{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.8552787663107946,
  "eval_steps": 3000,
  "global_step": 39000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 7.151007413864136,
      "epoch": 0.0004942665085013839,
      "grad_norm": 1.078125,
      "learning_rate": 2e-06,
      "loss": 7.279,
      "mean_token_accuracy": 0.08845784291625022,
      "num_tokens": 11289.0,
      "step": 5
    },
    {
      "entropy": 7.142992544174194,
      "epoch": 0.0009885330170027679,
      "grad_norm": 1.1015625,
      "learning_rate": 4.5e-06,
      "loss": 7.2577,
      "mean_token_accuracy": 0.08933138325810433,
      "num_tokens": 21541.0,
      "step": 10
    },
    {
      "entropy": 7.182202959060669,
      "epoch": 0.0014827995255041518,
      "grad_norm": 1.0703125,
      "learning_rate": 7e-06,
      "loss": 7.2317,
      "mean_token_accuracy": 0.0934896469116211,
      "num_tokens": 32257.0,
      "step": 15
    },
    {
      "entropy": 7.244844675064087,
      "epoch": 0.0019770660340055358,
      "grad_norm": 0.9140625,
      "learning_rate": 9.5e-06,
      "loss": 7.2492,
      "mean_token_accuracy": 0.09298909306526185,
      "num_tokens": 44176.0,
      "step": 20
    },
    {
      "entropy": 7.262534332275391,
      "epoch": 0.0024713325425069197,
      "grad_norm": 0.98828125,
      "learning_rate": 1.2e-05,
      "loss": 7.1678,
      "mean_token_accuracy": 0.09061622098088265,
      "num_tokens": 56115.0,
      "step": 25
    },
    {
      "entropy": 7.378157377243042,
      "epoch": 0.0029655990510083037,
      "grad_norm": 0.94921875,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 7.2162,
      "mean_token_accuracy": 0.09287815466523171,
      "num_tokens": 68828.0,
      "step": 30
    },
    {
      "entropy": 7.399503707885742,
      "epoch": 0.0034598655595096876,
      "grad_norm": 0.98828125,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 7.2093,
      "mean_token_accuracy": 0.09124612733721733,
      "num_tokens": 80097.0,
      "step": 35
    },
    {
      "entropy": 7.373995113372803,
      "epoch": 0.0039541320680110716,
      "grad_norm": 0.83203125,
      "learning_rate": 1.95e-05,
      "loss": 7.1903,
      "mean_token_accuracy": 0.09260638207197189,
      "num_tokens": 91510.0,
      "step": 40
    },
    {
      "entropy": 7.330695772171021,
      "epoch": 0.004448398576512456,
      "grad_norm": 0.8046875,
      "learning_rate": 2.2e-05,
      "loss": 7.234,
      "mean_token_accuracy": 0.09660193100571632,
      "num_tokens": 103443.0,
      "step": 45
    },
    {
      "entropy": 7.33756799697876,
      "epoch": 0.0049426650850138395,
      "grad_norm": 0.83984375,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 7.2565,
      "mean_token_accuracy": 0.09188243448734283,
      "num_tokens": 114629.0,
      "step": 50
    },
    {
      "entropy": 7.324716234207154,
      "epoch": 0.005436931593515224,
      "grad_norm": 0.875,
      "learning_rate": 2.7e-05,
      "loss": 7.2396,
      "mean_token_accuracy": 0.09266308695077896,
      "num_tokens": 125012.0,
      "step": 55
    },
    {
      "entropy": 7.393645143508911,
      "epoch": 0.005931198102016607,
      "grad_norm": 0.87890625,
      "learning_rate": 2.95e-05,
      "loss": 7.1731,
      "mean_token_accuracy": 0.09935653060674668,
      "num_tokens": 135834.0,
      "step": 60
    },
    {
      "entropy": 7.348312473297119,
      "epoch": 0.006425464610517992,
      "grad_norm": 0.89453125,
      "learning_rate": 3.2e-05,
      "loss": 7.0755,
      "mean_token_accuracy": 0.10009140819311142,
      "num_tokens": 147413.0,
      "step": 65
    },
    {
      "entropy": 7.368169450759888,
      "epoch": 0.006919731119019375,
      "grad_norm": 0.9375,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 7.2672,
      "mean_token_accuracy": 0.09652369320392609,
      "num_tokens": 158046.0,
      "step": 70
    },
    {
      "entropy": 7.3670432567596436,
      "epoch": 0.00741399762752076,
      "grad_norm": 0.99609375,
      "learning_rate": 3.7e-05,
      "loss": 7.1792,
      "mean_token_accuracy": 0.09774387776851653,
      "num_tokens": 168564.0,
      "step": 75
    },
    {
      "entropy": 7.313370609283448,
      "epoch": 0.007908264136022143,
      "grad_norm": 0.8515625,
      "learning_rate": 3.95e-05,
      "loss": 7.1885,
      "mean_token_accuracy": 0.09465310201048852,
      "num_tokens": 181168.0,
      "step": 80
    },
    {
      "entropy": 7.35333480834961,
      "epoch": 0.008402530644523527,
      "grad_norm": 0.9296875,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 7.1552,
      "mean_token_accuracy": 0.09859942272305489,
      "num_tokens": 192666.0,
      "step": 85
    },
    {
      "entropy": 7.388514804840088,
      "epoch": 0.008896797153024912,
      "grad_norm": 0.81640625,
      "learning_rate": 4.45e-05,
      "loss": 7.2204,
      "mean_token_accuracy": 0.09482975378632545,
      "num_tokens": 203688.0,
      "step": 90
    },
    {
      "entropy": 7.303425407409668,
      "epoch": 0.009391063661526295,
      "grad_norm": 0.9140625,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 7.0729,
      "mean_token_accuracy": 0.10156495124101639,
      "num_tokens": 214399.0,
      "step": 95
    },
    {
      "entropy": 7.27442512512207,
      "epoch": 0.009885330170027679,
      "grad_norm": 0.765625,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 7.1681,
      "mean_token_accuracy": 0.09274221584200859,
      "num_tokens": 227549.0,
      "step": 100
    },
    {
      "entropy": 7.317437267303466,
      "epoch": 0.010379596678529062,
      "grad_norm": 0.765625,
      "learning_rate": 5.2e-05,
      "loss": 7.0679,
      "mean_token_accuracy": 0.11123422309756278,
      "num_tokens": 239612.0,
      "step": 105
    },
    {
      "entropy": 7.363053894042968,
      "epoch": 0.010873863187030448,
      "grad_norm": 0.96484375,
      "learning_rate": 5.45e-05,
      "loss": 7.1978,
      "mean_token_accuracy": 0.09392422363162041,
      "num_tokens": 251163.0,
      "step": 110
    },
    {
      "entropy": 7.261190605163574,
      "epoch": 0.011368129695531831,
      "grad_norm": 0.87890625,
      "learning_rate": 5.7e-05,
      "loss": 7.0645,
      "mean_token_accuracy": 0.10441159456968307,
      "num_tokens": 263995.0,
      "step": 115
    },
    {
      "entropy": 7.40733699798584,
      "epoch": 0.011862396204033215,
      "grad_norm": 1.0390625,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 7.2354,
      "mean_token_accuracy": 0.08496984019875527,
      "num_tokens": 275103.0,
      "step": 120
    },
    {
      "entropy": 7.273692512512207,
      "epoch": 0.012356662712534598,
      "grad_norm": 0.8515625,
      "learning_rate": 6.2e-05,
      "loss": 7.1369,
      "mean_token_accuracy": 0.0968512661755085,
      "num_tokens": 286838.0,
      "step": 125
    },
    {
      "entropy": 7.423242712020874,
      "epoch": 0.012850929221035983,
      "grad_norm": 1.046875,
      "learning_rate": 6.450000000000001e-05,
      "loss": 7.1665,
      "mean_token_accuracy": 0.09364857673645019,
      "num_tokens": 298750.0,
      "step": 130
    },
    {
      "entropy": 7.250139808654785,
      "epoch": 0.013345195729537367,
      "grad_norm": 0.953125,
      "learning_rate": 6.7e-05,
      "loss": 7.0329,
      "mean_token_accuracy": 0.10625170543789864,
      "num_tokens": 310451.0,
      "step": 135
    },
    {
      "entropy": 7.348578405380249,
      "epoch": 0.01383946223803875,
      "grad_norm": 0.95703125,
      "learning_rate": 6.950000000000001e-05,
      "loss": 7.1935,
      "mean_token_accuracy": 0.09695170894265175,
      "num_tokens": 321232.0,
      "step": 140
    },
    {
      "entropy": 7.488092279434204,
      "epoch": 0.014333728746540134,
      "grad_norm": 0.83984375,
      "learning_rate": 7.2e-05,
      "loss": 7.106,
      "mean_token_accuracy": 0.09741687774658203,
      "num_tokens": 332531.0,
      "step": 145
    },
    {
      "entropy": 7.19592809677124,
      "epoch": 0.01482799525504152,
      "grad_norm": 0.94140625,
      "learning_rate": 7.45e-05,
      "loss": 7.1008,
      "mean_token_accuracy": 0.09623056724667549,
      "num_tokens": 346002.0,
      "step": 150
    },
    {
      "entropy": 7.401388263702392,
      "epoch": 0.015322261763542903,
      "grad_norm": 1.03125,
      "learning_rate": 7.7e-05,
      "loss": 7.0874,
      "mean_token_accuracy": 0.09739502891898155,
      "num_tokens": 356691.0,
      "step": 155
    },
    {
      "entropy": 7.243306970596313,
      "epoch": 0.015816528272044286,
      "grad_norm": 0.94140625,
      "learning_rate": 7.950000000000001e-05,
      "loss": 7.0938,
      "mean_token_accuracy": 0.1009882666170597,
      "num_tokens": 368749.0,
      "step": 160
    },
    {
      "entropy": 7.3369566917419435,
      "epoch": 0.01631079478054567,
      "grad_norm": 1.03125,
      "learning_rate": 8.2e-05,
      "loss": 7.0241,
      "mean_token_accuracy": 0.10501249730587006,
      "num_tokens": 379484.0,
      "step": 165
    },
    {
      "entropy": 7.281102800369263,
      "epoch": 0.016805061289047053,
      "grad_norm": 0.953125,
      "learning_rate": 8.450000000000001e-05,
      "loss": 7.0703,
      "mean_token_accuracy": 0.09354055374860763,
      "num_tokens": 389751.0,
      "step": 170
    },
    {
      "entropy": 7.325433778762817,
      "epoch": 0.017299327797548437,
      "grad_norm": 0.99609375,
      "learning_rate": 8.7e-05,
      "loss": 7.0892,
      "mean_token_accuracy": 0.09684126302599907,
      "num_tokens": 400771.0,
      "step": 175
    },
    {
      "entropy": 7.323175096511841,
      "epoch": 0.017793594306049824,
      "grad_norm": 0.90234375,
      "learning_rate": 8.95e-05,
      "loss": 7.1061,
      "mean_token_accuracy": 0.09781083017587662,
      "num_tokens": 413575.0,
      "step": 180
    },
    {
      "entropy": 7.328048992156982,
      "epoch": 0.018287860814551207,
      "grad_norm": 0.890625,
      "learning_rate": 9.2e-05,
      "loss": 7.0467,
      "mean_token_accuracy": 0.09532680064439773,
      "num_tokens": 424825.0,
      "step": 185
    },
    {
      "entropy": 7.333641099929809,
      "epoch": 0.01878212732305259,
      "grad_norm": 0.93359375,
      "learning_rate": 9.45e-05,
      "loss": 7.1064,
      "mean_token_accuracy": 0.10553916320204734,
      "num_tokens": 435610.0,
      "step": 190
    },
    {
      "entropy": 7.328917694091797,
      "epoch": 0.019276393831553974,
      "grad_norm": 1.0625,
      "learning_rate": 9.7e-05,
      "loss": 7.1187,
      "mean_token_accuracy": 0.0999750517308712,
      "num_tokens": 445855.0,
      "step": 195
    },
    {
      "entropy": 7.2166337966918945,
      "epoch": 0.019770660340055358,
      "grad_norm": 0.98046875,
      "learning_rate": 9.95e-05,
      "loss": 7.0063,
      "mean_token_accuracy": 0.10225539430975913,
      "num_tokens": 456895.0,
      "step": 200
    },
    {
      "entropy": 7.476543855667114,
      "epoch": 0.02026492684855674,
      "grad_norm": 1.234375,
      "learning_rate": 0.000102,
      "loss": 7.1414,
      "mean_token_accuracy": 0.09506701603531838,
      "num_tokens": 467446.0,
      "step": 205
    },
    {
      "entropy": 7.272785186767578,
      "epoch": 0.020759193357058125,
      "grad_norm": 0.921875,
      "learning_rate": 0.00010449999999999999,
      "loss": 7.0607,
      "mean_token_accuracy": 0.09756421521306038,
      "num_tokens": 478988.0,
      "step": 210
    },
    {
      "entropy": 7.290947103500367,
      "epoch": 0.02125345986555951,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000107,
      "loss": 7.0557,
      "mean_token_accuracy": 0.10137153342366219,
      "num_tokens": 491053.0,
      "step": 215
    },
    {
      "entropy": 7.357240819931031,
      "epoch": 0.021747726374060895,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0001095,
      "loss": 7.1334,
      "mean_token_accuracy": 0.09304505437612534,
      "num_tokens": 502279.0,
      "step": 220
    },
    {
      "entropy": 7.198823118209839,
      "epoch": 0.02224199288256228,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000112,
      "loss": 7.0177,
      "mean_token_accuracy": 0.10008044987916946,
      "num_tokens": 513470.0,
      "step": 225
    },
    {
      "entropy": 7.373158121109009,
      "epoch": 0.022736259391063662,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0001145,
      "loss": 7.0871,
      "mean_token_accuracy": 0.096877521276474,
      "num_tokens": 524380.0,
      "step": 230
    },
    {
      "entropy": 7.324968671798706,
      "epoch": 0.023230525899565046,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.0885,
      "mean_token_accuracy": 0.10163207277655602,
      "num_tokens": 535798.0,
      "step": 235
    },
    {
      "entropy": 7.273069143295288,
      "epoch": 0.02372479240806643,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.0313,
      "mean_token_accuracy": 0.10062324851751328,
      "num_tokens": 545720.0,
      "step": 240
    },
    {
      "entropy": 7.329104614257813,
      "epoch": 0.024219058916567813,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000122,
      "loss": 7.0595,
      "mean_token_accuracy": 0.09972514137625695,
      "num_tokens": 557110.0,
      "step": 245
    },
    {
      "entropy": 7.1850278854370115,
      "epoch": 0.024713325425069196,
      "grad_norm": 1.109375,
      "learning_rate": 0.0001245,
      "loss": 7.016,
      "mean_token_accuracy": 0.10394479855895042,
      "num_tokens": 568966.0,
      "step": 250
    },
    {
      "entropy": 7.360903692245484,
      "epoch": 0.02520759193357058,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000127,
      "loss": 7.0469,
      "mean_token_accuracy": 0.10300094187259674,
      "num_tokens": 580441.0,
      "step": 255
    },
    {
      "entropy": 7.301402139663696,
      "epoch": 0.025701858442071967,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0001295,
      "loss": 7.0452,
      "mean_token_accuracy": 0.09699605703353882,
      "num_tokens": 592698.0,
      "step": 260
    },
    {
      "entropy": 7.190503692626953,
      "epoch": 0.02619612495057335,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000132,
      "loss": 7.0198,
      "mean_token_accuracy": 0.10279446840286255,
      "num_tokens": 604787.0,
      "step": 265
    },
    {
      "entropy": 7.302949237823486,
      "epoch": 0.026690391459074734,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.013,
      "mean_token_accuracy": 0.09963318407535553,
      "num_tokens": 616445.0,
      "step": 270
    },
    {
      "entropy": 7.2669110774993895,
      "epoch": 0.027184657967576117,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00013700000000000002,
      "loss": 7.0808,
      "mean_token_accuracy": 0.1002419799566269,
      "num_tokens": 627723.0,
      "step": 275
    },
    {
      "entropy": 7.345779991149902,
      "epoch": 0.0276789244760775,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0001395,
      "loss": 7.1067,
      "mean_token_accuracy": 0.10031414031982422,
      "num_tokens": 638987.0,
      "step": 280
    },
    {
      "entropy": 7.279949617385864,
      "epoch": 0.028173190984578884,
      "grad_norm": 1.296875,
      "learning_rate": 0.00014199999999999998,
      "loss": 6.9384,
      "mean_token_accuracy": 0.10166889876127243,
      "num_tokens": 650626.0,
      "step": 285
    },
    {
      "entropy": 7.149828386306763,
      "epoch": 0.028667457493080268,
      "grad_norm": 1.109375,
      "learning_rate": 0.0001445,
      "loss": 6.9604,
      "mean_token_accuracy": 0.10975496619939804,
      "num_tokens": 661911.0,
      "step": 290
    },
    {
      "entropy": 7.256064605712891,
      "epoch": 0.02916172400158165,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000147,
      "loss": 7.0137,
      "mean_token_accuracy": 0.10477377250790595,
      "num_tokens": 674298.0,
      "step": 295
    },
    {
      "entropy": 7.236632680892944,
      "epoch": 0.02965599051008304,
      "grad_norm": 0.96875,
      "learning_rate": 0.0001495,
      "loss": 6.9431,
      "mean_token_accuracy": 0.10994272455573081,
      "num_tokens": 686000.0,
      "step": 300
    },
    {
      "entropy": 7.216502523422241,
      "epoch": 0.030150257018584422,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000152,
      "loss": 6.9926,
      "mean_token_accuracy": 0.10297218561172486,
      "num_tokens": 698464.0,
      "step": 305
    },
    {
      "entropy": 7.215946102142334,
      "epoch": 0.030644523527085805,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00015450000000000001,
      "loss": 6.9883,
      "mean_token_accuracy": 0.09871925041079521,
      "num_tokens": 709702.0,
      "step": 310
    },
    {
      "entropy": 7.40031623840332,
      "epoch": 0.03113879003558719,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000157,
      "loss": 7.1009,
      "mean_token_accuracy": 0.09918646588921547,
      "num_tokens": 722234.0,
      "step": 315
    },
    {
      "entropy": 7.222815370559692,
      "epoch": 0.03163305654408857,
      "grad_norm": 1.09375,
      "learning_rate": 0.0001595,
      "loss": 7.0443,
      "mean_token_accuracy": 0.10143259838223458,
      "num_tokens": 733001.0,
      "step": 320
    },
    {
      "entropy": 7.251048564910889,
      "epoch": 0.03212732305258996,
      "grad_norm": 1.0625,
      "learning_rate": 0.000162,
      "loss": 7.0047,
      "mean_token_accuracy": 0.09862519353628159,
      "num_tokens": 743315.0,
      "step": 325
    },
    {
      "entropy": 7.253071928024292,
      "epoch": 0.03262158956109134,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00016450000000000001,
      "loss": 6.9549,
      "mean_token_accuracy": 0.09725747108459473,
      "num_tokens": 754386.0,
      "step": 330
    },
    {
      "entropy": 7.24435601234436,
      "epoch": 0.033115856069592726,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00016700000000000002,
      "loss": 7.0694,
      "mean_token_accuracy": 0.10413578823208809,
      "num_tokens": 766504.0,
      "step": 335
    },
    {
      "entropy": 7.090495014190674,
      "epoch": 0.033610122578094107,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00016950000000000003,
      "loss": 6.9017,
      "mean_token_accuracy": 0.10729243233799934,
      "num_tokens": 777039.0,
      "step": 340
    },
    {
      "entropy": 7.284985685348511,
      "epoch": 0.034104389086595494,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00017199999999999998,
      "loss": 7.0025,
      "mean_token_accuracy": 0.10153071135282517,
      "num_tokens": 787890.0,
      "step": 345
    },
    {
      "entropy": 7.234701490402221,
      "epoch": 0.034598655595096874,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00017449999999999999,
      "loss": 6.9532,
      "mean_token_accuracy": 0.10262288972735405,
      "num_tokens": 800129.0,
      "step": 350
    },
    {
      "entropy": 7.139390373229981,
      "epoch": 0.03509292210359826,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000177,
      "loss": 7.0596,
      "mean_token_accuracy": 0.0944197230041027,
      "num_tokens": 812517.0,
      "step": 355
    },
    {
      "entropy": 7.259590196609497,
      "epoch": 0.03558718861209965,
      "grad_norm": 1.0,
      "learning_rate": 0.0001795,
      "loss": 6.9787,
      "mean_token_accuracy": 0.10446251705288886,
      "num_tokens": 824570.0,
      "step": 360
    },
    {
      "entropy": 7.1810768127441404,
      "epoch": 0.03608145512060103,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000182,
      "loss": 6.9823,
      "mean_token_accuracy": 0.1032750204205513,
      "num_tokens": 836506.0,
      "step": 365
    },
    {
      "entropy": 7.251035213470459,
      "epoch": 0.036575721629102415,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0001845,
      "loss": 6.9702,
      "mean_token_accuracy": 0.10457680597901345,
      "num_tokens": 846841.0,
      "step": 370
    },
    {
      "entropy": 7.27037239074707,
      "epoch": 0.037069988137603795,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000187,
      "loss": 7.0054,
      "mean_token_accuracy": 0.106174336373806,
      "num_tokens": 858770.0,
      "step": 375
    },
    {
      "entropy": 7.1212574481964115,
      "epoch": 0.03756425464610518,
      "grad_norm": 1.125,
      "learning_rate": 0.0001895,
      "loss": 6.8786,
      "mean_token_accuracy": 0.10385922640562058,
      "num_tokens": 870184.0,
      "step": 380
    },
    {
      "entropy": 7.172485733032227,
      "epoch": 0.03805852115460656,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000192,
      "loss": 6.9197,
      "mean_token_accuracy": 0.10581747069954872,
      "num_tokens": 881843.0,
      "step": 385
    },
    {
      "entropy": 7.2427778244018555,
      "epoch": 0.03855278766310795,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0001945,
      "loss": 6.9489,
      "mean_token_accuracy": 0.10129198804497719,
      "num_tokens": 893453.0,
      "step": 390
    },
    {
      "entropy": 7.205210494995117,
      "epoch": 0.03904705417160933,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00019700000000000002,
      "loss": 7.0049,
      "mean_token_accuracy": 0.09680559858679771,
      "num_tokens": 904417.0,
      "step": 395
    },
    {
      "entropy": 7.202549695968628,
      "epoch": 0.039541320680110716,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00019950000000000002,
      "loss": 6.9462,
      "mean_token_accuracy": 0.0999035358428955,
      "num_tokens": 915134.0,
      "step": 400
    },
    {
      "entropy": 7.133806133270264,
      "epoch": 0.0400355871886121,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000202,
      "loss": 6.9589,
      "mean_token_accuracy": 0.1034577377140522,
      "num_tokens": 926361.0,
      "step": 405
    },
    {
      "entropy": 7.144542980194092,
      "epoch": 0.04052985369711348,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00020449999999999998,
      "loss": 6.8841,
      "mean_token_accuracy": 0.10176018252968788,
      "num_tokens": 937474.0,
      "step": 410
    },
    {
      "entropy": 7.3199704647064205,
      "epoch": 0.04102412020561487,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000207,
      "loss": 6.947,
      "mean_token_accuracy": 0.10589171275496483,
      "num_tokens": 948897.0,
      "step": 415
    },
    {
      "entropy": 7.157266998291016,
      "epoch": 0.04151838671411625,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002095,
      "loss": 6.9643,
      "mean_token_accuracy": 0.1008329525589943,
      "num_tokens": 960669.0,
      "step": 420
    },
    {
      "entropy": 7.024852180480957,
      "epoch": 0.04201265322261764,
      "grad_norm": 1.0,
      "learning_rate": 0.000212,
      "loss": 6.8152,
      "mean_token_accuracy": 0.10260056778788566,
      "num_tokens": 973001.0,
      "step": 425
    },
    {
      "entropy": 7.119008874893188,
      "epoch": 0.04250691973111902,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002145,
      "loss": 6.8859,
      "mean_token_accuracy": 0.10822405442595481,
      "num_tokens": 986214.0,
      "step": 430
    },
    {
      "entropy": 7.105495595932007,
      "epoch": 0.043001186239620404,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00021700000000000002,
      "loss": 6.8937,
      "mean_token_accuracy": 0.10061700344085693,
      "num_tokens": 997363.0,
      "step": 435
    },
    {
      "entropy": 7.120448970794678,
      "epoch": 0.04349545274812179,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002195,
      "loss": 6.8093,
      "mean_token_accuracy": 0.11288737356662751,
      "num_tokens": 1009543.0,
      "step": 440
    },
    {
      "entropy": 7.0541246891021725,
      "epoch": 0.04398971925662317,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000222,
      "loss": 6.9215,
      "mean_token_accuracy": 0.10269275158643723,
      "num_tokens": 1020561.0,
      "step": 445
    },
    {
      "entropy": 7.185726308822632,
      "epoch": 0.04448398576512456,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002245,
      "loss": 6.8725,
      "mean_token_accuracy": 0.10224331989884376,
      "num_tokens": 1032587.0,
      "step": 450
    },
    {
      "entropy": 7.159319496154785,
      "epoch": 0.04497825227362594,
      "grad_norm": 1.140625,
      "learning_rate": 0.00022700000000000002,
      "loss": 6.8784,
      "mean_token_accuracy": 0.1019284375011921,
      "num_tokens": 1044249.0,
      "step": 455
    },
    {
      "entropy": 7.076584959030152,
      "epoch": 0.045472518782127325,
      "grad_norm": 1.328125,
      "learning_rate": 0.00022950000000000002,
      "loss": 6.9215,
      "mean_token_accuracy": 0.10558762773871422,
      "num_tokens": 1055415.0,
      "step": 460
    },
    {
      "entropy": 7.1408079147338865,
      "epoch": 0.045966785290628705,
      "grad_norm": 1.03125,
      "learning_rate": 0.00023200000000000003,
      "loss": 6.8942,
      "mean_token_accuracy": 0.10558660700917244,
      "num_tokens": 1066593.0,
      "step": 465
    },
    {
      "entropy": 7.15621657371521,
      "epoch": 0.04646105179913009,
      "grad_norm": 1.109375,
      "learning_rate": 0.00023449999999999998,
      "loss": 6.8924,
      "mean_token_accuracy": 0.10691260844469071,
      "num_tokens": 1076966.0,
      "step": 470
    },
    {
      "entropy": 7.067881631851196,
      "epoch": 0.04695531830763147,
      "grad_norm": 1.03125,
      "learning_rate": 0.000237,
      "loss": 6.9491,
      "mean_token_accuracy": 0.09959092512726783,
      "num_tokens": 1089927.0,
      "step": 475
    },
    {
      "entropy": 7.091880464553833,
      "epoch": 0.04744958481613286,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002395,
      "loss": 6.8869,
      "mean_token_accuracy": 0.10327862426638604,
      "num_tokens": 1101920.0,
      "step": 480
    },
    {
      "entropy": 7.182226657867432,
      "epoch": 0.047943851324634246,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000242,
      "loss": 6.9376,
      "mean_token_accuracy": 0.10818986669182777,
      "num_tokens": 1115078.0,
      "step": 485
    },
    {
      "entropy": 7.180230808258057,
      "epoch": 0.048438117833135626,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002445,
      "loss": 7.0315,
      "mean_token_accuracy": 0.0969414807856083,
      "num_tokens": 1127138.0,
      "step": 490
    },
    {
      "entropy": 7.088583707809448,
      "epoch": 0.04893238434163701,
      "grad_norm": 1.1875,
      "learning_rate": 0.000247,
      "loss": 6.815,
      "mean_token_accuracy": 0.10144753381609917,
      "num_tokens": 1138390.0,
      "step": 495
    },
    {
      "entropy": 7.003582715988159,
      "epoch": 0.04942665085013839,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002495,
      "loss": 6.78,
      "mean_token_accuracy": 0.11048796996474267,
      "num_tokens": 1149124.0,
      "step": 500
    },
    {
      "entropy": 7.101625585556031,
      "epoch": 0.04992091735863978,
      "grad_norm": 1.109375,
      "learning_rate": 0.000252,
      "loss": 6.8585,
      "mean_token_accuracy": 0.10951191484928131,
      "num_tokens": 1161091.0,
      "step": 505
    },
    {
      "entropy": 7.118922424316406,
      "epoch": 0.05041518386714116,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002545,
      "loss": 6.8355,
      "mean_token_accuracy": 0.10461836084723472,
      "num_tokens": 1171919.0,
      "step": 510
    },
    {
      "entropy": 7.051087188720703,
      "epoch": 0.05090945037564255,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000257,
      "loss": 6.8633,
      "mean_token_accuracy": 0.10401171222329139,
      "num_tokens": 1183688.0,
      "step": 515
    },
    {
      "entropy": 7.0168451309204105,
      "epoch": 0.051403716884143934,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002595,
      "loss": 6.7819,
      "mean_token_accuracy": 0.10449134409427643,
      "num_tokens": 1195323.0,
      "step": 520
    },
    {
      "entropy": 7.09809103012085,
      "epoch": 0.051897983392645314,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000262,
      "loss": 6.8999,
      "mean_token_accuracy": 0.10218222662806511,
      "num_tokens": 1206782.0,
      "step": 525
    },
    {
      "entropy": 6.965949535369873,
      "epoch": 0.0523922499011467,
      "grad_norm": 1.0625,
      "learning_rate": 0.00026450000000000003,
      "loss": 6.8013,
      "mean_token_accuracy": 0.10973932594060898,
      "num_tokens": 1218102.0,
      "step": 530
    },
    {
      "entropy": 7.017365598678589,
      "epoch": 0.05288651640964808,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.7659,
      "mean_token_accuracy": 0.10771052315831184,
      "num_tokens": 1230199.0,
      "step": 535
    },
    {
      "entropy": 7.0721368312835695,
      "epoch": 0.05338078291814947,
      "grad_norm": 1.03125,
      "learning_rate": 0.00026950000000000005,
      "loss": 6.8896,
      "mean_token_accuracy": 0.10399755015969277,
      "num_tokens": 1242127.0,
      "step": 540
    },
    {
      "entropy": 7.027034282684326,
      "epoch": 0.05387504942665085,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.7655,
      "mean_token_accuracy": 0.10667505711317063,
      "num_tokens": 1252329.0,
      "step": 545
    },
    {
      "entropy": 6.865018033981324,
      "epoch": 0.054369315935152235,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002745,
      "loss": 6.7666,
      "mean_token_accuracy": 0.11077538877725601,
      "num_tokens": 1264832.0,
      "step": 550
    },
    {
      "entropy": 7.053894948959351,
      "epoch": 0.054863582443653615,
      "grad_norm": 1.0625,
      "learning_rate": 0.000277,
      "loss": 6.8414,
      "mean_token_accuracy": 0.10259129181504249,
      "num_tokens": 1276708.0,
      "step": 555
    },
    {
      "entropy": 7.017140913009643,
      "epoch": 0.055357848952155,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002795,
      "loss": 6.7446,
      "mean_token_accuracy": 0.11817902401089668,
      "num_tokens": 1288089.0,
      "step": 560
    },
    {
      "entropy": 7.102619647979736,
      "epoch": 0.05585211546065639,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.9011,
      "mean_token_accuracy": 0.10063079670071602,
      "num_tokens": 1299411.0,
      "step": 565
    },
    {
      "entropy": 7.038193321228027,
      "epoch": 0.05634638196915777,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002845,
      "loss": 6.8678,
      "mean_token_accuracy": 0.10497712716460228,
      "num_tokens": 1311368.0,
      "step": 570
    },
    {
      "entropy": 6.996581506729126,
      "epoch": 0.056840648477659156,
      "grad_norm": 1.140625,
      "learning_rate": 0.000287,
      "loss": 6.7798,
      "mean_token_accuracy": 0.1104745551943779,
      "num_tokens": 1321962.0,
      "step": 575
    },
    {
      "entropy": 6.984814691543579,
      "epoch": 0.057334914986160536,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002895,
      "loss": 6.7845,
      "mean_token_accuracy": 0.10723584592342376,
      "num_tokens": 1335150.0,
      "step": 580
    },
    {
      "entropy": 7.0060882568359375,
      "epoch": 0.05782918149466192,
      "grad_norm": 1.015625,
      "learning_rate": 0.000292,
      "loss": 6.8106,
      "mean_token_accuracy": 0.10954170823097228,
      "num_tokens": 1346986.0,
      "step": 585
    },
    {
      "entropy": 7.04237003326416,
      "epoch": 0.0583234480031633,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002945,
      "loss": 6.7456,
      "mean_token_accuracy": 0.11029022559523582,
      "num_tokens": 1357820.0,
      "step": 590
    },
    {
      "entropy": 6.898483562469482,
      "epoch": 0.05881771451166469,
      "grad_norm": 1.203125,
      "learning_rate": 0.000297,
      "loss": 6.7381,
      "mean_token_accuracy": 0.1155846431851387,
      "num_tokens": 1368269.0,
      "step": 595
    },
    {
      "entropy": 6.914667129516602,
      "epoch": 0.05931198102016608,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002995,
      "loss": 6.7699,
      "mean_token_accuracy": 0.10781971290707588,
      "num_tokens": 1380364.0,
      "step": 600
    },
    {
      "entropy": 6.9423737049102785,
      "epoch": 0.05980624752866746,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000302,
      "loss": 6.7359,
      "mean_token_accuracy": 0.11458505317568779,
      "num_tokens": 1390667.0,
      "step": 605
    },
    {
      "entropy": 6.890728282928467,
      "epoch": 0.060300514037168844,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003045,
      "loss": 6.6885,
      "mean_token_accuracy": 0.10985699743032455,
      "num_tokens": 1401452.0,
      "step": 610
    },
    {
      "entropy": 6.968847942352295,
      "epoch": 0.060794780545670224,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000307,
      "loss": 6.6936,
      "mean_token_accuracy": 0.11100370958447456,
      "num_tokens": 1412947.0,
      "step": 615
    },
    {
      "entropy": 6.867507553100586,
      "epoch": 0.06128904705417161,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003095,
      "loss": 6.7476,
      "mean_token_accuracy": 0.10398394390940666,
      "num_tokens": 1425123.0,
      "step": 620
    },
    {
      "entropy": 6.928701877593994,
      "epoch": 0.06178331356267299,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000312,
      "loss": 6.7218,
      "mean_token_accuracy": 0.10982921719551086,
      "num_tokens": 1436538.0,
      "step": 625
    },
    {
      "entropy": 6.933251285552979,
      "epoch": 0.06227758007117438,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003145,
      "loss": 6.77,
      "mean_token_accuracy": 0.10577303692698478,
      "num_tokens": 1448269.0,
      "step": 630
    },
    {
      "entropy": 6.900527620315552,
      "epoch": 0.06277184657967576,
      "grad_norm": 1.1875,
      "learning_rate": 0.000317,
      "loss": 6.7432,
      "mean_token_accuracy": 0.11349172964692116,
      "num_tokens": 1459729.0,
      "step": 635
    },
    {
      "entropy": 6.960564184188843,
      "epoch": 0.06326611308817714,
      "grad_norm": 1.125,
      "learning_rate": 0.0003195,
      "loss": 6.7581,
      "mean_token_accuracy": 0.10724205747246743,
      "num_tokens": 1470684.0,
      "step": 640
    },
    {
      "entropy": 6.889400863647461,
      "epoch": 0.06376037959667853,
      "grad_norm": 0.9375,
      "learning_rate": 0.000322,
      "loss": 6.7232,
      "mean_token_accuracy": 0.10671663433313369,
      "num_tokens": 1483571.0,
      "step": 645
    },
    {
      "entropy": 6.8848834991455075,
      "epoch": 0.06425464610517992,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.6891,
      "mean_token_accuracy": 0.11692152470350266,
      "num_tokens": 1494671.0,
      "step": 650
    },
    {
      "entropy": 6.905150508880615,
      "epoch": 0.06474891261368129,
      "grad_norm": 1.296875,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.7095,
      "mean_token_accuracy": 0.1126905307173729,
      "num_tokens": 1505435.0,
      "step": 655
    },
    {
      "entropy": 6.948776197433472,
      "epoch": 0.06524317912218268,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.803,
      "mean_token_accuracy": 0.10531418174505233,
      "num_tokens": 1518510.0,
      "step": 660
    },
    {
      "entropy": 6.9253791809082035,
      "epoch": 0.06573744563068407,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.7145,
      "mean_token_accuracy": 0.11463218927383423,
      "num_tokens": 1530218.0,
      "step": 665
    },
    {
      "entropy": 6.91641206741333,
      "epoch": 0.06623171213918545,
      "grad_norm": 1.078125,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.7582,
      "mean_token_accuracy": 0.11466718763113022,
      "num_tokens": 1542486.0,
      "step": 670
    },
    {
      "entropy": 6.8444969177246096,
      "epoch": 0.06672597864768683,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000337,
      "loss": 6.7533,
      "mean_token_accuracy": 0.10930409133434296,
      "num_tokens": 1553873.0,
      "step": 675
    },
    {
      "entropy": 6.942787122726441,
      "epoch": 0.06722024515618821,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003395,
      "loss": 6.7047,
      "mean_token_accuracy": 0.1153480589389801,
      "num_tokens": 1565343.0,
      "step": 680
    },
    {
      "entropy": 6.903558826446533,
      "epoch": 0.0677145116646896,
      "grad_norm": 1.078125,
      "learning_rate": 0.000342,
      "loss": 6.7608,
      "mean_token_accuracy": 0.10652155056595802,
      "num_tokens": 1577061.0,
      "step": 685
    },
    {
      "entropy": 6.872377204895019,
      "epoch": 0.06820877817319099,
      "grad_norm": 1.0625,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.7529,
      "mean_token_accuracy": 0.11244799494743347,
      "num_tokens": 1588083.0,
      "step": 690
    },
    {
      "entropy": 6.879619789123535,
      "epoch": 0.06870304468169237,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000347,
      "loss": 6.7159,
      "mean_token_accuracy": 0.10822473615407943,
      "num_tokens": 1599796.0,
      "step": 695
    },
    {
      "entropy": 6.90326623916626,
      "epoch": 0.06919731119019375,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003495,
      "loss": 6.7681,
      "mean_token_accuracy": 0.10771963074803352,
      "num_tokens": 1611994.0,
      "step": 700
    },
    {
      "entropy": 6.872850322723389,
      "epoch": 0.06969157769869513,
      "grad_norm": 1.0625,
      "learning_rate": 0.000352,
      "loss": 6.7022,
      "mean_token_accuracy": 0.10573551878333091,
      "num_tokens": 1623280.0,
      "step": 705
    },
    {
      "entropy": 6.8690471172332765,
      "epoch": 0.07018584420719652,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003545,
      "loss": 6.7681,
      "mean_token_accuracy": 0.10785022675991059,
      "num_tokens": 1634989.0,
      "step": 710
    },
    {
      "entropy": 6.947969198226929,
      "epoch": 0.07068011071569791,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000357,
