropy": 5.470980644226074,
      "epoch": 1.0085359690009548,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004909395184505564,
      "loss": 5.3339,
      "mean_token_accuracy": 0.18901901245117186,
      "num_tokens": 20041493.0,
      "step": 8980
    },
    {
      "entropy": 5.435040616989136,
      "epoch": 1.0090975459089122,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004909282364353445,
      "loss": 5.3651,
      "mean_token_accuracy": 0.18890712708234786,
      "num_tokens": 20052489.0,
      "step": 8985
    },
    {
      "entropy": 5.398237705230713,
      "epoch": 1.0096591228168699,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004909169475448938,
      "loss": 5.3157,
      "mean_token_accuracy": 0.18595487326383592,
      "num_tokens": 20064252.0,
      "step": 8990
    },
    {
      "entropy": 5.533651542663574,
      "epoch": 1.0102206997248273,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004909056517795638,
      "loss": 5.274,
      "mean_token_accuracy": 0.1855074107646942,
      "num_tokens": 20075171.0,
      "step": 8995
    },
    {
      "entropy": 5.44556622505188,
      "epoch": 1.010782276632785,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004908943491397141,
      "loss": 5.2649,
      "mean_token_accuracy": 0.18751833140850066,
      "num_tokens": 20085746.0,
      "step": 9000
    },
    {
      "epoch": 1.010782276632785,
      "eval_entropy": 5.286550342652304,
      "eval_loss": 5.491282939910889,
      "eval_mean_token_accuracy": 0.18775901345015802,
      "eval_num_tokens": 20085746.0,
      "eval_runtime": 23.8717,
      "eval_samples_per_second": 1299.154,
      "eval_steps_per_second": 162.41,
      "step": 9000
    },
    {
      "entropy": 5.4073874950408936,
      "epoch": 1.0113438535407424,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004908830396257045,
      "loss": 5.3186,
      "mean_token_accuracy": 0.18990394920110704,
      "num_tokens": 20097435.0,
      "step": 9005
    },
    {
      "entropy": 5.492513847351074,
      "epoch": 1.0119054304486998,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004908717232378953,
      "loss": 5.3493,
      "mean_token_accuracy": 0.18201406449079513,
      "num_tokens": 20108100.0,
      "step": 9010
    },
    {
      "entropy": 5.525619745254517,
      "epoch": 1.0124670073566575,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004908603999766467,
      "loss": 5.3397,
      "mean_token_accuracy": 0.18651466816663742,
      "num_tokens": 20119042.0,
      "step": 9015
    },
    {
      "entropy": 5.489761209487915,
      "epoch": 1.013028584264615,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004908490698423191,
      "loss": 5.3944,
      "mean_token_accuracy": 0.18759676814079285,
      "num_tokens": 20130275.0,
      "step": 9020
    },
    {
      "entropy": 5.460607671737671,
      "epoch": 1.0135901611725726,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004908377328352736,
      "loss": 5.35,
      "mean_token_accuracy": 0.18441600054502488,
      "num_tokens": 20142523.0,
      "step": 9025
    },
    {
      "entropy": 5.445541906356811,
      "epoch": 1.01415173808053,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004908263889558709,
      "loss": 5.3224,
      "mean_token_accuracy": 0.1882548913359642,
      "num_tokens": 20153732.0,
      "step": 9030
    },
    {
      "entropy": 5.5418047428131105,
      "epoch": 1.0147133149884877,
      "grad_norm": 1.25,
      "learning_rate": 0.0004908150382044722,
      "loss": 5.3976,
      "mean_token_accuracy": 0.18300312012434006,
      "num_tokens": 20165136.0,
      "step": 9035
    },
    {
      "entropy": 5.481294870376587,
      "epoch": 1.0152748918964452,
      "grad_norm": 1.265625,
      "learning_rate": 0.000490803680581439,
      "loss": 5.2971,
      "mean_token_accuracy": 0.18985533863306045,
      "num_tokens": 20175326.0,
      "step": 9040
    },
    {
      "entropy": 5.442216873168945,
      "epoch": 1.0158364688044028,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000490792316087133,
      "loss": 5.258,
      "mean_token_accuracy": 0.19479164034128188,
      "num_tokens": 20186678.0,
      "step": 9045
    },
    {
      "entropy": 5.425194454193115,
      "epoch": 1.0163980457123603,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004907809447219158,
      "loss": 5.2838,
      "mean_token_accuracy": 0.1908227398991585,
      "num_tokens": 20196696.0,
      "step": 9050
    },
    {
      "entropy": 5.555895137786865,
      "epoch": 1.016959622620318,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004907695664861498,
      "loss": 5.2985,
      "mean_token_accuracy": 0.18947342038154602,
      "num_tokens": 20207566.0,
      "step": 9055
    },
    {
      "entropy": 5.422399044036865,
      "epoch": 1.0175211995282754,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004907581813801971,
      "loss": 5.3206,
      "mean_token_accuracy": 0.18712590187788009,
      "num_tokens": 20218233.0,
      "step": 9060
    },
    {
      "entropy": 5.459586524963379,
      "epoch": 1.018082776436233,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004907467894044202,
      "loss": 5.2385,
      "mean_token_accuracy": 0.18630337864160537,
      "num_tokens": 20228524.0,
      "step": 9065
    },
    {
      "entropy": 5.458630132675171,
      "epoch": 1.0186443533441905,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004907353905591817,
      "loss": 5.3362,
      "mean_token_accuracy": 0.18609109222888948,
      "num_tokens": 20240821.0,
      "step": 9070
    },
    {
      "entropy": 5.448675918579101,
      "epoch": 1.0192059302521481,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000490723984844845,
      "loss": 5.3284,
      "mean_token_accuracy": 0.18641338795423507,
      "num_tokens": 20252282.0,
      "step": 9075
    },
    {
      "entropy": 5.474467420578003,
      "epoch": 1.0197675071601056,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004907125722617727,
      "loss": 5.3314,
      "mean_token_accuracy": 0.1869756266474724,
      "num_tokens": 20263462.0,
      "step": 9080
    },
    {
      "entropy": 5.479928398132325,
      "epoch": 1.0203290840680632,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004907011528103285,
      "loss": 5.3611,
      "mean_token_accuracy": 0.18182427138090135,
      "num_tokens": 20275062.0,
      "step": 9085
    },
    {
      "entropy": 5.499647808074951,
      "epoch": 1.0208906609760207,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004906897264908759,
      "loss": 5.3262,
      "mean_token_accuracy": 0.18446984589099885,
      "num_tokens": 20286243.0,
      "step": 9090
    },
    {
      "entropy": 5.499685525894165,
      "epoch": 1.0214522378839783,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004906782933037789,
      "loss": 5.3455,
      "mean_token_accuracy": 0.18770674467086793,
      "num_tokens": 20296929.0,
      "step": 9095
    },
    {
      "entropy": 5.476659679412842,
      "epoch": 1.0220138147919358,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004906668532494012,
      "loss": 5.2324,
      "mean_token_accuracy": 0.18749124109745025,
      "num_tokens": 20308354.0,
      "step": 9100
    },
    {
      "entropy": 5.454270744323731,
      "epoch": 1.0225753916998932,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004906554063281074,
      "loss": 5.415,
      "mean_token_accuracy": 0.18400025218725205,
      "num_tokens": 20319487.0,
      "step": 9105
    },
    {
      "entropy": 5.4176102638244625,
      "epoch": 1.0231369686078509,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004906439525402618,
      "loss": 5.3375,
      "mean_token_accuracy": 0.18842825144529343,
      "num_tokens": 20329967.0,
      "step": 9110
    },
    {
      "entropy": 5.583118724822998,
      "epoch": 1.0236985455158083,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004906324918862291,
      "loss": 5.4093,
      "mean_token_accuracy": 0.1840923935174942,
      "num_tokens": 20341442.0,
      "step": 9115
    },
    {
      "entropy": 5.573734378814697,
      "epoch": 1.024260122423766,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004906210243663742,
      "loss": 5.4928,
      "mean_token_accuracy": 0.17829691171646117,
      "num_tokens": 20352818.0,
      "step": 9120
    },
    {
      "entropy": 5.578510284423828,
      "epoch": 1.0248216993317234,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004906095499810622,
      "loss": 5.408,
      "mean_token_accuracy": 0.18446095585823058,
      "num_tokens": 20364787.0,
      "step": 9125
    },
    {
      "entropy": 5.532891845703125,
      "epoch": 1.025383276239681,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004905980687306586,
      "loss": 5.2918,
      "mean_token_accuracy": 0.19612229168415068,
      "num_tokens": 20375491.0,
      "step": 9130
    },
    {
      "entropy": 5.490852355957031,
      "epoch": 1.0259448531476385,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004905865806155288,
      "loss": 5.3548,
      "mean_token_accuracy": 0.19025377780199051,
      "num_tokens": 20386270.0,
      "step": 9135
    },
    {
      "entropy": 5.548947334289551,
      "epoch": 1.0265064300555962,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004905750856360387,
      "loss": 5.4515,
      "mean_token_accuracy": 0.1789783701300621,
      "num_tokens": 20397953.0,
      "step": 9140
    },
    {
      "entropy": 5.558340930938721,
      "epoch": 1.0270680069635536,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004905635837925543,
      "loss": 5.4212,
      "mean_token_accuracy": 0.18149974793195725,
      "num_tokens": 20409847.0,
      "step": 9145
    },
    {
      "entropy": 5.490560483932495,
      "epoch": 1.0276295838715113,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004905520750854418,
      "loss": 5.4218,
      "mean_token_accuracy": 0.18184102326631546,
      "num_tokens": 20420737.0,
      "step": 9150
    },
    {
      "entropy": 5.4451171875,
      "epoch": 1.0281911607794687,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004905405595150674,
      "loss": 5.3073,
      "mean_token_accuracy": 0.18362639397382735,
      "num_tokens": 20432395.0,
      "step": 9155
    },
    {
      "entropy": 5.439031314849854,
      "epoch": 1.0287527376874264,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004905290370817982,
      "loss": 5.2987,
      "mean_token_accuracy": 0.1919223576784134,
      "num_tokens": 20442964.0,
      "step": 9160
    },
    {
      "entropy": 5.476463031768799,
      "epoch": 1.0293143145953838,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000490517507786001,
      "loss": 5.3436,
      "mean_token_accuracy": 0.1777918443083763,
      "num_tokens": 20454050.0,
      "step": 9165
    },
    {
      "entropy": 5.547261142730713,
      "epoch": 1.0298758915033415,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004905059716280425,
      "loss": 5.3633,
      "mean_token_accuracy": 0.1803059086203575,
      "num_tokens": 20465335.0,
      "step": 9170
    },
    {
      "entropy": 5.490376901626587,
      "epoch": 1.030437468411299,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004904944286082904,
      "loss": 5.3376,
      "mean_token_accuracy": 0.18705192357301711,
      "num_tokens": 20476781.0,
      "step": 9175
    },
    {
      "entropy": 5.470410776138306,
      "epoch": 1.0309990453192566,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004904828787271121,
      "loss": 5.3572,
      "mean_token_accuracy": 0.18612927198410034,
      "num_tokens": 20488599.0,
      "step": 9180
    },
    {
      "entropy": 5.4022469997406,
      "epoch": 1.031560622227214,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004904713219848753,
      "loss": 5.2585,
      "mean_token_accuracy": 0.19425875395536424,
      "num_tokens": 20498215.0,
      "step": 9185
    },
    {
      "entropy": 5.4949523448944095,
      "epoch": 1.0321221991351717,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004904597583819481,
      "loss": 5.2789,
      "mean_token_accuracy": 0.19382501244544983,
      "num_tokens": 20509764.0,
      "step": 9190
    },
    {
      "entropy": 5.649295949935913,
      "epoch": 1.032683776043129,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004904481879186987,
      "loss": 5.5116,
      "mean_token_accuracy": 0.1846738949418068,
      "num_tokens": 20521582.0,
      "step": 9195
    },
    {
      "entropy": 5.484695863723755,
      "epoch": 1.0332453529510865,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004904366105954954,
      "loss": 5.3918,
      "mean_token_accuracy": 0.1853424534201622,
      "num_tokens": 20532050.0,
      "step": 9200
    },
    {
      "entropy": 5.489978265762329,
      "epoch": 1.0338069298590442,
      "grad_norm": 1.25,
      "learning_rate": 0.0004904250264127067,
      "loss": 5.3694,
      "mean_token_accuracy": 0.18200583308935164,
      "num_tokens": 20543733.0,
      "step": 9205
    },
    {
      "entropy": 5.5596178531646725,
      "epoch": 1.0343685067670016,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004904134353707018,
      "loss": 5.3992,
      "mean_token_accuracy": 0.18875092417001724,
      "num_tokens": 20554898.0,
      "step": 9210
    },
    {
      "entropy": 5.423420000076294,
      "epoch": 1.0349300836749593,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004904018374698496,
      "loss": 5.2712,
      "mean_token_accuracy": 0.19046754539012908,
      "num_tokens": 20565243.0,
      "step": 9215
    },
    {
      "entropy": 5.468940591812133,
      "epoch": 1.0354916605829168,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004903902327105191,
      "loss": 5.2742,
      "mean_token_accuracy": 0.1936987340450287,
      "num_tokens": 20575321.0,
      "step": 9220
    },
    {
      "entropy": 5.523394393920898,
      "epoch": 1.0360532374908744,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004903786210930803,
      "loss": 5.4516,
      "mean_token_accuracy": 0.18120960742235184,
      "num_tokens": 20586450.0,
      "step": 9225
    },
    {
      "entropy": 5.552796983718872,
      "epoch": 1.0366148143988319,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004903670026179025,
      "loss": 5.4154,
      "mean_token_accuracy": 0.19785406291484833,
      "num_tokens": 20596996.0,
      "step": 9230
    },
    {
      "entropy": 5.468369054794311,
      "epoch": 1.0371763913067895,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000490355377285356,
      "loss": 5.33,
      "mean_token_accuracy": 0.183905166387558,
      "num_tokens": 20609136.0,
      "step": 9235
    },
    {
      "entropy": 5.507441711425781,
      "epoch": 1.037737968214747,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004903437450958106,
      "loss": 5.4003,
      "mean_token_accuracy": 0.1859529957175255,
      "num_tokens": 20620246.0,
      "step": 9240
    },
    {
      "entropy": 5.493914699554443,
      "epoch": 1.0382995451227046,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004903321060496369,
      "loss": 5.3412,
      "mean_token_accuracy": 0.18463215976953506,
      "num_tokens": 20631678.0,
      "step": 9245
    },
    {
      "entropy": 5.475243663787841,
      "epoch": 1.038861122030662,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004903204601472054,
      "loss": 5.2758,
      "mean_token_accuracy": 0.1933049112558365,
      "num_tokens": 20643287.0,
      "step": 9250
    },
    {
      "entropy": 5.465200328826905,
      "epoch": 1.0394226989386197,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004903088073888869,
      "loss": 5.327,
      "mean_token_accuracy": 0.1912533923983574,
      "num_tokens": 20655091.0,
      "step": 9255
    },
    {
      "entropy": 5.502106428146362,
      "epoch": 1.0399842758465772,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004902971477750526,
      "loss": 5.355,
      "mean_token_accuracy": 0.1882336840033531,
      "num_tokens": 20665043.0,
      "step": 9260
    },
    {
      "entropy": 5.421684837341308,
      "epoch": 1.0405458527545348,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004902854813060736,
      "loss": 5.2739,
      "mean_token_accuracy": 0.1954953908920288,
      "num_tokens": 20675225.0,
      "step": 9265
    },
    {
      "entropy": 5.453634405136109,
      "epoch": 1.0411074296624923,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004902738079823213,
      "loss": 5.3776,
      "mean_token_accuracy": 0.1887046992778778,
      "num_tokens": 20686542.0,
      "step": 9270
    },
    {
      "entropy": 5.46964430809021,
      "epoch": 1.04166900657045,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004902621278041674,
      "loss": 5.284,
      "mean_token_accuracy": 0.18962110877037047,
      "num_tokens": 20697926.0,
      "step": 9275
    },
    {
      "entropy": 5.508344745635986,
      "epoch": 1.0422305834784074,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004902504407719839,
      "loss": 5.4123,
      "mean_token_accuracy": 0.18312690854072572,
      "num_tokens": 20707890.0,
      "step": 9280
    },
    {
      "entropy": 5.473344898223877,
      "epoch": 1.042792160386365,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004902387468861429,
      "loss": 5.3331,
      "mean_token_accuracy": 0.18250956386327744,
      "num_tokens": 20719554.0,
      "step": 9285
    },
    {
      "entropy": 5.482661008834839,
      "epoch": 1.0433537372943225,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004902270461470169,
      "loss": 5.294,
      "mean_token_accuracy": 0.1863250270485878,
      "num_tokens": 20730899.0,
      "step": 9290
    },
    {
      "entropy": 5.489888477325439,
      "epoch": 1.04391531420228,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000490215338554978,
      "loss": 5.4287,
      "mean_token_accuracy": 0.18589134514331818,
      "num_tokens": 20741839.0,
      "step": 9295
    },
    {
      "entropy": 5.521352100372314,
      "epoch": 1.0444768911102376,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004902036241103994,
      "loss": 5.3755,
      "mean_token_accuracy": 0.1874153882265091,
      "num_tokens": 20753199.0,
      "step": 9300
    },
    {
      "entropy": 5.5476010799407955,
      "epoch": 1.045038468018195,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004901919028136539,
      "loss": 5.4575,
      "mean_token_accuracy": 0.18416406810283661,
      "num_tokens": 20764827.0,
      "step": 9305
    },
    {
      "entropy": 5.545746612548828,
      "epoch": 1.0456000449261527,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004901801746651147,
      "loss": 5.3301,
      "mean_token_accuracy": 0.19145667999982835,
      "num_tokens": 20774953.0,
      "step": 9310
    },
    {
      "entropy": 5.493142747879029,
      "epoch": 1.04616162183411,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004901684396651554,
      "loss": 5.342,
      "mean_token_accuracy": 0.18823156356811524,
      "num_tokens": 20784428.0,
      "step": 9315
    },
    {
      "entropy": 5.4824645042419435,
      "epoch": 1.0467231987420678,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004901566978141494,
      "loss": 5.4148,
      "mean_token_accuracy": 0.19024739563465118,
      "num_tokens": 20795672.0,
      "step": 9320
    },
    {
      "entropy": 5.532974481582642,
      "epoch": 1.0472847756500252,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004901449491124708,
      "loss": 5.3966,
      "mean_token_accuracy": 0.1851976051926613,
      "num_tokens": 20806972.0,
      "step": 9325
    },
    {
      "entropy": 5.543832063674927,
      "epoch": 1.0478463525579829,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004901331935604936,
      "loss": 5.427,
      "mean_token_accuracy": 0.18406199365854264,
      "num_tokens": 20817867.0,
      "step": 9330
    },
    {
      "entropy": 5.424962902069092,
      "epoch": 1.0484079294659403,
      "grad_norm": 1.125,
      "learning_rate": 0.000490121431158592,
      "loss": 5.2578,
      "mean_token_accuracy": 0.19031718671321868,
      "num_tokens": 20829205.0,
      "step": 9335
    },
    {
      "entropy": 5.488347911834717,
      "epoch": 1.048969506373898,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004901096619071407,
      "loss": 5.3799,
      "mean_token_accuracy": 0.18812948018312453,
      "num_tokens": 20840478.0,
      "step": 9340
    },
    {
      "entropy": 5.52170991897583,
      "epoch": 1.0495310832818554,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004900978858065142,
      "loss": 5.3403,
      "mean_token_accuracy": 0.18497686833143234,
      "num_tokens": 20850752.0,
      "step": 9345
    },
    {
      "entropy": 5.518002510070801,
      "epoch": 1.050092660189813,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004900861028570877,
      "loss": 5.3523,
      "mean_token_accuracy": 0.18545917123556138,
      "num_tokens": 20862261.0,
      "step": 9350
    },
    {
      "entropy": 5.53676929473877,
      "epoch": 1.0506542370977705,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004900743130592362,
      "loss": 5.2879,
      "mean_token_accuracy": 0.19013646095991135,
      "num_tokens": 20872604.0,
      "step": 9355
    },
    {
      "entropy": 5.372616958618164,
      "epoch": 1.0512158140057282,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004900625164133351,
      "loss": 5.2465,
      "mean_token_accuracy": 0.20051155835390091,
      "num_tokens": 20882751.0,
      "step": 9360
    },
    {
      "entropy": 5.442980480194092,
      "epoch": 1.0517773909136856,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004900507129197601,
      "loss": 5.3483,
      "mean_token_accuracy": 0.18516869693994523,
      "num_tokens": 20893809.0,
      "step": 9365
    },
    {
      "entropy": 5.546369504928589,
      "epoch": 1.0523389678216433,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004900389025788871,
      "loss": 5.2598,
      "mean_token_accuracy": 0.19659538567066193,
      "num_tokens": 20904201.0,
      "step": 9370
    },
    {
      "entropy": 5.509145307540893,
      "epoch": 1.0529005447296007,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000490027085391092,
      "loss": 5.3552,
      "mean_token_accuracy": 0.18328656107187272,
      "num_tokens": 20915226.0,
      "step": 9375
    },
    {
      "entropy": 5.458776330947876,
      "epoch": 1.0534621216375584,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000490015261356751,
      "loss": 5.3241,
      "mean_token_accuracy": 0.18109127432107924,
      "num_tokens": 20926040.0,
      "step": 9380
    },
    {
      "entropy": 5.503231716156006,
      "epoch": 1.0540236985455158,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004900034304762408,
      "loss": 5.3571,
      "mean_token_accuracy": 0.19169080555438994,
      "num_tokens": 20937051.0,
      "step": 9385
    },
    {
      "entropy": 5.558363580703736,
      "epoch": 1.0545852754534732,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000489991592749938,
      "loss": 5.388,
      "mean_token_accuracy": 0.18837668746709824,
      "num_tokens": 20947810.0,
      "step": 9390
    },
    {
      "entropy": 5.495450162887574,
      "epoch": 1.055146852361431,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004899797481782193,
      "loss": 5.3336,
      "mean_token_accuracy": 0.18289926648139954,
      "num_tokens": 20958307.0,
      "step": 9395
    },
    {
      "entropy": 5.423851680755615,
      "epoch": 1.0557084292693883,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004899678967614623,
      "loss": 5.2449,
      "mean_token_accuracy": 0.19056120216846467,
      "num_tokens": 20969493.0,
      "step": 9400
    },
    {
      "entropy": 5.466188621520996,
      "epoch": 1.056270006177346,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004899560385000439,
      "loss": 5.4091,
      "mean_token_accuracy": 0.1896325781941414,
      "num_tokens": 20980224.0,
      "step": 9405
    },
    {
      "entropy": 5.509534215927124,
      "epoch": 1.0568315830853034,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000489944173394342,
      "loss": 5.3582,
      "mean_token_accuracy": 0.18891672790050507,
      "num_tokens": 20991282.0,
      "step": 9410
    },
    {
      "entropy": 5.521426725387573,
      "epoch": 1.0573931599932611,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004899323014447342,
      "loss": 5.388,
      "mean_token_accuracy": 0.18168681710958481,
      "num_tokens": 21002395.0,
      "step": 9415
    },
    {
      "entropy": 5.449054384231568,
      "epoch": 1.0579547369012186,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004899204226515985,
      "loss": 5.2799,
      "mean_token_accuracy": 0.18943409621715546,
      "num_tokens": 21014136.0,
      "step": 9420
    },
    {
      "entropy": 5.506497240066528,
      "epoch": 1.0585163138091762,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004899085370153133,
      "loss": 5.3953,
      "mean_token_accuracy": 0.18619657605886458,
      "num_tokens": 21025637.0,
      "step": 9425
    },
    {
      "entropy": 5.68226637840271,
      "epoch": 1.0590778907171337,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004898966445362569,
      "loss": 5.4754,
      "mean_token_accuracy": 0.17564401924610137,
      "num_tokens": 21036393.0,
      "step": 9430
    },
    {
      "entropy": 5.588928604125977,
      "epoch": 1.0596394676250913,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004898847452148079,
      "loss": 5.4115,
      "mean_token_accuracy": 0.18799329847097396,
      "num_tokens": 21047339.0,
      "step": 9435
    },
    {
      "entropy": 5.5453239440917965,
      "epoch": 1.0602010445330488,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004898728390513454,
      "loss": 5.5179,
      "mean_token_accuracy": 0.1782512605190277,
      "num_tokens": 21059427.0,
      "step": 9440
    },
    {
      "entropy": 5.556725740432739,
      "epoch": 1.0607626214410064,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004898609260462483,
      "loss": 5.3416,
      "mean_token_accuracy": 0.18254511505365373,
      "num_tokens": 21068666.0,
      "step": 9445
    },
    {
      "entropy": 5.535867691040039,
      "epoch": 1.0613241983489639,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004898490061998961,
      "loss": 5.3766,
      "mean_token_accuracy": 0.18224382847547532,
      "num_tokens": 21080161.0,
      "step": 9450
    },
    {
      "entropy": 5.4877927780151365,
      "epoch": 1.0618857752569215,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004898370795126681,
      "loss": 5.4165,
      "mean_token_accuracy": 0.1775943249464035,
      "num_tokens": 21092363.0,
      "step": 9455
    },
    {
      "entropy": 5.441190528869629,
      "epoch": 1.062447352164879,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004898251459849441,
      "loss": 5.3338,
      "mean_token_accuracy": 0.18905844390392304,
      "num_tokens": 21103683.0,
      "step": 9460
    },
    {
      "entropy": 5.448581075668335,
      "epoch": 1.0630089290728366,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004898132056171041,
      "loss": 5.2913,
      "mean_token_accuracy": 0.18911735117435455,
      "num_tokens": 21115110.0,
      "step": 9465
    },
    {
      "entropy": 5.511330318450928,
      "epoch": 1.063570505980794,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004898012584095285,
      "loss": 5.2742,
      "mean_token_accuracy": 0.19287243634462356,
      "num_tokens": 21124327.0,
      "step": 9470
    },
    {
      "entropy": 5.412844610214234,
      "epoch": 1.0641320828887517,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004897893043625975,
      "loss": 5.4237,
      "mean_token_accuracy": 0.1861091822385788,
      "num_tokens": 21135542.0,
      "step": 9475
    },
    {
      "entropy": 5.442606210708618,
      "epoch": 1.0646936597967092,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004897773434766917,
      "loss": 5.2901,
      "mean_token_accuracy": 0.1920531362295151,
      "num_tokens": 21147621.0,
      "step": 9480
    },
    {
      "entropy": 5.519318437576294,
      "epoch": 1.0652552367046666,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000489765375752192,
      "loss": 5.3194,
      "mean_token_accuracy": 0.18672218769788743,
      "num_tokens": 21159725.0,
      "step": 9485
    },
    {
      "entropy": 5.462702512741089,
      "epoch": 1.0658168136126243,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004897534011894793,
      "loss": 5.3576,
      "mean_token_accuracy": 0.18832917064428328,
      "num_tokens": 21171303.0,
      "step": 9490
    },
    {
      "entropy": 5.593023252487183,
      "epoch": 1.0663783905205817,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004897414197889352,
      "loss": 5.454,
      "mean_token_accuracy": 0.18444048762321472,
      "num_tokens": 21183376.0,
      "step": 9495
    },
    {
      "entropy": 5.509600639343262,
      "epoch": 1.0669399674285394,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000489729431550941,
      "loss": 5.3305,
      "mean_token_accuracy": 0.1814745768904686,
      "num_tokens": 21194365.0,
      "step": 9500
    },
    {
      "entropy": 5.362879800796509,
      "epoch": 1.0675015443364968,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004897174364758783,
      "loss": 5.195,
      "mean_token_accuracy": 0.19651224464178085,
      "num_tokens": 21206440.0,
      "step": 9505
    },
    {
      "entropy": 5.3682225227355955,
      "epoch": 1.0680631212444545,
      "grad_norm": 1.125,
      "learning_rate": 0.0004897054345641291,
      "loss": 5.2499,
      "mean_token_accuracy": 0.19196462631225586,
      "num_tokens": 21217188.0,
      "step": 9510
    },
    {
      "entropy": 5.439888525009155,
      "epoch": 1.068624698152412,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004896934258160757,
      "loss": 5.344,
      "mean_token_accuracy": 0.18494148105382918,
      "num_tokens": 21227716.0,
      "step": 9515
    },
    {
      "entropy": 5.528606653213501,
      "epoch": 1.0691862750603696,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004896814102321004,
      "loss": 5.4522,
      "mean_token_accuracy": 0.17632656395435334,
      "num_tokens": 21241258.0,
      "step": 9520
    },
    {
      "entropy": 5.468914937973023,
      "epoch": 1.069747851968327,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004896693878125857,
      "loss": 5.2693,
      "mean_token_accuracy": 0.18632973432540895,
      "num_tokens": 21252543.0,
      "step": 9525
    },
    {
      "entropy": 5.511707973480225,
      "epoch": 1.0703094288762847,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004896573585579144,
      "loss": 5.3836,
      "mean_token_accuracy": 0.17754755467176436,
      "num_tokens": 21264829.0,
      "step": 9530
    },
    {
      "entropy": 5.56954140663147,
      "epoch": 1.070871005784242,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004896453224684695,
      "loss": 5.394,
      "mean_token_accuracy": 0.1829053446650505,
      "num_tokens": 21276240.0,
      "step": 9535
    },
    {
      "entropy": 5.545070028305053,
      "epoch": 1.0714325826921998,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004896332795446345,
      "loss": 5.4113,
      "mean_token_accuracy": 0.1789473384618759,
      "num_tokens": 21288877.0,
      "step": 9540
    },
    {
      "entropy": 5.539117288589478,
      "epoch": 1.0719941596001572,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004896212297867924,
      "loss": 5.4396,
      "mean_token_accuracy": 0.18080787658691405,
      "num_tokens": 21300147.0,
      "step": 9545
    },
    {
      "entropy": 5.454386186599732,
      "epoch": 1.0725557365081149,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004896091731953271,
      "loss": 5.2405,
      "mean_token_accuracy": 0.19084898382425308,
      "num_tokens": 21309731.0,
      "step": 9550
    },
    {
      "entropy": 5.4623034477233885,
      "epoch": 1.0731173134160723,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004895971097706227,
      "loss": 5.327,
      "mean_token_accuracy": 0.18598544597625732,
      "num_tokens": 21320358.0,
      "step": 9555
    },
    {
      "entropy": 5.416333198547363,
      "epoch": 1.07367889032403,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004895850395130629,
      "loss": 5.2156,
      "mean_token_accuracy": 0.19267586022615432,
      "num_tokens": 21331077.0,
      "step": 9560
    },
    {
      "entropy": 5.499848461151123,
      "epoch": 1.0742404672319874,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004895729624230323,
      "loss": 5.4231,
      "mean_token_accuracy": 0.18569728285074233,
      "num_tokens": 21341304.0,
      "step": 9565
    },
    {
      "entropy": 5.5434314727783205,
      "epoch": 1.074802044139945,
      "grad_norm": 1.125,
      "learning_rate": 0.0004895608785009153,
      "loss": 5.3763,
      "mean_token_accuracy": 0.1845643013715744,
      "num_tokens": 21353663.0,
      "step": 9570
    },
    {
      "entropy": 5.474525213241577,
      "epoch": 1.0753636210479025,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004895487877470969,
      "loss": 5.3597,
      "mean_token_accuracy": 0.18633938133716582,
      "num_tokens": 21364987.0,
      "step": 9575
    },
    {
      "entropy": 5.433635282516479,
      "epoch": 1.07592519795586,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004895366901619617,
      "loss": 5.3338,
      "mean_token_accuracy": 0.18123296201229094,
      "num_tokens": 21376169.0,
      "step": 9580
    },
    {
      "entropy": 5.486388158798218,
      "epoch": 1.0764867748638176,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004895245857458951,
      "loss": 5.297,
      "mean_token_accuracy": 0.18666536509990692,
      "num_tokens": 21387141.0,
      "step": 9585
    },
    {
      "entropy": 5.481012916564941,
      "epoch": 1.077048351771775,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004895124744992825,
      "loss": 5.3006,
      "mean_token_accuracy": 0.18934306204319,
      "num_tokens": 21396804.0,
      "step": 9590
    },
    {
      "entropy": 5.384679889678955,
      "epoch": 1.0776099286797327,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004895003564225095,
      "loss": 5.3148,
      "mean_token_accuracy": 0.19204550087451935,
      "num_tokens": 21406948.0,
      "step": 9595
    },
    {
      "entropy": 5.476761388778686,
      "epoch": 1.0781715055876901,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004894882315159621,
      "loss": 5.3088,
      "mean_token_accuracy": 0.18533989191055297,
      "num_tokens": 21417560.0,
      "step": 9600
    },
    {
      "entropy": 5.477229213714599,
      "epoch": 1.0787330824956478,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000489476099780026,
      "loss": 5.2872,
      "mean_token_accuracy": 0.19165398329496383,
      "num_tokens": 21428596.0,
      "step": 9605
    },
    {
      "entropy": 5.511223983764649,
      "epoch": 1.0792946594036052,
      "grad_norm": 1.25,
      "learning_rate": 0.0004894639612150879,
      "loss": 5.3865,
      "mean_token_accuracy": 0.17936157435178757,
      "num_tokens": 21440004.0,
      "step": 9610
    },
    {
      "entropy": 5.432472085952758,
      "epoch": 1.079856236311563,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004894518158215339,
      "loss": 5.3414,
      "mean_token_accuracy": 0.18925511986017227,
      "num_tokens": 21451423.0,
      "step": 9615
    },
    {
      "entropy": 5.529710006713867,
      "epoch": 1.0804178132195204,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000489439663599751,
      "loss": 5.3909,
      "mean_token_accuracy": 0.1860866591334343,
      "num_tokens": 21462065.0,
      "step": 9620
    },
    {
      "entropy": 5.535223531723022,
      "epoch": 1.080979390127478,
      "grad_norm": 1.140625,
      "learning_rate": 0.000489427504550126,
      "loss": 5.3191,
      "mean_token_accuracy": 0.18701447248458863,
      "num_tokens": 21473476.0,
      "step": 9625
    },
    {
      "entropy": 5.542981195449829,
      "epoch": 1.0815409670354355,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004894153386730462,
      "loss": 5.4396,
      "mean_token_accuracy": 0.18573597371578215,
      "num_tokens": 21485591.0,
      "step": 9630
    },
    {
      "entropy": 5.564949369430542,
      "epoch": 1.0821025439433931,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004894031659688988,
      "loss": 5.5279,
      "mean_token_accuracy": 0.17762351036071777,
      "num_tokens": 21497016.0,
      "step": 9635
    },
    {
      "entropy": 5.498974466323853,
      "epoch": 1.0826641208513506,
      "grad_norm": 1.125,
      "learning_rate": 0.0004893909864380714,
      "loss": 5.3352,
      "mean_token_accuracy": 0.1890597641468048,
      "num_tokens": 21508179.0,
      "step": 9640
    },
    {
      "entropy": 5.535704708099365,
      "epoch": 1.0832256977593082,
      "grad_norm": 1.21875,
      "learning_rate": 0.000489378800080952,
      "loss": 5.4439,
      "mean_token_accuracy": 0.17539966106414795,
      "num_tokens": 21520053.0,
      "step": 9645
    },
    {
      "entropy": 5.538521766662598,
      "epoch": 1.0837872746672657,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004893666068979283,
      "loss": 5.358,
      "mean_token_accuracy": 0.1889518156647682,
      "num_tokens": 21530746.0,
      "step": 9650
    },
    {
      "entropy": 5.482185125350952,
      "epoch": 1.0843488515752233,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004893544068893887,
      "loss": 5.3679,
      "mean_token_accuracy": 0.182982437312603,
      "num_tokens": 21540829.0,
      "step": 9655
    },
    {
      "entropy": 5.451089572906494,
      "epoch": 1.0849104284831808,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004893422000557217,
      "loss": 5.3002,
      "mean_token_accuracy": 0.1920873314142227,
      "num_tokens": 21552337.0,
      "step": 9660
    },
    {
      "entropy": 5.4669698715209964,
      "epoch": 1.0854720053911384,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004893299863973159,
      "loss": 5.3191,
      "mean_token_accuracy": 0.18807172924280166,
      "num_tokens": 21563752.0,
      "step": 9665
    },
    {
      "entropy": 5.566503429412842,
      "epoch": 1.0860335822990959,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004893177659145601,
      "loss": 5.4463,
      "mean_token_accuracy": 0.1807871088385582,
      "num_tokens": 21574834.0,
      "step": 9670
    },
    {
      "entropy": 5.502054738998413,
      "epoch": 1.0865951592070533,
      "grad_norm": 1.125,
      "learning_rate": 0.0004893055386078438,
      "loss": 5.3574,
      "mean_token_accuracy": 0.1802908882498741,
      "num_tokens": 21586674.0,
      "step": 9675
    },
    {
      "entropy": 5.544361639022827,
      "epoch": 1.087156736115011,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004892933044775558,
      "loss": 5.2792,
      "mean_token_accuracy": 0.1942461535334587,
      "num_tokens": 21595998.0,
      "step": 9680
    },
    {
      "entropy": 5.533473062515259,
      "epoch": 1.0877183130229684,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004892810635240859,
      "loss": 5.3958,
      "mean_token_accuracy": 0.18526570945978166,
      "num_tokens": 21607122.0,
      "step": 9685
    },
    {
      "entropy": 5.459864473342895,
      "epoch": 1.088279889930926,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004892688157478238,
      "loss": 5.2935,
      "mean_token_accuracy": 0.18829833567142487,
      "num_tokens": 21618780.0,
      "step": 9690
    },
    {
      "entropy": 5.480999851226807,
      "epoch": 1.0888414668388835,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004892565611491597,
      "loss": 5.3619,
      "mean_token_accuracy": 0.1869007781147957,
      "num_tokens": 21628868.0,
      "step": 9695
    },
    {
      "entropy": 5.510925531387329,
      "epoch": 1.0894030437468412,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004892442997284834,
      "loss": 5.2297,
      "mean_token_accuracy": 0.19964221864938736,
      "num_tokens": 21640189.0,
      "step": 9700
    },
    {
      "entropy": 5.557855939865112,
      "epoch": 1.0899646206547986,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004892320314861855,
      "loss": 5.4975,
      "mean_token_accuracy": 0.18298663645982743,
      "num_tokens": 21651452.0,
      "step": 9705
    },
    {
      "entropy": 5.384308624267578,
      "epoch": 1.0905261975627563,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004892197564226568,
      "loss": 5.2276,
      "mean_token_accuracy": 0.1897611975669861,
      "num_tokens": 21662555.0,
      "step": 9710
    },
    {
      "entropy": 5.4625331401824955,
      "epoch": 1.0910877744707137,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004892074745382878,
      "loss": 5.2707,
      "mean_token_accuracy": 0.19439806342124938,
      "num_tokens": 21673087.0,
      "step": 9715
    },
    {
      "entropy": 5.5283345699310305,
      "epoch": 1.0916493513786714,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004891951858334698,
      "loss": 5.3534,
      "mean_token_accuracy": 0.18657973557710647,
      "num_tokens": 21684084.0,
      "step": 9720
    },
    {
      "entropy": 5.418667507171631,
      "epoch": 1.0922109282866288,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004891828903085939,
      "loss": 5.2913,
      "mean_token_accuracy": 0.19195762276649475,
      "num_tokens": 21695765.0,
      "step": 9725
    },
    {
      "entropy": 5.554813194274902,
      "epoch": 1.0927725051945865,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004891705879640518,
      "loss": 5.3889,
      "mean_token_accuracy": 0.1861494615674019,
      "num_tokens": 21707396.0,
      "step": 9730
    },
    {
      "entropy": 5.507980728149414,
      "epoch": 1.093334082102544,
      "grad_norm": 1.125,
      "learning_rate": 0.0004891582788002351,
      "loss": 5.3023,
      "mean_token_accuracy": 0.19269267618656158,
      "num_tokens": 21718330.0,
      "step": 9735
    },
    {
      "entropy": 5.491648578643799,
      "epoch": 1.0938956590105016,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004891459628175358,
      "loss": 5.4428,
      "mean_token_accuracy": 0.18146657794713975,
      "num_tokens": 21729841.0,
      "step": 9740
    },
    {
      "entropy": 5.513806962966919,
      "epoch": 1.094457235918459,
      "grad_norm": 1.296875,
      "learning_rate": 0.000489133640016346,
      "loss": 5.3766,
      "mean_token_accuracy": 0.1862921565771103,
      "num_tokens": 21740403.0,
      "step": 9745
    },
    {
      "entropy": 5.550662279129028,
      "epoch": 1.0950188128264167,
      "grad_norm": 1.125,
      "learning_rate": 0.0004891213103970579,
      "loss": 5.3544,
      "mean_token_accuracy": 0.18901250064373015,
      "num_tokens": 21750696.0,
      "step": 9750
    },
    {
      "entropy": 5.421404933929443,
      "epoch": 1.095580389734374,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004891089739600643,
      "loss": 5.2603,
      "mean_token_accuracy": 0.19166276454925538,
      "num_tokens": 21761324.0,
      "step": 9755
    },
    {
      "entropy": 5.481038618087768,
      "epoch": 1.0961419666423318,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004890966307057581,
      "loss": 5.3452,
      "mean_token_accuracy": 0.18238668739795685,
      "num_tokens": 21774046.0,
      "step": 9760
    },
    {
      "entropy": 5.508424425125122,
      "epoch": 1.0967035435502892,
      "grad_norm": 1.125,
      "learning_rate": 0.000489084280634532,
      "loss": 5.3462,
      "mean_token_accuracy": 0.18711432069540024,
      "num_tokens": 21784625.0,
      "step": 9765
    },
    {
      "entropy": 5.556131744384766,
      "epoch": 1.0972651204582466,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004890719237467794,
      "loss": 5.439,
      "mean_token_accuracy": 0.17785610258579254,
      "num_tokens": 21795971.0,
      "step": 9770
    },
    {
      "entropy": 5.559793519973755,
      "epoch": 1.0978266973662043,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004890595600428937,
      "loss": 5.4059,
      "mean_token_accuracy": 0.1868266135454178,
      "num_tokens": 21806624.0,
      "step": 9775
    },
    {
      "entropy": 5.4982444763183596,
      "epoch": 1.0983882742741617,
      "grad_norm": 1.125,
      "learning_rate": 0.0004890471895232684,
      "loss": 5.3433,
      "mean_token_accuracy": 0.18478710800409318,
      "num_tokens": 21817830.0,
      "step": 9780
    },
    {
      "entropy": 5.457978248596191,
      "epoch": 1.0989498511821194,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004890348121882978,
      "loss": 5.3292,
      "mean_token_accuracy": 0.19188831448554994,
      "num_tokens": 21829540.0,
      "step": 9785
    },
    {
      "entropy": 5.554062986373902,
      "epoch": 1.0995114280900768,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004890224280383758,
      "loss": 5.4992,
      "mean_token_accuracy": 0.18281646817922592,
      "num_tokens": 21840550.0,
      "step": 9790
    },
    {
      "entropy": 5.589440536499024,
      "epoch": 1.1000730049980345,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004890100370738966,
      "loss": 5.4414,
      "mean_token_accuracy": 0.18216482549905777,
      "num_tokens": 21853366.0,
      "step": 9795
    },
    {
      "entropy": 5.521340847015381,
      "epoch": 1.100634581905992,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004889976392952548,
      "loss": 5.3551,
      "mean_token_accuracy": 0.19158031046390533,
      "num_tokens": 21863522.0,
      "step": 9800
    },
    {
      "entropy": 5.495182943344116,
      "epoch": 1.1011961588139496,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004889852347028452,
      "loss": 5.3183,
      "mean_token_accuracy": 0.18832553625106813,
      "num_tokens": 21873988.0,
      "step": 9805
    },
    {
      "entropy": 5.430131196975708,
      "epoch": 1.101757735721907,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004889728232970626,
      "loss": 5.2532,
      "mean_token_accuracy": 0.19541654884815216,
      "num_tokens": 21884595.0,
      "step": 9810
    },
    {
      "entropy": 5.45057225227356,
      "epoch": 1.1023193126298647,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004889604050783025,
      "loss": 5.2926,
      "mean_token_accuracy": 0.19278643429279327,
      "num_tokens": 21894643.0,
      "step": 9815
    },
    {
      "entropy": 5.475078916549682,
      "epoch": 1.1028808895378222,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048894798004696,
      "loss": 5.2953,
      "mean_token_accuracy": 0.18918010592460632,
      "num_tokens": 21906608.0,
      "step": 9820
    },
    {
      "entropy": 5.55851149559021,
      "epoch": 1.1034424664457798,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004889355482034309,
      "loss": 5.4275,
      "mean_token_accuracy": 0.1857927054166794,
      "num_tokens": 21919386.0,
      "step": 9825
    },
    {
      "entropy": 5.3658825874328615,
      "epoch": 1.1040040433537373,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004889231095481111,
      "loss": 5.298,
      "mean_token_accuracy": 0.19070663601160048,
      "num_tokens": 21930554.0,
      "step": 9830
    },
    {
      "entropy": 5.5421754837036135,
      "epoch": 1.104565620261695,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004889106640813964,
      "loss": 5.4016,
      "mean_token_accuracy": 0.19191469848155976,
      "num_tokens": 21941764.0,
      "step": 9835
    },
    {
      "entropy": 5.4746252536773685,
      "epoch": 1.1051271971696524,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004888982118036831,
      "loss": 5.2551,
      "mean_token_accuracy": 0.19387619197368622,
      "num_tokens": 21951804.0,
      "step": 9840
    },
    {
      "entropy": 5.4817784309387205,
      "epoch": 1.10568877407761,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004888857527153679,
      "loss": 5.4139,
      "mean_token_accuracy": 0.18176181316375734,
      "num_tokens": 21963356.0,
      "step": 9845
    },
    {
      "entropy": 5.483180236816406,
      "epoch": 1.1062503509855675,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004888732868168473,
      "loss": 5.3175,
      "mean_token_accuracy": 0.1873909577727318,
      "num_tokens": 21973785.0,
      "step": 9850
    },
    {
      "entropy": 5.508083295822144,
      "epoch": 1.1068119278935251,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004888608141085183,
      "loss": 5.3303,
      "mean_token_accuracy": 0.1844068631529808,
      "num_tokens": 21985057.0,
      "step": 9855
    },
    {
      "entropy": 5.504779720306397,
      "epoch": 1.1073735048014826,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004888483345907781,
      "loss": 5.3774,
      "mean_token_accuracy": 0.18658454418182374,
      "num_tokens": 21996554.0,
      "step": 9860
    },
    {
      "entropy": 5.4759446144104,
      "epoch": 1.10793508170944,
      "grad_norm": 1.234375,
      "learning_rate": 0.000488835848264024,
      "loss": 5.3161,
      "mean_token_accuracy": 0.18562721759080886,
      "num_tokens": 22008284.0,
      "step": 9865
    },
    {
      "entropy": 5.440015411376953,
      "epoch": 1.1084966586173977,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004888233551286534,
      "loss": 5.3864,
      "mean_token_accuracy": 0.18118107467889785,
      "num_tokens": 22019322.0,
      "step": 9870
    },
    {
      "entropy": 5.56392822265625,
      "epoch": 1.109058235525355,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004888108551850644,
      "loss": 5.4432,
      "mean_token_accuracy": 0.17740663141012192,
      "num_tokens": 22031444.0,
      "step": 9875
    },
    {
      "entropy": 5.474537038803101,
      "epoch": 1.1096198124333128,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004887983484336548,
      "loss": 5.261,
      "mean_token_accuracy": 0.19392662048339843,
      "num_tokens": 22042704.0,
      "step": 9880
    },
    {
      "entropy": 5.476427602767944,
      "epoch": 1.1101813893412702,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004887858348748229,
      "loss": 5.3748,
      "mean_token_accuracy": 0.18840647637844085,
      "num_tokens": 22052721.0,
      "step": 9885
    },
    {
      "entropy": 5.534831142425537,
      "epoch": 1.1107429662492279,
      "grad_norm": 1.171875,
      "learning_rate": 0.000488773314508967,
      "loss": 5.3806,
      "mean_token_accuracy": 0.18577093333005906,
      "num_tokens": 22063857.0,
      "step": 9890
    },
    {
      "entropy": 5.463254165649414,
      "epoch": 1.1113045431571853,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004887607873364859,
      "loss": 5.3255,
      "mean_token_accuracy": 0.18929167836904526,
      "num_tokens": 22074777.0,
      "step": 9895
    },
    {
      "entropy": 5.534868049621582,
      "epoch": 1.111866120065143,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004887482533577786,
      "loss": 5.3961,
      "mean_token_accuracy": 0.17663394808769226,
      "num_tokens": 22085301.0,
      "step": 9900
    },
    {
      "entropy": 5.489732456207276,
      "epoch": 1.1124276969731004,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004887357125732438,
      "loss": 5.3111,
      "mean_token_accuracy": 0.1892144724726677,
      "num_tokens": 22095635.0,
      "step": 9905
    },
    {
      "entropy": 5.530811166763305,
      "epoch": 1.112989273881058,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000488723164983281,
      "loss": 5.3203,
      "mean_token_accuracy": 0.18342317789793014,
      "num_tokens": 22105965.0,
      "step": 9910
    },
    {
      "entropy": 5.454964208602905,
      "epoch": 1.1135508507890155,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004887106105882898,
      "loss": 5.2603,
      "mean_token_accuracy": 0.1890107050538063,
      "num_tokens": 22117122.0,
      "step": 9915
    },
    {
      "entropy": 5.517485427856445,
      "epoch": 1.1141124276969732,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004886980493886699,
      "loss": 5.3218,
      "mean_token_accuracy": 0.1956054076552391,
      "num_tokens": 22128642.0,
      "step": 9920
    },
    {
      "entropy": 5.4373931884765625,
      "epoch": 1.1146740046049306,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004886854813848212,
      "loss": 5.3062,
      "mean_token_accuracy": 0.19211024045944214,
      "num_tokens": 22140260.0,
      "step": 9925
    },
    {
      "entropy": 5.460125350952149,
      "epoch": 1.1152355815128883,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004886729065771439,
      "loss": 5.4444,
      "mean_token_accuracy": 0.18107014298439025,
      "num_tokens": 22152460.0,
      "step": 9930
    },
    {
      "entropy": 5.46171817779541,
      "epoch": 1.1157971584208457,
      "grad_norm": 1.125,
      "learning_rate": 0.0004886603249660384,
      "loss": 5.1847,
      "mean_token_accuracy": 0.1943551480770111,
      "num_tokens": 22163134.0,
      "step": 9935
    },
    {
      "entropy": 5.476504564285278,
      "epoch": 1.1163587353288034,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004886477365519052,
      "loss": 5.3525,
      "mean_token_accuracy": 0.18250470012426376,
      "num_tokens": 22173956.0,
      "step": 9940
    },
    {
      "entropy": 5.436368322372436,
      "epoch": 1.1169203122367608,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004886351413351453,
      "loss": 5.3361,
      "mean_token_accuracy": 0.18535599261522293,
      "num_tokens": 22185506.0,
      "step": 9945
    },
    {
      "entropy": 5.500861310958863,
      "epoch": 1.1174818891447185,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004886225393161596,
      "loss": 5.3287,
      "mean_token_accuracy": 0.1871826171875,
      "num_tokens": 22196934.0,
      "step": 9950
    },
    {
      "entropy": 5.488682842254638,
      "epoch": 1.118043466052676,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004886099304953495,
      "loss": 5.3399,
      "mean_token_accuracy": 0.18429410308599473,
      "num_tokens": 22209013.0,
      "step": 9955
    },
    {
      "entropy": 5.521560192108154,
      "epoch": 1.1186050429606333,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004885973148731162,
      "loss": 5.4665,
      "mean_token_accuracy": 0.18359299153089523,
      "num_tokens": 22220890.0,
      "step": 9960
    },
    {
      "entropy": 5.544044351577758,
      "epoch": 1.119166619868591,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004885846924498617,
      "loss": 5.2899,
      "mean_token_accuracy": 0.19229473322629928,
      "num_tokens": 22232899.0,
      "step": 9965
    },
    {
      "entropy": 5.4252172946929935,
      "epoch": 1.1197281967765484,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004885720632259876,
      "loss": 5.2999,
      "mean_token_accuracy": 0.18870778381824493,
      "num_tokens": 22243839.0,
      "step": 9970
    },
    {
      "entropy": 5.540792465209961,
      "epoch": 1.120289773684506,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004885594272018963,
      "loss": 5.344,
      "mean_token_accuracy": 0.18891971409320832,
      "num_tokens": 22253970.0,
      "step": 9975
    },
    {
      "entropy": 5.405992794036865,
      "epoch": 1.1208513505924635,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048854678437799,
      "loss": 5.2236,
      "mean_token_accuracy": 0.19757343232631683,
      "num_tokens": 22263598.0,
      "step": 9980
    },
    {
      "entropy": 5.403408193588257,
      "epoch": 1.1214129275004212,
      "grad_norm": 1.125,
      "learning_rate": 0.0004885341347546713,
      "loss": 5.4363,
      "mean_token_accuracy": 0.1842990607023239,
      "num_tokens": 22274385.0,
      "step": 9985
    },
    {
      "entropy": 5.5041444301605225,
      "epoch": 1.1219745044083786,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004885214783323428,
      "loss": 5.2953,
      "mean_token_accuracy": 0.18946027904748916,
      "num_tokens": 22285837.0,
      "step": 9990
    },
    {
      "entropy": 5.566063785552979,
      "epoch": 1.1225360813163363,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004885088151114076,
      "loss": 5.368,
      "mean_token_accuracy": 0.18568748980760574,
      "num_tokens": 22296882.0,
      "step": 9995
    },
    {
      "entropy": 5.408493709564209,
      "epoch": 1.1230976582242937,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004884961450922691,
      "loss": 5.2945,
      "mean_token_accuracy": 0.190052030980587,
      "num_tokens": 22308110.0,
      "step": 10000
    },
    {
      "entropy": 5.438411045074463,
      "epoch": 1.1236592351322514,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004884834682753305,
      "loss": 5.3343,
      "mean_token_accuracy": 0.18975699841976165,
      "num_tokens": 22318541.0,
      "step": 10005
    },
    {
      "entropy": 5.473949146270752,
      "epoch": 1.1242208120402089,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004884707846609953,
      "loss": 5.3091,
      "mean_token_accuracy": 0.18497483432292938,
      "num_tokens": 22328713.0,
      "step": 10010
    },
    {
      "entropy": 5.556177711486816,
      "epoch": 1.1247823889481665,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004884580942496677,
      "loss": 5.3755,
      "mean_token_accuracy": 0.18221824169158934,
      "num_tokens": 22341050.0,
      "step": 10015
    },
    {
      "entropy": 5.4650740146636965,
      "epoch": 1.125343965856124,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004884453970417515,
      "loss": 5.2464,
      "mean_token_accuracy": 0.19214784651994704,
      "num_tokens": 22351184.0,
      "step": 10020
    },
    {
      "entropy": 5.4825136184692385,
      "epoch": 1.1259055427640816,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000488432693037651,
      "loss": 5.322,
      "mean_token_accuracy": 0.1871219679713249,
      "num_tokens": 22361651.0,
      "step": 10025
    },
    {
      "entropy": 5.573782873153687,
      "epoch": 1.126467119672039,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000488419982237771,
      "loss": 5.4251,
      "mean_token_accuracy": 0.1805936187505722,
      "num_tokens": 22373672.0,
      "step": 10030
    },
    {
      "entropy": 5.469729852676392,
      "epoch": 1.1270286965799967,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004884072646425157,
      "loss": 5.4057,
      "mean_token_accuracy": 0.18561234325170517,
      "num_tokens": 22384924.0,
      "step": 10035
    },
    {
      "entropy": 5.477024555206299,
      "epoch": 1.1275902734879542,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004883945402522907,
      "loss": 5.3169,
      "mean_token_accuracy": 0.18978175073862075,
      "num_tokens": 22396091.0,
      "step": 10040
    },
    {
      "entropy": 5.5944616317749025,
      "epoch": 1.1281518503959118,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004883818090675005,
      "loss": 5.3951,
      "mean_token_accuracy": 0.18209601938724518,
      "num_tokens": 22407925.0,
      "step": 10045
    },
    {
      "entropy": 5.491181802749634,
      "epoch": 1.1287134273038693,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004883690710885507,
      "loss": 5.3178,
      "mean_token_accuracy": 0.18729194700717927,
      "num_tokens": 22417917.0,
      "step": 10050
    },
    {
      "entropy": 5.533471775054932,
      "epoch": 1.1292750042118267,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004883563263158471,
      "loss": 5.3958,
      "mean_token_accuracy": 0.18572633415460588,
      "num_tokens": 22429836.0,
      "step": 10055
    },
    {
      "entropy": 5.589436340332031,
      "epoch": 1.1298365811197844,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004883435747497952,
      "loss": 5.492,
      "mean_token_accuracy": 0.1830290287733078,
      "num_tokens": 22440720.0,
      "step": 10060
    },
    {
      "entropy": 5.376164340972901,
      "epoch": 1.130398158027742,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004883308163908012,
      "loss": 5.2055,
      "mean_token_accuracy": 0.1916962146759033,
      "num_tokens": 22452587.0,
      "step": 10065
    },
    {
      "entropy": 5.45840744972229,
      "epoch": 1.1309597349356995,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004883180512392714,
      "loss": 5.3898,
      "mean_token_accuracy": 0.18569206595420837,
      "num_tokens": 22464063.0,
      "step": 10070
    },
    {
      "entropy": 5.478792190551758,
      "epoch": 1.131521311843657,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004883052792956119,
      "loss": 5.3238,
      "mean_token_accuracy": 0.18444016575813293,
      "num_tokens": 22474956.0,
      "step": 10075
    },
    {
      "entropy": 5.465976572036743,
      "epoch": 1.1320828887516146,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004882925005602296,
      "loss": 5.2978,
      "mean_token_accuracy": 0.18617224991321563,
      "num_tokens": 22486499.0,
      "step": 10080
    },
    {
      "entropy": 5.4245329856872555,
      "epoch": 1.132644465659572,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004882797150335314,
      "loss": 5.2204,
      "mean_token_accuracy": 0.1951652154326439,
      "num_tokens": 22497736.0,
      "step": 10085
    },
    {
      "entropy": 5.446517896652222,
      "epoch": 1.1332060425675297,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048826692271592434,
      "loss": 5.3196,
      "mean_token_accuracy": 0.1885177344083786,
      "num_tokens": 22509997.0,
      "step": 10090
    },
    {
      "entropy": 5.46684889793396,
      "epoch": 1.133767619475487,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004882541236078158,
      "loss": 5.3991,
      "mean_token_accuracy": 0.1880730375647545,
      "num_tokens": 22519721.0,
      "step": 10095
    },
    {
      "entropy": 5.532308721542359,
      "epoch": 1.1343291963834448,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048824131770961316,
      "loss": 5.2952,
      "mean_token_accuracy": 0.18371496796607972,
      "num_tokens": 22530219.0,
      "step": 10100
    },
    {
      "entropy": 5.474745416641236,
      "epoch": 1.1348907732914022,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048822850502172437,
      "loss": 5.3238,
      "mean_token_accuracy": 0.18468173891305922,
      "num_tokens": 22541373.0,
      "step": 10105
    },
    {
      "entropy": 5.352654027938843,
      "epoch": 1.1354523501993599,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004882156855445571,
      "loss": 5.2473,
      "mean_token_accuracy": 0.18527161777019502,
      "num_tokens": 22552974.0,
      "step": 10110
    },
    {
      "entropy": 5.49869933128357,
      "epoch": 1.1360139271073173,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004882028592785199,
      "loss": 5.2763,
      "mean_token_accuracy": 0.1881103277206421,
      "num_tokens": 22563468.0,
      "step": 10115
    },
    {
      "entropy": 5.485534906387329,
      "epoch": 1.136575504015275,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048819002622402095,
      "loss": 5.3997,
      "mean_token_accuracy": 0.18482900857925416,
      "num_tokens": 22574386.0,
      "step": 10120
    },
    {
      "entropy": 5.479231309890747,
      "epoch": 1.1371370809232324,
      "grad_norm": 1.125,
      "learning_rate": 0.0004881771863814688,
      "loss": 5.3075,
      "mean_token_accuracy": 0.19373198896646499,
      "num_tokens": 22585369.0,
      "step": 10125
    },
    {
      "entropy": 5.411350870132447,
      "epoch": 1.13769865783119,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048816433975127244,
      "loss": 5.3188,
      "mean_token_accuracy": 0.1893681988120079,
      "num_tokens": 22595853.0,
      "step": 10130
    },
    {
      "entropy": 5.542398595809937,
      "epoch": 1.1382602347391475,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004881514863338407,
      "loss": 5.3878,
      "mean_token_accuracy": 0.1882244274020195,
      "num_tokens": 22606596.0,
      "step": 10135
    },
    {
      "entropy": 5.5100569248199465,
      "epoch": 1.1388218116471052,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048813862612958307,
      "loss": 5.2772,
      "mean_token_accuracy": 0.18693946450948715,
      "num_tokens": 22617196.0,
      "step": 10140
    },
    {
      "entropy": 5.348569011688232,
      "epoch": 1.1393833885550626,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004881257591389089,
      "loss": 5.2365,
      "mean_token_accuracy": 0.19013486802577972,
      "num_tokens": 22627817.0,
      "step": 10145
    },
    {
      "entropy": 5.462574577331543,
      "epoch": 1.13994496546302,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048811288536222794,
      "loss": 5.3335,
      "mean_token_accuracy": 0.18206507414579393,
      "num_tokens": 22637877.0,
      "step": 10150
    },
    {
      "entropy": 5.578246212005615,
      "epoch": 1.1405065423709777,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048810000479995,
      "loss": 5.3662,
      "mean_token_accuracy": 0.18668453246355057,
      "num_tokens": 22648770.0,
      "step": 10155
    },
    {
      "entropy": 5.616941642761231,
      "epoch": 1.1410681192789354,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048808711745248524,
      "loss": 5.419,
      "mean_token_accuracy": 0.17958504259586333,
      "num_tokens": 22659626.0,
      "step": 10160
    },
    {
      "entropy": 5.511206769943238,
      "epoch": 1.1416296961868928,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048807422332024404,
      "loss": 5.4051,
      "mean_token_accuracy": 0.18367373049259186,
      "num_tokens": 22669035.0,
      "step": 10165
    },
    {
      "entropy": 5.510091018676758,
      "epoch": 1.1421912730948502,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004880613224036369,
      "loss": 5.3833,
      "mean_token_accuracy": 0.18738051801919936,
      "num_tokens": 22680536.0,
      "step": 10170
    },
    {
      "entropy": 5.533512258529663,
      "epoch": 1.142752850002808,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048804841470307466,
      "loss": 5.3251,
      "mean_token_accuracy": 0.19318681359291076,
      "num_tokens": 22692732.0,
      "step": 10175
    },
    {
      "entropy": 5.548362588882446,
      "epoch": 1.1433144269107653,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048803550021896824,
      "loss": 5.4657,
      "mean_token_accuracy": 0.18072652220726013,
      "num_tokens": 22703677.0,
      "step": 10180
    },
    {
      "entropy": 5.545753860473633,
      "epoch": 1.143876003818723,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004880225789517289,
      "loss": 5.3779,
      "mean_token_accuracy": 0.18104286640882492,
      "num_tokens": 22714826.0,
      "step": 10185
    },
    {
      "entropy": 5.51541166305542,
      "epoch": 1.1444375807266804,
      "grad_norm": 1.125,
      "learning_rate": 0.000488009650901768,
      "loss": 5.445,
      "mean_token_accuracy": 0.1825494334101677,
      "num_tokens": 22726013.0,
      "step": 10190
    },
    {
      "entropy": 5.501227807998657,
      "epoch": 1.144999157634638,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048799671606949724,
      "loss": 5.2876,
      "mean_token_accuracy": 0.18399654924869538,
      "num_tokens": 22737494.0,
      "step": 10195
    },
    {
      "entropy": 5.468310308456421,
      "epoch": 1.1455607345425955,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048798377445532837,
      "loss": 5.2285,
      "mean_token_accuracy": 0.1922479972243309,
      "num_tokens": 22746916.0,
      "step": 10200
    },
    {
      "entropy": 5.513650226593017,
      "epoch": 1.1461223114505532,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004879708260596736,
      "loss": 5.443,
      "mean_token_accuracy": 0.17282115668058395,
      "num_tokens": 22759197.0,
      "step": 10205
    },
    {
      "entropy": 5.573303174972534,
      "epoch": 1.1466838883585107,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004879578708829451,
      "loss": 5.41,
      "mean_token_accuracy": 0.1833725869655609,
      "num_tokens": 22770744.0,
      "step": 10210
    },
    {
      "entropy": 5.481962871551514,
      "epoch": 1.1472454652664683,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004879449089255553,
      "loss": 5.3291,
      "mean_token_accuracy": 0.1838076815009117,
      "num_tokens": 22781989.0,
      "step": 10215
    },
    {
      "entropy": 5.403723478317261,
      "epoch": 1.1478070421744258,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004879319401879171,
      "loss": 5.2456,
      "mean_token_accuracy": 0.1970457911491394,
      "num_tokens": 22792972.0,
      "step": 10220
    },
    {
      "entropy": 5.459436893463135,
      "epoch": 1.1483686190823834,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048791896467044335,
      "loss": 5.337,
      "mean_token_accuracy": 0.19282135963439942,
      "num_tokens": 22802623.0,
      "step": 10225
    },
    {
      "entropy": 5.3735795497894285,
      "epoch": 1.1489301959903409,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048790598237354716,
      "loss": 5.2622,
      "mean_token_accuracy": 0.19431663006544114,
      "num_tokens": 22813564.0,
      "step": 10230
    },
    {
      "entropy": 5.4960364818573,
      "epoch": 1.1494917728982985,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004878929932976419,
      "loss": 5.4233,
      "mean_token_accuracy": 0.18893394619226456,
      "num_tokens": 22824681.0,
      "step": 10235
    },
    {
      "entropy": 5.425031280517578,
      "epoch": 1.150053349806256,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004878799974431412,
      "loss": 5.251,
      "mean_token_accuracy": 0.19384222030639647,
      "num_tokens": 22836030.0,
      "step": 10240
    },
    {
      "entropy": 5.455839061737061,
      "epoch": 1.1506149267142134,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004878669948104588,
      "loss": 5.2532,
      "mean_token_accuracy": 0.1923241898417473,
      "num_tokens": 22847213.0,
      "step": 10245
    },
    {
      "entropy": 5.487225198745728,
      "epoch": 1.151176503622171,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048785398540000867,
      "loss": 5.2681,
      "mean_token_accuracy": 0.19509278237819672,
      "num_tokens": 22857904.0,
      "step": 10250
    },
    {
      "entropy": 5.466386175155639,
      "epoch": 1.1517380805301287,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048784096921220514,
      "loss": 5.4076,
      "mean_token_accuracy": 0.18061592876911164,
      "num_tokens": 22868975.0,
      "step": 10255
    },
    {
      "entropy": 5.570178365707397,
      "epoch": 1.1522996574380862,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004878279462474627,
      "loss": 5.3719,
      "mean_token_accuracy": 0.18787189722061157,
      "num_tokens": 22879463.0,
      "step": 10260
    },
    {
      "entropy": 5.52513861656189,
      "epoch": 1.1528612343460436,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048781491650619574,
      "loss": 5.2885,
      "mean_token_accuracy": 0.18793953210115433,
      "num_tokens": 22890382.0,
      "step": 10265
    },
    {
      "entropy": 5.475088691711425,
      "epoch": 1.1534228112540013,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048780187998881944,
      "loss": 5.3808,
      "mean_token_accuracy": 0.1865217462182045,
      "num_tokens": 22901968.0,
      "step": 10270
    },
    {
      "entropy": 5.574156856536865,
      "epoch": 1.1539843881619587,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004877888366957486,
      "loss": 5.4153,
      "mean_token_accuracy": 0.18759825229644775,
      "num_tokens": 22912147.0,
      "step": 10275
    },
    {
      "entropy": 5.4690141677856445,
      "epoch": 1.1545459650699164,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004877757866273988,
      "loss": 5.3231,
      "mean_token_accuracy": 0.18450030982494353,
      "num_tokens": 22923744.0,
      "step": 10280
    },
    {
      "entropy": 5.489623594284057,
      "epoch": 1.1551075419778738,
      "grad_norm": 1.25,
      "learning_rate": 0.0004877627297841853,
      "loss": 5.2723,
      "mean_token_accuracy": 0.18999790996313096,
      "num_tokens": 22935487.0,
      "step": 10285
    },
    {
      "entropy": 5.422398471832276,
      "epoch": 1.1556691188858315,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004877496661665241,
      "loss": 5.3156,
      "mean_token_accuracy": 0.19071489125490187,
      "num_tokens": 22946379.0,
      "step": 10290
    },
    {
      "entropy": 5.45652403831482,
      "epoch": 1.156230695793789,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000487736595774831,
      "loss": 5.3279,
      "mean_token_accuracy": 0.18550100773572922,
      "num_tokens": 22957219.0,
      "step": 10295
    },
    {
      "entropy": 5.4492370128631595,
      "epoch": 1.1567922727017466,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048772351860952215,
      "loss": 5.2743,
      "mean_token_accuracy": 0.19157078266143798,
      "num_tokens": 22967749.0,
      "step": 10300
    },
    {
      "entropy": 5.46882472038269,
      "epoch": 1.157353849609704,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004877104346710139,
      "loss": 5.2852,
      "mean_token_accuracy": 0.1868071138858795,
      "num_tokens": 22978317.0,
      "step": 10305
    },
    {
      "entropy": 5.4460334300994875,
      "epoch": 1.1579154265176617,
      "grad_norm": 1.265625,
      "learning_rate": 0.000487697343959723,
      "loss": 5.3263,
      "mean_token_accuracy": 0.18757355958223343,
      "num_tokens": 22988793.0,
      "step": 10310
    },
    {
      "entropy": 5.50702428817749,
      "epoch": 1.158477003425619,
      "grad_norm": 1.375,
      "learning_rate": 0.0004876842464760661,
      "loss": 5.3834,
      "mean_token_accuracy": 0.1833086222410202,
      "num_tokens": 23000073.0,
      "step": 10315
    },
    {
      "entropy": 5.544207620620727,
      "epoch": 1.1590385803335768,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048767114222046036,
      "loss": 5.3474,
      "mean_token_accuracy": 0.18358590006828307,
      "num_tokens": 23011260.0,
      "step": 10320
    },
    {
      "entropy": 5.469492626190186,
      "epoch": 1.1596001572415342,
      "grad_norm": 1.28125,
      "learning_rate": 0.000487658031193323,
      "loss": 5.3227,
      "mean_token_accuracy": 0.186710387468338,
      "num_tokens": 23022129.0,
      "step": 10325
    },
    {
      "entropy": 5.41064338684082,
      "epoch": 1.1601617341494919,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004876449133950714,
      "loss": 5.2791,
      "mean_token_accuracy": 0.1852422446012497,
      "num_tokens": 23033666.0,
      "step": 10330
    },
    {
      "entropy": 5.434285593032837,
      "epoch": 1.1607233110574493,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004876317888261233,
      "loss": 5.2874,
      "mean_token_accuracy": 0.1917390838265419,
      "num_tokens": 23044133.0,
      "step": 10335
    },
    {
      "entropy": 5.4549706935882565,
      "epoch": 1.1612848879654067,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048761865748689657,
      "loss": 5.3024,
      "mean_token_accuracy": 0.18544703274965285,
      "num_tokens": 23055877.0,
      "step": 10340
    },
    {
      "entropy": 5.548912143707275,
      "epoch": 1.1618464648733644,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048760551937780927,
      "loss": 5.4236,
      "mean_token_accuracy": 0.1831132650375366,
      "num_tokens": 23066811.0,
      "step": 10345
    },
    {
      "entropy": 5.530440139770508,
      "epoch": 1.162408041781322,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048759237449927986,
      "loss": 5.3596,
      "mean_token_accuracy": 0.18338708877563475,
      "num_tokens": 23078928.0,
      "step": 10350
    },
    {
      "entropy": 5.440915250778199,
      "epoch": 1.1629696186892795,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048757922285172674,
      "loss": 5.3443,
      "mean_token_accuracy": 0.18304735869169236,
      "num_tokens": 23090685.0,
      "step": 10355
    },
    {
      "entropy": 5.414602136611938,
      "epoch": 1.163531195597237,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048756606443556874,
      "loss": 5.2718,
      "mean_token_accuracy": 0.19179436564445496,
      "num_tokens": 23101659.0,
      "step": 10360
    },
    {
      "entropy": 5.438999938964844,
      "epoch": 1.1640927725051946,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048755289925122474,
      "loss": 5.3422,
      "mean_token_accuracy": 0.18727674782276155,
      "num_tokens": 23112970.0,
      "step": 10365
    },
    {
      "entropy": 5.495529222488403,
      "epoch": 1.164654349413152,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004875397272991141,
      "loss": 5.3602,
      "mean_token_accuracy": 0.18496147394180298,
      "num_tokens": 23125391.0,
      "step": 10370
    },
    {
      "entropy": 5.553882884979248,
      "epoch": 1.1652159263211097,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000487526548579656,
      "loss": 5.4008,
      "mean_token_accuracy": 0.18643703162670136,
      "num_tokens": 23138195.0,
      "step": 10375
    },
    {
      "entropy": 5.55526180267334,
      "epoch": 1.1657775032290671,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048751336309327026,
      "loss": 5.3912,
      "mean_token_accuracy": 0.184760682284832,
      "num_tokens": 23149378.0,
      "step": 10380
    },
    {
      "entropy": 5.383782052993775,
      "epoch": 1.1663390801370248,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004875001708403766,
      "loss": 5.2323,
      "mean_token_accuracy": 0.19789338111877441,
      "num_tokens": 23159394.0,
      "step": 10385
    },
    {
      "entropy": 5.5619439601898195,
      "epoch": 1.1669006570449822,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048748697182139503,
      "loss": 5.4636,
      "mean_token_accuracy": 0.18453750312328338,
      "num_tokens": 23171512.0,
      "step": 10390
    },
    {
      "entropy": 5.436984252929688,
      "epoch": 1.16746223395294,
      "grad_norm": 1.125,
      "learning_rate": 0.0004874737660367459,
      "loss": 5.3089,
      "mean_token_accuracy": 0.18689813911914827,
      "num_tokens": 23181635.0,
      "step": 10395
    },
    {
      "entropy": 5.587057828903198,
      "epoch": 1.1680238108608973,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048746055348684967,
      "loss": 5.5656,
      "mean_token_accuracy": 0.17790596634149553,
      "num_tokens": 23194510.0,
      "step": 10400
    },
    {
      "entropy": 5.6394671440124515,
      "epoch": 1.168585387768855,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000487447334172127,
      "loss": 5.4294,
      "mean_token_accuracy": 0.1896474540233612,
      "num_tokens": 23205146.0,
      "step": 10405
    },
    {
      "entropy": 5.449761629104614,
      "epoch": 1.1691469646768125,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048743410809299877,
      "loss": 5.3379,
      "mean_token_accuracy": 0.17951782643795014,
      "num_tokens": 23216621.0,
      "step": 10410
    },
    {
      "entropy": 5.443548965454101,
      "epoch": 1.1697085415847701,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004874208752498862,
      "loss": 5.3109,
      "mean_token_accuracy": 0.18659743219614028,
      "num_tokens": 23226401.0,
      "step": 10415
    },
    {
      "entropy": 5.527456617355346,
      "epoch": 1.1702701184927276,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004874076356432106,
      "loss": 5.3454,
      "mean_token_accuracy": 0.17964297086000441,
      "num_tokens": 23237921.0,
      "step": 10420
    },
    {
      "entropy": 5.513091087341309,
      "epoch": 1.1708316954006852,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048739438927339346,
      "loss": 5.3079,
      "mean_token_accuracy": 0.18994258791208268,
      "num_tokens": 23250015.0,
      "step": 10425
    },
    {
      "entropy": 5.496845388412476,
      "epoch": 1.1713932723086427,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048738113614085656,
      "loss": 5.3496,
      "mean_token_accuracy": 0.1837267830967903,
      "num_tokens": 23261243.0,
      "step": 10430
    },
    {
      "entropy": 5.540515804290772,
      "epoch": 1.1719548492166,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000487367876246022,
      "loss": 5.5015,
      "mean_token_accuracy": 0.1787189483642578,
      "num_tokens": 23273264.0,
      "step": 10435
    },
    {
      "entropy": 5.654577732086182,
      "epoch": 1.1725164261245578,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004873546095893118,
      "loss": 5.5113,
      "mean_token_accuracy": 0.1793489396572113,
      "num_tokens": 23285031.0,
      "step": 10440
    },
    {
      "entropy": 5.485291433334351,
      "epoch": 1.1730780030325154,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004873413361711485,
      "loss": 5.2816,
      "mean_token_accuracy": 0.19496330320835115,
      "num_tokens": 23296387.0,
      "step": 10445
    },
    {
      "entropy": 5.509927701950073,
      "epoch": 1.1736395799404729,
      "grad_norm": 1.125,
      "learning_rate": 0.00048732805599195465,
      "loss": 5.3672,
      "mean_token_accuracy": 0.18849920332431794,
      "num_tokens": 23307950.0,
      "step": 10450
    },
    {
      "entropy": 5.3654686450958256,
      "epoch": 1.1742011568484303,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048731476905215326,
      "loss": 5.1948,
      "mean_token_accuracy": 0.19201835691928865,
      "num_tokens": 23318424.0,
      "step": 10455
    },
    {
      "entropy": 5.494434976577759,
      "epoch": 1.174762733756388,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004873014753521673,
      "loss": 5.4667,
      "mean_token_accuracy": 0.1801562041044235,
      "num_tokens": 23330186.0,
      "step": 10460
    },
    {
      "entropy": 5.554242563247681,
      "epoch": 1.1753243106643454,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004872881748924199,
      "loss": 5.2258,
      "mean_token_accuracy": 0.19798149019479752,
      "num_tokens": 23341878.0,
      "step": 10465
    },
    {
      "entropy": 5.51068754196167,
      "epoch": 1.175885887572303,
      "grad_norm": 1.125,
      "learning_rate": 0.0004872748676733347,
      "loss": 5.3702,
      "mean_token_accuracy": 0.1835189864039421,
      "num_tokens": 23352471.0,
      "step": 10470
    },
    {
      "entropy": 5.487906217575073,
      "epoch": 1.1764474644802605,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048726155369533547,
      "loss": 5.3759,
      "mean_token_accuracy": 0.1909515604376793,
      "num_tokens": 23363871.0,
      "step": 10475
    },
    {
      "entropy": 5.560474634170532,
      "epoch": 1.1770090413882182,
      "grad_norm": 1.171875,
      "learning_rate": 0.000487248232958846,
      "loss": 5.3864,
      "mean_token_accuracy": 0.1841189295053482,
      "num_tokens": 23376072.0,
      "step": 10480
    },
    {
      "entropy": 5.345824718475342,
      "epoch": 1.1775706182961756,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004872349054642904,
      "loss": 5.2718,
      "mean_token_accuracy": 0.1938830569386482,
      "num_tokens": 23386670.0,
      "step": 10485
    },
    {
      "entropy": 5.499839448928833,
      "epoch": 1.1781321952041333,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004872215712120932,
      "loss": 5.3198,
      "mean_token_accuracy": 0.1952613562345505,
      "num_tokens": 23396499.0,
      "step": 10490
    },
    {
      "entropy": 5.505800008773804,
      "epoch": 1.1786937721120907,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004872082302026788,
      "loss": 5.3469,
      "mean_token_accuracy": 0.19099795818328857,
      "num_tokens": 23407898.0,
      "step": 10495
    },
    {
      "entropy": 5.425271320343017,
      "epoch": 1.1792553490200484,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004871948824364721,
      "loss": 5.3466,
      "mean_token_accuracy": 0.18085567355155946,
      "num_tokens": 23418877.0,
      "step": 10500
    },
    {
      "entropy": 5.465935897827149,
      "epoch": 1.1798169259280058,
      "grad_norm": 1.125,
      "learning_rate": 0.000487181527913898,
      "loss": 5.3039,
      "mean_token_accuracy": 0.18846290409564972,
      "num_tokens": 23431378.0,
      "step": 10505
    },
    {
      "entropy": 5.508903455734253,
      "epoch": 1.1803785028359635,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048716816663538177,
      "loss": 5.3154,
      "mean_token_accuracy": 0.1882656916975975,
      "num_tokens": 23442590.0,
      "step": 10510
    },
    {
      "entropy": 5.5088136196136475,
      "epoch": 1.180940079743921,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004871547986013488,
      "loss": 5.4007,
      "mean_token_accuracy": 0.18411462903022766,
      "num_tokens": 23454139.0,
      "step": 10515
    },
    {
      "entropy": 5.466353845596314,
      "epoch": 1.1815016566518786,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004871414238122248,
      "loss": 5.3368,
      "mean_token_accuracy": 0.19195101410150528,
      "num_tokens": 23464908.0,
      "step": 10520
    },
    {
      "entropy": 5.404631090164185,
      "epoch": 1.182063233559836,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004871280422684355,
      "loss": 5.2904,
      "mean_token_accuracy": 0.19446847289800645,
      "num_tokens": 23475436.0,
      "step": 10525
    },
    {
      "entropy": 5.562547636032105,
      "epoch": 1.1826248104677934,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048711465397040726,
      "loss": 5.4479,
      "mean_token_accuracy": 0.18536293804645537,
      "num_tokens": 23487384.0,
      "step": 10530
    },
    {
      "entropy": 5.597505283355713,
      "epoch": 1.183186387375751,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048710125891856604,
      "loss": 5.3958,
      "mean_token_accuracy": 0.1824021339416504,
      "num_tokens": 23498978.0,
      "step": 10535
    },
    {
      "entropy": 5.4043584823608395,
      "epoch": 1.1837479642837088,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048708785711333845,
      "loss": 5.2793,
      "mean_token_accuracy": 0.19243059009313584,
      "num_tokens": 23510825.0,
      "step": 10540
    },
    {
      "entropy": 5.378804779052734,
      "epoch": 1.1843095411916662,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048707444855515124,
      "loss": 5.2891,
      "mean_token_accuracy": 0.19109508246183396,
      "num_tokens": 23522383.0,
      "step": 10545
    },
    {
      "entropy": 5.5603917121887205,
      "epoch": 1.1848711180996236,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004870610332444314,
      "loss": 5.4606,
      "mean_token_accuracy": 0.18625949770212175,
      "num_tokens": 23534255.0,
      "step": 10550
    },
    {
      "entropy": 5.5891139030456545,
      "epoch": 1.1854326950075813,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004870476111816059,
      "loss": 5.3759,
      "mean_token_accuracy": 0.1831232264637947,
      "num_tokens": 23545888.0,
      "step": 10555
    },
    {
      "entropy": 5.438169002532959,
      "epoch": 1.1859942719155387,
      "grad_norm": 1.25,
      "learning_rate": 0.00048703418236710234,
      "loss": 5.3719,
      "mean_token_accuracy": 0.1819407522678375,
      "num_tokens": 23556776.0,
      "step": 10560
    },
    {
      "entropy": 5.555518674850464,
      "epoch": 1.1865558488234964,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004870207468013481,
      "loss": 5.2951,
      "mean_token_accuracy": 0.1916578531265259,
      "num_tokens": 23569971.0,
      "step": 10565
    },
    {
      "entropy": 5.527983379364014,
      "epoch": 1.1871174257314538,
      "grad_norm": 1.375,
      "learning_rate": 0.0004870073044847711,
      "loss": 5.4254,
      "mean_token_accuracy": 0.18327248841524124,
      "num_tokens": 23580751.0,
      "step": 10570
    },
    {
      "entropy": 5.431289482116699,
      "epoch": 1.1876790026394115,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004869938554177993,
      "loss": 5.2801,
      "mean_token_accuracy": 0.19805623590946198,
      "num_tokens": 23591423.0,
      "step": 10575
    },
    {
      "entropy": 5.491920852661133,
      "epoch": 1.188240579547369,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048698039960086084,
      "loss": 5.2558,
      "mean_token_accuracy": 0.18322913646697997,
      "num_tokens": 23601552.0,
      "step": 10580
    },
    {
      "entropy": 5.411893844604492,
      "epoch": 1.1888021564553266,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004869669370343843,
      "loss": 5.3691,
      "mean_token_accuracy": 0.1931927040219307,
      "num_tokens": 23613080.0,
      "step": 10585
    },
    {
      "entropy": 5.484119415283203,
      "epoch": 1.189363733363284,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004869534677187983,
      "loss": 5.259,
      "mean_token_accuracy": 0.19166787564754487,
      "num_tokens": 23623559.0,
      "step": 10590
    },
    {
      "entropy": 5.444127368927002,
      "epoch": 1.1899253102712417,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004869399916545317,
      "loss": 5.2624,
      "mean_token_accuracy": 0.19548728466033935,
      "num_tokens": 23634113.0,
      "step": 10595
    },
    {
      "entropy": 5.420895195007324,
      "epoch": 1.1904868871791991,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004869265088420135,
      "loss": 5.1833,
      "mean_token_accuracy": 0.19952433854341506,
      "num_tokens": 23645203.0,
      "step": 10600
    },
    {
      "entropy": 5.406381273269654,
      "epoch": 1.1910484640871568,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004869130192816731,
      "loss": 5.3143,
      "mean_token_accuracy": 0.19205373227596284,
      "num_tokens": 23654880.0,
      "step": 10605
    },
    {
      "entropy": 5.403260803222656,
      "epoch": 1.1916100409951143,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048689952297394,
      "loss": 5.2323,
      "mean_token_accuracy": 0.19772869497537612,
      "num_tokens": 23664962.0,
      "step": 10610
    },
    {
      "entropy": 5.4483644485473635,
      "epoch": 1.192171617903072,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004868860199192438,
      "loss": 5.2683,
      "mean_token_accuracy": 0.19701651334762574,
      "num_tokens": 23675288.0,
      "step": 10615
    },
    {
      "entropy": 5.4260498046875,
      "epoch": 1.1927331948110294,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004868725101180146,
      "loss": 5.2826,
      "mean_token_accuracy": 0.18938666135072707,
      "num_tokens": 23685948.0,
      "step": 10620
    },
    {
      "entropy": 5.537334775924682,
      "epoch": 1.1932947717189868,
      "grad_norm": 1.25,
      "learning_rate": 0.0004868589935706826,
      "loss": 5.3535,
      "mean_token_accuracy": 0.18100430071353912,
      "num_tokens": 23696980.0,
      "step": 10625
    },
    {
      "entropy": 5.538110160827637,
      "epoch": 1.1938563486269445,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048684547027767793,
      "loss": 5.3791,
      "mean_token_accuracy": 0.18451775908470153,
      "num_tokens": 23708644.0,
      "step": 10630
    },
    {
      "entropy": 5.464564609527588,
      "epoch": 1.1944179255349021,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004868319402394314,
      "loss": 5.3662,
      "mean_token_accuracy": 0.18750886619091034,
      "num_tokens": 23719652.0,
      "step": 10635
    },
    {
      "entropy": 5.4488729476928714,
      "epoch": 1.1949795024428596,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048681840345637375,
      "loss": 5.3276,
      "mean_token_accuracy": 0.18725836724042894,
      "num_tokens": 23731906.0,
      "step": 10640
    },
    {
      "entropy": 5.499956798553467,
      "epoch": 1.195541079350817,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048680485992893597,
      "loss": 5.385,
      "mean_token_accuracy": 0.18707500398159027,
      "num_tokens": 23744184.0,
      "step": 10645
    },
    {
      "entropy": 5.488353252410889,
      "epoch": 1.1961026562587747,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00048679130965754926,
      "loss": 5.1886,
      "mean_token_accuracy": 0.19169432073831558,
      "num_tokens": 23754807.0,
      "step": 10650
    },
    {
      "entropy": 5.382348537445068,
      "epoch": 1.196664233166732,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004867777526426451,
      "loss": 5.2685,
      "mean_token_accuracy": 0.18901412785053254,
      "num_tokens": 23765907.0,
      "step": 10655
    },
    {
      "entropy": 5.437325525283813,
      "epoch": 1.1972258100746898,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048676418888465525,
      "loss": 5.3214,
      "mean_token_accuracy": 0.19033824652433395,
      "num_tokens": 23776217.0,
      "step": 10660
    },
    {
      "entropy": 5.4633416652679445,
      "epoch": 1.1977873869826472,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004867506183840114,
      "loss": 5.26,
      "mean_token_accuracy": 0.19328122287988664,
      "num_tokens": 23786170.0,
      "step": 10665
    },
    {
      "entropy": 5.544186687469482,
      "epoch": 1.1983489638906049,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004867370411411459,
      "loss": 5.392,
      "mean_token_accuracy": 0.1785945013165474,
      "num_tokens": 23796491.0,
      "step": 10670
    },
    {
      "entropy": 5.4780525207519535,
      "epoch": 1.1989105407985623,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048672345715649066,
      "loss": 5.2648,
      "mean_token_accuracy": 0.1928586021065712,
      "num_tokens": 23807718.0,
      "step": 10675
    },
    {
      "entropy": 5.431114244461059,
      "epoch": 1.19947211770652,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004867098664304785,
      "loss": 5.4077,
      "mean_token_accuracy": 0.18344444632530213,
      "num_tokens": 23818623.0,
      "step": 10680
    },
    {
      "entropy": 5.425208902359008,
      "epoch": 1.2000336946144774,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004866962689635422,
      "loss": 5.3307,
      "mean_token_accuracy": 0.19101500362157822,
      "num_tokens": 23829173.0,
      "step": 10685
    },
    {
      "entropy": 5.467258739471435,
      "epoch": 1.200595271522435,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048668266475611447,
      "loss": 5.2364,
      "mean_token_accuracy": 0.18817202299833297,
      "num_tokens": 23840689.0,
      "step": 10690
    },
    {
      "entropy": 5.571165895462036,
      "epoch": 1.2011568484303925,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004866690538086286,
      "loss": 5.4641,
      "mean_token_accuracy": 0.18146587610244752,
      "num_tokens": 23852858.0,
      "step": 10695
    },
    {
      "entropy": 5.4639472484588625,
      "epoch": 1.2017184253383502,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048665543612151803,
      "loss": 5.2788,
      "mean_token_accuracy": 0.19116113781929017,
      "num_tokens": 23864137.0,
      "step": 10700
    },
    {
      "entropy": 5.458691787719727,
      "epoch": 1.2022800022463076,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048664181169521625,
      "loss": 5.3335,
      "mean_token_accuracy": 0.18668832778930664,
      "num_tokens": 23875087.0,
      "step": 10705
    },
    {
      "entropy": 5.405934143066406,
      "epoch": 1.2028415791542653,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004866281805301571,
      "loss": 5.2063,
      "mean_token_accuracy": 0.19864078611135483,
      "num_tokens": 23886521.0,
      "step": 10710
    },
    {
      "entropy": 5.444202518463134,
      "epoch": 1.2034031560622227,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004866145426267746,
      "loss": 5.3249,
      "mean_token_accuracy": 0.18816700726747512,
      "num_tokens": 23897660.0,
      "step": 10715
    },
    {
      "entropy": 5.492811441421509,
      "epoch": 1.2039647329701804,
      "grad_norm": 1.140625,
      "learning_rate": 0.000486600897985503,
      "loss": 5.3609,
      "mean_token_accuracy": 0.18554053753614425,
      "num_tokens": 23910543.0,
      "step": 10720
    },
    {
      "entropy": 5.471459007263183,
      "epoch": 1.2045263098781378,
      "grad_norm": 1.234375,
      "learning_rate": 0.00048658724660677674,
      "loss": 5.236,
      "mean_token_accuracy": 0.19174371659755707,
      "num_tokens": 23921687.0,
      "step": 10725
    },
    {
      "entropy": 5.421480751037597,
      "epoch": 1.2050878867860955,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004865735884910305,
      "loss": 5.3474,
      "mean_token_accuracy": 0.18721396327018738,
      "num_tokens": 23933717.0,
      "step": 10730
    },
    {
      "entropy": 5.444752931594849,
      "epoch": 1.205649463694053,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004865599236386991,
      "loss": 5.3934,
      "mean_token_accuracy": 0.18448521792888642,
      "num_tokens": 23947764.0,
      "step": 10735
    },
    {
      "entropy": 5.503841304779053,
      "epoch": 1.2062110406020103,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004865462520502177,
      "loss": 5.277,
      "mean_token_accuracy": 0.18383033126592635,
      "num_tokens": 23960096.0,
      "step": 10740
    },
    {
      "entropy": 5.507016801834107,
      "epoch": 1.206772617509968,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004865325737260216,
      "loss": 5.2837,
      "mean_token_accuracy": 0.1872893303632736,
      "num_tokens": 23971573.0,
      "step": 10745
    },
    {
      "entropy": 5.4688958644866945,
      "epoch": 1.2073341944179254,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048651888866654624,
      "loss": 5.256,
      "mean_token_accuracy": 0.1969495877623558,
      "num_tokens": 23982710.0,
      "step": 10750
    },
    {
      "entropy": 5.341468286514282,
      "epoch": 1.207895771325883,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004865051968722275,
      "loss": 5.2078,
      "mean_token_accuracy": 0.19535233080387115,
      "num_tokens": 23993663.0,
      "step": 10755
    },
    {
      "entropy": 5.4898618221282955,
      "epoch": 1.2084573482338405,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048649149834350126,
      "loss": 5.467,
      "mean_token_accuracy": 0.18701529502868652,
      "num_tokens": 24005112.0,
      "step": 10760
    },
    {
      "entropy": 5.420898962020874,
      "epoch": 1.2090189251417982,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004864777930808037,
      "loss": 5.2127,
      "mean_token_accuracy": 0.19623795598745347,
      "num_tokens": 24015855.0,
      "step": 10765
    },
    {
      "entropy": 5.38665885925293,
      "epoch": 1.2095805020497556,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004864640810845711,
      "loss": 5.2542,
      "mean_token_accuracy": 0.19129853397607804,
      "num_tokens": 24027232.0,
      "step": 10770
    },
    {
      "entropy": 5.469411563873291,
      "epoch": 1.2101420789577133,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004864503623552402,
      "loss": 5.2422,
      "mean_token_accuracy": 0.1964159667491913,
      "num_tokens": 24039127.0,
      "step": 10775
    },
    {
      "entropy": 5.548190975189209,
      "epoch": 1.2107036558656707,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004864366368932478,
      "loss": 5.3942,
      "mean_token_accuracy": 0.18704016655683517,
      "num_tokens": 24049821.0,
      "step": 10780
    },
    {
      "entropy": 5.488651990890503,
      "epoch": 1.2112652327736284,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048642290469903083,
      "loss": 5.4235,
      "mean_token_accuracy": 0.18653962016105652,
      "num_tokens": 24060305.0,
      "step": 10785
    },
    {
      "entropy": 5.480674076080322,
      "epoch": 1.2118268096815858,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048640916577302664,
      "loss": 5.2757,
      "mean_token_accuracy": 0.18868815451860427,
      "num_tokens": 24071193.0,
      "step": 10790
    },
    {
      "entropy": 5.371815156936646,
      "epoch": 1.2123883865895435,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048639542011567243,
      "loss": 5.1486,
      "mean_token_accuracy": 0.21288779526948928,
      "num_tokens": 24082380.0,
      "step": 10795
    },
    {
      "entropy": 5.5373869895935055,
      "epoch": 1.212949963497501,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004863816677274062,
      "loss": 5.4214,
      "mean_token_accuracy": 0.19051570296287537,
      "num_tokens": 24093168.0,
      "step": 10800
    },
    {
      "entropy": 5.440759801864624,
      "epoch": 1.2135115404054586,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048636790860866576,
      "loss": 5.2813,
      "mean_token_accuracy": 0.19197696298360825,
      "num_tokens": 24105355.0,
      "step": 10805
    },
    {
      "entropy": 5.428806591033935,
      "epoch": 1.214073117313416,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004863541427598889,
      "loss": 5.2454,
      "mean_token_accuracy": 0.1933158367872238,
      "num_tokens": 24117062.0,
      "step": 10810
    },
    {
      "entropy": 5.480834102630615,
      "epoch": 1.2146346942213737,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004863403701815144,
      "loss": 5.3867,
      "mean_token_accuracy": 0.18442770540714265,
      "num_tokens": 24127770.0,
      "step": 10815
    },
    {
      "entropy": 5.5059820175170895,
      "epoch": 1.2151962711293312,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004863265908739805,
      "loss": 5.46,
      "mean_token_accuracy": 0.17701869308948517,
      "num_tokens": 24139894.0,
      "step": 10820
    },
    {
      "entropy": 5.537369251251221,
      "epoch": 1.2157578480372888,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004863128048377259,
      "loss": 5.2605,
      "mean_token_accuracy": 0.199373559653759,
      "num_tokens": 24150799.0,
      "step": 10825
    },
    {
      "entropy": 5.491186857223511,
      "epoch": 1.2163194249452463,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004862990120731896,
      "loss": 5.3171,
      "mean_token_accuracy": 0.18599345833063125,
      "num_tokens": 24161406.0,
      "step": 10830
    },
    {
      "entropy": 5.300119781494141,
      "epoch": 1.2168810018532037,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048628521258081096,
      "loss": 5.2062,
      "mean_token_accuracy": 0.18894344717264175,
      "num_tokens": 24171884.0,
      "step": 10835
    },
    {
      "entropy": 5.415779256820679,
      "epoch": 1.2174425787611614,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048627140636102906,
      "loss": 5.2961,
      "mean_token_accuracy": 0.19503391534090042,
      "num_tokens": 24183873.0,
      "step": 10840
    },
    {
      "entropy": 5.570285129547119,
      "epoch": 1.2180041556691188,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004862575934142837,
      "loss": 5.327,
      "mean_token_accuracy": 0.1936239406466484,
      "num_tokens": 24194689.0,
      "step": 10845
    },
    {
      "entropy": 5.4427313804626465,
      "epoch": 1.2185657325770765,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004862437737410147,
      "loss": 5.2372,
      "mean_token_accuracy": 0.1886942520737648,
      "num_tokens": 24206675.0,
      "step": 10850
    },
    {
      "entropy": 5.395162343978882,
      "epoch": 1.219127309485034,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004862299473416619,
      "loss": 5.3246,
      "mean_token_accuracy": 0.18189102709293364,
      "num_tokens": 24218845.0,
      "step": 10855
    },
    {
      "entropy": 5.4743907928466795,
      "epoch": 1.2196888863929916,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048621611421666573,
      "loss": 5.3414,
      "mean_token_accuracy": 0.18609097152948378,
      "num_tokens": 24229889.0,
      "step": 10860
    },
    {
      "entropy": 5.528650188446045,
      "epoch": 1.220250463300949,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004862022743664665,
      "loss": 5.4129,
      "mean_token_accuracy": 0.18775719553232192,
      "num_tokens": 24240729.0,
      "step": 10865
    },
    {
      "entropy": 5.563079786300659,
      "epoch": 1.2208120402089067,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048618842779150494,
      "loss": 5.3968,
      "mean_token_accuracy": 0.1900452420115471,
      "num_tokens": 24251765.0,
      "step": 10870
    },
    {
      "entropy": 5.489863443374634,
      "epoch": 1.221373617116864,
      "grad_norm": 1.234375,
      "learning_rate": 0.00048617457449222205,
      "loss": 5.3251,
      "mean_token_accuracy": 0.18707325905561448,
      "num_tokens": 24262273.0,
      "step": 10875
    },
    {
      "entropy": 5.462723684310913,
      "epoch": 1.2219351940248218,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004861607144690587,
      "loss": 5.3194,
      "mean_token_accuracy": 0.18819813430309296,
      "num_tokens": 24272313.0,
      "step": 10880
    },
    {
      "entropy": 5.524364709854126,
      "epoch": 1.2224967709327792,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004861468477224563,
      "loss": 5.3932,
      "mean_token_accuracy": 0.18200551271438598,
      "num_tokens": 24283971.0,
      "step": 10885
    },
    {
      "entropy": 5.457604217529297,
      "epoch": 1.2230583478407369,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048613297425285635,
      "loss": 5.3259,
      "mean_token_accuracy": 0.18704751431941985,
      "num_tokens": 24294357.0,
      "step": 10890
    },
    {
      "entropy": 5.396797513961792,
      "epoch": 1.2236199247486943,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004861190940607007,
      "loss": 5.1948,
      "mean_token_accuracy": 0.20139901340007782,
      "num_tokens": 24305894.0,
      "step": 10895
    },
    {
      "entropy": 5.458518648147583,
      "epoch": 1.224181501656652,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004861052071464312,
      "loss": 5.2912,
      "mean_token_accuracy": 0.19687610417604445,
      "num_tokens": 24318506.0,
      "step": 10900
    },
    {
      "entropy": 5.427565050125122,
      "epoch": 1.2247430785646094,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048609131351048994,
      "loss": 5.2806,
      "mean_token_accuracy": 0.18720776587724686,
      "num_tokens": 24330169.0,
      "step": 10905
    },
    {
      "entropy": 5.427713775634766,
      "epoch": 1.225304655472567,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004860774131533195,
      "loss": 5.2358,
      "mean_token_accuracy": 0.20510795414447786,
      "num_tokens": 24340063.0,
      "step": 10910
    },
    {
      "entropy": 5.477045488357544,
      "epoch": 1.2258662323805245,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048606350607536235,
      "loss": 5.3265,
      "mean_token_accuracy": 0.19268265962600709,
      "num_tokens": 24351506.0,
      "step": 10915
    },
    {
      "entropy": 5.381949520111084,
      "epoch": 1.2264278092884822,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004860495922770612,
      "loss": 5.2848,
      "mean_token_accuracy": 0.18845685869455336,
      "num_tokens": 24362168.0,
      "step": 10920
    },
    {
      "entropy": 5.410154962539673,
      "epoch": 1.2269893861964396,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048603567175885926,
      "loss": 5.2512,
      "mean_token_accuracy": 0.1914696291089058,
      "num_tokens": 24372269.0,
      "step": 10925
    },
    {
      "entropy": 5.485535383224487,
      "epoch": 1.227550963104397,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004860217445211997,
      "loss": 5.3306,
      "mean_token_accuracy": 0.18908344954252243,
      "num_tokens": 24384368.0,
      "step": 10930
    },
    {
      "entropy": 5.466315841674804,
      "epoch": 1.2281125400123547,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004860078105645259,
      "loss": 5.2635,
      "mean_token_accuracy": 0.19214505553245545,
      "num_tokens": 24395395.0,
      "step": 10935
    },
    {
      "entropy": 5.386342000961304,
      "epoch": 1.2286741169203121,
      "grad_norm": 1.375,
      "learning_rate": 0.0004859938698892816,
      "loss": 5.226,
      "mean_token_accuracy": 0.1923302248120308,
      "num_tokens": 24406709.0,
      "step": 10940
    },
    {
      "entropy": 5.430661344528199,
      "epoch": 1.2292356938282698,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048597992249591065,
      "loss": 5.3773,
      "mean_token_accuracy": 0.19163648635149003,
      "num_tokens": 24418574.0,
      "step": 10945
    },
    {
      "entropy": 5.535741853713989,
      "epoch": 1.2297972707362272,
      "grad_norm": 1.125,
      "learning_rate": 0.00048596596838485704,
      "loss": 5.338,
      "mean_token_accuracy": 0.18751323074102402,
      "num_tokens": 24429418.0,
      "step": 10950
    },
    {
      "entropy": 5.526889133453369,
      "epoch": 1.230358847644185,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004859520075565654,
      "loss": 5.3988,
      "mean_token_accuracy": 0.18943314403295516,
      "num_tokens": 24440301.0,
      "step": 10955
    },
    {
      "entropy": 5.542314958572388,
      "epoch": 1.2309204245521423,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048593804001147986,
      "loss": 5.4153,
      "mean_token_accuracy": 0.18399468660354615,
      "num_tokens": 24451987.0,
      "step": 10960
    },
    {
      "entropy": 5.508684158325195,
      "epoch": 1.2314820014601,
      "grad_norm": 1.125,
      "learning_rate": 0.0004859240657500453,
      "loss": 5.2963,
      "mean_token_accuracy": 0.1933690696954727,
      "num_tokens": 24461971.0,
      "step": 10965
    },
    {
      "entropy": 5.382048749923706,
      "epoch": 1.2320435783680574,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048591008477270675,
      "loss": 5.2008,
      "mean_token_accuracy": 0.1976775735616684,
      "num_tokens": 24472193.0,
      "step": 10970
    },
    {
      "entropy": 5.39286618232727,
      "epoch": 1.232605155276015,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048589609707990925,
      "loss": 5.2932,
      "mean_token_accuracy": 0.18966082930564881,
      "num_tokens": 24483075.0,
      "step": 10975
    },
    {
      "entropy": 5.464377164840698,
      "epoch": 1.2331667321839725,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048588210267209823,
      "loss": 5.2361,
      "mean_token_accuracy": 0.19926120042800904,
      "num_tokens": 24493070.0,
      "step": 10980
    },
    {
      "entropy": 5.4815339088439945,
      "epoch": 1.2337283090919302,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048586810154971935,
      "loss": 5.3577,
      "mean_token_accuracy": 0.18928852528333664,
      "num_tokens": 24502866.0,
      "step": 10985
    },
    {
      "entropy": 5.392825937271118,
      "epoch": 1.2342898859998876,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048585409371321826,
      "loss": 5.22,
      "mean_token_accuracy": 0.1960653305053711,
      "num_tokens": 24514169.0,
      "step": 10990
    },
    {
      "entropy": 5.472100496292114,
      "epoch": 1.2348514629078453,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048584007916304106,
      "loss": 5.4076,
      "mean_token_accuracy": 0.186797533929348,
      "num_tokens": 24525552.0,
      "step": 10995
    },
    {
      "entropy": 5.526369237899781,
      "epoch": 1.2354130398158027,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048582605789963395,
      "loss": 5.3327,
      "mean_token_accuracy": 0.18830710649490356,
      "num_tokens": 24536370.0,
      "step": 11000
    },
    {
      "entropy": 5.503993272781372,
      "epoch": 1.2359746167237604,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004858120299234434,
      "loss": 5.3655,
      "mean_token_accuracy": 0.18914033621549606,
      "num_tokens": 24546868.0,
      "step": 11005
    },
    {
      "entropy": 5.519772005081177,
      "epoch": 1.2365361936317179,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048579799523491607,
      "loss": 5.3091,
      "mean_token_accuracy": 0.1827283799648285,
      "num_tokens": 24558279.0,
      "step": 11010
    },
    {
      "entropy": 5.425114536285401,
      "epoch": 1.2370977705396755,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004857839538344988,
      "loss": 5.2464,
      "mean_token_accuracy": 0.19218159317970276,
      "num_tokens": 24568321.0,
      "step": 11015
    },
    {
      "entropy": 5.448210668563843,
      "epoch": 1.237659347447633,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048576990572263875,
      "loss": 5.3071,
      "mean_token_accuracy": 0.18126791566610337,
      "num_tokens": 24579137.0,
      "step": 11020
    },
    {
      "entropy": 5.516415071487427,
      "epoch": 1.2382209243555904,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004857558508997831,
      "loss": 5.3483,
      "mean_token_accuracy": 0.18391966819763184,
      "num_tokens": 24591523.0,
      "step": 11025
    },
    {
      "entropy": 5.504225969314575,
      "epoch": 1.238782501263548,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004857417893663794,
      "loss": 5.2838,
      "mean_token_accuracy": 0.19196339696645737,
      "num_tokens": 24601998.0,
      "step": 11030
    },
    {
      "entropy": 5.423813676834106,
      "epoch": 1.2393440781715055,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004857277211228754,
      "loss": 5.265,
      "mean_token_accuracy": 0.19372736811637878,
      "num_tokens": 24612371.0,
      "step": 11035
    },
    {
      "entropy": 5.496390247344971,
      "epoch": 1.2399056550794632,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004857136461697191,
      "loss": 5.3475,
      "mean_token_accuracy": 0.18874841928482056,
      "num_tokens": 24623350.0,
      "step": 11040
    },
    {
      "entropy": 5.475476694107056,
      "epoch": 1.2404672319874206,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048569956450735853,
      "loss": 5.3432,
      "mean_token_accuracy": 0.18772679418325425,
      "num_tokens": 24634365.0,
      "step": 11045
    },
    {
      "entropy": 5.5366668701171875,
      "epoch": 1.2410288088953783,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048568547613624206,
      "loss": 5.44,
      "mean_token_accuracy": 0.18956054002046585,
      "num_tokens": 24645227.0,
      "step": 11050
    },
    {
      "entropy": 5.564941501617431,
      "epoch": 1.2415903858033357,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048567138105681833,
      "loss": 5.373,
      "mean_token_accuracy": 0.18783759623765944,
      "num_tokens": 24656807.0,
      "step": 11055
    },
    {
      "entropy": 5.467393398284912,
      "epoch": 1.2421519627112934,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048565727926953617,
      "loss": 5.3768,
      "mean_token_accuracy": 0.184771303832531,
      "num_tokens": 24669021.0,
      "step": 11060
    },
    {
      "entropy": 5.507412385940552,
      "epoch": 1.2427135396192508,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004856431707748445,
      "loss": 5.2864,
      "mean_token_accuracy": 0.1981239378452301,
      "num_tokens": 24678802.0,
      "step": 11065
    },
    {
      "entropy": 5.483022975921631,
      "epoch": 1.2432751165272085,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004856290555731926,
      "loss": 5.2883,
      "mean_token_accuracy": 0.1897783249616623,
      "num_tokens": 24688449.0,
      "step": 11070
    },
    {
      "entropy": 5.414370155334472,
      "epoch": 1.243836693435166,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004856149336650299,
      "loss": 5.3322,
      "mean_token_accuracy": 0.18270452916622162,
      "num_tokens": 24700272.0,
      "step": 11075
    },
    {
      "entropy": 5.5557033061981205,
      "epoch": 1.2443982703431236,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004856008050508059,
      "loss": 5.485,
      "mean_token_accuracy": 0.18108724802732468,
      "num_tokens": 24711949.0,
      "step": 11080
    },
    {
      "entropy": 5.57469220161438,
      "epoch": 1.244959847251081,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004855866697309707,
      "loss": 5.298,
      "mean_token_accuracy": 0.1914838582277298,
      "num_tokens": 24723535.0,
      "step": 11085
    },
    {
      "entropy": 5.405101108551025,
      "epoch": 1.2455214241590387,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004855725277059742,
      "loss": 5.3275,
      "mean_token_accuracy": 0.18285240978002548,
      "num_tokens": 24734915.0,
      "step": 11090
    },
    {
      "entropy": 5.443416166305542,
      "epoch": 1.246083001066996,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004855583789762668,
      "loss": 5.2489,
      "mean_token_accuracy": 0.19913407415151596,
      "num_tokens": 24745744.0,
      "step": 11095
    },
    {
      "entropy": 5.5102623462677,
      "epoch": 1.2466445779749538,
      "grad_norm": 1.21875,
      "learning_rate": 0.000485544223542299,
      "loss": 5.4461,
      "mean_token_accuracy": 0.18992600589990616,
      "num_tokens": 24758158.0,
      "step": 11100
    },
    {
      "entropy": 5.520338916778565,
      "epoch": 1.2472061548829112,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004855300614045214,
      "loss": 5.3019,
      "mean_token_accuracy": 0.19414403289556503,
      "num_tokens": 24768607.0,
      "step": 11105
    },
    {
      "entropy": 5.4876473903656,
      "epoch": 1.2477677317908689,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048551589256338497,
      "loss": 5.2914,
      "mean_token_accuracy": 0.19670223742723464,
      "num_tokens": 24779845.0,
      "step": 11110
    },
    {
      "entropy": 5.407142877578735,
      "epoch": 1.2483293086988263,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048550171701934087,
      "loss": 5.2655,
      "mean_token_accuracy": 0.18386080712080002,
      "num_tokens": 24790049.0,
      "step": 11115
    },
    {
      "entropy": 5.541993713378906,
      "epoch": 1.2488908856067837,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048548753477284053,
      "loss": 5.321,
      "mean_token_accuracy": 0.1838960662484169,
      "num_tokens": 24802112.0,
      "step": 11120
    },
    {
      "entropy": 5.426882123947143,
      "epoch": 1.2494524625147414,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004854733458243354,
      "loss": 5.2584,
      "mean_token_accuracy": 0.1877560794353485,
      "num_tokens": 24812630.0,
      "step": 11125
    },
    {
      "entropy": 5.384331703186035,
      "epoch": 1.250014039422699,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004854591501742773,
      "loss": 5.2501,
      "mean_token_accuracy": 0.19661716371774673,
      "num_tokens": 24823346.0,
      "step": 11130
    },
    {
      "entropy": 5.461381912231445,
      "epoch": 1.2505756163306565,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048544494782311825,
      "loss": 5.3354,
      "mean_token_accuracy": 0.18701744675636292,
      "num_tokens": 24834374.0,
      "step": 11135
    },
    {
      "entropy": 5.442190504074096,
      "epoch": 1.251137193238614,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048543073877131037,
      "loss": 5.1782,
      "mean_token_accuracy": 0.19452890157699584,
      "num_tokens": 24845474.0,
      "step": 11140
    },
    {
      "entropy": 5.454979085922242,
      "epoch": 1.2516987701465716,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048541652301930627,
      "loss": 5.2473,
      "mean_token_accuracy": 0.19900314658880233,
      "num_tokens": 24857427.0,
      "step": 11145
    },
    {
      "entropy": 5.464489078521728,
      "epoch": 1.252260347054529,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004854023005675585,
      "loss": 5.2661,
      "mean_token_accuracy": 0.19213441908359527,
      "num_tokens": 24868352.0,
      "step": 11150
    },
    {
      "entropy": 5.410154151916504,
      "epoch": 1.2528219239624867,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004853880714165197,
      "loss": 5.3345,
      "mean_token_accuracy": 0.1893559977412224,
      "num_tokens": 24878499.0,
      "step": 11155
    },
    {
      "entropy": 5.384410429000854,
      "epoch": 1.2533835008704441,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048537383556664315,
      "loss": 5.3034,
      "mean_token_accuracy": 0.19334885179996492,
      "num_tokens": 24888290.0,
      "step": 11160
    },
    {
      "entropy": 5.453674173355102,
      "epoch": 1.2539450777784018,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004853595930183822,
      "loss": 5.3685,
      "mean_token_accuracy": 0.19106528908014297,
      "num_tokens": 24900201.0,
      "step": 11165
    },
    {
      "entropy": 5.403820753097534,
      "epoch": 1.2545066546863592,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004853453437721901,
      "loss": 5.2668,
      "mean_token_accuracy": 0.18901242911815644,
      "num_tokens": 24910895.0,
      "step": 11170
    },
    {
      "entropy": 5.404502820968628,
      "epoch": 1.255068231594317,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004853310878285208,
      "loss": 5.1999,
      "mean_token_accuracy": 0.1949177637696266,
      "num_tokens": 24922079.0,
      "step": 11175
    },
    {
      "entropy": 5.5890247344970705,
      "epoch": 1.2556298085022743,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048531682518782797,
      "loss": 5.5061,
      "mean_token_accuracy": 0.18496251851320267,
      "num_tokens": 24935093.0,
      "step": 11180
    },
    {
      "entropy": 5.595027256011963,
      "epoch": 1.256191385410232,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048530255585056594,
      "loss": 5.4573,
      "mean_token_accuracy": 0.18300241827964783,
      "num_tokens": 24947316.0,
      "step": 11185
    },
    {
      "entropy": 5.413075256347656,
      "epoch": 1.2567529623181894,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004852882798171889,
      "loss": 5.1862,
      "mean_token_accuracy": 0.1983880117535591,
      "num_tokens": 24957255.0,
      "step": 11190
    },
    {
      "entropy": 5.409970331192016,
      "epoch": 1.2573145392261469,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004852739970881515,
      "loss": 5.3016,
      "mean_token_accuracy": 0.19258214682340621,
      "num_tokens": 24967646.0,
      "step": 11195
    },
    {
      "entropy": 5.52261791229248,
      "epoch": 1.2578761161341045,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048525970766390847,
      "loss": 5.4438,
      "mean_token_accuracy": 0.18403465449810028,
      "num_tokens": 24979748.0,
      "step": 11200
    },
    {
      "entropy": 5.583576583862305,
      "epoch": 1.2584376930420622,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004852454115449148,
      "loss": 5.4279,
      "mean_token_accuracy": 0.1899477645754814,
      "num_tokens": 24990727.0,
      "step": 11205
    },
    {
      "entropy": 5.430970239639282,
      "epoch": 1.2589992699500197,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004852311087316257,
      "loss": 5.2419,
      "mean_token_accuracy": 0.1907973498106003,
      "num_tokens": 25001912.0,
      "step": 11210
    },
    {
      "entropy": 5.484019470214844,
      "epoch": 1.259560846857977,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004852167992244966,
      "loss": 5.3405,
      "mean_token_accuracy": 0.185517418384552,
      "num_tokens": 25012952.0,
      "step": 11215
    },
    {
      "entropy": 5.416284084320068,
      "epoch": 1.2601224237659348,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000485202483023983,
      "loss": 5.2495,
      "mean_token_accuracy": 0.19014496952295304,
      "num_tokens": 25022851.0,
      "step": 11220
    },
    {
      "entropy": 5.4342276096344,
      "epoch": 1.2606840006738924,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048518816013054083,
      "loss": 5.2723,
      "mean_token_accuracy": 0.19107894748449325,
      "num_tokens": 25033177.0,
      "step": 11225
    },
    {
      "entropy": 5.4692412376403805,
      "epoch": 1.2612455775818499,
      "grad_norm": 1.25,
      "learning_rate": 0.00048517383054462617,
      "loss": 5.2253,
      "mean_token_accuracy": 0.19505978375673294,
      "num_tokens": 25044140.0,
      "step": 11230
    },
    {
      "entropy": 5.4517604351043705,
      "epoch": 1.2618071544898073,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004851594942666952,
      "loss": 5.273,
      "mean_token_accuracy": 0.18973116725683212,
      "num_tokens": 25054480.0,
      "step": 11235
    },
    {
      "entropy": 5.468903207778931,
      "epoch": 1.262368731397765,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048514515129720445,
      "loss": 5.3296,
      "mean_token_accuracy": 0.1831263229250908,
      "num_tokens": 25065673.0,
      "step": 11240
    },
    {
      "entropy": 5.4786396503448485,
      "epoch": 1.2629303083057224,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004851308016366105,
      "loss": 5.3196,
      "mean_token_accuracy": 0.19155306071043016,
      "num_tokens": 25076691.0,
      "step": 11245
    },
    {
      "entropy": 5.512743902206421,
      "epoch": 1.26349188521368,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048511644528537043,
      "loss": 5.3021,
      "mean_token_accuracy": 0.1940959244966507,
      "num_tokens": 25087882.0,
      "step": 11250
    },
    {
      "entropy": 5.493298435211182,
      "epoch": 1.2640534621216375,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004851020822439412,
      "loss": 5.4113,
      "mean_token_accuracy": 0.181867752969265,
      "num_tokens": 25099325.0,
      "step": 11255
    },
    {
      "entropy": 5.340627288818359,
      "epoch": 1.2646150390295952,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048508771251278015,
      "loss": 5.1891,
      "mean_token_accuracy": 0.19315120577812195,
      "num_tokens": 25110304.0,
      "step": 11260
    },
    {
      "entropy": 5.4338020324707035,
      "epoch": 1.2651766159375526,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004850733360923449,
      "loss": 5.3434,
      "mean_token_accuracy": 0.19013926088809968,
      "num_tokens": 25121597.0,
      "step": 11265
    },
    {
      "entropy": 5.501984977722168,
      "epoch": 1.2657381928455103,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004850589529830931,
      "loss": 5.3204,
      "mean_token_accuracy": 0.18620147705078124,
      "num_tokens": 25132817.0,
      "step": 11270
    },
    {
      "entropy": 5.539655876159668,
      "epoch": 1.2662997697534677,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048504456318548285,
      "loss": 5.3576,
      "mean_token_accuracy": 0.18903307467699051,
      "num_tokens": 25143740.0,
      "step": 11275
    },
    {
      "entropy": 5.484044313430786,
      "epoch": 1.2668613466614254,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004850301666999722,
      "loss": 5.3034,
      "mean_token_accuracy": 0.187858647108078,
      "num_tokens": 25155889.0,
      "step": 11280
    },
    {
      "entropy": 5.4342155933380125,
      "epoch": 1.2674229235693828,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004850157635270196,
      "loss": 5.4132,
      "mean_token_accuracy": 0.1899063691496849,
      "num_tokens": 25166043.0,
      "step": 11285
    },
    {
      "entropy": 5.4646679878234865,
      "epoch": 1.2679845004773402,
      "grad_norm": 1.375,
      "learning_rate": 0.0004850013536670836,
      "loss": 5.3047,
      "mean_token_accuracy": 0.19433421939611434,
      "num_tokens": 25176906.0,
      "step": 11290
    },
    {
      "entropy": 5.430424165725708,
      "epoch": 1.268546077385298,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004849869371206231,
      "loss": 5.275,
      "mean_token_accuracy": 0.18700045347213745,
      "num_tokens": 25187803.0,
      "step": 11295
    },
    {
      "entropy": 5.40882453918457,
      "epoch": 1.2691076542932556,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048497251388809697,
      "loss": 5.2202,
      "mean_token_accuracy": 0.18885144591331482,
      "num_tokens": 25199067.0,
      "step": 11300
    },
    {
      "entropy": 5.4471231460571286,
      "epoch": 1.269669231201213,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004849580839699647,
      "loss": 5.3026,
      "mean_token_accuracy": 0.188690884411335,
      "num_tokens": 25210768.0,
      "step": 11305
    },
    {
      "entropy": 5.529627513885498,
      "epoch": 1.2702308081091704,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048494364736668554,
      "loss": 5.3506,
      "mean_token_accuracy": 0.19032243490219117,
      "num_tokens": 25221602.0,
      "step": 11310
    },
    {
      "entropy": 5.472881603240967,
      "epoch": 1.270792385017128,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004849292040787192,
      "loss": 5.3372,
      "mean_token_accuracy": 0.18918239325284958,
      "num_tokens": 25232791.0,
      "step": 11315
    },
    {
      "entropy": 5.43222827911377,
      "epoch": 1.2713539619250858,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004849147541065256,
      "loss": 5.2869,
      "mean_token_accuracy": 0.19072076231241225,
      "num_tokens": 25243962.0,
      "step": 11320
    },
    {
      "entropy": 5.462640047073364,
      "epoch": 1.2719155388330432,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004849002974505649,
      "loss": 5.2705,
      "mean_token_accuracy": 0.19021091610193253,
      "num_tokens": 25254636.0,
      "step": 11325
    },
    {
      "entropy": 5.413700008392334,
      "epoch": 1.2724771157410006,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048488583411129723,
      "loss": 5.2618,
      "mean_token_accuracy": 0.19617535322904586,
      "num_tokens": 25265255.0,
      "step": 11330
    },
    {
      "entropy": 5.49795708656311,
      "epoch": 1.2730386926489583,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004848713640891832,
      "loss": 5.3766,
      "mean_token_accuracy": 0.18917652517557143,
      "num_tokens": 25276721.0,
      "step": 11335
    },
    {
      "entropy": 5.452847099304199,
      "epoch": 1.2736002695569157,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004848568873846835,
      "loss": 5.2871,
      "mean_token_accuracy": 0.19696080982685088,
      "num_tokens": 25287451.0,
      "step": 11340
    },
    {
      "entropy": 5.417770433425903,
      "epoch": 1.2741618464648734,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048484240399825925,
      "loss": 5.2412,
      "mean_token_accuracy": 0.1925506889820099,
      "num_tokens": 25298609.0,
      "step": 11345
    },
    {
      "entropy": 5.49193000793457,
      "epoch": 1.2747234233728308,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048482791393037134,
      "loss": 5.2898,
      "mean_token_accuracy": 0.19458225667476653,
      "num_tokens": 25310058.0,
      "step": 11350
    },
    {
      "entropy": 5.428733158111572,
      "epoch": 1.2752850002807885,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048481341718148137,
      "loss": 5.2872,
      "mean_token_accuracy": 0.1992921695113182,
      "num_tokens": 25321249.0,
      "step": 11355
    },
    {
      "entropy": 5.436584424972534,
      "epoch": 1.275846577188746,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048479891375205074,
      "loss": 5.2678,
      "mean_token_accuracy": 0.1908138170838356,
      "num_tokens": 25332603.0,
      "step": 11360
    },
    {
      "entropy": 5.372139739990234,
      "epoch": 1.2764081540967036,
      "grad_norm": 1.8515625,
      "learning_rate": 0.00048478440364254136,
      "loss": 5.2709,
      "mean_token_accuracy": 0.20115242749452591,
      "num_tokens": 25344955.0,
      "step": 11365
    },
    {
      "entropy": 5.525250196456909,
      "epoch": 1.276969731004661,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004847698868534152,
      "loss": 5.4159,
      "mean_token_accuracy": 0.1842200830578804,
      "num_tokens": 25356125.0,
      "step": 11370
    },
    {
      "entropy": 5.493772315979004,
      "epoch": 1.2775313079126187,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004847553633851345,
      "loss": 5.2456,
      "mean_token_accuracy": 0.19680748432874678,
      "num_tokens": 25366969.0,
      "step": 11375
    },
    {
      "entropy": 5.465518951416016,
      "epoch": 1.2780928848205761,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004847408332381616,
      "loss": 5.309,
      "mean_token_accuracy": 0.19478828608989715,
      "num_tokens": 25378815.0,
      "step": 11380
    },
    {
      "entropy": 5.326076030731201,
      "epoch": 1.2786544617285336,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048472629641295933,
      "loss": 5.2358,
      "mean_token_accuracy": 0.19650703221559523,
      "num_tokens": 25389969.0,
      "step": 11385
    },
    {
      "entropy": 5.391957330703735,
      "epoch": 1.2792160386364912,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004847117529099904,
      "loss": 5.2926,
      "mean_token_accuracy": 0.18488344550132751,
      "num_tokens": 25401443.0,
      "step": 11390
    },
    {
      "entropy": 5.516592264175415,
      "epoch": 1.279777615544449,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048469720272971797,
      "loss": 5.3257,
      "mean_token_accuracy": 0.1891709089279175,
      "num_tokens": 25412857.0,
      "step": 11395
    },
    {
      "entropy": 5.474941825866699,
      "epoch": 1.2803391924524063,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004846826458726052,
      "loss": 5.2112,
      "mean_token_accuracy": 0.1938275247812271,
      "num_tokens": 25423804.0,
      "step": 11400
    },
    {
      "entropy": 5.417476081848145,
      "epoch": 1.2809007693603638,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048466808233911567,
      "loss": 5.376,
      "mean_token_accuracy": 0.18660081475973128,
      "num_tokens": 25435686.0,
      "step": 11405
    },
    {
      "entropy": 5.494922685623169,
      "epoch": 1.2814623462683215,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048465351212971313,
      "loss": 5.3234,
      "mean_token_accuracy": 0.1888633906841278,
      "num_tokens": 25446012.0,
      "step": 11410
    },
    {
      "entropy": 5.511262464523315,
      "epoch": 1.2820239231762791,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004846389352448614,
      "loss": 5.3532,
      "mean_token_accuracy": 0.19125574380159377,
      "num_tokens": 25457768.0,
      "step": 11415
    },
    {
      "entropy": 5.441014528274536,
      "epoch": 1.2825855000842366,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004846243516850247,
      "loss": 5.2955,
      "mean_token_accuracy": 0.18980908244848252,
      "num_tokens": 25468706.0,
      "step": 11420
    },
    {
      "entropy": 5.493085336685181,
      "epoch": 1.283147076992194,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004846097614506673,
      "loss": 5.3581,
      "mean_token_accuracy": 0.18504564464092255,
      "num_tokens": 25479446.0,
      "step": 11425
    },
    {
      "entropy": 5.454481410980224,
      "epoch": 1.2837086539001517,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048459516454225384,
      "loss": 5.2895,
      "mean_token_accuracy": 0.19178877174854278,
      "num_tokens": 25491090.0,
      "step": 11430
    },
    {
      "entropy": 5.491308116912842,
      "epoch": 1.284270230808109,
      "grad_norm": 1.15625,
      "learning_rate": 0.000484580560960249,
      "loss": 5.3927,
      "mean_token_accuracy": 0.19185578525066377,
      "num_tokens": 25502105.0,
      "step": 11435
    },
    {
      "entropy": 5.6139665126800535,
      "epoch": 1.2848318077160668,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004845659507051178,
      "loss": 5.4558,
      "mean_token_accuracy": 0.1842873528599739,
      "num_tokens": 25513970.0,
      "step": 11440
    },
    {
      "entropy": 5.518034315109253,
      "epoch": 1.2853933846240242,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004845513337773255,
      "loss": 5.2963,
      "mean_token_accuracy": 0.18923108875751496,
      "num_tokens": 25524508.0,
      "step": 11445
    },
    {
      "entropy": 5.42512788772583,
      "epoch": 1.2859549615319819,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048453671017733755,
      "loss": 5.2547,
      "mean_token_accuracy": 0.18632541000843048,
      "num_tokens": 25536006.0,
      "step": 11450
    },
    {
      "entropy": 5.443130540847778,
      "epoch": 1.2865165384399393,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004845220799056194,
      "loss": 5.3053,
      "mean_token_accuracy": 0.1895155817270279,
      "num_tokens": 25545726.0,
      "step": 11455
    },
    {
      "entropy": 5.411314821243286,
      "epoch": 1.287078115347897,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048450744296263685,
      "loss": 5.2414,
      "mean_token_accuracy": 0.1947297364473343,
      "num_tokens": 25555914.0,
      "step": 11460
    },
    {
      "entropy": 5.537866449356079,
      "epoch": 1.2876396922558544,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048449279934885625,
      "loss": 5.4241,
      "mean_token_accuracy": 0.1824166089296341,
      "num_tokens": 25568021.0,
      "step": 11465
    },
    {
      "entropy": 5.527729845046997,
      "epoch": 1.288201269163812,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004844781490647435,
      "loss": 5.3169,
      "mean_token_accuracy": 0.18916783034801482,
      "num_tokens": 25578788.0,
      "step": 11470
    },
    {
      "entropy": 5.458562183380127,
      "epoch": 1.2887628460717695,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004844634921107654,
      "loss": 5.3312,
      "mean_token_accuracy": 0.19275230318307876,
      "num_tokens": 25590997.0,
      "step": 11475
    },
    {
      "entropy": 5.512963199615479,
      "epoch": 1.289324422979727,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048444882848738835,
      "loss": 5.3917,
      "mean_token_accuracy": 0.18336481899023055,
      "num_tokens": 25602810.0,
      "step": 11480
    },
    {
      "entropy": 5.606393146514892,
      "epoch": 1.2898859998876846,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048443415819507944,
      "loss": 5.396,
      "mean_token_accuracy": 0.18886154443025588,
      "num_tokens": 25614251.0,
      "step": 11485
    },
    {
      "entropy": 5.512076377868652,
      "epoch": 1.2904475767956423,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004844194812343057,
      "loss": 5.3844,
      "mean_token_accuracy": 0.18469516932964325,
      "num_tokens": 25626162.0,
      "step": 11490
    },
    {
      "entropy": 5.528716707229615,
      "epoch": 1.2910091537035997,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004844047976055344,
      "loss": 5.3646,
      "mean_token_accuracy": 0.19066940397024154,
      "num_tokens": 25637598.0,
      "step": 11495
    },
    {
      "entropy": 5.462852954864502,
      "epoch": 1.2915707306115571,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048439010730923313,
      "loss": 5.294,
      "mean_token_accuracy": 0.19428147673606871,
      "num_tokens": 25648297.0,
      "step": 11500
    },
    {
      "entropy": 5.467259120941162,
      "epoch": 1.2921323075195148,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048437541034586964,
      "loss": 5.3596,
      "mean_token_accuracy": 0.1839531645178795,
      "num_tokens": 25660139.0,
      "step": 11505
    },
    {
      "entropy": 5.50232195854187,
      "epoch": 1.2926938844274725,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048436070671591195,
      "loss": 5.2965,
      "mean_token_accuracy": 0.19321619868278503,
      "num_tokens": 25671947.0,
      "step": 11510
    },
    {
      "entropy": 5.438428783416748,
      "epoch": 1.29325546133543,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048434599641982806,
      "loss": 5.2307,
      "mean_token_accuracy": 0.19336870163679123,
      "num_tokens": 25682059.0,
      "step": 11515
    },
    {
      "entropy": 5.448292446136475,
      "epoch": 1.2938170382433873,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00048433127945808645,
      "loss": 5.3341,
      "mean_token_accuracy": 0.1864840567111969,
      "num_tokens": 25692792.0,
      "step": 11520
    },
    {
      "entropy": 5.512326622009278,
      "epoch": 1.294378615151345,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004843165558311558,
      "loss": 5.3545,
      "mean_token_accuracy": 0.18703494518995284,
      "num_tokens": 25703253.0,
      "step": 11525
    },
    {
      "entropy": 5.524423313140869,
      "epoch": 1.2949401920593024,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004843018255395048,
      "loss": 5.2873,
      "mean_token_accuracy": 0.19686329662799834,
      "num_tokens": 25714169.0,
      "step": 11530
    },
    {
      "entropy": 5.471404552459717,
      "epoch": 1.29550176896726,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048428708858360244,
      "loss": 5.3751,
      "mean_token_accuracy": 0.1883690521121025,
      "num_tokens": 25725827.0,
      "step": 11535
    },
    {
      "entropy": 5.447461366653442,
      "epoch": 1.2960633458752175,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048427234496391816,
      "loss": 5.33,
      "mean_token_accuracy": 0.18871326446533204,
      "num_tokens": 25737951.0,
      "step": 11540
    },
    {
      "entropy": 5.43828239440918,
      "epoch": 1.2966249227831752,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048425759468092113,
      "loss": 5.343,
      "mean_token_accuracy": 0.18574848920106887,
      "num_tokens": 25749676.0,
      "step": 11545
    },
    {
      "entropy": 5.520509433746338,
      "epoch": 1.2971864996911326,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004842428377350811,
      "loss": 5.3363,
      "mean_token_accuracy": 0.19716714024543763,
      "num_tokens": 25760005.0,
      "step": 11550
    },
    {
      "entropy": 5.560976886749268,
      "epoch": 1.2977480765990903,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048422807412686796,
      "loss": 5.398,
      "mean_token_accuracy": 0.17945507168769836,
      "num_tokens": 25770595.0,
      "step": 11555
    },
    {
      "entropy": 5.572739458084106,
      "epoch": 1.2983096535070477,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004842133038567519,
      "loss": 5.422,
      "mean_token_accuracy": 0.18912518173456191,
      "num_tokens": 25782301.0,
      "step": 11560
    },
    {
      "entropy": 5.49431791305542,
      "epoch": 1.2988712304150054,
      "grad_norm": 1.140625,
      "learning_rate": 0.000484198526925203,
      "loss": 5.2614,
      "mean_token_accuracy": 0.19575177282094955,
      "num_tokens": 25792904.0,
      "step": 11565
    },
    {
      "entropy": 5.467157173156738,
      "epoch": 1.2994328073229628,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004841837433326919,
      "loss": 5.3926,
      "mean_token_accuracy": 0.18410500437021254,
      "num_tokens": 25804594.0,
      "step": 11570
    },
    {
      "entropy": 5.444460678100586,
      "epoch": 1.2999943842309205,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004841689530796893,
      "loss": 5.2923,
      "mean_token_accuracy": 0.18940672725439073,
      "num_tokens": 25816496.0,
      "step": 11575
    },
    {
      "entropy": 5.458776235580444,
      "epoch": 1.300555961138878,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004841541561666661,
      "loss": 5.2347,
      "mean_token_accuracy": 0.1920021429657936,
      "num_tokens": 25827271.0,
      "step": 11580
    },
    {
      "entropy": 5.488854598999024,
      "epoch": 1.3011175380468356,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048413935259409343,
      "loss": 5.3082,
      "mean_token_accuracy": 0.1871398240327835,
      "num_tokens": 25837869.0,
      "step": 11585
    },
    {
      "entropy": 5.419495916366577,
      "epoch": 1.301679114954793,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048412454236244267,
      "loss": 5.2488,
      "mean_token_accuracy": 0.19229496866464615,
      "num_tokens": 25850244.0,
      "step": 11590
    },
    {
      "entropy": 5.462931442260742,
      "epoch": 1.3022406918627505,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048410972547218545,
      "loss": 5.3441,
      "mean_token_accuracy": 0.18771826326847077,
      "num_tokens": 25859981.0,
      "step": 11595
    },
    {
      "entropy": 5.41279411315918,
      "epoch": 1.3028022687707081,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048409490192379337,
      "loss": 5.209,
      "mean_token_accuracy": 0.2058693900704384,
      "num_tokens": 25870393.0,
      "step": 11600
    },
    {
      "entropy": 5.497046232223511,
      "epoch": 1.3033638456786658,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004840800717177386,
      "loss": 5.3702,
      "mean_token_accuracy": 0.19023812413215638,
      "num_tokens": 25882517.0,
      "step": 11605
    },
    {
      "entropy": 5.493133640289306,
      "epoch": 1.3039254225866233,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004840652348544931,
      "loss": 5.3034,
      "mean_token_accuracy": 0.19558178186416625,
      "num_tokens": 25893628.0,
      "step": 11610
    },
    {
      "entropy": 5.482105493545532,
      "epoch": 1.3044869994945807,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004840503913345296,
      "loss": 5.3446,
      "mean_token_accuracy": 0.18301226049661637,
      "num_tokens": 25904891.0,
      "step": 11615
    },
    {
      "entropy": 5.53426775932312,
      "epoch": 1.3050485764025384,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004840355411583205,
      "loss": 5.3278,
      "mean_token_accuracy": 0.18413127809762955,
      "num_tokens": 25914598.0,
      "step": 11620
    },
    {
      "entropy": 5.538898372650147,
      "epoch": 1.3056101533104958,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048402068432633864,
      "loss": 5.2983,
      "mean_token_accuracy": 0.1908440351486206,
      "num_tokens": 25924412.0,
      "step": 11625
    },
    {
      "entropy": 5.387416791915894,
      "epoch": 1.3061717302184535,
      "grad_norm": 1.25,
      "learning_rate": 0.0004840058208390572,
      "loss": 5.3127,
      "mean_token_accuracy": 0.18722869902849198,
      "num_tokens": 25935583.0,
      "step": 11630
    },
    {
      "entropy": 5.399539422988892,
      "epoch": 1.306733307126411,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004839909506969494,
      "loss": 5.2157,
      "mean_token_accuracy": 0.19659731686115264,
      "num_tokens": 25946949.0,
      "step": 11635
    },
    {
      "entropy": 5.41435432434082,
      "epoch": 1.3072948840343686,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004839760739004886,
      "loss": 5.2348,
      "mean_token_accuracy": 0.19048150926828383,
      "num_tokens": 25957070.0,
      "step": 11640
    },
    {
      "entropy": 5.443178510665893,
      "epoch": 1.307856460942326,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048396119045014853,
      "loss": 5.3515,
      "mean_token_accuracy": 0.18485526740550995,
      "num_tokens": 25968375.0,
      "step": 11645
    },
    {
      "entropy": 5.585588264465332,
      "epoch": 1.3084180378502837,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004839463003464032,
      "loss": 5.3446,
      "mean_token_accuracy": 0.1881679281592369,
      "num_tokens": 25979690.0,
      "step": 11650
    },
    {
      "entropy": 5.578572940826416,
      "epoch": 1.308979614758241,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004839314035897265,
      "loss": 5.4571,
      "mean_token_accuracy": 0.178991961479187,
      "num_tokens": 25990660.0,
      "step": 11655
    },
    {
      "entropy": 5.346743249893189,
      "epoch": 1.3095411916661988,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000483916500180593,
      "loss": 5.1925,
      "mean_token_accuracy": 0.19470360577106477,
      "num_tokens": 26002158.0,
      "step": 11660
    },
    {
      "entropy": 5.437980842590332,
      "epoch": 1.3101027685741562,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048390159011947706,
      "loss": 5.2241,
      "mean_token_accuracy": 0.19745472818613052,
      "num_tokens": 26011890.0,
      "step": 11665
    },
    {
      "entropy": 5.4943560600280765,
      "epoch": 1.3106643454821139,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004838866734068535,
      "loss": 5.3355,
      "mean_token_accuracy": 0.1872969835996628,
      "num_tokens": 26023548.0,
      "step": 11670
    },
    {
      "entropy": 5.470469570159912,
      "epoch": 1.3112259223900713,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048387175004319723,
      "loss": 5.3498,
      "mean_token_accuracy": 0.18960476368665696,
      "num_tokens": 26034178.0,
      "step": 11675
    },
    {
      "entropy": 5.45514030456543,
      "epoch": 1.311787499298029,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004838568200289834,
      "loss": 5.2778,
      "mean_token_accuracy": 0.1877224028110504,
      "num_tokens": 26045068.0,
      "step": 11680
    },
    {
      "entropy": 5.472187900543213,
      "epoch": 1.3123490762059864,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004838418833646875,
      "loss": 5.4003,
      "mean_token_accuracy": 0.18189065009355546,
      "num_tokens": 26057400.0,
      "step": 11685
    },
    {
      "entropy": 5.425566816329956,
      "epoch": 1.3129106531139438,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048382694005078485,
      "loss": 5.3053,
      "mean_token_accuracy": 0.19235040247440338,
      "num_tokens": 26068981.0,
      "step": 11690
    },
    {
      "entropy": 5.536303472518921,
      "epoch": 1.3134722300219015,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048381199008775165,
      "loss": 5.2673,
      "mean_token_accuracy": 0.197248113155365,
      "num_tokens": 26078902.0,
      "step": 11695
    },
    {
      "entropy": 5.359229850769043,
      "epoch": 1.3140338069298592,
      "grad_norm": 1.25,
      "learning_rate": 0.0004837970334760635,
      "loss": 5.1565,
      "mean_token_accuracy": 0.19453386813402176,
      "num_tokens": 26089891.0,
      "step": 11700
    },
    {
      "entropy": 5.372521781921387,
      "epoch": 1.3145953838378166,
      "grad_norm": 1.171875,
      "learning_rate": 0.000483782070216197,
      "loss": 5.2831,
      "mean_token_accuracy": 0.19459430426359176,
      "num_tokens": 26100701.0,
      "step": 11705
    },
    {
      "entropy": 5.466365098953247,
      "epoch": 1.315156960745774,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004837671003086283,
      "loss": 5.2748,
      "mean_token_accuracy": 0.19886948466300963,
      "num_tokens": 26110496.0,
      "step": 11710
    },
    {
      "entropy": 5.417109537124634,
      "epoch": 1.3157185376537317,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004837521237538342,
      "loss": 5.2507,
      "mean_token_accuracy": 0.19624001234769822,
      "num_tokens": 26120580.0,
      "step": 11715
    },
    {
      "entropy": 5.426248693466187,
      "epoch": 1.3162801145616891,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004837371405522915,
      "loss": 5.3249,
      "mean_token_accuracy": 0.19427521526813507,
      "num_tokens": 26130936.0,
      "step": 11720
    },
    {
      "entropy": 5.5378927230834964,
      "epoch": 1.3168416914696468,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048372215070447734,
      "loss": 5.289,
      "mean_token_accuracy": 0.19295212924480437,
      "num_tokens": 26143027.0,
      "step": 11725
    },
    {
      "entropy": 5.439666032791138,
      "epoch": 1.3174032683776042,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048370715421086884,
      "loss": 5.1951,
      "mean_token_accuracy": 0.19331690669059753,
      "num_tokens": 26152874.0,
      "step": 11730
    },
    {
      "entropy": 5.356834793090821,
      "epoch": 1.317964845285562,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004836921510719438,
      "loss": 5.1757,
      "mean_token_accuracy": 0.18876721560955048,
      "num_tokens": 26162781.0,
      "step": 11735
    },
    {
      "entropy": 5.461295080184937,
      "epoch": 1.3185264221935193,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004836771412881795,
      "loss": 5.2651,
      "mean_token_accuracy": 0.19408902674913406,
      "num_tokens": 26173498.0,
      "step": 11740
    },
    {
      "entropy": 5.449552154541015,
      "epoch": 1.319087999101477,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004836621248600542,
      "loss": 5.2693,
      "mean_token_accuracy": 0.1975193127989769,
      "num_tokens": 26185090.0,
      "step": 11745
    },
    {
      "entropy": 5.519197177886963,
      "epoch": 1.3196495760094344,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004836471017880459,
      "loss": 5.4928,
      "mean_token_accuracy": 0.1816697433590889,
      "num_tokens": 26197048.0,
      "step": 11750
    },
    {
      "entropy": 5.496648454666138,
      "epoch": 1.320211152917392,
      "grad_norm": 1.125,
      "learning_rate": 0.0004836320720726329,
      "loss": 5.3186,
      "mean_token_accuracy": 0.19342129677534103,
      "num_tokens": 26208037.0,
      "step": 11755
    },
    {
      "entropy": 5.525182294845581,
      "epoch": 1.3207727298253495,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004836170357142939,
      "loss": 5.3377,
      "mean_token_accuracy": 0.19096971303224564,
      "num_tokens": 26219345.0,
      "step": 11760
    },
    {
      "entropy": 5.49712028503418,
      "epoch": 1.3213343067333072,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004836019927135076,
      "loss": 5.3463,
      "mean_token_accuracy": 0.19398267716169357,
      "num_tokens": 26230299.0,
      "step": 11765
    },
    {
      "entropy": 5.521836042404175,
      "epoch": 1.3218958836412646,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004835869430707529,
      "loss": 5.4122,
      "mean_token_accuracy": 0.1814396783709526,
      "num_tokens": 26241427.0,
      "step": 11770
    },
    {
      "entropy": 5.42009391784668,
      "epoch": 1.3224574605492223,
      "grad_norm": 1.234375,
      "learning_rate": 0.00048357188678650905,
      "loss": 5.3587,
      "mean_token_accuracy": 0.19001218378543855,
      "num_tokens": 26252051.0,
      "step": 11775
    },
    {
      "entropy": 5.476691818237304,
      "epoch": 1.3230190374571797,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004835568238612553,
      "loss": 5.294,
      "mean_token_accuracy": 0.1859893649816513,
      "num_tokens": 26263640.0,
      "step": 11780
    },
    {
      "entropy": 5.624814701080322,
      "epoch": 1.3235806143651372,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048354175429547155,
      "loss": 5.4727,
      "mean_token_accuracy": 0.18206336945295334,
      "num_tokens": 26274996.0,
      "step": 11785
    },
    {
      "entropy": 5.594439601898193,
      "epoch": 1.3241421912730948,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004835266780896374,
      "loss": 5.4336,
      "mean_token_accuracy": 0.1842691108584404,
      "num_tokens": 26287859.0,
      "step": 11790
    },
    {
      "entropy": 5.493433809280395,
      "epoch": 1.3247037681810525,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004835115952442329,
      "loss": 5.2312,
      "mean_token_accuracy": 0.19297788441181182,
      "num_tokens": 26298302.0,
      "step": 11795
    },
    {
      "entropy": 5.473080444335937,
      "epoch": 1.32526534508901,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048349650575973843,
      "loss": 5.3369,
      "mean_token_accuracy": 0.19186369180679322,
      "num_tokens": 26310755.0,
      "step": 11800
    },
    {
      "entropy": 5.517804718017578,
      "epoch": 1.3258269219969674,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00048348140963663436,
      "loss": 5.4144,
      "mean_token_accuracy": 0.18497879207134246,
      "num_tokens": 26322041.0,
      "step": 11805
    },
    {
      "entropy": 5.534500217437744,
      "epoch": 1.326388498904925,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004834663068754012,
      "loss": 5.3455,
      "mean_token_accuracy": 0.1901350200176239,
      "num_tokens": 26334028.0,
      "step": 11810
    },
    {
      "entropy": 5.436474180221557,
      "epoch": 1.3269500758128827,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004834511974765202,
      "loss": 5.2805,
      "mean_token_accuracy": 0.19087714701890945,
      "num_tokens": 26344834.0,
      "step": 11815
    },
    {
      "entropy": 5.422639179229736,
      "epoch": 1.3275116527208402,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048343608144047194,
      "loss": 5.2913,
      "mean_token_accuracy": 0.18985347002744674,
      "num_tokens": 26354896.0,
      "step": 11820
    },
    {
      "entropy": 5.504586982727051,
      "epoch": 1.3280732296287976,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048342095876773825,
      "loss": 5.3563,
      "mean_token_accuracy": 0.18516043871641158,
      "num_tokens": 26367487.0,
      "step": 11825
    },
    {
      "entropy": 5.486867046356201,
      "epoch": 1.3286348065367553,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004834058294588003,
      "loss": 5.3569,
      "mean_token_accuracy": 0.18751164823770522,
      "num_tokens": 26379384.0,
      "step": 11830
    },
    {
      "entropy": 5.501225280761719,
      "epoch": 1.3291963834447127,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048339069351413993,
      "loss": 5.3101,
      "mean_token_accuracy": 0.1915382444858551,
      "num_tokens": 26390268.0,
      "step": 11835
    },
    {
      "entropy": 5.4992982864379885,
      "epoch": 1.3297579603526704,
      "grad_norm": 1.234375,
      "learning_rate": 0.000483375550934239,
      "loss": 5.4126,
      "mean_token_accuracy": 0.18186776340007782,
      "num_tokens": 26401348.0,
      "step": 11840
    },
    {
      "entropy": 5.435050916671753,
      "epoch": 1.3303195372606278,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004833604017195798,
      "loss": 5.2708,
      "mean_token_accuracy": 0.18476839512586593,
      "num_tokens": 26411955.0,
      "step": 11845
    },
    {
      "entropy": 5.410328388214111,
      "epoch": 1.3308811141685855,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004833452458706445,
      "loss": 5.2221,
      "mean_token_accuracy": 0.19947464317083358,
      "num_tokens": 26422200.0,
      "step": 11850
    },
    {
      "entropy": 5.47580304145813,
      "epoch": 1.331442691076543,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004833300833879159,
      "loss": 5.2355,
      "mean_token_accuracy": 0.19590162634849548,
      "num_tokens": 26433292.0,
      "step": 11855
    },
    {
      "entropy": 5.460307598114014,
      "epoch": 1.3320042679845006,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004833149142718765,
      "loss": 5.2674,
      "mean_token_accuracy": 0.19321906566619873,
      "num_tokens": 26444003.0,
      "step": 11860
    },
    {
      "entropy": 5.374373960494995,
      "epoch": 1.332565844892458,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004832997385230096,
      "loss": 5.27,
      "mean_token_accuracy": 0.2020948961377144,
      "num_tokens": 26455743.0,
      "step": 11865
    },
    {
      "entropy": 5.390793132781982,
      "epoch": 1.3331274218004157,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004832845561417981,
      "loss": 5.2635,
      "mean_token_accuracy": 0.18586072474718093,
      "num_tokens": 26466032.0,
      "step": 11870
    },
    {
      "entropy": 5.476644802093506,
      "epoch": 1.333688998708373,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004832693671287256,
      "loss": 5.2807,
      "mean_token_accuracy": 0.18713297843933105,
      "num_tokens": 26477660.0,
      "step": 11875
    },
    {
      "entropy": 5.407657337188721,
      "epoch": 1.3342505756163305,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004832541714842757,
      "loss": 5.2711,
      "mean_token_accuracy": 0.19107944667339324,
      "num_tokens": 26488778.0,
      "step": 11880
    },
    {
      "entropy": 5.527314186096191,
      "epoch": 1.3348121525242882,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004832389692089321,
      "loss": 5.3342,
      "mean_token_accuracy": 0.18387504518032075,
      "num_tokens": 26500771.0,
      "step": 11885
    },
    {
      "entropy": 5.422005271911621,
      "epoch": 1.3353737294322459,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048322376030317905,
      "loss": 5.2518,
      "mean_token_accuracy": 0.19556079506874086,
      "num_tokens": 26511786.0,
      "step": 11890
    },
    {
      "entropy": 5.409602928161621,
      "epoch": 1.3359353063402033,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004832085447675007,
      "loss": 5.3237,
      "mean_token_accuracy": 0.19361615180969238,
      "num_tokens": 26522273.0,
      "step": 11895
    },
    {
      "entropy": 5.576294946670532,
      "epoch": 1.3364968832481607,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048319332260238153,
      "loss": 5.5939,
      "mean_token_accuracy": 0.17889005988836287,
      "num_tokens": 26536619.0,
      "step": 11900
    },
    {
      "entropy": 5.607589912414551,
      "epoch": 1.3370584601561184,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048317809380830624,
      "loss": 5.3507,
      "mean_token_accuracy": 0.18198300749063492,
      "num_tokens": 26547807.0,
      "step": 11905
    },
    {
      "entropy": 5.525276327133179,
      "epoch": 1.337620037064076,
      "grad_norm": 1.125,
      "learning_rate": 0.0004831628583857597,
      "loss": 5.3465,
      "mean_token_accuracy": 0.18411613553762435,
      "num_tokens": 26558924.0,
      "step": 11910
    },
    {
      "entropy": 5.454687023162842,
      "epoch": 1.3381816139720335,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048314761633522696,
      "loss": 5.2938,
      "mean_token_accuracy": 0.19262010604143143,
      "num_tokens": 26570123.0,
      "step": 11915
    },
    {
      "entropy": 5.491355514526367,
      "epoch": 1.338743190879991,
      "grad_norm": 1.125,
      "learning_rate": 0.0004831323676571934,
      "loss": 5.3691,
      "mean_token_accuracy": 0.18942122161388397,
      "num_tokens": 26582841.0,
      "step": 11920
    },
    {
      "entropy": 5.435934686660767,
      "epoch": 1.3393047677879486,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004831171123521446,
      "loss": 5.2789,
      "mean_token_accuracy": 0.1932838350534439,
      "num_tokens": 26594397.0,
      "step": 11925
    },
    {
      "entropy": 5.465637588500977,
      "epoch": 1.339866344695906,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004831018504205662,
      "loss": 5.3562,
      "mean_token_accuracy": 0.18684175461530686,
      "num_tokens": 26605055.0,
      "step": 11930
    },
    {
      "entropy": 5.540536403656006,
      "epoch": 1.3404279216038637,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048308658186294414,
      "loss": 5.4042,
      "mean_token_accuracy": 0.17936392426490783,
      "num_tokens": 26616110.0,
      "step": 11935
    },
    {
      "entropy": 5.528040027618408,
      "epoch": 1.3409894985118211,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004830713066797646,
      "loss": 5.2711,
      "mean_token_accuracy": 0.19870582073926926,
      "num_tokens": 26627196.0,
      "step": 11940
    },
    {
      "entropy": 5.514052104949951,
      "epoch": 1.3415510754197788,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048305602487151396,
      "loss": 5.3798,
      "mean_token_accuracy": 0.1901874288916588,
      "num_tokens": 26637962.0,
      "step": 11945
    },
    {
      "entropy": 5.421580171585083,
      "epoch": 1.3421126523277362,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004830407364386789,
      "loss": 5.2867,
      "mean_token_accuracy": 0.18893011808395385,
      "num_tokens": 26649366.0,
      "step": 11950
    },
    {
      "entropy": 5.631961393356323,
      "epoch": 1.342674229235694,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000483025441381746,
      "loss": 5.4868,
      "mean_token_accuracy": 0.1802490994334221,
      "num_tokens": 26662455.0,
      "step": 11955
    },
    {
      "entropy": 5.492560863494873,
      "epoch": 1.3432358061436513,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048301013970120235,
      "loss": 5.3668,
      "mean_token_accuracy": 0.1894997239112854,
      "num_tokens": 26673467.0,
      "step": 11960
    },
    {
      "entropy": 5.419187450408936,
      "epoch": 1.343797383051609,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048299483139753517,
      "loss": 5.2257,
      "mean_token_accuracy": 0.19564353823661804,
      "num_tokens": 26684803.0,
      "step": 11965
    },
    {
      "entropy": 5.463820362091065,
      "epoch": 1.3443589599595664,
      "grad_norm": 2.5,
      "learning_rate": 0.0004829795164712319,
      "loss": 5.3344,
      "mean_token_accuracy": 0.18733816742897033,
      "num_tokens": 26696886.0,
      "step": 11970
    },
    {
      "entropy": 5.307031631469727,
      "epoch": 1.3449205368675239,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048296419492278023,
      "loss": 5.1659,
      "mean_token_accuracy": 0.20012296736240387,
      "num_tokens": 26708489.0,
      "step": 11975
    },
    {
      "entropy": 5.502926206588745,
      "epoch": 1.3454821137754815,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004829488667526679,
      "loss": 5.3166,
      "mean_token_accuracy": 0.18866093903779985,
      "num_tokens": 26719599.0,
      "step": 11980
    },
    {
      "entropy": 5.435163879394532,
      "epoch": 1.3460436906834392,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048293353196138294,
      "loss": 5.2664,
      "mean_token_accuracy": 0.19601985365152358,
      "num_tokens": 26730359.0,
      "step": 11985
    },
    {
      "entropy": 5.511836194992066,
      "epoch": 1.3466052675913966,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048291819054941375,
      "loss": 5.4475,
      "mean_token_accuracy": 0.17989581674337388,
      "num_tokens": 26742363.0,
      "step": 11990
    },
    {
      "entropy": 5.54780797958374,
      "epoch": 1.347166844499354,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004829028425172487,
      "loss": 5.4193,
      "mean_token_accuracy": 0.18575457781553267,
      "num_tokens": 26754058.0,
      "step": 11995
    },
    {
      "entropy": 5.486512088775635,
      "epoch": 1.3477284214073117,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048288748786537656,
      "loss": 5.3356,
      "mean_token_accuracy": 0.1865041434764862,
      "num_tokens": 26765602.0,
      "step": 12000
    },
    {
      "epoch": 1.3477284214073117,
      "eval_entropy": 5.366885109110102,
      "eval_loss": 5.411596775054932,
      "eval_mean_token_accuracy": 0.19360563530557112,
      "eval_num_tokens": 26765602.0,
      "eval_runtime": 23.9774,
      "eval_samples_per_second": 1293.426,
      "eval_steps_per_second": 161.694,
      "step": 12000
    },
    {
      "entropy": 5.464175891876221,
      "epoch": 1.3482899983152694,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004828721265942861,
      "loss": 5.2517,
      "mean_token_accuracy": 0.19497989416122435,
      "num_tokens": 26776867.0,
      "step": 12005
    },
    {
      "entropy": 5.426329565048218,
      "epoch": 1.3488515752232269,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048285675870446653,
      "loss": 5.2493,
      "mean_token_accuracy": 0.19001496583223343,
      "num_tokens": 26786383.0,
      "step": 12010
    },
    {
      "entropy": 5.554459953308106,
      "epoch": 1.3494131521311843,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048284138419640724,
      "loss": 5.393,
      "mean_token_accuracy": 0.1808932304382324,
      "num_tokens": 26797647.0,
      "step": 12015
    },
    {
      "entropy": 5.515417289733887,
      "epoch": 1.349974729039142,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004828260030705976,
      "loss": 5.2522,
      "mean_token_accuracy": 0.18739502280950546,
      "num_tokens": 26808674.0,
      "step": 12020
    },
    {
      "entropy": 5.400568151473999,
      "epoch": 1.3505363059470994,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048281061532752734,
      "loss": 5.1614,
      "mean_token_accuracy": 0.1933847740292549,
      "num_tokens": 26820858.0,
      "step": 12025
    },
    {
      "entropy": 5.4576746940612795,
      "epoch": 1.351097882855057,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004827952209676866,
      "loss": 5.2393,
      "mean_token_accuracy": 0.19495184272527694,
      "num_tokens": 26831432.0,
      "step": 12030
    },
    {
      "entropy": 5.484373235702515,
      "epoch": 1.3516594597630145,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048277981999156534,
      "loss": 5.3615,
      "mean_token_accuracy": 0.19081514924764634,
      "num_tokens": 26843240.0,
      "step": 12035
    },
    {
      "entropy": 5.5573502540588375,
      "epoch": 1.3522210366709722,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048276441239965414,
      "loss": 5.334,
      "mean_token_accuracy": 0.188915453851223,
      "num_tokens": 26853556.0,
      "step": 12040
    },
    {
      "entropy": 5.427993631362915,
      "epoch": 1.3527826135789296,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004827489981924434,
      "loss": 5.2944,
      "mean_token_accuracy": 0.19146947413682938,
      "num_tokens": 26866040.0,
      "step": 12045
    },
    {
      "entropy": 5.347304010391236,
      "epoch": 1.3533441904868873,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048273357737042393,
      "loss": 5.1493,
      "mean_token_accuracy": 0.19735122770071029,
      "num_tokens": 26876796.0,
      "step": 12050
    },
    {
      "entropy": 5.439880847930908,
      "epoch": 1.3539057673948447,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048271814993408694,
      "loss": 5.3336,
      "mean_token_accuracy": 0.18864540457725526,
      "num_tokens": 26888081.0,
      "step": 12055
    },
    {
      "entropy": 5.490010356903076,
      "epoch": 1.3544673443028024,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004827027158839233,
      "loss": 5.3209,
      "mean_token_accuracy": 0.18834337145090102,
      "num_tokens": 26898511.0,
      "step": 12060
    },
    {
      "entropy": 5.528523969650268,
      "epoch": 1.3550289212107598,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004826872752204248,
      "loss": 5.3352,
      "mean_token_accuracy": 0.19045519083738327,
      "num_tokens": 26909525.0,
      "step": 12065
    },
    {
      "entropy": 5.5093615531921385,
      "epoch": 1.3555904981187172,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00048267182794408276,
      "loss": 5.3544,
      "mean_token_accuracy": 0.19533660709857942,
      "num_tokens": 26920267.0,
      "step": 12070
    },
    {
      "entropy": 5.463819980621338,
      "epoch": 1.356152075026675,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004826563740553893,
      "loss": 5.3115,
      "mean_token_accuracy": 0.1902007758617401,
      "num_tokens": 26931351.0,
      "step": 12075
    },
    {
      "entropy": 5.3886007308959964,
      "epoch": 1.3567136519346326,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004826409135548363,
      "loss": 5.2617,
      "mean_token_accuracy": 0.1883338376879692,
      "num_tokens": 26942235.0,
      "step": 12080
    },
    {
      "entropy": 5.562569570541382,
      "epoch": 1.35727522884259,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004826254464429161,
      "loss": 5.4084,
      "mean_token_accuracy": 0.18839237987995147,
      "num_tokens": 26953375.0,
      "step": 12085
    },
    {
      "entropy": 5.518659687042236,
      "epoch": 1.3578368057505474,
      "grad_norm": 1.25,
      "learning_rate": 0.00048260997272012114,
      "loss": 5.3193,
      "mean_token_accuracy": 0.1958278387784958,
      "num_tokens": 26965081.0,
      "step": 12090
    },
    {
      "entropy": 5.4991106510162355,
      "epoch": 1.358398382658505,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004825944923869441,
      "loss": 5.3514,
      "mean_token_accuracy": 0.18456209003925322,
      "num_tokens": 26975428.0,
      "step": 12095
    },
    {
      "entropy": 5.461251592636108,
      "epoch": 1.3589599595664628,
      "grad_norm": 1.125,
      "learning_rate": 0.00048257900544387795,
      "loss": 5.2554,
      "mean_token_accuracy": 0.19516672044992447,
      "num_tokens": 26987040.0,
      "step": 12100
    },
    {
      "entropy": 5.497143173217774,
      "epoch": 1.3595215364744202,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048256351189141574,
      "loss": 5.3709,
      "mean_token_accuracy": 0.18420893847942352,
      "num_tokens": 26997931.0,
      "step": 12105
    },
    {
      "entropy": 5.507317352294922,
      "epoch": 1.3600831133823776,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004825480117300508,
      "loss": 5.2712,
      "mean_token_accuracy": 0.19465894401073455,
      "num_tokens": 27009528.0,
      "step": 12110
    },
    {
      "entropy": 5.390457105636597,
      "epoch": 1.3606446902903353,
      "grad_norm": 1.234375,
      "learning_rate": 0.00048253250496027675,
      "loss": 5.1769,
      "mean_token_accuracy": 0.2018658548593521,
      "num_tokens": 27020161.0,
      "step": 12115
    },
    {
      "entropy": 5.474413681030273,
      "epoch": 1.3612062671982927,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048251699158258714,
      "loss": 5.2678,
      "mean_token_accuracy": 0.19033357203006745,
      "num_tokens": 27032246.0,
      "step": 12120
    },
    {
      "entropy": 5.47997670173645,
      "epoch": 1.3617678441062504,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004825014715974761,
      "loss": 5.359,
      "mean_token_accuracy": 0.18714495301246642,
      "num_tokens": 27042903.0,
      "step": 12125
    },
    {
      "entropy": 5.5505718231201175,
      "epoch": 1.3623294210142078,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048248594500543767,
      "loss": 5.4628,
      "mean_token_accuracy": 0.1821836918592453,
      "num_tokens": 27054854.0,
      "step": 12130
    },
    {
      "entropy": 5.51246395111084,
      "epoch": 1.3628909979221655,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048247041180696633,
      "loss": 5.2849,
      "mean_token_accuracy": 0.1972144514322281,
      "num_tokens": 27065021.0,
      "step": 12135
    },
    {
      "entropy": 5.438337898254394,
      "epoch": 1.363452574830123,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004824548720025566,
      "loss": 5.3348,
      "mean_token_accuracy": 0.18295012414455414,
      "num_tokens": 27075717.0,
      "step": 12140
    },
    {
      "entropy": 5.532647037506104,
      "epoch": 1.3640141517380806,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004824393255927033,
      "loss": 5.4403,
      "mean_token_accuracy": 0.1860695704817772,
      "num_tokens": 27087310.0,
      "step": 12145
    },
    {
      "entropy": 5.5660223960876465,
      "epoch": 1.364575728646038,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048242377257790145,
      "loss": 5.425,
      "mean_token_accuracy": 0.1824367955327034,
      "num_tokens": 27100218.0,
      "step": 12150
    },
    {
      "entropy": 5.395139932632446,
      "epoch": 1.3651373055539957,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048240821295864614,
      "loss": 5.2062,
      "mean_token_accuracy": 0.19678564220666886,
      "num_tokens": 27111548.0,
      "step": 12155
    },
    {
      "entropy": 5.467675399780274,
      "epoch": 1.3656988824619531,
      "grad_norm": 1.359375,
      "learning_rate": 0.000482392646735433,
      "loss": 5.3577,
      "mean_token_accuracy": 0.18890367001295089,
      "num_tokens": 27123434.0,
      "step": 12160
    },
    {
      "entropy": 5.49817795753479,
      "epoch": 1.3662604593699106,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004823770739087574,
      "loss": 5.3768,
      "mean_token_accuracy": 0.18898044377565384,
      "num_tokens": 27135936.0,
      "step": 12165
    },
    {
      "entropy": 5.61887845993042,
      "epoch": 1.3668220362778682,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004823614944791154,
      "loss": 5.531,
      "mean_token_accuracy": 0.17838797271251677,
      "num_tokens": 27148240.0,
      "step": 12170
    },
    {
      "entropy": 5.509736824035644,
      "epoch": 1.367383613185826,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004823459084470031,
      "loss": 5.291,
      "mean_token_accuracy": 0.1909016564488411,
      "num_tokens": 27159057.0,
      "step": 12175
    },
    {
      "entropy": 5.439469289779663,
      "epoch": 1.3679451900937833,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004823303158129165,
      "loss": 5.2831,
      "mean_token_accuracy": 0.1931535556912422,
      "num_tokens": 27169189.0,
      "step": 12180
    },
    {
      "entropy": 5.429803991317749,
      "epoch": 1.3685067670017408,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004823147165773523,
      "loss": 5.2415,
      "mean_token_accuracy": 0.1967167466878891,
      "num_tokens": 27180777.0,
      "step": 12185
    },
    {
      "entropy": 5.436133003234863,
      "epoch": 1.3690683439096984,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048229911074080715,
      "loss": 5.325,
      "mean_token_accuracy": 0.2004435569047928,
      "num_tokens": 27190991.0,
      "step": 12190
    },
    {
      "entropy": 5.476318645477295,
      "epoch": 1.369629920817656,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048228349830377787,
      "loss": 5.324,
      "mean_token_accuracy": 0.19127963334321976,
      "num_tokens": 27203776.0,
      "step": 12195
    },
    {
      "entropy": 5.415596342086792,
      "epoch": 1.3701914977256135,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048226787926676165,
      "loss": 5.2447,
      "mean_token_accuracy": 0.19237148761749268,
      "num_tokens": 27213533.0,
      "step": 12200
    },
    {
      "entropy": 5.471880578994751,
      "epoch": 1.370753074633571,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048225225363025576,
      "loss": 5.382,
      "mean_token_accuracy": 0.18520113080739975,
      "num_tokens": 27224832.0,
      "step": 12205
    },
    {
      "entropy": 5.52721848487854,
      "epoch": 1.3713146515415287,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048223662139475773,
      "loss": 5.4064,
      "mean_token_accuracy": 0.18114156872034073,
      "num_tokens": 27237313.0,
      "step": 12210
    },
    {
      "entropy": 5.457478857040405,
      "epoch": 1.371876228449486,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048222098256076523,
      "loss": 5.2093,
      "mean_token_accuracy": 0.19222808480262757,
      "num_tokens": 27248159.0,
      "step": 12215
    },
    {
      "entropy": 5.4940704822540285,
      "epoch": 1.3724378053574438,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004822053371287763,
      "loss": 5.3566,
      "mean_token_accuracy": 0.18893896639347077,
      "num_tokens": 27259723.0,
      "step": 12220
    },
    {
      "entropy": 5.481373929977417,
      "epoch": 1.3729993822654012,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004821896850992891,
      "loss": 5.3087,
      "mean_token_accuracy": 0.18606701046228408,
      "num_tokens": 27270207.0,
      "step": 12225
    },
    {
      "entropy": 5.458189535140991,
      "epoch": 1.3735609591733589,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000482174026472802,
      "loss": 5.3403,
      "mean_token_accuracy": 0.19502801895141603,
      "num_tokens": 27281243.0,
      "step": 12230
    },
    {
      "entropy": 5.3786454677581785,
      "epoch": 1.3741225360813163,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004821583612498135,
      "loss": 5.1567,
      "mean_token_accuracy": 0.19473931342363357,
      "num_tokens": 27292816.0,
      "step": 12235
    },
    {
      "entropy": 5.418706321716309,
      "epoch": 1.374684112989274,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048214268943082246,
      "loss": 5.3185,
      "mean_token_accuracy": 0.18923725336790084,
      "num_tokens": 27304254.0,
      "step": 12240
    },
    {
      "entropy": 5.406972169876099,
      "epoch": 1.3752456898972314,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004821270110163278,
      "loss": 5.2818,
      "mean_token_accuracy": 0.18516048043966293,
      "num_tokens": 27314974.0,
      "step": 12245
    },
    {
      "entropy": 5.488144016265869,
      "epoch": 1.375807266805189,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004821113260068288,
      "loss": 5.3508,
      "mean_token_accuracy": 0.18806078881025315,
      "num_tokens": 27326051.0,
      "step": 12250
    },
    {
      "entropy": 5.416076278686523,
      "epoch": 1.3763688437131465,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004820956344028249,
      "loss": 5.2189,
      "mean_token_accuracy": 0.19107295870780944,
      "num_tokens": 27337589.0,
      "step": 12255
    },
    {
      "entropy": 5.430997037887574,
      "epoch": 1.376930420621104,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048207993620481566,
      "loss": 5.2922,
      "mean_token_accuracy": 0.1930110588669777,
      "num_tokens": 27348581.0,
      "step": 12260
    },
    {
      "entropy": 5.530972099304199,
      "epoch": 1.3774919975290616,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004820642314133009,
      "loss": 5.3753,
      "mean_token_accuracy": 0.19005978852510452,
      "num_tokens": 27359963.0,
      "step": 12265
    },
    {
      "entropy": 5.495496320724487,
      "epoch": 1.3780535744370193,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048204852002878073,
      "loss": 5.2489,
      "mean_token_accuracy": 0.1893151134252548,
      "num_tokens": 27370149.0,
      "step": 12270
    },
    {
      "entropy": 5.36246771812439,
      "epoch": 1.3786151513449767,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048203280205175544,
      "loss": 5.2461,
      "mean_token_accuracy": 0.19967357516288758,
      "num_tokens": 27381657.0,
      "step": 12275
    },
    {
      "entropy": 5.473087978363037,
      "epoch": 1.3791767282529341,
      "grad_norm": 1.25,
      "learning_rate": 0.0004820170774827253,
      "loss": 5.2942,
      "mean_token_accuracy": 0.18093011975288392,
      "num_tokens": 27392679.0,
      "step": 12280
    },
    {
      "entropy": 5.4467456340789795,
      "epoch": 1.3797383051608918,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004820013463221912,
      "loss": 5.2891,
      "mean_token_accuracy": 0.19028362333774568,
      "num_tokens": 27404003.0,
      "step": 12285
    },
    {
      "entropy": 5.5077837944030765,
      "epoch": 1.3802998820688495,
      "grad_norm": 1.078125,
      "learning_rate": 0.000481985608570654,
      "loss": 5.3069,
      "mean_token_accuracy": 0.1885668084025383,
      "num_tokens": 27414593.0,
      "step": 12290
    },
    {
      "entropy": 5.514311599731445,
      "epoch": 1.380861458976807,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004819698642286147,
      "loss": 5.446,
      "mean_token_accuracy": 0.1852269247174263,
      "num_tokens": 27426549.0,
      "step": 12295
    },
    {
      "entropy": 5.492937755584717,
      "epoch": 1.3814230358847643,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048195411329657464,
      "loss": 5.3935,
      "mean_token_accuracy": 0.18793376684188842,
      "num_tokens": 27439451.0,
      "step": 12300
    },
    {
      "entropy": 5.547123670578003,
      "epoch": 1.381984612792722,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004819383557750353,
      "loss": 5.3255,
      "mean_token_accuracy": 0.18728768676519394,
      "num_tokens": 27451153.0,
      "step": 12305
    },
    {
      "entropy": 5.453478622436523,
      "epoch": 1.3825461897006794,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048192259166449845,
      "loss": 5.2742,
      "mean_token_accuracy": 0.19076059013605118,
      "num_tokens": 27462229.0,
      "step": 12310
    },
    {
      "entropy": 5.421104669570923,
      "epoch": 1.383107766608637,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048190682096546606,
      "loss": 5.2641,
      "mean_token_accuracy": 0.19198974668979646,
      "num_tokens": 27473901.0,
      "step": 12315
    },
    {
      "entropy": 5.398230075836182,
      "epoch": 1.3836693435165945,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048189104367844034,
      "loss": 5.2707,
      "mean_token_accuracy": 0.1900333821773529,
      "num_tokens": 27485981.0,
      "step": 12320
    },
    {
      "entropy": 5.420984888076783,
      "epoch": 1.3842309204245522,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048187525980392343,
      "loss": 5.2015,
      "mean_token_accuracy": 0.19466772079467773,
      "num_tokens": 27495764.0,
      "step": 12325
    },
    {
      "entropy": 5.4703703880310055,
      "epoch": 1.3847924973325096,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000481859469342418,
      "loss": 5.2933,
      "mean_token_accuracy": 0.19381141662597656,
      "num_tokens": 27506602.0,
      "step": 12330
    },
    {
      "entropy": 5.378237676620484,
      "epoch": 1.3853540742404673,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004818436722944269,
      "loss": 5.2417,
      "mean_token_accuracy": 0.19462836384773255,
      "num_tokens": 27517482.0,
      "step": 12335
    },
    {
      "entropy": 5.458483505249023,
      "epoch": 1.3859156511484247,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004818278686604531,
      "loss": 5.3006,
      "mean_token_accuracy": 0.19507773965597153,
      "num_tokens": 27528560.0,
      "step": 12340
    },
    {
      "entropy": 5.472741317749024,
      "epoch": 1.3864772280563824,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004818120584409996,
      "loss": 5.2895,
      "mean_token_accuracy": 0.19037635624408722,
      "num_tokens": 27540019.0,
      "step": 12345
    },
    {
      "entropy": 5.508252429962158,
      "epoch": 1.3870388049643398,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004817962416365701,
      "loss": 5.3444,
      "mean_token_accuracy": 0.19066003412008287,
      "num_tokens": 27550553.0,
      "step": 12350
    },
    {
      "entropy": 5.488602590560913,
      "epoch": 1.3876003818722973,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048178041824766796,