      "loss": 6.7925,
      "mean_token_accuracy": 0.10478616431355477,
      "num_tokens": 1647042.0,
      "step": 715
    },
    {
      "entropy": 6.809376001358032,
      "epoch": 0.0711743772241993,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003595,
      "loss": 6.6855,
      "mean_token_accuracy": 0.1154635414481163,
      "num_tokens": 1659245.0,
      "step": 720
    },
    {
      "entropy": 6.81630311012268,
      "epoch": 0.07166864373270067,
      "grad_norm": 1.140625,
      "learning_rate": 0.000362,
      "loss": 6.6819,
      "mean_token_accuracy": 0.10983811244368553,
      "num_tokens": 1670271.0,
      "step": 725
    },
    {
      "entropy": 6.873712110519409,
      "epoch": 0.07216291024120206,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003645,
      "loss": 6.735,
      "mean_token_accuracy": 0.11080555841326714,
      "num_tokens": 1682648.0,
      "step": 730
    },
    {
      "entropy": 6.822260570526123,
      "epoch": 0.07265717674970344,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000367,
      "loss": 6.7573,
      "mean_token_accuracy": 0.1052017480134964,
      "num_tokens": 1694542.0,
      "step": 735
    },
    {
      "entropy": 6.849833726882935,
      "epoch": 0.07315144325820483,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003695,
      "loss": 6.6493,
      "mean_token_accuracy": 0.10704348310828209,
      "num_tokens": 1705737.0,
      "step": 740
    },
    {
      "entropy": 6.862112379074096,
      "epoch": 0.0736457097667062,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000372,
      "loss": 6.71,
      "mean_token_accuracy": 0.10965840369462967,
      "num_tokens": 1718474.0,
      "step": 745
    },
    {
      "entropy": 6.826199579238891,
      "epoch": 0.07413997627520759,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003745,
      "loss": 6.6873,
      "mean_token_accuracy": 0.1082968644797802,
      "num_tokens": 1729360.0,
      "step": 750
    },
    {
      "entropy": 6.8160795211792,
      "epoch": 0.07463424278370898,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000377,
      "loss": 6.6964,
      "mean_token_accuracy": 0.1107470378279686,
      "num_tokens": 1739654.0,
      "step": 755
    },
    {
      "entropy": 6.836647272109985,
      "epoch": 0.07512850929221036,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003795,
      "loss": 6.6939,
      "mean_token_accuracy": 0.11239867880940438,
      "num_tokens": 1752158.0,
      "step": 760
    },
    {
      "entropy": 6.834174489974975,
      "epoch": 0.07562277580071175,
      "grad_norm": 1.171875,
      "learning_rate": 0.000382,
      "loss": 6.6923,
      "mean_token_accuracy": 0.11612112373113632,
      "num_tokens": 1764545.0,
      "step": 765
    },
    {
      "entropy": 6.821158313751221,
      "epoch": 0.07611704230921312,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003845,
      "loss": 6.6814,
      "mean_token_accuracy": 0.1128039114177227,
      "num_tokens": 1775422.0,
      "step": 770
    },
    {
      "entropy": 6.840568494796753,
      "epoch": 0.07661130881771451,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.7088,
      "mean_token_accuracy": 0.10114984959363937,
      "num_tokens": 1787833.0,
      "step": 775
    },
    {
      "entropy": 6.74714674949646,
      "epoch": 0.0771055753262159,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.5723,
      "mean_token_accuracy": 0.1219009131193161,
      "num_tokens": 1798508.0,
      "step": 780
    },
    {
      "entropy": 6.753849983215332,
      "epoch": 0.07759984183471728,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.6209,
      "mean_token_accuracy": 0.11469837054610252,
      "num_tokens": 1809873.0,
      "step": 785
    },
    {
      "entropy": 6.821059226989746,
      "epoch": 0.07809410834321866,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.7014,
      "mean_token_accuracy": 0.11237634271383286,
      "num_tokens": 1820876.0,
      "step": 790
    },
    {
      "entropy": 6.7968470573425295,
      "epoch": 0.07858837485172004,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.663,
      "mean_token_accuracy": 0.10904834121465683,
      "num_tokens": 1833233.0,
      "step": 795
    },
    {
      "entropy": 6.812273120880127,
      "epoch": 0.07908264136022143,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003995,
      "loss": 6.7481,
      "mean_token_accuracy": 0.10785818174481392,
      "num_tokens": 1844520.0,
      "step": 800
    },
    {
      "entropy": 6.765308666229248,
      "epoch": 0.07957690786872282,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000402,
      "loss": 6.6292,
      "mean_token_accuracy": 0.1142186164855957,
      "num_tokens": 1856814.0,
      "step": 805
    },
    {
      "entropy": 6.732009649276733,
      "epoch": 0.0800711743772242,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004045,
      "loss": 6.6196,
      "mean_token_accuracy": 0.11811698004603385,
      "num_tokens": 1867581.0,
      "step": 810
    },
    {
      "entropy": 6.84562578201294,
      "epoch": 0.08056544088572558,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.6907,
      "mean_token_accuracy": 0.10590304806828499,
      "num_tokens": 1879302.0,
      "step": 815
    },
    {
      "entropy": 6.756962633132934,
      "epoch": 0.08105970739422697,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004095,
      "loss": 6.6571,
      "mean_token_accuracy": 0.10640354081988335,
      "num_tokens": 1891697.0,
      "step": 820
    },
    {
      "entropy": 6.759096002578735,
      "epoch": 0.08155397390272835,
      "grad_norm": 1.078125,
      "learning_rate": 0.000412,
      "loss": 6.5979,
      "mean_token_accuracy": 0.11474953293800354,
      "num_tokens": 1903544.0,
      "step": 825
    },
    {
      "entropy": 6.774603033065796,
      "epoch": 0.08204824041122974,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004145,
      "loss": 6.6545,
      "mean_token_accuracy": 0.11101827323436737,
      "num_tokens": 1915286.0,
      "step": 830
    },
    {
      "entropy": 6.688508129119873,
      "epoch": 0.08254250691973111,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000417,
      "loss": 6.5346,
      "mean_token_accuracy": 0.11861704438924789,
      "num_tokens": 1926361.0,
      "step": 835
    },
    {
      "entropy": 6.751745128631592,
      "epoch": 0.0830367734282325,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004195,
      "loss": 6.6299,
      "mean_token_accuracy": 0.11083293855190277,
      "num_tokens": 1938540.0,
      "step": 840
    },
    {
      "entropy": 6.763704109191894,
      "epoch": 0.08353103993673389,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000422,
      "loss": 6.6603,
      "mean_token_accuracy": 0.10724062249064445,
      "num_tokens": 1950532.0,
      "step": 845
    },
    {
      "entropy": 6.8038843154907225,
      "epoch": 0.08402530644523527,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004245,
      "loss": 6.6778,
      "mean_token_accuracy": 0.10919453650712967,
      "num_tokens": 1960899.0,
      "step": 850
    },
    {
      "entropy": 6.68118052482605,
      "epoch": 0.08451957295373666,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000427,
      "loss": 6.6496,
      "mean_token_accuracy": 0.11730073392391205,
      "num_tokens": 1972278.0,
      "step": 855
    },
    {
      "entropy": 6.725075197219849,
      "epoch": 0.08501383946223803,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004295,
      "loss": 6.6692,
      "mean_token_accuracy": 0.1092053584754467,
      "num_tokens": 1985079.0,
      "step": 860
    },
    {
      "entropy": 6.635829877853394,
      "epoch": 0.08550810597073942,
      "grad_norm": 1.15625,
      "learning_rate": 0.000432,
      "loss": 6.7065,
      "mean_token_accuracy": 0.10706321522593498,
      "num_tokens": 1996467.0,
      "step": 865
    },
    {
      "entropy": 6.8284436702728275,
      "epoch": 0.08600237247924081,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004345,
      "loss": 6.5832,
      "mean_token_accuracy": 0.11370615139603615,
      "num_tokens": 2007339.0,
      "step": 870
    },
    {
      "entropy": 6.677596378326416,
      "epoch": 0.0864966389877422,
      "grad_norm": 1.015625,
      "learning_rate": 0.000437,
      "loss": 6.5879,
      "mean_token_accuracy": 0.1196245089173317,
      "num_tokens": 2019962.0,
      "step": 875
    },
    {
      "entropy": 6.747453260421753,
      "epoch": 0.08699090549624358,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004395,
      "loss": 6.5866,
      "mean_token_accuracy": 0.11758355721831322,
      "num_tokens": 2031251.0,
      "step": 880
    },
    {
      "entropy": 6.661469602584839,
      "epoch": 0.08748517200474495,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000442,
      "loss": 6.6013,
      "mean_token_accuracy": 0.11161932870745658,
      "num_tokens": 2042412.0,
      "step": 885
    },
    {
      "entropy": 6.6397772312164305,
      "epoch": 0.08797943851324634,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004445,
      "loss": 6.6221,
      "mean_token_accuracy": 0.11210491955280304,
      "num_tokens": 2054423.0,
      "step": 890
    },
    {
      "entropy": 6.686139965057373,
      "epoch": 0.08847370502174773,
      "grad_norm": 1.046875,
      "learning_rate": 0.000447,
      "loss": 6.595,
      "mean_token_accuracy": 0.10692231729626656,
      "num_tokens": 2065657.0,
      "step": 895
    },
    {
      "entropy": 6.717379331588745,
      "epoch": 0.08896797153024912,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.5827,
      "mean_token_accuracy": 0.11325386464595795,
      "num_tokens": 2076305.0,
      "step": 900
    },
    {
      "entropy": 6.697859430313111,
      "epoch": 0.08946223803875049,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.5783,
      "mean_token_accuracy": 0.11316154897212982,
      "num_tokens": 2087734.0,
      "step": 905
    },
    {
      "entropy": 6.595232200622559,
      "epoch": 0.08995650454725188,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.4953,
      "mean_token_accuracy": 0.12035880610346794,
      "num_tokens": 2098016.0,
      "step": 910
    },
    {
      "entropy": 6.665337467193604,
      "epoch": 0.09045077105575326,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.6678,
      "mean_token_accuracy": 0.10886707603931427,
      "num_tokens": 2110293.0,
      "step": 915
    },
    {
      "entropy": 6.673768377304077,
      "epoch": 0.09094503756425465,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.4485,
      "mean_token_accuracy": 0.11682527139782906,
      "num_tokens": 2120441.0,
      "step": 920
    },
    {
      "entropy": 6.539246511459351,
      "epoch": 0.09143930407275604,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000462,
      "loss": 6.5683,
      "mean_token_accuracy": 0.11887063086032867,
      "num_tokens": 2131520.0,
      "step": 925
    },
    {
      "entropy": 6.664091300964356,
      "epoch": 0.09193357058125741,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004645,
      "loss": 6.5796,
      "mean_token_accuracy": 0.11637971252202987,
      "num_tokens": 2145005.0,
      "step": 930
    },
    {
      "entropy": 6.605041790008545,
      "epoch": 0.0924278370897588,
      "grad_norm": 1.0625,
      "learning_rate": 0.000467,
      "loss": 6.5338,
      "mean_token_accuracy": 0.1211809828877449,
      "num_tokens": 2156608.0,
      "step": 935
    },
    {
      "entropy": 6.601588439941406,
      "epoch": 0.09292210359826018,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004695,
      "loss": 6.5318,
      "mean_token_accuracy": 0.11969754993915557,
      "num_tokens": 2168341.0,
      "step": 940
    },
    {
      "entropy": 6.625442409515381,
      "epoch": 0.09341637010676157,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000472,
      "loss": 6.5258,
      "mean_token_accuracy": 0.11236160099506379,
      "num_tokens": 2178704.0,
      "step": 945
    },
    {
      "entropy": 6.704224395751953,
      "epoch": 0.09391063661526294,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004745,
      "loss": 6.6906,
      "mean_token_accuracy": 0.10928287953138352,
      "num_tokens": 2190884.0,
      "step": 950
    },
    {
      "entropy": 6.682482194900513,
      "epoch": 0.09440490312376433,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000477,
      "loss": 6.598,
      "mean_token_accuracy": 0.1167626291513443,
      "num_tokens": 2201822.0,
      "step": 955
    },
    {
      "entropy": 6.629432439804077,
      "epoch": 0.09489916963226572,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004795,
      "loss": 6.5926,
      "mean_token_accuracy": 0.11254436299204826,
      "num_tokens": 2213362.0,
      "step": 960
    },
    {
      "entropy": 6.692664337158203,
      "epoch": 0.0953934361407671,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000482,
      "loss": 6.602,
      "mean_token_accuracy": 0.11311232969164849,
      "num_tokens": 2224366.0,
      "step": 965
    },
    {
      "entropy": 6.46690821647644,
      "epoch": 0.09588770264926849,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004845,
      "loss": 6.4535,
      "mean_token_accuracy": 0.1190331868827343,
      "num_tokens": 2234645.0,
      "step": 970
    },
    {
      "entropy": 6.678108978271484,
      "epoch": 0.09638196915776986,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000487,
      "loss": 6.5498,
      "mean_token_accuracy": 0.12018159404397011,
      "num_tokens": 2245904.0,
      "step": 975
    },
    {
      "entropy": 6.4710277080535885,
      "epoch": 0.09687623566627125,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004895,
      "loss": 6.49,
      "mean_token_accuracy": 0.11942838132381439,
      "num_tokens": 2257753.0,
      "step": 980
    },
    {
      "entropy": 6.551316976547241,
      "epoch": 0.09737050217477264,
      "grad_norm": 1.03125,
      "learning_rate": 0.000492,
      "loss": 6.4953,
      "mean_token_accuracy": 0.1244705080986023,
      "num_tokens": 2269009.0,
      "step": 985
    },
    {
      "entropy": 6.693281459808349,
      "epoch": 0.09786476868327403,
      "grad_norm": 1.0,
      "learning_rate": 0.0004945,
      "loss": 6.5493,
      "mean_token_accuracy": 0.11529032737016678,
      "num_tokens": 2279840.0,
      "step": 990
    },
    {
      "entropy": 6.533772134780884,
      "epoch": 0.0983590351917754,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000497,
      "loss": 6.5154,
      "mean_token_accuracy": 0.11877169013023377,
      "num_tokens": 2290577.0,
      "step": 995
    },
    {
      "entropy": 6.553924179077148,
      "epoch": 0.09885330170027679,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995,
      "loss": 6.5274,
      "mean_token_accuracy": 0.12048136815428734,
      "num_tokens": 2301341.0,
      "step": 1000
    },
    {
      "entropy": 6.566124486923218,
      "epoch": 0.09934756820877817,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999999982287888,
      "loss": 6.5896,
      "mean_token_accuracy": 0.11639286801218987,
      "num_tokens": 2312505.0,
      "step": 1005
    },
    {
      "entropy": 6.718512153625488,
      "epoch": 0.09984183471727956,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999999910332434,
      "loss": 6.5236,
      "mean_token_accuracy": 0.11419004276394844,
      "num_tokens": 2323443.0,
      "step": 1010
    },
    {
      "entropy": 6.4791271686553955,
      "epoch": 0.10033610122578095,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999999783026635,
      "loss": 6.5439,
      "mean_token_accuracy": 0.11673406213521957,
      "num_tokens": 2335686.0,
      "step": 1015
    },
    {
      "entropy": 6.514018297195435,
      "epoch": 0.10083036773428232,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999999600370491,
      "loss": 6.5178,
      "mean_token_accuracy": 0.11814445331692695,
      "num_tokens": 2347273.0,
      "step": 1020
    },
    {
      "entropy": 6.552177906036377,
      "epoch": 0.1013246342427837,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999999362364008,
      "loss": 6.4427,
      "mean_token_accuracy": 0.12624893859028816,
      "num_tokens": 2358011.0,
      "step": 1025
    },
    {
      "entropy": 6.5004723072052,
      "epoch": 0.1018189007512851,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999999069007191,
      "loss": 6.5815,
      "mean_token_accuracy": 0.11536967605352402,
      "num_tokens": 2371390.0,
      "step": 1030
    },
    {
      "entropy": 6.596110677719116,
      "epoch": 0.10231316725978648,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999998720300048,
      "loss": 6.5271,
      "mean_token_accuracy": 0.12005563005805016,
      "num_tokens": 2381887.0,
      "step": 1035
    },
    {
      "entropy": 6.551437664031982,
      "epoch": 0.10280743376828787,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999998316242589,
      "loss": 6.4663,
      "mean_token_accuracy": 0.11999196782708169,
      "num_tokens": 2393632.0,
      "step": 1040
    },
    {
      "entropy": 6.530908870697021,
      "epoch": 0.10330170027678924,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999997856834821,
      "loss": 6.4778,
      "mean_token_accuracy": 0.12028345614671707,
      "num_tokens": 2405686.0,
      "step": 1045
    },
    {
      "entropy": 6.457987642288208,
      "epoch": 0.10379596678529063,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999997342076757,
      "loss": 6.455,
      "mean_token_accuracy": 0.11674800738692284,
      "num_tokens": 2417459.0,
      "step": 1050
    },
    {
      "entropy": 6.636799621582031,
      "epoch": 0.10429023329379201,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999996771968408,
      "loss": 6.4995,
      "mean_token_accuracy": 0.11853041127324104,
      "num_tokens": 2428689.0,
      "step": 1055
    },
    {
      "entropy": 6.509030675888061,
      "epoch": 0.1047844998022934,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999996146509791,
      "loss": 6.5295,
      "mean_token_accuracy": 0.11135807186365128,
      "num_tokens": 2439437.0,
      "step": 1060
    },
    {
      "entropy": 6.580507850646972,
      "epoch": 0.10527876631079477,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999995465700919,
      "loss": 6.5104,
      "mean_token_accuracy": 0.11756298467516899,
      "num_tokens": 2450576.0,
      "step": 1065
    },
    {
      "entropy": 6.587932443618774,
      "epoch": 0.10577303281929616,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999994729541809,
      "loss": 6.6509,
      "mean_token_accuracy": 0.11502083018422127,
      "num_tokens": 2462684.0,
      "step": 1070
    },
    {
      "entropy": 6.633237361907959,
      "epoch": 0.10626729932779755,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004999993938032481,
      "loss": 6.5244,
      "mean_token_accuracy": 0.12519583106040955,
      "num_tokens": 2474193.0,
      "step": 1075
    },
    {
      "entropy": 6.394119358062744,
      "epoch": 0.10676156583629894,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999993091172952,
      "loss": 6.4655,
      "mean_token_accuracy": 0.11539505198597907,
      "num_tokens": 2486264.0,
      "step": 1080
    },
    {
      "entropy": 6.555998420715332,
      "epoch": 0.10725583234480032,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999992188963245,
      "loss": 6.4541,
      "mean_token_accuracy": 0.12652762085199357,
      "num_tokens": 2497668.0,
      "step": 1085
    },
    {
      "entropy": 6.488944244384766,
      "epoch": 0.1077500988533017,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499999123140338,
      "loss": 6.5113,
      "mean_token_accuracy": 0.11666428670287132,
      "num_tokens": 2509635.0,
      "step": 1090
    },
    {
      "entropy": 6.545294523239136,
      "epoch": 0.10824436536180308,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999990218493382,
      "loss": 6.5051,
      "mean_token_accuracy": 0.12337316423654557,
      "num_tokens": 2521980.0,
      "step": 1095
    },
    {
      "entropy": 6.497365427017212,
      "epoch": 0.10873863187030447,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999989150233276,
      "loss": 6.4639,
      "mean_token_accuracy": 0.11867155879735947,
      "num_tokens": 2533750.0,
      "step": 1100
    },
    {
      "entropy": 6.565949201583862,
      "epoch": 0.10923289837880586,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999988026623088,
      "loss": 6.5618,
      "mean_token_accuracy": 0.11584960371255874,
      "num_tokens": 2545620.0,
      "step": 1105
    },
    {
      "entropy": 6.454707479476928,
      "epoch": 0.10972716488730723,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999986847662843,
      "loss": 6.3302,
      "mean_token_accuracy": 0.13472548872232437,
      "num_tokens": 2556463.0,
      "step": 1110
    },
    {
      "entropy": 6.503189325332642,
      "epoch": 0.11022143139580862,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999985613352576,
      "loss": 6.5358,
      "mean_token_accuracy": 0.12201443389058113,
      "num_tokens": 2567396.0,
      "step": 1115
    },
    {
      "entropy": 6.544538831710815,
      "epoch": 0.11071569790431,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999984323692311,
      "loss": 6.48,
      "mean_token_accuracy": 0.123195381462574,
      "num_tokens": 2579208.0,
      "step": 1120
    },
    {
      "entropy": 6.412357044219971,
      "epoch": 0.11120996441281139,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999982978682085,
      "loss": 6.4248,
      "mean_token_accuracy": 0.12385009601712227,
      "num_tokens": 2590940.0,
      "step": 1125
    },
    {
      "entropy": 6.558205413818359,
      "epoch": 0.11170423092131278,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999981578321927,
      "loss": 6.51,
      "mean_token_accuracy": 0.1274452067911625,
      "num_tokens": 2602491.0,
      "step": 1130
    },
    {
      "entropy": 6.439068698883057,
      "epoch": 0.11219849742981415,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004999980122611874,
      "loss": 6.3438,
      "mean_token_accuracy": 0.12809207662940025,
      "num_tokens": 2612139.0,
      "step": 1135
    },
    {
      "entropy": 6.445886325836182,
      "epoch": 0.11269276393831554,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499997861155196,
      "loss": 6.5104,
      "mean_token_accuracy": 0.12374393939971924,
      "num_tokens": 2624906.0,
      "step": 1140
    },
    {
      "entropy": 6.474299335479737,
      "epoch": 0.11318703044681692,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999977045142223,
      "loss": 6.4204,
      "mean_token_accuracy": 0.12185963913798332,
      "num_tokens": 2636046.0,
      "step": 1145
    },
    {
      "entropy": 6.521690130233765,
      "epoch": 0.11368129695531831,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999975423382703,
      "loss": 6.4288,
      "mean_token_accuracy": 0.13111222088336943,
      "num_tokens": 2645617.0,
      "step": 1150
    },
    {
      "entropy": 6.534403610229492,
      "epoch": 0.11417556346381968,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999973746273438,
      "loss": 6.4529,
      "mean_token_accuracy": 0.12487476840615272,
      "num_tokens": 2656780.0,
      "step": 1155
    },
    {
      "entropy": 6.454224348068237,
      "epoch": 0.11466982997232107,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499997201381447,
      "loss": 6.438,
      "mean_token_accuracy": 0.1251967266201973,
      "num_tokens": 2668078.0,
      "step": 1160
    },
    {
      "entropy": 6.491998386383057,
      "epoch": 0.11516409648082246,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999970226005841,
      "loss": 6.5242,
      "mean_token_accuracy": 0.11767346188426017,
      "num_tokens": 2679049.0,
      "step": 1165
    },
    {
      "entropy": 6.53035717010498,
      "epoch": 0.11565836298932385,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999968382847595,
      "loss": 6.4624,
      "mean_token_accuracy": 0.12308069095015525,
      "num_tokens": 2690165.0,
      "step": 1170
    },
    {
      "entropy": 6.536982250213623,
      "epoch": 0.11615262949782523,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999966484339778,
      "loss": 6.5002,
      "mean_token_accuracy": 0.116629458963871,
      "num_tokens": 2701965.0,
      "step": 1175
    },
    {
      "entropy": 6.402723407745361,
      "epoch": 0.1166468960063266,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999964530482437,
      "loss": 6.421,
      "mean_token_accuracy": 0.12439692988991738,
      "num_tokens": 2713479.0,
      "step": 1180
    },
    {
      "entropy": 6.465710687637329,
      "epoch": 0.11714116251482799,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999962521275619,
      "loss": 6.436,
      "mean_token_accuracy": 0.12075797766447068,
      "num_tokens": 2725629.0,
      "step": 1185
    },
    {
      "entropy": 6.268183088302612,
      "epoch": 0.11763542902332938,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999960456719374,
      "loss": 6.3492,
      "mean_token_accuracy": 0.12263342142105102,
      "num_tokens": 2736620.0,
      "step": 1190
    },
    {
      "entropy": 6.464852428436279,
      "epoch": 0.11812969553183077,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999958336813753,
      "loss": 6.3325,
      "mean_token_accuracy": 0.1303306147456169,
      "num_tokens": 2747118.0,
      "step": 1195
    },
    {
      "entropy": 6.372828674316406,
      "epoch": 0.11862396204033215,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999956161558807,
      "loss": 6.3799,
      "mean_token_accuracy": 0.1274876520037651,
      "num_tokens": 2758371.0,
      "step": 1200
    },
    {
      "entropy": 6.391465854644776,
      "epoch": 0.11911822854883353,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999953930954591,
      "loss": 6.3852,
      "mean_token_accuracy": 0.12252886295318603,
      "num_tokens": 2770575.0,
      "step": 1205
    },
    {
      "entropy": 6.448058319091797,
      "epoch": 0.11961249505733491,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999951645001159,
      "loss": 6.3455,
      "mean_token_accuracy": 0.1301012769341469,
      "num_tokens": 2783103.0,
      "step": 1210
    },
    {
      "entropy": 6.436516189575196,
      "epoch": 0.1201067615658363,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004999949303698568,
      "loss": 6.4665,
      "mean_token_accuracy": 0.12414569854736328,
      "num_tokens": 2796636.0,
      "step": 1215
    },
    {
      "entropy": 6.406609773635864,
      "epoch": 0.12060102807433769,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999946907046875,
      "loss": 6.322,
      "mean_token_accuracy": 0.13122917115688323,
      "num_tokens": 2807798.0,
      "step": 1220
    },
    {
      "entropy": 6.362497234344483,
      "epoch": 0.12109529458283906,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999944455046139,
      "loss": 6.439,
      "mean_token_accuracy": 0.12562692314386367,
      "num_tokens": 2819255.0,
      "step": 1225
    },
    {
      "entropy": 6.454278612136841,
      "epoch": 0.12158956109134045,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499994194769642,
      "loss": 6.4012,
      "mean_token_accuracy": 0.1260200709104538,
      "num_tokens": 2829952.0,
      "step": 1230
    },
    {
      "entropy": 6.518359088897705,
      "epoch": 0.12208382759984183,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499993938499778,
      "loss": 6.4773,
      "mean_token_accuracy": 0.1282096892595291,
      "num_tokens": 2841056.0,
      "step": 1235
    },
    {
      "entropy": 6.377371597290039,
      "epoch": 0.12257809410834322,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999936766950283,
      "loss": 6.3282,
      "mean_token_accuracy": 0.13203255608677864,
      "num_tokens": 2852045.0,
      "step": 1240
    },
    {
      "entropy": 6.4568657875061035,
      "epoch": 0.12307236061684461,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999934093553992,
      "loss": 6.434,
      "mean_token_accuracy": 0.12442487627267837,
      "num_tokens": 2863440.0,
      "step": 1245
    },
    {
      "entropy": 6.430741882324218,
      "epoch": 0.12356662712534598,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999931364808972,
      "loss": 6.4369,
      "mean_token_accuracy": 0.1217712052166462,
      "num_tokens": 2874017.0,
      "step": 1250
    },
    {
      "entropy": 6.364642333984375,
      "epoch": 0.12406089363384737,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999928580715293,
      "loss": 6.3657,
      "mean_token_accuracy": 0.1264870472252369,
      "num_tokens": 2885677.0,
      "step": 1255
    },
    {
      "entropy": 6.433004665374756,
      "epoch": 0.12455516014234876,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499992574127302,
      "loss": 6.4812,
      "mean_token_accuracy": 0.1225816085934639,
      "num_tokens": 2897825.0,
      "step": 1260
    },
    {
      "entropy": 6.524859094619751,
      "epoch": 0.12504942665085014,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999922846482226,
      "loss": 6.5257,
      "mean_token_accuracy": 0.1220851257443428,
      "num_tokens": 2909304.0,
      "step": 1265
    },
    {
      "entropy": 6.429646253585815,
      "epoch": 0.12554369315935152,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999919896342981,
      "loss": 6.4588,
      "mean_token_accuracy": 0.12074072510004044,
      "num_tokens": 2920918.0,
      "step": 1270
    },
    {
      "entropy": 6.501003408432007,
      "epoch": 0.12603795966785292,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999916890855358,
      "loss": 6.4073,
      "mean_token_accuracy": 0.12426696866750717,
      "num_tokens": 2932294.0,
      "step": 1275
    },
    {
      "entropy": 6.363391876220703,
      "epoch": 0.1265322261763543,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999913830019431,
      "loss": 6.3237,
      "mean_token_accuracy": 0.13384533002972604,
      "num_tokens": 2943372.0,
      "step": 1280
    },
    {
      "entropy": 6.311191463470459,
      "epoch": 0.12702649268485566,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999910713835272,
      "loss": 6.3047,
      "mean_token_accuracy": 0.13015900626778604,
      "num_tokens": 2953807.0,
      "step": 1285
    },
    {
      "entropy": 6.342788410186768,
      "epoch": 0.12752075919335706,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999907542302962,
      "loss": 6.2438,
      "mean_token_accuracy": 0.13418981581926345,
      "num_tokens": 2964372.0,
      "step": 1290
    },
    {
      "entropy": 6.376524639129639,
      "epoch": 0.12801502570185844,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999904315422577,
      "loss": 6.3365,
      "mean_token_accuracy": 0.1274750992655754,
      "num_tokens": 2975537.0,
      "step": 1295
    },
    {
      "entropy": 6.296396970748901,
      "epoch": 0.12850929221035984,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999901033194198,
      "loss": 6.3172,
      "mean_token_accuracy": 0.12614937424659728,
      "num_tokens": 2986094.0,
      "step": 1300
    },
    {
      "entropy": 6.412167549133301,
      "epoch": 0.1290035587188612,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999897695617904,
      "loss": 6.4284,
      "mean_token_accuracy": 0.12951384037733077,
      "num_tokens": 2996023.0,
      "step": 1305
    },
    {
      "entropy": 6.477496385574341,
      "epoch": 0.12949782522736258,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999894302693778,
      "loss": 6.4201,
      "mean_token_accuracy": 0.12168471738696099,
      "num_tokens": 3007584.0,
      "step": 1310
    },
    {
      "entropy": 6.470839118957519,
      "epoch": 0.12999209173586398,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999890854421902,
      "loss": 6.4722,
      "mean_token_accuracy": 0.11812516897916794,
      "num_tokens": 3018995.0,
      "step": 1315
    },
    {
      "entropy": 6.263270092010498,
      "epoch": 0.13048635824436536,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999887350802362,
      "loss": 6.2577,
      "mean_token_accuracy": 0.13282620459794997,
      "num_tokens": 3030576.0,
      "step": 1320
    },
    {
      "entropy": 6.371448612213134,
      "epoch": 0.13098062475286676,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999883791835246,
      "loss": 6.3224,
      "mean_token_accuracy": 0.12991052865982056,
      "num_tokens": 3040765.0,
      "step": 1325
    },
    {
      "entropy": 6.39703049659729,
      "epoch": 0.13147489126136813,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999880177520637,
      "loss": 6.3713,
      "mean_token_accuracy": 0.12290074527263642,
      "num_tokens": 3053912.0,
      "step": 1330
    },
    {
      "entropy": 6.269509172439575,
      "epoch": 0.1319691577698695,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999876507858628,
      "loss": 6.358,
      "mean_token_accuracy": 0.12943168208003045,
      "num_tokens": 3065921.0,
      "step": 1335
    },
    {
      "entropy": 6.491255092620849,
      "epoch": 0.1324634242783709,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999872782849308,
      "loss": 6.4118,
      "mean_token_accuracy": 0.1217479057610035,
      "num_tokens": 3078643.0,
      "step": 1340
    },
    {
      "entropy": 6.297278165817261,
      "epoch": 0.13295769078687228,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999869002492768,
      "loss": 6.3262,
      "mean_token_accuracy": 0.12477339208126068,
      "num_tokens": 3090300.0,
      "step": 1345
    },
    {
      "entropy": 6.313536787033081,
      "epoch": 0.13345195729537365,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999865166789102,
      "loss": 6.2574,
      "mean_token_accuracy": 0.13334870636463164,
      "num_tokens": 3102307.0,
      "step": 1350
    },
    {
      "entropy": 6.430685234069824,
      "epoch": 0.13394622380387505,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999861275738404,
      "loss": 6.4038,
      "mean_token_accuracy": 0.12200917750597,
      "num_tokens": 3113817.0,
      "step": 1355
    },
    {
      "entropy": 6.360499048233033,
      "epoch": 0.13444049031237643,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999857329340769,
      "loss": 6.2516,
      "mean_token_accuracy": 0.12839845269918443,
      "num_tokens": 3124719.0,
      "step": 1360
    },
    {
      "entropy": 6.246741867065429,
      "epoch": 0.13493475682087783,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999853327596295,
      "loss": 6.2873,
      "mean_token_accuracy": 0.1282256707549095,
      "num_tokens": 3136136.0,
      "step": 1365
    },
    {
      "entropy": 6.370300054550171,
      "epoch": 0.1354290233293792,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499984927050508,
      "loss": 6.3012,
      "mean_token_accuracy": 0.1290183626115322,
      "num_tokens": 3148977.0,
      "step": 1370
    },
    {
      "entropy": 6.377018070220947,
      "epoch": 0.13592328983788057,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999845158067223,
      "loss": 6.3579,
      "mean_token_accuracy": 0.1252500332891941,
      "num_tokens": 3160520.0,
      "step": 1375
    },
    {
      "entropy": 6.366725778579712,
      "epoch": 0.13641755634638197,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999840990282828,
      "loss": 6.3426,
      "mean_token_accuracy": 0.12962477132678032,
      "num_tokens": 3172895.0,
      "step": 1380
    },
    {
      "entropy": 6.324902153015136,
      "epoch": 0.13691182285488335,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999836767151994,
      "loss": 6.3544,
      "mean_token_accuracy": 0.13052070513367653,
      "num_tokens": 3183960.0,
      "step": 1385
    },
    {
      "entropy": 6.330108976364135,
      "epoch": 0.13740608936338475,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999832488674827,
      "loss": 6.3886,
      "mean_token_accuracy": 0.1273997098207474,
      "num_tokens": 3194276.0,
      "step": 1390
    },
    {
      "entropy": 6.366259002685547,
      "epoch": 0.13790035587188612,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999828154851432,
      "loss": 6.2558,
      "mean_token_accuracy": 0.13061744794249536,
      "num_tokens": 3206084.0,
      "step": 1395
    },
    {
      "entropy": 6.301025104522705,
      "epoch": 0.1383946223803875,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999823765681915,
      "loss": 6.2972,
      "mean_token_accuracy": 0.13314774185419082,
      "num_tokens": 3217941.0,
      "step": 1400
    },
    {
      "entropy": 6.338549423217773,
      "epoch": 0.1388888888888889,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999819321166385,
      "loss": 6.2742,
      "mean_token_accuracy": 0.13190385103225707,
      "num_tokens": 3229088.0,
      "step": 1405
    },
    {
      "entropy": 6.25566258430481,
      "epoch": 0.13938315539739027,
      "grad_norm": 1.0,
      "learning_rate": 0.000499981482130495,
      "loss": 6.2205,
      "mean_token_accuracy": 0.13236010000109671,
      "num_tokens": 3240708.0,
      "step": 1410
    },
    {
      "entropy": 6.294620752334595,
      "epoch": 0.13987742190589167,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999810266097721,
      "loss": 6.3444,
      "mean_token_accuracy": 0.1264272078871727,
      "num_tokens": 3251040.0,
      "step": 1415
    },
    {
      "entropy": 6.291800880432129,
      "epoch": 0.14037168841439304,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999805655544811,
      "loss": 6.2912,
      "mean_token_accuracy": 0.13168664127588273,
      "num_tokens": 3262655.0,
      "step": 1420
    },
    {
      "entropy": 6.291258907318115,
      "epoch": 0.14086595492289442,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999800989646333,
      "loss": 6.2228,
      "mean_token_accuracy": 0.13803994432091712,
      "num_tokens": 3273160.0,
      "step": 1425
    },
    {
      "entropy": 6.300457239151001,
      "epoch": 0.14136022143139582,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999796268402401,
      "loss": 6.2783,
      "mean_token_accuracy": 0.1296892285346985,
      "num_tokens": 3285932.0,
      "step": 1430
    },
    {
      "entropy": 6.249773216247559,
      "epoch": 0.1418544879398972,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999791491813132,
      "loss": 6.2971,
      "mean_token_accuracy": 0.12768914625048639,
      "num_tokens": 3298834.0,
      "step": 1435
    },
    {
      "entropy": 6.409033393859863,
      "epoch": 0.1423487544483986,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999786659878643,
      "loss": 6.2799,
      "mean_token_accuracy": 0.12664505764842032,
      "num_tokens": 3310473.0,
      "step": 1440
    },
    {
      "entropy": 6.182092523574829,
      "epoch": 0.14284302095689996,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999781772599055,
      "loss": 6.2497,
      "mean_token_accuracy": 0.13217792585492133,
      "num_tokens": 3322662.0,
      "step": 1445
    },
    {
      "entropy": 6.391679096221924,
      "epoch": 0.14333728746540134,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999776829974484,
      "loss": 6.3475,
      "mean_token_accuracy": 0.132525734603405,
      "num_tokens": 3334321.0,
      "step": 1450
    },
    {
      "entropy": 6.246479558944702,
      "epoch": 0.14383155397390274,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999771832005055,
      "loss": 6.2512,
      "mean_token_accuracy": 0.13095758333802224,
      "num_tokens": 3345130.0,
      "step": 1455
    },
    {
      "entropy": 6.2692060470581055,
      "epoch": 0.1443258204824041,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499976677869089,
      "loss": 6.2271,
      "mean_token_accuracy": 0.13088599741458892,
      "num_tokens": 3355601.0,
      "step": 1460
    },
    {
      "entropy": 6.194102716445923,
      "epoch": 0.14482008699090548,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999761670032115,
      "loss": 6.2295,
      "mean_token_accuracy": 0.13360995799303055,
      "num_tokens": 3367075.0,
      "step": 1465
    },
    {
      "entropy": 6.32319974899292,
      "epoch": 0.14531435349940688,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999756506028851,
      "loss": 6.1937,
      "mean_token_accuracy": 0.13329043239355087,
      "num_tokens": 3378888.0,
      "step": 1470
    },
    {
      "entropy": 6.248707866668701,
      "epoch": 0.14580862000790826,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999751286681229,
      "loss": 6.2325,
      "mean_token_accuracy": 0.1328907147049904,
      "num_tokens": 3390705.0,
      "step": 1475
    },
    {
      "entropy": 6.288841915130615,
      "epoch": 0.14630288651640966,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999746011989376,
      "loss": 6.249,
      "mean_token_accuracy": 0.1373384565114975,
      "num_tokens": 3401710.0,
      "step": 1480
    },
    {
      "entropy": 6.346017694473266,
      "epoch": 0.14679715302491103,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999740681953422,
      "loss": 6.3789,
      "mean_token_accuracy": 0.1264251612126827,
      "num_tokens": 3413467.0,
      "step": 1485
    },
    {
      "entropy": 6.331368350982666,
      "epoch": 0.1472914195334124,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999735296573499,
      "loss": 6.3007,
      "mean_token_accuracy": 0.13292821869254112,
      "num_tokens": 3425885.0,
      "step": 1490
    },
    {
      "entropy": 6.329574871063232,
      "epoch": 0.1477856860419138,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004999729855849738,
      "loss": 6.3151,
      "mean_token_accuracy": 0.12490678578615189,
      "num_tokens": 3437952.0,
      "step": 1495
    },
    {
      "entropy": 6.267677640914917,
      "epoch": 0.14827995255041518,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999724359782274,
      "loss": 6.1911,
      "mean_token_accuracy": 0.1455492317676544,
      "num_tokens": 3449680.0,
      "step": 1500
    },
    {
      "entropy": 6.317681789398193,
      "epoch": 0.14877421905891658,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999718808371241,
      "loss": 6.3432,
      "mean_token_accuracy": 0.12537492960691451,
      "num_tokens": 3460699.0,
      "step": 1505
    },
    {
      "entropy": 6.338581228256226,
      "epoch": 0.14926848556741795,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999713201616776,
      "loss": 6.2238,
      "mean_token_accuracy": 0.13232341259717942,
      "num_tokens": 3471919.0,
      "step": 1510
    },
    {
      "entropy": 6.166346454620362,
      "epoch": 0.14976275207591933,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499970753951902,
      "loss": 6.3229,
      "mean_token_accuracy": 0.12744686827063562,
      "num_tokens": 3484110.0,
      "step": 1515
    },
    {
      "entropy": 6.412110614776611,
      "epoch": 0.15025701858442073,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999701822078107,
      "loss": 6.2352,
      "mean_token_accuracy": 0.12983164712786674,
      "num_tokens": 3496103.0,
      "step": 1520
    },
    {
      "entropy": 6.272561883926391,
      "epoch": 0.1507512850929221,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499969604929418,
      "loss": 6.3098,
      "mean_token_accuracy": 0.13165246173739434,
      "num_tokens": 3507291.0,
      "step": 1525
    },
    {
      "entropy": 6.302027225494385,
      "epoch": 0.1512455516014235,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999690221167382,
      "loss": 6.3173,
      "mean_token_accuracy": 0.12720877900719643,
      "num_tokens": 3519564.0,
      "step": 1530
    },
    {
      "entropy": 6.204291439056396,
      "epoch": 0.15173981810992487,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999684337697855,
      "loss": 6.1724,
      "mean_token_accuracy": 0.13716719225049018,
      "num_tokens": 3530679.0,
      "step": 1535
    },
    {
      "entropy": 6.2982782363891605,
      "epoch": 0.15223408461842625,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999678398885746,
      "loss": 6.2473,
      "mean_token_accuracy": 0.1318476304411888,
      "num_tokens": 3542662.0,
      "step": 1540
    },
    {
      "entropy": 6.265805768966675,
      "epoch": 0.15272835112692765,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004999672404731197,
      "loss": 6.3499,
      "mean_token_accuracy": 0.12811536714434624,
      "num_tokens": 3554760.0,
      "step": 1545
    },
    {
      "entropy": 6.249004793167114,
      "epoch": 0.15322261763542902,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499966635523436,
      "loss": 6.1602,
      "mean_token_accuracy": 0.13934924975037574,
      "num_tokens": 3565604.0,
      "step": 1550
    },
    {
      "entropy": 6.277536392211914,
      "epoch": 0.1537168841439304,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999660250395381,
      "loss": 6.3141,
      "mean_token_accuracy": 0.12867220491170883,
      "num_tokens": 3577535.0,
      "step": 1555
    },
    {
      "entropy": 6.295968151092529,
      "epoch": 0.1542111506524318,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499965409021441,
      "loss": 6.3235,
      "mean_token_accuracy": 0.12822821363806725,
      "num_tokens": 3590458.0,
      "step": 1560
    },
    {
      "entropy": 6.375429725646972,
      "epoch": 0.15470541716093317,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00049996478746916,
      "loss": 6.2851,
      "mean_token_accuracy": 0.13392977118492128,
      "num_tokens": 3600916.0,
      "step": 1565
    },
    {
      "entropy": 6.222699928283691,
      "epoch": 0.15519968366943457,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999641603827104,
      "loss": 6.3037,
      "mean_token_accuracy": 0.12990911155939103,
      "num_tokens": 3612794.0,
      "step": 1570
    },
    {
      "entropy": 6.27130880355835,
      "epoch": 0.15569395017793594,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999635277621073,
      "loss": 6.2037,
      "mean_token_accuracy": 0.1420078307390213,
      "num_tokens": 3624186.0,
      "step": 1575
    },
    {
      "entropy": 6.199323654174805,
      "epoch": 0.15618821668643731,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999628896073668,
      "loss": 6.2448,
      "mean_token_accuracy": 0.12877988070249557,
      "num_tokens": 3636041.0,
      "step": 1580
    },
    {
      "entropy": 6.304298067092896,
      "epoch": 0.15668248319493872,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999622459185042,
      "loss": 6.3071,
      "mean_token_accuracy": 0.12442853599786759,
      "num_tokens": 3646859.0,
      "step": 1585
    },
    {
      "entropy": 6.2107603549957275,
      "epoch": 0.1571767497034401,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999615966955354,
      "loss": 6.119,
      "mean_token_accuracy": 0.14032085835933686,
      "num_tokens": 3657853.0,
      "step": 1590
    },
    {
      "entropy": 6.324536895751953,
      "epoch": 0.1576710162119415,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999609419384764,
      "loss": 6.2253,
      "mean_token_accuracy": 0.134918150305748,
      "num_tokens": 3668341.0,
      "step": 1595
    },
    {
      "entropy": 6.249271965026855,
      "epoch": 0.15816528272044286,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999602816473434,
      "loss": 6.2563,
      "mean_token_accuracy": 0.13581479713320732,
      "num_tokens": 3679020.0,
      "step": 1600
    },
    {
      "entropy": 6.1892296314239506,
      "epoch": 0.15865954922894424,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999596158221526,
      "loss": 6.1613,
      "mean_token_accuracy": 0.13848758935928346,
      "num_tokens": 3691121.0,
      "step": 1605
    },
    {
      "entropy": 6.234239339828491,
      "epoch": 0.15915381573744564,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999589444629203,
      "loss": 6.2286,
      "mean_token_accuracy": 0.13591036275029184,
      "num_tokens": 3702584.0,
      "step": 1610
    },
    {
      "entropy": 6.20807032585144,
      "epoch": 0.159648082245947,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999582675696631,
      "loss": 6.177,
      "mean_token_accuracy": 0.1353621505200863,
      "num_tokens": 3713085.0,
      "step": 1615
    },
    {
      "entropy": 6.320692300796509,
      "epoch": 0.1601423487544484,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999575851423976,
      "loss": 6.2929,
      "mean_token_accuracy": 0.12485695034265518,
      "num_tokens": 3724318.0,
      "step": 1620
    },
    {
      "entropy": 6.2754554271698,
      "epoch": 0.16063661526294978,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999568971811407,
      "loss": 6.1999,
      "mean_token_accuracy": 0.14148797765374183,
      "num_tokens": 3734288.0,
      "step": 1625
    },
    {
      "entropy": 6.201929521560669,
      "epoch": 0.16113088177145116,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499956203685909,
      "loss": 6.1941,
      "mean_token_accuracy": 0.13511522933840753,
      "num_tokens": 3744645.0,
      "step": 1630
    },
    {
      "entropy": 6.282017421722412,
      "epoch": 0.16162514827995256,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00049995550465672,
      "loss": 6.1854,
      "mean_token_accuracy": 0.13890446051955224,
      "num_tokens": 3755660.0,
      "step": 1635
    },
    {
      "entropy": 6.125180339813232,
      "epoch": 0.16211941478845393,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999548000935906,
      "loss": 6.1599,
      "mean_token_accuracy": 0.14167107045650482,
      "num_tokens": 3767614.0,
      "step": 1640
    },
    {
      "entropy": 6.254905891418457,
      "epoch": 0.16261368129695533,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999540899965383,
      "loss": 6.308,
      "mean_token_accuracy": 0.1335205040872097,
      "num_tokens": 3780750.0,
      "step": 1645
    },
    {
      "entropy": 6.242093515396118,
      "epoch": 0.1631079478054567,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999533743655803,
      "loss": 6.2965,
      "mean_token_accuracy": 0.1281206913292408,
      "num_tokens": 3793195.0,
      "step": 1650
    },
    {
      "entropy": 6.24614839553833,
      "epoch": 0.16360221431395808,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999526532007346,
      "loss": 6.2606,
      "mean_token_accuracy": 0.13350182473659516,
      "num_tokens": 3804256.0,
      "step": 1655
    },
    {
      "entropy": 6.289435911178589,
      "epoch": 0.16409648082245948,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999519265020187,
      "loss": 6.2237,
      "mean_token_accuracy": 0.12859297171235085,
      "num_tokens": 3815403.0,
      "step": 1660
    },
    {
      "entropy": 6.08995943069458,
      "epoch": 0.16459074733096085,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999511942694504,
      "loss": 6.128,
      "mean_token_accuracy": 0.14284673556685448,
      "num_tokens": 3826682.0,
      "step": 1665
    },
    {
      "entropy": 6.215432214736938,
      "epoch": 0.16508501383946222,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999504565030479,
      "loss": 6.139,
      "mean_token_accuracy": 0.13716638311743737,
      "num_tokens": 3837780.0,
      "step": 1670
    },
    {
      "entropy": 6.164974164962769,
      "epoch": 0.16557928034796363,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999497132028293,
      "loss": 6.2072,
      "mean_token_accuracy": 0.1344188041985035,
      "num_tokens": 3848782.0,
      "step": 1675
    },
    {
      "entropy": 6.244879055023193,
      "epoch": 0.166073546856465,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999489643688127,
      "loss": 6.1649,
      "mean_token_accuracy": 0.13939150497317315,
      "num_tokens": 3859220.0,
      "step": 1680
    },
    {
      "entropy": 6.102788019180298,
      "epoch": 0.1665678133649664,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999482100010169,
      "loss": 6.0808,
      "mean_token_accuracy": 0.1425606593489647,
      "num_tokens": 3869794.0,
      "step": 1685
    },
    {
      "entropy": 6.161266660690307,
      "epoch": 0.16706207987346777,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999474500994601,
      "loss": 6.2184,
      "mean_token_accuracy": 0.1334006331861019,
      "num_tokens": 3881133.0,
      "step": 1690
    },
    {
      "entropy": 6.2152763366699215,
      "epoch": 0.16755634638196915,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999466846641611,
      "loss": 6.1567,
      "mean_token_accuracy": 0.1364052601158619,
      "num_tokens": 3891600.0,
      "step": 1695
    },
    {
      "entropy": 6.224546241760254,
      "epoch": 0.16805061289047055,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999459136951389,
      "loss": 6.2529,
      "mean_token_accuracy": 0.13441457897424697,
      "num_tokens": 3903287.0,
      "step": 1700
    },
    {
      "entropy": 6.245505905151367,
      "epoch": 0.16854487939897192,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999451371924123,
      "loss": 6.1635,
      "mean_token_accuracy": 0.14323566406965255,
      "num_tokens": 3914497.0,
      "step": 1705
    },
    {
      "entropy": 6.2008990287780765,
      "epoch": 0.16903914590747332,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999443551560003,
      "loss": 6.1996,
      "mean_token_accuracy": 0.13277414664626122,
      "num_tokens": 3925536.0,
      "step": 1710
    },
    {
      "entropy": 6.179309701919555,
      "epoch": 0.1695334124159747,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999435675859224,
      "loss": 6.1432,
      "mean_token_accuracy": 0.1426541492342949,
      "num_tokens": 3936705.0,
      "step": 1715
    },
    {
      "entropy": 6.117032384872436,
      "epoch": 0.17002767892447607,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999427744821977,
      "loss": 6.0382,
      "mean_token_accuracy": 0.1389991022646427,
      "num_tokens": 3948422.0,
      "step": 1720
    },
    {
      "entropy": 6.144522857666016,
      "epoch": 0.17052194543297747,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499941975844846,
      "loss": 6.1085,
      "mean_token_accuracy": 0.14725376516580582,
      "num_tokens": 3958697.0,
      "step": 1725
    },
    {
      "entropy": 6.112959909439087,
      "epoch": 0.17101621194147884,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999411716738866,
      "loss": 6.1208,
      "mean_token_accuracy": 0.13783455789089202,
      "num_tokens": 3970255.0,
      "step": 1730
    },
    {
      "entropy": 6.206806802749634,
      "epoch": 0.17151047844998024,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999403619693397,
      "loss": 6.2172,
      "mean_token_accuracy": 0.13266652897000314,
      "num_tokens": 3980892.0,
      "step": 1735
    },
    {
      "entropy": 6.101211071014404,
      "epoch": 0.17200474495848161,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999395467312249,
      "loss": 6.1532,
      "mean_token_accuracy": 0.13273056447505951,
      "num_tokens": 3991375.0,
      "step": 1740
    },
    {
      "entropy": 6.304367208480835,
      "epoch": 0.172499011466983,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999387259595623,
      "loss": 6.3042,
      "mean_token_accuracy": 0.12940156012773513,
      "num_tokens": 4003352.0,
      "step": 1745
    },
    {
      "entropy": 6.152772092819214,
      "epoch": 0.1729932779754844,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999378996543723,
      "loss": 6.1745,
      "mean_token_accuracy": 0.13783681988716126,
      "num_tokens": 4014429.0,
      "step": 1750
    },
    {
      "entropy": 6.22617883682251,
      "epoch": 0.17348754448398576,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999370678156749,
      "loss": 6.1056,
      "mean_token_accuracy": 0.14252740070223807,
      "num_tokens": 4025213.0,
      "step": 1755
    },
    {
      "entropy": 6.1182598114013675,
      "epoch": 0.17398181099248716,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999362304434909,
      "loss": 6.0983,
      "mean_token_accuracy": 0.1411632165312767,
      "num_tokens": 4036244.0,
      "step": 1760
    },
    {
      "entropy": 6.218455600738525,
      "epoch": 0.17447607750098854,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999353875378406,
      "loss": 6.1392,
      "mean_token_accuracy": 0.13833620101213456,
      "num_tokens": 4047944.0,
      "step": 1765
    },
    {
      "entropy": 6.125010585784912,
      "epoch": 0.1749703440094899,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499934539098745,
      "loss": 6.1127,
      "mean_token_accuracy": 0.13938926681876182,
      "num_tokens": 4058865.0,
      "step": 1770
    },
    {
      "entropy": 6.112168598175049,
      "epoch": 0.1754646105179913,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999336851262247,
      "loss": 6.1605,
      "mean_token_accuracy": 0.13652415424585343,
      "num_tokens": 4070627.0,
      "step": 1775
    },
    {
      "entropy": 6.221715450286865,
      "epoch": 0.17595887702649268,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999328256203009,
      "loss": 6.2029,
      "mean_token_accuracy": 0.13283938243985177,
      "num_tokens": 4081406.0,
      "step": 1780
    },
    {
      "entropy": 6.202430009841919,
      "epoch": 0.17645314353499406,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999319605809947,
      "loss": 6.1763,
      "mean_token_accuracy": 0.1280331313610077,
      "num_tokens": 4092957.0,
      "step": 1785
    },
    {
      "entropy": 6.209701681137085,
      "epoch": 0.17694741004349546,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999310900083274,
      "loss": 6.232,
      "mean_token_accuracy": 0.13681847751140594,
      "num_tokens": 4104381.0,
      "step": 1790
    },
    {
      "entropy": 6.180554962158203,
      "epoch": 0.17744167655199683,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999302139023204,
      "loss": 6.1452,
      "mean_token_accuracy": 0.13721615672111512,
      "num_tokens": 4117190.0,
      "step": 1795
    },
    {
      "entropy": 6.11452693939209,
      "epoch": 0.17793594306049823,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999293322629951,
      "loss": 6.0752,
      "mean_token_accuracy": 0.14162272289395333,
      "num_tokens": 4127870.0,
      "step": 1800
    },
    {
      "entropy": 6.098488092422485,
      "epoch": 0.1784302095689996,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999284450903736,
      "loss": 6.1272,
      "mean_token_accuracy": 0.14127160534262656,
      "num_tokens": 4139240.0,
      "step": 1805
    },
    {
      "entropy": 6.223893737792968,
      "epoch": 0.17892447607750098,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999275523844772,
      "loss": 6.0978,
      "mean_token_accuracy": 0.14117365032434465,
      "num_tokens": 4149885.0,
      "step": 1810
    },
    {
      "entropy": 6.183619403839112,
      "epoch": 0.17941874258600238,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999266541453283,
      "loss": 6.2727,
      "mean_token_accuracy": 0.1341182231903076,
      "num_tokens": 4161824.0,
      "step": 1815
    },
    {
      "entropy": 6.192480278015137,
      "epoch": 0.17991300909450375,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999257503729487,
      "loss": 6.2305,
      "mean_token_accuracy": 0.13566601723432542,
      "num_tokens": 4173581.0,
      "step": 1820
    },
    {
      "entropy": 6.188820171356201,
      "epoch": 0.18040727560300515,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999248410673607,
      "loss": 6.1086,
      "mean_token_accuracy": 0.14354413226246834,
      "num_tokens": 4184843.0,
      "step": 1825
    },
    {
      "entropy": 6.199634456634522,
      "epoch": 0.18090154211150652,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999239262285869,
      "loss": 6.196,
      "mean_token_accuracy": 0.12894793301820756,
      "num_tokens": 4195016.0,
      "step": 1830
    },
    {
      "entropy": 6.183595895767212,
      "epoch": 0.1813958086200079,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999230058566495,
      "loss": 6.149,
      "mean_token_accuracy": 0.13868849724531174,
      "num_tokens": 4206798.0,
      "step": 1835
    },
    {
      "entropy": 6.20193657875061,
      "epoch": 0.1818900751285093,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999220799515713,
      "loss": 6.1197,
      "mean_token_accuracy": 0.13616999015212058,
      "num_tokens": 4218784.0,
      "step": 1840
    },
    {
      "entropy": 6.098097658157348,
      "epoch": 0.18238434163701067,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499921148513375,
      "loss": 6.1495,
      "mean_token_accuracy": 0.13528309762477875,
      "num_tokens": 4229875.0,
      "step": 1845
    },
    {
      "entropy": 6.120871686935425,
      "epoch": 0.18287860814551207,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999202115420836,
      "loss": 6.1274,
      "mean_token_accuracy": 0.13959604278206825,
      "num_tokens": 4241801.0,
      "step": 1850
    },
    {
      "entropy": 6.136850357055664,
      "epoch": 0.18337287465401345,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999192690377201,
      "loss": 6.112,
      "mean_token_accuracy": 0.13544971346855164,
      "num_tokens": 4253168.0,
      "step": 1855
    },
    {
      "entropy": 6.151134204864502,
      "epoch": 0.18386714116251482,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999183210003076,
      "loss": 6.1371,
      "mean_token_accuracy": 0.13654277250170707,
      "num_tokens": 4264791.0,
      "step": 1860
    },
    {
      "entropy": 6.127550983428955,
      "epoch": 0.18436140767101622,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999173674298696,
      "loss": 6.0802,
      "mean_token_accuracy": 0.1429269403219223,
      "num_tokens": 4275816.0,
      "step": 1865
    },
    {
      "entropy": 6.2200188636779785,
      "epoch": 0.1848556741795176,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999164083264294,
      "loss": 6.1638,
      "mean_token_accuracy": 0.13557607755064965,
      "num_tokens": 4287806.0,
      "step": 1870
    },
    {
      "entropy": 6.1365429878234865,
      "epoch": 0.185349940688019,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999154436900108,
      "loss": 6.2061,
      "mean_token_accuracy": 0.13776358291506768,
      "num_tokens": 4300319.0,
      "step": 1875
    },
    {
      "entropy": 6.261188077926636,
      "epoch": 0.18584420719652037,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999144735206373,
      "loss": 6.2191,
      "mean_token_accuracy": 0.13050539270043374,
      "num_tokens": 4310987.0,
      "step": 1880
    },
    {
      "entropy": 6.134886789321899,
      "epoch": 0.18633847370502174,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999134978183328,
      "loss": 6.1589,
      "mean_token_accuracy": 0.14048967361450196,
      "num_tokens": 4322438.0,
      "step": 1885
    },
    {
      "entropy": 6.216308641433716,
      "epoch": 0.18683274021352314,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999125165831215,
      "loss": 6.194,
      "mean_token_accuracy": 0.13951309472322465,
      "num_tokens": 4333242.0,
      "step": 1890
    },
    {
      "entropy": 6.1126538753509525,
      "epoch": 0.1873270067220245,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999115298150273,
      "loss": 6.0433,
      "mean_token_accuracy": 0.14102249965071678,
      "num_tokens": 4343660.0,
      "step": 1895
    },
    {
      "entropy": 6.163443613052368,
      "epoch": 0.1878212732305259,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999105375140746,
      "loss": 6.1246,
      "mean_token_accuracy": 0.1411570593714714,
      "num_tokens": 4353382.0,
      "step": 1900
    },
    {
      "entropy": 6.103315830230713,
      "epoch": 0.1883155397390273,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999095396802878,
      "loss": 6.1044,
      "mean_token_accuracy": 0.13109617531299592,
      "num_tokens": 4364614.0,
      "step": 1905
    },
    {
      "entropy": 6.2517712116241455,
      "epoch": 0.18880980624752866,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999085363136913,
      "loss": 6.1812,
      "mean_token_accuracy": 0.1353604979813099,
      "num_tokens": 4376001.0,
      "step": 1910
    },
    {
      "entropy": 6.128733491897583,
      "epoch": 0.18930407275603006,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00049990752741431,
      "loss": 6.1848,
      "mean_token_accuracy": 0.13217584118247033,
      "num_tokens": 4387083.0,
      "step": 1915
    },
    {
      "entropy": 6.139519214630127,
      "epoch": 0.18979833926453143,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999065129821687,
      "loss": 6.1356,
      "mean_token_accuracy": 0.13750590831041337,
      "num_tokens": 4399977.0,
      "step": 1920
    },
    {
      "entropy": 6.0825254917144775,
      "epoch": 0.1902926057730328,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999054930172923,
      "loss": 6.0592,
      "mean_token_accuracy": 0.14218987300992011,
      "num_tokens": 4411609.0,
      "step": 1925
    },
    {
      "entropy": 6.175016450881958,
      "epoch": 0.1907868722815342,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999044675197058,
      "loss": 6.1817,
      "mean_token_accuracy": 0.13701564595103263,
      "num_tokens": 4423567.0,
      "step": 1930
    },
    {
      "entropy": 6.147806310653687,
      "epoch": 0.19128113879003558,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999034364894345,
      "loss": 6.1472,
      "mean_token_accuracy": 0.13876281678676605,
      "num_tokens": 4435410.0,
      "step": 1935
    },
    {
      "entropy": 6.18219690322876,
      "epoch": 0.19177540529853698,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999023999265039,
      "loss": 6.1697,
      "mean_token_accuracy": 0.1378607526421547,
      "num_tokens": 4445279.0,
      "step": 1940
    },
    {
      "entropy": 6.077285814285278,
      "epoch": 0.19226967180703836,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999013578309392,
      "loss": 6.0385,
      "mean_token_accuracy": 0.14334971085190773,
      "num_tokens": 4456017.0,
      "step": 1945
    },
    {
      "entropy": 6.126475811004639,
      "epoch": 0.19276393831553973,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999003102027663,
      "loss": 6.0311,
      "mean_token_accuracy": 0.14338691681623458,
      "num_tokens": 4466759.0,
      "step": 1950
    },
    {
      "entropy": 6.117775201797485,
      "epoch": 0.19325820482404113,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499899257042011,
      "loss": 6.1139,
      "mean_token_accuracy": 0.1352181151509285,
      "num_tokens": 4478316.0,
      "step": 1955
    },
    {
      "entropy": 6.208990001678467,
      "epoch": 0.1937524713325425,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998981983486988,
      "loss": 6.1751,
      "mean_token_accuracy": 0.1322692520916462,
      "num_tokens": 4490775.0,
      "step": 1960
    },
    {
      "entropy": 6.064915752410888,
      "epoch": 0.1942467378410439,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998971341228563,
      "loss": 6.039,
      "mean_token_accuracy": 0.1443925455212593,
      "num_tokens": 4501622.0,
      "step": 1965
    },
    {
      "entropy": 6.149889087677002,
      "epoch": 0.19474100434954528,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998960643645093,
      "loss": 6.1338,
      "mean_token_accuracy": 0.13769015669822693,
      "num_tokens": 4512279.0,
      "step": 1970
    },
    {
      "entropy": 6.167524814605713,
      "epoch": 0.19523527085804665,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998949890736841,
      "loss": 6.1239,
      "mean_token_accuracy": 0.14070178642868997,
      "num_tokens": 4524087.0,
      "step": 1975
    },
    {
      "entropy": 6.139242649078369,
      "epoch": 0.19572953736654805,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998939082504075,
      "loss": 6.1075,
      "mean_token_accuracy": 0.13732974529266356,
      "num_tokens": 4534816.0,
      "step": 1980
    },
    {
      "entropy": 6.103002882003784,
      "epoch": 0.19622380387504942,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998928218947057,
      "loss": 6.0825,
      "mean_token_accuracy": 0.13870148733258247,
      "num_tokens": 4545725.0,
      "step": 1985
    },
    {
      "entropy": 6.218700933456421,
      "epoch": 0.1967180703835508,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004998917300066056,
      "loss": 6.1897,
      "mean_token_accuracy": 0.13517609983682632,
      "num_tokens": 4557003.0,
      "step": 1990
    },
    {
      "entropy": 6.099284267425537,
      "epoch": 0.1972123368920522,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998906325861342,
      "loss": 6.1023,
      "mean_token_accuracy": 0.13518477007746696,
      "num_tokens": 4568768.0,
      "step": 1995
    },
    {
      "entropy": 6.103130531311035,
      "epoch": 0.19770660340055357,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998895296333183,
      "loss": 6.0715,
      "mean_token_accuracy": 0.14329598397016524,
      "num_tokens": 4581116.0,
      "step": 2000
    },
    {
      "entropy": 6.170775890350342,
      "epoch": 0.19820086990905497,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000499888421148185,
      "loss": 6.1516,
      "mean_token_accuracy": 0.13928177580237389,
      "num_tokens": 4592660.0,
      "step": 2005
    },
    {
      "entropy": 6.074905729293823,
      "epoch": 0.19869513641755635,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998873071307617,
      "loss": 6.1321,
      "mean_token_accuracy": 0.13926766365766524,
      "num_tokens": 4603057.0,
      "step": 2010
    },
    {
      "entropy": 6.111081838607788,
      "epoch": 0.19918940292605772,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004998861875810758,
      "loss": 6.0145,
      "mean_token_accuracy": 0.14594943076372147,
      "num_tokens": 4613546.0,
      "step": 2015
    },
    {
      "entropy": 6.130213642120362,
      "epoch": 0.19968366943455912,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998850624991548,
      "loss": 6.0932,
      "mean_token_accuracy": 0.13966753333806992,
      "num_tokens": 4624613.0,
      "step": 2020
    },
    {
      "entropy": 6.070725202560425,
      "epoch": 0.2001779359430605,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998839318850263,
      "loss": 6.0581,
      "mean_token_accuracy": 0.14254929646849632,
      "num_tokens": 4636004.0,
      "step": 2025
    },
    {
      "entropy": 6.094811868667603,
      "epoch": 0.2006722024515619,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998827957387182,
      "loss": 6.1194,
      "mean_token_accuracy": 0.1468837432563305,
      "num_tokens": 4646450.0,
      "step": 2030
    },
    {
      "entropy": 6.105382490158081,
      "epoch": 0.20116646896006327,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998816540602584,
      "loss": 6.0993,
      "mean_token_accuracy": 0.14319370239973067,
      "num_tokens": 4658928.0,
      "step": 2035
    },
    {
      "entropy": 6.133113861083984,
      "epoch": 0.20166073546856464,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998805068496752,
      "loss": 6.1279,
      "mean_token_accuracy": 0.13475392609834672,
      "num_tokens": 4670392.0,
      "step": 2040
    },
    {
      "entropy": 6.141398954391479,
      "epoch": 0.20215500197706604,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998793541069964,
      "loss": 6.108,
      "mean_token_accuracy": 0.13688452690839767,
      "num_tokens": 4680605.0,
      "step": 2045
    },
    {
      "entropy": 6.069099378585816,
      "epoch": 0.2026492684855674,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998781958322507,
      "loss": 6.0134,
      "mean_token_accuracy": 0.14460673555731773,
      "num_tokens": 4694250.0,
      "step": 2050
    },
    {
      "entropy": 6.110835599899292,
      "epoch": 0.20314353499406881,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998770320254666,
      "loss": 6.1661,
      "mean_token_accuracy": 0.13903781324625014,
      "num_tokens": 4705754.0,
      "step": 2055
    },
    {
      "entropy": 6.213265514373779,
      "epoch": 0.2036378015025702,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998758626866725,
      "loss": 6.0439,
      "mean_token_accuracy": 0.14397211521863937,
      "num_tokens": 4716393.0,
      "step": 2060
    },
    {
      "entropy": 5.992143011093139,
      "epoch": 0.20413206801107156,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998746878158973,
      "loss": 6.055,
      "mean_token_accuracy": 0.14380352646112443,
      "num_tokens": 4726894.0,
      "step": 2065
    },
    {
      "entropy": 6.137889289855957,
      "epoch": 0.20462633451957296,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998735074131699,
      "loss": 6.082,
      "mean_token_accuracy": 0.13775463551282882,
      "num_tokens": 4737488.0,
      "step": 2070
    },
    {
      "entropy": 6.157661390304566,
      "epoch": 0.20512060102807433,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998723214785194,
      "loss": 6.1528,
      "mean_token_accuracy": 0.1327391304075718,
      "num_tokens": 4749131.0,
      "step": 2075
    },
    {
      "entropy": 6.090861511230469,
      "epoch": 0.20561486753657574,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998711300119749,
      "loss": 6.0322,
      "mean_token_accuracy": 0.1410945810377598,
      "num_tokens": 4760503.0,
      "step": 2080
    },
    {
      "entropy": 6.128068828582764,
      "epoch": 0.2061091340450771,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998699330135656,
      "loss": 6.1567,
      "mean_token_accuracy": 0.1358460932970047,
      "num_tokens": 4771965.0,
      "step": 2085
    },
    {
      "entropy": 6.096466827392578,
      "epoch": 0.20660340055357848,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998687304833212,
      "loss": 6.1254,
      "mean_token_accuracy": 0.13651231974363326,
      "num_tokens": 4784215.0,
      "step": 2090
    },
    {
      "entropy": 6.102608966827392,
      "epoch": 0.20709766706207988,
      "grad_norm": 1.078125,
      "learning_rate": 0.000499867522421271,
      "loss": 6.0457,
      "mean_token_accuracy": 0.1427777349948883,
      "num_tokens": 4796410.0,
      "step": 2095
    },
    {
      "entropy": 6.065385389328003,
      "epoch": 0.20759193357058126,
      "grad_norm": 1.109375,
      "learning_rate": 0.000499866308827445,
      "loss": 6.0463,
      "mean_token_accuracy": 0.1456620767712593,
      "num_tokens": 4806990.0,
      "step": 2100
    },
    {
      "entropy": 6.113207626342773,
      "epoch": 0.20808620007908263,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998650897018729,
      "loss": 6.0949,
      "mean_token_accuracy": 0.14114938974380492,
      "num_tokens": 4818903.0,
      "step": 2105
    },
    {
      "entropy": 6.091670751571655,
      "epoch": 0.20858046658758403,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998638650445847,
      "loss": 6.0222,
      "mean_token_accuracy": 0.14184736907482148,
      "num_tokens": 4829130.0,
      "step": 2110
    },
    {
      "entropy": 6.0775425910949705,
      "epoch": 0.2090747330960854,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998626348556105,
      "loss": 6.0507,
      "mean_token_accuracy": 0.1536749005317688,
      "num_tokens": 4840424.0,
      "step": 2115
    },
    {
      "entropy": 6.098685884475708,