      "loss": 5.2698,
      "mean_token_accuracy": 0.19139335602521895,
      "num_tokens": 27562169.0,
      "step": 12355
    },
    {
      "entropy": 5.4327497482299805,
      "epoch": 1.388161958780255,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004817645882747971,
      "loss": 5.2314,
      "mean_token_accuracy": 0.19290353655815123,
      "num_tokens": 27571732.0,
      "step": 12360
    },
    {
      "entropy": 5.386249685287476,
      "epoch": 1.3887235356882126,
      "grad_norm": 1.25,
      "learning_rate": 0.0004817487517184616,
      "loss": 5.2559,
      "mean_token_accuracy": 0.19271651059389114,
      "num_tokens": 27584394.0,
      "step": 12365
    },
    {
      "entropy": 5.395462465286255,
      "epoch": 1.38928511259617,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004817329085791657,
      "loss": 5.2561,
      "mean_token_accuracy": 0.2011577770113945,
      "num_tokens": 27594813.0,
      "step": 12370
    },
    {
      "entropy": 5.394870567321777,
      "epoch": 1.3898466895041275,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004817170588574138,
      "loss": 5.2045,
      "mean_token_accuracy": 0.20346917659044267,
      "num_tokens": 27605375.0,
      "step": 12375
    },
    {
      "entropy": 5.415407371520996,
      "epoch": 1.3904082664120851,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004817012025537105,
      "loss": 5.2268,
      "mean_token_accuracy": 0.19692109525203705,
      "num_tokens": 27617066.0,
      "step": 12380
    },
    {
      "entropy": 5.5027923583984375,
      "epoch": 1.3909698433200428,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004816853396685608,
      "loss": 5.3627,
      "mean_token_accuracy": 0.19065181016921998,
      "num_tokens": 27628667.0,
      "step": 12385
    },
    {
      "entropy": 5.457526826858521,
      "epoch": 1.3915314202280002,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048166947020246973,
      "loss": 5.3492,
      "mean_token_accuracy": 0.1884267285466194,
      "num_tokens": 27638858.0,
      "step": 12390
    },
    {
      "entropy": 5.541528415679932,
      "epoch": 1.3920929971359577,
      "grad_norm": 1.234375,
      "learning_rate": 0.00048165359415594245,
      "loss": 5.3937,
      "mean_token_accuracy": 0.19223472028970717,
      "num_tokens": 27650518.0,
      "step": 12395
    },
    {
      "entropy": 5.4594865322113035,
      "epoch": 1.3926545740439153,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048163771152948473,
      "loss": 5.3342,
      "mean_token_accuracy": 0.18858733475208284,
      "num_tokens": 27661781.0,
      "step": 12400
    },
    {
      "entropy": 5.421994209289551,
      "epoch": 1.3932161509518728,
      "grad_norm": 1.1875,
      "learning_rate": 0.000481621822323602,
      "loss": 5.2889,
      "mean_token_accuracy": 0.19110580533742905,
      "num_tokens": 27673040.0,
      "step": 12405
    },
    {
      "entropy": 5.48698992729187,
      "epoch": 1.3937777278598305,
      "grad_norm": 1.375,
      "learning_rate": 0.0004816059265388004,
      "loss": 5.3465,
      "mean_token_accuracy": 0.1948383167386055,
      "num_tokens": 27683541.0,
      "step": 12410
    },
    {
      "entropy": 5.555398225784302,
      "epoch": 1.394339304767788,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00048159002417558596,
      "loss": 5.3775,
      "mean_token_accuracy": 0.1871667191386223,
      "num_tokens": 27693648.0,
      "step": 12415
    },
    {
      "entropy": 5.446379566192627,
      "epoch": 1.3949008816757456,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004815741152344651,
      "loss": 5.2253,
      "mean_token_accuracy": 0.19445260912179946,
      "num_tokens": 27703920.0,
      "step": 12420
    },
    {
      "entropy": 5.557613277435303,
      "epoch": 1.395462458583703,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004815581997159442,
      "loss": 5.3345,
      "mean_token_accuracy": 0.1813591465353966,
      "num_tokens": 27715869.0,
      "step": 12425
    },
    {
      "entropy": 5.4196161270141605,
      "epoch": 1.3960240354916607,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048154227762053003,
      "loss": 5.2884,
      "mean_token_accuracy": 0.19719033688306808,
      "num_tokens": 27726727.0,
      "step": 12430
    },
    {
      "entropy": 5.451911735534668,
      "epoch": 1.396585612399618,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004815263489487298,
      "loss": 5.2821,
      "mean_token_accuracy": 0.19062285423278807,
      "num_tokens": 27738217.0,
      "step": 12435
    },
    {
      "entropy": 5.484431171417237,
      "epoch": 1.3971471893075758,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004815104137010504,
      "loss": 5.2725,
      "mean_token_accuracy": 0.1906019404530525,
      "num_tokens": 27749022.0,
      "step": 12440
    },
    {
      "entropy": 5.406584787368774,
      "epoch": 1.3977087662155332,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004814944718779993,
      "loss": 5.2726,
      "mean_token_accuracy": 0.1882222533226013,
      "num_tokens": 27760589.0,
      "step": 12445
    },
    {
      "entropy": 5.576084756851197,
      "epoch": 1.3982703431234906,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004814785234800843,
      "loss": 5.3401,
      "mean_token_accuracy": 0.19314385801553727,
      "num_tokens": 27772460.0,
      "step": 12450
    },
    {
      "entropy": 5.400389575958252,
      "epoch": 1.3988319200314483,
      "grad_norm": 1.125,
      "learning_rate": 0.00048146256850781287,
      "loss": 5.2083,
      "mean_token_accuracy": 0.19437296390533448,
      "num_tokens": 27783921.0,
      "step": 12455
    },
    {
      "entropy": 5.439860486984253,
      "epoch": 1.399393496939406,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048144660696169327,
      "loss": 5.3614,
      "mean_token_accuracy": 0.1894972413778305,
      "num_tokens": 27794843.0,
      "step": 12460
    },
    {
      "entropy": 5.49439754486084,
      "epoch": 1.3999550738473634,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048143063884223365,
      "loss": 5.2451,
      "mean_token_accuracy": 0.1918090671300888,
      "num_tokens": 27805973.0,
      "step": 12465
    },
    {
      "entropy": 5.435964918136596,
      "epoch": 1.4005166507553208,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004814146641499424,
      "loss": 5.3101,
      "mean_token_accuracy": 0.18675086349248887,
      "num_tokens": 27816916.0,
      "step": 12470
    },
    {
      "entropy": 5.439694261550903,
      "epoch": 1.4010782276632785,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048139868288532816,
      "loss": 5.3133,
      "mean_token_accuracy": 0.18654188960790635,
      "num_tokens": 27828077.0,
      "step": 12475
    },
    {
      "entropy": 5.4364197731018065,
      "epoch": 1.4016398045712362,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004813826950488998,
      "loss": 5.1998,
      "mean_token_accuracy": 0.19505369663238525,
      "num_tokens": 27838502.0,
      "step": 12480
    },
    {
      "entropy": 5.465064144134521,
      "epoch": 1.4022013814791936,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048136670064116645,
      "loss": 5.3265,
      "mean_token_accuracy": 0.18985458612442016,
      "num_tokens": 27850125.0,
      "step": 12485
    },
    {
      "entropy": 5.480968284606933,
      "epoch": 1.402762958387151,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004813506996626373,
      "loss": 5.3007,
      "mean_token_accuracy": 0.18942525684833528,
      "num_tokens": 27861099.0,
      "step": 12490
    },
    {
      "entropy": 5.487314748764038,
      "epoch": 1.4033245352951087,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004813346921138218,
      "loss": 5.2812,
      "mean_token_accuracy": 0.19160969704389572,
      "num_tokens": 27872769.0,
      "step": 12495
    },
    {
      "entropy": 5.490968656539917,
      "epoch": 1.4038861122030661,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004813186779952297,
      "loss": 5.3583,
      "mean_token_accuracy": 0.1895208701491356,
      "num_tokens": 27884448.0,
      "step": 12500
    },
    {
      "entropy": 5.409013891220093,
      "epoch": 1.4044476891110238,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004813026573073708,
      "loss": 5.2497,
      "mean_token_accuracy": 0.1837848722934723,
      "num_tokens": 27894074.0,
      "step": 12505
    },
    {
      "entropy": 5.573706340789795,
      "epoch": 1.4050092660189812,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004812866300507554,
      "loss": 5.3933,
      "mean_token_accuracy": 0.19490398913621904,
      "num_tokens": 27905150.0,
      "step": 12510
    },
    {
      "entropy": 5.498510360717773,
      "epoch": 1.405570842926939,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004812705962258936,
      "loss": 5.3173,
      "mean_token_accuracy": 0.18638655543327332,
      "num_tokens": 27916471.0,
      "step": 12515
    },
    {
      "entropy": 5.427494812011719,
      "epoch": 1.4061324198348963,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004812545558332961,
      "loss": 5.2659,
      "mean_token_accuracy": 0.19297430366277696,
      "num_tokens": 27928276.0,
      "step": 12520
    },
    {
      "entropy": 5.4732873916625975,
      "epoch": 1.406693996742854,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048123850887347346,
      "loss": 5.2929,
      "mean_token_accuracy": 0.1838265150785446,
      "num_tokens": 27940417.0,
      "step": 12525
    },
    {
      "entropy": 5.399258708953857,
      "epoch": 1.4072555736508114,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048122245534693676,
      "loss": 5.1699,
      "mean_token_accuracy": 0.1988804817199707,
      "num_tokens": 27950490.0,
      "step": 12530
    },
    {
      "entropy": 5.508321332931518,
      "epoch": 1.407817150558769,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004812063952541971,
      "loss": 5.3699,
      "mean_token_accuracy": 0.1812337040901184,
      "num_tokens": 27962593.0,
      "step": 12535
    },
    {
      "entropy": 5.589501810073853,
      "epoch": 1.4083787274667265,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048119032859576577,
      "loss": 5.4199,
      "mean_token_accuracy": 0.1831713080406189,
      "num_tokens": 27973567.0,
      "step": 12540
    },
    {
      "entropy": 5.417863464355468,
      "epoch": 1.408940304374684,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004811742553721545,
      "loss": 5.219,
      "mean_token_accuracy": 0.19738027155399324,
      "num_tokens": 27984383.0,
      "step": 12545
    },
    {
      "entropy": 5.533534860610962,
      "epoch": 1.4095018812826416,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048115817558387494,
      "loss": 5.3515,
      "mean_token_accuracy": 0.18330678194761277,
      "num_tokens": 27995735.0,
      "step": 12550
    },
    {
      "entropy": 5.501014375686646,
      "epoch": 1.4100634581905993,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004811420892314391,
      "loss": 5.3986,
      "mean_token_accuracy": 0.17909621745347976,
      "num_tokens": 28008261.0,
      "step": 12555
    },
    {
      "entropy": 5.454143714904785,
      "epoch": 1.4106250350985567,
      "grad_norm": 1.25,
      "learning_rate": 0.0004811259963153592,
      "loss": 5.2442,
      "mean_token_accuracy": 0.19303920567035676,
      "num_tokens": 28019424.0,
      "step": 12560
    },
    {
      "entropy": 5.510006999969482,
      "epoch": 1.4111866120065142,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004811098968361476,
      "loss": 5.3722,
      "mean_token_accuracy": 0.19383841902017593,
      "num_tokens": 28030961.0,
      "step": 12565
    },
    {
      "entropy": 5.493874168395996,
      "epoch": 1.4117481889144718,
      "grad_norm": 1.265625,
      "learning_rate": 0.000481093790794317,
      "loss": 5.3286,
      "mean_token_accuracy": 0.18856671154499055,
      "num_tokens": 28042580.0,
      "step": 12570
    },
    {
      "entropy": 5.681679153442383,
      "epoch": 1.4123097658224295,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004810776781903801,
      "loss": 5.4426,
      "mean_token_accuracy": 0.18635336458683013,
      "num_tokens": 28053435.0,
      "step": 12575
    },
    {
      "entropy": 5.52832670211792,
      "epoch": 1.412871342730387,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048106155902485003,
      "loss": 5.4275,
      "mean_token_accuracy": 0.1847854033112526,
      "num_tokens": 28065195.0,
      "step": 12580
    },
    {
      "entropy": 5.4907256126403805,
      "epoch": 1.4134329196383444,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004810454332982401,
      "loss": 5.3726,
      "mean_token_accuracy": 0.18728686422109603,
      "num_tokens": 28077457.0,
      "step": 12585
    },
    {
      "entropy": 5.464640140533447,
      "epoch": 1.413994496546302,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004810293010110635,
      "loss": 5.2525,
      "mean_token_accuracy": 0.192541965842247,
      "num_tokens": 28087598.0,
      "step": 12590
    },
    {
      "entropy": 5.401106595993042,
      "epoch": 1.4145560734542595,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004810131621638342,
      "loss": 5.2265,
      "mean_token_accuracy": 0.19532348960638046,
      "num_tokens": 28097232.0,
      "step": 12595
    },
    {
      "entropy": 5.4576328754425045,
      "epoch": 1.4151176503622171,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048099701675706583,
      "loss": 5.3376,
      "mean_token_accuracy": 0.19434461146593093,
      "num_tokens": 28107333.0,
      "step": 12600
    },
    {
      "entropy": 5.5152403831481935,
      "epoch": 1.4156792272701746,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004809808647912725,
      "loss": 5.3309,
      "mean_token_accuracy": 0.18604649156332015,
      "num_tokens": 28117886.0,
      "step": 12605
    },
    {
      "entropy": 5.528912687301636,
      "epoch": 1.4162408041781323,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004809647062669686,
      "loss": 5.3162,
      "mean_token_accuracy": 0.18966701477766038,
      "num_tokens": 28128884.0,
      "step": 12610
    },
    {
      "entropy": 5.495925760269165,
      "epoch": 1.4168023810860897,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048094854118466856,
      "loss": 5.3934,
      "mean_token_accuracy": 0.18094490319490433,
      "num_tokens": 28139859.0,
      "step": 12615
    },
    {
      "entropy": 5.621742534637451,
      "epoch": 1.4173639579940474,
      "grad_norm": 1.375,
      "learning_rate": 0.00048093236954488715,
      "loss": 5.3316,
      "mean_token_accuracy": 0.1940074384212494,
      "num_tokens": 28150398.0,
      "step": 12620
    },
    {
      "entropy": 5.468520593643189,
      "epoch": 1.4179255349020048,
      "grad_norm": 1.25,
      "learning_rate": 0.0004809161913481392,
      "loss": 5.2265,
      "mean_token_accuracy": 0.19564438462257386,
      "num_tokens": 28160766.0,
      "step": 12625
    },
    {
      "entropy": 5.431835174560547,
      "epoch": 1.4184871118099625,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048090000659493985,
      "loss": 5.3344,
      "mean_token_accuracy": 0.1860590472817421,
      "num_tokens": 28171583.0,
      "step": 12630
    },
    {
      "entropy": 5.411688375473022,
      "epoch": 1.41904868871792,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048088381528580445,
      "loss": 5.1819,
      "mean_token_accuracy": 0.1947830066084862,
      "num_tokens": 28183598.0,
      "step": 12635
    },
    {
      "entropy": 5.461697816848755,
      "epoch": 1.4196102656258773,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048086761742124846,
      "loss": 5.3068,
      "mean_token_accuracy": 0.1827881932258606,
      "num_tokens": 28193377.0,
      "step": 12640
    },
    {
      "entropy": 5.413489103317261,
      "epoch": 1.420171842533835,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048085141300178774,
      "loss": 5.2191,
      "mean_token_accuracy": 0.20018460005521774,
      "num_tokens": 28204536.0,
      "step": 12645
    },
    {
      "entropy": 5.40965166091919,
      "epoch": 1.4207334194417927,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048083520202793815,
      "loss": 5.3763,
      "mean_token_accuracy": 0.19241825938224794,
      "num_tokens": 28215806.0,
      "step": 12650
    },
    {
      "entropy": 5.522747945785523,
      "epoch": 1.42129499634975,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004808189845002159,
      "loss": 5.3069,
      "mean_token_accuracy": 0.18906259983778,
      "num_tokens": 28226765.0,
      "step": 12655
    },
    {
      "entropy": 5.640943956375122,
      "epoch": 1.4218565732577075,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004808027604191374,
      "loss": 5.4737,
      "mean_token_accuracy": 0.1846621036529541,
      "num_tokens": 28238077.0,
      "step": 12660
    },
    {
      "entropy": 5.470943832397461,
      "epoch": 1.4224181501656652,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048078652978521914,
      "loss": 5.3259,
      "mean_token_accuracy": 0.19031595289707184,
      "num_tokens": 28248992.0,
      "step": 12665
    },
    {
      "entropy": 5.452727031707764,
      "epoch": 1.4229797270736229,
      "grad_norm": 1.234375,
      "learning_rate": 0.00048077029259897797,
      "loss": 5.2817,
      "mean_token_accuracy": 0.193523009121418,
      "num_tokens": 28259818.0,
      "step": 12670
    },
    {
      "entropy": 5.422079801559448,
      "epoch": 1.4235413039815803,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048075404886093094,
      "loss": 5.2699,
      "mean_token_accuracy": 0.1927775889635086,
      "num_tokens": 28271139.0,
      "step": 12675
    },
    {
      "entropy": 5.423200559616089,
      "epoch": 1.4241028808895377,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048073779857159504,
      "loss": 5.2933,
      "mean_token_accuracy": 0.19640138149261474,
      "num_tokens": 28282093.0,
      "step": 12680
    },
    {
      "entropy": 5.428907632827759,
      "epoch": 1.4246644577974954,
      "grad_norm": 1.203125,
      "learning_rate": 0.000480721541731488,
      "loss": 5.2005,
      "mean_token_accuracy": 0.197603215277195,
      "num_tokens": 28292701.0,
      "step": 12685
    },
    {
      "entropy": 5.351401138305664,
      "epoch": 1.4252260347054528,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048070527834112727,
      "loss": 5.2495,
      "mean_token_accuracy": 0.19207984060049058,
      "num_tokens": 28303585.0,
      "step": 12690
    },
    {
      "entropy": 5.415815305709839,
      "epoch": 1.4257876116134105,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004806890084010306,
      "loss": 5.1951,
      "mean_token_accuracy": 0.19424072206020354,
      "num_tokens": 28314583.0,
      "step": 12695
    },
    {
      "entropy": 5.40502438545227,
      "epoch": 1.426349188521368,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048067273191171615,
      "loss": 5.2621,
      "mean_token_accuracy": 0.1918814092874527,
      "num_tokens": 28325895.0,
      "step": 12700
    },
    {
      "entropy": 5.547545909881592,
      "epoch": 1.4269107654293256,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004806564488737022,
      "loss": 5.335,
      "mean_token_accuracy": 0.18603358119726182,
      "num_tokens": 28337011.0,
      "step": 12705
    },
    {
      "entropy": 5.5984639644622805,
      "epoch": 1.427472342337283,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004806401592875071,
      "loss": 5.3701,
      "mean_token_accuracy": 0.18326019793748854,
      "num_tokens": 28347468.0,
      "step": 12710
    },
    {
      "entropy": 5.446509790420532,
      "epoch": 1.4280339192452407,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048062386315364964,
      "loss": 5.2755,
      "mean_token_accuracy": 0.19473423659801484,
      "num_tokens": 28357287.0,
      "step": 12715
    },
    {
      "entropy": 5.4957598686218265,
      "epoch": 1.4285954961531981,
      "grad_norm": 1.125,
      "learning_rate": 0.0004806075604726485,
      "loss": 5.3115,
      "mean_token_accuracy": 0.19761179238557816,
      "num_tokens": 28368508.0,
      "step": 12720
    },
    {
      "entropy": 5.459147500991821,
      "epoch": 1.4291570730611558,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048059125124502304,
      "loss": 5.2695,
      "mean_token_accuracy": 0.19363388121128083,
      "num_tokens": 28378211.0,
      "step": 12725
    },
    {
      "entropy": 5.385713863372803,
      "epoch": 1.4297186499691132,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004805749354712924,
      "loss": 5.2275,
      "mean_token_accuracy": 0.19853050261735916,
      "num_tokens": 28389083.0,
      "step": 12730
    },
    {
      "entropy": 5.486315202713013,
      "epoch": 1.4302802268770707,
      "grad_norm": 1.203125,
      "learning_rate": 0.000480558613151976,
      "loss": 5.2794,
      "mean_token_accuracy": 0.19317805469036103,
      "num_tokens": 28399786.0,
      "step": 12735
    },
    {
      "entropy": 5.437857294082642,
      "epoch": 1.4308418037850283,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004805422842875936,
      "loss": 5.3235,
      "mean_token_accuracy": 0.18600660860538482,
      "num_tokens": 28410650.0,
      "step": 12740
    },
    {
      "entropy": 5.532534217834472,
      "epoch": 1.431403380692986,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048052594887866524,
      "loss": 5.3487,
      "mean_token_accuracy": 0.19088319540023804,
      "num_tokens": 28421485.0,
      "step": 12745
    },
    {
      "entropy": 5.492064905166626,
      "epoch": 1.4319649576009434,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048050960692571086,
      "loss": 5.2855,
      "mean_token_accuracy": 0.18370473384857178,
      "num_tokens": 28432469.0,
      "step": 12750
    },
    {
      "entropy": 5.427737092971801,
      "epoch": 1.4325265345089009,
      "grad_norm": 1.171875,
      "learning_rate": 0.000480493258429251,
      "loss": 5.2401,
      "mean_token_accuracy": 0.19238171279430388,
      "num_tokens": 28443628.0,
      "step": 12755
    },
    {
      "entropy": 5.533960294723511,
      "epoch": 1.4330881114168585,
      "grad_norm": 1.5859375,
      "learning_rate": 0.000480476903389806,
      "loss": 5.3529,
      "mean_token_accuracy": 0.1820556640625,
      "num_tokens": 28455106.0,
      "step": 12760
    },
    {
      "entropy": 5.518792390823364,
      "epoch": 1.4336496883248162,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004804605418078968,
      "loss": 5.3022,
      "mean_token_accuracy": 0.1893300622701645,
      "num_tokens": 28466326.0,
      "step": 12765
    },
    {
      "entropy": 5.519400691986084,
      "epoch": 1.4342112652327736,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00048044417368404426,
      "loss": 5.4604,
      "mean_token_accuracy": 0.18598620295524598,
      "num_tokens": 28477490.0,
      "step": 12770
    },
    {
      "entropy": 5.463613557815552,
      "epoch": 1.434772842140731,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048042779901876946,
      "loss": 5.221,
      "mean_token_accuracy": 0.19350895285606384,
      "num_tokens": 28489780.0,
      "step": 12775
    },
    {
      "entropy": 5.553389072418213,
      "epoch": 1.4353344190486887,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004804114178125939,
      "loss": 5.3656,
      "mean_token_accuracy": 0.18867199271917343,
      "num_tokens": 28500335.0,
      "step": 12780
    },
    {
      "entropy": 5.32941427230835,
      "epoch": 1.4358959959566462,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004803950300660391,
      "loss": 5.1532,
      "mean_token_accuracy": 0.20113422274589537,
      "num_tokens": 28510680.0,
      "step": 12785
    },
    {
      "entropy": 5.389677286148071,
      "epoch": 1.4364575728646038,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004803786357796269,
      "loss": 5.2061,
      "mean_token_accuracy": 0.2009651854634285,
      "num_tokens": 28521719.0,
      "step": 12790
    },
    {
      "entropy": 5.460920238494873,
      "epoch": 1.4370191497725613,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048036223495387934,
      "loss": 5.3023,
      "mean_token_accuracy": 0.18776607513427734,
      "num_tokens": 28531935.0,
      "step": 12795
    },
    {
      "entropy": 5.518677997589111,
      "epoch": 1.437580726680519,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004803458275893185,
      "loss": 5.2958,
      "mean_token_accuracy": 0.18790262192487717,
      "num_tokens": 28542529.0,
      "step": 12800
    },
    {
      "entropy": 5.462746000289917,
      "epoch": 1.4381423035884764,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004803294136864669,
      "loss": 5.281,
      "mean_token_accuracy": 0.19504371136426926,
      "num_tokens": 28554724.0,
      "step": 12805
    },
    {
      "entropy": 5.482596206665039,
      "epoch": 1.438703880496434,
      "grad_norm": 1.25,
      "learning_rate": 0.00048031299324584716,
      "loss": 5.2453,
      "mean_token_accuracy": 0.19673299193382263,
      "num_tokens": 28565781.0,
      "step": 12810
    },
    {
      "entropy": 5.5013813972473145,
      "epoch": 1.4392654574043915,
      "grad_norm": 1.265625,
      "learning_rate": 0.000480296566267982,
      "loss": 5.2603,
      "mean_token_accuracy": 0.19467213600873948,
      "num_tokens": 28576186.0,
      "step": 12815
    },
    {
      "entropy": 5.388833951950073,
      "epoch": 1.4398270343123492,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004802801327533947,
      "loss": 5.2844,
      "mean_token_accuracy": 0.1888722836971283,
      "num_tokens": 28588240.0,
      "step": 12820
    },
    {
      "entropy": 5.496490144729615,
      "epoch": 1.4403886112203066,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048026369270260814,
      "loss": 5.3147,
      "mean_token_accuracy": 0.18569540828466416,
      "num_tokens": 28598240.0,
      "step": 12825
    },
    {
      "entropy": 5.529063892364502,
      "epoch": 1.440950188128264,
      "grad_norm": 1.375,
      "learning_rate": 0.0004802472461161462,
      "loss": 5.3417,
      "mean_token_accuracy": 0.19033762365579604,
      "num_tokens": 28610091.0,
      "step": 12830
    },
    {
      "entropy": 5.526371097564697,
      "epoch": 1.4415117650362217,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004802307929945322,
      "loss": 5.3333,
      "mean_token_accuracy": 0.189434415102005,
      "num_tokens": 28621060.0,
      "step": 12835
    },
    {
      "entropy": 5.544085168838501,
      "epoch": 1.4420733419441794,
      "grad_norm": 3.03125,
      "learning_rate": 0.00048021433333829015,
      "loss": 5.3593,
      "mean_token_accuracy": 0.18866870254278184,
      "num_tokens": 28632027.0,
      "step": 12840
    },
    {
      "entropy": 5.561224889755249,
      "epoch": 1.4426349188521368,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048019786714794424,
      "loss": 5.4099,
      "mean_token_accuracy": 0.19027109146118165,
      "num_tokens": 28643959.0,
      "step": 12845
    },
    {
      "entropy": 5.530548858642578,
      "epoch": 1.4431964957600942,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004801813944240186,
      "loss": 5.3138,
      "mean_token_accuracy": 0.1917598769068718,
      "num_tokens": 28654166.0,
      "step": 12850
    },
    {
      "entropy": 5.448145627975464,
      "epoch": 1.443758072668052,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004801649151670378,
      "loss": 5.2634,
      "mean_token_accuracy": 0.1889197662472725,
      "num_tokens": 28667975.0,
      "step": 12855
    },
    {
      "entropy": 5.420803451538086,
      "epoch": 1.4443196495760096,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004801484293775265,
      "loss": 5.2689,
      "mean_token_accuracy": 0.18978217840194703,
      "num_tokens": 28679333.0,
      "step": 12860
    },
    {
      "entropy": 5.457807111740112,
      "epoch": 1.444881226483967,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004801319370560097,
      "loss": 5.1837,
      "mean_token_accuracy": 0.1893686056137085,
      "num_tokens": 28689333.0,
      "step": 12865
    },
    {
      "entropy": 5.462051343917847,
      "epoch": 1.4454428033919244,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004801154382030124,
      "loss": 5.3675,
      "mean_token_accuracy": 0.17773723006248474,
      "num_tokens": 28700625.0,
      "step": 12870
    },
    {
      "entropy": 5.523106908798217,
      "epoch": 1.446004380299882,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048009893281906004,
      "loss": 5.3669,
      "mean_token_accuracy": 0.18397089540958406,
      "num_tokens": 28711768.0,
      "step": 12875
    },
    {
      "entropy": 5.478731727600097,
      "epoch": 1.4465659572078395,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048008242090467803,
      "loss": 5.371,
      "mean_token_accuracy": 0.18609944134950637,
      "num_tokens": 28722521.0,
      "step": 12880
    },
    {
      "entropy": 5.5220263481140135,
      "epoch": 1.4471275341157972,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004800659024603923,
      "loss": 5.3134,
      "mean_token_accuracy": 0.19136330336332322,
      "num_tokens": 28733293.0,
      "step": 12885
    },
    {
      "entropy": 5.544143915176392,
      "epoch": 1.4476891110237546,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048004937748672864,
      "loss": 5.303,
      "mean_token_accuracy": 0.18705746084451674,
      "num_tokens": 28744351.0,
      "step": 12890
    },
    {
      "entropy": 5.429519462585449,
      "epoch": 1.4482506879317123,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048003284598421327,
      "loss": 5.2363,
      "mean_token_accuracy": 0.18990573585033416,
      "num_tokens": 28755140.0,
      "step": 12895
    },
    {
      "entropy": 5.419640016555786,
      "epoch": 1.4488122648396697,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004800163079533726,
      "loss": 5.3514,
      "mean_token_accuracy": 0.187733057141304,
      "num_tokens": 28767161.0,
      "step": 12900
    },
    {
      "entropy": 5.51209454536438,
      "epoch": 1.4493738417476274,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004799997633947332,
      "loss": 5.2873,
      "mean_token_accuracy": 0.19328074604272844,
      "num_tokens": 28779501.0,
      "step": 12905
    },
    {
      "entropy": 5.533173274993897,
      "epoch": 1.4499354186555848,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047998321230882174,
      "loss": 5.4425,
      "mean_token_accuracy": 0.1884545937180519,
      "num_tokens": 28792021.0,
      "step": 12910
    },
    {
      "entropy": 5.405355978012085,
      "epoch": 1.4504969955635425,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047996665469616536,
      "loss": 5.2493,
      "mean_token_accuracy": 0.188119837641716,
      "num_tokens": 28802445.0,
      "step": 12915
    },
    {
      "entropy": 5.457556486129761,
      "epoch": 1.4510585724715,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047995009055729114,
      "loss": 5.2457,
      "mean_token_accuracy": 0.1897529259324074,
      "num_tokens": 28813821.0,
      "step": 12920
    },
    {
      "entropy": 5.457045459747315,
      "epoch": 1.4516201493794574,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004799335198927266,
      "loss": 5.2819,
      "mean_token_accuracy": 0.1892861992120743,
      "num_tokens": 28824764.0,
      "step": 12925
    },
    {
      "entropy": 5.5071197032928465,
      "epoch": 1.452181726287415,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004799169427029993,
      "loss": 5.3209,
      "mean_token_accuracy": 0.18950579017400743,
      "num_tokens": 28836720.0,
      "step": 12930
    },
    {
      "entropy": 5.468385410308838,
      "epoch": 1.4527433031953727,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047990035898863706,
      "loss": 5.2254,
      "mean_token_accuracy": 0.19476556926965713,
      "num_tokens": 28848846.0,
      "step": 12935
    },
    {
      "entropy": 5.439984846115112,
      "epoch": 1.4533048801033301,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004798837687501679,
      "loss": 5.2657,
      "mean_token_accuracy": 0.19292198568582536,
      "num_tokens": 28859453.0,
      "step": 12940
    },
    {
      "entropy": 5.4562098503112795,
      "epoch": 1.4538664570112876,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004798671719881201,
      "loss": 5.3502,
      "mean_token_accuracy": 0.18741936832666398,
      "num_tokens": 28869447.0,
      "step": 12945
    },
    {
      "entropy": 5.531781005859375,
      "epoch": 1.4544280339192452,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004798505687030221,
      "loss": 5.38,
      "mean_token_accuracy": 0.18734290450811386,
      "num_tokens": 28881660.0,
      "step": 12950
    },
    {
      "entropy": 5.56375207901001,
      "epoch": 1.454989610827203,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004798339588954025,
      "loss": 5.3898,
      "mean_token_accuracy": 0.18610263466835023,
      "num_tokens": 28893128.0,
      "step": 12955
    },
    {
      "entropy": 5.5075404167175295,
      "epoch": 1.4555511877351603,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004798173425657903,
      "loss": 5.287,
      "mean_token_accuracy": 0.1924218401312828,
      "num_tokens": 28904323.0,
      "step": 12960
    },
    {
      "entropy": 5.492358064651489,
      "epoch": 1.4561127646431178,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004798007197147144,
      "loss": 5.326,
      "mean_token_accuracy": 0.19349563419818877,
      "num_tokens": 28915672.0,
      "step": 12965
    },
    {
      "entropy": 5.556574010848999,
      "epoch": 1.4566743415510754,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004797840903427042,
      "loss": 5.4085,
      "mean_token_accuracy": 0.18541398495435715,
      "num_tokens": 28926657.0,
      "step": 12970
    },
    {
      "entropy": 5.380950689315796,
      "epoch": 1.4572359184590329,
      "grad_norm": 1.125,
      "learning_rate": 0.00047976745445028906,
      "loss": 5.2249,
      "mean_token_accuracy": 0.1930977523326874,
      "num_tokens": 28938787.0,
      "step": 12975
    },
    {
      "entropy": 5.377897834777832,
      "epoch": 1.4577974953669905,
      "grad_norm": 1.25,
      "learning_rate": 0.00047975081203799883,
      "loss": 5.3122,
      "mean_token_accuracy": 0.1889840245246887,
      "num_tokens": 28949836.0,
      "step": 12980
    },
    {
      "entropy": 5.446573305130005,
      "epoch": 1.458359072274948,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047973416310636327,
      "loss": 5.2506,
      "mean_token_accuracy": 0.1908799558877945,
      "num_tokens": 28960053.0,
      "step": 12985
    },
    {
      "entropy": 5.604538536071777,
      "epoch": 1.4589206491829056,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004797175076559126,
      "loss": 5.4265,
      "mean_token_accuracy": 0.19144188910722731,
      "num_tokens": 28972698.0,
      "step": 12990
    },
    {
      "entropy": 5.5056578636169435,
      "epoch": 1.459482226090863,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047970084568717703,
      "loss": 5.4017,
      "mean_token_accuracy": 0.18662042170763016,
      "num_tokens": 28984756.0,
      "step": 12995
    },
    {
      "entropy": 5.474940347671509,
      "epoch": 1.4600438029988208,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004796841772006872,
      "loss": 5.3203,
      "mean_token_accuracy": 0.18327830582857133,
      "num_tokens": 28996915.0,
      "step": 13000
    },
    {
      "entropy": 5.447759580612183,
      "epoch": 1.4606053799067782,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047966750219697376,
      "loss": 5.2529,
      "mean_token_accuracy": 0.19027094841003417,
      "num_tokens": 29007462.0,
      "step": 13005
    },
    {
      "entropy": 5.499920654296875,
      "epoch": 1.4611669568147359,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004796508206765677,
      "loss": 5.3333,
      "mean_token_accuracy": 0.1889558807015419,
      "num_tokens": 29017790.0,
      "step": 13010
    },
    {
      "entropy": 5.570965528488159,
      "epoch": 1.4617285337226933,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047963413264000005,
      "loss": 5.4109,
      "mean_token_accuracy": 0.1815931335091591,
      "num_tokens": 29029536.0,
      "step": 13015
    },
    {
      "entropy": 5.45293698310852,
      "epoch": 1.4622901106306507,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004796174380878023,
      "loss": 5.2134,
      "mean_token_accuracy": 0.1960282638669014,
      "num_tokens": 29039840.0,
      "step": 13020
    },
    {
      "entropy": 5.539623832702636,
      "epoch": 1.4628516875386084,
      "grad_norm": 1.234375,
      "learning_rate": 0.000479600737020506,
      "loss": 5.3613,
      "mean_token_accuracy": 0.18212379217147828,
      "num_tokens": 29052769.0,
      "step": 13025
    },
    {
      "entropy": 5.419940423965454,
      "epoch": 1.463413264446566,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004795840294386429,
      "loss": 5.2431,
      "mean_token_accuracy": 0.1946727454662323,
      "num_tokens": 29063645.0,
      "step": 13030
    },
    {
      "entropy": 5.4768593311309814,
      "epoch": 1.4639748413545235,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047956731534274483,
      "loss": 5.3513,
      "mean_token_accuracy": 0.18697307556867598,
      "num_tokens": 29075241.0,
      "step": 13035
    },
    {
      "entropy": 5.562857580184937,
      "epoch": 1.464536418262481,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047955059473334425,
      "loss": 5.315,
      "mean_token_accuracy": 0.19321728497743607,
      "num_tokens": 29087592.0,
      "step": 13040
    },
    {
      "entropy": 5.451340770721435,
      "epoch": 1.4650979951704386,
      "grad_norm": 1.125,
      "learning_rate": 0.0004795338676109733,
      "loss": 5.2726,
      "mean_token_accuracy": 0.19074058681726455,
      "num_tokens": 29099314.0,
      "step": 13045
    },
    {
      "entropy": 5.410186004638672,
      "epoch": 1.4656595720783963,
      "grad_norm": 1.25,
      "learning_rate": 0.0004795171339761647,
      "loss": 5.332,
      "mean_token_accuracy": 0.19200712740421294,
      "num_tokens": 29111916.0,
      "step": 13050
    },
    {
      "entropy": 5.486436367034912,
      "epoch": 1.4662211489863537,
      "grad_norm": 1.421875,
      "learning_rate": 0.00047950039382945126,
      "loss": 5.3294,
      "mean_token_accuracy": 0.1879766285419464,
      "num_tokens": 29122871.0,
      "step": 13055
    },
    {
      "entropy": 5.443666458129883,
      "epoch": 1.4667827258943111,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047948364717136587,
      "loss": 5.3142,
      "mean_token_accuracy": 0.18931028842926026,
      "num_tokens": 29133873.0,
      "step": 13060
    },
    {
      "entropy": 5.411843109130859,
      "epoch": 1.4673443028022688,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047946689400244196,
      "loss": 5.2125,
      "mean_token_accuracy": 0.19386772215366363,
      "num_tokens": 29144995.0,
      "step": 13065
    },
    {
      "entropy": 5.557354497909546,
      "epoch": 1.4679058797102262,
      "grad_norm": 1.5625,
      "learning_rate": 0.00047945013432321284,
      "loss": 5.3699,
      "mean_token_accuracy": 0.1885412886738777,
      "num_tokens": 29156801.0,
      "step": 13070
    },
    {
      "entropy": 5.503143692016602,
      "epoch": 1.468467456618184,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047943336813421205,
      "loss": 5.3958,
      "mean_token_accuracy": 0.18250905871391296,
      "num_tokens": 29168072.0,
      "step": 13075
    },
    {
      "entropy": 5.451800870895386,
      "epoch": 1.4690290335261413,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004794165954359736,
      "loss": 5.2648,
      "mean_token_accuracy": 0.1901470422744751,
      "num_tokens": 29179314.0,
      "step": 13080
    },
    {
      "entropy": 5.514900732040405,
      "epoch": 1.469590610434099,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004793998162290315,
      "loss": 5.3561,
      "mean_token_accuracy": 0.18320836275815963,
      "num_tokens": 29190808.0,
      "step": 13085
    },
    {
      "entropy": 5.562315177917481,
      "epoch": 1.4701521873420564,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004793830305139199,
      "loss": 5.4342,
      "mean_token_accuracy": 0.18613820225000383,
      "num_tokens": 29202557.0,
      "step": 13090
    },
    {
      "entropy": 5.421227931976318,
      "epoch": 1.470713764250014,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004793662382911734,
      "loss": 5.1998,
      "mean_token_accuracy": 0.19936956018209456,
      "num_tokens": 29213959.0,
      "step": 13095
    },
    {
      "entropy": 5.499758958816528,
      "epoch": 1.4712753411579715,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004793494395613265,
      "loss": 5.3336,
      "mean_token_accuracy": 0.19254908412694932,
      "num_tokens": 29226069.0,
      "step": 13100
    },
    {
      "entropy": 5.551686954498291,
      "epoch": 1.4718369180659292,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004793326343249144,
      "loss": 5.3646,
      "mean_token_accuracy": 0.18708494305610657,
      "num_tokens": 29237882.0,
      "step": 13105
    },
    {
      "entropy": 5.475058603286743,
      "epoch": 1.4723984949738866,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004793158225824718,
      "loss": 5.2987,
      "mean_token_accuracy": 0.19884727746248246,
      "num_tokens": 29248642.0,
      "step": 13110
    },
    {
      "entropy": 5.4112154006958,
      "epoch": 1.4729600718818443,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047929900433453423,
      "loss": 5.2685,
      "mean_token_accuracy": 0.19126975834369658,
      "num_tokens": 29260632.0,
      "step": 13115
    },
    {
      "entropy": 5.580027151107788,
      "epoch": 1.4735216487898017,
      "grad_norm": 1.25,
      "learning_rate": 0.00047928217958163703,
      "loss": 5.4106,
      "mean_token_accuracy": 0.183598031103611,
      "num_tokens": 29272563.0,
      "step": 13120
    },
    {
      "entropy": 5.562701797485351,
      "epoch": 1.4740832256977594,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047926534832431615,
      "loss": 5.3531,
      "mean_token_accuracy": 0.18901700675487518,
      "num_tokens": 29284307.0,
      "step": 13125
    },
    {
      "entropy": 5.430919790267945,
      "epoch": 1.4746448026057168,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004792485105631073,
      "loss": 5.2282,
      "mean_token_accuracy": 0.19300856292247773,
      "num_tokens": 29295413.0,
      "step": 13130
    },
    {
      "entropy": 5.541763162612915,
      "epoch": 1.4752063795136743,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004792316662985467,
      "loss": 5.3401,
      "mean_token_accuracy": 0.19652038514614106,
      "num_tokens": 29306612.0,
      "step": 13135
    },
    {
      "entropy": 5.467514610290527,
      "epoch": 1.475767956421632,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047921481553117053,
      "loss": 5.2767,
      "mean_token_accuracy": 0.19746398478746413,
      "num_tokens": 29316168.0,
      "step": 13140
    },
    {
      "entropy": 5.4918793678283695,
      "epoch": 1.4763295333295896,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047919795826151544,
      "loss": 5.3909,
      "mean_token_accuracy": 0.1901333749294281,
      "num_tokens": 29326949.0,
      "step": 13145
    },
    {
      "entropy": 5.495761442184448,
      "epoch": 1.476891110237547,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004791810944901183,
      "loss": 5.2561,
      "mean_token_accuracy": 0.19033121466636657,
      "num_tokens": 29336744.0,
      "step": 13150
    },
    {
      "entropy": 5.504734945297241,
      "epoch": 1.4774526871455045,
      "grad_norm": 1.25,
      "learning_rate": 0.0004791642242175158,
      "loss": 5.2837,
      "mean_token_accuracy": 0.19450800865888596,
      "num_tokens": 29346964.0,
      "step": 13155
    },
    {
      "entropy": 5.4534751892089846,
      "epoch": 1.4780142640534621,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047914734744424517,
      "loss": 5.2804,
      "mean_token_accuracy": 0.18906400799751283,
      "num_tokens": 29358689.0,
      "step": 13160
    },
    {
      "entropy": 5.418648815155029,
      "epoch": 1.4785758409614196,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047913046417084393,
      "loss": 5.3254,
      "mean_token_accuracy": 0.1903253436088562,
      "num_tokens": 29371243.0,
      "step": 13165
    },
    {
      "entropy": 5.42978801727295,
      "epoch": 1.4791374178693772,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00047911357439784947,
      "loss": 5.3231,
      "mean_token_accuracy": 0.19153652787208558,
      "num_tokens": 29383185.0,
      "step": 13170
    },
    {
      "entropy": 5.36575927734375,
      "epoch": 1.4796989947773347,
      "grad_norm": 1.453125,
      "learning_rate": 0.00047909667812579964,
      "loss": 5.1168,
      "mean_token_accuracy": 0.20103544145822524,
      "num_tokens": 29393986.0,
      "step": 13175
    },
    {
      "entropy": 5.399748182296753,
      "epoch": 1.4802605716852923,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047907977535523253,
      "loss": 5.3319,
      "mean_token_accuracy": 0.18958683758974076,
      "num_tokens": 29405407.0,
      "step": 13180
    },
    {
      "entropy": 5.445173263549805,
      "epoch": 1.4808221485932498,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004790628660866861,
      "loss": 5.2238,
      "mean_token_accuracy": 0.18947108536958696,
      "num_tokens": 29416352.0,
      "step": 13185
    },
    {
      "entropy": 5.444636964797974,
      "epoch": 1.4813837255012074,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004790459503206989,
      "loss": 5.1823,
      "mean_token_accuracy": 0.19593167304992676,
      "num_tokens": 29428369.0,
      "step": 13190
    },
    {
      "entropy": 5.4412566184997555,
      "epoch": 1.4819453024091649,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004790290280578095,
      "loss": 5.2808,
      "mean_token_accuracy": 0.19021789729595184,
      "num_tokens": 29439276.0,
      "step": 13195
    },
    {
      "entropy": 5.505595541000366,
      "epoch": 1.4825068793171226,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047901209929855675,
      "loss": 5.2957,
      "mean_token_accuracy": 0.1869095131754875,
      "num_tokens": 29451361.0,
      "step": 13200
    },
    {
      "entropy": 5.448739099502563,
      "epoch": 1.48306845622508,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004789951640434796,
      "loss": 5.3297,
      "mean_token_accuracy": 0.18787485361099243,
      "num_tokens": 29462977.0,
      "step": 13205
    },
    {
      "entropy": 5.4503566265106205,
      "epoch": 1.4836300331330377,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004789782222931173,
      "loss": 5.2653,
      "mean_token_accuracy": 0.1935327798128128,
      "num_tokens": 29474641.0,
      "step": 13210
    },
    {
      "entropy": 5.431048059463501,
      "epoch": 1.484191610040995,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047896127404800935,
      "loss": 5.1537,
      "mean_token_accuracy": 0.19886730313301088,
      "num_tokens": 29485560.0,
      "step": 13215
    },
    {
      "entropy": 5.385657739639282,
      "epoch": 1.4847531869489528,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004789443193086953,
      "loss": 5.2792,
      "mean_token_accuracy": 0.19575762003660202,
      "num_tokens": 29495872.0,
      "step": 13220
    },
    {
      "entropy": 5.371059370040894,
      "epoch": 1.4853147638569102,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047892735807571505,
      "loss": 5.2117,
      "mean_token_accuracy": 0.19819064885377885,
      "num_tokens": 29506704.0,
      "step": 13225
    },
    {
      "entropy": 5.459822082519532,
      "epoch": 1.4858763407648676,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004789103903496086,
      "loss": 5.2542,
      "mean_token_accuracy": 0.18846206068992616,
      "num_tokens": 29518226.0,
      "step": 13230
    },
    {
      "entropy": 5.460194253921509,
      "epoch": 1.4864379176728253,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004788934161309161,
      "loss": 5.2222,
      "mean_token_accuracy": 0.1979835346341133,
      "num_tokens": 29529176.0,
      "step": 13235
    },
    {
      "entropy": 5.345615434646606,
      "epoch": 1.486999494580783,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047887643542017836,
      "loss": 5.1047,
      "mean_token_accuracy": 0.19681110978126526,
      "num_tokens": 29539268.0,
      "step": 13240
    },
    {
      "entropy": 5.459047794342041,
      "epoch": 1.4875610714887404,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047885944821793574,
      "loss": 5.4115,
      "mean_token_accuracy": 0.1844659686088562,
      "num_tokens": 29550491.0,
      "step": 13245
    },
    {
      "entropy": 5.450409746170044,
      "epoch": 1.4881226483966978,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047884245452472926,
      "loss": 5.2215,
      "mean_token_accuracy": 0.1957974538207054,
      "num_tokens": 29561520.0,
      "step": 13250
    },
    {
      "entropy": 5.462947368621826,
      "epoch": 1.4886842253046555,
      "grad_norm": 1.25,
      "learning_rate": 0.0004788254543410999,
      "loss": 5.2886,
      "mean_token_accuracy": 0.1886413037776947,
      "num_tokens": 29573974.0,
      "step": 13255
    },
    {
      "entropy": 5.3695742130279545,
      "epoch": 1.489245802212613,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047880844766758914,
      "loss": 5.1808,
      "mean_token_accuracy": 0.2010017901659012,
      "num_tokens": 29584059.0,
      "step": 13260
    },
    {
      "entropy": 5.532934856414795,
      "epoch": 1.4898073791205706,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047879143450473824,
      "loss": 5.4159,
      "mean_token_accuracy": 0.1893548920750618,
      "num_tokens": 29595726.0,
      "step": 13265
    },
    {
      "entropy": 5.449132061004638,
      "epoch": 1.490368956028528,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047877441485308917,
      "loss": 5.3186,
      "mean_token_accuracy": 0.18625374883413315,
      "num_tokens": 29607354.0,
      "step": 13270
    },
    {
      "entropy": 5.474122190475464,
      "epoch": 1.4909305329364857,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047875738871318356,
      "loss": 5.2475,
      "mean_token_accuracy": 0.19682285338640212,
      "num_tokens": 29618698.0,
      "step": 13275
    },
    {
      "entropy": 5.378784942626953,
      "epoch": 1.4914921098444431,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004787403560855637,
      "loss": 5.2135,
      "mean_token_accuracy": 0.19360013157129288,
      "num_tokens": 29629578.0,
      "step": 13280
    },
    {
      "entropy": 5.385878276824951,
      "epoch": 1.4920536867524008,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004787233169707719,
      "loss": 5.0773,
      "mean_token_accuracy": 0.2017005756497383,
      "num_tokens": 29640277.0,
      "step": 13285
    },
    {
      "entropy": 5.445515203475952,
      "epoch": 1.4926152636603582,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004787062713693507,
      "loss": 5.3106,
      "mean_token_accuracy": 0.1915894240140915,
      "num_tokens": 29652029.0,
      "step": 13290
    },
    {
      "entropy": 5.461283254623413,
      "epoch": 1.493176840568316,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047868921928184276,
      "loss": 5.3091,
      "mean_token_accuracy": 0.1898849904537201,
      "num_tokens": 29663191.0,
      "step": 13295
    },
    {
      "entropy": 5.475704145431519,
      "epoch": 1.4937384174762733,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004786721607087911,
      "loss": 5.3369,
      "mean_token_accuracy": 0.1876775324344635,
      "num_tokens": 29674238.0,
      "step": 13300
    },
    {
      "entropy": 5.412631177902222,
      "epoch": 1.494299994384231,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004786550956507389,
      "loss": 5.2626,
      "mean_token_accuracy": 0.19538378864526748,
      "num_tokens": 29687255.0,
      "step": 13305
    },
    {
      "entropy": 5.428766298294067,
      "epoch": 1.4948615712921884,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004786380241082294,
      "loss": 5.2121,
      "mean_token_accuracy": 0.193307563662529,
      "num_tokens": 29699769.0,
      "step": 13310
    },
    {
      "entropy": 5.538261747360229,
      "epoch": 1.495423148200146,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047862094608180623,
      "loss": 5.3754,
      "mean_token_accuracy": 0.19091795533895492,
      "num_tokens": 29710281.0,
      "step": 13315
    },
    {
      "entropy": 5.4063859462738035,
      "epoch": 1.4959847251081035,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004786038615720132,
      "loss": 5.1657,
      "mean_token_accuracy": 0.1878063827753067,
      "num_tokens": 29720647.0,
      "step": 13320
    },
    {
      "entropy": 5.44906587600708,
      "epoch": 1.496546302016061,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047858677057939427,
      "loss": 5.3346,
      "mean_token_accuracy": 0.19098743945360183,
      "num_tokens": 29731565.0,
      "step": 13325
    },
    {
      "entropy": 5.424046754837036,
      "epoch": 1.4971078789240186,
      "grad_norm": 1.5625,
      "learning_rate": 0.00047856967310449356,
      "loss": 5.2325,
      "mean_token_accuracy": 0.19227646738290788,
      "num_tokens": 29743361.0,
      "step": 13330
    },
    {
      "entropy": 5.473336410522461,
      "epoch": 1.4976694558319763,
      "grad_norm": 1.4375,
      "learning_rate": 0.00047855256914785556,
      "loss": 5.2843,
      "mean_token_accuracy": 0.18884260654449464,
      "num_tokens": 29754799.0,
      "step": 13335
    },
    {
      "entropy": 5.528604459762573,
      "epoch": 1.4982310327399337,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047853545871002476,
      "loss": 5.4083,
      "mean_token_accuracy": 0.18437430113554001,
      "num_tokens": 29766401.0,
      "step": 13340
    },
    {
      "entropy": 5.398160219192505,
      "epoch": 1.4987926096478912,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047851834179154595,
      "loss": 5.1633,
      "mean_token_accuracy": 0.20011865794658662,
      "num_tokens": 29778268.0,
      "step": 13345
    },
    {
      "entropy": 5.381930828094482,
      "epoch": 1.4993541865558488,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004785012183929643,
      "loss": 5.2215,
      "mean_token_accuracy": 0.19724785834550856,
      "num_tokens": 29788551.0,
      "step": 13350
    },
    {
      "entropy": 5.443346929550171,
      "epoch": 1.4999157634638065,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004784840885148249,
      "loss": 5.2287,
      "mean_token_accuracy": 0.19136023968458177,
      "num_tokens": 29799726.0,
      "step": 13355
    },
    {
      "entropy": 5.471234130859375,
      "epoch": 1.500477340371764,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004784669521576732,
      "loss": 5.3421,
      "mean_token_accuracy": 0.18854429423809052,
      "num_tokens": 29810615.0,
      "step": 13360
    },
    {
      "entropy": 5.491786193847656,
      "epoch": 1.5010389172797214,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004784498093220549,
      "loss": 5.2587,
      "mean_token_accuracy": 0.18946825414896012,
      "num_tokens": 29821766.0,
      "step": 13365
    },
    {
      "entropy": 5.422983312606812,
      "epoch": 1.501600494187679,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004784326600085157,
      "loss": 5.3139,
      "mean_token_accuracy": 0.1891363114118576,
      "num_tokens": 29833724.0,
      "step": 13370
    },
    {
      "entropy": 5.380463457107544,
      "epoch": 1.5021620710956367,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047841550421760164,
      "loss": 5.1974,
      "mean_token_accuracy": 0.19609702080488206,
      "num_tokens": 29845057.0,
      "step": 13375
    },
    {
      "entropy": 5.556083297729492,
      "epoch": 1.502723648003594,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047839834194985916,
      "loss": 5.3362,
      "mean_token_accuracy": 0.1937363177537918,
      "num_tokens": 29855649.0,
      "step": 13380
    },
    {
      "entropy": 5.371402168273926,
      "epoch": 1.5032852249115516,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004783811732058345,
      "loss": 5.1867,
      "mean_token_accuracy": 0.1990133911371231,
      "num_tokens": 29866873.0,
      "step": 13385
    },
    {
      "entropy": 5.397166919708252,
      "epoch": 1.5038468018195092,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047836399798607446,
      "loss": 5.2299,
      "mean_token_accuracy": 0.19500789046287537,
      "num_tokens": 29877690.0,
      "step": 13390
    },
    {
      "entropy": 5.452762603759766,
      "epoch": 1.5044083787274667,
      "grad_norm": 1.25,
      "learning_rate": 0.00047834681629112583,
      "loss": 5.2498,
      "mean_token_accuracy": 0.20224375277757645,
      "num_tokens": 29888461.0,
      "step": 13395
    },
    {
      "entropy": 5.420311117172242,
      "epoch": 1.5049699556354241,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004783296281215357,
      "loss": 5.2665,
      "mean_token_accuracy": 0.19086920619010925,
      "num_tokens": 29900559.0,
      "step": 13400
    },
    {
      "entropy": 5.422214126586914,
      "epoch": 1.5055315325433818,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004783124334778513,
      "loss": 5.1837,
      "mean_token_accuracy": 0.19653283059597015,
      "num_tokens": 29910815.0,
      "step": 13405
    },
    {
      "entropy": 5.472454595565796,
      "epoch": 1.5060931094513395,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004782952323606202,
      "loss": 5.4214,
      "mean_token_accuracy": 0.19157439321279526,
      "num_tokens": 29923262.0,
      "step": 13410
    },
    {
      "entropy": 5.36757264137268,
      "epoch": 1.506654686359297,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004782780247703901,
      "loss": 5.2516,
      "mean_token_accuracy": 0.19677075147628784,
      "num_tokens": 29934744.0,
      "step": 13415
    },
    {
      "entropy": 5.575308418273925,
      "epoch": 1.5072162632672543,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004782608107077088,
      "loss": 5.4136,
      "mean_token_accuracy": 0.18802115172147751,
      "num_tokens": 29945828.0,
      "step": 13420
    },
    {
      "entropy": 5.37007884979248,
      "epoch": 1.507777840175212,
      "grad_norm": 1.453125,
      "learning_rate": 0.00047824359017312445,
      "loss": 5.2185,
      "mean_token_accuracy": 0.19271572828292846,
      "num_tokens": 29956964.0,
      "step": 13425
    },
    {
      "entropy": 5.453942060470581,
      "epoch": 1.5083394170831697,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004782263631671854,
      "loss": 5.2848,
      "mean_token_accuracy": 0.19152913987636566,
      "num_tokens": 29968004.0,
      "step": 13430
    },
    {
      "entropy": 5.481569242477417,
      "epoch": 1.508900993991127,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047820912969044014,
      "loss": 5.2406,
      "mean_token_accuracy": 0.19876619726419448,
      "num_tokens": 29978237.0,
      "step": 13435
    },
    {
      "entropy": 5.423660373687744,
      "epoch": 1.5094625708990845,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047819188974343734,
      "loss": 5.3029,
      "mean_token_accuracy": 0.1893783152103424,
      "num_tokens": 29988379.0,
      "step": 13440
    },
    {
      "entropy": 5.498622751235962,
      "epoch": 1.5100241478070422,
      "grad_norm": 1.25,
      "learning_rate": 0.0004781746433267259,
      "loss": 5.3382,
      "mean_token_accuracy": 0.18958016186952592,
      "num_tokens": 29999024.0,
      "step": 13445
    },
    {
      "entropy": 5.466541528701782,
      "epoch": 1.5105857247149999,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047815739044085515,
      "loss": 5.2499,
      "mean_token_accuracy": 0.1946989893913269,
      "num_tokens": 30009411.0,
      "step": 13450
    },
    {
      "entropy": 5.419334554672242,
      "epoch": 1.5111473016229573,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047814013108637423,
      "loss": 5.2714,
      "mean_token_accuracy": 0.19390248954296113,
      "num_tokens": 30020746.0,
      "step": 13455
    },
    {
      "entropy": 5.491099500656128,
      "epoch": 1.5117088785309147,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047812286526383274,
      "loss": 5.236,
      "mean_token_accuracy": 0.19364815056324006,
      "num_tokens": 30032568.0,
      "step": 13460
    },
    {
      "entropy": 5.482397270202637,
      "epoch": 1.5122704554388724,
      "grad_norm": 1.125,
      "learning_rate": 0.0004781055929737804,
      "loss": 5.2859,
      "mean_token_accuracy": 0.19000679701566697,
      "num_tokens": 30042772.0,
      "step": 13465
    },
    {
      "entropy": 5.447984600067139,
      "epoch": 1.51283203234683,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047808831421676724,
      "loss": 5.3416,
      "mean_token_accuracy": 0.19633647799491882,
      "num_tokens": 30053842.0,
      "step": 13470
    },
    {
      "entropy": 5.536101150512695,
      "epoch": 1.5133936092547875,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004780710289933435,
      "loss": 5.3034,
      "mean_token_accuracy": 0.18809012174606324,
      "num_tokens": 30064744.0,
      "step": 13475
    },
    {
      "entropy": 5.503149175643921,
      "epoch": 1.513955186162745,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004780537373040593,
      "loss": 5.2828,
      "mean_token_accuracy": 0.19019837975502013,
      "num_tokens": 30075475.0,
      "step": 13480
    },
    {
      "entropy": 5.446236038208008,
      "epoch": 1.5145167630707026,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047803643914946537,
      "loss": 5.2883,
      "mean_token_accuracy": 0.19319882541894912,
      "num_tokens": 30087189.0,
      "step": 13485
    },
    {
      "entropy": 5.464937305450439,
      "epoch": 1.51507833997866,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004780191345301125,
      "loss": 5.3825,
      "mean_token_accuracy": 0.18436428755521775,
      "num_tokens": 30099263.0,
      "step": 13490
    },
    {
      "entropy": 5.455781126022339,
      "epoch": 1.5156399168866175,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047800182344655165,
      "loss": 5.3017,
      "mean_token_accuracy": 0.18823737800121307,
      "num_tokens": 30110925.0,
      "step": 13495
    },
    {
      "entropy": 5.492401742935181,
      "epoch": 1.5162014937945751,
      "grad_norm": 1.546875,
      "learning_rate": 0.000477984505899334,
      "loss": 5.3169,
      "mean_token_accuracy": 0.18528926372528076,
      "num_tokens": 30122708.0,
      "step": 13500
    },
    {
      "entropy": 5.391092157363891,
      "epoch": 1.5167630707025328,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047796718188901095,
      "loss": 5.1731,
      "mean_token_accuracy": 0.19694525599479676,
      "num_tokens": 30133624.0,
      "step": 13505
    },
    {
      "entropy": 5.375802755355835,
      "epoch": 1.5173246476104902,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047794985141613413,
      "loss": 5.2763,
      "mean_token_accuracy": 0.1938489109277725,
      "num_tokens": 30144501.0,
      "step": 13510
    },
    {
      "entropy": 5.447692203521728,
      "epoch": 1.5178862245184477,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047793251448125526,
      "loss": 5.2334,
      "mean_token_accuracy": 0.20159218311309815,
      "num_tokens": 30156351.0,
      "step": 13515
    },
    {
      "entropy": 5.481799364089966,
      "epoch": 1.5184478014264053,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004779151710849265,
      "loss": 5.3137,
      "mean_token_accuracy": 0.190145780146122,
      "num_tokens": 30167445.0,
      "step": 13520
    },
    {
      "entropy": 5.4703268051147464,
      "epoch": 1.519009378334363,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004778978212276999,
      "loss": 5.3667,
      "mean_token_accuracy": 0.18589168787002563,
      "num_tokens": 30179054.0,
      "step": 13525
    },
    {
      "entropy": 5.5348122119903564,
      "epoch": 1.5195709552423204,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004778804649101281,
      "loss": 5.2949,
      "mean_token_accuracy": 0.19022188782691957,
      "num_tokens": 30189412.0,
      "step": 13530
    },
    {
      "entropy": 5.4573283195495605,
      "epoch": 1.5201325321502779,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004778631021327636,
      "loss": 5.2956,
      "mean_token_accuracy": 0.18819875866174698,
      "num_tokens": 30200775.0,
      "step": 13535
    },
    {
      "entropy": 5.456742668151856,
      "epoch": 1.5206941090582355,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004778457328961592,
      "loss": 5.2351,
      "mean_token_accuracy": 0.19426926374435424,
      "num_tokens": 30212121.0,
      "step": 13540
    },
    {
      "entropy": 5.41021957397461,
      "epoch": 1.5212556859661932,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047782835720086797,
      "loss": 5.1952,
      "mean_token_accuracy": 0.19774212688207626,
      "num_tokens": 30222814.0,
      "step": 13545
    },
    {
      "entropy": 5.46501407623291,
      "epoch": 1.5218172628741506,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004778109750474432,
      "loss": 5.3083,
      "mean_token_accuracy": 0.18862183541059493,
      "num_tokens": 30232414.0,
      "step": 13550
    },
    {
      "entropy": 5.511292743682861,
      "epoch": 1.522378839782108,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004777935864364383,
      "loss": 5.3979,
      "mean_token_accuracy": 0.18790566623210908,
      "num_tokens": 30243479.0,
      "step": 13555
    },
    {
      "entropy": 5.506114387512207,
      "epoch": 1.5229404166900657,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000477776191368407,
      "loss": 5.3274,
      "mean_token_accuracy": 0.1919953092932701,
      "num_tokens": 30254721.0,
      "step": 13560
    },
    {
      "entropy": 5.467269420623779,
      "epoch": 1.5235019935980234,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004777587898439031,
      "loss": 5.2388,
      "mean_token_accuracy": 0.19289160817861556,
      "num_tokens": 30266962.0,
      "step": 13565
    },
    {
      "entropy": 5.454935455322266,
      "epoch": 1.5240635705059808,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047774138186348073,
      "loss": 5.26,
      "mean_token_accuracy": 0.19219527691602706,
      "num_tokens": 30278773.0,
      "step": 13570
    },
    {
      "entropy": 5.361927509307861,
      "epoch": 1.5246251474139383,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004777239674276941,
      "loss": 5.1746,
      "mean_token_accuracy": 0.20370774418115617,
      "num_tokens": 30289631.0,
      "step": 13575
    },
    {
      "entropy": 5.425937175750732,
      "epoch": 1.525186724321896,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004777065465370977,
      "loss": 5.289,
      "mean_token_accuracy": 0.18958547562360764,
      "num_tokens": 30301433.0,
      "step": 13580
    },
    {
      "entropy": 5.513584280014038,
      "epoch": 1.5257483012298534,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004776891191922463,
      "loss": 5.2366,
      "mean_token_accuracy": 0.19540890604257583,
      "num_tokens": 30312695.0,
      "step": 13585
    },
    {
      "entropy": 5.404104709625244,
      "epoch": 1.5263098781378108,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004776716853936947,
      "loss": 5.3017,
      "mean_token_accuracy": 0.19057883322238922,
      "num_tokens": 30324254.0,
      "step": 13590
    },
    {
      "entropy": 5.495215129852295,
      "epoch": 1.5268714550457685,
      "grad_norm": 1.1875,
      "learning_rate": 0.000477654245141998,
      "loss": 5.3482,
      "mean_token_accuracy": 0.1910526528954506,
      "num_tokens": 30335389.0,
      "step": 13595
    },
    {
      "entropy": 5.454211473464966,
      "epoch": 1.5274330319537262,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004776367984377116,
      "loss": 5.2309,
      "mean_token_accuracy": 0.19881890267133712,
      "num_tokens": 30346627.0,
      "step": 13600
    },
    {
      "entropy": 5.427598476409912,
      "epoch": 1.5279946088616836,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047761934528139086,
      "loss": 5.3265,
      "mean_token_accuracy": 0.18906411081552504,
      "num_tokens": 30358215.0,
      "step": 13605
    },
    {
      "entropy": 5.437365245819092,
      "epoch": 1.528556185769641,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004776018856735916,
      "loss": 5.2087,
      "mean_token_accuracy": 0.1910713493824005,
      "num_tokens": 30369793.0,
      "step": 13610
    },
    {
      "entropy": 5.450023746490478,
      "epoch": 1.5291177626775987,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004775844196148698,
      "loss": 5.1687,
      "mean_token_accuracy": 0.19769091308116912,
      "num_tokens": 30380200.0,
      "step": 13615
    },
    {
      "entropy": 5.366498184204102,
      "epoch": 1.5296793395855564,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047756694710578146,
      "loss": 5.3466,
      "mean_token_accuracy": 0.18621691763401033,
      "num_tokens": 30392282.0,
      "step": 13620
    },
    {
      "entropy": 5.615476131439209,
      "epoch": 1.5302409164935138,
      "grad_norm": 1.125,
      "learning_rate": 0.000477549468146883,
      "loss": 5.503,
      "mean_token_accuracy": 0.1757819816470146,
      "num_tokens": 30405365.0,
      "step": 13625
    },
    {
      "entropy": 5.574750185012817,
      "epoch": 1.5308024934014712,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004775319827387309,
      "loss": 5.3089,
      "mean_token_accuracy": 0.19490093141794204,
      "num_tokens": 30416513.0,
      "step": 13630
    },
    {
      "entropy": 5.375363969802857,
      "epoch": 1.531364070309429,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004775144908818818,
      "loss": 5.1784,
      "mean_token_accuracy": 0.19661412090063096,
      "num_tokens": 30427245.0,
      "step": 13635
    },
    {
      "entropy": 5.496862936019897,
      "epoch": 1.5319256472173866,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047749699257689296,
      "loss": 5.3139,
      "mean_token_accuracy": 0.1938263490796089,
      "num_tokens": 30438437.0,
      "step": 13640
    },
    {
      "entropy": 5.4636803150177,
      "epoch": 1.532487224125344,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004774794878243212,
      "loss": 5.273,
      "mean_token_accuracy": 0.1933994248509407,
      "num_tokens": 30449726.0,
      "step": 13645
    },
    {
      "entropy": 5.491350173950195,
      "epoch": 1.5330488010333014,
      "grad_norm": 1.25,
      "learning_rate": 0.000477461976624724,
      "loss": 5.3146,
      "mean_token_accuracy": 0.18998759090900422,
      "num_tokens": 30460024.0,
      "step": 13650
    },
    {
      "entropy": 5.526596307754517,
      "epoch": 1.533610377941259,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047744445897865905,
      "loss": 5.4007,
      "mean_token_accuracy": 0.18470858931541442,
      "num_tokens": 30471981.0,
      "step": 13655
    },
    {
      "entropy": 5.429630422592163,
      "epoch": 1.5341719548492168,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000477426934886684,
      "loss": 5.2114,
      "mean_token_accuracy": 0.2044860377907753,
      "num_tokens": 30483143.0,
      "step": 13660
    },
    {
      "entropy": 5.406258630752563,
      "epoch": 1.5347335317571742,
      "grad_norm": 1.125,
      "learning_rate": 0.00047740940434935677,
      "loss": 5.2386,
      "mean_token_accuracy": 0.1919554814696312,
      "num_tokens": 30495049.0,
      "step": 13665
    },
    {
      "entropy": 5.483577060699463,
      "epoch": 1.5352951086651316,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004773918673672356,
      "loss": 5.3508,
      "mean_token_accuracy": 0.19450055211782455,
      "num_tokens": 30506736.0,
      "step": 13670
    },
    {
      "entropy": 5.629188823699951,
      "epoch": 1.5358566855730893,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047737432394087887,
      "loss": 5.4345,
      "mean_token_accuracy": 0.18511937707662582,
      "num_tokens": 30520130.0,
      "step": 13675
    },
    {
      "entropy": 5.537339210510254,
      "epoch": 1.5364182624810467,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047735677407084525,
      "loss": 5.4218,
      "mean_token_accuracy": 0.18245114833116532,
      "num_tokens": 30531480.0,
      "step": 13680
    },
    {
      "entropy": 5.423575067520142,
      "epoch": 1.5369798393890042,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004773392177576934,
      "loss": 5.2523,
      "mean_token_accuracy": 0.18773407191038133,
      "num_tokens": 30541760.0,
      "step": 13685
    },
    {
      "entropy": 5.507688808441162,
      "epoch": 1.5375414162969618,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004773216550019824,
      "loss": 5.2761,
      "mean_token_accuracy": 0.19846475422382354,
      "num_tokens": 30551162.0,
      "step": 13690
    },
    {
      "entropy": 5.474565601348877,
      "epoch": 1.5381029932049195,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047730408580427136,
      "loss": 5.2491,
      "mean_token_accuracy": 0.19227711856365204,
      "num_tokens": 30562601.0,
      "step": 13695
    },
    {
      "entropy": 5.475502872467041,
      "epoch": 1.538664570112877,
      "grad_norm": 1.125,
      "learning_rate": 0.00047728651016511994,
      "loss": 5.3154,
      "mean_token_accuracy": 0.19412962943315507,
      "num_tokens": 30573923.0,
      "step": 13700
    },
    {
      "entropy": 5.496020746231079,
      "epoch": 1.5392261470208344,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047726892808508744,
      "loss": 5.3805,
      "mean_token_accuracy": 0.18748225420713424,
      "num_tokens": 30586400.0,
      "step": 13705
    },
    {
      "entropy": 5.531232070922852,
      "epoch": 1.539787723928792,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004772513395647339,
      "loss": 5.3567,
      "mean_token_accuracy": 0.18611519485712053,
      "num_tokens": 30598317.0,
      "step": 13710
    },
    {
      "entropy": 5.49639368057251,
      "epoch": 1.5403493008367497,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004772337446046192,
      "loss": 5.2962,
      "mean_token_accuracy": 0.19507516622543336,
      "num_tokens": 30609413.0,
      "step": 13715
    },
    {
      "entropy": 5.521975517272949,
      "epoch": 1.5409108777447071,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047721614320530366,
      "loss": 5.2434,
      "mean_token_accuracy": 0.1902339369058609,
      "num_tokens": 30619252.0,
      "step": 13720
    },
    {
      "entropy": 5.44996829032898,
      "epoch": 1.5414724546526646,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047719853536734764,
      "loss": 5.2416,
      "mean_token_accuracy": 0.19175414890050888,
      "num_tokens": 30630477.0,
      "step": 13725
    },
    {
      "entropy": 5.423915815353394,
      "epoch": 1.5420340315606222,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004771809210913119,
      "loss": 5.314,
      "mean_token_accuracy": 0.19146976470947266,
      "num_tokens": 30641977.0,
      "step": 13730
    },
    {
      "entropy": 5.503075313568115,
      "epoch": 1.54259560846858,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047716330037775724,
      "loss": 5.2247,
      "mean_token_accuracy": 0.19324810951948165,
      "num_tokens": 30653547.0,
      "step": 13735
    },
    {
      "entropy": 5.457059907913208,
      "epoch": 1.5431571853765373,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004771456732272446,
      "loss": 5.2535,
      "mean_token_accuracy": 0.19154204577207565,
      "num_tokens": 30664270.0,
      "step": 13740
    },
    {
      "entropy": 5.567150449752807,
      "epoch": 1.5437187622844948,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047712803964033545,
      "loss": 5.475,
      "mean_token_accuracy": 0.17565417438745498,
      "num_tokens": 30677172.0,
      "step": 13745
    },
    {
      "entropy": 5.475821447372437,
      "epoch": 1.5442803391924524,
      "grad_norm": 1.125,
      "learning_rate": 0.00047711039961759107,
      "loss": 5.3367,
      "mean_token_accuracy": 0.1930075019598007,
      "num_tokens": 30688657.0,
      "step": 13750
    },
    {
      "entropy": 5.500812196731568,
      "epoch": 1.54484191610041,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004770927531595731,
      "loss": 5.3378,
      "mean_token_accuracy": 0.18685498982667922,
      "num_tokens": 30699103.0,
      "step": 13755
    },
    {
      "entropy": 5.490113925933838,
      "epoch": 1.5454034930083675,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047707510026684353,
      "loss": 5.2734,
      "mean_token_accuracy": 0.19076599627733232,
      "num_tokens": 30711352.0,
      "step": 13760
    },
    {
      "entropy": 5.541574430465698,
      "epoch": 1.545965069916325,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047705744093996447,
      "loss": 5.3555,
      "mean_token_accuracy": 0.18642870038747789,
      "num_tokens": 30722224.0,
      "step": 13765
    },
    {
      "entropy": 5.562950897216797,
      "epoch": 1.5465266468242826,
      "grad_norm": 1.21875,
      "learning_rate": 0.000477039775179498,
      "loss": 5.3292,
      "mean_token_accuracy": 0.18807446956634521,
      "num_tokens": 30733113.0,
      "step": 13770
    },
    {
      "entropy": 5.459945726394653,
      "epoch": 1.54708822373224,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047702210298600684,
      "loss": 5.3318,
      "mean_token_accuracy": 0.19380075633525848,
      "num_tokens": 30744903.0,
      "step": 13775
    },
    {
      "entropy": 5.4853616714477536,
      "epoch": 1.5476498006401975,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047700442436005356,
      "loss": 5.3211,
      "mean_token_accuracy": 0.18780923783779144,
      "num_tokens": 30756034.0,
      "step": 13780
    },
    {
      "entropy": 5.487456512451172,
      "epoch": 1.5482113775481552,
      "grad_norm": 1.265625,
      "learning_rate": 0.000476986739302201,
      "loss": 5.2364,
      "mean_token_accuracy": 0.1966862142086029,
      "num_tokens": 30766139.0,
      "step": 13785
    },
    {
      "entropy": 5.465761947631836,
      "epoch": 1.5487729544561128,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047696904781301233,
      "loss": 5.3245,
      "mean_token_accuracy": 0.18206168562173844,
      "num_tokens": 30777505.0,
      "step": 13790
    },
    {
      "entropy": 5.353411912918091,
      "epoch": 1.5493345313640703,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004769513498930508,
      "loss": 5.2027,
      "mean_token_accuracy": 0.19901520907878875,
      "num_tokens": 30788520.0,
      "step": 13795
    },
    {
      "entropy": 5.449715709686279,
      "epoch": 1.5498961082720277,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047693364554288,
      "loss": 5.208,
      "mean_token_accuracy": 0.19517993777990342,
      "num_tokens": 30799051.0,
      "step": 13800
    },
    {
      "entropy": 5.455475902557373,
      "epoch": 1.5504576851799854,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047691593476306373,
      "loss": 5.2275,
      "mean_token_accuracy": 0.204378379881382,
      "num_tokens": 30809375.0,
      "step": 13805
    },
    {
      "entropy": 5.428431701660156,
      "epoch": 1.551019262087943,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004768982175541656,
      "loss": 5.2257,
      "mean_token_accuracy": 0.1969002902507782,
      "num_tokens": 30821330.0,
      "step": 13810
    },
    {
      "entropy": 5.534437751770019,
      "epoch": 1.5515808389959005,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047688049391675006,
      "loss": 5.368,
      "mean_token_accuracy": 0.1871736854314804,
      "num_tokens": 30832366.0,
      "step": 13815
    },
    {
      "entropy": 5.482333183288574,
      "epoch": 1.552142415903858,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004768627638513813,
      "loss": 5.3016,
      "mean_token_accuracy": 0.18738782554864883,
      "num_tokens": 30843550.0,
      "step": 13820
    },
    {
      "entropy": 5.372391176223755,
      "epoch": 1.5527039928118156,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004768450273586237,
      "loss": 5.1525,
      "mean_token_accuracy": 0.1979817271232605,
      "num_tokens": 30854727.0,
      "step": 13825
    },
    {
      "entropy": 5.497431325912475,
      "epoch": 1.5532655697197733,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047682728443904223,
      "loss": 5.2617,
      "mean_token_accuracy": 0.1946622312068939,
      "num_tokens": 30865594.0,
      "step": 13830
    },
    {
      "entropy": 5.494666481018067,
      "epoch": 1.5538271466277307,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047680953509320176,
      "loss": 5.2837,
      "mean_token_accuracy": 0.19296862930059433,
      "num_tokens": 30877386.0,
      "step": 13835
    },
    {
      "entropy": 5.570396089553833,
      "epoch": 1.5543887235356881,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004767917793216674,
      "loss": 5.366,
      "mean_token_accuracy": 0.19092869758605957,
      "num_tokens": 30889052.0,
      "step": 13840
    },
    {
      "entropy": 5.477635049819947,
      "epoch": 1.5549503004436458,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047677401712500455,
      "loss": 5.2301,
      "mean_token_accuracy": 0.1887301117181778,
      "num_tokens": 30900524.0,
      "step": 13845
    },
    {
      "entropy": 5.458941411972046,
      "epoch": 1.5555118773516035,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004767562485037787,
      "loss": 5.3572,
      "mean_token_accuracy": 0.18743327111005784,
      "num_tokens": 30911770.0,
      "step": 13850
    },
    {
      "entropy": 5.4122730731964115,
      "epoch": 1.556073454259561,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047673847345855566,
      "loss": 5.1544,
      "mean_token_accuracy": 0.19762712121009826,
      "num_tokens": 30921662.0,
      "step": 13855
    },
    {
      "entropy": 5.591034746170044,
      "epoch": 1.5566350311675183,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004767206919899014,
      "loss": 5.4347,
      "mean_token_accuracy": 0.1897072896361351,
      "num_tokens": 30934138.0,
      "step": 13860
    },
    {
      "entropy": 5.498001670837402,
      "epoch": 1.557196608075476,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000476702904098382,
      "loss": 5.2659,
      "mean_token_accuracy": 0.19060425013303756,
      "num_tokens": 30945561.0,
      "step": 13865
    },
    {
      "entropy": 5.383753681182862,
      "epoch": 1.5577581849834334,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000476685109784564,
      "loss": 5.1583,
      "mean_token_accuracy": 0.19964545220136642,
      "num_tokens": 30957210.0,
      "step": 13870
    },
    {
      "entropy": 5.438637161254883,
      "epoch": 1.5583197618913909,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047666730904901377,
      "loss": 5.2787,
      "mean_token_accuracy": 0.1871353104710579,
      "num_tokens": 30968782.0,
      "step": 13875
    },
    {
      "entropy": 5.4773211002349855,
      "epoch": 1.5588813387993485,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047664950189229833,
      "loss": 5.2911,
      "mean_token_accuracy": 0.18637899905443192,
      "num_tokens": 30981239.0,
      "step": 13880
    },
    {
      "entropy": 5.518683528900146,
      "epoch": 1.5594429157073062,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004766316883149845,
      "loss": 5.3356,
      "mean_token_accuracy": 0.19096245616674423,
      "num_tokens": 30991773.0,
      "step": 13885
    },
    {
      "entropy": 5.5196739673614506,
      "epoch": 1.5600044926152636,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047661386831763955,
      "loss": 5.3321,
      "mean_token_accuracy": 0.19574459344148637,
      "num_tokens": 31002646.0,
      "step": 13890
    },
    {
      "entropy": 5.443702125549317,
      "epoch": 1.560566069523221,
      "grad_norm": 1.8046875,
      "learning_rate": 0.00047659604190083076,
      "loss": 5.2921,
      "mean_token_accuracy": 0.193208746612072,
      "num_tokens": 31014405.0,
      "step": 13895
    },
    {
      "entropy": 5.507214546203613,
      "epoch": 1.5611276464311787,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047657820906512585,
      "loss": 5.2218,
      "mean_token_accuracy": 0.19418345391750336,
      "num_tokens": 31025529.0,
      "step": 13900
    },
    {
      "entropy": 5.5514867305755615,
      "epoch": 1.5616892233391364,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047656036981109256,
      "loss": 5.3795,
      "mean_token_accuracy": 0.1843936875462532,
      "num_tokens": 31037558.0,
      "step": 13905
    },
    {
      "entropy": 5.337411165237427,
      "epoch": 1.5622508002470938,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047654252413929896,
      "loss": 5.147,
      "mean_token_accuracy": 0.20180606096982956,
      "num_tokens": 31048918.0,
      "step": 13910
    },
    {
      "entropy": 5.396874713897705,
      "epoch": 1.5628123771550513,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047652467205031317,
      "loss": 5.2374,
      "mean_token_accuracy": 0.19719748497009276,
      "num_tokens": 31060149.0,
      "step": 13915
    },
    {
      "entropy": 5.416375398635864,
      "epoch": 1.563373954063009,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047650681354470363,
      "loss": 5.2587,
      "mean_token_accuracy": 0.1959821715950966,
      "num_tokens": 31071197.0,
      "step": 13920
    },
    {
      "entropy": 5.520791292190552,
      "epoch": 1.5639355309709666,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047648894862303907,
      "loss": 5.3623,
      "mean_token_accuracy": 0.19651337414979936,
      "num_tokens": 31083385.0,
      "step": 13925
    },
    {
      "entropy": 5.477852535247803,
      "epoch": 1.564497107878924,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047647107728588826,
      "loss": 5.2949,
      "mean_token_accuracy": 0.18428505659103395,
      "num_tokens": 31094807.0,
      "step": 13930
    },
    {
      "entropy": 5.462962007522583,
      "epoch": 1.5650586847868815,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047645319953382024,
      "loss": 5.1889,
      "mean_token_accuracy": 0.19819659739732742,
      "num_tokens": 31105622.0,
      "step": 13935
    },
    {
      "entropy": 5.507463264465332,
      "epoch": 1.5656202616948391,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047643531536740405,
      "loss": 5.2197,
      "mean_token_accuracy": 0.19653423875570297,
      "num_tokens": 31115684.0,
      "step": 13940
    },
    {
      "entropy": 5.376924514770508,
      "epoch": 1.5661818386027968,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004764174247872094,
      "loss": 5.191,
      "mean_token_accuracy": 0.19938337057828903,
      "num_tokens": 31126449.0,
      "step": 13945
    },
    {
      "entropy": 5.431071853637695,
      "epoch": 1.5667434155107542,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047639952779380586,
      "loss": 5.2171,
      "mean_token_accuracy": 0.20294841825962068,
      "num_tokens": 31136894.0,
      "step": 13950
    },
    {
      "entropy": 5.470840835571289,
      "epoch": 1.5673049924187117,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004763816243877632,
      "loss": 5.3244,
      "mean_token_accuracy": 0.19289205521345137,
      "num_tokens": 31147131.0,
      "step": 13955
    },
    {
      "entropy": 5.407017087936401,
      "epoch": 1.5678665693266693,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004763637145696515,
      "loss": 5.1536,
      "mean_token_accuracy": 0.1944940909743309,
      "num_tokens": 31157903.0,
      "step": 13960
    },
    {
      "entropy": 5.4850743293762205,
      "epoch": 1.5684281462346268,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000476345798340041,
      "loss": 5.3247,
      "mean_token_accuracy": 0.1908445179462433,
      "num_tokens": 31169087.0,
      "step": 13965
    },
    {
      "entropy": 5.567381381988525,
      "epoch": 1.5689897231425842,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004763278756995022,
      "loss": 5.3732,
      "mean_token_accuracy": 0.18631485104560852,
      "num_tokens": 31180253.0,
      "step": 13970
    },
    {
      "entropy": 5.532321500778198,
      "epoch": 1.5695513000505419,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047630994664860576,
      "loss": 5.3198,
      "mean_token_accuracy": 0.19080108553171157,
      "num_tokens": 31191857.0,
      "step": 13975
    },
    {
      "entropy": 5.495871114730835,
      "epoch": 1.5701128769584995,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004762920111879226,
      "loss": 5.3544,
      "mean_token_accuracy": 0.1833937257528305,
      "num_tokens": 31201868.0,
      "step": 13980
    },
    {
      "entropy": 5.515967941284179,
      "epoch": 1.570674453866457,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004762740693180237,
      "loss": 5.2911,
      "mean_token_accuracy": 0.1966903477907181,
      "num_tokens": 31212716.0,
      "step": 13985
    },
    {
      "entropy": 5.469483280181885,
      "epoch": 1.5712360307744144,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047625612103948035,
      "loss": 5.3763,
      "mean_token_accuracy": 0.18465807139873505,
      "num_tokens": 31224823.0,
      "step": 13990
    },
    {
      "entropy": 5.486267852783203,
      "epoch": 1.571797607682372,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004762381663528641,
      "loss": 5.2576,
      "mean_token_accuracy": 0.1944064050912857,
      "num_tokens": 31236798.0,
      "step": 13995
    },
    {
      "entropy": 5.497094249725341,
      "epoch": 1.5723591845903298,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004762202052587466,
      "loss": 5.2839,
      "mean_token_accuracy": 0.1845656543970108,
      "num_tokens": 31248687.0,
      "step": 14000
    },
    {
      "entropy": 5.4764666080474855,
      "epoch": 1.5729207614982872,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004762022377576996,
      "loss": 5.3515,
      "mean_token_accuracy": 0.1868732675909996,
      "num_tokens": 31258385.0,
      "step": 14005
    },
    {
      "entropy": 5.551736307144165,
      "epoch": 1.5734823384062446,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004761842638502954,
      "loss": 5.3399,
      "mean_token_accuracy": 0.1913232311606407,
      "num_tokens": 31269308.0,
      "step": 14010
    },
    {
      "entropy": 5.53936014175415,
      "epoch": 1.5740439153142023,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004761662835371062,
      "loss": 5.2824,
      "mean_token_accuracy": 0.19593806862831115,
      "num_tokens": 31279900.0,
      "step": 14015
    },
    {
      "entropy": 5.420303153991699,
      "epoch": 1.57460549222216,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004761482968187045,
      "loss": 5.2762,
      "mean_token_accuracy": 0.20072393268346786,
      "num_tokens": 31290456.0,
      "step": 14020
    },
    {
      "entropy": 5.409570646286011,
      "epoch": 1.5751670691301174,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004761303036956631,
      "loss": 5.2384,
      "mean_token_accuracy": 0.1903083935379982,
      "num_tokens": 31301398.0,
      "step": 14025
    },
    {
      "entropy": 5.595342063903809,
      "epoch": 1.5757286460380748,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004761123041685547,
      "loss": 5.3471,
      "mean_token_accuracy": 0.19225261211395264,
      "num_tokens": 31311149.0,
      "step": 14030
    },
    {
      "entropy": 5.372718000411988,
      "epoch": 1.5762902229460325,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004760942982379526,
      "loss": 5.176,
      "mean_token_accuracy": 0.19950869232416152,
      "num_tokens": 31321099.0,
      "step": 14035
    },
    {
      "entropy": 5.412204122543335,
      "epoch": 1.5768517998539902,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047607628590443004,
      "loss": 5.3,
      "mean_token_accuracy": 0.18769247978925704,
      "num_tokens": 31331872.0,
      "step": 14040
    },
    {
      "entropy": 5.5768355369567875,
      "epoch": 1.5774133767619476,
      "grad_norm": 1.9140625,
      "learning_rate": 0.00047605826716856056,
      "loss": 5.337,
      "mean_token_accuracy": 0.18533363044261933,
      "num_tokens": 31342024.0,
      "step": 14045
    },
    {
      "entropy": 5.505996179580689,
      "epoch": 1.577974953669905,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004760402420309178,
      "loss": 5.1876,
      "mean_token_accuracy": 0.20012515187263488,
      "num_tokens": 31351857.0,
      "step": 14050
    },
    {
      "entropy": 5.460606908798217,
      "epoch": 1.5785365305778627,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004760222104920759,
      "loss": 5.3027,
      "mean_token_accuracy": 0.18984030783176423,
      "num_tokens": 31362609.0,
      "step": 14055
    },
    {
      "entropy": 5.395387363433838,
      "epoch": 1.5790981074858201,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004760041725526088,
      "loss": 5.2094,
      "mean_token_accuracy": 0.19440712928771972,
      "num_tokens": 31373129.0,
      "step": 14060
    },
    {
      "entropy": 5.457586860656738,
      "epoch": 1.5796596843937776,
      "grad_norm": 1.125,
      "learning_rate": 0.00047598612821309077,
      "loss": 5.1746,
      "mean_token_accuracy": 0.19568442702293395,
      "num_tokens": 31383109.0,
      "step": 14065
    },
    {
      "entropy": 5.388409423828125,
      "epoch": 1.5802212613017352,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004759680774740966,
      "loss": 5.1676,
      "mean_token_accuracy": 0.19534103125333785,
      "num_tokens": 31393603.0,
      "step": 14070
    },
    {
      "entropy": 5.464620733261109,
      "epoch": 1.580782838209693,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004759500203362008,
      "loss": 5.2553,
      "mean_token_accuracy": 0.1912466749548912,
      "num_tokens": 31404011.0,
      "step": 14075
    },
    {
      "entropy": 5.430010080337524,
      "epoch": 1.5813444151176503,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004759319567999785,
      "loss": 5.1964,
      "mean_token_accuracy": 0.19172051250934602,
      "num_tokens": 31414239.0,
      "step": 14080
    },
    {
      "entropy": 5.400055646896362,
      "epoch": 1.5819059920256078,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004759138868660048,
      "loss": 5.306,
      "mean_token_accuracy": 0.18827413767576218,
      "num_tokens": 31425055.0,
      "step": 14085
    },
    {
      "entropy": 5.491458940505981,
      "epoch": 1.5824675689335654,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047589581053485494,
      "loss": 5.2705,
      "mean_token_accuracy": 0.19474028348922728,
      "num_tokens": 31435537.0,
      "step": 14090
    },
    {
      "entropy": 5.5705078125,
      "epoch": 1.583029145841523,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047587772780710455,
      "loss": 5.3315,
      "mean_token_accuracy": 0.19486101865768432,
      "num_tokens": 31446056.0,
      "step": 14095
    },
    {
      "entropy": 5.497781133651733,
      "epoch": 1.5835907227494805,
      "grad_norm": 1.40625,
      "learning_rate": 0.00047585963868332945,
      "loss": 5.3008,
      "mean_token_accuracy": 0.19586532413959504,
      "num_tokens": 31457452.0,
      "step": 14100
    },
    {
      "entropy": 5.395171451568603,
      "epoch": 1.584152299657438,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004758415431641055,
      "loss": 5.252,
      "mean_token_accuracy": 0.19545141905546187,
      "num_tokens": 31469038.0,
      "step": 14105
    },
    {
      "entropy": 5.539952659606934,
      "epoch": 1.5847138765653956,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004758234412500089,
      "loss": 5.3486,
      "mean_token_accuracy": 0.18950719088315965,
      "num_tokens": 31481069.0,
      "step": 14110
    },
    {
      "entropy": 5.495297622680664,
      "epoch": 1.5852754534733533,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047580533294161616,
      "loss": 5.3072,
      "mean_token_accuracy": 0.18851478099823,
      "num_tokens": 31491393.0,
      "step": 14115
    },
    {
      "entropy": 5.4381715774536135,
      "epoch": 1.5858370303813107,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004757872182395036,
      "loss": 5.3042,
      "mean_token_accuracy": 0.1880606859922409,
      "num_tokens": 31502627.0,
      "step": 14120
    },
    {
      "entropy": 5.595828485488892,
      "epoch": 1.5863986072892682,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047576909714424823,
      "loss": 5.3147,
      "mean_token_accuracy": 0.18846513330936432,
      "num_tokens": 31513995.0,
      "step": 14125
    },
    {
      "entropy": 5.449325132369995,
      "epoch": 1.5869601841972258,
      "grad_norm": 1.25,
      "learning_rate": 0.00047575096965642687,
      "loss": 5.2598,
      "mean_token_accuracy": 0.19191921800374984,
      "num_tokens": 31524625.0,
      "step": 14130
    },
    {
      "entropy": 5.3978783130645756,
      "epoch": 1.5875217611051835,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047573283577661684,
      "loss": 5.2277,
      "mean_token_accuracy": 0.19620107114315033,
      "num_tokens": 31535940.0,
      "step": 14135
    },
    {
      "entropy": 5.529709148406982,
      "epoch": 1.588083338013141,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004757146955053955,
      "loss": 5.2804,
      "mean_token_accuracy": 0.19309903383255006,
      "num_tokens": 31546557.0,
      "step": 14140
    },
    {
      "entropy": 5.447588586807251,
      "epoch": 1.5886449149210984,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047569654884334036,
      "loss": 5.3091,
      "mean_token_accuracy": 0.18750424832105636,
      "num_tokens": 31559335.0,
      "step": 14145
    },
    {
      "entropy": 5.467824649810791,
      "epoch": 1.589206491829056,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004756783957910292,
      "loss": 5.2496,
      "mean_token_accuracy": 0.1899348571896553,
      "num_tokens": 31570698.0,
      "step": 14150
    },
    {
      "entropy": 5.474661731719971,
      "epoch": 1.5897680687370135,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047566023634904015,
      "loss": 5.3229,
      "mean_token_accuracy": 0.1926109239459038,
      "num_tokens": 31582524.0,
      "step": 14155
    },
    {
      "entropy": 5.4988922595977785,
      "epoch": 1.590329645644971,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004756420705179513,
      "loss": 5.2604,
      "mean_token_accuracy": 0.19487905651330947,
      "num_tokens": 31593399.0,
      "step": 14160
    },
    {
      "entropy": 5.447623300552368,
      "epoch": 1.5908912225529286,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004756238982983412,
      "loss": 5.2792,
      "mean_token_accuracy": 0.1858401343226433,
      "num_tokens": 31604169.0,
      "step": 14165
    },
    {
      "entropy": 5.461736965179443,
      "epoch": 1.5914527994608862,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004756057196907883,
      "loss": 5.2101,
      "mean_token_accuracy": 0.19511415511369706,
      "num_tokens": 31614671.0,
      "step": 14170
    },
    {
      "entropy": 5.398337459564209,
      "epoch": 1.5920143763688437,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047558753469587154,
      "loss": 5.2191,
      "mean_token_accuracy": 0.19570247530937196,
      "num_tokens": 31625793.0,
      "step": 14175
    },
    {
      "entropy": 5.431139659881592,
      "epoch": 1.5925759532768011,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047556934331416986,
      "loss": 5.2222,
      "mean_token_accuracy": 0.19889807999134063,
      "num_tokens": 31637632.0,
      "step": 14180
    },
    {
      "entropy": 5.4438300132751465,
      "epoch": 1.5931375301847588,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004755511455462624,
      "loss": 5.2198,
      "mean_token_accuracy": 0.19716380387544633,
      "num_tokens": 31648314.0,
      "step": 14185
    },
    {
      "entropy": 5.475461483001709,
      "epoch": 1.5936991070927164,
      "grad_norm": 1.25,
      "learning_rate": 0.0004755329413927287,
      "loss": 5.2704,
      "mean_token_accuracy": 0.198131063580513,
      "num_tokens": 31659152.0,
      "step": 14190
    },
    {
      "entropy": 5.465002870559692,
      "epoch": 1.5942606840006739,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004755147308541485,
      "loss": 5.2073,
      "mean_token_accuracy": 0.19407492578029634,
      "num_tokens": 31669837.0,
      "step": 14195
    },
    {
      "entropy": 5.420690250396729,
      "epoch": 1.5948222609086313,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047549651393110145,
      "loss": 5.2661,
      "mean_token_accuracy": 0.1941886469721794,
      "num_tokens": 31680663.0,
      "step": 14200
    },
    {
      "entropy": 5.47969160079956,
      "epoch": 1.595383837816589,
      "grad_norm": 1.375,
      "learning_rate": 0.00047547829062416757,
      "loss": 5.2449,
      "mean_token_accuracy": 0.1896169051527977,
      "num_tokens": 31691484.0,
      "step": 14205
    },
    {
      "entropy": 5.447725629806518,
      "epoch": 1.5959454147245467,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047546006093392726,
      "loss": 5.2251,
      "mean_token_accuracy": 0.19016451090574266,
      "num_tokens": 31701642.0,
      "step": 14210
    },
    {
      "entropy": 5.423415756225586,
      "epoch": 1.596506991632504,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004754418248609608,
      "loss": 5.2674,
      "mean_token_accuracy": 0.1946104273200035,
      "num_tokens": 31714203.0,
      "step": 14215
    },
    {
      "entropy": 5.505327463150024,
      "epoch": 1.5970685685404615,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004754235824058489,
      "loss": 5.2521,
      "mean_token_accuracy": 0.19306391179561616,
      "num_tokens": 31725890.0,
      "step": 14220
    },
    {
      "entropy": 5.4156678199768065,
      "epoch": 1.5976301454484192,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004754053335691724,
      "loss": 5.1889,
      "mean_token_accuracy": 0.20101438462734222,
      "num_tokens": 31738058.0,
      "step": 14225
    },
    {
      "entropy": 5.426040697097778,
      "epoch": 1.5981917223563769,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047538707835151237,
      "loss": 5.2753,
      "mean_token_accuracy": 0.19283527582883836,
      "num_tokens": 31748983.0,
      "step": 14230
    },
    {
      "entropy": 5.500078582763672,
      "epoch": 1.5987532992643343,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047536881675345,
      "loss": 5.2517,
      "mean_token_accuracy": 0.19759280681610109,
      "num_tokens": 31760083.0,
      "step": 14235
    },
    {
      "entropy": 5.433893489837646,
      "epoch": 1.5993148761722917,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004753505487755669,
      "loss": 5.1971,
      "mean_token_accuracy": 0.19693499952554702,
      "num_tokens": 31771627.0,
      "step": 14240
    },
    {
      "entropy": 5.458097457885742,
      "epoch": 1.5998764530802494,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047533227441844453,
      "loss": 5.3228,
      "mean_token_accuracy": 0.19433703422546386,
      "num_tokens": 31783476.0,
      "step": 14245
    },
    {
      "entropy": 5.516911506652832,
      "epoch": 1.6004380299882068,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047531399368266485,
      "loss": 5.3162,
      "mean_token_accuracy": 0.18814231008291243,
      "num_tokens": 31794072.0,
      "step": 14250
    },
    {
      "entropy": 5.446543645858765,
      "epoch": 1.6009996068961643,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004752957065688099,
      "loss": 5.202,
      "mean_token_accuracy": 0.1968376263976097,
      "num_tokens": 31803613.0,
      "step": 14255
    },
    {
      "entropy": 5.547620534896851,
      "epoch": 1.601561183804122,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000475277413077462,
      "loss": 5.3245,
      "mean_token_accuracy": 0.19501662254333496,
      "num_tokens": 31815018.0,
      "step": 14260
    },
    {
      "entropy": 5.36879301071167,
      "epoch": 1.6021227607120796,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047525911320920363,
      "loss": 5.1735,
      "mean_token_accuracy": 0.19654709547758104,
      "num_tokens": 31825603.0,
      "step": 14265
    },
    {
      "entropy": 5.400392007827759,
      "epoch": 1.602684337620037,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004752408069646174,
      "loss": 5.2311,
      "mean_token_accuracy": 0.20013221502304077,
      "num_tokens": 31835988.0,
      "step": 14270
    },
    {
      "entropy": 5.422975730895996,
      "epoch": 1.6032459145279945,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00047522249434428623,
      "loss": 5.2825,
      "mean_token_accuracy": 0.19159705489873885,
      "num_tokens": 31846378.0,
      "step": 14275
    },
    {
      "entropy": 5.566601085662842,
      "epoch": 1.6038074914359521,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004752041753487931,
      "loss": 5.3408,
      "mean_token_accuracy": 0.18455637395381927,
      "num_tokens": 31858836.0,
      "step": 14280
    },
    {
      "entropy": 5.429432916641235,
      "epoch": 1.6043690683439098,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004751858499787215,
      "loss": 5.2091,
      "mean_token_accuracy": 0.19151537716388703,
      "num_tokens": 31869315.0,
      "step": 14285
    },
    {
      "entropy": 5.5195516586303714,
      "epoch": 1.6049306452518672,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004751675182346547,
      "loss": 5.4163,
      "mean_token_accuracy": 0.18376508057117463,
      "num_tokens": 31880527.0,
      "step": 14290
    },
    {
      "entropy": 5.490113639831543,
      "epoch": 1.6054922221598247,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047514918011717656,
      "loss": 5.2787,
      "mean_token_accuracy": 0.19512311965227128,
      "num_tokens": 31891890.0,
      "step": 14295
    },
    {
      "entropy": 5.503733348846436,
      "epoch": 1.6060537990677823,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047513083562687085,
      "loss": 5.2905,
      "mean_token_accuracy": 0.19429044872522355,
      "num_tokens": 31902735.0,
      "step": 14300
    },
    {
      "entropy": 5.434233570098877,
      "epoch": 1.60661537597574,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004751124847643217,
      "loss": 5.3456,
      "mean_token_accuracy": 0.1877756506204605,
      "num_tokens": 31914504.0,
      "step": 14305
    },
    {
      "entropy": 5.515947961807251,
      "epoch": 1.6071769528836974,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004750941275301135,
      "loss": 5.3032,
      "mean_token_accuracy": 0.1959688037633896,
      "num_tokens": 31925287.0,
      "step": 14310
    },
    {
      "entropy": 5.4531231880187985,
      "epoch": 1.6077385297916549,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004750757639248306,
      "loss": 5.219,
      "mean_token_accuracy": 0.20041099637746812,
      "num_tokens": 31935845.0,
      "step": 14315
    },
    {
      "entropy": 5.484633255004883,
      "epoch": 1.6083001066996125,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047505739394905776,
      "loss": 5.3076,
      "mean_token_accuracy": 0.1961750492453575,
      "num_tokens": 31946771.0,
      "step": 14320
    },
    {
      "entropy": 5.3933892250061035,
      "epoch": 1.6088616836075702,
      "grad_norm": 1.40625,
      "learning_rate": 0.00047503901760338003,
      "loss": 5.2757,
      "mean_token_accuracy": 0.18924074321985246,
      "num_tokens": 31957733.0,
      "step": 14325
    },
    {
      "entropy": 5.420727014541626,
      "epoch": 1.6094232605155276,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047502063488838225,
      "loss": 5.1752,
      "mean_token_accuracy": 0.20114617496728898,
      "num_tokens": 31967809.0,
      "step": 14330
    },
    {
      "entropy": 5.359949111938477,
      "epoch": 1.609984837423485,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047500224580464994,
      "loss": 5.1523,
      "mean_token_accuracy": 0.19856377094984054,
      "num_tokens": 31978673.0,
      "step": 14335
    },
    {
      "entropy": 5.39651665687561,
      "epoch": 1.6105464143314427,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047498385035276856,
      "loss": 5.2581,
      "mean_token_accuracy": 0.19494690895080566,
      "num_tokens": 31990535.0,
      "step": 14340
    },
    {
      "entropy": 5.473408460617065,
      "epoch": 1.6111079912394002,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004749654485333238,
      "loss": 5.324,
      "mean_token_accuracy": 0.19301400929689408,
      "num_tokens": 32002164.0,
      "step": 14345
    },
    {
      "entropy": 5.572358417510986,
      "epoch": 1.6116695681473576,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047494704034690163,
      "loss": 5.3712,
      "mean_token_accuracy": 0.1857966125011444,
      "num_tokens": 32013651.0,
      "step": 14350
    },
    {
      "entropy": 5.464628887176514,
      "epoch": 1.6122311450553153,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047492862579408816,
      "loss": 5.2712,
      "mean_token_accuracy": 0.1945928603410721,
      "num_tokens": 32025128.0,
      "step": 14355
    },
    {
      "entropy": 5.47275857925415,
      "epoch": 1.612792721963273,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047491020487546966,
      "loss": 5.298,
      "mean_token_accuracy": 0.1810721844434738,
      "num_tokens": 32036359.0,
      "step": 14360
    },
    {
      "entropy": 5.5123685836792,
      "epoch": 1.6133542988712304,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004748917775916327,
      "loss": 5.1947,
      "mean_token_accuracy": 0.20041578263044357,
      "num_tokens": 32047803.0,
      "step": 14365
    },
    {
      "entropy": 5.459750556945801,
      "epoch": 1.6139158757791878,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047487334394316405,
      "loss": 5.4016,
      "mean_token_accuracy": 0.18833719193935394,
      "num_tokens": 32060156.0,
      "step": 14370
    },
    {
      "entropy": 5.452860021591187,
      "epoch": 1.6144774526871455,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004748549039306506,
      "loss": 5.272,
      "mean_token_accuracy": 0.1957177698612213,
      "num_tokens": 32071398.0,
      "step": 14375
    },
    {
      "entropy": 5.379015970230102,
      "epoch": 1.6150390295951031,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004748364575546795,
      "loss": 5.1107,
      "mean_token_accuracy": 0.2025664985179901,
      "num_tokens": 32082472.0,
      "step": 14380
    },
    {
      "entropy": 5.4582263946533205,
      "epoch": 1.6156006065030606,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047481800481583806,
      "loss": 5.2859,
      "mean_token_accuracy": 0.19029460549354554,
      "num_tokens": 32093103.0,
      "step": 14385
    },
    {
      "entropy": 5.51110610961914,
      "epoch": 1.616162183411018,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004747995457147139,
      "loss": 5.3037,
      "mean_token_accuracy": 0.19805677384138107,
      "num_tokens": 32104526.0,
      "step": 14390
    },
    {
      "entropy": 5.48813681602478,
      "epoch": 1.6167237603189757,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047478108025189467,
      "loss": 5.3341,
      "mean_token_accuracy": 0.18239841163158416,
      "num_tokens": 32116018.0,
      "step": 14395
    },
    {
      "entropy": 5.390604257583618,
      "epoch": 1.6172853372269334,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047476260842796837,
      "loss": 5.1883,
      "mean_token_accuracy": 0.1991129621863365,
      "num_tokens": 32127385.0,
      "step": 14400
    },
    {
      "entropy": 5.458072710037231,
      "epoch": 1.6178469141348908,
      "grad_norm": 1.125,
      "learning_rate": 0.0004747441302435231,
      "loss": 5.3227,
      "mean_token_accuracy": 0.19236428290605545,
      "num_tokens": 32137376.0,
      "step": 14405
    },
    {
      "entropy": 5.503872632980347,
      "epoch": 1.6184084910428482,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047472564569914726,
      "loss": 5.3127,
      "mean_token_accuracy": 0.19224690198898314,
      "num_tokens": 32148263.0,
      "step": 14410
    },
    {
      "entropy": 5.428544044494629,
      "epoch": 1.618970067950806,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047470715479542945,
      "loss": 5.2475,
      "mean_token_accuracy": 0.19046173989772797,
      "num_tokens": 32161323.0,
      "step": 14415
    },
    {
      "entropy": 5.518643188476562,
      "epoch": 1.6195316448587636,
      "grad_norm": 1.125,
      "learning_rate": 0.0004746886575329584,
      "loss": 5.2976,
      "mean_token_accuracy": 0.1846601352095604,
      "num_tokens": 32172294.0,
      "step": 14420
    },
    {
      "entropy": 5.417508935928344,
      "epoch": 1.620093221766721,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047467015391232305,
      "loss": 5.2539,
      "mean_token_accuracy": 0.1844158113002777,
      "num_tokens": 32182619.0,
      "step": 14425
    },
    {
      "entropy": 5.371473169326782,
      "epoch": 1.6206547986746784,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047465164393411245,
      "loss": 5.2378,
      "mean_token_accuracy": 0.19385784715414048,
      "num_tokens": 32193078.0,
      "step": 14430
    },
    {
      "entropy": 5.4527122497558596,
      "epoch": 1.621216375582636,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047463312759891615,
      "loss": 5.2426,
      "mean_token_accuracy": 0.19074202477931976,
      "num_tokens": 32204134.0,
      "step": 14435
    },
    {
      "entropy": 5.542316150665283,
      "epoch": 1.6217779524905935,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004746146049073236,
      "loss": 5.3022,
      "mean_token_accuracy": 0.19227657616138458,
      "num_tokens": 32215136.0,
      "step": 14440
    },
    {
      "entropy": 5.447862815856934,
      "epoch": 1.622339529398551,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047459607585992466,
      "loss": 5.303,
      "mean_token_accuracy": 0.19188439548015596,
      "num_tokens": 32225727.0,
      "step": 14445
    },
    {
      "entropy": 5.449442958831787,
      "epoch": 1.6229011063065086,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047457754045730923,
      "loss": 5.2657,
      "mean_token_accuracy": 0.18790579587221146,
      "num_tokens": 32236376.0,
      "step": 14450
    },
    {
      "entropy": 5.433148717880249,
      "epoch": 1.6234626832144663,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047455899870006745,
      "loss": 5.2191,
      "mean_token_accuracy": 0.20057521015405655,
      "num_tokens": 32247202.0,
      "step": 14455
    },
    {
      "entropy": 5.382509994506836,
      "epoch": 1.6240242601224237,
      "grad_norm": 1.25,
      "learning_rate": 0.0004745404505887899,
      "loss": 5.3474,
      "mean_token_accuracy": 0.1836336463689804,
      "num_tokens": 32258625.0,
      "step": 14460
    },
    {
      "entropy": 5.472654962539673,
      "epoch": 1.6245858370303812,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004745218961240669,
      "loss": 5.2449,
      "mean_token_accuracy": 0.19256425350904466,
      "num_tokens": 32269288.0,
      "step": 14465
    },
    {
      "entropy": 5.522455167770386,
      "epoch": 1.6251474139383388,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004745033353064893,
      "loss": 5.2767,
      "mean_token_accuracy": 0.19322887063026428,
      "num_tokens": 32279888.0,
      "step": 14470
    },
    {
      "entropy": 5.476666259765625,
      "epoch": 1.6257089908462965,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004744847681366482,
      "loss": 5.283,
      "mean_token_accuracy": 0.19855345487594606,
      "num_tokens": 32291153.0,
      "step": 14475
    },
    {
      "entropy": 5.365588712692261,
      "epoch": 1.626270567754254,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047446619461513466,
      "loss": 5.1878,
      "mean_token_accuracy": 0.1974964588880539,
      "num_tokens": 32302291.0,
      "step": 14480
    },
    {
      "entropy": 5.439507246017456,
      "epoch": 1.6268321446622114,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004744476147425402,
      "loss": 5.1946,
      "mean_token_accuracy": 0.1998571202158928,
      "num_tokens": 32312015.0,
      "step": 14485
    },
    {
      "entropy": 5.461084413528442,
      "epoch": 1.627393721570169,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047442902851945634,
      "loss": 5.2916,
      "mean_token_accuracy": 0.18781921416521072,
      "num_tokens": 32323967.0,
      "step": 14490
    },
    {
      "entropy": 5.447282314300537,
      "epoch": 1.6279552984781267,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047441043594647486,
      "loss": 5.3055,
      "mean_token_accuracy": 0.1869613990187645,
      "num_tokens": 32336181.0,
      "step": 14495
    },
    {
      "entropy": 5.480967950820923,
      "epoch": 1.6285168753860841,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047439183702418765,
      "loss": 5.2424,
      "mean_token_accuracy": 0.19703955352306365,
      "num_tokens": 32346843.0,
      "step": 14500
    },
    {
      "entropy": 5.52333173751831,
      "epoch": 1.6290784522940416,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047437323175318716,
      "loss": 5.315,
      "mean_token_accuracy": 0.19371241182088852,
      "num_tokens": 32358511.0,
      "step": 14505
    },
    {
      "entropy": 5.434536790847778,
      "epoch": 1.6296400292019992,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047435462013406554,
      "loss": 5.2472,
      "mean_token_accuracy": 0.19466399550437927,
      "num_tokens": 32370319.0,
      "step": 14510
    },
    {
      "entropy": 5.420909738540649,
      "epoch": 1.630201606109957,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004743360021674155,
      "loss": 5.3126,
      "mean_token_accuracy": 0.19242074638605117,
      "num_tokens": 32381455.0,
      "step": 14515
    },
    {
      "entropy": 5.497006177902222,
      "epoch": 1.6307631830179143,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047431737785382996,
      "loss": 5.2994,
      "mean_token_accuracy": 0.19316228926181794,
      "num_tokens": 32392553.0,
      "step": 14520
    },
    {
      "entropy": 5.6221105575561525,
      "epoch": 1.6313247599258718,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047429874719390165,
      "loss": 5.3596,
      "mean_token_accuracy": 0.18911121189594268,
      "num_tokens": 32404736.0,
      "step": 14525
    },
    {
      "entropy": 5.47715744972229,
      "epoch": 1.6318863368338294,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047428011018822404,
      "loss": 5.2448,
      "mean_token_accuracy": 0.19873294830322266,
      "num_tokens": 32415089.0,
      "step": 14530
    },
    {
      "entropy": 5.408508396148681,
      "epoch": 1.6324479137417869,
      "grad_norm": 1.25,
      "learning_rate": 0.00047426146683739043,
      "loss": 5.2501,
      "mean_token_accuracy": 0.1852736294269562,
      "num_tokens": 32426814.0,
      "step": 14535
    },
    {
      "entropy": 5.410169458389282,
      "epoch": 1.6330094906497443,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047424281714199436,
      "loss": 5.3202,
      "mean_token_accuracy": 0.1818560019135475,
      "num_tokens": 32437173.0,
      "step": 14540
    },
    {
      "entropy": 5.494043064117432,
      "epoch": 1.633571067557702,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047422416110262975,
      "loss": 5.2857,
      "mean_token_accuracy": 0.19224082678556442,
      "num_tokens": 32448597.0,
      "step": 14545
    },
    {
      "entropy": 5.625814199447632,
      "epoch": 1.6341326444656596,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047420549871989053,
      "loss": 5.4576,
      "mean_token_accuracy": 0.1856563836336136,
      "num_tokens": 32460906.0,
      "step": 14550
    },
    {
      "entropy": 5.548081541061402,
      "epoch": 1.634694221373617,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000474186829994371,
      "loss": 5.2461,
      "mean_token_accuracy": 0.1920793980360031,
      "num_tokens": 32473756.0,
      "step": 14555
    },
    {
      "entropy": 5.447436809539795,
      "epoch": 1.6352557982815745,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047416815492666564,
      "loss": 5.3306,
      "mean_token_accuracy": 0.1863094761967659,
      "num_tokens": 32483912.0,
      "step": 14560
    },
    {
      "entropy": 5.446155214309693,
      "epoch": 1.6358173751895322,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004741494735173689,
      "loss": 5.2689,
      "mean_token_accuracy": 0.18848029375076295,
      "num_tokens": 32493899.0,
      "step": 14565
    },
    {
      "entropy": 5.480790519714356,
      "epoch": 1.6363789520974898,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004741307857670757,
      "loss": 5.2926,
      "mean_token_accuracy": 0.19494600892066954,
      "num_tokens": 32505215.0,
      "step": 14570
    },
    {
      "entropy": 5.5060957908630375,
      "epoch": 1.6369405290054473,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00047411209167638106,
      "loss": 5.3509,
      "mean_token_accuracy": 0.18806742280721664,
      "num_tokens": 32517288.0,
      "step": 14575
    },
    {
      "entropy": 5.532022333145141,
      "epoch": 1.6375021059134047,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004740933912458801,
      "loss": 5.3193,
      "mean_token_accuracy": 0.18988440036773682,
      "num_tokens": 32529705.0,
      "step": 14580
    },
    {
      "entropy": 5.588199949264526,
      "epoch": 1.6380636828213624,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047407468447616845,
      "loss": 5.327,
      "mean_token_accuracy": 0.19298471808433532,
      "num_tokens": 32540868.0,
      "step": 14585
    },
    {
      "entropy": 5.413400936126709,
      "epoch": 1.63862525972932,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047405597136784166,
      "loss": 5.2673,
      "mean_token_accuracy": 0.19524504691362382,
      "num_tokens": 32552129.0,
      "step": 14590
    },
    {
      "entropy": 5.432618141174316,
      "epoch": 1.6391868366372775,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004740372519214955,
      "loss": 5.2054,
      "mean_token_accuracy": 0.1949848398566246,
      "num_tokens": 32563059.0,
      "step": 14595
    },
    {
      "entropy": 5.468929481506348,
      "epoch": 1.639748413545235,
      "grad_norm": 1.375,
      "learning_rate": 0.00047401852613772604,
      "loss": 5.3719,
      "mean_token_accuracy": 0.17867764085531235,
      "num_tokens": 32576628.0,
      "step": 14600
    },
    {
      "entropy": 5.547284460067749,
      "epoch": 1.6403099904531926,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004739997940171295,
      "loss": 5.3274,
      "mean_token_accuracy": 0.1882872998714447,
      "num_tokens": 32589138.0,
      "step": 14605
    },
    {
      "entropy": 5.542290496826172,
      "epoch": 1.6408715673611503,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047398105556030234,
      "loss": 5.3074,
      "mean_token_accuracy": 0.1895511269569397,
      "num_tokens": 32601865.0,
      "step": 14610
    },
    {
      "entropy": 5.5082597732543945,
      "epoch": 1.6414331442691077,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047396231076784123,
      "loss": 5.2776,
      "mean_token_accuracy": 0.19095823764801026,
      "num_tokens": 32612439.0,
      "step": 14615
    },
    {
      "entropy": 5.488807916641235,
      "epoch": 1.6419947211770651,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004739435596403429,
      "loss": 5.2922,
      "mean_token_accuracy": 0.18706877827644347,
      "num_tokens": 32623002.0,
      "step": 14620
    },
    {
      "entropy": 5.516695737838745,
      "epoch": 1.6425562980850228,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004739248021784046,
      "loss": 5.2725,
      "mean_token_accuracy": 0.1955020934343338,
      "num_tokens": 32632529.0,
      "step": 14625
    },
    {
      "entropy": 5.465663766860962,
      "epoch": 1.6431178749929805,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004739060383826234,
      "loss": 5.2738,
      "mean_token_accuracy": 0.1882711485028267,
      "num_tokens": 32643584.0,
      "step": 14630
    },
    {
      "entropy": 5.499310111999511,
      "epoch": 1.6436794519009377,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047388726825359673,
      "loss": 5.3496,
      "mean_token_accuracy": 0.19002342820167542,
      "num_tokens": 32655899.0,
      "step": 14635
    },
    {
      "entropy": 5.487424564361572,
      "epoch": 1.6442410288088953,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047386849179192235,
      "loss": 5.2861,
      "mean_token_accuracy": 0.18974341154098512,
      "num_tokens": 32666747.0,
      "step": 14640
    },
    {
      "entropy": 5.48378734588623,
      "epoch": 1.644802605716853,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047384970899819813,
      "loss": 5.2459,
      "mean_token_accuracy": 0.1957755133509636,
      "num_tokens": 32678025.0,
      "step": 14645
    },
    {
      "entropy": 5.423769807815551,
      "epoch": 1.6453641826248104,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004738309198730219,
      "loss": 5.2039,
      "mean_token_accuracy": 0.19665048867464066,
      "num_tokens": 32688894.0,
      "step": 14650
    },
    {
      "entropy": 5.505620908737183,
      "epoch": 1.6459257595327679,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047381212441699217,
      "loss": 5.2925,
      "mean_token_accuracy": 0.19041657000780104,
      "num_tokens": 32699530.0,
      "step": 14655
    },
    {
      "entropy": 5.53219838142395,
      "epoch": 1.6464873364407255,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004737933226307072,
      "loss": 5.3699,
      "mean_token_accuracy": 0.18907047510147096,
      "num_tokens": 32710969.0,
      "step": 14660
    },
    {
      "entropy": 5.475825691223145,
      "epoch": 1.6470489133486832,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047377451451476586,
      "loss": 5.1977,
      "mean_token_accuracy": 0.19761187136173247,
      "num_tokens": 32723715.0,
      "step": 14665
    },
    {
      "entropy": 5.438703107833862,
      "epoch": 1.6476104902566406,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004737557000697668,
      "loss": 5.2933,
      "mean_token_accuracy": 0.1900079607963562,
      "num_tokens": 32735406.0,
      "step": 14670
    },
    {
      "entropy": 5.441463613510132,
      "epoch": 1.648172067164598,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004737368792963092,
      "loss": 5.2911,
      "mean_token_accuracy": 0.18695611208677293,
      "num_tokens": 32746123.0,
      "step": 14675
    },
    {
      "entropy": 5.4754539966583256,
      "epoch": 1.6487336440725557,
      "grad_norm": 1.25,
      "learning_rate": 0.00047371805219499223,
      "loss": 5.2547,
      "mean_token_accuracy": 0.1885870799422264,
      "num_tokens": 32757172.0,
      "step": 14680
    },
    {
      "entropy": 5.456832218170166,
      "epoch": 1.6492952209805134,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047369921876641547,
      "loss": 5.1529,
      "mean_token_accuracy": 0.1981355518102646,
      "num_tokens": 32767810.0,
      "step": 14685
    },
    {
      "entropy": 5.413317060470581,
      "epoch": 1.6498567978884708,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047368037901117854,
      "loss": 5.2594,
      "mean_token_accuracy": 0.19225677400827407,
      "num_tokens": 32779723.0,
      "step": 14690
    },
    {
      "entropy": 5.611428070068359,
      "epoch": 1.6504183747964283,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004736615329298812,
      "loss": 5.5039,
      "mean_token_accuracy": 0.17668303549289704,
      "num_tokens": 32793378.0,
      "step": 14695
    },
    {
      "entropy": 5.524914121627807,
      "epoch": 1.650979951704386,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047364268052312367,
      "loss": 5.2855,
      "mean_token_accuracy": 0.19185836762189865,
      "num_tokens": 32805643.0,
      "step": 14700
    },
    {
      "entropy": 5.473713397979736,
      "epoch": 1.6515415286123436,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004736238217915061,
      "loss": 5.2464,
      "mean_token_accuracy": 0.19162303805351258,
      "num_tokens": 32818275.0,
      "step": 14705
    },
    {
      "entropy": 5.327887296676636,
      "epoch": 1.652103105520301,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047360495673562906,
      "loss": 5.1398,
      "mean_token_accuracy": 0.19743950963020324,
      "num_tokens": 32829548.0,
      "step": 14710
    },
    {
      "entropy": 5.401415157318115,
      "epoch": 1.6526646824282585,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004735860853560931,
      "loss": 5.3238,
      "mean_token_accuracy": 0.19296661168336868,
      "num_tokens": 32841472.0,
      "step": 14715
    },
    {
      "entropy": 5.580569744110107,
      "epoch": 1.6532262593362161,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047356720765349926,
      "loss": 5.3453,
      "mean_token_accuracy": 0.19090296179056168,
      "num_tokens": 32852589.0,
      "step": 14720
    },
    {
      "entropy": 5.4584063529968265,
      "epoch": 1.6537878362441738,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004735483236284485,
      "loss": 5.2768,
      "mean_token_accuracy": 0.19118585884571077,
      "num_tokens": 32863100.0,
      "step": 14725
    },
    {
      "entropy": 5.492718935012817,
      "epoch": 1.654349413152131,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047352943328154203,
      "loss": 5.2774,
      "mean_token_accuracy": 0.1917508527636528,
      "num_tokens": 32874912.0,
      "step": 14730
    },
    {
      "entropy": 5.399691152572632,
      "epoch": 1.6549109900600887,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047351053661338145,
      "loss": 5.212,
      "mean_token_accuracy": 0.1933803901076317,
      "num_tokens": 32886249.0,
      "step": 14735
    },
    {
      "entropy": 5.393367147445678,
      "epoch": 1.6554725669680463,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004734916336245684,
      "loss": 5.2966,
      "mean_token_accuracy": 0.18313236385583878,
      "num_tokens": 32898010.0,
      "step": 14740
    },
    {
      "entropy": 5.405088424682617,
      "epoch": 1.6560341438760038,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004734727243157047,
      "loss": 5.1429,
      "mean_token_accuracy": 0.19624395817518234,
      "num_tokens": 32909103.0,
      "step": 14745
    },
    {
      "entropy": 5.375612068176269,
      "epoch": 1.6565957207839612,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047345380868739253,
      "loss": 5.2396,
      "mean_token_accuracy": 0.19741520881652833,
      "num_tokens": 32919612.0,
      "step": 14750
    },
    {
      "entropy": 5.364291143417359,
      "epoch": 1.6571572976919189,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004734348867402341,
      "loss": 5.1947,
      "mean_token_accuracy": 0.19381499141454697,
      "num_tokens": 32930072.0,
      "step": 14755
    },
    {
      "entropy": 5.555651092529297,
      "epoch": 1.6577188745998765,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004734159584748319,
      "loss": 5.39,
      "mean_token_accuracy": 0.1905780777335167,
      "num_tokens": 32941984.0,
      "step": 14760
    },
    {
      "entropy": 5.485587310791016,
      "epoch": 1.658280451507834,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047339702389178874,
      "loss": 5.2554,
      "mean_token_accuracy": 0.20035102516412734,
      "num_tokens": 32952767.0,
      "step": 14765
    },
    {
      "entropy": 5.421601486206055,
      "epoch": 1.6588420284157914,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047337808299170725,
      "loss": 5.2931,
      "mean_token_accuracy": 0.1885802611708641,
      "num_tokens": 32964553.0,
      "step": 14770
    },
    {
      "entropy": 5.442891693115234,
      "epoch": 1.659403605323749,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047335913577519075,
      "loss": 5.2196,
      "mean_token_accuracy": 0.1917058050632477,
      "num_tokens": 32975526.0,
      "step": 14775
    },
    {
      "entropy": 5.433455610275269,
      "epoch": 1.6599651822317067,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047334018224284244,
      "loss": 5.2977,
      "mean_token_accuracy": 0.19370029270648956,
      "num_tokens": 32987025.0,
      "step": 14780
    },
    {
      "entropy": 5.553694343566894,
      "epoch": 1.6605267591396642,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004733212223952657,
      "loss": 5.2878,
      "mean_token_accuracy": 0.19418825805187226,
      "num_tokens": 32997447.0,
      "step": 14785
    },
    {
      "entropy": 5.516057109832763,
      "epoch": 1.6610883360476216,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047330225623306443,
      "loss": 5.3047,
      "mean_token_accuracy": 0.1951941505074501,
      "num_tokens": 33007432.0,
      "step": 14790
    },
    {
      "entropy": 5.455147647857666,
      "epoch": 1.6616499129555793,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047328328375684237,
      "loss": 5.2997,
      "mean_token_accuracy": 0.19755344837903976,
      "num_tokens": 33018547.0,
      "step": 14795
    },
    {
      "entropy": 5.468757915496826,
      "epoch": 1.662211489863537,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004732643049672036,
      "loss": 5.2165,
      "mean_token_accuracy": 0.19927789866924286,
      "num_tokens": 33029092.0,
      "step": 14800
    },
    {
      "entropy": 5.409808731079101,
      "epoch": 1.6627730667714944,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004732453198647525,
      "loss": 5.2541,
      "mean_token_accuracy": 0.19813369810581208,
      "num_tokens": 33040498.0,
      "step": 14805
    },
    {
      "entropy": 5.545165872573852,
      "epoch": 1.6633346436794518,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004732263284500936,
      "loss": 5.3074,
      "mean_token_accuracy": 0.18597964346408843,
      "num_tokens": 33052203.0,
      "step": 14810
    },
    {
      "entropy": 5.4575512409210205,
      "epoch": 1.6638962205874095,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004732073307238314,
      "loss": 5.2723,
      "mean_token_accuracy": 0.19583217352628707,
      "num_tokens": 33063453.0,
      "step": 14815
    },
    {
      "entropy": 5.458451843261718,
      "epoch": 1.6644577974953672,
      "grad_norm": 1.21875,
      "learning_rate": 0.000473188326686571,
      "loss": 5.3476,
      "mean_token_accuracy": 0.19455185532569885,
      "num_tokens": 33075511.0,
      "step": 14820
    },
    {
      "entropy": 5.484337615966797,
      "epoch": 1.6650193744033244,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047316931633891736,
      "loss": 5.3059,
      "mean_token_accuracy": 0.19140402972698212,
      "num_tokens": 33086094.0,
      "step": 14825
    },
    {
      "entropy": 5.5458417415618895,
      "epoch": 1.665580951311282,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004731502996814759,
      "loss": 5.3764,
      "mean_token_accuracy": 0.19079845696687697,
      "num_tokens": 33096430.0,
      "step": 14830
    },
    {
      "entropy": 5.452405595779419,
      "epoch": 1.6661425282192397,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047313127671485194,
      "loss": 5.2518,
      "mean_token_accuracy": 0.19055817127227784,
      "num_tokens": 33108040.0,
      "step": 14835
    },
    {
      "entropy": 5.4227996349334715,
      "epoch": 1.6667041051271971,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004731122474396514,
      "loss": 5.2297,
      "mean_token_accuracy": 0.19288623631000518,
      "num_tokens": 33118282.0,
      "step": 14840
    },
    {
      "entropy": 5.451448726654053,
      "epoch": 1.6672656820351546,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047309321185648,
      "loss": 5.3106,
      "mean_token_accuracy": 0.19101365208625792,
      "num_tokens": 33130308.0,
      "step": 14845
    },
    {
      "entropy": 5.5709551811218265,
      "epoch": 1.6678272589431122,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047307416996594386,
      "loss": 5.3635,
      "mean_token_accuracy": 0.18882355988025665,
      "num_tokens": 33140260.0,
      "step": 14850
    },
    {
      "entropy": 5.459329414367676,
      "epoch": 1.66838883585107,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004730551217686495,
      "loss": 5.2573,
      "mean_token_accuracy": 0.19243746399879455,
      "num_tokens": 33152064.0,
      "step": 14855
    },
    {
      "entropy": 5.504673099517822,
      "epoch": 1.6689504127590273,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004730360672652031,
      "loss": 5.3085,
      "mean_token_accuracy": 0.19186588227748871,
      "num_tokens": 33164057.0,
      "step": 14860
    },
    {
      "entropy": 5.4617712020874025,
      "epoch": 1.6695119896669848,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047301700645621154,
      "loss": 5.2853,
      "mean_token_accuracy": 0.19473952054977417,
      "num_tokens": 33175281.0,
      "step": 14865
    },
    {
      "entropy": 5.413696670532227,
      "epoch": 1.6700735665749424,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004729979393422817,
      "loss": 5.2277,
      "mean_token_accuracy": 0.1927478387951851,
      "num_tokens": 33186615.0,
      "step": 14870
    },
    {
      "entropy": 5.48008360862732,
      "epoch": 1.6706351434829,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004729788659240206,
      "loss": 5.3393,
      "mean_token_accuracy": 0.19185318797826767,
      "num_tokens": 33198693.0,
      "step": 14875
    },
    {
      "entropy": 5.523104238510132,
      "epoch": 1.6711967203908575,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004729597862020358,
      "loss": 5.2993,
      "mean_token_accuracy": 0.19602399915456772,
      "num_tokens": 33209743.0,
      "step": 14880
    },
    {
      "entropy": 5.4746442317962645,
      "epoch": 1.671758297298815,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047294070017693453,
      "loss": 5.1986,
      "mean_token_accuracy": 0.2000429078936577,
      "num_tokens": 33220907.0,
      "step": 14885
    },
    {
      "entropy": 5.542150497436523,
      "epoch": 1.6723198742067726,
      "grad_norm": 1.421875,
      "learning_rate": 0.00047292160784932454,
      "loss": 5.3599,
      "mean_token_accuracy": 0.18650929778814315,
      "num_tokens": 33232278.0,
      "step": 14890
    },
    {
      "entropy": 5.526180028915405,
      "epoch": 1.6728814511147303,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047290250921981386,
      "loss": 5.2718,
      "mean_token_accuracy": 0.1904658704996109,
      "num_tokens": 33243920.0,
      "step": 14895
    },
    {
      "entropy": 5.459878778457641,
      "epoch": 1.6734430280226877,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047288340428901053,
      "loss": 5.3234,
      "mean_token_accuracy": 0.19086100459098815,
      "num_tokens": 33254467.0,
      "step": 14900
    },
    {
      "entropy": 5.398480319976807,
      "epoch": 1.6740046049306452,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047286429305752283,
      "loss": 5.2647,
      "mean_token_accuracy": 0.19002762734889983,
      "num_tokens": 33265536.0,
      "step": 14905
    },
    {
      "entropy": 5.492729711532593,
      "epoch": 1.6745661818386028,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047284517552595933,
      "loss": 5.2804,
      "mean_token_accuracy": 0.18326881825923919,
      "num_tokens": 33277233.0,
      "step": 14910
    },
    {
      "entropy": 5.429772663116455,
      "epoch": 1.6751277587465605,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004728260516949287,
      "loss": 5.1518,
      "mean_token_accuracy": 0.1960465908050537,
      "num_tokens": 33287873.0,
      "step": 14915
    },
    {
      "entropy": 5.450742244720459,
      "epoch": 1.6756893356545177,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004728069215650399,
      "loss": 5.2856,
      "mean_token_accuracy": 0.18781570941209794,
      "num_tokens": 33298888.0,
      "step": 14920
    },
    {
      "entropy": 5.332154989242554,
      "epoch": 1.6762509125624754,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004727877851369019,
      "loss": 5.1564,
      "mean_token_accuracy": 0.19918228387832643,
      "num_tokens": 33309607.0,
      "step": 14925
    },
    {
      "entropy": 5.42673282623291,
      "epoch": 1.676812489470433,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047276864241112417,
      "loss": 5.1848,
      "mean_token_accuracy": 0.19833206832408906,
      "num_tokens": 33320582.0,
      "step": 14930
    },
    {
      "entropy": 5.422622632980347,
      "epoch": 1.6773740663783905,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004727494933883161,
      "loss": 5.2361,
      "mean_token_accuracy": 0.19378088116645814,
      "num_tokens": 33332145.0,
      "step": 14935
    },
    {
      "entropy": 5.552988862991333,
      "epoch": 1.677935643286348,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004727303380690876,
      "loss": 5.4893,
      "mean_token_accuracy": 0.18411082178354263,
      "num_tokens": 33343963.0,
      "step": 14940
    },
    {
      "entropy": 5.468435621261596,
      "epoch": 1.6784972201943056,
      "grad_norm": 1.25,
      "learning_rate": 0.0004727111764540484,
      "loss": 5.2936,
      "mean_token_accuracy": 0.19818158447742462,
      "num_tokens": 33354931.0,
      "step": 14945
    },
    {
      "entropy": 5.432682132720947,
      "epoch": 1.6790587971022632,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047269200854380854,
      "loss": 5.2964,
      "mean_token_accuracy": 0.19380978047847747,
      "num_tokens": 33365686.0,
      "step": 14950
    },
    {
      "entropy": 5.5454638481140135,
      "epoch": 1.6796203740102207,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047267283433897854,
      "loss": 5.2912,
      "mean_token_accuracy": 0.19802167415618896,
      "num_tokens": 33377099.0,
      "step": 14955
    },
    {
      "entropy": 5.524987125396729,
      "epoch": 1.6801819509181781,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047265365384016887,
      "loss": 5.3473,
      "mean_token_accuracy": 0.1849740281701088,
      "num_tokens": 33389477.0,
      "step": 14960
    },
    {
      "entropy": 5.466532897949219,
      "epoch": 1.6807435278261358,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047263446704799017,
      "loss": 5.3639,
      "mean_token_accuracy": 0.19132375419139863,
      "num_tokens": 33401970.0,
      "step": 14965
    },
    {
      "entropy": 5.451760196685791,
      "epoch": 1.6813051047340934,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047261527396305343,
      "loss": 5.2272,
      "mean_token_accuracy": 0.19191225320100785,
      "num_tokens": 33412986.0,
      "step": 14970
    },
    {
      "entropy": 5.506747674942017,
      "epoch": 1.6818666816420509,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047259607458596966,
      "loss": 5.3553,
      "mean_token_accuracy": 0.19107459485530853,
      "num_tokens": 33424717.0,
      "step": 14975
    },
    {
      "entropy": 5.391724109649658,
      "epoch": 1.6824282585500083,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047257686891735023,
      "loss": 5.1896,
      "mean_token_accuracy": 0.19591644406318665,
      "num_tokens": 33436713.0,
      "step": 14980
    },
    {
      "entropy": 5.576106309890747,
      "epoch": 1.682989835457966,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047255765695780665,
      "loss": 5.37,
      "mean_token_accuracy": 0.1877459853887558,
      "num_tokens": 33447532.0,
      "step": 14985
    },
    {
      "entropy": 5.40876955986023,
      "epoch": 1.6835514123659236,
      "grad_norm": 1.25,
      "learning_rate": 0.0004725384387079507,
      "loss": 5.1754,
      "mean_token_accuracy": 0.1924012839794159,
      "num_tokens": 33458530.0,
      "step": 14990
    },
    {
      "entropy": 5.469767045974732,
      "epoch": 1.684112989273881,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047251921416839425,
      "loss": 5.2954,
      "mean_token_accuracy": 0.18626859933137893,
      "num_tokens": 33469617.0,
      "step": 14995
    },
    {
      "entropy": 5.485902929306031,
      "epoch": 1.6846745661818385,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004724999833397494,
      "loss": 5.2113,
      "mean_token_accuracy": 0.20267764776945113,
      "num_tokens": 33479622.0,
      "step": 15000
    },
    {
      "epoch": 1.6846745661818385,
      "eval_entropy": 5.328831116747912,
      "eval_loss": 5.356277942657471,
      "eval_mean_token_accuracy": 0.19730421132822457,
      "eval_num_tokens": 33479622.0,
      "eval_runtime": 23.8806,
      "eval_samples_per_second": 1298.667,
      "eval_steps_per_second": 162.349,
      "step": 15000
    },
    {
      "entropy": 5.3665824890136715,
      "epoch": 1.6852361430897962,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004724807462226285,
      "loss": 5.1873,
      "mean_token_accuracy": 0.20435822606086732,
      "num_tokens": 33490907.0,
      "step": 15005
    },
    {
      "entropy": 5.4181475162506105,
      "epoch": 1.6857977199977539,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000472461502817644,
      "loss": 5.2858,
      "mean_token_accuracy": 0.19380941838026047,
      "num_tokens": 33501884.0,
      "step": 15010
    },
    {
      "entropy": 5.417263078689575,
      "epoch": 1.6863592969057113,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004724422531254087,
      "loss": 5.175,
      "mean_token_accuracy": 0.19577735811471939,
      "num_tokens": 33512106.0,
      "step": 15015
    },
    {
      "entropy": 5.449825286865234,
      "epoch": 1.6869208738136687,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004724229971465355,
      "loss": 5.2883,
      "mean_token_accuracy": 0.19318007081747054,
      "num_tokens": 33522889.0,
      "step": 15020
    },
    {
      "entropy": 5.489127922058105,
      "epoch": 1.6874824507216264,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004724037348816375,
      "loss": 5.256,
      "mean_token_accuracy": 0.1946806401014328,
      "num_tokens": 33534601.0,
      "step": 15025
    },
    {
      "entropy": 5.483654642105103,
      "epoch": 1.6880440276295838,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000472384466331328,
      "loss": 5.3234,
      "mean_token_accuracy": 0.1949715197086334,
      "num_tokens": 33545300.0,
      "step": 15030
    },
    {
      "entropy": 5.569486427307129,
      "epoch": 1.6886056045375413,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004723651914962205,
      "loss": 5.2979,
      "mean_token_accuracy": 0.19008689224720002,
      "num_tokens": 33555581.0,
      "step": 15035
    },
    {
      "entropy": 5.515807247161865,
      "epoch": 1.689167181445499,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047234591037692884,
      "loss": 5.2798,
      "mean_token_accuracy": 0.19053123891353607,
      "num_tokens": 33566902.0,
      "step": 15040
    },
    {
      "entropy": 5.48316125869751,
      "epoch": 1.6897287583534566,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047232662297406683,
      "loss": 5.2923,
      "mean_token_accuracy": 0.19330027252435683,
      "num_tokens": 33578316.0,
      "step": 15045
    },
    {
      "entropy": 5.449962949752807,
      "epoch": 1.690290335261414,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004723073292882485,
      "loss": 5.3104,
      "mean_token_accuracy": 0.18745491802692413,
      "num_tokens": 33590242.0,
      "step": 15050
    },
    {
      "entropy": 5.447853899002075,
      "epoch": 1.6908519121693715,
      "grad_norm": 1.125,
      "learning_rate": 0.00047228802932008836,
      "loss": 5.2087,
      "mean_token_accuracy": 0.1942342311143875,
      "num_tokens": 33601277.0,
      "step": 15055
    },
    {
      "entropy": 5.430882453918457,
      "epoch": 1.6914134890773291,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004722687230702009,
      "loss": 5.2609,
      "mean_token_accuracy": 0.1930315226316452,
      "num_tokens": 33612423.0,
      "step": 15060
    },
    {
      "entropy": 5.535700178146362,
      "epoch": 1.6919750659852868,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004722494105392007,
      "loss": 5.3284,
      "mean_token_accuracy": 0.18896596282720565,
      "num_tokens": 33624250.0,
      "step": 15065
    },
    {
      "entropy": 5.512123346328735,
      "epoch": 1.6925366428932442,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004722300917277029,
      "loss": 5.2778,
      "mean_token_accuracy": 0.19357503056526185,
      "num_tokens": 33636379.0,
      "step": 15070
    },
    {
      "entropy": 5.437799024581909,
      "epoch": 1.6930982198012017,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004722107666363224,
      "loss": 5.2773,
      "mean_token_accuracy": 0.18782281279563903,
      "num_tokens": 33648296.0,
      "step": 15075
    },
    {
      "entropy": 5.3667152404785154,
      "epoch": 1.6936597967091593,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004721914352656746,
      "loss": 5.2821,
      "mean_token_accuracy": 0.18598387837409974,
      "num_tokens": 33659438.0,
      "step": 15080
    },
    {
      "entropy": 5.548566579818726,
      "epoch": 1.694221373617117,
      "grad_norm": 1.21875,
      "learning_rate": 0.000472172097616375,
      "loss": 5.3764,
      "mean_token_accuracy": 0.18941598534584045,
      "num_tokens": 33670893.0,
      "step": 15085
    },
    {
      "entropy": 5.537064075469971,
      "epoch": 1.6947829505250744,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004721527536890394,
      "loss": 5.2499,
      "mean_token_accuracy": 0.19386748522520064,
      "num_tokens": 33681487.0,
      "step": 15090
    },
    {
      "entropy": 5.416791534423828,
      "epoch": 1.6953445274330319,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004721334034842836,
      "loss": 5.2699,
      "mean_token_accuracy": 0.2034762144088745,
      "num_tokens": 33692925.0,
      "step": 15095
    },
    {
      "entropy": 5.446263647079467,
      "epoch": 1.6959061043409895,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004721140470027238,
      "loss": 5.2586,
      "mean_token_accuracy": 0.19316800385713578,
      "num_tokens": 33703559.0,
      "step": 15100
    },
    {
      "entropy": 5.429151916503907,
      "epoch": 1.6964676812489472,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047209468424497627,
      "loss": 5.2487,
      "mean_token_accuracy": 0.1935478687286377,
      "num_tokens": 33715041.0,
      "step": 15105
    },
    {
      "entropy": 5.43087911605835,
      "epoch": 1.6970292581569046,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004720753152116576,
      "loss": 5.1893,
      "mean_token_accuracy": 0.19459699839353561,
      "num_tokens": 33726334.0,
      "step": 15110
    },
    {
      "entropy": 5.485027265548706,
      "epoch": 1.697590835064862,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047205593990338443,
      "loss": 5.2289,
      "mean_token_accuracy": 0.19392191767692565,
      "num_tokens": 33738663.0,
      "step": 15115
    },
    {
      "entropy": 5.398070430755615,
      "epoch": 1.6981524119728197,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004720365583207737,
      "loss": 5.1858,
      "mean_token_accuracy": 0.20194293558597565,
      "num_tokens": 33749202.0,
      "step": 15120
    },
    {
      "entropy": 5.4027293682098385,
      "epoch": 1.6987139888807772,
      "grad_norm": 1.125,
      "learning_rate": 0.0004720171704644425,
      "loss": 5.166,
      "mean_token_accuracy": 0.20751358121633529,
      "num_tokens": 33759809.0,
      "step": 15125
    },
    {
      "entropy": 5.378641557693482,
      "epoch": 1.6992755657887346,
      "grad_norm": 1.375,
      "learning_rate": 0.0004719977763350083,
      "loss": 5.2495,
      "mean_token_accuracy": 0.19495665282011032,
      "num_tokens": 33771374.0,
      "step": 15130
    },
    {
      "entropy": 5.428633356094361,
      "epoch": 1.6998371426966923,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047197837593308833,
      "loss": 5.2638,
      "mean_token_accuracy": 0.19419437050819396,
      "num_tokens": 33781645.0,
      "step": 15135
    },
    {
      "entropy": 5.456284141540527,
      "epoch": 1.70039871960465,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004719589692593006,
      "loss": 5.2316,
      "mean_token_accuracy": 0.19654669612646103,
      "num_tokens": 33793032.0,
      "step": 15140
    },
    {
      "entropy": 5.570026397705078,
      "epoch": 1.7009602965126074,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004719395563142629,
      "loss": 5.4449,
      "mean_token_accuracy": 0.18291118741035461,
      "num_tokens": 33805116.0,
      "step": 15145
    },
    {
      "entropy": 5.53737440109253,
      "epoch": 1.7015218734205648,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004719201370985933,
      "loss": 5.288,
      "mean_token_accuracy": 0.19232716858386995,
      "num_tokens": 33817135.0,
      "step": 15150
    },
    {
      "entropy": 5.443862676620483,
      "epoch": 1.7020834503285225,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00047190071161291007,
      "loss": 5.2527,
      "mean_token_accuracy": 0.1925329104065895,
      "num_tokens": 33830187.0,
      "step": 15155
    },
    {
      "entropy": 5.415344381332398,
      "epoch": 1.7026450272364801,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047188127985783195,
      "loss": 5.2824,
      "mean_token_accuracy": 0.18780236840248107,
      "num_tokens": 33841297.0,
      "step": 15160
    },
    {
      "entropy": 5.508278322219849,
      "epoch": 1.7032066041444376,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047186184183397754,
      "loss": 5.2669,
      "mean_token_accuracy": 0.19643225967884065,
      "num_tokens": 33852888.0,
      "step": 15165
    },
    {
      "entropy": 5.460748243331909,
      "epoch": 1.703768181052395,
      "grad_norm": 1.515625,
      "learning_rate": 0.00047184239754196564,
      "loss": 5.3231,
      "mean_token_accuracy": 0.18794745206832886,
      "num_tokens": 33865165.0,
      "step": 15170
    },
    {
      "entropy": 5.412379932403565,
      "epoch": 1.7043297579603527,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004718229469824155,
      "loss": 5.1375,
      "mean_token_accuracy": 0.19809648245573044,
      "num_tokens": 33875865.0,
      "step": 15175
    },
    {
      "entropy": 5.464021015167236,
      "epoch": 1.7048913348683103,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047180349015594634,
      "loss": 5.3754,
      "mean_token_accuracy": 0.1836199626326561,
      "num_tokens": 33887519.0,
      "step": 15180
    },
    {
      "entropy": 5.393822479248047,
      "epoch": 1.7054529117762678,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004717840270631778,
      "loss": 5.1879,
      "mean_token_accuracy": 0.2034103736281395,
      "num_tokens": 33898762.0,
      "step": 15185
    },
    {
      "entropy": 5.3856912612915036,
      "epoch": 1.7060144886842252,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047176455770472946,
      "loss": 5.1858,
      "mean_token_accuracy": 0.1980983316898346,
      "num_tokens": 33911109.0,
      "step": 15190
    },
    {
      "entropy": 5.485199451446533,
      "epoch": 1.706576065592183,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047174508208122135,
      "loss": 5.4453,
      "mean_token_accuracy": 0.18300354480743408,
      "num_tokens": 33922741.0,
      "step": 15195
    },
    {
      "entropy": 5.493468236923218,
      "epoch": 1.7071376425001406,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004717256001932735,
      "loss": 5.2667,
      "mean_token_accuracy": 0.1886766657233238,
      "num_tokens": 33933234.0,
      "step": 15200
    },
    {
      "entropy": 5.453039884567261,
      "epoch": 1.707699219408098,
      "grad_norm": 2.359375,
      "learning_rate": 0.00047170611204150624,
      "loss": 5.1933,
      "mean_token_accuracy": 0.19519753754138947,
      "num_tokens": 33943667.0,
      "step": 15205
    },
    {
      "entropy": 5.434653902053833,
      "epoch": 1.7082607963160554,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004716866176265401,
      "loss": 5.1604,
      "mean_token_accuracy": 0.20574185848236085,
      "num_tokens": 33954499.0,
      "step": 15210
    },
    {
      "entropy": 5.364242506027222,
      "epoch": 1.708822373224013,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047166711694899573,
      "loss": 5.2434,
      "mean_token_accuracy": 0.1991499975323677,
      "num_tokens": 33964700.0,
      "step": 15215
    },
    {
      "entropy": 5.383185529708863,
      "epoch": 1.7093839501319705,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047164761000949404,
      "loss": 5.2897,
      "mean_token_accuracy": 0.1941818505525589,
      "num_tokens": 33977759.0,
      "step": 15220
    },
    {
      "entropy": 5.550089406967163,
      "epoch": 1.709945527039928,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004716280968086563,
      "loss": 5.3559,
      "mean_token_accuracy": 0.18744521588087082,
      "num_tokens": 33990019.0,
      "step": 15225
    },
    {
      "entropy": 5.50252332687378,
      "epoch": 1.7105071039478856,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004716085773471036,
      "loss": 5.2537,
      "mean_token_accuracy": 0.19737190306186675,
      "num_tokens": 34000663.0,
      "step": 15230
    },
    {
      "entropy": 5.341259002685547,
      "epoch": 1.7110686808558433,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047158905162545756,
      "loss": 5.1501,
      "mean_token_accuracy": 0.20010828375816345,
      "num_tokens": 34012768.0,
      "step": 15235
    },
    {
      "entropy": 5.359045839309692,
      "epoch": 1.7116302577638007,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047156951964433984,
      "loss": 5.1774,
      "mean_token_accuracy": 0.19690180867910384,
      "num_tokens": 34025801.0,
      "step": 15240
    },
    {
      "entropy": 5.375544309616089,
      "epoch": 1.7121918346717582,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047154998140437246,
      "loss": 5.1043,
      "mean_token_accuracy": 0.21326058357954025,
      "num_tokens": 34036553.0,
      "step": 15245
    },
    {
      "entropy": 5.450237894058228,
      "epoch": 1.7127534115797158,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047153043690617737,
      "loss": 5.2853,
      "mean_token_accuracy": 0.1939275234937668,
      "num_tokens": 34048352.0,
      "step": 15250
    },
    {
      "entropy": 5.418494939804077,
      "epoch": 1.7133149884876735,
      "grad_norm": 1.09375,
      "learning_rate": 0.000471510886150377,
      "loss": 5.2344,
      "mean_token_accuracy": 0.20148492157459258,
      "num_tokens": 34059617.0,
      "step": 15255
    },
    {
      "entropy": 5.438854312896728,
      "epoch": 1.713876565395631,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004714913291375938,
      "loss": 5.2565,
      "mean_token_accuracy": 0.18967977464199065,
      "num_tokens": 34071176.0,
      "step": 15260
    },
    {
      "entropy": 5.39390983581543,
      "epoch": 1.7144381423035884,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047147176586845036,
      "loss": 5.2318,
      "mean_token_accuracy": 0.19420790076255798,
      "num_tokens": 34081883.0,
      "step": 15265
    },
    {
      "entropy": 5.429718399047852,
      "epoch": 1.714999719211546,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004714521963435698,
      "loss": 5.2703,
      "mean_token_accuracy": 0.19103458523750305,
      "num_tokens": 34092938.0,
      "step": 15270
    },
    {
      "entropy": 5.470838356018066,
      "epoch": 1.7155612961195037,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047143262056357506,
      "loss": 5.2212,
      "mean_token_accuracy": 0.19646950513124467,
      "num_tokens": 34104758.0,
      "step": 15275
    },
    {
      "entropy": 5.417604684829712,
      "epoch": 1.7161228730274611,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047141303852908965,
      "loss": 5.1406,
      "mean_token_accuracy": 0.2002304181456566,
      "num_tokens": 34114752.0,
      "step": 15280
    },
    {
      "entropy": 5.44685926437378,
      "epoch": 1.7166844499354186,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004713934502407367,
      "loss": 5.2906,
      "mean_token_accuracy": 0.18836827725172042,
      "num_tokens": 34125547.0,
      "step": 15285
    },
    {
      "entropy": 5.559075260162354,
      "epoch": 1.7172460268433762,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004713738556991402,
      "loss": 5.3418,
      "mean_token_accuracy": 0.1941390559077263,
      "num_tokens": 34136085.0,
      "step": 15290
    },
    {
      "entropy": 5.4588734149932865,
      "epoch": 1.717807603751334,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000471354254904924,
      "loss": 5.2711,
      "mean_token_accuracy": 0.1996251806616783,
      "num_tokens": 34147536.0,
      "step": 15295
    },
    {
      "entropy": 5.46059741973877,
      "epoch": 1.7183691806592913,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047133464785871216,
      "loss": 5.3902,
      "mean_token_accuracy": 0.19982053339481354,
      "num_tokens": 34158238.0,
      "step": 15300
    },
    {
      "entropy": 5.364193630218506,
      "epoch": 1.7189307575672488,
      "grad_norm": 1.234375,
      "learning_rate": 0.000471315034561129,
      "loss": 5.0992,
      "mean_token_accuracy": 0.20570196062326432,
      "num_tokens": 34169793.0,
      "step": 15305
    },
    {
      "entropy": 5.507556390762329,
      "epoch": 1.7194923344752064,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047129541501279897,
      "loss": 5.2804,
      "mean_token_accuracy": 0.2013788029551506,
      "num_tokens": 34179753.0,
      "step": 15310
    },
    {
      "entropy": 5.499370956420899,
      "epoch": 1.7200539113831639,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004712757892143468,
      "loss": 5.3774,
      "mean_token_accuracy": 0.19002106934785842,
      "num_tokens": 34189947.0,
      "step": 15315
    },
    {
      "entropy": 5.484274578094483,
      "epoch": 1.7206154882911213,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047125615716639744,
      "loss": 5.3245,
      "mean_token_accuracy": 0.19940817654132842,
      "num_tokens": 34201197.0,
      "step": 15320
    },
    {
      "entropy": 5.476716756820679,
      "epoch": 1.721177065199079,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004712365188695758,
      "loss": 5.3149,
      "mean_token_accuracy": 0.2012427493929863,
      "num_tokens": 34212762.0,
      "step": 15325
    },
    {
      "entropy": 5.4344940185546875,
      "epoch": 1.7217386421070366,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004712168743245074,
      "loss": 5.2803,
      "mean_token_accuracy": 0.19580717384815216,
      "num_tokens": 34223928.0,
      "step": 15330
    },
    {
      "entropy": 5.358058452606201,
      "epoch": 1.722300219014994,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047119722353181754,
      "loss": 5.1666,
      "mean_token_accuracy": 0.2024077668786049,
      "num_tokens": 34233888.0,
      "step": 15335
    },
    {
      "entropy": 5.520869207382202,
      "epoch": 1.7228617959229515,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047117756649213204,
      "loss": 5.2789,
      "mean_token_accuracy": 0.19239142835140227,
      "num_tokens": 34245894.0,
      "step": 15340
    },
    {
      "entropy": 5.461813592910767,
      "epoch": 1.7234233728309092,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004711579032060767,
      "loss": 5.1878,
      "mean_token_accuracy": 0.20102292001247407,
      "num_tokens": 34256288.0,
      "step": 15345
    },
    {
      "entropy": 5.4622434139251705,
      "epoch": 1.7239849497388668,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004711382336742777,
      "loss": 5.2407,
      "mean_token_accuracy": 0.1859392285346985,
      "num_tokens": 34266515.0,
      "step": 15350
    },
    {
      "entropy": 5.466305541992187,
      "epoch": 1.7245465266468243,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004711185578973612,
      "loss": 5.3102,
      "mean_token_accuracy": 0.18945770412683488,
      "num_tokens": 34278681.0,
      "step": 15355
    },
    {
      "entropy": 5.46079797744751,
      "epoch": 1.7251081035547817,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004710988758759538,
      "loss": 5.2402,
      "mean_token_accuracy": 0.1935290589928627,
      "num_tokens": 34289161.0,
      "step": 15360
    },
    {
      "entropy": 5.387932395935058,
      "epoch": 1.7256696804627394,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004710791876106821,
      "loss": 5.1996,
      "mean_token_accuracy": 0.19181107133626937,
      "num_tokens": 34300323.0,
      "step": 15365
    },
    {
      "entropy": 5.4188438892364506,
      "epoch": 1.726231257370697,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047105949310217304,
      "loss": 5.1828,
      "mean_token_accuracy": 0.19072486609220504,
      "num_tokens": 34311558.0,
      "step": 15370
    },
    {
      "entropy": 5.407376909255982,
      "epoch": 1.7267928342786545,
      "grad_norm": 1.25,
      "learning_rate": 0.0004710397923510537,
      "loss": 5.2946,
      "mean_token_accuracy": 0.18948574364185333,
      "num_tokens": 34323573.0,
      "step": 15375
    },
    {
      "entropy": 5.447926473617554,
      "epoch": 1.727354411186612,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047102008535795134,
      "loss": 5.2351,
      "mean_token_accuracy": 0.19117340296506882,
      "num_tokens": 34334698.0,
      "step": 15380
    },
    {
      "entropy": 5.463991260528564,
      "epoch": 1.7279159880945696,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004710003721234934,
      "loss": 5.2203,
      "mean_token_accuracy": 0.20594229996204377,
      "num_tokens": 34345948.0,
      "step": 15385
    },
    {
      "entropy": 5.43141655921936,
      "epoch": 1.7284775650025272,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047098065264830754,
      "loss": 5.2796,
      "mean_token_accuracy": 0.19364367574453353,
      "num_tokens": 34359067.0,
      "step": 15390
    },
    {
      "entropy": 5.326291799545288,
      "epoch": 1.7290391419104847,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047096092693302174,
      "loss": 5.146,
      "mean_token_accuracy": 0.19796561002731322,
      "num_tokens": 34370895.0,
      "step": 15395
    },
    {
      "entropy": 5.341482639312744,
      "epoch": 1.7296007188184421,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047094119497826406,
      "loss": 5.2542,
      "mean_token_accuracy": 0.19545391201972961,
      "num_tokens": 34382039.0,
      "step": 15400
    },
    {
      "entropy": 5.5028304100036625,
      "epoch": 1.7301622957263998,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047092145678466273,
      "loss": 5.3174,
      "mean_token_accuracy": 0.1881623923778534,
      "num_tokens": 34392939.0,
      "step": 15405
    },
    {
      "entropy": 5.572155380249024,
      "epoch": 1.7307238726343572,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004709017123528462,
      "loss": 5.336,
      "mean_token_accuracy": 0.19490597397089005,
      "num_tokens": 34404402.0,
      "step": 15410
    },
    {
      "entropy": 5.386563301086426,
      "epoch": 1.7312854495423147,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00047088196168344315,
      "loss": 5.231,
      "mean_token_accuracy": 0.19838883280754088,
      "num_tokens": 34415563.0,
      "step": 15415
    },
    {
      "entropy": 5.496076059341431,
      "epoch": 1.7318470264502723,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047086220477708256,
      "loss": 5.3426,
      "mean_token_accuracy": 0.1853834345936775,
      "num_tokens": 34426993.0,
      "step": 15420
    },
    {
      "entropy": 5.530219888687133,
      "epoch": 1.73240860335823,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047084244163439335,
      "loss": 5.3059,
      "mean_token_accuracy": 0.19982729256153106,
      "num_tokens": 34436697.0,
      "step": 15425
    },
    {
      "entropy": 5.510669851303101,
      "epoch": 1.7329701802661874,
      "grad_norm": 1.25,
      "learning_rate": 0.0004708226722560049,
      "loss": 5.2606,
      "mean_token_accuracy": 0.19673889130353928,
      "num_tokens": 34448076.0,
      "step": 15430
    },
    {
      "entropy": 5.4359307289123535,
      "epoch": 1.7335317571741449,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004708028966425465,
      "loss": 5.2906,
      "mean_token_accuracy": 0.19205243438482283,
      "num_tokens": 34460369.0,
      "step": 15435
    },
    {
      "entropy": 5.513742208480835,
      "epoch": 1.7340933340821025,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047078311479464797,
      "loss": 5.3442,
      "mean_token_accuracy": 0.18287210166454315,
      "num_tokens": 34472693.0,
      "step": 15440
    },
    {
      "entropy": 5.566083908081055,
      "epoch": 1.7346549109900602,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047076332671293917,
      "loss": 5.2854,
      "mean_token_accuracy": 0.19626031070947647,
      "num_tokens": 34485258.0,
      "step": 15445
    },
    {
      "entropy": 5.504504871368408,
      "epoch": 1.7352164878980176,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047074353239805005,
      "loss": 5.2698,
      "mean_token_accuracy": 0.19245284646749497,
      "num_tokens": 34496450.0,
      "step": 15450
    },
    {
      "entropy": 5.432540845870972,
      "epoch": 1.735778064805975,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004707237318506111,
      "loss": 5.3098,
      "mean_token_accuracy": 0.19632777869701384,
      "num_tokens": 34508062.0,
      "step": 15455
    },
    {
      "entropy": 5.47511100769043,
      "epoch": 1.7363396417139327,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047070392507125245,
      "loss": 5.262,
      "mean_token_accuracy": 0.2027345582842827,
      "num_tokens": 34519817.0,
      "step": 15460
    },
    {
      "entropy": 5.451759958267212,
      "epoch": 1.7369012186218904,
      "grad_norm": 1.828125,
      "learning_rate": 0.000470684112060605,
      "loss": 5.2508,
      "mean_token_accuracy": 0.1945961058139801,
      "num_tokens": 34532628.0,
      "step": 15465
    },
    {
      "entropy": 5.53095498085022,
      "epoch": 1.7374627955298478,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047066429281929944,
      "loss": 5.315,
      "mean_token_accuracy": 0.19680583626031875,
      "num_tokens": 34543558.0,
      "step": 15470
    },
    {
      "entropy": 5.570545768737793,
      "epoch": 1.7380243724378053,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000470644467347967,
      "loss": 5.3602,
      "mean_token_accuracy": 0.18485882580280305,
      "num_tokens": 34553720.0,
      "step": 15475
    },
    {
      "entropy": 5.465522718429566,
      "epoch": 1.738585949345763,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047062463564723884,
      "loss": 5.2408,
      "mean_token_accuracy": 0.19406428039073945,
      "num_tokens": 34564641.0,
      "step": 15480
    },
    {
      "entropy": 5.44709997177124,
      "epoch": 1.7391475262537206,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004706047977177464,
      "loss": 5.3277,
      "mean_token_accuracy": 0.1913145139813423,
      "num_tokens": 34576470.0,
      "step": 15485
    },
    {
      "entropy": 5.494604969024659,
      "epoch": 1.739709103161678,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047058495356012135,
      "loss": 5.3548,
      "mean_token_accuracy": 0.1924049213528633,
      "num_tokens": 34589158.0,
      "step": 15490
    },
    {
      "entropy": 5.406576871871948,
      "epoch": 1.7402706800696355,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047056510317499545,
      "loss": 5.1549,
      "mean_token_accuracy": 0.20320963561534883,
      "num_tokens": 34599192.0,
      "step": 15495
    },
    {
      "entropy": 5.430755376815796,
      "epoch": 1.7408322569775931,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047054524656300087,
      "loss": 5.2477,
      "mean_token_accuracy": 0.19713538736104966,
      "num_tokens": 34611922.0,
      "step": 15500
    },
    {
      "entropy": 5.5108397006988525,
      "epoch": 1.7413938338855506,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047052538372476985,
      "loss": 5.3069,
      "mean_token_accuracy": 0.19628826081752776,
      "num_tokens": 34623618.0,
      "step": 15505
    },
    {
      "entropy": 5.5127240180969235,
      "epoch": 1.741955410793508,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047050551466093466,
      "loss": 5.3462,
      "mean_token_accuracy": 0.18780522495508195,
      "num_tokens": 34635057.0,
      "step": 15510
    },
    {
      "entropy": 5.50947790145874,
      "epoch": 1.7425169877014657,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004704856393721281,
      "loss": 5.2759,
      "mean_token_accuracy": 0.19313253164291383,
      "num_tokens": 34647527.0,
      "step": 15515
    },
    {
      "entropy": 5.4397367477417,
      "epoch": 1.7430785646094233,
      "grad_norm": 1.25,
      "learning_rate": 0.00047046575785898297,
      "loss": 5.2412,
      "mean_token_accuracy": 0.19165327847003938,
      "num_tokens": 34658615.0,
      "step": 15520
    },
    {
      "entropy": 5.469088983535767,
      "epoch": 1.7436401415173808,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004704458701221323,
      "loss": 5.2786,
      "mean_token_accuracy": 0.1915829986333847,
      "num_tokens": 34668485.0,
      "step": 15525
    },
    {
      "entropy": 5.439208364486694,
      "epoch": 1.7442017184253382,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004704259761622094,
      "loss": 5.2399,
      "mean_token_accuracy": 0.1955853670835495,
      "num_tokens": 34681181.0,
      "step": 15530
    },
    {
      "entropy": 5.398151540756226,
      "epoch": 1.7447632953332959,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047040607597984745,
      "loss": 5.2284,
      "mean_token_accuracy": 0.19208595156669617,
      "num_tokens": 34692954.0,
      "step": 15535
    },
    {
      "entropy": 5.376310396194458,
      "epoch": 1.7453248722412535,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004703861695756804,
      "loss": 5.155,
      "mean_token_accuracy": 0.1993495136499405,
      "num_tokens": 34705120.0,
      "step": 15540
    },
    {
      "entropy": 5.495149183273315,
      "epoch": 1.745886449149211,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004703662569503419,
      "loss": 5.2674,
      "mean_token_accuracy": 0.19221558421850204,
      "num_tokens": 34716589.0,
      "step": 15545
    },
    {
      "entropy": 5.5011889934539795,
      "epoch": 1.7464480260571684,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047034633810446604,
      "loss": 5.253,
      "mean_token_accuracy": 0.1960187092423439,
      "num_tokens": 34727551.0,
      "step": 15550
    },
    {
      "entropy": 5.4215833187103275,
      "epoch": 1.747009602965126,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004703264130386869,
      "loss": 5.2782,
      "mean_token_accuracy": 0.19679175615310668,
      "num_tokens": 34738887.0,
      "step": 15555
    },
    {
      "entropy": 5.401621961593628,
      "epoch": 1.7475711798730837,
      "grad_norm": 1.40625,
      "learning_rate": 0.00047030648175363913,
      "loss": 5.2206,
      "mean_token_accuracy": 0.19658039957284928,
      "num_tokens": 34750264.0,
      "step": 15560
    },
    {
      "entropy": 5.5161881923675535,
      "epoch": 1.7481327567810412,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004702865442499572,
      "loss": 5.2763,
      "mean_token_accuracy": 0.19220732748508454,
      "num_tokens": 34761335.0,
      "step": 15565
    },
    {
      "entropy": 5.399585914611817,
      "epoch": 1.7486943336889986,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004702666005282759,
      "loss": 5.2656,
      "mean_token_accuracy": 0.1974201261997223,
      "num_tokens": 34772237.0,
      "step": 15570
    },
    {
      "entropy": 5.391970920562744,
      "epoch": 1.7492559105969563,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004702466505892304,
      "loss": 5.161,
      "mean_token_accuracy": 0.1986350581049919,
      "num_tokens": 34784648.0,
      "step": 15575
    },
    {
      "entropy": 5.514903831481933,
      "epoch": 1.749817487504914,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047022669443345575,
      "loss": 5.2862,
      "mean_token_accuracy": 0.19479808807373047,
      "num_tokens": 34794505.0,
      "step": 15580
    },
    {
      "entropy": 5.47529091835022,
      "epoch": 1.7503790644128714,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00047020673206158743,
      "loss": 5.2614,
      "mean_token_accuracy": 0.19599003046751023,
      "num_tokens": 34804887.0,
      "step": 15585
    },
    {
      "entropy": 5.515818786621094,
      "epoch": 1.7509406413208288,
      "grad_norm": 1.765625,
      "learning_rate": 0.000470186763474261,
      "loss": 5.276,
      "mean_token_accuracy": 0.19477153122425078,
      "num_tokens": 34816814.0,
      "step": 15590
    },
    {
      "entropy": 5.431199359893799,
      "epoch": 1.7515022182287865,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004701667886721125,
      "loss": 5.1885,
      "mean_token_accuracy": 0.2029449760913849,
      "num_tokens": 34827251.0,
      "step": 15595
    },
    {
      "entropy": 5.446490240097046,
      "epoch": 1.752063795136744,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004701468076557776,
      "loss": 5.3374,
      "mean_token_accuracy": 0.18841575235128402,
      "num_tokens": 34839010.0,
      "step": 15600
    },
    {
      "entropy": 5.50555534362793,
      "epoch": 1.7526253720447014,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047012682042589277,
      "loss": 5.3067,
      "mean_token_accuracy": 0.19196352511644363,
      "num_tokens": 34850435.0,
      "step": 15605
    },
    {
      "entropy": 5.404800128936768,
      "epoch": 1.753186948952659,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047010682698309423,
      "loss": 5.1616,
      "mean_token_accuracy": 0.20154355764389037,
      "num_tokens": 34861222.0,
      "step": 15610
    },
    {
      "entropy": 5.497649431228638,
      "epoch": 1.7537485258606167,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004700868273280186,
      "loss": 5.4099,
      "mean_token_accuracy": 0.1770594149827957,
      "num_tokens": 34872650.0,
      "step": 15615
    },
    {
      "entropy": 5.513383054733277,
      "epoch": 1.7543101027685741,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047006682146130273,
      "loss": 5.2038,
      "mean_token_accuracy": 0.20277346074581146,
      "num_tokens": 34883201.0,
      "step": 15620
    },
    {
      "entropy": 5.577050971984863,
      "epoch": 1.7548716796765316,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004700468093835837,
      "loss": 5.407,
      "mean_token_accuracy": 0.18148791044950485,
      "num_tokens": 34895936.0,
      "step": 15625
    },
    {
      "entropy": 5.438853073120117,
      "epoch": 1.7554332565844892,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047002679109549864,
      "loss": 5.2574,
      "mean_token_accuracy": 0.20195642560720445,
      "num_tokens": 34906491.0,
      "step": 15630
    },
    {
      "entropy": 5.410351514816284,
      "epoch": 1.755994833492447,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047000676659768477,
      "loss": 5.2679,
      "mean_token_accuracy": 0.19131840169429778,
      "num_tokens": 34918300.0,
      "step": 15635
    },
    {
      "entropy": 5.404550266265869,
      "epoch": 1.7565564104004043,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004699867358907799,
      "loss": 5.2076,
      "mean_token_accuracy": 0.1899575784802437,
      "num_tokens": 34928573.0,
      "step": 15640
    },
    {
      "entropy": 5.481291627883911,
      "epoch": 1.7571179873083618,
      "grad_norm": 1.125,
      "learning_rate": 0.0004699666989754217,
      "loss": 5.2721,
      "mean_token_accuracy": 0.1949351668357849,
      "num_tokens": 34940378.0,
      "step": 15645
    },
    {
      "entropy": 5.43005862236023,
      "epoch": 1.7576795642163194,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004699466558522482,
      "loss": 5.317,
      "mean_token_accuracy": 0.1905704990029335,
      "num_tokens": 34951627.0,
      "step": 15650
    },
    {
      "entropy": 5.463701820373535,
      "epoch": 1.758241141124277,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046992660652189755,
      "loss": 5.2445,
      "mean_token_accuracy": 0.1912709429860115,
      "num_tokens": 34961428.0,
      "step": 15655
    },
    {
      "entropy": 5.564219856262207,
      "epoch": 1.7588027180322345,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046990655098500824,
      "loss": 5.37,
      "mean_token_accuracy": 0.18633789271116258,
      "num_tokens": 34973736.0,
      "step": 15660
    },
    {
      "entropy": 5.477871131896973,
      "epoch": 1.759364294940192,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046988648924221866,
      "loss": 5.2395,
      "mean_token_accuracy": 0.20398525446653365,
      "num_tokens": 34984510.0,
      "step": 15665
    },
    {
      "entropy": 5.456082534790039,
      "epoch": 1.7599258718481496,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004698664212941677,
      "loss": 5.2476,
      "mean_token_accuracy": 0.19665072113275528,
      "num_tokens": 34995218.0,
      "step": 15670
    },
    {
      "entropy": 5.2973627090454105,
      "epoch": 1.7604874487561073,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046984634714149437,
      "loss": 5.0502,
      "mean_token_accuracy": 0.20763936042785644,
      "num_tokens": 35006681.0,
      "step": 15675
    },
    {
      "entropy": 5.4444938659667965,
      "epoch": 1.7610490256640647,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004698262667848377,
      "loss": 5.3,
      "mean_token_accuracy": 0.19175804555416107,
      "num_tokens": 35017341.0,
      "step": 15680
    },
    {
      "entropy": 5.433314990997315,
      "epoch": 1.7616106025720222,
      "grad_norm": 1.125,
      "learning_rate": 0.00046980618022483714,
      "loss": 5.1912,
      "mean_token_accuracy": 0.19930598735809327,
      "num_tokens": 35027718.0,
      "step": 15685
    },
    {
      "entropy": 5.520192003250122,
      "epoch": 1.7621721794799798,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046978608746213224,
      "loss": 5.259,
      "mean_token_accuracy": 0.19887492656707764,
      "num_tokens": 35038100.0,
      "step": 15690
    },
    {
      "entropy": 5.437778377532959,
      "epoch": 1.7627337563879373,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004697659884973628,
      "loss": 5.3025,
      "mean_token_accuracy": 0.19494782835245134,
      "num_tokens": 35049077.0,
      "step": 15695
    },
    {
      "entropy": 5.554211902618408,
      "epoch": 1.7632953332958947,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004697458833311687,
      "loss": 5.4193,
      "mean_token_accuracy": 0.19036051034927368,
      "num_tokens": 35060685.0,
      "step": 15700
    },
    {
      "entropy": 5.429080390930176,
      "epoch": 1.7638569102038524,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004697257719641902,
      "loss": 5.2139,
      "mean_token_accuracy": 0.19999346286058425,
      "num_tokens": 35072122.0,
      "step": 15705
    },
    {
      "entropy": 5.462600231170654,
      "epoch": 1.76441848711181,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004697056543970676,
      "loss": 5.2475,
      "mean_token_accuracy": 0.19505400061607361,
      "num_tokens": 35082427.0,
      "step": 15710
    },
    {
      "entropy": 5.5030198097229,
      "epoch": 1.7649800640197675,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046968553063044146,
      "loss": 5.2635,
      "mean_token_accuracy": 0.1970132037997246,
      "num_tokens": 35093061.0,
      "step": 15715
    },
    {
      "entropy": 5.510655355453491,
      "epoch": 1.765541640927725,
      "grad_norm": 1.25,
      "learning_rate": 0.00046966540066495247,
      "loss": 5.2865,
      "mean_token_accuracy": 0.19103433191776276,
      "num_tokens": 35103226.0,
      "step": 15720
    },
    {
      "entropy": 5.510529565811157,
      "epoch": 1.7661032178356826,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046964526450124164,
      "loss": 5.2699,
      "mean_token_accuracy": 0.19318871051073075,
      "num_tokens": 35113977.0,
      "step": 15725
    },
    {
      "entropy": 5.432564067840576,
      "epoch": 1.7666647947436402,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046962512213995,
      "loss": 5.2584,
      "mean_token_accuracy": 0.1915440320968628,
      "num_tokens": 35125132.0,
      "step": 15730
    },
    {
      "entropy": 5.4160041332244875,
      "epoch": 1.7672263716515977,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046960497358171913,
      "loss": 5.187,
      "mean_token_accuracy": 0.1970403626561165,
      "num_tokens": 35135266.0,
      "step": 15735
    },
    {
      "entropy": 5.490897941589355,
      "epoch": 1.7677879485595551,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046958481882719035,
      "loss": 5.2637,
      "mean_token_accuracy": 0.20451557487249375,
      "num_tokens": 35146268.0,
      "step": 15740
    },
    {
      "entropy": 5.427066612243652,
      "epoch": 1.7683495254675128,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046956465787700554,
      "loss": 5.1905,
      "mean_token_accuracy": 0.1946908324956894,
      "num_tokens": 35157046.0,
      "step": 15745
    },
    {
      "entropy": 5.526792812347412,
      "epoch": 1.7689111023754704,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046954449073180644,
      "loss": 5.3718,
      "mean_token_accuracy": 0.19114665389060975,
      "num_tokens": 35167624.0,
      "step": 15750
    },
    {
      "entropy": 5.53805980682373,
      "epoch": 1.7694726792834279,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004695243173922354,
      "loss": 5.3386,
      "mean_token_accuracy": 0.19464825093746185,
      "num_tokens": 35178987.0,
      "step": 15755
    },
    {
      "entropy": 5.433460474014282,
      "epoch": 1.7700342561913853,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046950413785893465,
      "loss": 5.1723,
      "mean_token_accuracy": 0.19622041434049606,
      "num_tokens": 35189804.0,
      "step": 15760
    },
    {
      "entropy": 5.497469568252564,
      "epoch": 1.770595833099343,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046948395213254667,
      "loss": 5.2386,
      "mean_token_accuracy": 0.1927215963602066,
      "num_tokens": 35200677.0,
      "step": 15765
    },
    {
      "entropy": 5.3624138832092285,
      "epoch": 1.7711574100073006,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004694637602137143,
      "loss": 5.2254,
      "mean_token_accuracy": 0.19677232950925827,
      "num_tokens": 35211570.0,
      "step": 15770
    },
    {
      "entropy": 5.337009334564209,
      "epoch": 1.771718986915258,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004694435621030803,
      "loss": 5.279,
      "mean_token_accuracy": 0.1915513664484024,
      "num_tokens": 35222451.0,
      "step": 15775
    },
    {
      "entropy": 5.565487957000732,
      "epoch": 1.7722805638232155,
      "grad_norm": 1.328125,
      "learning_rate": 0.00046942335780128785,
      "loss": 5.3737,
      "mean_token_accuracy": 0.1884576991200447,
      "num_tokens": 35233588.0,
      "step": 15780
    },
    {
      "entropy": 5.581005859375,
      "epoch": 1.7728421407311732,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046940314730898035,
      "loss": 5.3168,
      "mean_token_accuracy": 0.18655344545841218,
      "num_tokens": 35245766.0,
      "step": 15785
    },
    {
      "entropy": 5.3906886100769045,
      "epoch": 1.7734037176391306,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004693829306268012,
      "loss": 5.2574,
      "mean_token_accuracy": 0.1976647049188614,
      "num_tokens": 35255635.0,
      "step": 15790
    },
    {
      "entropy": 5.472668600082398,
      "epoch": 1.773965294547088,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004693627077553942,
      "loss": 5.2406,
      "mean_token_accuracy": 0.19204183518886567,
      "num_tokens": 35269342.0,
      "step": 15795
    },
    {
      "entropy": 5.523164367675781,
      "epoch": 1.7745268714550457,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004693424786954031,
      "loss": 5.2669,
      "mean_token_accuracy": 0.18786533921957016,
      "num_tokens": 35280735.0,
      "step": 15800
    },
    {
      "entropy": 5.447016096115112,
      "epoch": 1.7750884483630034,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046932224344747224,
      "loss": 5.2759,
      "mean_token_accuracy": 0.1893369436264038,
      "num_tokens": 35290803.0,
      "step": 15805
    },
    {
      "entropy": 5.458083248138427,
      "epoch": 1.7756500252709608,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046930200201224573,
      "loss": 5.2599,
      "mean_token_accuracy": 0.19553462862968446,
      "num_tokens": 35301811.0,
      "step": 15810
    },
    {
      "entropy": 5.538077068328858,
      "epoch": 1.7762116021789183,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004692817543903681,
      "loss": 5.3526,
      "mean_token_accuracy": 0.1851778119802475,
      "num_tokens": 35313374.0,
      "step": 15815
    },
    {
      "entropy": 5.5086558818817135,
      "epoch": 1.776773179086876,
      "grad_norm": 1.171875,
      "learning_rate": 0.000469261500582484,
      "loss": 5.213,
      "mean_token_accuracy": 0.20190156549215316,
      "num_tokens": 35323272.0,
      "step": 15820
    },
    {
      "entropy": 5.5644166469573975,
      "epoch": 1.7773347559948336,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00046924124058923843,
      "loss": 5.32,
      "mean_token_accuracy": 0.17922704070806503,
      "num_tokens": 35335720.0,
      "step": 15825
    },
    {
      "entropy": 5.477693367004394,
      "epoch": 1.777896332902791,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004692209744112765,
      "loss": 5.3681,
      "mean_token_accuracy": 0.19396947771310807,
      "num_tokens": 35347621.0,
      "step": 15830
    },
    {
      "entropy": 5.555389738082885,
      "epoch": 1.7784579098107485,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004692007020492433,
      "loss": 5.4638,
      "mean_token_accuracy": 0.1788490265607834,
      "num_tokens": 35360611.0,
      "step": 15835
    },
    {
      "entropy": 5.5429612159729,
      "epoch": 1.7790194867187061,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004691804235037845,
      "loss": 5.3547,
      "mean_token_accuracy": 0.1898273453116417,
      "num_tokens": 35372242.0,
      "step": 15840
    },
    {
      "entropy": 5.4759053707122805,
      "epoch": 1.7795810636266638,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004691601387755456,
      "loss": 5.2509,
      "mean_token_accuracy": 0.19679318964481354,
      "num_tokens": 35385718.0,
      "step": 15845
    },
    {
      "entropy": 5.487194538116455,
      "epoch": 1.7801426405346212,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046913984786517275,
      "loss": 5.3745,
      "mean_token_accuracy": 0.1855766624212265,
      "num_tokens": 35398298.0,
      "step": 15850
    },
    {
      "entropy": 5.487040424346924,
      "epoch": 1.7807042174425787,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00046911955077331177,
      "loss": 5.1505,
      "mean_token_accuracy": 0.19208363592624664,
      "num_tokens": 35408765.0,
      "step": 15855
    },
    {
      "entropy": 5.427100563049317,
      "epoch": 1.7812657943505363,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046909924750060893,
      "loss": 5.1662,
      "mean_token_accuracy": 0.1982302784919739,
      "num_tokens": 35418547.0,
      "step": 15860
    },
    {
      "entropy": 5.471491003036499,
      "epoch": 1.781827371258494,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004690789380477109,
      "loss": 5.2714,
      "mean_token_accuracy": 0.19071896225214005,
      "num_tokens": 35428743.0,
      "step": 15865
    },
    {
      "entropy": 5.367099571228027,
      "epoch": 1.7823889481664514,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004690586224152641,
      "loss": 5.2055,
      "mean_token_accuracy": 0.1951374292373657,
      "num_tokens": 35439557.0,
      "step": 15870
    },
    {
      "entropy": 5.481493902206421,
      "epoch": 1.7829505250744089,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004690383006039156,
      "loss": 5.3425,
      "mean_token_accuracy": 0.1849189057946205,
      "num_tokens": 35452041.0,
      "step": 15875
    },
    {
      "entropy": 5.562601947784424,
      "epoch": 1.7835121019823665,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004690179726143123,
      "loss": 5.2652,
      "mean_token_accuracy": 0.19939381033182144,
      "num_tokens": 35463283.0,
      "step": 15880
    },
    {
      "entropy": 5.515988540649414,
      "epoch": 1.784073678890324,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004689976384471014,
      "loss": 5.3788,
      "mean_token_accuracy": 0.20150773823261262,
      "num_tokens": 35475802.0,
      "step": 15885
    },
    {
      "entropy": 5.380010890960693,
      "epoch": 1.7846352557982814,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004689772981029306,
      "loss": 5.1754,
      "mean_token_accuracy": 0.20239955484867095,
      "num_tokens": 35486322.0,
      "step": 15890
    },
    {
      "entropy": 5.448879528045654,
      "epoch": 1.785196832706239,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004689569515824473,
      "loss": 5.2597,
      "mean_token_accuracy": 0.19475621730089188,
      "num_tokens": 35497021.0,
      "step": 15895
    },
    {
      "entropy": 5.466486406326294,
      "epoch": 1.7857584096141967,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004689365988862995,
      "loss": 5.2111,
      "mean_token_accuracy": 0.20281143933534623,
      "num_tokens": 35507579.0,
      "step": 15900
    },
    {
      "entropy": 5.406865119934082,
      "epoch": 1.7863199865221542,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046891624001513515,
      "loss": 5.1635,
      "mean_token_accuracy": 0.1920129492878914,
      "num_tokens": 35518131.0,
      "step": 15905
    },
    {
      "entropy": 5.468714427947998,
      "epoch": 1.7868815634301116,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046889587496960257,
      "loss": 5.3025,
      "mean_token_accuracy": 0.18764114677906035,
      "num_tokens": 35529577.0,
      "step": 15910
    },
    {
      "entropy": 5.474948310852051,
      "epoch": 1.7874431403380693,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046887550375034997,
      "loss": 5.2265,
      "mean_token_accuracy": 0.19944961369037628,
      "num_tokens": 35541555.0,
      "step": 15915
    },
    {
      "entropy": 5.438261842727661,
      "epoch": 1.788004717246027,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046885512635802627,
      "loss": 5.1769,
      "mean_token_accuracy": 0.19830718040466308,
      "num_tokens": 35552596.0,
      "step": 15920
    },
    {
      "entropy": 5.3410584926605225,
      "epoch": 1.7885662941539844,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004688347427932801,
      "loss": 5.2552,
      "mean_token_accuracy": 0.18777724653482436,
      "num_tokens": 35564962.0,
      "step": 15925
    },
    {
      "entropy": 5.542367458343506,
      "epoch": 1.7891278710619418,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046881435305676055,
      "loss": 5.3408,
      "mean_token_accuracy": 0.1888684645295143,
      "num_tokens": 35576436.0,
      "step": 15930
    },
    {
      "entropy": 5.487543392181396,
      "epoch": 1.7896894479698995,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046879395714911686,
      "loss": 5.2766,
      "mean_token_accuracy": 0.19473420828580856,
      "num_tokens": 35587467.0,
      "step": 15935
    },
    {
      "entropy": 5.520612287521362,
      "epoch": 1.7902510248778571,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046877355507099846,
      "loss": 5.3235,
      "mean_token_accuracy": 0.1848833367228508,
      "num_tokens": 35600302.0,
      "step": 15940
    },
    {
      "entropy": 5.510443449020386,
      "epoch": 1.7908126017858146,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004687531468230549,
      "loss": 5.2712,
      "mean_token_accuracy": 0.19405514895915985,
      "num_tokens": 35611757.0,
      "step": 15945
    },
    {
      "entropy": 5.407525062561035,
      "epoch": 1.791374178693772,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004687327324059359,
      "loss": 5.3179,
      "mean_token_accuracy": 0.19429656416177749,
      "num_tokens": 35621991.0,
      "step": 15950
    },
    {
      "entropy": 5.433898305892944,
      "epoch": 1.7919357556017297,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004687123118202916,
      "loss": 5.2912,
      "mean_token_accuracy": 0.19195396155118943,
      "num_tokens": 35633744.0,
      "step": 15955
    },
    {
      "entropy": 5.4222269535064695,
      "epoch": 1.7924973325096873,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004686918850667722,
      "loss": 5.1766,
      "mean_token_accuracy": 0.19886623024940492,
      "num_tokens": 35645523.0,
      "step": 15960
    },
    {
      "entropy": 5.522074365615845,
      "epoch": 1.7930589094176448,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000468671452146028,
      "loss": 5.3365,
      "mean_token_accuracy": 0.19130499362945558,
      "num_tokens": 35657670.0,
      "step": 15965
    },
    {
      "entropy": 5.420242738723755,
      "epoch": 1.7936204863256022,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046865101305870965,
      "loss": 5.1443,
      "mean_token_accuracy": 0.19701539427042009,
      "num_tokens": 35668149.0,
      "step": 15970
    },
    {
      "entropy": 5.419379568099975,
      "epoch": 1.7941820632335599,
      "grad_norm": 1.09375,
      "learning_rate": 0.000468630567805468,
      "loss": 5.2201,
      "mean_token_accuracy": 0.19425840377807618,
      "num_tokens": 35680225.0,
      "step": 15975
    },
    {
      "entropy": 5.366401529312133,
      "epoch": 1.7947436401415173,
      "grad_norm": 1.125,
      "learning_rate": 0.000468610116386954,
      "loss": 5.1348,
      "mean_token_accuracy": 0.20213534832000732,
      "num_tokens": 35691042.0,
      "step": 15980
    },
    {
      "entropy": 5.442798137664795,
      "epoch": 1.7953052170494748,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004685896588038187,
      "loss": 5.2413,
      "mean_token_accuracy": 0.18850397169589997,
      "num_tokens": 35702895.0,
      "step": 15985
    },
    {
      "entropy": 5.469142055511474,
      "epoch": 1.7958667939574324,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004685691950567136,
      "loss": 5.1865,
      "mean_token_accuracy": 0.1984875127673149,
      "num_tokens": 35714765.0,
      "step": 15990
    },
    {
      "entropy": 5.506585168838501,
      "epoch": 1.79642837086539,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046854872514629035,
      "loss": 5.3701,
      "mean_token_accuracy": 0.18735479861497878,
      "num_tokens": 35725832.0,
      "step": 15995
    },
    {
      "entropy": 5.393569755554199,
      "epoch": 1.7969899477733475,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004685282490732005,
      "loss": 5.1912,
      "mean_token_accuracy": 0.2007139205932617,
      "num_tokens": 35736619.0,
      "step": 16000
    },
    {
      "entropy": 5.400776386260986,
      "epoch": 1.797551524681305,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046850776683809623,
      "loss": 5.1473,
      "mean_token_accuracy": 0.1935574010014534,
      "num_tokens": 35748243.0,
      "step": 16005
    },
    {
      "entropy": 5.444024991989136,
      "epoch": 1.7981131015892626,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004684872784416295,
      "loss": 5.259,
      "mean_token_accuracy": 0.19114589393138887,
      "num_tokens": 35759126.0,
      "step": 16010
    },
    {
      "entropy": 5.532953977584839,
      "epoch": 1.7986746784972203,
      "grad_norm": 1.125,
      "learning_rate": 0.0004684667838844529,
      "loss": 5.2666,
      "mean_token_accuracy": 0.19306623488664626,
      "num_tokens": 35770998.0,
      "step": 16015
    },
    {
      "entropy": 5.431976842880249,
      "epoch": 1.7992362554051777,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004684462831672188,
      "loss": 5.2123,
      "mean_token_accuracy": 0.1986193597316742,
      "num_tokens": 35782835.0,
      "step": 16020
    },
    {
      "entropy": 5.451713752746582,
      "epoch": 1.7997978323131352,
      "grad_norm": 1.125,
      "learning_rate": 0.00046842577629058006,
      "loss": 5.2149,
      "mean_token_accuracy": 0.191050124168396,
      "num_tokens": 35792828.0,
      "step": 16025
    },
    {
      "entropy": 5.5196390628814695,
      "epoch": 1.8003594092210928,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004684052632551896,
      "loss": 5.408,
      "mean_token_accuracy": 0.1905906543135643,
      "num_tokens": 35804449.0,
      "step": 16030
    },
    {
      "entropy": 5.490135622024536,
      "epoch": 1.8009209861290505,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046838474406170046,
      "loss": 5.1893,
      "mean_token_accuracy": 0.19972139745950698,
      "num_tokens": 35815903.0,
      "step": 16035
    },
    {
      "entropy": 5.369788122177124,
      "epoch": 1.801482563037008,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046836421871076616,
      "loss": 5.1809,
      "mean_token_accuracy": 0.19678596556186675,
      "num_tokens": 35827872.0,
      "step": 16040
    },
    {
      "entropy": 5.412657117843628,
      "epoch": 1.8020441399449654,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004683436872030401,
      "loss": 5.1959,
      "mean_token_accuracy": 0.20014331340789795,
      "num_tokens": 35838864.0,
      "step": 16045
    },
    {
      "entropy": 5.406856966018677,
      "epoch": 1.802605716852923,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046832314953917605,
      "loss": 5.1053,
      "mean_token_accuracy": 0.19173719733953476,
      "num_tokens": 35850178.0,
      "step": 16050
    },
    {
      "entropy": 5.413277101516724,
      "epoch": 1.8031672937608807,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046830260571982795,
      "loss": 5.2721,
      "mean_token_accuracy": 0.19639455676078796,
      "num_tokens": 35861904.0,
      "step": 16055
    },
    {
      "entropy": 5.328383874893189,
      "epoch": 1.8037288706688381,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004682820557456499,
      "loss": 5.1408,
      "mean_token_accuracy": 0.20161622017621994,
      "num_tokens": 35872977.0,
      "step": 16060
    },
    {
      "entropy": 5.549663257598877,
      "epoch": 1.8042904475767956,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046826149961729625,
      "loss": 5.3317,
      "mean_token_accuracy": 0.19230997413396836,
      "num_tokens": 35886094.0,
      "step": 16065
    },
    {
      "entropy": 5.364774179458618,
      "epoch": 1.8048520244847532,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046824093733542144,
      "loss": 5.1652,
      "mean_token_accuracy": 0.20203492492437364,
      "num_tokens": 35897557.0,
      "step": 16070
    },
    {
      "entropy": 5.528737115859985,
      "epoch": 1.8054136013927107,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046822036890068033,
      "loss": 5.3644,
      "mean_token_accuracy": 0.1872581958770752,
      "num_tokens": 35910605.0,
      "step": 16075
    },
    {
      "entropy": 5.460405969619751,
      "epoch": 1.8059751783006681,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046819979431372773,
      "loss": 5.1599,
      "mean_token_accuracy": 0.19832293689250946,
      "num_tokens": 35921062.0,
      "step": 16080
    },
    {
      "entropy": 5.441573286056519,
      "epoch": 1.8065367552086258,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046817921357521865,
      "loss": 5.2379,
      "mean_token_accuracy": 0.1932335987687111,
      "num_tokens": 35932385.0,
      "step": 16085
    },
    {
      "entropy": 5.381153249740601,
      "epoch": 1.8070983321165834,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046815862668580854,
      "loss": 5.2394,
      "mean_token_accuracy": 0.202308289706707,
      "num_tokens": 35943628.0,
      "step": 16090
    },
    {
      "entropy": 5.493329858779907,
      "epoch": 1.8076599090245409,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004681380336461528,
      "loss": 5.3063,
      "mean_token_accuracy": 0.18229345679283143,
      "num_tokens": 35955074.0,
      "step": 16095
    },
    {
      "entropy": 5.516821718215942,
      "epoch": 1.8082214859324983,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046811743445690724,
      "loss": 5.2119,
      "mean_token_accuracy": 0.19212692677974702,
      "num_tokens": 35966200.0,
      "step": 16100
    },
    {
      "entropy": 5.369517183303833,
      "epoch": 1.808783062840456,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004680968291187276,
      "loss": 5.1218,
      "mean_token_accuracy": 0.20503711998462676,
      "num_tokens": 35975638.0,
      "step": 16105
    },
    {
      "entropy": 5.384795665740967,
      "epoch": 1.8093446397484136,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004680762176322701,
      "loss": 5.2458,
      "mean_token_accuracy": 0.19826361685991287,
      "num_tokens": 35986486.0,
      "step": 16110
    },
    {
      "entropy": 5.400921392440796,
      "epoch": 1.809906216656371,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046805559999819093,
      "loss": 5.165,
      "mean_token_accuracy": 0.20081044882535934,
      "num_tokens": 35996612.0,
      "step": 16115
    },
    {
      "entropy": 5.363110828399658,
      "epoch": 1.8104677935643285,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004680349762171465,
      "loss": 5.2366,
      "mean_token_accuracy": 0.19668970108032227,
      "num_tokens": 36006936.0,
      "step": 16120
    },
    {
      "entropy": 5.556390571594238,
      "epoch": 1.8110293704722862,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004680143462897936,
      "loss": 5.3438,
      "mean_token_accuracy": 0.18927601724863052,
      "num_tokens": 36019562.0,
      "step": 16125
    },
    {
      "entropy": 5.427904033660889,
      "epoch": 1.8115909473802438,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004679937102167891,
      "loss": 5.1511,
      "mean_token_accuracy": 0.19522274285554886,
      "num_tokens": 36029585.0,
      "step": 16130
    },
    {
      "entropy": 5.399410057067871,
      "epoch": 1.8121525242882013,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046797306799879,
      "loss": 5.2203,
      "mean_token_accuracy": 0.20020872801542283,
      "num_tokens": 36041495.0,
      "step": 16135
    },
    {
      "entropy": 5.427336263656616,
      "epoch": 1.8127141011961587,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046795241963645356,
      "loss": 5.2729,
      "mean_token_accuracy": 0.1923805519938469,
      "num_tokens": 36052087.0,
      "step": 16140
    },
    {
      "entropy": 5.481294631958008,
      "epoch": 1.8132756781041164,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004679317651304373,
      "loss": 5.2324,
      "mean_token_accuracy": 0.19829607158899307,
      "num_tokens": 36065011.0,
      "step": 16145
    },
    {
      "entropy": 5.445209741592407,
      "epoch": 1.813837255012074,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004679111044813987,
      "loss": 5.2912,
      "mean_token_accuracy": 0.20142097622156144,
      "num_tokens": 36075912.0,
      "step": 16150
    },
    {
      "entropy": 5.4640837669372555,
      "epoch": 1.8143988319200315,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046789043768999573,
      "loss": 5.2521,
      "mean_token_accuracy": 0.19256698042154313,
      "num_tokens": 36086757.0,
      "step": 16155
    },
    {
      "entropy": 5.4217077732086185,
      "epoch": 1.814960408827989,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004678697647568865,
      "loss": 5.1554,
      "mean_token_accuracy": 0.20215559303760527,
      "num_tokens": 36097319.0,
      "step": 16160
    },
    {
      "entropy": 5.455932903289795,
      "epoch": 1.8155219857359466,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046784908568272903,
      "loss": 5.2996,
      "mean_token_accuracy": 0.1942580074071884,
      "num_tokens": 36108416.0,
      "step": 16165
    },
    {
      "entropy": 5.367696619033813,
      "epoch": 1.8160835626439042,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000467828400468182,
      "loss": 5.2192,
      "mean_token_accuracy": 0.1924705058336258,
      "num_tokens": 36120833.0,
      "step": 16170
    },
    {
      "entropy": 5.487527990341187,
      "epoch": 1.8166451395518615,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046780770911390387,
      "loss": 5.3519,
      "mean_token_accuracy": 0.19003746211528777,
      "num_tokens": 36132348.0,
      "step": 16175
    },
    {
      "entropy": 5.462514352798462,
      "epoch": 1.8172067164598191,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004677870116205534,
      "loss": 5.1243,
      "mean_token_accuracy": 0.20007028430700302,
      "num_tokens": 36142829.0,
      "step": 16180
    },
    {
      "entropy": 5.4965400218963625,
      "epoch": 1.8177682933677768,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046776630798878985,
      "loss": 5.3935,
      "mean_token_accuracy": 0.18777074068784713,
      "num_tokens": 36155931.0,
      "step": 16185
    },
    {
      "entropy": 5.4444843292236325,
      "epoch": 1.8183298702757342,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004677455982192723,
      "loss": 5.2897,
      "mean_token_accuracy": 0.18703046888113023,
      "num_tokens": 36167591.0,
      "step": 16190
    },
    {
      "entropy": 5.451463985443115,
      "epoch": 1.8188914471836917,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046772488231266004,
      "loss": 5.1795,
      "mean_token_accuracy": 0.19847027957439423,
      "num_tokens": 36178867.0,
      "step": 16195
    },
    {
      "entropy": 5.324956607818604,
      "epoch": 1.8194530240916493,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046770416026961284,
      "loss": 5.1321,
      "mean_token_accuracy": 0.20928287357091904,
      "num_tokens": 36190691.0,
      "step": 16200
    },
    {
      "entropy": 5.5048281192779545,
      "epoch": 1.820014600999607,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004676834320907904,
      "loss": 5.3739,
      "mean_token_accuracy": 0.18715667724609375,
      "num_tokens": 36202222.0,
      "step": 16205
    },
    {
      "entropy": 5.433983898162841,
      "epoch": 1.8205761779075644,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004676626977768528,
      "loss": 5.141,
      "mean_token_accuracy": 0.20337275713682174,
      "num_tokens": 36214303.0,
      "step": 16210
    },
    {
      "entropy": 5.419623231887817,
      "epoch": 1.8211377548155219,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046764195732846013,
      "loss": 5.1753,
      "mean_token_accuracy": 0.19822729229927064,
      "num_tokens": 36224440.0,
      "step": 16215
    },
    {
      "entropy": 5.332764339447022,
      "epoch": 1.8216993317234795,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046762121074627285,
      "loss": 5.1698,
      "mean_token_accuracy": 0.2013513743877411,
      "num_tokens": 36235445.0,
      "step": 16220
    },
    {
      "entropy": 5.422706937789917,
      "epoch": 1.8222609086314372,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046760045803095153,
      "loss": 5.1947,
      "mean_token_accuracy": 0.19901381582021713,
      "num_tokens": 36247024.0,
      "step": 16225
    },
    {
      "entropy": 5.499009370803833,
      "epoch": 1.8228224855393946,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004675796991831569,
      "loss": 5.2299,
      "mean_token_accuracy": 0.19093571305274964,
      "num_tokens": 36257406.0,
      "step": 16230
    },
    {
      "entropy": 5.452253007888794,
      "epoch": 1.823384062447352,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046755893420355,
      "loss": 5.2381,
      "mean_token_accuracy": 0.19847143888473512,
      "num_tokens": 36270266.0,
      "step": 16235
    },
    {
      "entropy": 5.467320680618286,
      "epoch": 1.8239456393553097,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004675381630927919,
      "loss": 5.3224,
      "mean_token_accuracy": 0.1926882192492485,
      "num_tokens": 36281470.0,
      "step": 16240
    },
    {
      "entropy": 5.412930631637574,
      "epoch": 1.8245072162632674,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046751738585154405,
      "loss": 5.2041,
      "mean_token_accuracy": 0.19541314393281936,
      "num_tokens": 36293315.0,
      "step": 16245
    },
    {
      "entropy": 5.491446208953858,
      "epoch": 1.8250687931712248,
      "grad_norm": 1.25,
      "learning_rate": 0.00046749660248046794,
      "loss": 5.2565,
      "mean_token_accuracy": 0.1963328778743744,
      "num_tokens": 36304688.0,
      "step": 16250
    },
    {
      "entropy": 5.516882276535034,
      "epoch": 1.8256303700791823,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046747581298022537,
      "loss": 5.3112,
      "mean_token_accuracy": 0.19934389293193816,
      "num_tokens": 36315773.0,
      "step": 16255
    },
    {
      "entropy": 5.512779998779297,
      "epoch": 1.82619194698714,
      "grad_norm": 1.375,
      "learning_rate": 0.0004674550173514782,
      "loss": 5.2409,
      "mean_token_accuracy": 0.19458285272121428,
      "num_tokens": 36327914.0,
      "step": 16260
    },
    {
      "entropy": 5.472334384918213,
      "epoch": 1.8267535238950976,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046743421559488875,
      "loss": 5.3255,
      "mean_token_accuracy": 0.18716292232275009,
      "num_tokens": 36339617.0,
      "step": 16265
    },
    {
      "entropy": 5.534595632553101,
      "epoch": 1.8273151008030548,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046741340771111917,
      "loss": 5.3825,
      "mean_token_accuracy": 0.1893995776772499,
      "num_tokens": 36350768.0,
      "step": 16270
    },
    {
      "entropy": 5.417033958435058,
      "epoch": 1.8278766777110125,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046739259370083205,
      "loss": 5.2495,
      "mean_token_accuracy": 0.19959975481033326,
      "num_tokens": 36361267.0,
      "step": 16275
    },
    {
      "entropy": 5.437523984909058,
      "epoch": 1.8284382546189701,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004673717735646902,
      "loss": 5.2799,
      "mean_token_accuracy": 0.19092872738838196,
      "num_tokens": 36373259.0,
      "step": 16280
    },
    {
      "entropy": 5.464200639724732,
      "epoch": 1.8289998315269276,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004673509473033564,
      "loss": 5.2285,
      "mean_token_accuracy": 0.19183885902166367,
      "num_tokens": 36384295.0,
      "step": 16285
    },
    {
      "entropy": 5.447967958450318,
      "epoch": 1.829561408434885,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046733011491749383,
      "loss": 5.2059,
      "mean_token_accuracy": 0.20471107959747314,
      "num_tokens": 36394928.0,
      "step": 16290
    },
    {
      "entropy": 5.372482109069824,
      "epoch": 1.8301229853428427,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004673092764077658,
      "loss": 5.2578,
      "mean_token_accuracy": 0.20096465349197387,
      "num_tokens": 36407038.0,
      "step": 16295
    },
    {
      "entropy": 5.4954423904418945,
      "epoch": 1.8306845622508003,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004672884317748359,
      "loss": 5.2334,
      "mean_token_accuracy": 0.19265564978122712,
      "num_tokens": 36418221.0,
      "step": 16300
    },
    {
      "entropy": 5.441134881973267,
      "epoch": 1.8312461391587578,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004672675810193677,
      "loss": 5.1916,
      "mean_token_accuracy": 0.19624388068914414,
      "num_tokens": 36429036.0,
      "step": 16305
    },
    {
      "entropy": 5.416842937469482,
      "epoch": 1.8318077160667152,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046724672414202514,
      "loss": 5.2433,
      "mean_token_accuracy": 0.1970800518989563,
      "num_tokens": 36439851.0,
      "step": 16310
    },
    {
      "entropy": 5.446098899841308,
      "epoch": 1.8323692929746729,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004672258611434724,
      "loss": 5.2629,
      "mean_token_accuracy": 0.19566724300384522,
      "num_tokens": 36451612.0,
      "step": 16315
    },
    {
      "entropy": 5.523445892333984,
      "epoch": 1.8329308698826305,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046720499202437354,
      "loss": 5.3144,
      "mean_token_accuracy": 0.1889577701687813,
      "num_tokens": 36462883.0,
      "step": 16320
    },
    {
      "entropy": 5.395450925827026,
      "epoch": 1.833492446790588,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004671841167853933,
      "loss": 5.2105,
      "mean_token_accuracy": 0.19854866415262223,
      "num_tokens": 36473609.0,
      "step": 16325
    },
    {
      "entropy": 5.414622402191162,
      "epoch": 1.8340540236985454,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004671632354271962,
      "loss": 5.2057,
      "mean_token_accuracy": 0.19414829164743425,
      "num_tokens": 36484058.0,
      "step": 16330
    },
    {
      "entropy": 5.5043737411499025,
      "epoch": 1.834615600606503,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046714234795044724,
      "loss": 5.2345,
      "mean_token_accuracy": 0.19537086188793182,
      "num_tokens": 36496445.0,
      "step": 16335
    },
    {
      "entropy": 5.3957444190979,
      "epoch": 1.8351771775144607,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004671214543558113,
      "loss": 5.1865,
      "mean_token_accuracy": 0.1979435920715332,
      "num_tokens": 36507553.0,
      "step": 16340
    },
    {
      "entropy": 5.417101621627808,
      "epoch": 1.8357387544224182,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004671005546439538,
      "loss": 5.2216,
      "mean_token_accuracy": 0.19782491624355317,
      "num_tokens": 36517887.0,
      "step": 16345
    },
    {
      "entropy": 5.424105262756347,
      "epoch": 1.8363003313303756,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046707964881554005,
      "loss": 5.1789,
      "mean_token_accuracy": 0.19635697603225707,
      "num_tokens": 36529185.0,
      "step": 16350
    },
    {
      "entropy": 5.406076908111572,
      "epoch": 1.8368619082383333,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046705873687123586,
      "loss": 5.2692,
      "mean_token_accuracy": 0.19579024016857147,
      "num_tokens": 36539473.0,
      "step": 16355
    },
    {
      "entropy": 5.506750631332397,
      "epoch": 1.837423485146291,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046703781881170703,
      "loss": 5.2763,
      "mean_token_accuracy": 0.19624565839767455,
      "num_tokens": 36549603.0,
      "step": 16360
    },
    {
      "entropy": 5.45345516204834,
      "epoch": 1.8379850620542482,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00046701689463761947,
      "loss": 5.2289,
      "mean_token_accuracy": 0.1998908191919327,
      "num_tokens": 36560394.0,
      "step": 16365
    },
    {
      "entropy": 5.484287929534912,
      "epoch": 1.8385466389622058,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046699596434963953,
      "loss": 5.2519,
      "mean_token_accuracy": 0.19336451292037965,
      "num_tokens": 36570657.0,
      "step": 16370
    },
    {
      "entropy": 5.375856018066406,
      "epoch": 1.8391082158701635,
      "grad_norm": 1.25,
      "learning_rate": 0.0004669750279484337,
      "loss": 5.1248,
      "mean_token_accuracy": 0.2009825363755226,
      "num_tokens": 36580810.0,
      "step": 16375
    },
    {
      "entropy": 5.3637425899505615,
      "epoch": 1.839669792778121,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004669540854346685,
      "loss": 5.148,
      "mean_token_accuracy": 0.20142463147640227,
      "num_tokens": 36592981.0,
      "step": 16380
    },
    {
      "entropy": 5.317112445831299,
      "epoch": 1.8402313696860784,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046693313680901073,
      "loss": 5.1575,
      "mean_token_accuracy": 0.20143572986125946,
      "num_tokens": 36604130.0,
      "step": 16385
    },
    {
      "entropy": 5.344953393936157,
      "epoch": 1.840792946594036,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046691218207212744,
      "loss": 5.1858,
      "mean_token_accuracy": 0.19324650317430497,
      "num_tokens": 36614833.0,
      "step": 16390
    },
    {
      "entropy": 5.529646110534668,
      "epoch": 1.8413545235019937,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004668912212246859,
      "loss": 5.1902,
      "mean_token_accuracy": 0.198896324634552,
      "num_tokens": 36625306.0,
      "step": 16395
    },
    {
      "entropy": 5.582696199417114,
      "epoch": 1.8419161004099511,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046687025426735335,
      "loss": 5.4318,
      "mean_token_accuracy": 0.1826240062713623,
      "num_tokens": 36637396.0,
      "step": 16400
    },
    {
      "entropy": 5.4382189273834225,
      "epoch": 1.8424776773179086,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004668492812007976,
      "loss": 5.2266,
      "mean_token_accuracy": 0.18789196610450745,
      "num_tokens": 36648175.0,
      "step": 16405
    },
    {
      "entropy": 5.4672997951507565,
      "epoch": 1.8430392542258662,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004668283020256863,
      "loss": 5.2878,
      "mean_token_accuracy": 0.1911124989390373,
      "num_tokens": 36659876.0,
      "step": 16410
    },
    {
      "entropy": 5.494023990631104,
      "epoch": 1.843600831133824,
      "grad_norm": 1.59375,
      "learning_rate": 0.00046680731674268744,
      "loss": 5.2595,
      "mean_token_accuracy": 0.1967497706413269,
      "num_tokens": 36671168.0,
      "step": 16415
    },
    {
      "entropy": 5.4021769046783445,
      "epoch": 1.8441624080417813,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004667863253524692,
      "loss": 5.2042,
      "mean_token_accuracy": 0.20788257271051408,
      "num_tokens": 36680987.0,
      "step": 16420
    },
    {
      "entropy": 5.435606527328491,
      "epoch": 1.8447239849497388,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046676532785569995,
      "loss": 5.2616,
      "mean_token_accuracy": 0.1980789467692375,
      "num_tokens": 36692396.0,
      "step": 16425
    },
    {
      "entropy": 5.409848022460937,
      "epoch": 1.8452855618576964,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004667443242530484,
      "loss": 5.1488,
      "mean_token_accuracy": 0.2015661284327507,
      "num_tokens": 36702696.0,
      "step": 16430
    },
    {
      "entropy": 5.378023862838745,
      "epoch": 1.845847138765654,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00046672331454518314,
      "loss": 5.1736,
      "mean_token_accuracy": 0.20051533877849578,
      "num_tokens": 36713582.0,
      "step": 16435
    },
    {
      "entropy": 5.404054880142212,
      "epoch": 1.8464087156736115,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004667022987327732,
      "loss": 5.1785,
      "mean_token_accuracy": 0.20067718625068665,
      "num_tokens": 36724324.0,
      "step": 16440
    },
    {
      "entropy": 5.466361236572266,
      "epoch": 1.846970292581569,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004666812768164877,
      "loss": 5.2521,
      "mean_token_accuracy": 0.20517339259386064,
      "num_tokens": 36736369.0,
      "step": 16445
    },
    {
      "entropy": 5.492540168762207,
      "epoch": 1.8475318694895266,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046666024879699606,
      "loss": 5.3012,
      "mean_token_accuracy": 0.19992835223674774,
      "num_tokens": 36748047.0,
      "step": 16450
    },
    {
      "entropy": 5.489131164550781,
      "epoch": 1.8480934463974843,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046663921467496766,
      "loss": 5.2303,
      "mean_token_accuracy": 0.20209592431783677,
      "num_tokens": 36758943.0,
      "step": 16455
    },
    {
      "entropy": 5.4915131568908695,
      "epoch": 1.8486550233054415,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004666181744510724,
      "loss": 5.272,
      "mean_token_accuracy": 0.1926899239420891,
      "num_tokens": 36769807.0,
      "step": 16460
    },
    {
      "entropy": 5.347870922088623,
      "epoch": 1.8492166002133992,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046659712812598017,
      "loss": 5.1578,
      "mean_token_accuracy": 0.20521916598081588,
      "num_tokens": 36780370.0,
      "step": 16465
    },
    {
      "entropy": 5.374262905120849,
      "epoch": 1.8497781771213568,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046657607570036103,
      "loss": 5.1584,
      "mean_token_accuracy": 0.20322951227426528,
      "num_tokens": 36789757.0,
      "step": 16470
    },
    {
      "entropy": 5.4533239841461185,
      "epoch": 1.8503397540293143,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046655501717488533,
      "loss": 5.2039,
      "mean_token_accuracy": 0.19609831869602204,
      "num_tokens": 36800531.0,
      "step": 16475
    },
    {
      "entropy": 5.454979181289673,
      "epoch": 1.8509013309372717,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004665339525502236,
      "loss": 5.2944,
      "mean_token_accuracy": 0.19065287709236145,
      "num_tokens": 36811666.0,
      "step": 16480
    },
    {
      "entropy": 5.429889678955078,
      "epoch": 1.8514629078452294,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046651288182704653,
      "loss": 5.1556,
      "mean_token_accuracy": 0.19791501611471177,
      "num_tokens": 36822443.0,
      "step": 16485
    },
    {
      "entropy": 5.442723321914673,
      "epoch": 1.852024484753187,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004664918050060251,
      "loss": 5.2248,
      "mean_token_accuracy": 0.19585804790258407,
      "num_tokens": 36835517.0,
      "step": 16490
    },
    {
      "entropy": 5.390813398361206,
      "epoch": 1.8525860616611445,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004664707220878302,
      "loss": 5.2449,
      "mean_token_accuracy": 0.20134783834218978,
      "num_tokens": 36845177.0,
      "step": 16495
    },
    {
      "entropy": 5.374290657043457,
      "epoch": 1.853147638569102,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004664496330731334,
      "loss": 5.1765,
      "mean_token_accuracy": 0.19992644786834718,
      "num_tokens": 36855821.0,
      "step": 16500
    },
    {
      "entropy": 5.468858003616333,
      "epoch": 1.8537092154770596,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004664285379626059,
      "loss": 5.2663,
      "mean_token_accuracy": 0.1844902217388153,
      "num_tokens": 36866254.0,
      "step": 16505
    },
    {
      "entropy": 5.354598093032837,
      "epoch": 1.8542707923850172,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046640743675691954,
      "loss": 5.1183,
      "mean_token_accuracy": 0.19865940511226654,
      "num_tokens": 36876348.0,
      "step": 16510
    },
    {
      "entropy": 5.456766176223755,
      "epoch": 1.8548323692929747,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004663863294567462,
      "loss": 5.2741,
      "mean_token_accuracy": 0.19141979664564132,
      "num_tokens": 36888088.0,
      "step": 16515
    },
    {
      "entropy": 5.589468717575073,
      "epoch": 1.8553939462009321,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004663652160627579,
      "loss": 5.371,
      "mean_token_accuracy": 0.19015635550022125,
      "num_tokens": 36900203.0,
      "step": 16520
    },
    {
      "entropy": 5.4709391593933105,
      "epoch": 1.8559555231088898,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046634409657562683,
      "loss": 5.2319,
      "mean_token_accuracy": 0.19422696232795716,
      "num_tokens": 36911321.0,
      "step": 16525
    },
    {
      "entropy": 5.460640859603882,
      "epoch": 1.8565171000168474,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004663229709960255,
      "loss": 5.3209,
      "mean_token_accuracy": 0.19540603458881378,
      "num_tokens": 36922562.0,
      "step": 16530
    },
    {
      "entropy": 5.448884057998657,
      "epoch": 1.8570786769248049,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046630183932462664,
      "loss": 5.1793,
      "mean_token_accuracy": 0.19607226997613908,
      "num_tokens": 36933923.0,
      "step": 16535
    },
    {
      "entropy": 5.42947735786438,
      "epoch": 1.8576402538327623,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004662807015621031,
      "loss": 5.201,
      "mean_token_accuracy": 0.1955616682767868,
      "num_tokens": 36945203.0,
      "step": 16540
    },
    {
      "entropy": 5.401079988479614,
      "epoch": 1.85820183074072,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046625955770912764,
      "loss": 5.1926,
      "mean_token_accuracy": 0.19551490247249603,
      "num_tokens": 36956094.0,
      "step": 16545
    },
    {
      "entropy": 5.4257652759552,
      "epoch": 1.8587634076486776,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046623840776637384,
      "loss": 5.249,
      "mean_token_accuracy": 0.19914478659629822,
      "num_tokens": 36967375.0,
      "step": 16550
    },
    {
      "entropy": 5.346981334686279,
      "epoch": 1.859324984556635,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004662172517345149,
      "loss": 5.1446,
      "mean_token_accuracy": 0.19924600124359132,
      "num_tokens": 36977674.0,
      "step": 16555
    },
    {
      "entropy": 5.44625563621521,
      "epoch": 1.8598865614645925,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004661960896142245,
      "loss": 5.2723,
      "mean_token_accuracy": 0.1989913135766983,
      "num_tokens": 36990241.0,
      "step": 16560
    },
    {
      "entropy": 5.400309371948242,
      "epoch": 1.8604481383725502,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046617492140617654,
      "loss": 5.2497,
      "mean_token_accuracy": 0.18825239092111587,
      "num_tokens": 37002322.0,
      "step": 16565
    },
    {
      "entropy": 5.469271945953369,
      "epoch": 1.8610097152805076,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004661537471110448,
      "loss": 5.2546,
      "mean_token_accuracy": 0.19534233212471008,
      "num_tokens": 37014524.0,
      "step": 16570
    },
    {
      "entropy": 5.420941495895386,
      "epoch": 1.861571292188465,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004661325667295037,
      "loss": 5.1184,
      "mean_token_accuracy": 0.20517282485961913,
      "num_tokens": 37025057.0,
      "step": 16575
    },
    {
      "entropy": 5.430609893798828,
      "epoch": 1.8621328690964227,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004661113802622275,
      "loss": 5.278,
      "mean_token_accuracy": 0.1973417192697525,
      "num_tokens": 37035311.0,
      "step": 16580
    },
    {
      "entropy": 5.4676745414733885,
      "epoch": 1.8626944460043804,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004660901877098908,
      "loss": 5.276,
      "mean_token_accuracy": 0.1964414522051811,
      "num_tokens": 37046275.0,
      "step": 16585
    },
    {
      "entropy": 5.433700656890869,
      "epoch": 1.8632560229123378,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004660689890731685,
      "loss": 5.1125,
      "mean_token_accuracy": 0.198580202460289,
      "num_tokens": 37057658.0,
      "step": 16590
    },
    {
      "entropy": 5.413163948059082,
      "epoch": 1.8638175998202953,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046604778435273546,
      "loss": 5.1404,
      "mean_token_accuracy": 0.21062769144773483,
      "num_tokens": 37067946.0,
      "step": 16595
    },
    {
      "entropy": 5.384222984313965,
      "epoch": 1.864379176728253,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046602657354926686,
      "loss": 5.1559,
      "mean_token_accuracy": 0.2003291666507721,
      "num_tokens": 37078881.0,
      "step": 16600
    },
    {
      "entropy": 5.4461586475372314,
      "epoch": 1.8649407536362106,
      "grad_norm": 1.203125,
      "learning_rate": 0.000466005356663438,
      "loss": 5.2837,
      "mean_token_accuracy": 0.18932868987321855,
      "num_tokens": 37090474.0,
      "step": 16605
    },
    {
      "entropy": 5.493841552734375,
      "epoch": 1.865502330544168,
      "grad_norm": 1.296875,
      "learning_rate": 0.00046598413369592463,
      "loss": 5.3352,
      "mean_token_accuracy": 0.1964936524629593,
      "num_tokens": 37101527.0,
      "step": 16610
    },
    {
      "entropy": 5.462956523895263,
      "epoch": 1.8660639074521255,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004659629046474023,
      "loss": 5.2386,
      "mean_token_accuracy": 0.19062045812606812,
      "num_tokens": 37112615.0,
      "step": 16615
    },
    {
      "entropy": 5.406205177307129,
      "epoch": 1.8666254843600831,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000465941669518547,
      "loss": 5.1388,
      "mean_token_accuracy": 0.20119487047195433,
      "num_tokens": 37122922.0,
      "step": 16620
    },
    {
      "entropy": 5.383013677597046,
      "epoch": 1.8671870612680408,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004659204283100349,
      "loss": 5.1643,
      "mean_token_accuracy": 0.20362237244844436,
      "num_tokens": 37133215.0,
      "step": 16625
    },
    {
      "entropy": 5.302173280715943,
      "epoch": 1.8677486381759982,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00046589918102254237,
      "loss": 5.0885,
      "mean_token_accuracy": 0.20118670016527176,
      "num_tokens": 37144292.0,
      "step": 16630
    },
    {
      "entropy": 5.390112447738647,
      "epoch": 1.8683102150839557,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004658779276567458,
      "loss": 5.242,
      "mean_token_accuracy": 0.19739776998758315,
      "num_tokens": 37155623.0,
      "step": 16635
    },
    {
      "entropy": 5.542662048339844,
      "epoch": 1.8688717919919133,
      "grad_norm": 1.484375,
      "learning_rate": 0.00046585666821332203,
      "loss": 5.2483,
      "mean_token_accuracy": 0.1941982701420784,
      "num_tokens": 37165406.0,
      "step": 16640
    },
    {
      "entropy": 5.532507038116455,
      "epoch": 1.869433368899871,
      "grad_norm": 1.328125,
      "learning_rate": 0.00046583540269294794,
      "loss": 5.2536,
      "mean_token_accuracy": 0.20177050679922104,
      "num_tokens": 37175522.0,
      "step": 16645
    },
    {
      "entropy": 5.429593420028686,
      "epoch": 1.8699949458078284,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004658141310963006,
      "loss": 5.2844,
      "mean_token_accuracy": 0.19495324939489364,
      "num_tokens": 37186745.0,
      "step": 16650
    },
    {
      "entropy": 5.365559816360474,
      "epoch": 1.8705565227157859,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004657928534240573,
      "loss": 5.1524,
      "mean_token_accuracy": 0.19985704571008683,
      "num_tokens": 37197357.0,
      "step": 16655
    },
    {
      "entropy": 5.484663391113282,
      "epoch": 1.8711180996237435,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004657715696768955,
      "loss": 5.2733,
      "mean_token_accuracy": 0.19708819687366486,
      "num_tokens": 37209839.0,
      "step": 16660
    },
    {
      "entropy": 5.498227214813232,
      "epoch": 1.871679676531701,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000465750279855493,
      "loss": 5.3189,
      "mean_token_accuracy": 0.18525779694318772,
      "num_tokens": 37221314.0,
      "step": 16665
    },
    {
      "entropy": 5.450242328643799,
      "epoch": 1.8722412534396584,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004657289839605276,
      "loss": 5.162,
      "mean_token_accuracy": 0.2004223048686981,
      "num_tokens": 37233372.0,
      "step": 16670
    },
    {
      "entropy": 5.461801481246948,
      "epoch": 1.872802830347616,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004657076819926774,
      "loss": 5.2575,
      "mean_token_accuracy": 0.1905590072274208,
      "num_tokens": 37244758.0,
      "step": 16675
    },
    {
      "entropy": 5.474262952804565,
      "epoch": 1.8733644072555737,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004656863739526206,
      "loss": 5.2199,
      "mean_token_accuracy": 0.1941241666674614,
      "num_tokens": 37255848.0,
      "step": 16680
    },
    {
      "entropy": 5.427585458755493,
      "epoch": 1.8739259841635312,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046566505984103575,
      "loss": 5.1669,
      "mean_token_accuracy": 0.20163838863372802,
      "num_tokens": 37266010.0,
      "step": 16685
    },
    {
      "entropy": 5.311138486862182,
      "epoch": 1.8744875610714886,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004656437396586014,
      "loss": 5.1879,
      "mean_token_accuracy": 0.19384952783584594,
      "num_tokens": 37278445.0,
      "step": 16690
    },
    {
      "entropy": 5.505255031585693,
      "epoch": 1.8750491379794463,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004656224134059964,
      "loss": 5.3141,
      "mean_token_accuracy": 0.1874196335673332,
      "num_tokens": 37291181.0,
      "step": 16695
    },
    {
      "entropy": 5.584687232971191,
      "epoch": 1.875610714887404,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046560108108389985,
      "loss": 5.4112,
      "mean_token_accuracy": 0.18684365302324296,
      "num_tokens": 37302333.0,
      "step": 16700
    },
    {
      "entropy": 5.4045435905456545,
      "epoch": 1.8761722917953614,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000465579742692991,
      "loss": 5.1625,
      "mean_token_accuracy": 0.19999684244394303,
      "num_tokens": 37312754.0,
      "step": 16705
    },
    {
      "entropy": 5.49976544380188,
      "epoch": 1.8767338687033188,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004655583982339492,
      "loss": 5.2433,
      "mean_token_accuracy": 0.2025919422507286,
      "num_tokens": 37323335.0,
      "step": 16710
    },
    {
      "entropy": 5.4840387344360355,
      "epoch": 1.8772954456112765,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004655370477074541,
      "loss": 5.2341,
      "mean_token_accuracy": 0.19807248562574387,
      "num_tokens": 37334332.0,
      "step": 16715
    },
    {
      "entropy": 5.416347360610962,
      "epoch": 1.8778570225192341,
      "grad_norm": 1.125,
      "learning_rate": 0.0004655156911141855,
      "loss": 5.2452,
      "mean_token_accuracy": 0.1973089247941971,
      "num_tokens": 37344770.0,
      "step": 16720
    },
    {
      "entropy": 5.433119297027588,
      "epoch": 1.8784185994271916,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004654943284548234,
      "loss": 5.1702,
      "mean_token_accuracy": 0.19550320208072663,
      "num_tokens": 37356009.0,
      "step": 16725
    },
    {
      "entropy": 5.34617772102356,
      "epoch": 1.878980176335149,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046547295973004807,
      "loss": 5.1594,
      "mean_token_accuracy": 0.20490709245204924,
      "num_tokens": 37367823.0,
      "step": 16730
    },
    {
      "entropy": 5.42624683380127,
      "epoch": 1.8795417532431067,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046545158494053966,
      "loss": 5.1821,
      "mean_token_accuracy": 0.1982982024550438,
      "num_tokens": 37378156.0,
      "step": 16735
    },
    {
      "entropy": 5.495731210708618,
      "epoch": 1.8801033301510643,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046543020408697905,
      "loss": 5.2415,
      "mean_token_accuracy": 0.1919936016201973,
      "num_tokens": 37388378.0,
      "step": 16740
    },
    {
      "entropy": 5.395629119873047,
      "epoch": 1.8806649070590218,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046540881717004685,
      "loss": 5.1395,
      "mean_token_accuracy": 0.19621462374925613,
      "num_tokens": 37398457.0,
      "step": 16745
    },
    {
      "entropy": 5.458871984481812,
      "epoch": 1.8812264839669792,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004653874241904241,
      "loss": 5.3096,
      "mean_token_accuracy": 0.1923062190413475,
      "num_tokens": 37409341.0,
      "step": 16750
    },
    {
      "entropy": 5.418334007263184,
      "epoch": 1.8817880608749369,
      "grad_norm": 1.125,
      "learning_rate": 0.0004653660251487919,
      "loss": 5.1362,
      "mean_token_accuracy": 0.20190441608428955,
      "num_tokens": 37420081.0,
      "step": 16755
    },
    {
      "entropy": 5.40406060218811,
      "epoch": 1.8823496377828943,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046534462004583154,
      "loss": 5.1192,
      "mean_token_accuracy": 0.2129804879426956,
      "num_tokens": 37430707.0,
      "step": 16760
    },
    {
      "entropy": 5.469652366638184,
      "epoch": 1.8829112146908518,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004653232088822248,
      "loss": 5.2351,
      "mean_token_accuracy": 0.19375534653663634,
      "num_tokens": 37441250.0,
      "step": 16765
    },
    {
      "entropy": 5.437140655517578,
      "epoch": 1.8834727915988094,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004653017916586532,
      "loss": 5.2771,
      "mean_token_accuracy": 0.18888781666755677,
      "num_tokens": 37452909.0,
      "step": 16770
    },
    {
      "entropy": 5.497714471817017,
      "epoch": 1.884034368506767,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004652803683757987,
      "loss": 5.2231,
      "mean_token_accuracy": 0.1941472887992859,
      "num_tokens": 37463833.0,
      "step": 16775
    },
    {
      "entropy": 5.42803463935852,
      "epoch": 1.8845959454147245,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004652589390343436,
      "loss": 5.1758,
      "mean_token_accuracy": 0.19998481273651122,
      "num_tokens": 37473967.0,
      "step": 16780
    },
    {
      "entropy": 5.445745277404785,
      "epoch": 1.885157522322682,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046523750363497,
      "loss": 5.1963,
      "mean_token_accuracy": 0.2068118005990982,
      "num_tokens": 37484748.0,
      "step": 16785
    },
    {
      "entropy": 5.398991203308105,
      "epoch": 1.8857190992306396,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004652160621783605,
      "loss": 5.2141,
      "mean_token_accuracy": 0.19281625598669053,
      "num_tokens": 37495757.0,
      "step": 16790
    },
    {
      "entropy": 5.397913455963135,
      "epoch": 1.8862806761385973,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046519461466519785,
      "loss": 5.2379,
      "mean_token_accuracy": 0.18754149675369264,
      "num_tokens": 37506612.0,
      "step": 16795
    },
    {
      "entropy": 5.364954853057862,
      "epoch": 1.8868422530465547,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004651731610961649,
      "loss": 5.1312,
      "mean_token_accuracy": 0.20453477203845977,
      "num_tokens": 37518639.0,
      "step": 16800
    },
    {
      "entropy": 5.432919836044311,
      "epoch": 1.8874038299545122,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004651517014719446,
      "loss": 5.194,
      "mean_token_accuracy": 0.19433584064245224,
      "num_tokens": 37530020.0,
      "step": 16805
    },
    {
      "entropy": 5.391685247421265,
      "epoch": 1.8879654068624698,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004651302357932206,
      "loss": 5.1592,
      "mean_token_accuracy": 0.19571730494499207,
      "num_tokens": 37540737.0,
      "step": 16810
    },
    {
      "entropy": 5.350575733184814,
      "epoch": 1.8885269837704275,
      "grad_norm": 1.125,
      "learning_rate": 0.000465108764060676,
      "loss": 5.1702,
      "mean_token_accuracy": 0.19372555762529373,
      "num_tokens": 37551091.0,
      "step": 16815
    },
    {
      "entropy": 5.45450496673584,
      "epoch": 1.889088560678385,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046508728627499464,
      "loss": 5.2808,
      "mean_token_accuracy": 0.19491389095783235,
      "num_tokens": 37561904.0,
      "step": 16820
    },
    {
      "entropy": 5.52643461227417,
      "epoch": 1.8896501375863424,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004650658024368604,
      "loss": 5.3183,
      "mean_token_accuracy": 0.18911023437976837,
      "num_tokens": 37573532.0,
      "step": 16825
    },
    {
      "entropy": 5.540226650238037,
      "epoch": 1.8902117144943,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004650443125469573,
      "loss": 5.2947,
      "mean_token_accuracy": 0.19715383201837539,
      "num_tokens": 37584886.0,
      "step": 16830
    },
    {
      "entropy": 5.441531181335449,
      "epoch": 1.8907732914022577,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004650228166059695,
      "loss": 5.2351,
      "mean_token_accuracy": 0.19639926999807358,
      "num_tokens": 37595190.0,
      "step": 16835
    },
    {
      "entropy": 5.359212207794189,
      "epoch": 1.8913348683102151,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004650013146145816,
      "loss": 5.1744,
      "mean_token_accuracy": 0.19449964463710784,
      "num_tokens": 37605776.0,
      "step": 16840
    },
    {
      "entropy": 5.502131938934326,
      "epoch": 1.8918964452181726,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004649798065734781,
      "loss": 5.2915,
      "mean_token_accuracy": 0.19160082191228867,
      "num_tokens": 37616882.0,
      "step": 16845
    },
    {
      "entropy": 5.5570464611053465,
      "epoch": 1.8924580221261302,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046495829248334393,
      "loss": 5.3406,
      "mean_token_accuracy": 0.19252358824014665,
      "num_tokens": 37628528.0,
      "step": 16850
    },
    {
      "entropy": 5.5022765636444095,
      "epoch": 1.8930195990340877,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004649367723448641,
      "loss": 5.2515,
      "mean_token_accuracy": 0.19324158877134323,
      "num_tokens": 37640502.0,
      "step": 16855
    },
    {
      "entropy": 5.3829357624053955,
      "epoch": 1.8935811759420451,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046491524615872366,
      "loss": 5.1914,
      "mean_token_accuracy": 0.19664485901594161,
      "num_tokens": 37652845.0,
      "step": 16860
    },
    {
      "entropy": 5.473104047775268,
      "epoch": 1.8941427528500028,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046489371392560816,
      "loss": 5.347,
      "mean_token_accuracy": 0.19013507515192032,
      "num_tokens": 37665418.0,
      "step": 16865
    },
    {
      "entropy": 5.441142511367798,
      "epoch": 1.8947043297579604,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046487217564620317,
      "loss": 5.2686,
      "mean_token_accuracy": 0.19214784651994704,
      "num_tokens": 37676429.0,
      "step": 16870
    },
    {
      "entropy": 5.4757873058319095,
      "epoch": 1.8952659066659179,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004648506313211945,
      "loss": 5.2587,
      "mean_token_accuracy": 0.19430779963731765,
      "num_tokens": 37686779.0,
      "step": 16875
    },
    {
      "entropy": 5.464501619338989,
      "epoch": 1.8958274835738753,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046482908095126803,
      "loss": 5.2955,
      "mean_token_accuracy": 0.1926749438047409,
      "num_tokens": 37697665.0,
      "step": 16880
    },
    {
      "entropy": 5.531804990768433,
      "epoch": 1.896389060481833,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046480752453711,
      "loss": 5.2767,
      "mean_token_accuracy": 0.19173129498958588,
      "num_tokens": 37708987.0,
      "step": 16885
    },
    {
      "entropy": 5.386486911773682,
      "epoch": 1.8969506373897906,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004647859620794068,
      "loss": 5.1199,
      "mean_token_accuracy": 0.20364757776260375,
      "num_tokens": 37720335.0,
      "step": 16890
    },
    {
      "entropy": 5.479136943817139,
      "epoch": 1.897512214297748,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046476439357884497,
      "loss": 5.3053,
      "mean_token_accuracy": 0.1922602817416191,
      "num_tokens": 37731586.0,
      "step": 16895
    },
    {
      "entropy": 5.453959894180298,
      "epoch": 1.8980737912057055,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046474281903611126,
      "loss": 5.2003,
      "mean_token_accuracy": 0.20177598297595978,
      "num_tokens": 37742907.0,
      "step": 16900
    },
    {
      "entropy": 5.504015350341797,
      "epoch": 1.8986353681136632,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046472123845189253,
      "loss": 5.2853,
      "mean_token_accuracy": 0.19715773463249206,
      "num_tokens": 37753787.0,
      "step": 16905
    },
    {
      "entropy": 5.404870223999024,
      "epoch": 1.8991969450216208,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000464699651826876,
      "loss": 5.2013,
      "mean_token_accuracy": 0.19498239010572432,
      "num_tokens": 37766047.0,
      "step": 16910
    },
    {
      "entropy": 5.442248392105102,
      "epoch": 1.8997585219295783,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046467805916174907,
      "loss": 5.1846,
      "mean_token_accuracy": 0.19347760528326036,
      "num_tokens": 37777651.0,
      "step": 16915
    },
    {
      "entropy": 5.375590085983276,
      "epoch": 1.9003200988375357,
      "grad_norm": 1.53125,
      "learning_rate": 0.00046465646045719915,
      "loss": 5.232,
      "mean_token_accuracy": 0.19427147507667542,
      "num_tokens": 37788527.0,
      "step": 16920
    },
    {
      "entropy": 5.366548919677735,
      "epoch": 1.9008816757454934,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004646348557139139,
      "loss": 5.1477,
      "mean_token_accuracy": 0.19297089129686357,
      "num_tokens": 37800727.0,
      "step": 16925
    },
    {
      "entropy": 5.447894477844239,
      "epoch": 1.901443252653451,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046461324493258133,
      "loss": 5.1598,
      "mean_token_accuracy": 0.19702094048261642,
      "num_tokens": 37812332.0,
      "step": 16930
    },
    {
      "entropy": 5.4182517528533936,
      "epoch": 1.9020048295614085,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046459162811388955,
      "loss": 5.2557,
      "mean_token_accuracy": 0.19768508523702621,
      "num_tokens": 37823665.0,
      "step": 16935
    },
    {
      "entropy": 5.359939050674439,
      "epoch": 1.902566406469366,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004645700052585267,
      "loss": 5.0774,
      "mean_token_accuracy": 0.21142406314611434,
      "num_tokens": 37833793.0,
      "step": 16940
    },
    {
      "entropy": 5.369958448410034,
      "epoch": 1.9031279833773236,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004645483763671815,
      "loss": 5.1142,
      "mean_token_accuracy": 0.20716991275548935,
      "num_tokens": 37844104.0,
      "step": 16945
    },
    {
      "entropy": 5.471955299377441,
      "epoch": 1.903689560285281,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046452674144054236,
      "loss": 5.1978,
      "mean_token_accuracy": 0.19857761710882188,
      "num_tokens": 37853920.0,
      "step": 16950
    },
    {
      "entropy": 5.405372571945191,
      "epoch": 1.9042511371932385,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046450510047929826,
      "loss": 5.2591,
      "mean_token_accuracy": 0.19485348463058472,
      "num_tokens": 37865749.0,
      "step": 16955
    },
    {
      "entropy": 5.346922159194946,
      "epoch": 1.9048127141011961,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046448345348413833,
      "loss": 5.2157,
      "mean_token_accuracy": 0.1961336001753807,
      "num_tokens": 37876259.0,
      "step": 16960
    },
    {
      "entropy": 5.490972805023193,
      "epoch": 1.9053742910091538,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004644618004557517,
      "loss": 5.2672,
      "mean_token_accuracy": 0.19992337822914125,
      "num_tokens": 37887186.0,
      "step": 16965
    },
    {
      "entropy": 5.3797101974487305,
      "epoch": 1.9059358679171112,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004644401413948279,
      "loss": 5.0949,
      "mean_token_accuracy": 0.20302311182022095,
      "num_tokens": 37898210.0,
      "step": 16970
    },
    {
      "entropy": 5.361583948135376,
      "epoch": 1.9064974448250687,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004644184763020565,
      "loss": 5.2108,
      "mean_token_accuracy": 0.1912228584289551,
      "num_tokens": 37909332.0,
      "step": 16975
    },
    {
      "entropy": 5.411702394485474,
      "epoch": 1.9070590217330263,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004643968051781273,
      "loss": 5.1765,
      "mean_token_accuracy": 0.19546093195676803,
      "num_tokens": 37919796.0,
      "step": 16980
    },
    {
      "entropy": 5.480267000198364,
      "epoch": 1.907620598640984,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004643751280237305,
      "loss": 5.3321,
      "mean_token_accuracy": 0.1884612888097763,
      "num_tokens": 37931135.0,
      "step": 16985
    },
    {
      "entropy": 5.543706178665161,
      "epoch": 1.9081821755489414,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004643534448395561,
      "loss": 5.304,
      "mean_token_accuracy": 0.19644029587507247,
      "num_tokens": 37942302.0,
      "step": 16990
    },
    {
      "entropy": 5.374786424636841,
      "epoch": 1.9087437524568989,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004643317556262946,
      "loss": 5.2274,
      "mean_token_accuracy": 0.19282005876302719,
      "num_tokens": 37954188.0,
      "step": 16995
    },
    {
      "entropy": 5.419363737106323,
      "epoch": 1.9093053293648565,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046431006038463653,
      "loss": 5.255,
      "mean_token_accuracy": 0.1871391788125038,
      "num_tokens": 37964778.0,
      "step": 17000
    },
    {
      "entropy": 5.537145471572876,
      "epoch": 1.9098669062728142,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004642883591152727,
      "loss": 5.2713,
      "mean_token_accuracy": 0.18857947289943694,
      "num_tokens": 37976384.0,
      "step": 17005
    },
    {
      "entropy": 5.498571109771729,
      "epoch": 1.9104284831807716,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004642666518188942,
      "loss": 5.2601,
      "mean_token_accuracy": 0.18965450078248977,
      "num_tokens": 37986975.0,
      "step": 17010
    },
    {
      "entropy": 5.563908243179322,
      "epoch": 1.910990060088729,
      "grad_norm": 1.125,
      "learning_rate": 0.00046424493849619205,
      "loss": 5.37,
      "mean_token_accuracy": 0.19417090862989425,
      "num_tokens": 37999255.0,
      "step": 17015
    },
    {
      "entropy": 5.52285623550415,
      "epoch": 1.9115516369966867,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046422321914785757,
      "loss": 5.3204,
      "mean_token_accuracy": 0.1885424554347992,
      "num_tokens": 38010436.0,
      "step": 17020
    },
    {
      "entropy": 5.312445592880249,
      "epoch": 1.9121132139046444,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046420149377458255,
      "loss": 5.1274,
      "mean_token_accuracy": 0.2034679740667343,
      "num_tokens": 38021537.0,
      "step": 17025
    },
    {
      "entropy": 5.4133302688598635,
      "epoch": 1.9126747908126018,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046417976237705846,
      "loss": 5.2043,
      "mean_token_accuracy": 0.19637694358825683,
      "num_tokens": 38033396.0,
      "step": 17030
    },
    {
      "entropy": 5.437874507904053,
      "epoch": 1.9132363677205593,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046415802495597744,
      "loss": 5.1219,
      "mean_token_accuracy": 0.2110572040081024,
      "num_tokens": 38044917.0,
      "step": 17035
    },
    {
      "entropy": 5.389221239089966,
      "epoch": 1.913797944628517,
      "grad_norm": 1.125,
      "learning_rate": 0.00046413628151203154,
      "loss": 5.2819,
      "mean_token_accuracy": 0.20049762427806855,
      "num_tokens": 38057002.0,
      "step": 17040
    },
    {
      "entropy": 5.302631282806397,
      "epoch": 1.9143595215364744,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046411453204591303,
      "loss": 5.0597,
      "mean_token_accuracy": 0.20481076687574387,
      "num_tokens": 38068023.0,
      "step": 17045
    },
    {
      "entropy": 5.531670236587525,
      "epoch": 1.9149210984444318,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046409277655831455,
      "loss": 5.3441,
      "mean_token_accuracy": 0.1887894317507744,
      "num_tokens": 38080137.0,
      "step": 17050
    },
    {
      "entropy": 5.508067512512207,
      "epoch": 1.9154826753523895,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004640710150499287,
      "loss": 5.1671,
      "mean_token_accuracy": 0.20041078478097915,
      "num_tokens": 38090078.0,
      "step": 17055
    },
    {
      "entropy": 5.4494102001190186,
      "epoch": 1.9160442522603471,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046404924752144834,
      "loss": 5.2901,
      "mean_token_accuracy": 0.1950506255030632,
      "num_tokens": 38101410.0,
      "step": 17060
    },
    {
      "entropy": 5.414707040786743,
      "epoch": 1.9166058291683046,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046402747397356666,
      "loss": 5.1806,
      "mean_token_accuracy": 0.20311455875635148,
      "num_tokens": 38111610.0,
      "step": 17065
    },
    {
      "entropy": 5.406502342224121,
      "epoch": 1.917167406076262,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004640056944069768,
      "loss": 5.1994,
      "mean_token_accuracy": 0.1951902523636818,
      "num_tokens": 38122947.0,
      "step": 17070
    },
    {
      "entropy": 5.359834289550781,
      "epoch": 1.9177289829842197,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004639839088223725,
      "loss": 5.1625,
      "mean_token_accuracy": 0.20344320982694625,
      "num_tokens": 38133083.0,
      "step": 17075
    },
    {
      "entropy": 5.445795679092408,
      "epoch": 1.9182905598921773,
      "grad_norm": 1.125,
      "learning_rate": 0.00046396211722044716,
      "loss": 5.227,
      "mean_token_accuracy": 0.2015004947781563,
      "num_tokens": 38142941.0,
      "step": 17080
    },
    {
      "entropy": 5.500422859191895,
      "epoch": 1.9188521368001348,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046394031960189473,
      "loss": 5.2278,
      "mean_token_accuracy": 0.19490229189395905,
      "num_tokens": 38154269.0,
      "step": 17085
    },
    {
      "entropy": 5.385607719421387,
      "epoch": 1.9194137137080922,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004639185159674093,
      "loss": 5.1621,
      "mean_token_accuracy": 0.19849398881196975,
      "num_tokens": 38164153.0,
      "step": 17090
    },
    {
      "entropy": 5.431947898864746,
      "epoch": 1.9199752906160499,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046389670631768494,
      "loss": 5.3086,
      "mean_token_accuracy": 0.19017583578824998,
      "num_tokens": 38174704.0,
      "step": 17095
    },
    {
      "entropy": 5.576982879638672,
      "epoch": 1.9205368675240075,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004638748906534162,
      "loss": 5.322,
      "mean_token_accuracy": 0.19010902047157288,
      "num_tokens": 38186292.0,
      "step": 17100
    },
    {
      "entropy": 5.4989039421081545,
      "epoch": 1.921098444431965,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004638530689752977,
      "loss": 5.2645,
      "mean_token_accuracy": 0.19639619290828705,
      "num_tokens": 38197419.0,
      "step": 17105
    },
    {
      "entropy": 5.4381475925445555,
      "epoch": 1.9216600213399224,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046383124128402433,
      "loss": 5.2469,
      "mean_token_accuracy": 0.1944597139954567,
      "num_tokens": 38208276.0,
      "step": 17110
    },
    {
      "entropy": 5.352316665649414,
      "epoch": 1.92222159824788,
      "grad_norm": 1.125,
      "learning_rate": 0.00046380940758029095,
      "loss": 5.1557,
      "mean_token_accuracy": 0.20276721715927123,
      "num_tokens": 38220276.0,
      "step": 17115
    },
    {
      "entropy": 5.407376623153686,
      "epoch": 1.9227831751558377,
      "grad_norm": 1.125,
      "learning_rate": 0.00046378756786479274,
      "loss": 5.2615,
      "mean_token_accuracy": 0.1980969026684761,
      "num_tokens": 38230760.0,
      "step": 17120
    },
    {
      "entropy": 5.385149240493774,
      "epoch": 1.9233447520637952,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004637657221382252,
      "loss": 5.1431,
      "mean_token_accuracy": 0.2013310343027115,
      "num_tokens": 38241769.0,
      "step": 17125
    },
    {
      "entropy": 5.480165290832519,
      "epoch": 1.9239063289717526,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046374387040128384,
      "loss": 5.2418,
      "mean_token_accuracy": 0.1965601250529289,
      "num_tokens": 38252800.0,
      "step": 17130
    },
    {
      "entropy": 5.356214761734009,
      "epoch": 1.9244679058797103,
      "grad_norm": 1.25,
      "learning_rate": 0.00046372201265466444,
      "loss": 5.1212,
      "mean_token_accuracy": 0.20400320440530778,
      "num_tokens": 38264319.0,
      "step": 17135
    },
    {
      "entropy": 5.375586986541748,
      "epoch": 1.9250294827876677,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000463700148899063,
      "loss": 5.1994,
      "mean_token_accuracy": 0.20054273009300233,
      "num_tokens": 38275118.0,
      "step": 17140
    },
    {
      "entropy": 5.516795110702515,
      "epoch": 1.9255910596956252,
      "grad_norm": 1.125,
      "learning_rate": 0.0004636782791351756,
      "loss": 5.3481,
      "mean_token_accuracy": 0.19195895791053771,
      "num_tokens": 38286264.0,
      "step": 17145
    },
    {
      "entropy": 5.444025135040283,
      "epoch": 1.9261526366035828,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004636564033636987,
      "loss": 5.1447,
      "mean_token_accuracy": 0.20469896644353866,
      "num_tokens": 38297603.0,
      "step": 17150
    },
    {
      "entropy": 5.443168306350708,
      "epoch": 1.9267142135115405,
      "grad_norm": 1.125,
      "learning_rate": 0.00046363452158532857,
      "loss": 5.2887,
      "mean_token_accuracy": 0.1880109876394272,
      "num_tokens": 38308760.0,
      "step": 17155
    },
    {
      "entropy": 5.497552680969238,
      "epoch": 1.927275790419498,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004636126338007622,
      "loss": 5.2541,
      "mean_token_accuracy": 0.19806658029556273,
      "num_tokens": 38320807.0,
      "step": 17160
    },
    {
      "entropy": 5.397411108016968,
      "epoch": 1.9278373673274554,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004635907400106965,
      "loss": 5.2248,
      "mean_token_accuracy": 0.1977235868573189,
      "num_tokens": 38331087.0,
      "step": 17165
    },
    {
      "entropy": 5.385605812072754,
      "epoch": 1.928398944235413,
      "grad_norm": 1.125,
      "learning_rate": 0.0004635688402158285,
      "loss": 5.1844,
      "mean_token_accuracy": 0.20038879215717315,
      "num_tokens": 38341659.0,
      "step": 17170
    },
    {
      "entropy": 5.385630321502686,
      "epoch": 1.9289605211433707,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046354693441685537,
      "loss": 5.1475,
      "mean_token_accuracy": 0.20378071069717407,
      "num_tokens": 38351371.0,
      "step": 17175
    },
    {
      "entropy": 5.35546727180481,
      "epoch": 1.9295220980513281,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046352502261447475,
      "loss": 5.1557,
      "mean_token_accuracy": 0.2089747130870819,
      "num_tokens": 38362677.0,
      "step": 17180
    },
    {
      "entropy": 5.405532789230347,
      "epoch": 1.9300836749592856,
      "grad_norm": 1.375,
      "learning_rate": 0.0004635031048093843,
      "loss": 5.1861,
      "mean_token_accuracy": 0.20287054479122163,
      "num_tokens": 38372882.0,
      "step": 17185
    },
    {
      "entropy": 5.512738656997681,
      "epoch": 1.9306452518672432,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004634811810022819,
      "loss": 5.2851,
      "mean_token_accuracy": 0.19239278733730317,
      "num_tokens": 38383401.0,
      "step": 17190
    },
    {
      "entropy": 5.4630577087402346,
      "epoch": 1.931206828775201,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046345925119386556,
      "loss": 5.3341,
      "mean_token_accuracy": 0.1909775733947754,
      "num_tokens": 38394733.0,
      "step": 17195
    },
    {
      "entropy": 5.524986124038696,
      "epoch": 1.9317684056831583,
      "grad_norm": 1.125,
      "learning_rate": 0.0004634373153848336,
      "loss": 5.2805,
      "mean_token_accuracy": 0.19990334361791612,
      "num_tokens": 38406482.0,
      "step": 17200
    },
    {
      "entropy": 5.472387838363647,
      "epoch": 1.9323299825911158,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004634153735758845,
      "loss": 5.2021,
      "mean_token_accuracy": 0.19552544206380845,
      "num_tokens": 38417366.0,
      "step": 17205
    },
    {
      "entropy": 5.402789783477783,
      "epoch": 1.9328915594990734,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004633934257677167,
      "loss": 5.2709,
      "mean_token_accuracy": 0.2022368237376213,
      "num_tokens": 38428440.0,
      "step": 17210
    },
    {
      "entropy": 5.450861263275146,
      "epoch": 1.933453136407031,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004633714719610292,
      "loss": 5.2191,
      "mean_token_accuracy": 0.1953435242176056,
      "num_tokens": 38438471.0,
      "step": 17215
    },
    {
      "entropy": 5.480303859710693,
      "epoch": 1.9340147133149885,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046334951215652097,
      "loss": 5.1922,
      "mean_token_accuracy": 0.20012564659118653,
      "num_tokens": 38448990.0,
      "step": 17220
    },
    {
      "entropy": 5.50822606086731,
      "epoch": 1.934576290222946,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046332754635489126,
      "loss": 5.2662,
      "mean_token_accuracy": 0.19486362636089324,
      "num_tokens": 38460990.0,
      "step": 17225
    },
    {
      "entropy": 5.478990840911865,
      "epoch": 1.9351378671309036,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004633055745568394,
      "loss": 5.2882,
      "mean_token_accuracy": 0.18443589955568312,
      "num_tokens": 38471672.0,
      "step": 17230
    },
    {
      "entropy": 5.433164072036743,
      "epoch": 1.935699444038861,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046328359676306506,
      "loss": 5.1622,
      "mean_token_accuracy": 0.20360776036977768,
      "num_tokens": 38483248.0,
      "step": 17235
    },
    {
      "entropy": 5.452275657653809,
      "epoch": 1.9362610209468185,
      "grad_norm": 1.03125,
      "learning_rate": 0.000463261612974268,
      "loss": 5.2054,
      "mean_token_accuracy": 0.20055924952030182,
      "num_tokens": 38494795.0,
      "step": 17240
    },
    {
      "entropy": 5.353795671463013,
      "epoch": 1.9368225978547762,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004632396231911481,
      "loss": 5.2114,
      "mean_token_accuracy": 0.1964734226465225,
      "num_tokens": 38505276.0,
      "step": 17245
    },
    {
      "entropy": 5.5354588508605955,
      "epoch": 1.9373841747627338,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004632176274144056,
      "loss": 5.2579,
      "mean_token_accuracy": 0.1912876084446907,
      "num_tokens": 38515536.0,
      "step": 17250
    },
    {
      "entropy": 5.432906532287598,
      "epoch": 1.9379457516706913,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046319562564474083,
      "loss": 5.2339,
      "mean_token_accuracy": 0.19609032571315765,
      "num_tokens": 38525579.0,
      "step": 17255
    },
    {
      "entropy": 5.510797357559204,
      "epoch": 1.9385073285786487,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046317361788285436,
      "loss": 5.3612,
      "mean_token_accuracy": 0.1941184103488922,
      "num_tokens": 38536963.0,
      "step": 17260
    },
    {
      "entropy": 5.407287979125977,
      "epoch": 1.9390689054866064,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004631516041294469,
      "loss": 5.2779,
      "mean_token_accuracy": 0.18720667511224748,
      "num_tokens": 38547497.0,
      "step": 17265
    },
    {
      "entropy": 5.487315654754639,
      "epoch": 1.939630482394564,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004631295843852194,
      "loss": 5.1683,
      "mean_token_accuracy": 0.19586486518383026,
      "num_tokens": 38558840.0,
      "step": 17270
    },
    {
      "entropy": 5.463495683670044,
      "epoch": 1.9401920593025215,
      "grad_norm": 1.15625,
      "learning_rate": 0.000463107558650873,
      "loss": 5.1553,
      "mean_token_accuracy": 0.20114171802997588,
      "num_tokens": 38568670.0,
      "step": 17275
    },
    {
      "entropy": 5.441161727905273,
      "epoch": 1.940753636210479,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046308552692710895,
      "loss": 5.2052,
      "mean_token_accuracy": 0.20004095733165742,
      "num_tokens": 38578565.0,
      "step": 17280
    },
    {
      "entropy": 5.365658807754516,
      "epoch": 1.9413152131184366,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046306348921462876,
      "loss": 5.179,
      "mean_token_accuracy": 0.1935104951262474,
      "num_tokens": 38589126.0,
      "step": 17285
    },
    {
      "entropy": 5.449799823760986,
      "epoch": 1.9418767900263942,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004630414455141342,
      "loss": 5.296,
      "mean_token_accuracy": 0.1890251263976097,
      "num_tokens": 38601008.0,
      "step": 17290
    },
    {
      "entropy": 5.485762119293213,
      "epoch": 1.9424383669343517,
      "grad_norm": 1.09375,
      "learning_rate": 0.000463019395826327,
      "loss": 5.2727,
      "mean_token_accuracy": 0.19694801419973373,
      "num_tokens": 38611756.0,
      "step": 17295
    },
    {
      "entropy": 5.4574767589569095,
      "epoch": 1.9429999438423091,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046299734015190934,
      "loss": 5.1786,
      "mean_token_accuracy": 0.19253744035959244,
      "num_tokens": 38623156.0,
      "step": 17300
    },
    {
      "entropy": 5.501694869995117,
      "epoch": 1.9435615207502668,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046297527849158344,
      "loss": 5.2527,
      "mean_token_accuracy": 0.19997863471508026,
      "num_tokens": 38633920.0,
      "step": 17305
    },
    {
      "entropy": 5.436491012573242,
      "epoch": 1.9441230976582244,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004629532108460519,
      "loss": 5.1646,
      "mean_token_accuracy": 0.19582826793193817,
      "num_tokens": 38644744.0,
      "step": 17310
    },
    {
      "entropy": 5.386834716796875,
      "epoch": 1.9446846745661819,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046293113721601706,
      "loss": 5.1706,
      "mean_token_accuracy": 0.1978539764881134,
      "num_tokens": 38656524.0,
      "step": 17315
    },
    {
      "entropy": 5.389536142349243,
      "epoch": 1.9452462514741393,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046290905760218193,
      "loss": 5.2268,
      "mean_token_accuracy": 0.20254576057195664,
      "num_tokens": 38668386.0,
      "step": 17320
    },
    {
      "entropy": 5.385469627380371,
      "epoch": 1.945807828382097,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046288697200524964,
      "loss": 5.135,
      "mean_token_accuracy": 0.19812023937702178,
      "num_tokens": 38678498.0,
      "step": 17325
    },
    {
      "entropy": 5.550500106811524,
      "epoch": 1.9463694052900544,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046286488042592325,
      "loss": 5.3627,
      "mean_token_accuracy": 0.18431078642606735,
      "num_tokens": 38690084.0,
      "step": 17330
    },
    {
      "entropy": 5.40110764503479,
      "epoch": 1.9469309821980119,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004628427828649062,
      "loss": 5.2231,
      "mean_token_accuracy": 0.19438739567995073,
      "num_tokens": 38701282.0,
      "step": 17335
    },
    {
      "entropy": 5.366098928451538,
      "epoch": 1.9474925591059695,
      "grad_norm": 1.125,
      "learning_rate": 0.00046282067932290204,
      "loss": 5.0904,
      "mean_token_accuracy": 0.19872173368930818,
      "num_tokens": 38713012.0,
      "step": 17340
    },
    {
      "entropy": 5.361801433563232,
      "epoch": 1.9480541360139272,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004627985698006146,
      "loss": 5.1482,
      "mean_token_accuracy": 0.20405038595199584,
      "num_tokens": 38723179.0,
      "step": 17345
    },
    {
      "entropy": 5.458559179306031,
      "epoch": 1.9486157129218846,
      "grad_norm": 1.140625,
      "learning_rate": 0.000462776454298748,
      "loss": 5.1916,
      "mean_token_accuracy": 0.19987673163414002,
      "num_tokens": 38733729.0,
      "step": 17350
    },
    {
      "entropy": 5.4300459861755375,
      "epoch": 1.949177289829842,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046275433281800613,
      "loss": 5.1458,
      "mean_token_accuracy": 0.2052153766155243,
      "num_tokens": 38743139.0,
      "step": 17355
    },
    {
      "entropy": 5.437872123718262,
      "epoch": 1.9497388667377997,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004627322053590935,
      "loss": 5.2134,
      "mean_token_accuracy": 0.18721825033426284,
      "num_tokens": 38754697.0,
      "step": 17360
    },
    {
      "entropy": 5.3867326259613035,
      "epoch": 1.9503004436457574,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004627100719227147,
      "loss": 5.2533,
      "mean_token_accuracy": 0.1948298454284668,
      "num_tokens": 38765999.0,
      "step": 17365
    },
    {
      "entropy": 5.471641921997071,
      "epoch": 1.9508620205537148,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046268793250957437,
      "loss": 5.2587,
      "mean_token_accuracy": 0.1956437960267067,
      "num_tokens": 38776912.0,
      "step": 17370
    },
    {
      "entropy": 5.512712907791138,
      "epoch": 1.9514235974616723,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046266578712037747,
      "loss": 5.2458,
      "mean_token_accuracy": 0.19348823577165603,
      "num_tokens": 38787296.0,
      "step": 17375
    },
    {
      "entropy": 5.445453071594239,
      "epoch": 1.95198517436963,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046264363575582916,
      "loss": 5.2521,
      "mean_token_accuracy": 0.18699941039085388,
      "num_tokens": 38798013.0,
      "step": 17380
    },
    {
      "entropy": 5.400072813034058,
      "epoch": 1.9525467512775876,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004626214784166347,
      "loss": 5.1873,
      "mean_token_accuracy": 0.19722518026828767,
      "num_tokens": 38808842.0,
      "step": 17385
    },
    {
      "entropy": 5.449928092956543,
      "epoch": 1.953108328185545,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004625993151034995,
      "loss": 5.1238,
      "mean_token_accuracy": 0.2016418382525444,
      "num_tokens": 38819094.0,
      "step": 17390
    },
    {
      "entropy": 5.4233910083770756,
      "epoch": 1.9536699050935025,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004625771458171294,
      "loss": 5.2155,
      "mean_token_accuracy": 0.19327747076749802,
      "num_tokens": 38830348.0,
      "step": 17395
    },
    {
      "entropy": 5.46439790725708,
      "epoch": 1.9542314820014601,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046255497055823025,
      "loss": 5.2338,
      "mean_token_accuracy": 0.1972393274307251,
      "num_tokens": 38841927.0,
      "step": 17400
    },
    {
      "entropy": 5.308208560943603,
      "epoch": 1.9547930589094178,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004625327893275082,
      "loss": 5.039,
      "mean_token_accuracy": 0.20316302478313447,
      "num_tokens": 38852817.0,
      "step": 17405
    },
    {
      "entropy": 5.283917808532715,
      "epoch": 1.9553546358173752,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046251060212566927,
      "loss": 5.166,
      "mean_token_accuracy": 0.19269182085990905,
      "num_tokens": 38864181.0,
      "step": 17410
    },
    {
      "entropy": 5.430478906631469,
      "epoch": 1.9559162127253327,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004624884089534201,
      "loss": 5.2082,
      "mean_token_accuracy": 0.1992849037051201,
      "num_tokens": 38873827.0,
      "step": 17415
    },
    {
      "entropy": 5.470493030548096,
      "epoch": 1.9564777896332903,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004624662098114672,
      "loss": 5.2284,
      "mean_token_accuracy": 0.19971708357334136,
      "num_tokens": 38884540.0,
      "step": 17420
    },
    {
      "entropy": 5.336352872848511,
      "epoch": 1.9570393665412478,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046244400470051754,
      "loss": 5.0661,
      "mean_token_accuracy": 0.20701975822448732,
      "num_tokens": 38895600.0,
      "step": 17425
    },
    {
      "entropy": 5.390346908569336,
      "epoch": 1.9576009434492052,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000462421793621278,
      "loss": 5.1567,
      "mean_token_accuracy": 0.20227255523204804,
      "num_tokens": 38906166.0,
      "step": 17430
    },
    {
      "entropy": 5.316209125518799,
      "epoch": 1.9581625203571629,
      "grad_norm": 1.15625,
      "learning_rate": 0.000462399576574456,
      "loss": 5.0675,
      "mean_token_accuracy": 0.20503122061491014,
      "num_tokens": 38916247.0,
      "step": 17435
    },
    {
      "entropy": 5.454859828948974,
      "epoch": 1.9587240972651205,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004623773535607587,
      "loss": 5.2832,
      "mean_token_accuracy": 0.19425395280122756,
      "num_tokens": 38928276.0,
      "step": 17440
    },
    {
      "entropy": 5.4641804695129395,
      "epoch": 1.959285674173078,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004623551245808938,
      "loss": 5.1962,
      "mean_token_accuracy": 0.20096850842237474,
      "num_tokens": 38939460.0,
      "step": 17445
    },
    {
      "entropy": 5.396817350387574,
      "epoch": 1.9598472510810354,
      "grad_norm": 2.234375,
      "learning_rate": 0.0004623328896355691,
      "loss": 5.1473,
      "mean_token_accuracy": 0.1969576045870781,
      "num_tokens": 38950381.0,
      "step": 17450
    },
    {
      "entropy": 5.347562074661255,
      "epoch": 1.960408827988993,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046231064872549244,
      "loss": 5.146,
      "mean_token_accuracy": 0.20016928762197495,
      "num_tokens": 38961312.0,
      "step": 17455
    },
    {
      "entropy": 5.414300107955933,
      "epoch": 1.9609704048969507,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046228840185137215,
      "loss": 5.2327,
      "mean_token_accuracy": 0.19685864448547363,
      "num_tokens": 38974030.0,
      "step": 17460
    },
    {
      "entropy": 5.488665294647217,
      "epoch": 1.9615319818049082,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004622661490139165,
      "loss": 5.278,
      "mean_token_accuracy": 0.19315640330314637,
      "num_tokens": 38984640.0,
      "step": 17465
    },
    {
      "entropy": 5.352644443511963,
      "epoch": 1.9620935587128656,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004622438902138341,
      "loss": 5.1991,
      "mean_token_accuracy": 0.19610744714736938,
      "num_tokens": 38996206.0,
      "step": 17470
    },
    {
      "entropy": 5.346266984939575,
      "epoch": 1.9626551356208233,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046222162545183357,
      "loss": 5.0995,
      "mean_token_accuracy": 0.21035393476486205,
      "num_tokens": 39007765.0,
      "step": 17475
    },
    {
      "entropy": 5.4519562244415285,
      "epoch": 1.963216712528781,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046219935472862386,
      "loss": 5.2307,
      "mean_token_accuracy": 0.1965882495045662,
      "num_tokens": 39020691.0,
      "step": 17480
    },
    {
      "entropy": 5.3657083988189695,
      "epoch": 1.9637782894367384,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046217707804491404,
      "loss": 5.1738,
      "mean_token_accuracy": 0.20511361360549926,
      "num_tokens": 39032003.0,
      "step": 17485
    },
    {
      "entropy": 5.434570217132569,
      "epoch": 1.9643398663446958,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004621547954014135,
      "loss": 5.2438,
      "mean_token_accuracy": 0.19939392507076265,
      "num_tokens": 39042783.0,
      "step": 17490
    },
    {
      "entropy": 5.413248014450073,
      "epoch": 1.9649014432526535,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004621325067988317,
      "loss": 5.1432,
      "mean_token_accuracy": 0.20081115514039993,
      "num_tokens": 39054395.0,
      "step": 17495
    },
    {
      "entropy": 5.351507759094238,
      "epoch": 1.9654630201606111,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046211021223787827,
      "loss": 5.1158,
      "mean_token_accuracy": 0.21153762191534042,
      "num_tokens": 39064619.0,
      "step": 17500
    },
    {
      "entropy": 5.489558696746826,
      "epoch": 1.9660245970685686,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046208791171926315,
      "loss": 5.2713,
      "mean_token_accuracy": 0.19293173402547836,
      "num_tokens": 39076112.0,
      "step": 17505
    },
    {
      "entropy": 5.429635334014892,
      "epoch": 1.966586173976526,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046206560524369636,
      "loss": 5.1698,
      "mean_token_accuracy": 0.19933320730924606,
      "num_tokens": 39087669.0,
      "step": 17510
    },
    {
      "entropy": 5.418504095077514,
      "epoch": 1.9671477508844837,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046204329281188816,
      "loss": 5.1905,
      "mean_token_accuracy": 0.1965690553188324,
      "num_tokens": 39098506.0,
      "step": 17515
    },
    {
      "entropy": 5.409231090545655,
      "epoch": 1.9677093277924411,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004620209744245489,
      "loss": 5.2023,
      "mean_token_accuracy": 0.205171200633049,
      "num_tokens": 39109441.0,
      "step": 17520
    },
    {
      "entropy": 5.39208664894104,
      "epoch": 1.9682709047003986,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046199865008238937,
      "loss": 5.133,
      "mean_token_accuracy": 0.2046850100159645,
      "num_tokens": 39119765.0,
      "step": 17525
    },
    {
      "entropy": 5.317610836029052,
      "epoch": 1.9688324816083562,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004619763197861203,
      "loss": 5.0893,
      "mean_token_accuracy": 0.20616369247436522,
      "num_tokens": 39129794.0,
      "step": 17530
    },
    {
      "entropy": 5.360634136199951,
      "epoch": 1.9693940585163139,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046195398353645256,
      "loss": 5.1215,
      "mean_token_accuracy": 0.20423976182937623,
      "num_tokens": 39140095.0,
      "step": 17535
    },
    {
      "entropy": 5.448483896255493,
      "epoch": 1.9699556354242713,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004619316413340976,
      "loss": 5.2217,
      "mean_token_accuracy": 0.19715238362550735,
      "num_tokens": 39150855.0,
      "step": 17540
    },
    {
      "entropy": 5.418335962295532,
      "epoch": 1.9705172123322288,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046190929317976664,
      "loss": 5.137,
      "mean_token_accuracy": 0.20449523329734803,
      "num_tokens": 39161455.0,
      "step": 17545
    },
    {
      "entropy": 5.435212278366089,
      "epoch": 1.9710787892401864,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004618869390741713,
      "loss": 5.3339,
      "mean_token_accuracy": 0.19435828626155854,
      "num_tokens": 39173199.0,
      "step": 17550
    },
    {
      "entropy": 5.483836317062378,
      "epoch": 1.971640366148144,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004618645790180233,
      "loss": 5.2275,
      "mean_token_accuracy": 0.20099535286426545,
      "num_tokens": 39183145.0,
      "step": 17555
    },
    {
      "entropy": 5.459251117706299,
      "epoch": 1.9722019430561015,
      "grad_norm": 1.125,
      "learning_rate": 0.0004618422130120347,
      "loss": 5.2206,
      "mean_token_accuracy": 0.19523046016693116,
      "num_tokens": 39194900.0,
      "step": 17560
    },
    {
      "entropy": 5.474074029922486,
      "epoch": 1.972763519964059,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004618198410569175,
      "loss": 5.2843,
      "mean_token_accuracy": 0.19154127240180968,
      "num_tokens": 39206651.0,
      "step": 17565
    },
    {
      "entropy": 5.393407726287842,
      "epoch": 1.9733250968720166,
      "grad_norm": 1.125,
      "learning_rate": 0.00046179746315338414,
      "loss": 5.1813,
      "mean_token_accuracy": 0.19838632345199586,
      "num_tokens": 39217635.0,
      "step": 17570
    },
    {
      "entropy": 5.436306142807007,
      "epoch": 1.9738866737799743,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046177507930214714,
      "loss": 5.2052,
      "mean_token_accuracy": 0.20139442533254623,
      "num_tokens": 39227937.0,
      "step": 17575
    },
    {
      "entropy": 5.422332096099853,
      "epoch": 1.9744482506879317,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004617526895039191,
      "loss": 5.2009,
      "mean_token_accuracy": 0.20861320048570633,
      "num_tokens": 39238776.0,
      "step": 17580
    },
    {
      "entropy": 5.418473768234253,
      "epoch": 1.9750098275958892,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046173029375941305,
      "loss": 5.2292,
      "mean_token_accuracy": 0.1956673964858055,
      "num_tokens": 39250043.0,
      "step": 17585
    },
    {
      "entropy": 5.372447204589844,
      "epoch": 1.9755714045038468,
      "grad_norm": 1.109375,
      "learning_rate": 0.000461707892069342,
      "loss": 5.1753,
      "mean_token_accuracy": 0.20192056745290757,
      "num_tokens": 39261440.0,
      "step": 17590
    },
    {
      "entropy": 5.5058777809143065,
      "epoch": 1.9761329814118045,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046168548443441926,
      "loss": 5.3821,
      "mean_token_accuracy": 0.19442095905542373,
      "num_tokens": 39272151.0,
      "step": 17595
    },
    {
      "entropy": 5.559530305862427,
      "epoch": 1.976694558319762,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004616630708553583,
      "loss": 5.3532,
      "mean_token_accuracy": 0.18842432051897048,
      "num_tokens": 39284794.0,
      "step": 17600
    },
    {
      "entropy": 5.398608875274658,
      "epoch": 1.9772561352277194,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046164065133287276,
      "loss": 5.2847,
      "mean_token_accuracy": 0.19264792650938034,
      "num_tokens": 39296660.0,
      "step": 17605
    },
    {
      "entropy": 5.49310998916626,
      "epoch": 1.977817712135677,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004616182258676765,
      "loss": 5.2855,
      "mean_token_accuracy": 0.19437632262706755,
      "num_tokens": 39308609.0,
      "step": 17610
    },
    {
      "entropy": 5.409449815750122,
      "epoch": 1.9783792890436345,
      "grad_norm": 6.59375,
      "learning_rate": 0.0004615957944604835,
      "loss": 5.0926,
      "mean_token_accuracy": 0.20506742745637893,
      "num_tokens": 39319295.0,
      "step": 17615
    },
    {
      "entropy": 5.406455993652344,
      "epoch": 1.978940865951592,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004615733571120081,
      "loss": 5.1923,
      "mean_token_accuracy": 0.2004172086715698,
      "num_tokens": 39330874.0,
      "step": 17620
    },
    {
      "entropy": 5.3828857421875,
      "epoch": 1.9795024428595496,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004615509138229646,
      "loss": 5.2091,
      "mean_token_accuracy": 0.19369926750659944,
      "num_tokens": 39343574.0,
      "step": 17625
    },
    {
      "entropy": 5.4363292217254635,
      "epoch": 1.9800640197675072,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046152846459406765,
      "loss": 5.2185,
      "mean_token_accuracy": 0.19346231520175933,
      "num_tokens": 39355761.0,
      "step": 17630
    },
    {
      "entropy": 5.47263650894165,
      "epoch": 1.9806255966754647,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004615060094260321,
      "loss": 5.3388,
      "mean_token_accuracy": 0.18521904498338698,
      "num_tokens": 39368116.0,
      "step": 17635
    },
    {
      "entropy": 5.442290496826172,
      "epoch": 1.9811871735834221,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046148354831957285,
      "loss": 5.1926,
      "mean_token_accuracy": 0.20586927682161332,
      "num_tokens": 39378619.0,
      "step": 17640
    },
    {
      "entropy": 5.536953639984131,
      "epoch": 1.9817487504913798,
      "grad_norm": 1.125,
      "learning_rate": 0.00046146108127540506,
      "loss": 5.314,
      "mean_token_accuracy": 0.18941621780395507,
      "num_tokens": 39391595.0,
      "step": 17645
    },
    {
      "entropy": 5.496724367141724,
      "epoch": 1.9823103273993374,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004614386082942442,
      "loss": 5.2867,
      "mean_token_accuracy": 0.18969347327947617,
      "num_tokens": 39402494.0,
      "step": 17650
    },
    {
      "entropy": 5.370987606048584,
      "epoch": 1.9828719043072949,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004614161293768057,
      "loss": 5.1944,
      "mean_token_accuracy": 0.19882066249847413,
      "num_tokens": 39413371.0,
      "step": 17655
    },
    {
      "entropy": 5.389461326599121,
      "epoch": 1.9834334812152523,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046139364452380536,
      "loss": 5.2001,
      "mean_token_accuracy": 0.19597023576498032,
      "num_tokens": 39424170.0,
      "step": 17660
    },
    {
      "entropy": 5.540799951553344,
      "epoch": 1.98399505812321,
      "grad_norm": 1.140625,
      "learning_rate": 0.000461371153735959,
      "loss": 5.2697,
      "mean_token_accuracy": 0.20007153898477553,
      "num_tokens": 39434438.0,
      "step": 17665
    },
    {
      "entropy": 5.527892112731934,
      "epoch": 1.9845566350311676,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046134865701398287,
      "loss": 5.2148,
      "mean_token_accuracy": 0.19637876898050308,
      "num_tokens": 39445116.0,
      "step": 17670
    },
    {
      "entropy": 5.449889421463013,
      "epoch": 1.985118211939125,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046132615435859326,
      "loss": 5.1892,
      "mean_token_accuracy": 0.1988799288868904,
      "num_tokens": 39457296.0,
      "step": 17675
    },
    {
      "entropy": 5.393123388290405,
      "epoch": 1.9856797888470825,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046130364577050655,
      "loss": 5.256,
      "mean_token_accuracy": 0.19823174327611923,
      "num_tokens": 39468499.0,
      "step": 17680
    },
    {
      "entropy": 5.375853538513184,
      "epoch": 1.9862413657550402,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046128113125043957,
      "loss": 5.1531,
      "mean_token_accuracy": 0.20142264664173126,
      "num_tokens": 39479594.0,
      "step": 17685
    },
    {
      "entropy": 5.439415454864502,
      "epoch": 1.9868029426629978,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046125861079910913,
      "loss": 5.2297,
      "mean_token_accuracy": 0.19938703626394272,
      "num_tokens": 39490048.0,
      "step": 17690
    },
    {
      "entropy": 5.38686089515686,
      "epoch": 1.9873645195709553,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004612360844172322,
      "loss": 5.2111,
      "mean_token_accuracy": 0.2002210021018982,
      "num_tokens": 39501454.0,
      "step": 17695
    },
    {
      "entropy": 5.405000638961792,
      "epoch": 1.9879260964789127,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046121355210552626,
      "loss": 5.1571,
      "mean_token_accuracy": 0.2055862843990326,
      "num_tokens": 39512782.0,
      "step": 17700
    },
    {
      "entropy": 5.3808104515075685,
      "epoch": 1.9884876733868704,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004611910138647085,
      "loss": 5.2252,
      "mean_token_accuracy": 0.20097501426935196,
      "num_tokens": 39524342.0,
      "step": 17705
    },
    {
      "entropy": 5.401195907592774,
      "epoch": 1.989049250294828,
      "grad_norm": 1.125,
      "learning_rate": 0.00046116846969549666,
      "loss": 5.1212,
      "mean_token_accuracy": 0.20421625673770905,
      "num_tokens": 39534744.0,
      "step": 17710
    },
    {
      "entropy": 5.500453853607178,
      "epoch": 1.9896108272027853,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004611459195986085,
      "loss": 5.2614,
      "mean_token_accuracy": 0.19228100925683975,
      "num_tokens": 39547729.0,
      "step": 17715
    },
    {
      "entropy": 5.398198366165161,
      "epoch": 1.990172404110743,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046112336357476217,
      "loss": 5.1477,
      "mean_token_accuracy": 0.19643860906362534,
      "num_tokens": 39558194.0,
      "step": 17720
    },
    {
      "entropy": 5.3414960384368895,
      "epoch": 1.9907339810187006,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004611008016246757,
      "loss": 5.0932,
      "mean_token_accuracy": 0.20330063104629517,
      "num_tokens": 39568265.0,
      "step": 17725
    },
    {
      "entropy": 5.447497081756592,
      "epoch": 1.991295557926658,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046107823374906754,
      "loss": 5.2979,
      "mean_token_accuracy": 0.1921747848391533,
      "num_tokens": 39578936.0,
      "step": 17730
    },
    {
      "entropy": 5.488422870635986,
      "epoch": 1.9918571348346155,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046105565994865614,
      "loss": 5.2755,
      "mean_token_accuracy": 0.19129739701747894,
      "num_tokens": 39590180.0,
      "step": 17735
    },
    {
      "entropy": 5.429736137390137,
      "epoch": 1.9924187117425731,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004610330802241605,
      "loss": 5.2249,
      "mean_token_accuracy": 0.19575858116149902,
      "num_tokens": 39601226.0,
      "step": 17740
    },
    {
      "entropy": 5.382750844955444,
      "epoch": 1.9929802886505308,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004610104945762994,
      "loss": 5.2463,
      "mean_token_accuracy": 0.19801606833934784,
      "num_tokens": 39613119.0,
      "step": 17745
    },
    {
      "entropy": 5.474017190933227,
      "epoch": 1.9935418655584882,
      "grad_norm": 1.328125,
      "learning_rate": 0.000460987903005792,
      "loss": 5.2301,
      "mean_token_accuracy": 0.1990464672446251,
      "num_tokens": 39625269.0,
      "step": 17750
    },
    {
      "entropy": 5.444179916381836,
      "epoch": 1.9941034424664457,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004609653055133576,
      "loss": 5.1127,
      "mean_token_accuracy": 0.2091587945818901,
      "num_tokens": 39635883.0,
      "step": 17755
    },
    {
      "entropy": 5.385455083847046,
      "epoch": 1.9946650193744033,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046094270209971576,
      "loss": 5.1417,
      "mean_token_accuracy": 0.20454081743955613,
      "num_tokens": 39646425.0,
      "step": 17760
    },
    {
      "entropy": 5.438606119155883,
      "epoch": 1.995226596282361,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004609200927655862,
      "loss": 5.2588,
      "mean_token_accuracy": 0.1964820608496666,
      "num_tokens": 39658874.0,
      "step": 17765
    },
    {
      "entropy": 5.390671920776367,
      "epoch": 1.9957881731903184,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004608974775116888,
      "loss": 5.2203,
      "mean_token_accuracy": 0.1992528259754181,
      "num_tokens": 39670401.0,
      "step": 17770
    },
    {
      "entropy": 5.416931772232056,
      "epoch": 1.9963497500982759,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004608748563387436,
      "loss": 5.1469,
      "mean_token_accuracy": 0.2006738781929016,
      "num_tokens": 39681005.0,
      "step": 17775
    },
    {
      "entropy": 5.55496392250061,
      "epoch": 1.9969113270062335,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004608522292474708,
      "loss": 5.3045,
      "mean_token_accuracy": 0.18995154947042464,
      "num_tokens": 39692311.0,
      "step": 17780
    },
    {
      "entropy": 5.40487961769104,
      "epoch": 1.9974729039141912,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004608295962385911,
      "loss": 5.1175,
      "mean_token_accuracy": 0.21269765347242356,
      "num_tokens": 39702782.0,
      "step": 17785
    },
    {
      "entropy": 5.395885229110718,
      "epoch": 1.9980344808221486,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046080695731282484,
      "loss": 5.2716,
      "mean_token_accuracy": 0.19224880188703536,
      "num_tokens": 39713497.0,
      "step": 17790
    },
    {
      "entropy": 5.416759395599366,
      "epoch": 1.998596057730106,
      "grad_norm": 1.140625,
      "learning_rate": 0.000460784312470893,
      "loss": 5.1896,
      "mean_token_accuracy": 0.19332774579524994,
      "num_tokens": 39724355.0,
      "step": 17795
    },
    {
      "entropy": 5.50033130645752,
      "epoch": 1.9991576346380637,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046076166171351673,
      "loss": 5.254,
      "mean_token_accuracy": 0.1935026928782463,
      "num_tokens": 39736523.0,
      "step": 17800
    },
    {
      "entropy": 5.407515144348144,
      "epoch": 1.9997192115460214,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000460739005041417,
      "loss": 5.2011,
      "mean_token_accuracy": 0.19503271728754043,
      "num_tokens": 39746297.0,
      "step": 17805
    },
    {
      "entropy": 5.311057090759277,
      "epoch": 2.000224630763183,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004607163424553153,
      "loss": 5.0054,
      "mean_token_accuracy": 0.21303185158305699,
      "num_tokens": 39755163.0,
      "step": 17810
    },
    {
      "entropy": 5.411541080474853,
      "epoch": 2.0007862076711405,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046069367395593325,
      "loss": 5.2266,
      "mean_token_accuracy": 0.19763481616973877,
      "num_tokens": 39766444.0,
      "step": 17815
    },
    {
      "entropy": 5.480181074142456,
      "epoch": 2.001347784579098,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004606709995439926,
      "loss": 5.1321,
      "mean_token_accuracy": 0.20577976554632188,
      "num_tokens": 39778051.0,
      "step": 17820
    },
    {
      "entropy": 5.426545095443726,
      "epoch": 2.001909361487056,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046064831922021534,
      "loss": 5.1579,
      "mean_token_accuracy": 0.19427043199539185,
      "num_tokens": 39787850.0,
      "step": 17825
    },
    {
      "entropy": 5.4235352039337155,
      "epoch": 2.002470938395013,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004606256329853235,
      "loss": 5.1011,
      "mean_token_accuracy": 0.20248273462057115,
      "num_tokens": 39799428.0,
      "step": 17830
    },
    {
      "entropy": 5.514731311798096,
      "epoch": 2.0030325153029707,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004606029408400396,
      "loss": 5.2839,
      "mean_token_accuracy": 0.19303908199071884,
      "num_tokens": 39811688.0,
      "step": 17835
    },
    {
      "entropy": 5.394984292984009,
      "epoch": 2.0035940922109283,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004605802427850861,
      "loss": 5.0606,
      "mean_token_accuracy": 0.20326649248600007,
      "num_tokens": 39822243.0,
      "step": 17840
    },
    {
      "entropy": 5.370920944213867,
      "epoch": 2.004155669118886,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046055753882118564,
      "loss": 5.18,
      "mean_token_accuracy": 0.2025688886642456,
      "num_tokens": 39833699.0,
      "step": 17845
    },
    {
      "entropy": 5.550212717056274,
      "epoch": 2.004717246026843,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004605348289490611,
      "loss": 5.2984,
      "mean_token_accuracy": 0.19178436696529388,
      "num_tokens": 39844177.0,
      "step": 17850
    },
    {
      "entropy": 5.4014733791351315,
      "epoch": 2.005278822934801,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004605121131694358,
      "loss": 5.1617,
      "mean_token_accuracy": 0.20219145864248275,
      "num_tokens": 39856008.0,
      "step": 17855
    },
    {
      "entropy": 5.36871304512024,
      "epoch": 2.0058403998427585,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004604893914830328,
      "loss": 5.0531,
      "mean_token_accuracy": 0.19985663741827012,
      "num_tokens": 39866209.0,
      "step": 17860
    },
    {
      "entropy": 5.487306785583496,
      "epoch": 2.006401976750716,
      "grad_norm": 1.359375,
      "learning_rate": 0.00046046666389057563,
      "loss": 5.1166,
      "mean_token_accuracy": 0.19955774992704392,
      "num_tokens": 39876986.0,
      "step": 17865
    },
    {
      "entropy": 5.361792850494385,
      "epoch": 2.0069635536586734,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000460443930392788,
      "loss": 5.0788,
      "mean_token_accuracy": 0.20094774067401885,
      "num_tokens": 39886994.0,
      "step": 17870
    },
    {
      "entropy": 5.42294921875,
      "epoch": 2.007525130566631,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046042119099039367,
      "loss": 5.1453,
      "mean_token_accuracy": 0.19731851667165756,
      "num_tokens": 39897514.0,
      "step": 17875
    },
    {
      "entropy": 5.560348033905029,
      "epoch": 2.0080867074745887,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046039844568411667,
      "loss": 5.3461,
      "mean_token_accuracy": 0.19194342344999313,
      "num_tokens": 39909102.0,
      "step": 17880
    },
    {
      "entropy": 5.44173493385315,
      "epoch": 2.0086482843825464,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004603756944746813,
      "loss": 5.0559,
      "mean_token_accuracy": 0.20365308672189714,
      "num_tokens": 39919450.0,
      "step": 17885
    },
    {
      "entropy": 5.329760122299194,
      "epoch": 2.0092098612905036,
      "grad_norm": 1.125,
      "learning_rate": 0.00046035293736281185,
      "loss": 5.0941,
      "mean_token_accuracy": 0.19537930339574813,
      "num_tokens": 39930362.0,
      "step": 17890
    },
    {
      "entropy": 5.354328393936157,
      "epoch": 2.0097714381984613,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004603301743492331,
      "loss": 5.1941,
      "mean_token_accuracy": 0.1986897349357605,
      "num_tokens": 39942047.0,
      "step": 17895
    },
    {
      "entropy": 5.4792482376098635,
      "epoch": 2.010333015106419,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004603074054346696,
      "loss": 5.1531,
      "mean_token_accuracy": 0.20480251759290696,
      "num_tokens": 39954870.0,
      "step": 17900
    },
    {
      "entropy": 5.452645206451416,
      "epoch": 2.0108945920143766,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004602846306198465,
      "loss": 5.1566,
      "mean_token_accuracy": 0.20421427339315415,
      "num_tokens": 39966345.0,
      "step": 17905
    },
    {
      "entropy": 5.32283763885498,
      "epoch": 2.011456168922334,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004602618499054889,
      "loss": 5.0121,
      "mean_token_accuracy": 0.2103532910346985,
      "num_tokens": 39976453.0,
      "step": 17910
    },
    {
      "entropy": 5.412225008010864,
      "epoch": 2.0120177458302915,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004602390632923221,
      "loss": 5.1871,
      "mean_token_accuracy": 0.20009363442659378,
      "num_tokens": 39988990.0,
      "step": 17915
    },
    {
      "entropy": 5.424643230438233,
      "epoch": 2.012579322738249,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046021627078107177,
      "loss": 5.157,
      "mean_token_accuracy": 0.19944593906402588,
      "num_tokens": 39999131.0,
      "step": 17920
    },
    {
      "entropy": 5.469965839385987,
      "epoch": 2.0131408996462063,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046019347237246343,
      "loss": 5.1455,
      "mean_token_accuracy": 0.2024523213505745,
      "num_tokens": 40010197.0,
      "step": 17925
    },
    {
      "entropy": 5.353540706634521,
      "epoch": 2.013702476554164,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004601706680672232,
      "loss": 4.9982,
      "mean_token_accuracy": 0.20985756665468216,
      "num_tokens": 40021388.0,
      "step": 17930
    },
    {
      "entropy": 5.396045732498169,
      "epoch": 2.0142640534621217,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004601478578660771,
      "loss": 5.1278,
      "mean_token_accuracy": 0.20765674412250518,
      "num_tokens": 40032485.0,
      "step": 17935
    },
    {
      "entropy": 5.370942687988281,
      "epoch": 2.0148256303700793,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004601250417697513,
      "loss": 5.2273,
      "mean_token_accuracy": 0.19406579285860062,
      "num_tokens": 40043465.0,
      "step": 17940
    },
    {
      "entropy": 5.409340953826904,
      "epoch": 2.0153872072780366,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004601022197789724,
      "loss": 5.1046,
      "mean_token_accuracy": 0.20733041167259217,
      "num_tokens": 40054934.0,
      "step": 17945
    },
    {
      "entropy": 5.4164186954498295,
      "epoch": 2.015948784185994,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004600793918944671,
      "loss": 5.2087,
      "mean_token_accuracy": 0.20116138011217116,
      "num_tokens": 40066159.0,
      "step": 17950
    },
    {
      "entropy": 5.2966289043426515,
      "epoch": 2.016510361093952,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004600565581169621,
      "loss": 5.0397,
      "mean_token_accuracy": 0.21034094095230102,
      "num_tokens": 40076169.0,
      "step": 17955
    },
    {
      "entropy": 5.429509210586548,
      "epoch": 2.0170719380019095,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046003371844718455,
      "loss": 5.0973,
      "mean_token_accuracy": 0.20579639077186584,
      "num_tokens": 40086852.0,
      "step": 17960
    },
    {
      "entropy": 5.378335666656494,
      "epoch": 2.0176335149098668,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046001087288586164,
      "loss": 5.0849,
      "mean_token_accuracy": 0.1997901901602745,
      "num_tokens": 40097588.0,
      "step": 17965
    },
    {
      "entropy": 5.393562650680542,
      "epoch": 2.0181950918178244,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004599880214337208,
      "loss": 5.0399,
      "mean_token_accuracy": 0.20721816569566726,
      "num_tokens": 40108797.0,
      "step": 17970
    },
    {
      "entropy": 5.39578218460083,
      "epoch": 2.018756668725782,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004599651640914896,
      "loss": 5.179,
      "mean_token_accuracy": 0.19643018692731856,
      "num_tokens": 40119700.0,
      "step": 17975
    },
    {
      "entropy": 5.416951465606689,
      "epoch": 2.0193182456337397,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004599423008598957,
      "loss": 5.202,
      "mean_token_accuracy": 0.1995523378252983,
      "num_tokens": 40131611.0,
      "step": 17980
    },
    {
      "entropy": 5.387915420532226,
      "epoch": 2.019879822541697,
      "grad_norm": 1.125,
      "learning_rate": 0.00045991943173966744,
      "loss": 5.0713,
      "mean_token_accuracy": 0.21377059072256088,
      "num_tokens": 40142186.0,
      "step": 17985
    },
    {
      "entropy": 5.395058488845825,
      "epoch": 2.0204413994496546,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004598965567315326,
      "loss": 5.1208,
      "mean_token_accuracy": 0.20282603949308395,
      "num_tokens": 40153856.0,
      "step": 17990
    },
    {
      "entropy": 5.378701448440552,
      "epoch": 2.0210029763576123,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004598736758362198,
      "loss": 5.047,
      "mean_token_accuracy": 0.20751967132091523,
      "num_tokens": 40164158.0,
      "step": 17995
    },
    {
      "entropy": 5.384527063369751,
      "epoch": 2.02156455326557,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045985078905445734,
      "loss": 5.1871,
      "mean_token_accuracy": 0.20097993463277816,
      "num_tokens": 40176751.0,
      "step": 18000
    },
    {
      "epoch": 2.02156455326557,
      "eval_entropy": 5.213723817928138,
      "eval_loss": 5.295015811920166,
      "eval_mean_token_accuracy": 0.20172234415731266,
      "eval_num_tokens": 40176751.0,
      "eval_runtime": 24.1963,
      "eval_samples_per_second": 1281.726,
      "eval_steps_per_second": 160.231,
      "step": 18000
    },
    {
      "entropy": 5.43703670501709,
      "epoch": 2.022126130173527,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045982789638697414,
      "loss": 5.1731,
      "mean_token_accuracy": 0.19921893775463104,
      "num_tokens": 40188000.0,
      "step": 18005
    },
    {
      "entropy": 5.434857034683228,
      "epoch": 2.022687707081485,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045980499783449897,
      "loss": 5.1646,
      "mean_token_accuracy": 0.20498383045196533,
      "num_tokens": 40199177.0,
      "step": 18010
    },
    {
      "entropy": 5.446937990188599,
      "epoch": 2.0232492839894425,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045978209339776104,
      "loss": 5.1781,
      "mean_token_accuracy": 0.19953609704971315,
      "num_tokens": 40210082.0,
      "step": 18015
    },
    {
      "entropy": 5.437570428848266,
      "epoch": 2.0238108608973997,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045975918307748965,
      "loss": 5.157,
      "mean_token_accuracy": 0.19781923443078994,
      "num_tokens": 40222503.0,
      "step": 18020
    },
    {
      "entropy": 5.4052056789398195,
      "epoch": 2.0243724378053574,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004597362668744141,
      "loss": 5.1169,
      "mean_token_accuracy": 0.2043033331632614,
      "num_tokens": 40233690.0,
      "step": 18025
    },
    {
      "entropy": 5.483539676666259,
      "epoch": 2.024934014713315,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004597133447892642,
      "loss": 5.2318,
      "mean_token_accuracy": 0.19525859951972963,
      "num_tokens": 40246274.0,
      "step": 18030
    },
    {
      "entropy": 5.4171473503112795,
      "epoch": 2.0254955916212727,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004596904168227698,
      "loss": 5.1607,
      "mean_token_accuracy": 0.19733163863420486,
      "num_tokens": 40257469.0,
      "step": 18035
    },
    {
      "entropy": 5.381694221496582,
      "epoch": 2.02605716852923,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045966748297566085,
      "loss": 5.1464,
      "mean_token_accuracy": 0.20881411731243132,
      "num_tokens": 40268982.0,
      "step": 18040
    },
    {
      "entropy": 5.481057357788086,
      "epoch": 2.0266187454371876,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045964454324866767,
      "loss": 5.1806,
      "mean_token_accuracy": 0.20121414214372635,
      "num_tokens": 40280301.0,
      "step": 18045
    },
    {
      "entropy": 5.398323345184326,
      "epoch": 2.0271803223451452,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045962159764252055,
      "loss": 5.0922,
      "mean_token_accuracy": 0.207304385304451,
      "num_tokens": 40292058.0,
      "step": 18050
    },
    {
      "entropy": 5.369397115707398,
      "epoch": 2.027741899253103,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004595986461579502,
      "loss": 5.1495,
      "mean_token_accuracy": 0.19824391901493071,
      "num_tokens": 40301957.0,
      "step": 18055
    },
    {
      "entropy": 5.421345901489258,
      "epoch": 2.02830347616106,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045957568879568736,
      "loss": 5.1156,
      "mean_token_accuracy": 0.20401425212621688,
      "num_tokens": 40313645.0,
      "step": 18060
    },
    {
      "entropy": 5.413603591918945,
      "epoch": 2.0288650530690178,
      "grad_norm": 1.125,
      "learning_rate": 0.00045955272555646306,
      "loss": 5.1929,
      "mean_token_accuracy": 0.1979241967201233,
      "num_tokens": 40324416.0,
      "step": 18065
    },
    {
      "entropy": 5.534650993347168,
      "epoch": 2.0294266299769754,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004595297564410083,
      "loss": 5.2996,
      "mean_token_accuracy": 0.20027057379484176,
      "num_tokens": 40335596.0,
      "step": 18070
    },
    {
      "entropy": 5.486532878875733,
      "epoch": 2.029988206884933,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045950678145005456,
      "loss": 5.2627,
      "mean_token_accuracy": 0.1932991102337837,
      "num_tokens": 40346426.0,
      "step": 18075
    },
    {
      "entropy": 5.35315170288086,
      "epoch": 2.0305497837928903,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004594838005843333,
      "loss": 5.007,
      "mean_token_accuracy": 0.20781387835741044,
      "num_tokens": 40357435.0,
      "step": 18080
    },
    {
      "entropy": 5.38038477897644,
      "epoch": 2.031111360700848,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045946081384457633,
      "loss": 5.0927,
      "mean_token_accuracy": 0.204904505610466,
      "num_tokens": 40367332.0,
      "step": 18085
    },
    {
      "entropy": 5.378155517578125,
      "epoch": 2.0316729376088056,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045943782123151556,
      "loss": 5.0827,
      "mean_token_accuracy": 0.19926803857088088,
      "num_tokens": 40378571.0,
      "step": 18090
    },
    {
      "entropy": 5.393703460693359,
      "epoch": 2.0322345145167633,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004594148227458829,
      "loss": 5.1112,
      "mean_token_accuracy": 0.20203131139278413,
      "num_tokens": 40389467.0,
      "step": 18095
    },
    {
      "entropy": 5.407750844955444,
      "epoch": 2.0327960914247205,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004593918183884108,
      "loss": 5.1308,
      "mean_token_accuracy": 0.20328249335289,
      "num_tokens": 40399785.0,
      "step": 18100
    },
    {
      "entropy": 5.342033052444458,
      "epoch": 2.033357668332678,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004593688081598318,
      "loss": 5.1137,
      "mean_token_accuracy": 0.20046151131391526,
      "num_tokens": 40409862.0,
      "step": 18105
    },
    {
      "entropy": 5.4469459533691404,
      "epoch": 2.033919245240636,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045934579206087837,
      "loss": 5.2012,
      "mean_token_accuracy": 0.19689399898052215,
      "num_tokens": 40421627.0,
      "step": 18110
    },
    {
      "entropy": 5.454843902587891,
      "epoch": 2.034480822148593,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045932277009228336,
      "loss": 5.1771,
      "mean_token_accuracy": 0.19880998134613037,
      "num_tokens": 40433327.0,
      "step": 18115
    },
    {
      "entropy": 5.305614805221557,
      "epoch": 2.0350423990565507,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045929974225477986,
      "loss": 5.0545,
      "mean_token_accuracy": 0.20649130046367645,
      "num_tokens": 40444217.0,
      "step": 18120
    },
    {
      "entropy": 5.370799112319946,
      "epoch": 2.0356039759645084,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045927670854910117,
      "loss": 5.1389,
      "mean_token_accuracy": 0.2018886238336563,
      "num_tokens": 40454760.0,
      "step": 18125
    },
    {
      "entropy": 5.458308506011963,
      "epoch": 2.036165552872466,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004592536689759806,
      "loss": 5.1554,
      "mean_token_accuracy": 0.20088490098714828,
      "num_tokens": 40464638.0,
      "step": 18130
    },
    {
      "entropy": 5.463850450515747,
      "epoch": 2.0367271297804233,
      "grad_norm": 1.125,
      "learning_rate": 0.00045923062353615167,
      "loss": 5.1242,
      "mean_token_accuracy": 0.20152823477983475,
      "num_tokens": 40474911.0,
      "step": 18135
    },
    {
      "entropy": 5.38600869178772,
      "epoch": 2.037288706688381,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045920757223034833,
      "loss": 5.1396,
      "mean_token_accuracy": 0.19511234015226364,
      "num_tokens": 40485369.0,
      "step": 18140
    },
    {
      "entropy": 5.413777160644531,
      "epoch": 2.0378502835963386,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004591845150593044,
      "loss": 5.1659,
      "mean_token_accuracy": 0.19316488653421401,
      "num_tokens": 40495955.0,
      "step": 18145
    },
    {
      "entropy": 5.427492237091064,
      "epoch": 2.0384118605042962,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045916145202375407,
      "loss": 5.191,
      "mean_token_accuracy": 0.20139982998371125,
      "num_tokens": 40506633.0,
      "step": 18150
    },
    {
      "entropy": 5.465146064758301,
      "epoch": 2.0389734374122535,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004591383831244317,
      "loss": 5.1632,
      "mean_token_accuracy": 0.19729772508144378,
      "num_tokens": 40516868.0,
      "step": 18155
    },
    {
      "entropy": 5.3262919902801515,
      "epoch": 2.039535014320211,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045911530836207173,
      "loss": 5.0743,
      "mean_token_accuracy": 0.20908244401216508,
      "num_tokens": 40528873.0,
      "step": 18160
    },
    {
      "entropy": 5.411384963989258,
      "epoch": 2.0400965912281688,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045909222773740904,
      "loss": 5.1292,
      "mean_token_accuracy": 0.196569561958313,
      "num_tokens": 40539124.0,
      "step": 18165
    },
    {
      "entropy": 5.37464451789856,
      "epoch": 2.0406581681361264,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045906914125117833,
      "loss": 5.1231,
      "mean_token_accuracy": 0.2030426025390625,
      "num_tokens": 40550264.0,
      "step": 18170
    },
    {
      "entropy": 5.275174379348755,
      "epoch": 2.0412197450440837,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045904604890411476,
      "loss": 4.9763,
      "mean_token_accuracy": 0.2112727016210556,
      "num_tokens": 40562319.0,
      "step": 18175
    },
    {
      "entropy": 5.350495004653931,
      "epoch": 2.0417813219520413,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045902295069695373,
      "loss": 5.1794,
      "mean_token_accuracy": 0.202217735350132,
      "num_tokens": 40573962.0,
      "step": 18180
    },
    {
      "entropy": 5.390083837509155,
      "epoch": 2.042342898859999,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004589998466304305,
      "loss": 5.0567,
      "mean_token_accuracy": 0.20021899342536925,
      "num_tokens": 40584620.0,
      "step": 18185
    },
    {
      "entropy": 5.4081549644470215,
      "epoch": 2.0429044757679566,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045897673670528086,
      "loss": 5.1963,
      "mean_token_accuracy": 0.19877537190914155,
      "num_tokens": 40596670.0,
      "step": 18190
    },
    {
      "entropy": 5.393158721923828,
      "epoch": 2.043466052675914,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004589536209222404,
      "loss": 5.1134,
      "mean_token_accuracy": 0.2050512507557869,
      "num_tokens": 40607742.0,
      "step": 18195
    },
    {
      "entropy": 5.399907732009888,
      "epoch": 2.0440276295838715,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004589304992820454,
      "loss": 5.0485,
      "mean_token_accuracy": 0.20613530278205872,
      "num_tokens": 40618873.0,
      "step": 18200
    },
    {
      "entropy": 5.3731081008911135,
      "epoch": 2.044589206491829,
      "grad_norm": 1.25,
      "learning_rate": 0.000458907371785432,
      "loss": 5.1062,
      "mean_token_accuracy": 0.20110146701335907,
      "num_tokens": 40630327.0,
      "step": 18205
    },
    {
      "entropy": 5.302007484436035,
      "epoch": 2.0451507833997864,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004588842384331366,
      "loss": 5.0713,
      "mean_token_accuracy": 0.2047855630517006,
      "num_tokens": 40641515.0,
      "step": 18210
    },
    {
      "entropy": 5.410367250442505,
      "epoch": 2.045712360307744,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045886109922589564,
      "loss": 5.1092,
      "mean_token_accuracy": 0.2031296581029892,
      "num_tokens": 40652123.0,
      "step": 18215
    },
    {
      "entropy": 5.3820727348327635,
      "epoch": 2.0462739372157017,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045883795416444594,
      "loss": 5.084,
      "mean_token_accuracy": 0.20583404153585433,
      "num_tokens": 40662895.0,
      "step": 18220
    },
    {
      "entropy": 5.30221004486084,
      "epoch": 2.0468355141236594,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045881480324952455,
      "loss": 5.0189,
      "mean_token_accuracy": 0.21363647729158403,
      "num_tokens": 40673243.0,
      "step": 18225
    },
    {
      "entropy": 5.408434200286865,
      "epoch": 2.0473970910316166,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045879164648186853,
      "loss": 5.0864,
      "mean_token_accuracy": 0.2101287752389908,
      "num_tokens": 40683861.0,
      "step": 18230
    },
    {
      "entropy": 5.41359167098999,
      "epoch": 2.0479586679395743,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045876848386221513,
      "loss": 5.0575,
      "mean_token_accuracy": 0.20711664408445357,
      "num_tokens": 40694995.0,
      "step": 18235
    },
    {
      "entropy": 5.395422410964966,
      "epoch": 2.048520244847532,
      "grad_norm": 1.25,
      "learning_rate": 0.00045874531539130204,
      "loss": 5.1211,
      "mean_token_accuracy": 0.20387266725301742,
      "num_tokens": 40705203.0,
      "step": 18240
    },
    {
      "entropy": 5.40765872001648,
      "epoch": 2.0490818217554896,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045872214106986684,
      "loss": 5.1167,
      "mean_token_accuracy": 0.20410765260457991,
      "num_tokens": 40716439.0,
      "step": 18245
    },
    {
      "entropy": 5.335300540924072,
      "epoch": 2.049643398663447,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045869896089864733,
      "loss": 5.1038,
      "mean_token_accuracy": 0.2047787293791771,
      "num_tokens": 40727436.0,
      "step": 18250
    },
    {
      "entropy": 5.358970975875854,
      "epoch": 2.0502049755714045,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004586757748783817,
      "loss": 5.0707,
      "mean_token_accuracy": 0.2073178246617317,
      "num_tokens": 40738423.0,
      "step": 18255
    },
    {
      "entropy": 5.529004001617432,
      "epoch": 2.050766552479362,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045865258300980817,
      "loss": 5.2521,
      "mean_token_accuracy": 0.20093085318803788,
      "num_tokens": 40749506.0,
      "step": 18260
    },
    {
      "entropy": 5.425411558151245,
      "epoch": 2.05132812938732,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045862938529366514,
      "loss": 5.1006,
      "mean_token_accuracy": 0.21089959293603897,
      "num_tokens": 40762165.0,
      "step": 18265
    },
    {
      "entropy": 5.415039491653443,
      "epoch": 2.051889706295277,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004586061817306913,
      "loss": 5.1989,
      "mean_token_accuracy": 0.19133614003658295,
      "num_tokens": 40774043.0,
      "step": 18270
    },
    {
      "entropy": 5.299159812927246,
      "epoch": 2.0524512832032347,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045858297232162536,
      "loss": 5.0813,
      "mean_token_accuracy": 0.205808225274086,
      "num_tokens": 40784758.0,
      "step": 18275
    },
    {
      "entropy": 5.418612051010132,
      "epoch": 2.0530128601111923,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045855975706720644,
      "loss": 5.1089,
      "mean_token_accuracy": 0.20819225460290908,
      "num_tokens": 40794735.0,
      "step": 18280
    },
    {
      "entropy": 5.316957569122314,
      "epoch": 2.05357443701915,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004585365359681736,
      "loss": 4.9831,
      "mean_token_accuracy": 0.20653291791677475,
      "num_tokens": 40805491.0,
      "step": 18285
    },
    {
      "entropy": 5.438990068435669,
      "epoch": 2.054136013927107,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045851330902526633,
      "loss": 5.2431,
      "mean_token_accuracy": 0.1978488266468048,
      "num_tokens": 40816945.0,
      "step": 18290
    },
    {
      "entropy": 5.471668672561646,
      "epoch": 2.054697590835065,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045849007623922406,
      "loss": 5.171,
      "mean_token_accuracy": 0.20284661650657654,
      "num_tokens": 40828199.0,
      "step": 18295
    },
    {
      "entropy": 5.41905403137207,
      "epoch": 2.0552591677430225,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004584668376107867,
      "loss": 5.2148,
      "mean_token_accuracy": 0.19690203070640563,
      "num_tokens": 40839618.0,
      "step": 18300
    },
    {
      "entropy": 5.3349456787109375,
      "epoch": 2.0558207446509797,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000458443593140694,
      "loss": 5.0774,
      "mean_token_accuracy": 0.20294277966022492,
      "num_tokens": 40851078.0,
      "step": 18305
    },
    {
      "entropy": 5.398328495025635,
      "epoch": 2.0563823215589374,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045842034282968615,
      "loss": 5.1,
      "mean_token_accuracy": 0.2042147621512413,
      "num_tokens": 40862397.0,
      "step": 18310
    },
    {
      "entropy": 5.404021739959717,
      "epoch": 2.056943898466895,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004583970866785035,
      "loss": 5.0982,
      "mean_token_accuracy": 0.20357878655195236,
      "num_tokens": 40872739.0,
      "step": 18315
    },
    {
      "entropy": 5.473979902267456,
      "epoch": 2.0575054753748527,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045837382468788643,
      "loss": 5.1653,
      "mean_token_accuracy": 0.20331645607948304,
      "num_tokens": 40884723.0,
      "step": 18320
    },
    {
      "entropy": 5.400859022140503,
      "epoch": 2.05806705228281,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004583505568585757,
      "loss": 5.1657,
      "mean_token_accuracy": 0.1992650255560875,
      "num_tokens": 40895927.0,
      "step": 18325
    },
    {
      "entropy": 5.344442844390869,
      "epoch": 2.0586286291907676,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045832728319131205,
      "loss": 5.144,
      "mean_token_accuracy": 0.20040240436792373,
      "num_tokens": 40906504.0,
      "step": 18330
    },
    {
      "entropy": 5.396454334259033,
      "epoch": 2.0591902060987253,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004583040036868367,
      "loss": 5.025,
      "mean_token_accuracy": 0.2053947627544403,
      "num_tokens": 40918420.0,
      "step": 18335
    },
    {
      "entropy": 5.377739381790161,
      "epoch": 2.059751783006683,
      "grad_norm": 1.125,
      "learning_rate": 0.0004582807183458907,
      "loss": 5.1701,
      "mean_token_accuracy": 0.20263474732637404,
      "num_tokens": 40929665.0,
      "step": 18340
    },
    {
      "entropy": 5.375206184387207,
      "epoch": 2.06031335991464,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045825742716921553,
      "loss": 5.1143,
      "mean_token_accuracy": 0.19681310951709746,
      "num_tokens": 40941134.0,
      "step": 18345
    },
    {
      "entropy": 5.457425975799561,
      "epoch": 2.060874936822598,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004582341301575529,
      "loss": 5.1513,
      "mean_token_accuracy": 0.2029227316379547,
      "num_tokens": 40953230.0,
      "step": 18350
    },
    {
      "entropy": 5.355022954940796,
      "epoch": 2.0614365137305555,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045821082731164445,
      "loss": 5.0966,
      "mean_token_accuracy": 0.20658308863639832,
      "num_tokens": 40964398.0,
      "step": 18355
    },
    {
      "entropy": 5.413182497024536,
      "epoch": 2.061998090638513,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004581875186322322,
      "loss": 5.1568,
      "mean_token_accuracy": 0.20268506556749344,
      "num_tokens": 40975561.0,
      "step": 18360
    },
    {
      "entropy": 5.345660734176636,
      "epoch": 2.0625596675464704,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045816420412005826,
      "loss": 5.0437,
      "mean_token_accuracy": 0.2077670380473137,
      "num_tokens": 40986393.0,
      "step": 18365
    },
    {
      "entropy": 5.358331203460693,
      "epoch": 2.063121244454428,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045814088377586505,
      "loss": 5.1715,
      "mean_token_accuracy": 0.2059535041451454,
      "num_tokens": 40998143.0,
      "step": 18370
    },
    {
      "entropy": 5.4432251930236815,
      "epoch": 2.0636828213623857,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045811755760039513,
      "loss": 5.1712,
      "mean_token_accuracy": 0.20073002874851226,
      "num_tokens": 41009747.0,
      "step": 18375
    },
    {
      "entropy": 5.402835416793823,
      "epoch": 2.0642443982703433,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045809422559439113,
      "loss": 5.0729,
      "mean_token_accuracy": 0.2039320081472397,
      "num_tokens": 41023146.0,
      "step": 18380
    },
    {
      "entropy": 5.421822786331177,
      "epoch": 2.0648059751783006,
      "grad_norm": 1.125,
      "learning_rate": 0.00045807088775859595,
      "loss": 5.2145,
      "mean_token_accuracy": 0.19761134386062623,
      "num_tokens": 41034138.0,
      "step": 18385
    },
    {
      "entropy": 5.343217229843139,
      "epoch": 2.065367552086258,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004580475440937527,
      "loss": 5.012,
      "mean_token_accuracy": 0.2059514969587326,
      "num_tokens": 41045457.0,
      "step": 18390
    },
    {
      "entropy": 5.344277048110962,
      "epoch": 2.065929128994216,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045802419460060464,
      "loss": 5.0982,
      "mean_token_accuracy": 0.20653970986604692,
      "num_tokens": 41056264.0,
      "step": 18395
    },
    {
      "entropy": 5.480771541595459,
      "epoch": 2.066490705902173,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004580008392798953,
      "loss": 5.2678,
      "mean_token_accuracy": 0.19225547462701797,
      "num_tokens": 41067206.0,
      "step": 18400
    },
    {
      "entropy": 5.34266505241394,
      "epoch": 2.0670522828101308,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045797747813236815,
      "loss": 5.0075,
      "mean_token_accuracy": 0.20355408042669296,
      "num_tokens": 41077897.0,
      "step": 18405
    },
    {
      "entropy": 5.353272533416748,
      "epoch": 2.0676138597180884,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045795411115876725,
      "loss": 5.0911,
      "mean_token_accuracy": 0.20177918374538423,
      "num_tokens": 41088964.0,
      "step": 18410
    },
    {
      "entropy": 5.415457248687744,
      "epoch": 2.068175436626046,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045793073835983644,
      "loss": 5.033,
      "mean_token_accuracy": 0.20817486196756363,
      "num_tokens": 41098818.0,
      "step": 18415
    },
    {
      "entropy": 5.443577527999878,
      "epoch": 2.0687370135340033,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045790735973631996,
      "loss": 5.2272,
      "mean_token_accuracy": 0.19722471088171006,
      "num_tokens": 41111222.0,
      "step": 18420
    },
    {
      "entropy": 5.346242046356201,
      "epoch": 2.069298590441961,
      "grad_norm": 1.125,
      "learning_rate": 0.00045788397528896224,
      "loss": 5.1037,
      "mean_token_accuracy": 0.19928799271583558,
      "num_tokens": 41122875.0,
      "step": 18425
    },
    {
      "entropy": 5.376118469238281,
      "epoch": 2.0698601673499186,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045786058501850774,
      "loss": 5.1184,
      "mean_token_accuracy": 0.1959065467119217,
      "num_tokens": 41133869.0,
      "step": 18430
    },
    {
      "entropy": 5.440343809127808,
      "epoch": 2.0704217442578763,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045783718892570136,
      "loss": 5.0389,
      "mean_token_accuracy": 0.20243901908397674,
      "num_tokens": 41144482.0,
      "step": 18435
    },
    {
      "entropy": 5.314915895462036,
      "epoch": 2.0709833211658335,
      "grad_norm": 1.140625,
      "learning_rate": 0.000457813787011288,
      "loss": 5.0684,
      "mean_token_accuracy": 0.1957713857293129,
      "num_tokens": 41154997.0,
      "step": 18440
    },
    {
      "entropy": 5.295909023284912,
      "epoch": 2.071544898073791,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004577903792760127,
      "loss": 5.1066,
      "mean_token_accuracy": 0.2030698299407959,
      "num_tokens": 41167852.0,
      "step": 18445
    },
    {
      "entropy": 5.408872795104981,
      "epoch": 2.072106474981749,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004577669657206209,
      "loss": 5.0827,
      "mean_token_accuracy": 0.2056512713432312,
      "num_tokens": 41178364.0,
      "step": 18450
    },
    {
      "entropy": 5.4153101444244385,
      "epoch": 2.0726680518897065,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004577435463458579,
      "loss": 5.2064,
      "mean_token_accuracy": 0.1974971756339073,
      "num_tokens": 41190003.0,
      "step": 18455
    },
    {
      "entropy": 5.364392995834351,
      "epoch": 2.0732296287976637,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004577201211524696,
      "loss": 5.0358,
      "mean_token_accuracy": 0.20290035754442215,
      "num_tokens": 41201574.0,
      "step": 18460
    },
    {
      "entropy": 5.352954626083374,
      "epoch": 2.0737912057056214,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004576966901412018,
      "loss": 5.0163,
      "mean_token_accuracy": 0.21043792515993118,
      "num_tokens": 41211913.0,
      "step": 18465
    },
    {
      "entropy": 5.325253057479858,
      "epoch": 2.074352782613579,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045767325331280044,
      "loss": 5.0773,
      "mean_token_accuracy": 0.2065901756286621,
      "num_tokens": 41222626.0,
      "step": 18470
    },
    {
      "entropy": 5.423304557800293,
      "epoch": 2.0749143595215367,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004576498106680119,
      "loss": 5.2197,
      "mean_token_accuracy": 0.1958962321281433,
      "num_tokens": 41234321.0,
      "step": 18475
    },
    {
      "entropy": 5.4157373905181885,
      "epoch": 2.075475936429494,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004576263622075826,
      "loss": 5.0548,
      "mean_token_accuracy": 0.2094673588871956,
      "num_tokens": 41244889.0,
      "step": 18480
    },
    {
      "entropy": 5.468678712844849,
      "epoch": 2.0760375133374516,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000457602907932259,
      "loss": 5.2744,
      "mean_token_accuracy": 0.19212742894887924,
      "num_tokens": 41255710.0,
      "step": 18485
    },
    {
      "entropy": 5.435026359558106,
      "epoch": 2.0765990902454092,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045757944784278803,
      "loss": 5.1789,
      "mean_token_accuracy": 0.20542150735855103,
      "num_tokens": 41265786.0,
      "step": 18490
    },
    {
      "entropy": 5.3690355777740475,
      "epoch": 2.0771606671533664,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004575559819399166,
      "loss": 5.1015,
      "mean_token_accuracy": 0.20058395564556122,
      "num_tokens": 41276755.0,
      "step": 18495
    },
    {
      "entropy": 5.407320833206176,
      "epoch": 2.077722244061324,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045753251022439195,
      "loss": 5.0324,
      "mean_token_accuracy": 0.20316243469715117,
      "num_tokens": 41287219.0,
      "step": 18500
    },
    {
      "entropy": 5.387466859817505,
      "epoch": 2.0782838209692818,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004575090326969614,
      "loss": 5.1439,
      "mean_token_accuracy": 0.19994975626468658,
      "num_tokens": 41297356.0,
      "step": 18505
    },
    {
      "entropy": 5.275311088562011,
      "epoch": 2.0788453978772394,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004574855493583723,
      "loss": 5.047,
      "mean_token_accuracy": 0.21338758170604705,
      "num_tokens": 41307219.0,
      "step": 18510
    },
    {
      "entropy": 5.384287357330322,
      "epoch": 2.0794069747851966,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045746206020937276,
      "loss": 5.1614,
      "mean_token_accuracy": 0.19726884216070176,
      "num_tokens": 41318827.0,
      "step": 18515
    },
    {
      "entropy": 5.460938930511475,
      "epoch": 2.0799685516931543,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045743856525071034,
      "loss": 5.1793,
      "mean_token_accuracy": 0.19610237926244736,
      "num_tokens": 41329497.0,
      "step": 18520
    },
    {
      "entropy": 5.393164491653442,
      "epoch": 2.080530128601112,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045741506448313327,
      "loss": 5.1289,
      "mean_token_accuracy": 0.20450254529714584,
      "num_tokens": 41340577.0,
      "step": 18525
    },
    {
      "entropy": 5.403251123428345,
      "epoch": 2.0810917055090696,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045739155790738975,
      "loss": 5.1035,
      "mean_token_accuracy": 0.19617319405078887,
      "num_tokens": 41351685.0,
      "step": 18530
    },
    {
      "entropy": 5.5340827941894535,
      "epoch": 2.081653282417027,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045736804552422836,
      "loss": 5.2389,
      "mean_token_accuracy": 0.19586437344551086,
      "num_tokens": 41363053.0,
      "step": 18535
    },
    {
      "entropy": 5.448911380767822,
      "epoch": 2.0822148593249845,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004573445273343976,
      "loss": 5.2346,
      "mean_token_accuracy": 0.19701223373413085,
      "num_tokens": 41374847.0,
      "step": 18540
    },
    {
      "entropy": 5.374650859832764,
      "epoch": 2.082776436232942,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004573210033386464,
      "loss": 5.2038,
      "mean_token_accuracy": 0.19495317339897156,
      "num_tokens": 41386114.0,
      "step": 18545
    },
    {
      "entropy": 5.333445453643799,
      "epoch": 2.0833380131409,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004572974735377238,
      "loss": 5.0026,
      "mean_token_accuracy": 0.20739843249320983,
      "num_tokens": 41397842.0,
      "step": 18550
    },
    {
      "entropy": 5.388910007476807,
      "epoch": 2.083899590048857,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004572739379323788,
      "loss": 5.1264,
      "mean_token_accuracy": 0.20132465213537215,
      "num_tokens": 41409824.0,
      "step": 18555
    },
    {
      "entropy": 5.370268106460571,
      "epoch": 2.0844611669568147,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045725039652336105,
      "loss": 5.0861,
      "mean_token_accuracy": 0.19551468789577484,
      "num_tokens": 41422260.0,
      "step": 18560
    },
    {
      "entropy": 5.368318367004394,
      "epoch": 2.0850227438647724,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004572268493114199,
      "loss": 5.1142,
      "mean_token_accuracy": 0.204256471991539,
      "num_tokens": 41433068.0,
      "step": 18565
    },
    {
      "entropy": 5.390110635757447,
      "epoch": 2.08558432077273,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004572032962973052,
      "loss": 5.0741,
      "mean_token_accuracy": 0.20952352434396743,
      "num_tokens": 41443759.0,
      "step": 18570
    },
    {
      "entropy": 5.4095762252807615,
      "epoch": 2.0861458976806873,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045717973748176695,
      "loss": 5.1669,
      "mean_token_accuracy": 0.19900788515806198,
      "num_tokens": 41453876.0,
      "step": 18575
    },
    {
      "entropy": 5.3467199325561525,
      "epoch": 2.086707474588645,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045715617286555516,
      "loss": 5.0985,
      "mean_token_accuracy": 0.2006133571267128,
      "num_tokens": 41464731.0,
      "step": 18580
    },
    {
      "entropy": 5.4447534561157225,
      "epoch": 2.0872690514966026,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045713260244942017,
      "loss": 5.1819,
      "mean_token_accuracy": 0.1971881464123726,
      "num_tokens": 41476288.0,
      "step": 18585
    },
    {
      "entropy": 5.458415460586548,
      "epoch": 2.08783062840456,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004571090262341125,
      "loss": 5.0983,
      "mean_token_accuracy": 0.20746297538280487,
      "num_tokens": 41485968.0,
      "step": 18590
    },
    {
      "entropy": 5.394538688659668,
      "epoch": 2.0883922053125175,
      "grad_norm": 1.125,
      "learning_rate": 0.00045708544422038284,
      "loss": 5.1343,
      "mean_token_accuracy": 0.19838968962430953,
      "num_tokens": 41496393.0,
      "step": 18595
    },
    {
      "entropy": 5.368805360794068,
      "epoch": 2.088953782220475,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004570618564089819,
      "loss": 5.1152,
      "mean_token_accuracy": 0.2043430045247078,
      "num_tokens": 41507539.0,
      "step": 18600
    },
    {
      "entropy": 5.447133779525757,
      "epoch": 2.089515359128433,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045703826280066095,
      "loss": 5.2026,
      "mean_token_accuracy": 0.19173247069120408,
      "num_tokens": 41517845.0,
      "step": 18605
    },
    {
      "entropy": 5.370899963378906,
      "epoch": 2.09007693603639,
      "grad_norm": 1.125,
      "learning_rate": 0.0004570146633961711,
      "loss": 5.1098,
      "mean_token_accuracy": 0.20523688942193985,
      "num_tokens": 41528204.0,
      "step": 18610
    },
    {
      "entropy": 5.409570217132568,
      "epoch": 2.0906385129443477,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045699105819626374,
      "loss": 5.1427,
      "mean_token_accuracy": 0.20017992854118347,
      "num_tokens": 41538425.0,
      "step": 18615
    },
    {
      "entropy": 5.408479166030884,
      "epoch": 2.0912000898523053,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004569674472016904,
      "loss": 5.183,
      "mean_token_accuracy": 0.2004904180765152,
      "num_tokens": 41548762.0,
      "step": 18620
    },
    {
      "entropy": 5.362415742874146,
      "epoch": 2.091761666760263,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004569438304132031,
      "loss": 5.0607,
      "mean_token_accuracy": 0.21075141429901123,
      "num_tokens": 41559003.0,
      "step": 18625
    },
    {
      "entropy": 5.376693296432495,
      "epoch": 2.09232324366822,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045692020783155376,
      "loss": 5.0931,
      "mean_token_accuracy": 0.2045363590121269,
      "num_tokens": 41569630.0,
      "step": 18630
    },
    {
      "entropy": 5.440635442733765,
      "epoch": 2.092884820576178,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004568965794574943,
      "loss": 5.1793,
      "mean_token_accuracy": 0.20131716281175613,
      "num_tokens": 41580671.0,
      "step": 18635
    },
    {
      "entropy": 5.482160806655884,
      "epoch": 2.0934463974841355,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004568729452917772,
      "loss": 5.1818,
      "mean_token_accuracy": 0.19622886478900908,
      "num_tokens": 41592475.0,
      "step": 18640
    },
    {
      "entropy": 5.401454114913941,
      "epoch": 2.094007974392093,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045684930533515506,
      "loss": 5.1511,
      "mean_token_accuracy": 0.2002349317073822,
      "num_tokens": 41603675.0,
      "step": 18645
    },
    {
      "entropy": 5.3843354225158695,
      "epoch": 2.0945695513000504,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045682565958838045,
      "loss": 5.0866,
      "mean_token_accuracy": 0.2015704795718193,
      "num_tokens": 41614684.0,
      "step": 18650
    },
    {
      "entropy": 5.339318943023682,
      "epoch": 2.095131128208008,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004568020080522063,
      "loss": 5.1867,
      "mean_token_accuracy": 0.20150647014379502,
      "num_tokens": 41625368.0,
      "step": 18655
    },
    {
      "entropy": 5.418268442153931,
      "epoch": 2.0956927051159657,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004567783507273858,
      "loss": 5.1276,
      "mean_token_accuracy": 0.20380425006151198,
      "num_tokens": 41636542.0,
      "step": 18660
    },
    {
      "entropy": 5.49003849029541,
      "epoch": 2.0962542820239234,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045675468761467204,
      "loss": 5.1971,
      "mean_token_accuracy": 0.1976623609662056,
      "num_tokens": 41646654.0,
      "step": 18665
    },
    {
      "entropy": 5.492196798324585,
      "epoch": 2.0968158589318806,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045673101871481853,
      "loss": 5.2866,
      "mean_token_accuracy": 0.19260170310735703,
      "num_tokens": 41659160.0,
      "step": 18670
    },
    {
      "entropy": 5.4133275032043455,
      "epoch": 2.0973774358398383,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004567073440285789,
      "loss": 5.1268,
      "mean_token_accuracy": 0.20778571367263793,
      "num_tokens": 41670864.0,
      "step": 18675
    },
    {
      "entropy": 5.410419082641601,
      "epoch": 2.097939012747796,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045668366355670697,
      "loss": 5.0421,
      "mean_token_accuracy": 0.204177425801754,
      "num_tokens": 41681947.0,
      "step": 18680
    },
    {
      "entropy": 5.428155755996704,
      "epoch": 2.098500589655753,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004566599772999567,
      "loss": 5.1918,
      "mean_token_accuracy": 0.1920841008424759,
      "num_tokens": 41692625.0,
      "step": 18685
    },
    {
      "entropy": 5.530041074752807,
      "epoch": 2.099062166563711,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045663628525908237,
      "loss": 5.2085,
      "mean_token_accuracy": 0.19925596565008163,
      "num_tokens": 41703118.0,
      "step": 18690
    },
    {
      "entropy": 5.408686780929566,
      "epoch": 2.0996237434716685,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004566125874348382,
      "loss": 5.1343,
      "mean_token_accuracy": 0.20538390129804612,
      "num_tokens": 41713910.0,
      "step": 18695
    },
    {
      "entropy": 5.3539416790008545,
      "epoch": 2.100185320379626,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045658888382797886,
      "loss": 5.1332,
      "mean_token_accuracy": 0.2042943060398102,
      "num_tokens": 41725290.0,
      "step": 18700
    },
    {
      "entropy": 5.447580146789551,
      "epoch": 2.1007468972875833,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045656517443925897,
      "loss": 5.1617,
      "mean_token_accuracy": 0.2024991199374199,
      "num_tokens": 41736265.0,
      "step": 18705
    },
    {
      "entropy": 5.4621092796325685,
      "epoch": 2.101308474195541,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004565414592694336,
      "loss": 5.1781,
      "mean_token_accuracy": 0.19953703880310059,
      "num_tokens": 41746702.0,
      "step": 18710
    },
    {
      "entropy": 5.5182952880859375,
      "epoch": 2.1018700511034987,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004565177383192577,
      "loss": 5.2713,
      "mean_token_accuracy": 0.1989699497818947,
      "num_tokens": 41758570.0,
      "step": 18715
    },
    {
      "entropy": 5.464882516860962,
      "epoch": 2.1024316280114563,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045649401158948664,
      "loss": 5.179,
      "mean_token_accuracy": 0.21007685512304305,
      "num_tokens": 41769507.0,
      "step": 18720
    },
    {
      "entropy": 5.404947710037232,
      "epoch": 2.1029932049194136,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004564702790808758,
      "loss": 5.143,
      "mean_token_accuracy": 0.20690559446811677,
      "num_tokens": 41780558.0,
      "step": 18725
    },
    {
      "entropy": 5.3212165355682375,
      "epoch": 2.103554781827371,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004564465407941809,
      "loss": 5.0494,
      "mean_token_accuracy": 0.2096721574664116,
      "num_tokens": 41790518.0,
      "step": 18730
    },
    {
      "entropy": 5.440644884109497,
      "epoch": 2.104116358735329,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045642279673015787,
      "loss": 5.1727,
      "mean_token_accuracy": 0.19717319011688234,
      "num_tokens": 41800329.0,
      "step": 18735
    },
    {
      "entropy": 5.401043462753296,
      "epoch": 2.1046779356432865,
      "grad_norm": 1.125,
      "learning_rate": 0.0004563990468895625,
      "loss": 5.0108,
      "mean_token_accuracy": 0.20710598081350326,
      "num_tokens": 41811231.0,
      "step": 18740
    },
    {
      "entropy": 5.295556211471558,
      "epoch": 2.1052395125512438,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004563752912731512,
      "loss": 5.072,
      "mean_token_accuracy": 0.20868365466594696,
      "num_tokens": 41821544.0,
      "step": 18745
    },
    {
      "entropy": 5.293239450454712,
      "epoch": 2.1058010894592014,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004563515298816803,
      "loss": 5.1047,
      "mean_token_accuracy": 0.20148597657680511,
      "num_tokens": 41832371.0,
      "step": 18750
    },
    {
      "entropy": 5.447806406021118,
      "epoch": 2.106362666367159,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045632776271590624,
      "loss": 5.0908,
      "mean_token_accuracy": 0.20789908319711686,
      "num_tokens": 41842828.0,
      "step": 18755
    },
    {
      "entropy": 5.283441734313965,
      "epoch": 2.1069242432751167,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004563039897765859,
      "loss": 5.0557,
      "mean_token_accuracy": 0.2047991618514061,
      "num_tokens": 41855128.0,
      "step": 18760
    },
    {
      "entropy": 5.389542484283448,
      "epoch": 2.107485820183074,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004562802110644764,
      "loss": 5.1344,
      "mean_token_accuracy": 0.19791516214609145,
      "num_tokens": 41866258.0,
      "step": 18765
    },
    {
      "entropy": 5.373833322525025,
      "epoch": 2.1080473970910316,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045625642658033447,
      "loss": 5.126,
      "mean_token_accuracy": 0.2001771554350853,
      "num_tokens": 41877713.0,
      "step": 18770
    },
    {
      "entropy": 5.426022863388061,
      "epoch": 2.1086089739989893,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045623263632491766,
      "loss": 5.1172,
      "mean_token_accuracy": 0.2046523153781891,
      "num_tokens": 41888592.0,
      "step": 18775
    },
    {
      "entropy": 5.426513242721557,
      "epoch": 2.1091705509069465,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045620884029898344,
      "loss": 5.2087,
      "mean_token_accuracy": 0.1967288762331009,
      "num_tokens": 41901024.0,
      "step": 18780
    },
    {
      "entropy": 5.3880571842193605,
      "epoch": 2.109732127814904,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045618503850328954,
      "loss": 5.1306,
      "mean_token_accuracy": 0.199244923889637,
      "num_tokens": 41911184.0,
      "step": 18785
    },
    {
      "entropy": 5.416010951995849,
      "epoch": 2.110293704722862,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004561612309385936,
      "loss": 5.1656,
      "mean_token_accuracy": 0.19776888489723204,
      "num_tokens": 41922799.0,
      "step": 18790
    },
    {
      "entropy": 5.434871339797974,
      "epoch": 2.1108552816308195,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045613741760565385,
      "loss": 5.1385,
      "mean_token_accuracy": 0.2110702335834503,
      "num_tokens": 41933223.0,
      "step": 18795
    },
    {
      "entropy": 5.3352141857147215,
      "epoch": 2.1114168585387767,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004561135985052285,
      "loss": 5.1482,
      "mean_token_accuracy": 0.20093337446451187,
      "num_tokens": 41944288.0,
      "step": 18800
    },
    {
      "entropy": 5.357125091552734,
      "epoch": 2.1119784354467344,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000456089773638076,
      "loss": 5.0921,
      "mean_token_accuracy": 0.20546911805868148,
      "num_tokens": 41954312.0,
      "step": 18805
    },
    {
      "entropy": 5.411005353927612,
      "epoch": 2.112540012354692,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004560659430049548,
      "loss": 5.1729,
      "mean_token_accuracy": 0.19608519226312637,
      "num_tokens": 41966871.0,
      "step": 18810
    },
    {
      "entropy": 5.387155294418335,
      "epoch": 2.1131015892626497,
      "grad_norm": 1.25,
      "learning_rate": 0.0004560421066066237,
      "loss": 5.0646,
      "mean_token_accuracy": 0.20877329409122466,
      "num_tokens": 41979087.0,
      "step": 18815
    },
    {
      "entropy": 5.439654970169068,
      "epoch": 2.113663166170607,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045601826444384175,
      "loss": 5.1295,
      "mean_token_accuracy": 0.19994852989912032,
      "num_tokens": 41989867.0,
      "step": 18820
    },
    {
      "entropy": 5.366193866729736,
      "epoch": 2.1142247430785646,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045599441651736805,
      "loss": 5.1056,
      "mean_token_accuracy": 0.2086072400212288,
      "num_tokens": 42001205.0,
      "step": 18825
    },
    {
      "entropy": 5.464731311798095,
      "epoch": 2.1147863199865222,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045597056282796196,
      "loss": 5.2419,
      "mean_token_accuracy": 0.19247453957796096,
      "num_tokens": 42012466.0,
      "step": 18830
    },
    {
      "entropy": 5.222427606582642,
      "epoch": 2.11534789689448,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004559467033763829,
      "loss": 4.9014,
      "mean_token_accuracy": 0.21156173795461655,
      "num_tokens": 42022517.0,
      "step": 18835
    },
    {
      "entropy": 5.42097749710083,
      "epoch": 2.115909473802437,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004559228381633907,
      "loss": 5.0584,
      "mean_token_accuracy": 0.2108115941286087,
      "num_tokens": 42032253.0,
      "step": 18840
    },
    {
      "entropy": 5.4083030223846436,
      "epoch": 2.1164710507103948,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045589896718974517,
      "loss": 5.0862,
      "mean_token_accuracy": 0.20436487793922425,
      "num_tokens": 42042943.0,
      "step": 18845
    },
    {
      "entropy": 5.344696998596191,
      "epoch": 2.1170326276183524,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004558750904562063,
      "loss": 5.0361,
      "mean_token_accuracy": 0.21074390411376953,
      "num_tokens": 42054302.0,
      "step": 18850
    },
    {
      "entropy": 5.358214044570923,
      "epoch": 2.11759420452631,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004558512079635345,
      "loss": 5.0843,
      "mean_token_accuracy": 0.19336277097463608,
      "num_tokens": 42064810.0,
      "step": 18855
    },
    {
      "entropy": 5.371273326873779,
      "epoch": 2.1181557814342673,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004558273197124899,
      "loss": 5.1483,
      "mean_token_accuracy": 0.19439695179462432,
      "num_tokens": 42076468.0,
      "step": 18860
    },
    {
      "entropy": 5.352231407165528,
      "epoch": 2.118717358342225,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045580342570383347,
      "loss": 5.0833,
      "mean_token_accuracy": 0.20009465962648393,
      "num_tokens": 42087488.0,
      "step": 18865
    },
    {
      "entropy": 5.364237356185913,
      "epoch": 2.1192789352501826,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004557795259383258,
      "loss": 5.1767,
      "mean_token_accuracy": 0.19305503815412522,
      "num_tokens": 42099992.0,
      "step": 18870
    },
    {
      "entropy": 5.456837511062622,
      "epoch": 2.11984051215814,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045575562041672794,
      "loss": 5.1509,
      "mean_token_accuracy": 0.1969546839594841,
      "num_tokens": 42111073.0,
      "step": 18875
    },
    {
      "entropy": 5.380134105682373,
      "epoch": 2.1204020890660975,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045573170913980094,
      "loss": 5.1226,
      "mean_token_accuracy": 0.19720031768083573,
      "num_tokens": 42121413.0,
      "step": 18880
    },
    {
      "entropy": 5.458652353286743,
      "epoch": 2.120963665974055,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045570779210830627,
      "loss": 5.2,
      "mean_token_accuracy": 0.19613172113895416,
      "num_tokens": 42132239.0,
      "step": 18885
    },
    {
      "entropy": 5.393742942810059,
      "epoch": 2.121525242882013,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004556838693230054,
      "loss": 5.1492,
      "mean_token_accuracy": 0.20565389096736908,
      "num_tokens": 42143124.0,
      "step": 18890
    },
    {
      "entropy": 5.466341018676758,
      "epoch": 2.12208681978997,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045565994078466005,
      "loss": 5.1753,
      "mean_token_accuracy": 0.1938370645046234,
      "num_tokens": 42152653.0,
      "step": 18895
    },
    {
      "entropy": 5.494038534164429,
      "epoch": 2.1226483966979277,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045563600649403213,
      "loss": 5.181,
      "mean_token_accuracy": 0.20229168534278869,
      "num_tokens": 42162734.0,
      "step": 18900
    },
    {
      "entropy": 5.443363285064697,
      "epoch": 2.1232099736058854,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004556120664518838,
      "loss": 5.2394,
      "mean_token_accuracy": 0.19365810006856918,
      "num_tokens": 42175312.0,
      "step": 18905
    },
    {
      "entropy": 5.406123542785645,
      "epoch": 2.123771550513843,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004555881206589771,
      "loss": 5.1556,
      "mean_token_accuracy": 0.20204721987247468,
      "num_tokens": 42186512.0,
      "step": 18910
    },
    {
      "entropy": 5.377271413803101,
      "epoch": 2.1243331274218002,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004555641691160746,
      "loss": 5.192,
      "mean_token_accuracy": 0.19963553547859192,
      "num_tokens": 42197698.0,
      "step": 18915
    },
    {
      "entropy": 5.389962387084961,
      "epoch": 2.124894704329758,
      "grad_norm": 1.078125,
      "learning_rate": 0.000455540211823939,
      "loss": 5.1355,
      "mean_token_accuracy": 0.20783817619085312,
      "num_tokens": 42208761.0,
      "step": 18920
    },
    {
      "entropy": 5.419844484329223,
      "epoch": 2.1254562812377156,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045551624878333296,
      "loss": 5.1222,
      "mean_token_accuracy": 0.20362516343593598,
      "num_tokens": 42218920.0,
      "step": 18925
    },
    {
      "entropy": 5.415589046478272,
      "epoch": 2.1260178581456732,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045549227999501957,
      "loss": 5.1196,
      "mean_token_accuracy": 0.20875956267118453,
      "num_tokens": 42229819.0,
      "step": 18930
    },
    {
      "entropy": 5.335864543914795,
      "epoch": 2.1265794350536305,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045546830545976203,
      "loss": 5.1016,
      "mean_token_accuracy": 0.19718735218048095,
      "num_tokens": 42241189.0,
      "step": 18935
    },
    {
      "entropy": 5.361591005325318,
      "epoch": 2.127141011961588,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045544432517832356,
      "loss": 5.0464,
      "mean_token_accuracy": 0.2062685325741768,
      "num_tokens": 42252665.0,
      "step": 18940
    },
    {
      "entropy": 5.353673028945923,
      "epoch": 2.1277025888695458,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045542033915146787,
      "loss": 5.095,
      "mean_token_accuracy": 0.20272819995880126,
      "num_tokens": 42264514.0,
      "step": 18945
    },
    {
      "entropy": 5.426000785827637,
      "epoch": 2.1282641657775034,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045539634737995857,
      "loss": 5.1503,
      "mean_token_accuracy": 0.20168357491493225,
      "num_tokens": 42275229.0,
      "step": 18950
    },
    {
      "entropy": 5.4444849491119385,
      "epoch": 2.1288257426854607,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045537234986455964,
      "loss": 5.1347,
      "mean_token_accuracy": 0.2045433282852173,
      "num_tokens": 42286819.0,
      "step": 18955
    },
    {
      "entropy": 5.447703790664673,
      "epoch": 2.1293873195934183,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004553483466060351,
      "loss": 5.1708,
      "mean_token_accuracy": 0.2027878612279892,
      "num_tokens": 42298329.0,
      "step": 18960
    },
    {
      "entropy": 5.287286806106567,
      "epoch": 2.129948896501376,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045532433760514925,
      "loss": 5.027,
      "mean_token_accuracy": 0.20578677654266359,
      "num_tokens": 42308813.0,
      "step": 18965
    },
    {
      "entropy": 5.410430002212524,
      "epoch": 2.130510473409333,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004553003228626665,
      "loss": 5.1692,
      "mean_token_accuracy": 0.2024761453270912,
      "num_tokens": 42321264.0,
      "step": 18970
    },
    {
      "entropy": 5.336766719818115,
      "epoch": 2.131072050317291,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004552763023793517,
      "loss": 5.0565,
      "mean_token_accuracy": 0.1974918559193611,
      "num_tokens": 42332301.0,
      "step": 18975
    },
    {
      "entropy": 5.380697441101074,
      "epoch": 2.1316336272252485,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004552522761559694,
      "loss": 5.0335,
      "mean_token_accuracy": 0.20426950007677078,
      "num_tokens": 42343424.0,
      "step": 18980
    },
    {
      "entropy": 5.415702247619629,
      "epoch": 2.132195204133206,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045522824419328475,
      "loss": 5.1226,
      "mean_token_accuracy": 0.20281576067209245,
      "num_tokens": 42354350.0,
      "step": 18985
    },
    {
      "entropy": 5.341759061813354,
      "epoch": 2.1327567810411634,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045520420649206283,
      "loss": 5.0479,
      "mean_token_accuracy": 0.20580150485038756,
      "num_tokens": 42365110.0,
      "step": 18990
    },
    {
      "entropy": 5.40932068824768,
      "epoch": 2.133318357949121,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004551801630530692,
      "loss": 5.1846,
      "mean_token_accuracy": 0.19945698231458664,
      "num_tokens": 42376345.0,
      "step": 18995
    },
    {
      "entropy": 5.43607177734375,
      "epoch": 2.1338799348570787,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004551561138770693,
      "loss": 5.0391,
      "mean_token_accuracy": 0.20434303730726242,
      "num_tokens": 42387161.0,
      "step": 19000
    },
    {
      "entropy": 5.410964441299439,
      "epoch": 2.1344415117650364,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045513205896482875,
      "loss": 5.1059,
      "mean_token_accuracy": 0.20268671065568925,
      "num_tokens": 42397468.0,
      "step": 19005
    },
    {
      "entropy": 5.345913505554199,
      "epoch": 2.1350030886729936,
      "grad_norm": 1.125,
      "learning_rate": 0.00045510799831711366,
      "loss": 5.1226,
      "mean_token_accuracy": 0.2035314604640007,
      "num_tokens": 42408547.0,
      "step": 19010
    },
    {
      "entropy": 5.355752801895141,
      "epoch": 2.1355646655809513,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045508393193469,
      "loss": 5.0346,
      "mean_token_accuracy": 0.205076265335083,
      "num_tokens": 42420383.0,
      "step": 19015
    },
    {
      "entropy": 5.429932260513306,
      "epoch": 2.136126242488909,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004550598598183241,
      "loss": 5.1696,
      "mean_token_accuracy": 0.20086348950862884,
      "num_tokens": 42431723.0,
      "step": 19020
    },
    {
      "entropy": 5.424737405776978,
      "epoch": 2.1366878193968666,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004550357819687825,
      "loss": 5.1595,
      "mean_token_accuracy": 0.20134229809045792,
      "num_tokens": 42443740.0,
      "step": 19025
    },
    {
      "entropy": 5.314010238647461,
      "epoch": 2.137249396304824,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004550116983868317,
      "loss": 5.0845,
      "mean_token_accuracy": 0.20979016572237014,
      "num_tokens": 42454288.0,
      "step": 19030
    },
    {
      "entropy": 5.365576124191284,
      "epoch": 2.1378109732127815,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004549876090732386,
      "loss": 5.1171,
      "mean_token_accuracy": 0.20077276080846787,
      "num_tokens": 42467576.0,
      "step": 19035
    },
    {
      "entropy": 5.405465793609619,
      "epoch": 2.138372550120739,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004549635140287703,
      "loss": 5.0848,
      "mean_token_accuracy": 0.20437936037778853,
      "num_tokens": 42477839.0,
      "step": 19040
    },
    {
      "entropy": 5.348209714889526,
      "epoch": 2.138934127028697,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045493941325419385,
      "loss": 5.039,
      "mean_token_accuracy": 0.20213165134191513,
      "num_tokens": 42489875.0,
      "step": 19045
    },
    {
      "entropy": 5.4554276943206785,
      "epoch": 2.139495703936654,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004549153067502768,
      "loss": 5.1998,
      "mean_token_accuracy": 0.19841014742851257,
      "num_tokens": 42500946.0,
      "step": 19050
    },
    {
      "entropy": 5.384822940826416,
      "epoch": 2.1400572808446117,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045489119451778645,
      "loss": 5.0579,
      "mean_token_accuracy": 0.21067678779363633,
      "num_tokens": 42511503.0,
      "step": 19055
    },
    {
      "entropy": 5.521847295761108,
      "epoch": 2.1406188577525693,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004548670765574908,
      "loss": 5.2534,
      "mean_token_accuracy": 0.19207195043563843,
      "num_tokens": 42523436.0,
      "step": 19060
    },
    {
      "entropy": 5.384925651550293,
      "epoch": 2.1411804346605265,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045484295287015757,
      "loss": 5.1487,
      "mean_token_accuracy": 0.199361415207386,
      "num_tokens": 42534777.0,
      "step": 19065
    },
    {
      "entropy": 5.420631217956543,
      "epoch": 2.141742011568484,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004548188234565551,
      "loss": 5.1259,
      "mean_token_accuracy": 0.20389993190765382,
      "num_tokens": 42545803.0,
      "step": 19070
    },
    {
      "entropy": 5.4895610332489015,
      "epoch": 2.142303588476442,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045479468831745144,
      "loss": 5.1535,
      "mean_token_accuracy": 0.20183493345975875,
      "num_tokens": 42556803.0,
      "step": 19075
    },
    {
      "entropy": 5.461218690872192,
      "epoch": 2.1428651653843995,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004547705474536152,
      "loss": 5.2094,
      "mean_token_accuracy": 0.19324614256620407,
      "num_tokens": 42568357.0,
      "step": 19080
    },
    {
      "entropy": 5.36745777130127,
      "epoch": 2.1434267422923567,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045474640086581506,
      "loss": 5.0999,
      "mean_token_accuracy": 0.20423174053430557,
      "num_tokens": 42578990.0,
      "step": 19085
    },
    {
      "entropy": 5.431365633010865,
      "epoch": 2.1439883192003144,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045472224855481975,
      "loss": 5.1421,
      "mean_token_accuracy": 0.20195329189300537,
      "num_tokens": 42589823.0,
      "step": 19090
    },
    {
      "entropy": 5.378989315032959,
      "epoch": 2.144549896108272,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045469809052139844,
      "loss": 5.0853,
      "mean_token_accuracy": 0.20230145156383514,
      "num_tokens": 42600508.0,
      "step": 19095
    },
    {
      "entropy": 5.365769100189209,
      "epoch": 2.1451114730162297,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045467392676632016,
      "loss": 5.1407,
      "mean_token_accuracy": 0.1998974487185478,
      "num_tokens": 42611061.0,
      "step": 19100
    },
    {
      "entropy": 5.421975898742676,
      "epoch": 2.145673049924187,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004546497572903543,
      "loss": 5.0856,
      "mean_token_accuracy": 0.20445285737514496,
      "num_tokens": 42621936.0,
      "step": 19105
    },
    {
      "entropy": 5.372865772247314,
      "epoch": 2.1462346268321446,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004546255820942706,
      "loss": 5.1703,
      "mean_token_accuracy": 0.20332351177930832,
      "num_tokens": 42632880.0,
      "step": 19110
    },
    {