      "epoch": 0.2095689996045868,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998613991349807,
      "loss": 6.0283,
      "mean_token_accuracy": 0.143294495344162,
      "num_tokens": 4851674.0,
      "step": 2120
    },
    {
      "entropy": 5.9475749969482425,
      "epoch": 0.21006326611308818,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998601578827255,
      "loss": 5.9793,
      "mean_token_accuracy": 0.1536559596657753,
      "num_tokens": 4863884.0,
      "step": 2125
    },
    {
      "entropy": 6.033754873275757,
      "epoch": 0.21055753262158955,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998589110988755,
      "loss": 6.0494,
      "mean_token_accuracy": 0.14718157574534416,
      "num_tokens": 4875582.0,
      "step": 2130
    },
    {
      "entropy": 6.114961338043213,
      "epoch": 0.21105179913009095,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998576587834614,
      "loss": 6.0803,
      "mean_token_accuracy": 0.14199472591280937,
      "num_tokens": 4888325.0,
      "step": 2135
    },
    {
      "entropy": 6.098699426651001,
      "epoch": 0.21154606563859232,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499856400936514,
      "loss": 6.072,
      "mean_token_accuracy": 0.14264639914035798,
      "num_tokens": 4899631.0,
      "step": 2140
    },
    {
      "entropy": 6.029827070236206,
      "epoch": 0.21204033214709372,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998551375580643,
      "loss": 6.0712,
      "mean_token_accuracy": 0.13763367459177972,
      "num_tokens": 4910954.0,
      "step": 2145
    },
    {
      "entropy": 6.051994514465332,
      "epoch": 0.2125345986555951,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998538686481435,
      "loss": 6.0714,
      "mean_token_accuracy": 0.14147689938545227,
      "num_tokens": 4921952.0,
      "step": 2150
    },
    {
      "entropy": 6.0656555652618405,
      "epoch": 0.21302886516409647,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998525942067824,
      "loss": 6.0253,
      "mean_token_accuracy": 0.14360493123531343,
      "num_tokens": 4933984.0,
      "step": 2155
    },
    {
      "entropy": 6.164171361923218,
      "epoch": 0.21352313167259787,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998513142340128,
      "loss": 6.1657,
      "mean_token_accuracy": 0.14157587438821792,
      "num_tokens": 4945926.0,
      "step": 2160
    },
    {
      "entropy": 6.026093053817749,
      "epoch": 0.21401739818109924,
      "grad_norm": 1.125,
      "learning_rate": 0.0004998500287298657,
      "loss": 6.0878,
      "mean_token_accuracy": 0.14866259396076204,
      "num_tokens": 4958123.0,
      "step": 2165
    },
    {
      "entropy": 6.158975839614868,
      "epoch": 0.21451166468960065,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998487376943733,
      "loss": 6.1491,
      "mean_token_accuracy": 0.13338658437132836,
      "num_tokens": 4970409.0,
      "step": 2170
    },
    {
      "entropy": 6.122834396362305,
      "epoch": 0.21500593119810202,
      "grad_norm": 1.09375,
      "learning_rate": 0.000499847441127567,
      "loss": 6.0485,
      "mean_token_accuracy": 0.13795322477817534,
      "num_tokens": 4980429.0,
      "step": 2175
    },
    {
      "entropy": 6.053185510635376,
      "epoch": 0.2155001977066034,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998461390294787,
      "loss": 6.0711,
      "mean_token_accuracy": 0.14336730688810348,
      "num_tokens": 4991790.0,
      "step": 2180
    },
    {
      "entropy": 6.095970296859742,
      "epoch": 0.2159944642151048,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998448314001405,
      "loss": 6.0869,
      "mean_token_accuracy": 0.14493904113769532,
      "num_tokens": 5003611.0,
      "step": 2185
    },
    {
      "entropy": 6.176206493377686,
      "epoch": 0.21648873072360617,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004998435182395845,
      "loss": 6.1385,
      "mean_token_accuracy": 0.1309673383831978,
      "num_tokens": 5014824.0,
      "step": 2190
    },
    {
      "entropy": 6.142057228088379,
      "epoch": 0.21698299723210757,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998421995478432,
      "loss": 6.0831,
      "mean_token_accuracy": 0.143635593354702,
      "num_tokens": 5026937.0,
      "step": 2195
    },
    {
      "entropy": 6.116099071502686,
      "epoch": 0.21747726374060894,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998408753249489,
      "loss": 6.1134,
      "mean_token_accuracy": 0.1369356207549572,
      "num_tokens": 5037996.0,
      "step": 2200
    },
    {
      "entropy": 6.096542501449585,
      "epoch": 0.2179715302491103,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998395455709342,
      "loss": 6.0605,
      "mean_token_accuracy": 0.14553544148802758,
      "num_tokens": 5050944.0,
      "step": 2205
    },
    {
      "entropy": 6.150865888595581,
      "epoch": 0.2184657967576117,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998382102858317,
      "loss": 6.1,
      "mean_token_accuracy": 0.1378478489816189,
      "num_tokens": 5062129.0,
      "step": 2210
    },
    {
      "entropy": 6.0926353454589846,
      "epoch": 0.2189600632661131,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998368694696743,
      "loss": 6.0959,
      "mean_token_accuracy": 0.14366044774651526,
      "num_tokens": 5073318.0,
      "step": 2215
    },
    {
      "entropy": 6.064820623397827,
      "epoch": 0.21945432977461446,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998355231224952,
      "loss": 6.1278,
      "mean_token_accuracy": 0.13671059682965278,
      "num_tokens": 5084607.0,
      "step": 2220
    },
    {
      "entropy": 6.1097174167633055,
      "epoch": 0.21994859628311586,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998341712443273,
      "loss": 6.0165,
      "mean_token_accuracy": 0.14424162060022355,
      "num_tokens": 5096004.0,
      "step": 2225
    },
    {
      "entropy": 6.0761024951934814,
      "epoch": 0.22044286279161723,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499832813835204,
      "loss": 6.0296,
      "mean_token_accuracy": 0.14155133366584777,
      "num_tokens": 5107171.0,
      "step": 2230
    },
    {
      "entropy": 5.977806282043457,
      "epoch": 0.22093712930011863,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998314508951584,
      "loss": 5.9782,
      "mean_token_accuracy": 0.15077303797006608,
      "num_tokens": 5119249.0,
      "step": 2235
    },
    {
      "entropy": 6.04344334602356,
      "epoch": 0.22143139580862,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998300824242244,
      "loss": 5.9945,
      "mean_token_accuracy": 0.13899804279208183,
      "num_tokens": 5130798.0,
      "step": 2240
    },
    {
      "entropy": 6.146693372726441,
      "epoch": 0.22192566231712138,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004998287084224353,
      "loss": 6.1624,
      "mean_token_accuracy": 0.12782063856720924,
      "num_tokens": 5141376.0,
      "step": 2245
    },
    {
      "entropy": 5.979104852676391,
      "epoch": 0.22241992882562278,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998273288898252,
      "loss": 5.9599,
      "mean_token_accuracy": 0.1464287005364895,
      "num_tokens": 5152888.0,
      "step": 2250
    },
    {
      "entropy": 6.031632804870606,
      "epoch": 0.22291419533412415,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499825943826428,
      "loss": 6.0647,
      "mean_token_accuracy": 0.14004838690161706,
      "num_tokens": 5165720.0,
      "step": 2255
    },
    {
      "entropy": 6.069583225250244,
      "epoch": 0.22340846184262556,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998245532322776,
      "loss": 5.9363,
      "mean_token_accuracy": 0.14667778834700584,
      "num_tokens": 5177345.0,
      "step": 2260
    },
    {
      "entropy": 6.060588693618774,
      "epoch": 0.22390272835112693,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004998231571074083,
      "loss": 5.9634,
      "mean_token_accuracy": 0.14570290222764015,
      "num_tokens": 5187385.0,
      "step": 2265
    },
    {
      "entropy": 6.04728593826294,
      "epoch": 0.2243969948596283,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998217554518545,
      "loss": 6.0683,
      "mean_token_accuracy": 0.13824618086218834,
      "num_tokens": 5199552.0,
      "step": 2270
    },
    {
      "entropy": 6.039723873138428,
      "epoch": 0.2248912613681297,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998203482656507,
      "loss": 6.027,
      "mean_token_accuracy": 0.14244999885559081,
      "num_tokens": 5210316.0,
      "step": 2275
    },
    {
      "entropy": 6.0842948913574215,
      "epoch": 0.22538552787663108,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998189355488314,
      "loss": 6.1053,
      "mean_token_accuracy": 0.1472305215895176,
      "num_tokens": 5221595.0,
      "step": 2280
    },
    {
      "entropy": 6.140758895874024,
      "epoch": 0.22587979438513248,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998175173014313,
      "loss": 6.0722,
      "mean_token_accuracy": 0.13632547855377197,
      "num_tokens": 5232882.0,
      "step": 2285
    },
    {
      "entropy": 6.0675756454467775,
      "epoch": 0.22637406089363385,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998160935234857,
      "loss": 6.1709,
      "mean_token_accuracy": 0.13839787468314171,
      "num_tokens": 5244841.0,
      "step": 2290
    },
    {
      "entropy": 6.096714353561401,
      "epoch": 0.22686832740213522,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499814664215029,
      "loss": 5.9748,
      "mean_token_accuracy": 0.1458760157227516,
      "num_tokens": 5256162.0,
      "step": 2295
    },
    {
      "entropy": 6.00123872756958,
      "epoch": 0.22736259391063662,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004998132293760968,
      "loss": 5.9029,
      "mean_token_accuracy": 0.15427679419517518,
      "num_tokens": 5267130.0,
      "step": 2300
    },
    {
      "entropy": 5.969837474822998,
      "epoch": 0.227856860419138,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998117890067243,
      "loss": 6.0244,
      "mean_token_accuracy": 0.14411933422088624,
      "num_tokens": 5279394.0,
      "step": 2305
    },
    {
      "entropy": 5.981906747817993,
      "epoch": 0.22835112692763937,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998103431069469,
      "loss": 5.9975,
      "mean_token_accuracy": 0.1491586372256279,
      "num_tokens": 5291068.0,
      "step": 2310
    },
    {
      "entropy": 6.060888195037842,
      "epoch": 0.22884539343614077,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004998088916768002,
      "loss": 5.9579,
      "mean_token_accuracy": 0.14889078661799432,
      "num_tokens": 5300924.0,
      "step": 2315
    },
    {
      "entropy": 6.016797971725464,
      "epoch": 0.22933965994464214,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998074347163199,
      "loss": 5.9542,
      "mean_token_accuracy": 0.1480551391839981,
      "num_tokens": 5311998.0,
      "step": 2320
    },
    {
      "entropy": 6.06162691116333,
      "epoch": 0.22983392645314354,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998059722255418,
      "loss": 5.9899,
      "mean_token_accuracy": 0.14691204354166984,
      "num_tokens": 5322163.0,
      "step": 2325
    },
    {
      "entropy": 6.0275694847106935,
      "epoch": 0.23032819296164492,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004998045042045019,
      "loss": 5.996,
      "mean_token_accuracy": 0.1497631698846817,
      "num_tokens": 5331638.0,
      "step": 2330
    },
    {
      "entropy": 6.048976612091065,
      "epoch": 0.2308224594701463,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998030306532363,
      "loss": 5.9939,
      "mean_token_accuracy": 0.14958565384149553,
      "num_tokens": 5342399.0,
      "step": 2335
    },
    {
      "entropy": 6.055438804626465,
      "epoch": 0.2313167259786477,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998015515717813,
      "loss": 6.0695,
      "mean_token_accuracy": 0.14016990289092063,
      "num_tokens": 5353481.0,
      "step": 2340
    },
    {
      "entropy": 6.052402687072754,
      "epoch": 0.23181099248714906,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998000669601733,
      "loss": 5.9689,
      "mean_token_accuracy": 0.1473512254655361,
      "num_tokens": 5365167.0,
      "step": 2345
    },
    {
      "entropy": 6.002137184143066,
      "epoch": 0.23230525899565047,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997985768184488,
      "loss": 5.9967,
      "mean_token_accuracy": 0.1431095339357853,
      "num_tokens": 5375809.0,
      "step": 2350
    },
    {
      "entropy": 5.962526512145996,
      "epoch": 0.23279952550415184,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997970811466443,
      "loss": 5.9133,
      "mean_token_accuracy": 0.1538688361644745,
      "num_tokens": 5387372.0,
      "step": 2355
    },
    {
      "entropy": 6.187628889083863,
      "epoch": 0.2332937920126532,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004997955799447969,
      "loss": 6.0898,
      "mean_token_accuracy": 0.13684846088290215,
      "num_tokens": 5399210.0,
      "step": 2360
    },
    {
      "entropy": 5.985762119293213,
      "epoch": 0.2337880585211546,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997940732129432,
      "loss": 6.017,
      "mean_token_accuracy": 0.14093376472592353,
      "num_tokens": 5410969.0,
      "step": 2365
    },
    {
      "entropy": 5.984814739227295,
      "epoch": 0.23428232502965599,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004997925609511205,
      "loss": 5.9824,
      "mean_token_accuracy": 0.14780228435993195,
      "num_tokens": 5422569.0,
      "step": 2370
    },
    {
      "entropy": 6.031500720977784,
      "epoch": 0.2347765915381574,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004997910431593659,
      "loss": 5.9606,
      "mean_token_accuracy": 0.15011014714837073,
      "num_tokens": 5434592.0,
      "step": 2375
    },
    {
      "entropy": 6.0243175506591795,
      "epoch": 0.23527085804665876,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004997895198377168,
      "loss": 6.0545,
      "mean_token_accuracy": 0.14284207075834274,
      "num_tokens": 5446525.0,
      "step": 2380
    },
    {
      "entropy": 6.08224196434021,
      "epoch": 0.23576512455516013,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997879909862105,
      "loss": 6.025,
      "mean_token_accuracy": 0.14782738387584687,
      "num_tokens": 5457327.0,
      "step": 2385
    },
    {
      "entropy": 6.105248498916626,
      "epoch": 0.23625939106366153,
      "grad_norm": 1.078125,
      "learning_rate": 0.000499786456604885,
      "loss": 6.0642,
      "mean_token_accuracy": 0.1362588956952095,
      "num_tokens": 5468533.0,
      "step": 2390
    },
    {
      "entropy": 5.998384857177735,
      "epoch": 0.2367536575721629,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997849166937776,
      "loss": 5.9451,
      "mean_token_accuracy": 0.14417133554816247,
      "num_tokens": 5480652.0,
      "step": 2395
    },
    {
      "entropy": 6.012068605422973,
      "epoch": 0.2372479240806643,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997833712529265,
      "loss": 5.9707,
      "mean_token_accuracy": 0.14451036602258682,
      "num_tokens": 5491446.0,
      "step": 2400
    },
    {
      "entropy": 6.021755743026733,
      "epoch": 0.23774219058916568,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997818202823695,
      "loss": 5.9945,
      "mean_token_accuracy": 0.14959129244089125,
      "num_tokens": 5502886.0,
      "step": 2405
    },
    {
      "entropy": 6.0749561309814455,
      "epoch": 0.23823645709766705,
      "grad_norm": 1.125,
      "learning_rate": 0.0004997802637821449,
      "loss": 6.023,
      "mean_token_accuracy": 0.14053252562880517,
      "num_tokens": 5514305.0,
      "step": 2410
    },
    {
      "entropy": 5.98513970375061,
      "epoch": 0.23873072360616845,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997787017522909,
      "loss": 5.9437,
      "mean_token_accuracy": 0.15076548904180526,
      "num_tokens": 5525913.0,
      "step": 2415
    },
    {
      "entropy": 5.90565128326416,
      "epoch": 0.23922499011466983,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997771341928459,
      "loss": 5.9446,
      "mean_token_accuracy": 0.1544351950287819,
      "num_tokens": 5537069.0,
      "step": 2420
    },
    {
      "entropy": 6.022730112075806,
      "epoch": 0.2397192566231712,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004997755611038487,
      "loss": 6.0441,
      "mean_token_accuracy": 0.1400386117398739,
      "num_tokens": 5547988.0,
      "step": 2425
    },
    {
      "entropy": 6.105308675765992,
      "epoch": 0.2402135231316726,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997739824853378,
      "loss": 6.1088,
      "mean_token_accuracy": 0.1395256355404854,
      "num_tokens": 5560400.0,
      "step": 2430
    },
    {
      "entropy": 5.987665414810181,
      "epoch": 0.24070778964017397,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997723983373519,
      "loss": 5.9696,
      "mean_token_accuracy": 0.15029642432928086,
      "num_tokens": 5572291.0,
      "step": 2435
    },
    {
      "entropy": 6.100761747360229,
      "epoch": 0.24120205614867538,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997708086599303,
      "loss": 5.9938,
      "mean_token_accuracy": 0.14239495247602463,
      "num_tokens": 5583583.0,
      "step": 2440
    },
    {
      "entropy": 6.093548583984375,
      "epoch": 0.24169632265717675,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997692134531119,
      "loss": 6.0205,
      "mean_token_accuracy": 0.14349138885736465,
      "num_tokens": 5595499.0,
      "step": 2445
    },
    {
      "entropy": 5.937535381317138,
      "epoch": 0.24219058916567812,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499767612716936,
      "loss": 5.9486,
      "mean_token_accuracy": 0.1555730566382408,
      "num_tokens": 5606745.0,
      "step": 2450
    },
    {
      "entropy": 5.977556610107422,
      "epoch": 0.24268485567417952,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004997660064514419,
      "loss": 6.0333,
      "mean_token_accuracy": 0.1459631770849228,
      "num_tokens": 5618226.0,
      "step": 2455
    },
    {
      "entropy": 6.103812122344971,
      "epoch": 0.2431791221826809,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997643946566691,
      "loss": 6.0358,
      "mean_token_accuracy": 0.14382674545049667,
      "num_tokens": 5628941.0,
      "step": 2460
    },
    {
      "entropy": 5.948967981338501,
      "epoch": 0.2436733886911823,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997627773326574,
      "loss": 5.8499,
      "mean_token_accuracy": 0.15501064658164979,
      "num_tokens": 5641077.0,
      "step": 2465
    },
    {
      "entropy": 6.018094301223755,
      "epoch": 0.24416765519968367,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004997611544794465,
      "loss": 6.1555,
      "mean_token_accuracy": 0.1338454492390156,
      "num_tokens": 5652156.0,
      "step": 2470
    },
    {
      "entropy": 6.151158618927002,
      "epoch": 0.24466192170818504,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997595260970764,
      "loss": 6.0822,
      "mean_token_accuracy": 0.14347486943006516,
      "num_tokens": 5663988.0,
      "step": 2475
    },
    {
      "entropy": 5.977680206298828,
      "epoch": 0.24515618821668644,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499757892185587,
      "loss": 6.0018,
      "mean_token_accuracy": 0.14267092123627662,
      "num_tokens": 5676302.0,
      "step": 2480
    },
    {
      "entropy": 6.070641231536865,
      "epoch": 0.24565045472518782,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997562527450187,
      "loss": 6.046,
      "mean_token_accuracy": 0.14550234749913216,
      "num_tokens": 5686686.0,
      "step": 2485
    },
    {
      "entropy": 6.021747350692749,
      "epoch": 0.24614472123368922,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004997546077754116,
      "loss": 5.9525,
      "mean_token_accuracy": 0.15929894894361496,
      "num_tokens": 5698379.0,
      "step": 2490
    },
    {
      "entropy": 5.958306932449341,
      "epoch": 0.2466389877421906,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997529572768063,
      "loss": 5.9432,
      "mean_token_accuracy": 0.1498410865664482,
      "num_tokens": 5708891.0,
      "step": 2495
    },
    {
      "entropy": 5.981058549880982,
      "epoch": 0.24713325425069196,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004997513012492434,
      "loss": 5.9923,
      "mean_token_accuracy": 0.14332532063126563,
      "num_tokens": 5720326.0,
      "step": 2500
    },
    {
      "entropy": 6.052645683288574,
      "epoch": 0.24762752075919336,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997496396927636,
      "loss": 5.9112,
      "mean_token_accuracy": 0.15235294178128242,
      "num_tokens": 5731121.0,
      "step": 2505
    },
    {
      "entropy": 6.023140573501587,
      "epoch": 0.24812178726769474,
      "grad_norm": 1.125,
      "learning_rate": 0.0004997479726074077,
      "loss": 5.9597,
      "mean_token_accuracy": 0.15026197880506514,
      "num_tokens": 5741637.0,
      "step": 2510
    },
    {
      "entropy": 6.048560428619385,
      "epoch": 0.24861605377619614,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499746299993217,
      "loss": 6.0102,
      "mean_token_accuracy": 0.14581428989768028,
      "num_tokens": 5752659.0,
      "step": 2515
    },
    {
      "entropy": 5.998668670654297,
      "epoch": 0.2491103202846975,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997446218502324,
      "loss": 5.9728,
      "mean_token_accuracy": 0.14580842703580857,
      "num_tokens": 5763829.0,
      "step": 2520
    },
    {
      "entropy": 6.0419858455657955,
      "epoch": 0.24960458679319888,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997429381784952,
      "loss": 6.053,
      "mean_token_accuracy": 0.1448736622929573,
      "num_tokens": 5775848.0,
      "step": 2525
    },
    {
      "entropy": 5.984309673309326,
      "epoch": 0.2500988533017003,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004997412489780469,
      "loss": 6.0037,
      "mean_token_accuracy": 0.1499286875128746,
      "num_tokens": 5786595.0,
      "step": 2530
    },
    {
      "entropy": 5.964243793487549,
      "epoch": 0.25059311981020166,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499739554248929,
      "loss": 5.8929,
      "mean_token_accuracy": 0.15398191586136817,
      "num_tokens": 5797841.0,
      "step": 2535
    },
    {
      "entropy": 5.912666082382202,
      "epoch": 0.25108738631870303,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997378539911832,
      "loss": 5.9089,
      "mean_token_accuracy": 0.15088940635323525,
      "num_tokens": 5809106.0,
      "step": 2540
    },
    {
      "entropy": 5.973687314987183,
      "epoch": 0.2515816528272044,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997361482048513,
      "loss": 5.9215,
      "mean_token_accuracy": 0.1497866354882717,
      "num_tokens": 5820316.0,
      "step": 2545
    },
    {
      "entropy": 6.026650857925415,
      "epoch": 0.25207591933570583,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997344368899753,
      "loss": 6.0408,
      "mean_token_accuracy": 0.139362945407629,
      "num_tokens": 5833137.0,
      "step": 2550
    },
    {
      "entropy": 6.037950611114502,
      "epoch": 0.2525701858442072,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004997327200465972,
      "loss": 5.9694,
      "mean_token_accuracy": 0.14529137313365936,
      "num_tokens": 5844088.0,
      "step": 2555
    },
    {
      "entropy": 6.077861881256103,
      "epoch": 0.2530644523527086,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997309976747594,
      "loss": 5.9736,
      "mean_token_accuracy": 0.14326793625950812,
      "num_tokens": 5854546.0,
      "step": 2560
    },
    {
      "entropy": 6.011197853088379,
      "epoch": 0.25355871886120995,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004997292697745042,
      "loss": 5.9723,
      "mean_token_accuracy": 0.1466481104493141,
      "num_tokens": 5865326.0,
      "step": 2565
    },
    {
      "entropy": 5.999698734283447,
      "epoch": 0.2540529853697113,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997275363458741,
      "loss": 5.8594,
      "mean_token_accuracy": 0.1503140576183796,
      "num_tokens": 5876487.0,
      "step": 2570
    },
    {
      "entropy": 5.875597381591797,
      "epoch": 0.25454725187821275,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997257973889118,
      "loss": 5.9726,
      "mean_token_accuracy": 0.14843393564224244,
      "num_tokens": 5887846.0,
      "step": 2575
    },
    {
      "entropy": 6.101366996765137,
      "epoch": 0.25504151838671413,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004997240529036599,
      "loss": 5.9412,
      "mean_token_accuracy": 0.15172709226608277,
      "num_tokens": 5898572.0,
      "step": 2580
    },
    {
      "entropy": 5.881202983856201,
      "epoch": 0.2555357848952155,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004997223028901615,
      "loss": 6.0037,
      "mean_token_accuracy": 0.1475423127412796,
      "num_tokens": 5908927.0,
      "step": 2585
    },
    {
      "entropy": 6.037431478500366,
      "epoch": 0.2560300514037169,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004997205473484596,
      "loss": 5.9495,
      "mean_token_accuracy": 0.1498175874352455,
      "num_tokens": 5919807.0,
      "step": 2590
    },
    {
      "entropy": 5.920142316818238,
      "epoch": 0.25652431791221825,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997187862785974,
      "loss": 5.9779,
      "mean_token_accuracy": 0.14424620494246482,
      "num_tokens": 5931160.0,
      "step": 2595
    },
    {
      "entropy": 6.088014888763428,
      "epoch": 0.2570185844207197,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997170196806181,
      "loss": 6.0412,
      "mean_token_accuracy": 0.1426064506173134,
      "num_tokens": 5943694.0,
      "step": 2600
    },
    {
      "entropy": 5.981357526779175,
      "epoch": 0.25751285092922105,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997152475545652,
      "loss": 5.9319,
      "mean_token_accuracy": 0.14945385083556176,
      "num_tokens": 5954543.0,
      "step": 2605
    },
    {
      "entropy": 5.9636317729949955,
      "epoch": 0.2580071174377224,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997134699004825,
      "loss": 5.973,
      "mean_token_accuracy": 0.1513287492096424,
      "num_tokens": 5965080.0,
      "step": 2610
    },
    {
      "entropy": 6.036751079559326,
      "epoch": 0.2585013839462238,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004997116867184136,
      "loss": 5.9566,
      "mean_token_accuracy": 0.14630378186702728,
      "num_tokens": 5975314.0,
      "step": 2615
    },
    {
      "entropy": 6.02877459526062,
      "epoch": 0.25899565045472517,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004997098980084022,
      "loss": 5.9735,
      "mean_token_accuracy": 0.14686284810304642,
      "num_tokens": 5985927.0,
      "step": 2620
    },
    {
      "entropy": 5.9730428695678714,
      "epoch": 0.2594899169632266,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004997081037704926,
      "loss": 5.9986,
      "mean_token_accuracy": 0.145076747238636,
      "num_tokens": 5997311.0,
      "step": 2625
    },
    {
      "entropy": 5.986532354354859,
      "epoch": 0.25998418347172797,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997063040047288,
      "loss": 6.0236,
      "mean_token_accuracy": 0.13863110765814782,
      "num_tokens": 6008595.0,
      "step": 2630
    },
    {
      "entropy": 6.084332275390625,
      "epoch": 0.26047844998022934,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004997044987111548,
      "loss": 5.9346,
      "mean_token_accuracy": 0.1485560044646263,
      "num_tokens": 6019055.0,
      "step": 2635
    },
    {
      "entropy": 5.984119462966919,
      "epoch": 0.2609727164887307,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997026878898155,
      "loss": 6.0785,
      "mean_token_accuracy": 0.13726168423891066,
      "num_tokens": 6030029.0,
      "step": 2640
    },
    {
      "entropy": 6.064934492111206,
      "epoch": 0.2614669829972321,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997008715407552,
      "loss": 6.0149,
      "mean_token_accuracy": 0.14624494388699533,
      "num_tokens": 6041495.0,
      "step": 2645
    },
    {
      "entropy": 6.0098552227020265,
      "epoch": 0.2619612495057335,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996990496640185,
      "loss": 5.9223,
      "mean_token_accuracy": 0.14212577491998674,
      "num_tokens": 6052307.0,
      "step": 2650
    },
    {
      "entropy": 5.9654816627502445,
      "epoch": 0.2624555160142349,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996972222596505,
      "loss": 5.8811,
      "mean_token_accuracy": 0.14910161644220352,
      "num_tokens": 6062550.0,
      "step": 2655
    },
    {
      "entropy": 5.945908880233764,
      "epoch": 0.26294978252273626,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004996953893276958,
      "loss": 5.9301,
      "mean_token_accuracy": 0.14317954927682877,
      "num_tokens": 6074450.0,
      "step": 2660
    },
    {
      "entropy": 5.945242881774902,
      "epoch": 0.26344404903123764,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996935508681997,
      "loss": 5.9609,
      "mean_token_accuracy": 0.1495286300778389,
      "num_tokens": 6085749.0,
      "step": 2665
    },
    {
      "entropy": 6.005150699615479,
      "epoch": 0.263938315539739,
      "grad_norm": 1.125,
      "learning_rate": 0.0004996917068812075,
      "loss": 5.8905,
      "mean_token_accuracy": 0.1543781578540802,
      "num_tokens": 6096438.0,
      "step": 2670
    },
    {
      "entropy": 5.924571704864502,
      "epoch": 0.26443258204824044,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996898573667643,
      "loss": 5.9266,
      "mean_token_accuracy": 0.1454748436808586,
      "num_tokens": 6108386.0,
      "step": 2675
    },
    {
      "entropy": 6.008694839477539,
      "epoch": 0.2649268485567418,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004996880023249158,
      "loss": 5.8626,
      "mean_token_accuracy": 0.15368301123380662,
      "num_tokens": 6118771.0,
      "step": 2680
    },
    {
      "entropy": 5.974543571472168,
      "epoch": 0.2654211150652432,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996861417557076,
      "loss": 5.9496,
      "mean_token_accuracy": 0.14521803110837936,
      "num_tokens": 6130718.0,
      "step": 2685
    },
    {
      "entropy": 5.922163200378418,
      "epoch": 0.26591538157374456,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004996842756591855,
      "loss": 5.9631,
      "mean_token_accuracy": 0.14341218546032905,
      "num_tokens": 6142086.0,
      "step": 2690
    },
    {
      "entropy": 5.976787948608399,
      "epoch": 0.26640964808224593,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004996824040353953,
      "loss": 5.9613,
      "mean_token_accuracy": 0.1457512103021145,
      "num_tokens": 6153857.0,
      "step": 2695
    },
    {
      "entropy": 6.125744962692261,
      "epoch": 0.2669039145907473,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996805268843832,
      "loss": 6.0951,
      "mean_token_accuracy": 0.1432183265686035,
      "num_tokens": 6166796.0,
      "step": 2700
    },
    {
      "entropy": 5.963466644287109,
      "epoch": 0.26739818109924873,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996786442061952,
      "loss": 5.9169,
      "mean_token_accuracy": 0.15192901045084,
      "num_tokens": 6178046.0,
      "step": 2705
    },
    {
      "entropy": 5.9657057762146,
      "epoch": 0.2678924476077501,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996767560008777,
      "loss": 5.9593,
      "mean_token_accuracy": 0.14713699072599412,
      "num_tokens": 6190455.0,
      "step": 2710
    },
    {
      "entropy": 5.899187612533569,
      "epoch": 0.2683867141162515,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996748622684771,
      "loss": 5.8462,
      "mean_token_accuracy": 0.15217214077711105,
      "num_tokens": 6201659.0,
      "step": 2715
    },
    {
      "entropy": 5.924319171905518,
      "epoch": 0.26888098062475285,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996729630090401,
      "loss": 5.9552,
      "mean_token_accuracy": 0.1431477539241314,
      "num_tokens": 6213870.0,
      "step": 2720
    },
    {
      "entropy": 6.0065733909606935,
      "epoch": 0.2693752471332542,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996710582226133,
      "loss": 5.9012,
      "mean_token_accuracy": 0.1548033893108368,
      "num_tokens": 6224237.0,
      "step": 2725
    },
    {
      "entropy": 5.960059928894043,
      "epoch": 0.26986951364175565,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004996691479092437,
      "loss": 5.9433,
      "mean_token_accuracy": 0.14833316206932068,
      "num_tokens": 6235868.0,
      "step": 2730
    },
    {
      "entropy": 6.0247667789459225,
      "epoch": 0.270363780150257,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004996672320689782,
      "loss": 5.9961,
      "mean_token_accuracy": 0.14516259282827376,
      "num_tokens": 6247994.0,
      "step": 2735
    },
    {
      "entropy": 5.855693197250366,
      "epoch": 0.2708580466587584,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996653107018638,
      "loss": 5.8469,
      "mean_token_accuracy": 0.15463558286428453,
      "num_tokens": 6259654.0,
      "step": 2740
    },
    {
      "entropy": 5.977344179153443,
      "epoch": 0.2713523131672598,
      "grad_norm": 1.0,
      "learning_rate": 0.000499663383807948,
      "loss": 5.9049,
      "mean_token_accuracy": 0.1500097319483757,
      "num_tokens": 6270582.0,
      "step": 2745
    },
    {
      "entropy": 5.968403148651123,
      "epoch": 0.27184657967576115,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004996614513872781,
      "loss": 5.9016,
      "mean_token_accuracy": 0.15006383657455444,
      "num_tokens": 6281088.0,
      "step": 2750
    },
    {
      "entropy": 5.981626415252686,
      "epoch": 0.2723408461842626,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004996595134399017,
      "loss": 6.0555,
      "mean_token_accuracy": 0.140472724288702,
      "num_tokens": 6292920.0,
      "step": 2755
    },
    {
      "entropy": 6.020872259140015,
      "epoch": 0.27283511269276395,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996575699658664,
      "loss": 5.9111,
      "mean_token_accuracy": 0.15108120441436768,
      "num_tokens": 6304980.0,
      "step": 2760
    },
    {
      "entropy": 5.918898677825927,
      "epoch": 0.2733293792012653,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996556209652199,
      "loss": 5.8999,
      "mean_token_accuracy": 0.15111031085252763,
      "num_tokens": 6316175.0,
      "step": 2765
    },
    {
      "entropy": 5.976380968093872,
      "epoch": 0.2738236457097667,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996536664380104,
      "loss": 5.9517,
      "mean_token_accuracy": 0.15452203899621964,
      "num_tokens": 6327590.0,
      "step": 2770
    },
    {
      "entropy": 5.924459648132324,
      "epoch": 0.27431791221826807,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996517063842859,
      "loss": 5.9716,
      "mean_token_accuracy": 0.1498926803469658,
      "num_tokens": 6339066.0,
      "step": 2775
    },
    {
      "entropy": 6.0026692867279055,
      "epoch": 0.2748121787267695,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996497408040944,
      "loss": 5.9024,
      "mean_token_accuracy": 0.15032327100634574,
      "num_tokens": 6349485.0,
      "step": 2780
    },
    {
      "entropy": 5.844460916519165,
      "epoch": 0.27530644523527087,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004996477696974845,
      "loss": 5.8657,
      "mean_token_accuracy": 0.15661628544330597,
      "num_tokens": 6360775.0,
      "step": 2785
    },
    {
      "entropy": 5.9499735832214355,
      "epoch": 0.27580071174377224,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004996457930645046,
      "loss": 5.883,
      "mean_token_accuracy": 0.148551806807518,
      "num_tokens": 6371645.0,
      "step": 2790
    },
    {
      "entropy": 5.961523580551147,
      "epoch": 0.2762949782522736,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996438109052034,
      "loss": 5.9074,
      "mean_token_accuracy": 0.15085601136088372,
      "num_tokens": 6383877.0,
      "step": 2795
    },
    {
      "entropy": 5.9515691757202145,
      "epoch": 0.276789244760775,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996418232196296,
      "loss": 5.9828,
      "mean_token_accuracy": 0.14350538998842238,
      "num_tokens": 6394488.0,
      "step": 2800
    },
    {
      "entropy": 5.979497003555298,
      "epoch": 0.2772835112692764,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996398300078321,
      "loss": 5.9065,
      "mean_token_accuracy": 0.15064874663949013,
      "num_tokens": 6404948.0,
      "step": 2805
    },
    {
      "entropy": 5.974899482727051,
      "epoch": 0.2777777777777778,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996378312698598,
      "loss": 5.9471,
      "mean_token_accuracy": 0.1533238559961319,
      "num_tokens": 6416273.0,
      "step": 2810
    },
    {
      "entropy": 6.013521671295166,
      "epoch": 0.27827204428627916,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004996358270057622,
      "loss": 5.9557,
      "mean_token_accuracy": 0.1466079130768776,
      "num_tokens": 6427091.0,
      "step": 2815
    },
    {
      "entropy": 5.9516332149505615,
      "epoch": 0.27876631079478054,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004996338172155883,
      "loss": 5.9126,
      "mean_token_accuracy": 0.14914426058530808,
      "num_tokens": 6438254.0,
      "step": 2820
    },
    {
      "entropy": 5.89876914024353,
      "epoch": 0.2792605773032819,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996318018993876,
      "loss": 5.8239,
      "mean_token_accuracy": 0.1530715063214302,
      "num_tokens": 6450618.0,
      "step": 2825
    },
    {
      "entropy": 5.980556011199951,
      "epoch": 0.27975484381178334,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004996297810572098,
      "loss": 5.8956,
      "mean_token_accuracy": 0.1491038627922535,
      "num_tokens": 6462467.0,
      "step": 2830
    },
    {
      "entropy": 5.9045891761779785,
      "epoch": 0.2802491103202847,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996277546891045,
      "loss": 5.864,
      "mean_token_accuracy": 0.15746333003044127,
      "num_tokens": 6474206.0,
      "step": 2835
    },
    {
      "entropy": 5.942203712463379,
      "epoch": 0.2807433768287861,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996257227951216,
      "loss": 5.9691,
      "mean_token_accuracy": 0.1443627193570137,
      "num_tokens": 6486405.0,
      "step": 2840
    },
    {
      "entropy": 5.967880725860596,
      "epoch": 0.28123764333728746,
      "grad_norm": 1.09375,
      "learning_rate": 0.000499623685375311,
      "loss": 5.9336,
      "mean_token_accuracy": 0.14780732095241547,
      "num_tokens": 6496949.0,
      "step": 2845
    },
    {
      "entropy": 5.923089170455933,
      "epoch": 0.28173190984578883,
      "grad_norm": 1.0,
      "learning_rate": 0.000499621642429723,
      "loss": 5.874,
      "mean_token_accuracy": 0.1550042062997818,
      "num_tokens": 6508913.0,
      "step": 2850
    },
    {
      "entropy": 5.956121826171875,
      "epoch": 0.28222617635429026,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996195939584078,
      "loss": 5.9019,
      "mean_token_accuracy": 0.14797470420598985,
      "num_tokens": 6519837.0,
      "step": 2855
    },
    {
      "entropy": 5.941003894805908,
      "epoch": 0.28272044286279163,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996175399614156,
      "loss": 5.872,
      "mean_token_accuracy": 0.14837601631879807,
      "num_tokens": 6531409.0,
      "step": 2860
    },
    {
      "entropy": 5.937231063842773,
      "epoch": 0.283214709371293,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996154804387972,
      "loss": 5.9121,
      "mean_token_accuracy": 0.15122439116239547,
      "num_tokens": 6543433.0,
      "step": 2865
    },
    {
      "entropy": 5.91095666885376,
      "epoch": 0.2837089758797944,
      "grad_norm": 1.140625,
      "learning_rate": 0.000499613415390603,
      "loss": 5.9883,
      "mean_token_accuracy": 0.14575250148773194,
      "num_tokens": 6554511.0,
      "step": 2870
    },
    {
      "entropy": 6.029332876205444,
      "epoch": 0.28420324238829575,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000499611344816884,
      "loss": 5.9113,
      "mean_token_accuracy": 0.1500216841697693,
      "num_tokens": 6564591.0,
      "step": 2875
    },
    {
      "entropy": 5.943623399734497,
      "epoch": 0.2846975088967972,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499609268717691,
      "loss": 5.8826,
      "mean_token_accuracy": 0.15444658771157266,
      "num_tokens": 6575684.0,
      "step": 2880
    },
    {
      "entropy": 6.002884864807129,
      "epoch": 0.28519177540529855,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996071870930753,
      "loss": 6.0142,
      "mean_token_accuracy": 0.1460142984986305,
      "num_tokens": 6588504.0,
      "step": 2885
    },
    {
      "entropy": 5.843828010559082,
      "epoch": 0.2856860419137999,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996050999430878,
      "loss": 5.8341,
      "mean_token_accuracy": 0.1514135092496872,
      "num_tokens": 6599969.0,
      "step": 2890
    },
    {
      "entropy": 5.977163600921631,
      "epoch": 0.2861803084223013,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004996030072677802,
      "loss": 5.9088,
      "mean_token_accuracy": 0.15263523682951927,
      "num_tokens": 6611430.0,
      "step": 2895
    },
    {
      "entropy": 5.925257349014283,
      "epoch": 0.2866745749308027,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004996009090672037,
      "loss": 5.8906,
      "mean_token_accuracy": 0.14462630748748778,
      "num_tokens": 6623247.0,
      "step": 2900
    },
    {
      "entropy": 5.96600112915039,
      "epoch": 0.28716884143930405,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004995988053414098,
      "loss": 6.0225,
      "mean_token_accuracy": 0.14478808343410493,
      "num_tokens": 6635508.0,
      "step": 2905
    },
    {
      "entropy": 5.954598236083984,
      "epoch": 0.2876631079478055,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995966960904506,
      "loss": 5.9278,
      "mean_token_accuracy": 0.14939955547451972,
      "num_tokens": 6646020.0,
      "step": 2910
    },
    {
      "entropy": 5.978266286849975,
      "epoch": 0.28815737445630685,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995945813143778,
      "loss": 5.9568,
      "mean_token_accuracy": 0.14021606743335724,
      "num_tokens": 6658194.0,
      "step": 2915
    },
    {
      "entropy": 5.96915636062622,
      "epoch": 0.2886516409648082,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995924610132435,
      "loss": 5.9182,
      "mean_token_accuracy": 0.1475129932165146,
      "num_tokens": 6670582.0,
      "step": 2920
    },
    {
      "entropy": 5.928704261779785,
      "epoch": 0.2891459074733096,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004995903351870998,
      "loss": 5.8759,
      "mean_token_accuracy": 0.14914150834083556,
      "num_tokens": 6680836.0,
      "step": 2925
    },
    {
      "entropy": 6.073338317871094,
      "epoch": 0.28964017398181097,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000499588203835999,
      "loss": 5.9717,
      "mean_token_accuracy": 0.14328644201159477,
      "num_tokens": 6693045.0,
      "step": 2930
    },
    {
      "entropy": 5.933247041702271,
      "epoch": 0.2901344404903124,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004995860669599935,
      "loss": 5.9092,
      "mean_token_accuracy": 0.14393445253372192,
      "num_tokens": 6704006.0,
      "step": 2935
    },
    {
      "entropy": 5.837153053283691,
      "epoch": 0.29062870699881377,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499583924559136,
      "loss": 5.8793,
      "mean_token_accuracy": 0.14840964674949647,
      "num_tokens": 6717027.0,
      "step": 2940
    },
    {
      "entropy": 5.981049108505249,
      "epoch": 0.29112297350731514,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499581776633479,
      "loss": 5.8749,
      "mean_token_accuracy": 0.15013007670640946,
      "num_tokens": 6728107.0,
      "step": 2945
    },
    {
      "entropy": 5.88144817352295,
      "epoch": 0.2916172400158165,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004995796231830756,
      "loss": 5.8902,
      "mean_token_accuracy": 0.15416626930236815,
      "num_tokens": 6738628.0,
      "step": 2950
    },
    {
      "entropy": 5.934478712081909,
      "epoch": 0.2921115065243179,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004995774642079786,
      "loss": 5.8421,
      "mean_token_accuracy": 0.1537696972489357,
      "num_tokens": 6749416.0,
      "step": 2955
    },
    {
      "entropy": 5.905332374572754,
      "epoch": 0.2926057730328193,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499575299708241,
      "loss": 5.947,
      "mean_token_accuracy": 0.14088274016976357,
      "num_tokens": 6761663.0,
      "step": 2960
    },
    {
      "entropy": 6.020952272415161,
      "epoch": 0.2931000395413207,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004995731296839163,
      "loss": 5.8909,
      "mean_token_accuracy": 0.1538805305957794,
      "num_tokens": 6773557.0,
      "step": 2965
    },
    {
      "entropy": 5.81088809967041,
      "epoch": 0.29359430604982206,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004995709541350578,
      "loss": 5.8199,
      "mean_token_accuracy": 0.1593501538038254,
      "num_tokens": 6782832.0,
      "step": 2970
    },
    {
      "entropy": 5.9672423839569095,
      "epoch": 0.29408857255832344,
      "grad_norm": 1.109375,
      "learning_rate": 0.000499568773061719,
      "loss": 5.9335,
      "mean_token_accuracy": 0.14496245384216308,
      "num_tokens": 6795165.0,
      "step": 2975
    },
    {
      "entropy": 6.023906803131103,
      "epoch": 0.2945828390668248,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004995665864639534,
      "loss": 5.9486,
      "mean_token_accuracy": 0.1483907014131546,
      "num_tokens": 6807054.0,
      "step": 2980
    },
    {
      "entropy": 5.876065492630005,
      "epoch": 0.29507710557532624,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004995643943418149,
      "loss": 5.7332,
      "mean_token_accuracy": 0.15653893128037452,
      "num_tokens": 6817681.0,
      "step": 2985
    },
    {
      "entropy": 5.863463449478149,
      "epoch": 0.2955713720838276,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995621966953576,
      "loss": 5.9026,
      "mean_token_accuracy": 0.1518360435962677,
      "num_tokens": 6829387.0,
      "step": 2990
    },
    {
      "entropy": 5.936010360717773,
      "epoch": 0.296065638592329,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995599935246355,
      "loss": 5.9071,
      "mean_token_accuracy": 0.1520347461104393,
      "num_tokens": 6842023.0,
      "step": 2995
    },
    {
      "entropy": 5.92348346710205,
      "epoch": 0.29655990510083036,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995577848297026,
      "loss": 5.8992,
      "mean_token_accuracy": 0.1477331593632698,
      "num_tokens": 6853189.0,
      "step": 3000
    },
    {
      "epoch": 0.29655990510083036,
      "eval_entropy": 5.794283665472878,
      "eval_loss": 5.955329895019531,
      "eval_mean_token_accuracy": 0.15471349589865732,
      "eval_num_tokens": 6853189.0,
      "eval_runtime": 26.573,
      "eval_samples_per_second": 1223.537,
      "eval_steps_per_second": 152.975,
      "step": 3000
    },
    {
      "entropy": 5.963679313659668,
      "epoch": 0.29705417160933173,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004995555706106134,
      "loss": 5.9504,
      "mean_token_accuracy": 0.1432880811393261,
      "num_tokens": 6864338.0,
      "step": 3005
    },
    {
      "entropy": 5.9105650901794435,
      "epoch": 0.29754843811783316,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004995533508674223,
      "loss": 5.8262,
      "mean_token_accuracy": 0.15388680547475814,
      "num_tokens": 6875409.0,
      "step": 3010
    },
    {
      "entropy": 5.927374696731567,
      "epoch": 0.29804270462633453,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004995511256001839,
      "loss": 5.8561,
      "mean_token_accuracy": 0.14918482303619385,
      "num_tokens": 6885757.0,
      "step": 3015
    },
    {
      "entropy": 5.894947910308838,
      "epoch": 0.2985369711348359,
      "grad_norm": 1.125,
      "learning_rate": 0.000499548894808953,
      "loss": 5.9246,
      "mean_token_accuracy": 0.149386990070343,
      "num_tokens": 6896440.0,
      "step": 3020
    },
    {
      "entropy": 5.982046318054199,
      "epoch": 0.2990312376433373,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004995466584937846,
      "loss": 5.934,
      "mean_token_accuracy": 0.14567233771085739,
      "num_tokens": 6907319.0,
      "step": 3025
    },
    {
      "entropy": 5.953330326080322,
      "epoch": 0.29952550415183865,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004995444166547333,
      "loss": 5.8928,
      "mean_token_accuracy": 0.14992106705904007,
      "num_tokens": 6918683.0,
      "step": 3030
    },
    {
      "entropy": 5.879758739471436,
      "epoch": 0.3000197706603401,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004995421692918547,
      "loss": 5.8745,
      "mean_token_accuracy": 0.15101571381092072,
      "num_tokens": 6929974.0,
      "step": 3035
    },
    {
      "entropy": 5.998798656463623,
      "epoch": 0.30051403716884145,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004995399164052038,
      "loss": 5.8816,
      "mean_token_accuracy": 0.14776666909456254,
      "num_tokens": 6941213.0,
      "step": 3040
    },
    {
      "entropy": 6.010467910766602,
      "epoch": 0.3010083036773428,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004995376579948362,
      "loss": 5.9299,
      "mean_token_accuracy": 0.149107863008976,
      "num_tokens": 6952880.0,
      "step": 3045
    },
    {
      "entropy": 5.940840101242065,
      "epoch": 0.3015025701858442,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004995353940608074,
      "loss": 5.916,
      "mean_token_accuracy": 0.14906010404229164,
      "num_tokens": 6964344.0,
      "step": 3050
    },
    {
      "entropy": 5.867275619506836,
      "epoch": 0.30199683669434557,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499533124603173,
      "loss": 5.8496,
      "mean_token_accuracy": 0.15172645151615144,
      "num_tokens": 6976042.0,
      "step": 3055
    },
    {
      "entropy": 5.927947092056274,
      "epoch": 0.302491103202847,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499530849621989,
      "loss": 5.8254,
      "mean_token_accuracy": 0.1564239665865898,
      "num_tokens": 6988137.0,
      "step": 3060
    },
    {
      "entropy": 5.871209049224854,
      "epoch": 0.3029853697113484,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004995285691173111,
      "loss": 5.8455,
      "mean_token_accuracy": 0.15776384100317956,
      "num_tokens": 6999858.0,
      "step": 3065
    },
    {
      "entropy": 5.956633234024048,
      "epoch": 0.30347963621984975,
      "grad_norm": 1.125,
      "learning_rate": 0.0004995262830891957,
      "loss": 5.9192,
      "mean_token_accuracy": 0.1524971157312393,
      "num_tokens": 7012756.0,
      "step": 3070
    },
    {
      "entropy": 5.922817134857178,
      "epoch": 0.3039739027283511,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004995239915376988,
      "loss": 5.928,
      "mean_token_accuracy": 0.1458661213517189,
      "num_tokens": 7024769.0,
      "step": 3075
    },
    {
      "entropy": 5.925840139389038,
      "epoch": 0.3044681692368525,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004995216944628769,
      "loss": 5.9021,
      "mean_token_accuracy": 0.1427825465798378,
      "num_tokens": 7035422.0,
      "step": 3080
    },
    {
      "entropy": 5.907672977447509,
      "epoch": 0.3049624357453539,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004995193918647865,
      "loss": 5.8496,
      "mean_token_accuracy": 0.1537895068526268,
      "num_tokens": 7046923.0,
      "step": 3085
    },
    {
      "entropy": 6.0074094295501705,
      "epoch": 0.3054567022538553,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004995170837434844,
      "loss": 5.8896,
      "mean_token_accuracy": 0.1455712527036667,
      "num_tokens": 7058241.0,
      "step": 3090
    },
    {
      "entropy": 5.843660068511963,
      "epoch": 0.30595096876235667,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004995147700990269,
      "loss": 5.9194,
      "mean_token_accuracy": 0.15101224333047866,
      "num_tokens": 7070028.0,
      "step": 3095
    },
    {
      "entropy": 5.969692516326904,
      "epoch": 0.30644523527085804,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004995124509314714,
      "loss": 5.898,
      "mean_token_accuracy": 0.14975875839591027,
      "num_tokens": 7082641.0,
      "step": 3100
    },
    {
      "entropy": 5.934259939193725,
      "epoch": 0.3069395017793594,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004995101262408748,
      "loss": 5.8971,
      "mean_token_accuracy": 0.14581625610589982,
      "num_tokens": 7094462.0,
      "step": 3105
    },
    {
      "entropy": 5.879360914230347,
      "epoch": 0.3074337682878608,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004995077960272943,
      "loss": 5.8656,
      "mean_token_accuracy": 0.15483699291944503,
      "num_tokens": 7106598.0,
      "step": 3110
    },
    {
      "entropy": 5.908220291137695,
      "epoch": 0.3079280347963622,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004995054602907871,
      "loss": 5.7884,
      "mean_token_accuracy": 0.15013569295406343,
      "num_tokens": 7117006.0,
      "step": 3115
    },
    {
      "entropy": 5.8303180694580075,
      "epoch": 0.3084223013048636,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004995031190314107,
      "loss": 5.8671,
      "mean_token_accuracy": 0.15000214129686357,
      "num_tokens": 7128071.0,
      "step": 3120
    },
    {
      "entropy": 5.932742738723755,
      "epoch": 0.30891656781336496,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004995007722492228,
      "loss": 5.89,
      "mean_token_accuracy": 0.1478637784719467,
      "num_tokens": 7139896.0,
      "step": 3125
    },
    {
      "entropy": 5.860946559906006,
      "epoch": 0.30941083432186633,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994984199442811,
      "loss": 5.8406,
      "mean_token_accuracy": 0.1579609915614128,
      "num_tokens": 7151479.0,
      "step": 3130
    },
    {
      "entropy": 5.981082534790039,
      "epoch": 0.3099051008303677,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004994960621166434,
      "loss": 5.8752,
      "mean_token_accuracy": 0.1548340156674385,
      "num_tokens": 7162942.0,
      "step": 3135
    },
    {
      "entropy": 5.838286972045898,
      "epoch": 0.31039936733886914,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004994936987663678,
      "loss": 5.8426,
      "mean_token_accuracy": 0.16279832720756532,
      "num_tokens": 7174572.0,
      "step": 3140
    },
    {
      "entropy": 5.894599485397339,
      "epoch": 0.3108936338473705,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004994913298935124,
      "loss": 5.8984,
      "mean_token_accuracy": 0.14952836856245993,
      "num_tokens": 7186352.0,
      "step": 3145
    },
    {
      "entropy": 5.913238382339477,
      "epoch": 0.3113879003558719,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004994889554981353,
      "loss": 5.8677,
      "mean_token_accuracy": 0.1546166017651558,
      "num_tokens": 7197849.0,
      "step": 3150
    },
    {
      "entropy": 5.900243902206421,
      "epoch": 0.31188216686437326,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004994865755802953,
      "loss": 5.8454,
      "mean_token_accuracy": 0.14976087883114814,
      "num_tokens": 7207897.0,
      "step": 3155
    },
    {
      "entropy": 5.965626001358032,
      "epoch": 0.31237643337287463,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004994841901400505,
      "loss": 5.9604,
      "mean_token_accuracy": 0.14765686094760894,
      "num_tokens": 7219148.0,
      "step": 3160
    },
    {
      "entropy": 6.065278720855713,
      "epoch": 0.31287069988137606,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004994817991774599,
      "loss": 6.0262,
      "mean_token_accuracy": 0.13792147636413574,
      "num_tokens": 7230690.0,
      "step": 3165
    },
    {
      "entropy": 6.003130722045898,
      "epoch": 0.31336496638987743,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004994794026925822,
      "loss": 5.8691,
      "mean_token_accuracy": 0.15180785357952117,
      "num_tokens": 7240935.0,
      "step": 3170
    },
    {
      "entropy": 5.883154249191284,
      "epoch": 0.3138592328983788,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004994770006854763,
      "loss": 5.9264,
      "mean_token_accuracy": 0.15141893699765205,
      "num_tokens": 7253345.0,
      "step": 3175
    },
    {
      "entropy": 6.033541965484619,
      "epoch": 0.3143534994068802,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004994745931562015,
      "loss": 5.9278,
      "mean_token_accuracy": 0.1564277797937393,
      "num_tokens": 7265355.0,
      "step": 3180
    },
    {
      "entropy": 5.799361228942871,
      "epoch": 0.31484776591538155,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004994721801048168,
      "loss": 5.8203,
      "mean_token_accuracy": 0.15881146639585494,
      "num_tokens": 7277355.0,
      "step": 3185
    },
    {
      "entropy": 5.938270950317383,
      "epoch": 0.315342032423883,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004994697615313817,
      "loss": 5.884,
      "mean_token_accuracy": 0.15086214244365692,
      "num_tokens": 7288754.0,
      "step": 3190
    },
    {
      "entropy": 5.974021053314209,
      "epoch": 0.31583629893238435,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004994673374359556,
      "loss": 5.9535,
      "mean_token_accuracy": 0.14664196968078613,
      "num_tokens": 7301158.0,
      "step": 3195
    },
    {
      "entropy": 5.9225036144256595,
      "epoch": 0.3163305654408857,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004994649078185982,
      "loss": 5.9181,
      "mean_token_accuracy": 0.14927734285593033,
      "num_tokens": 7314420.0,
      "step": 3200
    },
    {
      "entropy": 5.934207391738892,
      "epoch": 0.3168248319493871,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004994624726793693,
      "loss": 5.8647,
      "mean_token_accuracy": 0.15222589820623397,
      "num_tokens": 7325459.0,
      "step": 3205
    },
    {
      "entropy": 5.925555992126465,
      "epoch": 0.31731909845788847,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004994600320183286,
      "loss": 5.8421,
      "mean_token_accuracy": 0.15424371659755706,
      "num_tokens": 7336016.0,
      "step": 3210
    },
    {
      "entropy": 5.891087865829467,
      "epoch": 0.3178133649663899,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004994575858355364,
      "loss": 5.8343,
      "mean_token_accuracy": 0.15168708860874175,
      "num_tokens": 7346716.0,
      "step": 3215
    },
    {
      "entropy": 5.9876645565032955,
      "epoch": 0.3183076314748913,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004994551341310526,
      "loss": 5.9459,
      "mean_token_accuracy": 0.1490081638097763,
      "num_tokens": 7357246.0,
      "step": 3220
    },
    {
      "entropy": 5.9510910987854,
      "epoch": 0.31880189798339265,
      "grad_norm": 1.25,
      "learning_rate": 0.0004994526769049379,
      "loss": 5.9666,
      "mean_token_accuracy": 0.1405780427157879,
      "num_tokens": 7369173.0,
      "step": 3225
    },
    {
      "entropy": 5.986663293838501,
      "epoch": 0.319296164491894,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004994502141572524,
      "loss": 5.9451,
      "mean_token_accuracy": 0.13844343274831772,
      "num_tokens": 7381118.0,
      "step": 3230
    },
    {
      "entropy": 5.9236208438873295,
      "epoch": 0.3197904310003954,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004994477458880568,
      "loss": 5.7938,
      "mean_token_accuracy": 0.15767068266868592,
      "num_tokens": 7391874.0,
      "step": 3235
    },
    {
      "entropy": 5.865980625152588,
      "epoch": 0.3202846975088968,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004994452720974119,
      "loss": 5.8872,
      "mean_token_accuracy": 0.15512483119964598,
      "num_tokens": 7403474.0,
      "step": 3240
    },
    {
      "entropy": 5.942899942398071,
      "epoch": 0.3207789640173982,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004994427927853785,
      "loss": 5.7869,
      "mean_token_accuracy": 0.15654317587614058,
      "num_tokens": 7415173.0,
      "step": 3245
    },
    {
      "entropy": 5.843418598175049,
      "epoch": 0.32127323052589957,
      "grad_norm": 1.125,
      "learning_rate": 0.0004994403079520175,
      "loss": 5.8195,
      "mean_token_accuracy": 0.15521105378866196,
      "num_tokens": 7426965.0,
      "step": 3250
    },
    {
      "entropy": 5.930048942565918,
      "epoch": 0.32176749703440094,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004994378175973902,
      "loss": 5.7899,
      "mean_token_accuracy": 0.15678596049547194,
      "num_tokens": 7438162.0,
      "step": 3255
    },
    {
      "entropy": 5.909450387954712,
      "epoch": 0.3222617635429023,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004994353217215577,
      "loss": 5.8716,
      "mean_token_accuracy": 0.1480770692229271,
      "num_tokens": 7450241.0,
      "step": 3260
    },
    {
      "entropy": 5.9811969757080075,
      "epoch": 0.32275603005140374,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004994328203245814,
      "loss": 5.9739,
      "mean_token_accuracy": 0.1421304538846016,
      "num_tokens": 7462236.0,
      "step": 3265
    },
    {
      "entropy": 5.981936073303222,
      "epoch": 0.3232502965599051,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004994303134065231,
      "loss": 5.8787,
      "mean_token_accuracy": 0.14774280935525894,
      "num_tokens": 7472811.0,
      "step": 3270
    },
    {
      "entropy": 5.823878812789917,
      "epoch": 0.3237445630684065,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004994278009674442,
      "loss": 5.7781,
      "mean_token_accuracy": 0.16131176203489303,
      "num_tokens": 7484837.0,
      "step": 3275
    },
    {
      "entropy": 5.996675300598144,
      "epoch": 0.32423882957690786,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004994252830074064,
      "loss": 5.8776,
      "mean_token_accuracy": 0.14760498106479644,
      "num_tokens": 7496712.0,
      "step": 3280
    },
    {
      "entropy": 5.824530220031738,
      "epoch": 0.32473309608540923,
      "grad_norm": 1.078125,
      "learning_rate": 0.000499422759526472,
      "loss": 5.8799,
      "mean_token_accuracy": 0.1516605421900749,
      "num_tokens": 7508014.0,
      "step": 3285
    },
    {
      "entropy": 5.916428327560425,
      "epoch": 0.32522736259391066,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004994202305247029,
      "loss": 5.9297,
      "mean_token_accuracy": 0.14520037919282913,
      "num_tokens": 7519466.0,
      "step": 3290
    },
    {
      "entropy": 6.085912036895752,
      "epoch": 0.32572162910241204,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004994176960021613,
      "loss": 5.9715,
      "mean_token_accuracy": 0.1455385245382786,
      "num_tokens": 7531661.0,
      "step": 3295
    },
    {
      "entropy": 5.927366018295288,
      "epoch": 0.3262158956109134,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004994151559589095,
      "loss": 5.8879,
      "mean_token_accuracy": 0.14917458668351175,
      "num_tokens": 7544096.0,
      "step": 3300
    },
    {
      "entropy": 5.924072742462158,
      "epoch": 0.3267101621194148,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004994126103950101,
      "loss": 5.863,
      "mean_token_accuracy": 0.14948750287294388,
      "num_tokens": 7556022.0,
      "step": 3305
    },
    {
      "entropy": 5.904584789276123,
      "epoch": 0.32720442862791616,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004994100593105257,
      "loss": 5.835,
      "mean_token_accuracy": 0.15778674483299254,
      "num_tokens": 7567399.0,
      "step": 3310
    },
    {
      "entropy": 5.845813608169555,
      "epoch": 0.3276986951364176,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004994075027055189,
      "loss": 5.9029,
      "mean_token_accuracy": 0.150156831741333,
      "num_tokens": 7580673.0,
      "step": 3315
    },
    {
      "entropy": 5.995374393463135,
      "epoch": 0.32819296164491896,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004994049405800529,
      "loss": 5.849,
      "mean_token_accuracy": 0.14884991571307182,
      "num_tokens": 7591799.0,
      "step": 3320
    },
    {
      "entropy": 5.906673097610474,
      "epoch": 0.32868722815342033,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004994023729341904,
      "loss": 5.8885,
      "mean_token_accuracy": 0.1490943245589733,
      "num_tokens": 7603663.0,
      "step": 3325
    },
    {
      "entropy": 5.905765390396118,
      "epoch": 0.3291814946619217,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004993997997679949,
      "loss": 5.8288,
      "mean_token_accuracy": 0.15654058754444122,
      "num_tokens": 7615651.0,
      "step": 3330
    },
    {
      "entropy": 5.823675203323364,
      "epoch": 0.3296757611704231,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004993972210815292,
      "loss": 5.8597,
      "mean_token_accuracy": 0.15760498493909836,
      "num_tokens": 7628033.0,
      "step": 3335
    },
    {
      "entropy": 5.996159982681275,
      "epoch": 0.33017002767892445,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004993946368748573,
      "loss": 5.9642,
      "mean_token_accuracy": 0.14904893189668655,
      "num_tokens": 7640227.0,
      "step": 3340
    },
    {
      "entropy": 5.973145627975464,
      "epoch": 0.3306642941874259,
      "grad_norm": 1.125,
      "learning_rate": 0.0004993920471480424,
      "loss": 5.8588,
      "mean_token_accuracy": 0.155265648663044,
      "num_tokens": 7651684.0,
      "step": 3345
    },
    {
      "entropy": 5.897155094146728,
      "epoch": 0.33115856069592725,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004993894519011484,
      "loss": 5.8992,
      "mean_token_accuracy": 0.15093035399913787,
      "num_tokens": 7662721.0,
      "step": 3350
    },
    {
      "entropy": 5.908133935928345,
      "epoch": 0.3316528272044286,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004993868511342391,
      "loss": 5.8487,
      "mean_token_accuracy": 0.15349284261465074,
      "num_tokens": 7674060.0,
      "step": 3355
    },
    {
      "entropy": 5.858733654022217,
      "epoch": 0.33214709371293,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004993842448473783,
      "loss": 5.8045,
      "mean_token_accuracy": 0.15487059578299522,
      "num_tokens": 7685054.0,
      "step": 3360
    },
    {
      "entropy": 5.886049079895019,
      "epoch": 0.33264136022143137,
      "grad_norm": 1.25,
      "learning_rate": 0.0004993816330406304,
      "loss": 5.9195,
      "mean_token_accuracy": 0.14408770725131034,
      "num_tokens": 7696554.0,
      "step": 3365
    },
    {
      "entropy": 6.02078652381897,
      "epoch": 0.3331356267299328,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004993790157140595,
      "loss": 5.7918,
      "mean_token_accuracy": 0.15052541792392732,
      "num_tokens": 7707000.0,
      "step": 3370
    },
    {
      "entropy": 5.9085273265838625,
      "epoch": 0.33362989323843417,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004993763928677299,
      "loss": 5.8716,
      "mean_token_accuracy": 0.15730493217706681,
      "num_tokens": 7718898.0,
      "step": 3375
    },
    {
      "entropy": 5.933470582962036,
      "epoch": 0.33412415974693555,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004993737645017064,
      "loss": 5.9046,
      "mean_token_accuracy": 0.15121666342020035,
      "num_tokens": 7730493.0,
      "step": 3380
    },
    {
      "entropy": 5.896996593475341,
      "epoch": 0.3346184262554369,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004993711306160534,
      "loss": 5.8152,
      "mean_token_accuracy": 0.15182201862335204,
      "num_tokens": 7742177.0,
      "step": 3385
    },
    {
      "entropy": 5.885739374160766,
      "epoch": 0.3351126927639383,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004993684912108358,
      "loss": 5.8901,
      "mean_token_accuracy": 0.1533852696418762,
      "num_tokens": 7753386.0,
      "step": 3390
    },
    {
      "entropy": 5.854862308502197,
      "epoch": 0.3356069592724397,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004993658462861184,
      "loss": 5.7926,
      "mean_token_accuracy": 0.15316681638360025,
      "num_tokens": 7764259.0,
      "step": 3395
    },
    {
      "entropy": 5.915011262893676,
      "epoch": 0.3361012257809411,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004993631958419666,
      "loss": 5.7892,
      "mean_token_accuracy": 0.15640075802803038,
      "num_tokens": 7775853.0,
      "step": 3400
    },
    {
      "entropy": 5.902416372299195,
      "epoch": 0.33659549228944247,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004993605398784451,
      "loss": 5.9146,
      "mean_token_accuracy": 0.139879260212183,
      "num_tokens": 7787116.0,
      "step": 3405
    },
    {
      "entropy": 5.956493091583252,
      "epoch": 0.33708975879794384,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004993578783956198,
      "loss": 5.9393,
      "mean_token_accuracy": 0.1419096127152443,
      "num_tokens": 7799666.0,
      "step": 3410
    },
    {
      "entropy": 5.90988245010376,
      "epoch": 0.3375840253064452,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004993552113935557,
      "loss": 5.8263,
      "mean_token_accuracy": 0.1583866521716118,
      "num_tokens": 7812454.0,
      "step": 3415
    },
    {
      "entropy": 5.916579532623291,
      "epoch": 0.33807829181494664,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004993525388723186,
      "loss": 5.8834,
      "mean_token_accuracy": 0.1530376985669136,
      "num_tokens": 7824427.0,
      "step": 3420
    },
    {
      "entropy": 5.901572895050049,
      "epoch": 0.338572558323448,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004993498608319742,
      "loss": 5.8844,
      "mean_token_accuracy": 0.15276393666863441,
      "num_tokens": 7835382.0,
      "step": 3425
    },
    {
      "entropy": 5.859151458740234,
      "epoch": 0.3390668248319494,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004993471772725886,
      "loss": 5.7989,
      "mean_token_accuracy": 0.15550214797258377,
      "num_tokens": 7846618.0,
      "step": 3430
    },
    {
      "entropy": 5.901988458633423,
      "epoch": 0.33956109134045076,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004993444881942274,
      "loss": 5.8146,
      "mean_token_accuracy": 0.15682211965322496,
      "num_tokens": 7859160.0,
      "step": 3435
    },
    {
      "entropy": 5.8803332328796385,
      "epoch": 0.34005535784895213,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004993417935969572,
      "loss": 5.8434,
      "mean_token_accuracy": 0.15452905595302582,
      "num_tokens": 7870628.0,
      "step": 3440
    },
    {
      "entropy": 5.94140362739563,
      "epoch": 0.34054962435745356,
      "grad_norm": 1.109375,
      "learning_rate": 0.000499339093480844,
      "loss": 5.8743,
      "mean_token_accuracy": 0.1545655608177185,
      "num_tokens": 7883016.0,
      "step": 3445
    },
    {
      "entropy": 5.867698383331299,
      "epoch": 0.34104389086595494,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004993363878459542,
      "loss": 5.8751,
      "mean_token_accuracy": 0.14724137187004088,
      "num_tokens": 7895940.0,
      "step": 3450
    },
    {
      "entropy": 5.9430724620819095,
      "epoch": 0.3415381573744563,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004993336766923546,
      "loss": 5.9059,
      "mean_token_accuracy": 0.14910922944545746,
      "num_tokens": 7907920.0,
      "step": 3455
    },
    {
      "entropy": 5.952860927581787,
      "epoch": 0.3420324238829577,
      "grad_norm": 1.375,
      "learning_rate": 0.0004993309600201117,
      "loss": 5.889,
      "mean_token_accuracy": 0.15425898805260657,
      "num_tokens": 7920181.0,
      "step": 3460
    },
    {
      "entropy": 5.888333415985107,
      "epoch": 0.34252669039145905,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004993282378292923,
      "loss": 5.8396,
      "mean_token_accuracy": 0.1592525601387024,
      "num_tokens": 7931135.0,
      "step": 3465
    },
    {
      "entropy": 5.885323476791382,
      "epoch": 0.3430209568999605,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004993255101199635,
      "loss": 5.8211,
      "mean_token_accuracy": 0.15970962271094322,
      "num_tokens": 7941002.0,
      "step": 3470
    },
    {
      "entropy": 5.942181777954102,
      "epoch": 0.34351522340846186,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004993227768921924,
      "loss": 5.792,
      "mean_token_accuracy": 0.1554045557975769,
      "num_tokens": 7950393.0,
      "step": 3475
    },
    {
      "entropy": 5.9116617202758786,
      "epoch": 0.34400948991696323,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004993200381460461,
      "loss": 5.8258,
      "mean_token_accuracy": 0.15245998203754424,
      "num_tokens": 7961468.0,
      "step": 3480
    },
    {
      "entropy": 5.817622375488281,
      "epoch": 0.3445037564254646,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004993172938815922,
      "loss": 5.6706,
      "mean_token_accuracy": 0.1642368882894516,
      "num_tokens": 7973783.0,
      "step": 3485
    },
    {
      "entropy": 5.789644384384156,
      "epoch": 0.344998022933966,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004993145440988979,
      "loss": 5.7316,
      "mean_token_accuracy": 0.16548056602478028,
      "num_tokens": 7984099.0,
      "step": 3490
    },
    {
      "entropy": 5.954108858108521,
      "epoch": 0.3454922894424674,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004993117887980311,