      "entropy": 5.392221403121948,
      "epoch": 2.1467962037401023,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004546014011788386,
      "loss": 5.0397,
      "mean_token_accuracy": 0.20965463668107986,
      "num_tokens": 42642819.0,
      "step": 19115
    },
    {
      "entropy": 5.390427970886231,
      "epoch": 2.14735778064806,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004545772145448284,
      "loss": 5.154,
      "mean_token_accuracy": 0.20142435282468796,
      "num_tokens": 42654581.0,
      "step": 19120
    },
    {
      "entropy": 5.395322847366333,
      "epoch": 2.147919357556017,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045455302219301,
      "loss": 5.2033,
      "mean_token_accuracy": 0.20176073014736176,
      "num_tokens": 42665323.0,
      "step": 19125
    },
    {
      "entropy": 5.497414922714233,
      "epoch": 2.148480934463975,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004545288241241538,
      "loss": 5.171,
      "mean_token_accuracy": 0.19757115244865417,
      "num_tokens": 42676465.0,
      "step": 19130
    },
    {
      "entropy": 5.4035888671875,
      "epoch": 2.1490425113719325,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004545046203390302,
      "loss": 5.1797,
      "mean_token_accuracy": 0.19823136627674104,
      "num_tokens": 42688242.0,
      "step": 19135
    },
    {
      "entropy": 5.383864641189575,
      "epoch": 2.14960408827989,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045448041083840973,
      "loss": 5.1592,
      "mean_token_accuracy": 0.20369527041912078,
      "num_tokens": 42699498.0,
      "step": 19140
    },
    {
      "entropy": 5.4497510433197025,
      "epoch": 2.1501656651878474,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045445619562306354,
      "loss": 5.1422,
      "mean_token_accuracy": 0.20674261897802354,
      "num_tokens": 42710330.0,
      "step": 19145
    },
    {
      "entropy": 5.416684484481811,
      "epoch": 2.150727242095805,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045443197469376243,
      "loss": 5.1457,
      "mean_token_accuracy": 0.20414287596940994,
      "num_tokens": 42721210.0,
      "step": 19150
    },
    {
      "entropy": 5.413285541534424,
      "epoch": 2.1512888190037627,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004544077480512776,
      "loss": 5.1411,
      "mean_token_accuracy": 0.19951637834310532,
      "num_tokens": 42732272.0,
      "step": 19155
    },
    {
      "entropy": 5.543734550476074,
      "epoch": 2.15185039591172,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045438351569638055,
      "loss": 5.2718,
      "mean_token_accuracy": 0.19083299040794371,
      "num_tokens": 42745737.0,
      "step": 19160
    },
    {
      "entropy": 5.439539003372192,
      "epoch": 2.1524119728196776,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045435927762984277,
      "loss": 5.1486,
      "mean_token_accuracy": 0.20136955827474595,
      "num_tokens": 42756405.0,
      "step": 19165
    },
    {
      "entropy": 5.418019151687622,
      "epoch": 2.152973549727635,
      "grad_norm": 1.09375,
      "learning_rate": 0.000454335033852436,
      "loss": 5.1707,
      "mean_token_accuracy": 0.19064721912145616,
      "num_tokens": 42767991.0,
      "step": 19170
    },
    {
      "entropy": 5.3960779190063475,
      "epoch": 2.153535126635593,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004543107843649322,
      "loss": 5.2162,
      "mean_token_accuracy": 0.19310415387153626,
      "num_tokens": 42779177.0,
      "step": 19175
    },
    {
      "entropy": 5.412387561798096,
      "epoch": 2.15409670354355,
      "grad_norm": 1.125,
      "learning_rate": 0.0004542865291681035,
      "loss": 5.1176,
      "mean_token_accuracy": 0.20465805977582932,
      "num_tokens": 42789748.0,
      "step": 19180
    },
    {
      "entropy": 5.493113088607788,
      "epoch": 2.1546582804515078,
      "grad_norm": 1.125,
      "learning_rate": 0.00045426226826272213,
      "loss": 5.159,
      "mean_token_accuracy": 0.19915618598461152,
      "num_tokens": 42802741.0,
      "step": 19185
    },
    {
      "entropy": 5.407629680633545,
      "epoch": 2.1552198573594654,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004542380016495607,
      "loss": 5.16,
      "mean_token_accuracy": 0.1962450161576271,
      "num_tokens": 42813091.0,
      "step": 19190
    },
    {
      "entropy": 5.430349349975586,
      "epoch": 2.155781434267423,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045421372932939177,
      "loss": 5.1491,
      "mean_token_accuracy": 0.20216340720653533,
      "num_tokens": 42823711.0,
      "step": 19195
    },
    {
      "entropy": 5.338369131088257,
      "epoch": 2.1563430111753803,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004541894513029881,
      "loss": 5.095,
      "mean_token_accuracy": 0.19413064420223236,
      "num_tokens": 42835057.0,
      "step": 19200
    },
    {
      "entropy": 5.509865665435791,
      "epoch": 2.156904588083338,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045416516757112286,
      "loss": 5.1933,
      "mean_token_accuracy": 0.20029471665620804,
      "num_tokens": 42846784.0,
      "step": 19205
    },
    {
      "entropy": 5.404611492156983,
      "epoch": 2.1574661649912956,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045414087813456925,
      "loss": 5.0837,
      "mean_token_accuracy": 0.2053150162100792,
      "num_tokens": 42857618.0,
      "step": 19210
    },
    {
      "entropy": 5.366456174850464,
      "epoch": 2.1580277418992533,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004541165829941006,
      "loss": 5.1442,
      "mean_token_accuracy": 0.20269455164670944,
      "num_tokens": 42869076.0,
      "step": 19215
    },
    {
      "entropy": 5.356043195724487,
      "epoch": 2.1585893188072105,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004540922821504904,
      "loss": 5.1204,
      "mean_token_accuracy": 0.20753590166568756,
      "num_tokens": 42879714.0,
      "step": 19220
    },
    {
      "entropy": 5.461410188674927,
      "epoch": 2.159150895715168,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004540679756045125,
      "loss": 5.1245,
      "mean_token_accuracy": 0.20590616762638092,
      "num_tokens": 42890187.0,
      "step": 19225
    },
    {
      "entropy": 5.363470935821534,
      "epoch": 2.159712472623126,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004540436633569407,
      "loss": 5.0748,
      "mean_token_accuracy": 0.20441191047430038,
      "num_tokens": 42900713.0,
      "step": 19230
    },
    {
      "entropy": 5.358960247039795,
      "epoch": 2.1602740495310835,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045401934540854937,
      "loss": 5.135,
      "mean_token_accuracy": 0.19577291011810302,
      "num_tokens": 42911703.0,
      "step": 19235
    },
    {
      "entropy": 5.367685794830322,
      "epoch": 2.1608356264390407,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004539950217601127,
      "loss": 5.0835,
      "mean_token_accuracy": 0.20934465229511262,
      "num_tokens": 42922684.0,
      "step": 19240
    },
    {
      "entropy": 5.396376037597657,
      "epoch": 2.1613972033469984,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045397069241240497,
      "loss": 5.1013,
      "mean_token_accuracy": 0.19645735919475554,
      "num_tokens": 42933095.0,
      "step": 19245
    },
    {
      "entropy": 5.412038707733155,
      "epoch": 2.161958780254956,
      "grad_norm": 1.09375,
      "learning_rate": 0.000453946357366201,
      "loss": 5.2147,
      "mean_token_accuracy": 0.20407461374998093,
      "num_tokens": 42944508.0,
      "step": 19250
    },
    {
      "entropy": 5.378571796417236,
      "epoch": 2.1625203571629132,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045392201662227565,
      "loss": 5.0713,
      "mean_token_accuracy": 0.2112127274274826,
      "num_tokens": 42954902.0,
      "step": 19255
    },
    {
      "entropy": 5.396368026733398,
      "epoch": 2.163081934070871,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045389767018140395,
      "loss": 5.0555,
      "mean_token_accuracy": 0.20733409225940705,
      "num_tokens": 42965165.0,
      "step": 19260
    },
    {
      "entropy": 5.370920562744141,
      "epoch": 2.1636435109788286,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045387331804436096,
      "loss": 5.0787,
      "mean_token_accuracy": 0.20907517820596694,
      "num_tokens": 42975025.0,
      "step": 19265
    },
    {
      "entropy": 5.467773723602295,
      "epoch": 2.1642050878867862,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004538489602119222,
      "loss": 5.2331,
      "mean_token_accuracy": 0.1992476612329483,
      "num_tokens": 42986666.0,
      "step": 19270
    },
    {
      "entropy": 5.411650943756103,
      "epoch": 2.1647666647947434,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004538245966848632,
      "loss": 5.176,
      "mean_token_accuracy": 0.19695753157138823,
      "num_tokens": 42998607.0,
      "step": 19275
    },
    {
      "entropy": 5.36742057800293,
      "epoch": 2.165328241702701,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045380022746395953,
      "loss": 5.1227,
      "mean_token_accuracy": 0.19864431023597717,
      "num_tokens": 43009969.0,
      "step": 19280
    },
    {
      "entropy": 5.5514952659606935,
      "epoch": 2.1658898186106588,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045377585254998743,
      "loss": 5.2689,
      "mean_token_accuracy": 0.19674205482006074,
      "num_tokens": 43023255.0,
      "step": 19285
    },
    {
      "entropy": 5.43598427772522,
      "epoch": 2.1664513955186164,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004537514719437228,
      "loss": 5.1145,
      "mean_token_accuracy": 0.1997988909482956,
      "num_tokens": 43033995.0,
      "step": 19290
    },
    {
      "entropy": 5.369700574874878,
      "epoch": 2.1670129724265736,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045372708564594196,
      "loss": 5.2208,
      "mean_token_accuracy": 0.1962608888745308,
      "num_tokens": 43046239.0,
      "step": 19295
    },
    {
      "entropy": 5.395691442489624,
      "epoch": 2.1675745493345313,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045370269365742143,
      "loss": 5.09,
      "mean_token_accuracy": 0.20518869459629058,
      "num_tokens": 43057262.0,
      "step": 19300
    },
    {
      "entropy": 5.4300378322601315,
      "epoch": 2.168136126242489,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045367829597893776,
      "loss": 5.1992,
      "mean_token_accuracy": 0.1962330162525177,
      "num_tokens": 43068495.0,
      "step": 19305
    },
    {
      "entropy": 5.37553186416626,
      "epoch": 2.1686977031504466,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045365389261126793,
      "loss": 5.0747,
      "mean_token_accuracy": 0.20501625686883926,
      "num_tokens": 43080706.0,
      "step": 19310
    },
    {
      "entropy": 5.28943018913269,
      "epoch": 2.169259280058404,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045362948355518883,
      "loss": 5.0816,
      "mean_token_accuracy": 0.21123532205820084,
      "num_tokens": 43091444.0,
      "step": 19315
    },
    {
      "entropy": 5.4722900390625,
      "epoch": 2.1698208569663615,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004536050688114777,
      "loss": 5.1701,
      "mean_token_accuracy": 0.2019159823656082,
      "num_tokens": 43103648.0,
      "step": 19320
    },
    {
      "entropy": 5.448609638214111,
      "epoch": 2.170382433874319,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045358064838091187,
      "loss": 5.1646,
      "mean_token_accuracy": 0.20055057108402252,
      "num_tokens": 43116183.0,
      "step": 19325
    },
    {
      "entropy": 5.390702629089356,
      "epoch": 2.170944010782277,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045355622226426903,
      "loss": 5.1554,
      "mean_token_accuracy": 0.20012110769748687,
      "num_tokens": 43126557.0,
      "step": 19330
    },
    {
      "entropy": 5.3991485118865965,
      "epoch": 2.171505587690234,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045353179046232674,
      "loss": 5.1504,
      "mean_token_accuracy": 0.2000217467546463,
      "num_tokens": 43137753.0,
      "step": 19335
    },
    {
      "entropy": 5.435703086853027,
      "epoch": 2.1720671645981917,
      "grad_norm": 0.984375,
      "learning_rate": 0.000453507352975863,
      "loss": 5.1681,
      "mean_token_accuracy": 0.20302040576934816,
      "num_tokens": 43148461.0,
      "step": 19340
    },
    {
      "entropy": 5.346978950500488,
      "epoch": 2.1726287415061494,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045348290980565595,
      "loss": 5.0798,
      "mean_token_accuracy": 0.2033142164349556,
      "num_tokens": 43161174.0,
      "step": 19345
    },
    {
      "entropy": 5.45029649734497,
      "epoch": 2.1731903184141066,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045345846095248376,
      "loss": 5.1909,
      "mean_token_accuracy": 0.1980501741170883,
      "num_tokens": 43172844.0,
      "step": 19350
    },
    {
      "entropy": 5.345496606826782,
      "epoch": 2.1737518953220643,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045343400641712503,
      "loss": 5.0663,
      "mean_token_accuracy": 0.2052164763212204,
      "num_tokens": 43183476.0,
      "step": 19355
    },
    {
      "entropy": 5.387218046188354,
      "epoch": 2.174313472230022,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045340954620035827,
      "loss": 5.1484,
      "mean_token_accuracy": 0.20048276633024215,
      "num_tokens": 43195563.0,
      "step": 19360
    },
    {
      "entropy": 5.395286226272583,
      "epoch": 2.1748750491379796,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045338508030296246,
      "loss": 5.18,
      "mean_token_accuracy": 0.20306676924228667,
      "num_tokens": 43206067.0,
      "step": 19365
    },
    {
      "entropy": 5.449159002304077,
      "epoch": 2.175436626045937,
      "grad_norm": 1.0,
      "learning_rate": 0.00045336060872571646,
      "loss": 5.089,
      "mean_token_accuracy": 0.2044341593980789,
      "num_tokens": 43216912.0,
      "step": 19370
    },
    {
      "entropy": 5.43939437866211,
      "epoch": 2.1759982029538945,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004533361314693994,
      "loss": 5.1383,
      "mean_token_accuracy": 0.199588605761528,
      "num_tokens": 43228746.0,
      "step": 19375
    },
    {
      "entropy": 5.369731807708741,
      "epoch": 2.176559779861852,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045331164853479084,
      "loss": 5.1247,
      "mean_token_accuracy": 0.20367695540189742,
      "num_tokens": 43241228.0,
      "step": 19380
    },
    {
      "entropy": 5.4361451148986815,
      "epoch": 2.17712135676981,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045328715992267024,
      "loss": 5.1984,
      "mean_token_accuracy": 0.20024345070123672,
      "num_tokens": 43252916.0,
      "step": 19385
    },
    {
      "entropy": 5.386648273468017,
      "epoch": 2.177682933677767,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045326266563381734,
      "loss": 5.0147,
      "mean_token_accuracy": 0.2126223102211952,
      "num_tokens": 43264357.0,
      "step": 19390
    },
    {
      "entropy": 5.346816444396973,
      "epoch": 2.1782445105857247,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045323816566901205,
      "loss": 5.1847,
      "mean_token_accuracy": 0.20313182473182678,
      "num_tokens": 43274547.0,
      "step": 19395
    },
    {
      "entropy": 5.315592861175537,
      "epoch": 2.1788060874936823,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004532136600290343,
      "loss": 5.0597,
      "mean_token_accuracy": 0.21028306782245637,
      "num_tokens": 43284225.0,
      "step": 19400
    },
    {
      "entropy": 5.4196244239807125,
      "epoch": 2.17936766440164,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004531891487146646,
      "loss": 5.1304,
      "mean_token_accuracy": 0.20272586047649382,
      "num_tokens": 43295218.0,
      "step": 19405
    },
    {
      "entropy": 5.321388101577758,
      "epoch": 2.179929241309597,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004531646317266833,
      "loss": 5.0909,
      "mean_token_accuracy": 0.20494987666606904,
      "num_tokens": 43305169.0,
      "step": 19410
    },
    {
      "entropy": 5.3746202945709225,
      "epoch": 2.180490818217555,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045314010906587096,
      "loss": 5.0924,
      "mean_token_accuracy": 0.20117725282907487,
      "num_tokens": 43317119.0,
      "step": 19415
    },
    {
      "entropy": 5.433563470840454,
      "epoch": 2.1810523951255125,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045311558073300853,
      "loss": 5.1276,
      "mean_token_accuracy": 0.20612337738275527,
      "num_tokens": 43327800.0,
      "step": 19420
    },
    {
      "entropy": 5.271383762359619,
      "epoch": 2.18161397203347,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004530910467288769,
      "loss": 5.0396,
      "mean_token_accuracy": 0.1986771419644356,
      "num_tokens": 43338607.0,
      "step": 19425
    },
    {
      "entropy": 5.391445589065552,
      "epoch": 2.1821755489414274,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045306650705425727,
      "loss": 5.1494,
      "mean_token_accuracy": 0.20843434929847718,
      "num_tokens": 43350621.0,
      "step": 19430
    },
    {
      "entropy": 5.386351537704468,
      "epoch": 2.182737125849385,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045304196170993093,
      "loss": 5.0992,
      "mean_token_accuracy": 0.20700636506080627,
      "num_tokens": 43360327.0,
      "step": 19435
    },
    {
      "entropy": 5.32601146697998,
      "epoch": 2.1832987027573427,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004530174106966795,
      "loss": 5.1159,
      "mean_token_accuracy": 0.20078652501106262,
      "num_tokens": 43371538.0,
      "step": 19440
    },
    {
      "entropy": 5.407646799087525,
      "epoch": 2.1838602796653,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045299285401528475,
      "loss": 5.1677,
      "mean_token_accuracy": 0.20063209682703018,
      "num_tokens": 43381818.0,
      "step": 19445
    },
    {
      "entropy": 5.399660301208496,
      "epoch": 2.1844218565732576,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004529682916665285,
      "loss": 5.11,
      "mean_token_accuracy": 0.20188453048467636,
      "num_tokens": 43391977.0,
      "step": 19450
    },
    {
      "entropy": 5.454909753799439,
      "epoch": 2.1849834334812153,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045294372365119276,
      "loss": 5.1718,
      "mean_token_accuracy": 0.19754231870174407,
      "num_tokens": 43403657.0,
      "step": 19455
    },
    {
      "entropy": 5.334466934204102,
      "epoch": 2.185545010389173,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004529191499700598,
      "loss": 5.0735,
      "mean_token_accuracy": 0.20237735658884048,
      "num_tokens": 43414325.0,
      "step": 19460
    },
    {
      "entropy": 5.4798441410064695,
      "epoch": 2.18610658729713,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004528945706239122,
      "loss": 5.2474,
      "mean_token_accuracy": 0.19383404403924942,
      "num_tokens": 43425864.0,
      "step": 19465
    },
    {
      "entropy": 5.423947286605835,
      "epoch": 2.186668164205088,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045286998561353236,
      "loss": 5.1792,
      "mean_token_accuracy": 0.19421925246715546,
      "num_tokens": 43437135.0,
      "step": 19470
    },
    {
      "entropy": 5.487023401260376,
      "epoch": 2.1872297411130455,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004528453949397033,
      "loss": 5.1896,
      "mean_token_accuracy": 0.20013256520032882,
      "num_tokens": 43448603.0,
      "step": 19475
    },
    {
      "entropy": 5.349306583404541,
      "epoch": 2.187791318021003,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045282079860320786,
      "loss": 5.0509,
      "mean_token_accuracy": 0.20713204145431519,
      "num_tokens": 43460316.0,
      "step": 19480
    },
    {
      "entropy": 5.323572778701783,
      "epoch": 2.1883528949289603,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045279619660482925,
      "loss": 5.1299,
      "mean_token_accuracy": 0.20601435899734497,
      "num_tokens": 43471869.0,
      "step": 19485
    },
    {
      "entropy": 5.426368856430054,
      "epoch": 2.188914471836918,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045277158894535075,
      "loss": 5.1783,
      "mean_token_accuracy": 0.1994830086827278,
      "num_tokens": 43484004.0,
      "step": 19490
    },
    {
      "entropy": 5.413016414642334,
      "epoch": 2.1894760487448757,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045274697562555584,
      "loss": 5.1586,
      "mean_token_accuracy": 0.1984018787741661,
      "num_tokens": 43494401.0,
      "step": 19495
    },
    {
      "entropy": 5.301991176605225,
      "epoch": 2.1900376256528333,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004527223566462284,
      "loss": 5.09,
      "mean_token_accuracy": 0.1981787219643593,
      "num_tokens": 43505437.0,
      "step": 19500
    },
    {
      "entropy": 5.365578508377075,
      "epoch": 2.1905992025607905,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004526977320081522,
      "loss": 5.0407,
      "mean_token_accuracy": 0.21004265993833543,
      "num_tokens": 43516177.0,
      "step": 19505
    },
    {
      "entropy": 5.432656574249267,
      "epoch": 2.191160779468748,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004526731017121111,
      "loss": 5.1392,
      "mean_token_accuracy": 0.2055959939956665,
      "num_tokens": 43527070.0,
      "step": 19510
    },
    {
      "entropy": 5.359364557266235,
      "epoch": 2.191722356376706,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045264846575888973,
      "loss": 5.1544,
      "mean_token_accuracy": 0.19991910010576247,
      "num_tokens": 43538247.0,
      "step": 19515
    },
    {
      "entropy": 5.420795631408692,
      "epoch": 2.1922839332846635,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045262382414927235,
      "loss": 5.1021,
      "mean_token_accuracy": 0.21089170575141908,
      "num_tokens": 43549994.0,
      "step": 19520
    },
    {
      "entropy": 5.4647503852844235,
      "epoch": 2.1928455101926208,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045259917688404336,
      "loss": 5.1956,
      "mean_token_accuracy": 0.1985134318470955,
      "num_tokens": 43562171.0,
      "step": 19525
    },
    {
      "entropy": 5.424193620681763,
      "epoch": 2.1934070871005784,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004525745239639877,
      "loss": 5.1083,
      "mean_token_accuracy": 0.19992561191320418,
      "num_tokens": 43572308.0,
      "step": 19530
    },
    {
      "entropy": 5.364195060729981,
      "epoch": 2.193968664008536,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045254986538989045,
      "loss": 5.1041,
      "mean_token_accuracy": 0.20344994813203812,
      "num_tokens": 43583565.0,
      "step": 19535
    },
    {
      "entropy": 5.361726331710815,
      "epoch": 2.1945302409164933,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004525252011625366,
      "loss": 5.2252,
      "mean_token_accuracy": 0.19757308214902877,
      "num_tokens": 43595921.0,
      "step": 19540
    },
    {
      "entropy": 5.458091449737549,
      "epoch": 2.195091817824451,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004525005312827114,
      "loss": 5.0942,
      "mean_token_accuracy": 0.20407532155513763,
      "num_tokens": 43606957.0,
      "step": 19545
    },
    {
      "entropy": 5.371614551544189,
      "epoch": 2.1956533947324086,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004524758557512005,
      "loss": 5.0401,
      "mean_token_accuracy": 0.21186549365520477,
      "num_tokens": 43616983.0,
      "step": 19550
    },
    {
      "entropy": 5.2470582008361815,
      "epoch": 2.1962149716403663,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004524511745687895,
      "loss": 4.995,
      "mean_token_accuracy": 0.2115111380815506,
      "num_tokens": 43626951.0,
      "step": 19555
    },
    {
      "entropy": 5.341251611709595,
      "epoch": 2.1967765485483235,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004524264877362643,
      "loss": 5.1033,
      "mean_token_accuracy": 0.1985931158065796,
      "num_tokens": 43638446.0,
      "step": 19560
    },
    {
      "entropy": 5.453473472595215,
      "epoch": 2.197338125456281,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045240179525441094,
      "loss": 5.1741,
      "mean_token_accuracy": 0.20073432773351668,
      "num_tokens": 43649646.0,
      "step": 19565
    },
    {
      "entropy": 5.378449392318726,
      "epoch": 2.197899702364239,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004523770971240156,
      "loss": 5.1768,
      "mean_token_accuracy": 0.1989554360508919,
      "num_tokens": 43661633.0,
      "step": 19570
    },
    {
      "entropy": 5.38330717086792,
      "epoch": 2.1984612792721965,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004523523933458647,
      "loss": 5.1379,
      "mean_token_accuracy": 0.20536414831876754,
      "num_tokens": 43672671.0,
      "step": 19575
    },
    {
      "entropy": 5.453365468978882,
      "epoch": 2.1990228561801537,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004523276839207448,
      "loss": 5.1281,
      "mean_token_accuracy": 0.20123444646596908,
      "num_tokens": 43683426.0,
      "step": 19580
    },
    {
      "entropy": 5.4247301578521725,
      "epoch": 2.1995844330881114,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045230296884944264,
      "loss": 5.1854,
      "mean_token_accuracy": 0.1990601196885109,
      "num_tokens": 43694460.0,
      "step": 19585
    },
    {
      "entropy": 5.333220148086548,
      "epoch": 2.200146009996069,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045227824813274517,
      "loss": 4.9807,
      "mean_token_accuracy": 0.2087511032819748,
      "num_tokens": 43706047.0,
      "step": 19590
    },
    {
      "entropy": 5.322297859191894,
      "epoch": 2.2007075869040267,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004522535217714396,
      "loss": 5.0979,
      "mean_token_accuracy": 0.20582454949617385,
      "num_tokens": 43715879.0,
      "step": 19595
    },
    {
      "entropy": 5.375786256790161,
      "epoch": 2.201269163811984,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004522287897663131,
      "loss": 5.1133,
      "mean_token_accuracy": 0.20442139655351638,
      "num_tokens": 43726870.0,
      "step": 19600
    },
    {
      "entropy": 5.385470724105835,
      "epoch": 2.2018307407199416,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045220405211815327,
      "loss": 5.1519,
      "mean_token_accuracy": 0.19960614144802094,
      "num_tokens": 43739116.0,
      "step": 19605
    },
    {
      "entropy": 5.393036413192749,
      "epoch": 2.202392317627899,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045217930882774754,
      "loss": 5.1338,
      "mean_token_accuracy": 0.20119705945253372,
      "num_tokens": 43750239.0,
      "step": 19610
    },
    {
      "entropy": 5.456850671768189,
      "epoch": 2.202953894535857,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045215455989588396,
      "loss": 5.2338,
      "mean_token_accuracy": 0.20052412897348404,
      "num_tokens": 43762391.0,
      "step": 19615
    },
    {
      "entropy": 5.393123579025269,
      "epoch": 2.203515471443814,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045212980532335055,
      "loss": 5.1425,
      "mean_token_accuracy": 0.2066319316625595,
      "num_tokens": 43773514.0,
      "step": 19620
    },
    {
      "entropy": 5.358402299880981,
      "epoch": 2.2040770483517718,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045210504511093534,
      "loss": 5.1039,
      "mean_token_accuracy": 0.19578895717859268,
      "num_tokens": 43784617.0,
      "step": 19625
    },
    {
      "entropy": 5.4838121891021725,
      "epoch": 2.2046386252597294,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004520802792594269,
      "loss": 5.2105,
      "mean_token_accuracy": 0.19914634078741072,
      "num_tokens": 43795396.0,
      "step": 19630
    },
    {
      "entropy": 5.429034328460693,
      "epoch": 2.2052002021676866,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045205550776961366,
      "loss": 5.148,
      "mean_token_accuracy": 0.1986371323466301,
      "num_tokens": 43805512.0,
      "step": 19635
    },
    {
      "entropy": 5.4502854347229,
      "epoch": 2.2057617790756443,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004520307306422844,
      "loss": 5.2626,
      "mean_token_accuracy": 0.1967759147286415,
      "num_tokens": 43815874.0,
      "step": 19640
    },
    {
      "entropy": 5.5178924083709715,
      "epoch": 2.206323355983602,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045200594787822795,
      "loss": 5.1892,
      "mean_token_accuracy": 0.19200937151908876,
      "num_tokens": 43827948.0,
      "step": 19645
    },
    {
      "entropy": 5.382894134521484,
      "epoch": 2.2068849328915596,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004519811594782335,
      "loss": 5.0435,
      "mean_token_accuracy": 0.20796052515506744,
      "num_tokens": 43838306.0,
      "step": 19650
    },
    {
      "entropy": 5.330549144744873,
      "epoch": 2.207446509799517,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045195636544309033,
      "loss": 5.0995,
      "mean_token_accuracy": 0.20548930764198303,
      "num_tokens": 43848716.0,
      "step": 19655
    },
    {
      "entropy": 5.449155664443969,
      "epoch": 2.2080080867074745,
      "grad_norm": 1.125,
      "learning_rate": 0.00045193156577358784,
      "loss": 5.1704,
      "mean_token_accuracy": 0.20312638431787491,
      "num_tokens": 43859707.0,
      "step": 19660
    },
    {
      "entropy": 5.3715626239776615,
      "epoch": 2.208569663615432,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004519067604705157,
      "loss": 5.0452,
      "mean_token_accuracy": 0.211484657227993,
      "num_tokens": 43870599.0,
      "step": 19665
    },
    {
      "entropy": 5.387804841995239,
      "epoch": 2.20913124052339,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045188194953466363,
      "loss": 5.1423,
      "mean_token_accuracy": 0.19970490634441376,
      "num_tokens": 43881843.0,
      "step": 19670
    },
    {
      "entropy": 5.347230863571167,
      "epoch": 2.209692817431347,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045185713296682175,
      "loss": 5.0667,
      "mean_token_accuracy": 0.2047081783413887,
      "num_tokens": 43892206.0,
      "step": 19675
    },
    {
      "entropy": 5.408244991302491,
      "epoch": 2.2102543943393047,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045183231076778005,
      "loss": 5.1284,
      "mean_token_accuracy": 0.20520488023757935,
      "num_tokens": 43905255.0,
      "step": 19680
    },
    {
      "entropy": 5.359394979476929,
      "epoch": 2.2108159712472624,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045180748293832906,
      "loss": 5.1822,
      "mean_token_accuracy": 0.19614837020635606,
      "num_tokens": 43917774.0,
      "step": 19685
    },
    {
      "entropy": 5.350852680206299,
      "epoch": 2.21137754815522,
      "grad_norm": 1.125,
      "learning_rate": 0.00045178264947925924,
      "loss": 5.0946,
      "mean_token_accuracy": 0.21021485775709153,
      "num_tokens": 43929127.0,
      "step": 19690
    },
    {
      "entropy": 5.470377683639526,
      "epoch": 2.2119391250631772,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045175781039136133,
      "loss": 5.1343,
      "mean_token_accuracy": 0.19852509945631028,
      "num_tokens": 43942558.0,
      "step": 19695
    },
    {
      "entropy": 5.345687532424927,
      "epoch": 2.212500701971135,
      "grad_norm": 1.125,
      "learning_rate": 0.0004517329656754261,
      "loss": 5.1372,
      "mean_token_accuracy": 0.20165610313415527,
      "num_tokens": 43952966.0,
      "step": 19700
    },
    {
      "entropy": 5.383357381820678,
      "epoch": 2.2130622788790926,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004517081153322448,
      "loss": 5.138,
      "mean_token_accuracy": 0.2008071169257164,
      "num_tokens": 43964725.0,
      "step": 19705
    },
    {
      "entropy": 5.384317302703858,
      "epoch": 2.2136238557870502,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004516832593626085,
      "loss": 5.0379,
      "mean_token_accuracy": 0.21041581481695176,
      "num_tokens": 43976356.0,
      "step": 19710
    },
    {
      "entropy": 5.263098287582397,
      "epoch": 2.2141854326950074,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004516583977673087,
      "loss": 5.0792,
      "mean_token_accuracy": 0.20420906096696853,
      "num_tokens": 43987732.0,
      "step": 19715
    },
    {
      "entropy": 5.412492322921753,
      "epoch": 2.214747009602965,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045163353054713704,
      "loss": 5.2168,
      "mean_token_accuracy": 0.19473868757486343,
      "num_tokens": 43998701.0,
      "step": 19720
    },
    {
      "entropy": 5.468985080718994,
      "epoch": 2.2153085865109228,
      "grad_norm": 1.125,
      "learning_rate": 0.0004516086577028852,
      "loss": 5.143,
      "mean_token_accuracy": 0.20163974314928054,
      "num_tokens": 44008978.0,
      "step": 19725
    },
    {
      "entropy": 5.366141319274902,
      "epoch": 2.21587016341888,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004515837792353453,
      "loss": 5.12,
      "mean_token_accuracy": 0.20525047481060027,
      "num_tokens": 44018928.0,
      "step": 19730
    },
    {
      "entropy": 5.35324068069458,
      "epoch": 2.2164317403268377,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004515588951453093,
      "loss": 5.1054,
      "mean_token_accuracy": 0.1964929521083832,
      "num_tokens": 44030044.0,
      "step": 19735
    },
    {
      "entropy": 5.28419132232666,
      "epoch": 2.2169933172347953,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004515340054335696,
      "loss": 5.052,
      "mean_token_accuracy": 0.204160375893116,
      "num_tokens": 44041851.0,
      "step": 19740
    },
    {
      "entropy": 5.345020055770874,
      "epoch": 2.217554894142753,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004515091101009188,
      "loss": 5.0385,
      "mean_token_accuracy": 0.20638213455677032,
      "num_tokens": 44051871.0,
      "step": 19745
    },
    {
      "entropy": 5.420204019546508,
      "epoch": 2.21811647105071,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004514842091481493,
      "loss": 5.2397,
      "mean_token_accuracy": 0.19367604404687883,
      "num_tokens": 44063222.0,
      "step": 19750
    },
    {
      "entropy": 5.5058349609375,
      "epoch": 2.218678047958668,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004514593025760543,
      "loss": 5.1926,
      "mean_token_accuracy": 0.19771205335855485,
      "num_tokens": 44075020.0,
      "step": 19755
    },
    {
      "entropy": 5.395500755310058,
      "epoch": 2.2192396248666255,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045143439038542653,
      "loss": 5.1141,
      "mean_token_accuracy": 0.19509703814983367,
      "num_tokens": 44086696.0,
      "step": 19760
    },
    {
      "entropy": 5.363218355178833,
      "epoch": 2.219801201774583,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004514094725770594,
      "loss": 5.0798,
      "mean_token_accuracy": 0.20398388504981996,
      "num_tokens": 44096768.0,
      "step": 19765
    },
    {
      "entropy": 5.329913902282715,
      "epoch": 2.2203627786825404,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045138454915174625,
      "loss": 5.0956,
      "mean_token_accuracy": 0.2093544825911522,
      "num_tokens": 44108008.0,
      "step": 19770
    },
    {
      "entropy": 5.418118143081665,
      "epoch": 2.220924355590498,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045135962011028064,
      "loss": 5.0616,
      "mean_token_accuracy": 0.20157649368047714,
      "num_tokens": 44119231.0,
      "step": 19775
    },
    {
      "entropy": 5.350447463989258,
      "epoch": 2.2214859324984557,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004513346854534564,
      "loss": 5.0611,
      "mean_token_accuracy": 0.20444795489311218,
      "num_tokens": 44129264.0,
      "step": 19780
    },
    {
      "entropy": 5.352105712890625,
      "epoch": 2.2220475094064134,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004513097451820673,
      "loss": 5.2004,
      "mean_token_accuracy": 0.19714549481868743,
      "num_tokens": 44141234.0,
      "step": 19785
    },
    {
      "entropy": 5.423063898086548,
      "epoch": 2.2226090863143706,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045128479929690756,
      "loss": 5.0433,
      "mean_token_accuracy": 0.21028388440608978,
      "num_tokens": 44151751.0,
      "step": 19790
    },
    {
      "entropy": 5.279069423675537,
      "epoch": 2.2231706632223283,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004512598477987714,
      "loss": 5.0016,
      "mean_token_accuracy": 0.2082223430275917,
      "num_tokens": 44162687.0,
      "step": 19795
    },
    {
      "entropy": 5.325049591064453,
      "epoch": 2.223732240130286,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045123489068845334,
      "loss": 5.0506,
      "mean_token_accuracy": 0.21059545129537582,
      "num_tokens": 44173758.0,
      "step": 19800
    },
    {
      "entropy": 5.364926052093506,
      "epoch": 2.2242938170382436,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045120992796674804,
      "loss": 5.0907,
      "mean_token_accuracy": 0.2044379934668541,
      "num_tokens": 44184176.0,
      "step": 19805
    },
    {
      "entropy": 5.408457660675049,
      "epoch": 2.224855393946201,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004511849596344503,
      "loss": 5.178,
      "mean_token_accuracy": 0.19322932958602906,
      "num_tokens": 44195155.0,
      "step": 19810
    },
    {
      "entropy": 5.505094003677368,
      "epoch": 2.2254169708541585,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045115998569235507,
      "loss": 5.1994,
      "mean_token_accuracy": 0.19846249669790267,
      "num_tokens": 44207139.0,
      "step": 19815
    },
    {
      "entropy": 5.401436901092529,
      "epoch": 2.225978547762116,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045113500614125756,
      "loss": 5.1731,
      "mean_token_accuracy": 0.2000483140349388,
      "num_tokens": 44218317.0,
      "step": 19820
    },
    {
      "entropy": 5.352754020690918,
      "epoch": 2.2265401246700733,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004511100209819532,
      "loss": 5.1009,
      "mean_token_accuracy": 0.20878102481365204,
      "num_tokens": 44228700.0,
      "step": 19825
    },
    {
      "entropy": 5.296424674987793,
      "epoch": 2.227101701578031,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045108503021523737,
      "loss": 5.0323,
      "mean_token_accuracy": 0.21434617340564727,
      "num_tokens": 44239991.0,
      "step": 19830
    },
    {
      "entropy": 5.264494132995606,
      "epoch": 2.2276632784859887,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045106003384190604,
      "loss": 4.9471,
      "mean_token_accuracy": 0.21130575388669967,
      "num_tokens": 44250224.0,
      "step": 19835
    },
    {
      "entropy": 5.3410697937011715,
      "epoch": 2.2282248553939463,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045103503186275486,
      "loss": 5.1304,
      "mean_token_accuracy": 0.2045440763235092,
      "num_tokens": 44261403.0,
      "step": 19840
    },
    {
      "entropy": 5.291489124298096,
      "epoch": 2.2287864323019035,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045101002427857997,
      "loss": 5.0572,
      "mean_token_accuracy": 0.19979802370071412,
      "num_tokens": 44272935.0,
      "step": 19845
    },
    {
      "entropy": 5.352489805221557,
      "epoch": 2.229348009209861,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045098501109017765,
      "loss": 5.0532,
      "mean_token_accuracy": 0.21259043663740157,
      "num_tokens": 44283458.0,
      "step": 19850
    },
    {
      "entropy": 5.344120836257934,
      "epoch": 2.229909586117819,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045095999229834437,
      "loss": 5.0804,
      "mean_token_accuracy": 0.20326650887727737,
      "num_tokens": 44294847.0,
      "step": 19855
    },
    {
      "entropy": 5.3839319229125975,
      "epoch": 2.2304711630257765,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004509349679038767,
      "loss": 5.1944,
      "mean_token_accuracy": 0.20121385753154755,
      "num_tokens": 44306395.0,
      "step": 19860
    },
    {
      "entropy": 5.360472011566162,
      "epoch": 2.2310327399337337,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045090993790757137,
      "loss": 5.0915,
      "mean_token_accuracy": 0.20900319069623946,
      "num_tokens": 44317452.0,
      "step": 19865
    },
    {
      "entropy": 5.337673139572144,
      "epoch": 2.2315943168416914,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004508849023102253,
      "loss": 5.0899,
      "mean_token_accuracy": 0.20763290375471116,
      "num_tokens": 44329862.0,
      "step": 19870
    },
    {
      "entropy": 5.409427213668823,
      "epoch": 2.232155893749649,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004508598611126359,
      "loss": 5.1671,
      "mean_token_accuracy": 0.20075237452983857,
      "num_tokens": 44340334.0,
      "step": 19875
    },
    {
      "entropy": 5.448436880111695,
      "epoch": 2.2327174706576067,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004508348143156002,
      "loss": 5.2298,
      "mean_token_accuracy": 0.19660582393407822,
      "num_tokens": 44351810.0,
      "step": 19880
    },
    {
      "entropy": 5.5019775390625,
      "epoch": 2.233279047565564,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045080976191991587,
      "loss": 5.2751,
      "mean_token_accuracy": 0.1884935602545738,
      "num_tokens": 44365143.0,
      "step": 19885
    },
    {
      "entropy": 5.411267900466919,
      "epoch": 2.2338406244735216,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004507847039263805,
      "loss": 5.1622,
      "mean_token_accuracy": 0.2050977349281311,
      "num_tokens": 44375895.0,
      "step": 19890
    },
    {
      "entropy": 5.371902799606323,
      "epoch": 2.2344022013814793,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000450759640335792,
      "loss": 5.0818,
      "mean_token_accuracy": 0.19852117896080018,
      "num_tokens": 44387378.0,
      "step": 19895
    },
    {
      "entropy": 5.365559530258179,
      "epoch": 2.234963778289437,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004507345711489484,
      "loss": 5.0354,
      "mean_token_accuracy": 0.21263857185840607,
      "num_tokens": 44398460.0,
      "step": 19900
    },
    {
      "entropy": 5.291798400878906,
      "epoch": 2.235525355197394,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045070949636664785,
      "loss": 5.0712,
      "mean_token_accuracy": 0.2069460391998291,
      "num_tokens": 44409885.0,
      "step": 19905
    },
    {
      "entropy": 5.304100036621094,
      "epoch": 2.236086932105352,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004506844159896888,
      "loss": 5.0016,
      "mean_token_accuracy": 0.20898811668157577,
      "num_tokens": 44420897.0,
      "step": 19910
    },
    {
      "entropy": 5.406945896148682,
      "epoch": 2.2366485090133095,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004506593300188698,
      "loss": 5.1894,
      "mean_token_accuracy": 0.19910797327756882,
      "num_tokens": 44432691.0,
      "step": 19915
    },
    {
      "entropy": 5.414349937438965,
      "epoch": 2.2372100859212667,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004506342384549897,
      "loss": 5.1106,
      "mean_token_accuracy": 0.19856184720993042,
      "num_tokens": 44444731.0,
      "step": 19920
    },
    {
      "entropy": 5.455244159698486,
      "epoch": 2.2377716628292244,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004506091412988472,
      "loss": 5.0967,
      "mean_token_accuracy": 0.21136207431554793,
      "num_tokens": 44456959.0,
      "step": 19925
    },
    {
      "entropy": 5.42680344581604,
      "epoch": 2.238333239737182,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045058403855124165,
      "loss": 5.1745,
      "mean_token_accuracy": 0.20232002586126327,
      "num_tokens": 44468970.0,
      "step": 19930
    },
    {
      "entropy": 5.334129810333252,
      "epoch": 2.2388948166451397,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004505589302129721,
      "loss": 5.062,
      "mean_token_accuracy": 0.20860576778650283,
      "num_tokens": 44480485.0,
      "step": 19935
    },
    {
      "entropy": 5.387248897552491,
      "epoch": 2.239456393553097,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004505338162848382,
      "loss": 5.209,
      "mean_token_accuracy": 0.2030231922864914,
      "num_tokens": 44492890.0,
      "step": 19940
    },
    {
      "entropy": 5.459932279586792,
      "epoch": 2.2400179704610546,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004505086967676395,
      "loss": 5.2764,
      "mean_token_accuracy": 0.18722792863845825,
      "num_tokens": 44504743.0,
      "step": 19945
    },
    {
      "entropy": 5.431375312805176,
      "epoch": 2.240579547369012,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004504835716621757,
      "loss": 5.0639,
      "mean_token_accuracy": 0.20350304394960403,
      "num_tokens": 44514998.0,
      "step": 19950
    },
    {
      "entropy": 5.338810110092163,
      "epoch": 2.24114112427697,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004504584409692471,
      "loss": 5.0637,
      "mean_token_accuracy": 0.2117594838142395,
      "num_tokens": 44527530.0,
      "step": 19955
    },
    {
      "entropy": 5.484750413894654,
      "epoch": 2.241702701184927,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045043330468965364,
      "loss": 5.2615,
      "mean_token_accuracy": 0.19735287874937057,
      "num_tokens": 44540939.0,
      "step": 19960
    },
    {
      "entropy": 5.426646661758423,
      "epoch": 2.2422642780928848,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045040816282419566,
      "loss": 5.1508,
      "mean_token_accuracy": 0.19989483207464218,
      "num_tokens": 44551557.0,
      "step": 19965
    },
    {
      "entropy": 5.339177560806275,
      "epoch": 2.2428258550008424,
      "grad_norm": 1.125,
      "learning_rate": 0.00045038301537367374,
      "loss": 5.0305,
      "mean_token_accuracy": 0.20837254375219344,
      "num_tokens": 44561969.0,
      "step": 19970
    },
    {
      "entropy": 5.296502876281738,
      "epoch": 2.2433874319088,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045035786233888864,
      "loss": 5.039,
      "mean_token_accuracy": 0.20229945331811905,
      "num_tokens": 44573542.0,
      "step": 19975
    },
    {
      "entropy": 5.444465827941895,
      "epoch": 2.2439490088167573,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045033270372064123,
      "loss": 5.1062,
      "mean_token_accuracy": 0.2069901168346405,
      "num_tokens": 44583720.0,
      "step": 19980
    },
    {
      "entropy": 5.384325456619263,
      "epoch": 2.244510585724715,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045030753951973243,
      "loss": 5.1805,
      "mean_token_accuracy": 0.20217148214578629,
      "num_tokens": 44595314.0,
      "step": 19985
    },
    {
      "entropy": 5.467894792556763,
      "epoch": 2.2450721626326726,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045028236973696363,
      "loss": 5.2949,
      "mean_token_accuracy": 0.19682769030332564,
      "num_tokens": 44608569.0,
      "step": 19990
    },
    {
      "entropy": 5.3731077194213865,
      "epoch": 2.2456337395406303,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004502571943731362,
      "loss": 5.119,
      "mean_token_accuracy": 0.19924229383468628,
      "num_tokens": 44619389.0,
      "step": 19995
    },
    {
      "entropy": 5.456809854507446,
      "epoch": 2.2461953164485875,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004502320134290517,
      "loss": 5.1985,
      "mean_token_accuracy": 0.1968458488583565,
      "num_tokens": 44629945.0,
      "step": 20000
    },
    {
      "entropy": 5.426967573165894,
      "epoch": 2.246756893356545,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045020682690551203,
      "loss": 5.219,
      "mean_token_accuracy": 0.19605760127305985,
      "num_tokens": 44640741.0,
      "step": 20005
    },
    {
      "entropy": 5.387755489349365,
      "epoch": 2.247318470264503,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045018163480331887,
      "loss": 5.0549,
      "mean_token_accuracy": 0.20975955128669738,
      "num_tokens": 44651312.0,
      "step": 20010
    },
    {
      "entropy": 5.310109186172485,
      "epoch": 2.24788004717246,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045015643712327465,
      "loss": 5.0992,
      "mean_token_accuracy": 0.2022841453552246,
      "num_tokens": 44663156.0,
      "step": 20015
    },
    {
      "entropy": 5.34509310722351,
      "epoch": 2.2484416240804177,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004501312338661815,
      "loss": 5.1037,
      "mean_token_accuracy": 0.20489188730716706,
      "num_tokens": 44673479.0,
      "step": 20020
    },
    {
      "entropy": 5.39595103263855,
      "epoch": 2.2490032009883754,
      "grad_norm": 1.0,
      "learning_rate": 0.0004501060250328419,
      "loss": 5.1172,
      "mean_token_accuracy": 0.2015745997428894,
      "num_tokens": 44684207.0,
      "step": 20025
    },
    {
      "entropy": 5.37456374168396,
      "epoch": 2.249564777896333,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004500808106240586,
      "loss": 5.0824,
      "mean_token_accuracy": 0.20756502598524093,
      "num_tokens": 44695611.0,
      "step": 20030
    },
    {
      "entropy": 5.392177581787109,
      "epoch": 2.2501263548042907,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004500555906406343,
      "loss": 5.151,
      "mean_token_accuracy": 0.19971729516983033,
      "num_tokens": 44707061.0,
      "step": 20035
    },
    {
      "entropy": 5.476693153381348,
      "epoch": 2.250687931712248,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004500303650833722,
      "loss": 5.229,
      "mean_token_accuracy": 0.19266939610242845,
      "num_tokens": 44719366.0,
      "step": 20040
    },
    {
      "entropy": 5.404720783233643,
      "epoch": 2.2512495086202056,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004500051339530754,
      "loss": 5.096,
      "mean_token_accuracy": 0.20842939168214797,
      "num_tokens": 44730578.0,
      "step": 20045
    },
    {
      "entropy": 5.3448793411254885,
      "epoch": 2.2518110855281632,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004499798972505472,
      "loss": 5.0556,
      "mean_token_accuracy": 0.20402756333351135,
      "num_tokens": 44740620.0,
      "step": 20050
    },
    {
      "entropy": 5.338052797317505,
      "epoch": 2.2523726624361204,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044995465497659127,
      "loss": 5.1141,
      "mean_token_accuracy": 0.2048641860485077,
      "num_tokens": 44751518.0,
      "step": 20055
    },
    {
      "entropy": 5.396733951568604,
      "epoch": 2.252934239344078,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004499294071320112,
      "loss": 5.1395,
      "mean_token_accuracy": 0.2032914087176323,
      "num_tokens": 44761563.0,
      "step": 20060
    },
    {
      "entropy": 5.430647182464599,
      "epoch": 2.2534958162520358,
      "grad_norm": 1.046875,
      "learning_rate": 0.000449904153717611,
      "loss": 5.1518,
      "mean_token_accuracy": 0.20310390144586563,
      "num_tokens": 44773330.0,
      "step": 20065
    },
    {
      "entropy": 5.390623092651367,
      "epoch": 2.2540573931599934,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004498788947341947,
      "loss": 5.1249,
      "mean_token_accuracy": 0.20284368246793746,
      "num_tokens": 44786198.0,
      "step": 20070
    },
    {
      "entropy": 5.2430610179901125,
      "epoch": 2.2546189700679506,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004498536301825665,
      "loss": 5.024,
      "mean_token_accuracy": 0.20852240771055222,
      "num_tokens": 44797643.0,
      "step": 20075
    },
    {
      "entropy": 5.3315552234649655,
      "epoch": 2.2551805469759083,
      "grad_norm": 1.0625,
      "learning_rate": 0.000449828360063531,
      "loss": 5.0423,
      "mean_token_accuracy": 0.21212098598480225,
      "num_tokens": 44808283.0,
      "step": 20080
    },
    {
      "entropy": 5.327229261398315,
      "epoch": 2.255742123883866,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044980308437789267,
      "loss": 4.9552,
      "mean_token_accuracy": 0.21080983132123948,
      "num_tokens": 44817699.0,
      "step": 20085
    },
    {
      "entropy": 5.321918630599976,
      "epoch": 2.2563037007918236,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004497778031264564,
      "loss": 5.1876,
      "mean_token_accuracy": 0.20064649283885955,
      "num_tokens": 44828428.0,
      "step": 20090
    },
    {
      "entropy": 5.279856443405151,
      "epoch": 2.256865277699781,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044975251631002694,
      "loss": 5.0817,
      "mean_token_accuracy": 0.2005918726325035,
      "num_tokens": 44839288.0,
      "step": 20095
    },
    {
      "entropy": 5.458464002609253,
      "epoch": 2.2574268546077385,
      "grad_norm": 2.0,
      "learning_rate": 0.0004497272239294097,
      "loss": 5.3002,
      "mean_token_accuracy": 0.19854089319705964,
      "num_tokens": 44850250.0,
      "step": 20100
    },
    {
      "entropy": 5.461380100250244,
      "epoch": 2.257988431515696,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004497019259854099,
      "loss": 5.1262,
      "mean_token_accuracy": 0.20368383526802064,
      "num_tokens": 44862186.0,
      "step": 20105
    },
    {
      "entropy": 5.432626962661743,
      "epoch": 2.2585500084236534,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044967662247883295,
      "loss": 5.1955,
      "mean_token_accuracy": 0.1978796139359474,
      "num_tokens": 44874266.0,
      "step": 20110
    },
    {
      "entropy": 5.43372073173523,
      "epoch": 2.259111585331611,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044965131341048454,
      "loss": 5.1602,
      "mean_token_accuracy": 0.20611430406570436,
      "num_tokens": 44885783.0,
      "step": 20115
    },
    {
      "entropy": 5.3811603546142575,
      "epoch": 2.2596731622395687,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004496259987811707,
      "loss": 5.0148,
      "mean_token_accuracy": 0.21075453609228134,
      "num_tokens": 44896972.0,
      "step": 20120
    },
    {
      "entropy": 5.363962411880493,
      "epoch": 2.2602347391475264,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004496006785916973,
      "loss": 5.105,
      "mean_token_accuracy": 0.20586262494325638,
      "num_tokens": 44907414.0,
      "step": 20125
    },
    {
      "entropy": 5.368867635726929,
      "epoch": 2.260796316055484,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044957535284287055,
      "loss": 5.0557,
      "mean_token_accuracy": 0.20357841551303862,
      "num_tokens": 44917150.0,
      "step": 20130
    },
    {
      "entropy": 5.359481000900269,
      "epoch": 2.2613578929634413,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044955002153549686,
      "loss": 5.0871,
      "mean_token_accuracy": 0.20869020968675614,
      "num_tokens": 44928927.0,
      "step": 20135
    },
    {
      "entropy": 5.374300289154053,
      "epoch": 2.261919469871399,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004495246846703828,
      "loss": 5.12,
      "mean_token_accuracy": 0.20586791187524794,
      "num_tokens": 44938987.0,
      "step": 20140
    },
    {
      "entropy": 5.37215256690979,
      "epoch": 2.2624810467793566,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044949934224833505,
      "loss": 5.0383,
      "mean_token_accuracy": 0.2049108028411865,
      "num_tokens": 44950395.0,
      "step": 20145
    },
    {
      "entropy": 5.3761156558990475,
      "epoch": 2.263042623687314,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004494739942701606,
      "loss": 5.0601,
      "mean_token_accuracy": 0.20769781172275542,
      "num_tokens": 44961389.0,
      "step": 20150
    },
    {
      "entropy": 5.406068420410156,
      "epoch": 2.2636042005952715,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044944864073666655,
      "loss": 5.2358,
      "mean_token_accuracy": 0.19917859733104706,
      "num_tokens": 44973142.0,
      "step": 20155
    },
    {
      "entropy": 5.339560651779175,
      "epoch": 2.264165777503229,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044942328164866,
      "loss": 5.0281,
      "mean_token_accuracy": 0.2050121784210205,
      "num_tokens": 44983831.0,
      "step": 20160
    },
    {
      "entropy": 5.371531438827515,
      "epoch": 2.2647273544111868,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004493979170069486,
      "loss": 5.1341,
      "mean_token_accuracy": 0.2060588553547859,
      "num_tokens": 44996386.0,
      "step": 20165
    },
    {
      "entropy": 5.363423109054565,
      "epoch": 2.265288931319144,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044937254681233984,
      "loss": 5.0841,
      "mean_token_accuracy": 0.2025669574737549,
      "num_tokens": 45007607.0,
      "step": 20170
    },
    {
      "entropy": 5.3483837127685545,
      "epoch": 2.2658505082271017,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044934717106564164,
      "loss": 5.0953,
      "mean_token_accuracy": 0.2053024485707283,
      "num_tokens": 45019637.0,
      "step": 20175
    },
    {
      "entropy": 5.354740190505981,
      "epoch": 2.2664120851350593,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044932178976766183,
      "loss": 5.0894,
      "mean_token_accuracy": 0.20830103605985642,
      "num_tokens": 45031329.0,
      "step": 20180
    },
    {
      "entropy": 5.383899927139282,
      "epoch": 2.266973662043017,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044929640291920865,
      "loss": 5.0723,
      "mean_token_accuracy": 0.20596210062503814,
      "num_tokens": 45042539.0,
      "step": 20185
    },
    {
      "entropy": 5.288854885101318,
      "epoch": 2.267535238950974,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004492710105210904,
      "loss": 4.9716,
      "mean_token_accuracy": 0.21215254664421082,
      "num_tokens": 45054021.0,
      "step": 20190
    },
    {
      "entropy": 5.426714372634888,
      "epoch": 2.268096815858932,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044924561257411554,
      "loss": 5.166,
      "mean_token_accuracy": 0.20182067155838013,
      "num_tokens": 45065912.0,
      "step": 20195
    },
    {
      "entropy": 5.391606378555298,
      "epoch": 2.2686583927668895,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004492202090790928,
      "loss": 5.0413,
      "mean_token_accuracy": 0.21461397111415864,
      "num_tokens": 45075965.0,
      "step": 20200
    },
    {
      "entropy": 5.346408033370972,
      "epoch": 2.2692199696748467,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004491948000368311,
      "loss": 5.0564,
      "mean_token_accuracy": 0.2054465591907501,
      "num_tokens": 45086676.0,
      "step": 20205
    },
    {
      "entropy": 5.356198787689209,
      "epoch": 2.2697815465828044,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044916938544813936,
      "loss": 5.0975,
      "mean_token_accuracy": 0.21046365648508072,
      "num_tokens": 45097747.0,
      "step": 20210
    },
    {
      "entropy": 5.41406283378601,
      "epoch": 2.270343123490762,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044914396531382686,
      "loss": 5.0526,
      "mean_token_accuracy": 0.21831392049789428,
      "num_tokens": 45107506.0,
      "step": 20215
    },
    {
      "entropy": 5.360881185531616,
      "epoch": 2.2709047003987197,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044911853963470295,
      "loss": 5.1166,
      "mean_token_accuracy": 0.20843956172466277,
      "num_tokens": 45119939.0,
      "step": 20220
    },
    {
      "entropy": 5.391731405258179,
      "epoch": 2.2714662773066774,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004490931084115772,
      "loss": 5.1806,
      "mean_token_accuracy": 0.196001772582531,
      "num_tokens": 45133749.0,
      "step": 20225
    },
    {
      "entropy": 5.441767168045044,
      "epoch": 2.2720278542146346,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044906767164525936,
      "loss": 5.1689,
      "mean_token_accuracy": 0.1956079050898552,
      "num_tokens": 45144696.0,
      "step": 20230
    },
    {
      "entropy": 5.3610412120819095,
      "epoch": 2.2725894311225923,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004490422293365594,
      "loss": 5.0263,
      "mean_token_accuracy": 0.20480331182479858,
      "num_tokens": 45154921.0,
      "step": 20235
    },
    {
      "entropy": 5.303757238388061,
      "epoch": 2.27315100803055,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004490167814862872,
      "loss": 5.0202,
      "mean_token_accuracy": 0.20810357928276063,
      "num_tokens": 45164769.0,
      "step": 20240
    },
    {
      "entropy": 5.44170823097229,
      "epoch": 2.273712584938507,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004489913280952533,
      "loss": 5.1821,
      "mean_token_accuracy": 0.2045654445886612,
      "num_tokens": 45176619.0,
      "step": 20245
    },
    {
      "entropy": 5.359971141815185,
      "epoch": 2.274274161846465,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004489658691642681,
      "loss": 5.1314,
      "mean_token_accuracy": 0.1993611201643944,
      "num_tokens": 45186508.0,
      "step": 20250
    },
    {
      "entropy": 5.495048522949219,
      "epoch": 2.2748357387544225,
      "grad_norm": 1.125,
      "learning_rate": 0.00044894040469414207,
      "loss": 5.2208,
      "mean_token_accuracy": 0.19484345614910126,
      "num_tokens": 45198213.0,
      "step": 20255
    },
    {
      "entropy": 5.389283752441406,
      "epoch": 2.27539731566238,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044891493468568607,
      "loss": 5.0899,
      "mean_token_accuracy": 0.19722937494516374,
      "num_tokens": 45209328.0,
      "step": 20260
    },
    {
      "entropy": 5.387155294418335,
      "epoch": 2.2759588925703373,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044888945913971106,
      "loss": 5.0753,
      "mean_token_accuracy": 0.2116628721356392,
      "num_tokens": 45219722.0,
      "step": 20265
    },
    {
      "entropy": 5.453005790710449,
      "epoch": 2.276520469478295,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044886397805702835,
      "loss": 5.1697,
      "mean_token_accuracy": 0.19668956100940704,
      "num_tokens": 45230377.0,
      "step": 20270
    },
    {
      "entropy": 5.406516361236572,
      "epoch": 2.2770820463862527,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044883849143844914,
      "loss": 5.064,
      "mean_token_accuracy": 0.20673507899045945,
      "num_tokens": 45241598.0,
      "step": 20275
    },
    {
      "entropy": 5.482073593139648,
      "epoch": 2.2776436232942103,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004488129992847848,
      "loss": 5.3091,
      "mean_token_accuracy": 0.19570596367120743,
      "num_tokens": 45253153.0,
      "step": 20280
    },
    {
      "entropy": 5.438924980163574,
      "epoch": 2.2782052002021675,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004487875015968473,
      "loss": 5.1684,
      "mean_token_accuracy": 0.20403917282819747,
      "num_tokens": 45264523.0,
      "step": 20285
    },
    {
      "entropy": 5.332516670227051,
      "epoch": 2.278766777110125,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004487619983754483,
      "loss": 5.0826,
      "mean_token_accuracy": 0.19761485308408738,
      "num_tokens": 45275320.0,
      "step": 20290
    },
    {
      "entropy": 5.37819242477417,
      "epoch": 2.279328354018083,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044873648962139983,
      "loss": 5.1015,
      "mean_token_accuracy": 0.20552792251110077,
      "num_tokens": 45286704.0,
      "step": 20295
    },
    {
      "entropy": 5.435221910476685,
      "epoch": 2.27988993092604,
      "grad_norm": 1.0,
      "learning_rate": 0.0004487109753355142,
      "loss": 5.1205,
      "mean_token_accuracy": 0.2048347756266594,
      "num_tokens": 45299098.0,
      "step": 20300
    },
    {
      "entropy": 5.4110502243042,
      "epoch": 2.2804515078339977,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004486854555186037,
      "loss": 5.0713,
      "mean_token_accuracy": 0.20433363020420076,
      "num_tokens": 45309995.0,
      "step": 20305
    },
    {
      "entropy": 5.455915021896362,
      "epoch": 2.2810130847419554,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004486599301714809,
      "loss": 5.2318,
      "mean_token_accuracy": 0.20042610317468643,
      "num_tokens": 45321711.0,
      "step": 20310
    },
    {
      "entropy": 5.356934261322022,
      "epoch": 2.281574661649913,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004486343992949587,
      "loss": 5.1483,
      "mean_token_accuracy": 0.20324225276708602,
      "num_tokens": 45332401.0,
      "step": 20315
    },
    {
      "entropy": 5.4901453971862795,
      "epoch": 2.2821362385578707,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044860886288984977,
      "loss": 5.2057,
      "mean_token_accuracy": 0.20343443155288696,
      "num_tokens": 45344138.0,
      "step": 20320
    },
    {
      "entropy": 5.392458295822143,
      "epoch": 2.282697815465828,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004485833209569674,
      "loss": 5.1196,
      "mean_token_accuracy": 0.2008978933095932,
      "num_tokens": 45355913.0,
      "step": 20325
    },
    {
      "entropy": 5.36622986793518,
      "epoch": 2.2832593923737856,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004485577734971247,
      "loss": 5.0503,
      "mean_token_accuracy": 0.21205074340105057,
      "num_tokens": 45367342.0,
      "step": 20330
    },
    {
      "entropy": 5.394284248352051,
      "epoch": 2.2838209692817433,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044853222051113525,
      "loss": 5.1101,
      "mean_token_accuracy": 0.20237682312726973,
      "num_tokens": 45377866.0,
      "step": 20335
    },
    {
      "entropy": 5.351701402664185,
      "epoch": 2.2843825461897005,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044850666199981244,
      "loss": 5.1281,
      "mean_token_accuracy": 0.20044913291931152,
      "num_tokens": 45389346.0,
      "step": 20340
    },
    {
      "entropy": 5.285377550125122,
      "epoch": 2.284944123097658,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044848109796397035,
      "loss": 5.0674,
      "mean_token_accuracy": 0.20663114339113237,
      "num_tokens": 45400112.0,
      "step": 20345
    },
    {
      "entropy": 5.330026340484619,
      "epoch": 2.285505700005616,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044845552840442276,
      "loss": 5.0608,
      "mean_token_accuracy": 0.20571451932191848,
      "num_tokens": 45411666.0,
      "step": 20350
    },
    {
      "entropy": 5.367145252227783,
      "epoch": 2.2860672769135735,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044842995332198393,
      "loss": 5.0329,
      "mean_token_accuracy": 0.21623374223709108,
      "num_tokens": 45423631.0,
      "step": 20355
    },
    {
      "entropy": 5.2478807926177975,
      "epoch": 2.2866288538215307,
      "grad_norm": 1.0,
      "learning_rate": 0.00044840437271746807,
      "loss": 5.0099,
      "mean_token_accuracy": 0.20629750043153763,
      "num_tokens": 45434697.0,
      "step": 20360
    },
    {
      "entropy": 5.3997878074646,
      "epoch": 2.2871904307294884,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004483787865916898,
      "loss": 5.1481,
      "mean_token_accuracy": 0.1986825242638588,
      "num_tokens": 45446637.0,
      "step": 20365
    },
    {
      "entropy": 5.362820291519165,
      "epoch": 2.287752007637446,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044835319494546366,
      "loss": 5.0839,
      "mean_token_accuracy": 0.20668365508317948,
      "num_tokens": 45456750.0,
      "step": 20370
    },
    {
      "entropy": 5.359393358230591,
      "epoch": 2.2883135845454037,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044832759777960456,
      "loss": 5.1185,
      "mean_token_accuracy": 0.20493456870317459,
      "num_tokens": 45469383.0,
      "step": 20375
    },
    {
      "entropy": 5.365430164337158,
      "epoch": 2.288875161453361,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004483019950949276,
      "loss": 5.0775,
      "mean_token_accuracy": 0.20621394366025925,
      "num_tokens": 45479668.0,
      "step": 20380
    },
    {
      "entropy": 5.345883893966675,
      "epoch": 2.2894367383613186,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004482763868922478,
      "loss": 5.1497,
      "mean_token_accuracy": 0.1997622326016426,
      "num_tokens": 45490888.0,
      "step": 20385
    },
    {
      "entropy": 5.429190683364868,
      "epoch": 2.289998315269276,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004482507731723807,
      "loss": 5.1392,
      "mean_token_accuracy": 0.19994169920682908,
      "num_tokens": 45501973.0,
      "step": 20390
    },
    {
      "entropy": 5.420404052734375,
      "epoch": 2.2905598921772334,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004482251539361418,
      "loss": 5.1187,
      "mean_token_accuracy": 0.20720255374908447,
      "num_tokens": 45513171.0,
      "step": 20395
    },
    {
      "entropy": 5.324348068237304,
      "epoch": 2.291121469085191,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004481995291843467,
      "loss": 5.0595,
      "mean_token_accuracy": 0.20472586899995804,
      "num_tokens": 45523642.0,
      "step": 20400
    },
    {
      "entropy": 5.42300443649292,
      "epoch": 2.2916830459931488,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004481738989178115,
      "loss": 5.147,
      "mean_token_accuracy": 0.20500423461198808,
      "num_tokens": 45534467.0,
      "step": 20405
    },
    {
      "entropy": 5.425853109359741,
      "epoch": 2.2922446229011064,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004481482631373522,
      "loss": 5.1166,
      "mean_token_accuracy": 0.1967497870326042,
      "num_tokens": 45545902.0,
      "step": 20410
    },
    {
      "entropy": 5.3463537216186525,
      "epoch": 2.292806199809064,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044812262184378503,
      "loss": 4.9967,
      "mean_token_accuracy": 0.21132127195596695,
      "num_tokens": 45556270.0,
      "step": 20415
    },
    {
      "entropy": 5.258968830108643,
      "epoch": 2.2933677767170213,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044809697503792647,
      "loss": 5.0489,
      "mean_token_accuracy": 0.21258210241794587,
      "num_tokens": 45566773.0,
      "step": 20420
    },
    {
      "entropy": 5.34417986869812,
      "epoch": 2.293929353624979,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00044807132272059306,
      "loss": 5.1523,
      "mean_token_accuracy": 0.20067519843578338,
      "num_tokens": 45578819.0,
      "step": 20425
    },
    {
      "entropy": 5.410848951339721,
      "epoch": 2.2944909305329366,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004480456648926016,
      "loss": 5.1139,
      "mean_token_accuracy": 0.19825557768344879,
      "num_tokens": 45589128.0,
      "step": 20430
    },
    {
      "entropy": 5.400263452529908,
      "epoch": 2.295052507440894,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004480200015547691,
      "loss": 5.151,
      "mean_token_accuracy": 0.20369257777929306,
      "num_tokens": 45600253.0,
      "step": 20435
    },
    {
      "entropy": 5.429322910308838,
      "epoch": 2.2956140843488515,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044799433270791255,
      "loss": 5.1257,
      "mean_token_accuracy": 0.2052313804626465,
      "num_tokens": 45611258.0,
      "step": 20440
    },
    {
      "entropy": 5.436296558380127,
      "epoch": 2.296175661256809,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004479686583528494,
      "loss": 5.0759,
      "mean_token_accuracy": 0.21046193838119506,
      "num_tokens": 45622720.0,
      "step": 20445
    },
    {
      "entropy": 5.277192163467407,
      "epoch": 2.296737238164767,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004479429784903971,
      "loss": 5.0717,
      "mean_token_accuracy": 0.20076282769441606,
      "num_tokens": 45634748.0,
      "step": 20450
    },
    {
      "entropy": 5.339889144897461,
      "epoch": 2.297298815072724,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004479172931213733,
      "loss": 5.0123,
      "mean_token_accuracy": 0.2068767249584198,
      "num_tokens": 45645735.0,
      "step": 20455
    },
    {
      "entropy": 5.308385753631592,
      "epoch": 2.2978603919806817,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004478916022465958,
      "loss": 5.0208,
      "mean_token_accuracy": 0.20858240276575088,
      "num_tokens": 45656309.0,
      "step": 20460
    },
    {
      "entropy": 5.324087381362915,
      "epoch": 2.2984219688886394,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004478659058668827,
      "loss": 5.1655,
      "mean_token_accuracy": 0.20656261295080186,
      "num_tokens": 45667925.0,
      "step": 20465
    },
    {
      "entropy": 5.381308555603027,
      "epoch": 2.298983545796597,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044784020398305204,
      "loss": 5.142,
      "mean_token_accuracy": 0.20486316084861755,
      "num_tokens": 45679352.0,
      "step": 20470
    },
    {
      "entropy": 5.399655055999756,
      "epoch": 2.2995451227045542,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004478144965959222,
      "loss": 5.129,
      "mean_token_accuracy": 0.20334452688694,
      "num_tokens": 45689723.0,
      "step": 20475
    },
    {
      "entropy": 5.474929428100586,
      "epoch": 2.300106699612512,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044778878370631183,
      "loss": 5.2391,
      "mean_token_accuracy": 0.20222919881343843,
      "num_tokens": 45701270.0,
      "step": 20480
    },
    {
      "entropy": 5.372686719894409,
      "epoch": 2.3006682765204696,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044776306531503957,
      "loss": 5.0089,
      "mean_token_accuracy": 0.22263590395450591,
      "num_tokens": 45711409.0,
      "step": 20485
    },
    {
      "entropy": 5.323312091827392,
      "epoch": 2.301229853428427,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044773734142292425,
      "loss": 5.0888,
      "mean_token_accuracy": 0.20661951154470443,
      "num_tokens": 45722032.0,
      "step": 20490
    },
    {
      "entropy": 5.337507009506226,
      "epoch": 2.3017914303363844,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004477116120307851,
      "loss": 5.0955,
      "mean_token_accuracy": 0.20888228863477706,
      "num_tokens": 45733211.0,
      "step": 20495
    },
    {
      "entropy": 5.479390859603882,
      "epoch": 2.302353007244342,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004476858771394412,
      "loss": 5.2248,
      "mean_token_accuracy": 0.19718378335237502,
      "num_tokens": 45744585.0,
      "step": 20500
    },
    {
      "entropy": 5.457705402374268,
      "epoch": 2.3029145841522998,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004476601367497119,
      "loss": 5.1742,
      "mean_token_accuracy": 0.19780077487230302,
      "num_tokens": 45756040.0,
      "step": 20505
    },
    {
      "entropy": 5.4125732421875,
      "epoch": 2.3034761610602574,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000447634390862417,
      "loss": 5.1287,
      "mean_token_accuracy": 0.20809611976146697,
      "num_tokens": 45767717.0,
      "step": 20510
    },
    {
      "entropy": 5.4121662139892575,
      "epoch": 2.3040377379682146,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044760863947837607,
      "loss": 5.1542,
      "mean_token_accuracy": 0.19980325698852539,
      "num_tokens": 45778326.0,
      "step": 20515
    },
    {
      "entropy": 5.385725212097168,
      "epoch": 2.3045993148761723,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044758288259840904,
      "loss": 5.0638,
      "mean_token_accuracy": 0.21286622434854507,
      "num_tokens": 45789780.0,
      "step": 20520
    },
    {
      "entropy": 5.369700241088867,
      "epoch": 2.30516089178413,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004475571202233363,
      "loss": 5.2227,
      "mean_token_accuracy": 0.1957785740494728,
      "num_tokens": 45801686.0,
      "step": 20525
    },
    {
      "entropy": 5.35234227180481,
      "epoch": 2.305722468692087,
      "grad_norm": 1.0,
      "learning_rate": 0.0004475313523539778,
      "loss": 5.0719,
      "mean_token_accuracy": 0.2043260872364044,
      "num_tokens": 45812123.0,
      "step": 20530
    },
    {
      "entropy": 5.4861914157867435,
      "epoch": 2.306284045600045,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044750557899115407,
      "loss": 5.1451,
      "mean_token_accuracy": 0.19930824339389802,
      "num_tokens": 45822863.0,
      "step": 20535
    },
    {
      "entropy": 5.40135006904602,
      "epoch": 2.3068456225080025,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004474798001356859,
      "loss": 5.1472,
      "mean_token_accuracy": 0.2048728048801422,
      "num_tokens": 45833585.0,
      "step": 20540
    },
    {
      "entropy": 5.2607588291168215,
      "epoch": 2.30740719941596,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000447454015788394,
      "loss": 4.9643,
      "mean_token_accuracy": 0.2110483780503273,
      "num_tokens": 45844745.0,
      "step": 20545
    },
    {
      "entropy": 5.332904863357544,
      "epoch": 2.3079687763239174,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004474282259500992,
      "loss": 5.0549,
      "mean_token_accuracy": 0.21292762756347655,
      "num_tokens": 45856018.0,
      "step": 20550
    },
    {
      "entropy": 5.478094148635864,
      "epoch": 2.308530353231875,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000447402430621623,
      "loss": 5.1542,
      "mean_token_accuracy": 0.2049010068178177,
      "num_tokens": 45867182.0,
      "step": 20555
    },
    {
      "entropy": 5.447095346450806,
      "epoch": 2.3090919301398327,
      "grad_norm": 1.0,
      "learning_rate": 0.0004473766298037865,
      "loss": 5.2324,
      "mean_token_accuracy": 0.19268191158771514,
      "num_tokens": 45880116.0,
      "step": 20560
    },
    {
      "entropy": 5.408196496963501,
      "epoch": 2.3096535070477904,
      "grad_norm": 1.0,
      "learning_rate": 0.00044735082349741116,
      "loss": 5.1812,
      "mean_token_accuracy": 0.19641530960798265,
      "num_tokens": 45891143.0,
      "step": 20565
    },
    {
      "entropy": 5.434721279144287,
      "epoch": 2.3102150839557476,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044732501170331886,
      "loss": 5.059,
      "mean_token_accuracy": 0.20880725979804993,
      "num_tokens": 45901443.0,
      "step": 20570
    },
    {
      "entropy": 5.455426263809204,
      "epoch": 2.3107766608637053,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004472991944223313,
      "loss": 5.1698,
      "mean_token_accuracy": 0.19662276208400725,
      "num_tokens": 45913313.0,
      "step": 20575
    },
    {
      "entropy": 5.421405601501465,
      "epoch": 2.311338237771663,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044727337165527056,
      "loss": 5.1938,
      "mean_token_accuracy": 0.19769566506147385,
      "num_tokens": 45925248.0,
      "step": 20580
    },
    {
      "entropy": 5.446941566467285,
      "epoch": 2.31189981467962,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044724754340295887,
      "loss": 5.2103,
      "mean_token_accuracy": 0.1956170305609703,
      "num_tokens": 45936366.0,
      "step": 20585
    },
    {
      "entropy": 5.343863868713379,
      "epoch": 2.312461391587578,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004472217096662186,
      "loss": 5.0583,
      "mean_token_accuracy": 0.21755491942167282,
      "num_tokens": 45947732.0,
      "step": 20590
    },
    {
      "entropy": 5.353420543670654,
      "epoch": 2.3130229684955355,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004471958704458723,
      "loss": 5.1004,
      "mean_token_accuracy": 0.20689903497695922,
      "num_tokens": 45958655.0,
      "step": 20595
    },
    {
      "entropy": 5.429808187484741,
      "epoch": 2.313584545403493,
      "grad_norm": 0.921875,
      "learning_rate": 0.00044717002574274267,
      "loss": 5.1971,
      "mean_token_accuracy": 0.20794346630573274,
      "num_tokens": 45970767.0,
      "step": 20600
    },
    {
      "entropy": 5.363553285598755,
      "epoch": 2.314146122311451,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004471441755576527,
      "loss": 5.0016,
      "mean_token_accuracy": 0.21409626603126525,
      "num_tokens": 45981063.0,
      "step": 20605
    },
    {
      "entropy": 5.27598934173584,
      "epoch": 2.314707699219408,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044711831989142534,
      "loss": 4.9374,
      "mean_token_accuracy": 0.21327233910560608,
      "num_tokens": 45991963.0,
      "step": 20610
    },
    {
      "entropy": 5.341286516189575,
      "epoch": 2.3152692761273657,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044709245874488405,
      "loss": 5.2098,
      "mean_token_accuracy": 0.19699079394340516,
      "num_tokens": 46003460.0,
      "step": 20615
    },
    {
      "entropy": 5.416177272796631,
      "epoch": 2.3158308530353233,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004470665921188519,
      "loss": 5.1025,
      "mean_token_accuracy": 0.2030985325574875,
      "num_tokens": 46014338.0,
      "step": 20620
    },
    {
      "entropy": 5.294857120513916,
      "epoch": 2.3163924299432805,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044704072001415285,
      "loss": 4.9685,
      "mean_token_accuracy": 0.2151375263929367,
      "num_tokens": 46025282.0,
      "step": 20625
    },
    {
      "entropy": 5.33125319480896,
      "epoch": 2.316954006851238,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044701484243161055,
      "loss": 5.144,
      "mean_token_accuracy": 0.19839617758989334,
      "num_tokens": 46035303.0,
      "step": 20630
    },
    {
      "entropy": 5.34207854270935,
      "epoch": 2.317515583759196,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000446988959372049,
      "loss": 5.1245,
      "mean_token_accuracy": 0.20761884301900863,
      "num_tokens": 46046336.0,
      "step": 20635
    },
    {
      "entropy": 5.2889360904693605,
      "epoch": 2.3180771606671535,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004469630708362922,
      "loss": 5.0097,
      "mean_token_accuracy": 0.2025742456316948,
      "num_tokens": 46056862.0,
      "step": 20640
    },
    {
      "entropy": 5.455788469314575,
      "epoch": 2.3186387375751107,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004469371768251646,
      "loss": 5.1389,
      "mean_token_accuracy": 0.20514186322689057,
      "num_tokens": 46067785.0,
      "step": 20645
    },
    {
      "entropy": 5.350670099258423,
      "epoch": 2.3192003144830684,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004469112773394906,
      "loss": 5.1328,
      "mean_token_accuracy": 0.1942281648516655,
      "num_tokens": 46080553.0,
      "step": 20650
    },
    {
      "entropy": 5.352980279922486,
      "epoch": 2.319761891391026,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004468853723800948,
      "loss": 5.0941,
      "mean_token_accuracy": 0.1994601771235466,
      "num_tokens": 46090985.0,
      "step": 20655
    },
    {
      "entropy": 5.418760967254639,
      "epoch": 2.3203234682989837,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004468594619478021,
      "loss": 5.1753,
      "mean_token_accuracy": 0.19926313757896424,
      "num_tokens": 46101990.0,
      "step": 20660
    },
    {
      "entropy": 5.2929033756256105,
      "epoch": 2.320885045206941,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004468335460434375,
      "loss": 5.0291,
      "mean_token_accuracy": 0.21100093573331832,
      "num_tokens": 46112853.0,
      "step": 20665
    },
    {
      "entropy": 5.338849210739136,
      "epoch": 2.3214466221148986,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004468076246678261,
      "loss": 5.1362,
      "mean_token_accuracy": 0.20487669110298157,
      "num_tokens": 46123508.0,
      "step": 20670
    },
    {
      "entropy": 5.541608810424805,
      "epoch": 2.3220081990228563,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044678169782179334,
      "loss": 5.2333,
      "mean_token_accuracy": 0.19712902456521988,
      "num_tokens": 46135394.0,
      "step": 20675
    },
    {
      "entropy": 5.461752986907959,
      "epoch": 2.3225697759308135,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044675576550616466,
      "loss": 5.1671,
      "mean_token_accuracy": 0.20687037706375122,
      "num_tokens": 46146838.0,
      "step": 20680
    },
    {
      "entropy": 5.450082206726075,
      "epoch": 2.323131352838771,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044672982772176584,
      "loss": 5.0567,
      "mean_token_accuracy": 0.20978581309318542,
      "num_tokens": 46157116.0,
      "step": 20685
    },
    {
      "entropy": 5.328959226608276,
      "epoch": 2.323692929746729,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004467038844694227,
      "loss": 5.0672,
      "mean_token_accuracy": 0.20279800295829772,
      "num_tokens": 46168405.0,
      "step": 20690
    },
    {
      "entropy": 5.33976902961731,
      "epoch": 2.3242545066546865,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004466779357499612,
      "loss": 5.1044,
      "mean_token_accuracy": 0.20674179196357728,
      "num_tokens": 46179907.0,
      "step": 20695
    },
    {
      "entropy": 5.295328760147095,
      "epoch": 2.324816083562644,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044665198156420774,
      "loss": 5.0403,
      "mean_token_accuracy": 0.2022988721728325,
      "num_tokens": 46190750.0,
      "step": 20700
    },
    {
      "entropy": 5.36085991859436,
      "epoch": 2.3253776604706013,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044662602191298856,
      "loss": 5.0246,
      "mean_token_accuracy": 0.214484141767025,
      "num_tokens": 46201364.0,
      "step": 20705
    },
    {
      "entropy": 5.372038888931274,
      "epoch": 2.325939237378559,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044660005679713027,
      "loss": 5.0499,
      "mean_token_accuracy": 0.20225207805633544,
      "num_tokens": 46212341.0,
      "step": 20710
    },
    {
      "entropy": 5.367149591445923,
      "epoch": 2.3265008142865167,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044657408621745964,
      "loss": 5.144,
      "mean_token_accuracy": 0.19937959015369416,
      "num_tokens": 46223286.0,
      "step": 20715
    },
    {
      "entropy": 5.294501686096192,
      "epoch": 2.327062391194474,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004465481101748036,
      "loss": 5.16,
      "mean_token_accuracy": 0.2056562766432762,
      "num_tokens": 46234787.0,
      "step": 20720
    },
    {
      "entropy": 5.373676443099976,
      "epoch": 2.3276239681024316,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044652212866998914,
      "loss": 5.0101,
      "mean_token_accuracy": 0.21225742399692535,
      "num_tokens": 46244845.0,
      "step": 20725
    },
    {
      "entropy": 5.400301218032837,
      "epoch": 2.328185545010389,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044649614170384354,
      "loss": 5.1273,
      "mean_token_accuracy": 0.2043696954846382,
      "num_tokens": 46256352.0,
      "step": 20730
    },
    {
      "entropy": 5.5320557117462155,
      "epoch": 2.328747121918347,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044647014927719436,
      "loss": 5.2793,
      "mean_token_accuracy": 0.19823544919490815,
      "num_tokens": 46268535.0,
      "step": 20735
    },
    {
      "entropy": 5.361215400695801,
      "epoch": 2.329308698826304,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044644415139086913,
      "loss": 5.0975,
      "mean_token_accuracy": 0.2076890364289284,
      "num_tokens": 46279164.0,
      "step": 20740
    },
    {
      "entropy": 5.31693959236145,
      "epoch": 2.3298702757342618,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004464181480456955,
      "loss": 4.9724,
      "mean_token_accuracy": 0.21731749176979065,
      "num_tokens": 46289379.0,
      "step": 20745
    },
    {
      "entropy": 5.2908162593841555,
      "epoch": 2.3304318526422194,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004463921392425017,
      "loss": 5.0166,
      "mean_token_accuracy": 0.21073711961507796,
      "num_tokens": 46301023.0,
      "step": 20750
    },
    {
      "entropy": 5.324965953826904,
      "epoch": 2.330993429550177,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044636612498211566,
      "loss": 5.114,
      "mean_token_accuracy": 0.2069528579711914,
      "num_tokens": 46312043.0,
      "step": 20755
    },
    {
      "entropy": 5.3092041015625,
      "epoch": 2.3315550064581343,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044634010526536575,
      "loss": 5.0656,
      "mean_token_accuracy": 0.19923676997423173,
      "num_tokens": 46322808.0,
      "step": 20760
    },
    {
      "entropy": 5.439116382598877,
      "epoch": 2.332116583366092,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004463140800930804,
      "loss": 5.141,
      "mean_token_accuracy": 0.19661039113998413,
      "num_tokens": 46334260.0,
      "step": 20765
    },
    {
      "entropy": 5.408546209335327,
      "epoch": 2.3326781602740496,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004462880494660883,
      "loss": 5.1013,
      "mean_token_accuracy": 0.20469922572374344,
      "num_tokens": 46344245.0,
      "step": 20770
    },
    {
      "entropy": 5.318436098098755,
      "epoch": 2.333239737182007,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044626201338521834,
      "loss": 5.0489,
      "mean_token_accuracy": 0.20724338740110398,
      "num_tokens": 46356105.0,
      "step": 20775
    },
    {
      "entropy": 5.279531002044678,
      "epoch": 2.3338013140899645,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004462359718512993,
      "loss": 5.0327,
      "mean_token_accuracy": 0.20768912136554718,
      "num_tokens": 46367591.0,
      "step": 20780
    },
    {
      "entropy": 5.32373456954956,
      "epoch": 2.334362890997922,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044620992486516073,
      "loss": 5.0443,
      "mean_token_accuracy": 0.2039280951023102,
      "num_tokens": 46379022.0,
      "step": 20785
    },
    {
      "entropy": 5.3038414478302,
      "epoch": 2.33492446790588,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004461838724276316,
      "loss": 5.0661,
      "mean_token_accuracy": 0.2138448730111122,
      "num_tokens": 46390988.0,
      "step": 20790
    },
    {
      "entropy": 5.422853660583496,
      "epoch": 2.3354860448138375,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044615781453954164,
      "loss": 5.1383,
      "mean_token_accuracy": 0.2061695486307144,
      "num_tokens": 46401576.0,
      "step": 20795
    },
    {
      "entropy": 5.353518342971801,
      "epoch": 2.3360476217217947,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044613175120172035,
      "loss": 5.0302,
      "mean_token_accuracy": 0.2067776545882225,
      "num_tokens": 46413114.0,
      "step": 20800
    },
    {
      "entropy": 5.2473245620727536,
      "epoch": 2.3366091986297524,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004461056824149979,
      "loss": 5.0089,
      "mean_token_accuracy": 0.2121535986661911,
      "num_tokens": 46424000.0,
      "step": 20805
    },
    {
      "entropy": 5.362704086303711,
      "epoch": 2.33717077553771,
      "grad_norm": 1.0625,
      "learning_rate": 0.000446079608180204,
      "loss": 5.0823,
      "mean_token_accuracy": 0.2055947333574295,
      "num_tokens": 46435392.0,
      "step": 20810
    },
    {
      "entropy": 5.304933309555054,
      "epoch": 2.3377323524456672,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004460535284981692,
      "loss": 4.9893,
      "mean_token_accuracy": 0.20650942176580428,
      "num_tokens": 46445876.0,
      "step": 20815
    },
    {
      "entropy": 5.392146110534668,
      "epoch": 2.338293929353625,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004460274433697235,
      "loss": 5.0772,
      "mean_token_accuracy": 0.20253685116767883,
      "num_tokens": 46457175.0,
      "step": 20820
    },
    {
      "entropy": 5.41105465888977,
      "epoch": 2.3388555062615826,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044600135279569784,
      "loss": 5.2254,
      "mean_token_accuracy": 0.19966565519571305,
      "num_tokens": 46469100.0,
      "step": 20825
    },
    {
      "entropy": 5.306105089187622,
      "epoch": 2.3394170831695402,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044597525677692273,
      "loss": 5.0345,
      "mean_token_accuracy": 0.20909464955329896,
      "num_tokens": 46481221.0,
      "step": 20830
    },
    {
      "entropy": 5.289729452133178,
      "epoch": 2.3399786600774974,
      "grad_norm": 1.125,
      "learning_rate": 0.0004459491553142291,
      "loss": 4.9796,
      "mean_token_accuracy": 0.21708090901374816,
      "num_tokens": 46492127.0,
      "step": 20835
    },
    {
      "entropy": 5.369559860229492,
      "epoch": 2.340540236985455,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044592304840844813,
      "loss": 5.0813,
      "mean_token_accuracy": 0.2004895880818367,
      "num_tokens": 46502880.0,
      "step": 20840
    },
    {
      "entropy": 5.342006254196167,
      "epoch": 2.3411018138934128,
      "grad_norm": 1.0625,
      "learning_rate": 0.000445896936060411,
      "loss": 5.0147,
      "mean_token_accuracy": 0.20563847422599793,
      "num_tokens": 46513483.0,
      "step": 20845
    },
    {
      "entropy": 5.339911699295044,
      "epoch": 2.3416633908013704,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044587081827094906,
      "loss": 5.1164,
      "mean_token_accuracy": 0.20118605047464372,
      "num_tokens": 46526110.0,
      "step": 20850
    },
    {
      "entropy": 5.346368360519409,
      "epoch": 2.3422249677093276,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044584469504089407,
      "loss": 5.0851,
      "mean_token_accuracy": 0.20454346388578415,
      "num_tokens": 46537832.0,
      "step": 20855
    },
    {
      "entropy": 5.306882715225219,
      "epoch": 2.3427865446172853,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004458185663710777,
      "loss": 5.0077,
      "mean_token_accuracy": 0.20749859511852264,
      "num_tokens": 46548553.0,
      "step": 20860
    },
    {
      "entropy": 5.4293944358825685,
      "epoch": 2.343348121525243,
      "grad_norm": 1.0,
      "learning_rate": 0.0004457924322623319,
      "loss": 5.1958,
      "mean_token_accuracy": 0.2043108746409416,
      "num_tokens": 46561095.0,
      "step": 20865
    },
    {
      "entropy": 5.466595268249511,
      "epoch": 2.3439096984332,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044576629271548884,
      "loss": 5.1681,
      "mean_token_accuracy": 0.20328537076711656,
      "num_tokens": 46572070.0,
      "step": 20870
    },
    {
      "entropy": 5.511287021636963,
      "epoch": 2.344471275341158,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004457401477313808,
      "loss": 5.2049,
      "mean_token_accuracy": 0.19908317923545837,
      "num_tokens": 46583355.0,
      "step": 20875
    },
    {
      "entropy": 5.332150268554687,
      "epoch": 2.3450328522491155,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004457139973108403,
      "loss": 5.0944,
      "mean_token_accuracy": 0.20120721012353898,
      "num_tokens": 46594909.0,
      "step": 20880
    },
    {
      "entropy": 5.3933454036712645,
      "epoch": 2.345594429157073,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004456878414546998,
      "loss": 5.1246,
      "mean_token_accuracy": 0.20797940492630004,
      "num_tokens": 46606245.0,
      "step": 20885
    },
    {
      "entropy": 5.3141035556793215,
      "epoch": 2.346156006065031,
      "grad_norm": 1.0,
      "learning_rate": 0.00044566168016379234,
      "loss": 5.0272,
      "mean_token_accuracy": 0.2066301017999649,
      "num_tokens": 46617761.0,
      "step": 20890
    },
    {
      "entropy": 5.3618896484375,
      "epoch": 2.346717582972988,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004456355134389507,
      "loss": 5.068,
      "mean_token_accuracy": 0.20076974630355834,
      "num_tokens": 46628947.0,
      "step": 20895
    },
    {
      "entropy": 5.366351127624512,
      "epoch": 2.3472791598809457,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044560934128100814,
      "loss": 5.0721,
      "mean_token_accuracy": 0.20674896836280823,
      "num_tokens": 46640243.0,
      "step": 20900
    },
    {
      "entropy": 5.276374101638794,
      "epoch": 2.3478407367889034,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044558316369079805,
      "loss": 5.0148,
      "mean_token_accuracy": 0.20418880134820938,
      "num_tokens": 46650328.0,
      "step": 20905
    },
    {
      "entropy": 5.288998699188232,
      "epoch": 2.3484023136968606,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044555698066915385,
      "loss": 5.0102,
      "mean_token_accuracy": 0.2062876433134079,
      "num_tokens": 46661770.0,
      "step": 20910
    },
    {
      "entropy": 5.328490352630615,
      "epoch": 2.3489638906048182,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004455307922169092,
      "loss": 4.9623,
      "mean_token_accuracy": 0.2121179297566414,
      "num_tokens": 46673075.0,
      "step": 20915
    },
    {
      "entropy": 5.380064582824707,
      "epoch": 2.349525467512776,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004455045983348979,
      "loss": 5.1158,
      "mean_token_accuracy": 0.19953811764717103,
      "num_tokens": 46684040.0,
      "step": 20920
    },
    {
      "entropy": 5.288325309753418,
      "epoch": 2.3500870444207336,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004454783990239542,
      "loss": 4.9766,
      "mean_token_accuracy": 0.2111085072159767,
      "num_tokens": 46696295.0,
      "step": 20925
    },
    {
      "entropy": 5.307677841186523,
      "epoch": 2.350648621328691,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044545219428491215,
      "loss": 5.0812,
      "mean_token_accuracy": 0.20409184396266938,
      "num_tokens": 46706399.0,
      "step": 20930
    },
    {
      "entropy": 5.402581357955933,
      "epoch": 2.3512101982366485,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044542598411860605,
      "loss": 5.1941,
      "mean_token_accuracy": 0.19729759097099303,
      "num_tokens": 46717644.0,
      "step": 20935
    },
    {
      "entropy": 5.50269455909729,
      "epoch": 2.351771775144606,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004453997685258706,
      "loss": 5.1528,
      "mean_token_accuracy": 0.20562772899866105,
      "num_tokens": 46728556.0,
      "step": 20940
    },
    {
      "entropy": 5.36421389579773,
      "epoch": 2.3523333520525638,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044537354750754033,
      "loss": 5.07,
      "mean_token_accuracy": 0.2082500070333481,
      "num_tokens": 46739730.0,
      "step": 20945
    },
    {
      "entropy": 5.424343490600586,
      "epoch": 2.352894928960521,
      "grad_norm": 1.125,
      "learning_rate": 0.0004453473210644503,
      "loss": 5.1691,
      "mean_token_accuracy": 0.20136818438768386,
      "num_tokens": 46752349.0,
      "step": 20950
    },
    {
      "entropy": 5.3920519828796385,
      "epoch": 2.3534565058684787,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004453210891974355,
      "loss": 5.1412,
      "mean_token_accuracy": 0.1967243328690529,
      "num_tokens": 46763771.0,
      "step": 20955
    },
    {
      "entropy": 5.363690519332886,
      "epoch": 2.3540180827764363,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044529485190733113,
      "loss": 5.059,
      "mean_token_accuracy": 0.20908091515302657,
      "num_tokens": 46773371.0,
      "step": 20960
    },
    {
      "entropy": 5.322858619689941,
      "epoch": 2.3545796596843935,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004452686091949727,
      "loss": 5.045,
      "mean_token_accuracy": 0.2101862370967865,
      "num_tokens": 46784244.0,
      "step": 20965
    },
    {
      "entropy": 5.366339254379272,
      "epoch": 2.355141236592351,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044524236106119563,
      "loss": 5.0939,
      "mean_token_accuracy": 0.20297334641218184,
      "num_tokens": 46795373.0,
      "step": 20970
    },
    {
      "entropy": 5.461482334136963,
      "epoch": 2.355702813500309,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044521610750683586,
      "loss": 5.1341,
      "mean_token_accuracy": 0.20529650896787643,
      "num_tokens": 46806342.0,
      "step": 20975
    },
    {
      "entropy": 5.312036180496216,
      "epoch": 2.3562643904082665,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004451898485327291,
      "loss": 5.0328,
      "mean_token_accuracy": 0.2112785443663597,
      "num_tokens": 46817576.0,
      "step": 20980
    },
    {
      "entropy": 5.309410810470581,
      "epoch": 2.356825967316224,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004451635841397116,
      "loss": 5.1125,
      "mean_token_accuracy": 0.20146515518426894,
      "num_tokens": 46829335.0,
      "step": 20985
    },
    {
      "entropy": 5.407531309127807,
      "epoch": 2.3573875442241814,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004451373143286195,
      "loss": 5.1378,
      "mean_token_accuracy": 0.2028006926178932,
      "num_tokens": 46841334.0,
      "step": 20990
    },
    {
      "entropy": 5.466504287719727,
      "epoch": 2.357949121132139,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044511103910028945,
      "loss": 5.2982,
      "mean_token_accuracy": 0.19948375523090361,
      "num_tokens": 46852418.0,
      "step": 20995
    },
    {
      "entropy": 5.42905626296997,
      "epoch": 2.3585106980400967,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004450847584555579,
      "loss": 5.1532,
      "mean_token_accuracy": 0.19869986474514006,
      "num_tokens": 46863743.0,
      "step": 21000
    },
    {
      "epoch": 2.3585106980400967,
      "eval_entropy": 5.238582449842428,
      "eval_loss": 5.2472052574157715,
      "eval_mean_token_accuracy": 0.20559493159533684,
      "eval_num_tokens": 46863743.0,
      "eval_runtime": 24.0974,
      "eval_samples_per_second": 1286.988,
      "eval_steps_per_second": 160.889,
      "step": 21000
    },
    {
      "entropy": 5.3374021530151365,
      "epoch": 2.359072274948054,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004450584723952617,
      "loss": 5.0062,
      "mean_token_accuracy": 0.2117014765739441,
      "num_tokens": 46873628.0,
      "step": 21005
    },
    {
      "entropy": 5.381721639633179,
      "epoch": 2.3596338518560116,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004450321809202377,
      "loss": 5.178,
      "mean_token_accuracy": 0.19399171620607375,
      "num_tokens": 46885288.0,
      "step": 21010
    },
    {
      "entropy": 5.411822843551636,
      "epoch": 2.3601954287639693,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004450058840313232,
      "loss": 5.1432,
      "mean_token_accuracy": 0.2032656654715538,
      "num_tokens": 46897027.0,
      "step": 21015
    },
    {
      "entropy": 5.340396213531494,
      "epoch": 2.360757005671927,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044497958172935536,
      "loss": 5.0077,
      "mean_token_accuracy": 0.21029412001371384,
      "num_tokens": 46907947.0,
      "step": 21020
    },
    {
      "entropy": 5.315897798538208,
      "epoch": 2.361318582579884,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004449532740151717,
      "loss": 5.074,
      "mean_token_accuracy": 0.21006755381822587,
      "num_tokens": 46918334.0,
      "step": 21025
    },
    {
      "entropy": 5.366293478012085,
      "epoch": 2.361880159487842,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044492696088960985,
      "loss": 5.1173,
      "mean_token_accuracy": 0.20512246191501618,
      "num_tokens": 46929710.0,
      "step": 21030
    },
    {
      "entropy": 5.368923807144165,
      "epoch": 2.3624417363957995,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004449006423535078,
      "loss": 5.081,
      "mean_token_accuracy": 0.20296206325292587,
      "num_tokens": 46941436.0,
      "step": 21035
    },
    {
      "entropy": 5.429467868804932,
      "epoch": 2.363003313303757,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044487431840770325,
      "loss": 5.1193,
      "mean_token_accuracy": 0.20308544039726256,
      "num_tokens": 46953673.0,
      "step": 21040
    },
    {
      "entropy": 5.281364107131958,
      "epoch": 2.3635648902117143,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004448479890530347,
      "loss": 5.0165,
      "mean_token_accuracy": 0.21024357825517653,
      "num_tokens": 46964412.0,
      "step": 21045
    },
    {
      "entropy": 5.3066023826599125,
      "epoch": 2.364126467119672,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044482165429034014,
      "loss": 5.0644,
      "mean_token_accuracy": 0.2077476754784584,
      "num_tokens": 46975159.0,
      "step": 21050
    },
    {
      "entropy": 5.381211042404175,
      "epoch": 2.3646880440276297,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004447953141204583,
      "loss": 5.0497,
      "mean_token_accuracy": 0.2119266077876091,
      "num_tokens": 46987165.0,
      "step": 21055
    },
    {
      "entropy": 5.386941003799438,
      "epoch": 2.365249620935587,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044476896854422785,
      "loss": 5.1645,
      "mean_token_accuracy": 0.1914224937558174,
      "num_tokens": 46999085.0,
      "step": 21060
    },
    {
      "entropy": 5.381426429748535,
      "epoch": 2.3658111978435445,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044474261756248745,
      "loss": 5.1416,
      "mean_token_accuracy": 0.19897047877311708,
      "num_tokens": 47010970.0,
      "step": 21065
    },
    {
      "entropy": 5.419819688796997,
      "epoch": 2.366372774751502,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004447162611760764,
      "loss": 5.0639,
      "mean_token_accuracy": 0.2126307189464569,
      "num_tokens": 47021396.0,
      "step": 21070
    },
    {
      "entropy": 5.42234435081482,
      "epoch": 2.36693435165946,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004446898993858338,
      "loss": 5.1077,
      "mean_token_accuracy": 0.2031111538410187,
      "num_tokens": 47033394.0,
      "step": 21075
    },
    {
      "entropy": 5.371638679504395,
      "epoch": 2.3674959285674175,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044466353219259884,
      "loss": 5.1087,
      "mean_token_accuracy": 0.2051181063055992,
      "num_tokens": 47044688.0,
      "step": 21080
    },
    {
      "entropy": 5.45648684501648,
      "epoch": 2.3680575054753747,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004446371595972114,
      "loss": 5.1574,
      "mean_token_accuracy": 0.202251698076725,
      "num_tokens": 47057483.0,
      "step": 21085
    },
    {
      "entropy": 5.407315063476562,
      "epoch": 2.3686190823833324,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004446107816005109,
      "loss": 5.0895,
      "mean_token_accuracy": 0.2069546952843666,
      "num_tokens": 47069031.0,
      "step": 21090
    },
    {
      "entropy": 5.311635875701905,
      "epoch": 2.36918065929129,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004445843982033373,
      "loss": 5.0885,
      "mean_token_accuracy": 0.2064369797706604,
      "num_tokens": 47080998.0,
      "step": 21095
    },
    {
      "entropy": 5.3121264457702635,
      "epoch": 2.3697422361992473,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004445580094065307,
      "loss": 5.0952,
      "mean_token_accuracy": 0.2056901052594185,
      "num_tokens": 47092065.0,
      "step": 21100
    },
    {
      "entropy": 5.394021797180176,
      "epoch": 2.370303813107205,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004445316152109313,
      "loss": 5.0858,
      "mean_token_accuracy": 0.19940893054008485,
      "num_tokens": 47102207.0,
      "step": 21105
    },
    {
      "entropy": 5.363934469223023,
      "epoch": 2.3708653900151626,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004445052156173796,
      "loss": 5.067,
      "mean_token_accuracy": 0.2071093812584877,
      "num_tokens": 47112675.0,
      "step": 21110
    },
    {
      "entropy": 5.372884225845337,
      "epoch": 2.3714269669231203,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004444788106267161,
      "loss": 5.1076,
      "mean_token_accuracy": 0.20639416873455046,
      "num_tokens": 47123728.0,
      "step": 21115
    },
    {
      "entropy": 5.3905822277069095,
      "epoch": 2.3719885438310775,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004444524002397815,
      "loss": 5.0971,
      "mean_token_accuracy": 0.19960644841194153,
      "num_tokens": 47136194.0,
      "step": 21120
    },
    {
      "entropy": 5.288156127929687,
      "epoch": 2.372550120739035,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004444259844574167,
      "loss": 4.9544,
      "mean_token_accuracy": 0.20850107818841934,
      "num_tokens": 47146539.0,
      "step": 21125
    },
    {
      "entropy": 5.376447343826294,
      "epoch": 2.373111697646993,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044439956328046296,
      "loss": 5.1525,
      "mean_token_accuracy": 0.20196831375360488,
      "num_tokens": 47158449.0,
      "step": 21130
    },
    {
      "entropy": 5.331133890151977,
      "epoch": 2.3736732745549505,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004443731367097613,
      "loss": 4.9846,
      "mean_token_accuracy": 0.21512770056724548,
      "num_tokens": 47169687.0,
      "step": 21135
    },
    {
      "entropy": 5.2088244438171385,
      "epoch": 2.3742348514629077,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044434670474615334,
      "loss": 4.8909,
      "mean_token_accuracy": 0.2164049044251442,
      "num_tokens": 47179731.0,
      "step": 21140
    },
    {
      "entropy": 5.3058586597442625,
      "epoch": 2.3747964283708654,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044432026739048064,
      "loss": 5.0438,
      "mean_token_accuracy": 0.21665866672992706,
      "num_tokens": 47189925.0,
      "step": 21145
    },
    {
      "entropy": 5.366073417663574,
      "epoch": 2.375358005278823,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000444293824643585,
      "loss": 5.0658,
      "mean_token_accuracy": 0.2092925563454628,
      "num_tokens": 47200637.0,
      "step": 21150
    },
    {
      "entropy": 5.359606838226318,
      "epoch": 2.3759195821867802,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004442673765063082,
      "loss": 5.0836,
      "mean_token_accuracy": 0.20663110613822938,
      "num_tokens": 47211969.0,
      "step": 21155
    },
    {
      "entropy": 5.4034665584564205,
      "epoch": 2.376481159094738,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004442409229794925,
      "loss": 5.2074,
      "mean_token_accuracy": 0.20302105993032454,
      "num_tokens": 47223150.0,
      "step": 21160
    },
    {
      "entropy": 5.426743698120117,
      "epoch": 2.3770427360026956,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004442144640639803,
      "loss": 5.1688,
      "mean_token_accuracy": 0.19774141162633896,
      "num_tokens": 47234403.0,
      "step": 21165
    },
    {
      "entropy": 5.363220167160034,
      "epoch": 2.377604312910653,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004441879997606138,
      "loss": 4.9913,
      "mean_token_accuracy": 0.21021130830049514,
      "num_tokens": 47245800.0,
      "step": 21170
    },
    {
      "entropy": 5.337976455688477,
      "epoch": 2.378165889818611,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004441615300702358,
      "loss": 5.0669,
      "mean_token_accuracy": 0.20077298134565352,
      "num_tokens": 47256109.0,
      "step": 21175
    },
    {
      "entropy": 5.328191471099854,
      "epoch": 2.378727466726568,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004441350549936891,
      "loss": 5.0056,
      "mean_token_accuracy": 0.21040431559085845,
      "num_tokens": 47266502.0,
      "step": 21180
    },
    {
      "entropy": 5.398164558410644,
      "epoch": 2.3792890436345258,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004441085745318167,
      "loss": 5.0824,
      "mean_token_accuracy": 0.2066683679819107,
      "num_tokens": 47278866.0,
      "step": 21185
    },
    {
      "entropy": 5.267224311828613,
      "epoch": 2.3798506205424834,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044408208868546164,
      "loss": 5.0423,
      "mean_token_accuracy": 0.2129942610859871,
      "num_tokens": 47289449.0,
      "step": 21190
    },
    {
      "entropy": 5.3325728416442875,
      "epoch": 2.3804121974504406,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004440555974554673,
      "loss": 5.1866,
      "mean_token_accuracy": 0.2049464464187622,
      "num_tokens": 47301026.0,
      "step": 21195
    },
    {
      "entropy": 5.407817649841308,
      "epoch": 2.3809737743583983,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044402910084267717,
      "loss": 5.0744,
      "mean_token_accuracy": 0.2051403269171715,
      "num_tokens": 47311579.0,
      "step": 21200
    },
    {
      "entropy": 5.379408264160157,
      "epoch": 2.381535351266356,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004440025988479349,
      "loss": 5.0679,
      "mean_token_accuracy": 0.20942533761262894,
      "num_tokens": 47322042.0,
      "step": 21205
    },
    {
      "entropy": 5.361328220367431,
      "epoch": 2.3820969281743136,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044397609147208436,
      "loss": 5.0352,
      "mean_token_accuracy": 0.21126195937395095,
      "num_tokens": 47332821.0,
      "step": 21210
    },
    {
      "entropy": 5.333816909790039,
      "epoch": 2.382658505082271,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004439495787159695,
      "loss": 5.0835,
      "mean_token_accuracy": 0.20604147762060165,
      "num_tokens": 47344221.0,
      "step": 21215
    },
    {
      "entropy": 5.480521249771118,
      "epoch": 2.3832200819902285,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044392306058043453,
      "loss": 5.2137,
      "mean_token_accuracy": 0.20230868458747864,
      "num_tokens": 47356424.0,
      "step": 21220
    },
    {
      "entropy": 5.375629234313965,
      "epoch": 2.383781658898186,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004438965370663238,
      "loss": 5.0432,
      "mean_token_accuracy": 0.21391202062368392,
      "num_tokens": 47367546.0,
      "step": 21225
    },
    {
      "entropy": 5.363240098953247,
      "epoch": 2.384343235806144,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004438700081744819,
      "loss": 5.1462,
      "mean_token_accuracy": 0.2024271696805954,
      "num_tokens": 47378281.0,
      "step": 21230
    },
    {
      "entropy": 5.311780071258545,
      "epoch": 2.384904812714101,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044384347390575337,
      "loss": 5.0383,
      "mean_token_accuracy": 0.2100851520895958,
      "num_tokens": 47389233.0,
      "step": 21235
    },
    {
      "entropy": 5.326208972930909,
      "epoch": 2.3854663896220587,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004438169342609831,
      "loss": 5.0831,
      "mean_token_accuracy": 0.205342698097229,
      "num_tokens": 47400114.0,
      "step": 21240
    },
    {
      "entropy": 5.495392274856568,
      "epoch": 2.3860279665300164,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044379038924101627,
      "loss": 5.2404,
      "mean_token_accuracy": 0.19837034344673157,
      "num_tokens": 47411926.0,
      "step": 21245
    },
    {
      "entropy": 5.445928573608398,
      "epoch": 2.3865895434379736,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004437638388466979,
      "loss": 5.1008,
      "mean_token_accuracy": 0.20677115470170976,
      "num_tokens": 47424048.0,
      "step": 21250
    },
    {
      "entropy": 5.297302722930908,
      "epoch": 2.3871511203459312,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044373728307887353,
      "loss": 5.0459,
      "mean_token_accuracy": 0.20979386419057847,
      "num_tokens": 47435581.0,
      "step": 21255
    },
    {
      "entropy": 5.340604686737061,
      "epoch": 2.387712697253889,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044371072193838857,
      "loss": 5.0405,
      "mean_token_accuracy": 0.21010734736919404,
      "num_tokens": 47446831.0,
      "step": 21260
    },
    {
      "entropy": 5.283810997009278,
      "epoch": 2.3882742741618466,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044368415542608874,
      "loss": 5.0128,
      "mean_token_accuracy": 0.2058903843164444,
      "num_tokens": 47457885.0,
      "step": 21265
    },
    {
      "entropy": 5.363905239105224,
      "epoch": 2.3888358510698042,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044365758354282005,
      "loss": 5.117,
      "mean_token_accuracy": 0.19727576076984404,
      "num_tokens": 47470195.0,
      "step": 21270
    },
    {
      "entropy": 5.3871876239776615,
      "epoch": 2.3893974279777614,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004436310062894285,
      "loss": 5.0701,
      "mean_token_accuracy": 0.21267375200986863,
      "num_tokens": 47481068.0,
      "step": 21275
    },
    {
      "entropy": 5.393402576446533,
      "epoch": 2.389959004885719,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004436044236667603,
      "loss": 5.0347,
      "mean_token_accuracy": 0.2109733283519745,
      "num_tokens": 47491409.0,
      "step": 21280
    },
    {
      "entropy": 5.315866613388062,
      "epoch": 2.3905205817936768,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004435778356756618,
      "loss": 5.1119,
      "mean_token_accuracy": 0.20527178049087524,
      "num_tokens": 47504440.0,
      "step": 21285
    },
    {
      "entropy": 5.330333280563354,
      "epoch": 2.391082158701634,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044355124231697967,
      "loss": 5.0112,
      "mean_token_accuracy": 0.2151612713932991,
      "num_tokens": 47514370.0,
      "step": 21290
    },
    {
      "entropy": 5.3594193935394285,
      "epoch": 2.3916437356095916,
      "grad_norm": 1.125,
      "learning_rate": 0.0004435246435915607,
      "loss": 5.0476,
      "mean_token_accuracy": 0.21239639967679977,
      "num_tokens": 47525415.0,
      "step": 21295
    },
    {
      "entropy": 5.354748868942261,
      "epoch": 2.3922053125175493,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004434980395002516,
      "loss": 5.1312,
      "mean_token_accuracy": 0.20504768043756486,
      "num_tokens": 47537029.0,
      "step": 21300
    },
    {
      "entropy": 5.374898624420166,
      "epoch": 2.392766889425507,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044347143004389955,
      "loss": 5.0367,
      "mean_token_accuracy": 0.20895216912031173,
      "num_tokens": 47550434.0,
      "step": 21305
    },
    {
      "entropy": 5.368861293792724,
      "epoch": 2.393328466333464,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004434448152233519,
      "loss": 5.0609,
      "mean_token_accuracy": 0.20924942195415497,
      "num_tokens": 47560985.0,
      "step": 21310
    },
    {
      "entropy": 5.385549116134643,
      "epoch": 2.393890043241422,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044341819503945594,
      "loss": 5.1376,
      "mean_token_accuracy": 0.20217866748571395,
      "num_tokens": 47573207.0,
      "step": 21315
    },
    {
      "entropy": 5.315194320678711,
      "epoch": 2.3944516201493795,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004433915694930594,
      "loss": 4.9663,
      "mean_token_accuracy": 0.21697862297296525,
      "num_tokens": 47584666.0,
      "step": 21320
    },
    {
      "entropy": 5.3793858051300045,
      "epoch": 2.395013197057337,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004433649385850099,
      "loss": 5.0702,
      "mean_token_accuracy": 0.20794228315353394,
      "num_tokens": 47598017.0,
      "step": 21325
    },
    {
      "entropy": 5.399866199493408,
      "epoch": 2.3955747739652944,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004433383023161555,
      "loss": 5.2324,
      "mean_token_accuracy": 0.19508096277713777,
      "num_tokens": 47610947.0,
      "step": 21330
    },
    {
      "entropy": 5.432730388641358,
      "epoch": 2.396136350873252,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004433116606873443,
      "loss": 5.1733,
      "mean_token_accuracy": 0.19949405044317245,
      "num_tokens": 47621393.0,
      "step": 21335
    },
    {
      "entropy": 5.478350353240967,
      "epoch": 2.3966979277812097,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044328501369942454,
      "loss": 5.1182,
      "mean_token_accuracy": 0.20150883495807648,
      "num_tokens": 47631615.0,
      "step": 21340
    },
    {
      "entropy": 5.333618974685669,
      "epoch": 2.3972595046891674,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004432583613532446,
      "loss": 5.0964,
      "mean_token_accuracy": 0.2039869949221611,
      "num_tokens": 47642813.0,
      "step": 21345
    },
    {
      "entropy": 5.518627595901489,
      "epoch": 2.3978210815971246,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004432317036496533,
      "loss": 5.164,
      "mean_token_accuracy": 0.20529311448335646,
      "num_tokens": 47654012.0,
      "step": 21350
    },
    {
      "entropy": 5.3348963260650635,
      "epoch": 2.3983826585050823,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044320504058949916,
      "loss": 5.0386,
      "mean_token_accuracy": 0.21518584638833999,
      "num_tokens": 47664575.0,
      "step": 21355
    },
    {
      "entropy": 5.317589473724365,
      "epoch": 2.39894423541304,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004431783721736314,
      "loss": 5.1066,
      "mean_token_accuracy": 0.20603702813386918,
      "num_tokens": 47674763.0,
      "step": 21360
    },
    {
      "entropy": 5.343067693710327,
      "epoch": 2.3995058123209976,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044315169840289896,
      "loss": 5.0596,
      "mean_token_accuracy": 0.20920476764440538,
      "num_tokens": 47685435.0,
      "step": 21365
    },
    {
      "entropy": 5.434263706207275,
      "epoch": 2.400067389228955,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004431250192781513,
      "loss": 5.2403,
      "mean_token_accuracy": 0.2019906908273697,
      "num_tokens": 47696634.0,
      "step": 21370
    },
    {
      "entropy": 5.3505446434021,
      "epoch": 2.4006289661369125,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004430983348002378,
      "loss": 5.1016,
      "mean_token_accuracy": 0.20051012188196182,
      "num_tokens": 47707737.0,
      "step": 21375
    },
    {
      "entropy": 5.456892108917236,
      "epoch": 2.40119054304487,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044307164497000813,
      "loss": 5.1061,
      "mean_token_accuracy": 0.20445989519357682,
      "num_tokens": 47718131.0,
      "step": 21380
    },
    {
      "entropy": 5.384807825088501,
      "epoch": 2.4017521199528273,
      "grad_norm": 1.25,
      "learning_rate": 0.00044304494978831204,
      "loss": 5.0771,
      "mean_token_accuracy": 0.20756487250328065,
      "num_tokens": 47730154.0,
      "step": 21385
    },
    {
      "entropy": 5.371616458892822,
      "epoch": 2.402313696860785,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004430182492559997,
      "loss": 5.0596,
      "mean_token_accuracy": 0.20688972771167755,
      "num_tokens": 47741893.0,
      "step": 21390
    },
    {
      "entropy": 5.3451500415802,
      "epoch": 2.4028752737687427,
      "grad_norm": 1.03125,
      "learning_rate": 0.000442991543373921,
      "loss": 5.1163,
      "mean_token_accuracy": 0.20658354908227922,
      "num_tokens": 47753430.0,
      "step": 21395
    },
    {
      "entropy": 5.297962665557861,
      "epoch": 2.4034368506767003,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004429648321429264,
      "loss": 5.0455,
      "mean_token_accuracy": 0.21113316863775253,
      "num_tokens": 47765181.0,
      "step": 21400
    },
    {
      "entropy": 5.454178667068481,
      "epoch": 2.4039984275846575,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044293811556386656,
      "loss": 5.2336,
      "mean_token_accuracy": 0.19440730661153793,
      "num_tokens": 47777211.0,
      "step": 21405
    },
    {
      "entropy": 5.342150402069092,
      "epoch": 2.404560004492615,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004429113936375918,
      "loss": 5.0627,
      "mean_token_accuracy": 0.20338311344385146,
      "num_tokens": 47788397.0,
      "step": 21410
    },
    {
      "entropy": 5.405908203125,
      "epoch": 2.405121581400573,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044288466636495323,
      "loss": 5.1504,
      "mean_token_accuracy": 0.20026626884937287,
      "num_tokens": 47799870.0,
      "step": 21415
    },
    {
      "entropy": 5.3871410369873045,
      "epoch": 2.4056831583085305,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044285793374680184,
      "loss": 5.1541,
      "mean_token_accuracy": 0.20365336388349534,
      "num_tokens": 47810696.0,
      "step": 21420
    },
    {
      "entropy": 5.402115058898926,
      "epoch": 2.4062447352164877,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044283119578398866,
      "loss": 5.1257,
      "mean_token_accuracy": 0.2062508076429367,
      "num_tokens": 47822579.0,
      "step": 21425
    },
    {
      "entropy": 5.3343689918518065,
      "epoch": 2.4068063121244454,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004428044524773651,
      "loss": 5.0393,
      "mean_token_accuracy": 0.20299503356218337,
      "num_tokens": 47832556.0,
      "step": 21430
    },
    {
      "entropy": 5.329568338394165,
      "epoch": 2.407367889032403,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004427777038277827,
      "loss": 5.1005,
      "mean_token_accuracy": 0.20297913253307343,
      "num_tokens": 47844409.0,
      "step": 21435
    },
    {
      "entropy": 5.378345966339111,
      "epoch": 2.4079294659403607,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044275094983609313,
      "loss": 5.0531,
      "mean_token_accuracy": 0.20710691064596176,
      "num_tokens": 47855666.0,
      "step": 21440
    },
    {
      "entropy": 5.269646120071411,
      "epoch": 2.408491042848318,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044272419050314833,
      "loss": 4.9852,
      "mean_token_accuracy": 0.2149094045162201,
      "num_tokens": 47866261.0,
      "step": 21445
    },
    {
      "entropy": 5.34073805809021,
      "epoch": 2.4090526197562756,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044269742582980017,
      "loss": 5.1811,
      "mean_token_accuracy": 0.1983111932873726,
      "num_tokens": 47878807.0,
      "step": 21450
    },
    {
      "entropy": 5.393631267547607,
      "epoch": 2.4096141966642333,
      "grad_norm": 1.09375,
      "learning_rate": 0.000442670655816901,
      "loss": 5.0359,
      "mean_token_accuracy": 0.2114022970199585,
      "num_tokens": 47888935.0,
      "step": 21455
    },
    {
      "entropy": 5.399390792846679,
      "epoch": 2.410175773572191,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044264388046530304,
      "loss": 5.0567,
      "mean_token_accuracy": 0.21181343495845795,
      "num_tokens": 47900208.0,
      "step": 21460
    },
    {
      "entropy": 5.43604793548584,
      "epoch": 2.410737350480148,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044261709977585893,
      "loss": 5.199,
      "mean_token_accuracy": 0.20511537790298462,
      "num_tokens": 47911131.0,
      "step": 21465
    },
    {
      "entropy": 5.382884120941162,
      "epoch": 2.411298927388106,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004425903137494214,
      "loss": 5.1642,
      "mean_token_accuracy": 0.19661177545785904,
      "num_tokens": 47922601.0,
      "step": 21470
    },
    {
      "entropy": 5.444476556777954,
      "epoch": 2.4118605042960635,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004425635223868433,
      "loss": 5.1194,
      "mean_token_accuracy": 0.20322322845458984,
      "num_tokens": 47933168.0,
      "step": 21475
    },
    {
      "entropy": 5.443022584915161,
      "epoch": 2.4124220812040207,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044253672568897754,
      "loss": 5.189,
      "mean_token_accuracy": 0.1964871034026146,
      "num_tokens": 47945433.0,
      "step": 21480
    },
    {
      "entropy": 5.425508117675781,
      "epoch": 2.4129836581119783,
      "grad_norm": 1.0,
      "learning_rate": 0.0004425099236566775,
      "loss": 5.1195,
      "mean_token_accuracy": 0.20166653394699097,
      "num_tokens": 47957513.0,
      "step": 21485
    },
    {
      "entropy": 5.38068151473999,
      "epoch": 2.413545235019936,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044248311629079657,
      "loss": 5.0587,
      "mean_token_accuracy": 0.20086897015571595,
      "num_tokens": 47968817.0,
      "step": 21490
    },
    {
      "entropy": 5.385633420944214,
      "epoch": 2.4141068119278937,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004424563035921882,
      "loss": 5.001,
      "mean_token_accuracy": 0.2105054721236229,
      "num_tokens": 47979034.0,
      "step": 21495
    },
    {
      "entropy": 5.414537763595581,
      "epoch": 2.414668388835851,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044242948556170634,
      "loss": 5.1063,
      "mean_token_accuracy": 0.20826312750577927,
      "num_tokens": 47990459.0,
      "step": 21500
    },
    {
      "entropy": 5.383299207687378,
      "epoch": 2.4152299657438085,
      "grad_norm": 1.71875,
      "learning_rate": 0.00044240266220020455,
      "loss": 5.1574,
      "mean_token_accuracy": 0.20459922701120375,
      "num_tokens": 48000007.0,
      "step": 21505
    },
    {
      "entropy": 5.4030293941497805,
      "epoch": 2.415791542651766,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004423758335085371,
      "loss": 5.0871,
      "mean_token_accuracy": 0.20226266235113144,
      "num_tokens": 48012962.0,
      "step": 21510
    },
    {
      "entropy": 5.353014850616455,
      "epoch": 2.416353119559724,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044234899948755827,
      "loss": 5.0886,
      "mean_token_accuracy": 0.2036088839173317,
      "num_tokens": 48025916.0,
      "step": 21515
    },
    {
      "entropy": 5.343579864501953,
      "epoch": 2.416914696467681,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044232216013812226,
      "loss": 5.0906,
      "mean_token_accuracy": 0.20479766726493837,
      "num_tokens": 48036255.0,
      "step": 21520
    },
    {
      "entropy": 5.3736248970031735,
      "epoch": 2.4174762733756388,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004422953154610838,
      "loss": 5.09,
      "mean_token_accuracy": 0.20907227098941802,
      "num_tokens": 48045943.0,
      "step": 21525
    },
    {
      "entropy": 5.326147603988647,
      "epoch": 2.4180378502835964,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004422684654572977,
      "loss": 5.0414,
      "mean_token_accuracy": 0.20977475494146347,
      "num_tokens": 48056562.0,
      "step": 21530
    },
    {
      "entropy": 5.317258358001709,
      "epoch": 2.418599427191554,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044224161012761874,
      "loss": 5.034,
      "mean_token_accuracy": 0.20596093982458114,
      "num_tokens": 48066997.0,
      "step": 21535
    },
    {
      "entropy": 5.419650268554688,
      "epoch": 2.4191610040995113,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044221474947290205,
      "loss": 5.1442,
      "mean_token_accuracy": 0.2015213966369629,
      "num_tokens": 48077810.0,
      "step": 21540
    },
    {
      "entropy": 5.413149023056031,
      "epoch": 2.419722581007469,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004421878834940029,
      "loss": 5.2018,
      "mean_token_accuracy": 0.1995735615491867,
      "num_tokens": 48089422.0,
      "step": 21545
    },
    {
      "entropy": 5.342988777160644,
      "epoch": 2.4202841579154266,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044216101219177667,
      "loss": 5.0517,
      "mean_token_accuracy": 0.20487606525421143,
      "num_tokens": 48101114.0,
      "step": 21550
    },
    {
      "entropy": 5.367941427230835,
      "epoch": 2.4208457348233843,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044213413556707905,
      "loss": 5.0446,
      "mean_token_accuracy": 0.21157099157571793,
      "num_tokens": 48110991.0,
      "step": 21555
    },
    {
      "entropy": 5.379810619354248,
      "epoch": 2.4214073117313415,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004421072536207656,
      "loss": 5.0926,
      "mean_token_accuracy": 0.2113562285900116,
      "num_tokens": 48122889.0,
      "step": 21560
    },
    {
      "entropy": 5.389226770401001,
      "epoch": 2.421968888639299,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044208036635369247,
      "loss": 5.0966,
      "mean_token_accuracy": 0.20552474409341812,
      "num_tokens": 48133586.0,
      "step": 21565
    },
    {
      "entropy": 5.328506422042847,
      "epoch": 2.422530465547257,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004420534737667157,
      "loss": 5.0336,
      "mean_token_accuracy": 0.21256475597620011,
      "num_tokens": 48144060.0,
      "step": 21570
    },
    {
      "entropy": 5.441243743896484,
      "epoch": 2.423092042455214,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044202657586069153,
      "loss": 5.1304,
      "mean_token_accuracy": 0.2100416600704193,
      "num_tokens": 48155197.0,
      "step": 21575
    },
    {
      "entropy": 5.4348310947418215,
      "epoch": 2.4236536193631717,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044199967263647635,
      "loss": 5.1896,
      "mean_token_accuracy": 0.20013038963079452,
      "num_tokens": 48166990.0,
      "step": 21580
    },
    {
      "entropy": 5.441673517227173,
      "epoch": 2.4242151962711294,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004419727640949268,
      "loss": 5.1783,
      "mean_token_accuracy": 0.19734276384115218,
      "num_tokens": 48179454.0,
      "step": 21585
    },
    {
      "entropy": 5.313477182388306,
      "epoch": 2.424776773179087,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004419458502368998,
      "loss": 5.0641,
      "mean_token_accuracy": 0.20587768405675888,
      "num_tokens": 48192009.0,
      "step": 21590
    },
    {
      "entropy": 5.46348729133606,
      "epoch": 2.4253383500870442,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044191893106325207,
      "loss": 5.1302,
      "mean_token_accuracy": 0.2052372947335243,
      "num_tokens": 48203111.0,
      "step": 21595
    },
    {
      "entropy": 5.466365623474121,
      "epoch": 2.425899926995002,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004418920065748409,
      "loss": 5.2845,
      "mean_token_accuracy": 0.19821547269821166,
      "num_tokens": 48214766.0,
      "step": 21600
    },
    {
      "entropy": 5.417569017410278,
      "epoch": 2.4264615039029596,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044186507677252345,
      "loss": 5.2101,
      "mean_token_accuracy": 0.2083445593714714,
      "num_tokens": 48225061.0,
      "step": 21605
    },
    {
      "entropy": 5.358213090896607,
      "epoch": 2.427023080810917,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004418381416571573,
      "loss": 5.0877,
      "mean_token_accuracy": 0.20468733757734298,
      "num_tokens": 48236101.0,
      "step": 21610
    },
    {
      "entropy": 5.383041620254517,
      "epoch": 2.4275846577188744,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004418112012296,
      "loss": 5.0606,
      "mean_token_accuracy": 0.2028485968708992,
      "num_tokens": 48248080.0,
      "step": 21615
    },
    {
      "entropy": 5.368525218963623,
      "epoch": 2.428146234626832,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044178425549070934,
      "loss": 5.0107,
      "mean_token_accuracy": 0.20657033175230027,
      "num_tokens": 48259177.0,
      "step": 21620
    },
    {
      "entropy": 5.35972113609314,
      "epoch": 2.4287078115347898,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044175730444134326,
      "loss": 5.0519,
      "mean_token_accuracy": 0.204851333796978,
      "num_tokens": 48271345.0,
      "step": 21625
    },
    {
      "entropy": 5.368231773376465,
      "epoch": 2.4292693884427474,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044173034808236,
      "loss": 5.1425,
      "mean_token_accuracy": 0.2020396575331688,
      "num_tokens": 48284219.0,
      "step": 21630
    },
    {
      "entropy": 5.341426372528076,
      "epoch": 2.4298309653507046,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004417033864146177,
      "loss": 5.0315,
      "mean_token_accuracy": 0.20887597203254699,
      "num_tokens": 48294626.0,
      "step": 21635
    },
    {
      "entropy": 5.29458441734314,
      "epoch": 2.4303925422586623,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004416764194389751,
      "loss": 5.0572,
      "mean_token_accuracy": 0.20354630202054977,
      "num_tokens": 48305459.0,
      "step": 21640
    },
    {
      "entropy": 5.390269470214844,
      "epoch": 2.43095411916662,
      "grad_norm": 1.125,
      "learning_rate": 0.0004416494471562904,
      "loss": 5.057,
      "mean_token_accuracy": 0.21215395331382753,
      "num_tokens": 48316232.0,
      "step": 21645
    },
    {
      "entropy": 5.363577270507813,
      "epoch": 2.4315156960745776,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004416224695674229,
      "loss": 5.0295,
      "mean_token_accuracy": 0.2065650850534439,
      "num_tokens": 48326711.0,
      "step": 21650
    },
    {
      "entropy": 5.21864333152771,
      "epoch": 2.432077272982535,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044159548667323126,
      "loss": 4.846,
      "mean_token_accuracy": 0.23145456314086915,
      "num_tokens": 48336860.0,
      "step": 21655
    },
    {
      "entropy": 5.2605931758880615,
      "epoch": 2.4326388498904925,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004415684984745747,
      "loss": 5.1699,
      "mean_token_accuracy": 0.20005354434251785,
      "num_tokens": 48348503.0,
      "step": 21660
    },
    {
      "entropy": 5.37547459602356,
      "epoch": 2.43320042679845,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004415415049723126,
      "loss": 5.1633,
      "mean_token_accuracy": 0.19843111485242843,
      "num_tokens": 48359855.0,
      "step": 21665
    },
    {
      "entropy": 5.46258134841919,
      "epoch": 2.4337620037064074,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004415145061673043,
      "loss": 5.132,
      "mean_token_accuracy": 0.19958877861499785,
      "num_tokens": 48370188.0,
      "step": 21670
    },
    {
      "entropy": 5.37732048034668,
      "epoch": 2.434323580614365,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004414875020604096,
      "loss": 5.0386,
      "mean_token_accuracy": 0.2171775847673416,
      "num_tokens": 48381152.0,
      "step": 21675
    },
    {
      "entropy": 5.36230993270874,
      "epoch": 2.4348851575223227,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004414604926524882,
      "loss": 5.0828,
      "mean_token_accuracy": 0.20780738145112992,
      "num_tokens": 48392721.0,
      "step": 21680
    },
    {
      "entropy": 5.312788295745849,
      "epoch": 2.4354467344302804,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004414334779444002,
      "loss": 5.0688,
      "mean_token_accuracy": 0.20699834674596787,
      "num_tokens": 48404071.0,
      "step": 21685
    },
    {
      "entropy": 5.361199235916137,
      "epoch": 2.4360083113382376,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044140645793700573,
      "loss": 5.0845,
      "mean_token_accuracy": 0.19984843879938125,
      "num_tokens": 48416001.0,
      "step": 21690
    },
    {
      "entropy": 5.383803844451904,
      "epoch": 2.4365698882461952,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044137943263116515,
      "loss": 5.0984,
      "mean_token_accuracy": 0.20274689197540283,
      "num_tokens": 48426880.0,
      "step": 21695
    },
    {
      "entropy": 5.455490350723267,
      "epoch": 2.437131465154153,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004413524020277388,
      "loss": 5.1734,
      "mean_token_accuracy": 0.20182285755872725,
      "num_tokens": 48437147.0,
      "step": 21700
    },
    {
      "entropy": 5.28728461265564,
      "epoch": 2.4376930420621106,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004413253661275875,
      "loss": 4.9876,
      "mean_token_accuracy": 0.21211253553628923,
      "num_tokens": 48448706.0,
      "step": 21705
    },
    {
      "entropy": 5.312035846710205,
      "epoch": 2.438254618970068,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044129832493157195,
      "loss": 4.9968,
      "mean_token_accuracy": 0.21103363186120988,
      "num_tokens": 48458873.0,
      "step": 21710
    },
    {
      "entropy": 5.368663120269775,
      "epoch": 2.4388161958780255,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044127127844055335,
      "loss": 5.1508,
      "mean_token_accuracy": 0.1996288225054741,
      "num_tokens": 48470726.0,
      "step": 21715
    },
    {
      "entropy": 5.343744325637817,
      "epoch": 2.439377772785983,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044124422665539273,
      "loss": 4.9832,
      "mean_token_accuracy": 0.21300038844347,
      "num_tokens": 48481912.0,
      "step": 21720
    },
    {
      "entropy": 5.402991533279419,
      "epoch": 2.4399393496939408,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004412171695769515,
      "loss": 5.1013,
      "mean_token_accuracy": 0.21003120988607407,
      "num_tokens": 48492048.0,
      "step": 21725
    },
    {
      "entropy": 5.395315217971802,
      "epoch": 2.440500926601898,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044119010720609106,
      "loss": 5.1509,
      "mean_token_accuracy": 0.20506772249937058,
      "num_tokens": 48503190.0,
      "step": 21730
    },
    {
      "entropy": 5.3729332447052,
      "epoch": 2.4410625035098557,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044116303954367314,
      "loss": 5.0461,
      "mean_token_accuracy": 0.20855064541101456,
      "num_tokens": 48514956.0,
      "step": 21735
    },
    {
      "entropy": 5.474702453613281,
      "epoch": 2.4416240804178133,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004411359665905596,
      "loss": 5.1789,
      "mean_token_accuracy": 0.20507064759731292,
      "num_tokens": 48526372.0,
      "step": 21740
    },
    {
      "entropy": 5.275550651550293,
      "epoch": 2.442185657325771,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004411088883476125,
      "loss": 5.0343,
      "mean_token_accuracy": 0.21101831048727035,
      "num_tokens": 48536985.0,
      "step": 21745
    },
    {
      "entropy": 5.347469329833984,
      "epoch": 2.442747234233728,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044108180481569387,
      "loss": 5.0841,
      "mean_token_accuracy": 0.2066688969731331,
      "num_tokens": 48547505.0,
      "step": 21750
    },
    {
      "entropy": 5.3747179985046385,
      "epoch": 2.443308811141686,
      "grad_norm": 1.125,
      "learning_rate": 0.00044105471599566626,
      "loss": 5.1164,
      "mean_token_accuracy": 0.2047415629029274,
      "num_tokens": 48560104.0,
      "step": 21755
    },
    {
      "entropy": 5.4537523746490475,
      "epoch": 2.4438703880496435,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000441027621888392,
      "loss": 5.1728,
      "mean_token_accuracy": 0.20691749304533005,
      "num_tokens": 48571203.0,
      "step": 21760
    },
    {
      "entropy": 5.347192001342774,
      "epoch": 2.4444319649576007,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044100052249473393,
      "loss": 4.989,
      "mean_token_accuracy": 0.2079248383641243,
      "num_tokens": 48581480.0,
      "step": 21765
    },
    {
      "entropy": 5.368220806121826,
      "epoch": 2.4449935418655584,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004409734178155548,
      "loss": 5.0782,
      "mean_token_accuracy": 0.20826440453529357,
      "num_tokens": 48592811.0,
      "step": 21770
    },
    {
      "entropy": 5.331868839263916,
      "epoch": 2.445555118773516,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004409463078517178,
      "loss": 5.0852,
      "mean_token_accuracy": 0.21106485575437545,
      "num_tokens": 48604245.0,
      "step": 21775
    },
    {
      "entropy": 5.3845141410827635,
      "epoch": 2.4461166956814737,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044091919260408585,
      "loss": 5.0495,
      "mean_token_accuracy": 0.20782522857189178,
      "num_tokens": 48615138.0,
      "step": 21780
    },
    {
      "entropy": 5.3348390579223635,
      "epoch": 2.446678272589431,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004408920720735225,
      "loss": 5.1283,
      "mean_token_accuracy": 0.20286808162927628,
      "num_tokens": 48626084.0,
      "step": 21785
    },
    {
      "entropy": 5.340631771087646,
      "epoch": 2.4472398494973886,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044086494626089127,
      "loss": 4.9883,
      "mean_token_accuracy": 0.21088313609361647,
      "num_tokens": 48636425.0,
      "step": 21790
    },
    {
      "entropy": 5.321488952636718,
      "epoch": 2.4478014264053463,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004408378151670558,
      "loss": 5.1216,
      "mean_token_accuracy": 0.19760634899139404,
      "num_tokens": 48646941.0,
      "step": 21795
    },
    {
      "entropy": 5.340544414520264,
      "epoch": 2.448363003313304,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044081067879288,
      "loss": 4.9905,
      "mean_token_accuracy": 0.21365498900413513,
      "num_tokens": 48657481.0,
      "step": 21800
    },
    {
      "entropy": 5.430232524871826,
      "epoch": 2.448924580221261,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004407835371392279,
      "loss": 5.1126,
      "mean_token_accuracy": 0.20486358255147935,
      "num_tokens": 48669306.0,
      "step": 21805
    },
    {
      "entropy": 5.313171625137329,
      "epoch": 2.449486157129219,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004407563902069636,
      "loss": 5.0525,
      "mean_token_accuracy": 0.2081127017736435,
      "num_tokens": 48680789.0,
      "step": 21810
    },
    {
      "entropy": 5.417173290252686,
      "epoch": 2.4500477340371765,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004407292379969515,
      "loss": 5.1428,
      "mean_token_accuracy": 0.20203507989645003,
      "num_tokens": 48691811.0,
      "step": 21815
    },
    {
      "entropy": 5.39336314201355,
      "epoch": 2.450609310945134,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004407020805100563,
      "loss": 5.1483,
      "mean_token_accuracy": 0.2046603262424469,
      "num_tokens": 48704248.0,
      "step": 21820
    },
    {
      "entropy": 5.34074215888977,
      "epoch": 2.4511708878530913,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004406749177471425,
      "loss": 5.0434,
      "mean_token_accuracy": 0.2133680373430252,
      "num_tokens": 48714452.0,
      "step": 21825
    },
    {
      "entropy": 5.343944025039673,
      "epoch": 2.451732464761049,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044064774970907516,
      "loss": 5.0432,
      "mean_token_accuracy": 0.20868532955646515,
      "num_tokens": 48725481.0,
      "step": 21830
    },
    {
      "entropy": 5.413015985488892,
      "epoch": 2.4522940416690067,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004406205763967193,
      "loss": 5.134,
      "mean_token_accuracy": 0.20841966718435287,
      "num_tokens": 48737704.0,
      "step": 21835
    },
    {
      "entropy": 5.405245685577393,
      "epoch": 2.4528556185769643,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004405933978109398,
      "loss": 5.131,
      "mean_token_accuracy": 0.2081867814064026,
      "num_tokens": 48749085.0,
      "step": 21840
    },
    {
      "entropy": 5.418499517440796,
      "epoch": 2.4534171954849215,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044056621395260247,
      "loss": 5.1105,
      "mean_token_accuracy": 0.20573025345802307,
      "num_tokens": 48759432.0,
      "step": 21845
    },
    {
      "entropy": 5.296266841888428,
      "epoch": 2.453978772392879,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044053902482257265,
      "loss": 4.9964,
      "mean_token_accuracy": 0.2076605513691902,
      "num_tokens": 48769549.0,
      "step": 21850
    },
    {
      "entropy": 5.301059627532959,
      "epoch": 2.454540349300837,
      "grad_norm": 1.046875,
      "learning_rate": 0.000440511830421716,
      "loss": 5.0289,
      "mean_token_accuracy": 0.21154962629079818,
      "num_tokens": 48779630.0,
      "step": 21855
    },
    {
      "entropy": 5.36486849784851,
      "epoch": 2.455101926208794,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004404846307508985,
      "loss": 5.1181,
      "mean_token_accuracy": 0.20281594693660737,
      "num_tokens": 48790489.0,
      "step": 21860
    },
    {
      "entropy": 5.461833906173706,
      "epoch": 2.4556635031167517,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044045742581098624,
      "loss": 5.1616,
      "mean_token_accuracy": 0.20898225605487825,
      "num_tokens": 48802310.0,
      "step": 21865
    },
    {
      "entropy": 5.370346021652222,
      "epoch": 2.4562250800247094,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004404302156028453,
      "loss": 5.0297,
      "mean_token_accuracy": 0.21559318900108337,
      "num_tokens": 48812066.0,
      "step": 21870
    },
    {
      "entropy": 5.430401563644409,
      "epoch": 2.456786656932667,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044040300012734217,
      "loss": 5.2241,
      "mean_token_accuracy": 0.19679941534996032,
      "num_tokens": 48824477.0,
      "step": 21875
    },
    {
      "entropy": 5.436723899841309,
      "epoch": 2.4573482338406243,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004403757793853434,
      "loss": 5.093,
      "mean_token_accuracy": 0.20222220718860626,
      "num_tokens": 48835147.0,
      "step": 21880
    },
    {
      "entropy": 5.3566302299499515,
      "epoch": 2.457909810748582,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004403485533777157,
      "loss": 5.1133,
      "mean_token_accuracy": 0.19944419711828232,
      "num_tokens": 48846147.0,
      "step": 21885
    },
    {
      "entropy": 5.466863203048706,
      "epoch": 2.4584713876565396,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004403213221053258,
      "loss": 5.1859,
      "mean_token_accuracy": 0.2014068454504013,
      "num_tokens": 48857340.0,
      "step": 21890
    },
    {
      "entropy": 5.33150086402893,
      "epoch": 2.4590329645644973,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044029408556904093,
      "loss": 5.0362,
      "mean_token_accuracy": 0.20978858470916747,
      "num_tokens": 48868256.0,
      "step": 21895
    },
    {
      "entropy": 5.476415729522705,
      "epoch": 2.4595945414724545,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004402668437697283,
      "loss": 5.1398,
      "mean_token_accuracy": 0.20274382084608078,
      "num_tokens": 48879346.0,
      "step": 21900
    },
    {
      "entropy": 5.400439071655273,
      "epoch": 2.460156118380412,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044023959670825533,
      "loss": 5.1673,
      "mean_token_accuracy": 0.19905785471200943,
      "num_tokens": 48890939.0,
      "step": 21905
    },
    {
      "entropy": 5.357732343673706,
      "epoch": 2.46071769528837,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004402123443854894,
      "loss": 5.1491,
      "mean_token_accuracy": 0.20459777265787124,
      "num_tokens": 48902239.0,
      "step": 21910
    },
    {
      "entropy": 5.407272100448608,
      "epoch": 2.4612792721963275,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004401850868022984,
      "loss": 5.1285,
      "mean_token_accuracy": 0.20706406086683274,
      "num_tokens": 48913810.0,
      "step": 21915
    },
    {
      "entropy": 5.334176397323608,
      "epoch": 2.4618408491042847,
      "grad_norm": 1.125,
      "learning_rate": 0.0004401578239595501,
      "loss": 4.9904,
      "mean_token_accuracy": 0.21035299748182296,
      "num_tokens": 48925180.0,
      "step": 21920
    },
    {
      "entropy": 5.333799505233765,
      "epoch": 2.4624024260122424,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044013055585811274,
      "loss": 5.1234,
      "mean_token_accuracy": 0.19647221714258195,
      "num_tokens": 48937067.0,
      "step": 21925
    },
    {
      "entropy": 5.3777460098266605,
      "epoch": 2.4629640029202,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044010328249885434,
      "loss": 5.1621,
      "mean_token_accuracy": 0.19942909926176072,
      "num_tokens": 48949220.0,
      "step": 21930
    },
    {
      "entropy": 5.3795393943786625,
      "epoch": 2.4635255798281577,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004400760038826434,
      "loss": 5.0778,
      "mean_token_accuracy": 0.20324401706457138,
      "num_tokens": 48960604.0,
      "step": 21935
    },
    {
      "entropy": 5.342543411254883,
      "epoch": 2.464087156736115,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044004872001034856,
      "loss": 5.0323,
      "mean_token_accuracy": 0.2151407778263092,
      "num_tokens": 48970561.0,
      "step": 21940
    },
    {
      "entropy": 5.380719614028931,
      "epoch": 2.4646487336440726,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004400214308828384,
      "loss": 5.1132,
      "mean_token_accuracy": 0.1998044356703758,
      "num_tokens": 48981639.0,
      "step": 21945
    },
    {
      "entropy": 5.439582586288452,
      "epoch": 2.46521031055203,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043999413650098193,
      "loss": 5.1167,
      "mean_token_accuracy": 0.20541362166404725,
      "num_tokens": 48992872.0,
      "step": 21950
    },
    {
      "entropy": 5.403166913986206,
      "epoch": 2.4657718874599874,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004399668368656481,
      "loss": 5.1049,
      "mean_token_accuracy": 0.2002684086561203,
      "num_tokens": 49004942.0,
      "step": 21955
    },
    {
      "entropy": 5.420056247711182,
      "epoch": 2.466333464367945,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004399395319777062,
      "loss": 5.0862,
      "mean_token_accuracy": 0.20795905590057373,
      "num_tokens": 49016493.0,
      "step": 21960
    },
    {
      "entropy": 5.323580598831176,
      "epoch": 2.4668950412759028,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043991222183802564,
      "loss": 5.0882,
      "mean_token_accuracy": 0.20489851385354996,
      "num_tokens": 49027780.0,
      "step": 21965
    },
    {
      "entropy": 5.309647512435913,
      "epoch": 2.4674566181838604,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000439884906447476,
      "loss": 5.0904,
      "mean_token_accuracy": 0.2058503210544586,
      "num_tokens": 49039441.0,
      "step": 21970
    },
    {
      "entropy": 5.345986175537109,
      "epoch": 2.4680181950918176,
      "grad_norm": 1.046875,
      "learning_rate": 0.000439857585806927,
      "loss": 5.1129,
      "mean_token_accuracy": 0.2073359966278076,
      "num_tokens": 49050740.0,
      "step": 21975
    },
    {
      "entropy": 5.351987552642822,
      "epoch": 2.4685797719997753,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043983025991724847,
      "loss": 5.0705,
      "mean_token_accuracy": 0.2085399627685547,
      "num_tokens": 49062216.0,
      "step": 21980
    },
    {
      "entropy": 5.323521947860717,
      "epoch": 2.469141348907733,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004398029287793106,
      "loss": 5.016,
      "mean_token_accuracy": 0.21476586610078813,
      "num_tokens": 49072848.0,
      "step": 21985
    },
    {
      "entropy": 5.320220041275024,
      "epoch": 2.4697029258156906,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043977559239398353,
      "loss": 4.9861,
      "mean_token_accuracy": 0.21397712379693984,
      "num_tokens": 49083382.0,
      "step": 21990
    },
    {
      "entropy": 5.319510984420776,
      "epoch": 2.470264502723648,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004397482507621376,
      "loss": 5.0316,
      "mean_token_accuracy": 0.20989374369382857,
      "num_tokens": 49095102.0,
      "step": 21995
    },
    {
      "entropy": 5.317521619796753,
      "epoch": 2.4708260796316055,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004397209038846436,
      "loss": 5.0746,
      "mean_token_accuracy": 0.20269681215286256,
      "num_tokens": 49105964.0,
      "step": 22000
    },
    {
      "entropy": 5.347790288925171,
      "epoch": 2.471387656539563,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043969355176237207,
      "loss": 5.1616,
      "mean_token_accuracy": 0.1953060358762741,
      "num_tokens": 49117748.0,
      "step": 22005
    },
    {
      "entropy": 5.276188659667969,
      "epoch": 2.471949233447521,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004396661943961941,
      "loss": 4.9118,
      "mean_token_accuracy": 0.21901594698429108,
      "num_tokens": 49128217.0,
      "step": 22010
    },
    {
      "entropy": 5.437936496734619,
      "epoch": 2.472510810355478,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004396388317869805,
      "loss": 5.1877,
      "mean_token_accuracy": 0.2027868628501892,
      "num_tokens": 49139272.0,
      "step": 22015
    },
    {
      "entropy": 5.396377944946289,
      "epoch": 2.4730723872634357,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004396114639356027,
      "loss": 5.0797,
      "mean_token_accuracy": 0.1969522461295128,
      "num_tokens": 49149570.0,
      "step": 22020
    },
    {
      "entropy": 5.275812005996704,
      "epoch": 2.4736339641713934,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000439584090842932,
      "loss": 5.0369,
      "mean_token_accuracy": 0.2095639631152153,
      "num_tokens": 49160892.0,
      "step": 22025
    },
    {
      "entropy": 5.307788324356079,
      "epoch": 2.474195541079351,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043955671250984005,
      "loss": 5.0347,
      "mean_token_accuracy": 0.20751384198665618,
      "num_tokens": 49172476.0,
      "step": 22030
    },
    {
      "entropy": 5.4316812515258786,
      "epoch": 2.4747571179873082,
      "grad_norm": 1.125,
      "learning_rate": 0.00043952932893719847,
      "loss": 5.1185,
      "mean_token_accuracy": 0.21251063346862792,
      "num_tokens": 49184052.0,
      "step": 22035
    },
    {
      "entropy": 5.369002485275269,
      "epoch": 2.475318694895266,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004395019401258794,
      "loss": 5.118,
      "mean_token_accuracy": 0.20304127037525177,
      "num_tokens": 49195441.0,
      "step": 22040
    },
    {
      "entropy": 5.290875816345215,
      "epoch": 2.4758802718032236,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043947454607675464,
      "loss": 5.0178,
      "mean_token_accuracy": 0.20719111561775208,
      "num_tokens": 49206395.0,
      "step": 22045
    },
    {
      "entropy": 5.403472375869751,
      "epoch": 2.476441848711181,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004394471467906966,
      "loss": 5.1055,
      "mean_token_accuracy": 0.20268541425466538,
      "num_tokens": 49217705.0,
      "step": 22050
    },
    {
      "entropy": 5.354883337020874,
      "epoch": 2.4770034256191384,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043941974226857763,
      "loss": 4.9656,
      "mean_token_accuracy": 0.21259523779153824,
      "num_tokens": 49228348.0,
      "step": 22055
    },
    {
      "entropy": 5.300539350509643,
      "epoch": 2.477565002527096,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004393923325112702,
      "loss": 5.0502,
      "mean_token_accuracy": 0.2041718065738678,
      "num_tokens": 49239031.0,
      "step": 22060
    },
    {
      "entropy": 5.405081796646118,
      "epoch": 2.4781265794350538,
      "grad_norm": 1.125,
      "learning_rate": 0.0004393649175196472,
      "loss": 5.1625,
      "mean_token_accuracy": 0.2000634789466858,
      "num_tokens": 49250918.0,
      "step": 22065
    },
    {
      "entropy": 5.33428111076355,
      "epoch": 2.478688156343011,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004393374972945815,
      "loss": 5.1242,
      "mean_token_accuracy": 0.20519136935472487,
      "num_tokens": 49262305.0,
      "step": 22070
    },
    {
      "entropy": 5.306736469268799,
      "epoch": 2.4792497332509686,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043931007183694616,
      "loss": 5.0418,
      "mean_token_accuracy": 0.20353138595819473,
      "num_tokens": 49273106.0,
      "step": 22075
    },
    {
      "entropy": 5.299357843399048,
      "epoch": 2.4798113101589263,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043928264114761437,
      "loss": 5.0086,
      "mean_token_accuracy": 0.2149631232023239,
      "num_tokens": 49284519.0,
      "step": 22080
    },
    {
      "entropy": 5.294533681869507,
      "epoch": 2.480372887066884,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043925520522745953,
      "loss": 5.025,
      "mean_token_accuracy": 0.20687436163425446,
      "num_tokens": 49295753.0,
      "step": 22085
    },
    {
      "entropy": 5.348375082015991,
      "epoch": 2.480934463974841,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004392277640773553,
      "loss": 5.1213,
      "mean_token_accuracy": 0.2129235327243805,
      "num_tokens": 49307682.0,
      "step": 22090
    },
    {
      "entropy": 5.338239431381226,
      "epoch": 2.481496040882799,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043920031769817543,
      "loss": 5.0573,
      "mean_token_accuracy": 0.20371610224246978,
      "num_tokens": 49318445.0,
      "step": 22095
    },
    {
      "entropy": 5.375595951080323,
      "epoch": 2.4820576177907565,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043917286609079364,
      "loss": 5.197,
      "mean_token_accuracy": 0.20433878898620605,
      "num_tokens": 49328559.0,
      "step": 22100
    },
    {
      "entropy": 5.328129625320434,
      "epoch": 2.482619194698714,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043914540925608426,
      "loss": 5.0754,
      "mean_token_accuracy": 0.20363164693117142,
      "num_tokens": 49340177.0,
      "step": 22105
    },
    {
      "entropy": 5.382462501525879,
      "epoch": 2.4831807716066714,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004391179471949213,
      "loss": 5.0264,
      "mean_token_accuracy": 0.21539869755506516,
      "num_tokens": 49351214.0,
      "step": 22110
    },
    {
      "entropy": 5.375477886199951,
      "epoch": 2.483742348514629,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043909047990817917,
      "loss": 5.0861,
      "mean_token_accuracy": 0.21147505044937134,
      "num_tokens": 49362697.0,
      "step": 22115
    },
    {
      "entropy": 5.381542158126831,
      "epoch": 2.4843039254225867,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043906300739673265,
      "loss": 5.0742,
      "mean_token_accuracy": 0.20800778418779373,
      "num_tokens": 49374439.0,
      "step": 22120
    },
    {
      "entropy": 5.296262979507446,
      "epoch": 2.4848655023305444,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043903552966145624,
      "loss": 4.9758,
      "mean_token_accuracy": 0.2067525014281273,
      "num_tokens": 49385834.0,
      "step": 22125
    },
    {
      "entropy": 5.308356142044067,
      "epoch": 2.4854270792385016,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000439008046703225,
      "loss": 5.0146,
      "mean_token_accuracy": 0.213261578977108,
      "num_tokens": 49396562.0,
      "step": 22130
    },
    {
      "entropy": 5.3230949401855465,
      "epoch": 2.4859886561464593,
      "grad_norm": 1.03125,
      "learning_rate": 0.000438980558522914,
      "loss": 5.0506,
      "mean_token_accuracy": 0.20426190048456191,
      "num_tokens": 49407339.0,
      "step": 22135
    },
    {
      "entropy": 5.351638126373291,
      "epoch": 2.486550233054417,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004389530651213983,
      "loss": 5.0479,
      "mean_token_accuracy": 0.21136243492364884,
      "num_tokens": 49419190.0,
      "step": 22140
    },
    {
      "entropy": 5.478767538070679,
      "epoch": 2.487111809962374,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043892556649955354,
      "loss": 5.1956,
      "mean_token_accuracy": 0.19224753677845002,
      "num_tokens": 49430559.0,
      "step": 22145
    },
    {
      "entropy": 5.385961675643921,
      "epoch": 2.487673386870332,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004388980626582551,
      "loss": 5.1217,
      "mean_token_accuracy": 0.20676996260881425,
      "num_tokens": 49441862.0,
      "step": 22150
    },
    {
      "entropy": 5.350239610671997,
      "epoch": 2.4882349637782895,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004388705535983788,
      "loss": 5.0449,
      "mean_token_accuracy": 0.19933124631643295,
      "num_tokens": 49452839.0,
      "step": 22155
    },
    {
      "entropy": 5.32871527671814,
      "epoch": 2.488796540686247,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004388430393208005,
      "loss": 5.0556,
      "mean_token_accuracy": 0.2121903881430626,
      "num_tokens": 49463683.0,
      "step": 22160
    },
    {
      "entropy": 5.335798549652099,
      "epoch": 2.4893581175942043,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004388155198263963,
      "loss": 5.0257,
      "mean_token_accuracy": 0.21363726258277893,
      "num_tokens": 49474338.0,
      "step": 22165
    },
    {
      "entropy": 5.324603271484375,
      "epoch": 2.489919694502162,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004387879951160424,
      "loss": 5.1239,
      "mean_token_accuracy": 0.20395575612783431,
      "num_tokens": 49485888.0,
      "step": 22170
    },
    {
      "entropy": 5.478207206726074,
      "epoch": 2.4904812714101197,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004387604651906152,
      "loss": 5.1435,
      "mean_token_accuracy": 0.20672149062156678,
      "num_tokens": 49497270.0,
      "step": 22175
    },
    {
      "entropy": 5.45848913192749,
      "epoch": 2.4910428483180773,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004387329300509913,
      "loss": 5.1142,
      "mean_token_accuracy": 0.2026048943400383,
      "num_tokens": 49509099.0,
      "step": 22180
    },
    {
      "entropy": 5.306802558898926,
      "epoch": 2.4916044252260345,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043870538969804733,
      "loss": 5.0382,
      "mean_token_accuracy": 0.20852340012788773,
      "num_tokens": 49520110.0,
      "step": 22185
    },
    {
      "entropy": 5.321734142303467,
      "epoch": 2.492166002133992,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004386778441326603,
      "loss": 5.0143,
      "mean_token_accuracy": 0.2123237133026123,
      "num_tokens": 49531866.0,
      "step": 22190
    },
    {
      "entropy": 5.414988899230957,
      "epoch": 2.49272757904195,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043865029335570725,
      "loss": 5.1102,
      "mean_token_accuracy": 0.20211392790079116,
      "num_tokens": 49541022.0,
      "step": 22195
    },
    {
      "entropy": 5.351059150695801,
      "epoch": 2.4932891559499075,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004386227373680655,
      "loss": 5.0846,
      "mean_token_accuracy": 0.20638477355241774,
      "num_tokens": 49551908.0,
      "step": 22200
    },
    {
      "entropy": 5.383871078491211,
      "epoch": 2.4938507328578647,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004385951761706122,
      "loss": 5.0246,
      "mean_token_accuracy": 0.21004384011030197,
      "num_tokens": 49562475.0,
      "step": 22205
    },
    {
      "entropy": 5.347735834121704,
      "epoch": 2.4944123097658224,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043856760976422506,
      "loss": 4.9932,
      "mean_token_accuracy": 0.2107164144515991,
      "num_tokens": 49572916.0,
      "step": 22210
    },
    {
      "entropy": 5.381935501098633,
      "epoch": 2.49497388667378,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043854003814978174,
      "loss": 5.08,
      "mean_token_accuracy": 0.20797484815120698,
      "num_tokens": 49584584.0,
      "step": 22215
    },
    {
      "entropy": 5.314293670654297,
      "epoch": 2.4955354635817377,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043851246132816025,
      "loss": 5.0701,
      "mean_token_accuracy": 0.20608883649110793,
      "num_tokens": 49595792.0,
      "step": 22220
    },
    {
      "entropy": 5.356944847106933,
      "epoch": 2.496097040489695,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004384848793002385,
      "loss": 5.1122,
      "mean_token_accuracy": 0.20641910880804062,
      "num_tokens": 49607596.0,
      "step": 22225
    },
    {
      "entropy": 5.402005529403686,
      "epoch": 2.4966586173976526,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004384572920668949,
      "loss": 5.0443,
      "mean_token_accuracy": 0.20303504168987274,
      "num_tokens": 49617990.0,
      "step": 22230
    },
    {
      "entropy": 5.429148960113525,
      "epoch": 2.4972201943056103,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004384296996290076,
      "loss": 5.1446,
      "mean_token_accuracy": 0.20163006335496902,
      "num_tokens": 49629458.0,
      "step": 22235
    },
    {
      "entropy": 5.412051820755005,
      "epoch": 2.4977817712135675,
      "grad_norm": 1.0,
      "learning_rate": 0.00043840210198745533,
      "loss": 5.1022,
      "mean_token_accuracy": 0.2047921046614647,
      "num_tokens": 49641227.0,
      "step": 22240
    },
    {
      "entropy": 5.360210657119751,
      "epoch": 2.498343348121525,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043837449914311676,
      "loss": 5.1622,
      "mean_token_accuracy": 0.19733384847640992,
      "num_tokens": 49652655.0,
      "step": 22245
    },
    {
      "entropy": 5.38282995223999,
      "epoch": 2.498904925029483,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043834689109687077,
      "loss": 5.1006,
      "mean_token_accuracy": 0.20836657732725145,
      "num_tokens": 49664308.0,
      "step": 22250
    },
    {
      "entropy": 5.364755630493164,
      "epoch": 2.4994665019374405,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004383192778495964,
      "loss": 5.0609,
      "mean_token_accuracy": 0.20921119898557664,
      "num_tokens": 49676656.0,
      "step": 22255
    },
    {
      "entropy": 5.3267130851745605,
      "epoch": 2.500028078845398,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000438291659402173,
      "loss": 5.0115,
      "mean_token_accuracy": 0.21085827350616454,
      "num_tokens": 49687706.0,
      "step": 22260
    },
    {
      "entropy": 5.346769762039185,
      "epoch": 2.5005896557533553,
      "grad_norm": 1.125,
      "learning_rate": 0.00043826403575547974,
      "loss": 5.0295,
      "mean_token_accuracy": 0.21668460667133332,
      "num_tokens": 49699139.0,
      "step": 22265
    },
    {
      "entropy": 5.23520975112915,
      "epoch": 2.501151232661313,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004382364069103962,
      "loss": 5.0378,
      "mean_token_accuracy": 0.21124025136232377,
      "num_tokens": 49710690.0,
      "step": 22270
    },
    {
      "entropy": 5.332208681106567,
      "epoch": 2.5017128095692707,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043820877286780224,
      "loss": 5.0811,
      "mean_token_accuracy": 0.2082766368985176,
      "num_tokens": 49721920.0,
      "step": 22275
    },
    {
      "entropy": 5.436357116699218,
      "epoch": 2.502274386477228,
      "grad_norm": 1.0,
      "learning_rate": 0.0004381811336285776,
      "loss": 5.0839,
      "mean_token_accuracy": 0.21116328686475755,
      "num_tokens": 49732923.0,
      "step": 22280
    },
    {
      "entropy": 5.2671746730804445,
      "epoch": 2.5028359633851855,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004381534891936024,
      "loss": 4.9707,
      "mean_token_accuracy": 0.2093236118555069,
      "num_tokens": 49744755.0,
      "step": 22285
    },
    {
      "entropy": 5.329423618316651,
      "epoch": 2.503397540293143,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004381258395637568,
      "loss": 5.0462,
      "mean_token_accuracy": 0.21032188683748246,
      "num_tokens": 49756126.0,
      "step": 22290
    },
    {
      "entropy": 5.359637928009033,
      "epoch": 2.5039591172011004,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004380981847399212,
      "loss": 5.0382,
      "mean_token_accuracy": 0.20962809920310974,
      "num_tokens": 49767137.0,
      "step": 22295
    },
    {
      "entropy": 5.360326719284058,
      "epoch": 2.504520694109058,
      "grad_norm": 1.125,
      "learning_rate": 0.00043807052472297626,
      "loss": 5.1093,
      "mean_token_accuracy": 0.20454536080360414,
      "num_tokens": 49778533.0,
      "step": 22300
    },
    {
      "entropy": 5.346648025512695,
      "epoch": 2.5050822710170157,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004380428595138024,
      "loss": 5.0683,
      "mean_token_accuracy": 0.21252695322036744,
      "num_tokens": 49789494.0,
      "step": 22305
    },
    {
      "entropy": 5.393642282485962,
      "epoch": 2.5056438479249734,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043801518911328074,
      "loss": 5.0712,
      "mean_token_accuracy": 0.20391117930412292,
      "num_tokens": 49799546.0,
      "step": 22310
    },
    {
      "entropy": 5.299918174743652,
      "epoch": 2.506205424832931,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043798751352229227,
      "loss": 4.978,
      "mean_token_accuracy": 0.21577945947647095,
      "num_tokens": 49810643.0,
      "step": 22315
    },
    {
      "entropy": 5.3350687503814695,
      "epoch": 2.5067670017408883,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043795983274171806,
      "loss": 5.1085,
      "mean_token_accuracy": 0.2046674072742462,
      "num_tokens": 49821410.0,
      "step": 22320
    },
    {
      "entropy": 5.3392938613891605,
      "epoch": 2.507328578648846,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043793214677243967,
      "loss": 5.121,
      "mean_token_accuracy": 0.20575061440467834,
      "num_tokens": 49832911.0,
      "step": 22325
    },
    {
      "entropy": 5.375244951248169,
      "epoch": 2.5078901555568036,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004379044556153384,
      "loss": 5.1074,
      "mean_token_accuracy": 0.20146720707416535,
      "num_tokens": 49845458.0,
      "step": 22330
    },
    {
      "entropy": 5.426929187774658,
      "epoch": 2.508451732464761,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043787675927129625,
      "loss": 5.2083,
      "mean_token_accuracy": 0.1952795296907425,
      "num_tokens": 49857054.0,
      "step": 22335
    },
    {
      "entropy": 5.3521788120269775,
      "epoch": 2.5090133093727185,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043784905774119483,
      "loss": 4.9982,
      "mean_token_accuracy": 0.20974872261285782,
      "num_tokens": 49867223.0,
      "step": 22340
    },
    {
      "entropy": 5.3058422088623045,
      "epoch": 2.509574886280676,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043782135102591626,
      "loss": 5.058,
      "mean_token_accuracy": 0.20529340356588363,
      "num_tokens": 49877801.0,
      "step": 22345
    },
    {
      "entropy": 5.302078676223755,
      "epoch": 2.510136463188634,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004377936391263427,
      "loss": 5.0691,
      "mean_token_accuracy": 0.2028646945953369,
      "num_tokens": 49888853.0,
      "step": 22350
    },
    {
      "entropy": 5.379345464706421,
      "epoch": 2.5106980400965915,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043776592204335655,
      "loss": 5.0752,
      "mean_token_accuracy": 0.20383783131837846,
      "num_tokens": 49899989.0,
      "step": 22355
    },
    {
      "entropy": 5.357038068771362,
      "epoch": 2.5112596170045487,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004377381997778403,
      "loss": 5.0359,
      "mean_token_accuracy": 0.21198545545339584,
      "num_tokens": 49910993.0,
      "step": 22360
    },
    {
      "entropy": 5.355185270309448,
      "epoch": 2.5118211939125064,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004377104723306767,
      "loss": 5.1068,
      "mean_token_accuracy": 0.20211372077465056,
      "num_tokens": 49922169.0,
      "step": 22365
    },
    {
      "entropy": 5.404388809204102,
      "epoch": 2.512382770820464,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004376827397027485,
      "loss": 5.1589,
      "mean_token_accuracy": 0.1981235459446907,
      "num_tokens": 49933700.0,
      "step": 22370
    },
    {
      "entropy": 5.396981620788575,
      "epoch": 2.5129443477284212,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004376550018949388,
      "loss": 5.0883,
      "mean_token_accuracy": 0.20319758951663972,
      "num_tokens": 49944689.0,
      "step": 22375
    },
    {
      "entropy": 5.392091178894043,
      "epoch": 2.513505924636379,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043762725890813083,
      "loss": 5.1049,
      "mean_token_accuracy": 0.20440608859062195,
      "num_tokens": 49955818.0,
      "step": 22380
    },
    {
      "entropy": 5.294638824462891,
      "epoch": 2.5140675015443366,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004375995107432077,
      "loss": 4.9949,
      "mean_token_accuracy": 0.2153933599591255,
      "num_tokens": 49965799.0,
      "step": 22385
    },
    {
      "entropy": 5.336868524551392,
      "epoch": 2.5146290784522938,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043757175740105316,
      "loss": 5.0953,
      "mean_token_accuracy": 0.2030978709459305,
      "num_tokens": 49977742.0,
      "step": 22390
    },
    {
      "entropy": 5.3681416511535645,
      "epoch": 2.5151906553602514,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004375439988825508,
      "loss": 5.0563,
      "mean_token_accuracy": 0.20869150310754775,
      "num_tokens": 49988382.0,
      "step": 22395
    },
    {
      "entropy": 5.420800352096558,
      "epoch": 2.515752232268209,
      "grad_norm": 1.0,
      "learning_rate": 0.0004375162351885845,
      "loss": 5.1873,
      "mean_token_accuracy": 0.20308376103639603,
      "num_tokens": 50000190.0,
      "step": 22400
    },
    {
      "entropy": 5.421718168258667,
      "epoch": 2.5163138091761668,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043748846632003823,
      "loss": 5.0874,
      "mean_token_accuracy": 0.21099355816841125,
      "num_tokens": 50011297.0,
      "step": 22405
    },
    {
      "entropy": 5.343901252746582,
      "epoch": 2.5168753860841244,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043746069227779615,
      "loss": 5.1139,
      "mean_token_accuracy": 0.20297098755836487,
      "num_tokens": 50023165.0,
      "step": 22410
    },
    {
      "entropy": 5.387256813049317,
      "epoch": 2.5174369629920816,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004374329130627427,
      "loss": 5.0882,
      "mean_token_accuracy": 0.2127643942832947,
      "num_tokens": 50035803.0,
      "step": 22415
    },
    {
      "entropy": 5.358915090560913,
      "epoch": 2.5179985399000393,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043740512867576224,
      "loss": 4.9988,
      "mean_token_accuracy": 0.2098386913537979,
      "num_tokens": 50046599.0,
      "step": 22420
    },
    {
      "entropy": 5.348342943191528,
      "epoch": 2.518560116807997,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004373773391177395,
      "loss": 5.2031,
      "mean_token_accuracy": 0.198530712723732,
      "num_tokens": 50058170.0,
      "step": 22425
    },
    {
      "entropy": 5.309135818481446,
      "epoch": 2.519121693715954,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004373495443895593,
      "loss": 5.0351,
      "mean_token_accuracy": 0.2151987925171852,
      "num_tokens": 50068818.0,
      "step": 22430
    },
    {
      "entropy": 5.412164402008057,
      "epoch": 2.519683270623912,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004373217444921066,
      "loss": 5.186,
      "mean_token_accuracy": 0.19728488773107528,
      "num_tokens": 50080533.0,
      "step": 22435
    },
    {
      "entropy": 5.350225925445557,
      "epoch": 2.5202448475318695,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043729393942626665,
      "loss": 5.0064,
      "mean_token_accuracy": 0.2111138731241226,
      "num_tokens": 50090897.0,
      "step": 22440
    },
    {
      "entropy": 5.246854066848755,
      "epoch": 2.520806424439827,
      "grad_norm": 1.0,
      "learning_rate": 0.00043726612919292476,
      "loss": 4.984,
      "mean_token_accuracy": 0.2154233291745186,
      "num_tokens": 50101594.0,
      "step": 22445
    },
    {
      "entropy": 5.278415536880493,
      "epoch": 2.521368001347785,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004372383137929663,
      "loss": 5.0341,
      "mean_token_accuracy": 0.21164635568857193,
      "num_tokens": 50111923.0,
      "step": 22450
    },
    {
      "entropy": 5.382044935226441,
      "epoch": 2.521929578255742,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043721049322727705,
      "loss": 5.1176,
      "mean_token_accuracy": 0.2101084679365158,
      "num_tokens": 50123295.0,
      "step": 22455
    },
    {
      "entropy": 5.424056768417358,
      "epoch": 2.5224911551636997,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004371826674967428,
      "loss": 5.1741,
      "mean_token_accuracy": 0.20775683224201202,
      "num_tokens": 50133245.0,
      "step": 22460
    },
    {
      "entropy": 5.307792663574219,
      "epoch": 2.5230527320716574,
      "grad_norm": 1.0,
      "learning_rate": 0.00043715483660224943,
      "loss": 5.0316,
      "mean_token_accuracy": 0.2081843689084053,
      "num_tokens": 50144827.0,
      "step": 22465
    },
    {
      "entropy": 5.440236139297485,
      "epoch": 2.5236143089796146,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004371270005446833,
      "loss": 5.1612,
      "mean_token_accuracy": 0.19821587055921555,
      "num_tokens": 50156244.0,
      "step": 22470
    },
    {
      "entropy": 5.438194847106933,
      "epoch": 2.5241758858875722,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004370991593249305,
      "loss": 5.1233,
      "mean_token_accuracy": 0.20539896190166473,
      "num_tokens": 50168363.0,
      "step": 22475
    },
    {
      "entropy": 5.350019788742065,
      "epoch": 2.52473746279553,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043707131294387766,
      "loss": 5.0655,
      "mean_token_accuracy": 0.2036655679345131,
      "num_tokens": 50179370.0,
      "step": 22480
    },
    {
      "entropy": 5.415399694442749,
      "epoch": 2.525299039703487,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004370434614024113,
      "loss": 5.0939,
      "mean_token_accuracy": 0.21581622809171677,
      "num_tokens": 50191583.0,
      "step": 22485
    },
    {
      "entropy": 5.375579452514648,
      "epoch": 2.525860616611445,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043701560470141824,
      "loss": 5.0437,
      "mean_token_accuracy": 0.20309269279241562,
      "num_tokens": 50202008.0,
      "step": 22490
    },
    {
      "entropy": 5.327255010604858,
      "epoch": 2.5264221935194024,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004369877428417856,
      "loss": 5.118,
      "mean_token_accuracy": 0.19655717611312867,