      "loss": 5.8921,
      "mean_token_accuracy": 0.15117180049419404,
      "num_tokens": 7995115.0,
      "step": 3495
    },
    {
      "entropy": 5.920248556137085,
      "epoch": 0.3459865559509688,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004993090279790596,
      "loss": 5.8406,
      "mean_token_accuracy": 0.15169650912284852,
      "num_tokens": 8006103.0,
      "step": 3500
    },
    {
      "entropy": 5.7916463851928714,
      "epoch": 0.34648082245947015,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000499306261642051,
      "loss": 5.8067,
      "mean_token_accuracy": 0.14806041717529297,
      "num_tokens": 8018228.0,
      "step": 3505
    },
    {
      "entropy": 5.884153890609741,
      "epoch": 0.3469750889679715,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004993034897870737,
      "loss": 5.7563,
      "mean_token_accuracy": 0.15497301369905472,
      "num_tokens": 8029808.0,
      "step": 3510
    },
    {
      "entropy": 5.843921661376953,
      "epoch": 0.3474693554764729,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004993007124141958,
      "loss": 5.8536,
      "mean_token_accuracy": 0.14964404553174973,
      "num_tokens": 8041136.0,
      "step": 3515
    },
    {
      "entropy": 5.920845794677734,
      "epoch": 0.3479636219849743,
      "grad_norm": 1.25,
      "learning_rate": 0.0004992979295234854,
      "loss": 5.9013,
      "mean_token_accuracy": 0.14759098887443542,
      "num_tokens": 8053013.0,
      "step": 3520
    },
    {
      "entropy": 5.940861701965332,
      "epoch": 0.3484578884934757,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004992951411150112,
      "loss": 5.8713,
      "mean_token_accuracy": 0.14649871438741685,
      "num_tokens": 8065139.0,
      "step": 3525
    },
    {
      "entropy": 5.860837078094482,
      "epoch": 0.34895215500197707,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004992923471888418,
      "loss": 5.7316,
      "mean_token_accuracy": 0.16246522814035416,
      "num_tokens": 8075719.0,
      "step": 3530
    },
    {
      "entropy": 5.876574087142944,
      "epoch": 0.34944642151047844,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004992895477450459,
      "loss": 5.8621,
      "mean_token_accuracy": 0.15280634313821792,
      "num_tokens": 8087821.0,
      "step": 3535
    },
    {
      "entropy": 5.838806915283203,
      "epoch": 0.3499406880189798,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004992867427836923,
      "loss": 5.755,
      "mean_token_accuracy": 0.15870047360658646,
      "num_tokens": 8099861.0,
      "step": 3540
    },
    {
      "entropy": 6.010531711578369,
      "epoch": 0.3504349545274812,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004992839323048501,
      "loss": 5.9882,
      "mean_token_accuracy": 0.14447013214230536,
      "num_tokens": 8111671.0,
      "step": 3545
    },
    {
      "entropy": 6.0363317966461185,
      "epoch": 0.3509292210359826,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004992811163085882,
      "loss": 5.9404,
      "mean_token_accuracy": 0.150308321416378,
      "num_tokens": 8123569.0,
      "step": 3550
    },
    {
      "entropy": 5.9858965396881105,
      "epoch": 0.351423487544484,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004992782947949761,
      "loss": 5.9541,
      "mean_token_accuracy": 0.15137761533260347,
      "num_tokens": 8135621.0,
      "step": 3555
    },
    {
      "entropy": 5.8561443328857425,
      "epoch": 0.35191775405298537,
      "grad_norm": 1.25,
      "learning_rate": 0.0004992754677640831,
      "loss": 5.7978,
      "mean_token_accuracy": 0.15685692727565764,
      "num_tokens": 8147012.0,
      "step": 3560
    },
    {
      "entropy": 5.908249092102051,
      "epoch": 0.35241202056148674,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004992726352159789,
      "loss": 5.7851,
      "mean_token_accuracy": 0.15655775964260102,
      "num_tokens": 8158686.0,
      "step": 3565
    },
    {
      "entropy": 5.85140061378479,
      "epoch": 0.3529062870699881,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004992697971507331,
      "loss": 5.7901,
      "mean_token_accuracy": 0.15729555934667588,
      "num_tokens": 8170188.0,
      "step": 3570
    },
    {
      "entropy": 5.890997982025146,
      "epoch": 0.35340055357848954,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004992669535684155,
      "loss": 5.8717,
      "mean_token_accuracy": 0.15248715728521348,
      "num_tokens": 8182467.0,
      "step": 3575
    },
    {
      "entropy": 5.957807207107544,
      "epoch": 0.3538948200869909,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004992641044690959,
      "loss": 5.9638,
      "mean_token_accuracy": 0.15112252160906792,
      "num_tokens": 8193433.0,
      "step": 3580
    },
    {
      "entropy": 5.893219804763794,
      "epoch": 0.3543890865954923,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004992612498528447,
      "loss": 5.7478,
      "mean_token_accuracy": 0.1555414766073227,
      "num_tokens": 8204220.0,
      "step": 3585
    },
    {
      "entropy": 5.810119009017944,
      "epoch": 0.35488335310399366,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004992583897197319,
      "loss": 5.7642,
      "mean_token_accuracy": 0.16225289255380632,
      "num_tokens": 8214203.0,
      "step": 3590
    },
    {
      "entropy": 5.9189142227172855,
      "epoch": 0.35537761961249503,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004992555240698279,
      "loss": 5.9038,
      "mean_token_accuracy": 0.14603765159845353,
      "num_tokens": 8225143.0,
      "step": 3595
    },
    {
      "entropy": 5.91175103187561,
      "epoch": 0.35587188612099646,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004992526529032032,
      "loss": 5.7585,
      "mean_token_accuracy": 0.15905050486326217,
      "num_tokens": 8237715.0,
      "step": 3600
    },
    {
      "entropy": 5.805362415313721,
      "epoch": 0.35636615262949783,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004992497762199285,
      "loss": 5.8196,
      "mean_token_accuracy": 0.15873343497514725,
      "num_tokens": 8249645.0,
      "step": 3605
    },
    {
      "entropy": 5.945392084121704,
      "epoch": 0.3568604191379992,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004992468940200745,
      "loss": 5.7897,
      "mean_token_accuracy": 0.1611865848302841,
      "num_tokens": 8260624.0,
      "step": 3610
    },
    {
      "entropy": 5.910581159591675,
      "epoch": 0.3573546856465006,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004992440063037121,
      "loss": 5.8894,
      "mean_token_accuracy": 0.1492436170578003,
      "num_tokens": 8272526.0,
      "step": 3615
    },
    {
      "entropy": 5.777053737640381,
      "epoch": 0.35784895215500195,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004992411130709124,
      "loss": 5.7908,
      "mean_token_accuracy": 0.15661219209432603,
      "num_tokens": 8284270.0,
      "step": 3620
    },
    {
      "entropy": 5.969174575805664,
      "epoch": 0.3583432186635034,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004992382143217466,
      "loss": 5.9021,
      "mean_token_accuracy": 0.15277914106845855,
      "num_tokens": 8296407.0,
      "step": 3625
    },
    {
      "entropy": 5.841032934188843,
      "epoch": 0.35883748517200476,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004992353100562858,
      "loss": 5.805,
      "mean_token_accuracy": 0.16218566447496413,
      "num_tokens": 8307695.0,
      "step": 3630
    },
    {
      "entropy": 5.794250535964966,
      "epoch": 0.35933175168050613,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004992324002746016,
      "loss": 5.7942,
      "mean_token_accuracy": 0.15522915124893188,
      "num_tokens": 8319458.0,
      "step": 3635
    },
    {
      "entropy": 5.872878932952881,
      "epoch": 0.3598260181890075,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004992294849767656,
      "loss": 5.7111,
      "mean_token_accuracy": 0.16143833845853806,
      "num_tokens": 8330284.0,
      "step": 3640
    },
    {
      "entropy": 5.853314065933228,
      "epoch": 0.3603202846975089,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004992265641628495,
      "loss": 5.827,
      "mean_token_accuracy": 0.15445810109376906,
      "num_tokens": 8343622.0,
      "step": 3645
    },
    {
      "entropy": 5.894721937179566,
      "epoch": 0.3608145512060103,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004992236378329252,
      "loss": 5.7864,
      "mean_token_accuracy": 0.1555868223309517,
      "num_tokens": 8354079.0,
      "step": 3650
    },
    {
      "entropy": 5.8952436447143555,
      "epoch": 0.3613088177145117,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004992207059870645,
      "loss": 5.83,
      "mean_token_accuracy": 0.15822544246912001,
      "num_tokens": 8364876.0,
      "step": 3655
    },
    {
      "entropy": 5.934147882461548,
      "epoch": 0.36180308422301305,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004992177686253396,
      "loss": 5.7813,
      "mean_token_accuracy": 0.15549880415201187,
      "num_tokens": 8375095.0,
      "step": 3660
    },
    {
      "entropy": 5.7936859130859375,
      "epoch": 0.3622973507315144,
      "grad_norm": 1.15625,
      "learning_rate": 0.000499214825747823,
      "loss": 5.7544,
      "mean_token_accuracy": 0.15899945050477982,
      "num_tokens": 8385934.0,
      "step": 3665
    },
    {
      "entropy": 5.8251642227172855,
      "epoch": 0.3627916172400158,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004992118773545868,
      "loss": 5.7247,
      "mean_token_accuracy": 0.1615453451871872,
      "num_tokens": 8397018.0,
      "step": 3670
    },
    {
      "entropy": 5.841867876052857,
      "epoch": 0.3632858837485172,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004992089234457036,
      "loss": 5.7961,
      "mean_token_accuracy": 0.1640404224395752,
      "num_tokens": 8406668.0,
      "step": 3675
    },
    {
      "entropy": 5.953409671783447,
      "epoch": 0.3637801502570186,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004992059640212461,
      "loss": 5.8298,
      "mean_token_accuracy": 0.15458669289946556,
      "num_tokens": 8417471.0,
      "step": 3680
    },
    {
      "entropy": 5.856666088104248,
      "epoch": 0.36427441676551997,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004992029990812872,
      "loss": 5.9011,
      "mean_token_accuracy": 0.14456111043691636,
      "num_tokens": 8430073.0,
      "step": 3685
    },
    {
      "entropy": 5.893247652053833,
      "epoch": 0.36476868327402134,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004992000286258997,
      "loss": 5.8086,
      "mean_token_accuracy": 0.15492908284068108,
      "num_tokens": 8442573.0,
      "step": 3690
    },
    {
      "entropy": 5.940594816207886,
      "epoch": 0.3652629497825227,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004991970526551567,
      "loss": 5.8576,
      "mean_token_accuracy": 0.15471384078264236,
      "num_tokens": 8453920.0,
      "step": 3695
    },
    {
      "entropy": 5.851522207260132,
      "epoch": 0.36575721629102415,
      "grad_norm": 1.125,
      "learning_rate": 0.0004991940711691314,
      "loss": 5.8945,
      "mean_token_accuracy": 0.15315308719873427,
      "num_tokens": 8466652.0,
      "step": 3700
    },
    {
      "entropy": 5.93485655784607,
      "epoch": 0.3662514827995255,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004991910841678972,
      "loss": 5.9105,
      "mean_token_accuracy": 0.14977106899023057,
      "num_tokens": 8480203.0,
      "step": 3705
    },
    {
      "entropy": 5.864224863052368,
      "epoch": 0.3667457493080269,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004991880916515275,
      "loss": 5.8174,
      "mean_token_accuracy": 0.15368005335330964,
      "num_tokens": 8491396.0,
      "step": 3710
    },
    {
      "entropy": 5.918771553039551,
      "epoch": 0.36724001581652826,
      "grad_norm": 1.15625,
      "learning_rate": 0.000499185093620096,
      "loss": 5.7838,
      "mean_token_accuracy": 0.15006800889968872,
      "num_tokens": 8501838.0,
      "step": 3715
    },
    {
      "entropy": 5.798188638687134,
      "epoch": 0.36773428232502964,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004991820900736765,
      "loss": 5.6723,
      "mean_token_accuracy": 0.1582162246108055,
      "num_tokens": 8512113.0,
      "step": 3720
    },
    {
      "entropy": 5.831536865234375,
      "epoch": 0.36822854883353107,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004991790810123427,
      "loss": 5.7507,
      "mean_token_accuracy": 0.15408645272254945,
      "num_tokens": 8523888.0,
      "step": 3725
    },
    {
      "entropy": 5.79871392250061,
      "epoch": 0.36872281534203244,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004991760664361688,
      "loss": 5.6859,
      "mean_token_accuracy": 0.1632385030388832,
      "num_tokens": 8536078.0,
      "step": 3730
    },
    {
      "entropy": 5.837257719039917,
      "epoch": 0.3692170818505338,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004991730463452288,
      "loss": 5.8265,
      "mean_token_accuracy": 0.15056723803281785,
      "num_tokens": 8547131.0,
      "step": 3735
    },
    {
      "entropy": 5.847607707977295,
      "epoch": 0.3697113483590352,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004991700207395971,
      "loss": 5.6923,
      "mean_token_accuracy": 0.16566429436206817,
      "num_tokens": 8557269.0,
      "step": 3740
    },
    {
      "entropy": 5.921040964126587,
      "epoch": 0.37020561486753656,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004991669896193482,
      "loss": 5.8815,
      "mean_token_accuracy": 0.146759133040905,
      "num_tokens": 8569193.0,
      "step": 3745
    },
    {
      "entropy": 5.870231437683105,
      "epoch": 0.370699881376038,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004991639529845565,
      "loss": 5.7533,
      "mean_token_accuracy": 0.15931029468774796,
      "num_tokens": 8580888.0,
      "step": 3750
    },
    {
      "entropy": 5.840953636169433,
      "epoch": 0.37119414788453936,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004991609108352968,
      "loss": 5.836,
      "mean_token_accuracy": 0.1561110332608223,
      "num_tokens": 8591868.0,
      "step": 3755
    },
    {
      "entropy": 5.83723406791687,
      "epoch": 0.37168841439304073,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004991578631716438,
      "loss": 5.7962,
      "mean_token_accuracy": 0.15383539870381355,
      "num_tokens": 8602338.0,
      "step": 3760
    },
    {
      "entropy": 5.8507262706756595,
      "epoch": 0.3721826809015421,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004991548099936727,
      "loss": 5.7276,
      "mean_token_accuracy": 0.15699041187763213,
      "num_tokens": 8611698.0,
      "step": 3765
    },
    {
      "entropy": 5.836185693740845,
      "epoch": 0.3726769474100435,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004991517513014585,
      "loss": 5.7327,
      "mean_token_accuracy": 0.1610111728310585,
      "num_tokens": 8623537.0,
      "step": 3770
    },
    {
      "entropy": 5.787095975875855,
      "epoch": 0.37317121391854485,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004991486870950765,
      "loss": 5.7996,
      "mean_token_accuracy": 0.1559009499847889,
      "num_tokens": 8634678.0,
      "step": 3775
    },
    {
      "entropy": 5.864709520339966,
      "epoch": 0.3736654804270463,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004991456173746021,
      "loss": 5.7777,
      "mean_token_accuracy": 0.16089840829372407,
      "num_tokens": 8646207.0,
      "step": 3780
    },
    {
      "entropy": 5.955631351470947,
      "epoch": 0.37415974693554765,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004991425421401105,
      "loss": 5.839,
      "mean_token_accuracy": 0.15324022620916367,
      "num_tokens": 8657069.0,
      "step": 3785
    },
    {
      "entropy": 5.766555738449097,
      "epoch": 0.374654013444049,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004991394613916777,
      "loss": 5.6965,
      "mean_token_accuracy": 0.16205490231513978,
      "num_tokens": 8668296.0,
      "step": 3790
    },
    {
      "entropy": 5.816662788391113,
      "epoch": 0.3751482799525504,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004991363751293795,
      "loss": 5.8514,
      "mean_token_accuracy": 0.15196674913167954,
      "num_tokens": 8680626.0,
      "step": 3795
    },
    {
      "entropy": 5.868446969985962,
      "epoch": 0.3756425464610518,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004991332833532915,
      "loss": 5.7609,
      "mean_token_accuracy": 0.16015817373991012,
      "num_tokens": 8691571.0,
      "step": 3800
    },
    {
      "entropy": 5.863410568237304,
      "epoch": 0.3761368129695532,
      "grad_norm": 1.140625,
      "learning_rate": 0.00049913018606349,
      "loss": 5.814,
      "mean_token_accuracy": 0.16047453731298447,
      "num_tokens": 8703262.0,
      "step": 3805
    },
    {
      "entropy": 5.922005653381348,
      "epoch": 0.3766310794780546,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004991270832600512,
      "loss": 5.9128,
      "mean_token_accuracy": 0.1440843015909195,
      "num_tokens": 8714799.0,
      "step": 3810
    },
    {
      "entropy": 5.948727369308472,
      "epoch": 0.37712534598655595,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004991239749430514,
      "loss": 5.8848,
      "mean_token_accuracy": 0.14975858181715013,
      "num_tokens": 8726417.0,
      "step": 3815
    },
    {
      "entropy": 5.77113709449768,
      "epoch": 0.3776196124950573,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000499120861112567,
      "loss": 5.7504,
      "mean_token_accuracy": 0.16276681572198867,
      "num_tokens": 8738572.0,
      "step": 3820
    },
    {
      "entropy": 5.987140226364136,
      "epoch": 0.3781138790035587,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004991177417686747,
      "loss": 5.9179,
      "mean_token_accuracy": 0.1479107543826103,
      "num_tokens": 8750557.0,
      "step": 3825
    },
    {
      "entropy": 5.889439392089844,
      "epoch": 0.3786081455120601,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004991146169114511,
      "loss": 5.7055,
      "mean_token_accuracy": 0.1570228233933449,
      "num_tokens": 8761921.0,
      "step": 3830
    },
    {
      "entropy": 5.750439548492432,
      "epoch": 0.3791024120205615,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004991114865409732,
      "loss": 5.7991,
      "mean_token_accuracy": 0.15767900198698043,
      "num_tokens": 8772705.0,
      "step": 3835
    },
    {
      "entropy": 5.842856407165527,
      "epoch": 0.37959667852906287,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000499108350657318,
      "loss": 5.8153,
      "mean_token_accuracy": 0.1552213490009308,
      "num_tokens": 8783563.0,
      "step": 3840
    },
    {
      "entropy": 5.858430480957031,
      "epoch": 0.38009094503756424,
      "grad_norm": 1.25,
      "learning_rate": 0.0004991052092605627,
      "loss": 5.7152,
      "mean_token_accuracy": 0.16301678866147995,
      "num_tokens": 8794171.0,
      "step": 3845
    },
    {
      "entropy": 5.80806884765625,
      "epoch": 0.3805852115460656,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004991020623507842,
      "loss": 5.8158,
      "mean_token_accuracy": 0.1521042175590992,
      "num_tokens": 8806487.0,
      "step": 3850
    },
    {
      "entropy": 5.895160150527954,
      "epoch": 0.38107947805456704,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004990989099280604,
      "loss": 5.8211,
      "mean_token_accuracy": 0.1585649460554123,
      "num_tokens": 8817819.0,
      "step": 3855
    },
    {
      "entropy": 5.874980354309082,
      "epoch": 0.3815737445630684,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004990957519924686,
      "loss": 5.764,
      "mean_token_accuracy": 0.15150784254074096,
      "num_tokens": 8828485.0,
      "step": 3860
    },
    {
      "entropy": 5.821585559844971,
      "epoch": 0.3820680110715698,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004990925885440865,
      "loss": 5.8205,
      "mean_token_accuracy": 0.15444822385907173,
      "num_tokens": 8840179.0,
      "step": 3865
    },
    {
      "entropy": 5.953197860717774,
      "epoch": 0.38256227758007116,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004990894195829921,
      "loss": 5.8494,
      "mean_token_accuracy": 0.1579738438129425,
      "num_tokens": 8851813.0,
      "step": 3870
    },
    {
      "entropy": 5.844118452072143,
      "epoch": 0.38305654408857254,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004990862451092631,
      "loss": 5.7774,
      "mean_token_accuracy": 0.15051404684782027,
      "num_tokens": 8863072.0,
      "step": 3875
    },
    {
      "entropy": 5.880052471160889,
      "epoch": 0.38355081059707397,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004990830651229776,
      "loss": 5.7813,
      "mean_token_accuracy": 0.15792119652032852,
      "num_tokens": 8873510.0,
      "step": 3880
    },
    {
      "entropy": 5.8031305313110355,
      "epoch": 0.38404507710557534,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004990798796242142,
      "loss": 5.7356,
      "mean_token_accuracy": 0.15857678204774855,
      "num_tokens": 8884296.0,
      "step": 3885
    },
    {
      "entropy": 5.851049709320068,
      "epoch": 0.3845393436140767,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004990766886130508,
      "loss": 5.7761,
      "mean_token_accuracy": 0.16058651506900787,
      "num_tokens": 8894208.0,
      "step": 3890
    },
    {
      "entropy": 5.988053703308106,
      "epoch": 0.3850336101225781,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004990734920895662,
      "loss": 5.9627,
      "mean_token_accuracy": 0.14808114171028136,
      "num_tokens": 8905703.0,
      "step": 3895
    },
    {
      "entropy": 5.8592668056488035,
      "epoch": 0.38552787663107946,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004990702900538388,
      "loss": 5.7484,
      "mean_token_accuracy": 0.15851551443338394,
      "num_tokens": 8918077.0,
      "step": 3900
    },
    {
      "entropy": 5.788652181625366,
      "epoch": 0.3860221431395809,
      "grad_norm": 1.125,
      "learning_rate": 0.0004990670825059477,
      "loss": 5.7267,
      "mean_token_accuracy": 0.15688996613025666,
      "num_tokens": 8930240.0,
      "step": 3905
    },
    {
      "entropy": 5.900861597061157,
      "epoch": 0.38651640964808226,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004990638694459715,
      "loss": 5.7986,
      "mean_token_accuracy": 0.15847810059785844,
      "num_tokens": 8941421.0,
      "step": 3910
    },
    {
      "entropy": 5.872343826293945,
      "epoch": 0.38701067615658363,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004990606508739893,
      "loss": 5.8115,
      "mean_token_accuracy": 0.15537682622671128,
      "num_tokens": 8952296.0,
      "step": 3915
    },
    {
      "entropy": 5.869143915176392,
      "epoch": 0.387504942665085,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004990574267900804,
      "loss": 5.8186,
      "mean_token_accuracy": 0.15669666230678558,
      "num_tokens": 8963167.0,
      "step": 3920
    },
    {
      "entropy": 5.86326756477356,
      "epoch": 0.3879992091735864,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004990541971943241,
      "loss": 5.787,
      "mean_token_accuracy": 0.15738223493099213,
      "num_tokens": 8974405.0,
      "step": 3925
    },
    {
      "entropy": 5.871088266372681,
      "epoch": 0.3884934756820878,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004990509620867997,
      "loss": 5.7112,
      "mean_token_accuracy": 0.16412873268127443,
      "num_tokens": 8986012.0,
      "step": 3930
    },
    {
      "entropy": 5.720556116104126,
      "epoch": 0.3889877421905892,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004990477214675868,
      "loss": 5.694,
      "mean_token_accuracy": 0.16026492416858673,
      "num_tokens": 8996759.0,
      "step": 3935
    },
    {
      "entropy": 5.849727296829224,
      "epoch": 0.38948200869909055,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004990444753367653,
      "loss": 5.829,
      "mean_token_accuracy": 0.15216880887746811,
      "num_tokens": 9007215.0,
      "step": 3940
    },
    {
      "entropy": 5.791629219055176,
      "epoch": 0.3899762752075919,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004990412236944149,
      "loss": 5.7292,
      "mean_token_accuracy": 0.17175416946411132,
      "num_tokens": 9018281.0,
      "step": 3945
    },
    {
      "entropy": 5.912288618087769,
      "epoch": 0.3904705417160933,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004990379665406157,
      "loss": 5.8764,
      "mean_token_accuracy": 0.15223623141646386,
      "num_tokens": 9029312.0,
      "step": 3950
    },
    {
      "entropy": 5.94501895904541,
      "epoch": 0.39096480822459473,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004990347038754477,
      "loss": 5.8566,
      "mean_token_accuracy": 0.14973756968975066,
      "num_tokens": 9040273.0,
      "step": 3955
    },
    {
      "entropy": 5.866108179092407,
      "epoch": 0.3914590747330961,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004990314356989912,
      "loss": 5.7922,
      "mean_token_accuracy": 0.15092711076140403,
      "num_tokens": 9052571.0,
      "step": 3960
    },
    {
      "entropy": 5.806940126419067,
      "epoch": 0.3919533412415975,
      "grad_norm": 1.5,
      "learning_rate": 0.0004990281620113268,
      "loss": 5.7422,
      "mean_token_accuracy": 0.15850640684366227,
      "num_tokens": 9063531.0,
      "step": 3965
    },
    {
      "entropy": 5.925866508483887,
      "epoch": 0.39244760775009885,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004990248828125347,
      "loss": 5.7836,
      "mean_token_accuracy": 0.15395828783512117,
      "num_tokens": 9075404.0,
      "step": 3970
    },
    {
      "entropy": 5.757915782928467,
      "epoch": 0.3929418742586002,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004990215981026958,
      "loss": 5.8031,
      "mean_token_accuracy": 0.15948758646845818,
      "num_tokens": 9086570.0,
      "step": 3975
    },
    {
      "entropy": 5.876324081420899,
      "epoch": 0.3934361407671016,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004990183078818908,
      "loss": 5.6828,
      "mean_token_accuracy": 0.1598816230893135,
      "num_tokens": 9097993.0,
      "step": 3980
    },
    {
      "entropy": 5.838431692123413,
      "epoch": 0.393930407275603,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004990150121502007,
      "loss": 5.7628,
      "mean_token_accuracy": 0.16408235877752303,
      "num_tokens": 9108783.0,
      "step": 3985
    },
    {
      "entropy": 5.788733673095703,
      "epoch": 0.3944246737841044,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004990117109077066,
      "loss": 5.8128,
      "mean_token_accuracy": 0.15555327832698823,
      "num_tokens": 9120039.0,
      "step": 3990
    },
    {
      "entropy": 5.8930316925048825,
      "epoch": 0.39491894029260577,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004990084041544895,
      "loss": 5.7652,
      "mean_token_accuracy": 0.15779706388711928,
      "num_tokens": 9131543.0,
      "step": 3995
    },
    {
      "entropy": 5.815483951568604,
      "epoch": 0.39541320680110714,
      "grad_norm": 1.078125,
      "learning_rate": 0.000499005091890631,
      "loss": 5.7806,
      "mean_token_accuracy": 0.15455961227416992,
      "num_tokens": 9142968.0,
      "step": 4000
    },
    {
      "entropy": 5.782220888137817,
      "epoch": 0.3959074733096085,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004990017741162125,
      "loss": 5.8476,
      "mean_token_accuracy": 0.15507417023181916,
      "num_tokens": 9155705.0,
      "step": 4005
    },
    {
      "entropy": 5.916092014312744,
      "epoch": 0.39640173981810994,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004989984508313156,
      "loss": 5.8555,
      "mean_token_accuracy": 0.14767262265086173,
      "num_tokens": 9167575.0,
      "step": 4010
    },
    {
      "entropy": 5.864837169647217,
      "epoch": 0.3968960063266113,
      "grad_norm": 1.203125,
      "learning_rate": 0.000498995122036022,
      "loss": 5.8303,
      "mean_token_accuracy": 0.15483421534299852,
      "num_tokens": 9178493.0,
      "step": 4015
    },
    {
      "entropy": 5.9377106666564945,
      "epoch": 0.3973902728351127,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004989917877304138,
      "loss": 5.8824,
      "mean_token_accuracy": 0.1505347326397896,
      "num_tokens": 9189677.0,
      "step": 4020
    },
    {
      "entropy": 5.827444219589234,
      "epoch": 0.39788453934361406,
      "grad_norm": 1.25,
      "learning_rate": 0.0004989884479145727,
      "loss": 5.7078,
      "mean_token_accuracy": 0.15970520824193954,
      "num_tokens": 9199179.0,
      "step": 4025
    },
    {
      "entropy": 5.905575847625732,
      "epoch": 0.39837880585211544,
      "grad_norm": 1.125,
      "learning_rate": 0.0004989851025885812,
      "loss": 5.778,
      "mean_token_accuracy": 0.15371362268924713,
      "num_tokens": 9210689.0,
      "step": 4030
    },
    {
      "entropy": 5.812503290176392,
      "epoch": 0.39887307236061686,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004989817517525212,
      "loss": 5.7441,
      "mean_token_accuracy": 0.15978539288043975,
      "num_tokens": 9221710.0,
      "step": 4035
    },
    {
      "entropy": 5.845702409744263,
      "epoch": 0.39936733886911824,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004989783954064755,
      "loss": 5.84,
      "mean_token_accuracy": 0.15523956567049027,
      "num_tokens": 9234079.0,
      "step": 4040
    },
    {
      "entropy": 5.842787218093872,
      "epoch": 0.3998616053776196,
      "grad_norm": 1.125,
      "learning_rate": 0.0004989750335505265,
      "loss": 5.7661,
      "mean_token_accuracy": 0.15682925432920455,
      "num_tokens": 9245050.0,
      "step": 4045
    },
    {
      "entropy": 5.9115033626556395,
      "epoch": 0.400355871886121,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004989716661847571,
      "loss": 5.8554,
      "mean_token_accuracy": 0.15244610458612443,
      "num_tokens": 9256455.0,
      "step": 4050
    },
    {
      "entropy": 5.779877328872681,
      "epoch": 0.40085013839462236,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004989682933092497,
      "loss": 5.7513,
      "mean_token_accuracy": 0.16125411689281463,
      "num_tokens": 9267722.0,
      "step": 4055
    },
    {
      "entropy": 5.817647457122803,
      "epoch": 0.4013444049031238,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004989649149240877,
      "loss": 5.7853,
      "mean_token_accuracy": 0.15072275847196578,
      "num_tokens": 9277533.0,
      "step": 4060
    },
    {
      "entropy": 5.887001609802246,
      "epoch": 0.40183867141162516,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000498961531029354,
      "loss": 5.8136,
      "mean_token_accuracy": 0.16053701788187028,
      "num_tokens": 9290052.0,
      "step": 4065
    },
    {
      "entropy": 5.850340270996094,
      "epoch": 0.40233293792012653,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000498958141625132,
      "loss": 5.8021,
      "mean_token_accuracy": 0.15744038969278334,
      "num_tokens": 9302201.0,
      "step": 4070
    },
    {
      "entropy": 5.914964580535889,
      "epoch": 0.4028272044286279,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004989547467115047,
      "loss": 5.8788,
      "mean_token_accuracy": 0.1513068199157715,
      "num_tokens": 9313215.0,
      "step": 4075
    },
    {
      "entropy": 5.910677099227906,
      "epoch": 0.4033214709371293,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004989513462885561,
      "loss": 5.917,
      "mean_token_accuracy": 0.145223967730999,
      "num_tokens": 9324098.0,
      "step": 4080
    },
    {
      "entropy": 5.964850568771363,
      "epoch": 0.4038157374456307,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004989479403563696,
      "loss": 5.7984,
      "mean_token_accuracy": 0.15427584648132325,
      "num_tokens": 9336048.0,
      "step": 4085
    },
    {
      "entropy": 5.874447202682495,
      "epoch": 0.4043100039541321,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004989445289150289,
      "loss": 5.7562,
      "mean_token_accuracy": 0.15428639203310013,
      "num_tokens": 9347723.0,
      "step": 4090
    },
    {
      "entropy": 5.81725811958313,
      "epoch": 0.40480427046263345,
      "grad_norm": 1.0,
      "learning_rate": 0.0004989411119646183,
      "loss": 5.802,
      "mean_token_accuracy": 0.15105294287204743,
      "num_tokens": 9360588.0,
      "step": 4095
    },
    {
      "entropy": 5.881053924560547,
      "epoch": 0.4052985369711348,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004989376895052214,
      "loss": 5.7714,
      "mean_token_accuracy": 0.1561678320169449,
      "num_tokens": 9371798.0,
      "step": 4100
    },
    {
      "entropy": 5.816941499710083,
      "epoch": 0.4057928034796362,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004989342615369226,
      "loss": 5.7775,
      "mean_token_accuracy": 0.16338589042425156,
      "num_tokens": 9382421.0,
      "step": 4105
    },
    {
      "entropy": 5.828975009918213,
      "epoch": 0.40628706998813763,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004989308280598063,
      "loss": 5.7102,
      "mean_token_accuracy": 0.16223904639482498,
      "num_tokens": 9392832.0,
      "step": 4110
    },
    {
      "entropy": 5.789451217651367,
      "epoch": 0.406781336496639,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004989273890739569,
      "loss": 5.7038,
      "mean_token_accuracy": 0.16221638768911362,
      "num_tokens": 9405309.0,
      "step": 4115
    },
    {
      "entropy": 5.795983219146729,
      "epoch": 0.4072756030051404,
      "grad_norm": 1.125,
      "learning_rate": 0.000498923944579459,
      "loss": 5.7135,
      "mean_token_accuracy": 0.16932716518640517,
      "num_tokens": 9416707.0,
      "step": 4120
    },
    {
      "entropy": 5.821607875823974,
      "epoch": 0.40776986951364175,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004989204945763972,
      "loss": 5.8453,
      "mean_token_accuracy": 0.1561842828989029,
      "num_tokens": 9427684.0,
      "step": 4125
    },
    {
      "entropy": 5.83395471572876,
      "epoch": 0.4082641360221431,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004989170390648567,
      "loss": 5.8027,
      "mean_token_accuracy": 0.15728633105754852,
      "num_tokens": 9439073.0,
      "step": 4130
    },
    {
      "entropy": 5.800190210342407,
      "epoch": 0.40875840253064455,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004989135780449222,
      "loss": 5.7427,
      "mean_token_accuracy": 0.15998244285583496,
      "num_tokens": 9450266.0,
      "step": 4135
    },
    {
      "entropy": 5.776017856597901,
      "epoch": 0.4092526690391459,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004989101115166789,
      "loss": 5.706,
      "mean_token_accuracy": 0.16565122753381728,
      "num_tokens": 9462662.0,
      "step": 4140
    },
    {
      "entropy": 5.86929087638855,
      "epoch": 0.4097469355476473,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004989066394802122,
      "loss": 5.8024,
      "mean_token_accuracy": 0.15239896178245543,
      "num_tokens": 9474338.0,
      "step": 4145
    },
    {
      "entropy": 5.845749139785767,
      "epoch": 0.41024120205614867,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004989031619356074,
      "loss": 5.7838,
      "mean_token_accuracy": 0.16020090132951736,
      "num_tokens": 9486353.0,
      "step": 4150
    },
    {
      "entropy": 5.8469836711883545,
      "epoch": 0.41073546856465004,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00049889967888295,
      "loss": 5.7535,
      "mean_token_accuracy": 0.1544842854142189,
      "num_tokens": 9497464.0,
      "step": 4155
    },
    {
      "entropy": 5.81313943862915,
      "epoch": 0.41122973507315147,
      "grad_norm": 1.125,
      "learning_rate": 0.0004988961903223258,
      "loss": 5.7288,
      "mean_token_accuracy": 0.1628125235438347,
      "num_tokens": 9508037.0,
      "step": 4160
    },
    {
      "entropy": 5.773452091217041,
      "epoch": 0.41172400158165284,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004988926962538207,
      "loss": 5.7676,
      "mean_token_accuracy": 0.15996645092964173,
      "num_tokens": 9518653.0,
      "step": 4165
    },
    {
      "entropy": 5.854199171066284,
      "epoch": 0.4122182680901542,
      "grad_norm": 1.25,
      "learning_rate": 0.0004988891966775204,
      "loss": 5.7609,
      "mean_token_accuracy": 0.16049931049346924,
      "num_tokens": 9529225.0,
      "step": 4170
    },
    {
      "entropy": 5.8694295406341555,
      "epoch": 0.4127125345986556,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004988856915935112,
      "loss": 5.7343,
      "mean_token_accuracy": 0.15607544779777527,
      "num_tokens": 9541010.0,
      "step": 4175
    },
    {
      "entropy": 5.747597646713257,
      "epoch": 0.41320680110715696,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004988821810018791,
      "loss": 5.8116,
      "mean_token_accuracy": 0.15240168645977975,
      "num_tokens": 9552061.0,
      "step": 4180
    },
    {
      "entropy": 5.953191566467285,
      "epoch": 0.41370106761565834,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004988786649027108,
      "loss": 5.9074,
      "mean_token_accuracy": 0.14590952843427657,
      "num_tokens": 9563970.0,
      "step": 4185
    },
    {
      "entropy": 5.872267389297486,
      "epoch": 0.41419533412415976,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004988751432960926,
      "loss": 5.6962,
      "mean_token_accuracy": 0.16297404766082763,
      "num_tokens": 9575381.0,
      "step": 4190
    },
    {
      "entropy": 5.843984842300415,
      "epoch": 0.41468960063266114,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004988716161821111,
      "loss": 5.7267,
      "mean_token_accuracy": 0.16292373985052108,
      "num_tokens": 9586821.0,
      "step": 4195
    },
    {
      "entropy": 5.843246555328369,
      "epoch": 0.4151838671411625,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004988680835608531,
      "loss": 5.8082,
      "mean_token_accuracy": 0.15058472603559495,
      "num_tokens": 9599580.0,
      "step": 4200
    },
    {
      "entropy": 5.7772629737854,
      "epoch": 0.4156781336496639,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004988645454324055,
      "loss": 5.8068,
      "mean_token_accuracy": 0.15320300832390785,
      "num_tokens": 9611495.0,
      "step": 4205
    },
    {
      "entropy": 5.92116961479187,
      "epoch": 0.41617240015816526,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004988610017968554,
      "loss": 5.7728,
      "mean_token_accuracy": 0.15595485121011735,
      "num_tokens": 9623713.0,
      "step": 4210
    },
    {
      "entropy": 5.768342113494873,
      "epoch": 0.4166666666666667,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004988574526542899,
      "loss": 5.7569,
      "mean_token_accuracy": 0.1648036479949951,
      "num_tokens": 9633922.0,
      "step": 4215
    },
    {
      "entropy": 5.831979560852051,
      "epoch": 0.41716093317516806,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004988538980047963,
      "loss": 5.8424,
      "mean_token_accuracy": 0.15677017271518706,
      "num_tokens": 9645641.0,
      "step": 4220
    },
    {
      "entropy": 5.851514863967895,
      "epoch": 0.41765519968366943,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004988503378484622,
      "loss": 5.7467,
      "mean_token_accuracy": 0.16601389199495314,
      "num_tokens": 9656927.0,
      "step": 4225
    },
    {
      "entropy": 5.79203519821167,
      "epoch": 0.4181494661921708,
      "grad_norm": 1.25,
      "learning_rate": 0.000498846772185375,
      "loss": 5.7172,
      "mean_token_accuracy": 0.1635448768734932,
      "num_tokens": 9669148.0,
      "step": 4230
    },
    {
      "entropy": 5.891840267181396,
      "epoch": 0.4186437327006722,
      "grad_norm": 1.125,
      "learning_rate": 0.0004988432010156224,
      "loss": 5.838,
      "mean_token_accuracy": 0.1492895156145096,
      "num_tokens": 9680620.0,
      "step": 4235
    },
    {
      "entropy": 5.9005063533782955,
      "epoch": 0.4191379992091736,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004988396243392924,
      "loss": 5.7901,
      "mean_token_accuracy": 0.15724762678146362,
      "num_tokens": 9690984.0,
      "step": 4240
    },
    {
      "entropy": 5.832825088500977,
      "epoch": 0.419632265717675,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000498836042156473,
      "loss": 5.768,
      "mean_token_accuracy": 0.16585759073495865,
      "num_tokens": 9701561.0,
      "step": 4245
    },
    {
      "entropy": 5.790317058563232,
      "epoch": 0.42012653222617635,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004988324544672521,
      "loss": 5.7601,
      "mean_token_accuracy": 0.1544044718146324,
      "num_tokens": 9712780.0,
      "step": 4250
    },
    {
      "entropy": 5.848903608322144,
      "epoch": 0.4206207987346777,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004988288612717182,
      "loss": 5.7547,
      "mean_token_accuracy": 0.1562611922621727,
      "num_tokens": 9724547.0,
      "step": 4255
    },
    {
      "entropy": 5.8364794731140135,
      "epoch": 0.4211150652431791,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004988252625699594,
      "loss": 5.8111,
      "mean_token_accuracy": 0.15629005581140518,
      "num_tokens": 9736897.0,
      "step": 4260
    },
    {
      "entropy": 5.8446488857269285,
      "epoch": 0.4216093317516805,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004988216583620647,
      "loss": 5.7584,
      "mean_token_accuracy": 0.16204439848661423,
      "num_tokens": 9748214.0,
      "step": 4265
    },
    {
      "entropy": 5.886655187606811,
      "epoch": 0.4221035982601819,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004988180486481223,
      "loss": 5.7894,
      "mean_token_accuracy": 0.15823814868927003,
      "num_tokens": 9759602.0,
      "step": 4270
    },
    {
      "entropy": 5.827601718902588,
      "epoch": 0.4225978647686833,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004988144334282211,
      "loss": 5.7425,
      "mean_token_accuracy": 0.1589537337422371,
      "num_tokens": 9771645.0,
      "step": 4275
    },
    {
      "entropy": 5.853240966796875,
      "epoch": 0.42309213127718465,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004988108127024503,
      "loss": 5.7505,
      "mean_token_accuracy": 0.1587558016180992,
      "num_tokens": 9783504.0,
      "step": 4280
    },
    {
      "entropy": 5.891699171066284,
      "epoch": 0.423586397785686,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004988071864708987,
      "loss": 5.8804,
      "mean_token_accuracy": 0.15425789207220078,
      "num_tokens": 9795572.0,
      "step": 4285
    },
    {
      "entropy": 5.83182110786438,
      "epoch": 0.42408066429418745,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004988035547336555,
      "loss": 5.7336,
      "mean_token_accuracy": 0.15293268710374833,
      "num_tokens": 9806398.0,
      "step": 4290
    },
    {
      "entropy": 5.82321310043335,
      "epoch": 0.4245749308026888,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004987999174908104,
      "loss": 5.7342,
      "mean_token_accuracy": 0.1546010985970497,
      "num_tokens": 9817171.0,
      "step": 4295
    },
    {
      "entropy": 5.818284416198731,
      "epoch": 0.4250691973111902,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004987962747424524,
      "loss": 5.7763,
      "mean_token_accuracy": 0.15528347045183183,
      "num_tokens": 9827535.0,
      "step": 4300
    },
    {
      "entropy": 5.79456467628479,
      "epoch": 0.42556346381969157,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004987926264886713,
      "loss": 5.7551,
      "mean_token_accuracy": 0.1570479989051819,
      "num_tokens": 9839619.0,
      "step": 4305
    },
    {
      "entropy": 5.840618705749511,
      "epoch": 0.42605773032819294,
      "grad_norm": 1.28125,
      "learning_rate": 0.000498788972729557,
      "loss": 5.7544,
      "mean_token_accuracy": 0.15230125635862352,
      "num_tokens": 9850100.0,
      "step": 4310
    },
    {
      "entropy": 5.917747354507446,
      "epoch": 0.42655199683669437,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004987853134651992,
      "loss": 5.8327,
      "mean_token_accuracy": 0.14419816955924034,
      "num_tokens": 9862228.0,
      "step": 4315
    },
    {
      "entropy": 5.803866004943847,
      "epoch": 0.42704626334519574,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000498781648695688,
      "loss": 5.7229,
      "mean_token_accuracy": 0.1623837560415268,
      "num_tokens": 9873080.0,
      "step": 4320
    },
    {
      "entropy": 5.791725206375122,
      "epoch": 0.4275405298536971,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004987779784211135,
      "loss": 5.7442,
      "mean_token_accuracy": 0.15227282643318177,
      "num_tokens": 9884735.0,
      "step": 4325
    },
    {
      "entropy": 5.879214715957642,
      "epoch": 0.4280347963621985,
      "grad_norm": 1.234375,
      "learning_rate": 0.000498774302641566,
      "loss": 5.7641,
      "mean_token_accuracy": 0.15441785752773285,
      "num_tokens": 9895018.0,
      "step": 4330
    },
    {
      "entropy": 5.850545740127563,
      "epoch": 0.42852906287069986,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004987706213571359,
      "loss": 5.7371,
      "mean_token_accuracy": 0.1566600888967514,
      "num_tokens": 9906838.0,
      "step": 4335
    },
    {
      "entropy": 5.844648790359497,
      "epoch": 0.4290233293792013,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004987669345679139,
      "loss": 5.7701,
      "mean_token_accuracy": 0.1565292850136757,
      "num_tokens": 9919556.0,
      "step": 4340
    },
    {
      "entropy": 5.781128025054931,
      "epoch": 0.42951759588770266,
      "grad_norm": 1.25,
      "learning_rate": 0.0004987632422739904,
      "loss": 5.6781,
      "mean_token_accuracy": 0.15627526938915254,
      "num_tokens": 9929783.0,
      "step": 4345
    },
    {
      "entropy": 5.772721195220948,
      "epoch": 0.43001186239620404,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004987595444754567,
      "loss": 5.7391,
      "mean_token_accuracy": 0.15912755578756332,
      "num_tokens": 9940894.0,
      "step": 4350
    },
    {
      "entropy": 5.80626540184021,
      "epoch": 0.4305061289047054,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004987558411724032,
      "loss": 5.718,
      "mean_token_accuracy": 0.16627187728881837,
      "num_tokens": 9951973.0,
      "step": 4355
    },
    {
      "entropy": 5.9029295444488525,
      "epoch": 0.4310003954132068,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004987521323649215,
      "loss": 5.7694,
      "mean_token_accuracy": 0.15462518632411956,
      "num_tokens": 9963304.0,
      "step": 4360
    },
    {
      "entropy": 5.838963317871094,
      "epoch": 0.4314946619217082,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004987484180531026,
      "loss": 5.7812,
      "mean_token_accuracy": 0.16214534044265747,
      "num_tokens": 9973981.0,
      "step": 4365
    },
    {
      "entropy": 5.8302850246429445,
      "epoch": 0.4319889284302096,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004987446982370379,
      "loss": 5.8036,
      "mean_token_accuracy": 0.1544625535607338,
      "num_tokens": 9986461.0,
      "step": 4370
    },
    {
      "entropy": 5.8808770179748535,
      "epoch": 0.43248319493871096,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000498740972916819,
      "loss": 5.7715,
      "mean_token_accuracy": 0.15348488688468934,
      "num_tokens": 9997973.0,
      "step": 4375
    },
    {
      "entropy": 5.828121280670166,
      "epoch": 0.43297746144721233,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004987372420925373,
      "loss": 5.78,
      "mean_token_accuracy": 0.15659529715776443,
      "num_tokens": 10009572.0,
      "step": 4380
    },
    {
      "entropy": 5.816518115997314,
      "epoch": 0.4334717279557137,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004987335057642847,
      "loss": 5.6502,
      "mean_token_accuracy": 0.1678347870707512,
      "num_tokens": 10020437.0,
      "step": 4385
    },
    {
      "entropy": 5.809616422653198,
      "epoch": 0.43396599446421513,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004987297639321533,
      "loss": 5.7808,
      "mean_token_accuracy": 0.15647988617420197,
      "num_tokens": 10031827.0,
      "step": 4390
    },
    {
      "entropy": 5.821065092086792,
      "epoch": 0.4344602609727165,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004987260165962349,
      "loss": 5.7813,
      "mean_token_accuracy": 0.15543572455644608,
      "num_tokens": 10043833.0,
      "step": 4395
    },
    {
      "entropy": 5.8089728355407715,
      "epoch": 0.4349545274812179,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004987222637566218,
      "loss": 5.7225,
      "mean_token_accuracy": 0.1572750225663185,
      "num_tokens": 10054645.0,
      "step": 4400
    },
    {
      "entropy": 5.7818284034729,
      "epoch": 0.43544879398971925,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004987185054134064,
      "loss": 5.6724,
      "mean_token_accuracy": 0.1580430954694748,
      "num_tokens": 10065178.0,
      "step": 4405
    },
    {
      "entropy": 5.825777530670166,
      "epoch": 0.4359430604982206,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000498714741566681,
      "loss": 5.706,
      "mean_token_accuracy": 0.16232621520757676,
      "num_tokens": 10076701.0,
      "step": 4410
    },
    {
      "entropy": 5.735870885848999,
      "epoch": 0.436437327006722,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004987109722165383,
      "loss": 5.7263,
      "mean_token_accuracy": 0.15729656368494033,
      "num_tokens": 10087959.0,
      "step": 4415
    },
    {
      "entropy": 5.88615665435791,
      "epoch": 0.4369315935152234,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004987071973630708,
      "loss": 5.7156,
      "mean_token_accuracy": 0.15740345194935798,
      "num_tokens": 10099463.0,
      "step": 4420
    },
    {
      "entropy": 5.822025537490845,
      "epoch": 0.4374258600237248,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004987034170063717,
      "loss": 5.8005,
      "mean_token_accuracy": 0.15701815336942673,
      "num_tokens": 10111384.0,
      "step": 4425
    },
    {
      "entropy": 5.821445465087891,
      "epoch": 0.4379201265322262,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004986996311465338,
      "loss": 5.7321,
      "mean_token_accuracy": 0.15710116550326347,
      "num_tokens": 10121838.0,
      "step": 4430
    },
    {
      "entropy": 5.8637402057647705,
      "epoch": 0.43841439304072755,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004986958397836503,
      "loss": 5.6928,
      "mean_token_accuracy": 0.1640600949525833,
      "num_tokens": 10132225.0,
      "step": 4435
    },
    {
      "entropy": 5.729298543930054,
      "epoch": 0.4389086595492289,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004986920429178144,
      "loss": 5.6631,
      "mean_token_accuracy": 0.15749302357435227,
      "num_tokens": 10143114.0,
      "step": 4440
    },
    {
      "entropy": 5.893101644515991,
      "epoch": 0.43940292605773035,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004986882405491195,
      "loss": 5.8626,
      "mean_token_accuracy": 0.15157403349876403,
      "num_tokens": 10155057.0,
      "step": 4445
    },
    {
      "entropy": 5.75573582649231,
      "epoch": 0.4398971925662317,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004986844326776593,
      "loss": 5.6229,
      "mean_token_accuracy": 0.1684441089630127,
      "num_tokens": 10166816.0,
      "step": 4450
    },
    {
      "entropy": 5.734735536575317,
      "epoch": 0.4403914590747331,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004986806193035272,
      "loss": 5.695,
      "mean_token_accuracy": 0.1676469177007675,
      "num_tokens": 10179433.0,
      "step": 4455
    },
    {
      "entropy": 5.8267802715301515,
      "epoch": 0.44088572558323447,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004986768004268173,
      "loss": 5.693,
      "mean_token_accuracy": 0.16008566021919252,
      "num_tokens": 10190784.0,
      "step": 4460
    },
    {
      "entropy": 5.836114597320557,
      "epoch": 0.44137999209173584,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004986729760476234,
      "loss": 5.7523,
      "mean_token_accuracy": 0.15255020558834076,
      "num_tokens": 10201941.0,
      "step": 4465
    },
    {
      "entropy": 5.808204221725464,
      "epoch": 0.44187425860023727,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004986691461660394,
      "loss": 5.7176,
      "mean_token_accuracy": 0.16116027235984803,
      "num_tokens": 10214029.0,
      "step": 4470
    },
    {
      "entropy": 5.776810598373413,
      "epoch": 0.44236852510873864,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004986653107821599,
      "loss": 5.7602,
      "mean_token_accuracy": 0.16111526936292647,
      "num_tokens": 10225585.0,
      "step": 4475
    },
    {
      "entropy": 5.78863844871521,
      "epoch": 0.44286279161724,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004986614698960789,
      "loss": 5.7031,
      "mean_token_accuracy": 0.16227239221334458,
      "num_tokens": 10237753.0,
      "step": 4480
    },
    {
      "entropy": 5.912133312225341,
      "epoch": 0.4433570581257414,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004986576235078913,
      "loss": 5.9483,
      "mean_token_accuracy": 0.1438101887702942,
      "num_tokens": 10251522.0,
      "step": 4485
    },
    {
      "entropy": 5.831998538970947,
      "epoch": 0.44385132463424276,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004986537716176912,
      "loss": 5.6578,
      "mean_token_accuracy": 0.16271060109138488,
      "num_tokens": 10263319.0,
      "step": 4490
    },
    {
      "entropy": 5.715137338638305,
      "epoch": 0.4443455911427442,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004986499142255738,
      "loss": 5.7045,
      "mean_token_accuracy": 0.15843060836195946,
      "num_tokens": 10274899.0,
      "step": 4495
    },
    {
      "entropy": 5.836452913284302,
      "epoch": 0.44483985765124556,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004986460513316338,
      "loss": 5.676,
      "mean_token_accuracy": 0.16559335291385652,
      "num_tokens": 10286110.0,
      "step": 4500
    },
    {
      "entropy": 5.78662633895874,
      "epoch": 0.44533412415974694,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004986421829359662,
      "loss": 5.8238,
      "mean_token_accuracy": 0.15288708209991456,
      "num_tokens": 10298870.0,
      "step": 4505
    },
    {
      "entropy": 5.843674945831299,
      "epoch": 0.4458283906682483,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004986383090386661,
      "loss": 5.744,
      "mean_token_accuracy": 0.15693433284759523,
      "num_tokens": 10310326.0,
      "step": 4510
    },
    {
      "entropy": 5.83041410446167,
      "epoch": 0.4463226571767497,
      "grad_norm": 1.265625,
      "learning_rate": 0.000498634429639829,
      "loss": 5.6721,
      "mean_token_accuracy": 0.15878720134496688,
      "num_tokens": 10320899.0,
      "step": 4515
    },
    {
      "entropy": 5.792200517654419,
      "epoch": 0.4468169236852511,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004986305447395502,
      "loss": 5.7275,
      "mean_token_accuracy": 0.1623765215277672,
      "num_tokens": 10331669.0,
      "step": 4520
    },
    {
      "entropy": 5.829644346237183,
      "epoch": 0.4473111901937525,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004986266543379254,
      "loss": 5.7114,
      "mean_token_accuracy": 0.1589057058095932,
      "num_tokens": 10343120.0,
      "step": 4525
    },
    {
      "entropy": 5.772820949554443,
      "epoch": 0.44780545670225386,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004986227584350501,
      "loss": 5.7149,
      "mean_token_accuracy": 0.1589616447687149,
      "num_tokens": 10354806.0,
      "step": 4530
    },
    {
      "entropy": 5.848662900924682,
      "epoch": 0.44829972321075523,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004986188570310203,
      "loss": 5.7359,
      "mean_token_accuracy": 0.1575729265809059,
      "num_tokens": 10366453.0,
      "step": 4535
    },
    {
      "entropy": 5.86790361404419,
      "epoch": 0.4487939897192566,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004986149501259319,
      "loss": 5.8281,
      "mean_token_accuracy": 0.15670044720172882,
      "num_tokens": 10378095.0,
      "step": 4540
    },
    {
      "entropy": 5.732289028167725,
      "epoch": 0.44928825622775803,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004986110377198812,
      "loss": 5.6931,
      "mean_token_accuracy": 0.16421767175197602,
      "num_tokens": 10389620.0,
      "step": 4545
    },
    {
      "entropy": 5.868263101577758,
      "epoch": 0.4497825227362594,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000498607119812964,
      "loss": 5.8237,
      "mean_token_accuracy": 0.15202950835227966,
      "num_tokens": 10400166.0,
      "step": 4550
    },
    {
      "entropy": 5.809820365905762,
      "epoch": 0.4502767892447608,
      "grad_norm": 1.25,
      "learning_rate": 0.0004986031964052772,
      "loss": 5.7247,
      "mean_token_accuracy": 0.15717918127775193,
      "num_tokens": 10411116.0,
      "step": 4555
    },
    {
      "entropy": 5.741772174835205,
      "epoch": 0.45077105575326215,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004985992674969169,
      "loss": 5.7477,
      "mean_token_accuracy": 0.15264249593019485,
      "num_tokens": 10423484.0,
      "step": 4560
    },
    {
      "entropy": 5.902524757385254,
      "epoch": 0.4512653222617635,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004985953330879801,
      "loss": 5.813,
      "mean_token_accuracy": 0.14781858772039413,
      "num_tokens": 10435636.0,
      "step": 4565
    },
    {
      "entropy": 5.847655820846557,
      "epoch": 0.45175958877026495,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004985913931785633,
      "loss": 5.7489,
      "mean_token_accuracy": 0.1599886104464531,
      "num_tokens": 10447586.0,
      "step": 4570
    },
    {
      "entropy": 5.782388496398926,
      "epoch": 0.4522538552787663,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004985874477687636,
      "loss": 5.8058,
      "mean_token_accuracy": 0.1555984526872635,
      "num_tokens": 10460678.0,
      "step": 4575
    },
    {
      "entropy": 5.818998098373413,
      "epoch": 0.4527481217872677,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004985834968586779,
      "loss": 5.7685,
      "mean_token_accuracy": 0.1549448549747467,
      "num_tokens": 10472738.0,
      "step": 4580
    },
    {
      "entropy": 5.85989031791687,
      "epoch": 0.45324238829576907,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004985795404484037,
      "loss": 5.6418,
      "mean_token_accuracy": 0.16428616791963577,
      "num_tokens": 10483716.0,
      "step": 4585
    },
    {
      "entropy": 5.747600412368774,
      "epoch": 0.45373665480427045,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004985755785380379,
      "loss": 5.7724,
      "mean_token_accuracy": 0.15435491055250167,
      "num_tokens": 10495287.0,
      "step": 4590
    },
    {
      "entropy": 5.795434284210205,
      "epoch": 0.4542309213127719,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004985716111276781,
      "loss": 5.7235,
      "mean_token_accuracy": 0.16080528646707534,
      "num_tokens": 10506332.0,
      "step": 4595
    },
    {
      "entropy": 5.858145904541016,
      "epoch": 0.45472518782127325,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004985676382174223,
      "loss": 5.7538,
      "mean_token_accuracy": 0.15550379902124406,
      "num_tokens": 10517574.0,
      "step": 4600
    },
    {
      "entropy": 5.77792444229126,
      "epoch": 0.4552194543297746,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004985636598073677,
      "loss": 5.6777,
      "mean_token_accuracy": 0.1682252198457718,
      "num_tokens": 10528998.0,
      "step": 4605
    },
    {
      "entropy": 5.829964351654053,
      "epoch": 0.455713720838276,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004985596758976125,
      "loss": 5.7456,
      "mean_token_accuracy": 0.15128750652074813,
      "num_tokens": 10539669.0,
      "step": 4610
    },
    {
      "entropy": 5.807875728607177,
      "epoch": 0.45620798734677737,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004985556864882545,
      "loss": 5.7253,
      "mean_token_accuracy": 0.16209764927625656,
      "num_tokens": 10550005.0,
      "step": 4615
    },
    {
      "entropy": 5.745830821990967,
      "epoch": 0.45670225385527874,
      "grad_norm": 1.171875,
      "learning_rate": 0.000498551691579392,
      "loss": 5.6994,
      "mean_token_accuracy": 0.16218698173761367,
      "num_tokens": 10561122.0,
      "step": 4620
    },
    {
      "entropy": 5.817157936096192,
      "epoch": 0.45719652036378017,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004985476911711232,
      "loss": 5.6775,
      "mean_token_accuracy": 0.15966929942369462,
      "num_tokens": 10572081.0,
      "step": 4625
    },
    {
      "entropy": 5.656571245193481,
      "epoch": 0.45769078687228154,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004985436852635465,
      "loss": 5.6708,
      "mean_token_accuracy": 0.1661362901329994,
      "num_tokens": 10582297.0,
      "step": 4630
    },
    {
      "entropy": 5.8948314666748045,
      "epoch": 0.4581850533807829,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004985396738567604,
      "loss": 5.7712,
      "mean_token_accuracy": 0.15315597653388976,
      "num_tokens": 10593161.0,
      "step": 4635
    },
    {
      "entropy": 5.829736757278442,
      "epoch": 0.4586793198892843,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004985356569508638,
      "loss": 5.7647,
      "mean_token_accuracy": 0.1625633120536804,
      "num_tokens": 10603900.0,
      "step": 4640
    },
    {
      "entropy": 5.788417482376099,
      "epoch": 0.45917358639778566,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004985316345459551,
      "loss": 5.7583,
      "mean_token_accuracy": 0.16019802242517472,
      "num_tokens": 10614946.0,
      "step": 4645
    },
    {
      "entropy": 5.819199132919311,
      "epoch": 0.4596678529062871,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004985276066421338,
      "loss": 5.7609,
      "mean_token_accuracy": 0.1518936887383461,
      "num_tokens": 10625073.0,
      "step": 4650
    },
    {
      "entropy": 5.856818628311157,
      "epoch": 0.46016211941478846,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004985235732394986,
      "loss": 5.8623,
      "mean_token_accuracy": 0.15111786425113677,
      "num_tokens": 10636942.0,
      "step": 4655
    },
    {
      "entropy": 5.893058109283447,
      "epoch": 0.46065638592328984,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004985195343381487,
      "loss": 5.7983,
      "mean_token_accuracy": 0.1530800074338913,
      "num_tokens": 10647608.0,
      "step": 4660
    },
    {
      "entropy": 5.854513692855835,
      "epoch": 0.4611506524317912,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004985154899381837,
      "loss": 5.8287,
      "mean_token_accuracy": 0.15122355073690413,
      "num_tokens": 10659532.0,
      "step": 4665
    },
    {
      "entropy": 5.853507041931152,
      "epoch": 0.4616449189402926,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004985114400397029,
      "loss": 5.7571,
      "mean_token_accuracy": 0.16026072353124618,
      "num_tokens": 10670342.0,
      "step": 4670
    },
    {
      "entropy": 5.8415283203125,
      "epoch": 0.462139185448794,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004985073846428061,
      "loss": 5.7877,
      "mean_token_accuracy": 0.15555054545402527,
      "num_tokens": 10682984.0,
      "step": 4675
    },
    {
      "entropy": 5.841789770126343,
      "epoch": 0.4626334519572954,
      "grad_norm": 1.125,
      "learning_rate": 0.0004985033237475929,
      "loss": 5.773,
      "mean_token_accuracy": 0.15423207432031633,
      "num_tokens": 10694235.0,
      "step": 4680
    },
    {
      "entropy": 5.926882314682007,
      "epoch": 0.46312771846579676,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004984992573541633,
      "loss": 5.7837,
      "mean_token_accuracy": 0.15645679384469985,
      "num_tokens": 10704508.0,
      "step": 4685
    },
    {
      "entropy": 5.676226234436035,
      "epoch": 0.46362198497429813,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004984951854626173,
      "loss": 5.6549,
      "mean_token_accuracy": 0.1645615130662918,
      "num_tokens": 10716927.0,
      "step": 4690
    },
    {
      "entropy": 5.752277278900147,
      "epoch": 0.4641162514827995,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004984911080730551,
      "loss": 5.6912,
      "mean_token_accuracy": 0.16231297999620437,
      "num_tokens": 10728396.0,
      "step": 4695
    },
    {
      "entropy": 5.783680486679077,
      "epoch": 0.46461051799130093,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004984870251855769,
      "loss": 5.7673,
      "mean_token_accuracy": 0.16363271474838256,
      "num_tokens": 10739001.0,
      "step": 4700
    },
    {
      "entropy": 5.851905822753906,
      "epoch": 0.4651047844998023,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004984829368002831,
      "loss": 5.7344,
      "mean_token_accuracy": 0.15897003412246705,
      "num_tokens": 10748768.0,
      "step": 4705
    },
    {
      "entropy": 5.860234689712525,
      "epoch": 0.4655990510083037,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004984788429172745,
      "loss": 5.7698,
      "mean_token_accuracy": 0.1561150312423706,
      "num_tokens": 10760582.0,
      "step": 4710
    },
    {
      "entropy": 5.760329294204712,
      "epoch": 0.46609331751680505,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004984747435366517,
      "loss": 5.6788,
      "mean_token_accuracy": 0.16664675623178482,
      "num_tokens": 10772537.0,
      "step": 4715
    },
    {
      "entropy": 5.772965812683106,
      "epoch": 0.4665875840253064,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004984706386585156,
      "loss": 5.6206,
      "mean_token_accuracy": 0.168980173766613,
      "num_tokens": 10783740.0,
      "step": 4720
    },
    {
      "entropy": 5.840044403076172,
      "epoch": 0.46708185053380785,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000498466528282967,
      "loss": 5.77,
      "mean_token_accuracy": 0.15619183331727982,
      "num_tokens": 10794793.0,
      "step": 4725
    },
    {
      "entropy": 5.840643644332886,
      "epoch": 0.4675761170423092,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004984624124101072,
      "loss": 5.7317,
      "mean_token_accuracy": 0.1579558730125427,
      "num_tokens": 10805944.0,
      "step": 4730
    },
    {
      "entropy": 5.8369344711303714,
      "epoch": 0.4680703835508106,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004984582910400374,
      "loss": 5.7556,
      "mean_token_accuracy": 0.15298434495925903,
      "num_tokens": 10818362.0,
      "step": 4735
    },
    {
      "entropy": 5.780828523635864,
      "epoch": 0.46856465005931197,
      "grad_norm": 1.109375,
      "learning_rate": 0.000498454164172859,
      "loss": 5.7502,
      "mean_token_accuracy": 0.16281691044569016,
      "num_tokens": 10831306.0,
      "step": 4740
    },
    {
      "entropy": 5.807979011535645,
      "epoch": 0.46905891656781334,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004984500318086733,
      "loss": 5.7066,
      "mean_token_accuracy": 0.16005800068378448,
      "num_tokens": 10841671.0,
      "step": 4745
    },
    {
      "entropy": 5.77095742225647,
      "epoch": 0.4695531830763148,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004984458939475824,
      "loss": 5.6916,
      "mean_token_accuracy": 0.1572381556034088,
      "num_tokens": 10852754.0,
      "step": 4750
    },
    {
      "entropy": 5.815678262710572,
      "epoch": 0.47004744958481615,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004984417505896877,
      "loss": 5.7662,
      "mean_token_accuracy": 0.156097312271595,
      "num_tokens": 10864812.0,
      "step": 4755
    },
    {
      "entropy": 5.857409429550171,
      "epoch": 0.4705417160933175,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004984376017350913,
      "loss": 5.7473,
      "mean_token_accuracy": 0.1609667122364044,
      "num_tokens": 10875967.0,
      "step": 4760
    },
    {
      "entropy": 5.79653754234314,
      "epoch": 0.4710359826018189,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004984334473838953,
      "loss": 5.7227,
      "mean_token_accuracy": 0.16150201112031937,
      "num_tokens": 10889575.0,
      "step": 4765
    },
    {
      "entropy": 5.82901701927185,
      "epoch": 0.47153024911032027,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004984292875362018,
      "loss": 5.804,
      "mean_token_accuracy": 0.1580755218863487,
      "num_tokens": 10900751.0,
      "step": 4770
    },
    {
      "entropy": 5.878472757339478,
      "epoch": 0.4720245156188217,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004984251221921131,
      "loss": 5.7473,
      "mean_token_accuracy": 0.15800192058086396,
      "num_tokens": 10912618.0,
      "step": 4775
    },
    {
      "entropy": 5.765990495681763,
      "epoch": 0.47251878212732307,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004984209513517317,
      "loss": 5.7156,
      "mean_token_accuracy": 0.15646711066365243,
      "num_tokens": 10925286.0,
      "step": 4780
    },
    {
      "entropy": 5.829292821884155,
      "epoch": 0.47301304863582444,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004984167750151603,
      "loss": 5.7242,
      "mean_token_accuracy": 0.16060231178998946,
      "num_tokens": 10936033.0,
      "step": 4785
    },
    {
      "entropy": 5.779935884475708,
      "epoch": 0.4735073151443258,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004984125931825018,
      "loss": 5.6637,
      "mean_token_accuracy": 0.16456516981124877,
      "num_tokens": 10947055.0,
      "step": 4790
    },
    {
      "entropy": 5.793975830078125,
      "epoch": 0.4740015816528272,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004984084058538585,
      "loss": 5.7137,
      "mean_token_accuracy": 0.1572008863091469,
      "num_tokens": 10957468.0,
      "step": 4795
    },
    {
      "entropy": 5.861984443664551,
      "epoch": 0.4744958481613286,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004984042130293339,
      "loss": 5.8362,
      "mean_token_accuracy": 0.1582918718457222,
      "num_tokens": 10970273.0,
      "step": 4800
    },
    {
      "entropy": 5.86493616104126,
      "epoch": 0.47499011466983,
      "grad_norm": 1.375,
      "learning_rate": 0.0004984000147090311,
      "loss": 5.8211,
      "mean_token_accuracy": 0.1545470640063286,
      "num_tokens": 10981842.0,
      "step": 4805
    },
    {
      "entropy": 5.888898468017578,
      "epoch": 0.47548438117833136,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004983958108930531,
      "loss": 5.7169,
      "mean_token_accuracy": 0.15943304374814032,
      "num_tokens": 10993615.0,
      "step": 4810
    },
    {
      "entropy": 5.83760347366333,
      "epoch": 0.47597864768683273,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004983916015815036,
      "loss": 5.8064,
      "mean_token_accuracy": 0.15529759526252745,
      "num_tokens": 11004451.0,
      "step": 4815
    },
    {
      "entropy": 5.901088237762451,
      "epoch": 0.4764729141953341,
      "grad_norm": 1.53125,
      "learning_rate": 0.000498387386774486,
      "loss": 5.745,
      "mean_token_accuracy": 0.15757741779088974,
      "num_tokens": 11014834.0,
      "step": 4820
    },
    {
      "entropy": 5.731997919082642,
      "epoch": 0.4769671807038355,
      "grad_norm": 1.375,
      "learning_rate": 0.000498383166472104,
      "loss": 5.6858,
      "mean_token_accuracy": 0.15810937881469728,
      "num_tokens": 11026116.0,
      "step": 4825
    },
    {
      "entropy": 5.807245063781738,
      "epoch": 0.4774614472123369,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004983789406744615,
      "loss": 5.7472,
      "mean_token_accuracy": 0.15772093832492828,
      "num_tokens": 11038780.0,
      "step": 4830
    },
    {
      "entropy": 5.878645515441894,
      "epoch": 0.4779557137208383,
      "grad_norm": 1.375,
      "learning_rate": 0.0004983747093816625,
      "loss": 5.7909,
      "mean_token_accuracy": 0.1571904167532921,
      "num_tokens": 11049824.0,
      "step": 4835
    },
    {
      "entropy": 5.7510171890258786,
      "epoch": 0.47844998022933966,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004983704725938108,
      "loss": 5.6787,
      "mean_token_accuracy": 0.1669359490275383,
      "num_tokens": 11060863.0,
      "step": 4840
    },
    {
      "entropy": 5.7988502979278564,
      "epoch": 0.47894424673784103,
      "grad_norm": 1.28125,
      "learning_rate": 0.000498366230311011,
      "loss": 5.6829,
      "mean_token_accuracy": 0.1632984146475792,
      "num_tokens": 11072533.0,
      "step": 4845
    },
    {
      "entropy": 5.72292914390564,
      "epoch": 0.4794385132463424,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004983619825333671,
      "loss": 5.6217,
      "mean_token_accuracy": 0.1602368801832199,
      "num_tokens": 11083627.0,
      "step": 4850
    },
    {
      "entropy": 5.8456309795379635,
      "epoch": 0.47993277975484383,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004983577292609839,
      "loss": 5.8346,
      "mean_token_accuracy": 0.14952256679534912,
      "num_tokens": 11096421.0,
      "step": 4855
    },
    {
      "entropy": 5.796829748153686,
      "epoch": 0.4804270462633452,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004983534704939659,
      "loss": 5.6678,
      "mean_token_accuracy": 0.15885301381349565,
      "num_tokens": 11109053.0,
      "step": 4860
    },
    {
      "entropy": 5.730153894424438,
      "epoch": 0.4809213127718466,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004983492062324179,
      "loss": 5.6214,
      "mean_token_accuracy": 0.16144478619098662,
      "num_tokens": 11120431.0,
      "step": 4865
    },
    {
      "entropy": 5.791682338714599,
      "epoch": 0.48141557928034795,
      "grad_norm": 1.375,