      "num_tokens": 50212632.0,
      "step": 22495
    },
    {
      "entropy": 5.371885919570923,
      "epoch": 2.52698377042736,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004369598758244003,
      "loss": 5.0982,
      "mean_token_accuracy": 0.21122310161590577,
      "num_tokens": 50223485.0,
      "step": 22500
    },
    {
      "entropy": 5.4226966381073,
      "epoch": 2.5275453473353178,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043693200365014985,
      "loss": 5.1609,
      "mean_token_accuracy": 0.2045956864953041,
      "num_tokens": 50235516.0,
      "step": 22505
    },
    {
      "entropy": 5.3421173095703125,
      "epoch": 2.528106924243275,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043690412631992154,
      "loss": 5.0205,
      "mean_token_accuracy": 0.21143171042203904,
      "num_tokens": 50246846.0,
      "step": 22510
    },
    {
      "entropy": 5.3275457382202145,
      "epoch": 2.5286685011512327,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00043687624383460306,
      "loss": 5.0848,
      "mean_token_accuracy": 0.19968171417713165,
      "num_tokens": 50258433.0,
      "step": 22515
    },
    {
      "entropy": 5.444024658203125,
      "epoch": 2.5292300780591903,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043684835619508215,
      "loss": 5.1881,
      "mean_token_accuracy": 0.20230540931224822,
      "num_tokens": 50269111.0,
      "step": 22520
    },
    {
      "entropy": 5.388620662689209,
      "epoch": 2.5297916549671475,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004368204634022469,
      "loss": 5.2097,
      "mean_token_accuracy": 0.19952683299779891,
      "num_tokens": 50280137.0,
      "step": 22525
    },
    {
      "entropy": 5.3832440853118895,
      "epoch": 2.530353231875105,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004367925654569852,
      "loss": 5.0383,
      "mean_token_accuracy": 0.21466072499752045,
      "num_tokens": 50291717.0,
      "step": 22530
    },
    {
      "entropy": 5.388839864730835,
      "epoch": 2.530914808783063,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004367646623601856,
      "loss": 5.1234,
      "mean_token_accuracy": 0.20211385190486908,
      "num_tokens": 50303107.0,
      "step": 22535
    },
    {
      "entropy": 5.278520154953003,
      "epoch": 2.5314763856910205,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004367367541127364,
      "loss": 5.0467,
      "mean_token_accuracy": 0.2078275576233864,
      "num_tokens": 50313997.0,
      "step": 22540
    },
    {
      "entropy": 5.355160713195801,
      "epoch": 2.532037962598978,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004367088407155261,
      "loss": 5.0442,
      "mean_token_accuracy": 0.21278139650821687,
      "num_tokens": 50324327.0,
      "step": 22545
    },
    {
      "entropy": 5.2998779773712155,
      "epoch": 2.5325995395069354,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004366809221694437,
      "loss": 4.982,
      "mean_token_accuracy": 0.2186443626880646,
      "num_tokens": 50334787.0,
      "step": 22550
    },
    {
      "entropy": 5.282994556427002,
      "epoch": 2.533161116414893,
      "grad_norm": 1.0,
      "learning_rate": 0.00043665299847537805,
      "loss": 4.9653,
      "mean_token_accuracy": 0.21710716634988786,
      "num_tokens": 50345418.0,
      "step": 22555
    },
    {
      "entropy": 5.303112888336182,
      "epoch": 2.5337226933228507,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043662506963421815,
      "loss": 5.0624,
      "mean_token_accuracy": 0.21014363765716554,
      "num_tokens": 50357047.0,
      "step": 22560
    },
    {
      "entropy": 5.2941858768463135,
      "epoch": 2.534284270230808,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004365971356468534,
      "loss": 5.0382,
      "mean_token_accuracy": 0.21629771441221238,
      "num_tokens": 50368385.0,
      "step": 22565
    },
    {
      "entropy": 5.36540150642395,
      "epoch": 2.5348458471387656,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043656919651417314,
      "loss": 5.1078,
      "mean_token_accuracy": 0.20719412863254547,
      "num_tokens": 50379389.0,
      "step": 22570
    },
    {
      "entropy": 5.384038019180298,
      "epoch": 2.5354074240467233,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000436541252237067,
      "loss": 5.078,
      "mean_token_accuracy": 0.2131941497325897,
      "num_tokens": 50390124.0,
      "step": 22575
    },
    {
      "entropy": 5.37825198173523,
      "epoch": 2.5359690009546805,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004365133028164248,
      "loss": 5.0829,
      "mean_token_accuracy": 0.2105655699968338,
      "num_tokens": 50401385.0,
      "step": 22580
    },
    {
      "entropy": 5.33890700340271,
      "epoch": 2.536530577862638,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004364853482531363,
      "loss": 5.1249,
      "mean_token_accuracy": 0.19879220724105834,
      "num_tokens": 50412890.0,
      "step": 22585
    },
    {
      "entropy": 5.363964462280274,
      "epoch": 2.537092154770596,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043645738854809177,
      "loss": 5.13,
      "mean_token_accuracy": 0.20506908595561982,
      "num_tokens": 50424403.0,
      "step": 22590
    },
    {
      "entropy": 5.449519872665405,
      "epoch": 2.5376537316785535,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004364294237021813,
      "loss": 5.0996,
      "mean_token_accuracy": 0.20289226770401,
      "num_tokens": 50434598.0,
      "step": 22595
    },
    {
      "entropy": 5.415514850616455,
      "epoch": 2.538215308586511,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043640145371629544,
      "loss": 5.1581,
      "mean_token_accuracy": 0.20111916959285736,
      "num_tokens": 50446661.0,
      "step": 22600
    },
    {
      "entropy": 5.450956344604492,
      "epoch": 2.5387768854944683,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004363734785913246,
      "loss": 5.1639,
      "mean_token_accuracy": 0.19975059628486633,
      "num_tokens": 50458188.0,
      "step": 22605
    },
    {
      "entropy": 5.407372665405274,
      "epoch": 2.539338462402426,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004363454983281597,
      "loss": 5.1381,
      "mean_token_accuracy": 0.20601719617843628,
      "num_tokens": 50469305.0,
      "step": 22610
    },
    {
      "entropy": 5.377738428115845,
      "epoch": 2.5399000393103837,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004363175129276915,
      "loss": 5.0893,
      "mean_token_accuracy": 0.20348107367753981,
      "num_tokens": 50480450.0,
      "step": 22615
    },
    {
      "entropy": 5.296732425689697,
      "epoch": 2.540461616218341,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043628952239081113,
      "loss": 5.0589,
      "mean_token_accuracy": 0.20992061346769333,
      "num_tokens": 50490397.0,
      "step": 22620
    },
    {
      "entropy": 5.269248533248901,
      "epoch": 2.5410231931262985,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043626152671840984,
      "loss": 5.0218,
      "mean_token_accuracy": 0.21269934624433517,
      "num_tokens": 50501068.0,
      "step": 22625
    },
    {
      "entropy": 5.408339023590088,
      "epoch": 2.541584770034256,
      "grad_norm": 1.0,
      "learning_rate": 0.000436233525911379,
      "loss": 5.281,
      "mean_token_accuracy": 0.19299737811088563,
      "num_tokens": 50513268.0,
      "step": 22630
    },
    {
      "entropy": 5.345050954818726,
      "epoch": 2.542146346942214,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004362055199706101,
      "loss": 4.9697,
      "mean_token_accuracy": 0.21364488005638121,
      "num_tokens": 50523806.0,
      "step": 22635
    },
    {
      "entropy": 5.35197639465332,
      "epoch": 2.5427079238501715,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043617750889699496,
      "loss": 5.0554,
      "mean_token_accuracy": 0.20641474276781083,
      "num_tokens": 50535000.0,
      "step": 22640
    },
    {
      "entropy": 5.375270795822144,
      "epoch": 2.5432695007581287,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043614949269142546,
      "loss": 5.1504,
      "mean_token_accuracy": 0.19856346994638444,
      "num_tokens": 50546042.0,
      "step": 22645
    },
    {
      "entropy": 5.475940179824829,
      "epoch": 2.5438310776660864,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043612147135479356,
      "loss": 5.1256,
      "mean_token_accuracy": 0.2018774628639221,
      "num_tokens": 50556645.0,
      "step": 22650
    },
    {
      "entropy": 5.436217403411865,
      "epoch": 2.544392654574044,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004360934448879916,
      "loss": 5.1373,
      "mean_token_accuracy": 0.20553362518548965,
      "num_tokens": 50567904.0,
      "step": 22655
    },
    {
      "entropy": 5.26300950050354,
      "epoch": 2.5449542314820013,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004360654132919118,
      "loss": 4.9582,
      "mean_token_accuracy": 0.21463854759931564,
      "num_tokens": 50578280.0,
      "step": 22660
    },
    {
      "entropy": 5.288274431228638,
      "epoch": 2.545515808389959,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004360373765674467,
      "loss": 5.0653,
      "mean_token_accuracy": 0.20973957628011702,
      "num_tokens": 50590232.0,
      "step": 22665
    },
    {
      "entropy": 5.332500267028808,
      "epoch": 2.5460773852979166,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043600933471548927,
      "loss": 4.9636,
      "mean_token_accuracy": 0.2139849230647087,
      "num_tokens": 50601469.0,
      "step": 22670
    },
    {
      "entropy": 5.296098709106445,
      "epoch": 2.546638962205874,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004359812877369321,
      "loss": 4.9522,
      "mean_token_accuracy": 0.2098448544740677,
      "num_tokens": 50612090.0,
      "step": 22675
    },
    {
      "entropy": 5.249640274047851,
      "epoch": 2.5472005391138315,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043595323563266815,
      "loss": 5.072,
      "mean_token_accuracy": 0.2051727756857872,
      "num_tokens": 50623834.0,
      "step": 22680
    },
    {
      "entropy": 5.337622547149659,
      "epoch": 2.547762116021789,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043592517840359093,
      "loss": 5.1317,
      "mean_token_accuracy": 0.203157040476799,
      "num_tokens": 50633966.0,
      "step": 22685
    },
    {
      "entropy": 5.343468046188354,
      "epoch": 2.548323692929747,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004358971160505936,
      "loss": 4.973,
      "mean_token_accuracy": 0.2127581611275673,
      "num_tokens": 50644354.0,
      "step": 22690
    },
    {
      "entropy": 5.305032634735108,
      "epoch": 2.5488852698377045,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043586904857456963,
      "loss": 4.9625,
      "mean_token_accuracy": 0.2070537120103836,
      "num_tokens": 50655525.0,
      "step": 22695
    },
    {
      "entropy": 5.384944772720337,
      "epoch": 2.5494468467456617,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00043584097597641275,
      "loss": 5.198,
      "mean_token_accuracy": 0.19856284260749818,
      "num_tokens": 50667879.0,
      "step": 22700
    },
    {
      "entropy": 5.27630581855774,
      "epoch": 2.5500084236536193,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043581289825701683,
      "loss": 4.9374,
      "mean_token_accuracy": 0.21263434290885924,
      "num_tokens": 50679568.0,
      "step": 22705
    },
    {
      "entropy": 5.332930326461792,
      "epoch": 2.550570000561577,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004357848154172759,
      "loss": 5.1173,
      "mean_token_accuracy": 0.20907909721136092,
      "num_tokens": 50690172.0,
      "step": 22710
    },
    {
      "entropy": 5.323129177093506,
      "epoch": 2.5511315774695342,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043575672745808404,
      "loss": 5.0194,
      "mean_token_accuracy": 0.20844885259866713,
      "num_tokens": 50701150.0,
      "step": 22715
    },
    {
      "entropy": 5.358335256576538,
      "epoch": 2.551693154377492,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043572863438033556,
      "loss": 5.1106,
      "mean_token_accuracy": 0.21047590672969818,
      "num_tokens": 50712407.0,
      "step": 22720
    },
    {
      "entropy": 5.325708961486816,
      "epoch": 2.5522547312854496,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004357005361849251,
      "loss": 5.0441,
      "mean_token_accuracy": 0.20480820089578627,
      "num_tokens": 50723122.0,
      "step": 22725
    },
    {
      "entropy": 5.409446811676025,
      "epoch": 2.552816308193407,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043567243287274726,
      "loss": 5.1019,
      "mean_token_accuracy": 0.20430083721876144,
      "num_tokens": 50734479.0,
      "step": 22730
    },
    {
      "entropy": 5.327328443527222,
      "epoch": 2.553377885101365,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043564432444469686,
      "loss": 5.047,
      "mean_token_accuracy": 0.2154715031385422,
      "num_tokens": 50745098.0,
      "step": 22735
    },
    {
      "entropy": 5.339509439468384,
      "epoch": 2.553939462009322,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043561621090166885,
      "loss": 5.1052,
      "mean_token_accuracy": 0.20289770811796187,
      "num_tokens": 50755793.0,
      "step": 22740
    },
    {
      "entropy": 5.330500078201294,
      "epoch": 2.5545010389172798,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004355880922445584,
      "loss": 5.11,
      "mean_token_accuracy": 0.19896526634693146,
      "num_tokens": 50767685.0,
      "step": 22745
    },
    {
      "entropy": 5.449169874191284,
      "epoch": 2.5550626158252374,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043555996847426077,
      "loss": 5.1939,
      "mean_token_accuracy": 0.19696093052625657,
      "num_tokens": 50778583.0,
      "step": 22750
    },
    {
      "entropy": 5.415363502502442,
      "epoch": 2.5556241927331946,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043553183959167154,
      "loss": 5.0605,
      "mean_token_accuracy": 0.20648893266916274,
      "num_tokens": 50789651.0,
      "step": 22755
    },
    {
      "entropy": 5.436167669296265,
      "epoch": 2.5561857696411523,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043550370559768614,
      "loss": 5.2022,
      "mean_token_accuracy": 0.20034205615520478,
      "num_tokens": 50800846.0,
      "step": 22760
    },
    {
      "entropy": 5.3435486316680905,
      "epoch": 2.55674734654911,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004354755664932006,
      "loss": 4.9488,
      "mean_token_accuracy": 0.21430336833000183,
      "num_tokens": 50812532.0,
      "step": 22765
    },
    {
      "entropy": 5.373645496368408,
      "epoch": 2.557308923457067,
      "grad_norm": 0.9375,
      "learning_rate": 0.00043544742227911076,
      "loss": 5.1275,
      "mean_token_accuracy": 0.20356102883815766,
      "num_tokens": 50823592.0,
      "step": 22770
    },
    {
      "entropy": 5.325995302200317,
      "epoch": 2.557870500365025,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043541927295631276,
      "loss": 5.0132,
      "mean_token_accuracy": 0.2024658814072609,
      "num_tokens": 50834392.0,
      "step": 22775
    },
    {
      "entropy": 5.415244197845459,
      "epoch": 2.5584320772729825,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043539111852570296,
      "loss": 5.1439,
      "mean_token_accuracy": 0.2063068687915802,
      "num_tokens": 50845371.0,
      "step": 22780
    },
    {
      "entropy": 5.403409051895141,
      "epoch": 2.55899365418094,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043536295898817767,
      "loss": 5.1445,
      "mean_token_accuracy": 0.20398071259260178,
      "num_tokens": 50857748.0,
      "step": 22785
    },
    {
      "entropy": 5.3058274269104,
      "epoch": 2.559555231088898,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004353347943446336,
      "loss": 4.9509,
      "mean_token_accuracy": 0.20971325635910035,
      "num_tokens": 50868646.0,
      "step": 22790
    },
    {
      "entropy": 5.290994930267334,
      "epoch": 2.560116807996855,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043530662459596744,
      "loss": 4.9781,
      "mean_token_accuracy": 0.21417610645294188,
      "num_tokens": 50880566.0,
      "step": 22795
    },
    {
      "entropy": 5.350013637542725,
      "epoch": 2.5606783849048127,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004352784497430762,
      "loss": 5.0869,
      "mean_token_accuracy": 0.20841921716928483,
      "num_tokens": 50891487.0,
      "step": 22800
    },
    {
      "entropy": 5.381331157684326,
      "epoch": 2.5612399618127704,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043525026978685706,
      "loss": 5.1347,
      "mean_token_accuracy": 0.19636826813220978,
      "num_tokens": 50902522.0,
      "step": 22805
    },
    {
      "entropy": 5.455000972747802,
      "epoch": 2.5618015387207276,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043522208472820706,
      "loss": 5.1022,
      "mean_token_accuracy": 0.20308726727962495,
      "num_tokens": 50913363.0,
      "step": 22810
    },
    {
      "entropy": 5.406198930740357,
      "epoch": 2.5623631156286852,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004351938945680238,
      "loss": 5.0754,
      "mean_token_accuracy": 0.2074052259325981,
      "num_tokens": 50923724.0,
      "step": 22815
    },
    {
      "entropy": 5.391058015823364,
      "epoch": 2.562924692536643,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004351656993072048,
      "loss": 5.0813,
      "mean_token_accuracy": 0.20719061195850372,
      "num_tokens": 50933575.0,
      "step": 22820
    },
    {
      "entropy": 5.322501611709595,
      "epoch": 2.5634862694446006,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004351374989466477,
      "loss": 5.0521,
      "mean_token_accuracy": 0.20933962166309356,
      "num_tokens": 50944619.0,
      "step": 22825
    },
    {
      "entropy": 5.3295060157775875,
      "epoch": 2.5640478463525582,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043510929348725065,
      "loss": 5.0743,
      "mean_token_accuracy": 0.2131260633468628,
      "num_tokens": 50955701.0,
      "step": 22830
    },
    {
      "entropy": 5.308396577835083,
      "epoch": 2.5646094232605154,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043508108292991156,
      "loss": 4.9193,
      "mean_token_accuracy": 0.21261988431215287,
      "num_tokens": 50966544.0,
      "step": 22835
    },
    {
      "entropy": 5.341888284683227,
      "epoch": 2.565171000168473,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004350528672755287,
      "loss": 5.138,
      "mean_token_accuracy": 0.20841312408447266,
      "num_tokens": 50976648.0,
      "step": 22840
    },
    {
      "entropy": 5.413437461853027,
      "epoch": 2.5657325770764308,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004350246465250005,
      "loss": 5.0832,
      "mean_token_accuracy": 0.20249558240175247,
      "num_tokens": 50988173.0,
      "step": 22845
    },
    {
      "entropy": 5.356917238235473,
      "epoch": 2.566294153984388,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043499642067922544,
      "loss": 4.9694,
      "mean_token_accuracy": 0.21420007348060607,
      "num_tokens": 50999274.0,
      "step": 22850
    },
    {
      "entropy": 5.2861647605896,
      "epoch": 2.5668557308923456,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043496818973910235,
      "loss": 5.0316,
      "mean_token_accuracy": 0.21074003726243973,
      "num_tokens": 51010213.0,
      "step": 22855
    },
    {
      "entropy": 5.317293453216553,
      "epoch": 2.5674173078003033,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043493995370553,
      "loss": 5.0843,
      "mean_token_accuracy": 0.20244687646627427,
      "num_tokens": 51021460.0,
      "step": 22860
    },
    {
      "entropy": 5.450330638885498,
      "epoch": 2.5679788847082605,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043491171257940755,
      "loss": 5.0857,
      "mean_token_accuracy": 0.20442305207252504,
      "num_tokens": 51031713.0,
      "step": 22865
    },
    {
      "entropy": 5.42491569519043,
      "epoch": 2.568540461616218,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043488346636163405,
      "loss": 5.1762,
      "mean_token_accuracy": 0.20066711604595183,
      "num_tokens": 51043026.0,
      "step": 22870
    },
    {
      "entropy": 5.456652879714966,
      "epoch": 2.569102038524176,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004348552150531091,
      "loss": 5.1143,
      "mean_token_accuracy": 0.20488483160734178,
      "num_tokens": 51054838.0,
      "step": 22875
    },
    {
      "entropy": 5.357372379302978,
      "epoch": 2.5696636154321335,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043482695865473207,
      "loss": 4.9539,
      "mean_token_accuracy": 0.21211812347173692,
      "num_tokens": 51065574.0,
      "step": 22880
    },
    {
      "entropy": 5.290013027191162,
      "epoch": 2.570225192340091,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004347986971674026,
      "loss": 5.0181,
      "mean_token_accuracy": 0.21177231669425964,
      "num_tokens": 51076814.0,
      "step": 22885
    },
    {
      "entropy": 5.317489337921143,
      "epoch": 2.5707867692480484,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043477043059202077,
      "loss": 5.119,
      "mean_token_accuracy": 0.19911332577466964,
      "num_tokens": 51089075.0,
      "step": 22890
    },
    {
      "entropy": 5.388493251800537,
      "epoch": 2.571348346156006,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004347421589294864,
      "loss": 5.1488,
      "mean_token_accuracy": 0.20322621911764144,
      "num_tokens": 51100148.0,
      "step": 22895
    },
    {
      "entropy": 5.46496090888977,
      "epoch": 2.5719099230639637,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004347138821806997,
      "loss": 5.1873,
      "mean_token_accuracy": 0.20448700934648514,
      "num_tokens": 51111148.0,
      "step": 22900
    },
    {
      "entropy": 5.510583114624024,
      "epoch": 2.572471499971921,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000434685600346561,
      "loss": 5.168,
      "mean_token_accuracy": 0.20173385590314866,
      "num_tokens": 51122563.0,
      "step": 22905
    },
    {
      "entropy": 5.287107276916504,
      "epoch": 2.5730330768798786,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000434657313427971,
      "loss": 5.042,
      "mean_token_accuracy": 0.21202314347028733,
      "num_tokens": 51132676.0,
      "step": 22910
    },
    {
      "entropy": 5.319718742370606,
      "epoch": 2.5735946537878363,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004346290214258301,
      "loss": 5.0489,
      "mean_token_accuracy": 0.2025744140148163,
      "num_tokens": 51144317.0,
      "step": 22915
    },
    {
      "entropy": 5.493367052078247,
      "epoch": 2.574156230695794,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00043460072434103934,
      "loss": 5.2619,
      "mean_token_accuracy": 0.1939291775226593,
      "num_tokens": 51156181.0,
      "step": 22920
    },
    {
      "entropy": 5.313303232192993,
      "epoch": 2.5747178076037516,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004345724221744996,
      "loss": 4.9985,
      "mean_token_accuracy": 0.21352035403251649,
      "num_tokens": 51166825.0,
      "step": 22925
    },
    {
      "entropy": 5.4008873462677,
      "epoch": 2.575279384511709,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043454411492711197,
      "loss": 5.2222,
      "mean_token_accuracy": 0.20298070311546326,
      "num_tokens": 51178839.0,
      "step": 22930
    },
    {
      "entropy": 5.310602140426636,
      "epoch": 2.5758409614196665,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043451580259977793,
      "loss": 5.0795,
      "mean_token_accuracy": 0.20235954821109772,
      "num_tokens": 51190811.0,
      "step": 22935
    },
    {
      "entropy": 5.4252196788787845,
      "epoch": 2.576402538327624,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043448748519339875,
      "loss": 5.0699,
      "mean_token_accuracy": 0.20337700843811035,
      "num_tokens": 51201991.0,
      "step": 22940
    },
    {
      "entropy": 5.415936040878296,
      "epoch": 2.5769641152355813,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043445916270887626,
      "loss": 5.1048,
      "mean_token_accuracy": 0.20999229103326797,
      "num_tokens": 51213017.0,
      "step": 22945
    },
    {
      "entropy": 5.3535895347595215,
      "epoch": 2.577525692143539,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004344308351471123,
      "loss": 5.0564,
      "mean_token_accuracy": 0.20494825690984725,
      "num_tokens": 51224412.0,
      "step": 22950
    },
    {
      "entropy": 5.397398471832275,
      "epoch": 2.5780872690514967,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043440250250900854,
      "loss": 5.1227,
      "mean_token_accuracy": 0.20350587815046312,
      "num_tokens": 51235043.0,
      "step": 22955
    },
    {
      "entropy": 5.356507968902588,
      "epoch": 2.578648845959454,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004343741647954673,
      "loss": 5.1502,
      "mean_token_accuracy": 0.20113606452941896,
      "num_tokens": 51246182.0,
      "step": 22960
    },
    {
      "entropy": 5.416786813735962,
      "epoch": 2.5792104228674115,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004343458220073909,
      "loss": 5.206,
      "mean_token_accuracy": 0.20547375380992888,
      "num_tokens": 51259791.0,
      "step": 22965
    },
    {
      "entropy": 5.458284711837768,
      "epoch": 2.579771999775369,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004343174741456817,
      "loss": 5.1969,
      "mean_token_accuracy": 0.2006654217839241,
      "num_tokens": 51270530.0,
      "step": 22970
    },
    {
      "entropy": 5.330371570587158,
      "epoch": 2.580333576683327,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004342891212112423,
      "loss": 5.0547,
      "mean_token_accuracy": 0.2050156384706497,
      "num_tokens": 51281969.0,
      "step": 22975
    },
    {
      "entropy": 5.295652389526367,
      "epoch": 2.5808951535912845,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043426076320497557,
      "loss": 4.948,
      "mean_token_accuracy": 0.2116953268647194,
      "num_tokens": 51292790.0,
      "step": 22980
    },
    {
      "entropy": 5.320979022979737,
      "epoch": 2.5814567304992417,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043423240012778437,
      "loss": 5.0597,
      "mean_token_accuracy": 0.20204729735851287,
      "num_tokens": 51303949.0,
      "step": 22985
    },
    {
      "entropy": 5.3227867603302,
      "epoch": 2.5820183074071994,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004342040319805717,
      "loss": 5.0134,
      "mean_token_accuracy": 0.2106783390045166,
      "num_tokens": 51314570.0,
      "step": 22990
    },
    {
      "entropy": 5.221020317077636,
      "epoch": 2.582579884315157,
      "grad_norm": 0.953125,
      "learning_rate": 0.000434175658764241,
      "loss": 4.9554,
      "mean_token_accuracy": 0.21475682109594346,
      "num_tokens": 51325210.0,
      "step": 22995
    },
    {
      "entropy": 5.305114603042602,
      "epoch": 2.5831414612231143,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043414728047969553,
      "loss": 4.9291,
      "mean_token_accuracy": 0.21331031918525695,
      "num_tokens": 51335871.0,
      "step": 23000
    },
    {
      "entropy": 5.374608850479126,
      "epoch": 2.583703038131072,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043411889712783897,
      "loss": 5.1089,
      "mean_token_accuracy": 0.20967003107070922,
      "num_tokens": 51346821.0,
      "step": 23005
    },
    {
      "entropy": 5.379583406448364,
      "epoch": 2.5842646150390296,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043409050870957505,
      "loss": 5.0895,
      "mean_token_accuracy": 0.1967412993311882,
      "num_tokens": 51358746.0,
      "step": 23010
    },
    {
      "entropy": 5.399767827987671,
      "epoch": 2.5848261919469873,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004340621152258075,
      "loss": 5.1221,
      "mean_token_accuracy": 0.2015463173389435,
      "num_tokens": 51369628.0,
      "step": 23015
    },
    {
      "entropy": 5.461532402038574,
      "epoch": 2.585387768854945,
      "grad_norm": 1.0,
      "learning_rate": 0.00043403371667744055,
      "loss": 5.095,
      "mean_token_accuracy": 0.2115330293774605,
      "num_tokens": 51382160.0,
      "step": 23020
    },
    {
      "entropy": 5.374148511886597,
      "epoch": 2.585949345762902,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004340053130653785,
      "loss": 5.1136,
      "mean_token_accuracy": 0.20409150123596193,
      "num_tokens": 51394843.0,
      "step": 23025
    },
    {
      "entropy": 5.355620670318603,
      "epoch": 2.58651092267086,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004339769043905255,
      "loss": 5.0777,
      "mean_token_accuracy": 0.2091650351881981,
      "num_tokens": 51405714.0,
      "step": 23030
    },
    {
      "entropy": 5.371368646621704,
      "epoch": 2.5870724995788175,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043394849065378624,
      "loss": 5.0426,
      "mean_token_accuracy": 0.20763597190380095,
      "num_tokens": 51416292.0,
      "step": 23035
    },
    {
      "entropy": 5.336013889312744,
      "epoch": 2.5876340764867747,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043392007185606533,
      "loss": 4.9953,
      "mean_token_accuracy": 0.21139849126338958,
      "num_tokens": 51426993.0,
      "step": 23040
    },
    {
      "entropy": 5.289787721633911,
      "epoch": 2.5881956533947323,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004338916479982678,
      "loss": 5.1425,
      "mean_token_accuracy": 0.2042524993419647,
      "num_tokens": 51439493.0,
      "step": 23045
    },
    {
      "entropy": 5.3841626167297365,
      "epoch": 2.58875723030269,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043386321908129856,
      "loss": 5.0434,
      "mean_token_accuracy": 0.20693590342998505,
      "num_tokens": 51451705.0,
      "step": 23050
    },
    {
      "entropy": 5.441543006896973,
      "epoch": 2.589318807210647,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00043383478510606276,
      "loss": 5.0455,
      "mean_token_accuracy": 0.21044283658266066,
      "num_tokens": 51463044.0,
      "step": 23055
    },
    {
      "entropy": 5.262734317779541,
      "epoch": 2.589880384118605,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004338063460734658,
      "loss": 5.0731,
      "mean_token_accuracy": 0.20718219727277756,
      "num_tokens": 51475009.0,
      "step": 23060
    },
    {
      "entropy": 5.3030548095703125,
      "epoch": 2.5904419610265625,
      "grad_norm": 1.0,
      "learning_rate": 0.0004337779019844132,
      "loss": 5.0839,
      "mean_token_accuracy": 0.20573724955320358,
      "num_tokens": 51486068.0,
      "step": 23065
    },
    {
      "entropy": 5.437864542007446,
      "epoch": 2.59100353793452,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043374945283981066,
      "loss": 5.1227,
      "mean_token_accuracy": 0.2065598875284195,
      "num_tokens": 51497642.0,
      "step": 23070
    },
    {
      "entropy": 5.447027349472046,
      "epoch": 2.591565114842478,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004337209986405639,
      "loss": 5.1278,
      "mean_token_accuracy": 0.20126843750476836,
      "num_tokens": 51511013.0,
      "step": 23075
    },
    {
      "entropy": 5.449158525466919,
      "epoch": 2.592126691750435,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000433692539387579,
      "loss": 5.2429,
      "mean_token_accuracy": 0.1969563126564026,
      "num_tokens": 51521687.0,
      "step": 23080
    },
    {
      "entropy": 5.417449331283569,
      "epoch": 2.5926882686583927,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004336640750817621,
      "loss": 5.1247,
      "mean_token_accuracy": 0.20303906202316285,
      "num_tokens": 51532288.0,
      "step": 23085
    },
    {
      "entropy": 5.4335267543792725,
      "epoch": 2.5932498455663504,
      "grad_norm": 1.125,
      "learning_rate": 0.00043363560572401953,
      "loss": 5.0989,
      "mean_token_accuracy": 0.20949408859014512,
      "num_tokens": 51543101.0,
      "step": 23090
    },
    {
      "entropy": 5.331538820266724,
      "epoch": 2.5938114224743076,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043360713131525774,
      "loss": 5.1694,
      "mean_token_accuracy": 0.20160995572805404,
      "num_tokens": 51554729.0,
      "step": 23095
    },
    {
      "entropy": 5.339231729507446,
      "epoch": 2.5943729993822653,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004335786518563834,
      "loss": 5.0739,
      "mean_token_accuracy": 0.20566402673721312,
      "num_tokens": 51567232.0,
      "step": 23100
    },
    {
      "entropy": 5.429230833053589,
      "epoch": 2.594934576290223,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004335501673483033,
      "loss": 5.0649,
      "mean_token_accuracy": 0.20836181491613387,
      "num_tokens": 51577679.0,
      "step": 23105
    },
    {
      "entropy": 5.352954721450805,
      "epoch": 2.5954961531981806,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043352167779192423,
      "loss": 4.9972,
      "mean_token_accuracy": 0.21185696125030518,
      "num_tokens": 51588656.0,
      "step": 23110
    },
    {
      "entropy": 5.401973485946655,
      "epoch": 2.5960577301061383,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043349318318815363,
      "loss": 5.0865,
      "mean_token_accuracy": 0.207855923473835,
      "num_tokens": 51599016.0,
      "step": 23115
    },
    {
      "entropy": 5.201404333114624,
      "epoch": 2.5966193070140955,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043346468353789853,
      "loss": 4.8711,
      "mean_token_accuracy": 0.21768977493047714,
      "num_tokens": 51610288.0,
      "step": 23120
    },
    {
      "entropy": 5.358903264999389,
      "epoch": 2.597180883922053,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004334361788420664,
      "loss": 5.1635,
      "mean_token_accuracy": 0.2031911239027977,
      "num_tokens": 51622450.0,
      "step": 23125
    },
    {
      "entropy": 5.392947435379028,
      "epoch": 2.597742460830011,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043340766910156497,
      "loss": 4.9964,
      "mean_token_accuracy": 0.2089812934398651,
      "num_tokens": 51635111.0,
      "step": 23130
    },
    {
      "entropy": 5.40368242263794,
      "epoch": 2.598304037737968,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043337915431730195,
      "loss": 5.1204,
      "mean_token_accuracy": 0.2045867472887039,
      "num_tokens": 51646088.0,
      "step": 23135
    },
    {
      "entropy": 5.359442281723022,
      "epoch": 2.5988656146459257,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043335063449018515,
      "loss": 5.0459,
      "mean_token_accuracy": 0.20769044160842895,
      "num_tokens": 51659337.0,
      "step": 23140
    },
    {
      "entropy": 5.421291208267212,
      "epoch": 2.5994271915538834,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004333221096211227,
      "loss": 5.1882,
      "mean_token_accuracy": 0.20331063270568847,
      "num_tokens": 51670954.0,
      "step": 23145
    },
    {
      "entropy": 5.315551614761352,
      "epoch": 2.599988768461841,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043329357971102296,
      "loss": 5.0242,
      "mean_token_accuracy": 0.2091248854994774,
      "num_tokens": 51681104.0,
      "step": 23150
    },
    {
      "entropy": 5.245476198196411,
      "epoch": 2.6005503453697982,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043326504476079425,
      "loss": 4.9116,
      "mean_token_accuracy": 0.2172659918665886,
      "num_tokens": 51691337.0,
      "step": 23155
    },
    {
      "entropy": 5.29286470413208,
      "epoch": 2.601111922277756,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004332365047713451,
      "loss": 5.0374,
      "mean_token_accuracy": 0.20359603464603424,
      "num_tokens": 51702488.0,
      "step": 23160
    },
    {
      "entropy": 5.4405616283416744,
      "epoch": 2.6016734991857136,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004332079597435842,
      "loss": 5.1181,
      "mean_token_accuracy": 0.21350227147340775,
      "num_tokens": 51714439.0,
      "step": 23165
    },
    {
      "entropy": 5.506386184692383,
      "epoch": 2.602235076093671,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004331794096784206,
      "loss": 5.1633,
      "mean_token_accuracy": 0.20367325395345687,
      "num_tokens": 51726423.0,
      "step": 23170
    },
    {
      "entropy": 5.414096689224243,
      "epoch": 2.6027966530016284,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004331508545767633,
      "loss": 5.1477,
      "mean_token_accuracy": 0.21351937502622603,
      "num_tokens": 51736198.0,
      "step": 23175
    },
    {
      "entropy": 5.284890079498291,
      "epoch": 2.603358229909586,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004331222944395213,
      "loss": 5.019,
      "mean_token_accuracy": 0.21264111250638962,
      "num_tokens": 51746993.0,
      "step": 23180
    },
    {
      "entropy": 5.328834009170532,
      "epoch": 2.6039198068175438,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004330937292676042,
      "loss": 5.0385,
      "mean_token_accuracy": 0.20635722130537032,
      "num_tokens": 51756981.0,
      "step": 23185
    },
    {
      "entropy": 5.310337877273559,
      "epoch": 2.604481383725501,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004330651590619214,
      "loss": 5.0603,
      "mean_token_accuracy": 0.20785594433546067,
      "num_tokens": 51768192.0,
      "step": 23190
    },
    {
      "entropy": 5.346359825134277,
      "epoch": 2.6050429606334586,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004330365838233827,
      "loss": 5.0918,
      "mean_token_accuracy": 0.20464052110910416,
      "num_tokens": 51778996.0,
      "step": 23195
    },
    {
      "entropy": 5.392290019989014,
      "epoch": 2.6056045375414163,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043300800355289786,
      "loss": 5.1236,
      "mean_token_accuracy": 0.19716514647006989,
      "num_tokens": 51790179.0,
      "step": 23200
    },
    {
      "entropy": 5.411373281478882,
      "epoch": 2.606166114449374,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043297941825137684,
      "loss": 5.1348,
      "mean_token_accuracy": 0.20872581899166107,
      "num_tokens": 51801312.0,
      "step": 23205
    },
    {
      "entropy": 5.36368498802185,
      "epoch": 2.6067276913573316,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043295082791972993,
      "loss": 5.0493,
      "mean_token_accuracy": 0.21198984384536743,
      "num_tokens": 51811960.0,
      "step": 23210
    },
    {
      "entropy": 5.3701560497283936,
      "epoch": 2.607289268265289,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043292223255886737,
      "loss": 5.1409,
      "mean_token_accuracy": 0.20090073943138123,
      "num_tokens": 51823166.0,
      "step": 23215
    },
    {
      "entropy": 5.353069305419922,
      "epoch": 2.6078508451732465,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043289363216969973,
      "loss": 5.0305,
      "mean_token_accuracy": 0.20262852311134338,
      "num_tokens": 51835517.0,
      "step": 23220
    },
    {
      "entropy": 5.329936599731445,
      "epoch": 2.608412422081204,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043286502675313757,
      "loss": 5.0525,
      "mean_token_accuracy": 0.21196455359458924,
      "num_tokens": 51846690.0,
      "step": 23225
    },
    {
      "entropy": 5.2875665664672855,
      "epoch": 2.6089739989891614,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004328364163100917,
      "loss": 5.0387,
      "mean_token_accuracy": 0.21192826926708222,
      "num_tokens": 51857234.0,
      "step": 23230
    },
    {
      "entropy": 5.375195550918579,
      "epoch": 2.609535575897119,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004328078008414731,
      "loss": 5.1193,
      "mean_token_accuracy": 0.20915010273456575,
      "num_tokens": 51867913.0,
      "step": 23235
    },
    {
      "entropy": 5.36847448348999,
      "epoch": 2.6100971528050767,
      "grad_norm": 1.046875,
      "learning_rate": 0.000432779180348193,
      "loss": 5.0792,
      "mean_token_accuracy": 0.2148212105035782,
      "num_tokens": 51878252.0,
      "step": 23240
    },
    {
      "entropy": 5.281154012680053,
      "epoch": 2.6106587297130344,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043275055483116254,
      "loss": 4.9941,
      "mean_token_accuracy": 0.20693284273147583,
      "num_tokens": 51888407.0,
      "step": 23245
    },
    {
      "entropy": 5.365239810943604,
      "epoch": 2.6112203066209916,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004327219242912933,
      "loss": 5.0769,
      "mean_token_accuracy": 0.20607816129922868,
      "num_tokens": 51900042.0,
      "step": 23250
    },
    {
      "entropy": 5.378627157211303,
      "epoch": 2.6117818835289492,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004326932887294967,
      "loss": 5.0794,
      "mean_token_accuracy": 0.20439748764038085,
      "num_tokens": 51909691.0,
      "step": 23255
    },
    {
      "entropy": 5.434157943725586,
      "epoch": 2.612343460436907,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004326646481466848,
      "loss": 5.1267,
      "mean_token_accuracy": 0.20319681167602538,
      "num_tokens": 51921691.0,
      "step": 23260
    },
    {
      "entropy": 5.301544809341431,
      "epoch": 2.6129050373448646,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004326360025437692,
      "loss": 5.0185,
      "mean_token_accuracy": 0.20760179460048675,
      "num_tokens": 51934050.0,
      "step": 23265
    },
    {
      "entropy": 5.338772773742676,
      "epoch": 2.613466614252822,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043260735192166217,
      "loss": 5.0518,
      "mean_token_accuracy": 0.20997365117073058,
      "num_tokens": 51944566.0,
      "step": 23270
    },
    {
      "entropy": 5.359357023239136,
      "epoch": 2.6140281911607794,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000432578696281276,
      "loss": 4.9823,
      "mean_token_accuracy": 0.21775347143411636,
      "num_tokens": 51955485.0,
      "step": 23275
    },
    {
      "entropy": 5.290650749206543,
      "epoch": 2.614589768068737,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000432550035623523,
      "loss": 5.0605,
      "mean_token_accuracy": 0.21034972071647645,
      "num_tokens": 51965124.0,
      "step": 23280
    },
    {
      "entropy": 5.432386684417724,
      "epoch": 2.6151513449766943,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004325213699493157,
      "loss": 5.0899,
      "mean_token_accuracy": 0.21329840123653412,
      "num_tokens": 51976362.0,
      "step": 23285
    },
    {
      "entropy": 5.291180992126465,
      "epoch": 2.615712921884652,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004324926992595669,
      "loss": 4.9804,
      "mean_token_accuracy": 0.20828547477722167,
      "num_tokens": 51988068.0,
      "step": 23290
    },
    {
      "entropy": 5.314297437667847,
      "epoch": 2.6162744987926096,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004324640235551895,
      "loss": 5.0284,
      "mean_token_accuracy": 0.2091020777821541,
      "num_tokens": 51998392.0,
      "step": 23295
    },
    {
      "entropy": 5.278060102462769,
      "epoch": 2.6168360757005673,
      "grad_norm": 1.0,
      "learning_rate": 0.0004324353428370965,
      "loss": 4.9912,
      "mean_token_accuracy": 0.21019038707017898,
      "num_tokens": 52009416.0,
      "step": 23300
    },
    {
      "entropy": 5.379085493087769,
      "epoch": 2.617397652608525,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004324066571062012,
      "loss": 5.0688,
      "mean_token_accuracy": 0.2096051275730133,
      "num_tokens": 52020554.0,
      "step": 23305
    },
    {
      "entropy": 5.340315532684326,
      "epoch": 2.617959229516482,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004323779663634168,
      "loss": 5.0724,
      "mean_token_accuracy": 0.20562739670276642,
      "num_tokens": 52031301.0,
      "step": 23310
    },
    {
      "entropy": 5.334989595413208,
      "epoch": 2.61852080642444,
      "grad_norm": 1.015625,
      "learning_rate": 0.000432349270609657,
      "loss": 5.0236,
      "mean_token_accuracy": 0.20727073699235915,
      "num_tokens": 52041720.0,
      "step": 23315
    },
    {
      "entropy": 5.276964950561523,
      "epoch": 2.6190823833323975,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043232056984583527,
      "loss": 4.9579,
      "mean_token_accuracy": 0.21574908047914504,
      "num_tokens": 52051911.0,
      "step": 23320
    },
    {
      "entropy": 5.33335394859314,
      "epoch": 2.6196439602403547,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004322918640728656,
      "loss": 5.0419,
      "mean_token_accuracy": 0.20607040971517562,
      "num_tokens": 52062842.0,
      "step": 23325
    },
    {
      "entropy": 5.355015659332276,
      "epoch": 2.6202055371483124,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043226315329166203,
      "loss": 5.1696,
      "mean_token_accuracy": 0.2012346252799034,
      "num_tokens": 52075034.0,
      "step": 23330
    },
    {
      "entropy": 5.402558422088623,
      "epoch": 2.62076711405627,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043223443750313873,
      "loss": 4.9848,
      "mean_token_accuracy": 0.2141025498509407,
      "num_tokens": 52084782.0,
      "step": 23335
    },
    {
      "entropy": 5.247109842300415,
      "epoch": 2.6213286909642277,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004322057167082099,
      "loss": 5.0019,
      "mean_token_accuracy": 0.2062771812081337,
      "num_tokens": 52095623.0,
      "step": 23340
    },
    {
      "entropy": 5.294002676010132,
      "epoch": 2.621890267872185,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043217699090779006,
      "loss": 5.0433,
      "mean_token_accuracy": 0.2139701172709465,
      "num_tokens": 52107240.0,
      "step": 23345
    },
    {
      "entropy": 5.364339303970337,
      "epoch": 2.6224518447801426,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043214826010279383,
      "loss": 5.1544,
      "mean_token_accuracy": 0.20749560743570328,
      "num_tokens": 52118157.0,
      "step": 23350
    },
    {
      "entropy": 5.4658205032348635,
      "epoch": 2.6230134216881003,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004321195242941361,
      "loss": 5.1718,
      "mean_token_accuracy": 0.19401295483112335,
      "num_tokens": 52130578.0,
      "step": 23355
    },
    {
      "entropy": 5.388621854782104,
      "epoch": 2.623574998596058,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043209078348273175,
      "loss": 5.018,
      "mean_token_accuracy": 0.20971461087465287,
      "num_tokens": 52142057.0,
      "step": 23360
    },
    {
      "entropy": 5.408728361129761,
      "epoch": 2.624136575504015,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004320620376694959,
      "loss": 5.1519,
      "mean_token_accuracy": 0.20398398041725158,
      "num_tokens": 52152370.0,
      "step": 23365
    },
    {
      "entropy": 5.301637935638428,
      "epoch": 2.624698152411973,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043203328685534396,
      "loss": 5.0433,
      "mean_token_accuracy": 0.21166160553693772,
      "num_tokens": 52162139.0,
      "step": 23370
    },
    {
      "entropy": 5.3518774032592775,
      "epoch": 2.6252597293199305,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004320045310411912,
      "loss": 5.1202,
      "mean_token_accuracy": 0.20423512756824494,
      "num_tokens": 52172820.0,
      "step": 23375
    },
    {
      "entropy": 5.290776872634888,
      "epoch": 2.6258213062278877,
      "grad_norm": 1.0,
      "learning_rate": 0.00043197577022795327,
      "loss": 4.9519,
      "mean_token_accuracy": 0.21418640911579132,
      "num_tokens": 52183315.0,
      "step": 23380
    },
    {
      "entropy": 5.261458063125611,
      "epoch": 2.6263828831358453,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004319470044165459,
      "loss": 5.0222,
      "mean_token_accuracy": 0.21670226603746415,
      "num_tokens": 52194776.0,
      "step": 23385
    },
    {
      "entropy": 5.475599336624145,
      "epoch": 2.626944460043803,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004319182336078852,
      "loss": 5.2046,
      "mean_token_accuracy": 0.20121908634901048,
      "num_tokens": 52207370.0,
      "step": 23390
    },
    {
      "entropy": 5.348545694351197,
      "epoch": 2.6275060369517607,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000431889457802887,
      "loss": 5.0025,
      "mean_token_accuracy": 0.2146325007081032,
      "num_tokens": 52218287.0,
      "step": 23395
    },
    {
      "entropy": 5.340466642379761,
      "epoch": 2.6280676138597183,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004318606770024675,
      "loss": 5.1464,
      "mean_token_accuracy": 0.20035791248083115,
      "num_tokens": 52229418.0,
      "step": 23400
    },
    {
      "entropy": 5.449384880065918,
      "epoch": 2.6286291907676755,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004318318912075433,
      "loss": 5.0763,
      "mean_token_accuracy": 0.20615590512752532,
      "num_tokens": 52240382.0,
      "step": 23405
    },
    {
      "entropy": 5.419957828521729,
      "epoch": 2.629190767675633,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004318031004190308,
      "loss": 5.1155,
      "mean_token_accuracy": 0.20560937970876694,
      "num_tokens": 52252608.0,
      "step": 23410
    },
    {
      "entropy": 5.310688924789429,
      "epoch": 2.629752344583591,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043177430463784685,
      "loss": 4.9971,
      "mean_token_accuracy": 0.22276368737220764,
      "num_tokens": 52262940.0,
      "step": 23415
    },
    {
      "entropy": 5.240532636642456,
      "epoch": 2.630313921491548,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043174550386490816,
      "loss": 4.964,
      "mean_token_accuracy": 0.21494892984628677,
      "num_tokens": 52273270.0,
      "step": 23420
    },
    {
      "entropy": 5.278834486007691,
      "epoch": 2.6308754983995057,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004317166981011319,
      "loss": 5.0392,
      "mean_token_accuracy": 0.20881633162498475,
      "num_tokens": 52284433.0,
      "step": 23425
    },
    {
      "entropy": 5.403772687911987,
      "epoch": 2.6314370753074634,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043168788734743516,
      "loss": 5.0534,
      "mean_token_accuracy": 0.21475321799516678,
      "num_tokens": 52296088.0,
      "step": 23430
    },
    {
      "entropy": 5.3656634330749515,
      "epoch": 2.631998652215421,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043165907160473525,
      "loss": 5.1284,
      "mean_token_accuracy": 0.19868915528059006,
      "num_tokens": 52307683.0,
      "step": 23435
    },
    {
      "entropy": 5.33123254776001,
      "epoch": 2.6325602291233783,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004316302508739499,
      "loss": 5.1072,
      "mean_token_accuracy": 0.2036604478955269,
      "num_tokens": 52319845.0,
      "step": 23440
    },
    {
      "entropy": 5.375166606903076,
      "epoch": 2.633121806031336,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004316014251559965,
      "loss": 5.0598,
      "mean_token_accuracy": 0.20869748294353485,
      "num_tokens": 52330196.0,
      "step": 23445
    },
    {
      "entropy": 5.459926128387451,
      "epoch": 2.6336833829392936,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00043157259445179304,
      "loss": 5.2469,
      "mean_token_accuracy": 0.19826640337705612,
      "num_tokens": 52341510.0,
      "step": 23450
    },
    {
      "entropy": 5.3643793106079105,
      "epoch": 2.6342449598472513,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004315437587622574,
      "loss": 5.0015,
      "mean_token_accuracy": 0.2080070808529854,
      "num_tokens": 52351848.0,
      "step": 23455
    },
    {
      "entropy": 5.374601602554321,
      "epoch": 2.6348065367552085,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043151491808830777,
      "loss": 5.0589,
      "mean_token_accuracy": 0.20673497915267944,
      "num_tokens": 52363698.0,
      "step": 23460
    },
    {
      "entropy": 5.378453397750855,
      "epoch": 2.635368113663166,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043148607243086255,
      "loss": 5.0358,
      "mean_token_accuracy": 0.20849082320928575,
      "num_tokens": 52374433.0,
      "step": 23465
    },
    {
      "entropy": 5.291690397262573,
      "epoch": 2.635929690571124,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004314572217908399,
      "loss": 5.091,
      "mean_token_accuracy": 0.20648332685232162,
      "num_tokens": 52385762.0,
      "step": 23470
    },
    {
      "entropy": 5.304075431823731,
      "epoch": 2.636491267479081,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004314283661691587,
      "loss": 4.8622,
      "mean_token_accuracy": 0.21787068396806716,
      "num_tokens": 52397058.0,
      "step": 23475
    },
    {
      "entropy": 5.407605743408203,
      "epoch": 2.6370528443870387,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004313995055667377,
      "loss": 5.1404,
      "mean_token_accuracy": 0.2010448932647705,
      "num_tokens": 52407671.0,
      "step": 23480
    },
    {
      "entropy": 5.342394161224365,
      "epoch": 2.6376144212949963,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004313706399844958,
      "loss": 5.1109,
      "mean_token_accuracy": 0.20849163979291915,
      "num_tokens": 52419511.0,
      "step": 23485
    },
    {
      "entropy": 5.268547201156617,
      "epoch": 2.638175998202954,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000431341769423352,
      "loss": 4.9956,
      "mean_token_accuracy": 0.20771500319242478,
      "num_tokens": 52431311.0,
      "step": 23490
    },
    {
      "entropy": 5.37665023803711,
      "epoch": 2.6387375751109117,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004313128938842257,
      "loss": 5.0335,
      "mean_token_accuracy": 0.20625521540641784,
      "num_tokens": 52442184.0,
      "step": 23495
    },
    {
      "entropy": 5.3839939594268795,
      "epoch": 2.639299152018869,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043128401336803616,
      "loss": 5.0877,
      "mean_token_accuracy": 0.2024039536714554,
      "num_tokens": 52454837.0,
      "step": 23500
    },
    {
      "entropy": 5.382292985916138,
      "epoch": 2.6398607289268265,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043125512787570305,
      "loss": 5.0926,
      "mean_token_accuracy": 0.20668212324380875,
      "num_tokens": 52465378.0,
      "step": 23505
    },
    {
      "entropy": 5.37150821685791,
      "epoch": 2.640422305834784,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004312262374081461,
      "loss": 5.1251,
      "mean_token_accuracy": 0.2077893301844597,
      "num_tokens": 52476025.0,
      "step": 23510
    },
    {
      "entropy": 5.432770109176635,
      "epoch": 2.6409838827427414,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00043119734196628507,
      "loss": 5.0564,
      "mean_token_accuracy": 0.20585641115903855,
      "num_tokens": 52487408.0,
      "step": 23515
    },
    {
      "entropy": 5.349102735519409,
      "epoch": 2.641545459650699,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043116844155104025,
      "loss": 4.9965,
      "mean_token_accuracy": 0.21467709243297578,
      "num_tokens": 52497801.0,
      "step": 23520
    },
    {
      "entropy": 5.313938617706299,
      "epoch": 2.6421070365586568,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004311395361633316,
      "loss": 5.0623,
      "mean_token_accuracy": 0.20701874941587448,
      "num_tokens": 52508918.0,
      "step": 23525
    },
    {
      "entropy": 5.307626533508301,
      "epoch": 2.6426686134666144,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004311106258040795,
      "loss": 5.0057,
      "mean_token_accuracy": 0.2129124402999878,
      "num_tokens": 52519951.0,
      "step": 23530
    },
    {
      "entropy": 5.223710918426514,
      "epoch": 2.643230190374572,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043108171047420463,
      "loss": 5.0202,
      "mean_token_accuracy": 0.20800579935312272,
      "num_tokens": 52530710.0,
      "step": 23535
    },
    {
      "entropy": 5.237263345718384,
      "epoch": 2.6437917672825293,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004310527901746275,
      "loss": 4.9614,
      "mean_token_accuracy": 0.21652065068483353,
      "num_tokens": 52542188.0,
      "step": 23540
    },
    {
      "entropy": 5.3791567325592045,
      "epoch": 2.644353344190487,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000431023864906269,
      "loss": 5.1009,
      "mean_token_accuracy": 0.21036977767944337,
      "num_tokens": 52553848.0,
      "step": 23545
    },
    {
      "entropy": 5.342021656036377,
      "epoch": 2.6449149210984446,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043099493467005015,
      "loss": 5.0406,
      "mean_token_accuracy": 0.20821016579866408,
      "num_tokens": 52563338.0,
      "step": 23550
    },
    {
      "entropy": 5.40296630859375,
      "epoch": 2.645476498006402,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000430965999466892,
      "loss": 5.1731,
      "mean_token_accuracy": 0.2054514229297638,
      "num_tokens": 52576369.0,
      "step": 23555
    },
    {
      "entropy": 5.393935489654541,
      "epoch": 2.6460380749143595,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043093705929771604,
      "loss": 5.0868,
      "mean_token_accuracy": 0.21455593705177306,
      "num_tokens": 52587409.0,
      "step": 23560
    },
    {
      "entropy": 5.29620590209961,
      "epoch": 2.646599651822317,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004309081141634436,
      "loss": 5.0132,
      "mean_token_accuracy": 0.2041884869337082,
      "num_tokens": 52599398.0,
      "step": 23565
    },
    {
      "entropy": 5.316946601867675,
      "epoch": 2.6471612287302744,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043087916406499635,
      "loss": 5.0205,
      "mean_token_accuracy": 0.20966537147760392,
      "num_tokens": 52609565.0,
      "step": 23570
    },
    {
      "entropy": 5.361834526062012,
      "epoch": 2.647722805638232,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043085020900329604,
      "loss": 5.0856,
      "mean_token_accuracy": 0.20847289264202118,
      "num_tokens": 52621150.0,
      "step": 23575
    },
    {
      "entropy": 5.3488171100616455,
      "epoch": 2.6482843825461897,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043082124897926463,
      "loss": 5.0616,
      "mean_token_accuracy": 0.20821574479341506,
      "num_tokens": 52632030.0,
      "step": 23580
    },
    {
      "entropy": 5.396046113967896,
      "epoch": 2.6488459594541474,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004307922839938242,
      "loss": 5.1747,
      "mean_token_accuracy": 0.20198758989572524,
      "num_tokens": 52643447.0,
      "step": 23585
    },
    {
      "entropy": 5.400092840194702,
      "epoch": 2.649407536362105,
      "grad_norm": 0.953125,
      "learning_rate": 0.00043076331404789703,
      "loss": 5.0473,
      "mean_token_accuracy": 0.21251001209020615,
      "num_tokens": 52655017.0,
      "step": 23590
    },
    {
      "entropy": 5.412901210784912,
      "epoch": 2.6499691132700622,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004307343391424055,
      "loss": 5.0852,
      "mean_token_accuracy": 0.20645769387483598,
      "num_tokens": 52665674.0,
      "step": 23595
    },
    {
      "entropy": 5.277818918228149,
      "epoch": 2.65053069017802,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004307053592782723,
      "loss": 5.0352,
      "mean_token_accuracy": 0.2087297484278679,
      "num_tokens": 52678177.0,
      "step": 23600
    },
    {
      "entropy": 5.351116180419922,
      "epoch": 2.6510922670859776,
      "grad_norm": 1.125,
      "learning_rate": 0.00043067637445642004,
      "loss": 5.1277,
      "mean_token_accuracy": 0.20456790030002595,
      "num_tokens": 52689326.0,
      "step": 23605
    },
    {
      "entropy": 5.414654922485352,
      "epoch": 2.651653843993935,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043064738467777166,
      "loss": 5.1486,
      "mean_token_accuracy": 0.19570215046405792,
      "num_tokens": 52701686.0,
      "step": 23610
    },
    {
      "entropy": 5.352618789672851,
      "epoch": 2.6522154209018924,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043061838994325007,
      "loss": 5.0573,
      "mean_token_accuracy": 0.2095184415578842,
      "num_tokens": 52713422.0,
      "step": 23615
    },
    {
      "entropy": 5.365526914596558,
      "epoch": 2.65277699780985,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043058939025377867,
      "loss": 5.0954,
      "mean_token_accuracy": 0.2135357066988945,
      "num_tokens": 52724438.0,
      "step": 23620
    },
    {
      "entropy": 5.33053822517395,
      "epoch": 2.6533385747178078,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004305603856102808,
      "loss": 5.0616,
      "mean_token_accuracy": 0.21326029896736146,
      "num_tokens": 52734657.0,
      "step": 23625
    },
    {
      "entropy": 5.331798553466797,
      "epoch": 2.6539001516257654,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004305313760136799,
      "loss": 5.0441,
      "mean_token_accuracy": 0.2144949585199356,
      "num_tokens": 52745832.0,
      "step": 23630
    },
    {
      "entropy": 5.378699684143067,
      "epoch": 2.6544617285337226,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00043050236146489955,
      "loss": 5.1834,
      "mean_token_accuracy": 0.20224661529064178,
      "num_tokens": 52758795.0,
      "step": 23635
    },
    {
      "entropy": 5.385630559921265,
      "epoch": 2.6550233054416803,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004304733419648638,
      "loss": 5.0876,
      "mean_token_accuracy": 0.2118660107254982,
      "num_tokens": 52769154.0,
      "step": 23640
    },
    {
      "entropy": 5.345931720733643,
      "epoch": 2.655584882349638,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043044431751449656,
      "loss": 5.0471,
      "mean_token_accuracy": 0.21084111630916597,
      "num_tokens": 52779189.0,
      "step": 23645
    },
    {
      "entropy": 5.411244201660156,
      "epoch": 2.656146459257595,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004304152881147219,
      "loss": 5.1647,
      "mean_token_accuracy": 0.20879255831241608,
      "num_tokens": 52789545.0,
      "step": 23650
    },
    {
      "entropy": 5.492417335510254,
      "epoch": 2.656708036165553,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004303862537664642,
      "loss": 5.2642,
      "mean_token_accuracy": 0.1949790358543396,
      "num_tokens": 52802410.0,
      "step": 23655
    },
    {
      "entropy": 5.347533750534057,
      "epoch": 2.6572696130735105,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000430357214470648,
      "loss": 5.0736,
      "mean_token_accuracy": 0.2082818940281868,
      "num_tokens": 52814223.0,
      "step": 23660
    },
    {
      "entropy": 5.306902360916138,
      "epoch": 2.6578311899814677,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004303281702281977,
      "loss": 5.0224,
      "mean_token_accuracy": 0.21131232976913453,
      "num_tokens": 52824554.0,
      "step": 23665
    },
    {
      "entropy": 5.42786602973938,
      "epoch": 2.6583927668894254,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043029912104003827,
      "loss": 5.2079,
      "mean_token_accuracy": 0.20774156749248504,
      "num_tokens": 52836058.0,
      "step": 23670
    },
    {
      "entropy": 5.368735504150391,
      "epoch": 2.658954343797383,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043027006690709457,
      "loss": 5.1438,
      "mean_token_accuracy": 0.20539306700229645,
      "num_tokens": 52847220.0,
      "step": 23675
    },
    {
      "entropy": 5.379433584213257,
      "epoch": 2.6595159207053407,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004302410078302918,
      "loss": 4.986,
      "mean_token_accuracy": 0.2108723610639572,
      "num_tokens": 52858242.0,
      "step": 23680
    },
    {
      "entropy": 5.283656740188599,
      "epoch": 2.6600774976132984,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043021194381055505,
      "loss": 4.9408,
      "mean_token_accuracy": 0.2112239271402359,
      "num_tokens": 52869439.0,
      "step": 23685
    },
    {
      "entropy": 5.453416538238526,
      "epoch": 2.6606390745212556,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043018287484880986,
      "loss": 5.2871,
      "mean_token_accuracy": 0.1957633912563324,
      "num_tokens": 52879893.0,
      "step": 23690
    },
    {
      "entropy": 5.430699586868286,
      "epoch": 2.6612006514292132,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004301538009459817,
      "loss": 5.101,
      "mean_token_accuracy": 0.2103486031293869,
      "num_tokens": 52891352.0,
      "step": 23695
    },
    {
      "entropy": 5.33960976600647,
      "epoch": 2.661762228337171,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004301247221029964,
      "loss": 5.1168,
      "mean_token_accuracy": 0.20088522732257844,
      "num_tokens": 52904202.0,
      "step": 23700
    },
    {
      "entropy": 5.355183410644531,
      "epoch": 2.662323805245128,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043009563832077977,
      "loss": 5.0118,
      "mean_token_accuracy": 0.2130415618419647,
      "num_tokens": 52914495.0,
      "step": 23705
    },
    {
      "entropy": 5.351727294921875,
      "epoch": 2.662885382153086,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004300665496002578,
      "loss": 5.0997,
      "mean_token_accuracy": 0.20406445860862732,
      "num_tokens": 52925930.0,
      "step": 23710
    },
    {
      "entropy": 5.4174267768859865,
      "epoch": 2.6634469590610435,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043003745594235687,
      "loss": 5.0583,
      "mean_token_accuracy": 0.20529289543628693,
      "num_tokens": 52936310.0,
      "step": 23715
    },
    {
      "entropy": 5.380947923660278,
      "epoch": 2.664008535969001,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004300083573480031,
      "loss": 5.1569,
      "mean_token_accuracy": 0.20312414467334747,
      "num_tokens": 52948801.0,
      "step": 23720
    },
    {
      "entropy": 5.412672424316407,
      "epoch": 2.6645701128769588,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042997925381812313,
      "loss": 5.2252,
      "mean_token_accuracy": 0.20632123202085495,
      "num_tokens": 52960541.0,
      "step": 23725
    },
    {
      "entropy": 5.400978136062622,
      "epoch": 2.665131689784916,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042995014535364363,
      "loss": 5.0905,
      "mean_token_accuracy": 0.20851745158433915,
      "num_tokens": 52971973.0,
      "step": 23730
    },
    {
      "entropy": 5.3363443374633786,
      "epoch": 2.6656932666928737,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004299210319554914,
      "loss": 5.0515,
      "mean_token_accuracy": 0.2235025092959404,
      "num_tokens": 52983124.0,
      "step": 23735
    },
    {
      "entropy": 5.335207557678222,
      "epoch": 2.6662548436008313,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042989191362459347,
      "loss": 5.0093,
      "mean_token_accuracy": 0.21560687124729155,
      "num_tokens": 52994402.0,
      "step": 23740
    },
    {
      "entropy": 5.3945355892181395,
      "epoch": 2.6668164205087885,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004298627903618769,
      "loss": 5.13,
      "mean_token_accuracy": 0.19864969849586486,
      "num_tokens": 53005253.0,
      "step": 23745
    },
    {
      "entropy": 5.423642873764038,
      "epoch": 2.667377997416746,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042983366216826907,
      "loss": 5.1413,
      "mean_token_accuracy": 0.2047628790140152,
      "num_tokens": 53016599.0,
      "step": 23750
    },
    {
      "entropy": 5.355526399612427,
      "epoch": 2.667939574324704,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004298045290446973,
      "loss": 5.069,
      "mean_token_accuracy": 0.21071497946977616,
      "num_tokens": 53027412.0,
      "step": 23755
    },
    {
      "entropy": 5.35480523109436,
      "epoch": 2.668501151232661,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042977539099208937,
      "loss": 5.0974,
      "mean_token_accuracy": 0.2067239537835121,
      "num_tokens": 53039490.0,
      "step": 23760
    },
    {
      "entropy": 5.399639987945557,
      "epoch": 2.6690627281406187,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042974624801137283,
      "loss": 5.0881,
      "mean_token_accuracy": 0.2048841431736946,
      "num_tokens": 53049837.0,
      "step": 23765
    },
    {
      "entropy": 5.317218494415283,
      "epoch": 2.6696243050485764,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042971710010347594,
      "loss": 5.0104,
      "mean_token_accuracy": 0.21715941727161409,
      "num_tokens": 53060187.0,
      "step": 23770
    },
    {
      "entropy": 5.298475980758667,
      "epoch": 2.670185881956534,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042968794726932634,
      "loss": 5.0395,
      "mean_token_accuracy": 0.20881342738866807,
      "num_tokens": 53072050.0,
      "step": 23775
    },
    {
      "entropy": 5.362100076675415,
      "epoch": 2.6707474588644917,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004296587895098527,
      "loss": 5.1118,
      "mean_token_accuracy": 0.1990792080760002,
      "num_tokens": 53083235.0,
      "step": 23780
    },
    {
      "entropy": 5.464153623580932,
      "epoch": 2.671309035772449,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042962962682598316,
      "loss": 5.1278,
      "mean_token_accuracy": 0.20357729196548463,
      "num_tokens": 53093615.0,
      "step": 23785
    },
    {
      "entropy": 5.417017459869385,
      "epoch": 2.6718706126804066,
      "grad_norm": 1.0,
      "learning_rate": 0.00042960045921864627,
      "loss": 5.1723,
      "mean_token_accuracy": 0.20092887431383133,
      "num_tokens": 53105611.0,
      "step": 23790
    },
    {
      "entropy": 5.369257688522339,
      "epoch": 2.6724321895883643,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004295712866887709,
      "loss": 4.9263,
      "mean_token_accuracy": 0.21216228902339934,
      "num_tokens": 53116343.0,
      "step": 23795
    },
    {
      "entropy": 5.379072093963623,
      "epoch": 2.6729937664963215,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042954210923728565,
      "loss": 5.0875,
      "mean_token_accuracy": 0.21580112725496292,
      "num_tokens": 53128754.0,
      "step": 23800
    },
    {
      "entropy": 5.271940851211548,
      "epoch": 2.673555343404279,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004295129268651198,
      "loss": 4.9725,
      "mean_token_accuracy": 0.21834630370140076,
      "num_tokens": 53139508.0,
      "step": 23805
    },
    {
      "entropy": 5.317148923873901,
      "epoch": 2.674116920312237,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004294837395732025,
      "loss": 4.9809,
      "mean_token_accuracy": 0.21650561541318894,
      "num_tokens": 53149925.0,
      "step": 23810
    },
    {
      "entropy": 5.313843059539795,
      "epoch": 2.6746784972201945,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042945454736246293,
      "loss": 5.049,
      "mean_token_accuracy": 0.20954806208610535,
      "num_tokens": 53160053.0,
      "step": 23815
    },
    {
      "entropy": 5.368851804733277,
      "epoch": 2.675240074128152,
      "grad_norm": 1.0,
      "learning_rate": 0.00042942535023383075,
      "loss": 5.0655,
      "mean_token_accuracy": 0.20884327590465546,
      "num_tokens": 53170723.0,
      "step": 23820
    },
    {
      "entropy": 5.357493352890015,
      "epoch": 2.6758016510361093,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042939614818823547,
      "loss": 5.0896,
      "mean_token_accuracy": 0.20579399764537812,
      "num_tokens": 53182120.0,
      "step": 23825
    },
    {
      "entropy": 5.380192279815674,
      "epoch": 2.676363227944067,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004293669412266069,
      "loss": 5.0755,
      "mean_token_accuracy": 0.20492709279060364,
      "num_tokens": 53192481.0,
      "step": 23830
    },
    {
      "entropy": 5.357753276824951,
      "epoch": 2.6769248048520247,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004293377293498751,
      "loss": 5.0659,
      "mean_token_accuracy": 0.2051623597741127,
      "num_tokens": 53203710.0,
      "step": 23835
    },
    {
      "entropy": 5.3474424362182615,
      "epoch": 2.677486381759982,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042930851255897014,
      "loss": 4.9952,
      "mean_token_accuracy": 0.21120841205120086,
      "num_tokens": 53214051.0,
      "step": 23840
    },
    {
      "entropy": 5.29975471496582,
      "epoch": 2.6780479586679395,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004292792908548223,
      "loss": 5.0199,
      "mean_token_accuracy": 0.21328401416540146,
      "num_tokens": 53224981.0,
      "step": 23845
    },
    {
      "entropy": 5.23189959526062,
      "epoch": 2.678609535575897,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042925006423836197,
      "loss": 4.8745,
      "mean_token_accuracy": 0.22238829284906386,
      "num_tokens": 53236336.0,
      "step": 23850
    },
    {
      "entropy": 5.354677534103393,
      "epoch": 2.6791711124838544,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00042922083271051975,
      "loss": 5.1008,
      "mean_token_accuracy": 0.2034292683005333,
      "num_tokens": 53248212.0,
      "step": 23855
    },
    {
      "entropy": 5.3845288276672365,
      "epoch": 2.679732689391812,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004291915962722265,
      "loss": 5.0844,
      "mean_token_accuracy": 0.20816956907510759,
      "num_tokens": 53258910.0,
      "step": 23860
    },
    {
      "entropy": 5.424533033370972,
      "epoch": 2.6802942662997697,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004291623549244129,
      "loss": 5.1223,
      "mean_token_accuracy": 0.20433868914842607,
      "num_tokens": 53269445.0,
      "step": 23865
    },
    {
      "entropy": 5.367903327941894,
      "epoch": 2.6808558432077274,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042913310866801024,
      "loss": 5.0406,
      "mean_token_accuracy": 0.21101130843162536,
      "num_tokens": 53280119.0,
      "step": 23870
    },
    {
      "entropy": 5.3810083866119385,
      "epoch": 2.681417420115685,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00042910385750394964,
      "loss": 5.0773,
      "mean_token_accuracy": 0.2125815212726593,
      "num_tokens": 53291496.0,
      "step": 23875
    },
    {
      "entropy": 5.358265018463134,
      "epoch": 2.6819789970236423,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004290746014331623,
      "loss": 5.049,
      "mean_token_accuracy": 0.2010360538959503,
      "num_tokens": 53303470.0,
      "step": 23880
    },
    {
      "entropy": 5.427646780014038,
      "epoch": 2.6825405739316,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042904534045658,
      "loss": 5.1477,
      "mean_token_accuracy": 0.20355771332979203,
      "num_tokens": 53313552.0,
      "step": 23885
    },
    {
      "entropy": 5.393313407897949,
      "epoch": 2.6831021508395576,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004290160745751343,
      "loss": 5.1627,
      "mean_token_accuracy": 0.20382462292909623,
      "num_tokens": 53325280.0,
      "step": 23890
    },
    {
      "entropy": 5.287936162948609,
      "epoch": 2.683663727747515,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004289868037897571,
      "loss": 4.9454,
      "mean_token_accuracy": 0.218405519425869,
      "num_tokens": 53334982.0,
      "step": 23895
    },
    {
      "entropy": 5.379474830627442,
      "epoch": 2.6842253046554725,
      "grad_norm": 1.125,
      "learning_rate": 0.0004289575281013802,
      "loss": 5.1054,
      "mean_token_accuracy": 0.20113924592733384,
      "num_tokens": 53346367.0,
      "step": 23900
    },
    {
      "entropy": 5.344570159912109,
      "epoch": 2.68478688156343,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000428928247510936,
      "loss": 5.0503,
      "mean_token_accuracy": 0.20767495185136794,
      "num_tokens": 53356907.0,
      "step": 23905
    },
    {
      "entropy": 5.327908086776733,
      "epoch": 2.685348458471388,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004288989620193567,
      "loss": 5.0178,
      "mean_token_accuracy": 0.20781024992465974,
      "num_tokens": 53366619.0,
      "step": 23910
    },
    {
      "entropy": 5.397247266769409,
      "epoch": 2.6859100353793455,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042886967162757475,
      "loss": 5.0837,
      "mean_token_accuracy": 0.20937106013298035,
      "num_tokens": 53377578.0,
      "step": 23915
    },
    {
      "entropy": 5.372220802307129,
      "epoch": 2.6864716122873027,
      "grad_norm": 0.9375,
      "learning_rate": 0.00042884037633652276,
      "loss": 5.0677,
      "mean_token_accuracy": 0.20842819809913635,
      "num_tokens": 53389416.0,
      "step": 23920
    },
    {
      "entropy": 5.319972562789917,
      "epoch": 2.6870331891952604,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042881107614713355,
      "loss": 4.9063,
      "mean_token_accuracy": 0.220466148853302,
      "num_tokens": 53399453.0,
      "step": 23925
    },
    {
      "entropy": 5.308684062957764,
      "epoch": 2.687594766103218,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042878177106034,
      "loss": 5.0758,
      "mean_token_accuracy": 0.205148346722126,
      "num_tokens": 53409933.0,
      "step": 23930
    },
    {
      "entropy": 5.382736110687256,
      "epoch": 2.6881563430111752,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004287524610770752,
      "loss": 5.1181,
      "mean_token_accuracy": 0.20699744671583176,
      "num_tokens": 53420943.0,
      "step": 23935
    },
    {
      "entropy": 5.278658246994018,
      "epoch": 2.688717919919133,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042872314619827234,
      "loss": 4.9379,
      "mean_token_accuracy": 0.2145269528031349,
      "num_tokens": 53431450.0,
      "step": 23940
    },
    {
      "entropy": 5.2472772121429445,
      "epoch": 2.6892794968270906,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000428693826424865,
      "loss": 4.8996,
      "mean_token_accuracy": 0.21876748502254487,
      "num_tokens": 53442198.0,
      "step": 23945
    },
    {
      "entropy": 5.309199953079224,
      "epoch": 2.6898410737350478,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004286645017577866,
      "loss": 4.9529,
      "mean_token_accuracy": 0.20794051736593247,
      "num_tokens": 53452411.0,
      "step": 23950
    },
    {
      "entropy": 5.260905647277832,
      "epoch": 2.6904026506430054,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042863517219797085,
      "loss": 5.02,
      "mean_token_accuracy": 0.21511953324079514,
      "num_tokens": 53462644.0,
      "step": 23955
    },
    {
      "entropy": 5.278943920135498,
      "epoch": 2.690964227550963,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042860583774635165,
      "loss": 5.0194,
      "mean_token_accuracy": 0.20631268471479416,
      "num_tokens": 53473258.0,
      "step": 23960
    },
    {
      "entropy": 5.373235702514648,
      "epoch": 2.6915258044589208,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000428576498403863,
      "loss": 5.0212,
      "mean_token_accuracy": 0.2099225863814354,
      "num_tokens": 53484011.0,
      "step": 23965
    },
    {
      "entropy": 5.370475482940674,
      "epoch": 2.6920873813668784,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000428547154171439,
      "loss": 5.0508,
      "mean_token_accuracy": 0.21701299399137497,
      "num_tokens": 53495549.0,
      "step": 23970
    },
    {
      "entropy": 5.333487462997437,
      "epoch": 2.6926489582748356,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004285178050500142,
      "loss": 5.1157,
      "mean_token_accuracy": 0.20898982733488083,
      "num_tokens": 53506721.0,
      "step": 23975
    },
    {
      "entropy": 5.439150333404541,
      "epoch": 2.6932105351827933,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004284884510405228,
      "loss": 5.2295,
      "mean_token_accuracy": 0.1964177444577217,
      "num_tokens": 53518363.0,
      "step": 23980
    },
    {
      "entropy": 5.311640882492066,
      "epoch": 2.693772112090751,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004284590921438997,
      "loss": 4.9697,
      "mean_token_accuracy": 0.21039088815450668,
      "num_tokens": 53529140.0,
      "step": 23985
    },
    {
      "entropy": 5.254176235198974,
      "epoch": 2.694333688998708,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042842972836107954,
      "loss": 5.0441,
      "mean_token_accuracy": 0.2041165456175804,
      "num_tokens": 53540707.0,
      "step": 23990
    },
    {
      "entropy": 5.338412189483643,
      "epoch": 2.694895265906666,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004284003596929973,
      "loss": 5.0322,
      "mean_token_accuracy": 0.21043017357587815,
      "num_tokens": 53551297.0,
      "step": 23995
    },
    {
      "entropy": 5.337515449523925,
      "epoch": 2.6954568428146235,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042837098614058815,
      "loss": 4.9379,
      "mean_token_accuracy": 0.21133470684289932,
      "num_tokens": 53562207.0,
      "step": 24000
    },
    {
      "epoch": 2.6954568428146235,
      "eval_entropy": 5.1799222616796365,
      "eval_loss": 5.191533088684082,
      "eval_mean_token_accuracy": 0.21026900977949398,
      "eval_num_tokens": 53562207.0,
      "eval_runtime": 24.1489,
      "eval_samples_per_second": 1284.242,
      "eval_steps_per_second": 160.546,
      "step": 24000
    },
    {
      "entropy": 5.335341119766236,
      "epoch": 2.696018419722581,
      "grad_norm": 1.0,
      "learning_rate": 0.00042834160770478726,
      "loss": 5.1193,
      "mean_token_accuracy": 0.20370425581932067,
      "num_tokens": 53573823.0,
      "step": 24005
    },
    {
      "entropy": 5.30514235496521,
      "epoch": 2.696579996630539,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042831222438653016,
      "loss": 4.9709,
      "mean_token_accuracy": 0.21689033061265944,
      "num_tokens": 53584724.0,
      "step": 24010
    },
    {
      "entropy": 5.444219589233398,
      "epoch": 2.697141573538496,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004282828361867524,
      "loss": 5.114,
      "mean_token_accuracy": 0.20354236513376237,
      "num_tokens": 53596212.0,
      "step": 24015
    },
    {
      "entropy": 5.383861112594604,
      "epoch": 2.6977031504464537,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042825344310638966,
      "loss": 5.0959,
      "mean_token_accuracy": 0.20757583975791932,
      "num_tokens": 53608341.0,
      "step": 24020
    },
    {
      "entropy": 5.352255249023438,
      "epoch": 2.6982647273544114,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042822404514637785,
      "loss": 5.0593,
      "mean_token_accuracy": 0.2110473945736885,
      "num_tokens": 53619598.0,
      "step": 24025
    },
    {
      "entropy": 5.305366086959839,
      "epoch": 2.6988263042623686,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000428194642307653,
      "loss": 5.0095,
      "mean_token_accuracy": 0.20907191634178163,
      "num_tokens": 53630174.0,
      "step": 24030
    },
    {
      "entropy": 5.391218280792236,
      "epoch": 2.6993878811703262,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004281652345911513,
      "loss": 5.09,
      "mean_token_accuracy": 0.20734569132328035,
      "num_tokens": 53641929.0,
      "step": 24035
    },
    {
      "entropy": 5.319969987869262,
      "epoch": 2.699949458078284,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042813582199780913,
      "loss": 4.9769,
      "mean_token_accuracy": 0.207482185959816,
      "num_tokens": 53652262.0,
      "step": 24040
    },
    {
      "entropy": 5.418235349655151,
      "epoch": 2.700511034986241,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004281064045285631,
      "loss": 5.1154,
      "mean_token_accuracy": 0.20837373584508895,
      "num_tokens": 53663351.0,
      "step": 24045
    },
    {
      "entropy": 5.258724069595337,
      "epoch": 2.701072611894199,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042807698218434964,
      "loss": 4.9831,
      "mean_token_accuracy": 0.21085526794195175,
      "num_tokens": 53674632.0,
      "step": 24050
    },
    {
      "entropy": 5.287291622161865,
      "epoch": 2.7016341888021564,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042804755496610576,
      "loss": 5.0947,
      "mean_token_accuracy": 0.2093037724494934,
      "num_tokens": 53686346.0,
      "step": 24055
    },
    {
      "entropy": 5.43842658996582,
      "epoch": 2.702195765710114,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042801812287476836,
      "loss": 5.1018,
      "mean_token_accuracy": 0.20815750658512117,
      "num_tokens": 53697519.0,
      "step": 24060
    },
    {
      "entropy": 5.361558866500855,
      "epoch": 2.7027573426180718,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004279886859112746,
      "loss": 5.0447,
      "mean_token_accuracy": 0.2063666746020317,
      "num_tokens": 53707858.0,
      "step": 24065
    },
    {
      "entropy": 5.2730762481689455,
      "epoch": 2.703318919526029,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042795924407656175,
      "loss": 5.0013,
      "mean_token_accuracy": 0.2121836245059967,
      "num_tokens": 53719168.0,
      "step": 24070
    },
    {
      "entropy": 5.345731210708618,
      "epoch": 2.7038804964339866,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00042792979737156717,
      "loss": 4.9975,
      "mean_token_accuracy": 0.2152164176106453,
      "num_tokens": 53730601.0,
      "step": 24075
    },
    {
      "entropy": 5.323926210403442,
      "epoch": 2.7044420733419443,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004279003457972286,
      "loss": 5.0125,
      "mean_token_accuracy": 0.2105117544531822,
      "num_tokens": 53741410.0,
      "step": 24080
    },
    {
      "entropy": 5.271752882003784,
      "epoch": 2.7050036502499015,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004278708893544837,
      "loss": 4.9994,
      "mean_token_accuracy": 0.21429509967565535,
      "num_tokens": 53752585.0,
      "step": 24085
    },
    {
      "entropy": 5.326888227462769,
      "epoch": 2.705565227157859,
      "grad_norm": 1.0,
      "learning_rate": 0.0004278414280442704,
      "loss": 5.0317,
      "mean_token_accuracy": 0.2096115380525589,
      "num_tokens": 53762604.0,
      "step": 24090
    },
    {
      "entropy": 5.270272874832154,
      "epoch": 2.706126804065817,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042781196186752676,
      "loss": 5.0513,
      "mean_token_accuracy": 0.2004112794995308,
      "num_tokens": 53773938.0,
      "step": 24095
    },
    {
      "entropy": 5.437699508666992,
      "epoch": 2.7066883809737745,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004277824908251909,
      "loss": 5.1368,
      "mean_token_accuracy": 0.2078937128186226,
      "num_tokens": 53785142.0,
      "step": 24100
    },
    {
      "entropy": 5.466690444946289,
      "epoch": 2.707249957881732,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004277530149182014,
      "loss": 5.1351,
      "mean_token_accuracy": 0.20292258709669114,
      "num_tokens": 53797352.0,
      "step": 24105
    },
    {
      "entropy": 5.397158479690551,
      "epoch": 2.7078115347896894,
      "grad_norm": 1.0,
      "learning_rate": 0.00042772353414749655,
      "loss": 5.0687,
      "mean_token_accuracy": 0.20530294477939606,
      "num_tokens": 53808027.0,
      "step": 24110
    },
    {
      "entropy": 5.236713600158692,
      "epoch": 2.708373111697647,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042769404851401524,
      "loss": 4.9625,
      "mean_token_accuracy": 0.21033705174922943,
      "num_tokens": 53818520.0,
      "step": 24115
    },
    {
      "entropy": 5.323527050018311,
      "epoch": 2.7089346886056047,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004276645580186961,
      "loss": 5.0267,
      "mean_token_accuracy": 0.21036600023508073,
      "num_tokens": 53829234.0,
      "step": 24120
    },
    {
      "entropy": 5.355731964111328,
      "epoch": 2.709496265513562,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042763506266247834,
      "loss": 4.9563,
      "mean_token_accuracy": 0.21330944895744325,
      "num_tokens": 53839097.0,
      "step": 24125
    },
    {
      "entropy": 5.299003744125367,
      "epoch": 2.7100578424215196,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004276055624463009,
      "loss": 5.0365,
      "mean_token_accuracy": 0.20563182085752488,
      "num_tokens": 53850375.0,
      "step": 24130
    },
    {
      "entropy": 5.272104024887085,
      "epoch": 2.7106194193294773,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042757605737110324,
      "loss": 4.9883,
      "mean_token_accuracy": 0.2231307476758957,
      "num_tokens": 53861711.0,
      "step": 24135
    },
    {
      "entropy": 5.33498249053955,
      "epoch": 2.7111809962374345,
      "grad_norm": 1.0,
      "learning_rate": 0.0004275465474378246,
      "loss": 5.0324,
      "mean_token_accuracy": 0.20660230219364167,
      "num_tokens": 53872081.0,
      "step": 24140
    },
    {
      "entropy": 5.345885992050171,
      "epoch": 2.711742573145392,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004275170326474049,
      "loss": 5.0148,
      "mean_token_accuracy": 0.20987122505903244,
      "num_tokens": 53882591.0,
      "step": 24145
    },
    {
      "entropy": 5.364799451828003,
      "epoch": 2.71230415005335,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042748751300078367,
      "loss": 5.0019,
      "mean_token_accuracy": 0.22104429006576537,
      "num_tokens": 53893246.0,
      "step": 24150
    },
    {
      "entropy": 5.232566165924072,
      "epoch": 2.7128657269613075,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004274579884989008,
      "loss": 4.9393,
      "mean_token_accuracy": 0.2250444322824478,
      "num_tokens": 53904579.0,
      "step": 24155
    },
    {
      "entropy": 5.233660268783569,
      "epoch": 2.713427303869265,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042742845914269645,
      "loss": 4.978,
      "mean_token_accuracy": 0.20911311060190202,
      "num_tokens": 53916342.0,
      "step": 24160
    },
    {
      "entropy": 5.356160020828247,
      "epoch": 2.7139888807772223,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042739892493311095,
      "loss": 5.0134,
      "mean_token_accuracy": 0.2114718034863472,
      "num_tokens": 53927414.0,
      "step": 24165
    },
    {
      "entropy": 5.346490001678466,
      "epoch": 2.71455045768518,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042736938587108447,
      "loss": 5.0583,
      "mean_token_accuracy": 0.21181682348251343,
      "num_tokens": 53938679.0,
      "step": 24170
    },
    {
      "entropy": 5.36653151512146,
      "epoch": 2.7151120345931377,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004273398419575577,
      "loss": 5.1082,
      "mean_token_accuracy": 0.20485886037349701,
      "num_tokens": 53951364.0,
      "step": 24175
    },
    {
      "entropy": 5.413525199890136,
      "epoch": 2.715673611501095,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004273102931934713,
      "loss": 5.1499,
      "mean_token_accuracy": 0.2023979604244232,
      "num_tokens": 53963590.0,
      "step": 24180
    },
    {
      "entropy": 5.464923763275147,
      "epoch": 2.7162351884090525,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000427280739579766,
      "loss": 5.128,
      "mean_token_accuracy": 0.20522390007972718,
      "num_tokens": 53974315.0,
      "step": 24185
    },
    {
      "entropy": 5.356031513214111,
      "epoch": 2.71679676531701,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042725118111738305,
      "loss": 5.1154,
      "mean_token_accuracy": 0.20495956987142563,
      "num_tokens": 53985058.0,
      "step": 24190
    },
    {
      "entropy": 5.335622644424438,
      "epoch": 2.717358342224968,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004272216178072633,
      "loss": 5.014,
      "mean_token_accuracy": 0.21741616427898408,
      "num_tokens": 53994876.0,
      "step": 24195
    },
    {
      "entropy": 5.370824241638184,
      "epoch": 2.7179199191329255,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004271920496503483,
      "loss": 5.0939,
      "mean_token_accuracy": 0.20403338223695755,
      "num_tokens": 54005705.0,
      "step": 24200
    },
    {
      "entropy": 5.362220668792725,
      "epoch": 2.7184814960408827,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042716247664757934,
      "loss": 5.0233,
      "mean_token_accuracy": 0.2142615497112274,
      "num_tokens": 54015787.0,
      "step": 24205
    },
    {
      "entropy": 5.381118059158325,
      "epoch": 2.7190430729488404,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042713289879989816,
      "loss": 5.0819,
      "mean_token_accuracy": 0.20687510669231415,
      "num_tokens": 54027545.0,
      "step": 24210
    },
    {
      "entropy": 5.409627103805542,
      "epoch": 2.719604649856798,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004271033161082464,
      "loss": 5.1007,
      "mean_token_accuracy": 0.2109963923692703,
      "num_tokens": 54038560.0,
      "step": 24215
    },
    {
      "entropy": 5.4384928226470945,
      "epoch": 2.7201662267647553,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042707372857356615,
      "loss": 5.1237,
      "mean_token_accuracy": 0.20294771790504457,
      "num_tokens": 54049286.0,
      "step": 24220
    },
    {
      "entropy": 5.286927556991577,
      "epoch": 2.720727803672713,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004270441361967994,
      "loss": 4.9769,
      "mean_token_accuracy": 0.21635160893201827,
      "num_tokens": 54059894.0,
      "step": 24225
    },
    {
      "entropy": 5.37004771232605,
      "epoch": 2.7212893805806706,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004270145389788883,
      "loss": 5.118,
      "mean_token_accuracy": 0.2098048985004425,
      "num_tokens": 54070084.0,
      "step": 24230
    },
    {
      "entropy": 5.318173027038574,
      "epoch": 2.721850957488628,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042698493692077533,
      "loss": 5.0207,
      "mean_token_accuracy": 0.2139551430940628,
      "num_tokens": 54081607.0,
      "step": 24235
    },
    {
      "entropy": 5.348332214355469,
      "epoch": 2.7224125343965855,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004269553300234031,
      "loss": 5.1264,
      "mean_token_accuracy": 0.20698614418506622,
      "num_tokens": 54093364.0,
      "step": 24240
    },
    {
      "entropy": 5.313727188110351,
      "epoch": 2.722974111304543,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00042692571828771423,
      "loss": 4.9712,
      "mean_token_accuracy": 0.2152891904115677,
      "num_tokens": 54105199.0,
      "step": 24245
    },
    {
      "entropy": 5.284808540344239,
      "epoch": 2.723535688212501,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042689610171465147,
      "loss": 4.9481,
      "mean_token_accuracy": 0.22189362347126007,
      "num_tokens": 54116013.0,
      "step": 24250
    },
    {
      "entropy": 5.381634140014649,
      "epoch": 2.7240972651204585,
      "grad_norm": 0.96875,
      "learning_rate": 0.000426866480305158,
      "loss": 5.1021,
      "mean_token_accuracy": 0.2078578069806099,
      "num_tokens": 54128457.0,
      "step": 24255
    },
    {
      "entropy": 5.322844552993774,
      "epoch": 2.7246588420284157,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004268368540601769,
      "loss": 5.0521,
      "mean_token_accuracy": 0.2089722990989685,
      "num_tokens": 54139865.0,
      "step": 24260
    },
    {
      "entropy": 5.447920322418213,
      "epoch": 2.7252204189363733,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004268072229806514,
      "loss": 5.2254,
      "mean_token_accuracy": 0.20354983657598497,
      "num_tokens": 54152264.0,
      "step": 24265
    },
    {
      "entropy": 5.406608772277832,
      "epoch": 2.725781995844331,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004267775870675251,
      "loss": 5.0789,
      "mean_token_accuracy": 0.21504413187503815,
      "num_tokens": 54164339.0,
      "step": 24270
    },
    {
      "entropy": 5.205030632019043,
      "epoch": 2.7263435727522882,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004267479463217415,
      "loss": 4.871,
      "mean_token_accuracy": 0.21823744028806685,
      "num_tokens": 54174041.0,
      "step": 24275
    },
    {
      "entropy": 5.340686464309693,
      "epoch": 2.726905149660246,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042671830074424453,
      "loss": 5.0236,
      "mean_token_accuracy": 0.21367223113775252,
      "num_tokens": 54186457.0,
      "step": 24280
    },
    {
      "entropy": 5.367058324813843,
      "epoch": 2.7274667265682035,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000426688650335978,
      "loss": 5.0509,
      "mean_token_accuracy": 0.21020673364400863,
      "num_tokens": 54197217.0,
      "step": 24285
    },
    {
      "entropy": 5.348050737380982,
      "epoch": 2.728028303476161,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00042665899509788593,
      "loss": 5.0711,
      "mean_token_accuracy": 0.20615342855453492,
      "num_tokens": 54209773.0,
      "step": 24290
    },
    {
      "entropy": 5.32188196182251,
      "epoch": 2.728589880384119,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004266293350309128,
      "loss": 5.0039,
      "mean_token_accuracy": 0.2100941941142082,
      "num_tokens": 54219772.0,
      "step": 24295
    },
    {
      "entropy": 5.28015022277832,
      "epoch": 2.729151457292076,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004265996701360026,
      "loss": 5.0284,
      "mean_token_accuracy": 0.20695804208517074,
      "num_tokens": 54231222.0,
      "step": 24300
    },
    {
      "entropy": 5.365170240402222,
      "epoch": 2.7297130342000337,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042657000041410026,
      "loss": 5.0384,
      "mean_token_accuracy": 0.21554714888334275,
      "num_tokens": 54242113.0,
      "step": 24305
    },
    {
      "entropy": 5.34452109336853,
      "epoch": 2.7302746111079914,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004265403258661503,
      "loss": 5.0715,
      "mean_token_accuracy": 0.2152536168694496,
      "num_tokens": 54252073.0,
      "step": 24310
    },
    {
      "entropy": 5.389655542373657,
      "epoch": 2.7308361880159486,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004265106464930976,
      "loss": 5.12,
      "mean_token_accuracy": 0.2074998825788498,
      "num_tokens": 54262920.0,
      "step": 24315
    },
    {
      "entropy": 5.389509010314941,
      "epoch": 2.7313977649239063,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042648096229588715,
      "loss": 5.0894,
      "mean_token_accuracy": 0.20961420834064484,
      "num_tokens": 54274525.0,
      "step": 24320
    },
    {
      "entropy": 5.3571349620819095,
      "epoch": 2.731959341831864,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042645127327546405,
      "loss": 4.9915,
      "mean_token_accuracy": 0.21635957807302475,
      "num_tokens": 54284312.0,
      "step": 24325
    },
    {
      "entropy": 5.285584831237793,
      "epoch": 2.732520918739821,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042642157943277375,
      "loss": 5.0032,
      "mean_token_accuracy": 0.21069179326295853,
      "num_tokens": 54295735.0,
      "step": 24330
    },
    {
      "entropy": 5.267446279525757,
      "epoch": 2.733082495647779,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004263918807687616,
      "loss": 4.9735,
      "mean_token_accuracy": 0.21473378688097,
      "num_tokens": 54306435.0,
      "step": 24335
    },
    {
      "entropy": 5.287841081619263,
      "epoch": 2.7336440725557365,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004263621772843733,
      "loss": 4.9383,
      "mean_token_accuracy": 0.2174173727631569,
      "num_tokens": 54316454.0,
      "step": 24340
    },
    {
      "entropy": 5.365663719177246,
      "epoch": 2.734205649463694,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004263324689805545,
      "loss": 5.0887,
      "mean_token_accuracy": 0.21185376197099687,
      "num_tokens": 54327270.0,
      "step": 24345
    },
    {
      "entropy": 5.371901273727417,
      "epoch": 2.734767226371652,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004263027558582512,
      "loss": 5.057,
      "mean_token_accuracy": 0.20800520479679108,
      "num_tokens": 54338788.0,
      "step": 24350
    },
    {
      "entropy": 5.469391822814941,
      "epoch": 2.735328803279609,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004262730379184095,
      "loss": 5.1628,
      "mean_token_accuracy": 0.20142074525356293,
      "num_tokens": 54350568.0,
      "step": 24355
    },
    {
      "entropy": 5.3193216800689695,
      "epoch": 2.7358903801875667,
      "grad_norm": 1.0,
      "learning_rate": 0.00042624331516197554,
      "loss": 5.0058,
      "mean_token_accuracy": 0.21614603698253632,
      "num_tokens": 54361318.0,
      "step": 24360
    },
    {
      "entropy": 5.308881664276123,
      "epoch": 2.7364519570955244,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042621358758989584,
      "loss": 4.9646,
      "mean_token_accuracy": 0.21037058681249618,
      "num_tokens": 54371548.0,
      "step": 24365
    },
    {
      "entropy": 5.362273836135865,
      "epoch": 2.7370135340034816,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004261838552031169,
      "loss": 5.1925,
      "mean_token_accuracy": 0.1993140384554863,
      "num_tokens": 54384485.0,
      "step": 24370
    },
    {
      "entropy": 5.434891128540039,
      "epoch": 2.7375751109114392,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042615411800258523,
      "loss": 5.0998,
      "mean_token_accuracy": 0.2053905487060547,
      "num_tokens": 54396155.0,
      "step": 24375
    },
    {
      "entropy": 5.312304925918579,
      "epoch": 2.738136687819397,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042612437598924793,
      "loss": 5.0656,
      "mean_token_accuracy": 0.20907787084579468,
      "num_tokens": 54406222.0,
      "step": 24380
    },
    {
      "entropy": 5.473263502120972,
      "epoch": 2.7386982647273546,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004260946291640519,
      "loss": 5.1703,
      "mean_token_accuracy": 0.2001417189836502,
      "num_tokens": 54419271.0,
      "step": 24385
    },
    {
      "entropy": 5.367524290084839,
      "epoch": 2.739259841635312,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042606487752794417,
      "loss": 5.1498,
      "mean_token_accuracy": 0.20546028912067413,
      "num_tokens": 54430517.0,
      "step": 24390
    },
    {
      "entropy": 5.373941469192505,
      "epoch": 2.7398214185432694,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042603512108187224,
      "loss": 5.0852,
      "mean_token_accuracy": 0.2022665560245514,
      "num_tokens": 54442273.0,
      "step": 24395
    },
    {
      "entropy": 5.30891170501709,
      "epoch": 2.740382995451227,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042600535982678347,
      "loss": 4.9323,
      "mean_token_accuracy": 0.21763429492712022,
      "num_tokens": 54452727.0,
      "step": 24400
    },
    {
      "entropy": 5.40637059211731,
      "epoch": 2.7409445723591848,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042597559376362544,
      "loss": 5.075,
      "mean_token_accuracy": 0.21365509778261185,
      "num_tokens": 54463133.0,
      "step": 24405
    },
    {
      "entropy": 5.383695077896118,
      "epoch": 2.741506149267142,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004259458228933459,
      "loss": 5.1319,
      "mean_token_accuracy": 0.19990667849779128,
      "num_tokens": 54475525.0,
      "step": 24410
    },
    {
      "entropy": 5.365835380554199,
      "epoch": 2.7420677261750996,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042591604721689297,
      "loss": 5.039,
      "mean_token_accuracy": 0.21112280637025832,
      "num_tokens": 54488012.0,
      "step": 24415
    },
    {
      "entropy": 5.267049312591553,
      "epoch": 2.7426293030830573,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004258862667352145,
      "loss": 4.9611,
      "mean_token_accuracy": 0.21549397259950637,
      "num_tokens": 54499485.0,
      "step": 24420
    },
    {
      "entropy": 5.374320840835571,
      "epoch": 2.7431908799910145,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004258564814492587,
      "loss": 5.1289,
      "mean_token_accuracy": 0.20742729902267457,
      "num_tokens": 54509883.0,
      "step": 24425
    },
    {
      "entropy": 5.401013565063477,
      "epoch": 2.743752456898972,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000425826691359974,
      "loss": 5.1439,
      "mean_token_accuracy": 0.2050498455762863,
      "num_tokens": 54521320.0,
      "step": 24430
    },
    {
      "entropy": 5.403294658660888,
      "epoch": 2.74431403380693,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004257968964683091,
      "loss": 5.1096,
      "mean_token_accuracy": 0.20704846680164338,
      "num_tokens": 54532537.0,
      "step": 24435
    },
    {
      "entropy": 5.358677244186401,
      "epoch": 2.7448756107148875,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004257670967752125,
      "loss": 5.0059,
      "mean_token_accuracy": 0.209369757771492,
      "num_tokens": 54542494.0,
      "step": 24440
    },
    {
      "entropy": 5.419699668884277,
      "epoch": 2.745437187622845,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000425737292281633,
      "loss": 5.1521,
      "mean_token_accuracy": 0.2008867785334587,
      "num_tokens": 54553583.0,
      "step": 24445
    },
    {
      "entropy": 5.324308395385742,
      "epoch": 2.7459987645308024,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004257074829885196,
      "loss": 5.0372,
      "mean_token_accuracy": 0.21120425462722778,
      "num_tokens": 54565655.0,
      "step": 24450
    },
    {
      "entropy": 5.260974550247193,
      "epoch": 2.74656034143876,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004256776688968216,
      "loss": 5.0295,
      "mean_token_accuracy": 0.20798840671777724,
      "num_tokens": 54578615.0,
      "step": 24455
    },
    {
      "entropy": 5.398751163482666,
      "epoch": 2.7471219183467177,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00042564785000748814,
      "loss": 5.0994,
      "mean_token_accuracy": 0.20679675191640853,
      "num_tokens": 54590645.0,
      "step": 24460
    },
    {
      "entropy": 5.460895824432373,
      "epoch": 2.747683495254675,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004256180263214687,
      "loss": 5.2047,
      "mean_token_accuracy": 0.20247151404619218,
      "num_tokens": 54602326.0,
      "step": 24465
    },
    {
      "entropy": 5.4429178714752195,
      "epoch": 2.7482450721626326,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00042558819783971295,
      "loss": 5.1784,
      "mean_token_accuracy": 0.2002517655491829,
      "num_tokens": 54614008.0,
      "step": 24470
    },
    {
      "entropy": 5.385930728912354,
      "epoch": 2.7488066490705902,
      "grad_norm": 1.0,
      "learning_rate": 0.0004255583645631705,
      "loss": 5.0091,
      "mean_token_accuracy": 0.2138516828417778,
      "num_tokens": 54625482.0,
      "step": 24475
    },
    {
      "entropy": 5.446760272979736,
      "epoch": 2.749368225978548,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004255285264927913,
      "loss": 5.1086,
      "mean_token_accuracy": 0.20748445242643357,
      "num_tokens": 54636423.0,
      "step": 24480
    },
    {
      "entropy": 5.368801927566528,
      "epoch": 2.7499298028865056,
      "grad_norm": 1.0,
      "learning_rate": 0.00042549868362952556,
      "loss": 5.1079,
      "mean_token_accuracy": 0.20529335290193557,
      "num_tokens": 54648152.0,
      "step": 24485
    },
    {
      "entropy": 5.316090869903564,
      "epoch": 2.750491379794463,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042546883597432327,
      "loss": 5.1007,
      "mean_token_accuracy": 0.21287831515073777,
      "num_tokens": 54661141.0,
      "step": 24490
    },
    {
      "entropy": 5.403056049346924,
      "epoch": 2.7510529567024204,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00042543898352813485,
      "loss": 5.0657,
      "mean_token_accuracy": 0.20128516852855682,
      "num_tokens": 54671180.0,
      "step": 24495
    },
    {
      "entropy": 5.291955661773682,
      "epoch": 2.751614533610378,
      "grad_norm": 1.0625,
      "learning_rate": 0.000425409126291911,
      "loss": 4.9505,
      "mean_token_accuracy": 0.22101855725049974,
      "num_tokens": 54681451.0,
      "step": 24500
    },
    {
      "entropy": 5.306322002410889,
      "epoch": 2.7521761105183353,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004253792642666021,
      "loss": 4.9611,
      "mean_token_accuracy": 0.2160378873348236,
      "num_tokens": 54691760.0,
      "step": 24505
    },
    {
      "entropy": 5.339082670211792,
      "epoch": 2.752737687426293,
      "grad_norm": 0.921875,
      "learning_rate": 0.00042534939745315914,
      "loss": 5.0422,
      "mean_token_accuracy": 0.2048217162489891,
      "num_tokens": 54703761.0,
      "step": 24510
    },
    {
      "entropy": 5.345145606994629,
      "epoch": 2.7532992643342507,
      "grad_norm": 1.015625,
      "learning_rate": 0.000425319525852533,
      "loss": 5.1008,
      "mean_token_accuracy": 0.20395825505256654,
      "num_tokens": 54715495.0,
      "step": 24515
    },
    {
      "entropy": 5.306663179397583,
      "epoch": 2.753860841242208,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004252896494656749,
      "loss": 4.9969,
      "mean_token_accuracy": 0.20854011476039885,
      "num_tokens": 54727061.0,
      "step": 24520
    },
    {
      "entropy": 5.363085412979126,
      "epoch": 2.7544224181501655,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004252597682935362,
      "loss": 4.9744,
      "mean_token_accuracy": 0.20876908600330352,
      "num_tokens": 54738368.0,
      "step": 24525
    },
    {
      "entropy": 5.336516380310059,
      "epoch": 2.754983995058123,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000425229882337068,
      "loss": 5.0786,
      "mean_token_accuracy": 0.20632562935352325,
      "num_tokens": 54749372.0,
      "step": 24530
    },
    {
      "entropy": 5.380286741256714,
      "epoch": 2.755545571966081,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042519999159722216,
      "loss": 5.1804,
      "mean_token_accuracy": 0.20077678859233855,
      "num_tokens": 54761309.0,
      "step": 24535
    },
    {
      "entropy": 5.35131425857544,
      "epoch": 2.7561071488740385,
      "grad_norm": 0.921875,
      "learning_rate": 0.00042517009607495035,
      "loss": 5.0994,
      "mean_token_accuracy": 0.20232448428869249,
      "num_tokens": 54773072.0,
      "step": 24540
    },
    {
      "entropy": 5.3574395179748535,
      "epoch": 2.7566687257819957,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042514019577120444,
      "loss": 5.0274,
      "mean_token_accuracy": 0.20717495679855347,
      "num_tokens": 54783603.0,
      "step": 24545
    },
    {
      "entropy": 5.353979396820068,
      "epoch": 2.7572303026899534,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004251102906869364,
      "loss": 5.0666,
      "mean_token_accuracy": 0.20679473429918288,
      "num_tokens": 54794706.0,
      "step": 24550
    },
    {
      "entropy": 5.405583095550537,
      "epoch": 2.757791879597911,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004250803808230986,
      "loss": 5.1581,
      "mean_token_accuracy": 0.20686376243829727,
      "num_tokens": 54804023.0,
      "step": 24555
    },
    {
      "entropy": 5.2946258068084715,
      "epoch": 2.7583534565058683,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042505046618064314,
      "loss": 5.0333,
      "mean_token_accuracy": 0.21565994918346404,
      "num_tokens": 54813616.0,
      "step": 24560
    },
    {
      "entropy": 5.353783416748047,
      "epoch": 2.758915033413826,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042502054676052273,
      "loss": 5.0293,
      "mean_token_accuracy": 0.20993560552597046,
      "num_tokens": 54825876.0,
      "step": 24565
    },
    {
      "entropy": 5.440632677078247,
      "epoch": 2.7594766103217836,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00042499062256368987,
      "loss": 5.1228,
      "mean_token_accuracy": 0.20666125863790513,
      "num_tokens": 54837207.0,
      "step": 24570
    },
    {
      "entropy": 5.374009037017823,
      "epoch": 2.7600381872297413,
      "grad_norm": 1.0,
      "learning_rate": 0.0004249606935910975,
      "loss": 5.0492,
      "mean_token_accuracy": 0.21378666907548904,
      "num_tokens": 54849155.0,
      "step": 24575
    },
    {
      "entropy": 5.326979875564575,
      "epoch": 2.760599764137699,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004249307598436984,
      "loss": 5.0275,
      "mean_token_accuracy": 0.2138698309659958,
      "num_tokens": 54860345.0,
      "step": 24580
    },
    {
      "entropy": 5.2914246082305905,
      "epoch": 2.761161341045656,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004249008213224458,
      "loss": 5.036,
      "mean_token_accuracy": 0.21170716136693954,
      "num_tokens": 54871676.0,
      "step": 24585
    },
    {
      "entropy": 5.33764910697937,
      "epoch": 2.761722917953614,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042487087802829275,
      "loss": 5.0429,
      "mean_token_accuracy": 0.20833947956562043,
      "num_tokens": 54882352.0,
      "step": 24590
    },
    {
      "entropy": 5.473085117340088,
      "epoch": 2.7622844948615715,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004248409299621931,
      "loss": 5.0677,
      "mean_token_accuracy": 0.21063538640737534,
      "num_tokens": 54893862.0,
      "step": 24595
    },
    {
      "entropy": 5.312916851043701,
      "epoch": 2.7628460717695287,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004248109771251,
      "loss": 5.0282,
      "mean_token_accuracy": 0.21142642199993134,
      "num_tokens": 54904851.0,
      "step": 24600
    },
    {
      "entropy": 5.3573974609375,
      "epoch": 2.7634076486774863,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042478101951796716,
      "loss": 5.1094,
      "mean_token_accuracy": 0.20617587119340897,
      "num_tokens": 54916472.0,
      "step": 24605
    },
    {
      "entropy": 5.368881320953369,
      "epoch": 2.763969225585444,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004247510571417487,
      "loss": 4.9509,
      "mean_token_accuracy": 0.21169255971908568,
      "num_tokens": 54927383.0,
      "step": 24610
    },
    {
      "entropy": 5.377057218551636,
      "epoch": 2.764530802493401,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004247210899973985,
      "loss": 5.0405,
      "mean_token_accuracy": 0.2110537588596344,
      "num_tokens": 54938157.0,
      "step": 24615
    },
    {
      "entropy": 5.367011499404907,
      "epoch": 2.765092379401359,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004246911180858707,
      "loss": 5.0816,
      "mean_token_accuracy": 0.21319883763790132,
      "num_tokens": 54949812.0,
      "step": 24620
    },
    {
      "entropy": 5.44423770904541,
      "epoch": 2.7656539563093165,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004246611414081197,
      "loss": 5.15,
      "mean_token_accuracy": 0.20262496918439865,
      "num_tokens": 54961201.0,
      "step": 24625
    },
    {
      "entropy": 5.3524816036224365,
      "epoch": 2.766215533217274,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042463115996509984,
      "loss": 5.0306,
      "mean_token_accuracy": 0.21450898200273513,
      "num_tokens": 54971071.0,
      "step": 24630
    },
    {
      "entropy": 5.302660417556763,
      "epoch": 2.766777110125232,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004246011737577658,
      "loss": 5.1087,
      "mean_token_accuracy": 0.2114376023411751,
      "num_tokens": 54981294.0,
      "step": 24635
    },
    {
      "entropy": 5.410831642150879,
      "epoch": 2.767338687033189,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004245711827870724,
      "loss": 5.134,
      "mean_token_accuracy": 0.20757617205381393,
      "num_tokens": 54994071.0,
      "step": 24640
    },
    {
      "entropy": 5.456339931488037,
      "epoch": 2.7679002639411467,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042454118705397454,
      "loss": 5.1373,
      "mean_token_accuracy": 0.2105777993798256,
      "num_tokens": 55005087.0,
      "step": 24645
    },
    {
      "entropy": 5.323210334777832,
      "epoch": 2.7684618408491044,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004245111865594273,
      "loss": 4.9649,
      "mean_token_accuracy": 0.2158615455031395,
      "num_tokens": 55016103.0,
      "step": 24650
    },
    {
      "entropy": 5.320210123062134,
      "epoch": 2.7690234177570616,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042448118130438583,
      "loss": 5.1272,
      "mean_token_accuracy": 0.20173705667257308,
      "num_tokens": 55028656.0,
      "step": 24655
    },
    {
      "entropy": 5.388666677474975,
      "epoch": 2.7695849946650193,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042445117128980553,
      "loss": 5.1771,
      "mean_token_accuracy": 0.2067439079284668,
      "num_tokens": 55038957.0,
      "step": 24660
    },
    {
      "entropy": 5.363309335708618,
      "epoch": 2.770146571572977,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042442115651664217,
      "loss": 5.0067,
      "mean_token_accuracy": 0.2164471849799156,
      "num_tokens": 55050407.0,
      "step": 24665
    },
    {
      "entropy": 5.3486724376678465,
      "epoch": 2.7707081484809346,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042439113698585103,
      "loss": 5.0675,
      "mean_token_accuracy": 0.20435316264629363,
      "num_tokens": 55060119.0,
      "step": 24670
    },
    {
      "entropy": 5.346675729751587,
      "epoch": 2.7712697253888923,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004243611126983882,
      "loss": 5.0857,
      "mean_token_accuracy": 0.2106426924467087,
      "num_tokens": 55072059.0,
      "step": 24675
    },
    {
      "entropy": 5.2947979927062985,
      "epoch": 2.7718313022968495,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042433108365520967,
      "loss": 5.0483,
      "mean_token_accuracy": 0.20933853089809418,
      "num_tokens": 55083640.0,
      "step": 24680
    },
    {
      "entropy": 5.377762699127198,
      "epoch": 2.772392879204807,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042430104985727155,
      "loss": 5.0288,
      "mean_token_accuracy": 0.2113464131951332,
      "num_tokens": 55093014.0,
      "step": 24685
    },
    {
      "entropy": 5.335366821289062,
      "epoch": 2.772954456112765,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042427101130553,
      "loss": 5.0378,
      "mean_token_accuracy": 0.21670205593109132,
      "num_tokens": 55102795.0,
      "step": 24690
    },
    {
      "entropy": 5.337248659133911,
      "epoch": 2.773516033020722,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004242409680009416,
      "loss": 5.1157,
      "mean_token_accuracy": 0.2021672636270523,
      "num_tokens": 55114389.0,
      "step": 24695
    },
    {
      "entropy": 5.421837377548218,
      "epoch": 2.7740776099286797,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004242109199444629,
      "loss": 5.1252,
      "mean_token_accuracy": 0.2097255066037178,
      "num_tokens": 55125078.0,
      "step": 24700
    },
    {
      "entropy": 5.374179935455322,
      "epoch": 2.7746391868366374,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004241808671370506,
      "loss": 5.0502,
      "mean_token_accuracy": 0.2078545793890953,
      "num_tokens": 55135151.0,
      "step": 24705
    },
    {
      "entropy": 5.3464844703674315,
      "epoch": 2.7752007637445946,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042415080957966176,
      "loss": 5.0654,
      "mean_token_accuracy": 0.21245260089635848,
      "num_tokens": 55146330.0,
      "step": 24710
    },
    {
      "entropy": 5.308850860595703,
      "epoch": 2.7757623406525522,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004241207472732531,
      "loss": 5.0305,
      "mean_token_accuracy": 0.21287765502929687,
      "num_tokens": 55156636.0,
      "step": 24715
    },
    {
      "entropy": 5.368315601348877,
      "epoch": 2.77632391756051,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004240906802187821,
      "loss": 5.1123,
      "mean_token_accuracy": 0.20546637624502181,
      "num_tokens": 55168761.0,
      "step": 24720
    },
    {
      "entropy": 5.324301671981812,
      "epoch": 2.7768854944684676,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004240606084172061,
      "loss": 5.0055,
      "mean_token_accuracy": 0.21703553944826126,
      "num_tokens": 55180747.0,
      "step": 24725
    },
    {
      "entropy": 5.4008362770080565,
      "epoch": 2.777447071376425,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042403053186948237,
      "loss": 5.1441,
      "mean_token_accuracy": 0.20204276889562606,
      "num_tokens": 55191645.0,
      "step": 24730
    },
    {
      "entropy": 5.351100730895996,
      "epoch": 2.7780086482843824,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004240004505765688,
      "loss": 5.0386,
      "mean_token_accuracy": 0.20273702293634416,
      "num_tokens": 55203623.0,
      "step": 24735
    },
    {
      "entropy": 5.439945363998413,
      "epoch": 2.77857022519234,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004239703645394231,
      "loss": 5.1679,
      "mean_token_accuracy": 0.20171007215976716,
      "num_tokens": 55214635.0,
      "step": 24740
    },
    {
      "entropy": 5.300849390029907,
      "epoch": 2.7791318021002978,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004239402737590032,
      "loss": 5.0081,
      "mean_token_accuracy": 0.20932772606611252,
      "num_tokens": 55225524.0,
      "step": 24745
    },
    {
      "entropy": 5.246091222763061,
      "epoch": 2.779693379008255,
      "grad_norm": 1.0,
      "learning_rate": 0.00042391017823626716,
      "loss": 5.0073,
      "mean_token_accuracy": 0.21637169718742372,
      "num_tokens": 55238360.0,
      "step": 24750
    },
    {
      "entropy": 5.458284950256347,
      "epoch": 2.7802549559162126,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004238800779721734,
      "loss": 5.1534,
      "mean_token_accuracy": 0.19897132068872453,
      "num_tokens": 55248524.0,
      "step": 24755
    },
    {
      "entropy": 5.416894197463989,
      "epoch": 2.7808165328241703,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00042384997296768007,
      "loss": 5.1663,
      "mean_token_accuracy": 0.2084077924489975,
      "num_tokens": 55260283.0,
      "step": 24760
    },
    {
      "entropy": 5.445787334442139,
      "epoch": 2.781378109732128,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000423819863223746,
      "loss": 5.0656,
      "mean_token_accuracy": 0.20748061686754227,
      "num_tokens": 55271928.0,
      "step": 24765
    },
    {
      "entropy": 5.369857740402222,
      "epoch": 2.7819396866400856,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042378974874132967,
      "loss": 5.0874,
      "mean_token_accuracy": 0.2079591006040573,
      "num_tokens": 55283015.0,
      "step": 24770
    },
    {
      "entropy": 5.315274620056153,
      "epoch": 2.782501263548043,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004237596295213901,
      "loss": 4.9499,
      "mean_token_accuracy": 0.21328247785568238,
      "num_tokens": 55293918.0,
      "step": 24775
    },
    {
      "entropy": 5.243858289718628,
      "epoch": 2.7830628404560005,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004237295055648862,
      "loss": 4.9995,
      "mean_token_accuracy": 0.20650495886802672,
      "num_tokens": 55305808.0,
      "step": 24780
    },
    {
      "entropy": 5.305925178527832,
      "epoch": 2.783624417363958,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042369937687277716,
      "loss": 5.099,
      "mean_token_accuracy": 0.20363037139177323,
      "num_tokens": 55318128.0,
      "step": 24785
    },
    {
      "entropy": 5.406986713409424,
      "epoch": 2.7841859942719154,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004236692434460223,
      "loss": 5.1127,
      "mean_token_accuracy": 0.2079336479306221,
      "num_tokens": 55329934.0,
      "step": 24790
    },
    {
      "entropy": 5.4955394744873045,
      "epoch": 2.784747571179873,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004236391052855811,
      "loss": 5.2441,
      "mean_token_accuracy": 0.20410149693489074,
      "num_tokens": 55340971.0,
      "step": 24795
    },
    {
      "entropy": 5.380181789398193,
      "epoch": 2.7853091480878307,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042360896239241306,
      "loss": 5.0997,
      "mean_token_accuracy": 0.20555775463581086,
      "num_tokens": 55352243.0,
      "step": 24800
    },
    {
      "entropy": 5.430887365341187,
      "epoch": 2.785870724995788,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042357881476747807,
      "loss": 5.0089,
      "mean_token_accuracy": 0.21246019154787063,
      "num_tokens": 55363694.0,
      "step": 24805
    },
    {
      "entropy": 5.361173105239868,
      "epoch": 2.7864323019037456,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000423548662411736,
      "loss": 5.2006,
      "mean_token_accuracy": 0.2064851015806198,
      "num_tokens": 55375720.0,
      "step": 24810
    },
    {
      "entropy": 5.386740779876709,
      "epoch": 2.7869938788117032,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004235185053261469,
      "loss": 5.0234,
      "mean_token_accuracy": 0.21174682527780533,
      "num_tokens": 55387493.0,
      "step": 24815
    },
    {
      "entropy": 5.445042943954467,
      "epoch": 2.787555455719661,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000423488343511671,
      "loss": 5.2136,
      "mean_token_accuracy": 0.20129734128713608,
      "num_tokens": 55399468.0,
      "step": 24820
    },
    {
      "entropy": 5.387980222702026,
      "epoch": 2.7881170326276186,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004234581769692686,
      "loss": 5.053,
      "mean_token_accuracy": 0.2045532464981079,
      "num_tokens": 55411526.0,
      "step": 24825
    },
    {
      "entropy": 5.359442949295044,
      "epoch": 2.788678609535576,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004234280056999003,
      "loss": 5.01,
      "mean_token_accuracy": 0.208513280749321,
      "num_tokens": 55422550.0,
      "step": 24830
    },
    {
      "entropy": 5.302337789535523,
      "epoch": 2.7892401864435334,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004233978297045267,
      "loss": 4.9844,
      "mean_token_accuracy": 0.21497288346290588,
      "num_tokens": 55433433.0,
      "step": 24835
    },
    {
      "entropy": 5.30279426574707,
      "epoch": 2.789801763351491,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004233676489841087,
      "loss": 5.027,
      "mean_token_accuracy": 0.2145466312766075,
      "num_tokens": 55444225.0,
      "step": 24840
    },
    {
      "entropy": 5.439083099365234,
      "epoch": 2.7903633402594483,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042333746353960723,
      "loss": 5.1755,
      "mean_token_accuracy": 0.2042992264032364,
      "num_tokens": 55456117.0,
      "step": 24845
    },
    {
      "entropy": 5.328064203262329,
      "epoch": 2.790924917167406,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004233072733719833,
      "loss": 5.0128,
      "mean_token_accuracy": 0.214436137676239,
      "num_tokens": 55467267.0,
      "step": 24850
    },
    {
      "entropy": 5.382145595550537,
      "epoch": 2.7914864940753636,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004232770784821984,
      "loss": 5.0774,
      "mean_token_accuracy": 0.2094866454601288,
      "num_tokens": 55478520.0,
      "step": 24855
    },
    {
      "entropy": 5.423856735229492,
      "epoch": 2.7920480709833213,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042324687887121366,
      "loss": 5.1504,
      "mean_token_accuracy": 0.2011175960302353,
      "num_tokens": 55489704.0,
      "step": 24860
    },
    {
      "entropy": 5.3674393653869625,
      "epoch": 2.792609647891279,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004232166745399909,
      "loss": 5.0098,
      "mean_token_accuracy": 0.21528191566467286,
      "num_tokens": 55500942.0,
      "step": 24865
    },
    {
      "entropy": 5.377057695388794,
      "epoch": 2.793171224799236,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042318646548949177,
      "loss": 5.2062,
      "mean_token_accuracy": 0.1931743636727333,
      "num_tokens": 55512691.0,
      "step": 24870
    },
    {
      "entropy": 5.345473194122315,
      "epoch": 2.793732801707194,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042315625172067794,
      "loss": 4.9823,
      "mean_token_accuracy": 0.21734073162078857,
      "num_tokens": 55523827.0,
      "step": 24875
    },
    {
      "entropy": 5.453054523468017,
      "epoch": 2.7942943786151515,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042312603323451173,
      "loss": 5.1594,
      "mean_token_accuracy": 0.20538645088672638,
      "num_tokens": 55534674.0,
      "step": 24880
    },
    {
      "entropy": 5.341828489303589,
      "epoch": 2.7948559555231087,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004230958100319551,
      "loss": 5.0038,
      "mean_token_accuracy": 0.21375340223312378,
      "num_tokens": 55545504.0,
      "step": 24885
    },
    {
      "entropy": 5.352264833450318,
      "epoch": 2.7954175324310664,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004230655821139705,
      "loss": 5.1724,
      "mean_token_accuracy": 0.19551487118005753,
      "num_tokens": 55557818.0,
      "step": 24890
    },
    {
      "entropy": 5.390302562713623,
      "epoch": 2.795979109339024,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004230353494815203,
      "loss": 5.0598,
      "mean_token_accuracy": 0.2143446087837219,
      "num_tokens": 55568143.0,
      "step": 24895
    },
    {
      "entropy": 5.463162136077881,
      "epoch": 2.7965406862469813,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004230051121355672,
      "loss": 5.143,
      "mean_token_accuracy": 0.20322568118572235,
      "num_tokens": 55579358.0,
      "step": 24900
    },
    {
      "entropy": 5.434896421432495,
      "epoch": 2.797102263154939,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042297487007707387,
      "loss": 5.2161,
      "mean_token_accuracy": 0.20239448994398118,
      "num_tokens": 55590430.0,
      "step": 24905
    },
    {
      "entropy": 5.401182174682617,
      "epoch": 2.7976638400628966,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042294462330700336,
      "loss": 5.1182,
      "mean_token_accuracy": 0.21561560183763503,
      "num_tokens": 55602377.0,
      "step": 24910
    },
    {
      "entropy": 5.343977785110473,
      "epoch": 2.7982254169708543,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004229143718263186,
      "loss": 5.0305,
      "mean_token_accuracy": 0.20903510153293609,
      "num_tokens": 55613261.0,
      "step": 24915
    },
    {
      "entropy": 5.234525728225708,
      "epoch": 2.798786993878812,
      "grad_norm": 1.125,
      "learning_rate": 0.0004228841156359829,
      "loss": 4.9916,
      "mean_token_accuracy": 0.20785183310508729,
      "num_tokens": 55624514.0,
      "step": 24920
    },
    {
      "entropy": 5.405308675765991,
      "epoch": 2.799348570786769,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004228538547369596,
      "loss": 5.0699,
      "mean_token_accuracy": 0.2057131424546242,
      "num_tokens": 55635899.0,
      "step": 24925
    },
    {
      "entropy": 5.386952257156372,
      "epoch": 2.799910147694727,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042282358913021224,
      "loss": 5.0562,
      "mean_token_accuracy": 0.20809140354394912,
      "num_tokens": 55646807.0,
      "step": 24930
    },
    {
      "entropy": 5.422532749176026,
      "epoch": 2.8004717246026845,
      "grad_norm": 1.0,
      "learning_rate": 0.00042279331881670446,
      "loss": 5.0816,
      "mean_token_accuracy": 0.21141209900379182,
      "num_tokens": 55658937.0,
      "step": 24935
    },
    {
      "entropy": 5.287449073791504,
      "epoch": 2.8010333015106417,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004227630437974,
      "loss": 5.0173,
      "mean_token_accuracy": 0.21649404764175414,
      "num_tokens": 55670216.0,
      "step": 24940
    },
    {
      "entropy": 5.401245403289795,
      "epoch": 2.8015948784185993,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00042273276407326304,
      "loss": 5.0885,
      "mean_token_accuracy": 0.20967592149972916,
      "num_tokens": 55682743.0,
      "step": 24945
    },
    {
      "entropy": 5.345742702484131,
      "epoch": 2.802156455326557,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042270247964525755,
      "loss": 5.0574,
      "mean_token_accuracy": 0.2032907336950302,
      "num_tokens": 55694338.0,
      "step": 24950
    },
    {
      "entropy": 5.371348190307617,
      "epoch": 2.8027180322345147,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004226721905143477,
      "loss": 5.075,
      "mean_token_accuracy": 0.20659867525100709,
      "num_tokens": 55704224.0,
      "step": 24955
    },
    {
      "entropy": 5.303373003005982,
      "epoch": 2.8032796091424723,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042264189668149817,
      "loss": 5.0225,
      "mean_token_accuracy": 0.21028662770986556,
      "num_tokens": 55714901.0,
      "step": 24960
    },
    {
      "entropy": 5.43898491859436,
      "epoch": 2.8038411860504295,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004226115981476733,
      "loss": 5.1849,
      "mean_token_accuracy": 0.20087548792362214,
      "num_tokens": 55725020.0,
      "step": 24965
    },
    {
      "entropy": 5.307449150085449,
      "epoch": 2.804402762958387,
      "grad_norm": 1.09375,
      "learning_rate": 0.000422581294913838,
      "loss": 4.9596,
      "mean_token_accuracy": 0.21484794914722444,
      "num_tokens": 55735643.0,
      "step": 24970
    },
    {
      "entropy": 5.314781951904297,
      "epoch": 2.804964339866345,
      "grad_norm": 1.03125,
      "learning_rate": 0.000422550986980957,
      "loss": 5.0112,
      "mean_token_accuracy": 0.20519519299268724,
      "num_tokens": 55745778.0,
      "step": 24975
    },
    {
      "entropy": 5.266573286056518,
      "epoch": 2.805525916774302,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042252067434999524,
      "loss": 5.0496,
      "mean_token_accuracy": 0.20236002057790756,
      "num_tokens": 55757055.0,
      "step": 24980
    },
    {
      "entropy": 5.345400619506836,
      "epoch": 2.8060874936822597,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004224903570219181,
      "loss": 5.0988,
      "mean_token_accuracy": 0.20863455832004546,
      "num_tokens": 55768193.0,
      "step": 24985
    },
    {
      "entropy": 5.3019286632537845,
      "epoch": 2.8066490705902174,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042246003499769077,
      "loss": 4.9842,
      "mean_token_accuracy": 0.21205989122390748,
      "num_tokens": 55779382.0,
      "step": 24990
    },
    {
      "entropy": 5.328254461288452,
      "epoch": 2.8072106474981746,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042242970827827874,
      "loss": 4.9653,
      "mean_token_accuracy": 0.21736496537923813,
      "num_tokens": 55790933.0,
      "step": 24995
    },
    {
      "entropy": 5.266987037658692,
      "epoch": 2.8077722244061323,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004223993768646476,
      "loss": 4.9396,
      "mean_token_accuracy": 0.2188433215022087,
      "num_tokens": 55803101.0,
      "step": 25000
    },
    {
      "entropy": 5.242200469970703,
      "epoch": 2.80833380131409,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004223690407577631,
      "loss": 4.9661,
      "mean_token_accuracy": 0.21989805698394777,
      "num_tokens": 55814651.0,
      "step": 25005
    },
    {
      "entropy": 5.366455745697022,
      "epoch": 2.8088953782220476,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042233869995859127,
      "loss": 5.066,
      "mean_token_accuracy": 0.2107747822999954,
      "num_tokens": 55825696.0,
      "step": 25010
    },
    {
      "entropy": 5.318101835250855,
      "epoch": 2.8094569551300053,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042230835446809803,
      "loss": 5.0357,
      "mean_token_accuracy": 0.21161027997732162,
      "num_tokens": 55836679.0,
      "step": 25015
    },
    {
      "entropy": 5.351746988296509,
      "epoch": 2.8100185320379625,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004222780042872496,
      "loss": 5.1022,
      "mean_token_accuracy": 0.20278820395469666,
      "num_tokens": 55848013.0,
      "step": 25020
    },
    {
      "entropy": 5.259368896484375,
      "epoch": 2.81058010894592,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004222476494170125,
      "loss": 4.9228,
      "mean_token_accuracy": 0.21595935076475142,
      "num_tokens": 55859310.0,
      "step": 25025
    },
    {
      "entropy": 5.38680739402771,
      "epoch": 2.811141685853878,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004222172898583531,
      "loss": 5.1203,
      "mean_token_accuracy": 0.19720055162906647,
      "num_tokens": 55870484.0,
      "step": 25030
    },
    {
      "entropy": 5.423621749877929,
      "epoch": 2.811703262761835,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004221869256122381,
      "loss": 5.1082,
      "mean_token_accuracy": 0.20472100824117662,
      "num_tokens": 55881279.0,
      "step": 25035
    },
    {
      "entropy": 5.312305593490601,
      "epoch": 2.8122648396697927,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004221565566796343,
      "loss": 4.972,
      "mean_token_accuracy": 0.21027495563030243,
      "num_tokens": 55893029.0,
      "step": 25040
    },
    {
      "entropy": 5.215112924575806,
      "epoch": 2.8128264165777503,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004221261830615086,
      "loss": 4.9318,
      "mean_token_accuracy": 0.22069261372089385,
      "num_tokens": 55904171.0,
      "step": 25045
    },
    {
      "entropy": 5.341006088256836,
      "epoch": 2.813387993485708,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004220958047588282,
      "loss": 5.0172,
      "mean_token_accuracy": 0.2147486090660095,
      "num_tokens": 55916121.0,
      "step": 25050
    },
    {
      "entropy": 5.3299308776855465,
      "epoch": 2.8139495703936657,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004220654217725603,
      "loss": 5.0765,
      "mean_token_accuracy": 0.2038099706172943,
      "num_tokens": 55925982.0,
      "step": 25055
    },
    {
      "entropy": 5.346964740753174,
      "epoch": 2.814511147301623,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042203503410367244,
      "loss": 5.0368,
      "mean_token_accuracy": 0.21243065297603608,
      "num_tokens": 55935768.0,
      "step": 25060
    },
    {
      "entropy": 5.391199827194214,
      "epoch": 2.8150727242095805,
      "grad_norm": 0.984375,
      "learning_rate": 0.000422004641753132,
      "loss": 5.0928,
      "mean_token_accuracy": 0.21418495923280717,
      "num_tokens": 55946742.0,
      "step": 25065
    },
    {
      "entropy": 5.337608671188354,
      "epoch": 2.815634301117538,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004219742447219068,
      "loss": 5.0046,
      "mean_token_accuracy": 0.2157620370388031,
      "num_tokens": 55957133.0,
      "step": 25070
    },
    {
      "entropy": 5.323590421676636,
      "epoch": 2.8161958780254954,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042194384301096457,
      "loss": 5.0426,
      "mean_token_accuracy": 0.2122401550412178,
      "num_tokens": 55967179.0,
      "step": 25075
    },
    {
      "entropy": 5.28327169418335,
      "epoch": 2.816757454933453,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042191343662127337,
      "loss": 4.9449,
      "mean_token_accuracy": 0.22082908898591996,
      "num_tokens": 55977713.0,
      "step": 25080
    },
    {
      "entropy": 5.397963619232177,
      "epoch": 2.8173190318414107,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042188302555380144,
      "loss": 5.0864,
      "mean_token_accuracy": 0.20528085231781007,
      "num_tokens": 55988650.0,
      "step": 25085
    },
    {
      "entropy": 5.307288980484008,
      "epoch": 2.817880608749368,
      "grad_norm": 1.03125,
      "learning_rate": 0.000421852609809517,
      "loss": 4.9561,
      "mean_token_accuracy": 0.20778716504573821,
      "num_tokens": 55999745.0,
      "step": 25090
    },
    {
      "entropy": 5.25607123374939,
      "epoch": 2.8184421856573256,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004218221893893884,
      "loss": 4.9865,
      "mean_token_accuracy": 0.21312991678714752,
      "num_tokens": 56011171.0,
      "step": 25095
    },
    {
      "entropy": 5.271456527709961,
      "epoch": 2.8190037625652833,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042179176429438454,
      "loss": 5.0197,
      "mean_token_accuracy": 0.21852748692035676,
      "num_tokens": 56022121.0,
      "step": 25100
    },
    {
      "entropy": 5.314577436447143,
      "epoch": 2.819565339473241,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004217613345254739,
      "loss": 5.0329,
      "mean_token_accuracy": 0.21862117648124696,
      "num_tokens": 56033769.0,
      "step": 25105
    },
    {
      "entropy": 5.353180980682373,
      "epoch": 2.8201269163811986,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042173090008362535,
      "loss": 4.9805,
      "mean_token_accuracy": 0.21461938619613646,
      "num_tokens": 56044691.0,
      "step": 25110
    },
    {
      "entropy": 5.257045936584473,
      "epoch": 2.820688493289156,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042170046096980807,
      "loss": 4.9666,
      "mean_token_accuracy": 0.2167877346277237,
      "num_tokens": 56055639.0,
      "step": 25115
    },
    {
      "entropy": 5.420208978652954,
      "epoch": 2.8212500701971135,
      "grad_norm": 1.0,
      "learning_rate": 0.00042167001718499124,
      "loss": 5.0944,
      "mean_token_accuracy": 0.20938827991485595,
      "num_tokens": 56066579.0,
      "step": 25120
    },
    {
      "entropy": 5.401353216171264,
      "epoch": 2.821811647105071,
      "grad_norm": 1.0,
      "learning_rate": 0.0004216395687301442,
      "loss": 5.0664,
      "mean_token_accuracy": 0.20054652392864228,
      "num_tokens": 56076871.0,
      "step": 25125
    },
    {
      "entropy": 5.38863353729248,
      "epoch": 2.8223732240130284,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004216091156062364,
      "loss": 5.1543,
      "mean_token_accuracy": 0.20848173648118973,
      "num_tokens": 56090022.0,
      "step": 25130
    },
    {
      "entropy": 5.205065441131592,
      "epoch": 2.822934800920986,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004215786578142375,
      "loss": 4.8418,
      "mean_token_accuracy": 0.2214314043521881,
      "num_tokens": 56100545.0,
      "step": 25135
    },
    {
      "entropy": 5.330967617034912,
      "epoch": 2.8234963778289437,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042154819535511733,
      "loss": 5.1323,
      "mean_token_accuracy": 0.2018802911043167,
      "num_tokens": 56112059.0,
      "step": 25140
    },
    {
      "entropy": 5.42360143661499,
      "epoch": 2.8240579547369014,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042151772822984567,
      "loss": 5.077,
      "mean_token_accuracy": 0.2146626204252243,
      "num_tokens": 56123162.0,
      "step": 25145
    },
    {
      "entropy": 5.494584751129151,
      "epoch": 2.824619531644859,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004214872564393928,
      "loss": 5.1859,
      "mean_token_accuracy": 0.20524044781923295,
      "num_tokens": 56136008.0,
      "step": 25150
    },
    {
      "entropy": 5.306376123428345,
      "epoch": 2.8251811085528162,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004214567799847289,
      "loss": 5.0176,
      "mean_token_accuracy": 0.21305887401103973,
      "num_tokens": 56147169.0,
      "step": 25155
    },
    {
      "entropy": 5.323599529266358,
      "epoch": 2.825742685460774,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004214262988668242,
      "loss": 5.0052,
      "mean_token_accuracy": 0.20680559873580934,
      "num_tokens": 56157644.0,
      "step": 25160
    },
    {
      "entropy": 5.415505886077881,
      "epoch": 2.8263042623687316,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004213958130866494,
      "loss": 5.1068,
      "mean_token_accuracy": 0.20488515198230745,
      "num_tokens": 56168248.0,
      "step": 25165
    },
    {
      "entropy": 5.423869705200195,
      "epoch": 2.8268658392766888,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00042136532264517527,
      "loss": 5.1374,
      "mean_token_accuracy": 0.20479087233543397,
      "num_tokens": 56179140.0,
      "step": 25170
    },
    {
      "entropy": 5.394404077529908,
      "epoch": 2.8274274161846464,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042133482754337236,
      "loss": 5.0177,
      "mean_token_accuracy": 0.20548575967550278,
      "num_tokens": 56190184.0,
      "step": 25175
    },
    {
      "entropy": 5.3638897895812985,
      "epoch": 2.827988993092604,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042130432778221186,
      "loss": 5.1097,
      "mean_token_accuracy": 0.21577054113149643,
      "num_tokens": 56201619.0,
      "step": 25180
    },
    {
      "entropy": 5.352601623535156,
      "epoch": 2.8285505700005613,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042127382336266467,
      "loss": 5.0105,
      "mean_token_accuracy": 0.20491111278533936,
      "num_tokens": 56212339.0,
      "step": 25185
    },
    {
      "entropy": 5.4060015201568605,
      "epoch": 2.829112146908519,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004212433142857024,
      "loss": 5.0526,
      "mean_token_accuracy": 0.21625359505414962,
      "num_tokens": 56224734.0,
      "step": 25190
    },
    {
      "entropy": 5.363369178771973,
      "epoch": 2.8296737238164766,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00042121280055229624,
      "loss": 4.9957,
      "mean_token_accuracy": 0.2130989298224449,
      "num_tokens": 56235594.0,
      "step": 25195
    },
    {
      "entropy": 5.309859848022461,
      "epoch": 2.8302353007244343,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004211822821634177,
      "loss": 4.9846,
      "mean_token_accuracy": 0.21481854617595672,
      "num_tokens": 56246932.0,
      "step": 25200
    },
    {
      "entropy": 5.279452705383301,
      "epoch": 2.830796877632392,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004211517591200387,
      "loss": 5.1075,
      "mean_token_accuracy": 0.21301871687173843,
      "num_tokens": 56258217.0,
      "step": 25205
    },
    {
      "entropy": 5.369362115859985,
      "epoch": 2.831358454540349,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004211212314231309,
      "loss": 5.0853,
      "mean_token_accuracy": 0.2104479432106018,
      "num_tokens": 56269226.0,
      "step": 25210
    },
    {
      "entropy": 5.352803087234497,
      "epoch": 2.831920031448307,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042109069907366646,
      "loss": 5.042,
      "mean_token_accuracy": 0.21089656949043273,
      "num_tokens": 56280411.0,
      "step": 25215
    },
    {
      "entropy": 5.403091335296631,
      "epoch": 2.8324816083562645,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042106016207261754,
      "loss": 5.1409,
      "mean_token_accuracy": 0.20433432459831238,
      "num_tokens": 56292048.0,
      "step": 25220
    },
    {
      "entropy": 5.343544816970825,
      "epoch": 2.8330431852642217,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042102962042095643,
      "loss": 5.0495,
      "mean_token_accuracy": 0.21246617138385773,
      "num_tokens": 56302470.0,
      "step": 25225
    },
    {
      "entropy": 5.35925784111023,
      "epoch": 2.8336047621721794,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042099907411965553,
      "loss": 5.0049,
      "mean_token_accuracy": 0.21154739260673522,
      "num_tokens": 56313871.0,
      "step": 25230
    },
    {
      "entropy": 5.248078060150147,
      "epoch": 2.834166339080137,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042096852316968756,
      "loss": 4.9984,
      "mean_token_accuracy": 0.21061795055866242,
      "num_tokens": 56325455.0,
      "step": 25235
    },
    {
      "entropy": 5.395863342285156,
      "epoch": 2.8347279159880947,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042093796757202513,
      "loss": 5.1382,
      "mean_token_accuracy": 0.2077445611357689,
      "num_tokens": 56336187.0,
      "step": 25240
    },
    {
      "entropy": 5.482807540893555,
      "epoch": 2.8352894928960524,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004209074073276412,
      "loss": 5.1521,
      "mean_token_accuracy": 0.20215410590171815,
      "num_tokens": 56348175.0,
      "step": 25245
    },
    {
      "entropy": 5.3528890132904055,
      "epoch": 2.8358510698040096,
      "grad_norm": 0.875,
      "learning_rate": 0.0004208768424375089,
      "loss": 5.0604,
      "mean_token_accuracy": 0.203012815117836,
      "num_tokens": 56359668.0,
      "step": 25250
    },
    {
      "entropy": 5.350370693206787,
      "epoch": 2.8364126467119672,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00042084627290260124,
      "loss": 5.0305,
      "mean_token_accuracy": 0.2135455086827278,
      "num_tokens": 56371160.0,
      "step": 25255
    },
    {
      "entropy": 5.27198805809021,
      "epoch": 2.836974223619925,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004208156987238918,
      "loss": 4.9334,
      "mean_token_accuracy": 0.2130702868103981,
      "num_tokens": 56382873.0,
      "step": 25260
    },
    {
      "entropy": 5.3433356285095215,
      "epoch": 2.837535800527882,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004207851199023539,
      "loss": 5.0191,
      "mean_token_accuracy": 0.214192795753479,
      "num_tokens": 56393562.0,
      "step": 25265
    },
    {
      "entropy": 5.410723733901977,
      "epoch": 2.83809737743584,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042075453643896125,
      "loss": 5.1523,
      "mean_token_accuracy": 0.2107653483748436,
      "num_tokens": 56405789.0,
      "step": 25270
    },
    {
      "entropy": 5.308949422836304,
      "epoch": 2.8386589543437974,
      "grad_norm": 1.0,
      "learning_rate": 0.00042072394833468765,
      "loss": 4.9918,
      "mean_token_accuracy": 0.21686058789491652,
      "num_tokens": 56417621.0,
      "step": 25275
    },
    {
      "entropy": 5.334698009490967,
      "epoch": 2.8392205312517547,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000420693355590507,
      "loss": 5.0881,
      "mean_token_accuracy": 0.20522015541791916,
      "num_tokens": 56430167.0,
      "step": 25280
    },
    {
      "entropy": 5.383414554595947,
      "epoch": 2.8397821081597123,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042066275820739333,
      "loss": 5.1144,
      "mean_token_accuracy": 0.20092668533325195,
      "num_tokens": 56441853.0,
      "step": 25285
    },
    {
      "entropy": 5.301263666152954,
      "epoch": 2.84034368506767,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000420632156186321,
      "loss": 4.9878,
      "mean_token_accuracy": 0.21107610911130906,
      "num_tokens": 56454504.0,
      "step": 25290
    },
    {
      "entropy": 5.353687572479248,
      "epoch": 2.8409052619756276,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004206015495282644,
      "loss": 5.0605,
      "mean_token_accuracy": 0.21707905530929567,
      "num_tokens": 56465503.0,
      "step": 25295
    },
    {
      "entropy": 5.3594255447387695,
      "epoch": 2.8414668388835853,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042057093823419795,
      "loss": 5.1119,
      "mean_token_accuracy": 0.2080554723739624,
      "num_tokens": 56475310.0,
      "step": 25300
    },
    {
      "entropy": 5.356772232055664,
      "epoch": 2.8420284157915425,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042054032230509636,
      "loss": 5.0181,
      "mean_token_accuracy": 0.2146414563059807,
      "num_tokens": 56486575.0,
      "step": 25305
    },
    {
      "entropy": 5.4506919384002686,
      "epoch": 2.8425899926995,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042050970174193437,
      "loss": 5.1226,
      "mean_token_accuracy": 0.21023194193840028,
      "num_tokens": 56496542.0,
      "step": 25310
    },
    {
      "entropy": 5.441633939743042,
      "epoch": 2.843151569607458,
      "grad_norm": 1.0,
      "learning_rate": 0.00042047907654568724,
      "loss": 5.0667,
      "mean_token_accuracy": 0.21179478466510773,
      "num_tokens": 56507124.0,
      "step": 25315
    },
    {
      "entropy": 5.336295652389526,
      "epoch": 2.843713146515415,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004204484467173297,
      "loss": 4.9789,
      "mean_token_accuracy": 0.2125703975558281,
      "num_tokens": 56518635.0,
      "step": 25320
    },
    {
      "entropy": 5.26924238204956,
      "epoch": 2.8442747234233727,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004204178122578373,
      "loss": 5.0085,
      "mean_token_accuracy": 0.21786385625600815,
      "num_tokens": 56529499.0,
      "step": 25325
    },
    {
      "entropy": 5.384020471572876,
      "epoch": 2.8448363003313304,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042038717316818534,
      "loss": 5.0809,
      "mean_token_accuracy": 0.21286399513483048,
      "num_tokens": 56542432.0,
      "step": 25330
    },
    {
      "entropy": 5.335423851013184,
      "epoch": 2.845397877239288,
      "grad_norm": 1.0,
      "learning_rate": 0.0004203565294493494,
      "loss": 5.082,
      "mean_token_accuracy": 0.20494160056114197,
      "num_tokens": 56552783.0,
      "step": 25335
    },
    {
      "entropy": 5.298158645629883,
      "epoch": 2.8459594541472457,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042032588110230525,
      "loss": 5.0177,
      "mean_token_accuracy": 0.2138516217470169,
      "num_tokens": 56563654.0,
      "step": 25340
    },
    {
      "entropy": 5.3878586769104,
      "epoch": 2.846521031055203,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004202952281280286,
      "loss": 5.0845,
      "mean_token_accuracy": 0.2110460579395294,
      "num_tokens": 56575541.0,
      "step": 25345
    },
    {
      "entropy": 5.463498735427857,
      "epoch": 2.8470826079631606,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004202645705274956,
      "loss": 5.182,
      "mean_token_accuracy": 0.20303021371364594,
      "num_tokens": 56588397.0,
      "step": 25350
    },
    {
      "entropy": 5.4037940979003904,
      "epoch": 2.8476441848711183,
      "grad_norm": 0.9375,
      "learning_rate": 0.00042023390830168224,
      "loss": 5.0778,
      "mean_token_accuracy": 0.20689986497163773,
      "num_tokens": 56599858.0,
      "step": 25355
    },
    {
      "entropy": 5.3700824737548825,
      "epoch": 2.8482057617790755,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000420203241451565,
      "loss": 5.0168,
      "mean_token_accuracy": 0.21812091022729874,
      "num_tokens": 56610972.0,
      "step": 25360
    },
    {
      "entropy": 5.295264673233032,
      "epoch": 2.848767338687033,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042017256997812016,
      "loss": 4.9612,
      "mean_token_accuracy": 0.21859341412782668,
      "num_tokens": 56621619.0,
      "step": 25365
    },
    {
      "entropy": 5.287691116333008,
      "epoch": 2.849328915594991,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042014189388232446,
      "loss": 5.069,
      "mean_token_accuracy": 0.20758555233478546,
      "num_tokens": 56632955.0,
      "step": 25370
    },
    {
      "entropy": 5.405336856842041,
      "epoch": 2.849890492502948,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004201112131651545,
      "loss": 5.1206,
      "mean_token_accuracy": 0.2111274316906929,
      "num_tokens": 56645896.0,
      "step": 25375
    },
    {
      "entropy": 5.296153116226196,
      "epoch": 2.8504520694109057,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004200805278275873,
      "loss": 4.9819,
      "mean_token_accuracy": 0.21018437594175338,
      "num_tokens": 56656974.0,
      "step": 25380
    },
    {
      "entropy": 5.423883390426636,
      "epoch": 2.8510136463188633,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004200498378705997,
      "loss": 5.2073,
      "mean_token_accuracy": 0.20477059334516526,
      "num_tokens": 56669251.0,
      "step": 25385
    },
    {
      "entropy": 5.397593927383423,
      "epoch": 2.851575223226821,
      "grad_norm": 1.0,
      "learning_rate": 0.0004200191432951692,
      "loss": 4.9563,
      "mean_token_accuracy": 0.2223512664437294,
      "num_tokens": 56679526.0,
      "step": 25390
    },
    {
      "entropy": 5.315218496322632,
      "epoch": 2.8521368001347787,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004199884441022728,
      "loss": 5.0141,
      "mean_token_accuracy": 0.21128101646900177,
      "num_tokens": 56690276.0,
      "step": 25395
    },
    {
      "entropy": 5.272818994522095,
      "epoch": 2.852698377042736,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00041995774029288807,
      "loss": 4.939,
      "mean_token_accuracy": 0.21955628097057342,
      "num_tokens": 56702509.0,
      "step": 25400
    },
    {
      "entropy": 5.346932458877563,
      "epoch": 2.8532599539506935,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041992703186799273,
      "loss": 5.0877,
      "mean_token_accuracy": 0.21310963481664658,
      "num_tokens": 56713562.0,
      "step": 25405
    },
    {
      "entropy": 5.301258230209351,
      "epoch": 2.853821530858651,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041989631882856447,
      "loss": 5.0194,
      "mean_token_accuracy": 0.21350958645343782,
      "num_tokens": 56724793.0,
      "step": 25410
    },
    {
      "entropy": 5.3546715259552,
      "epoch": 2.8543831077666084,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004198656011755812,
      "loss": 5.1272,
      "mean_token_accuracy": 0.20172636955976486,
      "num_tokens": 56736842.0,
      "step": 25415
    },
    {
      "entropy": 5.387082719802857,
      "epoch": 2.854944684674566,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004198348789100211,
      "loss": 5.0122,
      "mean_token_accuracy": 0.2027241587638855,
      "num_tokens": 56748220.0,
      "step": 25420
    },
    {
      "entropy": 5.256319427490235,
      "epoch": 2.8555062615825237,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041980415203286223,
      "loss": 4.9322,
      "mean_token_accuracy": 0.2172551319003105,
      "num_tokens": 56758420.0,
      "step": 25425
    },
    {
      "entropy": 5.2475830078125,
      "epoch": 2.8560678384904814,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000419773420545083,
      "loss": 4.9532,
      "mean_token_accuracy": 0.2176455169916153,
      "num_tokens": 56768801.0,
      "step": 25430
    },
    {
      "entropy": 5.417011308670044,
      "epoch": 2.856629415398439,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041974268444766187,
      "loss": 5.1329,
      "mean_token_accuracy": 0.20171410739421844,
      "num_tokens": 56780756.0,
      "step": 25435
    },
    {
      "entropy": 5.344172382354737,
      "epoch": 2.8571909923063963,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004197119437415776,
      "loss": 4.9598,
      "mean_token_accuracy": 0.21601435393095017,
      "num_tokens": 56790679.0,
      "step": 25440
    },
    {
      "entropy": 5.3480470180511475,
      "epoch": 2.857752569214354,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004196811984278089,
      "loss": 5.0473,
      "mean_token_accuracy": 0.2068769946694374,
      "num_tokens": 56802513.0,
      "step": 25445
    },
    {
      "entropy": 5.313390398025513,
      "epoch": 2.8583141461223116,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004196504485073347,
      "loss": 5.0552,
      "mean_token_accuracy": 0.21275153309106826,
      "num_tokens": 56813213.0,
      "step": 25450
    },
    {
      "entropy": 5.327061557769776,
      "epoch": 2.858875723030269,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041961969398113404,
      "loss": 5.0581,
      "mean_token_accuracy": 0.2064370632171631,
      "num_tokens": 56825097.0,
      "step": 25455
    },
    {
      "entropy": 5.371201515197754,
      "epoch": 2.8594372999382265,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041958893485018634,
      "loss": 5.1526,
      "mean_token_accuracy": 0.20595792829990386,
      "num_tokens": 56837185.0,
      "step": 25460
    },
    {
      "entropy": 5.32672381401062,
      "epoch": 2.859998876846184,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004195581711154709,
      "loss": 5.0542,
      "mean_token_accuracy": 0.20622363239526748,
      "num_tokens": 56848696.0,
      "step": 25465
    },
    {
      "entropy": 5.344236278533936,
      "epoch": 2.8605604537541414,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00041952740277796717,
      "loss": 5.108,
      "mean_token_accuracy": 0.20324507653713225,
      "num_tokens": 56860955.0,
      "step": 25470
    },
    {
      "entropy": 5.401370859146118,
      "epoch": 2.861122030662099,
      "grad_norm": 1.0,
      "learning_rate": 0.00041949662983865486,
      "loss": 5.1103,
      "mean_token_accuracy": 0.21178350895643233,
      "num_tokens": 56872375.0,
      "step": 25475
    },
    {
      "entropy": 5.394258308410644,
      "epoch": 2.8616836075700567,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004194658522985139,
      "loss": 5.1148,
      "mean_token_accuracy": 0.20408420264720917,
      "num_tokens": 56883644.0,
      "step": 25480
    },
    {
      "entropy": 5.380287027359008,
      "epoch": 2.8622451844780143,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004194350701585241,
      "loss": 5.0852,
      "mean_token_accuracy": 0.20355095118284225,
      "num_tokens": 56894803.0,
      "step": 25485
    },
    {
      "entropy": 5.380870151519775,
      "epoch": 2.862806761385972,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004194042834196656,
      "loss": 5.0659,
      "mean_token_accuracy": 0.20760650783777237,
      "num_tokens": 56906475.0,
      "step": 25490
    },
    {
      "entropy": 5.389796447753906,
      "epoch": 2.8633683382939292,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004193734920829188,
      "loss": 5.1413,
      "mean_token_accuracy": 0.208842833340168,
      "num_tokens": 56918086.0,
      "step": 25495
    },
    {
      "entropy": 5.333242034912109,
      "epoch": 2.863929915201887,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000419342696149264,
      "loss": 5.0282,
      "mean_token_accuracy": 0.21467415243387222,
      "num_tokens": 56929300.0,
      "step": 25500
    },
    {
      "entropy": 5.409069442749024,
      "epoch": 2.8644914921098446,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004193118956196818,
      "loss": 5.0697,
      "mean_token_accuracy": 0.21525777876377106,
      "num_tokens": 56941858.0,
      "step": 25505
    },
    {
      "entropy": 5.373847246170044,
      "epoch": 2.8650530690178018,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004192810904951528,
      "loss": 5.1559,
      "mean_token_accuracy": 0.2055509105324745,
      "num_tokens": 56953214.0,
      "step": 25510
    },
    {
      "entropy": 5.304107189178467,
      "epoch": 2.8656146459257594,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000419250280776658,
      "loss": 5.0027,
      "mean_token_accuracy": 0.21210146099328994,
      "num_tokens": 56963437.0,
      "step": 25515
    },
    {
      "entropy": 5.406508874893189,
      "epoch": 2.866176222833717,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004192194664651783,
      "loss": 5.0488,
      "mean_token_accuracy": 0.21414054483175277,
      "num_tokens": 56975771.0,
      "step": 25520
    },
    {
      "entropy": 5.272009801864624,
      "epoch": 2.8667377997416748,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004191886475616947,
      "loss": 4.9306,
      "mean_token_accuracy": 0.2201763302087784,
      "num_tokens": 56986084.0,
      "step": 25525
    },
    {
      "entropy": 5.186738681793213,
      "epoch": 2.8672993766496324,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041915782406718875,
      "loss": 4.855,
      "mean_token_accuracy": 0.2217966765165329,
      "num_tokens": 56996616.0,
      "step": 25530
    },
    {
      "entropy": 5.221186685562134,
      "epoch": 2.8678609535575896,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004191269959826418,
      "loss": 4.9681,
      "mean_token_accuracy": 0.21430294960737228,
      "num_tokens": 57008431.0,
      "step": 25535
    },
    {
      "entropy": 5.3199131965637205,
      "epoch": 2.8684225304655473,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041909616330903544,
      "loss": 5.0791,
      "mean_token_accuracy": 0.20984585583209991,
      "num_tokens": 57019265.0,
      "step": 25540
    },
    {
      "entropy": 5.33425464630127,
      "epoch": 2.868984107373505,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041906532604735127,
      "loss": 4.9969,
      "mean_token_accuracy": 0.2113575428724289,
      "num_tokens": 57029869.0,
      "step": 25545
    },
    {
      "entropy": 5.350223255157471,
      "epoch": 2.869545684281462,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004190344841985713,
      "loss": 5.0053,
      "mean_token_accuracy": 0.2171717494726181,
      "num_tokens": 57041168.0,
      "step": 25550
    },
    {
      "entropy": 5.244949054718018,
      "epoch": 2.87010726118942,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041900363776367747,
      "loss": 4.9013,
      "mean_token_accuracy": 0.22521180659532547,
      "num_tokens": 57052190.0,
      "step": 25555
    },
    {
      "entropy": 5.349984264373779,
      "epoch": 2.8706688380973775,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004189727867436519,
      "loss": 5.1059,
      "mean_token_accuracy": 0.20525414496660233,
      "num_tokens": 57063817.0,
      "step": 25560
    },
    {
      "entropy": 5.384402084350586,
      "epoch": 2.8712304150053347,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041894193113947703,
      "loss": 5.1087,
      "mean_token_accuracy": 0.20873149186372758,
      "num_tokens": 57074998.0,
      "step": 25565
    },
    {
      "entropy": 5.303041934967041,
      "epoch": 2.8717919919132924,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041891107095213523,
      "loss": 4.9726,
      "mean_token_accuracy": 0.2113891765475273,
      "num_tokens": 57085377.0,
      "step": 25570
    },
    {
      "entropy": 5.285383081436157,
      "epoch": 2.87235356882125,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041888020618260913,
      "loss": 5.0111,
      "mean_token_accuracy": 0.21417250782251357,
      "num_tokens": 57096698.0,
      "step": 25575
    },
    {
      "entropy": 5.264262390136719,
      "epoch": 2.8729151457292077,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004188493368318814,
      "loss": 4.9944,
      "mean_token_accuracy": 0.22833818048238755,
      "num_tokens": 57107559.0,
      "step": 25580
    },
    {
      "entropy": 5.366135597229004,
      "epoch": 2.8734767226371654,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000418818462900935,
      "loss": 5.0625,
      "mean_token_accuracy": 0.2155907407402992,
      "num_tokens": 57118252.0,
      "step": 25585
    },
    {
      "entropy": 5.42399001121521,
      "epoch": 2.8740382995451226,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004187875843907531,
      "loss": 5.0865,
      "mean_token_accuracy": 0.19933468699455262,
      "num_tokens": 57128635.0,
      "step": 25590
    },
    {
      "entropy": 5.330120325088501,
      "epoch": 2.8745998764530802,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041875670130231857,
      "loss": 5.0353,
      "mean_token_accuracy": 0.2048434406518936,
      "num_tokens": 57138139.0,
      "step": 25595
    },
    {
      "entropy": 5.361653900146484,
      "epoch": 2.875161453361038,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000418725813636615,
      "loss": 5.0806,
      "mean_token_accuracy": 0.21405237764120102,
      "num_tokens": 57149456.0,
      "step": 25600
    },
    {
      "entropy": 5.434307527542114,
      "epoch": 2.875723030268995,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041869492139462585,
      "loss": 5.1245,
      "mean_token_accuracy": 0.20957318246364592,
      "num_tokens": 57160710.0,
      "step": 25605
    },
    {
      "entropy": 5.385175037384033,
      "epoch": 2.876284607176953,
      "grad_norm": 1.0,
      "learning_rate": 0.00041866402457733454,
      "loss": 5.0246,
      "mean_token_accuracy": 0.2123329982161522,
      "num_tokens": 57172006.0,
      "step": 25610
    },
    {
      "entropy": 5.2984840869903564,
      "epoch": 2.8768461840849104,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041863312318572504,
      "loss": 5.0424,
      "mean_token_accuracy": 0.20814408510923385,
      "num_tokens": 57183665.0,
      "step": 25615
    },
    {
      "entropy": 5.300361680984497,
      "epoch": 2.877407760992868,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004186022172207812,
      "loss": 5.0873,
      "mean_token_accuracy": 0.200809483230114,
      "num_tokens": 57195711.0,
      "step": 25620
    },
    {
      "entropy": 5.354044437408447,
      "epoch": 2.8779693379008258,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00041857130668348705,
      "loss": 4.9882,
      "mean_token_accuracy": 0.21750750690698623,
      "num_tokens": 57206664.0,
      "step": 25625
    },
    {
      "entropy": 5.346395492553711,
      "epoch": 2.878530914808783,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004185403915748268,
      "loss": 5.0881,
      "mean_token_accuracy": 0.2041509583592415,
      "num_tokens": 57217407.0,
      "step": 25630
    },
    {
      "entropy": 5.308056974411011,
      "epoch": 2.8790924917167406,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004185094718957848,
      "loss": 4.9964,
      "mean_token_accuracy": 0.21262036859989167,
      "num_tokens": 57228707.0,
      "step": 25635
    },
    {
      "entropy": 5.275713205337524,
      "epoch": 2.8796540686246983,
      "grad_norm": 0.9375,
      "learning_rate": 0.00041847854764734556,
      "loss": 4.9786,
      "mean_token_accuracy": 0.2146846741437912,
      "num_tokens": 57240802.0,
      "step": 25640
    },
    {
      "entropy": 5.353298759460449,
      "epoch": 2.8802156455326555,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004184476188304937,
      "loss": 5.0945,
      "mean_token_accuracy": 0.2060813069343567,
      "num_tokens": 57251999.0,
      "step": 25645
    },
    {
      "entropy": 5.504148960113525,
      "epoch": 2.880777222440613,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00041841668544621396,
      "loss": 5.2312,
      "mean_token_accuracy": 0.19982937574386597,
      "num_tokens": 57264199.0,
      "step": 25650
    },
    {
      "entropy": 5.287557601928711,
      "epoch": 2.881338799348571,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004183857474954915,
      "loss": 4.9629,
      "mean_token_accuracy": 0.20950134098529816,
      "num_tokens": 57275234.0,
      "step": 25655
    },
    {
      "entropy": 5.325562906265259,
      "epoch": 2.881900376256528,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041835480497931114,
      "loss": 5.1079,
      "mean_token_accuracy": 0.20764128416776656,
      "num_tokens": 57286920.0,
      "step": 25660
    },
    {
      "entropy": 5.390311241149902,
      "epoch": 2.8824619531644857,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004183238578986582,
      "loss": 5.0446,
      "mean_token_accuracy": 0.2142785981297493,
      "num_tokens": 57298148.0,
      "step": 25665
    },
    {
      "entropy": 5.346717882156372,
      "epoch": 2.8830235300724434,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041829290625451806,
      "loss": 5.0345,
      "mean_token_accuracy": 0.2137853592634201,
      "num_tokens": 57308717.0,
      "step": 25670
    },
    {
      "entropy": 5.34081244468689,
      "epoch": 2.883585106980401,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041826195004787616,
      "loss": 5.009,
      "mean_token_accuracy": 0.20411043763160705,
      "num_tokens": 57319308.0,
      "step": 25675
    },
    {
      "entropy": 5.443326473236084,
      "epoch": 2.8841466838883587,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00041823098927971827,
      "loss": 5.143,
      "mean_token_accuracy": 0.19876517057418824,
      "num_tokens": 57330911.0,
      "step": 25680
    },
    {
      "entropy": 5.364762973785401,
      "epoch": 2.884708260796316,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004182000239510301,
      "loss": 5.0696,
      "mean_token_accuracy": 0.2043062061071396,
      "num_tokens": 57341624.0,
      "step": 25685
    },
    {
      "entropy": 5.343545579910279,
      "epoch": 2.8852698377042736,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00041816905406279756,
      "loss": 5.0274,
      "mean_token_accuracy": 0.21711026430130004,
      "num_tokens": 57353399.0,
      "step": 25690
    },
    {
      "entropy": 5.418311357498169,
      "epoch": 2.8858314146122312,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041813807961600685,
      "loss": 5.1298,
      "mean_token_accuracy": 0.20321349054574966,
      "num_tokens": 57364481.0,
      "step": 25695
    },
    {
      "entropy": 5.328123617172241,
      "epoch": 2.8863929915201885,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004181071006116442,
      "loss": 4.9523,
      "mean_token_accuracy": 0.2144824296236038,
      "num_tokens": 57375588.0,
      "step": 25700
    },
    {
      "entropy": 5.224964904785156,
      "epoch": 2.886954568428146,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041807611705069593,
      "loss": 5.0035,
      "mean_token_accuracy": 0.21191116720438002,
      "num_tokens": 57387048.0,
      "step": 25705
    },
    {
      "entropy": 5.282592010498047,
      "epoch": 2.887516145336104,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041804512893414865,
      "loss": 4.9537,
      "mean_token_accuracy": 0.22176286727190017,
      "num_tokens": 57397946.0,
      "step": 25710
    },
    {
      "entropy": 5.361367893218994,
      "epoch": 2.8880777222440615,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00041801413626298893,
      "loss": 5.1134,
      "mean_token_accuracy": 0.208761827647686,
      "num_tokens": 57409505.0,
      "step": 25715
    },
    {
      "entropy": 5.259291458129883,
      "epoch": 2.888639299152019,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004179831390382037,
      "loss": 5.0471,
      "mean_token_accuracy": 0.20747087597846986,
      "num_tokens": 57420745.0,
      "step": 25720
    },
    {
      "entropy": 5.359415817260742,
      "epoch": 2.8892008760599763,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004179521372607798,
      "loss": 5.1176,
      "mean_token_accuracy": 0.20789354145526887,
      "num_tokens": 57431557.0,
      "step": 25725
    },
    {
      "entropy": 5.345820808410645,
      "epoch": 2.889762452967934,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041792113093170447,
      "loss": 4.9768,
      "mean_token_accuracy": 0.20965595692396163,
      "num_tokens": 57442029.0,
      "step": 25730
    },
    {
      "entropy": 5.464095878601074,
      "epoch": 2.8903240298758917,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004178901200519649,
      "loss": 5.0978,
      "mean_token_accuracy": 0.20154281258583068,
      "num_tokens": 57452921.0,
      "step": 25735
    },
    {
      "entropy": 5.281476736068726,
      "epoch": 2.890885606783849,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004178591046225485,
      "loss": 4.9475,
      "mean_token_accuracy": 0.21033574342727662,
      "num_tokens": 57463068.0,
      "step": 25740
    },
    {
      "entropy": 5.31468653678894,
      "epoch": 2.8914471836918065,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041782808464444274,
      "loss": 5.0719,
      "mean_token_accuracy": 0.21508701145648956,
      "num_tokens": 57474064.0,
      "step": 25745
    },
    {
      "entropy": 5.290033578872681,
      "epoch": 2.892008760599764,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041779706011863546,
      "loss": 5.0361,
      "mean_token_accuracy": 0.2100598156452179,
      "num_tokens": 57485473.0,
      "step": 25750
    },
    {
      "entropy": 5.257337236404419,
      "epoch": 2.8925703375077214,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041776603104611427,
      "loss": 4.9199,
      "mean_token_accuracy": 0.2206255465745926,
      "num_tokens": 57495562.0,
      "step": 25755
    },
    {
      "entropy": 5.295083808898926,
      "epoch": 2.893131914415679,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041773499742786736,
      "loss": 5.045,
      "mean_token_accuracy": 0.21642139554023743,
      "num_tokens": 57506322.0,
      "step": 25760
    },
    {
      "entropy": 5.370458698272705,
      "epoch": 2.8936934913236367,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041770395926488276,
      "loss": 5.0458,
      "mean_token_accuracy": 0.214109343290329,
      "num_tokens": 57517353.0,
      "step": 25765
    },
    {
      "entropy": 5.420284128189087,
      "epoch": 2.8942550682315944,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004176729165581488,
      "loss": 5.0777,
      "mean_token_accuracy": 0.21396704763174057,
      "num_tokens": 57527273.0,
      "step": 25770
    },
    {
      "entropy": 5.267073440551758,
      "epoch": 2.894816645139552,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041764186930865377,
      "loss": 4.9748,
      "mean_token_accuracy": 0.21574198752641677,
      "num_tokens": 57538486.0,
      "step": 25775
    },
    {
      "entropy": 5.315969514846802,
      "epoch": 2.8953782220475093,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004176108175173863,
      "loss": 5.0299,
      "mean_token_accuracy": 0.21383726298809053,
      "num_tokens": 57549682.0,
      "step": 25780
    },
    {
      "entropy": 5.369179391860962,
      "epoch": 2.895939798955467,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004175797611853351,
      "loss": 5.0036,
      "mean_token_accuracy": 0.21228594034910203,
      "num_tokens": 57559397.0,
      "step": 25785
    },
    {
      "entropy": 5.378868913650512,
      "epoch": 2.8965013758634246,
      "grad_norm": 1.078125,
      "learning_rate": 0.000417548700313489,
      "loss": 5.0585,
      "mean_token_accuracy": 0.21544300615787507,
      "num_tokens": 57570568.0,
      "step": 25790
    },
    {
      "entropy": 5.2857685565948485,
      "epoch": 2.897062952771382,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000417517634902837,
      "loss": 4.934,
      "mean_token_accuracy": 0.2179110452532768,
      "num_tokens": 57581152.0,
      "step": 25795
    },
    {
      "entropy": 5.361731386184692,
      "epoch": 2.8976245296793395,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041748656495436834,
      "loss": 5.111,
      "mean_token_accuracy": 0.20986947417259216,
      "num_tokens": 57592227.0,
      "step": 25800
    },
    {
      "entropy": 5.339450359344482,
      "epoch": 2.898186106587297,
      "grad_norm": 1.0,
      "learning_rate": 0.00041745549046907205,
      "loss": 5.0374,
      "mean_token_accuracy": 0.2071199968457222,
      "num_tokens": 57604037.0,
      "step": 25805
    },
    {
      "entropy": 5.4430512428283695,
      "epoch": 2.898747683495255,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041742441144793774,
      "loss": 5.0838,
      "mean_token_accuracy": 0.21680572628974915,
      "num_tokens": 57613937.0,
      "step": 25810
    },
    {
      "entropy": 5.301461553573608,
      "epoch": 2.8993092604032125,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004173933278919549,
      "loss": 4.9644,
      "mean_token_accuracy": 0.20697468072175979,
      "num_tokens": 57624287.0,
      "step": 25815
    },
    {
      "entropy": 5.313282632827759,
      "epoch": 2.8998708373111697,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004173622398021133,
      "loss": 5.0989,
      "mean_token_accuracy": 0.20736879408359526,
      "num_tokens": 57636677.0,
      "step": 25820
    },
    {
      "entropy": 5.228234481811524,
      "epoch": 2.9004324142191273,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00041733114717940273,
      "loss": 4.9665,
      "mean_token_accuracy": 0.21880622506141661,
      "num_tokens": 57647848.0,
      "step": 25825
    },
    {
      "entropy": 5.257660627365112,
      "epoch": 2.900993991127085,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004173000500248133,
      "loss": 4.879,
      "mean_token_accuracy": 0.2203317865729332,
      "num_tokens": 57658946.0,
      "step": 25830
    },
    {
      "entropy": 5.305136871337891,
      "epoch": 2.901555568035042,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004172689483393351,
      "loss": 5.0591,
      "mean_token_accuracy": 0.21170557886362076,
      "num_tokens": 57670690.0,
      "step": 25835
    },
    {
      "entropy": 5.474815893173218,
      "epoch": 2.902117144943,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041723784212395834,
      "loss": 5.1346,
      "mean_token_accuracy": 0.20245403200387954,
      "num_tokens": 57682309.0,
      "step": 25840
    },
    {
      "entropy": 5.357135963439942,
      "epoch": 2.9026787218509575,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041720673137967356,
      "loss": 5.0976,
      "mean_token_accuracy": 0.20942550301551818,
      "num_tokens": 57692804.0,
      "step": 25845
    },
    {
      "entropy": 5.4469537258148195,
      "epoch": 2.9032402987589148,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041717561610747124,
      "loss": 5.1506,
      "mean_token_accuracy": 0.2092818424105644,
      "num_tokens": 57705824.0,
      "step": 25850
    },
    {
      "entropy": 5.365650701522827,
      "epoch": 2.9038018756668724,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041714449630834223,
      "loss": 5.0276,
      "mean_token_accuracy": 0.2151342123746872,
      "num_tokens": 57718983.0,
      "step": 25855
    },
    {
      "entropy": 5.348388671875,
      "epoch": 2.90436345257483,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004171133719832774,
      "loss": 4.9935,
      "mean_token_accuracy": 0.21126711666584014,
      "num_tokens": 57730727.0,
      "step": 25860
    },
    {
      "entropy": 5.320789575576782,
      "epoch": 2.9049250294827877,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041708224313326747,
      "loss": 5.1211,
      "mean_token_accuracy": 0.20156874209642411,
      "num_tokens": 57742474.0,
      "step": 25865
    },
    {
      "entropy": 5.43751368522644,
      "epoch": 2.9054866063907454,
      "grad_norm": 0.921875,
      "learning_rate": 0.000417051109759304,
      "loss": 5.0824,
      "mean_token_accuracy": 0.20963372588157653,
      "num_tokens": 57755094.0,
      "step": 25870
    },
    {
      "entropy": 5.394785070419312,
      "epoch": 2.9060481832987026,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000417019971862378,
      "loss": 5.106,
      "mean_token_accuracy": 0.2035985916852951,
      "num_tokens": 57765820.0,
      "step": 25875
    },
    {
      "entropy": 5.342053985595703,
      "epoch": 2.9066097602066603,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00041698882944348105,
      "loss": 5.0342,
      "mean_token_accuracy": 0.20583954006433486,
      "num_tokens": 57776749.0,
      "step": 25880
    },
    {
      "entropy": 5.2854986667633055,
      "epoch": 2.907171337114618,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00041695768250360466,
      "loss": 4.9547,
      "mean_token_accuracy": 0.21296027153730393,
      "num_tokens": 57787734.0,
      "step": 25885
    },
    {
      "entropy": 5.327765703201294,
      "epoch": 2.907732914022575,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004169265310437407,
      "loss": 5.0601,
      "mean_token_accuracy": 0.2084406942129135,
      "num_tokens": 57799799.0,
      "step": 25890
    },
    {
      "entropy": 5.225455570220947,
      "epoch": 2.908294490930533,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004168953750648809,
      "loss": 4.9369,
      "mean_token_accuracy": 0.21224041432142257,
      "num_tokens": 57810446.0,
      "step": 25895
    },
    {
      "entropy": 5.339599466323852,
      "epoch": 2.9088560678384905,
      "grad_norm": 1.0,
      "learning_rate": 0.00041686421456801724,
      "loss": 5.1546,
      "mean_token_accuracy": 0.2068501129746437,
      "num_tokens": 57822931.0,
      "step": 25900
    },
    {
      "entropy": 5.409865283966065,
      "epoch": 2.909417644746448,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004168330495541421,
      "loss": 5.1098,
      "mean_token_accuracy": 0.19969463646411895,
      "num_tokens": 57835250.0,
      "step": 25905
    },
    {
      "entropy": 5.336263418197632,
      "epoch": 2.909979221654406,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00041680188002424755,
      "loss": 5.0937,
      "mean_token_accuracy": 0.20981687754392625,
      "num_tokens": 57847726.0,
      "step": 25910
    },
    {
      "entropy": 5.343692684173584,
      "epoch": 2.910540798562363,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004167707059793262,
      "loss": 5.0688,
      "mean_token_accuracy": 0.21067297011613845,
      "num_tokens": 57858484.0,
      "step": 25915
    },
    {
      "entropy": 5.31429386138916,
      "epoch": 2.9111023754703207,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00041673952742037057,
      "loss": 4.989,
      "mean_token_accuracy": 0.22166004925966262,
      "num_tokens": 57870469.0,
      "step": 25920
    },
    {
      "entropy": 5.348054981231689,
      "epoch": 2.9116639523782784,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004167083443483734,
      "loss": 5.0679,
      "mean_token_accuracy": 0.2077491343021393,
      "num_tokens": 57880088.0,
      "step": 25925
    },
    {
      "entropy": 5.354591703414917,
      "epoch": 2.9122255292862356,
      "grad_norm": 1.0,
      "learning_rate": 0.0004166771567643276,
      "loss": 5.0459,
      "mean_token_accuracy": 0.20210457295179368,
      "num_tokens": 57890521.0,
      "step": 25930
    },
    {
      "entropy": 5.441933584213257,
      "epoch": 2.9127871061941932,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004166459646692262,
      "loss": 5.1852,
      "mean_token_accuracy": 0.2036396935582161,
      "num_tokens": 57904111.0,
      "step": 25935
    },
    {
      "entropy": 5.401638698577881,
      "epoch": 2.913348683102151,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004166147680640623,
      "loss": 5.0935,
      "mean_token_accuracy": 0.20744675993919373,
      "num_tokens": 57915854.0,
      "step": 25940
    },
    {
      "entropy": 5.3160652160644535,
      "epoch": 2.913910260010108,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041658356694982936,
      "loss": 4.9433,
      "mean_token_accuracy": 0.22015414088964463,
      "num_tokens": 57926822.0,
      "step": 25945
    },
    {
      "entropy": 5.3170829772949215,
      "epoch": 2.9144718369180658,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004165523613275207,
      "loss": 5.0777,
      "mean_token_accuracy": 0.21174923628568648,
      "num_tokens": 57939029.0,
      "step": 25950
    },
    {
      "entropy": 5.2961937427520756,
      "epoch": 2.9150334138260234,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004165211511981299,
      "loss": 5.0102,
      "mean_token_accuracy": 0.21199098378419876,
      "num_tokens": 57950053.0,
      "step": 25955
    },
    {
      "entropy": 5.312186145782471,
      "epoch": 2.915594990733981,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004164899365626507,
      "loss": 4.9325,
      "mean_token_accuracy": 0.22279451340436934,
      "num_tokens": 57960925.0,
      "step": 25960
    },
    {
      "entropy": 5.270992374420166,
      "epoch": 2.9161565676419388,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004164587174220771,
      "loss": 4.9923,
      "mean_token_accuracy": 0.21064579337835312,
      "num_tokens": 57971948.0,
      "step": 25965
    },
    {
      "entropy": 5.322525691986084,
      "epoch": 2.916718144549896,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041642749377740316,
      "loss": 5.093,
      "mean_token_accuracy": 0.20065228193998336,
      "num_tokens": 57984310.0,
      "step": 25970
    },
    {
      "entropy": 5.326031732559204,
      "epoch": 2.9172797214578536,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00041639626562962293,
      "loss": 4.9613,
      "mean_token_accuracy": 0.2163064405322075,
      "num_tokens": 57995194.0,
      "step": 25975
    },
    {
      "entropy": 5.3615233421325685,
      "epoch": 2.9178412983658113,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041636503297973074,
      "loss": 5.0914,
      "mean_token_accuracy": 0.20767486691474915,
      "num_tokens": 58006370.0,
      "step": 25980
    },
    {
      "entropy": 5.3638448238372805,
      "epoch": 2.9184028752737685,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000416333795828721,
      "loss": 5.0309,
      "mean_token_accuracy": 0.21984266638755798,
      "num_tokens": 58017635.0,
      "step": 25985
    },
    {
      "entropy": 5.360009670257568,
      "epoch": 2.918964452181726,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004163025541775885,
      "loss": 5.0228,
      "mean_token_accuracy": 0.22042160779237746,
      "num_tokens": 58028110.0,
      "step": 25990
    },
    {
      "entropy": 5.354009437561035,
      "epoch": 2.919526029089684,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004162713080273278,
      "loss": 5.1223,
      "mean_token_accuracy": 0.20028709322214128,
      "num_tokens": 58039715.0,
      "step": 25995
    },
    {
      "entropy": 5.431981182098388,
      "epoch": 2.9200876059976415,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041624005737893387,
      "loss": 5.0637,
      "mean_token_accuracy": 0.21514069885015488,
      "num_tokens": 58051052.0,
      "step": 26000
    },
    {
      "entropy": 5.297756147384644,
      "epoch": 2.920649182905599,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004162088022334018,
      "loss": 5.0625,
      "mean_token_accuracy": 0.20774129182100295,
      "num_tokens": 58062789.0,
      "step": 26005
    },
    {
      "entropy": 5.245779132843017,
      "epoch": 2.9212107598135564,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00041617754259172665,
      "loss": 4.9713,
      "mean_token_accuracy": 0.21295049637556077,
      "num_tokens": 58073896.0,
      "step": 26010
    },
    {
      "entropy": 5.496474266052246,
      "epoch": 2.921772336721514,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041614627845490377,
      "loss": 5.2421,
      "mean_token_accuracy": 0.20014094114303588,
      "num_tokens": 58084529.0,
      "step": 26015
    },
    {
      "entropy": 5.398008966445923,
      "epoch": 2.9223339136294717,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041611500982392867,
      "loss": 5.0677,
      "mean_token_accuracy": 0.20793922394514083,
      "num_tokens": 58095989.0,
      "step": 26020
    },
    {
      "entropy": 5.284086847305298,
      "epoch": 2.922895490537429,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041608373669979696,
      "loss": 4.9161,
      "mean_token_accuracy": 0.2210013046860695,
      "num_tokens": 58107001.0,
      "step": 26025
    },
    {
      "entropy": 5.299263000488281,
      "epoch": 2.9234570674453866,
      "grad_norm": 1.0,
      "learning_rate": 0.0004160524590835042,
      "loss": 4.9919,
      "mean_token_accuracy": 0.21370194852352142,
      "num_tokens": 58118224.0,
      "step": 26030
    },
    {
      "entropy": 5.308971834182739,
      "epoch": 2.9240186443533442,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041602117697604664,
      "loss": 4.9626,
      "mean_token_accuracy": 0.21929688900709152,
      "num_tokens": 58129675.0,
      "step": 26035
    },
    {
      "entropy": 5.336674690246582,
      "epoch": 2.9245802212613015,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041598989037842,
      "loss": 5.0427,
      "mean_token_accuracy": 0.20455956906080247,
      "num_tokens": 58140409.0,
      "step": 26040
    },
    {
      "entropy": 5.291958808898926,
      "epoch": 2.925141798169259,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004159585992916206,
      "loss": 4.9808,
      "mean_token_accuracy": 0.21754807978868484,
      "num_tokens": 58151139.0,
      "step": 26045
    },
    {
      "entropy": 5.392928838729858,
      "epoch": 2.925703375077217,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004159273037166447,
      "loss": 5.1249,
      "mean_token_accuracy": 0.20404072701931,
      "num_tokens": 58162953.0,
      "step": 26050
    },
    {
      "entropy": 5.371767473220825,
      "epoch": 2.9262649519851744,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041589600365448884,
      "loss": 4.9722,
      "mean_token_accuracy": 0.21044668555259705,
      "num_tokens": 58173480.0,
      "step": 26055
    },
    {
      "entropy": 5.391554975509644,
      "epoch": 2.926826528893132,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004158646991061495,
      "loss": 5.131,
      "mean_token_accuracy": 0.20800490081310272,
      "num_tokens": 58185392.0,
      "step": 26060
    },
    {
      "entropy": 5.321698141098023,
      "epoch": 2.9273881058010893,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004158333900726235,
      "loss": 4.9427,
      "mean_token_accuracy": 0.21968081146478652,
      "num_tokens": 58195984.0,
      "step": 26065
    },
    {
      "entropy": 5.392368316650391,
      "epoch": 2.927949682709047,
      "grad_norm": 1.0,
      "learning_rate": 0.00041580207655490776,
      "loss": 5.0908,
      "mean_token_accuracy": 0.20760823488235475,
      "num_tokens": 58207529.0,
      "step": 26070
    },
    {
      "entropy": 5.344250059127807,
      "epoch": 2.9285112596170046,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004157707585539994,
      "loss": 5.0263,
      "mean_token_accuracy": 0.21238889694213867,
      "num_tokens": 58217936.0,
      "step": 26075
    },
    {
      "entropy": 5.331900978088379,
      "epoch": 2.929072836524962,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041573943607089536,
      "loss": 5.0174,
      "mean_token_accuracy": 0.21718883514404297,
      "num_tokens": 58228144.0,
      "step": 26080
    },
    {
      "entropy": 5.271958875656128,
      "epoch": 2.9296344134329195,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004157081091065932,
      "loss": 5.0923,
      "mean_token_accuracy": 0.21507591307163237,
      "num_tokens": 58239130.0,
      "step": 26085
    },
    {
      "entropy": 5.3209099769592285,
      "epoch": 2.930195990340877,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004156767776620902,
      "loss": 5.0372,
      "mean_token_accuracy": 0.21300189346075057,
      "num_tokens": 58253119.0,
      "step": 26090
    },
    {
      "entropy": 5.402999782562256,
      "epoch": 2.930757567248835,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000415645441738384,
      "loss": 5.0653,
      "mean_token_accuracy": 0.20582000464200972,
      "num_tokens": 58265549.0,
      "step": 26095
    },
    {
      "entropy": 5.300690221786499,
      "epoch": 2.9313191441567925,
      "grad_norm": 1.0,
      "learning_rate": 0.00041561410133647244,
      "loss": 4.951,
      "mean_token_accuracy": 0.2230711132287979,
      "num_tokens": 58275021.0,
      "step": 26100
    },
    {
      "entropy": 5.335240077972412,
      "epoch": 2.9318807210647497,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041558275645735336,
      "loss": 5.0369,
      "mean_token_accuracy": 0.2102704241871834,
      "num_tokens": 58286191.0,
      "step": 26105
    },
    {
      "entropy": 5.259894752502442,
      "epoch": 2.9324422979727074,
      "grad_norm": 0.9375,
      "learning_rate": 0.00041555140710202484,
      "loss": 4.9133,
      "mean_token_accuracy": 0.21802672892808914,
      "num_tokens": 58297066.0,
      "step": 26110
    },
    {
      "entropy": 5.342387008666992,
      "epoch": 2.933003874880665,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041552005327148494,
      "loss": 5.0039,
      "mean_token_accuracy": 0.21951244920492172,
      "num_tokens": 58308008.0,
      "step": 26115
    },
    {
      "entropy": 5.36880316734314,
      "epoch": 2.9335654517886223,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004154886949667321,
      "loss": 5.1289,
      "mean_token_accuracy": 0.2078862965106964,
      "num_tokens": 58319584.0,
      "step": 26120
    },
    {
      "entropy": 5.405076694488526,
      "epoch": 2.93412702869658,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041545733218876466,
      "loss": 5.0047,
      "mean_token_accuracy": 0.21137510985136032,
      "num_tokens": 58328964.0,
      "step": 26125
    },
    {
      "entropy": 5.38420000076294,
      "epoch": 2.9346886056045376,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004154259649385813,
      "loss": 5.0555,
      "mean_token_accuracy": 0.2180003196001053,
      "num_tokens": 58339673.0,
      "step": 26130
    },
    {
      "entropy": 5.364515066146851,
      "epoch": 2.935250182512495,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041539459321718076,
      "loss": 5.0907,
      "mean_token_accuracy": 0.2103130891919136,
      "num_tokens": 58351453.0,
      "step": 26135
    },
    {
      "entropy": 5.424620819091797,
      "epoch": 2.9358117594204525,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004153632170255619,
      "loss": 5.1302,
      "mean_token_accuracy": 0.205039444565773,
      "num_tokens": 58362792.0,
      "step": 26140
    },
    {
      "entropy": 5.348247528076172,
      "epoch": 2.93637333632841,
      "grad_norm": 0.921875,
      "learning_rate": 0.00041533183636472385,
      "loss": 5.0938,
      "mean_token_accuracy": 0.21118698716163636,
      "num_tokens": 58372628.0,
      "step": 26145
    },
    {
      "entropy": 5.258068180084228,
      "epoch": 2.936934913236368,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004153004512356656,
      "loss": 4.9527,
      "mean_token_accuracy": 0.2157023921608925,
      "num_tokens": 58383723.0,
      "step": 26150
    },
    {
      "entropy": 5.304973554611206,
      "epoch": 2.9374964901443255,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004152690616393866,
      "loss": 5.0169,
      "mean_token_accuracy": 0.21480415463447572,
      "num_tokens": 58394547.0,
      "step": 26155
    },
    {
      "entropy": 5.3180042743682865,
      "epoch": 2.9380580670522827,
      "grad_norm": 1.0,
      "learning_rate": 0.00041523766757688634,
      "loss": 5.0344,
      "mean_token_accuracy": 0.21190262585878372,
      "num_tokens": 58405460.0,
      "step": 26160
    },
    {
      "entropy": 5.364919805526734,
      "epoch": 2.9386196439602403,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004152062690491643,
      "loss": 5.0561,
      "mean_token_accuracy": 0.21116422563791276,
      "num_tokens": 58416436.0,
      "step": 26165
    },
    {
      "entropy": 5.387572479248047,
      "epoch": 2.939181220868198,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004151748660572203,
      "loss": 4.9879,
      "mean_token_accuracy": 0.21332670897245407,
      "num_tokens": 58426920.0,
      "step": 26170
    },
    {
      "entropy": 5.283280992507935,
      "epoch": 2.939742797776155,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004151434586020541,
      "loss": 4.8902,
      "mean_token_accuracy": 0.221931329369545,
      "num_tokens": 58437525.0,
      "step": 26175
    },
    {
      "entropy": 5.301758098602295,
      "epoch": 2.940304374684113,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00041511204668466593,
      "loss": 5.1328,
      "mean_token_accuracy": 0.21099719107151033,
      "num_tokens": 58448955.0,
      "step": 26180
    },
    {
      "entropy": 5.2407372951507565,
      "epoch": 2.9408659515920705,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041508063030605585,
      "loss": 4.9092,
      "mean_token_accuracy": 0.21966032087802886,
      "num_tokens": 58459941.0,
      "step": 26185
    },
    {
      "entropy": 5.332545948028565,
      "epoch": 2.941427528500028,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041504920946722413,
      "loss": 4.985,
      "mean_token_accuracy": 0.21366311013698577,
      "num_tokens": 58470499.0,
      "step": 26190
    },
    {
      "entropy": 5.315614223480225,
      "epoch": 2.941989105407986,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004150177841691713,
      "loss": 5.0107,
      "mean_token_accuracy": 0.2129129499197006,
      "num_tokens": 58482969.0,
      "step": 26195
    },
    {
      "entropy": 5.410284328460693,
      "epoch": 2.942550682315943,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004149863544128979,
      "loss": 5.064,
      "mean_token_accuracy": 0.20868894904851915,
      "num_tokens": 58492929.0,
      "step": 26200
    },
    {
      "entropy": 5.314926195144653,
      "epoch": 2.9431122592239007,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00041495492019940467,
      "loss": 4.9907,
      "mean_token_accuracy": 0.2177803173661232,
      "num_tokens": 58503932.0,
      "step": 26205
    },
    {
      "entropy": 5.182077169418335,
      "epoch": 2.9436738361318584,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004149234815296925,
      "loss": 4.902,
      "mean_token_accuracy": 0.21629796028137208,
      "num_tokens": 58515282.0,
      "step": 26210
    },
    {
      "entropy": 5.343235015869141,
      "epoch": 2.9442354130398156,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041489203840476243,
      "loss": 5.1436,
      "mean_token_accuracy": 0.20679474472999573,
      "num_tokens": 58527218.0,
      "step": 26215
    },
    {
      "entropy": 5.398764419555664,
      "epoch": 2.9447969899477733,
      "grad_norm": 1.0,
      "learning_rate": 0.00041486059082561556,
      "loss": 5.0,
      "mean_token_accuracy": 0.21402018815279006,
      "num_tokens": 58538622.0,
      "step": 26220
    },
    {
      "entropy": 5.3185515880584715,
      "epoch": 2.945358566855731,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004148291387932533,
      "loss": 5.034,
      "mean_token_accuracy": 0.2180113300681114,
      "num_tokens": 58549747.0,
      "step": 26225
    },
    {
      "entropy": 5.284799671173095,
      "epoch": 2.9459201437636886,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041479768230867696,
      "loss": 5.071,
      "mean_token_accuracy": 0.20805447697639465,
      "num_tokens": 58560771.0,
      "step": 26230
    },
    {
      "entropy": 5.333605480194092,
      "epoch": 2.946481720671646,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041476622137288823,
      "loss": 4.9549,
      "mean_token_accuracy": 0.2100521594285965,
      "num_tokens": 58571184.0,
      "step": 26235
    },
    {
      "entropy": 5.403164625167847,
      "epoch": 2.9470432975796035,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041473475598688886,
      "loss": 5.0907,
      "mean_token_accuracy": 0.20486722141504288,
      "num_tokens": 58583075.0,
      "step": 26240
    },
    {
      "entropy": 5.3624587059021,
      "epoch": 2.947604874487561,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004147032861516805,
      "loss": 5.0005,
      "mean_token_accuracy": 0.2129198431968689,
      "num_tokens": 58593978.0,
      "step": 26245
    },
    {
      "entropy": 5.325455188751221,
      "epoch": 2.948166451395519,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00041467181186826546,
      "loss": 5.0237,
      "mean_token_accuracy": 0.20967066586017608,
      "num_tokens": 58604525.0,
      "step": 26250
    },
    {
      "entropy": 5.270766925811768,
      "epoch": 2.948728028303476,
      "grad_norm": 1.0,
      "learning_rate": 0.00041464033313764575,
      "loss": 5.0121,
      "mean_token_accuracy": 0.21648085713386536,
      "num_tokens": 58614965.0,
      "step": 26255
    },
    {
      "entropy": 5.301524114608765,
      "epoch": 2.9492896052114337,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041460884996082363,
      "loss": 5.0144,
      "mean_token_accuracy": 0.21362558007240295,
      "num_tokens": 58626254.0,
      "step": 26260
    },
    {
      "entropy": 5.337866020202637,
      "epoch": 2.9498511821193913,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041457736233880166,
      "loss": 5.0747,
      "mean_token_accuracy": 0.2102547273039818,
      "num_tokens": 58637474.0,
      "step": 26265
    },
    {
      "entropy": 5.393818998336792,
      "epoch": 2.9504127590273486,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004145458702725823,
      "loss": 5.0563,
      "mean_token_accuracy": 0.20888643115758895,
      "num_tokens": 58649560.0,
      "step": 26270
    },
    {
      "entropy": 5.298198127746582,
      "epoch": 2.9509743359353062,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041451437376316825,
      "loss": 4.9893,
      "mean_token_accuracy": 0.2096676543354988,
      "num_tokens": 58659944.0,
      "step": 26275
    },
    {
      "entropy": 5.37312798500061,
      "epoch": 2.951535912843264,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004144828728115624,
      "loss": 5.016,
      "mean_token_accuracy": 0.20840391516685486,
      "num_tokens": 58670636.0,
      "step": 26280
    },
    {
      "entropy": 5.343139171600342,
      "epoch": 2.9520974897512215,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004144513674187679,
      "loss": 5.0729,
      "mean_token_accuracy": 0.20852116346359253,
      "num_tokens": 58682832.0,
      "step": 26285
    },
    {
      "entropy": 5.338013124465943,
      "epoch": 2.952659066659179,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004144198575857878,
      "loss": 5.0513,
      "mean_token_accuracy": 0.21308449655771255,
      "num_tokens": 58694675.0,
      "step": 26290
    },
    {
      "entropy": 5.287676668167114,
      "epoch": 2.9532206435671364,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004143883433136252,
      "loss": 5.0275,
      "mean_token_accuracy": 0.21696840822696686,
      "num_tokens": 58705613.0,
      "step": 26295
    },
    {
      "entropy": 5.414187097549439,
      "epoch": 2.953782220475094,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004143568246032838,
      "loss": 5.0939,
      "mean_token_accuracy": 0.2092654213309288,
      "num_tokens": 58716951.0,
      "step": 26300
    },
    {
      "entropy": 5.320406293869018,
      "epoch": 2.9543437973830518,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004143253014557671,
      "loss": 5.106,
      "mean_token_accuracy": 0.20611971765756607,
      "num_tokens": 58727730.0,
      "step": 26305
    },
    {
      "entropy": 5.297557735443116,
      "epoch": 2.954905374291009,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041429377387207865,
      "loss": 4.9849,
      "mean_token_accuracy": 0.21345534473657607,
      "num_tokens": 58739183.0,
      "step": 26310
    },
    {
      "entropy": 5.411561489105225,
      "epoch": 2.9554669511989666,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041426224185322255,
      "loss": 5.0881,
      "mean_token_accuracy": 0.21326705664396287,
      "num_tokens": 58750031.0,
      "step": 26315
    },
    {
      "entropy": 5.389292573928833,
      "epoch": 2.9560285281069243,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004142307054002025,
      "loss": 5.064,
      "mean_token_accuracy": 0.21074432879686356,
      "num_tokens": 58760488.0,
      "step": 26320
    },
    {
      "entropy": 5.3934290409088135,
      "epoch": 2.956590105014882,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004141991645140228,
      "loss": 5.1134,
      "mean_token_accuracy": 0.20074480921030044,
      "num_tokens": 58771484.0,
      "step": 26325
    },
    {
      "entropy": 5.274192047119141,
      "epoch": 2.957151681922839,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041416761919568784,
      "loss": 4.9765,
      "mean_token_accuracy": 0.21077294051647186,
      "num_tokens": 58782208.0,
      "step": 26330
    },
    {
      "entropy": 5.345566511154175,
      "epoch": 2.957713258830797,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00041413606944620183,
      "loss": 5.0441,
      "mean_token_accuracy": 0.21075872629880904,
      "num_tokens": 58792556.0,
      "step": 26335
    },
    {
      "entropy": 5.291284990310669,
      "epoch": 2.9582748357387545,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004141045152665695,
      "loss": 4.9939,
      "mean_token_accuracy": 0.21843261420726776,
      "num_tokens": 58803531.0,
      "step": 26340
    },
    {
      "entropy": 5.379722690582275,
      "epoch": 2.958836412646712,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041407295665779533,
      "loss": 5.0527,
      "mean_token_accuracy": 0.20933543443679808,
      "num_tokens": 58813639.0,
      "step": 26345
    },
    {
      "entropy": 5.386694622039795,
      "epoch": 2.9593979895546694,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041404139362088434,
      "loss": 5.0321,
      "mean_token_accuracy": 0.21411840319633485,
      "num_tokens": 58824830.0,
      "step": 26350
    },
    {
      "entropy": 5.31787805557251,
      "epoch": 2.959959566462627,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004140098261568414,
      "loss": 5.0563,
      "mean_token_accuracy": 0.21120530217885972,
      "num_tokens": 58835664.0,
      "step": 26355
    },
    {
      "entropy": 5.335652780532837,
      "epoch": 2.9605211433705847,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041397825426667175,
      "loss": 5.0183,
      "mean_token_accuracy": 0.21146222352981567,
      "num_tokens": 58845872.0,
      "step": 26360
    },
    {
      "entropy": 5.262155818939209,
      "epoch": 2.961082720278542,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004139466779513805,
      "loss": 4.8608,
      "mean_token_accuracy": 0.2238917753100395,
      "num_tokens": 58857090.0,
      "step": 26365
    },
    {
      "entropy": 5.3513171672821045,
      "epoch": 2.9616442971864996,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004139150972119732,
      "loss": 5.0522,
      "mean_token_accuracy": 0.20936914086341857,
      "num_tokens": 58868814.0,
      "step": 26370
    },
    {
      "entropy": 5.2399321556091305,
      "epoch": 2.9622058740944572,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041388351204945534,
      "loss": 4.9709,
      "mean_token_accuracy": 0.2103540852665901,
      "num_tokens": 58879270.0,
      "step": 26375
    },
    {
      "entropy": 5.379551362991333,
      "epoch": 2.962767451002415,
      "grad_norm": 0.953125,
      "learning_rate": 0.00041385192246483257,
      "loss": 5.0727,
      "mean_token_accuracy": 0.2107746809720993,
      "num_tokens": 58891861.0,
      "step": 26380
    },
    {
      "entropy": 5.3613653659820555,
      "epoch": 2.9633290279103726,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041382032845911073,
      "loss": 5.0621,
      "mean_token_accuracy": 0.21044692695140838,
      "num_tokens": 58903779.0,
      "step": 26385
    },
    {
      "entropy": 5.421366024017334,
      "epoch": 2.9638906048183298,
      "grad_norm": 0.921875,
      "learning_rate": 0.00041378873003329575,
      "loss": 5.0905,
      "mean_token_accuracy": 0.2085845947265625,
      "num_tokens": 58915773.0,
      "step": 26390
    },
    {
      "entropy": 5.303115367889404,
      "epoch": 2.9644521817262874,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004137571271883938,
      "loss": 5.0167,
      "mean_token_accuracy": 0.21243275701999664,
      "num_tokens": 58926701.0,
      "step": 26395
    },
    {
      "entropy": 5.287404203414917,
      "epoch": 2.965013758634245,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004137255199254112,
      "loss": 4.9346,
      "mean_token_accuracy": 0.2194040060043335,
      "num_tokens": 58937606.0,
      "step": 26400
    },
    {
      "entropy": 5.403477334976197,
      "epoch": 2.9655753355422023,
      "grad_norm": 1.0,
      "learning_rate": 0.00041369390824535417,
      "loss": 5.083,
      "mean_token_accuracy": 0.21493049710988998,
      "num_tokens": 58948982.0,
      "step": 26405
    },
    {
      "entropy": 5.345158052444458,
      "epoch": 2.96613691245016,
      "grad_norm": 0.921875,
      "learning_rate": 0.00041366229214922926,
      "loss": 5.0171,
      "mean_token_accuracy": 0.21974304914474488,
      "num_tokens": 58961607.0,
      "step": 26410
    },
    {
      "entropy": 5.29091215133667,
      "epoch": 2.9666984893581176,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004136306716380432,
      "loss": 4.9733,
      "mean_token_accuracy": 0.21405725926160812,
      "num_tokens": 58973072.0,
      "step": 26415
    },
    {
      "entropy": 5.299877643585205,
      "epoch": 2.9672600662660753,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041359904671280285,
      "loss": 4.9854,
      "mean_token_accuracy": 0.21464429795742035,
      "num_tokens": 58983330.0,
      "step": 26420
    },
    {
      "entropy": 5.43519434928894,
      "epoch": 2.9678216431740325,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004135674173745151,
      "loss": 5.1785,
      "mean_token_accuracy": 0.2011588454246521,
      "num_tokens": 58995912.0,
      "step": 26425
    },
    {
      "entropy": 5.435634565353394,
      "epoch": 2.96838322008199,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041353578362418695,
      "loss": 5.1573,
      "mean_token_accuracy": 0.20858098119497298,
      "num_tokens": 59005964.0,
      "step": 26430
    },
    {
      "entropy": 5.421501398086548,
      "epoch": 2.968944796989948,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004135041454628258,
      "loss": 5.1495,
      "mean_token_accuracy": 0.20441606044769287,
      "num_tokens": 59017285.0,
      "step": 26435
    },
    {
      "entropy": 5.2298136234283445,
      "epoch": 2.9695063738979055,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004134725028914389,
      "loss": 4.9429,
      "mean_token_accuracy": 0.2222428098320961,
      "num_tokens": 59027751.0,
      "step": 26440
    },
    {
      "entropy": 5.327543163299561,
      "epoch": 2.9700679508058627,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041344085591103383,
      "loss": 5.0897,
      "mean_token_accuracy": 0.20767239332199097,
      "num_tokens": 59039568.0,
      "step": 26445
    },
    {
      "entropy": 5.376237487792968,
      "epoch": 2.9706295277138204,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004134092045226182,
      "loss": 5.027,
      "mean_token_accuracy": 0.20906975567340852,
      "num_tokens": 59051162.0,
      "step": 26450
    },
    {
      "entropy": 5.30674934387207,
      "epoch": 2.971191104621778,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041337754872719984,
      "loss": 4.9258,
      "mean_token_accuracy": 0.21368396878242493,
      "num_tokens": 59061381.0,
      "step": 26455
    },
    {
      "entropy": 5.2869462966918945,
      "epoch": 2.9717526815297353,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004133458885257866,
      "loss": 4.9289,
      "mean_token_accuracy": 0.2126012220978737,
      "num_tokens": 59071026.0,
      "step": 26460
    },
    {
      "entropy": 5.264104557037354,
      "epoch": 2.972314258437693,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041331422391938663,
      "loss": 4.9817,
      "mean_token_accuracy": 0.2103909894824028,
      "num_tokens": 59082793.0,
      "step": 26465
    },
    {
      "entropy": 5.453774356842041,
      "epoch": 2.9728758353456506,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004132825549090082,
      "loss": 5.2022,
      "mean_token_accuracy": 0.20695194005966186,
      "num_tokens": 59094247.0,
      "step": 26470
    },
    {
      "entropy": 5.272838020324707,
      "epoch": 2.9734374122536082,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004132508814956595,
      "loss": 4.9748,
      "mean_token_accuracy": 0.2104943349957466,
      "num_tokens": 59104936.0,
      "step": 26475
    },
    {
      "entropy": 5.333774518966675,
      "epoch": 2.973998989161566,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004132192036803491,
      "loss": 5.0148,
      "mean_token_accuracy": 0.21572229862213135,
      "num_tokens": 59116181.0,
      "step": 26480
    },
    {
      "entropy": 5.205937337875366,
      "epoch": 2.974560566069523,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041318752146408575,
      "loss": 4.9439,
      "mean_token_accuracy": 0.21616362482309343,
      "num_tokens": 59126953.0,
      "step": 26485
    },
    {
      "entropy": 5.3302473545074465,
      "epoch": 2.975122142977481,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004131558348478781,
      "loss": 5.1258,
      "mean_token_accuracy": 0.20943909585475923,
      "num_tokens": 59138460.0,
      "step": 26490
    },
    {
      "entropy": 5.4159019947052,
      "epoch": 2.9756837198854384,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041312414383273507,
      "loss": 5.0298,
      "mean_token_accuracy": 0.21508723348379136,
      "num_tokens": 59149003.0,
      "step": 26495
    },
    {
      "entropy": 5.378587770462036,
      "epoch": 2.9762452967933957,
      "grad_norm": 0.9375,
      "learning_rate": 0.00041309244841966565,
      "loss": 5.0211,
      "mean_token_accuracy": 0.21039779037237166,
      "num_tokens": 59160582.0,
      "step": 26500
    },
    {
      "entropy": 5.347186136245727,
      "epoch": 2.9768068737013533,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00041306074860967916,
      "loss": 5.1128,
      "mean_token_accuracy": 0.20167977660894393,
      "num_tokens": 59172115.0,
      "step": 26505
    },
    {
      "entropy": 5.284080123901367,
      "epoch": 2.977368450609311,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041302904440378494,
      "loss": 4.9584,
      "mean_token_accuracy": 0.2138582244515419,
      "num_tokens": 59183044.0,
      "step": 26510
    },
    {
      "entropy": 5.386540746688842,
      "epoch": 2.9779300275172687,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041299733580299236,
      "loss": 5.1462,
      "mean_token_accuracy": 0.19994013607501984,
      "num_tokens": 59194791.0,
      "step": 26515
    },
    {
      "entropy": 5.249656677246094,
      "epoch": 2.978491604425226,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00041296562280831106,
      "loss": 4.9839,
      "mean_token_accuracy": 0.21176043152809143,
      "num_tokens": 59206285.0,
      "step": 26520
    },
    {
      "entropy": 5.379307413101197,
      "epoch": 2.9790531813331835,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041293390542075095,
      "loss": 5.0879,
      "mean_token_accuracy": 0.20855608731508254,
      "num_tokens": 59217906.0,
      "step": 26525
    },
    {
      "entropy": 5.352666854858398,
      "epoch": 2.979614758241141,
      "grad_norm": 1.0,
      "learning_rate": 0.0004129021836413217,
      "loss": 5.0729,
      "mean_token_accuracy": 0.20929300785064697,
      "num_tokens": 59228163.0,
      "step": 26530
    },
    {
      "entropy": 5.278653526306153,
      "epoch": 2.980176335149099,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004128704574710334,
      "loss": 5.0314,
      "mean_token_accuracy": 0.21051753014326097,
      "num_tokens": 59238633.0,
      "step": 26535
    },
    {
      "entropy": 5.371733140945435,
      "epoch": 2.980737912057056,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004128387269108963,
      "loss": 5.0548,
      "mean_token_accuracy": 0.208394418656826,
      "num_tokens": 59249523.0,
      "step": 26540
    },
    {
      "entropy": 5.405754041671753,
      "epoch": 2.9812994889650137,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00041280699196192074,
      "loss": 5.1328,
      "mean_token_accuracy": 0.2035241261124611,
      "num_tokens": 59259636.0,
      "step": 26545
    },
    {
      "entropy": 5.325657892227173,
      "epoch": 2.9818610658729714,
      "grad_norm": 0.953125,
      "learning_rate": 0.00041277525262511705,
      "loss": 5.0274,
      "mean_token_accuracy": 0.21189655661582946,
      "num_tokens": 59270963.0,
      "step": 26550
    },
    {
      "entropy": 5.389918088912964,
      "epoch": 2.9824226427809286,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041274350890149584,
      "loss": 5.1077,
      "mean_token_accuracy": 0.20693572461605073,
      "num_tokens": 59282085.0,
      "step": 26555
    },
    {
      "entropy": 5.40252571105957,
      "epoch": 2.9829842196888863,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000412711760792068,
      "loss": 5.0663,
      "mean_token_accuracy": 0.21347056925296784,
      "num_tokens": 59292210.0,
      "step": 26560
    },
    {
      "entropy": 5.344774627685547,
      "epoch": 2.983545796596844,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004126800082978442,
      "loss": 5.087,
      "mean_token_accuracy": 0.2118708610534668,
      "num_tokens": 59303727.0,
      "step": 26565
    },
    {
      "entropy": 5.3288124084472654,
      "epoch": 2.9841073735048016,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041264825141983547,
      "loss": 4.9868,
      "mean_token_accuracy": 0.21567689776420593,
      "num_tokens": 59313922.0,
      "step": 26570
    },
    {
      "entropy": 5.306971597671509,
      "epoch": 2.9846689504127593,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00041261649015905314,
      "loss": 5.0058,
      "mean_token_accuracy": 0.2148689314723015,
      "num_tokens": 59324974.0,
      "step": 26575
    },
    {
      "entropy": 5.372596645355225,
      "epoch": 2.9852305273207165,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004125847245165084,
      "loss": 5.0461,
      "mean_token_accuracy": 0.20748182386159897,
      "num_tokens": 59335538.0,
      "step": 26580
    },
    {
      "entropy": 5.375613832473755,
      "epoch": 2.985792104228674,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004125529544932126,
      "loss": 5.1234,
      "mean_token_accuracy": 0.2091645672917366,
      "num_tokens": 59347116.0,
      "step": 26585
    },
    {
      "entropy": 5.350049018859863,
      "epoch": 2.986353681136632,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004125211800901774,
      "loss": 5.0404,
      "mean_token_accuracy": 0.21377675980329514,
      "num_tokens": 59358505.0,
      "step": 26590
    },
    {
      "entropy": 5.308620595932007,
      "epoch": 2.986915258044589,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004124894013084145,
      "loss": 4.9518,
      "mean_token_accuracy": 0.2191245898604393,
      "num_tokens": 59369897.0,
      "step": 26595
    },
    {
      "entropy": 5.326073026657104,
      "epoch": 2.9874768349525467,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041245761814893576,
      "loss": 5.048,
      "mean_token_accuracy": 0.21281567364931106,
      "num_tokens": 59381372.0,
      "step": 26600
    },
    {
      "entropy": 5.346283721923828,
      "epoch": 2.9880384118605043,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004124258306127531,
      "loss": 4.9922,
      "mean_token_accuracy": 0.21277335584163665,
      "num_tokens": 59391543.0,
      "step": 26605
    },
    {
      "entropy": 5.369219779968262,
      "epoch": 2.988599988768462,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004123940387008787,
      "loss": 5.0683,
      "mean_token_accuracy": 0.21338210701942445,
      "num_tokens": 59402474.0,
      "step": 26610
    },
    {
      "entropy": 5.320281744003296,
      "epoch": 2.9891615656764197,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004123622424143248,
      "loss": 4.9862,
      "mean_token_accuracy": 0.21388629972934722,
      "num_tokens": 59413970.0,
      "step": 26615
    },
    {
      "entropy": 5.255481481552124,
      "epoch": 2.989723142584377,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041233044175410387,
      "loss": 4.9661,
      "mean_token_accuracy": 0.2175861418247223,
      "num_tokens": 59424958.0,
      "step": 26620
    },
    {
      "entropy": 5.323951196670532,
      "epoch": 2.9902847194923345,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041229863672122844,
      "loss": 5.0406,
      "mean_token_accuracy": 0.21179508566856384,
      "num_tokens": 59436226.0,
      "step": 26625
    },
    {
      "entropy": 5.438098430633545,
      "epoch": 2.990846296400292,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004122668273167111,
      "loss": 5.1214,
      "mean_token_accuracy": 0.20454955101013184,
      "num_tokens": 59447607.0,
      "step": 26630
    },
    {
      "entropy": 5.249334907531738,
      "epoch": 2.9914078733082494,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004122350135415648,
      "loss": 4.9227,
      "mean_token_accuracy": 0.21553991436958314,
      "num_tokens": 59458120.0,
      "step": 26635
    },
    {
      "entropy": 5.334640741348267,
      "epoch": 2.991969450216207,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004122031953968024,
      "loss": 5.0749,
      "mean_token_accuracy": 0.20959312319755555,
      "num_tokens": 59468646.0,
      "step": 26640
    },
    {
      "entropy": 5.347740840911865,
      "epoch": 2.9925310271241647,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041217137288343704,
      "loss": 5.0409,
      "mean_token_accuracy": 0.21030667126178743,
      "num_tokens": 59478887.0,
      "step": 26645
    },
    {
      "entropy": 5.383488750457763,
      "epoch": 2.993092604032122,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000412139546002482,
      "loss": 5.0837,
      "mean_token_accuracy": 0.2099412888288498,
      "num_tokens": 59490501.0,
      "step": 26650
    },
    {
      "entropy": 5.385051727294922,
      "epoch": 2.9936541809400796,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041210771475495064,
      "loss": 5.135,
      "mean_token_accuracy": 0.20791876912117005,
      "num_tokens": 59500873.0,
      "step": 26655
    },
    {
      "entropy": 5.2837423324584964,
      "epoch": 2.9942157578480373,
      "grad_norm": 1.0,
      "learning_rate": 0.0004120758791418564,
      "loss": 4.9697,
      "mean_token_accuracy": 0.2144717574119568,
      "num_tokens": 59511793.0,
      "step": 26660
    },
    {
      "entropy": 5.369166994094849,
      "epoch": 2.994777334755995,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004120440391642131,
      "loss": 5.1028,
      "mean_token_accuracy": 0.20607990473508836,
      "num_tokens": 59523978.0,
      "step": 26665
    },
    {
      "entropy": 5.3786773681640625,
      "epoch": 2.9953389116639526,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004120121948230343,
      "loss": 5.1241,
      "mean_token_accuracy": 0.2093312621116638,
      "num_tokens": 59534575.0,
      "step": 26670
    },
    {
      "entropy": 5.3693195343017575,
      "epoch": 2.99590048857191,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004119803461193342,
      "loss": 5.0576,
      "mean_token_accuracy": 0.20249320119619368,
      "num_tokens": 59545542.0,
      "step": 26675
    },
    {
      "entropy": 5.361839294433594,
      "epoch": 2.9964620654798675,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004119484930541267,
      "loss": 5.0119,
      "mean_token_accuracy": 0.2148972913622856,
      "num_tokens": 59556482.0,
      "step": 26680
    },
    {
      "entropy": 5.3414225578308105,
      "epoch": 2.997023642387825,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041191663562842605,
      "loss": 5.029,
      "mean_token_accuracy": 0.2117735043168068,
      "num_tokens": 59568941.0,
      "step": 26685
    },
    {
      "entropy": 5.367746496200562,
      "epoch": 2.9975852192957824,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004118847738432467,
      "loss": 5.1002,
      "mean_token_accuracy": 0.2058937981724739,
      "num_tokens": 59579690.0,
      "step": 26690
    },
    {
      "entropy": 5.337344217300415,
      "epoch": 2.99814679620374,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004118529076996029,
      "loss": 5.0637,
      "mean_token_accuracy": 0.2090964511036873,
      "num_tokens": 59589831.0,
      "step": 26695
    },
    {
      "entropy": 5.309319496154785,
      "epoch": 2.9987083731116977,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041182103719850967,
      "loss": 5.0336,
      "mean_token_accuracy": 0.20450652986764908,
      "num_tokens": 59600939.0,
      "step": 26700
    },
    {
      "entropy": 5.247100877761841,
      "epoch": 2.9992699500196554,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00041178916234098134,
      "loss": 4.9057,
      "mean_token_accuracy": 0.22068696469068527,
      "num_tokens": 59612084.0,
      "step": 26705
    },
    {
      "entropy": 5.351404047012329,
      "epoch": 2.999831526927613,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004117572831280332,
      "loss": 5.0746,
      "mean_token_accuracy": 0.2036201313138008,
      "num_tokens": 59624359.0,
      "step": 26710
    },
    {
      "entropy": 5.4379922019110785,
      "epoch": 3.0003369461447744,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00041172539956068,
      "loss": 5.0497,
      "mean_token_accuracy": 0.21474692225456238,
      "num_tokens": 59633030.0,
      "step": 26715
    },
    {
      "entropy": 5.269414567947388,
      "epoch": 3.000898523052732,
      "grad_norm": 0.953125,
      "learning_rate": 0.00041169351163993705,
      "loss": 4.8971,
      "mean_token_accuracy": 0.2179547667503357,
      "num_tokens": 59644107.0,
      "step": 26720
    },
    {
      "entropy": 5.241611909866333,
      "epoch": 3.0014600999606897,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041166161936681973,
      "loss": 4.7974,
      "mean_token_accuracy": 0.226563560962677,
      "num_tokens": 59654166.0,
      "step": 26725
    },
    {
      "entropy": 5.313516283035279,
      "epoch": 3.002021676868647,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004116297227423435,
      "loss": 4.9341,
      "mean_token_accuracy": 0.21424570530653,
      "num_tokens": 59664785.0,
      "step": 26730
    },
    {
      "entropy": 5.253862190246582,
      "epoch": 3.0025832537766046,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041159782176752375,
      "loss": 4.8355,
      "mean_token_accuracy": 0.2253906711935997,
      "num_tokens": 59675634.0,
      "step": 26735
    },
    {
      "entropy": 5.286987352371216,
      "epoch": 3.0031448306845623,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004115659164433765,
      "loss": 4.9826,
      "mean_token_accuracy": 0.2203440099954605,
      "num_tokens": 59687636.0,
      "step": 26740
    },
    {
      "entropy": 5.345514297485352,
      "epoch": 3.00370640759252,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00041153400677091745,
      "loss": 4.9426,
      "mean_token_accuracy": 0.2128476694226265,
      "num_tokens": 59700389.0,
      "step": 26745
    },
    {
      "entropy": 5.283603382110596,
      "epoch": 3.004267984500477,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041150209275116266,
      "loss": 4.8784,
      "mean_token_accuracy": 0.2250046342611313,
      "num_tokens": 59710898.0,
      "step": 26750
    },
    {
      "entropy": 5.326199007034302,
      "epoch": 3.004829561408435,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041147017438512834,
      "loss": 4.9571,
      "mean_token_accuracy": 0.21079932898283005,
      "num_tokens": 59722398.0,
      "step": 26755
    },
    {
      "entropy": 5.31957483291626,
      "epoch": 3.0053911383163925,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041143825167383073,
      "loss": 4.94,
      "mean_token_accuracy": 0.21636795848608018,
      "num_tokens": 59732504.0,
      "step": 26760
    },
    {
      "entropy": 5.328311824798584,
      "epoch": 3.00595271522435,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004114063246182863,
      "loss": 4.9556,
      "mean_token_accuracy": 0.21731124818325043,
      "num_tokens": 59742919.0,
      "step": 26765
    },
    {
      "entropy": 5.3198826789855955,
      "epoch": 3.0065142921323074,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041137439321951147,
      "loss": 4.9602,
      "mean_token_accuracy": 0.21630615293979644,
      "num_tokens": 59754316.0,
      "step": 26770
    },
    {
      "entropy": 5.369644689559936,
      "epoch": 3.007075869040265,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041134245747852314,
      "loss": 5.033,
      "mean_token_accuracy": 0.2077586144208908,
      "num_tokens": 59765031.0,
      "step": 26775
    },
    {
      "entropy": 5.323687648773193,
      "epoch": 3.0076374459482227,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004113105173963382,
      "loss": 4.9836,
      "mean_token_accuracy": 0.21194494664669036,
      "num_tokens": 59776338.0,
      "step": 26780
    },
    {
      "entropy": 5.32881350517273,
      "epoch": 3.0081990228561803,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004112785729739733,
      "loss": 4.9246,
      "mean_token_accuracy": 0.22019748836755754,
      "num_tokens": 59788544.0,
      "step": 26785
    },
    {
      "entropy": 5.313954257965088,
      "epoch": 3.0087605997641376,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041124662421244593,
      "loss": 4.9231,
      "mean_token_accuracy": 0.2096757173538208,
      "num_tokens": 59800131.0,
      "step": 26790
    },
    {
      "entropy": 5.320450067520142,
      "epoch": 3.0093221766720952,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004112146711127732,
      "loss": 4.957,
      "mean_token_accuracy": 0.2164208933711052,
      "num_tokens": 59811986.0,
      "step": 26795
    },
    {
      "entropy": 5.352895927429199,
      "epoch": 3.009883753580053,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004111827136759725,
      "loss": 5.0157,
      "mean_token_accuracy": 0.21345575004816056,
      "num_tokens": 59824758.0,
      "step": 26800
    },
    {
      "entropy": 5.233724975585938,
      "epoch": 3.0104453304880106,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004111507519030613,
      "loss": 4.9298,
      "mean_token_accuracy": 0.21804398149251938,
      "num_tokens": 59836317.0,
      "step": 26805
    },
    {
      "entropy": 5.361664056777954,
      "epoch": 3.0110069073959678,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004111187857950574,
      "loss": 5.007,
      "mean_token_accuracy": 0.21567562371492385,
      "num_tokens": 59847743.0,
      "step": 26810
    },
    {
      "entropy": 5.355305147171021,
      "epoch": 3.0115684843039254,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004110868153529786,
      "loss": 4.9624,
      "mean_token_accuracy": 0.2203311488032341,
      "num_tokens": 59859442.0,
      "step": 26815
    },
    {
      "entropy": 5.26543698310852,
      "epoch": 3.012130061211883,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004110548405778428,
      "loss": 4.869,
      "mean_token_accuracy": 0.2195791184902191,
      "num_tokens": 59869949.0,
      "step": 26820
    },
    {
      "entropy": 5.267998552322387,
      "epoch": 3.0126916381198403,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004110228614706681,
      "loss": 4.9163,
      "mean_token_accuracy": 0.21699672937393188,
      "num_tokens": 59881818.0,
      "step": 26825
    },
    {
      "entropy": 5.30213680267334,
      "epoch": 3.013253215027798,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041099087803247276,
      "loss": 4.8841,
      "mean_token_accuracy": 0.21668126434087753,
      "num_tokens": 59893152.0,
      "step": 26830
    },
    {
      "entropy": 5.399004983901977,
      "epoch": 3.0138147919357556,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004109588902642751,
      "loss": 5.0301,
      "mean_token_accuracy": 0.20301027745008468,
      "num_tokens": 59904082.0,
      "step": 26835
    },
    {
      "entropy": 5.316755056381226,
      "epoch": 3.0143763688437133,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004109268981670936,
      "loss": 4.9698,
      "mean_token_accuracy": 0.21860734969377518,
      "num_tokens": 59916116.0,
      "step": 26840
    },
    {
      "entropy": 5.194122409820556,
      "epoch": 3.0149379457516705,
      "grad_norm": 1.0,
      "learning_rate": 0.000410894901741947,
      "loss": 4.8327,
      "mean_token_accuracy": 0.22336750626564025,
      "num_tokens": 59926399.0,
      "step": 26845
    },
    {
      "entropy": 5.288911437988281,
      "epoch": 3.015499522659628,
      "grad_norm": 1.015625,
      "learning_rate": 0.000410862900989854,
      "loss": 4.9175,
      "mean_token_accuracy": 0.21900468617677687,
      "num_tokens": 59936986.0,
      "step": 26850
    },
    {
      "entropy": 5.37878007888794,
      "epoch": 3.016061099567586,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004108308959118336,
      "loss": 5.0127,
      "mean_token_accuracy": 0.20667029023170472,
      "num_tokens": 59948926.0,
      "step": 26855
    },
    {
      "entropy": 5.235599422454834,
      "epoch": 3.0166226764755435,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004107988865089047,
      "loss": 4.8641,
      "mean_token_accuracy": 0.22187516391277312,
      "num_tokens": 59960468.0,
      "step": 26860
    },
    {
      "entropy": 5.31623272895813,
      "epoch": 3.0171842533835007,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041076687278208664,
      "loss": 4.9489,
      "mean_token_accuracy": 0.21648863106966018,
      "num_tokens": 59971151.0,
      "step": 26865
    },
    {
      "entropy": 5.279011011123657,
      "epoch": 3.0177458302914584,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004107348547323987,
      "loss": 4.871,
      "mean_token_accuracy": 0.22052381783723832,
      "num_tokens": 59982205.0,
      "step": 26870
    },
    {
      "entropy": 5.232743263244629,
      "epoch": 3.018307407199416,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041070283236086024,
      "loss": 4.8989,
      "mean_token_accuracy": 0.2235160529613495,
      "num_tokens": 59993612.0,
      "step": 26875
    },
    {
      "entropy": 5.417162561416626,
      "epoch": 3.0188689841073737,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004106708056684911,
      "loss": 5.074,
      "mean_token_accuracy": 0.20721004754304886,
      "num_tokens": 60004284.0,
      "step": 26880
    },
    {
      "entropy": 5.41718635559082,
      "epoch": 3.019430561015331,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00041063877465631084,
      "loss": 5.0573,
      "mean_token_accuracy": 0.21042075455188752,
      "num_tokens": 60016276.0,
      "step": 26885
    },
    {
      "entropy": 5.2669905662536625,
      "epoch": 3.0199921379232886,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004106067393253394,
      "loss": 4.9737,
      "mean_token_accuracy": 0.21721131801605226,
      "num_tokens": 60029155.0,
      "step": 26890
    },
    {
      "entropy": 5.2647796154022215,
      "epoch": 3.0205537148312462,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004105746996765968,
      "loss": 4.9523,
      "mean_token_accuracy": 0.2105598658323288,
      "num_tokens": 60041042.0,
      "step": 26895
    },
    {
      "entropy": 5.389530897140503,
      "epoch": 3.021115291739204,
      "grad_norm": 1.0,
      "learning_rate": 0.0004105426557111031,
      "loss": 4.9945,
      "mean_token_accuracy": 0.21618689596652985,
      "num_tokens": 60052025.0,
      "step": 26900
    },
    {
      "entropy": 5.304889345169068,
      "epoch": 3.021676868647161,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004105106074298787,
      "loss": 4.9498,
      "mean_token_accuracy": 0.21734809279441833,
      "num_tokens": 60063260.0,
      "step": 26905
    },
    {
      "entropy": 5.3016040325164795,
      "epoch": 3.022238445555119,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000410478554833944,
      "loss": 4.979,
      "mean_token_accuracy": 0.20921106934547423,
      "num_tokens": 60074721.0,
      "step": 26910
    },
    {
      "entropy": 5.298282957077026,
      "epoch": 3.0228000224630764,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041044649792431963,
      "loss": 4.9085,
      "mean_token_accuracy": 0.2187381938099861,
      "num_tokens": 60085028.0,
      "step": 26915
    },
    {
      "entropy": 5.260143136978149,
      "epoch": 3.0233615993710337,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041041443670202616,
      "loss": 4.7967,
      "mean_token_accuracy": 0.22405229657888412,
      "num_tokens": 60095587.0,
      "step": 26920
    },
    {
      "entropy": 5.237364816665649,
      "epoch": 3.0239231762789913,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004103823711680846,
      "loss": 4.8037,
      "mean_token_accuracy": 0.22293294072151185,
      "num_tokens": 60106574.0,
      "step": 26925
    },
    {
      "entropy": 5.223855686187744,
      "epoch": 3.024484753186949,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00041035030132351575,
      "loss": 4.8556,
      "mean_token_accuracy": 0.2216453269124031,
      "num_tokens": 60116821.0,
      "step": 26930
    },
    {
      "entropy": 5.314420604705811,
      "epoch": 3.0250463300949066,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00041031822716934086,
      "loss": 4.9919,
      "mean_token_accuracy": 0.2140338107943535,
      "num_tokens": 60128229.0,
      "step": 26935
    },
    {
      "entropy": 5.297655391693115,
      "epoch": 3.025607907002864,
      "grad_norm": 1.125,
      "learning_rate": 0.00041028614870658106,
      "loss": 4.8958,
      "mean_token_accuracy": 0.21322874277830123,
      "num_tokens": 60139416.0,
      "step": 26940
    },
    {
      "entropy": 5.292050552368164,
      "epoch": 3.0261694839108215,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004102540659362579,
      "loss": 4.8961,
      "mean_token_accuracy": 0.21729740351438523,
      "num_tokens": 60150225.0,
      "step": 26945
    },
    {
      "entropy": 5.314477396011353,
      "epoch": 3.026731060818779,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041022197885939285,
      "loss": 4.8311,
      "mean_token_accuracy": 0.2150799348950386,
      "num_tokens": 60160065.0,
      "step": 26950
    },
    {
      "entropy": 5.343855190277099,
      "epoch": 3.027292637726737,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041018988747700745,
      "loss": 4.9456,
      "mean_token_accuracy": 0.2239023208618164,
      "num_tokens": 60170396.0,
      "step": 26955
    },
    {
      "entropy": 5.257840585708618,
      "epoch": 3.027854214634694,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004101577917901236,
      "loss": 4.9309,
      "mean_token_accuracy": 0.21504322737455367,
      "num_tokens": 60181938.0,
      "step": 26960
    },
    {
      "entropy": 5.267040967941284,
      "epoch": 3.0284157915426517,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041012569179976326,
      "loss": 4.8868,
      "mean_token_accuracy": 0.214192296564579,
      "num_tokens": 60192388.0,
      "step": 26965
    },
    {
      "entropy": 5.359966897964478,
      "epoch": 3.0289773684506094,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004100935875069485,
      "loss": 4.984,
      "mean_token_accuracy": 0.2132430598139763,
      "num_tokens": 60203195.0,
      "step": 26970
    },
    {
      "entropy": 5.338379812240601,
      "epoch": 3.029538945358567,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004100614789127015,
      "loss": 4.9979,
      "mean_token_accuracy": 0.21412320137023927,
      "num_tokens": 60213659.0,
      "step": 26975
    },
    {
      "entropy": 5.22256064414978,
      "epoch": 3.0301005222665243,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004100293660180447,
      "loss": 4.9375,
      "mean_token_accuracy": 0.21520572453737258,
      "num_tokens": 60224795.0,
      "step": 26980
    },
    {
      "entropy": 5.217825698852539,
      "epoch": 3.030662099174482,
      "grad_norm": 1.0,
      "learning_rate": 0.00040999724882400046,
      "loss": 4.8315,
      "mean_token_accuracy": 0.22389816343784333,
      "num_tokens": 60236188.0,
      "step": 26985
    },
    {
      "entropy": 5.308682298660278,
      "epoch": 3.0312236760824396,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004099651273315915,
      "loss": 4.9373,
      "mean_token_accuracy": 0.21589406579732895,
      "num_tokens": 60247003.0,
      "step": 26990
    },
    {
      "entropy": 5.297772407531738,
      "epoch": 3.0317852529903973,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004099330015418405,
      "loss": 4.9382,
      "mean_token_accuracy": 0.21247612982988356,
      "num_tokens": 60258317.0,
      "step": 26995
    },
    {
      "entropy": 5.365457010269165,
      "epoch": 3.0323468298983545,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040990087145577044,
      "loss": 4.9639,
      "mean_token_accuracy": 0.2177238568663597,
      "num_tokens": 60269000.0,
      "step": 27000
    },
    {
      "epoch": 3.0323468298983545,
      "eval_entropy": 5.150028362205295,
      "eval_loss": 5.143547534942627,
      "eval_mean_token_accuracy": 0.21435529654684649,
      "eval_num_tokens": 60269000.0,
      "eval_runtime": 24.082,
      "eval_samples_per_second": 1287.81,
      "eval_steps_per_second": 160.992,
      "step": 27000
    },
    {
      "entropy": 5.230336427688599,
      "epoch": 3.032908406806312,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004098687370744042,
      "loss": 4.8294,
      "mean_token_accuracy": 0.21852117627859116,
      "num_tokens": 60280411.0,
      "step": 27005
    },
    {
      "entropy": 5.3599577903747555,
      "epoch": 3.03346998371427,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004098365983987652,
      "loss": 4.9959,
      "mean_token_accuracy": 0.20946189314126967,
      "num_tokens": 60290197.0,
      "step": 27010
    },
    {
      "entropy": 5.366162824630737,
      "epoch": 3.034031560622227,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040980445542987653,
      "loss": 4.8936,
      "mean_token_accuracy": 0.21951605677604674,
      "num_tokens": 60300927.0,
      "step": 27015
    },
    {
      "entropy": 5.243457698822022,
      "epoch": 3.0345931375301847,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00040977230816876176,
      "loss": 4.8818,
      "mean_token_accuracy": 0.2215030774474144,
      "num_tokens": 60312697.0,
      "step": 27020
    },
    {
      "entropy": 5.195396947860718,
      "epoch": 3.0351547144381423,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040974015661644433,
      "loss": 4.8216,
      "mean_token_accuracy": 0.22851794511079787,
      "num_tokens": 60323779.0,
      "step": 27025
    },
    {
      "entropy": 5.30742769241333,
      "epoch": 3.0357162913461,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004097080007739481,
      "loss": 4.9616,
      "mean_token_accuracy": 0.2185770332813263,
      "num_tokens": 60335036.0,
      "step": 27030
    },
    {
      "entropy": 5.289138221740723,
      "epoch": 3.036277868254057,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040967584064229695,
      "loss": 4.9298,
      "mean_token_accuracy": 0.20884300619363785,
      "num_tokens": 60345600.0,
      "step": 27035
    },
    {
      "entropy": 5.402020978927612,
      "epoch": 3.036839445162015,
      "grad_norm": 0.953125,
      "learning_rate": 0.00040964367622251473,
      "loss": 4.9837,
      "mean_token_accuracy": 0.21180517077445984,
      "num_tokens": 60357241.0,
      "step": 27040
    },
    {
      "entropy": 5.3151263236999515,
      "epoch": 3.0374010220699725,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004096115075156257,
      "loss": 4.9132,
      "mean_token_accuracy": 0.2132827490568161,
      "num_tokens": 60369036.0,
      "step": 27045
    },
    {
      "entropy": 5.350251579284668,
      "epoch": 3.03796259897793,
      "grad_norm": 1.0,
      "learning_rate": 0.0004095793345226541,
      "loss": 5.0245,
      "mean_token_accuracy": 0.21397764682769777,
      "num_tokens": 60380310.0,
      "step": 27050
    },
    {
      "entropy": 5.264749479293823,
      "epoch": 3.0385241758858874,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004095471572446242,
      "loss": 4.8575,
      "mean_token_accuracy": 0.22144055217504502,
      "num_tokens": 60390962.0,
      "step": 27055
    },
    {
      "entropy": 5.413425159454346,
      "epoch": 3.039085752793845,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004095149756825606,
      "loss": 4.9806,
      "mean_token_accuracy": 0.21541517674922944,
      "num_tokens": 60400342.0,
      "step": 27060
    },
    {
      "entropy": 5.252880239486695,
      "epoch": 3.0396473297018027,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004094827898374881,
      "loss": 4.8843,
      "mean_token_accuracy": 0.22313380986452103,
      "num_tokens": 60411921.0,
      "step": 27065
    },
    {
      "entropy": 5.426985836029052,
      "epoch": 3.0402089066097604,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004094505997104314,
      "loss": 5.0233,
      "mean_token_accuracy": 0.2073667272925377,
      "num_tokens": 60424185.0,
      "step": 27070
    },
    {
      "entropy": 5.145985078811646,
      "epoch": 3.0407704835177176,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040941840530241537,
      "loss": 4.8424,
      "mean_token_accuracy": 0.21536551117897035,
      "num_tokens": 60433729.0,
      "step": 27075
    },
    {
      "entropy": 5.356107091903686,
      "epoch": 3.0413320604256753,
      "grad_norm": 1.125,
      "learning_rate": 0.00040938620661446526,
      "loss": 4.982,
      "mean_token_accuracy": 0.21535283774137498,
      "num_tokens": 60443846.0,
      "step": 27080
    },
    {
      "entropy": 5.319244146347046,
      "epoch": 3.041893637333633,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004093540036476062,
      "loss": 4.9675,
      "mean_token_accuracy": 0.21455367058515548,
      "num_tokens": 60455183.0,
      "step": 27085
    },
    {
      "entropy": 5.263688707351685,
      "epoch": 3.0424552142415906,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040932179640286356,
      "loss": 4.9258,
      "mean_token_accuracy": 0.2141054630279541,
      "num_tokens": 60465769.0,
      "step": 27090
    },
    {
      "entropy": 5.2442834854125975,
      "epoch": 3.043016791149548,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004092895848812628,
      "loss": 4.8246,
      "mean_token_accuracy": 0.22411901205778123,
      "num_tokens": 60476252.0,
      "step": 27095
    },
    {
      "entropy": 5.327517080307007,
      "epoch": 3.0435783680575055,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040925736908382953,
      "loss": 4.9062,
      "mean_token_accuracy": 0.21649682223796846,
      "num_tokens": 60485939.0,
      "step": 27100
    },
    {
      "entropy": 5.211687231063843,
      "epoch": 3.044139944965463,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004092251490115896,
      "loss": 4.9441,
      "mean_token_accuracy": 0.21656700372695922,
      "num_tokens": 60497994.0,
      "step": 27105
    },
    {
      "entropy": 5.464140796661377,
      "epoch": 3.0447015218734204,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040919292466556895,
      "loss": 5.0786,
      "mean_token_accuracy": 0.21322320997714997,
      "num_tokens": 60508944.0,
      "step": 27110
    },
    {
      "entropy": 5.368603658676148,
      "epoch": 3.045263098781378,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004091606960467934,
      "loss": 4.9909,
      "mean_token_accuracy": 0.22192550599575042,
      "num_tokens": 60519928.0,
      "step": 27115
    },
    {
      "entropy": 5.282523155212402,
      "epoch": 3.0458246756893357,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040912846315628934,
      "loss": 4.8981,
      "mean_token_accuracy": 0.22321904599666595,
      "num_tokens": 60530417.0,
      "step": 27120
    },
    {
      "entropy": 5.3004913330078125,
      "epoch": 3.0463862525972933,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000409096225995083,
      "loss": 4.9875,
      "mean_token_accuracy": 0.21457773000001906,
      "num_tokens": 60542140.0,
      "step": 27125
    },
    {
      "entropy": 5.30202579498291,
      "epoch": 3.0469478295052506,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00040906398456420077,
      "loss": 4.8776,
      "mean_token_accuracy": 0.2261667102575302,
      "num_tokens": 60553032.0,
      "step": 27130
    },
    {
      "entropy": 5.304340934753418,
      "epoch": 3.047509406413208,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040903173886466934,
      "loss": 4.9901,
      "mean_token_accuracy": 0.22290573865175248,
      "num_tokens": 60563773.0,
      "step": 27135
    },
    {
      "entropy": 5.202450370788574,
      "epoch": 3.048070983321166,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040899948889751525,
      "loss": 4.87,
      "mean_token_accuracy": 0.219879387319088,
      "num_tokens": 60574354.0,
      "step": 27140
    },
    {
      "entropy": 5.322264385223389,
      "epoch": 3.0486325602291235,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040896723466376557,
      "loss": 4.9243,
      "mean_token_accuracy": 0.21532915085554122,
      "num_tokens": 60584760.0,
      "step": 27145
    },
    {
      "entropy": 5.139884519577026,
      "epoch": 3.0491941371370808,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040893497616444714,
      "loss": 4.8065,
      "mean_token_accuracy": 0.22429149448871613,
      "num_tokens": 60595322.0,
      "step": 27150
    },
    {
      "entropy": 5.3862865447998045,
      "epoch": 3.0497557140450384,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004089027134005871,
      "loss": 5.1033,
      "mean_token_accuracy": 0.20547993183135987,
      "num_tokens": 60607897.0,
      "step": 27155
    },
    {
      "entropy": 5.286964750289917,
      "epoch": 3.050317290952996,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004088704463732127,
      "loss": 4.897,
      "mean_token_accuracy": 0.21553023308515548,
      "num_tokens": 60618882.0,
      "step": 27160
    },
    {
      "entropy": 5.306805276870728,
      "epoch": 3.0508788678609537,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004088381750833514,
      "loss": 4.9576,
      "mean_token_accuracy": 0.2113136574625969,
      "num_tokens": 60630431.0,
      "step": 27165
    },
    {
      "entropy": 5.410397672653199,
      "epoch": 3.051440444768911,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004088058995320307,
      "loss": 4.9902,
      "mean_token_accuracy": 0.21239527612924575,
      "num_tokens": 60642479.0,
      "step": 27170
    },
    {
      "entropy": 5.270603513717651,
      "epoch": 3.0520020216768686,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004087736197202784,
      "loss": 4.8477,
      "mean_token_accuracy": 0.2217095285654068,
      "num_tokens": 60652922.0,
      "step": 27175
    },
    {
      "entropy": 5.275615644454956,
      "epoch": 3.0525635985848263,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040874133564912195,
      "loss": 4.9155,
      "mean_token_accuracy": 0.21909305304288865,
      "num_tokens": 60665125.0,
      "step": 27180
    },
    {
      "entropy": 5.421450996398926,
      "epoch": 3.053125175492784,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00040870904731958965,
      "loss": 5.0459,
      "mean_token_accuracy": 0.21330163031816482,
      "num_tokens": 60675807.0,
      "step": 27185
    },
    {
      "entropy": 5.270182371139526,
      "epoch": 3.053686752400741,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004086767547327094,
      "loss": 4.8709,
      "mean_token_accuracy": 0.22407983392477035,
      "num_tokens": 60686841.0,
      "step": 27190
    },
    {
      "entropy": 5.283323955535889,
      "epoch": 3.054248329308699,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00040864445788950935,
      "loss": 4.9576,
      "mean_token_accuracy": 0.21665271371603012,
      "num_tokens": 60698965.0,
      "step": 27195
    },
    {
      "entropy": 5.35181565284729,
      "epoch": 3.0548099062166565,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000408612156791018,
      "loss": 5.0374,
      "mean_token_accuracy": 0.21865007728338243,
      "num_tokens": 60709962.0,
      "step": 27200
    },
    {
      "entropy": 5.304647397994995,
      "epoch": 3.0553714831246137,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004085798514382637,
      "loss": 4.8222,
      "mean_token_accuracy": 0.22066616714000703,
      "num_tokens": 60721267.0,
      "step": 27205
    },
    {
      "entropy": 5.253386163711548,
      "epoch": 3.0559330600325714,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040854754183227516,
      "loss": 4.8787,
      "mean_token_accuracy": 0.2241024851799011,
      "num_tokens": 60731892.0,
      "step": 27210
    },
    {
      "entropy": 5.274100589752197,
      "epoch": 3.056494636940529,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004085152279740811,
      "loss": 4.9517,
      "mean_token_accuracy": 0.20822966396808623,
      "num_tokens": 60742875.0,
      "step": 27215
    },
    {
      "entropy": 5.380665874481201,
      "epoch": 3.0570562138484867,
      "grad_norm": 1.125,
      "learning_rate": 0.00040848290986471037,
      "loss": 4.9502,
      "mean_token_accuracy": 0.2127927303314209,
      "num_tokens": 60752716.0,
      "step": 27220
    },
    {
      "entropy": 5.30497465133667,
      "epoch": 3.057617790756444,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040845058750519205,
      "loss": 5.034,
      "mean_token_accuracy": 0.20735326856374742,
      "num_tokens": 60765466.0,
      "step": 27225
    },
    {
      "entropy": 5.335328245162964,
      "epoch": 3.0581793676644016,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040841826089655525,
      "loss": 4.9767,
      "mean_token_accuracy": 0.2067878484725952,
      "num_tokens": 60776488.0,
      "step": 27230
    },
    {
      "entropy": 5.366192817687988,
      "epoch": 3.0587409445723592,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040838593003982926,
      "loss": 4.961,
      "mean_token_accuracy": 0.2156498685479164,
      "num_tokens": 60786510.0,
      "step": 27235
    },
    {
      "entropy": 5.265503072738648,
      "epoch": 3.059302521480317,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004083535949360436,
      "loss": 4.883,
      "mean_token_accuracy": 0.2297059878706932,
      "num_tokens": 60798300.0,
      "step": 27240
    },
    {
      "entropy": 5.1637578964233395,
      "epoch": 3.059864098388274,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004083212555862277,
      "loss": 4.8529,
      "mean_token_accuracy": 0.21547932028770447,
      "num_tokens": 60809949.0,
      "step": 27245
    },
    {
      "entropy": 5.267165088653565,
      "epoch": 3.0604256752962318,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004082889119914113,
      "loss": 4.8978,
      "mean_token_accuracy": 0.2151651233434677,
      "num_tokens": 60819960.0,
      "step": 27250
    },
    {
      "entropy": 5.375799083709717,
      "epoch": 3.0609872522041894,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004082565641526242,
      "loss": 4.9282,
      "mean_token_accuracy": 0.22332373708486558,
      "num_tokens": 60830934.0,
      "step": 27255
    },
    {
      "entropy": 5.240466737747193,
      "epoch": 3.061548829112147,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00040822421207089654,
      "loss": 4.9136,
      "mean_token_accuracy": 0.21303803473711014,
      "num_tokens": 60842161.0,
      "step": 27260
    },
    {
      "entropy": 5.321337413787842,
      "epoch": 3.0621104060201043,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004081918557472582,
      "loss": 4.9735,
      "mean_token_accuracy": 0.21772877275943756,
      "num_tokens": 60854220.0,
      "step": 27265
    },
    {
      "entropy": 5.2937671661376955,
      "epoch": 3.062671982928062,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004081594951827396,
      "loss": 4.9208,
      "mean_token_accuracy": 0.22723539173603058,
      "num_tokens": 60864769.0,
      "step": 27270
    },
    {
      "entropy": 5.27568359375,
      "epoch": 3.0632335598360196,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040812713037837096,
      "loss": 4.8635,
      "mean_token_accuracy": 0.2268289655447006,
      "num_tokens": 60874865.0,
      "step": 27275
    },
    {
      "entropy": 5.3268890380859375,
      "epoch": 3.0637951367439773,
      "grad_norm": 1.0,
      "learning_rate": 0.00040809476133518284,
      "loss": 4.9415,
      "mean_token_accuracy": 0.21088396459817887,
      "num_tokens": 60885099.0,
      "step": 27280
    },
    {
      "entropy": 5.3378276348114015,
      "epoch": 3.0643567136519345,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040806238805420595,
      "loss": 5.0058,
      "mean_token_accuracy": 0.212557652592659,
      "num_tokens": 60896071.0,
      "step": 27285
    },
    {
      "entropy": 5.359554052352905,
      "epoch": 3.064918290559892,
      "grad_norm": 0.953125,
      "learning_rate": 0.000408030010536471,
      "loss": 5.0301,
      "mean_token_accuracy": 0.2107207730412483,
      "num_tokens": 60907255.0,
      "step": 27290
    },
    {
      "entropy": 5.280068349838257,
      "epoch": 3.06547986746785,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040799762878300895,
      "loss": 4.8272,
      "mean_token_accuracy": 0.22804595530033112,
      "num_tokens": 60918740.0,
      "step": 27295
    },
    {
      "entropy": 5.220046091079712,
      "epoch": 3.066041444375807,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00040796524279485084,
      "loss": 4.8425,
      "mean_token_accuracy": 0.22506999969482422,
      "num_tokens": 60929327.0,
      "step": 27300
    },
    {
      "entropy": 5.253331804275513,
      "epoch": 3.0666030212837647,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040793285257302776,
      "loss": 4.9024,
      "mean_token_accuracy": 0.21883469372987746,
      "num_tokens": 60939987.0,
      "step": 27305
    },
    {
      "entropy": 5.419541931152343,
      "epoch": 3.0671645981917224,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000407900458118571,
      "loss": 5.0449,
      "mean_token_accuracy": 0.21678775697946548,
      "num_tokens": 60951924.0,
      "step": 27310
    },
    {
      "entropy": 5.337277793884278,
      "epoch": 3.06772617509968,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004078680594325123,
      "loss": 4.9673,
      "mean_token_accuracy": 0.2141156405210495,
      "num_tokens": 60963706.0,
      "step": 27315
    },
    {
      "entropy": 5.3058006286621096,
      "epoch": 3.0682877520076373,
      "grad_norm": 1.125,
      "learning_rate": 0.000407835656515883,
      "loss": 4.9673,
      "mean_token_accuracy": 0.21544882655143738,
      "num_tokens": 60974842.0,
      "step": 27320
    },
    {
      "entropy": 5.261927270889283,
      "epoch": 3.068849328915595,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040780324936971486,
      "loss": 4.8756,
      "mean_token_accuracy": 0.2226244330406189,
      "num_tokens": 60984921.0,
      "step": 27325
    },
    {
      "entropy": 5.306805801391602,
      "epoch": 3.0694109058235526,
      "grad_norm": 1.0,
      "learning_rate": 0.0004077708379950397,
      "loss": 4.9659,
      "mean_token_accuracy": 0.21855118572711946,
      "num_tokens": 60995648.0,
      "step": 27330
    },
    {
      "entropy": 5.255483341217041,
      "epoch": 3.0699724827315102,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004077384223928896,
      "loss": 4.9452,
      "mean_token_accuracy": 0.21209691911935807,
      "num_tokens": 61008020.0,
      "step": 27335
    },
    {
      "entropy": 5.3065159797668455,
      "epoch": 3.0705340596394675,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004077060025642967,
      "loss": 4.8797,
      "mean_token_accuracy": 0.21952322125434875,
      "num_tokens": 61017735.0,
      "step": 27340
    },
    {
      "entropy": 5.286941432952881,
      "epoch": 3.071095636547425,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004076735785102931,
      "loss": 5.0253,
      "mean_token_accuracy": 0.20989513099193574,
      "num_tokens": 61029111.0,
      "step": 27345
    },
    {
      "entropy": 5.334425067901611,
      "epoch": 3.071657213455383,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004076411502319114,
      "loss": 4.9199,
      "mean_token_accuracy": 0.2171369656920433,
      "num_tokens": 61039392.0,
      "step": 27350
    },
    {
      "entropy": 5.264988899230957,
      "epoch": 3.0722187903633404,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040760871773018405,
      "loss": 4.9171,
      "mean_token_accuracy": 0.22324508875608445,
      "num_tokens": 61049681.0,
      "step": 27355
    },
    {
      "entropy": 5.288323831558228,
      "epoch": 3.0727803672712977,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004075762810061437,
      "loss": 4.8855,
      "mean_token_accuracy": 0.21921051740646363,
      "num_tokens": 61059910.0,
      "step": 27360
    },
    {
      "entropy": 5.3519594192504885,
      "epoch": 3.0733419441792553,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000407543840060823,
      "loss": 4.9979,
      "mean_token_accuracy": 0.21527708768844606,
      "num_tokens": 61071074.0,
      "step": 27365
    },
    {
      "entropy": 5.310376071929932,
      "epoch": 3.073903521087213,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004075113948952551,
      "loss": 4.8924,
      "mean_token_accuracy": 0.2206498146057129,
      "num_tokens": 61082051.0,
      "step": 27370
    },
    {
      "entropy": 5.314424896240235,
      "epoch": 3.0744650979951706,
      "grad_norm": 1.0,
      "learning_rate": 0.0004074789455104731,
      "loss": 4.9546,
      "mean_token_accuracy": 0.21640318483114243,
      "num_tokens": 61093578.0,
      "step": 27375
    },
    {
      "entropy": 5.258527040481567,
      "epoch": 3.075026674903128,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040744649190750995,
      "loss": 4.9067,
      "mean_token_accuracy": 0.217627677321434,
      "num_tokens": 61105192.0,
      "step": 27380
    },
    {
      "entropy": 5.304828453063965,
      "epoch": 3.0755882518110855,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004074140340873992,
      "loss": 4.9935,
      "mean_token_accuracy": 0.21506838649511337,
      "num_tokens": 61116659.0,
      "step": 27385
    },
    {
      "entropy": 5.369470024108887,
      "epoch": 3.076149828719043,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004073815720511743,
      "loss": 4.9383,
      "mean_token_accuracy": 0.21842813342809678,
      "num_tokens": 61127761.0,
      "step": 27390
    },
    {
      "entropy": 5.3375328540802,
      "epoch": 3.0767114056270004,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00040734910579986875,
      "loss": 4.8576,
      "mean_token_accuracy": 0.22861295640468599,
      "num_tokens": 61139039.0,
      "step": 27395
    },
    {
      "entropy": 5.301487684249878,
      "epoch": 3.077272982534958,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040731663533451624,
      "loss": 4.9338,
      "mean_token_accuracy": 0.21790089458227158,
      "num_tokens": 61149877.0,
      "step": 27400
    },
    {
      "entropy": 5.363093566894531,
      "epoch": 3.0778345594429157,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040728416065615087,
      "loss": 4.9501,
      "mean_token_accuracy": 0.2167642056941986,
      "num_tokens": 61160806.0,
      "step": 27405
    },
    {
      "entropy": 5.329862642288208,
      "epoch": 3.0783961363508734,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004072516817658065,
      "loss": 5.0646,
      "mean_token_accuracy": 0.20687309503555298,
      "num_tokens": 61171471.0,
      "step": 27410
    },
    {
      "entropy": 5.331740760803223,
      "epoch": 3.0789577132588306,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040721919866451724,
      "loss": 4.9557,
      "mean_token_accuracy": 0.21108760684728622,
      "num_tokens": 61182140.0,
      "step": 27415
    },
    {
      "entropy": 5.2758911609649655,
      "epoch": 3.0795192901667883,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004071867113533174,
      "loss": 4.9161,
      "mean_token_accuracy": 0.21684859842061996,
      "num_tokens": 61192532.0,
      "step": 27420
    },
    {
      "entropy": 5.313822174072266,
      "epoch": 3.080080867074746,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00040715421983324147,
      "loss": 5.0152,
      "mean_token_accuracy": 0.21687484979629518,
      "num_tokens": 61205207.0,
      "step": 27425
    },
    {
      "entropy": 5.406329107284546,
      "epoch": 3.0806424439827036,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040712172410532397,
      "loss": 5.0284,
      "mean_token_accuracy": 0.20764925926923752,
      "num_tokens": 61218039.0,
      "step": 27430
    },
    {
      "entropy": 5.392036819458008,
      "epoch": 3.081204020890661,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004070892241705994,
      "loss": 4.9869,
      "mean_token_accuracy": 0.2152500256896019,
      "num_tokens": 61230343.0,
      "step": 27435
    },
    {
      "entropy": 5.235243272781372,
      "epoch": 3.0817655977986185,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040705672003010275,
      "loss": 4.8648,
      "mean_token_accuracy": 0.21943992525339126,
      "num_tokens": 61240781.0,
      "step": 27440
    },
    {
      "entropy": 5.236502552032471,
      "epoch": 3.082327174706576,
      "grad_norm": 1.0,
      "learning_rate": 0.000407024211684869,
      "loss": 4.9209,
      "mean_token_accuracy": 0.21854648143053054,
      "num_tokens": 61250638.0,
      "step": 27445
    },
    {
      "entropy": 5.334246015548706,
      "epoch": 3.082888751614534,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040699169913593306,
      "loss": 4.9494,
      "mean_token_accuracy": 0.21322983205318452,
      "num_tokens": 61261431.0,
      "step": 27450
    },
    {
      "entropy": 5.374189233779907,
      "epoch": 3.083450328522491,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004069591823843302,
      "loss": 5.0479,
      "mean_token_accuracy": 0.2074557363986969,
      "num_tokens": 61272726.0,
      "step": 27455
    },
    {
      "entropy": 5.383172416687012,
      "epoch": 3.0840119054304487,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004069266614310959,
      "loss": 4.9696,
      "mean_token_accuracy": 0.2147562712430954,
      "num_tokens": 61283779.0,
      "step": 27460
    },
    {
      "entropy": 5.3639589786529545,
      "epoch": 3.0845734823384063,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00040689413627726544,
      "loss": 5.0137,
      "mean_token_accuracy": 0.21950871646404266,
      "num_tokens": 61294477.0,
      "step": 27465
    },
    {
      "entropy": 5.371457481384278,
      "epoch": 3.085135059246364,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040686160692387455,
      "loss": 5.0609,
      "mean_token_accuracy": 0.20952224135398864,
      "num_tokens": 61305302.0,
      "step": 27470
    },
    {
      "entropy": 5.303634595870972,
      "epoch": 3.085696636154321,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040682907337195894,
      "loss": 4.9151,
      "mean_token_accuracy": 0.21847373247146606,
      "num_tokens": 61315499.0,
      "step": 27475
    },
    {
      "entropy": 5.376663398742676,
      "epoch": 3.086258213062279,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004067965356225545,
      "loss": 4.9547,
      "mean_token_accuracy": 0.20652424693107604,
      "num_tokens": 61326654.0,
      "step": 27480
    },
    {
      "entropy": 5.325640487670898,
      "epoch": 3.0868197899702365,
      "grad_norm": 0.921875,
      "learning_rate": 0.00040676399367669733,
      "loss": 4.9758,
      "mean_token_accuracy": 0.21220100671052933,
      "num_tokens": 61338475.0,
      "step": 27485
    },
    {
      "entropy": 5.311359786987305,
      "epoch": 3.0873813668781938,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004067314475354234,
      "loss": 5.0425,
      "mean_token_accuracy": 0.2121629998087883,
      "num_tokens": 61348748.0,
      "step": 27490
    },
    {
      "entropy": 5.2938995361328125,
      "epoch": 3.0879429437861514,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00040669889719976916,
      "loss": 4.9465,
      "mean_token_accuracy": 0.23018036782741547,
      "num_tokens": 61359466.0,
      "step": 27495
    },
    {
      "entropy": 5.291392230987549,
      "epoch": 3.088504520694109,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004066663426707709,
      "loss": 4.8645,
      "mean_token_accuracy": 0.228829924762249,
      "num_tokens": 61370136.0,
      "step": 27500
    },
    {
      "entropy": 5.255514001846313,
      "epoch": 3.0890660976020667,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040663378394946534,
      "loss": 4.9237,
      "mean_token_accuracy": 0.2132097899913788,
      "num_tokens": 61381012.0,
      "step": 27505
    },
    {
      "entropy": 5.296189641952514,
      "epoch": 3.089627674510024,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004066012210368889,
      "loss": 4.901,
      "mean_token_accuracy": 0.21901082545518874,
      "num_tokens": 61393530.0,
      "step": 27510
    },
    {
      "entropy": 5.31511435508728,
      "epoch": 3.0901892514179816,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004065686539340786,
      "loss": 4.9536,
      "mean_token_accuracy": 0.21372055560350417,
      "num_tokens": 61405753.0,
      "step": 27515
    },
    {
      "entropy": 5.241207408905029,
      "epoch": 3.0907508283259393,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040653608264207134,
      "loss": 4.8744,
      "mean_token_accuracy": 0.21866302341222763,
      "num_tokens": 61416688.0,
      "step": 27520
    },
    {
      "entropy": 5.408501148223877,
      "epoch": 3.091312405233897,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040650350716190417,
      "loss": 5.0008,
      "mean_token_accuracy": 0.21564358323812485,
      "num_tokens": 61427377.0,
      "step": 27525
    },
    {
      "entropy": 5.304213428497315,
      "epoch": 3.091873982141854,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004064709274946143,
      "loss": 4.8794,
      "mean_token_accuracy": 0.22299712002277375,
      "num_tokens": 61438457.0,
      "step": 27530
    },
    {
      "entropy": 5.169082403182983,
      "epoch": 3.092435559049812,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004064383436412392,
      "loss": 4.9007,
      "mean_token_accuracy": 0.2192373052239418,
      "num_tokens": 61449380.0,
      "step": 27535
    },
    {
      "entropy": 5.305188608169556,
      "epoch": 3.0929971359577695,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004064057556028162,
      "loss": 4.9737,
      "mean_token_accuracy": 0.21509650945663453,
      "num_tokens": 61460846.0,
      "step": 27540
    },
    {
      "entropy": 5.3858559131622314,
      "epoch": 3.093558712865727,
      "grad_norm": 1.171875,
      "learning_rate": 0.000406373163380383,
      "loss": 4.9329,
      "mean_token_accuracy": 0.21567509025335313,
      "num_tokens": 61471593.0,
      "step": 27545
    },
    {
      "entropy": 5.294015502929687,
      "epoch": 3.0941202897736844,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040634056697497726,
      "loss": 4.9651,
      "mean_token_accuracy": 0.21430791318416595,
      "num_tokens": 61483643.0,
      "step": 27550
    },
    {
      "entropy": 5.262836265563965,
      "epoch": 3.094681866681642,
      "grad_norm": 1.015625,
      "learning_rate": 0.000406307966387637,
      "loss": 4.8851,
      "mean_token_accuracy": 0.2160701870918274,
      "num_tokens": 61493220.0,
      "step": 27555
    },
    {
      "entropy": 5.311472320556641,
      "epoch": 3.0952434435895997,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040627536161940016,
      "loss": 4.9262,
      "mean_token_accuracy": 0.21581867039203645,
      "num_tokens": 61504372.0,
      "step": 27560
    },
    {
      "entropy": 5.254740858078003,
      "epoch": 3.0958050204975573,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004062427526713049,
      "loss": 4.9306,
      "mean_token_accuracy": 0.21410764753818512,
      "num_tokens": 61515335.0,
      "step": 27565
    },
    {
      "entropy": 5.260566520690918,
      "epoch": 3.0963665974055146,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004062101395443894,
      "loss": 4.8938,
      "mean_token_accuracy": 0.21586652100086212,
      "num_tokens": 61526265.0,
      "step": 27570
    },
    {
      "entropy": 5.359550142288208,
      "epoch": 3.0969281743134722,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004061775222396922,
      "loss": 4.9612,
      "mean_token_accuracy": 0.2086852341890335,
      "num_tokens": 61538075.0,
      "step": 27575
    },
    {
      "entropy": 5.267929410934448,
      "epoch": 3.09748975122143,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004061449007582519,
      "loss": 4.9004,
      "mean_token_accuracy": 0.21796487718820573,
      "num_tokens": 61548612.0,
      "step": 27580
    },
    {
      "entropy": 5.23541579246521,
      "epoch": 3.098051328129387,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000406112275101107,
      "loss": 4.9303,
      "mean_token_accuracy": 0.2177943617105484,
      "num_tokens": 61559702.0,
      "step": 27585
    },
    {
      "entropy": 5.34126091003418,
      "epoch": 3.0986129050373448,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040607964526929655,
      "loss": 4.9166,
      "mean_token_accuracy": 0.21958177983760835,
      "num_tokens": 61569815.0,
      "step": 27590
    },
    {
      "entropy": 5.301848316192627,
      "epoch": 3.0991744819453024,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00040604701126385924,
      "loss": 4.9887,
      "mean_token_accuracy": 0.21827514767646788,
      "num_tokens": 61580050.0,
      "step": 27595
    },
    {
      "entropy": 5.321314525604248,
      "epoch": 3.09973605885326,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004060143730858342,
      "loss": 4.9986,
      "mean_token_accuracy": 0.21409133076667786,
      "num_tokens": 61590885.0,
      "step": 27600
    },
    {
      "entropy": 5.312752342224121,
      "epoch": 3.1002976357612173,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00040598173073626086,
      "loss": 4.9627,
      "mean_token_accuracy": 0.21902069598436355,
      "num_tokens": 61601633.0,
      "step": 27605
    },
    {
      "entropy": 5.383961009979248,
      "epoch": 3.100859212669175,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004059490842161783,
      "loss": 5.0744,
      "mean_token_accuracy": 0.20906985998153688,
      "num_tokens": 61613244.0,
      "step": 27610
    },
    {
      "entropy": 5.295055913925171,
      "epoch": 3.1014207895771326,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00040591643352662615,
      "loss": 4.9242,
      "mean_token_accuracy": 0.22546230256557465,
      "num_tokens": 61624939.0,
      "step": 27615
    },
    {
      "entropy": 5.227590656280517,
      "epoch": 3.1019823664850903,
      "grad_norm": 1.015625,
      "learning_rate": 0.000405883778668644,
      "loss": 4.7921,
      "mean_token_accuracy": 0.22409431040287017,
      "num_tokens": 61635480.0,
      "step": 27620
    },
    {
      "entropy": 5.291620254516602,
      "epoch": 3.1025439433930475,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040585111964327157,
      "loss": 4.9441,
      "mean_token_accuracy": 0.22119781970977784,
      "num_tokens": 61645500.0,
      "step": 27625
    },
    {
      "entropy": 5.208774995803833,
      "epoch": 3.103105520301005,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040581845645154877,
      "loss": 4.8913,
      "mean_token_accuracy": 0.22542984038591385,
      "num_tokens": 61656880.0,
      "step": 27630
    },
    {
      "entropy": 5.360792350769043,
      "epoch": 3.103667097208963,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004057857890945155,
      "loss": 4.9888,
      "mean_token_accuracy": 0.2229280725121498,
      "num_tokens": 61667451.0,
      "step": 27635
    },
    {
      "entropy": 5.39846363067627,
      "epoch": 3.1042286741169205,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00040575311757321203,
      "loss": 5.1077,
      "mean_token_accuracy": 0.21059718430042268,
      "num_tokens": 61679356.0,
      "step": 27640
    },
    {
      "entropy": 5.320131731033325,
      "epoch": 3.1047902510248777,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040572044188867863,
      "loss": 4.9357,
      "mean_token_accuracy": 0.21448320597410203,
      "num_tokens": 61690154.0,
      "step": 27645
    },
    {
      "entropy": 5.31276912689209,
      "epoch": 3.1053518279328354,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004056877620419556,
      "loss": 5.0159,
      "mean_token_accuracy": 0.21490059047937393,
      "num_tokens": 61702321.0,
      "step": 27650
    },
    {
      "entropy": 5.331673097610474,
      "epoch": 3.105913404840793,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004056550780340835,
      "loss": 4.9478,
      "mean_token_accuracy": 0.21139991879463196,
      "num_tokens": 61713498.0,
      "step": 27655
    },
    {
      "entropy": 5.474890232086182,
      "epoch": 3.1064749817487507,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004056223898661031,
      "loss": 5.0819,
      "mean_token_accuracy": 0.21090321540832518,
      "num_tokens": 61725009.0,
      "step": 27660
    },
    {
      "entropy": 5.315106678009033,
      "epoch": 3.107036558656708,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040558969753905514,
      "loss": 4.8437,
      "mean_token_accuracy": 0.22967462688684465,
      "num_tokens": 61735652.0,
      "step": 27665
    },
    {
      "entropy": 5.139658069610595,
      "epoch": 3.1075981355646656,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004055570010539806,
      "loss": 4.8215,
      "mean_token_accuracy": 0.22613579630851746,
      "num_tokens": 61746435.0,
      "step": 27670
    },
    {
      "entropy": 5.2013976097106935,
      "epoch": 3.1081597124726232,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004055243004119204,
      "loss": 4.8407,
      "mean_token_accuracy": 0.223257577419281,
      "num_tokens": 61757523.0,
      "step": 27675
    },
    {
      "entropy": 5.420965194702148,
      "epoch": 3.108721289380581,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040549159561391595,
      "loss": 5.0024,
      "mean_token_accuracy": 0.20849621891975403,
      "num_tokens": 61769309.0,
      "step": 27680
    },
    {
      "entropy": 5.312799692153931,
      "epoch": 3.109282866288538,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004054588866610085,
      "loss": 5.0091,
      "mean_token_accuracy": 0.20942789167165757,
      "num_tokens": 61781093.0,
      "step": 27685
    },
    {
      "entropy": 5.2909510135650635,
      "epoch": 3.1098444431964958,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040542617355423947,
      "loss": 4.9035,
      "mean_token_accuracy": 0.22232095450162886,
      "num_tokens": 61792837.0,
      "step": 27690
    },
    {
      "entropy": 5.360822582244873,
      "epoch": 3.1104060201044534,
      "grad_norm": 1.0,
      "learning_rate": 0.00040539345629465043,
      "loss": 4.9964,
      "mean_token_accuracy": 0.21217221468687059,
      "num_tokens": 61804212.0,
      "step": 27695
    },
    {
      "entropy": 5.33679838180542,
      "epoch": 3.1109675970124107,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040536073488328325,
      "loss": 4.97,
      "mean_token_accuracy": 0.2107241004705429,
      "num_tokens": 61815436.0,
      "step": 27700
    },
    {
      "entropy": 5.371153497695923,
      "epoch": 3.1115291739203683,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004053280093211796,
      "loss": 5.003,
      "mean_token_accuracy": 0.20977836549282075,
      "num_tokens": 61829292.0,
      "step": 27705
    },
    {
      "entropy": 5.373329210281372,
      "epoch": 3.112090750828326,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004052952796093817,
      "loss": 5.0233,
      "mean_token_accuracy": 0.21894402354955672,
      "num_tokens": 61841440.0,
      "step": 27710
    },
    {
      "entropy": 5.262772464752198,
      "epoch": 3.1126523277362836,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040526254574893154,
      "loss": 4.9505,
      "mean_token_accuracy": 0.21602265685796737,
      "num_tokens": 61853864.0,
      "step": 27715
    },
    {
      "entropy": 5.316644906997681,
      "epoch": 3.113213904644241,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040522980774087143,
      "loss": 4.9506,
      "mean_token_accuracy": 0.2213927537202835,
      "num_tokens": 61864739.0,
      "step": 27720
    },
    {
      "entropy": 5.284361219406128,
      "epoch": 3.1137754815521985,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004051970655862437,
      "loss": 4.9652,
      "mean_token_accuracy": 0.2192248985171318,
      "num_tokens": 61875940.0,
      "step": 27725
    },
    {
      "entropy": 5.308916139602661,
      "epoch": 3.114337058460156,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004051643192860908,
      "loss": 4.9164,
      "mean_token_accuracy": 0.21739348024129868,
      "num_tokens": 61887650.0,
      "step": 27730
    },
    {
      "entropy": 5.363607501983642,
      "epoch": 3.114898635368114,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040513156884145564,
      "loss": 5.0494,
      "mean_token_accuracy": 0.20939431935548783,
      "num_tokens": 61898624.0,
      "step": 27735
    },
    {
      "entropy": 5.314692878723145,
      "epoch": 3.115460212276071,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004050988142533808,
      "loss": 5.027,
      "mean_token_accuracy": 0.20567891299724578,
      "num_tokens": 61911017.0,
      "step": 27740
    },
    {
      "entropy": 5.333507919311524,
      "epoch": 3.1160217891840287,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004050660555229093,
      "loss": 4.9901,
      "mean_token_accuracy": 0.20920861512422562,
      "num_tokens": 61922381.0,
      "step": 27745
    },
    {
      "entropy": 5.430763912200928,
      "epoch": 3.1165833660919864,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004050332926510841,
      "loss": 5.052,
      "mean_token_accuracy": 0.21436668634414674,
      "num_tokens": 61932876.0,
      "step": 27750
    },
    {
      "entropy": 5.260071897506714,
      "epoch": 3.117144942999944,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004050005256389484,
      "loss": 4.8592,
      "mean_token_accuracy": 0.2262026399374008,
      "num_tokens": 61943121.0,
      "step": 27755
    },
    {
      "entropy": 5.20931921005249,
      "epoch": 3.1177065199079013,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004049677544875456,
      "loss": 4.8157,
      "mean_token_accuracy": 0.21980841159820558,
      "num_tokens": 61953988.0,
      "step": 27760
    },
    {
      "entropy": 5.374913597106934,
      "epoch": 3.118268096815859,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040493497919791906,
      "loss": 4.9741,
      "mean_token_accuracy": 0.21803593039512634,
      "num_tokens": 61965111.0,
      "step": 27765
    },
    {
      "entropy": 5.387731599807739,
      "epoch": 3.1188296737238166,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040490219977111237,
      "loss": 5.0198,
      "mean_token_accuracy": 0.22130236476659776,
      "num_tokens": 61976193.0,
      "step": 27770
    },
    {
      "entropy": 5.346433353424072,
      "epoch": 3.1193912506317742,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004048694162081693,
      "loss": 4.9996,
      "mean_token_accuracy": 0.21469646841287612,
      "num_tokens": 61987554.0,
      "step": 27775
    },
    {
      "entropy": 5.401660919189453,
      "epoch": 3.1199528275397315,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004048366285101336,
      "loss": 5.0436,
      "mean_token_accuracy": 0.20831758975982667,
      "num_tokens": 61998472.0,
      "step": 27780
    },
    {
      "entropy": 5.3193018436431885,
      "epoch": 3.120514404447689,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004048038366780494,
      "loss": 4.9212,
      "mean_token_accuracy": 0.215060992538929,
      "num_tokens": 62009368.0,
      "step": 27785
    },
    {
      "entropy": 5.278522920608521,
      "epoch": 3.121075981355647,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004047710407129606,
      "loss": 4.9258,
      "mean_token_accuracy": 0.2171493351459503,
      "num_tokens": 62021510.0,
      "step": 27790
    },
    {
      "entropy": 5.338250780105591,
      "epoch": 3.121637558263604,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004047382406159115,
      "loss": 5.0125,
      "mean_token_accuracy": 0.2174329310655594,
      "num_tokens": 62032057.0,
      "step": 27795
    },
    {
      "entropy": 5.263273429870606,
      "epoch": 3.1221991351715617,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004047054363879466,
      "loss": 4.9081,
      "mean_token_accuracy": 0.22684977054595948,
      "num_tokens": 62043185.0,
      "step": 27800
    },
    {
      "entropy": 5.415730667114258,
      "epoch": 3.1227607120795193,
      "grad_norm": 1.0,
      "learning_rate": 0.00040467262803011025,
      "loss": 4.999,
      "mean_token_accuracy": 0.2148878440260887,
      "num_tokens": 62053483.0,
      "step": 27805
    },
    {
      "entropy": 5.34845118522644,
      "epoch": 3.123322288987477,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040463981554344716,
      "loss": 5.049,
      "mean_token_accuracy": 0.20764924138784407,
      "num_tokens": 62065721.0,
      "step": 27810
    },
    {
      "entropy": 5.289376878738404,
      "epoch": 3.123883865895434,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040460699892900197,
      "loss": 4.9193,
      "mean_token_accuracy": 0.224539615213871,
      "num_tokens": 62076251.0,
      "step": 27815
    },
    {
      "entropy": 5.312393856048584,
      "epoch": 3.124445442803392,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004045741781878198,
      "loss": 4.937,
      "mean_token_accuracy": 0.22161087691783904,
      "num_tokens": 62086791.0,
      "step": 27820
    },
    {
      "entropy": 5.2697412967681885,
      "epoch": 3.1250070197113495,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040454135332094543,
      "loss": 4.8006,
      "mean_token_accuracy": 0.2244540974497795,
      "num_tokens": 62097054.0,
      "step": 27825
    },
    {
      "entropy": 5.278116750717163,
      "epoch": 3.125568596619307,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004045085243294242,
      "loss": 4.9486,
      "mean_token_accuracy": 0.22188085019588472,
      "num_tokens": 62108304.0,
      "step": 27830
    },
    {
      "entropy": 5.301598930358887,
      "epoch": 3.1261301735272644,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004044756912143013,
      "loss": 5.011,
      "mean_token_accuracy": 0.21143115013837815,
      "num_tokens": 62119724.0,
      "step": 27835
    },
    {
      "entropy": 5.258556222915649,
      "epoch": 3.126691750435222,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00040444285397662225,
      "loss": 4.875,
      "mean_token_accuracy": 0.22809136062860488,
      "num_tokens": 62131021.0,
      "step": 27840
    },
    {
      "entropy": 5.38786473274231,
      "epoch": 3.1272533273431797,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004044100126174324,
      "loss": 5.0102,
      "mean_token_accuracy": 0.20755075961351394,
      "num_tokens": 62141846.0,
      "step": 27845
    },
    {
      "entropy": 5.406996870040894,
      "epoch": 3.1278149042511374,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040437716713777756,
      "loss": 5.0749,
      "mean_token_accuracy": 0.21134277135133744,
      "num_tokens": 62152922.0,
      "step": 27850
    },
    {
      "entropy": 5.366918563842773,
      "epoch": 3.1283764811590946,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004043443175387036,
      "loss": 5.0279,
      "mean_token_accuracy": 0.2126530274748802,
      "num_tokens": 62164645.0,
      "step": 27855
    },
    {
      "entropy": 5.273736572265625,
      "epoch": 3.1289380580670523,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040431146382125644,
      "loss": 5.0201,
      "mean_token_accuracy": 0.20763671547174453,
      "num_tokens": 62175910.0,
      "step": 27860
    },
    {
      "entropy": 5.22418475151062,
      "epoch": 3.12949963497501,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000404278605986482,
      "loss": 4.8641,
      "mean_token_accuracy": 0.22282831072807313,
      "num_tokens": 62187411.0,
      "step": 27865
    },
    {
      "entropy": 5.236042594909668,
      "epoch": 3.1300612118829676,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004042457440354267,
      "loss": 4.8782,
      "mean_token_accuracy": 0.22392542362213136,
      "num_tokens": 62199316.0,
      "step": 27870
    },
    {
      "entropy": 5.434413433074951,
      "epoch": 3.130622788790925,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004042128779691368,
      "loss": 5.0351,
      "mean_token_accuracy": 0.21640237271785737,
      "num_tokens": 62211864.0,
      "step": 27875
    },
    {
      "entropy": 5.2917557716369625,
      "epoch": 3.1311843656988825,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004041800077886587,
      "loss": 4.8871,
      "mean_token_accuracy": 0.2264878049492836,
      "num_tokens": 62221715.0,
      "step": 27880
    },
    {
      "entropy": 5.304490566253662,
      "epoch": 3.13174594260684,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00040414713349503907,
      "loss": 5.0069,
      "mean_token_accuracy": 0.21608723402023317,
      "num_tokens": 62233198.0,
      "step": 27885
    },
    {
      "entropy": 5.310071849822998,
      "epoch": 3.1323075195147974,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040411425508932457,
      "loss": 4.922,
      "mean_token_accuracy": 0.21629306375980378,
      "num_tokens": 62244149.0,
      "step": 27890
    },
    {
      "entropy": 5.411935806274414,
      "epoch": 3.132869096422755,
      "grad_norm": 1.0,
      "learning_rate": 0.00040408137257256225,
      "loss": 4.9831,
      "mean_token_accuracy": 0.20838623493909836,
      "num_tokens": 62254994.0,
      "step": 27895
    },
    {
      "entropy": 5.374716520309448,
      "epoch": 3.1334306733307127,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004040484859457988,
      "loss": 5.0421,
      "mean_token_accuracy": 0.21223976612091064,
      "num_tokens": 62265685.0,
      "step": 27900
    },
    {
      "entropy": 5.25061411857605,
      "epoch": 3.1339922502386703,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040401559521008155,
      "loss": 4.9229,
      "mean_token_accuracy": 0.21944621652364732,
      "num_tokens": 62275765.0,
      "step": 27905
    },
    {
      "entropy": 5.225871944427491,
      "epoch": 3.1345538271466276,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040398270036645775,
      "loss": 4.9158,
      "mean_token_accuracy": 0.21517620980739594,
      "num_tokens": 62286020.0,
      "step": 27910
    },
    {
      "entropy": 5.329073047637939,
      "epoch": 3.135115404054585,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004039498014159748,
      "loss": 4.9261,
      "mean_token_accuracy": 0.21685650795698166,
      "num_tokens": 62296231.0,
      "step": 27915
    },
    {
      "entropy": 5.328163051605225,
      "epoch": 3.135676980962543,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040391689835968004,
      "loss": 4.9049,
      "mean_token_accuracy": 0.21769960820674897,
      "num_tokens": 62307243.0,
      "step": 27920
    },
    {
      "entropy": 5.304602146148682,
      "epoch": 3.1362385578705005,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004038839911986213,
      "loss": 4.9639,
      "mean_token_accuracy": 0.2127692297101021,
      "num_tokens": 62319061.0,
      "step": 27925
    },
    {
      "entropy": 5.358406209945679,
      "epoch": 3.1368001347784578,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004038510799338462,
      "loss": 5.0121,
      "mean_token_accuracy": 0.21512772142887115,
      "num_tokens": 62329477.0,
      "step": 27930
    },
    {
      "entropy": 5.32354507446289,
      "epoch": 3.1373617116864154,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004038181645664029,
      "loss": 4.9386,
      "mean_token_accuracy": 0.22190093696117402,
      "num_tokens": 62339170.0,
      "step": 27935
    },
    {
      "entropy": 5.391398286819458,
      "epoch": 3.137923288594373,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00040378524509733916,
      "loss": 5.0367,
      "mean_token_accuracy": 0.21365013867616653,
      "num_tokens": 62350548.0,
      "step": 27940
    },
    {
      "entropy": 5.236664819717407,
      "epoch": 3.1384848655023307,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004037523215277033,
      "loss": 4.8902,
      "mean_token_accuracy": 0.21546908915042878,
      "num_tokens": 62361105.0,
      "step": 27945
    },
    {
      "entropy": 5.242595100402832,
      "epoch": 3.139046442410288,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004037193938585436,
      "loss": 4.9508,
      "mean_token_accuracy": 0.2184287741780281,
      "num_tokens": 62372048.0,
      "step": 27950
    },
    {
      "entropy": 5.3218296527862545,
      "epoch": 3.1396080193182456,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004036864620909085,
      "loss": 4.9895,
      "mean_token_accuracy": 0.21113411039114,
      "num_tokens": 62382936.0,
      "step": 27955
    },
    {
      "entropy": 5.31208553314209,
      "epoch": 3.1401695962262033,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004036535262258465,
      "loss": 4.8707,
      "mean_token_accuracy": 0.22054540663957595,
      "num_tokens": 62393656.0,
      "step": 27960
    },
    {
      "entropy": 5.300800848007202,
      "epoch": 3.140731173134161,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040362058626440637,
      "loss": 4.9856,
      "mean_token_accuracy": 0.21813572943210602,
      "num_tokens": 62404234.0,
      "step": 27965
    },
    {
      "entropy": 5.352295541763306,
      "epoch": 3.141292750042118,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00040358764220763685,
      "loss": 4.9386,
      "mean_token_accuracy": 0.216964490711689,
      "num_tokens": 62416694.0,
      "step": 27970
    },
    {
      "entropy": 5.313505744934082,
      "epoch": 3.141854326950076,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00040355469405658687,
      "loss": 4.9352,
      "mean_token_accuracy": 0.21523544043302537,
      "num_tokens": 62429437.0,
      "step": 27975
    },
    {
      "entropy": 5.307492303848266,
      "epoch": 3.1424159038580335,
      "grad_norm": 1.0,
      "learning_rate": 0.0004035217418123056,
      "loss": 4.9136,
      "mean_token_accuracy": 0.22241295576095582,
      "num_tokens": 62440085.0,
      "step": 27980
    },
    {
      "entropy": 5.338489770889282,
      "epoch": 3.1429774807659907,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040348878547584227,
      "loss": 4.9652,
      "mean_token_accuracy": 0.21554743200540544,
      "num_tokens": 62451285.0,
      "step": 27985
    },
    {
      "entropy": 5.293754720687867,
      "epoch": 3.1435390576739484,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040345582504824624,
      "loss": 4.9394,
      "mean_token_accuracy": 0.2195811837911606,
      "num_tokens": 62461777.0,
      "step": 27990
    },
    {
      "entropy": 5.320055437088013,
      "epoch": 3.144100634581906,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004034228605305668,
      "loss": 4.9822,
      "mean_token_accuracy": 0.22391709089279174,
      "num_tokens": 62472463.0,
      "step": 27995
    },
    {
      "entropy": 5.239036941528321,
      "epoch": 3.1446622114898637,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040338989192385377,
      "loss": 4.9173,
      "mean_token_accuracy": 0.2166106715798378,
      "num_tokens": 62484168.0,
      "step": 28000
    },
    {
      "entropy": 5.3417500972747805,
      "epoch": 3.145223788397821,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004033569192291567,
      "loss": 4.9708,
      "mean_token_accuracy": 0.21552836149930954,
      "num_tokens": 62495187.0,
      "step": 28005
    },
    {
      "entropy": 5.335152006149292,
      "epoch": 3.1457853653057786,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004033239424475256,
      "loss": 5.046,
      "mean_token_accuracy": 0.20768278539180757,
      "num_tokens": 62505936.0,
      "step": 28010
    },
    {
      "entropy": 5.392810964584351,
      "epoch": 3.1463469422137362,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040329096158001033,
      "loss": 4.995,
      "mean_token_accuracy": 0.21215295046567917,
      "num_tokens": 62516486.0,
      "step": 28015
    },
    {
      "entropy": 5.310181903839111,
      "epoch": 3.146908519121694,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004032579766276612,
      "loss": 4.8848,
      "mean_token_accuracy": 0.22163988500833512,
      "num_tokens": 62528349.0,
      "step": 28020
    },
    {
      "entropy": 5.172423696517944,
      "epoch": 3.147470096029651,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040322498759152824,
      "loss": 4.887,
      "mean_token_accuracy": 0.21939153522253035,
      "num_tokens": 62538698.0,
      "step": 28025
    },
    {
      "entropy": 5.355631589889526,
      "epoch": 3.1480316729376088,
      "grad_norm": 1.0,
      "learning_rate": 0.00040319199447266205,
      "loss": 5.0803,
      "mean_token_accuracy": 0.20554948598146439,
      "num_tokens": 62550891.0,
      "step": 28030
    },
    {
      "entropy": 5.392551279067993,
      "epoch": 3.1485932498455664,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004031589972721129,
      "loss": 4.9348,
      "mean_token_accuracy": 0.21315138787031174,
      "num_tokens": 62560553.0,
      "step": 28035
    },
    {
      "entropy": 5.3573626518249515,
      "epoch": 3.149154826753524,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004031259959909317,
      "loss": 4.9185,
      "mean_token_accuracy": 0.21822001188993453,
      "num_tokens": 62571281.0,
      "step": 28040
    },
    {
      "entropy": 5.214624977111816,
      "epoch": 3.1497164036614813,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000403092990630169,
      "loss": 4.9195,
      "mean_token_accuracy": 0.21687147617340088,
      "num_tokens": 62583489.0,
      "step": 28045
    },
    {
      "entropy": 5.217158603668213,
      "epoch": 3.150277980569439,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004030599811908758,
      "loss": 4.8861,
      "mean_token_accuracy": 0.22636463344097138,
      "num_tokens": 62594155.0,
      "step": 28050
    },
    {
      "entropy": 5.301469373703003,
      "epoch": 3.1508395574773966,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004030269676741032,
      "loss": 4.8971,
      "mean_token_accuracy": 0.22320511937141418,
      "num_tokens": 62605464.0,
      "step": 28055
    },
    {
      "entropy": 5.281586456298828,
      "epoch": 3.1514011343853543,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004029939500809023,
      "loss": 4.9587,
      "mean_token_accuracy": 0.21641321927309037,
      "num_tokens": 62617343.0,
      "step": 28060
    },
    {
      "entropy": 5.298098754882813,
      "epoch": 3.1519627112933115,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004029609284123243,
      "loss": 4.9463,
      "mean_token_accuracy": 0.22181383073329924,
      "num_tokens": 62628461.0,
      "step": 28065
    },
    {
      "entropy": 5.320556020736694,
      "epoch": 3.152524288201269,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00040292790266942074,
      "loss": 4.9079,
      "mean_token_accuracy": 0.2221237912774086,
      "num_tokens": 62638947.0,
      "step": 28070
    },
    {
      "entropy": 5.212317609786988,
      "epoch": 3.153085865109227,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040289487285324324,
      "loss": 4.775,
      "mean_token_accuracy": 0.2276085615158081,
      "num_tokens": 62648840.0,
      "step": 28075
    },
    {
      "entropy": 5.333423852920532,
      "epoch": 3.153647442017184,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004028618389648433,
      "loss": 5.0411,
      "mean_token_accuracy": 0.2093634322285652,
      "num_tokens": 62659491.0,
      "step": 28080
    },
    {
      "entropy": 5.381044101715088,
      "epoch": 3.1542090189251417,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004028288010052728,
      "loss": 5.0074,
      "mean_token_accuracy": 0.20852453559637069,
      "num_tokens": 62669038.0,
      "step": 28085
    },
    {
      "entropy": 5.263435316085816,
      "epoch": 3.1547705958330994,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004027957589755837,
      "loss": 4.8828,
      "mean_token_accuracy": 0.22822125852108002,
      "num_tokens": 62680280.0,
      "step": 28090
    },
    {
      "entropy": 5.329967927932739,
      "epoch": 3.155332172741057,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040276271287682813,
      "loss": 4.9605,
      "mean_token_accuracy": 0.220234714448452,
      "num_tokens": 62691767.0,
      "step": 28095
    },
    {
      "entropy": 5.366234970092774,
      "epoch": 3.1558937496490143,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00040272966271005816,
      "loss": 4.9467,
      "mean_token_accuracy": 0.21550538390874863,
      "num_tokens": 62701721.0,
      "step": 28100
    },
    {
      "entropy": 5.3306293964385985,
      "epoch": 3.156455326556972,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040269660847632626,
      "loss": 4.9691,
      "mean_token_accuracy": 0.21246879547834396,
      "num_tokens": 62713238.0,
      "step": 28105
    },
    {
      "entropy": 5.217879581451416,
      "epoch": 3.1570169034649296,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00040266355017668475,
      "loss": 4.8421,
      "mean_token_accuracy": 0.2237108126282692,
      "num_tokens": 62724457.0,
      "step": 28110
    },
    {
      "entropy": 5.304288530349732,
      "epoch": 3.1575784803728872,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004026304878121864,
      "loss": 4.9797,
      "mean_token_accuracy": 0.22102373093366623,
      "num_tokens": 62735178.0,
      "step": 28115
    },
    {
      "entropy": 5.314140844345093,
      "epoch": 3.1581400572808445,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00040259742138388366,
      "loss": 4.9536,
      "mean_token_accuracy": 0.2191523566842079,
      "num_tokens": 62746767.0,
      "step": 28120
    },
    {
      "entropy": 5.334393501281738,
      "epoch": 3.158701634188802,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00040256435089282966,
      "loss": 4.9568,
      "mean_token_accuracy": 0.21930950433015822,
      "num_tokens": 62757963.0,
      "step": 28125
    },
    {
      "entropy": 5.35592131614685,
      "epoch": 3.15926321109676,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004025312763400772,
      "loss": 5.0479,
      "mean_token_accuracy": 0.2117084041237831,
      "num_tokens": 62769763.0,
      "step": 28130
    },
    {
      "entropy": 5.2040886878967285,
      "epoch": 3.1598247880047174,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004024981977266794,
      "loss": 4.8419,
      "mean_token_accuracy": 0.2274887278676033,
      "num_tokens": 62780947.0,
      "step": 28135
    },
    {
      "entropy": 5.332962608337402,
      "epoch": 3.1603863649126747,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004024651150536896,
      "loss": 5.0554,
      "mean_token_accuracy": 0.1991260305047035,
      "num_tokens": 62793003.0,
      "step": 28140
    },
    {
      "entropy": 5.334596633911133,
      "epoch": 3.1609479418206323,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004024320283221611,
      "loss": 4.937,
      "mean_token_accuracy": 0.21238436847925185,
      "num_tokens": 62804467.0,
      "step": 28145
    },
    {
      "entropy": 5.265217971801758,
      "epoch": 3.16150951872859,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00040239893753314735,
      "loss": 4.9116,
      "mean_token_accuracy": 0.22352216392755508,
      "num_tokens": 62815834.0,
      "step": 28150
    },
    {
      "entropy": 5.307755804061889,
      "epoch": 3.1620710956365476,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004023658426877021,
      "loss": 4.9765,
      "mean_token_accuracy": 0.21414149701595306,
      "num_tokens": 62827277.0,
      "step": 28155
    },
    {
      "entropy": 5.316776847839355,
      "epoch": 3.162632672544505,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040233274378687897,
      "loss": 4.8617,
      "mean_token_accuracy": 0.2231655403971672,
      "num_tokens": 62838652.0,
      "step": 28160
    },
    {
      "entropy": 5.349366188049316,
      "epoch": 3.1631942494524625,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040229964083173187,
      "loss": 5.0065,
      "mean_token_accuracy": 0.20976434648036957,
      "num_tokens": 62849448.0,
      "step": 28165
    },
    {
      "entropy": 5.343938589096069,
      "epoch": 3.16375582636042,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040226653382331484,
      "loss": 4.936,
      "mean_token_accuracy": 0.2231398969888687,
      "num_tokens": 62860840.0,
      "step": 28170
    },
    {
      "entropy": 5.361573076248169,
      "epoch": 3.1643174032683774,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000402233422762682,
      "loss": 4.9442,
      "mean_token_accuracy": 0.22188040167093276,
      "num_tokens": 62871321.0,
      "step": 28175
    },
    {
      "entropy": 5.209718036651611,
      "epoch": 3.164878980176335,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040220030765088764,
      "loss": 4.8739,
      "mean_token_accuracy": 0.22673251032829284,
      "num_tokens": 62882865.0,
      "step": 28180
    },
    {
      "entropy": 5.249802494049073,
      "epoch": 3.1654405570842927,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004021671884889862,
      "loss": 4.9424,
      "mean_token_accuracy": 0.2158096343278885,
      "num_tokens": 62894122.0,
      "step": 28185
    },
    {
      "entropy": 5.359518909454346,
      "epoch": 3.1660021339922504,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004021340652780321,
      "loss": 5.0159,
      "mean_token_accuracy": 0.20702804923057555,
      "num_tokens": 62904797.0,
      "step": 28190
    },
    {
      "entropy": 5.346257591247559,
      "epoch": 3.1665637109002076,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004021009380190801,
      "loss": 4.9297,
      "mean_token_accuracy": 0.21652252972126007,
      "num_tokens": 62914739.0,
      "step": 28195
    },
    {
      "entropy": 5.272299003601074,
      "epoch": 3.1671252878081653,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040206780671318495,
      "loss": 4.945,
      "mean_token_accuracy": 0.21526908576488496,
      "num_tokens": 62925183.0,
      "step": 28200
    },
    {
      "entropy": 5.242510032653809,
      "epoch": 3.167686864716123,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004020346713614015,
      "loss": 4.9231,
      "mean_token_accuracy": 0.21893549114465713,
      "num_tokens": 62936197.0,
      "step": 28205
    },
    {
      "entropy": 5.297391080856324,
      "epoch": 3.1682484416240806,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040200153196478486,
      "loss": 4.9653,
      "mean_token_accuracy": 0.22105251252651215,
      "num_tokens": 62947571.0,
      "step": 28210
    },
    {
      "entropy": 5.303922843933106,
      "epoch": 3.168810018532038,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004019683885243902,
      "loss": 4.971,
      "mean_token_accuracy": 0.21863656640052795,
      "num_tokens": 62957614.0,
      "step": 28215
    },
    {
      "entropy": 5.25503568649292,
      "epoch": 3.1693715954399955,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004019352410412728,
      "loss": 4.872,
      "mean_token_accuracy": 0.22083041071891785,
      "num_tokens": 62968840.0,
      "step": 28220
    },
    {
      "entropy": 5.273051738739014,
      "epoch": 3.169933172347953,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040190208951648807,
      "loss": 4.9611,
      "mean_token_accuracy": 0.21797753274440765,
      "num_tokens": 62979956.0,
      "step": 28225
    },
    {
      "entropy": 5.337050580978394,
      "epoch": 3.170494749255911,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004018689339510916,
      "loss": 4.8987,
      "mean_token_accuracy": 0.22278688102960587,
      "num_tokens": 62991352.0,
      "step": 28230
    },
    {
      "entropy": 5.430960083007813,
      "epoch": 3.171056326163868,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040183577434613904,
      "loss": 5.0901,
      "mean_token_accuracy": 0.21182891577482224,
      "num_tokens": 63002738.0,
      "step": 28235
    },
    {
      "entropy": 5.3498939037322994,
      "epoch": 3.1716179030718257,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004018026107026863,
      "loss": 5.032,
      "mean_token_accuracy": 0.21620649397373198,
      "num_tokens": 63013989.0,
      "step": 28240
    },
    {
      "entropy": 5.263688850402832,
      "epoch": 3.1721794799797833,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004017694430217892,
      "loss": 4.8839,
      "mean_token_accuracy": 0.21562898755073548,
      "num_tokens": 63026158.0,
      "step": 28245
    },
    {
      "entropy": 5.2281310081481935,
      "epoch": 3.172741056887741,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004017362713045039,
      "loss": 4.8418,
      "mean_token_accuracy": 0.2206060841679573,
      "num_tokens": 63036432.0,
      "step": 28250
    },
    {
      "entropy": 5.3542522430419925,
      "epoch": 3.173302633795698,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00040170309555188655,
      "loss": 4.9538,
      "mean_token_accuracy": 0.21693918555974961,
      "num_tokens": 63047151.0,
      "step": 28255
    },
    {
      "entropy": 5.292980718612671,
      "epoch": 3.173864210703656,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004016699157649936,
      "loss": 4.961,
      "mean_token_accuracy": 0.2146256998181343,
      "num_tokens": 63059514.0,
      "step": 28260
    },
    {
      "entropy": 5.407813453674317,
      "epoch": 3.1744257876116135,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004016367319448812,
      "loss": 5.1684,
      "mean_token_accuracy": 0.20842941701412201,
      "num_tokens": 63071037.0,
      "step": 28265
    },
    {
      "entropy": 5.334617805480957,
      "epoch": 3.1749873645195708,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040160354409260625,
      "loss": 4.947,
      "mean_token_accuracy": 0.21581348031759262,
      "num_tokens": 63083615.0,
      "step": 28270
    },
    {
      "entropy": 5.312992763519287,
      "epoch": 3.1755489414275284,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004015703522092254,
      "loss": 4.981,
      "mean_token_accuracy": 0.2194305866956711,
      "num_tokens": 63094377.0,
      "step": 28275
    },
    {
      "entropy": 5.275470113754272,
      "epoch": 3.176110518335486,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004015371562957954,
      "loss": 4.9357,
      "mean_token_accuracy": 0.21924457997083663,
      "num_tokens": 63105108.0,
      "step": 28280
    },
    {
      "entropy": 5.3608917713165285,
      "epoch": 3.1766720952434437,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004015039563533732,
      "loss": 5.0662,
      "mean_token_accuracy": 0.20824234038591385,
      "num_tokens": 63116576.0,
      "step": 28285
    },
    {
      "entropy": 5.3675422191619875,
      "epoch": 3.177233672151401,
      "grad_norm": 1.0,
      "learning_rate": 0.0004014707523830161,
      "loss": 5.0246,
      "mean_token_accuracy": 0.21262292861938475,
      "num_tokens": 63128343.0,
      "step": 28290
    },
    {
      "entropy": 5.4240124225616455,
      "epoch": 3.1777952490593586,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004014375443857811,
      "loss": 5.1555,
      "mean_token_accuracy": 0.205175045132637,
      "num_tokens": 63140309.0,
      "step": 28295
    },
    {
      "entropy": 5.382351589202881,
      "epoch": 3.1783568259673163,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004014043323627257,
      "loss": 5.0098,
      "mean_token_accuracy": 0.21110262274742125,
      "num_tokens": 63150825.0,
      "step": 28300
    },
    {
      "entropy": 5.3116868019104,
      "epoch": 3.178918402875274,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004013711163149072,
      "loss": 4.8974,
      "mean_token_accuracy": 0.2265123024582863,
      "num_tokens": 63161296.0,
      "step": 28305
    },
    {
      "entropy": 5.2861391544342045,
      "epoch": 3.179479979783231,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040133789624338353,
      "loss": 4.8974,
      "mean_token_accuracy": 0.21742161363363266,
      "num_tokens": 63171447.0,
      "step": 28310
    },
    {
      "entropy": 5.259005737304688,
      "epoch": 3.180041556691189,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040130467214921216,
      "loss": 4.9783,
      "mean_token_accuracy": 0.20392885953187942,
      "num_tokens": 63182860.0,
      "step": 28315
    },
    {
      "entropy": 5.338739538192749,
      "epoch": 3.1806031335991465,
      "grad_norm": 1.09375,
      "learning_rate": 0.000401271444033451,
      "loss": 4.9862,
      "mean_token_accuracy": 0.21406301707029343,
      "num_tokens": 63192894.0,
      "step": 28320
    },
    {
      "entropy": 5.369056510925293,
      "epoch": 3.181164710507104,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004012382118971582,
      "loss": 5.0091,
      "mean_token_accuracy": 0.21138260066509246,
      "num_tokens": 63204013.0,
      "step": 28325
    },
    {
      "entropy": 5.301993274688721,
      "epoch": 3.1817262874150614,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004012049757413917,
      "loss": 4.9083,
      "mean_token_accuracy": 0.2216563418507576,
      "num_tokens": 63215079.0,
      "step": 28330
    },
    {
      "entropy": 5.261045646667481,
      "epoch": 3.182287864323019,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040117173556720976,
      "loss": 4.9544,
      "mean_token_accuracy": 0.21556589007377625,
      "num_tokens": 63225185.0,
      "step": 28335
    },
    {
      "entropy": 5.394541931152344,
      "epoch": 3.1828494412309767,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040113849137567094,
      "loss": 4.9466,
      "mean_token_accuracy": 0.21592654287815094,
      "num_tokens": 63237490.0,
      "step": 28340
    },
    {
      "entropy": 5.304803371429443,
      "epoch": 3.1834110181389343,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004011052431678336,
      "loss": 4.869,
      "mean_token_accuracy": 0.22405995279550553,
      "num_tokens": 63247774.0,
      "step": 28345
    },
    {
      "entropy": 5.244814825057984,
      "epoch": 3.1839725950468916,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004010719909447564,
      "loss": 4.9532,
      "mean_token_accuracy": 0.2122838616371155,
      "num_tokens": 63258953.0,
      "step": 28350
    },
    {
      "entropy": 5.29204044342041,
      "epoch": 3.1845341719548492,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004010387347074981,
      "loss": 4.9309,
      "mean_token_accuracy": 0.21917369961738586,
      "num_tokens": 63269670.0,
      "step": 28355
    },
    {
      "entropy": 5.3113123893737795,
      "epoch": 3.185095748862807,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040100547445711765,
      "loss": 4.9498,
      "mean_token_accuracy": 0.21976223289966584,
      "num_tokens": 63280644.0,
      "step": 28360
    },
    {
      "entropy": 5.330310106277466,
      "epoch": 3.185657325770764,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040097221019467393,
      "loss": 4.9603,
      "mean_token_accuracy": 0.21570168286561966,
      "num_tokens": 63292328.0,
      "step": 28365
    },
    {
      "entropy": 5.187137460708618,
      "epoch": 3.1862189026787218,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004009389419212262,
      "loss": 4.8139,
      "mean_token_accuracy": 0.22437052577733993,
      "num_tokens": 63302849.0,
      "step": 28370
    },
    {
      "entropy": 5.2710853099823,
      "epoch": 3.1867804795866794,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040090566963783375,
      "loss": 4.9728,
      "mean_token_accuracy": 0.21900575757026672,
      "num_tokens": 63314460.0,
      "step": 28375
    },
    {
      "entropy": 5.320605897903443,
      "epoch": 3.187342056494637,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040087239334555603,
      "loss": 4.9538,
      "mean_token_accuracy": 0.21987587958574295,
      "num_tokens": 63325713.0,
      "step": 28380
    },
    {
      "entropy": 5.288709020614624,
      "epoch": 3.1879036334025943,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040083911304545234,
      "loss": 4.9862,
      "mean_token_accuracy": 0.21303053051233292,
      "num_tokens": 63338287.0,
      "step": 28385
    },
    {
      "entropy": 5.334245872497559,
      "epoch": 3.188465210310552,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00040080582873858254,
      "loss": 4.9307,
      "mean_token_accuracy": 0.2222326710820198,
      "num_tokens": 63349592.0,
      "step": 28390
    },
    {
      "entropy": 5.313800954818726,
      "epoch": 3.1890267872185096,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004007725404260064,
      "loss": 4.8723,
      "mean_token_accuracy": 0.22726139277219773,
      "num_tokens": 63359913.0,
      "step": 28395
    },
    {
      "entropy": 5.297017335891724,
      "epoch": 3.1895883641264673,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004007392481087837,
      "loss": 4.9818,
      "mean_token_accuracy": 0.21229254454374313,
      "num_tokens": 63370677.0,
      "step": 28400
    },
    {
      "entropy": 5.305945110321045,
      "epoch": 3.1901499410344245,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004007059517879747,
      "loss": 4.9014,
      "mean_token_accuracy": 0.22753550708293915,
      "num_tokens": 63381467.0,
      "step": 28405
    },
    {
      "entropy": 5.315435886383057,
      "epoch": 3.190711517942382,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004006726514646393,
      "loss": 5.0188,
      "mean_token_accuracy": 0.21287231743335724,
      "num_tokens": 63392808.0,
      "step": 28410
    },
    {
      "entropy": 5.253287935256958,
      "epoch": 3.19127309485034,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040063934713983803,
      "loss": 4.8727,
      "mean_token_accuracy": 0.21932893246412277,
      "num_tokens": 63405206.0,
      "step": 28415
    },
    {
      "entropy": 5.171057558059692,
      "epoch": 3.1918346717582975,
      "grad_norm": 0.953125,
      "learning_rate": 0.00040060603881463127,
      "loss": 4.8788,
      "mean_token_accuracy": 0.22559989839792252,
      "num_tokens": 63416923.0,
      "step": 28420
    },
    {
      "entropy": 5.289749526977539,
      "epoch": 3.1923962486662547,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040057272649007944,
      "loss": 4.9381,
      "mean_token_accuracy": 0.21651872098445893,
      "num_tokens": 63426681.0,
      "step": 28425
    },
    {
      "entropy": 5.369942998886108,
      "epoch": 3.1929578255742124,
      "grad_norm": 1.40625,
      "learning_rate": 0.00040053941016724326,
      "loss": 4.9015,
      "mean_token_accuracy": 0.21949184834957122,
      "num_tokens": 63437299.0,
      "step": 28430
    },
    {
      "entropy": 5.319213581085205,
      "epoch": 3.19351940248217,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040050608984718364,
      "loss": 5.0123,
      "mean_token_accuracy": 0.21838558167219163,
      "num_tokens": 63449075.0,
      "step": 28435
    },
    {
      "entropy": 5.30498776435852,
      "epoch": 3.1940809793901277,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004004727655309614,
      "loss": 4.9529,
      "mean_token_accuracy": 0.2170911356806755,
      "num_tokens": 63460592.0,
      "step": 28440
    },
    {
      "entropy": 5.330946063995361,
      "epoch": 3.194642556298085,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040043943721963766,
      "loss": 4.9755,
      "mean_token_accuracy": 0.21680078357458116,
      "num_tokens": 63471043.0,
      "step": 28445
    },
    {
      "entropy": 5.348067045211792,
      "epoch": 3.1952041332060426,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040040610491427357,
      "loss": 5.0671,
      "mean_token_accuracy": 0.2100324586033821,
      "num_tokens": 63482556.0,
      "step": 28450
    },
    {
      "entropy": 5.43840479850769,
      "epoch": 3.1957657101140002,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004003727686159305,
      "loss": 5.1024,
      "mean_token_accuracy": 0.20108622163534165,
      "num_tokens": 63493973.0,
      "step": 28455
    },
    {
      "entropy": 5.386784839630127,
      "epoch": 3.1963272870219575,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040033942832566984,
      "loss": 5.0624,
      "mean_token_accuracy": 0.20580572187900542,
      "num_tokens": 63504489.0,
      "step": 28460
    },
    {
      "entropy": 5.366659545898438,
      "epoch": 3.196888863929915,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004003060840445531,
      "loss": 4.9578,
      "mean_token_accuracy": 0.21408507823944092,
      "num_tokens": 63516688.0,
      "step": 28465
    },
    {
      "entropy": 5.410526323318481,
      "epoch": 3.1974504408378728,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004002727357736422,
      "loss": 5.0721,
      "mean_token_accuracy": 0.20650816857814788,
      "num_tokens": 63528680.0,
      "step": 28470
    },
    {
      "entropy": 5.327978658676147,
      "epoch": 3.1980120177458304,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040023938351399865,
      "loss": 5.0149,
      "mean_token_accuracy": 0.2138925611972809,
      "num_tokens": 63540489.0,
      "step": 28475
    },
    {
      "entropy": 5.326773643493652,
      "epoch": 3.1985735946537877,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004002060272666846,
      "loss": 5.0035,
      "mean_token_accuracy": 0.20974558293819429,
      "num_tokens": 63551974.0,
      "step": 28480
    },
    {
      "entropy": 5.388510894775391,
      "epoch": 3.1991351715617453,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004001726670327621,
      "loss": 5.0481,
      "mean_token_accuracy": 0.20837115198373796,
      "num_tokens": 63564437.0,
      "step": 28485
    },
    {
      "entropy": 5.325735235214234,
      "epoch": 3.199696748469703,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004001393028132933,
      "loss": 4.9108,
      "mean_token_accuracy": 0.22303489446640015,
      "num_tokens": 63575041.0,
      "step": 28490
    },
    {
      "entropy": 5.307061386108399,
      "epoch": 3.2002583253776606,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040010593460934065,
      "loss": 4.9469,
      "mean_token_accuracy": 0.21407704353332518,
      "num_tokens": 63587381.0,
      "step": 28495
    },
    {
      "entropy": 5.340592765808106,
      "epoch": 3.200819902285618,
      "grad_norm": 0.9375,
      "learning_rate": 0.00040007256242196644,
      "loss": 5.0195,
      "mean_token_accuracy": 0.2122740402817726,
      "num_tokens": 63599206.0,
      "step": 28500
    },
    {
      "entropy": 5.336346387863159,
      "epoch": 3.2013814791935755,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040003918625223335,
      "loss": 4.9572,
      "mean_token_accuracy": 0.21681152135133744,
      "num_tokens": 63609557.0,
      "step": 28505
    },
    {
      "entropy": 5.274911022186279,
      "epoch": 3.201943056101533,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040000580610120406,
      "loss": 4.926,
      "mean_token_accuracy": 0.2176375210285187,
      "num_tokens": 63620423.0,
      "step": 28510
    },
    {
      "entropy": 5.428939867019653,
      "epoch": 3.202504633009491,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003999724219699415,
      "loss": 5.0916,
      "mean_token_accuracy": 0.21168097853660583,
      "num_tokens": 63632222.0,
      "step": 28515
    },
    {
      "entropy": 5.318672227859497,
      "epoch": 3.203066209917448,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003999390338595085,
      "loss": 4.9542,
      "mean_token_accuracy": 0.21749294996261598,
      "num_tokens": 63643077.0,
      "step": 28520
    },
    {
      "entropy": 5.233269548416137,
      "epoch": 3.2036277868254057,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039990564177096817,
      "loss": 4.8601,
      "mean_token_accuracy": 0.2285235658288002,
      "num_tokens": 63653642.0,
      "step": 28525
    },
    {
      "entropy": 5.309992408752441,
      "epoch": 3.2041893637333634,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003998722457053838,
      "loss": 5.0057,
      "mean_token_accuracy": 0.2113835871219635,
      "num_tokens": 63665646.0,
      "step": 28530
    },
    {
      "entropy": 5.412748336791992,
      "epoch": 3.204750940641321,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039983884566381866,
      "loss": 5.0769,
      "mean_token_accuracy": 0.20852287262678146,
      "num_tokens": 63677928.0,
      "step": 28535
    },
    {
      "entropy": 5.339959716796875,
      "epoch": 3.2053125175492783,
      "grad_norm": 1.0,
      "learning_rate": 0.0003998054416473363,
      "loss": 4.951,
      "mean_token_accuracy": 0.21667619049549103,
      "num_tokens": 63688670.0,
      "step": 28540
    },
    {
      "entropy": 5.396267414093018,
      "epoch": 3.205874094457236,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039977203365700026,
      "loss": 5.0539,
      "mean_token_accuracy": 0.2157898098230362,
      "num_tokens": 63699963.0,
      "step": 28545
    },
    {
      "entropy": 5.357429075241089,
      "epoch": 3.2064356713651936,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003997386216938743,
      "loss": 5.0618,
      "mean_token_accuracy": 0.20863089859485626,
      "num_tokens": 63711333.0,
      "step": 28550
    },
    {
      "entropy": 5.312561702728272,
      "epoch": 3.206997248273151,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039970520575902216,
      "loss": 4.9168,
      "mean_token_accuracy": 0.21683055013418198,
      "num_tokens": 63721916.0,
      "step": 28555
    },
    {
      "entropy": 5.282412767410278,
      "epoch": 3.2075588251811085,
      "grad_norm": 0.96875,
      "learning_rate": 0.000399671785853508,
      "loss": 4.9844,
      "mean_token_accuracy": 0.21497683227062225,
      "num_tokens": 63733313.0,
      "step": 28560
    },
    {
      "entropy": 5.374819564819336,
      "epoch": 3.208120402089066,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039963836197839584,
      "loss": 4.9845,
      "mean_token_accuracy": 0.21397990882396697,
      "num_tokens": 63743249.0,
      "step": 28565
    },
    {
      "entropy": 5.290561819076538,
      "epoch": 3.208681978997024,
      "grad_norm": 1.0,
      "learning_rate": 0.00039960493413474977,
      "loss": 4.9775,
      "mean_token_accuracy": 0.21715268045663833,
      "num_tokens": 63754355.0,
      "step": 28570
    },
    {
      "entropy": 5.506867694854736,
      "epoch": 3.209243555904981,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039957150232363433,
      "loss": 5.1147,
      "mean_token_accuracy": 0.21010885536670684,
      "num_tokens": 63766620.0,
      "step": 28575
    },
    {
      "entropy": 5.343281984329224,
      "epoch": 3.2098051328129387,
      "grad_norm": 0.96875,
      "learning_rate": 0.000399538066546114,
      "loss": 4.8952,
      "mean_token_accuracy": 0.2225152149796486,
      "num_tokens": 63777195.0,
      "step": 28580
    },
    {
      "entropy": 5.2178761005401615,
      "epoch": 3.2103667097208963,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00039950462680325336,
      "loss": 4.9135,
      "mean_token_accuracy": 0.21881953924894332,
      "num_tokens": 63787653.0,
      "step": 28585
    },
    {
      "entropy": 5.277022981643677,
      "epoch": 3.210928286628854,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000399471183096117,
      "loss": 4.9781,
      "mean_token_accuracy": 0.21730617731809615,
      "num_tokens": 63799371.0,
      "step": 28590
    },
    {
      "entropy": 5.2430342674255375,
      "epoch": 3.211489863536811,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003994377354257698,
      "loss": 4.7972,
      "mean_token_accuracy": 0.22769564390182495,
      "num_tokens": 63809791.0,
      "step": 28595
    },
    {
      "entropy": 5.306268882751465,
      "epoch": 3.212051440444769,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039940428379327703,
      "loss": 4.9073,
      "mean_token_accuracy": 0.21811453849077225,
      "num_tokens": 63819972.0,
      "step": 28600
    },
    {
      "entropy": 5.22514762878418,
      "epoch": 3.2126130173527265,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003993708281997036,
      "loss": 4.9177,
      "mean_token_accuracy": 0.21792841106653213,
      "num_tokens": 63830144.0,
      "step": 28605
    },
    {
      "entropy": 5.342931652069092,
      "epoch": 3.213174594260684,
      "grad_norm": 1.0,
      "learning_rate": 0.0003993373686461147,
      "loss": 5.0066,
      "mean_token_accuracy": 0.21220150589942932,
      "num_tokens": 63842393.0,
      "step": 28610
    },
    {
      "entropy": 5.34925274848938,
      "epoch": 3.2137361711686414,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003993039051335758,
      "loss": 4.9389,
      "mean_token_accuracy": 0.21230369359254836,
      "num_tokens": 63853392.0,
      "step": 28615
    },
    {
      "entropy": 5.337741231918335,
      "epoch": 3.214297748076599,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003992704376631523,
      "loss": 4.9697,
      "mean_token_accuracy": 0.21390029191970825,
      "num_tokens": 63866037.0,
      "step": 28620
    },
    {
      "entropy": 5.327512931823731,
      "epoch": 3.2148593249845567,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003992369662359099,
      "loss": 4.9356,
      "mean_token_accuracy": 0.22495548129081727,
      "num_tokens": 63876264.0,
      "step": 28625
    },
    {
      "entropy": 5.2320881366729735,
      "epoch": 3.2154209018925144,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003992034908529143,
      "loss": 4.8522,
      "mean_token_accuracy": 0.22506452947854996,
      "num_tokens": 63887545.0,
      "step": 28630
    },
    {
      "entropy": 5.265501356124878,
      "epoch": 3.2159824788004716,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039917001151523144,
      "loss": 4.9852,
      "mean_token_accuracy": 0.21217832267284392,
      "num_tokens": 63897689.0,
      "step": 28635
    },
    {
      "entropy": 5.316146755218506,
      "epoch": 3.2165440557084293,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003991365282239272,
      "loss": 4.9073,
      "mean_token_accuracy": 0.21365359276533127,
      "num_tokens": 63908884.0,
      "step": 28640
    },
    {
      "entropy": 5.3090005874633786,
      "epoch": 3.217105632616387,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039910304098006775,
      "loss": 4.8927,
      "mean_token_accuracy": 0.2202077776193619,
      "num_tokens": 63918515.0,
      "step": 28645
    },
    {
      "entropy": 5.234101629257202,
      "epoch": 3.217667209524344,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003990695497847193,
      "loss": 4.9208,
      "mean_token_accuracy": 0.219472037255764,
      "num_tokens": 63928802.0,
      "step": 28650
    },
    {
      "entropy": 5.281265640258789,
      "epoch": 3.218228786432302,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003990360546389484,
      "loss": 4.925,
      "mean_token_accuracy": 0.2207740917801857,
      "num_tokens": 63940770.0,
      "step": 28655
    },
    {
      "entropy": 5.410409545898437,
      "epoch": 3.2187903633402595,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00039900255554382126,
      "loss": 4.9864,
      "mean_token_accuracy": 0.2173728883266449,
      "num_tokens": 63951851.0,
      "step": 28660
    },
    {
      "entropy": 5.294085454940796,
      "epoch": 3.219351940248217,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039896905250040473,
      "loss": 4.9562,
      "mean_token_accuracy": 0.22140593975782394,
      "num_tokens": 63962768.0,
      "step": 28665
    },
    {
      "entropy": 5.294540882110596,
      "epoch": 3.2199135171561744,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039893554550976545,
      "loss": 4.9817,
      "mean_token_accuracy": 0.20848191827535628,
      "num_tokens": 63973855.0,
      "step": 28670
    },
    {
      "entropy": 5.4220047950744625,
      "epoch": 3.220475094064132,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003989020345729705,
      "loss": 5.0939,
      "mean_token_accuracy": 0.2125639170408249,
      "num_tokens": 63986342.0,
      "step": 28675
    },
    {
      "entropy": 5.331042957305908,
      "epoch": 3.2210366709720897,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039886851969108653,
      "loss": 4.925,
      "mean_token_accuracy": 0.21933818459510804,
      "num_tokens": 63997080.0,
      "step": 28680
    },
    {
      "entropy": 5.244397783279419,
      "epoch": 3.2215982478800473,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039883500086518086,
      "loss": 4.9313,
      "mean_token_accuracy": 0.2174136146903038,
      "num_tokens": 64007373.0,
      "step": 28685
    },
    {
      "entropy": 5.318726634979248,
      "epoch": 3.2221598247880046,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00039880147809632084,
      "loss": 5.0356,
      "mean_token_accuracy": 0.21788446605205536,
      "num_tokens": 64018429.0,
      "step": 28690
    },
    {
      "entropy": 5.32415623664856,
      "epoch": 3.222721401695962,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039876795138557363,
      "loss": 4.9555,
      "mean_token_accuracy": 0.2194507122039795,
      "num_tokens": 64028633.0,
      "step": 28695
    },
    {
      "entropy": 5.387431192398071,
      "epoch": 3.22328297860392,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003987344207340069,
      "loss": 4.9858,
      "mean_token_accuracy": 0.21556867361068727,
      "num_tokens": 64040860.0,
      "step": 28700
    },
    {
      "entropy": 5.365358829498291,
      "epoch": 3.2238445555118775,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00039870088614268824,
      "loss": 5.087,
      "mean_token_accuracy": 0.2089954912662506,
      "num_tokens": 64053141.0,
      "step": 28705
    },
    {
      "entropy": 5.241791725158691,
      "epoch": 3.2244061324198348,
      "grad_norm": 0.953125,
      "learning_rate": 0.00039866734761268534,
      "loss": 4.8683,
      "mean_token_accuracy": 0.2239476338028908,
      "num_tokens": 64064583.0,
      "step": 28710
    },
    {
      "entropy": 5.398545551300049,
      "epoch": 3.2249677093277924,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039863380514506614,
      "loss": 5.0596,
      "mean_token_accuracy": 0.2117016851902008,
      "num_tokens": 64075368.0,
      "step": 28715
    },
    {
      "entropy": 5.36952772140503,
      "epoch": 3.22552928623575,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039860025874089865,
      "loss": 5.0128,
      "mean_token_accuracy": 0.21948052495718,
      "num_tokens": 64085815.0,
      "step": 28720
    },
    {
      "entropy": 5.265235805511475,
      "epoch": 3.2260908631437077,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000398566708401251,
      "loss": 4.9154,
      "mean_token_accuracy": 0.22399600595235825,
      "num_tokens": 64096825.0,
      "step": 28725
    },
    {
      "entropy": 5.3001172065734865,
      "epoch": 3.226652440051665,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039853315412719144,
      "loss": 4.967,
      "mean_token_accuracy": 0.219826577603817,
      "num_tokens": 64108010.0,
      "step": 28730
    },
    {
      "entropy": 5.359076690673828,
      "epoch": 3.2272140169596226,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003984995959197883,
      "loss": 4.9771,
      "mean_token_accuracy": 0.21046190559864045,
      "num_tokens": 64119829.0,
      "step": 28735
    },
    {
      "entropy": 5.296918630599976,
      "epoch": 3.2277755938675803,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003984660337801101,
      "loss": 4.8754,
      "mean_token_accuracy": 0.22201937437057495,
      "num_tokens": 64130138.0,
      "step": 28740
    },
    {
      "entropy": 5.288098621368408,
      "epoch": 3.2283371707755375,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00039843246770922557,
      "loss": 4.9426,
      "mean_token_accuracy": 0.21163656264543534,
      "num_tokens": 64141651.0,
      "step": 28745
    },
    {
      "entropy": 5.441115808486939,
      "epoch": 3.228898747683495,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003983988977082033,
      "loss": 5.0433,
      "mean_token_accuracy": 0.20935871452093124,
      "num_tokens": 64152718.0,
      "step": 28750
    },
    {
      "entropy": 5.333096313476562,
      "epoch": 3.229460324591453,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039836532377811237,
      "loss": 5.0069,
      "mean_token_accuracy": 0.21575986295938493,
      "num_tokens": 64163977.0,
      "step": 28755
    },
    {
      "entropy": 5.254445171356201,
      "epoch": 3.2300219014994105,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003983317459200217,
      "loss": 4.9644,
      "mean_token_accuracy": 0.22353567481040953,
      "num_tokens": 64175298.0,
      "step": 28760
    },
    {
      "entropy": 5.315092372894287,
      "epoch": 3.2305834784073677,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039829816413500035,
      "loss": 4.9944,
      "mean_token_accuracy": 0.21390583217144013,
      "num_tokens": 64186746.0,
      "step": 28765
    },
    {
      "entropy": 5.337050485610962,
      "epoch": 3.2311450553153254,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003982645784241176,
      "loss": 4.9926,
      "mean_token_accuracy": 0.21075412034988403,
      "num_tokens": 64198682.0,
      "step": 28770
    },
    {
      "entropy": 5.36042799949646,
      "epoch": 3.231706632223283,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003982309887884429,
      "loss": 4.9579,
      "mean_token_accuracy": 0.22017654925584793,
      "num_tokens": 64210425.0,
      "step": 28775
    },
    {
      "entropy": 5.289422416687012,
      "epoch": 3.2322682091312407,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039819739522904576,
      "loss": 4.8638,
      "mean_token_accuracy": 0.2197912558913231,
      "num_tokens": 64222839.0,
      "step": 28780
    },
    {
      "entropy": 5.23187952041626,
      "epoch": 3.232829786039198,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00039816379774699573,
      "loss": 4.9368,
      "mean_token_accuracy": 0.22343114167451858,
      "num_tokens": 64234235.0,
      "step": 28785
    },
    {
      "entropy": 5.313961553573608,
      "epoch": 3.2333913629471556,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039813019634336264,
      "loss": 4.9361,
      "mean_token_accuracy": 0.21649348437786103,
      "num_tokens": 64244607.0,
      "step": 28790
    },
    {
      "entropy": 5.312313890457153,
      "epoch": 3.2339529398551132,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003980965910192163,
      "loss": 4.9534,
      "mean_token_accuracy": 0.2121251717209816,
      "num_tokens": 64254434.0,
      "step": 28795
    },
    {
      "entropy": 5.228569793701172,
      "epoch": 3.234514516763071,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00039806298177562675,
      "loss": 4.9179,
      "mean_token_accuracy": 0.21516283303499223,
      "num_tokens": 64266693.0,
      "step": 28800
    },
    {
      "entropy": 5.32692141532898,
      "epoch": 3.235076093671028,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00039802936861366416,
      "loss": 4.9383,
      "mean_token_accuracy": 0.21986793726682663,
      "num_tokens": 64277901.0,
      "step": 28805
    },
    {
      "entropy": 5.393144035339356,
      "epoch": 3.2356376705789858,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003979957515343988,
      "loss": 4.9778,
      "mean_token_accuracy": 0.2167685404419899,
      "num_tokens": 64288583.0,
      "step": 28810
    },
    {
      "entropy": 5.306557559967041,
      "epoch": 3.2361992474869434,
      "grad_norm": 0.984375,
      "learning_rate": 0.000397962130538901,
      "loss": 4.8871,
      "mean_token_accuracy": 0.2185106545686722,
      "num_tokens": 64299688.0,
      "step": 28815
    },
    {
      "entropy": 5.240636730194092,
      "epoch": 3.236760824394901,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039792850562824124,
      "loss": 4.9282,
      "mean_token_accuracy": 0.22064827382564545,
      "num_tokens": 64311150.0,
      "step": 28820
    },
    {
      "entropy": 5.277690601348877,
      "epoch": 3.2373224013028583,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003978948768034902,
      "loss": 4.9479,
      "mean_token_accuracy": 0.21905473321676255,
      "num_tokens": 64322658.0,
      "step": 28825
    },
    {
      "entropy": 5.272652864456177,
      "epoch": 3.237883978210816,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039786124406571854,
      "loss": 4.8886,
      "mean_token_accuracy": 0.21909582763910293,
      "num_tokens": 64333350.0,
      "step": 28830
    },
    {
      "entropy": 5.371354675292968,
      "epoch": 3.2384455551187736,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039782760741599734,
      "loss": 5.0204,
      "mean_token_accuracy": 0.21583033204078675,
      "num_tokens": 64344689.0,
      "step": 28835
    },
    {
      "entropy": 5.259570074081421,
      "epoch": 3.239007132026731,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003977939668553973,
      "loss": 4.9797,
      "mean_token_accuracy": 0.21783460378646852,
      "num_tokens": 64356602.0,
      "step": 28840
    },
    {
      "entropy": 5.261673355102539,
      "epoch": 3.2395687089346885,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039776032238498995,
      "loss": 4.9235,
      "mean_token_accuracy": 0.2197751596570015,
      "num_tokens": 64368496.0,
      "step": 28845
    },
    {
      "entropy": 5.351871109008789,
      "epoch": 3.240130285842646,
      "grad_norm": 1.0,
      "learning_rate": 0.0003977266740058462,
      "loss": 4.9577,
      "mean_token_accuracy": 0.21757280230522155,
      "num_tokens": 64379374.0,
      "step": 28850
    },
    {
      "entropy": 5.384038591384888,
      "epoch": 3.240691862750604,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003976930217190376,
      "loss": 4.9318,
      "mean_token_accuracy": 0.22188475877046585,
      "num_tokens": 64391006.0,
      "step": 28855
    },
    {
      "entropy": 5.24977445602417,
      "epoch": 3.241253439658561,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003976593655256356,
      "loss": 4.9149,
      "mean_token_accuracy": 0.21828859001398088,
      "num_tokens": 64403100.0,
      "step": 28860
    },
    {
      "entropy": 5.210212898254395,
      "epoch": 3.2418150165665187,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039762570542671176,
      "loss": 4.8683,
      "mean_token_accuracy": 0.2209101364016533,
      "num_tokens": 64413763.0,
      "step": 28865
    },
    {
      "entropy": 5.2489001750946045,
      "epoch": 3.2423765934744764,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000397592041423338,
      "loss": 4.8298,
      "mean_token_accuracy": 0.22011463791131974,
      "num_tokens": 64425289.0,
      "step": 28870
    },
    {
      "entropy": 5.369996976852417,
      "epoch": 3.242938170382434,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039755837351658605,
      "loss": 4.9216,
      "mean_token_accuracy": 0.22386018335819244,
      "num_tokens": 64436080.0,
      "step": 28875
    },
    {
      "entropy": 5.275088882446289,
      "epoch": 3.2434997472903913,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003975247017075279,
      "loss": 4.9408,
      "mean_token_accuracy": 0.21987965553998948,
      "num_tokens": 64447036.0,
      "step": 28880
    },
    {
      "entropy": 5.387848043441773,
      "epoch": 3.244061324198349,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003974910259972358,
      "loss": 5.0873,
      "mean_token_accuracy": 0.2088085502386093,
      "num_tokens": 64458173.0,
      "step": 28885
    },
    {
      "entropy": 5.380909061431884,
      "epoch": 3.2446229011063066,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003974573463867819,
      "loss": 4.9607,
      "mean_token_accuracy": 0.2129489228129387,
      "num_tokens": 64468985.0,
      "step": 28890
    },
    {
      "entropy": 5.3020494937896725,
      "epoch": 3.2451844780142642,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039742366287723855,
      "loss": 4.9681,
      "mean_token_accuracy": 0.2192863792181015,
      "num_tokens": 64479802.0,
      "step": 28895
    },
    {
      "entropy": 5.25401611328125,
      "epoch": 3.2457460549222215,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039738997546967834,
      "loss": 4.9625,
      "mean_token_accuracy": 0.204403755068779,
      "num_tokens": 64490072.0,
      "step": 28900
    },
    {
      "entropy": 5.30256495475769,
      "epoch": 3.246307631830179,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039735628416517394,
      "loss": 4.9468,
      "mean_token_accuracy": 0.21722718179225922,
      "num_tokens": 64501096.0,
      "step": 28905
    },
    {
      "entropy": 5.343685245513916,
      "epoch": 3.246869208738137,
      "grad_norm": 0.9375,
      "learning_rate": 0.00039732258896479786,
      "loss": 4.9526,
      "mean_token_accuracy": 0.21988616585731507,
      "num_tokens": 64512562.0,
      "step": 28910
    },
    {
      "entropy": 5.208364486694336,
      "epoch": 3.2474307856460944,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003972888898696231,
      "loss": 4.8242,
      "mean_token_accuracy": 0.21957779079675674,
      "num_tokens": 64522955.0,
      "step": 28915
    },
    {
      "entropy": 5.304768896102905,
      "epoch": 3.2479923625540517,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039725518688072276,
      "loss": 4.9332,
      "mean_token_accuracy": 0.21707673519849777,
      "num_tokens": 64533805.0,
      "step": 28920
    },
    {
      "entropy": 5.274272060394287,
      "epoch": 3.2485539394620093,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00039722147999916976,
      "loss": 4.9912,
      "mean_token_accuracy": 0.20914304852485657,
      "num_tokens": 64545543.0,
      "step": 28925
    },
    {
      "entropy": 5.357356166839599,
      "epoch": 3.249115516369967,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003971877692260375,
      "loss": 4.9963,
      "mean_token_accuracy": 0.2080279842019081,
      "num_tokens": 64557109.0,
      "step": 28930
    },
    {
      "entropy": 5.361635398864746,
      "epoch": 3.249677093277924,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00039715405456239924,
      "loss": 5.0001,
      "mean_token_accuracy": 0.21495866030454636,
      "num_tokens": 64568880.0,
      "step": 28935
    },
    {
      "entropy": 5.336015939712524,
      "epoch": 3.250238670185882,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003971203360093286,
      "loss": 4.9871,
      "mean_token_accuracy": 0.21968231797218324,
      "num_tokens": 64579859.0,
      "step": 28940
    },
    {
      "entropy": 5.256529045104981,
      "epoch": 3.2508002470938395,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039708661356789896,
      "loss": 4.8637,
      "mean_token_accuracy": 0.23224980384111404,
      "num_tokens": 64589133.0,
      "step": 28945
    },
    {
      "entropy": 5.303734207153321,
      "epoch": 3.251361824001797,
      "grad_norm": 1.0,
      "learning_rate": 0.00039705288723918415,
      "loss": 4.9369,
      "mean_token_accuracy": 0.21793006211519242,
      "num_tokens": 64600040.0,
      "step": 28950
    },
    {
      "entropy": 5.193812608718872,
      "epoch": 3.2519234009097544,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003970191570242582,
      "loss": 4.8689,
      "mean_token_accuracy": 0.22500550448894502,
      "num_tokens": 64611756.0,
      "step": 28955
    },
    {
      "entropy": 5.24061689376831,
      "epoch": 3.252484977817712,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000396985422924195,
      "loss": 5.0013,
      "mean_token_accuracy": 0.2175721064209938,
      "num_tokens": 64622994.0,
      "step": 28960
    },
    {
      "entropy": 5.309499311447143,
      "epoch": 3.2530465547256697,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039695168494006857,
      "loss": 4.8569,
      "mean_token_accuracy": 0.2188839688897133,
      "num_tokens": 64633200.0,
      "step": 28965
    },
    {
      "entropy": 5.3397948265075685,
      "epoch": 3.2536081316336274,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003969179430729531,
      "loss": 5.0551,
      "mean_token_accuracy": 0.21211952865123748,
      "num_tokens": 64644412.0,
      "step": 28970
    },
    {
      "entropy": 5.283937072753906,
      "epoch": 3.2541697085415846,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003968841973239231,
      "loss": 4.8636,
      "mean_token_accuracy": 0.2169060930609703,
      "num_tokens": 64655986.0,
      "step": 28975
    },
    {
      "entropy": 5.168784236907959,
      "epoch": 3.2547312854495423,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000396850447694053,
      "loss": 4.8414,
      "mean_token_accuracy": 0.21684621274471283,
      "num_tokens": 64667770.0,
      "step": 28980
    },
    {
      "entropy": 5.41079044342041,
      "epoch": 3.2552928623575,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003968166941844174,
      "loss": 5.0576,
      "mean_token_accuracy": 0.21014534831047058,
      "num_tokens": 64679239.0,
      "step": 28985
    },
    {
      "entropy": 5.342538833618164,
      "epoch": 3.2558544392654576,
      "grad_norm": 1.015625,
      "learning_rate": 0.000396782936796091,
      "loss": 4.8535,
      "mean_token_accuracy": 0.21594661176204683,
      "num_tokens": 64690192.0,
      "step": 28990
    },
    {
      "entropy": 5.267762136459351,
      "epoch": 3.256416016173415,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003967491755301487,
      "loss": 4.9162,
      "mean_token_accuracy": 0.21840698271989822,
      "num_tokens": 64700692.0,
      "step": 28995
    },
    {
      "entropy": 5.29859356880188,
      "epoch": 3.2569775930813725,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003967154103876654,
      "loss": 5.0176,
      "mean_token_accuracy": 0.214374278485775,
      "num_tokens": 64711753.0,
      "step": 29000
    },
    {
      "entropy": 5.386541080474854,
      "epoch": 3.25753916998933,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039668164136971615,
      "loss": 5.075,
      "mean_token_accuracy": 0.20659082531929016,
      "num_tokens": 64722145.0,
      "step": 29005
    },
    {
      "entropy": 5.300726413726807,
      "epoch": 3.258100746897288,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003966478684773763,
      "loss": 4.894,
      "mean_token_accuracy": 0.21922819912433625,
      "num_tokens": 64732059.0,
      "step": 29010
    },
    {
      "entropy": 5.326655340194702,
      "epoch": 3.258662323805245,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039661409171172117,
      "loss": 4.9518,
      "mean_token_accuracy": 0.21591241657733917,
      "num_tokens": 64742811.0,
      "step": 29015
    },
    {
      "entropy": 5.195096588134765,
      "epoch": 3.2592239007132027,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00039658031107382615,
      "loss": 4.8723,
      "mean_token_accuracy": 0.22797658145427704,
      "num_tokens": 64753287.0,
      "step": 29020
    },
    {
      "entropy": 5.3972944736480715,
      "epoch": 3.2597854776211603,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039654652656476695,
      "loss": 5.1631,
      "mean_token_accuracy": 0.20333324521780013,
      "num_tokens": 64767739.0,
      "step": 29025
    },
    {
      "entropy": 5.33060884475708,
      "epoch": 3.2603470545291175,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003965127381856191,
      "loss": 4.9504,
      "mean_token_accuracy": 0.21543456614017487,
      "num_tokens": 64779217.0,
      "step": 29030
    },
    {
      "entropy": 5.334230089187622,
      "epoch": 3.260908631437075,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039647894593745865,
      "loss": 4.9144,
      "mean_token_accuracy": 0.2196929231286049,
      "num_tokens": 64790613.0,
      "step": 29035
    },
    {
      "entropy": 5.220885562896728,
      "epoch": 3.261470208345033,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039644514982136136,
      "loss": 4.8685,
      "mean_token_accuracy": 0.22136759161949157,
      "num_tokens": 64801256.0,
      "step": 29040
    },
    {
      "entropy": 5.2573741436004635,
      "epoch": 3.2620317852529905,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00039641134983840336,
      "loss": 4.8355,
      "mean_token_accuracy": 0.22716474682092666,
      "num_tokens": 64812938.0,
      "step": 29045
    },
    {
      "entropy": 5.36338529586792,
      "epoch": 3.2625933621609478,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000396377545989661,
      "loss": 4.9201,
      "mean_token_accuracy": 0.21785055249929428,
      "num_tokens": 64824041.0,
      "step": 29050
    },
    {
      "entropy": 5.21912956237793,
      "epoch": 3.2631549390689054,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039634373827621047,
      "loss": 4.816,
      "mean_token_accuracy": 0.22299732863903046,
      "num_tokens": 64833837.0,
      "step": 29055
    },
    {
      "entropy": 5.182550239562988,
      "epoch": 3.263716515976863,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003963099266991282,
      "loss": 4.8476,
      "mean_token_accuracy": 0.22088703215122224,
      "num_tokens": 64843937.0,
      "step": 29060
    },
    {
      "entropy": 5.279118871688842,
      "epoch": 3.2642780928848207,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039627611125949086,
      "loss": 4.9109,
      "mean_token_accuracy": 0.214684197306633,
      "num_tokens": 64855955.0,
      "step": 29065
    },
    {
      "entropy": 5.436502933502197,
      "epoch": 3.264839669792778,
      "grad_norm": 1.0,
      "learning_rate": 0.00039624229195837506,
      "loss": 5.0798,
      "mean_token_accuracy": 0.21205699145793916,
      "num_tokens": 64867505.0,
      "step": 29070
    },
    {
      "entropy": 5.24262261390686,
      "epoch": 3.2654012467007356,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039620846879685767,
      "loss": 4.855,
      "mean_token_accuracy": 0.22611407786607743,
      "num_tokens": 64878746.0,
      "step": 29075
    },
    {
      "entropy": 5.302385044097901,
      "epoch": 3.2659628236086933,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039617464177601564,
      "loss": 5.0331,
      "mean_token_accuracy": 0.2158088430762291,
      "num_tokens": 64889520.0,
      "step": 29080
    },
    {
      "entropy": 5.379261636734009,
      "epoch": 3.266524400516651,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039614081089692606,
      "loss": 5.0815,
      "mean_token_accuracy": 0.20859936773777008,
      "num_tokens": 64900380.0,
      "step": 29085
    },
    {
      "entropy": 5.36366720199585,
      "epoch": 3.267085977424608,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039610697616066604,
      "loss": 4.9693,
      "mean_token_accuracy": 0.22184129059314728,
      "num_tokens": 64910480.0,
      "step": 29090
    },
    {
      "entropy": 5.243341255187988,
      "epoch": 3.267647554332566,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00039607313756831296,
      "loss": 4.9244,
      "mean_token_accuracy": 0.22035108506679535,
      "num_tokens": 64920911.0,
      "step": 29095
    },
    {
      "entropy": 5.256927919387818,
      "epoch": 3.2682091312405235,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039603929512094417,
      "loss": 4.8371,
      "mean_token_accuracy": 0.22116660177707673,
      "num_tokens": 64930687.0,
      "step": 29100
    },
    {
      "entropy": 5.205454015731812,
      "epoch": 3.268770708148481,
      "grad_norm": 1.0,
      "learning_rate": 0.0003960054488196373,
      "loss": 4.8876,
      "mean_token_accuracy": 0.22316131591796876,
      "num_tokens": 64942143.0,
      "step": 29105
    },
    {
      "entropy": 5.274940299987793,
      "epoch": 3.2693322850564384,
      "grad_norm": 0.984375,
      "learning_rate": 0.00039597159866546996,
      "loss": 4.9162,
      "mean_token_accuracy": 0.21838511824607848,
      "num_tokens": 64952440.0,
      "step": 29110
    },
    {
      "entropy": 5.343635129928589,
      "epoch": 3.269893861964396,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003959377446595201,
      "loss": 4.9191,
      "mean_token_accuracy": 0.22323812544345856,
      "num_tokens": 64962752.0,
      "step": 29115
    },
    {
      "entropy": 5.268872499465942,
      "epoch": 3.2704554388723537,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00039590388680286543,
      "loss": 4.9418,
      "mean_token_accuracy": 0.21654542684555053,
      "num_tokens": 64975413.0,
      "step": 29120
    },
    {
      "entropy": 5.261586141586304,
      "epoch": 3.271017015780311,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003958700250965842,
      "loss": 4.9285,
      "mean_token_accuracy": 0.2167259067296982,
      "num_tokens": 64987922.0,
      "step": 29125
    },
    {
      "entropy": 5.210989761352539,
      "epoch": 3.2715785926882686,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003958361595417544,
      "loss": 4.932,
      "mean_token_accuracy": 0.2180864006280899,
      "num_tokens": 64999448.0,
      "step": 29130
    },
    {
      "entropy": 5.376282596588135,
      "epoch": 3.2721401695962262,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003958022901394545,
      "loss": 5.0648,
      "mean_token_accuracy": 0.20841967761516572,
      "num_tokens": 65010771.0,
      "step": 29135
    },
    {
      "entropy": 5.417096138000488,
      "epoch": 3.272701746504184,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003957684168907628,
      "loss": 5.0069,
      "mean_token_accuracy": 0.21280168890953063,
      "num_tokens": 65021368.0,
      "step": 29140
    },
    {
      "entropy": 5.294193696975708,
      "epoch": 3.273263323412141,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00039573453979675793,
      "loss": 4.8504,
      "mean_token_accuracy": 0.23173828572034835,
      "num_tokens": 65032622.0,
      "step": 29145
    },
    {
      "entropy": 5.16272006034851,
      "epoch": 3.2738249003200988,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039570065885851844,
      "loss": 4.8238,
      "mean_token_accuracy": 0.22127371430397033,
      "num_tokens": 65043655.0,
      "step": 29150
    },
    {
      "entropy": 5.304007244110108,
      "epoch": 3.2743864772280564,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003956667740771231,
      "loss": 5.0835,
      "mean_token_accuracy": 0.20698320120573044,
      "num_tokens": 65055846.0,
      "step": 29155
    },
    {
      "entropy": 5.429491901397705,
      "epoch": 3.274948054136014,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003956328854536509,
      "loss": 5.1149,
      "mean_token_accuracy": 0.20887255668640137,
      "num_tokens": 65067617.0,
      "step": 29160
    },
    {
      "entropy": 5.455239915847779,
      "epoch": 3.2755096310439713,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00039559899298918094,
      "loss": 5.0194,
      "mean_token_accuracy": 0.22537959218025208,
      "num_tokens": 65080111.0,
      "step": 29165
    },
    {
      "entropy": 5.2702539443969725,
      "epoch": 3.276071207951929,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003955650966847922,
      "loss": 4.8673,
      "mean_token_accuracy": 0.22765863686800003,
      "num_tokens": 65092284.0,
      "step": 29170
    },
    {
      "entropy": 5.330512285232544,
      "epoch": 3.2766327848598866,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000395531196541564,
      "loss": 4.9624,
      "mean_token_accuracy": 0.2147870346903801,
      "num_tokens": 65104133.0,
      "step": 29175
    },
    {
      "entropy": 5.24319429397583,
      "epoch": 3.2771943617678443,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039549729256057583,
      "loss": 4.8529,
      "mean_token_accuracy": 0.21909899413585662,
      "num_tokens": 65116065.0,
      "step": 29180
    },
    {
      "entropy": 5.2012897491455075,
      "epoch": 3.2777559386758015,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000395463384742907,
      "loss": 4.8452,
      "mean_token_accuracy": 0.2172808051109314,
      "num_tokens": 65127235.0,
      "step": 29185
    },
    {
      "entropy": 5.319353866577148,
      "epoch": 3.278317515583759,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039542947308963733,
      "loss": 5.0217,
      "mean_token_accuracy": 0.20778318345546723,
      "num_tokens": 65138186.0,
      "step": 29190
    },
    {
      "entropy": 5.268405437469482,
      "epoch": 3.278879092491717,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003953955576018466,
      "loss": 4.9248,
      "mean_token_accuracy": 0.21931843757629393,
      "num_tokens": 65149222.0,
      "step": 29195
    },
    {
      "entropy": 5.301938486099243,
      "epoch": 3.2794406693996745,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003953616382806145,
      "loss": 4.9254,
      "mean_token_accuracy": 0.21623367220163345,
      "num_tokens": 65160404.0,
      "step": 29200
    },
    {
      "entropy": 5.309209680557251,
      "epoch": 3.2800022463076317,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039532771512702124,
      "loss": 4.9176,
      "mean_token_accuracy": 0.20979031026363373,
      "num_tokens": 65171395.0,
      "step": 29205
    },
    {
      "entropy": 5.2678216934204105,
      "epoch": 3.2805638232155894,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0003952937881421468,
      "loss": 4.9694,
      "mean_token_accuracy": 0.21355675756931305,
      "num_tokens": 65183752.0,
      "step": 29210
    },
    {
      "entropy": 5.240690660476685,
      "epoch": 3.281125400123547,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003952598573270716,
      "loss": 4.8359,
      "mean_token_accuracy": 0.22610283493995667,
      "num_tokens": 65193807.0,
      "step": 29215
    },
    {
      "entropy": 5.30090012550354,
      "epoch": 3.2816869770315042,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003952259226828759,
      "loss": 4.968,
      "mean_token_accuracy": 0.21131884604692458,
      "num_tokens": 65203482.0,
      "step": 29220
    },
    {
      "entropy": 5.337068128585815,
      "epoch": 3.282248553939462,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039519198421064005,
      "loss": 4.9715,
      "mean_token_accuracy": 0.21169612556695938,
      "num_tokens": 65214965.0,
      "step": 29225
    },
    {
      "entropy": 5.291839599609375,
      "epoch": 3.2828101308474196,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039515804191144496,
      "loss": 4.8529,
      "mean_token_accuracy": 0.2207476407289505,
      "num_tokens": 65227791.0,
      "step": 29230
    },
    {
      "entropy": 5.414787244796753,
      "epoch": 3.2833717077553772,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039512409578637114,
      "loss": 5.019,
      "mean_token_accuracy": 0.21386122405529023,
      "num_tokens": 65238937.0,
      "step": 29235
    },
    {
      "entropy": 5.310895729064941,
      "epoch": 3.2839332846633345,
      "grad_norm": 1.0,
      "learning_rate": 0.00039509014583649947,
      "loss": 4.9886,
      "mean_token_accuracy": 0.21055995672941208,
      "num_tokens": 65250912.0,
      "step": 29240
    },
    {
      "entropy": 5.316776132583618,
      "epoch": 3.284494861571292,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039505619206291106,
      "loss": 4.9499,
      "mean_token_accuracy": 0.22364903688430787,
      "num_tokens": 65262094.0,
      "step": 29245
    },
    {
      "entropy": 5.202996110916137,
      "epoch": 3.2850564384792498,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039502223446668687,
      "loss": 4.8163,
      "mean_token_accuracy": 0.2250420480966568,
      "num_tokens": 65271860.0,
      "step": 29250
    },
    {
      "entropy": 5.3047102928161625,
      "epoch": 3.2856180153872074,
      "grad_norm": 1.0,
      "learning_rate": 0.0003949882730489082,
      "loss": 5.0081,
      "mean_token_accuracy": 0.2109309509396553,
      "num_tokens": 65282341.0,
      "step": 29255
    },
    {
      "entropy": 5.3224146366119385,
      "epoch": 3.2861795922951647,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039495430781065645,
      "loss": 4.9984,
      "mean_token_accuracy": 0.2172051653265953,
      "num_tokens": 65293570.0,
      "step": 29260
    },
    {
      "entropy": 5.280454015731811,
      "epoch": 3.2867411692031223,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000394920338753013,
      "loss": 4.8915,
      "mean_token_accuracy": 0.219457146525383,
      "num_tokens": 65305226.0,
      "step": 29265
    },
    {
      "entropy": 5.249974632263184,
      "epoch": 3.28730274611108,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003948863658770594,
      "loss": 4.8268,
      "mean_token_accuracy": 0.22449052780866624,
      "num_tokens": 65316309.0,
      "step": 29270
    },
    {
      "entropy": 5.283241891860962,
      "epoch": 3.2878643230190376,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003948523891838774,
      "loss": 4.8847,
      "mean_token_accuracy": 0.22324872612953187,
      "num_tokens": 65327491.0,
      "step": 29275
    },
    {
      "entropy": 5.2039166450500485,
      "epoch": 3.288425899926995,
      "grad_norm": 1.0,
      "learning_rate": 0.00039481840867454886,
      "loss": 4.7638,
      "mean_token_accuracy": 0.22708147317171096,
      "num_tokens": 65337908.0,
      "step": 29280
    },
    {
      "entropy": 5.138336610794068,
      "epoch": 3.2889874768349525,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003947844243501558,
      "loss": 4.8195,
      "mean_token_accuracy": 0.22995772510766982,
      "num_tokens": 65348209.0,
      "step": 29285
    },
    {
      "entropy": 5.2498551368713375,
      "epoch": 3.28954905374291,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039475043621178005,
      "loss": 4.9259,
      "mean_token_accuracy": 0.2208228215575218,
      "num_tokens": 65358648.0,
      "step": 29290
    },
    {
      "entropy": 5.228465843200683,
      "epoch": 3.290110630650868,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003947164442605041,
      "loss": 4.9306,
      "mean_token_accuracy": 0.2185330718755722,
      "num_tokens": 65369132.0,
      "step": 29295
    },
    {
      "entropy": 5.288647985458374,
      "epoch": 3.290672207558825,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003946824484974101,
      "loss": 4.9462,
      "mean_token_accuracy": 0.21715626269578933,
      "num_tokens": 65380334.0,
      "step": 29300
    },
    {
      "entropy": 5.485975980758667,
      "epoch": 3.2912337844667827,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003946484489235805,
      "loss": 5.0748,
      "mean_token_accuracy": 0.20528641641139983,
      "num_tokens": 65394680.0,
      "step": 29305
    },
    {
      "entropy": 5.376774787902832,
      "epoch": 3.2917953613747404,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00039461444554009787,
      "loss": 4.962,
      "mean_token_accuracy": 0.2203369379043579,
      "num_tokens": 65406029.0,
      "step": 29310
    },
    {
      "entropy": 5.261801052093506,
      "epoch": 3.2923569382826976,
      "grad_norm": 1.03125,
      "learning_rate": 0.000394580438348045,
      "loss": 4.9087,
      "mean_token_accuracy": 0.22125074714422227,
      "num_tokens": 65416626.0,
      "step": 29315
    },
    {
      "entropy": 5.261602783203125,
      "epoch": 3.2929185151906553,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039454642734850447,
      "loss": 4.9489,
      "mean_token_accuracy": 0.2225902035832405,
      "num_tokens": 65427579.0,
      "step": 29320
    },
    {
      "entropy": 5.271692705154419,
      "epoch": 3.293480092098613,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003945124125425594,
      "loss": 4.9547,
      "mean_token_accuracy": 0.21369752436876296,
      "num_tokens": 65439893.0,
      "step": 29325
    },
    {
      "entropy": 5.434156322479248,
      "epoch": 3.2940416690065706,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003944783939312928,
      "loss": 5.0734,
      "mean_token_accuracy": 0.21449653208255767,
      "num_tokens": 65451861.0,
      "step": 29330
    },
    {
      "entropy": 5.257406282424927,
      "epoch": 3.294603245914528,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003944443715157877,
      "loss": 4.8971,
      "mean_token_accuracy": 0.21803565323352814,
      "num_tokens": 65463764.0,
      "step": 29335
    },
    {
      "entropy": 5.181863117218017,
      "epoch": 3.2951648228224855,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003944103452971275,
      "loss": 4.8773,
      "mean_token_accuracy": 0.21447864919900894,
      "num_tokens": 65474367.0,
      "step": 29340
    },
    {
      "entropy": 5.283137035369873,
      "epoch": 3.295726399730443,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039437631527639553,
      "loss": 4.8553,
      "mean_token_accuracy": 0.23110000789165497,
      "num_tokens": 65485347.0,
      "step": 29345
    },
    {
      "entropy": 5.339414262771607,
      "epoch": 3.296287976638401,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00039434228145467554,
      "loss": 4.9899,
      "mean_token_accuracy": 0.2199369415640831,
      "num_tokens": 65495858.0,
      "step": 29350
    },
    {
      "entropy": 5.204788017272949,
      "epoch": 3.296849553546358,
      "grad_norm": 0.9375,
      "learning_rate": 0.000394308243833051,
      "loss": 4.7695,
      "mean_token_accuracy": 0.22865051925182342,
      "num_tokens": 65505649.0,
      "step": 29355
    },
    {
      "entropy": 5.2467429637908936,
      "epoch": 3.2974111304543157,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00039427420241260557,
      "loss": 4.9165,
      "mean_token_accuracy": 0.2213166132569313,
      "num_tokens": 65516437.0,
      "step": 29360
    },
    {
      "entropy": 5.190713548660279,
      "epoch": 3.2979727073622733,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00039424015719442335,
      "loss": 4.8675,
      "mean_token_accuracy": 0.2203418269753456,
      "num_tokens": 65528854.0,
      "step": 29365
    },
    {
      "entropy": 5.363630676269532,
      "epoch": 3.298534284270231,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039420610817958836,
      "loss": 4.9867,
      "mean_token_accuracy": 0.21775536835193635,
      "num_tokens": 65539459.0,
      "step": 29370
    },
    {
      "entropy": 5.340176105499268,
      "epoch": 3.299095861178188,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003941720553691846,
      "loss": 4.9636,
      "mean_token_accuracy": 0.21368683278560638,
      "num_tokens": 65550818.0,
      "step": 29375
    },
    {
      "entropy": 5.28668909072876,
      "epoch": 3.299657438086146,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003941379987642963,
      "loss": 4.8915,
      "mean_token_accuracy": 0.21873450875282288,
      "num_tokens": 65562572.0,
      "step": 29380
    },
    {
      "entropy": 5.25863995552063,
      "epoch": 3.3002190149941035,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000394103938366008,
      "loss": 4.9645,
      "mean_token_accuracy": 0.21726431995630263,
      "num_tokens": 65573764.0,
      "step": 29385
    },
    {
      "entropy": 5.355067014694214,
      "epoch": 3.300780591902061,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003940698741754041,
      "loss": 4.9581,
      "mean_token_accuracy": 0.2222646415233612,
      "num_tokens": 65585628.0,
      "step": 29390
    },
    {
      "entropy": 5.310615873336792,
      "epoch": 3.3013421688100184,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003940358061935692,
      "loss": 4.924,
      "mean_token_accuracy": 0.21774480491876602,
      "num_tokens": 65596142.0,
      "step": 29395
    },
    {
      "entropy": 5.386207103729248,
      "epoch": 3.301903745717976,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003940017344215881,
      "loss": 5.0836,
      "mean_token_accuracy": 0.206740665435791,
      "num_tokens": 65607824.0,
      "step": 29400
    },
    {
      "entropy": 5.287068939208984,
      "epoch": 3.3024653226259337,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039396765886054555,
      "loss": 4.9411,
      "mean_token_accuracy": 0.218595452606678,
      "num_tokens": 65617862.0,
      "step": 29405
    },
    {
      "entropy": 5.27408504486084,
      "epoch": 3.303026899533891,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00039393357951152667,
      "loss": 4.9208,
      "mean_token_accuracy": 0.2162480026483536,
      "num_tokens": 65629224.0,
      "step": 29410
    },
    {
      "entropy": 5.23193211555481,
      "epoch": 3.3035884764418486,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00039389949637561646,
      "loss": 4.8285,
      "mean_token_accuracy": 0.21933496445417405,
      "num_tokens": 65640419.0,
      "step": 29415
    },
    {
      "entropy": 5.308955192565918,
      "epoch": 3.3041500533498063,
      "grad_norm": 0.984375,
      "learning_rate": 0.00039386540945390017,
      "loss": 4.9467,
      "mean_token_accuracy": 0.21521318852901458,
      "num_tokens": 65651305.0,
      "step": 29420
    },
    {
      "entropy": 5.341869926452636,
      "epoch": 3.304711630257764,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003938313187474631,
      "loss": 5.0092,
      "mean_token_accuracy": 0.20905921310186387,
      "num_tokens": 65662702.0,
      "step": 29425
    },
    {
      "entropy": 5.2984339714050295,
      "epoch": 3.3052732071657216,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003937972242573908,
      "loss": 4.8936,
      "mean_token_accuracy": 0.21995642483234407,
      "num_tokens": 65673981.0,
      "step": 29430
    },
    {
      "entropy": 5.3143720626831055,
      "epoch": 3.305834784073679,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003937631259847688,
      "loss": 4.9427,
      "mean_token_accuracy": 0.2213476926088333,
      "num_tokens": 65684633.0,
      "step": 29435
    },
    {
      "entropy": 5.2875741481781,
      "epoch": 3.3063963609816365,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003937290239306828,
      "loss": 5.0182,
      "mean_token_accuracy": 0.21149489730596543,
      "num_tokens": 65696890.0,
      "step": 29440
    },
    {
      "entropy": 5.262507724761963,
      "epoch": 3.306957937889594,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039369491809621867,
      "loss": 4.899,
      "mean_token_accuracy": 0.2204977259039879,
      "num_tokens": 65708620.0,
      "step": 29445
    },
    {
      "entropy": 5.341099071502685,
      "epoch": 3.3075195147975514,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003936608084824623,
      "loss": 5.0009,
      "mean_token_accuracy": 0.21675412952899933,
      "num_tokens": 65720182.0,
      "step": 29450
    },
    {
      "entropy": 5.3373682498931885,
      "epoch": 3.308081091705509,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039362669509049975,
      "loss": 5.0266,
      "mean_token_accuracy": 0.20932487547397613,
      "num_tokens": 65731376.0,
      "step": 29455
    },
    {
      "entropy": 5.324780225753784,
      "epoch": 3.3086426686134667,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039359257792141717,
      "loss": 4.9376,
      "mean_token_accuracy": 0.2198083981871605,
      "num_tokens": 65743303.0,
      "step": 29460
    },
    {
      "entropy": 5.380134201049804,
      "epoch": 3.3092042455214243,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000393558456976301,
      "loss": 5.089,
      "mean_token_accuracy": 0.2080442115664482,
      "num_tokens": 65755486.0,
      "step": 29465
    },
    {
      "entropy": 5.247684526443481,
      "epoch": 3.3097658224293816,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003935243322562375,
      "loss": 4.9091,
      "mean_token_accuracy": 0.21362417489290236,
      "num_tokens": 65766855.0,
      "step": 29470
    },
    {
      "entropy": 5.281040239334106,
      "epoch": 3.310327399337339,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003934902037623133,
      "loss": 4.9129,
      "mean_token_accuracy": 0.21697085946798325,
      "num_tokens": 65778077.0,
      "step": 29475
    },
    {
      "entropy": 5.273123693466187,
      "epoch": 3.310888976245297,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00039345607149561513,
      "loss": 4.9483,
      "mean_token_accuracy": 0.21756281852722167,
      "num_tokens": 65791051.0,
      "step": 29480
    },
    {
      "entropy": 5.387137079238892,
      "epoch": 3.3114505531532545,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039342193545722966,
      "loss": 5.0501,
      "mean_token_accuracy": 0.2142600893974304,
      "num_tokens": 65802769.0,
      "step": 29485
    },
    {
      "entropy": 5.351698303222657,
      "epoch": 3.3120121300612118,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003933877956482438,
      "loss": 4.9354,
      "mean_token_accuracy": 0.22056228071451187,
      "num_tokens": 65813602.0,
      "step": 29490
    },
    {
      "entropy": 5.326882076263428,
      "epoch": 3.3125737069691694,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00039335365206974465,
      "loss": 5.0055,
      "mean_token_accuracy": 0.214094278216362,
      "num_tokens": 65826675.0,
      "step": 29495
    },
    {
      "entropy": 5.236273860931396,
      "epoch": 3.313135283877127,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003933195047228193,
      "loss": 4.8829,
      "mean_token_accuracy": 0.22622334957122803,
      "num_tokens": 65837643.0,
      "step": 29500
    },
    {
      "entropy": 5.362683868408203,
      "epoch": 3.3136968607850843,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039328535360855503,
      "loss": 4.9504,
      "mean_token_accuracy": 0.22372203171253205,
      "num_tokens": 65850184.0,
      "step": 29505
    },
    {
      "entropy": 5.328020238876343,
      "epoch": 3.314258437693042,
      "grad_norm": 1.0,
      "learning_rate": 0.00039325119872803924,
      "loss": 4.9621,
      "mean_token_accuracy": 0.21705697029829024,
      "num_tokens": 65861178.0,
      "step": 29510
    },
    {
      "entropy": 5.237962818145752,
      "epoch": 3.3148200146009996,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039321704008235943,
      "loss": 4.846,
      "mean_token_accuracy": 0.22844691723585128,
      "num_tokens": 65872458.0,
      "step": 29515
    },
    {
      "entropy": 5.237392377853394,
      "epoch": 3.3153815915089573,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039318287767260316,
      "loss": 4.8806,
      "mean_token_accuracy": 0.22113249450922012,
      "num_tokens": 65883152.0,
      "step": 29520
    },
    {
      "entropy": 5.267205429077149,
      "epoch": 3.315943168416915,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039314871149985834,
      "loss": 4.8438,
      "mean_token_accuracy": 0.2261584669351578,
      "num_tokens": 65894220.0,
      "step": 29525
    },
    {
      "entropy": 5.291937685012817,
      "epoch": 3.316504745324872,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003931145415652126,
      "loss": 4.9932,
      "mean_token_accuracy": 0.21053894162178038,
      "num_tokens": 65904159.0,
      "step": 29530
    },
    {
      "entropy": 5.398499536514282,
      "epoch": 3.31706632223283,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039308036786975416,
      "loss": 5.0264,
      "mean_token_accuracy": 0.20999467372894287,
      "num_tokens": 65915599.0,
      "step": 29535
    },
    {
      "entropy": 5.283306217193603,
      "epoch": 3.3176278991407875,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003930461904145709,
      "loss": 4.9476,
      "mean_token_accuracy": 0.21888077706098558,
      "num_tokens": 65926518.0,
      "step": 29540
    },
    {
      "entropy": 5.290559911727906,
      "epoch": 3.3181894760487447,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003930120092007512,
      "loss": 4.9674,
      "mean_token_accuracy": 0.2142360344529152,
      "num_tokens": 65938901.0,
      "step": 29545
    },
    {
      "entropy": 5.302507543563843,
      "epoch": 3.3187510529567024,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003929778242293834,
      "loss": 4.9209,
      "mean_token_accuracy": 0.21580000966787338,
      "num_tokens": 65949798.0,
      "step": 29550
    },
    {
      "entropy": 5.244705581665039,
      "epoch": 3.31931262986466,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003929436355015558,
      "loss": 4.8564,
      "mean_token_accuracy": 0.22844296246767043,
      "num_tokens": 65961222.0,
      "step": 29555
    },
    {
      "entropy": 5.378484630584717,
      "epoch": 3.3198742067726177,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00039290944301835723,
      "loss": 5.0391,
      "mean_token_accuracy": 0.20441722273826599,
      "num_tokens": 65972373.0,
      "step": 29560
    },
    {
      "entropy": 5.281479072570801,
      "epoch": 3.320435783680575,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039287524678087617,
      "loss": 4.9024,
      "mean_token_accuracy": 0.2224862813949585,
      "num_tokens": 65983716.0,
      "step": 29565
    },
    {
      "entropy": 5.373929023742676,
      "epoch": 3.3209973605885326,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039284104679020154,
      "loss": 5.0172,
      "mean_token_accuracy": 0.21907622814178468,
      "num_tokens": 65994827.0,
      "step": 29570
    },
    {
      "entropy": 5.292165899276734,
      "epoch": 3.3215589374964902,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003928068430474224,
      "loss": 4.8021,
      "mean_token_accuracy": 0.2274312421679497,
      "num_tokens": 66006249.0,
      "step": 29575
    },
    {
      "entropy": 5.305194473266601,
      "epoch": 3.322120514404448,
      "grad_norm": 1.0,
      "learning_rate": 0.0003927726355536275,
      "loss": 4.9277,
      "mean_token_accuracy": 0.2150719329714775,
      "num_tokens": 66017175.0,
      "step": 29580
    },
    {
      "entropy": 5.236071395874023,
      "epoch": 3.322682091312405,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00039273842430990634,
      "loss": 4.8552,
      "mean_token_accuracy": 0.22299974858760835,
      "num_tokens": 66028518.0,
      "step": 29585
    },
    {
      "entropy": 5.274035882949829,
      "epoch": 3.3232436682203628,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003927042093173481,
      "loss": 4.9278,
      "mean_token_accuracy": 0.21230306178331376,
      "num_tokens": 66038927.0,
      "step": 29590
    },
    {
      "entropy": 5.301504850387573,
      "epoch": 3.3238052451283204,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003926699905770421,
      "loss": 4.9013,
      "mean_token_accuracy": 0.2207379549741745,
      "num_tokens": 66050649.0,
      "step": 29595
    },
    {
      "entropy": 5.327165651321411,
      "epoch": 3.3243668220362776,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000392635768090078,
      "loss": 4.9603,
      "mean_token_accuracy": 0.21768250167369843,
      "num_tokens": 66061585.0,
      "step": 29600
    },
    {
      "entropy": 5.29525842666626,
      "epoch": 3.3249283989442353,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039260154185754545,
      "loss": 4.9039,
      "mean_token_accuracy": 0.22176303416490556,
      "num_tokens": 66071312.0,
      "step": 29605
    },
    {
      "entropy": 5.293570137023925,
      "epoch": 3.325489975852193,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003925673118805341,
      "loss": 4.9431,
      "mean_token_accuracy": 0.21961518377065659,
      "num_tokens": 66082643.0,
      "step": 29610
    },
    {
      "entropy": 5.250333261489868,
      "epoch": 3.3260515527601506,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000392533078160134,
      "loss": 4.829,
      "mean_token_accuracy": 0.22242804765701293,
      "num_tokens": 66094094.0,
      "step": 29615
    },
    {
      "entropy": 5.271414566040039,
      "epoch": 3.3266131296681083,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00039249884069743514,
      "loss": 4.9239,
      "mean_token_accuracy": 0.2191016986966133,
      "num_tokens": 66105365.0,
      "step": 29620
    },
    {
      "entropy": 5.3352045059204105,
      "epoch": 3.3271747065760655,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003924645994935276,
      "loss": 4.9696,
      "mean_token_accuracy": 0.2122862845659256,
      "num_tokens": 66115440.0,
      "step": 29625
    },
    {
      "entropy": 5.313116121292114,
      "epoch": 3.327736283484023,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003924303545495017,
      "loss": 4.9268,
      "mean_token_accuracy": 0.2160539746284485,
      "num_tokens": 66127213.0,
      "step": 29630
    },
    {
      "entropy": 5.280793857574463,
      "epoch": 3.328297860391981,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00039239610586644765,
      "loss": 4.8314,
      "mean_token_accuracy": 0.2249166563153267,
      "num_tokens": 66138016.0,
      "step": 29635
    },
    {
      "entropy": 5.193481206893921,
      "epoch": 3.328859437299938,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003923618534454561,
      "loss": 4.8486,
      "mean_token_accuracy": 0.2270383879542351,
      "num_tokens": 66149981.0,
      "step": 29640
    },
    {
      "entropy": 5.266609191894531,
      "epoch": 3.3294210142078957,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003923275972876176,
      "loss": 4.9462,
      "mean_token_accuracy": 0.22023501992225647,
      "num_tokens": 66161609.0,
      "step": 29645
    },
    {
      "entropy": 5.251606845855713,
      "epoch": 3.3299825911158534,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039229333739402303,
      "loss": 4.8484,
      "mean_token_accuracy": 0.22119930237531663,
      "num_tokens": 66171665.0,
      "step": 29650
    },
    {
      "entropy": 5.250096130371094,
      "epoch": 3.330544168023811,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000392259073765763,
      "loss": 4.8587,
      "mean_token_accuracy": 0.2197912573814392,
      "num_tokens": 66181773.0,
      "step": 29655
    },
    {
      "entropy": 5.316006755828857,
      "epoch": 3.3311057449317683,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039222480640392854,
      "loss": 4.9655,
      "mean_token_accuracy": 0.21996872872114182,
      "num_tokens": 66193578.0,
      "step": 29660
    },
    {
      "entropy": 5.367908906936646,
      "epoch": 3.331667321839726,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039219053530961087,
      "loss": 5.0,
      "mean_token_accuracy": 0.21073278933763503,
      "num_tokens": 66205185.0,
      "step": 29665
    },
    {
      "entropy": 5.22599606513977,
      "epoch": 3.3322288987476836,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039215626048390105,
      "loss": 4.882,
      "mean_token_accuracy": 0.22267740815877915,
      "num_tokens": 66217143.0,
      "step": 29670
    },
    {
      "entropy": 5.2528763771057125,
      "epoch": 3.3327904756556412,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003921219819278905,
      "loss": 4.9298,
      "mean_token_accuracy": 0.22581557780504227,
      "num_tokens": 66229022.0,
      "step": 29675
    },
    {
      "entropy": 5.3073896884918215,
      "epoch": 3.3333520525635985,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003920876996426706,
      "loss": 4.9845,
      "mean_token_accuracy": 0.21448229998350143,
      "num_tokens": 66240028.0,
      "step": 29680
    },
    {
      "entropy": 5.32898473739624,
      "epoch": 3.333913629471556,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039205341362933294,
      "loss": 4.8815,
      "mean_token_accuracy": 0.22534094750881195,
      "num_tokens": 66251656.0,
      "step": 29685
    },
    {
      "entropy": 5.261350345611572,
      "epoch": 3.334475206379514,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003920191238889691,
      "loss": 4.8933,
      "mean_token_accuracy": 0.22350490391254424,
      "num_tokens": 66262644.0,
      "step": 29690
    },
    {
      "entropy": 5.202312564849853,
      "epoch": 3.335036783287471,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003919848304226711,
      "loss": 4.8954,
      "mean_token_accuracy": 0.21879435926675797,
      "num_tokens": 66273374.0,
      "step": 29695
    },
    {
      "entropy": 5.298409509658813,
      "epoch": 3.3355983601954287,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003919505332315307,
      "loss": 4.9247,
      "mean_token_accuracy": 0.2206019401550293,
      "num_tokens": 66284490.0,
      "step": 29700
    },
    {
      "entropy": 5.364229774475097,
      "epoch": 3.3361599371033863,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039191623231663985,
      "loss": 4.9126,
      "mean_token_accuracy": 0.2184852421283722,
      "num_tokens": 66295555.0,
      "step": 29705
    },
    {
      "entropy": 5.267061710357666,
      "epoch": 3.336721514011344,
      "grad_norm": 0.984375,
      "learning_rate": 0.000391881927679091,
      "loss": 4.895,
      "mean_token_accuracy": 0.22059390246868132,
      "num_tokens": 66307234.0,
      "step": 29710
    },
    {
      "entropy": 5.331704616546631,
      "epoch": 3.3372830909193016,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039184761931997607,
      "loss": 5.0448,
      "mean_token_accuracy": 0.2067874163389206,
      "num_tokens": 66319239.0,
      "step": 29715
    },
    {
      "entropy": 5.3286376953125,
      "epoch": 3.337844667827259,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003918133072403878,
      "loss": 4.9941,
      "mean_token_accuracy": 0.21850164532661437,
      "num_tokens": 66332260.0,
      "step": 29720
    },
    {
      "entropy": 5.341793489456177,
      "epoch": 3.3384062447352165,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039177899144141835,
      "loss": 4.9841,
      "mean_token_accuracy": 0.2094061091542244,
      "num_tokens": 66343817.0,
      "step": 29725
    },
    {
      "entropy": 5.391280651092529,
      "epoch": 3.338967821643174,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003917446719241606,
      "loss": 4.9864,
      "mean_token_accuracy": 0.2120492249727249,
      "num_tokens": 66355724.0,
      "step": 29730
    },
    {
      "entropy": 5.275605773925781,
      "epoch": 3.3395293985511314,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039171034868970724,
      "loss": 4.8664,
      "mean_token_accuracy": 0.22136332988739013,
      "num_tokens": 66365994.0,
      "step": 29735
    },
    {
      "entropy": 5.284602785110474,
      "epoch": 3.340090975459089,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00039167602173915104,
      "loss": 5.0683,
      "mean_token_accuracy": 0.21000848710536957,
      "num_tokens": 66377929.0,
      "step": 29740
    },
    {
      "entropy": 5.276469421386719,
      "epoch": 3.3406525523670467,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003916416910735851,
      "loss": 4.8978,
      "mean_token_accuracy": 0.222618705034256,
      "num_tokens": 66389292.0,
      "step": 29745
    },
    {
      "entropy": 5.318738126754761,
      "epoch": 3.3412141292750044,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039160735669410246,
      "loss": 4.9199,
      "mean_token_accuracy": 0.21987851411104203,
      "num_tokens": 66399266.0,
      "step": 29750
    },
    {
      "entropy": 5.245822429656982,
      "epoch": 3.3417757061829616,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003915730186017964,
      "loss": 4.9497,
      "mean_token_accuracy": 0.21542029529809953,
      "num_tokens": 66410718.0,
      "step": 29755
    },
    {
      "entropy": 5.2777904033660885,
      "epoch": 3.3423372830909193,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00039153867679776023,
      "loss": 4.9801,
      "mean_token_accuracy": 0.21866793185472488,
      "num_tokens": 66420982.0,
      "step": 29760
    },
    {
      "entropy": 5.336616039276123,
      "epoch": 3.342898859998877,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003915043312830873,
      "loss": 4.9288,
      "mean_token_accuracy": 0.22403363287448883,
      "num_tokens": 66431378.0,
      "step": 29765
    },
    {
      "entropy": 5.336844491958618,
      "epoch": 3.3434604369068346,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003914699820588714,
      "loss": 4.9315,
      "mean_token_accuracy": 0.2159969240427017,
      "num_tokens": 66442033.0,
      "step": 29770
    },
    {
      "entropy": 5.2348237991333,
      "epoch": 3.344022013814792,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039143562912620603,
      "loss": 4.8641,
      "mean_token_accuracy": 0.2213260769844055,
      "num_tokens": 66453486.0,
      "step": 29775
    },
    {
      "entropy": 5.321174812316895,
      "epoch": 3.3445835907227495,
      "grad_norm": 1.03125,
      "learning_rate": 0.000391401272486185,
      "loss": 5.019,
      "mean_token_accuracy": 0.21273980289697647,
      "num_tokens": 66464931.0,
      "step": 29780
    },
    {
      "entropy": 5.510326337814331,
      "epoch": 3.345145167630707,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003913669121399024,
      "loss": 5.13,
      "mean_token_accuracy": 0.2092198207974434,
      "num_tokens": 66477187.0,
      "step": 29785
    },
    {
      "entropy": 5.3183714866638185,
      "epoch": 3.3457067445386643,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039133254808845216,
      "loss": 4.9714,
      "mean_token_accuracy": 0.2212025299668312,
      "num_tokens": 66488777.0,
      "step": 29790
    },
    {
      "entropy": 5.2676016330719,
      "epoch": 3.346268321446622,
      "grad_norm": 0.921875,
      "learning_rate": 0.00039129818033292843,
      "loss": 4.9317,
      "mean_token_accuracy": 0.21875543892383575,
      "num_tokens": 66499576.0,
      "step": 29795
    },
    {
      "entropy": 5.239945793151856,
      "epoch": 3.3468298983545797,
      "grad_norm": 1.609375,
      "learning_rate": 0.0003912638088744256,
      "loss": 4.759,
      "mean_token_accuracy": 0.24029300808906556,
      "num_tokens": 66510824.0,
      "step": 29800
    },
    {
      "entropy": 5.348616456985473,
      "epoch": 3.3473914752625373,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000391229433714038,
      "loss": 4.9849,
      "mean_token_accuracy": 0.21605535298585893,
      "num_tokens": 66523151.0,
      "step": 29805
    },
    {
      "entropy": 5.214786338806152,
      "epoch": 3.347953052170495,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003911950548528601,
      "loss": 4.8491,
      "mean_token_accuracy": 0.22497001439332961,
      "num_tokens": 66533921.0,
      "step": 29810
    },
    {
      "entropy": 5.219886922836304,
      "epoch": 3.348514629078452,
      "grad_norm": 0.859375,
      "learning_rate": 0.0003911606722919866,
      "loss": 4.8952,
      "mean_token_accuracy": 0.2193068355321884,
      "num_tokens": 66545867.0,
      "step": 29815
    },
    {
      "entropy": 5.299258613586426,
      "epoch": 3.34907620598641,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003911262860325123,
      "loss": 4.9012,
      "mean_token_accuracy": 0.22706687301397324,
      "num_tokens": 66557522.0,
      "step": 29820
    },
    {
      "entropy": 5.286866140365601,
      "epoch": 3.3496377828943675,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003910918960755319,
      "loss": 4.9233,
      "mean_token_accuracy": 0.2232730805873871,
      "num_tokens": 66568792.0,
      "step": 29825
    },
    {
      "entropy": 5.315064334869385,
      "epoch": 3.3501993598023248,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003910575024221406,
      "loss": 5.0068,
      "mean_token_accuracy": 0.2210424065589905,
      "num_tokens": 66580231.0,
      "step": 29830
    },
    {
      "entropy": 5.224500417709351,
      "epoch": 3.3507609367102824,
      "grad_norm": 0.953125,
      "learning_rate": 0.00039102310507343335,
      "loss": 4.8332,
      "mean_token_accuracy": 0.22472823560237884,
      "num_tokens": 66591171.0,
      "step": 29835
    },
    {
      "entropy": 5.287578916549682,
      "epoch": 3.35132251361824,
      "grad_norm": 1.125,
      "learning_rate": 0.0003909887040305055,
      "loss": 4.9682,
      "mean_token_accuracy": 0.2117022916674614,
      "num_tokens": 66602676.0,
      "step": 29840
    },
    {
      "entropy": 5.2708837509155275,
      "epoch": 3.3518840905261977,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003909542992944523,
      "loss": 4.9046,
      "mean_token_accuracy": 0.22028473317623137,
      "num_tokens": 66613526.0,
      "step": 29845
    },
    {
      "entropy": 5.248419666290284,
      "epoch": 3.352445667434155,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003909198908663692,
      "loss": 4.8787,
      "mean_token_accuracy": 0.22023431211709976,
      "num_tokens": 66624279.0,
      "step": 29850
    },
    {
      "entropy": 5.28839144706726,
      "epoch": 3.3530072443421126,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039088547874735187,
      "loss": 4.8839,
      "mean_token_accuracy": 0.21917184442281723,
      "num_tokens": 66636908.0,
      "step": 29855
    },
    {
      "entropy": 5.415254020690918,
      "epoch": 3.3535688212500703,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00039085106293849594,
      "loss": 5.0795,
      "mean_token_accuracy": 0.2036385491490364,
      "num_tokens": 66648545.0,
      "step": 29860
    },
    {
      "entropy": 5.33657169342041,
      "epoch": 3.354130398158028,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039081664344089724,
      "loss": 4.9374,
      "mean_token_accuracy": 0.21902300268411637,
      "num_tokens": 66659245.0,
      "step": 29865
    },
    {
      "entropy": 5.213165807723999,
      "epoch": 3.354691975065985,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003907822202556516,
      "loss": 4.848,
      "mean_token_accuracy": 0.2242389962077141,
      "num_tokens": 66670866.0,
      "step": 29870
    },
    {
      "entropy": 5.302959012985229,
      "epoch": 3.355253551973943,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003907477933838553,
      "loss": 4.9737,
      "mean_token_accuracy": 0.21618662923574447,
      "num_tokens": 66681970.0,
      "step": 29875
    },
    {
      "entropy": 5.388374280929566,
      "epoch": 3.3558151288819005,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039071336282660437,
      "loss": 5.0019,
      "mean_token_accuracy": 0.21536251306533813,
      "num_tokens": 66693252.0,
      "step": 29880
    },
    {
      "entropy": 5.259941339492798,
      "epoch": 3.3563767057898577,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00039067892858499496,
      "loss": 4.9096,
      "mean_token_accuracy": 0.21680019199848174,
      "num_tokens": 66703821.0,
      "step": 29885
    },
    {
      "entropy": 5.235723495483398,
      "epoch": 3.3569382826978154,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003906444906601237,
      "loss": 4.8716,
      "mean_token_accuracy": 0.2243185430765152,
      "num_tokens": 66715945.0,
      "step": 29890
    },
    {
      "entropy": 5.319207096099854,
      "epoch": 3.357499859605773,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000390610049053087,
      "loss": 4.9626,
      "mean_token_accuracy": 0.22012477219104767,
      "num_tokens": 66727551.0,
      "step": 29895
    },
    {
      "entropy": 5.322775840759277,
      "epoch": 3.3580614365137307,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039057560376498147,
      "loss": 4.9323,
      "mean_token_accuracy": 0.22021580189466478,
      "num_tokens": 66738257.0,
      "step": 29900
    },
    {
      "entropy": 5.412526559829712,
      "epoch": 3.3586230134216883,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003905411547969038,
      "loss": 5.0503,
      "mean_token_accuracy": 0.20330336987972258,
      "num_tokens": 66750305.0,
      "step": 29905
    },
    {
      "entropy": 5.354148960113525,
      "epoch": 3.3591845903296456,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003905067021499511,
      "loss": 4.9496,
      "mean_token_accuracy": 0.21244880706071853,
      "num_tokens": 66761659.0,
      "step": 29910
    },
    {
      "entropy": 5.277486085891724,
      "epoch": 3.359746167237603,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039047224582522015,
      "loss": 4.8248,
      "mean_token_accuracy": 0.23013193607330323,
      "num_tokens": 66771981.0,
      "step": 29915
    },
    {
      "entropy": 5.288273715972901,
      "epoch": 3.360307744145561,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003904377858238081,
      "loss": 5.0428,
      "mean_token_accuracy": 0.21086501479148864,
      "num_tokens": 66783915.0,
      "step": 29920
    },
    {
      "entropy": 5.257399845123291,
      "epoch": 3.360869321053518,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003904033221468122,
      "loss": 4.9211,
      "mean_token_accuracy": 0.2200329601764679,
      "num_tokens": 66795691.0,
      "step": 29925
    },
    {
      "entropy": 5.326707887649536,
      "epoch": 3.3614308979614758,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003903688547953297,
      "loss": 4.9866,
      "mean_token_accuracy": 0.21464476287364959,
      "num_tokens": 66805792.0,
      "step": 29930
    },
    {
      "entropy": 5.281932163238525,
      "epoch": 3.3619924748694334,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003903343837704582,
      "loss": 4.887,
      "mean_token_accuracy": 0.2233949452638626,
      "num_tokens": 66817657.0,
      "step": 29935
    },
    {
      "entropy": 5.248887014389038,
      "epoch": 3.362554051777391,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003902999090732951,
      "loss": 4.8153,
      "mean_token_accuracy": 0.22550482749938966,
      "num_tokens": 66828593.0,
      "step": 29940
    },
    {
      "entropy": 5.226264142990113,
      "epoch": 3.3631156286853483,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039026543070493823,
      "loss": 4.9224,
      "mean_token_accuracy": 0.22367295920848845,
      "num_tokens": 66839405.0,
      "step": 29945
    },
    {
      "entropy": 5.23192982673645,
      "epoch": 3.363677205593306,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003902309486664854,
      "loss": 4.841,
      "mean_token_accuracy": 0.21939224898815154,
      "num_tokens": 66849938.0,
      "step": 29950
    },
    {
      "entropy": 5.306934452056884,
      "epoch": 3.3642387825012636,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003901964629590344,
      "loss": 4.9148,
      "mean_token_accuracy": 0.21858424842357635,
      "num_tokens": 66861105.0,
      "step": 29955
    },
    {
      "entropy": 5.246570777893067,
      "epoch": 3.3648003594092213,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003901619735836833,
      "loss": 4.8815,
      "mean_token_accuracy": 0.2192879244685173,
      "num_tokens": 66873065.0,
      "step": 29960
    },
    {
      "entropy": 5.426685571670532,
      "epoch": 3.3653619363171785,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00039012748054153044,
      "loss": 5.0948,
      "mean_token_accuracy": 0.2083771273493767,
      "num_tokens": 66885639.0,
      "step": 29965
    },
    {
      "entropy": 5.3208507061004635,
      "epoch": 3.365923513225136,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003900929838336738,
      "loss": 4.9979,
      "mean_token_accuracy": 0.2084314629435539,
      "num_tokens": 66896576.0,
      "step": 29970
    },
    {
      "entropy": 5.271938943862915,
      "epoch": 3.366485090133094,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003900584834612119,
      "loss": 4.8692,
      "mean_token_accuracy": 0.21975520849227906,
      "num_tokens": 66907595.0,
      "step": 29975
    },
    {
      "entropy": 5.366647386550904,
      "epoch": 3.367046667041051,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003900239794252434,
      "loss": 4.9377,
      "mean_token_accuracy": 0.21536162495613098,
      "num_tokens": 66918231.0,
      "step": 29980
    },
    {
      "entropy": 5.385846710205078,
      "epoch": 3.3676082439490087,
      "grad_norm": 1.0,
      "learning_rate": 0.0003899894717268667,
      "loss": 5.1051,
      "mean_token_accuracy": 0.20358218103647233,
      "num_tokens": 66930929.0,
      "step": 29985
    },
    {
      "entropy": 5.305954837799073,
      "epoch": 3.3681698208569664,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003899549603671807,
      "loss": 4.9652,
      "mean_token_accuracy": 0.2187981516122818,
      "num_tokens": 66941714.0,
      "step": 29990
    },
    {
      "entropy": 5.318918085098266,
      "epoch": 3.368731397764924,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038992044534728404,
      "loss": 4.9975,
      "mean_token_accuracy": 0.2118672952055931,
      "num_tokens": 66951732.0,
      "step": 29995
    },
    {
      "entropy": 5.3340410709381105,
      "epoch": 3.3692929746728817,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038988592666827594,
      "loss": 4.9625,
      "mean_token_accuracy": 0.21413425207138062,
      "num_tokens": 66962941.0,
      "step": 30000
    },
    {
      "epoch": 3.3692929746728817,
      "eval_entropy": 5.1608845611377445,
      "eval_loss": 5.107177734375,
      "eval_mean_token_accuracy": 0.21762566324751004,
      "eval_num_tokens": 66962941.0,
      "eval_runtime": 23.8276,
      "eval_samples_per_second": 1301.558,
      "eval_steps_per_second": 162.71,
      "step": 30000
    },
    {
      "entropy": 5.400703048706054,
      "epoch": 3.369854551580839,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003898514043312553,
      "loss": 5.067,
      "mean_token_accuracy": 0.21500703394412995,
      "num_tokens": 66975096.0,
      "step": 30005
    },
    {
      "entropy": 5.347104406356811,
      "epoch": 3.3704161284887966,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038981687833732146,
      "loss": 4.9608,
      "mean_token_accuracy": 0.22076522558927536,
      "num_tokens": 66987621.0,
      "step": 30010
    },
    {
      "entropy": 5.335265111923218,
      "epoch": 3.3709777053967542,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003897823486875737,
      "loss": 5.0224,
      "mean_token_accuracy": 0.21627651602029802,
      "num_tokens": 66998801.0,
      "step": 30015
    },
    {
      "entropy": 5.353843021392822,
      "epoch": 3.3715392823047114,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003897478153831114,
      "loss": 5.0118,
      "mean_token_accuracy": 0.21219913512468339,
      "num_tokens": 67010698.0,
      "step": 30020
    },
    {
      "entropy": 5.263390970230103,
      "epoch": 3.372100859212669,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038971327842503414,
      "loss": 4.879,
      "mean_token_accuracy": 0.22527946680784225,
      "num_tokens": 67022487.0,
      "step": 30025
    },
    {
      "entropy": 5.253789234161377,
      "epoch": 3.3726624361206268,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038967873781444167,
      "loss": 4.9097,
      "mean_token_accuracy": 0.22390564531087875,
      "num_tokens": 67032517.0,
      "step": 30030
    },
    {
      "entropy": 5.267511415481567,
      "epoch": 3.3732240130285844,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038964419355243365,
      "loss": 4.8597,
      "mean_token_accuracy": 0.2252635881304741,
      "num_tokens": 67042466.0,
      "step": 30035
    },
    {
      "entropy": 5.347834825515747,
      "epoch": 3.3737855899365417,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003896096456401101,
      "loss": 4.9955,
      "mean_token_accuracy": 0.21758337914943696,
      "num_tokens": 67053090.0,
      "step": 30040
    },
    {
      "entropy": 5.273881149291992,
      "epoch": 3.3743471668444993,
      "grad_norm": 0.953125,
      "learning_rate": 0.00038957509407857097,
      "loss": 4.8295,
      "mean_token_accuracy": 0.22282977551221847,
      "num_tokens": 67064332.0,
      "step": 30045
    },
    {
      "entropy": 5.246058368682862,
      "epoch": 3.374908743752457,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00038954053886891635,
      "loss": 4.8824,
      "mean_token_accuracy": 0.2163597211241722,
      "num_tokens": 67074788.0,
      "step": 30050
    },
    {
      "entropy": 5.258951997756958,
      "epoch": 3.3754703206604146,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003895059800122466,
      "loss": 4.9418,
      "mean_token_accuracy": 0.21964870691299437,
      "num_tokens": 67086510.0,
      "step": 30055
    },
    {
      "entropy": 5.302891635894776,
      "epoch": 3.376031897568372,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000389471417509662,
      "loss": 4.9294,
      "mean_token_accuracy": 0.21326974928379058,
      "num_tokens": 67096561.0,
      "step": 30060
    },
    {
      "entropy": 5.2621214389801025,
      "epoch": 3.3765934744763295,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000389436851362263,
      "loss": 4.8781,
      "mean_token_accuracy": 0.22064272463321685,
      "num_tokens": 67107897.0,
      "step": 30065
    },
    {
      "entropy": 5.273105192184448,
      "epoch": 3.377155051384287,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038940228157115036,
      "loss": 5.0104,
      "mean_token_accuracy": 0.2090381056070328,
      "num_tokens": 67119686.0,
      "step": 30070
    },
    {
      "entropy": 5.323588466644287,
      "epoch": 3.3777166282922444,
      "grad_norm": 0.9375,
      "learning_rate": 0.00038936770813742473,
      "loss": 4.9181,
      "mean_token_accuracy": 0.2150008797645569,
      "num_tokens": 67131041.0,
      "step": 30075
    },
    {
      "entropy": 5.409162998199463,
      "epoch": 3.378278205200202,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00038933313106218694,
      "loss": 5.03,
      "mean_token_accuracy": 0.22077602446079253,
      "num_tokens": 67143065.0,
      "step": 30080
    },
    {
      "entropy": 5.262815237045288,
      "epoch": 3.3788397821081597,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038929855034653776,
      "loss": 4.8634,
      "mean_token_accuracy": 0.2219126507639885,
      "num_tokens": 67154173.0,
      "step": 30085
    },
    {
      "entropy": 5.273532199859619,
      "epoch": 3.3794013590161174,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038926396599157855,
      "loss": 4.9304,
      "mean_token_accuracy": 0.21300324201583862,
      "num_tokens": 67165909.0,
      "step": 30090
    },
    {
      "entropy": 5.269088840484619,
      "epoch": 3.379962935924075,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003892293779984103,
      "loss": 4.9041,
      "mean_token_accuracy": 0.2218851327896118,
      "num_tokens": 67177474.0,
      "step": 30095
    },
    {
      "entropy": 5.360263204574585,
      "epoch": 3.3805245128320323,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003891947863681344,
      "loss": 5.0003,
      "mean_token_accuracy": 0.21311191320419312,
      "num_tokens": 67187590.0,
      "step": 30100
    },
    {
      "entropy": 5.26072211265564,
      "epoch": 3.38108608973999,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003891601911018521,
      "loss": 4.8791,
      "mean_token_accuracy": 0.21495928913354873,
      "num_tokens": 67198579.0,
      "step": 30105
    },
    {
      "entropy": 5.271651744842529,
      "epoch": 3.3816476666479476,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003891255922006652,
      "loss": 4.8994,
      "mean_token_accuracy": 0.22605673521757125,
      "num_tokens": 67208503.0,
      "step": 30110
    },
    {
      "entropy": 5.232346487045288,
      "epoch": 3.382209243555905,
      "grad_norm": 1.015625,
      "learning_rate": 0.000389090989665675,
      "loss": 4.8686,
      "mean_token_accuracy": 0.21860363930463791,
      "num_tokens": 67219363.0,
      "step": 30115
    },
    {
      "entropy": 5.202801465988159,
      "epoch": 3.3827708204638625,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003890563834979835,
      "loss": 4.965,
      "mean_token_accuracy": 0.21529794186353685,
      "num_tokens": 67231030.0,
      "step": 30120
    },
    {
      "entropy": 5.325349569320679,
      "epoch": 3.38333239737182,
      "grad_norm": 1.0,
      "learning_rate": 0.0003890217736986926,
      "loss": 4.9189,
      "mean_token_accuracy": 0.21490842401981353,
      "num_tokens": 67242962.0,
      "step": 30125
    },
    {
      "entropy": 5.331257152557373,
      "epoch": 3.383893974279778,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00038898716026890414,
      "loss": 4.9363,
      "mean_token_accuracy": 0.21532754898071288,
      "num_tokens": 67253871.0,
      "step": 30130
    },
    {
      "entropy": 5.313115501403809,
      "epoch": 3.384455551187735,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003889525432097203,
      "loss": 4.9223,
      "mean_token_accuracy": 0.2191022664308548,
      "num_tokens": 67265875.0,
      "step": 30135
    },
    {
      "entropy": 5.273467063903809,
      "epoch": 3.3850171280956927,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003889179225222432,
      "loss": 4.9074,
      "mean_token_accuracy": 0.2213152155280113,
      "num_tokens": 67274920.0,
      "step": 30140
    },
    {
      "entropy": 5.240643739700317,
      "epoch": 3.3855787050036503,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00038888329820757536,
      "loss": 4.9255,
      "mean_token_accuracy": 0.22089312225580215,
      "num_tokens": 67286440.0,
      "step": 30145
    },
    {
      "entropy": 5.245520448684692,
      "epoch": 3.386140281911608,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038884867026681906,
      "loss": 4.8859,
      "mean_token_accuracy": 0.22610736638307571,
      "num_tokens": 67297803.0,
      "step": 30150
    },
    {
      "entropy": 5.177592849731445,
      "epoch": 3.386701858819565,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003888140387010771,
      "loss": 4.8023,
      "mean_token_accuracy": 0.22794388830661774,
      "num_tokens": 67309153.0,
      "step": 30155
    },
    {
      "entropy": 5.186595106124878,
      "epoch": 3.387263435727523,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038877940351145186,
      "loss": 4.886,
      "mean_token_accuracy": 0.22236080169677735,
      "num_tokens": 67319695.0,
      "step": 30160
    },
    {
      "entropy": 5.251593208312988,
      "epoch": 3.3878250126354805,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00038874476469904633,
      "loss": 4.8851,
      "mean_token_accuracy": 0.22634855508804322,
      "num_tokens": 67331615.0,
      "step": 30165
    },
    {
      "entropy": 5.251108312606812,
      "epoch": 3.3883865895434377,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003887101222649634,
      "loss": 4.7933,
      "mean_token_accuracy": 0.22959545701742173,
      "num_tokens": 67341898.0,
      "step": 30170
    },
    {
      "entropy": 5.328473806381226,
      "epoch": 3.3889481664513954,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000388675476210306,
      "loss": 4.9761,
      "mean_token_accuracy": 0.21892083138227464,
      "num_tokens": 67352677.0,
      "step": 30175
    },
    {
      "entropy": 5.24788727760315,
      "epoch": 3.389509743359353,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038864082653617744,
      "loss": 4.9144,
      "mean_token_accuracy": 0.2168678119778633,
      "num_tokens": 67363215.0,
      "step": 30180
    },
    {
      "entropy": 5.223884439468383,
      "epoch": 3.3900713202673107,
      "grad_norm": 0.921875,
      "learning_rate": 0.00038860617324368085,
      "loss": 4.9238,
      "mean_token_accuracy": 0.2192314624786377,
      "num_tokens": 67374489.0,
      "step": 30185
    },
    {
      "entropy": 5.330619525909424,
      "epoch": 3.3906328971752684,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003885715163339196,
      "loss": 4.9366,
      "mean_token_accuracy": 0.22265529930591582,
      "num_tokens": 67384890.0,
      "step": 30190
    },
    {
      "entropy": 5.275802373886108,
      "epoch": 3.3911944740832256,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00038853685580799726,
      "loss": 4.9024,
      "mean_token_accuracy": 0.21881698518991471,
      "num_tokens": 67395811.0,
      "step": 30195
    },
    {
      "entropy": 5.272500705718994,
      "epoch": 3.3917560509911833,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003885021916670174,
      "loss": 4.8475,
      "mean_token_accuracy": 0.22722146213054656,
      "num_tokens": 67407056.0,
      "step": 30200
    },
    {
      "entropy": 5.307015991210937,
      "epoch": 3.392317627899141,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003884675239120838,
      "loss": 4.9486,
      "mean_token_accuracy": 0.2176618054509163,
      "num_tokens": 67418371.0,
      "step": 30205
    },
    {
      "entropy": 5.286378049850464,
      "epoch": 3.392879204807098,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00038843285254430005,
      "loss": 4.9314,
      "mean_token_accuracy": 0.21653366684913636,
      "num_tokens": 67429165.0,
      "step": 30210
    },
    {
      "entropy": 5.473616027832032,
      "epoch": 3.393440781715056,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003883981775647705,
      "loss": 5.1452,
      "mean_token_accuracy": 0.2101510152220726,
      "num_tokens": 67442317.0,
      "step": 30215
    },
    {
      "entropy": 5.414137458801269,
      "epoch": 3.3940023586230135,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003883634989745988,
      "loss": 5.0283,
      "mean_token_accuracy": 0.22013490200042723,
      "num_tokens": 67454086.0,
      "step": 30220
    },
    {
      "entropy": 5.334450626373291,
      "epoch": 3.394563935530971,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038832881677488944,
      "loss": 4.9933,
      "mean_token_accuracy": 0.21190305352210997,
      "num_tokens": 67465066.0,
      "step": 30225
    },
    {
      "entropy": 5.205579423904419,
      "epoch": 3.3951255124389284,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038829413096674653,
      "loss": 4.8559,
      "mean_token_accuracy": 0.22687539011240004,
      "num_tokens": 67475456.0,
      "step": 30230
    },
    {
      "entropy": 5.275921726226807,
      "epoch": 3.395687089346886,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003882594415512746,
      "loss": 4.9038,
      "mean_token_accuracy": 0.21883978098630905,
      "num_tokens": 67486315.0,
      "step": 30235
    },
    {
      "entropy": 5.2521298885345455,
      "epoch": 3.3962486662548437,
      "grad_norm": 0.96875,
      "learning_rate": 0.00038822474852957817,
      "loss": 4.8364,
      "mean_token_accuracy": 0.22677484899759293,
      "num_tokens": 67497431.0,
      "step": 30240
    },
    {
      "entropy": 5.282762861251831,
      "epoch": 3.3968102431628013,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003881900519027618,
      "loss": 4.9384,
      "mean_token_accuracy": 0.22125106006860734,
      "num_tokens": 67508054.0,
      "step": 30245
    },
    {
      "entropy": 5.313575601577758,
      "epoch": 3.3973718200707586,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003881553516719303,
      "loss": 4.9735,
      "mean_token_accuracy": 0.21272627860307694,
      "num_tokens": 67519083.0,
      "step": 30250
    },
    {
      "entropy": 5.356269979476929,
      "epoch": 3.397933396978716,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00038812064783818847,
      "loss": 4.973,
      "mean_token_accuracy": 0.21806224137544633,
      "num_tokens": 67528909.0,
      "step": 30255
    },
    {
      "entropy": 5.2241730213165285,
      "epoch": 3.398494973886674,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038808594040264127,
      "loss": 4.7733,
      "mean_token_accuracy": 0.2365007370710373,
      "num_tokens": 67540069.0,
      "step": 30260
    },
    {
      "entropy": 5.220929145812988,
      "epoch": 3.399056550794631,
      "grad_norm": 0.96875,
      "learning_rate": 0.000388051229366394,
      "loss": 4.9093,
      "mean_token_accuracy": 0.22116651982069016,
      "num_tokens": 67550733.0,
      "step": 30265
    },
    {
      "entropy": 5.367205715179443,
      "epoch": 3.3996181277025888,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00038801651473055167,
      "loss": 5.0289,
      "mean_token_accuracy": 0.21674747318029403,
      "num_tokens": 67562118.0,
      "step": 30270
    },
    {
      "entropy": 5.282076644897461,
      "epoch": 3.4001797046105464,
      "grad_norm": 1.125,
      "learning_rate": 0.0003879817964962197,
      "loss": 4.8618,
      "mean_token_accuracy": 0.2256859451532364,
      "num_tokens": 67572408.0,
      "step": 30275
    },
    {
      "entropy": 5.3435461044311525,
      "epoch": 3.400741281518504,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003879470746645035,
      "loss": 5.0473,
      "mean_token_accuracy": 0.21574695706367492,
      "num_tokens": 67584125.0,
      "step": 30280
    },
    {
      "entropy": 5.313580131530761,
      "epoch": 3.4013028584264617,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038791234923650866,
      "loss": 4.9109,
      "mean_token_accuracy": 0.2194651335477829,
      "num_tokens": 67595093.0,
      "step": 30285
    },
    {
      "entropy": 5.333483600616455,
      "epoch": 3.401864435334419,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038787762021334077,
      "loss": 4.9879,
      "mean_token_accuracy": 0.21241582036018372,
      "num_tokens": 67606794.0,
      "step": 30290
    },
    {
      "entropy": 5.2101274013519285,
      "epoch": 3.4024260122423766,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003878428875961057,
      "loss": 4.7104,
      "mean_token_accuracy": 0.2315253034234047,
      "num_tokens": 67617207.0,
      "step": 30295
    },
    {
      "entropy": 5.178969001770019,
      "epoch": 3.4029875891503343,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00038780815138590935,
      "loss": 4.8492,
      "mean_token_accuracy": 0.22166278064250947,
      "num_tokens": 67628301.0,
      "step": 30300
    },
    {
      "entropy": 5.305446577072144,
      "epoch": 3.4035491660582915,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003877734115838577,
      "loss": 4.9588,
      "mean_token_accuracy": 0.21627208441495896,
      "num_tokens": 67640363.0,
      "step": 30305
    },
    {
      "entropy": 5.344685506820679,
      "epoch": 3.404110742966249,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00038773866819105684,
      "loss": 4.9686,
      "mean_token_accuracy": 0.2151542827486992,
      "num_tokens": 67651515.0,
      "step": 30310
    },
    {
      "entropy": 5.2783770084381105,
      "epoch": 3.404672319874207,
      "grad_norm": 1.0,
      "learning_rate": 0.0003877039212086131,
      "loss": 4.9006,
      "mean_token_accuracy": 0.22280460745096206,
      "num_tokens": 67661669.0,
      "step": 30315
    },
    {
      "entropy": 5.234801816940307,
      "epoch": 3.4052338967821645,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038766917063763287,
      "loss": 4.8422,
      "mean_token_accuracy": 0.22416077107191085,
      "num_tokens": 67671945.0,
      "step": 30320
    },
    {
      "entropy": 5.253852033615113,
      "epoch": 3.4057954736901217,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00038763441647922246,
      "loss": 4.8819,
      "mean_token_accuracy": 0.2201020434498787,
      "num_tokens": 67683090.0,
      "step": 30325
    },
    {
      "entropy": 5.301914358139038,
      "epoch": 3.4063570505980794,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003875996587344885,
      "loss": 4.9384,
      "mean_token_accuracy": 0.22330522388219834,
      "num_tokens": 67694036.0,
      "step": 30330
    },
    {
      "entropy": 5.301812028884887,
      "epoch": 3.406918627506037,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038756489740453785,
      "loss": 4.9126,
      "mean_token_accuracy": 0.22040043622255326,
      "num_tokens": 67704677.0,
      "step": 30335
    },
    {
      "entropy": 5.278276491165161,
      "epoch": 3.4074802044139947,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038753013249047706,
      "loss": 4.9559,
      "mean_token_accuracy": 0.21808010339736938,
      "num_tokens": 67715165.0,
      "step": 30340
    },
    {
      "entropy": 5.266901588439941,
      "epoch": 3.408041781321952,
      "grad_norm": 1.0,
      "learning_rate": 0.0003874953639934134,
      "loss": 4.8805,
      "mean_token_accuracy": 0.2314671605825424,
      "num_tokens": 67726226.0,
      "step": 30345
    },
    {
      "entropy": 5.293924570083618,
      "epoch": 3.4086033582299096,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003874605919144536,
      "loss": 4.8717,
      "mean_token_accuracy": 0.22526600658893586,
      "num_tokens": 67739564.0,
      "step": 30350
    },
    {
      "entropy": 5.253958606719971,
      "epoch": 3.4091649351378672,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003874258162547049,
      "loss": 5.0137,
      "mean_token_accuracy": 0.2181858316063881,
      "num_tokens": 67751683.0,
      "step": 30355
    },
    {
      "entropy": 5.304076719284057,
      "epoch": 3.4097265120458244,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003873910370152747,
      "loss": 4.9404,
      "mean_token_accuracy": 0.2217478945851326,
      "num_tokens": 67762782.0,
      "step": 30360
    },
    {
      "entropy": 5.321379518508911,
      "epoch": 3.410288088953782,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003873562541972702,
      "loss": 4.9276,
      "mean_token_accuracy": 0.21509518027305602,
      "num_tokens": 67773583.0,
      "step": 30365
    },
    {
      "entropy": 5.285723352432251,
      "epoch": 3.4108496658617398,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003873214678017991,
      "loss": 4.9036,
      "mean_token_accuracy": 0.22340618818998337,
      "num_tokens": 67785428.0,
      "step": 30370
    },
    {
      "entropy": 5.352050256729126,
      "epoch": 3.4114112427696974,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003872866778299688,
      "loss": 5.0226,
      "mean_token_accuracy": 0.21347611546516418,
      "num_tokens": 67797678.0,
      "step": 30375
    },
    {
      "entropy": 5.361484098434448,
      "epoch": 3.411972819677655,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003872518842828872,
      "loss": 4.9544,
      "mean_token_accuracy": 0.22525509148836137,
      "num_tokens": 67809314.0,
      "step": 30380
    },
    {
      "entropy": 5.261776113510132,
      "epoch": 3.4125343965856123,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00038721708716166203,
      "loss": 4.9253,
      "mean_token_accuracy": 0.2201200917363167,
      "num_tokens": 67820591.0,
      "step": 30385
    },
    {
      "entropy": 5.224807834625244,
      "epoch": 3.41309597349357,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003871822864674013,
      "loss": 4.8939,
      "mean_token_accuracy": 0.22561081647872924,
      "num_tokens": 67831436.0,
      "step": 30390
    },
    {
      "entropy": 5.328497314453125,
      "epoch": 3.4136575504015276,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038714748220121305,
      "loss": 4.9544,
      "mean_token_accuracy": 0.21979170739650727,
      "num_tokens": 67842813.0,
      "step": 30395
    },
    {
      "entropy": 5.308058834075927,
      "epoch": 3.414219127309485,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003871126743642055,
      "loss": 4.9503,
      "mean_token_accuracy": 0.2150189608335495,
      "num_tokens": 67853899.0,
      "step": 30400
    },
    {
      "entropy": 5.2980118751525875,
      "epoch": 3.4147807042174425,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00038707786295748693,
      "loss": 4.9045,
      "mean_token_accuracy": 0.22042475491762162,
      "num_tokens": 67864237.0,
      "step": 30405
    },
    {
      "entropy": 5.268946743011474,
      "epoch": 3.4153422811254,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00038704304798216566,
      "loss": 4.8795,
      "mean_token_accuracy": 0.21668624132871628,
      "num_tokens": 67876109.0,
      "step": 30410
    },
    {
      "entropy": 5.208331489562989,
      "epoch": 3.415903858033358,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003870082294393503,
      "loss": 4.8284,
      "mean_token_accuracy": 0.22200530022382736,
      "num_tokens": 67888461.0,
      "step": 30415
    },
    {
      "entropy": 5.296799993515014,
      "epoch": 3.416465434941315,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003869734073301495,
      "loss": 4.8962,
      "mean_token_accuracy": 0.22683072388172149,
      "num_tokens": 67899543.0,
      "step": 30420
    },
    {
      "entropy": 5.351187705993652,
      "epoch": 3.4170270118492727,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038693858165567195,
      "loss": 4.9936,
      "mean_token_accuracy": 0.21902961134910584,
      "num_tokens": 67910606.0,
      "step": 30425
    },
    {
      "entropy": 5.262405061721802,
      "epoch": 3.4175885887572304,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003869037524170266,
      "loss": 4.8473,
      "mean_token_accuracy": 0.22479843944311143,
      "num_tokens": 67921240.0,
      "step": 30430
    },
    {
      "entropy": 5.367140865325927,
      "epoch": 3.418150165665188,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00038686891961532227,
      "loss": 5.1365,
      "mean_token_accuracy": 0.20509537309408188,
      "num_tokens": 67935178.0,
      "step": 30435
    },
    {
      "entropy": 5.189346075057983,
      "epoch": 3.4187117425731453,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003868340832516683,
      "loss": 4.7878,
      "mean_token_accuracy": 0.23048375993967057,
      "num_tokens": 67947260.0,
      "step": 30440
    },
    {
      "entropy": 5.2622452735900875,
      "epoch": 3.419273319481103,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038679924332717363,
      "loss": 4.8491,
      "mean_token_accuracy": 0.224557264149189,
      "num_tokens": 67956665.0,
      "step": 30445
    },
    {
      "entropy": 5.292057514190674,
      "epoch": 3.4198348963890606,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00038676439984294776,
      "loss": 4.9321,
      "mean_token_accuracy": 0.2161068245768547,
      "num_tokens": 67967927.0,
      "step": 30450
    },
    {
      "entropy": 5.268556451797485,
      "epoch": 3.420396473297018,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00038672955280009996,
      "loss": 4.9116,
      "mean_token_accuracy": 0.22376132011413574,
      "num_tokens": 67979574.0,
      "step": 30455
    },
    {
      "entropy": 5.275531864166259,
      "epoch": 3.4209580502049755,
      "grad_norm": 0.9375,
      "learning_rate": 0.00038669470219973996,
      "loss": 4.9654,
      "mean_token_accuracy": 0.21852470189332962,
      "num_tokens": 67991671.0,
      "step": 30460
    },
    {
      "entropy": 5.365899705886841,
      "epoch": 3.421519627112933,
      "grad_norm": 0.953125,
      "learning_rate": 0.00038665984804297734,
      "loss": 4.9653,
      "mean_token_accuracy": 0.21242029815912247,
      "num_tokens": 68002443.0,
      "step": 30465
    },
    {
      "entropy": 5.302632856369018,
      "epoch": 3.4220812040208908,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003866249903309218,
      "loss": 4.9055,
      "mean_token_accuracy": 0.21186491698026658,
      "num_tokens": 68013681.0,
      "step": 30470
    },
    {
      "entropy": 5.329411888122559,
      "epoch": 3.4226427809288484,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003865901290646833,
      "loss": 4.914,
      "mean_token_accuracy": 0.2232861876487732,
      "num_tokens": 68023757.0,
      "step": 30475
    },
    {
      "entropy": 5.225091600418091,
      "epoch": 3.4232043578368057,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038655526424537175,
      "loss": 4.8938,
      "mean_token_accuracy": 0.22274692058563234,
      "num_tokens": 68035491.0,
      "step": 30480
    },
    {
      "entropy": 5.2084376335144045,
      "epoch": 3.4237659347447633,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00038652039587409734,
      "loss": 4.9049,
      "mean_token_accuracy": 0.2236247882246971,
      "num_tokens": 68047581.0,
      "step": 30485
    },
    {
      "entropy": 5.3160315990448,
      "epoch": 3.424327511652721,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003864855239519703,
      "loss": 4.9614,
      "mean_token_accuracy": 0.21244020164012908,
      "num_tokens": 68058232.0,
      "step": 30490
    },
    {
      "entropy": 5.486220455169677,
      "epoch": 3.424889088560678,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00038645064848010093,
      "loss": 5.1158,
      "mean_token_accuracy": 0.21011122912168503,
      "num_tokens": 68071602.0,
      "step": 30495
    },
    {
      "entropy": 5.337065124511719,
      "epoch": 3.425450665468636,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038641576945959964,
      "loss": 4.9188,
      "mean_token_accuracy": 0.22017821669578552,
      "num_tokens": 68083169.0,
      "step": 30500
    },
    {
      "entropy": 5.2887317657470705,
      "epoch": 3.4260122423765935,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00038638088689157713,
      "loss": 4.9493,
      "mean_token_accuracy": 0.21112601459026337,
      "num_tokens": 68093951.0,
      "step": 30505
    },
    {
      "entropy": 5.36361665725708,
      "epoch": 3.426573819284551,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00038634600077714395,
      "loss": 5.0263,
      "mean_token_accuracy": 0.2145601287484169,
      "num_tokens": 68105398.0,
      "step": 30510
    },
    {
      "entropy": 5.346930599212646,
      "epoch": 3.4271353961925084,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003863111111174109,
      "loss": 4.9057,
      "mean_token_accuracy": 0.2211119920015335,
      "num_tokens": 68116544.0,
      "step": 30515
    },
    {
      "entropy": 5.2952601432800295,
      "epoch": 3.427696973100466,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038627621791348886,
      "loss": 4.8526,
      "mean_token_accuracy": 0.23309947699308395,
      "num_tokens": 68126977.0,
      "step": 30520
    },
    {
      "entropy": 5.332938814163208,
      "epoch": 3.4282585500084237,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038624132116648897,
      "loss": 5.0257,
      "mean_token_accuracy": 0.210595965385437,
      "num_tokens": 68137884.0,
      "step": 30525
    },
    {
      "entropy": 5.378839588165283,
      "epoch": 3.4288201269163814,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003862064208775223,
      "loss": 5.0131,
      "mean_token_accuracy": 0.21725127547979356,
      "num_tokens": 68149026.0,
      "step": 30530
    },
    {
      "entropy": 5.290618181228638,
      "epoch": 3.4293817038243386,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00038617151704769994,
      "loss": 4.9396,
      "mean_token_accuracy": 0.2206128090620041,
      "num_tokens": 68159962.0,
      "step": 30535
    },
    {
      "entropy": 5.297018957138062,
      "epoch": 3.4299432807322963,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003861366096781334,
      "loss": 4.9416,
      "mean_token_accuracy": 0.22466456443071364,
      "num_tokens": 68170849.0,
      "step": 30540
    },
    {
      "entropy": 5.2626877784729,
      "epoch": 3.430504857640254,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003861016987699342,
      "loss": 4.8231,
      "mean_token_accuracy": 0.22341955751180648,
      "num_tokens": 68181623.0,
      "step": 30545
    },
    {
      "entropy": 5.23856029510498,
      "epoch": 3.431066434548211,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038606678432421366,
      "loss": 4.7991,
      "mean_token_accuracy": 0.23252814412117004,
      "num_tokens": 68191943.0,
      "step": 30550
    },
    {
      "entropy": 5.231379222869873,
      "epoch": 3.431628011456169,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003860318663420837,
      "loss": 4.8608,
      "mean_token_accuracy": 0.22036972343921662,
      "num_tokens": 68203833.0,
      "step": 30555
    },
    {
      "entropy": 5.241099405288696,
      "epoch": 3.4321895883641265,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038599694482465606,
      "loss": 4.9125,
      "mean_token_accuracy": 0.21320911645889282,
      "num_tokens": 68213992.0,
      "step": 30560
    },
    {
      "entropy": 5.261928510665894,
      "epoch": 3.432751165272084,
      "grad_norm": 0.953125,
      "learning_rate": 0.00038596201977304263,
      "loss": 4.9023,
      "mean_token_accuracy": 0.21991546005010604,
      "num_tokens": 68224575.0,
      "step": 30565
    },
    {
      "entropy": 5.276725387573242,
      "epoch": 3.433312742180042,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038592709118835547,
      "loss": 4.9334,
      "mean_token_accuracy": 0.21946476846933366,
      "num_tokens": 68235513.0,
      "step": 30570
    },
    {
      "entropy": 5.372237682342529,
      "epoch": 3.433874319087999,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038589215907170664,
      "loss": 4.9438,
      "mean_token_accuracy": 0.2226831704378128,
      "num_tokens": 68246560.0,
      "step": 30575
    },
    {
      "entropy": 5.19332971572876,
      "epoch": 3.4344358959959567,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003858572234242086,
      "loss": 4.8084,
      "mean_token_accuracy": 0.22678698003292083,
      "num_tokens": 68257760.0,
      "step": 30580
    },
    {
      "entropy": 5.186057996749878,
      "epoch": 3.4349974729039143,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038582228424697347,
      "loss": 4.8005,
      "mean_token_accuracy": 0.23841001987457275,
      "num_tokens": 68269631.0,
      "step": 30585
    },
    {
      "entropy": 5.33116512298584,
      "epoch": 3.4355590498118715,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003857873415411138,
      "loss": 4.9399,
      "mean_token_accuracy": 0.2226322665810585,
      "num_tokens": 68280198.0,
      "step": 30590
    },
    {
      "entropy": 5.324833059310913,
      "epoch": 3.436120626719829,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003857523953077423,
      "loss": 4.9758,
      "mean_token_accuracy": 0.2183331862092018,
      "num_tokens": 68291812.0,
      "step": 30595
    },
    {
      "entropy": 5.282559013366699,
      "epoch": 3.436682203627787,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038571744554797156,
      "loss": 4.9753,
      "mean_token_accuracy": 0.21661412119865417,
      "num_tokens": 68303202.0,
      "step": 30600
    },
    {
      "entropy": 5.227714109420776,
      "epoch": 3.4372437805357445,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038568249226291433,
      "loss": 4.834,
      "mean_token_accuracy": 0.22480027377605438,
      "num_tokens": 68313768.0,
      "step": 30605
    },
    {
      "entropy": 5.30202465057373,
      "epoch": 3.4378053574437017,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038564753545368375,
      "loss": 4.9002,
      "mean_token_accuracy": 0.22558102756738663,
      "num_tokens": 68325325.0,
      "step": 30610
    },
    {
      "entropy": 5.340759181976319,
      "epoch": 3.4383669343516594,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00038561257512139264,
      "loss": 5.0026,
      "mean_token_accuracy": 0.21709089279174804,
      "num_tokens": 68337166.0,
      "step": 30615
    },
    {
      "entropy": 5.35964617729187,
      "epoch": 3.438928511259617,
      "grad_norm": 1.0,
      "learning_rate": 0.00038557761126715426,
      "loss": 4.9105,
      "mean_token_accuracy": 0.22478995025157927,
      "num_tokens": 68348898.0,
      "step": 30620
    },
    {
      "entropy": 5.270451498031616,
      "epoch": 3.4394900881675747,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003855426438920818,
      "loss": 4.9641,
      "mean_token_accuracy": 0.21887124329805374,
      "num_tokens": 68360589.0,
      "step": 30625
    },
    {
      "entropy": 5.160590696334839,
      "epoch": 3.440051665075532,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038550767299728875,
      "loss": 4.799,
      "mean_token_accuracy": 0.23082064241170883,
      "num_tokens": 68371813.0,
      "step": 30630
    },
    {
      "entropy": 5.269250965118408,
      "epoch": 3.4406132419834896,
      "grad_norm": 1.0,
      "learning_rate": 0.00038547269858388853,
      "loss": 4.9951,
      "mean_token_accuracy": 0.21880663484334945,
      "num_tokens": 68382060.0,
      "step": 30635
    },
    {
      "entropy": 5.347521686553955,
      "epoch": 3.4411748188914473,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003854377206529947,
      "loss": 4.9851,
      "mean_token_accuracy": 0.22288960069417954,
      "num_tokens": 68395603.0,
      "step": 30640
    },
    {
      "entropy": 5.2537102699279785,
      "epoch": 3.4417363957994045,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000385402739205721,
      "loss": 4.8772,
      "mean_token_accuracy": 0.22527910470962526,
      "num_tokens": 68406725.0,
      "step": 30645
    },
    {
      "entropy": 5.311653709411621,
      "epoch": 3.442297972707362,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003853677542431813,
      "loss": 5.0223,
      "mean_token_accuracy": 0.21164242327213287,
      "num_tokens": 68418577.0,
      "step": 30650
    },
    {
      "entropy": 5.286208534240723,
      "epoch": 3.44285954961532,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003853327657664895,
      "loss": 4.882,
      "mean_token_accuracy": 0.21766053587198259,
      "num_tokens": 68429708.0,
      "step": 30655
    },
    {
      "entropy": 5.323845243453979,
      "epoch": 3.4434211265232775,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003852977737767596,
      "loss": 5.0197,
      "mean_token_accuracy": 0.21652918457984924,
      "num_tokens": 68442115.0,
      "step": 30660
    },
    {
      "entropy": 5.2579230785369875,
      "epoch": 3.443982703431235,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003852627782751059,
      "loss": 4.9042,
      "mean_token_accuracy": 0.22166054248809813,
      "num_tokens": 68453555.0,
      "step": 30665
    },
    {
      "entropy": 5.222078323364258,
      "epoch": 3.4445442803391924,
      "grad_norm": 1.0,
      "learning_rate": 0.0003852277792626425,
      "loss": 4.8159,
      "mean_token_accuracy": 0.226267671585083,
      "num_tokens": 68466859.0,
      "step": 30670
    },
    {
      "entropy": 5.28090124130249,
      "epoch": 3.44510585724715,
      "grad_norm": 1.0,
      "learning_rate": 0.0003851927767404839,
      "loss": 4.9079,
      "mean_token_accuracy": 0.22006739675998688,
      "num_tokens": 68477628.0,
      "step": 30675
    },
    {
      "entropy": 5.3234395503997805,
      "epoch": 3.4456674341551077,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003851577707097445,
      "loss": 4.9959,
      "mean_token_accuracy": 0.22864916324615478,
      "num_tokens": 68488764.0,
      "step": 30680
    },
    {
      "entropy": 5.379192352294922,
      "epoch": 3.446229011063065,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000385122761171539,
      "loss": 5.0311,
      "mean_token_accuracy": 0.21578186601400376,
      "num_tokens": 68499433.0,
      "step": 30685
    },
    {
      "entropy": 5.322414445877075,
      "epoch": 3.4467905879710226,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038508774812698207,
      "loss": 4.92,
      "mean_token_accuracy": 0.21599152386188508,
      "num_tokens": 68510832.0,
      "step": 30690
    },
    {
      "entropy": 5.19415168762207,
      "epoch": 3.44735216487898,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003850527315771885,
      "loss": 4.805,
      "mean_token_accuracy": 0.2194118544459343,
      "num_tokens": 68520946.0,
      "step": 30695
    },
    {
      "entropy": 5.286277341842651,
      "epoch": 3.447913741786938,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038501771152327333,
      "loss": 4.9886,
      "mean_token_accuracy": 0.2145747259259224,
      "num_tokens": 68531769.0,
      "step": 30700
    },
    {
      "entropy": 5.392726993560791,
      "epoch": 3.448475318694895,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038498268796635155,
      "loss": 4.9974,
      "mean_token_accuracy": 0.2124207228422165,
      "num_tokens": 68543172.0,
      "step": 30705
    },
    {
      "entropy": 5.254800319671631,
      "epoch": 3.4490368956028528,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00038494766090753824,
      "loss": 4.8531,
      "mean_token_accuracy": 0.22283474057912828,
      "num_tokens": 68554078.0,
      "step": 30710
    },
    {
      "entropy": 5.234576749801636,
      "epoch": 3.4495984725108104,
      "grad_norm": 1.0,
      "learning_rate": 0.00038491263034794886,
      "loss": 4.8342,
      "mean_token_accuracy": 0.2267376199364662,
      "num_tokens": 68564705.0,
      "step": 30715
    },
    {
      "entropy": 5.366774415969848,
      "epoch": 3.450160049418768,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003848775962886987,
      "loss": 5.127,
      "mean_token_accuracy": 0.2058712884783745,
      "num_tokens": 68577010.0,
      "step": 30720
    },
    {
      "entropy": 5.314423036575318,
      "epoch": 3.4507216263267253,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003848425587309032,
      "loss": 4.9464,
      "mean_token_accuracy": 0.22027731835842132,
      "num_tokens": 68588465.0,
      "step": 30725
    },
    {
      "entropy": 5.302283096313476,
      "epoch": 3.451283203234683,
      "grad_norm": 1.0,
      "learning_rate": 0.00038480751767567807,
      "loss": 4.9535,
      "mean_token_accuracy": 0.21887488067150115,
      "num_tokens": 68599666.0,
      "step": 30730
    },
    {
      "entropy": 5.237896680831909,
      "epoch": 3.4518447801426406,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000384772473124139,
      "loss": 4.9007,
      "mean_token_accuracy": 0.2274625599384308,
      "num_tokens": 68610530.0,
      "step": 30735
    },
    {
      "entropy": 5.171618509292602,
      "epoch": 3.452406357050598,
      "grad_norm": 1.0,
      "learning_rate": 0.0003847374250774018,
      "loss": 4.8713,
      "mean_token_accuracy": 0.22594252228736877,
      "num_tokens": 68621360.0,
      "step": 30740
    },
    {
      "entropy": 5.359643507003784,
      "epoch": 3.4529679339585555,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038470237353658237,
      "loss": 4.9549,
      "mean_token_accuracy": 0.2170094594359398,
      "num_tokens": 68632874.0,
      "step": 30745
    },
    {
      "entropy": 5.3147478103637695,
      "epoch": 3.453529510866513,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00038466731850279683,
      "loss": 4.929,
      "mean_token_accuracy": 0.21780410408973694,
      "num_tokens": 68643715.0,
      "step": 30750
    },
    {
      "entropy": 5.243458795547485,
      "epoch": 3.454091087774471,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00038463225997716137,
      "loss": 4.8998,
      "mean_token_accuracy": 0.21898628622293473,
      "num_tokens": 68655433.0,
      "step": 30755
    },
    {
      "entropy": 5.301352214813233,
      "epoch": 3.4546526646824285,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038459719796079227,
      "loss": 4.9565,
      "mean_token_accuracy": 0.2262657478451729,
      "num_tokens": 68667321.0,
      "step": 30760
    },
    {
      "entropy": 5.331048488616943,
      "epoch": 3.4552142415903857,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003845621324548059,
      "loss": 4.8733,
      "mean_token_accuracy": 0.2199128195643425,
      "num_tokens": 68677314.0,
      "step": 30765
    },
    {
      "entropy": 5.273228311538697,
      "epoch": 3.4557758184983434,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003845270634603186,
      "loss": 5.0074,
      "mean_token_accuracy": 0.2166917234659195,
      "num_tokens": 68688871.0,
      "step": 30770
    },
    {
      "entropy": 5.310706853866577,
      "epoch": 3.456337395406301,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00038449199097844724,
      "loss": 4.9964,
      "mean_token_accuracy": 0.22654688060283662,
      "num_tokens": 68699760.0,
      "step": 30775
    },
    {
      "entropy": 5.423220777511597,
      "epoch": 3.4568989723142582,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003844569150103084,
      "loss": 5.059,
      "mean_token_accuracy": 0.2077832415699959,
      "num_tokens": 68711113.0,
      "step": 30780
    },
    {
      "entropy": 5.352151393890381,
      "epoch": 3.457460549222216,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003844218355570189,
      "loss": 4.9152,
      "mean_token_accuracy": 0.22018228173255922,
      "num_tokens": 68722904.0,
      "step": 30785
    },
    {
      "entropy": 5.253110027313232,
      "epoch": 3.4580221261301736,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003843867526196957,
      "loss": 4.8487,
      "mean_token_accuracy": 0.22133223861455917,
      "num_tokens": 68733753.0,
      "step": 30790
    },
    {
      "entropy": 5.249821996688842,
      "epoch": 3.4585837030381312,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003843516661994559,
      "loss": 4.9403,
      "mean_token_accuracy": 0.2151384621858597,
      "num_tokens": 68744477.0,
      "step": 30795
    },
    {
      "entropy": 5.369223165512085,
      "epoch": 3.4591452799460884,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003843165762974167,
      "loss": 4.9829,
      "mean_token_accuracy": 0.21838703155517578,
      "num_tokens": 68755607.0,
      "step": 30800
    },
    {
      "entropy": 5.387865877151489,
      "epoch": 3.459706856854046,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00038428148291469515,
      "loss": 4.9688,
      "mean_token_accuracy": 0.2169134348630905,
      "num_tokens": 68766477.0,
      "step": 30805
    },
    {
      "entropy": 5.241168069839477,
      "epoch": 3.4602684337620038,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000384246386052409,
      "loss": 4.8138,
      "mean_token_accuracy": 0.23250671178102494,
      "num_tokens": 68776806.0,
      "step": 30810
    },
    {
      "entropy": 5.212116384506226,
      "epoch": 3.4608300106699614,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003842112857116754,
      "loss": 4.905,
      "mean_token_accuracy": 0.2234220489859581,
      "num_tokens": 68788011.0,
      "step": 30815
    },
    {
      "entropy": 5.252206516265869,
      "epoch": 3.4613915875779186,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003841761818936122,
      "loss": 4.8976,
      "mean_token_accuracy": 0.2218811631202698,
      "num_tokens": 68800370.0,
      "step": 30820
    },
    {
      "entropy": 5.355507040023804,
      "epoch": 3.4619531644858763,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000384141074599337,
      "loss": 5.0195,
      "mean_token_accuracy": 0.21106944233179092,
      "num_tokens": 68811644.0,
      "step": 30825
    },
    {
      "entropy": 5.28693528175354,
      "epoch": 3.462514741393834,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00038410596382996764,
      "loss": 4.8853,
      "mean_token_accuracy": 0.20412426441907883,
      "num_tokens": 68823672.0,
      "step": 30830
    },
    {
      "entropy": 5.274688386917115,
      "epoch": 3.463076318301791,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003840708495866221,
      "loss": 4.9443,
      "mean_token_accuracy": 0.21437640488147736,
      "num_tokens": 68835340.0,
      "step": 30835
    },
    {
      "entropy": 5.280823183059693,
      "epoch": 3.463637895209749,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003840357318704183,
      "loss": 4.8281,
      "mean_token_accuracy": 0.22436444759368895,
      "num_tokens": 68845903.0,
      "step": 30840
    },
    {
      "entropy": 5.2592590808868405,
      "epoch": 3.4641994721177065,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003840006106824746,
      "loss": 4.9353,
      "mean_token_accuracy": 0.2163148432970047,
      "num_tokens": 68857287.0,
      "step": 30845
    },
    {
      "entropy": 5.308160972595215,
      "epoch": 3.464761049025664,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003839654860239091,
      "loss": 5.0117,
      "mean_token_accuracy": 0.213688226044178,
      "num_tokens": 68868961.0,
      "step": 30850
    },
    {
      "entropy": 5.336117744445801,
      "epoch": 3.465322625933622,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003839303578958404,
      "loss": 4.9194,
      "mean_token_accuracy": 0.2188081383705139,
      "num_tokens": 68880275.0,
      "step": 30855
    },
    {
      "entropy": 5.339656829833984,
      "epoch": 3.465884202841579,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00038389522629938676,
      "loss": 4.893,
      "mean_token_accuracy": 0.22946717888116835,
      "num_tokens": 68891445.0,
      "step": 30860
    },
    {
      "entropy": 5.23143424987793,
      "epoch": 3.4664457797495367,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003838600912356669,
      "loss": 4.8669,
      "mean_token_accuracy": 0.2182082176208496,
      "num_tokens": 68903591.0,
      "step": 30865
    },
    {
      "entropy": 5.265141344070434,
      "epoch": 3.4670073566574944,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003838249527057995,
      "loss": 4.9139,
      "mean_token_accuracy": 0.22093510925769805,
      "num_tokens": 68914094.0,
      "step": 30870
    },
    {
      "entropy": 5.308651924133301,
      "epoch": 3.4675689335654516,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038378981071090346,
      "loss": 4.9694,
      "mean_token_accuracy": 0.22398405522108078,
      "num_tokens": 68925929.0,
      "step": 30875
    },
    {
      "entropy": 5.248938274383545,
      "epoch": 3.4681305104734093,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038375466525209764,
      "loss": 4.8692,
      "mean_token_accuracy": 0.22075880467891693,
      "num_tokens": 68936087.0,
      "step": 30880
    },
    {
      "entropy": 5.248407602310181,
      "epoch": 3.468692087381367,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038371951633050106,
      "loss": 4.8925,
      "mean_token_accuracy": 0.22508499324321746,
      "num_tokens": 68946878.0,
      "step": 30885
    },
    {
      "entropy": 5.347327995300293,
      "epoch": 3.4692536642893246,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00038368436394723294,
      "loss": 4.9423,
      "mean_token_accuracy": 0.21144128292798997,
      "num_tokens": 68957896.0,
      "step": 30890
    },
    {
      "entropy": 5.31224422454834,
      "epoch": 3.469815241197282,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038364920810341255,
      "loss": 4.9258,
      "mean_token_accuracy": 0.2195307046175003,
      "num_tokens": 68969173.0,
      "step": 30895
    },
    {
      "entropy": 5.343195486068725,
      "epoch": 3.4703768181052395,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003836140488001592,
      "loss": 5.0035,
      "mean_token_accuracy": 0.22263323068618773,
      "num_tokens": 68980537.0,
      "step": 30900
    },
    {
      "entropy": 5.307783508300782,
      "epoch": 3.470938395013197,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038357888603859236,
      "loss": 4.904,
      "mean_token_accuracy": 0.21905263364315034,
      "num_tokens": 68991736.0,
      "step": 30905
    },
    {
      "entropy": 5.253303813934326,
      "epoch": 3.471499971921155,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038354371981983175,
      "loss": 4.8818,
      "mean_token_accuracy": 0.22081284821033478,
      "num_tokens": 69002316.0,
      "step": 30910
    },
    {
      "entropy": 5.208464097976685,
      "epoch": 3.472061548829112,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003835085501449969,
      "loss": 4.9036,
      "mean_token_accuracy": 0.22031215727329254,
      "num_tokens": 69014204.0,
      "step": 30915
    },
    {
      "entropy": 5.284231042861938,
      "epoch": 3.4726231257370697,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00038347337701520785,
      "loss": 4.9367,
      "mean_token_accuracy": 0.2165558770298958,
      "num_tokens": 69026482.0,
      "step": 30920
    },
    {
      "entropy": 5.352337169647217,
      "epoch": 3.4731847026450273,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038343820043158436,
      "loss": 4.9129,
      "mean_token_accuracy": 0.2203999161720276,
      "num_tokens": 69036527.0,
      "step": 30925
    },
    {
      "entropy": 5.33596658706665,
      "epoch": 3.4737462795529845,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003834030203952465,
      "loss": 4.9437,
      "mean_token_accuracy": 0.2222582519054413,
      "num_tokens": 69047684.0,
      "step": 30930
    },
    {
      "entropy": 5.273388862609863,
      "epoch": 3.474307856460942,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003833678369073144,
      "loss": 4.9377,
      "mean_token_accuracy": 0.21771463602781296,
      "num_tokens": 69059236.0,
      "step": 30935
    },
    {
      "entropy": 5.206654214859009,
      "epoch": 3.4748694333689,
      "grad_norm": 1.0,
      "learning_rate": 0.0003833326499689084,
      "loss": 4.87,
      "mean_token_accuracy": 0.2218744158744812,
      "num_tokens": 69069713.0,
      "step": 30940
    },
    {
      "entropy": 5.30845160484314,
      "epoch": 3.4754310102768575,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003832974595811488,
      "loss": 4.9649,
      "mean_token_accuracy": 0.21389081776142121,
      "num_tokens": 69079046.0,
      "step": 30945
    },
    {
      "entropy": 5.39464750289917,
      "epoch": 3.475992587184815,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000383262265745156,
      "loss": 4.9657,
      "mean_token_accuracy": 0.21902954578399658,
      "num_tokens": 69090080.0,
      "step": 30950
    },
    {
      "entropy": 5.358093309402466,
      "epoch": 3.4765541640927724,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038322706846205073,
      "loss": 5.0061,
      "mean_token_accuracy": 0.21362896412611007,
      "num_tokens": 69102374.0,
      "step": 30955
    },
    {
      "entropy": 5.307491779327393,
      "epoch": 3.47711574100073,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038319186773295364,
      "loss": 4.9512,
      "mean_token_accuracy": 0.22228768616914749,
      "num_tokens": 69112688.0,
      "step": 30960
    },
    {
      "entropy": 5.321045160293579,
      "epoch": 3.4776773179086877,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003831566635589856,
      "loss": 4.9341,
      "mean_token_accuracy": 0.22230856120586395,
      "num_tokens": 69124861.0,
      "step": 30965
    },
    {
      "entropy": 5.321417665481567,
      "epoch": 3.478238894816645,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003831214559412674,
      "loss": 4.9964,
      "mean_token_accuracy": 0.2187824621796608,
      "num_tokens": 69135898.0,
      "step": 30970
    },
    {
      "entropy": 5.387098026275635,
      "epoch": 3.4788004717246026,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00038308624488092005,
      "loss": 5.0111,
      "mean_token_accuracy": 0.21435560286045074,
      "num_tokens": 69146692.0,
      "step": 30975
    },
    {
      "entropy": 5.2780296325683596,
      "epoch": 3.4793620486325603,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003830510303790647,
      "loss": 4.9162,
      "mean_token_accuracy": 0.22017544656991958,
      "num_tokens": 69157908.0,
      "step": 30980
    },
    {
      "entropy": 5.292721080780029,
      "epoch": 3.479923625540518,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00038301581243682263,
      "loss": 4.9327,
      "mean_token_accuracy": 0.22175292521715165,
      "num_tokens": 69168596.0,
      "step": 30985
    },
    {
      "entropy": 5.234196472167969,
      "epoch": 3.480485202448475,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003829805910553153,
      "loss": 4.8601,
      "mean_token_accuracy": 0.2250855579972267,
      "num_tokens": 69179261.0,
      "step": 30990
    },
    {
      "entropy": 5.235365486145019,
      "epoch": 3.481046779356433,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038294536623566407,
      "loss": 4.8687,
      "mean_token_accuracy": 0.21775831729173661,
      "num_tokens": 69190410.0,
      "step": 30995
    },
    {
      "entropy": 5.344401884078979,
      "epoch": 3.4816083562643905,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003829101379789905,
      "loss": 5.023,
      "mean_token_accuracy": 0.21843025982379913,
      "num_tokens": 69201114.0,
      "step": 31000
    },
    {
      "entropy": 5.319644165039063,
      "epoch": 3.482169933172348,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003828749062864162,
      "loss": 4.8667,
      "mean_token_accuracy": 0.22662801295518875,
      "num_tokens": 69212240.0,
      "step": 31005
    },
    {
      "entropy": 5.279863786697388,
      "epoch": 3.4827315100803053,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003828396711590631,
      "loss": 4.9728,
      "mean_token_accuracy": 0.2170567125082016,
      "num_tokens": 69223714.0,
      "step": 31010
    },
    {
      "entropy": 5.294331502914429,
      "epoch": 3.483293086988263,
      "grad_norm": 1.0,
      "learning_rate": 0.00038280443259805307,
      "loss": 4.9279,
      "mean_token_accuracy": 0.2203684538602829,
      "num_tokens": 69234364.0,
      "step": 31015
    },
    {
      "entropy": 5.256495714187622,
      "epoch": 3.4838546638962207,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003827691906045081,
      "loss": 4.8467,
      "mean_token_accuracy": 0.22483622580766677,
      "num_tokens": 69245584.0,
      "step": 31020
    },
    {
      "entropy": 5.296344995498657,
      "epoch": 3.484416240804178,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003827339451795503,
      "loss": 4.8652,
      "mean_token_accuracy": 0.21217818409204484,
      "num_tokens": 69256891.0,
      "step": 31025
    },
    {
      "entropy": 5.345409965515136,
      "epoch": 3.4849778177121356,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003826986963243018,
      "loss": 4.8698,
      "mean_token_accuracy": 0.22392058968544007,
      "num_tokens": 69266972.0,
      "step": 31030
    },
    {
      "entropy": 5.286230659484863,
      "epoch": 3.485539394620093,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003826634440398852,
      "loss": 4.9317,
      "mean_token_accuracy": 0.22680806815624238,
      "num_tokens": 69277993.0,
      "step": 31035
    },
    {
      "entropy": 5.217210817337036,
      "epoch": 3.486100971528051,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00038262818832742276,
      "loss": 4.8833,
      "mean_token_accuracy": 0.22232405245304107,
      "num_tokens": 69289720.0,
      "step": 31040
    },
    {
      "entropy": 5.345918893814087,
      "epoch": 3.4866625484360085,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00038259292918803696,
      "loss": 5.0142,
      "mean_token_accuracy": 0.21132430285215378,
      "num_tokens": 69301169.0,
      "step": 31045
    },
    {
      "entropy": 5.28996729850769,
      "epoch": 3.4872241253439658,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003825576666228506,
      "loss": 4.9175,
      "mean_token_accuracy": 0.21967144906520844,
      "num_tokens": 69311443.0,
      "step": 31050
    },
    {
      "entropy": 5.298004055023194,
      "epoch": 3.4877857022519234,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003825224006329864,
      "loss": 4.9349,
      "mean_token_accuracy": 0.2201033815741539,
      "num_tokens": 69322035.0,
      "step": 31055
    },
    {
      "entropy": 5.235646104812622,
      "epoch": 3.488347279159881,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003824871312195673,
      "loss": 4.7813,
      "mean_token_accuracy": 0.2348018079996109,
      "num_tokens": 69331955.0,
      "step": 31060
    },
    {
      "entropy": 5.272967052459717,
      "epoch": 3.4889088560678383,
      "grad_norm": 1.0,
      "learning_rate": 0.0003824518583837163,
      "loss": 4.9411,
      "mean_token_accuracy": 0.21937306523323058,
      "num_tokens": 69343897.0,
      "step": 31065
    },
    {
      "entropy": 5.216382026672363,
      "epoch": 3.489470432975796,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003824165821265564,
      "loss": 4.7884,
      "mean_token_accuracy": 0.23358898013830184,
      "num_tokens": 69353886.0,
      "step": 31070
    },
    {
      "entropy": 5.171055936813355,
      "epoch": 3.4900320098837536,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003823813024492108,
      "loss": 4.8661,
      "mean_token_accuracy": 0.2261333644390106,
      "num_tokens": 69365328.0,
      "step": 31075
    },
    {
      "entropy": 5.286994409561157,
      "epoch": 3.4905935867917113,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000382346019352803,
      "loss": 4.9975,
      "mean_token_accuracy": 0.21392251551151276,
      "num_tokens": 69376007.0,
      "step": 31080
    },
    {
      "entropy": 5.359319400787354,
      "epoch": 3.4911551636996685,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003823107328384562,
      "loss": 4.9956,
      "mean_token_accuracy": 0.21648463755846023,
      "num_tokens": 69386668.0,
      "step": 31085
    },
    {
      "entropy": 5.29227089881897,
      "epoch": 3.491716740607626,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003822754429072942,
      "loss": 4.7558,
      "mean_token_accuracy": 0.23871874660253525,
      "num_tokens": 69397827.0,
      "step": 31090
    },
    {
      "entropy": 5.275832796096802,
      "epoch": 3.492278317515584,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003822401495604403,
      "loss": 4.9291,
      "mean_token_accuracy": 0.22414256036281585,
      "num_tokens": 69409268.0,
      "step": 31095
    },
    {
      "entropy": 5.268617725372314,
      "epoch": 3.4928398944235415,
      "grad_norm": 0.921875,
      "learning_rate": 0.00038220485279901855,
      "loss": 4.8921,
      "mean_token_accuracy": 0.22611673325300216,
      "num_tokens": 69420995.0,
      "step": 31100
    },
    {
      "entropy": 5.290023517608643,
      "epoch": 3.4934014713314987,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00038216955262415273,
      "loss": 4.8964,
      "mean_token_accuracy": 0.22753616124391557,
      "num_tokens": 69432919.0,
      "step": 31105
    },
    {
      "entropy": 5.285727310180664,
      "epoch": 3.4939630482394564,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038213424903696673,
      "loss": 4.9434,
      "mean_token_accuracy": 0.2201964095234871,
      "num_tokens": 69443321.0,
      "step": 31110
    },
    {
      "entropy": 5.282668352127075,
      "epoch": 3.494524625147414,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00038209894203858477,
      "loss": 4.8927,
      "mean_token_accuracy": 0.2262163296341896,
      "num_tokens": 69454695.0,
      "step": 31115
    },
    {
      "entropy": 5.297403573989868,
      "epoch": 3.4950862020553712,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00038206363163013093,
      "loss": 4.8425,
      "mean_token_accuracy": 0.22614737451076508,
      "num_tokens": 69465617.0,
      "step": 31120
    },
    {
      "entropy": 5.18984899520874,
      "epoch": 3.495647778963329,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038202831781272944,
      "loss": 4.9023,
      "mean_token_accuracy": 0.22026312947273255,
      "num_tokens": 69476730.0,
      "step": 31125
    },
    {
      "entropy": 5.221196937561035,
      "epoch": 3.4962093558712866,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00038199300058750494,
      "loss": 4.8829,
      "mean_token_accuracy": 0.22067078202962875,
      "num_tokens": 69488488.0,
      "step": 31130
    },
    {
      "entropy": 5.39414119720459,
      "epoch": 3.4967709327792442,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038195767995558174,
      "loss": 4.9682,
      "mean_token_accuracy": 0.21531396061182023,
      "num_tokens": 69498718.0,
      "step": 31135
    },
    {
      "entropy": 5.301620626449585,
      "epoch": 3.497332509687202,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003819223559180845,
      "loss": 4.8824,
      "mean_token_accuracy": 0.2209479421377182,
      "num_tokens": 69509567.0,
      "step": 31140
    },
    {
      "entropy": 5.197951364517212,
      "epoch": 3.497894086595159,
      "grad_norm": 1.03125,
      "learning_rate": 0.000381887028476138,
      "loss": 4.9036,
      "mean_token_accuracy": 0.2214992582798004,
      "num_tokens": 69519392.0,
      "step": 31145
    },
    {
      "entropy": 5.387411117553711,
      "epoch": 3.4984556635031168,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00038185169763086703,
      "loss": 5.0237,
      "mean_token_accuracy": 0.20701181441545485,
      "num_tokens": 69531228.0,
      "step": 31150
    },
    {
      "entropy": 5.398252630233765,
      "epoch": 3.4990172404110744,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0003818163633833966,
      "loss": 4.9313,
      "mean_token_accuracy": 0.22571035027503966,
      "num_tokens": 69542888.0,
      "step": 31155
    },
    {
      "entropy": 5.214018821716309,
      "epoch": 3.4995788173190316,
      "grad_norm": 1.0,
      "learning_rate": 0.00038178102573485173,
      "loss": 4.8772,
      "mean_token_accuracy": 0.2243664503097534,
      "num_tokens": 69553481.0,
      "step": 31160
    },
    {
      "entropy": 5.251704216003418,
      "epoch": 3.5001403942269893,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003817456846863575,
      "loss": 4.9141,
      "mean_token_accuracy": 0.21955126374959946,
      "num_tokens": 69565314.0,
      "step": 31165
    },
    {
      "entropy": 5.276648092269897,
      "epoch": 3.500701971134947,
      "grad_norm": 0.953125,
      "learning_rate": 0.00038171034023903937,
      "loss": 4.881,
      "mean_token_accuracy": 0.21646792888641359,
      "num_tokens": 69575807.0,
      "step": 31170
    },
    {
      "entropy": 5.332695150375367,
      "epoch": 3.5012635480429046,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003816749923940226,
      "loss": 5.0109,
      "mean_token_accuracy": 0.21894629150629044,
      "num_tokens": 69587873.0,
      "step": 31175
    },
    {
      "entropy": 5.346682405471801,
      "epoch": 3.5018251249508623,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00038163964115243276,
      "loss": 4.9568,
      "mean_token_accuracy": 0.21911311149597168,
      "num_tokens": 69598015.0,
      "step": 31180
    },
    {
      "entropy": 5.3106166362762455,
      "epoch": 3.5023867018588195,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038160428651539535,
      "loss": 4.9042,
      "mean_token_accuracy": 0.22066877484321595,
      "num_tokens": 69609345.0,
      "step": 31185
    },
    {
      "entropy": 5.255890846252441,
      "epoch": 3.502948278766777,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00038156892848403606,
      "loss": 4.9263,
      "mean_token_accuracy": 0.20871613174676895,
      "num_tokens": 69620075.0,
      "step": 31190
    },
    {
      "entropy": 5.2988138675689695,
      "epoch": 3.503509855674735,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003815335670594808,
      "loss": 4.9217,
      "mean_token_accuracy": 0.22501251697540284,
      "num_tokens": 69630825.0,
      "step": 31195
    },
    {
      "entropy": 5.389105415344238,
      "epoch": 3.504071432582692,
      "grad_norm": 0.921875,
      "learning_rate": 0.00038149820224285546,
      "loss": 5.0131,
      "mean_token_accuracy": 0.22114652693271636,
      "num_tokens": 69642415.0,
      "step": 31200
    },
    {
      "entropy": 5.1864502906799315,
      "epoch": 3.5046330094906497,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000381462834035286,
      "loss": 4.8095,
      "mean_token_accuracy": 0.2289103627204895,
      "num_tokens": 69652735.0,
      "step": 31205
    },
    {
      "entropy": 5.231696891784668,
      "epoch": 3.5051945863986074,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003814274624378986,
      "loss": 4.91,
      "mean_token_accuracy": 0.2264372304081917,
      "num_tokens": 69664312.0,
      "step": 31210
    },
    {
      "entropy": 5.282000255584717,
      "epoch": 3.5057561633065646,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003813920874518196,
      "loss": 4.8749,
      "mean_token_accuracy": 0.21890823543071747,
      "num_tokens": 69674570.0,
      "step": 31215
    },
    {
      "entropy": 5.257593107223511,
      "epoch": 3.5063177402145222,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003813567090781752,
      "loss": 4.9253,
      "mean_token_accuracy": 0.21759086847305298,
      "num_tokens": 69686591.0,
      "step": 31220
    },
    {
      "entropy": 5.325746345520019,
      "epoch": 3.50687931712248,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000381321327318092,
      "loss": 4.9449,
      "mean_token_accuracy": 0.22676262110471726,
      "num_tokens": 69697104.0,
      "step": 31225
    },
    {
      "entropy": 5.335159111022949,
      "epoch": 3.5074408940304376,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003812859421726964,
      "loss": 4.8862,
      "mean_token_accuracy": 0.22630295306444168,
      "num_tokens": 69708827.0,
      "step": 31230
    },
    {
      "entropy": 5.263040542602539,
      "epoch": 3.5080024709383952,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00038125055364311526,
      "loss": 4.9345,
      "mean_token_accuracy": 0.22347561717033387,
      "num_tokens": 69720426.0,
      "step": 31235
    },
    {
      "entropy": 5.391822671890258,
      "epoch": 3.5085640478463525,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003812151617304753,
      "loss": 5.1217,
      "mean_token_accuracy": 0.2104395255446434,
      "num_tokens": 69733199.0,
      "step": 31240
    },
    {
      "entropy": 5.2312061309814455,
      "epoch": 3.50912562475431,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003811797664359033,
      "loss": 4.7681,
      "mean_token_accuracy": 0.22622035443782806,
      "num_tokens": 69744322.0,
      "step": 31245
    },
    {
      "entropy": 5.21504225730896,
      "epoch": 3.5096872016622678,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00038114436776052644,
      "loss": 4.8612,
      "mean_token_accuracy": 0.21841011047363282,
      "num_tokens": 69755662.0,
      "step": 31250
    },
    {
      "entropy": 5.1695033550262455,
      "epoch": 3.510248778570225,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003811089657054717,
      "loss": 4.7743,
      "mean_token_accuracy": 0.2294035479426384,
      "num_tokens": 69765663.0,
      "step": 31255
    },
    {
      "entropy": 5.303169536590576,
      "epoch": 3.5108103554781827,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003810735602718664,
      "loss": 5.0717,
      "mean_token_accuracy": 0.21663507521152497,
      "num_tokens": 69777483.0,
      "step": 31260
    },
    {
      "entropy": 5.397022676467896,
      "epoch": 3.5113719323861403,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003810381514608378,
      "loss": 4.9504,
      "mean_token_accuracy": 0.21627181619405747,
      "num_tokens": 69789034.0,
      "step": 31265
    },
    {
      "entropy": 5.350554466247559,
      "epoch": 3.511933509294098,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003810027392735133,
      "loss": 4.9204,
      "mean_token_accuracy": 0.2202489361166954,
      "num_tokens": 69801288.0,
      "step": 31270
    },
    {
      "entropy": 5.236014318466187,
      "epoch": 3.5124950862020556,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003809673237110204,
      "loss": 4.8677,
      "mean_token_accuracy": 0.2185143619775772,
      "num_tokens": 69811063.0,
      "step": 31275
    },
    {
      "entropy": 5.237791204452515,
      "epoch": 3.513056663110013,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003809319047744869,
      "loss": 4.8994,
      "mean_token_accuracy": 0.22358224540948868,
      "num_tokens": 69822572.0,
      "step": 31280
    },
    {
      "entropy": 5.195301103591919,
      "epoch": 3.5136182400179705,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038089648246504045,
      "loss": 4.8371,
      "mean_token_accuracy": 0.22304292470216752,
      "num_tokens": 69834082.0,
      "step": 31285
    },
    {
      "entropy": 5.258901643753052,
      "epoch": 3.514179816925928,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038086105678380896,
      "loss": 4.8959,
      "mean_token_accuracy": 0.2273969754576683,
      "num_tokens": 69844560.0,
      "step": 31290
    },
    {
      "entropy": 5.267577314376831,
      "epoch": 3.5147413938338854,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038082562773192026,
      "loss": 4.8934,
      "mean_token_accuracy": 0.2211877405643463,
      "num_tokens": 69854993.0,
      "step": 31295
    },
    {
      "entropy": 5.270772409439087,
      "epoch": 3.515302970741843,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038079019531050257,
      "loss": 4.8534,
      "mean_token_accuracy": 0.22488428503274918,
      "num_tokens": 69866074.0,
      "step": 31300
    },
    {
      "entropy": 5.2915984153747555,
      "epoch": 3.5158645476498007,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038075475952068405,
      "loss": 4.9193,
      "mean_token_accuracy": 0.21760949343442917,
      "num_tokens": 69877101.0,
      "step": 31305
    },
    {
      "entropy": 5.1930900573730465,
      "epoch": 3.516426124557758,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000380719320363593,
      "loss": 4.8253,
      "mean_token_accuracy": 0.23273673802614211,
      "num_tokens": 69887388.0,
      "step": 31310
    },
    {
      "entropy": 5.281467390060425,
      "epoch": 3.5169877014657156,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038068387784035774,
      "loss": 4.9226,
      "mean_token_accuracy": 0.2221325680613518,
      "num_tokens": 69898242.0,
      "step": 31315
    },
    {
      "entropy": 5.275081729888916,
      "epoch": 3.5175492783736733,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003806484319521069,
      "loss": 4.9296,
      "mean_token_accuracy": 0.22126188725233079,
      "num_tokens": 69909645.0,
      "step": 31320
    },
    {
      "entropy": 5.290223264694214,
      "epoch": 3.518110855281631,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00038061298269996893,
      "loss": 4.9051,
      "mean_token_accuracy": 0.22015096992254257,
      "num_tokens": 69920862.0,
      "step": 31325
    },
    {
      "entropy": 5.257872676849365,
      "epoch": 3.5186724321895886,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003805775300850727,
      "loss": 4.8386,
      "mean_token_accuracy": 0.22541294991970062,
      "num_tokens": 69932219.0,
      "step": 31330
    },
    {
      "entropy": 5.322806406021118,
      "epoch": 3.519234009097546,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003805420741085469,
      "loss": 4.926,
      "mean_token_accuracy": 0.219169457256794,
      "num_tokens": 69942642.0,
      "step": 31335
    },
    {
      "entropy": 5.363058996200562,
      "epoch": 3.5197955860055035,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003805066147715205,
      "loss": 5.0651,
      "mean_token_accuracy": 0.2031511753797531,
      "num_tokens": 69954336.0,
      "step": 31340
    },
    {
      "entropy": 5.289618396759034,
      "epoch": 3.520357162913461,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003804711520751226,
      "loss": 4.8856,
      "mean_token_accuracy": 0.2259404569864273,
      "num_tokens": 69965716.0,
      "step": 31345
    },
    {
      "entropy": 5.270514726638794,
      "epoch": 3.5209187398214183,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038043568602048227,
      "loss": 4.9244,
      "mean_token_accuracy": 0.218100206553936,
      "num_tokens": 69976545.0,
      "step": 31350
    },
    {
      "entropy": 5.3309797763824465,
      "epoch": 3.521480316729376,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003804002166087289,
      "loss": 4.9402,
      "mean_token_accuracy": 0.22156260758638383,
      "num_tokens": 69987846.0,
      "step": 31355
    },
    {
      "entropy": 5.2705864906311035,
      "epoch": 3.5220418936373337,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038036474384099174,
      "loss": 4.8365,
      "mean_token_accuracy": 0.22589399367570878,
      "num_tokens": 69998089.0,
      "step": 31360
    },
    {
      "entropy": 5.28804121017456,
      "epoch": 3.5226034705452913,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003803292677184002,
      "loss": 4.9918,
      "mean_token_accuracy": 0.21729234009981155,
      "num_tokens": 70008900.0,
      "step": 31365
    },
    {
      "entropy": 5.327318429946899,
      "epoch": 3.523165047453249,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00038029378824208394,
      "loss": 4.9496,
      "mean_token_accuracy": 0.22241339683532715,
      "num_tokens": 70019675.0,
      "step": 31370
    },
    {
      "entropy": 5.344908809661865,
      "epoch": 3.523726624361206,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003802583054131727,
      "loss": 5.0331,
      "mean_token_accuracy": 0.21678320318460464,
      "num_tokens": 70031077.0,
      "step": 31375
    },
    {
      "entropy": 5.374970579147339,
      "epoch": 3.524288201269164,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003802228192327961,
      "loss": 5.0314,
      "mean_token_accuracy": 0.2057809427380562,
      "num_tokens": 70044569.0,
      "step": 31380
    },
    {
      "entropy": 5.310085391998291,
      "epoch": 3.5248497781771215,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038018732970208414,
      "loss": 4.8996,
      "mean_token_accuracy": 0.2249251812696457,
      "num_tokens": 70056396.0,
      "step": 31385
    },
    {
      "entropy": 5.396480226516724,
      "epoch": 3.5254113550850787,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003801518368221668,
      "loss": 5.1103,
      "mean_token_accuracy": 0.21263699531555175,
      "num_tokens": 70068341.0,
      "step": 31390
    },
    {
      "entropy": 5.258814668655395,
      "epoch": 3.5259729319930364,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003801163405941743,
      "loss": 4.8247,
      "mean_token_accuracy": 0.2257183462381363,
      "num_tokens": 70079069.0,
      "step": 31395
    },
    {
      "entropy": 5.30569109916687,
      "epoch": 3.526534508900994,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00038008084101923665,
      "loss": 5.0004,
      "mean_token_accuracy": 0.2177607908844948,
      "num_tokens": 70090274.0,
      "step": 31400
    },
    {
      "entropy": 5.223629570007324,
      "epoch": 3.5270960858089513,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00038004533809848435,
      "loss": 4.8936,
      "mean_token_accuracy": 0.22501357644796371,
      "num_tokens": 70101144.0,
      "step": 31405
    },
    {
      "entropy": 5.2829780101776125,
      "epoch": 3.527657662716909,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038000983183304765,
      "loss": 4.9134,
      "mean_token_accuracy": 0.22200986444950105,
      "num_tokens": 70112210.0,
      "step": 31410
    },
    {
      "entropy": 5.370964670181275,
      "epoch": 3.5282192396248666,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003799743222240572,
      "loss": 5.0332,
      "mean_token_accuracy": 0.21579583138227462,
      "num_tokens": 70124302.0,
      "step": 31415
    },
    {
      "entropy": 5.337023305892944,
      "epoch": 3.5287808165328243,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003799388092726436,
      "loss": 4.999,
      "mean_token_accuracy": 0.2185861200094223,
      "num_tokens": 70136026.0,
      "step": 31420
    },
    {
      "entropy": 5.359019708633423,
      "epoch": 3.529342393440782,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00037990329297993763,
      "loss": 4.9335,
      "mean_token_accuracy": 0.21761567741632462,
      "num_tokens": 70147439.0,
      "step": 31425
    },
    {
      "entropy": 5.219349479675293,
      "epoch": 3.529903970348739,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003798677733470701,
      "loss": 4.7845,
      "mean_token_accuracy": 0.23501665145158768,
      "num_tokens": 70158367.0,
      "step": 31430
    },
    {
      "entropy": 5.224279069900513,
      "epoch": 3.530465547256697,
      "grad_norm": 0.953125,
      "learning_rate": 0.00037983225037517207,
      "loss": 5.0119,
      "mean_token_accuracy": 0.21247811764478683,
      "num_tokens": 70169226.0,
      "step": 31435
    },
    {
      "entropy": 5.308132934570312,
      "epoch": 3.5310271241646545,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037979672406537445,
      "loss": 4.8944,
      "mean_token_accuracy": 0.22221734672784804,
      "num_tokens": 70179440.0,
      "step": 31440
    },
    {
      "entropy": 5.326109075546265,
      "epoch": 3.5315887010726117,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003797611944188086,
      "loss": 4.9688,
      "mean_token_accuracy": 0.219835464656353,
      "num_tokens": 70191160.0,
      "step": 31445
    },
    {
      "entropy": 5.289790487289428,
      "epoch": 3.5321502779805694,
      "grad_norm": 0.953125,
      "learning_rate": 0.00037972566143660566,
      "loss": 4.9967,
      "mean_token_accuracy": 0.21525720208883287,
      "num_tokens": 70202143.0,
      "step": 31450
    },
    {
      "entropy": 5.309869480133057,
      "epoch": 3.532711854888527,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003796901251198969,
      "loss": 4.9015,
      "mean_token_accuracy": 0.2179402306675911,
      "num_tokens": 70213957.0,
      "step": 31455
    },
    {
      "entropy": 5.3137870788574215,
      "epoch": 3.5332734317964847,
      "grad_norm": 1.015625,
      "learning_rate": 0.000379654585469814,
      "loss": 4.9122,
      "mean_token_accuracy": 0.22079637497663498,
      "num_tokens": 70224153.0,
      "step": 31460
    },
    {
      "entropy": 5.293676710128784,
      "epoch": 3.5338350087044423,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003796190424874886,
      "loss": 4.9337,
      "mean_token_accuracy": 0.21726232022047043,
      "num_tokens": 70236093.0,
      "step": 31465
    },
    {
      "entropy": 5.260035514831543,
      "epoch": 3.5343965856123996,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003795834961740523,
      "loss": 4.9543,
      "mean_token_accuracy": 0.22103006839752198,
      "num_tokens": 70246865.0,
      "step": 31470
    },
    {
      "entropy": 5.289642190933227,
      "epoch": 3.534958162520357,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00037954794653063695,
      "loss": 4.9906,
      "mean_token_accuracy": 0.21200790256261826,
      "num_tokens": 70258601.0,
      "step": 31475
    },
    {
      "entropy": 5.313614273071289,
      "epoch": 3.535519739428315,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003795123935583743,
      "loss": 4.9094,
      "mean_token_accuracy": 0.22005391567945481,
      "num_tokens": 70269020.0,
      "step": 31480
    },
    {
      "entropy": 5.331886577606201,
      "epoch": 3.536081316336272,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003794768372583966,
      "loss": 4.9589,
      "mean_token_accuracy": 0.2102758601307869,
      "num_tokens": 70280582.0,
      "step": 31485
    },
    {
      "entropy": 5.372834205627441,
      "epoch": 3.5366428932442298,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037944127763183587,
      "loss": 4.9478,
      "mean_token_accuracy": 0.21840815246105194,
      "num_tokens": 70292848.0,
      "step": 31490
    },
    {
      "entropy": 5.3935534954071045,
      "epoch": 3.5372044701521874,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00037940571467982436,
      "loss": 4.9721,
      "mean_token_accuracy": 0.21753677129745483,
      "num_tokens": 70304892.0,
      "step": 31495
    },
    {
      "entropy": 5.24358901977539,
      "epoch": 3.5377660470601446,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003793701484034944,
      "loss": 4.8639,
      "mean_token_accuracy": 0.22965763211250306,
      "num_tokens": 70316244.0,
      "step": 31500
    },
    {
      "entropy": 5.30887770652771,
      "epoch": 3.5383276239681023,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037933457880397843,
      "loss": 4.9221,
      "mean_token_accuracy": 0.21632213592529298,
      "num_tokens": 70327037.0,
      "step": 31505
    },
    {
      "entropy": 5.268599367141723,
      "epoch": 3.53888920087606,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00037929900588240897,
      "loss": 4.8316,
      "mean_token_accuracy": 0.228417207300663,
      "num_tokens": 70338587.0,
      "step": 31510
    },
    {
      "entropy": 5.17706184387207,
      "epoch": 3.5394507777840176,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003792634296399187,
      "loss": 4.9407,
      "mean_token_accuracy": 0.22063326239585876,
      "num_tokens": 70349930.0,
      "step": 31515
    },
    {
      "entropy": 5.176056575775147,
      "epoch": 3.5400123546919753,
      "grad_norm": 1.0,
      "learning_rate": 0.00037922785007764043,
      "loss": 4.7576,
      "mean_token_accuracy": 0.22737986445426941,
      "num_tokens": 70360880.0,
      "step": 31520
    },
    {
      "entropy": 5.405176591873169,
      "epoch": 3.5405739315999325,
      "grad_norm": 0.90625,
      "learning_rate": 0.000379192267196707,
      "loss": 5.0411,
      "mean_token_accuracy": 0.21292384564876557,
      "num_tokens": 70372508.0,
      "step": 31525
    },
    {
      "entropy": 5.33902907371521,
      "epoch": 3.54113550850789,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037915668099825135,
      "loss": 4.9333,
      "mean_token_accuracy": 0.22344198375940322,
      "num_tokens": 70383960.0,
      "step": 31530
    },
    {
      "entropy": 5.274180698394775,
      "epoch": 3.541697085415848,
      "grad_norm": 0.96875,
      "learning_rate": 0.00037912109148340663,
      "loss": 4.925,
      "mean_token_accuracy": 0.22033117115497589,
      "num_tokens": 70395881.0,
      "step": 31535
    },
    {
      "entropy": 5.258727931976319,
      "epoch": 3.542258662323805,
      "grad_norm": 0.953125,
      "learning_rate": 0.00037908549865330596,
      "loss": 4.9308,
      "mean_token_accuracy": 0.2199956566095352,
      "num_tokens": 70407590.0,
      "step": 31540
    },
    {
      "entropy": 5.311266756057739,
      "epoch": 3.5428202392317627,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037904990250908256,
      "loss": 4.8992,
      "mean_token_accuracy": 0.23334798216819763,
      "num_tokens": 70418078.0,
      "step": 31545
    },
    {
      "entropy": 5.3207306385040285,
      "epoch": 3.5433818161397204,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037901430305187,
      "loss": 4.9058,
      "mean_token_accuracy": 0.22171232402324675,
      "num_tokens": 70428292.0,
      "step": 31550
    },
    {
      "entropy": 5.205560398101807,
      "epoch": 3.543943393047678,
      "grad_norm": 1.0,
      "learning_rate": 0.00037897870028280167,
      "loss": 4.8144,
      "mean_token_accuracy": 0.22544403821229936,
      "num_tokens": 70438481.0,
      "step": 31555
    },
    {
      "entropy": 5.268299627304077,
      "epoch": 3.5445049699556357,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0003789430942030112,
      "loss": 4.979,
      "mean_token_accuracy": 0.21436949372291564,
      "num_tokens": 70449949.0,
      "step": 31560
    },
    {
      "entropy": 5.392711734771728,
      "epoch": 3.545066546863593,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037890748481363216,
      "loss": 5.0169,
      "mean_token_accuracy": 0.21663267463445662,
      "num_tokens": 70461159.0,
      "step": 31565
    },
    {
      "entropy": 5.307415866851807,
      "epoch": 3.5456281237715506,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003788718721157987,
      "loss": 4.913,
      "mean_token_accuracy": 0.22063572853803634,
      "num_tokens": 70471212.0,
      "step": 31570
    },
    {
      "entropy": 5.294901418685913,
      "epoch": 3.5461897006795082,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003788362561106445,
      "loss": 4.9108,
      "mean_token_accuracy": 0.2228067249059677,
      "num_tokens": 70481587.0,
      "step": 31575
    },
    {
      "entropy": 5.396735382080078,
      "epoch": 3.5467512775874654,
      "grad_norm": 1.0,
      "learning_rate": 0.00037880063679930355,
      "loss": 4.9911,
      "mean_token_accuracy": 0.21835146695375443,
      "num_tokens": 70491952.0,
      "step": 31580
    },
    {
      "entropy": 5.30038890838623,
      "epoch": 3.547312854495423,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037876501418291015,
      "loss": 4.9022,
      "mean_token_accuracy": 0.21868725419044494,
      "num_tokens": 70503279.0,
      "step": 31585
    },
    {
      "entropy": 5.255746364593506,
      "epoch": 3.5478744314033808,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003787293882625984,
      "loss": 4.9744,
      "mean_token_accuracy": 0.21319018453359603,
      "num_tokens": 70514779.0,
      "step": 31590
    },
    {
      "entropy": 5.24497504234314,
      "epoch": 3.548436008311338,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003786937590395027,
      "loss": 4.8886,
      "mean_token_accuracy": 0.22268475443124772,
      "num_tokens": 70526478.0,
      "step": 31595
    },
    {
      "entropy": 5.324106168746948,
      "epoch": 3.5489975852192956,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003786581265147575,
      "loss": 4.9797,
      "mean_token_accuracy": 0.21215957403182983,
      "num_tokens": 70537293.0,
      "step": 31600
    },
    {
      "entropy": 5.334910202026367,
      "epoch": 3.5495591621272533,
      "grad_norm": 1.0,
      "learning_rate": 0.0003786224906894974,
      "loss": 4.8851,
      "mean_token_accuracy": 0.2243972361087799,
      "num_tokens": 70547357.0,
      "step": 31605
    },
    {
      "entropy": 5.272511577606201,
      "epoch": 3.550120739035211,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000378586851564857,
      "loss": 4.9427,
      "mean_token_accuracy": 0.21501441895961762,
      "num_tokens": 70558102.0,
      "step": 31610
    },
    {
      "entropy": 5.350009489059448,
      "epoch": 3.5506823159431686,
      "grad_norm": 1.078125,
      "learning_rate": 0.000378551209141971,
      "loss": 4.9703,
      "mean_token_accuracy": 0.22357803732156753,
      "num_tokens": 70569454.0,
      "step": 31615
    },
    {
      "entropy": 5.314974069595337,
      "epoch": 3.551243892851126,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003785155634219744,
      "loss": 4.8878,
      "mean_token_accuracy": 0.22112821489572526,
      "num_tokens": 70580590.0,
      "step": 31620
    },
    {
      "entropy": 5.255598163604736,
      "epoch": 3.5518054697590835,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003784799144060021,
      "loss": 4.9099,
      "mean_token_accuracy": 0.22415250837802886,
      "num_tokens": 70591233.0,
      "step": 31625
    },
    {
      "entropy": 5.2925270080566404,
      "epoch": 3.552367046667041,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00037844426209518916,
      "loss": 4.8911,
      "mean_token_accuracy": 0.23035763204097748,
      "num_tokens": 70603029.0,
      "step": 31630
    },
    {
      "entropy": 5.402775287628174,
      "epoch": 3.5529286235749984,
      "grad_norm": 0.9375,
      "learning_rate": 0.00037840860649067084,
      "loss": 5.0432,
      "mean_token_accuracy": 0.21393407732248307,
      "num_tokens": 70616021.0,
      "step": 31635
    },
    {
      "entropy": 5.284610652923584,
      "epoch": 3.553490200482956,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037837294759358235,
      "loss": 4.9323,
      "mean_token_accuracy": 0.2275229051709175,
      "num_tokens": 70627044.0,
      "step": 31640
    },
    {
      "entropy": 5.231058788299561,
      "epoch": 3.5540517773909137,
      "grad_norm": 1.0,
      "learning_rate": 0.00037833728540505914,
      "loss": 4.8449,
      "mean_token_accuracy": 0.22254250347614288,
      "num_tokens": 70637435.0,
      "step": 31645
    },
    {
      "entropy": 5.3077192306518555,
      "epoch": 3.5546133542988714,
      "grad_norm": 1.0,
      "learning_rate": 0.0003783016199262367,
      "loss": 4.9181,
      "mean_token_accuracy": 0.22577877789735795,
      "num_tokens": 70648106.0,
      "step": 31650
    },
    {
      "entropy": 5.233428239822388,
      "epoch": 3.555174931206829,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003782659511582505,
      "loss": 4.9248,
      "mean_token_accuracy": 0.21714639365673066,
      "num_tokens": 70661242.0,
      "step": 31655
    },
    {
      "entropy": 5.251920223236084,
      "epoch": 3.5557365081147863,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003782302791022365,
      "loss": 4.95,
      "mean_token_accuracy": 0.21748458445072175,
      "num_tokens": 70672758.0,
      "step": 31660
    },
    {
      "entropy": 5.3528767108917235,
      "epoch": 3.556298085022744,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037819460375933034,
      "loss": 4.9368,
      "mean_token_accuracy": 0.2174877181649208,
      "num_tokens": 70683365.0,
      "step": 31665
    },
    {
      "entropy": 5.234508323669433,
      "epoch": 3.5568596619307016,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003781589251306679,
      "loss": 4.8514,
      "mean_token_accuracy": 0.22670917958021164,
      "num_tokens": 70693984.0,
      "step": 31670
    },
    {
      "entropy": 5.262110042572021,
      "epoch": 3.557421238838659,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003781232432173853,
      "loss": 4.9158,
      "mean_token_accuracy": 0.2139906644821167,
      "num_tokens": 70704748.0,
      "step": 31675
    },
    {
      "entropy": 5.209932661056518,
      "epoch": 3.5579828157466165,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037808755802061874,
      "loss": 4.8666,
      "mean_token_accuracy": 0.22766690999269484,
      "num_tokens": 70715415.0,
      "step": 31680
    },
    {
      "entropy": 5.236859941482544,
      "epoch": 3.558544392654574,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003780518695415043,
      "loss": 4.9065,
      "mean_token_accuracy": 0.22406142801046372,
      "num_tokens": 70726646.0,
      "step": 31685
    },
    {
      "entropy": 5.404096603393555,
      "epoch": 3.5591059695625313,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00037801617778117835,
      "loss": 5.0939,
      "mean_token_accuracy": 0.21036745756864547,
      "num_tokens": 70737515.0,
      "step": 31690
    },
    {
      "entropy": 5.391033077239991,
      "epoch": 3.559667546470489,
      "grad_norm": 1.0,
      "learning_rate": 0.00037798048274077735,
      "loss": 4.9402,
      "mean_token_accuracy": 0.22106135338544847,
      "num_tokens": 70747751.0,
      "step": 31695
    },
    {
      "entropy": 5.325807523727417,
      "epoch": 3.5602291233784467,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037794478442143786,
      "loss": 4.8448,
      "mean_token_accuracy": 0.22885860353708268,
      "num_tokens": 70757951.0,
      "step": 31700
    },
    {
      "entropy": 5.318636465072632,
      "epoch": 3.5607907002864043,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00037790908282429647,
      "loss": 5.0297,
      "mean_token_accuracy": 0.20935260355472565,
      "num_tokens": 70769515.0,
      "step": 31705
    },
    {
      "entropy": 5.289512586593628,
      "epoch": 3.561352277194362,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037787337795049,
      "loss": 4.9521,
      "mean_token_accuracy": 0.21766523867845536,
      "num_tokens": 70780567.0,
      "step": 31710
    },
    {
      "entropy": 5.3386157035827635,
      "epoch": 3.561913854102319,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00037783766980115525,
      "loss": 5.0413,
      "mean_token_accuracy": 0.2120776429772377,
      "num_tokens": 70793250.0,
      "step": 31715
    },
    {
      "entropy": 5.394042015075684,
      "epoch": 3.562475431010277,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037780195837742927,
      "loss": 4.9667,
      "mean_token_accuracy": 0.21338024884462356,
      "num_tokens": 70804579.0,
      "step": 31720
    },
    {
      "entropy": 5.302375078201294,
      "epoch": 3.5630370079182345,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037776624368044895,
      "loss": 4.8579,
      "mean_token_accuracy": 0.2250915974378586,
      "num_tokens": 70815931.0,
      "step": 31725
    },
    {
      "entropy": 5.353155708312988,
      "epoch": 3.5635985848261917,
      "grad_norm": 1.8828125,
      "learning_rate": 0.00037773052571135157,
      "loss": 4.9908,
      "mean_token_accuracy": 0.21182172149419784,
      "num_tokens": 70828058.0,
      "step": 31730
    },
    {
      "entropy": 5.315308427810669,
      "epoch": 3.5641601617341494,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003776948044712745,
      "loss": 4.9456,
      "mean_token_accuracy": 0.21882034987211227,
      "num_tokens": 70838177.0,
      "step": 31735
    },
    {
      "entropy": 5.292368030548095,
      "epoch": 3.564721738642107,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000377659079961355,
      "loss": 4.9258,
      "mean_token_accuracy": 0.22500565350055696,
      "num_tokens": 70849888.0,
      "step": 31740
    },
    {
      "entropy": 5.285151147842408,
      "epoch": 3.5652833155500647,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00037762335218273057,
      "loss": 4.9034,
      "mean_token_accuracy": 0.22424452751874924,
      "num_tokens": 70859754.0,
      "step": 31745
    },
    {
      "entropy": 5.351862287521362,
      "epoch": 3.5658448924580224,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003775876211365388,
      "loss": 5.0057,
      "mean_token_accuracy": 0.2133370041847229,
      "num_tokens": 70870166.0,
      "step": 31750
    },
    {
      "entropy": 5.360077953338623,
      "epoch": 3.5664064693659796,
      "grad_norm": 1.0,
      "learning_rate": 0.0003775518868239174,
      "loss": 4.9177,
      "mean_token_accuracy": 0.21953082233667373,
      "num_tokens": 70882204.0,
      "step": 31755
    },
    {
      "entropy": 5.229487514495849,
      "epoch": 3.5669680462739373,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037751614924600417,
      "loss": 4.9284,
      "mean_token_accuracy": 0.2197223797440529,
      "num_tokens": 70894317.0,
      "step": 31760
    },
    {
      "entropy": 5.308276605606079,
      "epoch": 3.567529623181895,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000377480408403937,
      "loss": 4.9338,
      "mean_token_accuracy": 0.22132431119680404,
      "num_tokens": 70906224.0,
      "step": 31765
    },
    {
      "entropy": 5.37962532043457,
      "epoch": 3.568091200089852,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003774446642988538,
      "loss": 5.0396,
      "mean_token_accuracy": 0.20988627821207045,
      "num_tokens": 70918134.0,
      "step": 31770
    },
    {
      "entropy": 5.2447456359863285,
      "epoch": 3.56865277699781,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00037740891693189286,
      "loss": 4.9011,
      "mean_token_accuracy": 0.21907398104667664,
      "num_tokens": 70929695.0,
      "step": 31775
    },
    {
      "entropy": 5.360164022445678,
      "epoch": 3.5692143539057675,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003773731663041922,
      "loss": 4.9938,
      "mean_token_accuracy": 0.21376034170389174,
      "num_tokens": 70941204.0,
      "step": 31780
    },
    {
      "entropy": 5.394133615493774,
      "epoch": 3.5697759308137247,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037733741241689024,
      "loss": 5.0063,
      "mean_token_accuracy": 0.21859212964773178,
      "num_tokens": 70953221.0,
      "step": 31785
    },
    {
      "entropy": 5.32682957649231,
      "epoch": 3.5703375077216823,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003773016552711254,
      "loss": 4.9845,
      "mean_token_accuracy": 0.2180962532758713,
      "num_tokens": 70967081.0,
      "step": 31790
    },
    {
      "entropy": 5.236804723739624,
      "epoch": 3.57089908462964,
      "grad_norm": 0.953125,
      "learning_rate": 0.00037726589486803617,
      "loss": 4.8489,
      "mean_token_accuracy": 0.22227933555841445,
      "num_tokens": 70977911.0,
      "step": 31795
    },
    {
      "entropy": 5.306356906890869,
      "epoch": 3.5714606615375977,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037723013120876113,
      "loss": 4.8989,
      "mean_token_accuracy": 0.22476733326911927,
      "num_tokens": 70989001.0,
      "step": 31800
    },
    {
      "entropy": 5.320364046096802,
      "epoch": 3.5720222384455553,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037719436429443915,
      "loss": 5.0052,
      "mean_token_accuracy": 0.21253447234630585,
      "num_tokens": 70999698.0,
      "step": 31805
    },
    {
      "entropy": 5.18927412033081,
      "epoch": 3.5725838153535125,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00037715859412620895,
      "loss": 4.7877,
      "mean_token_accuracy": 0.2266393184661865,
      "num_tokens": 71010490.0,
      "step": 31810
    },
    {
      "entropy": 5.299437046051025,
      "epoch": 3.57314539226147,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003771228207052095,
      "loss": 4.9449,
      "mean_token_accuracy": 0.2263795703649521,
      "num_tokens": 71021581.0,
      "step": 31815
    },
    {
      "entropy": 5.334377431869507,
      "epoch": 3.573706969169428,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037708704403257974,
      "loss": 4.9573,
      "mean_token_accuracy": 0.21737345159053803,
      "num_tokens": 71033215.0,
      "step": 31820
    },
    {
      "entropy": 5.25474705696106,
      "epoch": 3.574268546077385,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000377051264109459,
      "loss": 4.8022,
      "mean_token_accuracy": 0.23153724670410156,
      "num_tokens": 71043484.0,
      "step": 31825
    },
    {
      "entropy": 5.254907894134521,
      "epoch": 3.5748301229853428,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003770154809369864,
      "loss": 4.8845,
      "mean_token_accuracy": 0.22478361427783966,
      "num_tokens": 71054101.0,
      "step": 31830
    },
    {
      "entropy": 5.267241191864014,
      "epoch": 3.5753916998933004,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037697969451630135,
      "loss": 4.9828,
      "mean_token_accuracy": 0.2121705561876297,
      "num_tokens": 71065543.0,
      "step": 31835
    },
    {
      "entropy": 5.257774591445923,
      "epoch": 3.575953276801258,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037694390484854323,
      "loss": 4.8282,
      "mean_token_accuracy": 0.217411507666111,
      "num_tokens": 71076450.0,
      "step": 31840
    },
    {
      "entropy": 5.1862787246704105,
      "epoch": 3.5765148537092157,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003769081119348517,
      "loss": 4.7425,
      "mean_token_accuracy": 0.22954797148704528,
      "num_tokens": 71087625.0,
      "step": 31845
    },
    {
      "entropy": 5.20180869102478,
      "epoch": 3.577076430617173,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00037687231577636626,
      "loss": 4.9176,
      "mean_token_accuracy": 0.21837166398763658,
      "num_tokens": 71098205.0,
      "step": 31850
    },
    {
      "entropy": 5.338868284225464,
      "epoch": 3.5776380075251306,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003768365163742269,
      "loss": 5.0185,
      "mean_token_accuracy": 0.21784589886665345,
      "num_tokens": 71109207.0,
      "step": 31855
    },
    {
      "entropy": 5.366505193710327,
      "epoch": 3.5781995844330883,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00037680071372957327,
      "loss": 4.9606,
      "mean_token_accuracy": 0.21800403147935868,
      "num_tokens": 71120911.0,
      "step": 31860
    },
    {
      "entropy": 5.3927229881286625,
      "epoch": 3.5787611613410455,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037676490784354543,
      "loss": 5.0091,
      "mean_token_accuracy": 0.217721888422966,
      "num_tokens": 71131398.0,
      "step": 31865
    },
    {
      "entropy": 5.290842437744141,
      "epoch": 3.579322738249003,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003767290987172835,
      "loss": 4.9578,
      "mean_token_accuracy": 0.22350935339927674,
      "num_tokens": 71142574.0,
      "step": 31870
    },
    {
      "entropy": 5.3581578731536865,
      "epoch": 3.579884315156961,
      "grad_norm": 1.0,
      "learning_rate": 0.00037669328635192757,
      "loss": 4.9568,
      "mean_token_accuracy": 0.21418482363224028,
      "num_tokens": 71154624.0,
      "step": 31875
    },
    {
      "entropy": 5.230339097976684,
      "epoch": 3.580445892064918,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000376657470748618,
      "loss": 4.8224,
      "mean_token_accuracy": 0.22009534239768982,
      "num_tokens": 71165375.0,
      "step": 31880
    },
    {
      "entropy": 5.233891344070434,
      "epoch": 3.5810074689728757,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003766216519084952,
      "loss": 4.8456,
      "mean_token_accuracy": 0.23039909601211547,
      "num_tokens": 71175835.0,
      "step": 31885
    },
    {
      "entropy": 5.287428712844848,
      "epoch": 3.5815690458808334,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003765858298326994,
      "loss": 4.9058,
      "mean_token_accuracy": 0.22159547954797745,
      "num_tokens": 71187173.0,
      "step": 31890
    },
    {
      "entropy": 5.315844917297364,
      "epoch": 3.582130622788791,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037655000452237154,
      "loss": 4.9513,
      "mean_token_accuracy": 0.2215362697839737,
      "num_tokens": 71198424.0,
      "step": 31895
    },
    {
      "entropy": 5.344527101516723,
      "epoch": 3.5826921996967487,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003765141759786521,
      "loss": 4.9515,
      "mean_token_accuracy": 0.2149422600865364,
      "num_tokens": 71210198.0,
      "step": 31900
    },
    {
      "entropy": 5.351575183868408,
      "epoch": 3.583253776604706,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00037647834420268195,
      "loss": 5.0082,
      "mean_token_accuracy": 0.21252341866493224,
      "num_tokens": 71221367.0,
      "step": 31905
    },
    {
      "entropy": 5.331216955184937,
      "epoch": 3.5838153535126636,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003764425091956019,
      "loss": 4.86,
      "mean_token_accuracy": 0.22243351340293885,
      "num_tokens": 71232413.0,
      "step": 31910
    },
    {
      "entropy": 5.303031301498413,
      "epoch": 3.584376930420621,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037640667095855307,
      "loss": 4.9533,
      "mean_token_accuracy": 0.21782980114221573,
      "num_tokens": 71244219.0,
      "step": 31915
    },
    {
      "entropy": 5.213309335708618,
      "epoch": 3.5849385073285784,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00037637082949267654,
      "loss": 4.8273,
      "mean_token_accuracy": 0.22610630989074706,
      "num_tokens": 71254452.0,
      "step": 31920
    },
    {
      "entropy": 5.303627777099609,
      "epoch": 3.585500084236536,
      "grad_norm": 1.0,
      "learning_rate": 0.0003763349847991134,
      "loss": 4.9456,
      "mean_token_accuracy": 0.22082816511392594,
      "num_tokens": 71265267.0,
      "step": 31925
    },
    {
      "entropy": 5.359746742248535,
      "epoch": 3.5860616611444938,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000376299136879005,
      "loss": 5.0092,
      "mean_token_accuracy": 0.2125110685825348,
      "num_tokens": 71277334.0,
      "step": 31930
    },
    {
      "entropy": 5.314145708084107,
      "epoch": 3.5866232380524514,
      "grad_norm": 1.0,
      "learning_rate": 0.00037626328573349296,
      "loss": 4.9236,
      "mean_token_accuracy": 0.2230032578110695,
      "num_tokens": 71289159.0,
      "step": 31935
    },
    {
      "entropy": 5.2570360660552975,
      "epoch": 3.587184814960409,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003762274313637185,
      "loss": 4.8715,
      "mean_token_accuracy": 0.22229373157024385,
      "num_tokens": 71300715.0,
      "step": 31940
    },
    {
      "entropy": 5.265485954284668,
      "epoch": 3.5877463918683663,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003761915737708234,
      "loss": 4.9083,
      "mean_token_accuracy": 0.22280561178922653,
      "num_tokens": 71310948.0,
      "step": 31945
    },
    {
      "entropy": 5.311934900283814,
      "epoch": 3.588307968776324,
      "grad_norm": 0.96875,
      "learning_rate": 0.00037615571295594936,
      "loss": 4.9898,
      "mean_token_accuracy": 0.21043619364500046,
      "num_tokens": 71321958.0,
      "step": 31950
    },
    {
      "entropy": 5.3406599998474125,
      "epoch": 3.5888695456842816,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003761198489202382,
      "loss": 4.9446,
      "mean_token_accuracy": 0.2223346307873726,
      "num_tokens": 71332522.0,
      "step": 31955
    },
    {
      "entropy": 5.377405738830566,
      "epoch": 3.589431122592239,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003760839816648317,
      "loss": 4.942,
      "mean_token_accuracy": 0.2222786396741867,
      "num_tokens": 71344198.0,
      "step": 31960
    },
    {
      "entropy": 5.28923921585083,
      "epoch": 3.5899926995001965,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037604811119087213,
      "loss": 4.9361,
      "mean_token_accuracy": 0.22648613154888153,
      "num_tokens": 71355094.0,
      "step": 31965
    },
    {
      "entropy": 5.327838182449341,
      "epoch": 3.590554276408154,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003760122374995015,
      "loss": 5.005,
      "mean_token_accuracy": 0.21647080332040786,
      "num_tokens": 71366596.0,
      "step": 31970
    },
    {
      "entropy": 5.295250797271729,
      "epoch": 3.5911158533161114,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003759763605918621,
      "loss": 4.8326,
      "mean_token_accuracy": 0.2240265667438507,
      "num_tokens": 71377607.0,
      "step": 31975
    },
    {
      "entropy": 5.255206489562989,
      "epoch": 3.591677430224069,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037594048046909616,
      "loss": 4.9012,
      "mean_token_accuracy": 0.22120285034179688,
      "num_tokens": 71388815.0,
      "step": 31980
    },
    {
      "entropy": 5.237357759475708,
      "epoch": 3.5922390071320267,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003759045971323462,
      "loss": 4.9301,
      "mean_token_accuracy": 0.21682320535182953,
      "num_tokens": 71399600.0,
      "step": 31985
    },
    {
      "entropy": 5.29844708442688,
      "epoch": 3.5928005840399844,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00037586871058275476,
      "loss": 4.8905,
      "mean_token_accuracy": 0.22776577919721602,
      "num_tokens": 71410652.0,
      "step": 31990
    },
    {
      "entropy": 5.299919748306275,
      "epoch": 3.593362160947942,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00037583282082146434,
      "loss": 4.9985,
      "mean_token_accuracy": 0.21938793808221818,
      "num_tokens": 71422656.0,
      "step": 31995
    },
    {
      "entropy": 5.256129932403565,
      "epoch": 3.5939237378558992,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037579692784961785,
      "loss": 4.8483,
      "mean_token_accuracy": 0.21764781475067138,
      "num_tokens": 71433677.0,
      "step": 32000
    },
    {
      "entropy": 5.286105489730835,
      "epoch": 3.594485314763857,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003757610316683581,
      "loss": 4.9403,
      "mean_token_accuracy": 0.21328798681497574,
      "num_tokens": 71445922.0,
      "step": 32005
    },
    {
      "entropy": 5.31122989654541,
      "epoch": 3.5950468916718146,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00037572513227882807,
      "loss": 4.86,
      "mean_token_accuracy": 0.225202314555645,
      "num_tokens": 71457317.0,
      "step": 32010
    },
    {
      "entropy": 5.19009165763855,
      "epoch": 3.595608468579772,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037568922968217065,
      "loss": 4.8325,
      "mean_token_accuracy": 0.2258998140692711,
      "num_tokens": 71468622.0,
      "step": 32015
    },
    {
      "entropy": 5.273753261566162,
      "epoch": 3.5961700454877294,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00037565332387952914,
      "loss": 4.9935,
      "mean_token_accuracy": 0.22313259094953536,
      "num_tokens": 71479254.0,
      "step": 32020
    },
    {
      "entropy": 5.308802127838135,
      "epoch": 3.596731622395687,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037561741487204684,
      "loss": 4.914,
      "mean_token_accuracy": 0.2272108882665634,
      "num_tokens": 71489694.0,
      "step": 32025
    },
    {
      "entropy": 5.373579311370849,
      "epoch": 3.5972931993036448,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003755815026608669,
      "loss": 4.9702,
      "mean_token_accuracy": 0.21833304911851883,
      "num_tokens": 71499983.0,
      "step": 32030
    },
    {
      "entropy": 5.240620565414429,
      "epoch": 3.5978547762116024,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037554558724713305,
      "loss": 4.8694,
      "mean_token_accuracy": 0.22765439599752427,
      "num_tokens": 71511628.0,
      "step": 32035
    },
    {
      "entropy": 5.440347576141358,
      "epoch": 3.5984163531195597,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0003755096686319886,
      "loss": 5.0443,
      "mean_token_accuracy": 0.2110971689224243,
      "num_tokens": 71524432.0,
      "step": 32040
    },
    {
      "entropy": 5.315503406524658,
      "epoch": 3.5989779300275173,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00037547374681657736,
      "loss": 4.9078,
      "mean_token_accuracy": 0.2239917203783989,
      "num_tokens": 71534858.0,
      "step": 32045
    },
    {
      "entropy": 5.277278137207031,
      "epoch": 3.599539506935475,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037543782180204303,
      "loss": 4.9976,
      "mean_token_accuracy": 0.21939639002084732,
      "num_tokens": 71547028.0,
      "step": 32050
    },
    {
      "entropy": 5.346483850479126,
      "epoch": 3.600101083843432,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00037540189358952955,
      "loss": 4.9271,
      "mean_token_accuracy": 0.2226929023861885,
      "num_tokens": 71557477.0,
      "step": 32055
    },
    {
      "entropy": 5.3135090351104735,
      "epoch": 3.60066266075139,
      "grad_norm": 0.875,
      "learning_rate": 0.0003753659621801808,
      "loss": 4.9273,
      "mean_token_accuracy": 0.22158978432416915,
      "num_tokens": 71568568.0,
      "step": 32060
    },
    {
      "entropy": 5.281470584869385,
      "epoch": 3.6012242376593475,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037533002757514095,
      "loss": 4.8528,
      "mean_token_accuracy": 0.2200552985072136,
      "num_tokens": 71579143.0,
      "step": 32065
    },
    {
      "entropy": 5.293504762649536,
      "epoch": 3.6017858145673047,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000375294089775554,
      "loss": 4.9008,
      "mean_token_accuracy": 0.2267071709036827,
      "num_tokens": 71590674.0,
      "step": 32070
    },
    {
      "entropy": 5.271864175796509,
      "epoch": 3.6023473914752624,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003752581487825644,
      "loss": 4.9195,
      "mean_token_accuracy": 0.2244152158498764,
      "num_tokens": 71602133.0,
      "step": 32075
    },
    {
      "entropy": 5.248800039291382,
      "epoch": 3.60290896838322,
      "grad_norm": 0.921875,
      "learning_rate": 0.00037522220459731644,
      "loss": 4.8862,
      "mean_token_accuracy": 0.22366160303354263,
      "num_tokens": 71613735.0,
      "step": 32080
    },
    {
      "entropy": 5.327818918228149,
      "epoch": 3.6034705452911777,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003751862572209546,
      "loss": 4.9063,
      "mean_token_accuracy": 0.2252167910337448,
      "num_tokens": 71625553.0,
      "step": 32085
    },
    {
      "entropy": 5.177931880950927,
      "epoch": 3.6040321221991354,
      "grad_norm": 0.859375,
      "learning_rate": 0.00037515030665462345,
      "loss": 4.7675,
      "mean_token_accuracy": 0.22836246192455292,
      "num_tokens": 71636677.0,
      "step": 32090
    },
    {
      "entropy": 5.272428226470947,
      "epoch": 3.6045936991070926,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00037511435289946775,
      "loss": 4.9491,
      "mean_token_accuracy": 0.21789927184581756,
      "num_tokens": 71648040.0,
      "step": 32095
    },
    {
      "entropy": 5.294327974319458,
      "epoch": 3.6051552760150503,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003750783959566322,
      "loss": 4.8969,
      "mean_token_accuracy": 0.21957550048828126,
      "num_tokens": 71659189.0,
      "step": 32100
    },
    {
      "entropy": 5.277444124221802,
      "epoch": 3.605716852923008,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003750424358272617,
      "loss": 4.7744,
      "mean_token_accuracy": 0.22555863708257676,
      "num_tokens": 71670262.0,
      "step": 32105
    },
    {
      "entropy": 5.202238655090332,
      "epoch": 3.606278429830965,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003750064725125013,
      "loss": 4.879,
      "mean_token_accuracy": 0.2192429080605507,
      "num_tokens": 71680271.0,
      "step": 32110
    },
    {
      "entropy": 5.295335054397583,
      "epoch": 3.606840006738923,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037497050601349605,
      "loss": 4.9459,
      "mean_token_accuracy": 0.22159641087055207,
      "num_tokens": 71691299.0,
      "step": 32115
    },
    {
      "entropy": 5.270822286605835,
      "epoch": 3.6074015836468805,
      "grad_norm": 0.953125,
      "learning_rate": 0.00037493453633139103,
      "loss": 4.8892,
      "mean_token_accuracy": 0.22303142994642258,
      "num_tokens": 71702517.0,
      "step": 32120
    },
    {
      "entropy": 5.241553926467896,
      "epoch": 3.607963160554838,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00037489856346733163,
      "loss": 4.9008,
      "mean_token_accuracy": 0.2276723191142082,
      "num_tokens": 71713386.0,
      "step": 32125
    },
    {
      "entropy": 5.316582822799683,
      "epoch": 3.608524737462796,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003748625874224632,
      "loss": 4.906,
      "mean_token_accuracy": 0.22406211197376252,
      "num_tokens": 71724361.0,
      "step": 32130
    },
    {
      "entropy": 5.323209428787232,
      "epoch": 3.609086314370753,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037482660819793134,
      "loss": 5.0129,
      "mean_token_accuracy": 0.21775518208742142,
      "num_tokens": 71735558.0,
      "step": 32135
    },
    {
      "entropy": 5.316735029220581,
      "epoch": 3.6096478912787107,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003747906257948814,
      "loss": 4.9745,
      "mean_token_accuracy": 0.2178767904639244,
      "num_tokens": 71747253.0,
      "step": 32140
    },
    {
      "entropy": 5.375102424621582,
      "epoch": 3.6102094681866683,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003747546402144594,
      "loss": 4.9892,
      "mean_token_accuracy": 0.21679691821336747,
      "num_tokens": 71760223.0,
      "step": 32145
    },
    {
      "entropy": 5.320663642883301,
      "epoch": 3.6107710450946255,
      "grad_norm": 0.984375,
      "learning_rate": 0.00037471865145781094,
      "loss": 4.8448,
      "mean_token_accuracy": 0.21920929998159408,
      "num_tokens": 71771976.0,
      "step": 32150
    },
    {
      "entropy": 5.241389179229737,
      "epoch": 3.611332622002583,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037468265952608196,
      "loss": 4.9331,
      "mean_token_accuracy": 0.21617104709148408,
      "num_tokens": 71784364.0,
      "step": 32155
    },
    {
      "entropy": 5.3200867652893065,
      "epoch": 3.611894198910541,
      "grad_norm": 1.0,
      "learning_rate": 0.0003746466644204183,
      "loss": 5.0092,
      "mean_token_accuracy": 0.2148905411362648,
      "num_tokens": 71795495.0,
      "step": 32160
    },
    {
      "entropy": 5.348194932937622,
      "epoch": 3.612455775818498,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003746106661419663,
      "loss": 4.953,
      "mean_token_accuracy": 0.2230401501059532,
      "num_tokens": 71806475.0,
      "step": 32165
    },
    {
      "entropy": 5.344534254074096,
      "epoch": 3.6130173527264557,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000374574664691872,
      "loss": 4.9467,
      "mean_token_accuracy": 0.21974625289440156,
      "num_tokens": 71816701.0,
      "step": 32170
    },
    {
      "entropy": 5.260061407089234,
      "epoch": 3.6135789296344134,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037453866007128176,
      "loss": 4.8829,
      "mean_token_accuracy": 0.22086270153522491,
      "num_tokens": 71827860.0,
      "step": 32175
    },
    {
      "entropy": 5.26414942741394,
      "epoch": 3.614140506542371,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037450265228134203,
      "loss": 4.8869,
      "mean_token_accuracy": 0.2265329033136368,
      "num_tokens": 71839631.0,
      "step": 32180
    },
    {
      "entropy": 5.312286472320556,
      "epoch": 3.6147020834503287,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037446664132319925,
      "loss": 4.8975,
      "mean_token_accuracy": 0.22459484487771988,
      "num_tokens": 71849926.0,
      "step": 32185
    },
    {
      "entropy": 5.214063405990601,
      "epoch": 3.615263660358286,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000374430627198,
      "loss": 4.8688,
      "mean_token_accuracy": 0.22545913904905318,
      "num_tokens": 71860815.0,
      "step": 32190
    },
    {
      "entropy": 5.3312177658081055,
      "epoch": 3.6158252372662436,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037439460990689106,
      "loss": 4.8792,
      "mean_token_accuracy": 0.22601479887962342,
      "num_tokens": 71871814.0,
      "step": 32195
    },
    {
      "entropy": 5.317677116394043,
      "epoch": 3.6163868141742013,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003743585894510192,
      "loss": 4.9054,
      "mean_token_accuracy": 0.22138668596744537,
      "num_tokens": 71881606.0,
      "step": 32200
    },
    {
      "entropy": 5.24078631401062,
      "epoch": 3.6169483910821585,
      "grad_norm": 0.96875,
      "learning_rate": 0.00037432256583153127,
      "loss": 4.9206,
      "mean_token_accuracy": 0.21587447822093964,
      "num_tokens": 71893371.0,
      "step": 32205
    },
    {
      "entropy": 5.306335973739624,
      "epoch": 3.617509967990116,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00037428653904957433,
      "loss": 4.9502,
      "mean_token_accuracy": 0.22455600947141646,
      "num_tokens": 71906068.0,
      "step": 32210
    },
    {
      "entropy": 5.3248790264129635,
      "epoch": 3.618071544898074,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003742505091062955,
      "loss": 4.9706,
      "mean_token_accuracy": 0.21502378433942795,
      "num_tokens": 71917450.0,
      "step": 32215
    },
    {
      "entropy": 5.340534305572509,
      "epoch": 3.6186331218060315,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003742144760028419,
      "loss": 4.8995,
      "mean_token_accuracy": 0.21637608706951142,
      "num_tokens": 71928333.0,
      "step": 32220
    },
    {
      "entropy": 5.26017804145813,
      "epoch": 3.619194698713989,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000374178439740361,
      "loss": 4.935,
      "mean_token_accuracy": 0.2204379618167877,
      "num_tokens": 71939836.0,
      "step": 32225
    },
    {
      "entropy": 5.346491575241089,
      "epoch": 3.6197562756219464,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003741424003200001,
      "loss": 4.9868,
      "mean_token_accuracy": 0.21904581040143967,
      "num_tokens": 71951120.0,
      "step": 32230
    },
    {
      "entropy": 5.328101301193238,
      "epoch": 3.620317852529904,
      "grad_norm": 1.0,
      "learning_rate": 0.00037410635774290667,
      "loss": 4.9245,
      "mean_token_accuracy": 0.22197225838899612,
      "num_tokens": 71961547.0,
      "step": 32235
    },
    {
      "entropy": 5.313261365890503,
      "epoch": 3.6208794294378617,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003740703120102284,
      "loss": 4.8813,
      "mean_token_accuracy": 0.22379183769226074,
      "num_tokens": 71973287.0,
      "step": 32240
    },
    {
      "entropy": 5.31168384552002,
      "epoch": 3.621441006345819,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000374034263123113,
      "loss": 4.9004,
      "mean_token_accuracy": 0.22908279150724412,
      "num_tokens": 71984205.0,
      "step": 32245
    },
    {
      "entropy": 5.193104934692383,
      "epoch": 3.6220025832537766,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00037399821108270823,
      "loss": 4.9279,
      "mean_token_accuracy": 0.21983107477426528,
      "num_tokens": 71995090.0,
      "step": 32250
    },
    {
      "entropy": 5.378453254699707,
      "epoch": 3.622564160161734,
      "grad_norm": 1.015625,
      "learning_rate": 0.000373962155890162,
      "loss": 5.0527,
      "mean_token_accuracy": 0.215772046148777,
      "num_tokens": 72005504.0,
      "step": 32255
    },
    {
      "entropy": 5.336942100524903,
      "epoch": 3.6231257370696914,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003739260975466223,
      "loss": 4.9059,
      "mean_token_accuracy": 0.22336162775754928,
      "num_tokens": 72017250.0,
      "step": 32260
    },
    {
      "entropy": 5.3354939937591555,
      "epoch": 3.623687313977649,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00037389003605323737,
      "loss": 4.8524,
      "mean_token_accuracy": 0.22502138316631318,
      "num_tokens": 72028089.0,
      "step": 32265
    },
    {
      "entropy": 5.304392242431641,
      "epoch": 3.6242488908856068,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00037385397141115535,
      "loss": 4.91,
      "mean_token_accuracy": 0.22094779908657075,
      "num_tokens": 72039140.0,
      "step": 32270
    },
    {
      "entropy": 5.292251968383789,
      "epoch": 3.6248104677935644,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00037381790362152453,
      "loss": 4.9377,
      "mean_token_accuracy": 0.21659452021121978,
      "num_tokens": 72049944.0,
      "step": 32275
    },
    {
      "entropy": 5.227715158462525,
      "epoch": 3.625372044701522,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00037378183268549326,
      "loss": 4.8678,
      "mean_token_accuracy": 0.2238489493727684,
      "num_tokens": 72061799.0,
      "step": 32280
    },
    {
      "entropy": 5.265416860580444,
      "epoch": 3.6259336216094793,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003737457586042102,
      "loss": 4.8734,
      "mean_token_accuracy": 0.22853825837373734,
      "num_tokens": 72073539.0,
      "step": 32285
    },
    {
      "entropy": 5.278962469100952,
      "epoch": 3.626495198517437,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003737096813788238,
      "loss": 4.9134,
      "mean_token_accuracy": 0.2162514865398407,
      "num_tokens": 72085268.0,
      "step": 32290
    },
    {
      "entropy": 5.373533535003662,
      "epoch": 3.6270567754253946,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003736736010104829,
      "loss": 5.097,
      "mean_token_accuracy": 0.20873973965644838,
      "num_tokens": 72096269.0,
      "step": 32295
    },
    {
      "entropy": 5.259831666946411,
      "epoch": 3.627618352333352,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003736375175003363,
      "loss": 4.8882,
      "mean_token_accuracy": 0.22586388736963273,
      "num_tokens": 72108934.0,
      "step": 32300
    },
    {
      "entropy": 5.290205097198486,
      "epoch": 3.6281799292413095,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037360143084953285,
      "loss": 4.8896,
      "mean_token_accuracy": 0.2241421550512314,
      "num_tokens": 72118371.0,
      "step": 32305
    },
    {
      "entropy": 5.235533809661865,
      "epoch": 3.628741506149267,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003735653410592217,
      "loss": 4.8198,
      "mean_token_accuracy": 0.231116446852684,
      "num_tokens": 72129187.0,
      "step": 32310
    },
    {
      "entropy": 5.250532150268555,
      "epoch": 3.629303083057225,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037352924813055185,
      "loss": 4.8816,
      "mean_token_accuracy": 0.22909190952777864,
      "num_tokens": 72141317.0,
      "step": 32315
    },
    {
      "entropy": 5.297877836227417,
      "epoch": 3.6298646599651825,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003734931520646724,
      "loss": 4.8898,
      "mean_token_accuracy": 0.22389761805534364,
      "num_tokens": 72152754.0,
      "step": 32320
    },
    {
      "entropy": 5.19847526550293,
      "epoch": 3.6304262368731397,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00037345705286273285,
      "loss": 4.7789,
      "mean_token_accuracy": 0.23565226495265962,
      "num_tokens": 72162115.0,
      "step": 32325
    },
    {
      "entropy": 5.201715040206909,
      "epoch": 3.6309878137810974,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003734209505258825,
      "loss": 4.8335,
      "mean_token_accuracy": 0.2253744274377823,
      "num_tokens": 72173633.0,
      "step": 32330
    },
    {
      "entropy": 5.296513271331787,
      "epoch": 3.631549390689055,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037338484505527103,
      "loss": 4.9287,
      "mean_token_accuracy": 0.21149974465370178,
      "num_tokens": 72185274.0,
      "step": 32335
    },
    {
      "entropy": 5.334851837158203,
      "epoch": 3.6321109675970122,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037334873645204785,
      "loss": 4.9867,
      "mean_token_accuracy": 0.2185519814491272,
      "num_tokens": 72195991.0,
      "step": 32340
    },
    {
      "entropy": 5.269070053100586,
      "epoch": 3.63267254450497,
      "grad_norm": 0.953125,
      "learning_rate": 0.00037331262471736285,
      "loss": 4.9036,
      "mean_token_accuracy": 0.22567045837640762,
      "num_tokens": 72207390.0,
      "step": 32345
    },
    {
      "entropy": 5.287621355056762,
      "epoch": 3.6332341214129276,
      "grad_norm": 0.921875,
      "learning_rate": 0.00037327650985236564,
      "loss": 4.95,
      "mean_token_accuracy": 0.22101799249649048,
      "num_tokens": 72218747.0,
      "step": 32350
    },
    {
      "entropy": 5.290278959274292,
      "epoch": 3.633795698320885,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003732403918582063,
      "loss": 4.9162,
      "mean_token_accuracy": 0.22070825845003128,
      "num_tokens": 72230623.0,
      "step": 32355
    },
    {
      "entropy": 5.293929624557495,
      "epoch": 3.6343572752288424,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00037320427073603475,
      "loss": 4.8724,
      "mean_token_accuracy": 0.22265358716249467,
      "num_tokens": 72242510.0,
      "step": 32360
    },
    {
      "entropy": 5.229230546951294,
      "epoch": 3.6349188521368,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037316814648700113,
      "loss": 4.8445,
      "mean_token_accuracy": 0.2251549169421196,
      "num_tokens": 72253498.0,
      "step": 32365
    },
    {
      "entropy": 5.248933887481689,
      "epoch": 3.6354804290447578,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003731320191122557,
      "loss": 4.9694,
      "mean_token_accuracy": 0.22171658426523208,
      "num_tokens": 72265698.0,
      "step": 32370
    },
    {
      "entropy": 5.253143692016602,
      "epoch": 3.6360420059527154,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003730958886129486,
      "loss": 4.7929,
      "mean_token_accuracy": 0.23728069216012954,
      "num_tokens": 72276563.0,
      "step": 32375
    },
    {
      "entropy": 5.250199556350708,
      "epoch": 3.6366035828606726,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00037305975499023046,
      "loss": 4.8792,
      "mean_token_accuracy": 0.2259607806801796,
      "num_tokens": 72286826.0,
      "step": 32380
    },
    {
      "entropy": 5.351262712478638,
      "epoch": 3.6371651597686303,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003730236182452517,
      "loss": 5.0351,
      "mean_token_accuracy": 0.21663929224014283,
      "num_tokens": 72299446.0,
      "step": 32385
    },
    {
      "entropy": 5.3811084747314455,
      "epoch": 3.637726736676588,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003729874783791629,
      "loss": 5.0563,
      "mean_token_accuracy": 0.21554765552282334,
      "num_tokens": 72310894.0,
      "step": 32390
    },
    {
      "entropy": 5.346879339218139,
      "epoch": 3.638288313584545,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003729513353931148,
      "loss": 4.9509,
      "mean_token_accuracy": 0.2241373121738434,
      "num_tokens": 72322810.0,
      "step": 32395
    },
    {
      "entropy": 5.325802516937256,
      "epoch": 3.638849890492503,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003729151892882581,
      "loss": 4.9626,
      "mean_token_accuracy": 0.2219771698117256,
      "num_tokens": 72334509.0,
      "step": 32400
    },
    {
      "entropy": 5.248663902282715,
      "epoch": 3.6394114674004605,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003728790400657439,
      "loss": 4.9178,
      "mean_token_accuracy": 0.22065912336111068,
      "num_tokens": 72345036.0,
      "step": 32405
    },
    {
      "entropy": 5.292475032806396,
      "epoch": 3.639973044308418,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00037284288772672315,
      "loss": 4.8737,
      "mean_token_accuracy": 0.2184360146522522,
      "num_tokens": 72356029.0,
      "step": 32410
    },
    {
      "entropy": 5.315659999847412,
      "epoch": 3.640534621216376,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037280673227234687,
      "loss": 4.9716,
      "mean_token_accuracy": 0.21762108504772187,
      "num_tokens": 72366548.0,
      "step": 32415
    },
    {
      "entropy": 5.225948524475098,
      "epoch": 3.641096198124333,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003727705737037662,
      "loss": 4.8735,
      "mean_token_accuracy": 0.2224797487258911,
      "num_tokens": 72377271.0,
      "step": 32420
    },
    {
      "entropy": 5.286157035827637,
      "epoch": 3.6416577750322907,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003727344120221326,
      "loss": 4.9369,
      "mean_token_accuracy": 0.22156209945678712,
      "num_tokens": 72388651.0,
      "step": 32425
    },
    {
      "entropy": 5.259539842605591,
      "epoch": 3.6422193519402484,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00037269824722859743,
      "loss": 4.9866,
      "mean_token_accuracy": 0.2184301346540451,
      "num_tokens": 72400955.0,
      "step": 32430
    },
    {
      "entropy": 5.342967319488525,
      "epoch": 3.6427809288482056,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003726620793243121,
      "loss": 4.9235,
      "mean_token_accuracy": 0.22027406543493272,
      "num_tokens": 72412372.0,
      "step": 32435
    },
    {
      "entropy": 5.325666618347168,
      "epoch": 3.6433425057561633,
      "grad_norm": 1.0,
      "learning_rate": 0.0003726259083104284,
      "loss": 4.8875,
      "mean_token_accuracy": 0.21628865003585815,
      "num_tokens": 72422664.0,
      "step": 32440
    },
    {
      "entropy": 5.2946614742279055,
      "epoch": 3.643904082664121,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00037258973418809784,
      "loss": 4.9407,
      "mean_token_accuracy": 0.21367895007133483,
      "num_tokens": 72434829.0,
      "step": 32445
    },
    {
      "entropy": 5.237254285812378,
      "epoch": 3.644465659572078,
      "grad_norm": 1.0,
      "learning_rate": 0.0003725535569584723,
      "loss": 4.8409,
      "mean_token_accuracy": 0.22360854148864745,
      "num_tokens": 72445140.0,
      "step": 32450
    },
    {
      "entropy": 5.254906749725341,
      "epoch": 3.645027236480036,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037251737662270367,
      "loss": 4.8348,
      "mean_token_accuracy": 0.23005046248435973,
      "num_tokens": 72455281.0,
      "step": 32455
    },
    {
      "entropy": 5.279285192489624,
      "epoch": 3.6455888133879935,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000372481193181944,
      "loss": 4.9411,
      "mean_token_accuracy": 0.21739633828401567,
      "num_tokens": 72466296.0,
      "step": 32460
    },
    {
      "entropy": 5.251202249526978,
      "epoch": 3.646150390295951,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037244500663734524,
      "loss": 4.8764,
      "mean_token_accuracy": 0.2234960362315178,
      "num_tokens": 72477390.0,
      "step": 32465
    },
    {
      "entropy": 5.25613374710083,
      "epoch": 3.6467119672039088,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003724088169900597,
      "loss": 4.8325,
      "mean_token_accuracy": 0.23173629641532897,
      "num_tokens": 72488241.0,
      "step": 32470
    },
    {
      "entropy": 5.35576868057251,
      "epoch": 3.647273544111866,
      "grad_norm": 0.953125,
      "learning_rate": 0.00037237262424123975,
      "loss": 4.9376,
      "mean_token_accuracy": 0.22006672620773315,
      "num_tokens": 72498823.0,
      "step": 32475
    },
    {
      "entropy": 5.327776098251343,
      "epoch": 3.6478351210198237,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00037233642839203763,
      "loss": 4.9871,
      "mean_token_accuracy": 0.21685864478349687,
      "num_tokens": 72509399.0,
      "step": 32480
    },
    {
      "entropy": 5.334390449523926,
      "epoch": 3.6483966979277813,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003723002294436059,
      "loss": 4.9865,
      "mean_token_accuracy": 0.21541919708251953,
      "num_tokens": 72519182.0,
      "step": 32485
    },
    {
      "entropy": 5.242059946060181,
      "epoch": 3.6489582748357385,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037226402739709704,
      "loss": 4.816,
      "mean_token_accuracy": 0.22491220235824586,
      "num_tokens": 72528852.0,
      "step": 32490
    },
    {
      "entropy": 5.271094512939453,
      "epoch": 3.649519851743696,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003722278222536639,
      "loss": 4.9285,
      "mean_token_accuracy": 0.21861349493265153,
      "num_tokens": 72540477.0,
      "step": 32495
    },
    {
      "entropy": 5.35116548538208,
      "epoch": 3.650081428651654,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037219161401445914,
      "loss": 4.9296,
      "mean_token_accuracy": 0.21478652507066726,
      "num_tokens": 72553404.0,
      "step": 32500
    },
    {
      "entropy": 5.349891519546508,
      "epoch": 3.6506430055596115,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00037215540268063573,
      "loss": 4.9497,
      "mean_token_accuracy": 0.2168440267443657,
      "num_tokens": 72564516.0,
      "step": 32505
    },
    {
      "entropy": 5.2468705654144285,
      "epoch": 3.651204582467569,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003721191882533466,
      "loss": 4.8373,
      "mean_token_accuracy": 0.22324019968509673,
      "num_tokens": 72575888.0,
      "step": 32510
    },
    {
      "entropy": 5.267144680023193,
      "epoch": 3.6517661593755264,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003720829707337449,
      "loss": 4.9007,
      "mean_token_accuracy": 0.21799544394016265,
      "num_tokens": 72586629.0,
      "step": 32515
    },
    {
      "entropy": 5.285119962692261,
      "epoch": 3.652327736283484,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003720467501229836,
      "loss": 4.905,
      "mean_token_accuracy": 0.22425225377082825,
      "num_tokens": 72597264.0,
      "step": 32520
    },
    {
      "entropy": 5.323624897003174,
      "epoch": 3.6528893131914417,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003720105264222163,
      "loss": 4.9018,
      "mean_token_accuracy": 0.2193514123558998,
      "num_tokens": 72608450.0,
      "step": 32525
    },
    {
      "entropy": 5.268409204483032,
      "epoch": 3.653450890099399,
      "grad_norm": 0.890625,
      "learning_rate": 0.00037197429963259615,
      "loss": 4.9536,
      "mean_token_accuracy": 0.2161368504166603,
      "num_tokens": 72620538.0,
      "step": 32530
    },
    {
      "entropy": 5.346345138549805,
      "epoch": 3.6540124670073566,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0003719380697552767,
      "loss": 4.9943,
      "mean_token_accuracy": 0.22106500416994096,
      "num_tokens": 72632589.0,
      "step": 32535
    },
    {
      "entropy": 5.310322093963623,
      "epoch": 3.6545740439153143,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003719018367914115,
      "loss": 4.906,
      "mean_token_accuracy": 0.220661661028862,
      "num_tokens": 72643749.0,
      "step": 32540
    },
    {
      "entropy": 5.281716060638428,
      "epoch": 3.6551356208232715,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003718656007421542,
      "loss": 4.9562,
      "mean_token_accuracy": 0.22074756324291228,
      "num_tokens": 72655485.0,
      "step": 32545
    },
    {
      "entropy": 5.353695440292358,
      "epoch": 3.655697197731229,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00037182936160865866,
      "loss": 4.972,
      "mean_token_accuracy": 0.2174240306019783,
      "num_tokens": 72666175.0,
      "step": 32550
    },
    {
      "entropy": 5.325708341598511,
      "epoch": 3.656258774639187,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003717931193920788,
      "loss": 4.8659,
      "mean_token_accuracy": 0.22015419155359267,
      "num_tokens": 72675858.0,
      "step": 32555
    },
    {
      "entropy": 5.206853723526001,
      "epoch": 3.6568203515471445,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003717568740935683,
      "loss": 4.7949,
      "mean_token_accuracy": 0.22563571035861968,
      "num_tokens": 72686623.0,
      "step": 32560
    },
    {
      "entropy": 5.259970617294312,
      "epoch": 3.657381928455102,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003717206257142815,
      "loss": 4.8966,
      "mean_token_accuracy": 0.22824667096138002,
      "num_tokens": 72698002.0,
      "step": 32565
    },
    {
      "entropy": 5.265815114974975,
      "epoch": 3.6579435053630593,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003716843742553724,
      "loss": 4.9252,
      "mean_token_accuracy": 0.22549088150262833,
      "num_tokens": 72708981.0,
      "step": 32570
    },
    {
      "entropy": 5.213521337509155,
      "epoch": 3.658505082271017,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003716481197179953,
      "loss": 4.9174,
      "mean_token_accuracy": 0.21938764452934265,
      "num_tokens": 72719764.0,
      "step": 32575
    },
    {
      "entropy": 5.463380146026611,
      "epoch": 3.6590666591789747,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037161186210330464,
      "loss": 5.1177,
      "mean_token_accuracy": 0.2109011396765709,
      "num_tokens": 72730915.0,
      "step": 32580
    },
    {
      "entropy": 5.381714105606079,
      "epoch": 3.659628236086932,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00037157560141245485,
      "loss": 4.9804,
      "mean_token_accuracy": 0.21646557599306107,
      "num_tokens": 72741869.0,
      "step": 32585
    },
    {
      "entropy": 5.30565996170044,
      "epoch": 3.6601898129948895,
      "grad_norm": 0.96875,
      "learning_rate": 0.00037153933764660034,
      "loss": 4.9257,
      "mean_token_accuracy": 0.22050319910049437,
      "num_tokens": 72753638.0,
      "step": 32590
    },
    {
      "entropy": 5.346259927749633,
      "epoch": 3.660751389902847,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000371503070806896,
      "loss": 4.9346,
      "mean_token_accuracy": 0.22235694527626038,
      "num_tokens": 72765505.0,
      "step": 32595
    },
    {
      "entropy": 5.285023403167725,
      "epoch": 3.661312966810805,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037146680089449643,
      "loss": 4.787,
      "mean_token_accuracy": 0.22715942561626434,
      "num_tokens": 72776348.0,
      "step": 32600
    },
    {
      "entropy": 5.278302001953125,
      "epoch": 3.6618745437187625,
      "grad_norm": 1.0,
      "learning_rate": 0.00037143052791055654,
      "loss": 4.9785,
      "mean_token_accuracy": 0.21298522204160691,
      "num_tokens": 72787720.0,
      "step": 32605
    },
    {
      "entropy": 5.22928032875061,
      "epoch": 3.6624361206267197,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003713942518562311,
      "loss": 4.8812,
      "mean_token_accuracy": 0.22096941173076629,
      "num_tokens": 72798947.0,
      "step": 32610
    },
    {
      "entropy": 5.349171829223633,
      "epoch": 3.6629976975346774,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003713579727326754,
      "loss": 5.0154,
      "mean_token_accuracy": 0.21486082971096038,
      "num_tokens": 72810859.0,
      "step": 32615
    },
    {
      "entropy": 5.269004249572754,
      "epoch": 3.663559274442635,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00037132169054104446,
      "loss": 4.8292,
      "mean_token_accuracy": 0.23035752326250075,
      "num_tokens": 72821089.0,
      "step": 32620
    },
    {
      "entropy": 5.212738084793091,
      "epoch": 3.6641208513505923,
      "grad_norm": 1.0,
      "learning_rate": 0.00037128540528249355,
      "loss": 4.8139,
      "mean_token_accuracy": 0.22339337319135666,
      "num_tokens": 72832226.0,
      "step": 32625
    },
    {
      "entropy": 5.283232975006103,
      "epoch": 3.66468242825855,
      "grad_norm": 0.984375,
      "learning_rate": 0.000371249116958178,
      "loss": 4.9189,
      "mean_token_accuracy": 0.22496186792850495,
      "num_tokens": 72843939.0,
      "step": 32630
    },
    {
      "entropy": 5.231316661834716,
      "epoch": 3.6652440051665076,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00037121282556925326,
      "loss": 4.7783,
      "mean_token_accuracy": 0.23150470554828645,
      "num_tokens": 72854332.0,
      "step": 32635
    },
    {
      "entropy": 5.3468687534332275,
      "epoch": 3.665805582074465,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003711765311168748,
      "loss": 5.0598,
      "mean_token_accuracy": 0.21167584657669067,
      "num_tokens": 72865530.0,
      "step": 32640
    },
    {
      "entropy": 5.362318229675293,
      "epoch": 3.6663671589824225,
      "grad_norm": 0.984375,
      "learning_rate": 0.00037114023360219824,
      "loss": 5.0026,
      "mean_token_accuracy": 0.21602636128664016,
      "num_tokens": 72877683.0,
      "step": 32645
    },
    {
      "entropy": 5.299330711364746,
      "epoch": 3.66692873589038,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037110393302637937,
      "loss": 4.9312,
      "mean_token_accuracy": 0.22305779457092284,
      "num_tokens": 72888926.0,
      "step": 32650
    },
    {
      "entropy": 5.303528499603272,
      "epoch": 3.667490312798338,
      "grad_norm": 0.921875,
      "learning_rate": 0.00037106762939057404,
      "loss": 4.968,
      "mean_token_accuracy": 0.21833346784114838,
      "num_tokens": 72901318.0,
      "step": 32655
    },
    {
      "entropy": 5.376237583160401,
      "epoch": 3.6680518897062955,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003710313226959381,
      "loss": 4.9641,
      "mean_token_accuracy": 0.22382885813713074,
      "num_tokens": 72913177.0,
      "step": 32660
    },
    {
      "entropy": 5.359752082824707,
      "epoch": 3.6686134666142527,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003709950129436276,
      "loss": 5.028,
      "mean_token_accuracy": 0.21154063940048218,
      "num_tokens": 72924322.0,
      "step": 32665
    },
    {
      "entropy": 5.281236886978149,
      "epoch": 3.6691750435222104,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003709587001347987,
      "loss": 4.9056,
      "mean_token_accuracy": 0.22522212117910384,
      "num_tokens": 72936287.0,
      "step": 32670
    },
    {
      "entropy": 5.3095921039581295,
      "epoch": 3.669736620430168,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00037092238427060756,
      "loss": 4.9739,
      "mean_token_accuracy": 0.2150696784257889,
      "num_tokens": 72949262.0,
      "step": 32675
    },
    {
      "entropy": 5.350678968429565,
      "epoch": 3.6702981973381252,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003708860653522104,
      "loss": 4.9922,
      "mean_token_accuracy": 0.21845998615026474,
      "num_tokens": 72959283.0,
      "step": 32680
    },
    {
      "entropy": 5.28519229888916,
      "epoch": 3.670859774246083,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003708497433807639,
      "loss": 4.8577,
      "mean_token_accuracy": 0.23097195029258727,
      "num_tokens": 72971063.0,
      "step": 32685
    },
    {
      "entropy": 5.285391521453858,
      "epoch": 3.6714213511540406,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037081341835742433,
      "loss": 4.8298,
      "mean_token_accuracy": 0.22815779000520706,
      "num_tokens": 72980775.0,
      "step": 32690
    },
    {
      "entropy": 5.207010746002197,
      "epoch": 3.671982928061998,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003707770902833483,
      "loss": 4.8781,
      "mean_token_accuracy": 0.22639305740594864,
      "num_tokens": 72992166.0,
      "step": 32695
    },
    {
      "entropy": 5.140190505981446,
      "epoch": 3.672544504969956,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003707407591596926,
      "loss": 4.8203,
      "mean_token_accuracy": 0.22188149839639665,
      "num_tokens": 73003194.0,
      "step": 32700
    },
    {
      "entropy": 5.278506517410278,
      "epoch": 3.673106081877913,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000370704424987614,
      "loss": 4.9178,
      "mean_token_accuracy": 0.22101584672927857,
      "num_tokens": 73013375.0,
      "step": 32705
    },
    {
      "entropy": 5.3337242126464846,
      "epoch": 3.6736676587858708,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003706680877682694,
      "loss": 4.9303,
      "mean_token_accuracy": 0.226487335562706,
      "num_tokens": 73024532.0,
      "step": 32710
    },
    {
      "entropy": 5.262618112564087,
      "epoch": 3.6742292356938284,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037063174750281573,
      "loss": 4.8726,
      "mean_token_accuracy": 0.22603003233671187,
      "num_tokens": 73034721.0,
      "step": 32715
    },
    {
      "entropy": 5.2352746486663815,
      "epoch": 3.6747908126017856,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003705954041924102,
      "loss": 4.8462,
      "mean_token_accuracy": 0.23028165102005005,
      "num_tokens": 73045894.0,
      "step": 32720
    },
    {
      "entropy": 5.1473000049591064,
      "epoch": 3.6753523895097433,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00037055905783821,
      "loss": 4.73,
      "mean_token_accuracy": 0.24466962218284607,
      "num_tokens": 73057370.0,
      "step": 32725
    },
    {
      "entropy": 5.318375539779663,
      "epoch": 3.675913966417701,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00037052270844137216,
      "loss": 5.0017,
      "mean_token_accuracy": 0.22056709975004196,
      "num_tokens": 73067700.0,
      "step": 32730
    },
    {
      "entropy": 5.226662349700928,
      "epoch": 3.676475543325658,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0003704863560030543,
      "loss": 4.8464,
      "mean_token_accuracy": 0.2241429567337036,
      "num_tokens": 73079612.0,
      "step": 32735
    },
    {
      "entropy": 5.316842937469483,
      "epoch": 3.677037120233616,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003704500005244139,
      "loss": 4.9305,
      "mean_token_accuracy": 0.22004917562007903,
      "num_tokens": 73091060.0,
      "step": 32740
    },
    {
      "entropy": 5.343201398849487,
      "epoch": 3.6775986971415735,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00037041364200660835,
      "loss": 4.9485,
      "mean_token_accuracy": 0.21891173869371414,
      "num_tokens": 73102218.0,
      "step": 32745
    },
    {
      "entropy": 5.35313138961792,
      "epoch": 3.678160274049531,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037037728045079553,
      "loss": 4.9926,
      "mean_token_accuracy": 0.21920866519212723,
      "num_tokens": 73113669.0,
      "step": 32750
    },
    {
      "entropy": 5.288056135177612,
      "epoch": 3.678721850957489,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000370340915858133,
      "loss": 4.845,
      "mean_token_accuracy": 0.22538647502660752,
      "num_tokens": 73124765.0,
      "step": 32755
    },
    {
      "entropy": 5.312839794158935,
      "epoch": 3.679283427865446,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003703045482297788,
      "loss": 4.9773,
      "mean_token_accuracy": 0.2163061261177063,
      "num_tokens": 73135278.0,
      "step": 32760
    },
    {
      "entropy": 5.195609188079834,
      "epoch": 3.6798450047734037,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003702681775668908,
      "loss": 4.7463,
      "mean_token_accuracy": 0.22953661680221557,
      "num_tokens": 73145176.0,
      "step": 32765
    },
    {
      "entropy": 5.399434041976929,
      "epoch": 3.6804065816813614,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003702318038706272,
      "loss": 5.026,
      "mean_token_accuracy": 0.20964398831129075,
      "num_tokens": 73156001.0,
      "step": 32770
    },
    {
      "entropy": 5.226847314834595,
      "epoch": 3.6809681585893186,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003701954271421459,
      "loss": 4.8589,
      "mean_token_accuracy": 0.21999670565128326,
      "num_tokens": 73167116.0,
      "step": 32775
    },
    {
      "entropy": 5.333441114425659,
      "epoch": 3.6815297354972762,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037015904738260533,
      "loss": 4.945,
      "mean_token_accuracy": 0.22576695680618286,
      "num_tokens": 73178750.0,
      "step": 32780
    },
    {
      "entropy": 5.260503149032592,
      "epoch": 3.682091312405234,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003701226645931637,
      "loss": 4.8684,
      "mean_token_accuracy": 0.22995112389326094,
      "num_tokens": 73189781.0,
      "step": 32785
    },
    {
      "entropy": 5.234999418258667,
      "epoch": 3.6826528893131916,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003700862787749796,
      "loss": 4.8238,
      "mean_token_accuracy": 0.23072213381528855,
      "num_tokens": 73200374.0,
      "step": 32790
    },
    {
      "entropy": 5.3441181659698485,
      "epoch": 3.6832144662211492,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003700498899292115,
      "loss": 4.9869,
      "mean_token_accuracy": 0.21026737242937088,
      "num_tokens": 73211613.0,
      "step": 32795
    },
    {
      "entropy": 5.277911710739136,
      "epoch": 3.6837760431291064,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003700134980570181,
      "loss": 4.8846,
      "mean_token_accuracy": 0.2327021762728691,
      "num_tokens": 73222309.0,
      "step": 32800
    },
    {
      "entropy": 5.406866073608398,
      "epoch": 3.684337620037064,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003699771031595579,
      "loss": 5.1343,
      "mean_token_accuracy": 0.20923101752996445,
      "num_tokens": 73233956.0,
      "step": 32805
    },
    {
      "entropy": 5.377783250808716,
      "epoch": 3.6848991969450218,
      "grad_norm": 1.0,
      "learning_rate": 0.00036994070523799005,
      "loss": 5.0128,
      "mean_token_accuracy": 0.21310620307922362,
      "num_tokens": 73245218.0,
      "step": 32810
    },
    {
      "entropy": 5.304779481887818,
      "epoch": 3.685460773852979,
      "grad_norm": 1.0,
      "learning_rate": 0.00036990430429347334,
      "loss": 4.8808,
      "mean_token_accuracy": 0.23065072894096375,
      "num_tokens": 73255830.0,
      "step": 32815
    },
    {
      "entropy": 5.232605981826782,
      "epoch": 3.6860223507609367,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00036986790032716673,
      "loss": 4.795,
      "mean_token_accuracy": 0.23119370490312577,
      "num_tokens": 73266384.0,
      "step": 32820
    },
    {
      "entropy": 5.2484776973724365,
      "epoch": 3.6865839276688943,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003698314933402292,
      "loss": 4.9038,
      "mean_token_accuracy": 0.22062620371580124,
      "num_tokens": 73276839.0,
      "step": 32825
    },
    {
      "entropy": 5.341525888442993,
      "epoch": 3.6871455045768515,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003697950833338204,
      "loss": 4.9827,
      "mean_token_accuracy": 0.22317744493484498,
      "num_tokens": 73287398.0,
      "step": 32830
    },
    {
      "entropy": 5.298176527023315,
      "epoch": 3.687707081484809,
      "grad_norm": 1.0,
      "learning_rate": 0.00036975867030909933,
      "loss": 4.8706,
      "mean_token_accuracy": 0.21931006014347076,
      "num_tokens": 73297432.0,
      "step": 32835
    },
    {
      "entropy": 5.229006290435791,
      "epoch": 3.688268658392767,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036972225426722545,
      "loss": 4.7632,
      "mean_token_accuracy": 0.2263340026140213,
      "num_tokens": 73308539.0,
      "step": 32840
    },
    {
      "entropy": 5.287004470825195,
      "epoch": 3.6888302353007245,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003696858352093582,
      "loss": 4.8969,
      "mean_token_accuracy": 0.22943985164165498,
      "num_tokens": 73319698.0,
      "step": 32845
    },
    {
      "entropy": 5.243289661407471,
      "epoch": 3.689391812208682,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036964941313665715,
      "loss": 4.9365,
      "mean_token_accuracy": 0.2195710599422455,
      "num_tokens": 73329747.0,
      "step": 32850
    },
    {
      "entropy": 5.3324957370758055,
      "epoch": 3.6899533891166394,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003696129880502823,
      "loss": 4.9201,
      "mean_token_accuracy": 0.22916085571050643,
      "num_tokens": 73340523.0,
      "step": 32855
    },
    {
      "entropy": 5.461028671264648,
      "epoch": 3.690514966024597,
      "grad_norm": 1.046875,
      "learning_rate": 0.000369576559951393,
      "loss": 4.9853,
      "mean_token_accuracy": 0.22127919793128967,
      "num_tokens": 73352219.0,
      "step": 32860
    },
    {
      "entropy": 5.308103704452515,
      "epoch": 3.6910765429325547,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00036954012884114946,
      "loss": 4.9359,
      "mean_token_accuracy": 0.2188747525215149,
      "num_tokens": 73363157.0,
      "step": 32865
    },
    {
      "entropy": 5.291367864608764,
      "epoch": 3.691638119840512,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003695036947207115,
      "loss": 4.9845,
      "mean_token_accuracy": 0.22107960432767867,
      "num_tokens": 73375387.0,
      "step": 32870
    },
    {
      "entropy": 5.372172737121582,
      "epoch": 3.6921996967484696,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00036946725759123936,
      "loss": 5.0535,
      "mean_token_accuracy": 0.21334247589111327,
      "num_tokens": 73388056.0,
      "step": 32875
    },
    {
      "entropy": 5.3368572235107425,
      "epoch": 3.6927612736564273,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003694308174538929,
      "loss": 5.003,
      "mean_token_accuracy": 0.22400177419185638,
      "num_tokens": 73398510.0,
      "step": 32880
    },
    {
      "entropy": 5.3316443920135494,
      "epoch": 3.693322850564385,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036939437430983274,
      "loss": 4.9076,
      "mean_token_accuracy": 0.2221526935696602,
      "num_tokens": 73409382.0,
      "step": 32885
    },
    {
      "entropy": 5.273379230499268,
      "epoch": 3.6938844274723426,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036935792816021905,
      "loss": 4.9053,
      "mean_token_accuracy": 0.22176304906606675,
      "num_tokens": 73419958.0,
      "step": 32890
    },
    {
      "entropy": 5.311949729919434,
      "epoch": 3.6944460043803,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036932147900621234,
      "loss": 4.8798,
      "mean_token_accuracy": 0.22285351157188416,
      "num_tokens": 73430982.0,
      "step": 32895
    },
    {
      "entropy": 5.283746814727783,
      "epoch": 3.6950075812882575,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036928502684897314,
      "loss": 4.8862,
      "mean_token_accuracy": 0.222919362783432,
      "num_tokens": 73442375.0,
      "step": 32900
    },
    {
      "entropy": 5.200436925888061,
      "epoch": 3.695569158196215,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00036924857168966216,
      "loss": 4.7559,
      "mean_token_accuracy": 0.23450357764959334,
      "num_tokens": 73452212.0,
      "step": 32905
    },
    {
      "entropy": 5.213329696655274,
      "epoch": 3.6961307351041723,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003692121135294401,
      "loss": 4.9061,
      "mean_token_accuracy": 0.22431111633777617,
      "num_tokens": 73462998.0,
      "step": 32910
    },
    {
      "entropy": 5.207977437973023,
      "epoch": 3.69669231201213,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00036917565236946776,
      "loss": 4.8362,
      "mean_token_accuracy": 0.2272441253066063,
      "num_tokens": 73474387.0,
      "step": 32915
    },
    {
      "entropy": 5.295939779281616,
      "epoch": 3.6972538889200877,
      "grad_norm": 0.984375,
      "learning_rate": 0.00036913918821090616,
      "loss": 4.8568,
      "mean_token_accuracy": 0.22813885807991027,
      "num_tokens": 73484524.0,
      "step": 32920
    },
    {
      "entropy": 5.269787406921386,
      "epoch": 3.697815465828045,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036910272105491625,
      "loss": 4.8764,
      "mean_token_accuracy": 0.2270350992679596,
      "num_tokens": 73495286.0,
      "step": 32925
    },
    {
      "entropy": 5.235397529602051,
      "epoch": 3.6983770427360025,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00036906625090265925,
      "loss": 4.9051,
      "mean_token_accuracy": 0.2197237566113472,
      "num_tokens": 73507366.0,
      "step": 32930
    },
    {
      "entropy": 5.239141464233398,
      "epoch": 3.69893861964396,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036902977775529636,
      "loss": 4.8325,
      "mean_token_accuracy": 0.23357400596141814,
      "num_tokens": 73517823.0,
      "step": 32935
    },
    {
      "entropy": 5.244951915740967,
      "epoch": 3.699500196551918,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003689933016139888,
      "loss": 4.914,
      "mean_token_accuracy": 0.2201544538140297,
      "num_tokens": 73529199.0,
      "step": 32940
    },
    {
      "entropy": 5.226585817337036,
      "epoch": 3.7000617734598755,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003689568224798981,
      "loss": 4.811,
      "mean_token_accuracy": 0.22055019587278366,
      "num_tokens": 73539845.0,
      "step": 32945
    },
    {
      "entropy": 5.326054620742798,
      "epoch": 3.7006233503678327,
      "grad_norm": 1.0,
      "learning_rate": 0.0003689203403541857,
      "loss": 4.9414,
      "mean_token_accuracy": 0.21536635011434554,
      "num_tokens": 73551331.0,
      "step": 32950
    },
    {
      "entropy": 5.295011186599732,
      "epoch": 3.7011849272757904,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00036888385523801325,
      "loss": 4.9438,
      "mean_token_accuracy": 0.22400058060884476,
      "num_tokens": 73562155.0,
      "step": 32955
    },
    {
      "entropy": 5.2311015129089355,
      "epoch": 3.701746504183748,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00036884736713254247,
      "loss": 4.9064,
      "mean_token_accuracy": 0.22389867752790452,
      "num_tokens": 73573014.0,
      "step": 32960
    },
    {
      "entropy": 5.250722217559814,
      "epoch": 3.7023080810917053,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036881087603893515,
      "loss": 4.9278,
      "mean_token_accuracy": 0.2191484585404396,
      "num_tokens": 73584919.0,
      "step": 32965
    },
    {
      "entropy": 5.288492870330811,
      "epoch": 3.702869657999663,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00036877438195835307,
      "loss": 4.8618,
      "mean_token_accuracy": 0.2212742820382118,
      "num_tokens": 73596015.0,
      "step": 32970
    },
    {
      "entropy": 5.257446765899658,
      "epoch": 3.7034312349076206,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003687378848919584,
      "loss": 4.8411,
      "mean_token_accuracy": 0.2239953801035881,
      "num_tokens": 73606312.0,
      "step": 32975
    },
    {
      "entropy": 5.364084959030151,
      "epoch": 3.7039928118155783,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036870138484091297,
      "loss": 5.098,
      "mean_token_accuracy": 0.2092813804745674,
      "num_tokens": 73617411.0,
      "step": 32980
    },
    {
      "entropy": 5.348041677474976,
      "epoch": 3.704554388723536,
      "grad_norm": 0.90625,
      "learning_rate": 0.00036866488180637926,
      "loss": 4.8624,
      "mean_token_accuracy": 0.21859039962291718,
      "num_tokens": 73627725.0,
      "step": 32985
    },
    {
      "entropy": 5.2923903465271,
      "epoch": 3.705115965631493,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003686283757895193,
      "loss": 4.8122,
      "mean_token_accuracy": 0.22377502620220185,
      "num_tokens": 73639130.0,
      "step": 32990
    },
    {
      "entropy": 5.179399871826172,
      "epoch": 3.705677542539451,
      "grad_norm": 0.9375,
      "learning_rate": 0.00036859186679149555,
      "loss": 4.8426,
      "mean_token_accuracy": 0.22492436617612838,
      "num_tokens": 73650107.0,
      "step": 32995
    },
    {
      "entropy": 5.256532621383667,
      "epoch": 3.7062391194474085,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003685553548134705,
      "loss": 4.8427,
      "mean_token_accuracy": 0.22892729789018632,
      "num_tokens": 73659735.0,
      "step": 33000
    },
    {
      "epoch": 3.7062391194474085,
      "eval_entropy": 5.128730094276017,
      "eval_loss": 5.066525459289551,
      "eval_mean_token_accuracy": 0.2219385903233531,
      "eval_num_tokens": 73659735.0,
      "eval_runtime": 23.797,
      "eval_samples_per_second": 1303.231,
      "eval_steps_per_second": 162.92,
      "step": 33000
    },
    {
      "entropy": 5.26786208152771,
      "epoch": 3.7068006963553657,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003685188398566066,
      "loss": 4.8613,
      "mean_token_accuracy": 0.22030798345804214,
      "num_tokens": 73670969.0,
      "step": 33005
    },
    {
      "entropy": 5.299601888656616,
      "epoch": 3.7073622732633233,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003684823219220666,
      "loss": 4.9353,
      "mean_token_accuracy": 0.2323679357767105,
      "num_tokens": 73681513.0,
      "step": 33010
    },
    {
      "entropy": 5.3257513523101805,
      "epoch": 3.707923850171281,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003684458010110133,
      "loss": 5.0058,
      "mean_token_accuracy": 0.21803582459688187,
      "num_tokens": 73692485.0,
      "step": 33015
    },
    {
      "entropy": 5.332743120193482,
      "epoch": 3.7084854270792382,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003684092771246094,
      "loss": 4.9136,
      "mean_token_accuracy": 0.22334067523479462,
      "num_tokens": 73702582.0,
      "step": 33020
    },
    {
      "entropy": 5.2419921398162845,
      "epoch": 3.709047003987196,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003683727502640178,
      "loss": 4.9065,
      "mean_token_accuracy": 0.22646551728248596,
      "num_tokens": 73713337.0,
      "step": 33025
    },
    {
      "entropy": 5.376405906677246,
      "epoch": 3.7096085808951536,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003683362204304017,
      "loss": 5.0491,
      "mean_token_accuracy": 0.21796755492687225,
      "num_tokens": 73724500.0,
      "step": 33030
    },
    {
      "entropy": 5.242242383956909,
      "epoch": 3.710170157803111,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003682996876249241,
      "loss": 4.776,
      "mean_token_accuracy": 0.23185500800609588,
      "num_tokens": 73734690.0,
      "step": 33035
    },
    {
      "entropy": 5.199083566665649,
      "epoch": 3.710731734711069,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003682631518487483,
      "loss": 4.7536,
      "mean_token_accuracy": 0.2273904487490654,
      "num_tokens": 73744887.0,
      "step": 33040
    },
    {
      "entropy": 5.317247295379639,
      "epoch": 3.711293311619026,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003682266131030375,
      "loss": 4.9853,
      "mean_token_accuracy": 0.21645215898752213,
      "num_tokens": 73757141.0,
      "step": 33045
    },
    {
      "entropy": 5.27277135848999,
      "epoch": 3.7118548885269838,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00036819007138895525,
      "loss": 4.9186,
      "mean_token_accuracy": 0.2289204105734825,
      "num_tokens": 73767802.0,
      "step": 33050
    },
    {
      "entropy": 5.27337532043457,
      "epoch": 3.7124164654349414,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000368153526707665,
      "loss": 4.9255,
      "mean_token_accuracy": 0.21750029176473618,
      "num_tokens": 73778278.0,
      "step": 33055
    },
    {
      "entropy": 5.223908948898315,
      "epoch": 3.7129780423428986,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003681169790603303,
      "loss": 4.8387,
      "mean_token_accuracy": 0.2222057119011879,
      "num_tokens": 73790065.0,
      "step": 33060
    },
    {
      "entropy": 5.293649578094483,
      "epoch": 3.7135396192508563,
      "grad_norm": 0.984375,
      "learning_rate": 0.00036808042844811484,
      "loss": 4.8953,
      "mean_token_accuracy": 0.23105957061052323,
      "num_tokens": 73800868.0,
      "step": 33065
    },
    {
      "entropy": 5.294353675842285,
      "epoch": 3.714101196158814,
      "grad_norm": 0.984375,
      "learning_rate": 0.00036804387487218254,
      "loss": 4.8221,
      "mean_token_accuracy": 0.22916160076856612,
      "num_tokens": 73811478.0,
      "step": 33070
    },
    {
      "entropy": 5.229940366744995,
      "epoch": 3.7146627730667716,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000368007318333697,
      "loss": 4.9196,
      "mean_token_accuracy": 0.2305711328983307,
      "num_tokens": 73823618.0,
      "step": 33075
    },
    {
      "entropy": 5.257183361053467,
      "epoch": 3.7152243499747293,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003679707588338225,
      "loss": 4.8919,
      "mean_token_accuracy": 0.21671504080295562,
      "num_tokens": 73835268.0,
      "step": 33080
    },
    {
      "entropy": 5.300042915344238,
      "epoch": 3.7157859268826865,
      "grad_norm": 0.984375,
      "learning_rate": 0.000367934196373723,
      "loss": 4.8296,
      "mean_token_accuracy": 0.2261858120560646,
      "num_tokens": 73845877.0,
      "step": 33085
    },
    {
      "entropy": 5.30859899520874,
      "epoch": 3.716347503790644,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036789763095456264,
      "loss": 4.9137,
      "mean_token_accuracy": 0.2214436113834381,
      "num_tokens": 73857038.0,
      "step": 33090
    },
    {
      "entropy": 5.281433629989624,
      "epoch": 3.716909080698602,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003678610625775056,
      "loss": 4.9276,
      "mean_token_accuracy": 0.22239402532577515,
      "num_tokens": 73868439.0,
      "step": 33095
    },
    {
      "entropy": 5.302278470993042,
      "epoch": 3.717470657606559,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036782449124371643,
      "loss": 4.9159,
      "mean_token_accuracy": 0.22101748436689378,
      "num_tokens": 73879427.0,
      "step": 33100
    },
    {
      "entropy": 5.298070335388184,
      "epoch": 3.7180322345145167,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036778791695435935,
      "loss": 4.8863,
      "mean_token_accuracy": 0.2266685903072357,
      "num_tokens": 73890369.0,
      "step": 33105
    },
    {
      "entropy": 5.301187419891358,
      "epoch": 3.7185938114224744,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036775133971059904,
      "loss": 4.9679,
      "mean_token_accuracy": 0.2181553825736046,
      "num_tokens": 73902524.0,
      "step": 33110
    },
    {
      "entropy": 5.216589689254761,
      "epoch": 3.7191553883304316,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003677147595136001,
      "loss": 4.8137,
      "mean_token_accuracy": 0.23258864730596543,
      "num_tokens": 73912163.0,
      "step": 33115
    },
    {
      "entropy": 5.217120742797851,
      "epoch": 3.7197169652383892,
      "grad_norm": 1.0,
      "learning_rate": 0.00036767817636452723,
      "loss": 4.8378,
      "mean_token_accuracy": 0.22506798505783082,
      "num_tokens": 73923238.0,
      "step": 33120
    },
    {
      "entropy": 5.279687070846558,
      "epoch": 3.720278542146347,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00036764159026454534,
      "loss": 4.9228,
      "mean_token_accuracy": 0.22710350453853606,
      "num_tokens": 73933833.0,
      "step": 33125
    },
    {
      "entropy": 5.339019918441773,
      "epoch": 3.7208401190543046,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003676050012148193,
      "loss": 4.9527,
      "mean_token_accuracy": 0.21923412680625914,
      "num_tokens": 73944850.0,
      "step": 33130
    },
    {
      "entropy": 5.325947570800781,
      "epoch": 3.7214016959622622,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000367568409216514,
      "loss": 4.9657,
      "mean_token_accuracy": 0.2180815815925598,
      "num_tokens": 73954490.0,
      "step": 33135
    },
    {
      "entropy": 5.326819610595703,
      "epoch": 3.7219632728702194,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003675318142707947,
      "loss": 4.9453,
      "mean_token_accuracy": 0.2181501492857933,
      "num_tokens": 73966250.0,
      "step": 33140
    },
    {
      "entropy": 5.286402654647827,
      "epoch": 3.722524849778177,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003674952163788266,
      "loss": 4.8098,
      "mean_token_accuracy": 0.22882833778858186,
      "num_tokens": 73976931.0,
      "step": 33145
    },
    {
      "entropy": 5.2614470481872555,
      "epoch": 3.7230864266861348,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00036745861554177486,
      "loss": 4.9789,
      "mean_token_accuracy": 0.2210617706179619,
      "num_tokens": 73989813.0,
      "step": 33150
    },
    {
      "entropy": 5.168765830993652,
      "epoch": 3.723648003594092,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036742201176080496,
      "loss": 4.771,
      "mean_token_accuracy": 0.23269301354885102,
      "num_tokens": 74000961.0,
      "step": 33155
    },
    {
      "entropy": 5.271940755844116,
      "epoch": 3.7242095805020496,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003673854050370824,
      "loss": 4.8647,
      "mean_token_accuracy": 0.22514603286981583,
      "num_tokens": 74011717.0,
      "step": 33160
    },
    {
      "entropy": 5.232967138290405,
      "epoch": 3.7247711574100073,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036734879537177276,
      "loss": 4.8441,
      "mean_token_accuracy": 0.22634206414222718,
      "num_tokens": 74022931.0,
      "step": 33165
    },
    {
      "entropy": 5.183789253234863,
      "epoch": 3.725332734317965,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003673121827660415,
      "loss": 4.7751,
      "mean_token_accuracy": 0.22672684341669083,
      "num_tokens": 74033110.0,
      "step": 33170
    },
    {
      "entropy": 5.238201570510864,
      "epoch": 3.7258943112259226,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00036727556722105465,
      "loss": 4.8241,
      "mean_token_accuracy": 0.22593133449554442,
      "num_tokens": 74043694.0,
      "step": 33175
    },
    {
      "entropy": 5.30288143157959,
      "epoch": 3.72645588813388,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00036723894873797786,
      "loss": 4.8162,
      "mean_token_accuracy": 0.2327468991279602,
      "num_tokens": 74053997.0,
      "step": 33180
    },
    {
      "entropy": 5.281660032272339,
      "epoch": 3.7270174650418375,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036720232731797726,
      "loss": 4.9753,
      "mean_token_accuracy": 0.2163172945380211,
      "num_tokens": 74065562.0,
      "step": 33185
    },
    {
      "entropy": 5.365810108184815,
      "epoch": 3.727579041949795,
      "grad_norm": 0.953125,
      "learning_rate": 0.00036716570296221873,
      "loss": 4.9973,
      "mean_token_accuracy": 0.2195315733551979,
      "num_tokens": 74075988.0,
      "step": 33190
    },
    {
      "entropy": 5.269671154022217,
      "epoch": 3.7281406188577524,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00036712907567186857,
      "loss": 4.8749,
      "mean_token_accuracy": 0.22592364996671677,
      "num_tokens": 74087644.0,
      "step": 33195
    },
    {
      "entropy": 5.3066481590271,
      "epoch": 3.72870219576571,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003670924454480928,
      "loss": 4.9435,
      "mean_token_accuracy": 0.22544198334217072,
      "num_tokens": 74097841.0,
      "step": 33200
    },
    {
      "entropy": 5.2374402523040775,
      "epoch": 3.7292637726736677,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003670558122920579,
      "loss": 4.8855,
      "mean_token_accuracy": 0.22174183279275894,
      "num_tokens": 74108164.0,
      "step": 33205
    },
    {
      "entropy": 5.28228988647461,
      "epoch": 3.729825349581625,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00036701917620493026,
      "loss": 4.9085,
      "mean_token_accuracy": 0.22316069155931473,
      "num_tokens": 74120456.0,
      "step": 33210
    },
    {
      "entropy": 5.317462921142578,
      "epoch": 3.7303869264895826,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003669825371878764,
      "loss": 4.9953,
      "mean_token_accuracy": 0.22098135501146315,
      "num_tokens": 74131023.0,
      "step": 33215
    },
    {
      "entropy": 5.32909140586853,
      "epoch": 3.7309485033975403,
      "grad_norm": 1.125,
      "learning_rate": 0.00036694589524206275,
      "loss": 4.9788,
      "mean_token_accuracy": 0.21976243555545807,
      "num_tokens": 74141654.0,
      "step": 33220
    },
    {
      "entropy": 5.251143741607666,
      "epoch": 3.731510080305498,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003669092503686562,
      "loss": 4.8763,
      "mean_token_accuracy": 0.2230052575469017,
      "num_tokens": 74152972.0,
      "step": 33225
    },
    {
      "entropy": 5.311899995803833,
      "epoch": 3.7320716572134556,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036687260256882347,
      "loss": 4.9004,
      "mean_token_accuracy": 0.21953137516975402,
      "num_tokens": 74164725.0,
      "step": 33230
    },
    {
      "entropy": 5.201064825057983,
      "epoch": 3.732633234121413,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00036683595184373147,
      "loss": 4.7896,
      "mean_token_accuracy": 0.23354719430208207,
      "num_tokens": 74175356.0,
      "step": 33235
    },
    {
      "entropy": 5.182060432434082,
      "epoch": 3.7331948110293705,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003667992981945471,
      "loss": 4.804,
      "mean_token_accuracy": 0.22659308910369874,
      "num_tokens": 74186418.0,
      "step": 33240
    },
    {
      "entropy": 5.1985259532928465,
      "epoch": 3.733756387937328,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00036676264162243744,
      "loss": 4.8576,
      "mean_token_accuracy": 0.21978849619627,
      "num_tokens": 74199236.0,
      "step": 33245
    },
    {
      "entropy": 5.249340200424195,
      "epoch": 3.7343179648452853,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003667259821285698,
      "loss": 4.8401,
      "mean_token_accuracy": 0.23124587237834932,
      "num_tokens": 74211475.0,
      "step": 33250
    },
    {
      "entropy": 5.230448246002197,
      "epoch": 3.734879541753243,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003666893197141112,
      "loss": 4.7989,
      "mean_token_accuracy": 0.234292696416378,
      "num_tokens": 74223909.0,
      "step": 33255
    },
    {
      "entropy": 5.365285396575928,
      "epoch": 3.7354411186612007,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003666526543802291,
      "loss": 4.9887,
      "mean_token_accuracy": 0.2154576987028122,
      "num_tokens": 74235060.0,
      "step": 33260
    },
    {
      "entropy": 5.254886960983276,
      "epoch": 3.7360026955691583,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000366615986128091,
      "loss": 4.8408,
      "mean_token_accuracy": 0.22340618222951888,
      "num_tokens": 74245930.0,
      "step": 33265
    },
    {
      "entropy": 5.279229688644409,
      "epoch": 3.736564272477116,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003665793149588644,
      "loss": 4.9355,
      "mean_token_accuracy": 0.2158684477210045,
      "num_tokens": 74257337.0,
      "step": 33270
    },
    {
      "entropy": 5.219858551025391,
      "epoch": 3.737125849385073,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00036654264087371686,
      "loss": 4.8165,
      "mean_token_accuracy": 0.2319710671901703,
      "num_tokens": 74268217.0,
      "step": 33275
    },
    {
      "entropy": 5.259803628921508,
      "epoch": 3.737687426293031,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00036650596387381615,
      "loss": 4.912,
      "mean_token_accuracy": 0.21641715914011,
      "num_tokens": 74279842.0,
      "step": 33280
    },
    {
      "entropy": 5.284705924987793,
      "epoch": 3.7382490032009885,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00036646928396033,
      "loss": 4.925,
      "mean_token_accuracy": 0.21905587464571,
      "num_tokens": 74292328.0,
      "step": 33285
    },
    {
      "entropy": 5.346007823944092,
      "epoch": 3.7388105801089457,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00036643260113442634,
      "loss": 4.9313,
      "mean_token_accuracy": 0.22348889708518982,
      "num_tokens": 74302706.0,
      "step": 33290
    },
    {
      "entropy": 5.409532737731934,
      "epoch": 3.7393721570169034,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003663959153972732,
      "loss": 4.9904,
      "mean_token_accuracy": 0.21279321014881133,
      "num_tokens": 74314320.0,
      "step": 33295
    },
    {
      "entropy": 5.203146266937256,
      "epoch": 3.739933733924861,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003663592267500388,
      "loss": 4.7622,
      "mean_token_accuracy": 0.23334677815437316,
      "num_tokens": 74325324.0,
      "step": 33300
    },
    {
      "entropy": 5.2623003959655765,
      "epoch": 3.7404953108328187,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003663225351938911,
      "loss": 4.9661,
      "mean_token_accuracy": 0.22221101075410843,
      "num_tokens": 74337623.0,
      "step": 33305
    },
    {
      "entropy": 5.241576147079468,
      "epoch": 3.741056887740776,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003662858407299984,
      "loss": 4.797,
      "mean_token_accuracy": 0.22457110732793809,
      "num_tokens": 74348971.0,
      "step": 33310
    },
    {
      "entropy": 5.256911706924439,
      "epoch": 3.7416184646487336,
      "grad_norm": 0.921875,
      "learning_rate": 0.00036624914335952924,
      "loss": 4.9258,
      "mean_token_accuracy": 0.21963575929403306,
      "num_tokens": 74360162.0,
      "step": 33315
    },
    {
      "entropy": 5.3895487785339355,
      "epoch": 3.7421800415566913,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036621244308365195,
      "loss": 5.0201,
      "mean_token_accuracy": 0.22032448202371596,
      "num_tokens": 74371180.0,
      "step": 33320
    },
    {
      "entropy": 5.266159009933472,
      "epoch": 3.742741618464649,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00036617573990353503,
      "loss": 4.8617,
      "mean_token_accuracy": 0.22600747495889664,
      "num_tokens": 74382854.0,
      "step": 33325
    },
    {
      "entropy": 5.29011082649231,
      "epoch": 3.743303195372606,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003661390338203472,
      "loss": 5.0319,
      "mean_token_accuracy": 0.20761062055826188,
      "num_tokens": 74394266.0,
      "step": 33330
    },
    {
      "entropy": 5.350609397888183,
      "epoch": 3.743864772280564,
      "grad_norm": 0.859375,
      "learning_rate": 0.0003661023248352572,
      "loss": 5.0202,
      "mean_token_accuracy": 0.21384213417768477,
      "num_tokens": 74406275.0,
      "step": 33335
    },
    {
      "entropy": 5.288472461700439,
      "epoch": 3.7444263491885215,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003660656129494339,
      "loss": 4.8464,
      "mean_token_accuracy": 0.22638914734125137,
      "num_tokens": 74417150.0,
      "step": 33340
    },
    {
      "entropy": 5.258701276779175,
      "epoch": 3.7449879260964787,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00036602889816404613,
      "loss": 4.8456,
      "mean_token_accuracy": 0.23260728865861893,
      "num_tokens": 74427997.0,
      "step": 33345
    },
    {
      "entropy": 5.318592596054077,
      "epoch": 3.7455495030044363,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00036599218048026294,
      "loss": 4.9879,
      "mean_token_accuracy": 0.22024301439523697,
      "num_tokens": 74439278.0,
      "step": 33350
    },
    {
      "entropy": 5.256107378005981,
      "epoch": 3.746111079912394,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00036595545989925345,
      "loss": 4.9326,
      "mean_token_accuracy": 0.22337716817855835,
      "num_tokens": 74450372.0,
      "step": 33355
    },
    {
      "entropy": 5.298558282852173,
      "epoch": 3.7466726568203517,
      "grad_norm": 1.0,
      "learning_rate": 0.00036591873642218676,
      "loss": 4.86,
      "mean_token_accuracy": 0.228847374022007,
      "num_tokens": 74460040.0,
      "step": 33360
    },
    {
      "entropy": 5.337679195404053,
      "epoch": 3.7472342337283093,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036588201005023223,
      "loss": 4.9281,
      "mean_token_accuracy": 0.22424606680870057,
      "num_tokens": 74472291.0,
      "step": 33365
    },
    {
      "entropy": 5.366052150726318,
      "epoch": 3.7477958106362665,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003658452807845593,
      "loss": 4.9106,
      "mean_token_accuracy": 0.22537636756896973,
      "num_tokens": 74482373.0,
      "step": 33370
    },
    {
      "entropy": 5.232218217849732,
      "epoch": 3.748357387544224,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00036580854862633747,
      "loss": 4.8936,
      "mean_token_accuracy": 0.23033847361803056,
      "num_tokens": 74494240.0,
      "step": 33375
    },
    {
      "entropy": 5.268816184997559,
      "epoch": 3.748918964452182,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003657718135767362,
      "loss": 4.9405,
      "mean_token_accuracy": 0.2223765507340431,
      "num_tokens": 74504505.0,
      "step": 33380
    },
    {
      "entropy": 5.302392911911011,
      "epoch": 3.749480541360139,
      "grad_norm": 1.125,
      "learning_rate": 0.0003657350756369251,
      "loss": 4.8375,
      "mean_token_accuracy": 0.23153853714466094,
      "num_tokens": 74515467.0,
      "step": 33385
    },
    {
      "entropy": 5.267843866348267,
      "epoch": 3.7500421182680967,
      "grad_norm": 1.0,
      "learning_rate": 0.00036569833480807403,
      "loss": 4.8731,
      "mean_token_accuracy": 0.22694872915744782,
      "num_tokens": 74527502.0,
      "step": 33390
    },
    {
      "entropy": 5.288361597061157,
      "epoch": 3.7506036951760544,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003656615910913528,
      "loss": 4.9301,
      "mean_token_accuracy": 0.22041190266609192,
      "num_tokens": 74539491.0,
      "step": 33395
    },
    {
      "entropy": 5.3131101608276365,
      "epoch": 3.751165272084012,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00036562484448793135,
      "loss": 4.9261,
      "mean_token_accuracy": 0.2158409222960472,
      "num_tokens": 74550847.0,
      "step": 33400
    },
    {
      "entropy": 5.324523973464966,
      "epoch": 3.7517268489919693,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003655880949989798,
      "loss": 4.8974,
      "mean_token_accuracy": 0.22026911675930022,
      "num_tokens": 74561180.0,
      "step": 33405
    },
    {
      "entropy": 5.3482279777526855,
      "epoch": 3.752288425899927,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036555134262566806,
      "loss": 4.959,
      "mean_token_accuracy": 0.22119443714618683,
      "num_tokens": 74571557.0,
      "step": 33410
    },
    {
      "entropy": 5.2686073780059814,
      "epoch": 3.7528500028078846,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003655145873691665,
      "loss": 4.9383,
      "mean_token_accuracy": 0.21512483209371566,
      "num_tokens": 74582614.0,
      "step": 33415
    },
    {
      "entropy": 5.247936010360718,
      "epoch": 3.7534115797158423,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036547782923064537,
      "loss": 4.8995,
      "mean_token_accuracy": 0.2230631023645401,
      "num_tokens": 74593176.0,
      "step": 33420
    },
    {
      "entropy": 5.286539793014526,
      "epoch": 3.7539731566237995,
      "grad_norm": 0.921875,
      "learning_rate": 0.00036544106821127505,
      "loss": 4.7981,
      "mean_token_accuracy": 0.23139649778604507,
      "num_tokens": 74604601.0,
      "step": 33425
    },
    {
      "entropy": 5.340945196151734,
      "epoch": 3.754534733531757,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000365404304312226,
      "loss": 4.9925,
      "mean_token_accuracy": 0.21863230913877488,
      "num_tokens": 74615591.0,
      "step": 33430
    },
    {
      "entropy": 5.261146497726441,
      "epoch": 3.755096310439715,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00036536753753466897,
      "loss": 4.8247,
      "mean_token_accuracy": 0.2259350061416626,
      "num_tokens": 74626905.0,
      "step": 33435
    },
    {
      "entropy": 5.290807008743286,
      "epoch": 3.755657887347672,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003653307678797744,
      "loss": 4.9296,
      "mean_token_accuracy": 0.21771447509527206,
      "num_tokens": 74638201.0,
      "step": 33440
    },
    {
      "entropy": 5.27585072517395,
      "epoch": 3.7562194642556297,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003652939953487131,
      "loss": 4.9738,
      "mean_token_accuracy": 0.2194276347756386,
      "num_tokens": 74649121.0,
      "step": 33445
    },
    {
      "entropy": 5.255408143997192,
      "epoch": 3.7567810411635874,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036525721994265606,
      "loss": 4.8014,
      "mean_token_accuracy": 0.22731202691793442,
      "num_tokens": 74659482.0,
      "step": 33450
    },
    {
      "entropy": 5.197954273223877,
      "epoch": 3.757342618071545,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003652204416627741,
      "loss": 4.8111,
      "mean_token_accuracy": 0.22821059823036194,
      "num_tokens": 74671587.0,
      "step": 33455
    },
    {
      "entropy": 5.298653173446655,
      "epoch": 3.7579041949795027,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036518366051023824,
      "loss": 4.8928,
      "mean_token_accuracy": 0.22390350252389907,
      "num_tokens": 74682753.0,
      "step": 33460
    },
    {
      "entropy": 5.260490751266479,
      "epoch": 3.75846577188746,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003651468764862197,
      "loss": 4.8507,
      "mean_token_accuracy": 0.2229623168706894,
      "num_tokens": 74692670.0,
      "step": 33465
    },
    {
      "entropy": 5.239466714859009,
      "epoch": 3.7590273487954176,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036511008959188966,
      "loss": 4.96,
      "mean_token_accuracy": 0.21689655780792236,
      "num_tokens": 74703832.0,
      "step": 33470
    },
    {
      "entropy": 5.2632135391235355,
      "epoch": 3.759588925703375,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00036507329982841935,
      "loss": 4.8213,
      "mean_token_accuracy": 0.23007698208093644,
      "num_tokens": 74716577.0,
      "step": 33475
    },
    {
      "entropy": 5.3329144477844235,
      "epoch": 3.7601505026113324,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003650365071969803,
      "loss": 4.9479,
      "mean_token_accuracy": 0.21957100480794906,
      "num_tokens": 74727458.0,
      "step": 33480
    },
    {
      "entropy": 5.223479986190796,
      "epoch": 3.76071207951929,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003649997116987439,
      "loss": 4.7857,
      "mean_token_accuracy": 0.2319221943616867,
      "num_tokens": 74739049.0,
      "step": 33485
    },
    {
      "entropy": 5.273102140426635,
      "epoch": 3.7612736564272478,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003649629133348818,
      "loss": 4.9407,
      "mean_token_accuracy": 0.2292320802807808,
      "num_tokens": 74749509.0,
      "step": 33490
    },
    {
      "entropy": 5.279347229003906,
      "epoch": 3.7618352333352054,
      "grad_norm": 0.90625,
      "learning_rate": 0.00036492611210656574,
      "loss": 4.8888,
      "mean_token_accuracy": 0.22775188088417053,
      "num_tokens": 74762067.0,
      "step": 33495
    },
    {
      "entropy": 5.305961990356446,
      "epoch": 3.7623968102431626,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003648893080149673,
      "loss": 4.911,
      "mean_token_accuracy": 0.23227352052927017,
      "num_tokens": 74773106.0,
      "step": 33500
    },
    {
      "entropy": 5.310882425308227,
      "epoch": 3.7629583871511203,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00036485250106125844,
      "loss": 4.98,
      "mean_token_accuracy": 0.21960017830133438,
      "num_tokens": 74785464.0,
      "step": 33505
    },
    {
      "entropy": 5.25408902168274,
      "epoch": 3.763519964059078,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000364815691246611,
      "loss": 4.8587,
      "mean_token_accuracy": 0.2288272961974144,
      "num_tokens": 74795939.0,
      "step": 33510
    },
    {
      "entropy": 5.269687461853027,
      "epoch": 3.7640815409670356,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036477887857219726,
      "loss": 4.8752,
      "mean_token_accuracy": 0.22884688526391983,
      "num_tokens": 74807036.0,
      "step": 33515
    },
    {
      "entropy": 5.297936582565308,
      "epoch": 3.764643117874993,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036474206303918914,
      "loss": 4.9259,
      "mean_token_accuracy": 0.2228242889046669,
      "num_tokens": 74818402.0,
      "step": 33520
    },
    {
      "entropy": 5.230245780944824,
      "epoch": 3.7652046947829505,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000364705244648759,
      "loss": 4.8851,
      "mean_token_accuracy": 0.2205903097987175,
      "num_tokens": 74830246.0,
      "step": 33525
    },
    {
      "entropy": 5.211356258392334,
      "epoch": 3.765766271690908,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003646684234020789,
      "loss": 4.846,
      "mean_token_accuracy": 0.22502476274967192,
      "num_tokens": 74840690.0,
      "step": 33530
    },
    {
      "entropy": 5.326031827926636,
      "epoch": 3.7663278485988654,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036463159930032163,
      "loss": 4.8729,
      "mean_token_accuracy": 0.2353584051132202,
      "num_tokens": 74851807.0,
      "step": 33535
    },
    {
      "entropy": 5.2571838855743405,
      "epoch": 3.766889425506823,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00036459477234465935,
      "loss": 4.8894,
      "mean_token_accuracy": 0.22187091857194902,
      "num_tokens": 74863028.0,
      "step": 33540
    },
    {
      "entropy": 5.2172871112823485,
      "epoch": 3.7674510024147807,
      "grad_norm": 0.890625,
      "learning_rate": 0.00036455794253626486,
      "loss": 4.8497,
      "mean_token_accuracy": 0.22832049280405045,
      "num_tokens": 74874674.0,
      "step": 33545
    },
    {
      "entropy": 5.31328763961792,
      "epoch": 3.7680125793227384,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036452110987631066,
      "loss": 4.9398,
      "mean_token_accuracy": 0.2195666566491127,
      "num_tokens": 74885791.0,
      "step": 33550
    },
    {
      "entropy": 5.296212911605835,
      "epoch": 3.768574156230696,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036448427436596966,
      "loss": 4.8921,
      "mean_token_accuracy": 0.227281790971756,
      "num_tokens": 74897422.0,
      "step": 33555
    },
    {
      "entropy": 5.316484594345093,
      "epoch": 3.7691357331386532,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003644474360064146,
      "loss": 4.8854,
      "mean_token_accuracy": 0.21629224568605424,
      "num_tokens": 74908680.0,
      "step": 33560
    },
    {
      "entropy": 5.294777822494507,
      "epoch": 3.769697310046611,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003644105947988185,
      "loss": 4.8806,
      "mean_token_accuracy": 0.22728199064731597,
      "num_tokens": 74919573.0,
      "step": 33565
    },
    {
      "entropy": 5.253656387329102,
      "epoch": 3.7702588869545686,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00036437375074435446,
      "loss": 4.8461,
      "mean_token_accuracy": 0.22906003445386885,
      "num_tokens": 74930507.0,
      "step": 33570
    },
    {
      "entropy": 5.251479721069336,
      "epoch": 3.770820463862526,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003643369038441955,
      "loss": 4.8418,
      "mean_token_accuracy": 0.2254742681980133,
      "num_tokens": 74942436.0,
      "step": 33575
    },
    {
      "entropy": 5.164855480194092,
      "epoch": 3.7713820407704834,
      "grad_norm": 1.0,
      "learning_rate": 0.00036430005409951476,
      "loss": 4.7687,
      "mean_token_accuracy": 0.23495590984821318,
      "num_tokens": 74953717.0,
      "step": 33580
    },
    {
      "entropy": 5.266802787780762,
      "epoch": 3.771943617678441,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036426320151148575,
      "loss": 4.9092,
      "mean_token_accuracy": 0.22720496654510497,
      "num_tokens": 74963878.0,
      "step": 33585
    },
    {
      "entropy": 5.285261535644532,
      "epoch": 3.7725051945863988,
      "grad_norm": 0.953125,
      "learning_rate": 0.00036422634608128184,
      "loss": 4.8257,
      "mean_token_accuracy": 0.23669052869081497,
      "num_tokens": 74976239.0,
      "step": 33590
    },
    {
      "entropy": 5.250201988220215,
      "epoch": 3.773066771494356,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036418948781007636,
      "loss": 4.929,
      "mean_token_accuracy": 0.2213844522833824,
      "num_tokens": 74987694.0,
      "step": 33595
    },
    {
      "entropy": 5.302253723144531,
      "epoch": 3.7736283484023136,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036415262669904305,
      "loss": 4.8605,
      "mean_token_accuracy": 0.21837455183267593,
      "num_tokens": 74998514.0,
      "step": 33600
    },
    {
      "entropy": 5.317055177688599,
      "epoch": 3.7741899253102713,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00036411576274935546,
      "loss": 4.8901,
      "mean_token_accuracy": 0.22609002739191056,
      "num_tokens": 75010438.0,
      "step": 33605
    },
    {
      "entropy": 5.223860168457032,
      "epoch": 3.774751502218229,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00036407889596218757,
      "loss": 4.8114,
      "mean_token_accuracy": 0.22479359358549117,
      "num_tokens": 75021280.0,
      "step": 33610
    },
    {
      "entropy": 5.188869142532349,
      "epoch": 3.775313079126186,
      "grad_norm": 0.984375,
      "learning_rate": 0.000364042026338713,
      "loss": 4.8242,
      "mean_token_accuracy": 0.2281291589140892,
      "num_tokens": 75032746.0,
      "step": 33615
    },
    {
      "entropy": 5.343181228637695,
      "epoch": 3.775874656034144,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00036400515388010574,
      "loss": 4.9912,
      "mean_token_accuracy": 0.21860844194889067,
      "num_tokens": 75044227.0,
      "step": 33620
    },
    {
      "entropy": 5.364304876327514,
      "epoch": 3.7764362329421015,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003639682785875399,
      "loss": 5.0121,
      "mean_token_accuracy": 0.21502550542354584,
      "num_tokens": 75053974.0,
      "step": 33625
    },
    {
      "entropy": 5.269173669815063,
      "epoch": 3.7769978098500587,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00036393140046218957,
      "loss": 4.8392,
      "mean_token_accuracy": 0.23223899453878402,
      "num_tokens": 75064539.0,
      "step": 33630
    },
    {
      "entropy": 5.2665627002716064,
      "epoch": 3.7775593867580164,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00036389451950522895,
      "loss": 4.8489,
      "mean_token_accuracy": 0.2316355288028717,
      "num_tokens": 75075229.0,
      "step": 33635
    },
    {
      "entropy": 5.325377416610718,
      "epoch": 3.778120963665974,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036385763571783236,
      "loss": 4.9444,
      "mean_token_accuracy": 0.22025225907564164,
      "num_tokens": 75085347.0,
      "step": 33640
    },
    {
      "entropy": 5.288340377807617,
      "epoch": 3.7786825405739317,
      "grad_norm": 1.0,
      "learning_rate": 0.0003638207491011742,
      "loss": 4.8888,
      "mean_token_accuracy": 0.22060669660568238,
      "num_tokens": 75096141.0,
      "step": 33645
    },
    {
      "entropy": 5.266131734848022,
      "epoch": 3.7792441174818894,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003637838596564289,
      "loss": 4.8488,
      "mean_token_accuracy": 0.2327826976776123,
      "num_tokens": 75107749.0,
      "step": 33650
    },
    {
      "entropy": 5.28335280418396,
      "epoch": 3.7798056943898466,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036374696738477105,
      "loss": 4.9099,
      "mean_token_accuracy": 0.2260517045855522,
      "num_tokens": 75118644.0,
      "step": 33655
    },
    {
      "entropy": 5.265323972702026,
      "epoch": 3.7803672712978043,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003637100722873754,
      "loss": 4.9346,
      "mean_token_accuracy": 0.21776852309703826,
      "num_tokens": 75130534.0,
      "step": 33660
    },
    {
      "entropy": 5.346036100387574,
      "epoch": 3.780928848205762,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003636731743654166,
      "loss": 4.8537,
      "mean_token_accuracy": 0.22068895846605302,
      "num_tokens": 75142159.0,
      "step": 33665
    },
    {
      "entropy": 5.289687967300415,
      "epoch": 3.781490425113719,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036363627362006955,
      "loss": 4.8392,
      "mean_token_accuracy": 0.22775519490242005,
      "num_tokens": 75153603.0,
      "step": 33670
    },
    {
      "entropy": 5.157969284057617,
      "epoch": 3.782052002021677,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003635993700525092,
      "loss": 4.8355,
      "mean_token_accuracy": 0.22897187769412994,
      "num_tokens": 75164538.0,
      "step": 33675
    },
    {
      "entropy": 5.259122323989868,
      "epoch": 3.7826135789296345,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00036356246366391054,
      "loss": 4.9441,
      "mean_token_accuracy": 0.22508665174245834,
      "num_tokens": 75175030.0,
      "step": 33680
    },
    {
      "entropy": 5.371172332763672,
      "epoch": 3.783175155837592,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036352555445544864,
      "loss": 4.9484,
      "mean_token_accuracy": 0.22682326883077622,
      "num_tokens": 75187116.0,
      "step": 33685
    },
    {
      "entropy": 5.328619289398193,
      "epoch": 3.78373673274555,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036348864242829874,
      "loss": 4.8772,
      "mean_token_accuracy": 0.22937259674072266,
      "num_tokens": 75197247.0,
      "step": 33690
    },
    {
      "entropy": 5.231230592727661,
      "epoch": 3.784298309653507,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036345172758363623,
      "loss": 4.9089,
      "mean_token_accuracy": 0.22449299395084382,
      "num_tokens": 75208531.0,
      "step": 33695
    },
    {
      "entropy": 5.27730975151062,
      "epoch": 3.7848598865614647,
      "grad_norm": 1.0,
      "learning_rate": 0.0003634148099226364,
      "loss": 4.9685,
      "mean_token_accuracy": 0.2181454122066498,
      "num_tokens": 75221061.0,
      "step": 33700
    },
    {
      "entropy": 5.258248567581177,
      "epoch": 3.7854214634694223,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00036337788944647466,
      "loss": 4.866,
      "mean_token_accuracy": 0.2245375081896782,
      "num_tokens": 75232897.0,
      "step": 33705
    },
    {
      "entropy": 5.369842100143432,
      "epoch": 3.7859830403773795,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003633409661563267,
      "loss": 4.9456,
      "mean_token_accuracy": 0.22060571163892745,
      "num_tokens": 75245228.0,
      "step": 33710
    },
    {
      "entropy": 5.308136129379273,
      "epoch": 3.786544617285337,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036330404005336807,
      "loss": 4.8503,
      "mean_token_accuracy": 0.22041118443012236,
      "num_tokens": 75256486.0,
      "step": 33715
    },
    {
      "entropy": 5.229057359695434,
      "epoch": 3.787106194193295,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003632671111387746,
      "loss": 4.8572,
      "mean_token_accuracy": 0.23638930022716523,
      "num_tokens": 75266439.0,
      "step": 33720
    },
    {
      "entropy": 5.290452909469605,
      "epoch": 3.787667771101252,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036323017941372203,
      "loss": 4.8825,
      "mean_token_accuracy": 0.22910776883363723,
      "num_tokens": 75276550.0,
      "step": 33725
    },
    {
      "entropy": 5.219924545288086,
      "epoch": 3.7882293480092097,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003631932448793864,
      "loss": 4.892,
      "mean_token_accuracy": 0.22275592684745787,
      "num_tokens": 75287061.0,
      "step": 33730
    },
    {
      "entropy": 5.291351079940796,
      "epoch": 3.7887909249171674,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003631563075369436,
      "loss": 4.9107,
      "mean_token_accuracy": 0.21930774599313735,
      "num_tokens": 75298393.0,
      "step": 33735
    },
    {
      "entropy": 5.260597133636475,
      "epoch": 3.789352501825125,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036311936738756977,
      "loss": 4.8302,
      "mean_token_accuracy": 0.22135710567235947,
      "num_tokens": 75308153.0,
      "step": 33740
    },
    {
      "entropy": 5.21954755783081,
      "epoch": 3.7899140787330827,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003630824244324411,
      "loss": 4.7987,
      "mean_token_accuracy": 0.22892561852931975,
      "num_tokens": 75319383.0,
      "step": 33745
    },
    {
      "entropy": 5.299035739898682,
      "epoch": 3.79047565564104,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003630454786727339,
      "loss": 4.8859,
      "mean_token_accuracy": 0.21952502131462098,
      "num_tokens": 75329736.0,
      "step": 33750
    },
    {
      "entropy": 5.383345985412598,
      "epoch": 3.7910372325489976,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003630085301096245,
      "loss": 4.995,
      "mean_token_accuracy": 0.21337319016456605,
      "num_tokens": 75341451.0,
      "step": 33755
    },
    {
      "entropy": 5.173659706115723,
      "epoch": 3.7915988094569553,
      "grad_norm": 0.9375,
      "learning_rate": 0.00036297157874428933,
      "loss": 4.7674,
      "mean_token_accuracy": 0.22874380499124528,
      "num_tokens": 75352664.0,
      "step": 33760
    },
    {
      "entropy": 5.2816853523254395,
      "epoch": 3.7921603863649125,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00036293462457790505,
      "loss": 4.9493,
      "mean_token_accuracy": 0.2200862467288971,
      "num_tokens": 75362847.0,
      "step": 33765
    },
    {
      "entropy": 5.281961059570312,
      "epoch": 3.79272196327287,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00036289766761164814,
      "loss": 4.9359,
      "mean_token_accuracy": 0.22171373814344406,
      "num_tokens": 75373344.0,
      "step": 33770
    },
    {
      "entropy": 5.249218130111695,
      "epoch": 3.793283540180828,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036286070784669543,
      "loss": 4.8943,
      "mean_token_accuracy": 0.22182975262403487,
      "num_tokens": 75384262.0,
      "step": 33775
    },
    {
      "entropy": 5.328280687332153,
      "epoch": 3.7938451170887855,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036282374528422373,
      "loss": 4.9166,
      "mean_token_accuracy": 0.2263138085603714,
      "num_tokens": 75395817.0,
      "step": 33780
    },
    {
      "entropy": 5.228666734695435,
      "epoch": 3.794406693996743,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00036278677992540993,
      "loss": 4.8084,
      "mean_token_accuracy": 0.23650040626525878,
      "num_tokens": 75406452.0,
      "step": 33785
    },
    {
      "entropy": 5.249700546264648,
      "epoch": 3.7949682709047003,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000362749811771431,
      "loss": 4.9055,
      "mean_token_accuracy": 0.2209149107336998,
      "num_tokens": 75418638.0,
      "step": 33790
    },
    {
      "entropy": 5.320770120620727,
      "epoch": 3.795529847812658,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003627128408234641,
      "loss": 4.9106,
      "mean_token_accuracy": 0.22692644745111465,
      "num_tokens": 75430024.0,
      "step": 33795
    },
    {
      "entropy": 5.383403491973877,
      "epoch": 3.7960914247206157,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003626758670826861,
      "loss": 5.0425,
      "mean_token_accuracy": 0.22082798331975936,
      "num_tokens": 75442641.0,
      "step": 33800
    },
    {
      "entropy": 5.347500419616699,
      "epoch": 3.796653001628573,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003626388905502747,
      "loss": 4.99,
      "mean_token_accuracy": 0.22424981296062468,
      "num_tokens": 75453245.0,
      "step": 33805
    },
    {
      "entropy": 5.256898403167725,
      "epoch": 3.7972145785365305,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000362601911227407,
      "loss": 4.8652,
      "mean_token_accuracy": 0.22000292539596558,
      "num_tokens": 75464243.0,
      "step": 33810
    },
    {
      "entropy": 5.289534425735473,
      "epoch": 3.797776155444488,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003625649291152604,
      "loss": 4.8878,
      "mean_token_accuracy": 0.2272087812423706,
      "num_tokens": 75474348.0,
      "step": 33815
    },
    {
      "entropy": 5.287720584869385,
      "epoch": 3.7983377323524454,
      "grad_norm": 1.0,
      "learning_rate": 0.00036252794421501257,
      "loss": 4.965,
      "mean_token_accuracy": 0.2250383675098419,
      "num_tokens": 75486903.0,
      "step": 33820
    },
    {
      "entropy": 5.281700706481933,
      "epoch": 3.798899309260403,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00036249095652784096,
      "loss": 4.8978,
      "mean_token_accuracy": 0.22855658531188966,
      "num_tokens": 75498198.0,
      "step": 33825
    },
    {
      "entropy": 5.281529760360717,
      "epoch": 3.7994608861683608,
      "grad_norm": 0.90625,
      "learning_rate": 0.00036245396605492344,
      "loss": 4.9338,
      "mean_token_accuracy": 0.21971203684806823,
      "num_tokens": 75509787.0,
      "step": 33830
    },
    {
      "entropy": 5.2190694332122805,
      "epoch": 3.8000224630763184,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003624169727974376,
      "loss": 4.8058,
      "mean_token_accuracy": 0.2344295158982277,
      "num_tokens": 75521706.0,
      "step": 33835
    },
    {
      "entropy": 5.242628860473633,
      "epoch": 3.800584039984276,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036237997675656167,
      "loss": 4.914,
      "mean_token_accuracy": 0.229352143406868,
      "num_tokens": 75532432.0,
      "step": 33840
    },
    {
      "entropy": 5.380811309814453,
      "epoch": 3.8011456168922333,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003623429779334732,
      "loss": 4.998,
      "mean_token_accuracy": 0.21237182915210723,
      "num_tokens": 75543473.0,
      "step": 33845
    },
    {
      "entropy": 5.3075480461120605,
      "epoch": 3.801707193800191,
      "grad_norm": 0.9375,
      "learning_rate": 0.00036230597632935053,
      "loss": 4.8618,
      "mean_token_accuracy": 0.232765331864357,
      "num_tokens": 75554065.0,
      "step": 33850
    },
    {
      "entropy": 5.211465120315552,
      "epoch": 3.8022687707081486,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003622689719453717,
      "loss": 4.777,
      "mean_token_accuracy": 0.23214301466941833,
      "num_tokens": 75564823.0,
      "step": 33855
    },
    {
      "entropy": 5.276632833480835,
      "epoch": 3.802830347616106,
      "grad_norm": 0.921875,
      "learning_rate": 0.00036223196478271495,
      "loss": 4.9635,
      "mean_token_accuracy": 0.22186739146709442,
      "num_tokens": 75576062.0,
      "step": 33860
    },
    {
      "entropy": 5.325357437133789,
      "epoch": 3.8033919245240635,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003621949548425585,
      "loss": 4.842,
      "mean_token_accuracy": 0.22394041270017623,
      "num_tokens": 75587038.0,
      "step": 33865
    },
    {
      "entropy": 5.247396326065063,
      "epoch": 3.803953501432021,
      "grad_norm": 1.015625,
      "learning_rate": 0.000362157942126081,
      "loss": 4.887,
      "mean_token_accuracy": 0.21706050336360933,
      "num_tokens": 75598306.0,
      "step": 33870
    },
    {
      "entropy": 5.229903316497802,
      "epoch": 3.804515078339979,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003621209266344608,
      "loss": 4.9004,
      "mean_token_accuracy": 0.22626238018274308,
      "num_tokens": 75609432.0,
      "step": 33875
    },
    {
      "entropy": 5.171363258361817,
      "epoch": 3.8050766552479365,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003620839083688765,
      "loss": 4.8189,
      "mean_token_accuracy": 0.2320956215262413,
      "num_tokens": 75619936.0,
      "step": 33880
    },
    {
      "entropy": 5.301234149932862,
      "epoch": 3.8056382321558937,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003620468873305068,
      "loss": 4.8472,
      "mean_token_accuracy": 0.2314567670226097,
      "num_tokens": 75630271.0,
      "step": 33885
    },
    {
      "entropy": 5.2635640621185305,
      "epoch": 3.8061998090638514,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036200986352053036,
      "loss": 4.8178,
      "mean_token_accuracy": 0.22719140499830245,
      "num_tokens": 75639859.0,
      "step": 33890
    },
    {
      "entropy": 5.255832576751709,
      "epoch": 3.806761385971809,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00036197283694012615,
      "loss": 4.8626,
      "mean_token_accuracy": 0.22842437624931336,
      "num_tokens": 75650610.0,
      "step": 33895
    },
    {
      "entropy": 5.293430471420288,
      "epoch": 3.8073229628797662,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003619358075904732,
      "loss": 4.8656,
      "mean_token_accuracy": 0.22546726018190383,
      "num_tokens": 75661604.0,
      "step": 33900
    },
    {
      "entropy": 5.297801733016968,
      "epoch": 3.807884539787724,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00036189877547275023,
      "loss": 4.9281,
      "mean_token_accuracy": 0.22298236638307573,
      "num_tokens": 75674173.0,
      "step": 33905
    },
    {
      "entropy": 5.293553590774536,
      "epoch": 3.8084461166956816,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036186174058813666,
      "loss": 4.8638,
      "mean_token_accuracy": 0.22681059390306474,
      "num_tokens": 75684579.0,
      "step": 33910
    },
    {
      "entropy": 5.255897426605225,
      "epoch": 3.8090076936036388,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00036182470293781154,
      "loss": 4.874,
      "mean_token_accuracy": 0.23143054842948912,
      "num_tokens": 75696022.0,
      "step": 33915
    },
    {
      "entropy": 5.336514663696289,
      "epoch": 3.8095692705115964,
      "grad_norm": 1.0,
      "learning_rate": 0.0003617876625229543,
      "loss": 4.9344,
      "mean_token_accuracy": 0.2248812809586525,
      "num_tokens": 75708255.0,
      "step": 33920
    },
    {
      "entropy": 5.267062139511109,
      "epoch": 3.810130847419554,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003617506193447442,
      "loss": 4.8172,
      "mean_token_accuracy": 0.2294149562716484,
      "num_tokens": 75719069.0,
      "step": 33925
    },
    {
      "entropy": 5.197224950790405,
      "epoch": 3.8106924243275118,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003617135734043606,
      "loss": 4.8198,
      "mean_token_accuracy": 0.23909565061330795,
      "num_tokens": 75730078.0,
      "step": 33930
    },
    {
      "entropy": 5.231386327743531,
      "epoch": 3.8112540012354694,
      "grad_norm": 1.0,
      "learning_rate": 0.0003616765247029833,
      "loss": 4.8401,
      "mean_token_accuracy": 0.2263010174036026,
      "num_tokens": 75741910.0,
      "step": 33935
    },
    {
      "entropy": 5.227885007858276,
      "epoch": 3.8118155781434266,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00036163947324179186,
      "loss": 4.9016,
      "mean_token_accuracy": 0.22157650142908097,
      "num_tokens": 75752261.0,
      "step": 33940
    },
    {
      "entropy": 5.284749460220337,
      "epoch": 3.8123771550513843,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036160241902196587,
      "loss": 4.8708,
      "mean_token_accuracy": 0.2228846400976181,
      "num_tokens": 75762442.0,
      "step": 33945
    },
    {
      "entropy": 5.286042356491089,
      "epoch": 3.812938731959342,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036156536204468535,
      "loss": 4.8203,
      "mean_token_accuracy": 0.22754324972629547,
      "num_tokens": 75773550.0,
      "step": 33950
    },
    {
      "entropy": 5.389838027954101,
      "epoch": 3.813500308867299,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036152830231113013,
      "loss": 4.9772,
      "mean_token_accuracy": 0.21976683288812637,
      "num_tokens": 75785730.0,
      "step": 33955
    },
    {
      "entropy": 5.196620512008667,
      "epoch": 3.814061885775257,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003614912398224802,
      "loss": 4.7012,
      "mean_token_accuracy": 0.23577404916286468,
      "num_tokens": 75795868.0,
      "step": 33960
    },
    {
      "entropy": 5.254671812057495,
      "epoch": 3.8146234626832145,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003614541745799156,
      "loss": 4.9899,
      "mean_token_accuracy": 0.21765154898166655,
      "num_tokens": 75808318.0,
      "step": 33965
    },
    {
      "entropy": 5.2572235584259035,
      "epoch": 3.815185039591172,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003614171065846166,
      "loss": 4.9232,
      "mean_token_accuracy": 0.22312377542257308,
      "num_tokens": 75819930.0,
      "step": 33970
    },
    {
      "entropy": 5.32566204071045,
      "epoch": 3.81574661649913,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003613800358377635,
      "loss": 4.9628,
      "mean_token_accuracy": 0.22164054661989213,
      "num_tokens": 75831814.0,
      "step": 33975
    },
    {
      "entropy": 5.316137027740479,
      "epoch": 3.816308193407087,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003613429623405364,
      "loss": 4.8891,
      "mean_token_accuracy": 0.22124878466129302,
      "num_tokens": 75842610.0,
      "step": 33980
    },
    {
      "entropy": 5.269616508483887,
      "epoch": 3.8168697703150447,
      "grad_norm": 0.9375,
      "learning_rate": 0.00036130588609411597,
      "loss": 4.9589,
      "mean_token_accuracy": 0.22080897092819213,
      "num_tokens": 75854582.0,
      "step": 33985
    },
    {
      "entropy": 5.251255941390991,
      "epoch": 3.8174313472230024,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00036126880709968264,
      "loss": 4.8415,
      "mean_token_accuracy": 0.2331311136484146,
      "num_tokens": 75865715.0,
      "step": 33990
    },
    {
      "entropy": 5.299725770950317,
      "epoch": 3.8179929241309596,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00036123172535841705,
      "loss": 4.8673,
      "mean_token_accuracy": 0.22950404584407808,
      "num_tokens": 75876348.0,
      "step": 33995
    },
    {
      "entropy": 5.310259056091309,
      "epoch": 3.8185545010389172,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003611946408714998,
      "loss": 4.9676,
      "mean_token_accuracy": 0.22008429169654847,
      "num_tokens": 75888316.0,
      "step": 34000
    },
    {
      "entropy": 5.257753372192383,
      "epoch": 3.819116077946875,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036115755364011186,
      "loss": 4.8897,
      "mean_token_accuracy": 0.2244138553738594,
      "num_tokens": 75898872.0,
      "step": 34005
    },
    {
      "entropy": 5.282508897781372,
      "epoch": 3.819677654854832,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000361120463665434,
      "loss": 4.9442,
      "mean_token_accuracy": 0.2215681716799736,
      "num_tokens": 75909768.0,
      "step": 34010
    },
    {
      "entropy": 5.343292284011841,
      "epoch": 3.82023923176279,
      "grad_norm": 0.875,
      "learning_rate": 0.0003610833709486471,
      "loss": 4.9321,
      "mean_token_accuracy": 0.2189019203186035,
      "num_tokens": 75921488.0,
      "step": 34015
    },
    {
      "entropy": 5.270740699768067,
      "epoch": 3.8208008086707475,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003610462754909323,
      "loss": 4.8853,
      "mean_token_accuracy": 0.21938182413578033,
      "num_tokens": 75933619.0,
      "step": 34020
    },
    {
      "entropy": 5.2610557079315186,
      "epoch": 3.821362385578705,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036100917729347073,
      "loss": 4.8206,
      "mean_token_accuracy": 0.22360656559467315,
      "num_tokens": 75943369.0,
      "step": 34025
    },
    {
      "entropy": 5.303887271881104,
      "epoch": 3.8219239624866628,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036097207635744363,
      "loss": 4.9785,
      "mean_token_accuracy": 0.21679448038339616,
      "num_tokens": 75954793.0,
      "step": 34030
    },
    {
      "entropy": 5.217080974578858,
      "epoch": 3.82248553939462,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036093497268403215,
      "loss": 4.8298,
      "mean_token_accuracy": 0.2309556171298027,
      "num_tokens": 75965765.0,
      "step": 34035
    },
    {
      "entropy": 5.262321996688843,
      "epoch": 3.8230471163025777,
      "grad_norm": 1.078125,
      "learning_rate": 0.000360897866274418,
      "loss": 4.8758,
      "mean_token_accuracy": 0.2279304876923561,
      "num_tokens": 75976435.0,
      "step": 34040
    },
    {
      "entropy": 5.308135747909546,
      "epoch": 3.8236086932105353,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036086075712978233,
      "loss": 4.8731,
      "mean_token_accuracy": 0.22374450415372849,
      "num_tokens": 75987379.0,
      "step": 34045
    },
    {
      "entropy": 5.342043733596801,
      "epoch": 3.8241702701184925,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00036082364525130684,
      "loss": 5.0255,
      "mean_token_accuracy": 0.21261365115642547,
      "num_tokens": 75999073.0,
      "step": 34050
    },
    {
      "entropy": 5.3119536399841305,
      "epoch": 3.82473184702645,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003607865306401732,
      "loss": 4.8746,
      "mean_token_accuracy": 0.22233077734708787,
      "num_tokens": 76009490.0,
      "step": 34055
    },
    {
      "entropy": 5.262546253204346,
      "epoch": 3.825293423934408,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00036074941329756327,
      "loss": 4.9065,
      "mean_token_accuracy": 0.2251956045627594,
      "num_tokens": 76021616.0,
      "step": 34060
    },
    {
      "entropy": 5.257856369018555,
      "epoch": 3.8258550008423655,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003607122932246587,
      "loss": 4.9109,
      "mean_token_accuracy": 0.22714192271232606,
      "num_tokens": 76031031.0,
      "step": 34065
    },
    {
      "entropy": 5.2051338195800785,
      "epoch": 3.826416577750323,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00036067517042264147,
      "loss": 4.7483,
      "mean_token_accuracy": 0.23150772154331206,
      "num_tokens": 76041433.0,
      "step": 34070
    },
    {
      "entropy": 5.300303506851196,
      "epoch": 3.8269781546582804,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003606380448926936,
      "loss": 4.9068,
      "mean_token_accuracy": 0.22581379264593124,
      "num_tokens": 76052359.0,
      "step": 34075
    },
    {
      "entropy": 5.272863674163818,
      "epoch": 3.827539731566238,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003606009166359972,
      "loss": 4.8642,
      "mean_token_accuracy": 0.22736063450574875,
      "num_tokens": 76062857.0,
      "step": 34080
    },
    {
      "entropy": 5.420915269851685,
      "epoch": 3.8281013084741957,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036056378565373434,
      "loss": 5.0581,
      "mean_token_accuracy": 0.21165815144777297,
      "num_tokens": 76074937.0,
      "step": 34085
    },
    {
      "entropy": 5.203970003128052,
      "epoch": 3.828662885382153,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003605266519470874,
      "loss": 4.8186,
      "mean_token_accuracy": 0.2244480550289154,
      "num_tokens": 76087434.0,
      "step": 34090
    },
    {
      "entropy": 5.25016188621521,
      "epoch": 3.8292244622901106,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003604895155172386,
      "loss": 4.865,
      "mean_token_accuracy": 0.2256707400083542,
      "num_tokens": 76098244.0,
      "step": 34095
    },
    {
      "entropy": 5.21964807510376,
      "epoch": 3.8297860391980683,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003604523763653706,
      "loss": 4.8348,
      "mean_token_accuracy": 0.22742789089679719,
      "num_tokens": 76108798.0,
      "step": 34100
    },
    {
      "entropy": 5.309400796890259,
      "epoch": 3.8303476161060255,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036041523449266573,
      "loss": 4.885,
      "mean_token_accuracy": 0.2266853541135788,
      "num_tokens": 76120776.0,
      "step": 34105
    },
    {
      "entropy": 5.289412307739258,
      "epoch": 3.830909193013983,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00036037808990030667,
      "loss": 4.9429,
      "mean_token_accuracy": 0.2251514658331871,
      "num_tokens": 76133562.0,
      "step": 34110
    },
    {
      "entropy": 5.21394100189209,
      "epoch": 3.831470769921941,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003603409425894761,
      "loss": 4.906,
      "mean_token_accuracy": 0.22493847757577895,
      "num_tokens": 76146071.0,
      "step": 34115
    },
    {
      "entropy": 5.277028894424438,
      "epoch": 3.8320323468298985,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003603037925613568,
      "loss": 4.8396,
      "mean_token_accuracy": 0.23269666731357574,
      "num_tokens": 76157568.0,
      "step": 34120
    },
    {
      "entropy": 5.249169063568115,
      "epoch": 3.832593923737856,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003602666398171317,
      "loss": 4.8183,
      "mean_token_accuracy": 0.23510532975196838,
      "num_tokens": 76168853.0,
      "step": 34125
    },
    {
      "entropy": 5.134120512008667,
      "epoch": 3.8331555006458133,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036022948435798373,
      "loss": 4.7686,
      "mean_token_accuracy": 0.22784252166748048,
      "num_tokens": 76178480.0,
      "step": 34130
    },
    {
      "entropy": 5.205069208145142,
      "epoch": 3.833717077553771,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000360192326185096,
      "loss": 4.7891,
      "mean_token_accuracy": 0.22901634275913238,
      "num_tokens": 76189031.0,
      "step": 34135
    },
    {
      "entropy": 5.148138475418091,
      "epoch": 3.8342786544617287,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003601551652996514,
      "loss": 4.7651,
      "mean_token_accuracy": 0.23046513944864272,
      "num_tokens": 76199345.0,
      "step": 34140
    },
    {
      "entropy": 5.281720304489136,
      "epoch": 3.834840231369686,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003601180017028334,
      "loss": 4.9781,
      "mean_token_accuracy": 0.2138596296310425,
      "num_tokens": 76211023.0,
      "step": 34145
    },
    {
      "entropy": 5.359538841247558,
      "epoch": 3.8354018082776435,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003600808353958252,
      "loss": 4.9556,
      "mean_token_accuracy": 0.2206903874874115,
      "num_tokens": 76221562.0,
      "step": 34150
    },
    {
      "entropy": 5.359281301498413,
      "epoch": 3.835963385185601,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003600436663798102,
      "loss": 5.0011,
      "mean_token_accuracy": 0.2153517484664917,
      "num_tokens": 76235113.0,
      "step": 34155
    },
    {
      "entropy": 5.312883901596069,
      "epoch": 3.836524962093559,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036000649465597185,
      "loss": 4.8898,
      "mean_token_accuracy": 0.22973314821720123,
      "num_tokens": 76245804.0,
      "step": 34160
    },
    {
      "entropy": 5.252270603179932,
      "epoch": 3.8370865390015165,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00035996932022549377,
      "loss": 4.8678,
      "mean_token_accuracy": 0.23283231258392334,
      "num_tokens": 76257275.0,
      "step": 34165
    },
    {
      "entropy": 5.216567707061768,
      "epoch": 3.8376481159094737,
      "grad_norm": 0.890625,
      "learning_rate": 0.00035993214308955955,
      "loss": 4.8789,
      "mean_token_accuracy": 0.22585703134536744,
      "num_tokens": 76268401.0,
      "step": 34170
    },
    {
      "entropy": 5.21506085395813,
      "epoch": 3.8382096928174314,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000359894963249353,
      "loss": 4.8064,
      "mean_token_accuracy": 0.22463005632162095,
      "num_tokens": 76278943.0,
      "step": 34175
    },
    {
      "entropy": 5.26408281326294,
      "epoch": 3.838771269725389,
      "grad_norm": 1.125,
      "learning_rate": 0.00035985778070605785,
      "loss": 4.8943,
      "mean_token_accuracy": 0.22610405534505845,
      "num_tokens": 76289899.0,
      "step": 34180
    },
    {
      "entropy": 5.325868225097656,
      "epoch": 3.8393328466333463,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003598205954608581,
      "loss": 4.9864,
      "mean_token_accuracy": 0.22094006836414337,
      "num_tokens": 76300574.0,
      "step": 34185
    },
    {
      "entropy": 5.285263013839722,
      "epoch": 3.839894423541304,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00035978340751493767,
      "loss": 4.9285,
      "mean_token_accuracy": 0.22236034721136094,
      "num_tokens": 76312789.0,
      "step": 34190
    },
    {
      "entropy": 5.2058055877685545,
      "epoch": 3.8404560004492616,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00035974621686948063,
      "loss": 4.8152,
      "mean_token_accuracy": 0.22481757402420044,
      "num_tokens": 76322911.0,
      "step": 34195
    },
    {
      "entropy": 5.2659296035766605,
      "epoch": 3.841017577357219,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003597090235256712,
      "loss": 4.9324,
      "mean_token_accuracy": 0.22719287723302842,
      "num_tokens": 76333699.0,
      "step": 34200
    },
    {
      "entropy": 5.179018640518189,
      "epoch": 3.8415791542651765,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003596718274846936,
      "loss": 4.8161,
      "mean_token_accuracy": 0.22690704911947251,
      "num_tokens": 76344405.0,
      "step": 34205
    },
    {
      "entropy": 5.25769624710083,
      "epoch": 3.842140731173134,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003596346287477322,
      "loss": 4.9027,
      "mean_token_accuracy": 0.22276989221572877,
      "num_tokens": 76356403.0,
      "step": 34210
    },
    {
      "entropy": 5.277002334594727,
      "epoch": 3.842702308081092,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003595974273159714,
      "loss": 4.8429,
      "mean_token_accuracy": 0.22135212123394013,
      "num_tokens": 76368216.0,
      "step": 34215
    },
    {
      "entropy": 5.3624931335449215,
      "epoch": 3.8432638849890495,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003595602231905957,
      "loss": 5.0954,
      "mean_token_accuracy": 0.20936744213104247,
      "num_tokens": 76379996.0,
      "step": 34220
    },
    {
      "entropy": 5.249874687194824,
      "epoch": 3.8438254618970067,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00035952301637278973,
      "loss": 4.849,
      "mean_token_accuracy": 0.23530532121658326,
      "num_tokens": 76391409.0,
      "step": 34225
    },
    {
      "entropy": 5.290356922149658,
      "epoch": 3.8443870388049644,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00035948580686373803,
      "loss": 4.9129,
      "mean_token_accuracy": 0.22467241883277894,
      "num_tokens": 76402111.0,
      "step": 34230
    },
    {
      "entropy": 5.2184998989105225,
      "epoch": 3.844948615712922,
      "grad_norm": 1.0,
      "learning_rate": 0.00035944859466462554,
      "loss": 4.7919,
      "mean_token_accuracy": 0.23326494991779329,
      "num_tokens": 76413225.0,
      "step": 34235
    },
    {
      "entropy": 5.294613790512085,
      "epoch": 3.8455101926208792,
      "grad_norm": 1.0,
      "learning_rate": 0.000359411379776637,
      "loss": 4.9061,
      "mean_token_accuracy": 0.22200678437948226,
      "num_tokens": 76423236.0,
      "step": 34240
    },
    {
      "entropy": 5.310553789138794,
      "epoch": 3.846071769528837,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00035937416220095745,
      "loss": 4.9287,
      "mean_token_accuracy": 0.23044774681329727,
      "num_tokens": 76435599.0,
      "step": 34245
    },
    {
      "entropy": 5.222480249404907,
      "epoch": 3.8466333464367946,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003593369419387718,
      "loss": 4.886,
      "mean_token_accuracy": 0.2263246074318886,
      "num_tokens": 76446182.0,
      "step": 34250
    },
    {
      "entropy": 5.244609022140503,
      "epoch": 3.847194923344752,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003592997189912652,
      "loss": 4.8498,
      "mean_token_accuracy": 0.225451697409153,
      "num_tokens": 76459151.0,
      "step": 34255
    },
    {
      "entropy": 5.309891891479492,
      "epoch": 3.84775650025271,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035926249335962296,
      "loss": 4.97,
      "mean_token_accuracy": 0.21851343363523484,
      "num_tokens": 76471235.0,
      "step": 34260
    },
    {
      "entropy": 5.234830570220947,
      "epoch": 3.848318077160667,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035922526504503016,
      "loss": 4.8563,
      "mean_token_accuracy": 0.2235744848847389,
      "num_tokens": 76481699.0,
      "step": 34265
    },
    {
      "entropy": 5.305647373199463,
      "epoch": 3.8488796540686248,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003591880340486723,
      "loss": 4.9546,
      "mean_token_accuracy": 0.2210097625851631,
      "num_tokens": 76493038.0,
      "step": 34270
    },
    {
      "entropy": 5.231347846984863,
      "epoch": 3.8494412309765824,
      "grad_norm": 1.0,
      "learning_rate": 0.0003591508003717347,
      "loss": 4.8564,
      "mean_token_accuracy": 0.23037817925214768,
      "num_tokens": 76503601.0,
      "step": 34275
    },
    {
      "entropy": 5.242006397247314,
      "epoch": 3.8500028078845396,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00035911356401540306,
      "loss": 4.8136,
      "mean_token_accuracy": 0.23216175734996797,
      "num_tokens": 76513914.0,
      "step": 34280
    },
    {
      "entropy": 5.265952444076538,
      "epoch": 3.8505643847924973,
      "grad_norm": 1.0,
      "learning_rate": 0.0003590763249808629,
      "loss": 4.8507,
      "mean_token_accuracy": 0.23462510406970977,
      "num_tokens": 76524924.0,
      "step": 34285
    },
    {
      "entropy": 5.19795184135437,
      "epoch": 3.851125961700455,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00035903908326929995,
      "loss": 4.8105,
      "mean_token_accuracy": 0.23200812935829163,
      "num_tokens": 76536468.0,
      "step": 34290
    },
    {
      "entropy": 5.328640937805176,
      "epoch": 3.851687538608412,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00035900183888190005,
      "loss": 4.9496,
      "mean_token_accuracy": 0.22454443871974944,
      "num_tokens": 76548320.0,
      "step": 34295
    },
    {
      "entropy": 5.300194597244262,
      "epoch": 3.85224911551637,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003589645918198489,
      "loss": 4.8856,
      "mean_token_accuracy": 0.22632029056549072,
      "num_tokens": 76560482.0,
      "step": 34300
    },
    {
      "entropy": 5.287341594696045,
      "epoch": 3.8528106924243275,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00035892734208433264,
      "loss": 4.9445,
      "mean_token_accuracy": 0.22188685089349747,
      "num_tokens": 76571743.0,
      "step": 34305
    },
    {
      "entropy": 5.242906093597412,
      "epoch": 3.853372269332285,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00035889008967653725,
      "loss": 4.9139,
      "mean_token_accuracy": 0.22409356981515885,
      "num_tokens": 76582545.0,
      "step": 34310
    },
    {
      "entropy": 5.326081323623657,
      "epoch": 3.853933846240243,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003588528345976489,
      "loss": 4.9728,
      "mean_token_accuracy": 0.22290417701005935,
      "num_tokens": 76594104.0,
      "step": 34315
    },
    {
      "entropy": 5.315026950836182,
      "epoch": 3.8544954231482,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003588155768488537,
      "loss": 4.9244,
      "mean_token_accuracy": 0.227963450551033,
      "num_tokens": 76603414.0,
      "step": 34320
    },
    {
      "entropy": 5.30792269706726,
      "epoch": 3.8550570000561577,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035877831643133803,
      "loss": 4.9419,
      "mean_token_accuracy": 0.22045070976018905,
      "num_tokens": 76614098.0,
      "step": 34325
    },
    {
      "entropy": 5.39159893989563,
      "epoch": 3.8556185769641154,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003587410533462883,
      "loss": 5.0554,
      "mean_token_accuracy": 0.218463034927845,
      "num_tokens": 76625179.0,
      "step": 34330
    },
    {
      "entropy": 5.265978574752808,
      "epoch": 3.8561801538720726,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00035870378759489104,
      "loss": 4.7736,
      "mean_token_accuracy": 0.2336119994521141,
      "num_tokens": 76636710.0,
      "step": 34335
    },
    {
      "entropy": 5.251077651977539,
      "epoch": 3.8567417307800302,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003586665191783326,
      "loss": 4.9212,
      "mean_token_accuracy": 0.22215411216020584,
      "num_tokens": 76647585.0,
      "step": 34340
    },
    {
      "entropy": 5.273920965194702,
      "epoch": 3.857303307687988,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003586292480977998,
      "loss": 4.9,
      "mean_token_accuracy": 0.21816809624433517,
      "num_tokens": 76658376.0,
      "step": 34345
    },
    {
      "entropy": 5.306312465667725,
      "epoch": 3.8578648845959456,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00035859197435447927,
      "loss": 4.8683,
      "mean_token_accuracy": 0.23076998442411423,
      "num_tokens": 76669237.0,
      "step": 34350
    },
    {
      "entropy": 5.343724012374878,
      "epoch": 3.8584264615039032,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00035855469794955787,
      "loss": 4.9923,
      "mean_token_accuracy": 0.2130566194653511,
      "num_tokens": 76681067.0,
      "step": 34355
    },
    {
      "entropy": 5.353814554214478,
      "epoch": 3.8589880384118604,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035851741888422254,
      "loss": 4.8901,
      "mean_token_accuracy": 0.23043228685855865,
      "num_tokens": 76690858.0,
      "step": 34360
    },
    {
      "entropy": 5.318914413452148,
      "epoch": 3.859549615319818,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003584801371596601,
      "loss": 4.9502,
      "mean_token_accuracy": 0.2187487542629242,
      "num_tokens": 76701571.0,
      "step": 34365
    },
    {
      "entropy": 5.281003093719482,
      "epoch": 3.8601111922277758,
      "grad_norm": 1.0,
      "learning_rate": 0.0003584428527770578,
      "loss": 4.8979,
      "mean_token_accuracy": 0.21535158306360244,
      "num_tokens": 76712565.0,
      "step": 34370
    },
    {
      "entropy": 5.2938305854797365,
      "epoch": 3.860672769135733,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00035840556573760266,
      "loss": 4.9229,
      "mean_token_accuracy": 0.22332206517457961,
      "num_tokens": 76724484.0,
      "step": 34375
    },
    {
      "entropy": 5.253723907470703,
      "epoch": 3.8612343460436906,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003583682760424819,
      "loss": 4.7911,
      "mean_token_accuracy": 0.22959929406642915,
      "num_tokens": 76735596.0,
      "step": 34380
    },
    {
      "entropy": 5.33608717918396,
      "epoch": 3.8617959229516483,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00035833098369288303,
      "loss": 4.9667,
      "mean_token_accuracy": 0.21849900484085083,
      "num_tokens": 76747160.0,
      "step": 34385
    },
    {
      "entropy": 5.234181594848633,
      "epoch": 3.8623574998596055,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035829368868999325,
      "loss": 4.8543,
      "mean_token_accuracy": 0.2188256040215492,
      "num_tokens": 76758136.0,
      "step": 34390
    },
    {
      "entropy": 5.241367626190185,
      "epoch": 3.862919076767563,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00035825639103500005,
      "loss": 4.886,
      "mean_token_accuracy": 0.23100064992904662,
      "num_tokens": 76768167.0,
      "step": 34395
    },
    {
      "entropy": 5.3940118789672855,
      "epoch": 3.863480653675521,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00035821909072909113,
      "loss": 5.0289,
      "mean_token_accuracy": 0.2218724563717842,
      "num_tokens": 76779717.0,
      "step": 34400
    },
    {
      "entropy": 5.198473310470581,
      "epoch": 3.8640422305834785,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00035818178777345417,
      "loss": 4.876,
      "mean_token_accuracy": 0.22140266597270966,
      "num_tokens": 76790837.0,
      "step": 34405
    },
    {
      "entropy": 5.27897686958313,
      "epoch": 3.864603807491436,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003581444821692767,
      "loss": 4.8794,
      "mean_token_accuracy": 0.22414804697036744,
      "num_tokens": 76801159.0,
      "step": 34410
    },
    {
      "entropy": 5.285088396072387,
      "epoch": 3.8651653843993934,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003581071739177467,
      "loss": 4.9075,
      "mean_token_accuracy": 0.2227723479270935,
      "num_tokens": 76812409.0,
      "step": 34415
    },
    {
      "entropy": 5.365723371505737,
      "epoch": 3.865726961307351,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035806986302005205,
      "loss": 4.9965,
      "mean_token_accuracy": 0.21641923636198043,
      "num_tokens": 76824360.0,
      "step": 34420
    },
    {
      "entropy": 5.336919927597046,
      "epoch": 3.8662885382153087,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003580325494773808,
      "loss": 4.8766,
      "mean_token_accuracy": 0.22377846091985704,
      "num_tokens": 76834725.0,
      "step": 34425
    },
    {
      "entropy": 5.217624092102051,
      "epoch": 3.866850115123266,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035799523329092093,
      "loss": 4.8851,
      "mean_token_accuracy": 0.22609542012214662,
      "num_tokens": 76846099.0,
      "step": 34430
    },
    {
      "entropy": 5.330995035171509,
      "epoch": 3.8674116920312236,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00035795791446186073,
      "loss": 4.9562,
      "mean_token_accuracy": 0.21585319936275482,
      "num_tokens": 76857229.0,
      "step": 34435
    },
    {
      "entropy": 5.335905075073242,
      "epoch": 3.8679732689391813,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003579205929913882,
      "loss": 4.9069,
      "mean_token_accuracy": 0.22917262613773345,
      "num_tokens": 76868609.0,
      "step": 34440
    },
    {
      "entropy": 5.336432695388794,
      "epoch": 3.868534845847139,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003578832688806919,
      "loss": 4.8953,
      "mean_token_accuracy": 0.22816008180379868,
      "num_tokens": 76879539.0,
      "step": 34445
    },
    {
      "entropy": 5.329161787033081,
      "epoch": 3.8690964227550966,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003578459421309602,
      "loss": 4.9311,
      "mean_token_accuracy": 0.2146620571613312,
      "num_tokens": 76889333.0,
      "step": 34450
    },
    {
      "entropy": 5.266942262649536,
      "epoch": 3.869657999663054,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003578086127433816,
      "loss": 4.9848,
      "mean_token_accuracy": 0.21876032948493956,
      "num_tokens": 76901334.0,
      "step": 34455
    },
    {
      "entropy": 5.307915210723877,
      "epoch": 3.8702195765710115,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00035777128071914454,
      "loss": 4.8587,
      "mean_token_accuracy": 0.2241705283522606,
      "num_tokens": 76913758.0,
      "step": 34460
    },
    {
      "entropy": 5.323215293884277,
      "epoch": 3.870781153478969,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003577339460594378,
      "loss": 4.9278,
      "mean_token_accuracy": 0.22010490149259568,
      "num_tokens": 76925317.0,
      "step": 34465
    },
    {
      "entropy": 5.329947376251221,
      "epoch": 3.8713427303869263,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003576966087654502,
      "loss": 4.9992,
      "mean_token_accuracy": 0.21932460516691207,
      "num_tokens": 76937169.0,
      "step": 34470
    },
    {
      "entropy": 5.239972496032715,
      "epoch": 3.871904307294884,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00035765926883837054,
      "loss": 4.879,
      "mean_token_accuracy": 0.22321043014526368,
      "num_tokens": 76948958.0,
      "step": 34475
    },
    {
      "entropy": 5.329688644409179,
      "epoch": 3.8724658842028417,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003576219262793877,
      "loss": 4.9837,
      "mean_token_accuracy": 0.21946621984243392,
      "num_tokens": 76960686.0,
      "step": 34480
    },
    {
      "entropy": 5.290989828109741,
      "epoch": 3.873027461110799,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00035758458108969056,
      "loss": 4.9006,
      "mean_token_accuracy": 0.22724266350269318,
      "num_tokens": 76971127.0,
      "step": 34485
    },
    {
      "entropy": 5.228529644012451,
      "epoch": 3.8735890380187565,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003575472332704684,
      "loss": 4.7273,
      "mean_token_accuracy": 0.23852065801620484,
      "num_tokens": 76982647.0,
      "step": 34490
    },
    {
      "entropy": 5.320918941497803,
      "epoch": 3.874150614926714,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00035750988282291037,
      "loss": 4.9559,
      "mean_token_accuracy": 0.2245274230837822,
      "num_tokens": 76994274.0,
      "step": 34495
    },
    {
      "entropy": 5.248554801940918,
      "epoch": 3.874712191834672,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00035747252974820565,
      "loss": 4.8571,
      "mean_token_accuracy": 0.229855777323246,
      "num_tokens": 77005435.0,
      "step": 34500
    },
    {
      "entropy": 5.235554265975952,
      "epoch": 3.8752737687426295,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035743517404754354,
      "loss": 4.8542,
      "mean_token_accuracy": 0.22735839784145356,
      "num_tokens": 77016913.0,
      "step": 34505
    },
    {
      "entropy": 5.279568767547607,
      "epoch": 3.8758353456505867,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003573978157221136,
      "loss": 4.916,
      "mean_token_accuracy": 0.22417614310979844,
      "num_tokens": 77027340.0,
      "step": 34510
    },
    {
      "entropy": 5.25700798034668,
      "epoch": 3.8763969225585444,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00035736045477310517,
      "loss": 4.8836,
      "mean_token_accuracy": 0.22270741015672685,
      "num_tokens": 77039465.0,
      "step": 34515
    },
    {
      "entropy": 5.280011701583862,
      "epoch": 3.876958499466502,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035732309120170786,
      "loss": 4.9322,
      "mean_token_accuracy": 0.22023397088050842,
      "num_tokens": 77050503.0,
      "step": 34520
    },
    {
      "entropy": 5.35649733543396,
      "epoch": 3.8775200763744593,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00035728572500911155,
      "loss": 4.8687,
      "mean_token_accuracy": 0.22560582011938096,
      "num_tokens": 77060212.0,
      "step": 34525
    },
    {
      "entropy": 5.333701324462891,
      "epoch": 3.878081653282417,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00035724835619650573,
      "loss": 4.9909,
      "mean_token_accuracy": 0.21963285505771638,
      "num_tokens": 77071134.0,
      "step": 34530
    },
    {
      "entropy": 5.331094074249267,
      "epoch": 3.8786432301903746,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003572109847650804,
      "loss": 4.9456,
      "mean_token_accuracy": 0.22893598675727844,
      "num_tokens": 77081844.0,
      "step": 34535
    },
    {
      "entropy": 5.325154113769531,
      "epoch": 3.8792048070983323,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00035717361071602546,
      "loss": 4.9051,
      "mean_token_accuracy": 0.22805575728416444,
      "num_tokens": 77092891.0,
      "step": 34540
    },
    {
      "entropy": 5.237735605239868,
      "epoch": 3.87976638400629,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00035713623405053083,
      "loss": 4.8446,
      "mean_token_accuracy": 0.2266984760761261,
      "num_tokens": 77104620.0,
      "step": 34545
    },
    {
      "entropy": 5.247331953048706,
      "epoch": 3.880327960914247,
      "grad_norm": 0.984375,
      "learning_rate": 0.00035709885476978666,
      "loss": 4.9322,
      "mean_token_accuracy": 0.21626074761152267,
      "num_tokens": 77115249.0,
      "step": 34550
    },
    {
      "entropy": 5.248649930953979,
      "epoch": 3.880889537822205,
      "grad_norm": 0.953125,
      "learning_rate": 0.00035706147287498303,
      "loss": 4.9061,
      "mean_token_accuracy": 0.2218364581465721,
      "num_tokens": 77127287.0,
      "step": 34555
    },
    {
      "entropy": 5.349998807907104,
      "epoch": 3.8814511147301625,
      "grad_norm": 0.9375,
      "learning_rate": 0.00035702408836731043,
      "loss": 4.8909,
      "mean_token_accuracy": 0.22706527411937713,
      "num_tokens": 77139022.0,
      "step": 34560
    },
    {
      "entropy": 5.252149391174316,
      "epoch": 3.8820126916381197,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000356986701247959,
      "loss": 4.8697,
      "mean_token_accuracy": 0.22500486820936202,
      "num_tokens": 77148594.0,
      "step": 34565
    },
    {
      "entropy": 5.28051061630249,
      "epoch": 3.8825742685460773,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003569493115181192,
      "loss": 4.8702,
      "mean_token_accuracy": 0.2278356209397316,
      "num_tokens": 77158882.0,
      "step": 34570
    },
    {
      "entropy": 5.26755166053772,
      "epoch": 3.883135845454035,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003569119191789815,
      "loss": 4.903,
      "mean_token_accuracy": 0.22252333909273148,
      "num_tokens": 77169721.0,
      "step": 34575
    },
    {
      "entropy": 5.310887432098388,
      "epoch": 3.8836974223619922,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00035687452423173655,
      "loss": 4.9111,
      "mean_token_accuracy": 0.220664681494236,
      "num_tokens": 77181223.0,
      "step": 34580
    },
    {
      "entropy": 5.2812761783599855,
      "epoch": 3.88425899926995,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000356837126677575,
      "loss": 4.8729,
      "mean_token_accuracy": 0.22135841548442842,
      "num_tokens": 77193000.0,
      "step": 34585
    },
    {
      "entropy": 5.255116558074951,
      "epoch": 3.8848205761779075,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003567997265176876,
      "loss": 4.8679,
      "mean_token_accuracy": 0.2258219301700592,
      "num_tokens": 77203721.0,
      "step": 34590
    },
    {
      "entropy": 5.201212596893311,
      "epoch": 3.885382153085865,
      "grad_norm": 0.984375,
      "learning_rate": 0.00035676232375326513,
      "loss": 4.7237,
      "mean_token_accuracy": 0.23440989404916762,
      "num_tokens": 77214601.0,
      "step": 34595
    },
    {
      "entropy": 5.304338836669922,
      "epoch": 3.885943729993823,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035672491838549867,
      "loss": 4.901,
      "mean_token_accuracy": 0.2216184616088867,
      "num_tokens": 77225478.0,
      "step": 34600
    },
    {
      "entropy": 5.317203664779663,
      "epoch": 3.88650530690178,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000356687510415579,
      "loss": 4.9666,
      "mean_token_accuracy": 0.2196904942393303,
      "num_tokens": 77237601.0,
      "step": 34605
    },
    {
      "entropy": 5.3245162010192875,
      "epoch": 3.8870668838097377,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003566500998446973,
      "loss": 4.9805,
      "mean_token_accuracy": 0.21804504245519638,
      "num_tokens": 77248137.0,
      "step": 34610
    },
    {
      "entropy": 5.3404988765716555,
      "epoch": 3.8876284607176954,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00035661268667404485,
      "loss": 4.9918,
      "mean_token_accuracy": 0.22293551862239838,
      "num_tokens": 77260366.0,
      "step": 34615
    },
    {
      "entropy": 5.390023994445801,
      "epoch": 3.8881900376256526,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003565752709048127,
      "loss": 4.9291,
      "mean_token_accuracy": 0.22071584314107895,
      "num_tokens": 77270301.0,
      "step": 34620
    },
    {
      "entropy": 5.338554525375367,
      "epoch": 3.8887516145336103,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003565378525381924,
      "loss": 4.956,
      "mean_token_accuracy": 0.22302247136831282,
      "num_tokens": 77282824.0,
      "step": 34625
    },
    {
      "entropy": 5.277181625366211,
      "epoch": 3.889313191441568,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003565004315753752,
      "loss": 4.8974,
      "mean_token_accuracy": 0.22718023955821992,
      "num_tokens": 77293532.0,
      "step": 34630
    },
    {
      "entropy": 5.323145484924316,
      "epoch": 3.8898747683495256,
      "grad_norm": 1.125,
      "learning_rate": 0.00035646300801755266,
      "loss": 4.8879,
      "mean_token_accuracy": 0.22404447793960572,
      "num_tokens": 77304713.0,
      "step": 34635
    },
    {
      "entropy": 5.268237924575805,
      "epoch": 3.8904363452574833,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00035642558186591635,
      "loss": 4.8794,
      "mean_token_accuracy": 0.22477043867111207,
      "num_tokens": 77317666.0,
      "step": 34640
    },
    {
      "entropy": 5.215662717819214,
      "epoch": 3.8909979221654405,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003563881531216579,
      "loss": 4.8783,
      "mean_token_accuracy": 0.2294680431485176,
      "num_tokens": 77328217.0,
      "step": 34645
    },
    {
      "entropy": 5.299503374099731,
      "epoch": 3.891559499073398,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003563507217859691,
      "loss": 4.9031,
      "mean_token_accuracy": 0.22598866522312164,
      "num_tokens": 77339607.0,
      "step": 34650
    },
    {
      "entropy": 5.390716457366944,
      "epoch": 3.892121075981356,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003563132878600419,
      "loss": 4.9677,
      "mean_token_accuracy": 0.2269105762243271,
      "num_tokens": 77351651.0,
      "step": 34655
    },
    {
      "entropy": 5.408884811401367,
      "epoch": 3.892682652889313,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003562758513450678,
      "loss": 5.022,
      "mean_token_accuracy": 0.21862683892250062,
      "num_tokens": 77364431.0,
      "step": 34660
    },
    {
      "entropy": 5.273098611831665,
      "epoch": 3.8932442297972707,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003562384122422393,
      "loss": 4.9234,
      "mean_token_accuracy": 0.22398302406072618,
      "num_tokens": 77375426.0,
      "step": 34665
    },
    {
      "entropy": 5.267915821075439,
      "epoch": 3.8938058067052284,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003562009705527481,
      "loss": 4.8887,
      "mean_token_accuracy": 0.22572114318609238,
      "num_tokens": 77386794.0,
      "step": 34670
    },
    {
      "entropy": 5.330571174621582,
      "epoch": 3.8943673836131856,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00035616352627778665,
      "loss": 4.8385,
      "mean_token_accuracy": 0.2270701050758362,
      "num_tokens": 77398155.0,
      "step": 34675
    },
    {
      "entropy": 5.379941082000732,
      "epoch": 3.8949289605211432,
      "grad_norm": 0.96875,
      "learning_rate": 0.00035612607941854695,
      "loss": 5.0513,
      "mean_token_accuracy": 0.21255846023559571,
      "num_tokens": 77408922.0,
      "step": 34680
    },
    {
      "entropy": 5.233512210845947,
      "epoch": 3.895490537429101,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003560886299762214,
      "loss": 4.8539,
      "mean_token_accuracy": 0.22943132370710373,
      "num_tokens": 77419084.0,
      "step": 34685
    },
    {
      "entropy": 5.191167640686035,
      "epoch": 3.8960521143370586,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003560511779520025,
      "loss": 4.7959,
      "mean_token_accuracy": 0.2349379301071167,
      "num_tokens": 77430141.0,
      "step": 34690
    },
    {
      "entropy": 5.3415594577789305,
      "epoch": 3.896613691245016,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003560137233470826,
      "loss": 5.0114,
      "mean_token_accuracy": 0.22436405569314957,
      "num_tokens": 77440754.0,
      "step": 34695
    },
    {
      "entropy": 5.278448915481567,
      "epoch": 3.8971752681529734,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035597626616265434,
      "loss": 4.8446,
      "mean_token_accuracy": 0.2282521277666092,
      "num_tokens": 77453532.0,
      "step": 34700
    },
    {
      "entropy": 5.282544565200806,
      "epoch": 3.897736845060931,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00035593880639991037,
      "loss": 4.8636,
      "mean_token_accuracy": 0.2314741924405098,
      "num_tokens": 77464987.0,
      "step": 34705
    },
    {
      "entropy": 5.252427577972412,
      "epoch": 3.8982984219688888,
      "grad_norm": 1.09375,
      "learning_rate": 0.00035590134406004336,
      "loss": 4.9066,
      "mean_token_accuracy": 0.2201730102300644,
      "num_tokens": 77477559.0,
      "step": 34710
    },
    {
      "entropy": 5.24111852645874,
      "epoch": 3.898859998876846,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003558638791442463,
      "loss": 4.8602,
      "mean_token_accuracy": 0.22155980616807938,
      "num_tokens": 77489863.0,
      "step": 34715
    },
    {
      "entropy": 5.223232126235962,
      "epoch": 3.8994215757848036,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003558264116537119,
      "loss": 4.8496,
      "mean_token_accuracy": 0.2224620670080185,
      "num_tokens": 77501119.0,
      "step": 34720
    },
    {
      "entropy": 5.297039127349853,
      "epoch": 3.8999831526927613,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003557889415896332,
      "loss": 4.9167,
      "mean_token_accuracy": 0.2337488830089569,
      "num_tokens": 77512530.0,
      "step": 34725
    },
    {
      "entropy": 5.304555130004883,
      "epoch": 3.900544729600719,
      "grad_norm": 1.0,
      "learning_rate": 0.00035575146895320327,
      "loss": 4.9153,
      "mean_token_accuracy": 0.2240309163928032,
      "num_tokens": 77524624.0,
      "step": 34730
    },
    {
      "entropy": 5.326170587539673,
      "epoch": 3.9011063065086766,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003557139937456152,
      "loss": 4.9133,
      "mean_token_accuracy": 0.22078305184841157,
      "num_tokens": 77536567.0,
      "step": 34735
    },
    {
      "entropy": 5.259147310256958,
      "epoch": 3.901667883416634,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035567651596806236,
      "loss": 4.8739,
      "mean_token_accuracy": 0.22617120295763016,
      "num_tokens": 77547061.0,
      "step": 34740
    },
    {
      "entropy": 5.300286054611206,
      "epoch": 3.9022294603245915,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00035563903562173786,
      "loss": 4.9305,
      "mean_token_accuracy": 0.22500188052654266,
      "num_tokens": 77558783.0,
      "step": 34745
    },
    {
      "entropy": 5.256566953659058,
      "epoch": 3.902791037232549,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00035560155270783527,
      "loss": 4.8658,
      "mean_token_accuracy": 0.22868071049451827,
      "num_tokens": 77570382.0,
      "step": 34750
    },
    {
      "entropy": 5.282086324691773,
      "epoch": 3.9033526141405064,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003555640672275479,
      "loss": 4.8889,
      "mean_token_accuracy": 0.2254563108086586,
      "num_tokens": 77581750.0,
      "step": 34755
    },
    {
      "entropy": 5.268329334259033,
      "epoch": 3.903914191048464,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003555265791820695,
      "loss": 4.9213,
      "mean_token_accuracy": 0.2255515217781067,
      "num_tokens": 77593784.0,
      "step": 34760
    },
    {
      "entropy": 5.201765823364258,
      "epoch": 3.9044757679564217,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003554890885725935,
      "loss": 4.7979,
      "mean_token_accuracy": 0.22752924263477325,
      "num_tokens": 77604993.0,
      "step": 34765
    },
    {
      "entropy": 5.223362398147583,
      "epoch": 3.905037344864379,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003554515954003137,
      "loss": 4.8836,
      "mean_token_accuracy": 0.22703205049037933,
      "num_tokens": 77615977.0,
      "step": 34770
    },
    {
      "entropy": 5.249632930755615,
      "epoch": 3.9055989217723366,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003554140996664239,
      "loss": 4.8659,
      "mean_token_accuracy": 0.2306675359606743,
      "num_tokens": 77625724.0,
      "step": 34775
    },
    {
      "entropy": 5.223937368392944,
      "epoch": 3.9061604986802942,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000355376601372118,
      "loss": 4.8728,
      "mean_token_accuracy": 0.22591926008462906,
      "num_tokens": 77635924.0,
      "step": 34780
    },
    {
      "entropy": 5.317987108230591,
      "epoch": 3.906722075588252,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003553391005185899,
      "loss": 4.9439,
      "mean_token_accuracy": 0.22220877259969712,
      "num_tokens": 77646142.0,
      "step": 34785
    },
    {
      "entropy": 5.268857431411743,
      "epoch": 3.9072836524962096,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003553015971070338,
      "loss": 4.8553,
      "mean_token_accuracy": 0.2256593346595764,
      "num_tokens": 77657500.0,
      "step": 34790
    },
    {
      "entropy": 5.3613160133361815,
      "epoch": 3.907845229404167,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003552640911386436,
      "loss": 4.9419,
      "mean_token_accuracy": 0.21891119480133056,
      "num_tokens": 77668916.0,
      "step": 34795
    },
    {
      "entropy": 5.320800924301148,
      "epoch": 3.9084068063121244,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00035522658261461366,
      "loss": 4.9333,
      "mean_token_accuracy": 0.2155911386013031,
      "num_tokens": 77680395.0,
      "step": 34800
    },
    {
      "entropy": 5.220264530181884,
      "epoch": 3.908968383220082,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003551890715361382,
      "loss": 4.7841,
      "mean_token_accuracy": 0.22765739858150483,
      "num_tokens": 77691309.0,
      "step": 34805
    },
    {
      "entropy": 5.26171498298645,
      "epoch": 3.9095299601280393,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003551515579044116,
      "loss": 4.8425,
      "mean_token_accuracy": 0.2327638864517212,
      "num_tokens": 77702922.0,
      "step": 34810
    },
    {
      "entropy": 5.287754821777344,
      "epoch": 3.910091537035997,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00035511404172062827,
      "loss": 4.88,
      "mean_token_accuracy": 0.22447886914014817,
      "num_tokens": 77714252.0,
      "step": 34815
    },
    {
      "entropy": 5.232956886291504,
      "epoch": 3.9106531139439547,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003550765229859828,
      "loss": 4.8105,
      "mean_token_accuracy": 0.22538516968488692,
      "num_tokens": 77724473.0,
      "step": 34820
    },
    {
      "entropy": 5.244296741485596,
      "epoch": 3.9112146908519123,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003550390017016698,
      "loss": 4.8616,
      "mean_token_accuracy": 0.2233276352286339,
      "num_tokens": 77735103.0,
      "step": 34825
    },
    {
      "entropy": 5.407916069030762,
      "epoch": 3.91177626775987,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00035500147786888387,
      "loss": 4.9902,
      "mean_token_accuracy": 0.21820642054080963,
      "num_tokens": 77747370.0,
      "step": 34830
    },
    {
      "entropy": 5.285176181793213,
      "epoch": 3.912337844667827,
      "grad_norm": 0.96875,
      "learning_rate": 0.00035496395148881994,
      "loss": 4.9015,
      "mean_token_accuracy": 0.21928659975528716,
      "num_tokens": 77759183.0,
      "step": 34835
    },
    {
      "entropy": 5.3133457660675045,
      "epoch": 3.912899421575785,
      "grad_norm": 1.078125,
      "learning_rate": 0.00035492642256267277,
      "loss": 4.8713,
      "mean_token_accuracy": 0.23064311742782592,
      "num_tokens": 77770130.0,
      "step": 34840
    },
    {
      "entropy": 5.197973012924194,
      "epoch": 3.9134609984837425,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00035488889109163725,
      "loss": 4.7532,
      "mean_token_accuracy": 0.24232198894023896,
      "num_tokens": 77780830.0,
      "step": 34845
    },
    {
      "entropy": 5.278346109390259,
      "epoch": 3.9140225753916997,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003548513570769085,
      "loss": 4.8616,
      "mean_token_accuracy": 0.2256148338317871,
      "num_tokens": 77791265.0,
      "step": 34850
    },
    {
      "entropy": 5.17408857345581,
      "epoch": 3.9145841522996574,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003548138205196816,
      "loss": 4.8146,
      "mean_token_accuracy": 0.2274922490119934,
      "num_tokens": 77802899.0,
      "step": 34855
    },
    {
      "entropy": 5.16103196144104,
      "epoch": 3.915145729207615,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003547762814211517,
      "loss": 4.8469,
      "mean_token_accuracy": 0.23138746172189711,
      "num_tokens": 77815034.0,
      "step": 34860
    },
    {
      "entropy": 5.269836521148681,
      "epoch": 3.9157073061155723,
      "grad_norm": 0.9375,
      "learning_rate": 0.000354738739782514,
      "loss": 4.9267,
      "mean_token_accuracy": 0.21881146878004074,
      "num_tokens": 77826305.0,
      "step": 34865
    },
    {
      "entropy": 5.280698537826538,
      "epoch": 3.91626888302353,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000354701195604964,
      "loss": 4.9125,
      "mean_token_accuracy": 0.22344717234373093,
      "num_tokens": 77837698.0,
      "step": 34870
    },
    {
      "entropy": 5.283184242248535,
      "epoch": 3.9168304599314876,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003546636488896969,
      "loss": 4.8851,
      "mean_token_accuracy": 0.22585351765155792,
      "num_tokens": 77848630.0,
      "step": 34875
    },
    {
      "entropy": 5.299853086471558,
      "epoch": 3.9173920368394453,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003546260996379083,
      "loss": 4.9661,
      "mean_token_accuracy": 0.21457189023494722,
      "num_tokens": 77859832.0,
      "step": 34880
    },
    {
      "entropy": 5.384355926513672,
      "epoch": 3.917953613747403,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000354588547850794,
      "loss": 5.0032,
      "mean_token_accuracy": 0.2231437310576439,
      "num_tokens": 77871071.0,
      "step": 34885
    },
    {
      "entropy": 5.281819820404053,
      "epoch": 3.91851519065536,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003545509935295493,
      "loss": 4.9005,
      "mean_token_accuracy": 0.2241417095065117,
      "num_tokens": 77881323.0,
      "step": 34890
    },
    {
      "entropy": 5.315157413482666,
      "epoch": 3.919076767563318,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003545134366753701,
      "loss": 4.9451,
      "mean_token_accuracy": 0.22569464892148972,
      "num_tokens": 77893754.0,
      "step": 34895
    },
    {
      "entropy": 5.273713684082031,
      "epoch": 3.9196383444712755,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003544758772894523,
      "loss": 4.8333,
      "mean_token_accuracy": 0.22816308885812758,
      "num_tokens": 77905408.0,
      "step": 34900
    },
    {
      "entropy": 5.1797933101654055,
      "epoch": 3.9201999213792327,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00035443831537299177,
      "loss": 4.7538,
      "mean_token_accuracy": 0.2384544774889946,
      "num_tokens": 77915839.0,
      "step": 34905
    },
    {
      "entropy": 5.212069654464722,
      "epoch": 3.9207614982871903,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003544007509271844,
      "loss": 4.8387,
      "mean_token_accuracy": 0.22585681974887847,
      "num_tokens": 77926626.0,
      "step": 34910
    },
    {
      "entropy": 5.303510236740112,
      "epoch": 3.921323075195148,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00035436318395322636,
      "loss": 5.0462,
      "mean_token_accuracy": 0.21014297157526016,
      "num_tokens": 77938781.0,
      "step": 34915
    },
    {
      "entropy": 5.354021596908569,
      "epoch": 3.9218846521031057,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003543256144523137,
      "loss": 4.9356,
      "mean_token_accuracy": 0.21908646076917648,
      "num_tokens": 77949370.0,
      "step": 34920
    },
    {
      "entropy": 5.254240465164185,
      "epoch": 3.9224462290110633,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003542880424256428,
      "loss": 4.8727,
      "mean_token_accuracy": 0.2189296454191208,
      "num_tokens": 77961396.0,
      "step": 34925
    },
    {
      "entropy": 5.342385196685791,
      "epoch": 3.9230078059190205,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003542504678744099,
      "loss": 4.9554,
      "mean_token_accuracy": 0.21731773316860198,
      "num_tokens": 77972811.0,
      "step": 34930
    },
    {
      "entropy": 5.263238525390625,
      "epoch": 3.923569382826978,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003542128907998113,
      "loss": 4.8489,
      "mean_token_accuracy": 0.23363059759140015,
      "num_tokens": 77984452.0,
      "step": 34935
    },
    {
      "entropy": 5.2173210144042965,
      "epoch": 3.924130959734936,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003541753112030435,
      "loss": 4.8459,
      "mean_token_accuracy": 0.22757040411233903,
      "num_tokens": 77995963.0,
      "step": 34940
    },
    {
      "entropy": 5.226803922653199,
      "epoch": 3.924692536642893,
      "grad_norm": 0.96875,
      "learning_rate": 0.00035413772908530317,
      "loss": 4.7774,
      "mean_token_accuracy": 0.23228265047073365,
      "num_tokens": 78006638.0,
      "step": 34945
    },
    {
      "entropy": 5.158686304092408,
      "epoch": 3.9252541135508507,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003541001444477869,
      "loss": 4.7914,
      "mean_token_accuracy": 0.2379720151424408,
      "num_tokens": 78017982.0,
      "step": 34950
    },
    {
      "entropy": 5.1616637229919435,
      "epoch": 3.9258156904588084,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00035406255729169124,
      "loss": 4.8402,
      "mean_token_accuracy": 0.22724001258611679,
      "num_tokens": 78028827.0,
      "step": 34955
    },
    {
      "entropy": 5.302624607086182,
      "epoch": 3.9263772673667656,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003540249676182131,
      "loss": 4.9009,
      "mean_token_accuracy": 0.2230764612555504,
      "num_tokens": 78040021.0,
      "step": 34960
    },
    {
      "entropy": 5.3357878684997555,
      "epoch": 3.9269388442747233,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035398737542854936,
      "loss": 4.9658,
      "mean_token_accuracy": 0.21650195866823196,
      "num_tokens": 78050827.0,
      "step": 34965
    },
    {
      "entropy": 5.297491884231567,
      "epoch": 3.927500421182681,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003539497807238969,
      "loss": 4.9161,
      "mean_token_accuracy": 0.2193453148007393,
      "num_tokens": 78062241.0,
      "step": 34970
    },
    {
      "entropy": 5.366393184661865,
      "epoch": 3.9280619980906386,
      "grad_norm": 1.03125,
      "learning_rate": 0.00035391218350545287,
      "loss": 4.9727,
      "mean_token_accuracy": 0.21766307204961777,
      "num_tokens": 78073611.0,
      "step": 34975
    },
    {
      "entropy": 5.270255517959595,
      "epoch": 3.9286235749985963,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035387458377441434,
      "loss": 4.8129,
      "mean_token_accuracy": 0.22743256390094757,
      "num_tokens": 78084212.0,
      "step": 34980
    },
    {
      "entropy": 5.23097333908081,
      "epoch": 3.9291851519065535,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003538369815319784,
      "loss": 4.8282,
      "mean_token_accuracy": 0.2287912905216217,
      "num_tokens": 78094696.0,
      "step": 34985
    },
    {
      "entropy": 5.206655979156494,
      "epoch": 3.929746728814511,
      "grad_norm": 1.109375,
      "learning_rate": 0.00035379937677934245,
      "loss": 4.8095,
      "mean_token_accuracy": 0.22930267900228501,
      "num_tokens": 78106204.0,
      "step": 34990
    },
    {
      "entropy": 5.358131456375122,
      "epoch": 3.930308305722469,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003537617695177037,
      "loss": 5.0348,
      "mean_token_accuracy": 0.21453662365674972,
      "num_tokens": 78118660.0,
      "step": 34995
    },
    {
      "entropy": 5.241956520080566,
      "epoch": 3.930869882630426,
      "grad_norm": 1.03125,
      "learning_rate": 0.00035372415974825974,
      "loss": 4.7925,
      "mean_token_accuracy": 0.22947833836078643,
      "num_tokens": 78129507.0,
      "step": 35000
    },
    {
      "entropy": 5.317421627044678,
      "epoch": 3.9314314595383837,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035368654747220795,
      "loss": 4.8939,
      "mean_token_accuracy": 0.22451305389404297,
      "num_tokens": 78140137.0,
      "step": 35005
    },
    {
      "entropy": 5.314343023300171,
      "epoch": 3.9319930364463413,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035364893269074606,
      "loss": 4.9708,
      "mean_token_accuracy": 0.21517615020275116,
      "num_tokens": 78152484.0,
      "step": 35010
    },
    {
      "entropy": 5.347742986679077,
      "epoch": 3.932554613354299,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00035361131540507157,
      "loss": 4.9886,
      "mean_token_accuracy": 0.2159101590514183,
      "num_tokens": 78164559.0,
      "step": 35015
    },
    {
      "entropy": 5.2627321720123295,
      "epoch": 3.9331161902622567,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003535736956163823,
      "loss": 4.864,
      "mean_token_accuracy": 0.2181944504380226,
      "num_tokens": 78175236.0,
      "step": 35020
    },
    {
      "entropy": 5.335025453567505,
      "epoch": 3.933677767170214,
      "grad_norm": 0.984375,
      "learning_rate": 0.00035353607332587613,
      "loss": 4.9629,
      "mean_token_accuracy": 0.21648606806993484,
      "num_tokens": 78186753.0,
      "step": 35025
    },
    {
      "entropy": 5.381012487411499,
      "epoch": 3.9342393440781716,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003534984485347509,
      "loss": 5.0782,
      "mean_token_accuracy": 0.21228013634681703,
      "num_tokens": 78199207.0,
      "step": 35030
    },
    {
      "entropy": 5.433023166656494,
      "epoch": 3.934800920986129,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003534608212442046,
      "loss": 5.1004,
      "mean_token_accuracy": 0.21541324853897095,
      "num_tokens": 78212177.0,
      "step": 35035
    },
    {
      "entropy": 5.336371660232544,
      "epoch": 3.9353624978940864,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00035342319145543533,
      "loss": 4.8848,
      "mean_token_accuracy": 0.21905239373445512,
      "num_tokens": 78223925.0,
      "step": 35040
    },
    {
      "entropy": 5.3075971603393555,
      "epoch": 3.935924074802044,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035338555916964126,
      "loss": 4.8976,
      "mean_token_accuracy": 0.21908605545759202,
      "num_tokens": 78234593.0,
      "step": 35045
    },
    {
      "entropy": 5.280201005935669,
      "epoch": 3.9364856517100018,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00035334792438802056,
      "loss": 4.9547,
      "mean_token_accuracy": 0.22086520195007325,
      "num_tokens": 78245457.0,
      "step": 35050
    },
    {
      "entropy": 5.288337278366089,
      "epoch": 3.937047228617959,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003533102871117716,
      "loss": 4.8892,
      "mean_token_accuracy": 0.22556538730859757,
      "num_tokens": 78256394.0,
      "step": 35055
    },
    {
      "entropy": 5.386278057098389,
      "epoch": 3.9376088055259166,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003532726473420927,
      "loss": 4.9515,
      "mean_token_accuracy": 0.21799045205116271,
      "num_tokens": 78268889.0,
      "step": 35060
    },
    {
      "entropy": 5.22689528465271,
      "epoch": 3.9381703824338743,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003532350050801823,
      "loss": 4.8333,
      "mean_token_accuracy": 0.22639999985694886,
      "num_tokens": 78280441.0,
      "step": 35065
    },
    {
      "entropy": 5.264405632019043,
      "epoch": 3.938731959341832,
      "grad_norm": 1.0625,
      "learning_rate": 0.00035319736032723905,
      "loss": 4.9361,
      "mean_token_accuracy": 0.2245262935757637,
      "num_tokens": 78291366.0,
      "step": 35070
    },
    {
      "entropy": 5.348208951950073,
      "epoch": 3.9392935362497896,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003531597130844616,
      "loss": 4.9428,
      "mean_token_accuracy": 0.21335278451442719,
      "num_tokens": 78301541.0,
      "step": 35075
    },
    {
      "entropy": 5.322513103485107,
      "epoch": 3.939855113157747,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00035312206335304843,
      "loss": 4.9374,
      "mean_token_accuracy": 0.22332670390605927,
      "num_tokens": 78311782.0,
      "step": 35080
    },
    {
      "entropy": 5.247590351104736,
      "epoch": 3.9404166900657045,
      "grad_norm": 1.109375,
      "learning_rate": 0.00035308441113419855,
      "loss": 4.8928,
      "mean_token_accuracy": 0.2160123035311699,
      "num_tokens": 78321636.0,
      "step": 35085
    },
    {
      "entropy": 5.394304847717285,
      "epoch": 3.940978266973662,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00035304675642911074,
      "loss": 4.9673,
      "mean_token_accuracy": 0.22498744279146193,
      "num_tokens": 78332882.0,
      "step": 35090
    },
    {
      "entropy": 5.395733594894409,
      "epoch": 3.9415398438816194,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003530090992389838,
      "loss": 5.0174,
      "mean_token_accuracy": 0.21799246221780777,
      "num_tokens": 78345122.0,
      "step": 35095
    },
    {
      "entropy": 5.3100762367248535,
      "epoch": 3.942101420789577,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000352971439565017,
      "loss": 4.9548,
      "mean_token_accuracy": 0.2235313758254051,
      "num_tokens": 78355730.0,
      "step": 35100
    },
    {
      "entropy": 5.2865949153900145,
      "epoch": 3.9426629976975347,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00035293377740840934,
      "loss": 4.8646,
      "mean_token_accuracy": 0.2274550825357437,
      "num_tokens": 78366651.0,
      "step": 35105
    },
    {
      "entropy": 5.275113248825074,
      "epoch": 3.9432245746054924,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00035289611277036,
      "loss": 4.8137,
      "mean_token_accuracy": 0.22858020216226577,
      "num_tokens": 78377633.0,
      "step": 35110
    },
    {
      "entropy": 5.2685136795043945,
      "epoch": 3.94378615151345,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003528584456520681,
      "loss": 4.892,
      "mean_token_accuracy": 0.22157983034849166,
      "num_tokens": 78389038.0,
      "step": 35115
    },
    {
      "entropy": 5.220474815368652,
      "epoch": 3.9443477284214072,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00035282077605473317,
      "loss": 4.9441,
      "mean_token_accuracy": 0.21887344419956206,
      "num_tokens": 78401045.0,
      "step": 35120
    },
    {
      "entropy": 5.305587387084961,
      "epoch": 3.944909305329365,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003527831039795545,
      "loss": 4.9633,
      "mean_token_accuracy": 0.21661159098148347,
      "num_tokens": 78413013.0,
      "step": 35125
    },
    {
      "entropy": 5.202114915847778,
      "epoch": 3.9454708822373226,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003527454294277316,
      "loss": 4.8458,
      "mean_token_accuracy": 0.23397735208272935,
      "num_tokens": 78423794.0,
      "step": 35130
    },
    {
      "entropy": 5.250099849700928,
      "epoch": 3.94603245914528,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003527077524004642,
      "loss": 4.8742,
      "mean_token_accuracy": 0.2296661838889122,
      "num_tokens": 78434902.0,
      "step": 35135
    },
    {
      "entropy": 5.288788604736328,
      "epoch": 3.9465940360532374,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035267007289895176,
      "loss": 4.8621,
      "mean_token_accuracy": 0.22578113824129104,
      "num_tokens": 78446114.0,
      "step": 35140
    },
    {
      "entropy": 5.250031805038452,
      "epoch": 3.947155612961195,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00035263239092439417,
      "loss": 4.8272,
      "mean_token_accuracy": 0.23507112860679627,
      "num_tokens": 78456882.0,
      "step": 35145
    },
    {
      "entropy": 5.263872957229614,
      "epoch": 3.9477171898691523,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00035259470647799104,
      "loss": 4.8311,
      "mean_token_accuracy": 0.2326715737581253,
      "num_tokens": 78467416.0,
      "step": 35150
    },
    {
      "entropy": 5.229687881469727,
      "epoch": 3.94827876677711,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00035255701956094236,
      "loss": 4.8333,
      "mean_token_accuracy": 0.23307756781578065,
      "num_tokens": 78477364.0,
      "step": 35155
    },
    {
      "entropy": 5.233487939834594,
      "epoch": 3.9488403436850676,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00035251933017444814,
      "loss": 4.8667,
      "mean_token_accuracy": 0.22628338187932967,
      "num_tokens": 78488396.0,
      "step": 35160
    },
    {
      "entropy": 5.261657285690307,
      "epoch": 3.9494019205930253,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035248163831970837,
      "loss": 4.8659,
      "mean_token_accuracy": 0.22224292308092117,
      "num_tokens": 78498586.0,
      "step": 35165
    },
    {
      "entropy": 5.328397178649903,
      "epoch": 3.949963497500983,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00035244394399792304,
      "loss": 4.8445,
      "mean_token_accuracy": 0.22888573110103608,
      "num_tokens": 78509664.0,
      "step": 35170
    },
    {
      "entropy": 5.424096918106079,
      "epoch": 3.95052507440894,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003524062472102927,
      "loss": 5.0665,
      "mean_token_accuracy": 0.21862461864948274,
      "num_tokens": 78521950.0,
      "step": 35175
    },
    {
      "entropy": 5.297977352142334,
      "epoch": 3.951086651316898,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003523685479580174,
      "loss": 4.9756,
      "mean_token_accuracy": 0.21882144659757613,
      "num_tokens": 78532943.0,
      "step": 35180
    },
    {
      "entropy": 5.206993150711059,
      "epoch": 3.9516482282248555,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003523308462422974,
      "loss": 4.7773,
      "mean_token_accuracy": 0.22868580222129822,
      "num_tokens": 78544042.0,
      "step": 35185
    },
    {
      "entropy": 5.216303968429566,
      "epoch": 3.9522098051328127,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003522931420643334,
      "loss": 4.819,
      "mean_token_accuracy": 0.22887776046991348,
      "num_tokens": 78554812.0,
      "step": 35190
    },
    {
      "entropy": 5.1714259624481205,
      "epoch": 3.9527713820407704,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003522554354253257,
      "loss": 4.8077,
      "mean_token_accuracy": 0.2290653333067894,
      "num_tokens": 78565036.0,
      "step": 35195
    },
    {
      "entropy": 5.16633152961731,
      "epoch": 3.953332958948728,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00035221772632647503,
      "loss": 4.7682,
      "mean_token_accuracy": 0.23225594609975814,
      "num_tokens": 78575420.0,
      "step": 35200
    },
    {
      "entropy": 5.37187819480896,
      "epoch": 3.9538945358566857,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00035218001476898194,
      "loss": 4.9355,
      "mean_token_accuracy": 0.22865281850099564,
      "num_tokens": 78586670.0,
      "step": 35205
    },
    {
      "entropy": 5.2274878978729244,
      "epoch": 3.9544561127646434,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003521423007540472,
      "loss": 4.815,
      "mean_token_accuracy": 0.22911221981048585,
      "num_tokens": 78598048.0,
      "step": 35210
    },
    {
      "entropy": 5.248955583572387,
      "epoch": 3.9550176896726006,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003521045842828718,
      "loss": 4.9237,
      "mean_token_accuracy": 0.22712473273277284,
      "num_tokens": 78610364.0,
      "step": 35215
    },
    {
      "entropy": 5.312531614303589,
      "epoch": 3.9555792665805583,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003520668653566565,
      "loss": 4.898,
      "mean_token_accuracy": 0.21671735793352126,
      "num_tokens": 78622117.0,
      "step": 35220
    },
    {
      "entropy": 5.2877068519592285,
      "epoch": 3.956140843488516,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003520291439766024,
      "loss": 4.8294,
      "mean_token_accuracy": 0.2269025906920433,
      "num_tokens": 78632894.0,
      "step": 35225
    },
    {
      "entropy": 5.274543285369873,
      "epoch": 3.956702420396473,
      "grad_norm": 0.984375,
      "learning_rate": 0.00035199142014391034,
      "loss": 4.8696,
      "mean_token_accuracy": 0.22920819818973542,
      "num_tokens": 78643851.0,
      "step": 35230
    },
    {
      "entropy": 5.278875827789307,
      "epoch": 3.957263997304431,
      "grad_norm": 1.0625,
      "learning_rate": 0.00035195369385978165,
      "loss": 4.916,
      "mean_token_accuracy": 0.21897276490926743,
      "num_tokens": 78655097.0,
      "step": 35235
    },
    {
      "entropy": 5.200615739822387,
      "epoch": 3.9578255742123885,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00035191596512541747,
      "loss": 4.8104,
      "mean_token_accuracy": 0.22750144451856613,
      "num_tokens": 78667572.0,
      "step": 35240
    },
    {
      "entropy": 5.344396734237671,
      "epoch": 3.9583871511203457,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003518782339420192,
      "loss": 4.9258,
      "mean_token_accuracy": 0.2212958127260208,
      "num_tokens": 78678858.0,
      "step": 35245
    },
    {
      "entropy": 5.247781419754029,
      "epoch": 3.9589487280283033,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00035184050031078815,
      "loss": 4.7457,
      "mean_token_accuracy": 0.23099232316017151,
      "num_tokens": 78688074.0,
      "step": 35250
    },
    {
      "entropy": 5.229494953155518,
      "epoch": 3.959510304936261,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035180276423292566,
      "loss": 4.8341,
      "mean_token_accuracy": 0.21939797848463058,
      "num_tokens": 78698805.0,
      "step": 35255
    },
    {
      "entropy": 5.141230964660645,
      "epoch": 3.9600718818442187,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003517650257096334,
      "loss": 4.7927,
      "mean_token_accuracy": 0.22836097627878188,
      "num_tokens": 78710636.0,
      "step": 35260
    },
    {
      "entropy": 5.209285593032837,
      "epoch": 3.9606334587521763,
      "grad_norm": 1.15625,
      "learning_rate": 0.00035172728474211306,
      "loss": 4.8829,
      "mean_token_accuracy": 0.2209864377975464,
      "num_tokens": 78721464.0,
      "step": 35265
    },
    {
      "entropy": 5.312131452560425,
      "epoch": 3.9611950356601335,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00035168954133156613,
      "loss": 4.8557,
      "mean_token_accuracy": 0.22617655694484712,
      "num_tokens": 78733306.0,
      "step": 35270
    },
    {
      "entropy": 5.342561197280884,
      "epoch": 3.961756612568091,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003516517954791944,
      "loss": 4.9938,
      "mean_token_accuracy": 0.21878737062215806,
      "num_tokens": 78745254.0,
      "step": 35275
    },
    {
      "entropy": 5.240498161315918,
      "epoch": 3.962318189476049,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00035161404718619986,
      "loss": 4.8199,
      "mean_token_accuracy": 0.23039366900920868,
      "num_tokens": 78755796.0,
      "step": 35280
    },
    {
      "entropy": 5.264400625228882,
      "epoch": 3.962879766384006,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003515762964537844,
      "loss": 4.8846,
      "mean_token_accuracy": 0.22190546244382858,
      "num_tokens": 78767394.0,
      "step": 35285
    },
    {
      "entropy": 5.3404810428619385,
      "epoch": 3.9634413432919637,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003515385432831499,
      "loss": 4.9291,
      "mean_token_accuracy": 0.232056824862957,
      "num_tokens": 78779017.0,
      "step": 35290
    },
    {
      "entropy": 5.329929256439209,
      "epoch": 3.9640029201999214,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003515007876754984,
      "loss": 4.9202,
      "mean_token_accuracy": 0.22034894227981566,
      "num_tokens": 78789964.0,
      "step": 35295
    },
    {
      "entropy": 5.267889070510864,
      "epoch": 3.964564497107879,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003514630296320323,
      "loss": 4.9039,
      "mean_token_accuracy": 0.22594827264547349,
      "num_tokens": 78801657.0,
      "step": 35300
    },
    {
      "entropy": 5.345570468902588,
      "epoch": 3.9651260740158367,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003514252691539536,
      "loss": 5.0247,
      "mean_token_accuracy": 0.22313400954008103,
      "num_tokens": 78812396.0,
      "step": 35305
    },
    {
      "entropy": 5.277970647811889,
      "epoch": 3.965687650923794,
      "grad_norm": 0.96875,
      "learning_rate": 0.00035138750624246466,
      "loss": 4.8888,
      "mean_token_accuracy": 0.21861737072467805,
      "num_tokens": 78823711.0,
      "step": 35310
    },
    {
      "entropy": 5.310520267486572,
      "epoch": 3.9662492278317516,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000351349740898768,
      "loss": 4.9343,
      "mean_token_accuracy": 0.2222214624285698,
      "num_tokens": 78835394.0,
      "step": 35315
    },
    {
      "entropy": 5.255678510665893,
      "epoch": 3.9668108047397093,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003513119731240659,
      "loss": 4.8713,
      "mean_token_accuracy": 0.23269438594579697,
      "num_tokens": 78846180.0,
      "step": 35320
    },
    {
      "entropy": 5.234151411056518,
      "epoch": 3.9673723816476665,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00035127420291956104,
      "loss": 4.8638,
      "mean_token_accuracy": 0.2286669507622719,
      "num_tokens": 78856537.0,
      "step": 35325
    },
    {
      "entropy": 5.304278230667114,
      "epoch": 3.967933958555624,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003512364302864559,
      "loss": 4.8984,
      "mean_token_accuracy": 0.2221909463405609,
      "num_tokens": 78867416.0,
      "step": 35330
    },
    {
      "entropy": 5.292031908035279,
      "epoch": 3.968495535463582,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003511986552259533,
      "loss": 4.8838,
      "mean_token_accuracy": 0.22251032739877702,
      "num_tokens": 78878725.0,
      "step": 35335
    },
    {
      "entropy": 5.27224817276001,
      "epoch": 3.969057112371539,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00035116087773925594,
      "loss": 4.894,
      "mean_token_accuracy": 0.22236318290233612,
      "num_tokens": 78889659.0,
      "step": 35340
    },
    {
      "entropy": 5.199196910858154,
      "epoch": 3.9696186892794967,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003511230978275667,
      "loss": 4.7656,
      "mean_token_accuracy": 0.23162130415439605,
      "num_tokens": 78900056.0,
      "step": 35345
    },
    {
      "entropy": 5.261955404281617,
      "epoch": 3.9701802661874543,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035108531549208846,
      "loss": 4.8774,
      "mean_token_accuracy": 0.22612771093845369,
      "num_tokens": 78911343.0,
      "step": 35350
    },
    {
      "entropy": 5.2976460456848145,
      "epoch": 3.970741843095412,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035104753073402433,
      "loss": 4.8522,
      "mean_token_accuracy": 0.22684813886880875,
      "num_tokens": 78921580.0,
      "step": 35355
    },
    {
      "entropy": 5.272527360916138,
      "epoch": 3.9713034200033697,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003510097435545774,
      "loss": 4.8797,
      "mean_token_accuracy": 0.22862574607133865,
      "num_tokens": 78932810.0,
      "step": 35360
    },
    {
      "entropy": 5.238641929626465,
      "epoch": 3.971864996911327,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00035097195395495057,
      "loss": 4.8627,
      "mean_token_accuracy": 0.22735609263181686,
      "num_tokens": 78945100.0,
      "step": 35365
    },
    {
      "entropy": 5.323822689056397,
      "epoch": 3.9724265738192845,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003509341619363474,
      "loss": 4.9018,
      "mean_token_accuracy": 0.22280891984701157,
      "num_tokens": 78957284.0,
      "step": 35370
    },
    {
      "entropy": 5.353881597518921,
      "epoch": 3.972988150727242,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00035089636749997095,
      "loss": 4.8943,
      "mean_token_accuracy": 0.23191314041614533,
      "num_tokens": 78968639.0,
      "step": 35375
    },
    {
      "entropy": 5.199768495559693,
      "epoch": 3.9735497276351994,
      "grad_norm": 1.078125,
      "learning_rate": 0.00035085857064702487,
      "loss": 4.8046,
      "mean_token_accuracy": 0.229317007958889,
      "num_tokens": 78978323.0,
      "step": 35380
    },
    {
      "entropy": 5.256274843215943,
      "epoch": 3.974111304543157,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003508207713787123,
      "loss": 4.9252,
      "mean_token_accuracy": 0.22157416492700577,
      "num_tokens": 78989193.0,
      "step": 35385
    },
    {
      "entropy": 5.228826570510864,
      "epoch": 3.9746728814511147,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00035078296969623716,
      "loss": 4.7503,
      "mean_token_accuracy": 0.2356490045785904,
      "num_tokens": 78999640.0,
      "step": 35390
    },
    {
      "entropy": 5.194830369949341,
      "epoch": 3.9752344583590724,
      "grad_norm": 0.984375,
      "learning_rate": 0.00035074516560080273,
      "loss": 4.7994,
      "mean_token_accuracy": 0.23463514894247056,
      "num_tokens": 79010468.0,
      "step": 35395
    },
    {
      "entropy": 5.240610408782959,
      "epoch": 3.97579603526703,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003507073590936129,
      "loss": 4.909,
      "mean_token_accuracy": 0.21869378834962844,
      "num_tokens": 79021399.0,
      "step": 35400
    },
    {
      "entropy": 5.242806673049927,
      "epoch": 3.9763576121749873,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00035066955017587135,
      "loss": 4.8288,
      "mean_token_accuracy": 0.2229751780629158,
      "num_tokens": 79031643.0,
      "step": 35405
    },
    {
      "entropy": 5.40739483833313,
      "epoch": 3.976919189082945,
      "grad_norm": 1.015625,
      "learning_rate": 0.000350631738848782,
      "loss": 5.033,
      "mean_token_accuracy": 0.21740177422761917,
      "num_tokens": 79041911.0,
      "step": 35410
    },
    {
      "entropy": 5.351735162734985,
      "epoch": 3.9774807659909026,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003505939251135488,
      "loss": 5.0078,
      "mean_token_accuracy": 0.21084133684635162,
      "num_tokens": 79054698.0,
      "step": 35415
    },
    {
      "entropy": 5.398605632781982,
      "epoch": 3.97804234289886,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003505561089713756,
      "loss": 5.0123,
      "mean_token_accuracy": 0.20977278500795365,
      "num_tokens": 79065855.0,
      "step": 35420
    },
    {
      "entropy": 5.409931707382202,
      "epoch": 3.9786039198068175,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003505182904234667,
      "loss": 4.9797,
      "mean_token_accuracy": 0.21914974898099898,
      "num_tokens": 79077332.0,
      "step": 35425
    },
    {
      "entropy": 5.3049064636230465,
      "epoch": 3.979165496714775,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003504804694710261,
      "loss": 4.914,
      "mean_token_accuracy": 0.22370981127023698,
      "num_tokens": 79088996.0,
      "step": 35430
    },
    {
      "entropy": 5.260089302062989,
      "epoch": 3.9797270736227324,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003504426461152581,
      "loss": 4.9191,
      "mean_token_accuracy": 0.22471291124820708,
      "num_tokens": 79100828.0,
      "step": 35435
    },
    {
      "entropy": 5.214281272888184,
      "epoch": 3.98028865053069,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003504048203573669,
      "loss": 4.8457,
      "mean_token_accuracy": 0.22939778715372086,
      "num_tokens": 79112297.0,
      "step": 35440
    },
    {
      "entropy": 5.274517059326172,
      "epoch": 3.9808502274386477,
      "grad_norm": 1.125,
      "learning_rate": 0.0003503669921985571,
      "loss": 4.895,
      "mean_token_accuracy": 0.22336256206035615,
      "num_tokens": 79122795.0,
      "step": 35445
    },
    {
      "entropy": 5.2721751689910885,
      "epoch": 3.9814118043466054,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000350329161640033,
      "loss": 4.8699,
      "mean_token_accuracy": 0.2266692712903023,
      "num_tokens": 79133652.0,
      "step": 35450
    },
    {
      "entropy": 5.279295921325684,
      "epoch": 3.981973381254563,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00035029132868299914,
      "loss": 4.8536,
      "mean_token_accuracy": 0.22911391407251358,
      "num_tokens": 79144107.0,
      "step": 35455
    },
    {
      "entropy": 5.240471315383911,
      "epoch": 3.9825349581625202,
      "grad_norm": 0.953125,
      "learning_rate": 0.00035025349332866017,
      "loss": 4.8746,
      "mean_token_accuracy": 0.22571857422590255,
      "num_tokens": 79155900.0,
      "step": 35460
    },
    {
      "entropy": 5.2921223640441895,
      "epoch": 3.983096535070478,
      "grad_norm": 0.921875,
      "learning_rate": 0.00035021565557822085,
      "loss": 4.9228,
      "mean_token_accuracy": 0.22713594287633895,
      "num_tokens": 79167425.0,
      "step": 35465
    },
    {
      "entropy": 5.331947755813599,
      "epoch": 3.9836581119784356,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003501778154328858,
      "loss": 4.9636,
      "mean_token_accuracy": 0.22355202883481978,
      "num_tokens": 79178880.0,
      "step": 35470
    },
    {
      "entropy": 5.265964078903198,
      "epoch": 3.9842196888863928,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003501399728938599,
      "loss": 4.8232,
      "mean_token_accuracy": 0.23459725379943847,
      "num_tokens": 79190210.0,
      "step": 35475
    },
    {
      "entropy": 5.25033860206604,
      "epoch": 3.9847812657943504,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035010212796234827,
      "loss": 4.8524,
      "mean_token_accuracy": 0.2296846777200699,
      "num_tokens": 79200497.0,
      "step": 35480
    },
    {
      "entropy": 5.230093145370484,
      "epoch": 3.985342842702308,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003500642806395557,
      "loss": 4.8816,
      "mean_token_accuracy": 0.23018314093351364,
      "num_tokens": 79212305.0,
      "step": 35485
    },
    {
      "entropy": 5.31821928024292,
      "epoch": 3.9859044196102658,
      "grad_norm": 1.0,
      "learning_rate": 0.0003500264309266873,
      "loss": 4.8785,
      "mean_token_accuracy": 0.22232018560171127,
      "num_tokens": 79222968.0,
      "step": 35490
    },
    {
      "entropy": 5.295018815994263,
      "epoch": 3.9864659965182234,
      "grad_norm": 1.0,
      "learning_rate": 0.00034998857882494823,
      "loss": 4.9304,
      "mean_token_accuracy": 0.22128220051527023,
      "num_tokens": 79235064.0,
      "step": 35495
    },
    {
      "entropy": 5.328885698318482,
      "epoch": 3.9870275734261806,
      "grad_norm": 1.0,
      "learning_rate": 0.00034995072433554377,
      "loss": 4.8682,
      "mean_token_accuracy": 0.22633276283740997,
      "num_tokens": 79245761.0,
      "step": 35500
    },
    {
      "entropy": 5.267241668701172,
      "epoch": 3.9875891503341383,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003499128674596792,
      "loss": 4.8153,
      "mean_token_accuracy": 0.23143135756254196,
      "num_tokens": 79256070.0,
      "step": 35505
    },
    {
      "entropy": 5.278869771957398,
      "epoch": 3.988150727242096,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00034987500819855976,
      "loss": 4.9188,
      "mean_token_accuracy": 0.2184286132454872,
      "num_tokens": 79266322.0,
      "step": 35510
    },
    {
      "entropy": 5.331770753860473,
      "epoch": 3.988712304150053,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00034983714655339115,
      "loss": 4.9521,
      "mean_token_accuracy": 0.22554160058498382,
      "num_tokens": 79277272.0,
      "step": 35515
    },
    {
      "entropy": 5.330368947982788,
      "epoch": 3.989273881058011,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003497992825253787,
      "loss": 4.9241,
      "mean_token_accuracy": 0.2210568070411682,
      "num_tokens": 79288463.0,
      "step": 35520
    },
    {
      "entropy": 5.312254953384399,
      "epoch": 3.9898354579659685,
      "grad_norm": 1.03125,
      "learning_rate": 0.00034976141611572817,
      "loss": 4.8696,
      "mean_token_accuracy": 0.2269513264298439,
      "num_tokens": 79298655.0,
      "step": 35525
    },
    {
      "entropy": 5.212372303009033,
      "epoch": 3.9903970348739257,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034972354732564506,
      "loss": 4.8485,
      "mean_token_accuracy": 0.23272888362407684,
      "num_tokens": 79309731.0,
      "step": 35530
    },
    {
      "entropy": 5.2812840938568115,
      "epoch": 3.9909586117818834,
      "grad_norm": 0.90625,
      "learning_rate": 0.00034968567615633516,
      "loss": 4.9218,
      "mean_token_accuracy": 0.22649964094161987,
      "num_tokens": 79321712.0,
      "step": 35535
    },
    {
      "entropy": 5.239543056488037,
      "epoch": 3.991520188689841,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00034964780260900445,
      "loss": 4.8518,
      "mean_token_accuracy": 0.23161873668432237,
      "num_tokens": 79333159.0,
      "step": 35540
    },
    {
      "entropy": 5.335504102706909,
      "epoch": 3.9920817655977987,
      "grad_norm": 1.015625,
      "learning_rate": 0.00034960992668485875,
      "loss": 4.8921,
      "mean_token_accuracy": 0.22722331434488297,
      "num_tokens": 79344012.0,
      "step": 35545
    },
    {
      "entropy": 5.297263479232788,
      "epoch": 3.9926433425057564,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00034957204838510406,
      "loss": 4.9466,
      "mean_token_accuracy": 0.2229917511343956,
      "num_tokens": 79355162.0,
      "step": 35550
    },
    {
      "entropy": 5.258846664428711,
      "epoch": 3.9932049194137136,
      "grad_norm": 1.015625,
      "learning_rate": 0.00034953416771094645,
      "loss": 4.8007,
      "mean_token_accuracy": 0.23033169209957122,
      "num_tokens": 79365605.0,
      "step": 35555
    },
    {
      "entropy": 5.230972337722778,
      "epoch": 3.9937664963216712,
      "grad_norm": 0.953125,
      "learning_rate": 0.000349496284663592,
      "loss": 4.879,
      "mean_token_accuracy": 0.2231452226638794,
      "num_tokens": 79376548.0,
      "step": 35560
    },
    {
      "entropy": 5.2674134254455565,
      "epoch": 3.994328073229629,
      "grad_norm": 1.078125,
      "learning_rate": 0.00034945839924424696,
      "loss": 4.8121,
      "mean_token_accuracy": 0.233462917804718,
      "num_tokens": 79388352.0,
      "step": 35565
    },
    {
      "entropy": 5.2701154232025145,
      "epoch": 3.994889650137586,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003494205114541176,
      "loss": 4.8629,
      "mean_token_accuracy": 0.22835576832294463,
      "num_tokens": 79399110.0,
      "step": 35570
    },
    {
      "entropy": 5.2828700065612795,
      "epoch": 3.995451227045544,
      "grad_norm": 1.03125,
      "learning_rate": 0.00034938262129441026,
      "loss": 4.9651,
      "mean_token_accuracy": 0.22055698037147523,
      "num_tokens": 79410350.0,
      "step": 35575
    },
    {
      "entropy": 5.279828023910523,
      "epoch": 3.9960128039535014,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003493447287663315,
      "loss": 4.8007,
      "mean_token_accuracy": 0.22350062280893326,
      "num_tokens": 79421015.0,
      "step": 35580
    },
    {
      "entropy": 5.297459506988526,
      "epoch": 3.996574380861459,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003493068338710877,
      "loss": 4.9596,
      "mean_token_accuracy": 0.22601325660943986,
      "num_tokens": 79432200.0,
      "step": 35585
    },
    {
      "entropy": 5.329850196838379,
      "epoch": 3.9971359577694168,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003492689366098855,
      "loss": 5.0024,
      "mean_token_accuracy": 0.2186721757054329,
      "num_tokens": 79445121.0,
      "step": 35590
    },
    {
      "entropy": 5.26259298324585,
      "epoch": 3.997697534677374,
      "grad_norm": 0.953125,
      "learning_rate": 0.00034923103698393147,
      "loss": 4.7913,
      "mean_token_accuracy": 0.22776474058628082,
      "num_tokens": 79456533.0,
      "step": 35595
    },
    {
      "entropy": 5.239823579788208,
      "epoch": 3.9982591115853316,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003491931349944325,
      "loss": 4.8199,
      "mean_token_accuracy": 0.2259430021047592,
      "num_tokens": 79466365.0,
      "step": 35600
    },
    {
      "entropy": 5.329803514480591,
      "epoch": 3.9988206884932893,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00034915523064259533,
      "loss": 4.9821,
      "mean_token_accuracy": 0.2149461567401886,
      "num_tokens": 79477203.0,
      "step": 35605
    },
    {
      "entropy": 5.2061601161956785,
      "epoch": 3.9993822654012465,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003491173239296268,
      "loss": 4.8677,
      "mean_token_accuracy": 0.2251120239496231,
      "num_tokens": 79490062.0,
      "step": 35610
    },
    {
      "entropy": 5.34804253578186,
      "epoch": 3.999943842309204,
      "grad_norm": 1.03125,
      "learning_rate": 0.000349079414856734,
      "loss": 4.9403,
      "mean_token_accuracy": 0.22231367379426956,
      "num_tokens": 79501727.0,
      "step": 35615
    },
    {
      "entropy": 5.217591073777941,
      "epoch": 4.000449261526366,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003490415034251239,
      "loss": 4.7002,
      "mean_token_accuracy": 0.2392729851934645,
      "num_tokens": 79510304.0,
      "step": 35620
    },
    {
      "entropy": 5.303219366073608,
      "epoch": 4.001010838434324,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003490035896360036,
      "loss": 4.85,
      "mean_token_accuracy": 0.2267559975385666,
      "num_tokens": 79521811.0,
      "step": 35625
    },
    {
      "entropy": 5.235414743423462,
      "epoch": 4.001572415342281,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00034896567349058027,
      "loss": 4.7318,
      "mean_token_accuracy": 0.24330737590789794,
      "num_tokens": 79532756.0,
      "step": 35630
    },
    {
      "entropy": 5.339402484893799,
      "epoch": 4.002133992250239,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00034892775499006126,
      "loss": 4.8308,
      "mean_token_accuracy": 0.2241450220346451,
      "num_tokens": 79544945.0,
      "step": 35635
    },
    {
      "entropy": 5.311671161651612,
      "epoch": 4.002695569158196,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003488898341356538,
      "loss": 4.9056,
      "mean_token_accuracy": 0.22416615039110183,
      "num_tokens": 79555374.0,
      "step": 35640
    },
    {
      "entropy": 5.275216579437256,
      "epoch": 4.0032571460661535,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00034885191092856545,
      "loss": 4.8055,
      "mean_token_accuracy": 0.23268196731805801,
      "num_tokens": 79566598.0,
      "step": 35645
    },
    {
      "entropy": 5.3141562938690186,
      "epoch": 4.003818722974112,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003488139853700036,
      "loss": 4.7935,
      "mean_token_accuracy": 0.2351074233651161,
      "num_tokens": 79577859.0,
      "step": 35650
    },
    {
      "entropy": 5.257867002487183,
      "epoch": 4.004380299882069,
      "grad_norm": 1.03125,
      "learning_rate": 0.00034877605746117576,
      "loss": 4.793,
      "mean_token_accuracy": 0.2299335092306137,
      "num_tokens": 79588704.0,
      "step": 35655
    },
    {
      "entropy": 5.130130863189697,
      "epoch": 4.004941876790026,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003487381272032897,
      "loss": 4.6404,
      "mean_token_accuracy": 0.24927979856729507,
      "num_tokens": 79599493.0,
      "step": 35660
    },
    {
      "entropy": 5.275761079788208,
      "epoch": 4.005503453697984,
      "grad_norm": 0.9375,
      "learning_rate": 0.000348700194597553,
      "loss": 4.8072,
      "mean_token_accuracy": 0.2326055258512497,
      "num_tokens": 79610305.0,
      "step": 35665
    },
    {
      "entropy": 5.346838521957397,
      "epoch": 4.006065030605941,
      "grad_norm": 0.890625,
      "learning_rate": 0.00034866225964517354,
      "loss": 4.9083,
      "mean_token_accuracy": 0.22570579200983049,
      "num_tokens": 79622077.0,
      "step": 35670
    },
    {
      "entropy": 5.361871767044067,
      "epoch": 4.006626607513899,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003486243223473592,
      "loss": 4.9331,
      "mean_token_accuracy": 0.22300866097211838,
      "num_tokens": 79634999.0,
      "step": 35675
    },
    {
      "entropy": 5.2221814632415775,
      "epoch": 4.007188184421857,
      "grad_norm": 1.0,
      "learning_rate": 0.00034858638270531786,
      "loss": 4.7204,
      "mean_token_accuracy": 0.23645230680704116,
      "num_tokens": 79646152.0,
      "step": 35680
    },
    {
      "entropy": 5.268419647216797,
      "epoch": 4.007749761329814,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00034854844072025755,
      "loss": 4.8406,
      "mean_token_accuracy": 0.23397001028060913,
      "num_tokens": 79657941.0,
      "step": 35685
    },
    {
      "entropy": 5.271525764465332,
      "epoch": 4.008311338237772,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00034851049639338637,
      "loss": 4.7205,
      "mean_token_accuracy": 0.2404434621334076,
      "num_tokens": 79667713.0,
      "step": 35690
    },
    {
      "entropy": 5.2748113632202145,
      "epoch": 4.008872915145729,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003484725497259125,
      "loss": 4.7848,
      "mean_token_accuracy": 0.23185040205717086,
      "num_tokens": 79679351.0,
      "step": 35695
    },
    {
      "entropy": 5.302718639373779,
      "epoch": 4.009434492053686,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034843460071904407,
      "loss": 4.9041,
      "mean_token_accuracy": 0.22310521602630615,
      "num_tokens": 79689406.0,
      "step": 35700
    },
    {
      "entropy": 5.231268072128296,
      "epoch": 4.0099960689616445,
      "grad_norm": 1.0625,
      "learning_rate": 0.00034839664937398964,
      "loss": 4.7235,
      "mean_token_accuracy": 0.2316025897860527,
      "num_tokens": 79700231.0,
      "step": 35705
    },
    {
      "entropy": 5.2478611946105955,
      "epoch": 4.010557645869602,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003483586956919572,
      "loss": 4.8101,
      "mean_token_accuracy": 0.23243284821510315,
      "num_tokens": 79711721.0,
      "step": 35710
    },
    {
      "entropy": 5.1654926300048825,
      "epoch": 4.011119222777559,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003483207396741556,
      "loss": 4.693,
      "mean_token_accuracy": 0.2347804754972458,
      "num_tokens": 79721795.0,
      "step": 35715
    },
    {
      "entropy": 5.348480319976806,
      "epoch": 4.011680799685517,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003482827813217932,
      "loss": 4.9232,
      "mean_token_accuracy": 0.22637430727481841,
      "num_tokens": 79734356.0,
      "step": 35720
    },
    {
      "entropy": 5.236114120483398,
      "epoch": 4.012242376593474,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00034824482063607856,
      "loss": 4.7751,
      "mean_token_accuracy": 0.23564398139715195,
      "num_tokens": 79745847.0,
      "step": 35725
    },
    {
      "entropy": 5.219485521316528,
      "epoch": 4.012803953501432,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003482068576182204,
      "loss": 4.8333,
      "mean_token_accuracy": 0.23120176941156387,
      "num_tokens": 79756580.0,
      "step": 35730
    },
    {
      "entropy": 5.300530672073364,
      "epoch": 4.01336553040939,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034816889226942755,
      "loss": 4.8694,
      "mean_token_accuracy": 0.23020175844430923,
      "num_tokens": 79767912.0,
      "step": 35735
    },
    {
      "entropy": 5.339720344543457,
      "epoch": 4.013927107317347,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003481309245909088,
      "loss": 4.9053,
      "mean_token_accuracy": 0.22861750572919845,
      "num_tokens": 79778572.0,
      "step": 35740
    },
    {
      "entropy": 5.274091720581055,
      "epoch": 4.014488684225305,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003480929545838731,
      "loss": 4.769,
      "mean_token_accuracy": 0.24010961204767228,
      "num_tokens": 79789237.0,
      "step": 35745
    },
    {
      "entropy": 5.335403871536255,
      "epoch": 4.015050261133262,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003480549822495293,
      "loss": 4.9747,
      "mean_token_accuracy": 0.22313235998153685,
      "num_tokens": 79800271.0,
      "step": 35750
    },
    {
      "entropy": 5.3076653480529785,
      "epoch": 4.015611838041219,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00034801700758908653,
      "loss": 4.837,
      "mean_token_accuracy": 0.22731812447309493,
      "num_tokens": 79811378.0,
      "step": 35755
    },
    {
      "entropy": 5.29190673828125,
      "epoch": 4.0161734149491775,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00034797903060375396,
      "loss": 4.8696,
      "mean_token_accuracy": 0.22705837935209275,
      "num_tokens": 79822795.0,
      "step": 35760
    },
    {
      "entropy": 5.313378477096558,
      "epoch": 4.016734991857135,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003479410512947407,
      "loss": 4.8729,
      "mean_token_accuracy": 0.22479652017354965,
      "num_tokens": 79833331.0,
      "step": 35765
    },
    {
      "entropy": 5.2206292152404785,
      "epoch": 4.017296568765093,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00034790306966325606,
      "loss": 4.7628,
      "mean_token_accuracy": 0.23483756482601165,
      "num_tokens": 79844079.0,
      "step": 35770
    },
    {
      "entropy": 5.255379009246826,
      "epoch": 4.01785814567305,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003478650857105095,
      "loss": 4.8153,
      "mean_token_accuracy": 0.23156358748674394,
      "num_tokens": 79855275.0,
      "step": 35775
    },
    {
      "entropy": 5.297036075592041,
      "epoch": 4.018419722581007,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003478270994377102,
      "loss": 4.8188,
      "mean_token_accuracy": 0.224640154838562,
      "num_tokens": 79866147.0,
      "step": 35780
    },
    {
      "entropy": 5.294878244400024,
      "epoch": 4.018981299488965,
      "grad_norm": 1.0625,
      "learning_rate": 0.00034778911084606793,
      "loss": 4.7894,
      "mean_token_accuracy": 0.2292419046163559,
      "num_tokens": 79876791.0,
      "step": 35785
    },
    {
      "entropy": 5.261782932281494,
      "epoch": 4.0195428763969225,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003477511199367922,
      "loss": 4.7944,
      "mean_token_accuracy": 0.23321970850229262,
      "num_tokens": 79887817.0,
      "step": 35790
    },
    {
      "entropy": 5.195351457595825,
      "epoch": 4.02010445330488,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00034771312671109256,
      "loss": 4.7868,
      "mean_token_accuracy": 0.22593356817960739,
      "num_tokens": 79899055.0,
      "step": 35795
    },
    {
      "entropy": 5.243024396896362,
      "epoch": 4.020666030212838,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003476751311701787,
      "loss": 4.7968,
      "mean_token_accuracy": 0.23253573328256608,
      "num_tokens": 79912066.0,
      "step": 35800
    },
    {
      "entropy": 5.372771263122559,
      "epoch": 4.021227607120795,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003476371333152605,
      "loss": 4.8963,
      "mean_token_accuracy": 0.22740176618099212,
      "num_tokens": 79922845.0,
      "step": 35805
    },
    {
      "entropy": 5.3195716381073,
      "epoch": 4.021789184028753,
      "grad_norm": 0.984375,
      "learning_rate": 0.00034759913314754776,
      "loss": 4.8575,
      "mean_token_accuracy": 0.22719766497611998,
      "num_tokens": 79934732.0,
      "step": 35810
    },
    {
      "entropy": 5.3377461433410645,
      "epoch": 4.02235076093671,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00034756113066825045,
      "loss": 4.8585,
      "mean_token_accuracy": 0.23113021105527878,
      "num_tokens": 79945182.0,
      "step": 35815
    },
    {
      "entropy": 5.264257192611694,
      "epoch": 4.022912337844668,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003475231258785787,
      "loss": 4.8932,
      "mean_token_accuracy": 0.2232726901769638,
      "num_tokens": 79956929.0,
      "step": 35820
    },
    {
      "entropy": 5.264922714233398,
      "epoch": 4.023473914752626,
      "grad_norm": 1.0,
      "learning_rate": 0.0003474851187797424,
      "loss": 4.8196,
      "mean_token_accuracy": 0.22456487119197846,
      "num_tokens": 79968361.0,
      "step": 35825
    },
    {
      "entropy": 5.286408996582031,
      "epoch": 4.024035491660583,
      "grad_norm": 1.015625,
      "learning_rate": 0.00034744710937295185,
      "loss": 4.8586,
      "mean_token_accuracy": 0.22477383762598038,
      "num_tokens": 79978833.0,
      "step": 35830
    },
    {
      "entropy": 5.246161222457886,
      "epoch": 4.02459706856854,
      "grad_norm": 1.015625,
      "learning_rate": 0.00034740909765941725,
      "loss": 4.8304,
      "mean_token_accuracy": 0.2265417143702507,
      "num_tokens": 79989997.0,
      "step": 35835
    },
    {
      "entropy": 5.189162874221802,
      "epoch": 4.025158645476498,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003473710836403489,
      "loss": 4.8072,
      "mean_token_accuracy": 0.22408731281757355,
      "num_tokens": 80001218.0,
      "step": 35840
    },
    {
      "entropy": 5.395273065567016,
      "epoch": 4.0257202223844555,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003473330673169572,
      "loss": 4.9637,
      "mean_token_accuracy": 0.22288746237754822,
      "num_tokens": 80013906.0,
      "step": 35845
    },
    {
      "entropy": 5.305909729003906,
      "epoch": 4.026281799292413,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003472950486904525,
      "loss": 4.8331,
      "mean_token_accuracy": 0.2203275665640831,
      "num_tokens": 80026384.0,
      "step": 35850
    },
    {
      "entropy": 5.255284786224365,
      "epoch": 4.026843376200371,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003472570277620455,
      "loss": 4.7698,
      "mean_token_accuracy": 0.23682335913181304,
      "num_tokens": 80037718.0,
      "step": 35855
    },
    {
      "entropy": 5.2675200462341305,
      "epoch": 4.027404953108328,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003472190045329466,
      "loss": 4.8599,
      "mean_token_accuracy": 0.2303346037864685,
      "num_tokens": 80049081.0,
      "step": 35860
    },
    {
      "entropy": 5.208883476257324,
      "epoch": 4.027966530016286,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003471809790043667,
      "loss": 4.7767,
      "mean_token_accuracy": 0.23235308825969697,
      "num_tokens": 80060435.0,
      "step": 35865
    },
    {
      "entropy": 5.317750740051269,
      "epoch": 4.028528106924243,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00034714295117751635,
      "loss": 4.8165,
      "mean_token_accuracy": 0.22974136024713515,
      "num_tokens": 80070657.0,
      "step": 35870
    },
    {
      "entropy": 5.240705442428589,
      "epoch": 4.029089683832201,
      "grad_norm": 0.984375,
      "learning_rate": 0.00034710492105360653,
      "loss": 4.7382,
      "mean_token_accuracy": 0.23876873403787613,
      "num_tokens": 80080265.0,
      "step": 35875
    },
    {
      "entropy": 5.273084545135498,
      "epoch": 4.029651260740159,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003470668886338481,
      "loss": 4.8101,
      "mean_token_accuracy": 0.23165100067853928,
      "num_tokens": 80091917.0,
      "step": 35880
    },
    {
      "entropy": 5.21268253326416,
      "epoch": 4.030212837648116,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00034702885391945184,
      "loss": 4.8342,
      "mean_token_accuracy": 0.22930820435285568,
      "num_tokens": 80103919.0,
      "step": 35885
    },
    {
      "entropy": 5.334311389923096,
      "epoch": 4.030774414556073,
      "grad_norm": 0.921875,
      "learning_rate": 0.000346990816911629,
      "loss": 4.8466,
      "mean_token_accuracy": 0.22880687415599824,
      "num_tokens": 80115926.0,
      "step": 35890
    },
    {
      "entropy": 5.278008460998535,
      "epoch": 4.031335991464031,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003469527776115906,
      "loss": 4.7626,
      "mean_token_accuracy": 0.23327454626560212,
      "num_tokens": 80127007.0,
      "step": 35895
    },
    {
      "entropy": 5.215300369262695,
      "epoch": 4.031897568371988,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034691473602054793,
      "loss": 4.8115,
      "mean_token_accuracy": 0.22583246380090713,
      "num_tokens": 80138442.0,
      "step": 35900
    },
    {
      "entropy": 5.28994460105896,
      "epoch": 4.032459145279946,
      "grad_norm": 1.0,
      "learning_rate": 0.0003468766921397121,
      "loss": 4.8374,
      "mean_token_accuracy": 0.2276547968387604,
      "num_tokens": 80148395.0,
      "step": 35905
    },
    {
      "entropy": 5.273488569259643,
      "epoch": 4.033020722187904,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003468386459702946,
      "loss": 4.8242,
      "mean_token_accuracy": 0.22503122985363005,
      "num_tokens": 80159348.0,
      "step": 35910
    },
    {
      "entropy": 5.31574068069458,
      "epoch": 4.033582299095861,
      "grad_norm": 1.015625,
      "learning_rate": 0.00034680059751350664,
      "loss": 4.8116,
      "mean_token_accuracy": 0.23480495661497117,
      "num_tokens": 80169772.0,
      "step": 35915
    },
    {
      "entropy": 5.259739112854004,
      "epoch": 4.034143876003819,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00034676254677055984,
      "loss": 4.7918,
      "mean_token_accuracy": 0.23463689237833024,
      "num_tokens": 80180538.0,
      "step": 35920
    },
    {
      "entropy": 5.326104927062988,
      "epoch": 4.034705452911776,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00034672449374266575,
      "loss": 4.9843,
      "mean_token_accuracy": 0.2129383787512779,
      "num_tokens": 80191693.0,
      "step": 35925
    },
    {
      "entropy": 5.251544952392578,
      "epoch": 4.0352670298197335,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003466864384310359,
      "loss": 4.7708,
      "mean_token_accuracy": 0.232445427775383,
      "num_tokens": 80201923.0,
      "step": 35930
    },
    {
      "entropy": 5.316903352737427,
      "epoch": 4.035828606727692,
      "grad_norm": 0.953125,
      "learning_rate": 0.00034664838083688223,
      "loss": 4.7661,
      "mean_token_accuracy": 0.23276143670082092,
      "num_tokens": 80213588.0,
      "step": 35935
    },
    {
      "entropy": 5.2479774951934814,
      "epoch": 4.036390183635649,
      "grad_norm": 1.078125,
      "learning_rate": 0.00034661032096141615,
      "loss": 4.79,
      "mean_token_accuracy": 0.23045438528060913,
      "num_tokens": 80224780.0,
      "step": 35940
    },
    {
      "entropy": 5.350775289535522,
      "epoch": 4.036951760543606,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003465722588058497,
      "loss": 4.9199,
      "mean_token_accuracy": 0.22351645827293395,
      "num_tokens": 80235399.0,
      "step": 35945
    },
    {
      "entropy": 5.280110692977905,
      "epoch": 4.037513337451564,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034653419437139475,
      "loss": 4.7714,
      "mean_token_accuracy": 0.23413630425930024,
      "num_tokens": 80245019.0,
      "step": 35950
    },
    {
      "entropy": 5.309762573242187,
      "epoch": 4.038074914359521,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003464961276592634,
      "loss": 4.8891,
      "mean_token_accuracy": 0.23116171211004258,
      "num_tokens": 80256650.0,
      "step": 35955
    },
    {
      "entropy": 5.241188478469849,
      "epoch": 4.0386364912674795,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00034645805867066756,
      "loss": 4.7702,
      "mean_token_accuracy": 0.22871480137109756,
      "num_tokens": 80267463.0,
      "step": 35960
    },
    {
      "entropy": 5.298409366607666,
      "epoch": 4.039198068175437,
      "grad_norm": 0.9375,
      "learning_rate": 0.00034641998740681933,
      "loss": 4.871,
      "mean_token_accuracy": 0.22519782036542893,
      "num_tokens": 80279751.0,
      "step": 35965
    },
    {
      "entropy": 5.274011039733887,
      "epoch": 4.039759645083394,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00034638191386893115,
      "loss": 4.7887,
      "mean_token_accuracy": 0.23524820506572724,
      "num_tokens": 80291249.0,
      "step": 35970
    },
    {
      "entropy": 5.266691207885742,
      "epoch": 4.040321221991352,
      "grad_norm": 1.0625,
      "learning_rate": 0.00034634383805821504,
      "loss": 4.7748,
      "mean_token_accuracy": 0.23837108314037322,
      "num_tokens": 80302082.0,
      "step": 35975
    },
    {
      "entropy": 5.159908390045166,
      "epoch": 4.040882798899309,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034630575997588353,
      "loss": 4.7112,
      "mean_token_accuracy": 0.2470157578587532,
      "num_tokens": 80312470.0,
      "step": 35980
    },
    {
      "entropy": 5.232155275344849,
      "epoch": 4.0414443758072665,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000346267679623149,
      "loss": 4.7527,
      "mean_token_accuracy": 0.2425937235355377,
      "num_tokens": 80322562.0,
      "step": 35985
    },
    {
      "entropy": 5.217661714553833,
      "epoch": 4.042005952715225,
      "grad_norm": 1.171875,
      "learning_rate": 0.00034622959700122383,
      "loss": 4.8285,
      "mean_token_accuracy": 0.2360434651374817,
      "num_tokens": 80333434.0,
      "step": 35990
    },
    {
      "entropy": 5.169496536254883,
      "epoch": 4.042567529623182,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003461915121113206,
      "loss": 4.7226,
      "mean_token_accuracy": 0.23881673216819763,
      "num_tokens": 80345028.0,
      "step": 35995
    },
    {
      "entropy": 5.181522941589355,
      "epoch": 4.04312910653114,
      "grad_norm": 1.0,
      "learning_rate": 0.0003461534249546521,
      "loss": 4.7403,
      "mean_token_accuracy": 0.23748517036437988,
      "num_tokens": 80356536.0,
      "step": 36000
    },
    {
      "epoch": 4.04312910653114,
      "eval_entropy": 5.064616212778433,
      "eval_loss": 5.036772727966309,
      "eval_mean_token_accuracy": 0.225197712053166,
      "eval_num_tokens": 80356536.0,
      "eval_runtime": 23.9798,
      "eval_samples_per_second": 1293.295,
      "eval_steps_per_second": 161.678,
      "step": 36000
    }
  ],
  "logging_steps": 5,
  "max_steps": 89030,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 3000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.26030064134528e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}