      "learning_rate": 0.0004983449364764447,
      "loss": 5.7241,
      "mean_token_accuracy": 0.1602831542491913,
      "num_tokens": 11132767.0,
      "step": 4870
    },
    {
      "entropy": 5.746629905700684,
      "epoch": 0.4819098457888493,
      "grad_norm": 1.375,
      "learning_rate": 0.0004983406612261514,
      "loss": 5.7184,
      "mean_token_accuracy": 0.15982515960931779,
      "num_tokens": 11144508.0,
      "step": 4875
    },
    {
      "entropy": 5.765431594848633,
      "epoch": 0.48240411229735075,
      "grad_norm": 1.25,
      "learning_rate": 0.0004983363804816433,
      "loss": 5.7026,
      "mean_token_accuracy": 0.16171906739473343,
      "num_tokens": 11156905.0,
      "step": 4880
    },
    {
      "entropy": 5.812859296798706,
      "epoch": 0.4828983788058521,
      "grad_norm": 1.25,
      "learning_rate": 0.0004983320942430256,
      "loss": 5.6916,
      "mean_token_accuracy": 0.16429043263196946,
      "num_tokens": 11169383.0,
      "step": 4885
    },
    {
      "entropy": 5.784932374954224,
      "epoch": 0.4833926453143535,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004983278025104036,
      "loss": 5.6305,
      "mean_token_accuracy": 0.17158691734075546,
      "num_tokens": 11179795.0,
      "step": 4890
    },
    {
      "entropy": 5.760914993286133,
      "epoch": 0.48388691182285487,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000498323505283883,
      "loss": 5.7569,
      "mean_token_accuracy": 0.15697767585515976,
      "num_tokens": 11190783.0,
      "step": 4895
    },
    {
      "entropy": 5.82587571144104,
      "epoch": 0.48438117833135624,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004983192025635698,
      "loss": 5.6969,
      "mean_token_accuracy": 0.15884145647287368,
      "num_tokens": 11203000.0,
      "step": 4900
    },
    {
      "entropy": 5.802449417114258,
      "epoch": 0.48487544483985767,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004983148943495693,
      "loss": 5.7132,
      "mean_token_accuracy": 0.1578467845916748,
      "num_tokens": 11213721.0,
      "step": 4905
    },
    {
      "entropy": 5.8142822742462155,
      "epoch": 0.48536971134835905,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004983105806419878,
      "loss": 5.7944,
      "mean_token_accuracy": 0.1531928800046444,
      "num_tokens": 11224490.0,
      "step": 4910
    },
    {
      "entropy": 5.801441812515259,
      "epoch": 0.4858639778568604,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004983062614409314,
      "loss": 5.6924,
      "mean_token_accuracy": 0.16117836385965348,
      "num_tokens": 11235022.0,
      "step": 4915
    },
    {
      "entropy": 5.850840520858765,
      "epoch": 0.4863582443653618,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004983019367465064,
      "loss": 5.7245,
      "mean_token_accuracy": 0.15540830790996552,
      "num_tokens": 11247485.0,
      "step": 4920
    },
    {
      "entropy": 5.75506625175476,
      "epoch": 0.48685251087386316,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000498297606558819,
      "loss": 5.6655,
      "mean_token_accuracy": 0.1630787342786789,
      "num_tokens": 11258488.0,
      "step": 4925
    },
    {
      "entropy": 5.736137628555298,
      "epoch": 0.4873467773823646,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004982932708779759,
      "loss": 5.7167,
      "mean_token_accuracy": 0.16207361966371536,
      "num_tokens": 11270786.0,
      "step": 4930
    },
    {
      "entropy": 5.90667872428894,
      "epoch": 0.48784104389086597,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004982889297040836,
      "loss": 5.7079,
      "mean_token_accuracy": 0.15878405645489693,
      "num_tokens": 11282434.0,
      "step": 4935
    },
    {
      "entropy": 5.791211462020874,
      "epoch": 0.48833531039936734,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004982845830372491,
      "loss": 5.694,
      "mean_token_accuracy": 0.1584634080529213,
      "num_tokens": 11292406.0,
      "step": 4940
    },
    {
      "entropy": 5.844303798675537,
      "epoch": 0.4888295769078687,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004982802308775791,
      "loss": 5.83,
      "mean_token_accuracy": 0.1521375820040703,
      "num_tokens": 11304491.0,
      "step": 4945
    },
    {
      "entropy": 5.850516319274902,
      "epoch": 0.4893238434163701,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000498275873225181,
      "loss": 5.7576,
      "mean_token_accuracy": 0.159835883975029,
      "num_tokens": 11317038.0,
      "step": 4950
    },
    {
      "entropy": 5.819036674499512,
      "epoch": 0.4898181099248715,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004982715100801616,
      "loss": 5.7688,
      "mean_token_accuracy": 0.1559791311621666,
      "num_tokens": 11329354.0,
      "step": 4955
    },
    {
      "entropy": 5.789994812011718,
      "epoch": 0.4903123764333729,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004982671414426285,
      "loss": 5.735,
      "mean_token_accuracy": 0.15645507723093033,
      "num_tokens": 11340508.0,
      "step": 4960
    },
    {
      "entropy": 5.792756795883179,
      "epoch": 0.49080664294187426,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004982627673126889,
      "loss": 5.7224,
      "mean_token_accuracy": 0.16280135810375213,
      "num_tokens": 11352608.0,
      "step": 4965
    },
    {
      "entropy": 5.749039602279663,
      "epoch": 0.49130090945037563,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004982583876904508,
      "loss": 5.6953,
      "mean_token_accuracy": 0.16312929540872573,
      "num_tokens": 11364089.0,
      "step": 4970
    },
    {
      "entropy": 5.7701911449432375,
      "epoch": 0.491795175958877,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004982540025760217,
      "loss": 5.6459,
      "mean_token_accuracy": 0.16562199592590332,
      "num_tokens": 11375521.0,
      "step": 4975
    },
    {
      "entropy": 5.782875490188599,
      "epoch": 0.49228944246737844,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004982496119695094,
      "loss": 5.7708,
      "mean_token_accuracy": 0.1639150306582451,
      "num_tokens": 11387721.0,
      "step": 4980
    },
    {
      "entropy": 5.809487771987915,
      "epoch": 0.4927837089758798,
      "grad_norm": 1.453125,
      "learning_rate": 0.000498245215871022,
      "loss": 5.7793,
      "mean_token_accuracy": 0.15731133073568343,
      "num_tokens": 11398688.0,
      "step": 4985
    },
    {
      "entropy": 5.778379106521607,
      "epoch": 0.4932779754843812,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004982408142806678,
      "loss": 5.6689,
      "mean_token_accuracy": 0.1573247194290161,
      "num_tokens": 11410420.0,
      "step": 4990
    },
    {
      "entropy": 5.766946649551391,
      "epoch": 0.49377224199288255,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004982364071985549,
      "loss": 5.6466,
      "mean_token_accuracy": 0.16381097137928008,
      "num_tokens": 11420521.0,
      "step": 4995
    },
    {
      "entropy": 5.80004506111145,
      "epoch": 0.4942665085013839,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004982319946247917,
      "loss": 5.7457,
      "mean_token_accuracy": 0.15540773868560792,
      "num_tokens": 11432550.0,
      "step": 5000
    },
    {
      "entropy": 5.876774835586548,
      "epoch": 0.49476077500988536,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004982275765594869,
      "loss": 5.7119,
      "mean_token_accuracy": 0.15528473407030105,
      "num_tokens": 11444602.0,
      "step": 5005
    },
    {
      "entropy": 5.723411178588867,
      "epoch": 0.49525504151838673,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004982231530027489,
      "loss": 5.5624,
      "mean_token_accuracy": 0.16830452531576157,
      "num_tokens": 11456187.0,
      "step": 5010
    },
    {
      "entropy": 5.714942979812622,
      "epoch": 0.4957493080268881,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004982187239546868,
      "loss": 5.7802,
      "mean_token_accuracy": 0.15819563418626786,
      "num_tokens": 11468615.0,
      "step": 5015
    },
    {
      "entropy": 5.8007875919342045,
      "epoch": 0.4962435745353895,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004982142894154095,
      "loss": 5.6822,
      "mean_token_accuracy": 0.1621628761291504,
      "num_tokens": 11480865.0,
      "step": 5020
    },
    {
      "entropy": 5.8547523021698,
      "epoch": 0.49673784104389085,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000498209849385026,
      "loss": 5.7576,
      "mean_token_accuracy": 0.1487893909215927,
      "num_tokens": 11492590.0,
      "step": 5025
    },
    {
      "entropy": 5.832790374755859,
      "epoch": 0.4972321075523923,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004982054038636456,
      "loss": 5.7379,
      "mean_token_accuracy": 0.16455563455820083,
      "num_tokens": 11503481.0,
      "step": 5030
    },
    {
      "entropy": 5.797842264175415,
      "epoch": 0.49772637406089365,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004982009528513776,
      "loss": 5.677,
      "mean_token_accuracy": 0.15781579166650772,
      "num_tokens": 11515083.0,
      "step": 5035
    },
    {
      "entropy": 5.786930131912231,
      "epoch": 0.498220640569395,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004981964963483316,
      "loss": 5.7479,
      "mean_token_accuracy": 0.15487340539693834,
      "num_tokens": 11527923.0,
      "step": 5040
    },
    {
      "entropy": 5.790785455703736,
      "epoch": 0.4987149070778964,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004981920343546171,
      "loss": 5.767,
      "mean_token_accuracy": 0.15837057381868364,
      "num_tokens": 11539456.0,
      "step": 5045
    },
    {
      "entropy": 5.834515476226807,
      "epoch": 0.49920917358639777,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000498187566870344,
      "loss": 5.6746,
      "mean_token_accuracy": 0.16797788292169571,
      "num_tokens": 11550809.0,
      "step": 5050
    },
    {
      "entropy": 5.907103157043457,
      "epoch": 0.49970344009489914,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004981830938956221,
      "loss": 5.8069,
      "mean_token_accuracy": 0.15619237422943116,
      "num_tokens": 11562545.0,
      "step": 5055
    },
    {
      "entropy": 5.729175996780396,
      "epoch": 0.5001977066034006,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004981786154305615,
      "loss": 5.6898,
      "mean_token_accuracy": 0.16165309697389602,
      "num_tokens": 11574479.0,
      "step": 5060
    },
    {
      "entropy": 5.8273388862609865,
      "epoch": 0.500691973111902,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004981741314752722,
      "loss": 5.6308,
      "mean_token_accuracy": 0.16577657908201218,
      "num_tokens": 11583942.0,
      "step": 5065
    },
    {
      "entropy": 5.768800449371338,
      "epoch": 0.5011862396204033,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004981696420298647,
      "loss": 5.7375,
      "mean_token_accuracy": 0.15999182015657426,
      "num_tokens": 11596745.0,
      "step": 5070
    },
    {
      "entropy": 5.69454779624939,
      "epoch": 0.5016805061289047,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004981651470944495,
      "loss": 5.6275,
      "mean_token_accuracy": 0.16875656843185424,
      "num_tokens": 11608950.0,
      "step": 5075
    },
    {
      "entropy": 5.708047723770141,
      "epoch": 0.5021747726374061,
      "grad_norm": 1.265625,
      "learning_rate": 0.000498160646669137,
      "loss": 5.7504,
      "mean_token_accuracy": 0.16267986744642257,
      "num_tokens": 11619996.0,
      "step": 5080
    },
    {
      "entropy": 5.8013287544250485,
      "epoch": 0.5026690391459074,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004981561407540379,
      "loss": 5.6509,
      "mean_token_accuracy": 0.16587038934230805,
      "num_tokens": 11632167.0,
      "step": 5085
    },
    {
      "entropy": 5.7611579418182375,
      "epoch": 0.5031633056544088,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004981516293492632,
      "loss": 5.6612,
      "mean_token_accuracy": 0.16275765299797057,
      "num_tokens": 11643722.0,
      "step": 5090
    },
    {
      "entropy": 5.723428201675415,
      "epoch": 0.5036575721629103,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004981471124549237,
      "loss": 5.6624,
      "mean_token_accuracy": 0.1652078226208687,
      "num_tokens": 11654890.0,
      "step": 5095
    },
    {
      "entropy": 5.806342506408692,
      "epoch": 0.5041518386714117,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004981425900711308,
      "loss": 5.7349,
      "mean_token_accuracy": 0.15667447149753572,
      "num_tokens": 11665485.0,
      "step": 5100
    },
    {
      "entropy": 5.781207275390625,
      "epoch": 0.504646105179913,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004981380621979954,
      "loss": 5.6345,
      "mean_token_accuracy": 0.1660120278596878,
      "num_tokens": 11676145.0,
      "step": 5105
    },
    {
      "entropy": 5.761055755615234,
      "epoch": 0.5051403716884144,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004981335288356291,
      "loss": 5.6519,
      "mean_token_accuracy": 0.15547278001904488,
      "num_tokens": 11687777.0,
      "step": 5110
    },
    {
      "entropy": 5.841087675094604,
      "epoch": 0.5056346381969158,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004981289899841434,
      "loss": 5.6841,
      "mean_token_accuracy": 0.1652728870511055,
      "num_tokens": 11699703.0,
      "step": 5115
    },
    {
      "entropy": 5.83178071975708,
      "epoch": 0.5061289047054172,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004981244456436499,
      "loss": 5.7897,
      "mean_token_accuracy": 0.14936670213937758,
      "num_tokens": 11711632.0,
      "step": 5120
    },
    {
      "entropy": 5.860979509353638,
      "epoch": 0.5066231712139185,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004981198958142604,
      "loss": 5.7706,
      "mean_token_accuracy": 0.15854335874319075,
      "num_tokens": 11723763.0,
      "step": 5125
    },
    {
      "entropy": 5.8356544971466064,
      "epoch": 0.5071174377224199,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004981153404960869,
      "loss": 5.7765,
      "mean_token_accuracy": 0.15582619458436966,
      "num_tokens": 11736061.0,
      "step": 5130
    },
    {
      "entropy": 5.736947250366211,
      "epoch": 0.5076117042309213,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004981107796892414,
      "loss": 5.7768,
      "mean_token_accuracy": 0.16211810559034348,
      "num_tokens": 11749342.0,
      "step": 5135
    },
    {
      "entropy": 5.870418214797974,
      "epoch": 0.5081059707394227,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004981062133938361,
      "loss": 5.7008,
      "mean_token_accuracy": 0.16118012517690658,
      "num_tokens": 11760441.0,
      "step": 5140
    },
    {
      "entropy": 5.736885452270508,
      "epoch": 0.508600237247924,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004981016416099834,
      "loss": 5.617,
      "mean_token_accuracy": 0.16642038524150848,
      "num_tokens": 11772191.0,
      "step": 5145
    },
    {
      "entropy": 5.8572736263275145,
      "epoch": 0.5090945037564255,
      "grad_norm": 1.25,
      "learning_rate": 0.0004980970643377956,
      "loss": 5.8781,
      "mean_token_accuracy": 0.14796603471040726,
      "num_tokens": 11786347.0,
      "step": 5150
    },
    {
      "entropy": 5.814970922470093,
      "epoch": 0.5095887702649269,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004980924815773855,
      "loss": 5.6264,
      "mean_token_accuracy": 0.16511821299791335,
      "num_tokens": 11797741.0,
      "step": 5155
    },
    {
      "entropy": 5.809056663513184,
      "epoch": 0.5100830367734283,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004980878933288657,
      "loss": 5.7587,
      "mean_token_accuracy": 0.15323774218559266,
      "num_tokens": 11809728.0,
      "step": 5160
    },
    {
      "entropy": 5.859042739868164,
      "epoch": 0.5105773032819296,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000498083299592349,
      "loss": 5.7098,
      "mean_token_accuracy": 0.1520225413143635,
      "num_tokens": 11822000.0,
      "step": 5165
    },
    {
      "entropy": 5.8006284713745115,
      "epoch": 0.511071569790431,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004980787003679486,
      "loss": 5.7056,
      "mean_token_accuracy": 0.15795185714960097,
      "num_tokens": 11834059.0,
      "step": 5170
    },
    {
      "entropy": 5.692893600463867,
      "epoch": 0.5115658362989324,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004980740956557776,
      "loss": 5.5905,
      "mean_token_accuracy": 0.1717328056693077,
      "num_tokens": 11844001.0,
      "step": 5175
    },
    {
      "entropy": 5.73483099937439,
      "epoch": 0.5120601028074337,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004980694854559493,
      "loss": 5.7625,
      "mean_token_accuracy": 0.16032130271196365,
      "num_tokens": 11855991.0,
      "step": 5180
    },
    {
      "entropy": 5.868142175674438,
      "epoch": 0.5125543693159351,
      "grad_norm": 1.25,
      "learning_rate": 0.000498064869768577,
      "loss": 5.7355,
      "mean_token_accuracy": 0.1608465313911438,
      "num_tokens": 11867975.0,
      "step": 5185
    },
    {
      "entropy": 5.761822128295899,
      "epoch": 0.5130486358244365,
      "grad_norm": 1.25,
      "learning_rate": 0.0004980602485937743,
      "loss": 5.6842,
      "mean_token_accuracy": 0.1614299662411213,
      "num_tokens": 11880034.0,
      "step": 5190
    },
    {
      "entropy": 5.815242528915405,
      "epoch": 0.5135429023329379,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004980556219316549,
      "loss": 5.7557,
      "mean_token_accuracy": 0.1584977239370346,
      "num_tokens": 11891490.0,
      "step": 5195
    },
    {
      "entropy": 5.83032603263855,
      "epoch": 0.5140371688414394,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004980509897823326,
      "loss": 5.6579,
      "mean_token_accuracy": 0.16577705144882202,
      "num_tokens": 11902988.0,
      "step": 5200
    },
    {
      "entropy": 5.741867351531982,
      "epoch": 0.5145314353499407,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004980463521459212,
      "loss": 5.7216,
      "mean_token_accuracy": 0.16082934737205506,
      "num_tokens": 11915088.0,
      "step": 5205
    },
    {
      "entropy": 5.786735105514526,
      "epoch": 0.5150257018584421,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004980417090225351,
      "loss": 5.7334,
      "mean_token_accuracy": 0.16337004601955413,
      "num_tokens": 11926789.0,
      "step": 5210
    },
    {
      "entropy": 5.9148640632629395,
      "epoch": 0.5155199683669435,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004980370604122883,
      "loss": 5.7611,
      "mean_token_accuracy": 0.15934159755706787,
      "num_tokens": 11938876.0,
      "step": 5215
    },
    {
      "entropy": 5.8335541725158695,
      "epoch": 0.5160142348754448,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004980324063152952,
      "loss": 5.7172,
      "mean_token_accuracy": 0.16266890615224838,
      "num_tokens": 11950099.0,
      "step": 5220
    },
    {
      "entropy": 5.842185688018799,
      "epoch": 0.5165085013839462,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004980277467316704,
      "loss": 5.7757,
      "mean_token_accuracy": 0.15510211884975433,
      "num_tokens": 11962585.0,
      "step": 5225
    },
    {
      "entropy": 5.820865440368652,
      "epoch": 0.5170027678924476,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004980230816615282,
      "loss": 5.7006,
      "mean_token_accuracy": 0.16341910213232042,
      "num_tokens": 11972579.0,
      "step": 5230
    },
    {
      "entropy": 5.80379729270935,
      "epoch": 0.517497034400949,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004980184111049837,
      "loss": 5.6335,
      "mean_token_accuracy": 0.16957199275493623,
      "num_tokens": 11983337.0,
      "step": 5235
    },
    {
      "entropy": 5.778374338150025,
      "epoch": 0.5179913009094503,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004980137350621518,
      "loss": 5.7555,
      "mean_token_accuracy": 0.1562938116490841,
      "num_tokens": 11995086.0,
      "step": 5240
    },
    {
      "entropy": 5.847301006317139,
      "epoch": 0.5184855674179517,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004980090535331473,
      "loss": 5.7757,
      "mean_token_accuracy": 0.1603434130549431,
      "num_tokens": 12007814.0,
      "step": 5245
    },
    {
      "entropy": 5.7690938949584964,
      "epoch": 0.5189798339264532,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004980043665180857,
      "loss": 5.6201,
      "mean_token_accuracy": 0.16814614981412887,
      "num_tokens": 12020714.0,
      "step": 5250
    },
    {
      "entropy": 5.707731676101685,
      "epoch": 0.5194741004349546,
      "grad_norm": 1.5,
      "learning_rate": 0.0004979996740170819,
      "loss": 5.6433,
      "mean_token_accuracy": 0.1608419969677925,
      "num_tokens": 12031994.0,
      "step": 5255
    },
    {
      "entropy": 5.841365575790405,
      "epoch": 0.5199683669434559,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004979949760302515,
      "loss": 5.682,
      "mean_token_accuracy": 0.15966767072677612,
      "num_tokens": 12042887.0,
      "step": 5260
    },
    {
      "entropy": 5.849974203109741,
      "epoch": 0.5204626334519573,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004979902725577101,
      "loss": 5.735,
      "mean_token_accuracy": 0.15892827957868577,
      "num_tokens": 12054426.0,
      "step": 5265
    },
    {
      "entropy": 5.810884571075439,
      "epoch": 0.5209568999604587,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004979855635995735,
      "loss": 5.7355,
      "mean_token_accuracy": 0.15427639484405517,
      "num_tokens": 12066360.0,
      "step": 5270
    },
    {
      "entropy": 5.898036861419678,
      "epoch": 0.5214511664689601,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004979808491559573,
      "loss": 5.7527,
      "mean_token_accuracy": 0.15606434941291808,
      "num_tokens": 12076651.0,
      "step": 5275
    },
    {
      "entropy": 5.781798648834228,
      "epoch": 0.5219454329774614,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004979761292269778,
      "loss": 5.6263,
      "mean_token_accuracy": 0.16331419348716736,
      "num_tokens": 12087522.0,
      "step": 5280
    },
    {
      "entropy": 5.740968227386475,
      "epoch": 0.5224396994859628,
      "grad_norm": 1.375,
      "learning_rate": 0.0004979714038127508,
      "loss": 5.7151,
      "mean_token_accuracy": 0.15970026403665544,
      "num_tokens": 12097999.0,
      "step": 5285
    },
    {
      "entropy": 5.736458444595337,
      "epoch": 0.5229339659944642,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004979666729133927,
      "loss": 5.6423,
      "mean_token_accuracy": 0.15874699205160142,
      "num_tokens": 12108080.0,
      "step": 5290
    },
    {
      "entropy": 5.913213014602661,
      "epoch": 0.5234282325029656,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004979619365290199,
      "loss": 5.8909,
      "mean_token_accuracy": 0.1450125604867935,
      "num_tokens": 12122705.0,
      "step": 5295
    },
    {
      "entropy": 5.923708438873291,
      "epoch": 0.523922499011467,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004979571946597488,
      "loss": 5.7064,
      "mean_token_accuracy": 0.1583932265639305,
      "num_tokens": 12134000.0,
      "step": 5300
    },
    {
      "entropy": 5.741547060012818,
      "epoch": 0.5244167655199684,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004979524473056961,
      "loss": 5.7182,
      "mean_token_accuracy": 0.1550974190235138,
      "num_tokens": 12144825.0,
      "step": 5305
    },
    {
      "entropy": 5.817003345489502,
      "epoch": 0.5249110320284698,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004979476944669787,
      "loss": 5.8085,
      "mean_token_accuracy": 0.15489696711301804,
      "num_tokens": 12156590.0,
      "step": 5310
    },
    {
      "entropy": 5.874768495559692,
      "epoch": 0.5254052985369712,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004979429361437134,
      "loss": 5.6488,
      "mean_token_accuracy": 0.16562367230653763,
      "num_tokens": 12167478.0,
      "step": 5315
    },
    {
      "entropy": 5.818252611160278,
      "epoch": 0.5258995650454725,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004979381723360173,
      "loss": 5.6482,
      "mean_token_accuracy": 0.16822979003190994,
      "num_tokens": 12178728.0,
      "step": 5320
    },
    {
      "entropy": 5.6653368949890135,
      "epoch": 0.5263938315539739,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004979334030440076,
      "loss": 5.6163,
      "mean_token_accuracy": 0.15801714062690736,
      "num_tokens": 12190061.0,
      "step": 5325
    },
    {
      "entropy": 5.708331680297851,
      "epoch": 0.5268880980624753,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004979286282678015,
      "loss": 5.7284,
      "mean_token_accuracy": 0.1623130276799202,
      "num_tokens": 12200867.0,
      "step": 5330
    },
    {
      "entropy": 5.832484483718872,
      "epoch": 0.5273823645709766,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004979238480075166,
      "loss": 5.6842,
      "mean_token_accuracy": 0.16192447394132614,
      "num_tokens": 12212681.0,
      "step": 5335
    },
    {
      "entropy": 5.8057342052459715,
      "epoch": 0.527876631079478,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004979190622632705,
      "loss": 5.727,
      "mean_token_accuracy": 0.16040914058685302,
      "num_tokens": 12224279.0,
      "step": 5340
    },
    {
      "entropy": 5.689106702804565,
      "epoch": 0.5283708975879794,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004979142710351809,
      "loss": 5.5774,
      "mean_token_accuracy": 0.17663966119289398,
      "num_tokens": 12235754.0,
      "step": 5345
    },
    {
      "entropy": 5.832288217544556,
      "epoch": 0.5288651640964809,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004979094743233656,
      "loss": 5.706,
      "mean_token_accuracy": 0.16103370785713195,
      "num_tokens": 12246014.0,
      "step": 5350
    },
    {
      "entropy": 5.761579513549805,
      "epoch": 0.5293594306049823,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004979046721279426,
      "loss": 5.6665,
      "mean_token_accuracy": 0.16219642460346223,
      "num_tokens": 12257519.0,
      "step": 5355
    },
    {
      "entropy": 5.7685311794281,
      "epoch": 0.5298536971134836,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004978998644490302,
      "loss": 5.7177,
      "mean_token_accuracy": 0.15667742639780044,
      "num_tokens": 12269806.0,
      "step": 5360
    },
    {
      "entropy": 5.7736852169036865,
      "epoch": 0.530347963621985,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004978950512867465,
      "loss": 5.7213,
      "mean_token_accuracy": 0.15609590262174605,
      "num_tokens": 12281612.0,
      "step": 5365
    },
    {
      "entropy": 5.835879468917847,
      "epoch": 0.5308422301304864,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004978902326412099,
      "loss": 5.6872,
      "mean_token_accuracy": 0.159187912940979,
      "num_tokens": 12293022.0,
      "step": 5370
    },
    {
      "entropy": 5.790759038925171,
      "epoch": 0.5313364966389877,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000497885408512539,
      "loss": 5.6714,
      "mean_token_accuracy": 0.16126388013362886,
      "num_tokens": 12304414.0,
      "step": 5375
    },
    {
      "entropy": 5.893075704574585,
      "epoch": 0.5318307631474891,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004978805789008526,
      "loss": 5.8077,
      "mean_token_accuracy": 0.15097813457250595,
      "num_tokens": 12315618.0,
      "step": 5380
    },
    {
      "entropy": 5.829635524749756,
      "epoch": 0.5323250296559905,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004978757438062692,
      "loss": 5.6497,
      "mean_token_accuracy": 0.16604820936918258,
      "num_tokens": 12328026.0,
      "step": 5385
    },
    {
      "entropy": 5.793191528320312,
      "epoch": 0.5328192961644919,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004978709032289081,
      "loss": 5.7214,
      "mean_token_accuracy": 0.1559169754385948,
      "num_tokens": 12339441.0,
      "step": 5390
    },
    {
      "entropy": 5.85589632987976,
      "epoch": 0.5333135626729932,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004978660571688881,
      "loss": 5.7425,
      "mean_token_accuracy": 0.15749634355306624,
      "num_tokens": 12351005.0,
      "step": 5395
    },
    {
      "entropy": 5.811450242996216,
      "epoch": 0.5338078291814946,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004978612056263286,
      "loss": 5.7471,
      "mean_token_accuracy": 0.16507457196712494,
      "num_tokens": 12363147.0,
      "step": 5400
    },
    {
      "entropy": 5.825579690933227,
      "epoch": 0.5343020956899961,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004978563486013488,
      "loss": 5.6776,
      "mean_token_accuracy": 0.16178106069564818,
      "num_tokens": 12373446.0,
      "step": 5405
    },
    {
      "entropy": 5.7309596061706545,
      "epoch": 0.5347963621984975,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004978514860940683,
      "loss": 5.6618,
      "mean_token_accuracy": 0.16514517068862916,
      "num_tokens": 12384591.0,
      "step": 5410
    },
    {
      "entropy": 5.761965560913086,
      "epoch": 0.5352906287069988,
      "grad_norm": 1.5,
      "learning_rate": 0.0004978466181046067,
      "loss": 5.6196,
      "mean_token_accuracy": 0.16532472819089888,
      "num_tokens": 12396117.0,
      "step": 5415
    },
    {
      "entropy": 5.76508355140686,
      "epoch": 0.5357848952155002,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004978417446330837,
      "loss": 5.6999,
      "mean_token_accuracy": 0.16554254442453384,
      "num_tokens": 12407400.0,
      "step": 5420
    },
    {
      "entropy": 5.7422675609588625,
      "epoch": 0.5362791617240016,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004978368656796191,
      "loss": 5.6184,
      "mean_token_accuracy": 0.16983277797698976,
      "num_tokens": 12417629.0,
      "step": 5425
    },
    {
      "entropy": 5.7121666431427,
      "epoch": 0.536773428232503,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004978319812443331,
      "loss": 5.6461,
      "mean_token_accuracy": 0.1678358569741249,
      "num_tokens": 12428522.0,
      "step": 5430
    },
    {
      "entropy": 5.803666687011718,
      "epoch": 0.5372676947410043,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004978270913273458,
      "loss": 5.7394,
      "mean_token_accuracy": 0.15472337305545808,
      "num_tokens": 12440390.0,
      "step": 5435
    },
    {
      "entropy": 5.783153104782104,
      "epoch": 0.5377619612495057,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004978221959287775,
      "loss": 5.6473,
      "mean_token_accuracy": 0.16296303272247314,
      "num_tokens": 12451846.0,
      "step": 5440
    },
    {
      "entropy": 5.743931531906128,
      "epoch": 0.5382562277580071,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004978172950487486,
      "loss": 5.6829,
      "mean_token_accuracy": 0.16433457732200624,
      "num_tokens": 12462051.0,
      "step": 5445
    },
    {
      "entropy": 5.716442108154297,
      "epoch": 0.5387504942665085,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004978123886873797,
      "loss": 5.6008,
      "mean_token_accuracy": 0.17102793008089065,
      "num_tokens": 12472730.0,
      "step": 5450
    },
    {
      "entropy": 5.751273155212402,
      "epoch": 0.5392447607750099,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004978074768447915,
      "loss": 5.6123,
      "mean_token_accuracy": 0.17399577647447587,
      "num_tokens": 12482719.0,
      "step": 5455
    },
    {
      "entropy": 5.761351537704468,
      "epoch": 0.5397390272835113,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004978025595211048,
      "loss": 5.6517,
      "mean_token_accuracy": 0.16074512004852295,
      "num_tokens": 12492659.0,
      "step": 5460
    },
    {
      "entropy": 5.882458209991455,
      "epoch": 0.5402332937920127,
      "grad_norm": 1.375,
      "learning_rate": 0.0004977976367164404,
      "loss": 5.7174,
      "mean_token_accuracy": 0.16112817525863649,
      "num_tokens": 12504381.0,
      "step": 5465
    },
    {
      "entropy": 5.807905769348144,
      "epoch": 0.540727560300514,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004977927084309198,
      "loss": 5.5634,
      "mean_token_accuracy": 0.17260986417531968,
      "num_tokens": 12515041.0,
      "step": 5470
    },
    {
      "entropy": 5.804184818267823,
      "epoch": 0.5412218268090154,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004977877746646639,
      "loss": 5.7184,
      "mean_token_accuracy": 0.15993063598871232,
      "num_tokens": 12526313.0,
      "step": 5475
    },
    {
      "entropy": 5.80208010673523,
      "epoch": 0.5417160933175168,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004977828354177941,
      "loss": 5.807,
      "mean_token_accuracy": 0.14883915781974794,
      "num_tokens": 12538268.0,
      "step": 5480
    },
    {
      "entropy": 5.808342599868775,
      "epoch": 0.5422103598260182,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004977778906904321,
      "loss": 5.7029,
      "mean_token_accuracy": 0.1601228341460228,
      "num_tokens": 12548365.0,
      "step": 5485
    },
    {
      "entropy": 5.879845809936524,
      "epoch": 0.5427046263345195,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004977729404826995,
      "loss": 5.8156,
      "mean_token_accuracy": 0.14491160213947296,
      "num_tokens": 12560652.0,
      "step": 5490
    },
    {
      "entropy": 5.794532537460327,
      "epoch": 0.5431988928430209,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004977679847947179,
      "loss": 5.6637,
      "mean_token_accuracy": 0.16358893662691115,
      "num_tokens": 12572079.0,
      "step": 5495
    },
    {
      "entropy": 5.745615434646607,
      "epoch": 0.5436931593515223,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004977630236266093,
      "loss": 5.5743,
      "mean_token_accuracy": 0.16720821112394332,
      "num_tokens": 12582601.0,
      "step": 5500
    },
    {
      "entropy": 5.836099672317505,
      "epoch": 0.5441874258600238,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004977580569784958,
      "loss": 5.7246,
      "mean_token_accuracy": 0.15426233559846877,
      "num_tokens": 12595176.0,
      "step": 5505
    },
    {
      "entropy": 5.81144642829895,
      "epoch": 0.5446816923685251,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004977530848504994,
      "loss": 5.7258,
      "mean_token_accuracy": 0.15800877958536147,
      "num_tokens": 12608161.0,
      "step": 5510
    },
    {
      "entropy": 5.777765703201294,
      "epoch": 0.5451759588770265,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004977481072427427,
      "loss": 5.7436,
      "mean_token_accuracy": 0.15919482260942458,
      "num_tokens": 12620109.0,
      "step": 5515
    },
    {
      "entropy": 5.854056882858276,
      "epoch": 0.5456702253855279,
      "grad_norm": 1.640625,
      "learning_rate": 0.000497743124155348,
      "loss": 5.7314,
      "mean_token_accuracy": 0.15875154584646226,
      "num_tokens": 12630339.0,
      "step": 5520
    },
    {
      "entropy": 5.824705743789673,
      "epoch": 0.5461644918940293,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004977381355884379,
      "loss": 5.7727,
      "mean_token_accuracy": 0.16132542490959167,
      "num_tokens": 12642101.0,
      "step": 5525
    },
    {
      "entropy": 5.889630699157715,
      "epoch": 0.5466587584025306,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004977331415421351,
      "loss": 5.8264,
      "mean_token_accuracy": 0.14514302015304564,
      "num_tokens": 12655436.0,
      "step": 5530
    },
    {
      "entropy": 5.740127992630005,
      "epoch": 0.547153024911032,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004977281420165624,
      "loss": 5.5126,
      "mean_token_accuracy": 0.1704614356160164,
      "num_tokens": 12667100.0,
      "step": 5535
    },
    {
      "entropy": 5.814797306060791,
      "epoch": 0.5476472914195334,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000497723137011843,
      "loss": 5.7994,
      "mean_token_accuracy": 0.15522741228342057,
      "num_tokens": 12678788.0,
      "step": 5540
    },
    {
      "entropy": 5.793706703186035,
      "epoch": 0.5481415579280348,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004977181265280997,
      "loss": 5.6823,
      "mean_token_accuracy": 0.16902316808700563,
      "num_tokens": 12690980.0,
      "step": 5545
    },
    {
      "entropy": 5.889812088012695,
      "epoch": 0.5486358244365361,
      "grad_norm": 1.453125,
      "learning_rate": 0.000497713110565456,
      "loss": 5.7941,
      "mean_token_accuracy": 0.15406299531459808,
      "num_tokens": 12702642.0,
      "step": 5550
    },
    {
      "entropy": 5.710815286636352,
      "epoch": 0.5491300909450376,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004977080891240352,
      "loss": 5.5668,
      "mean_token_accuracy": 0.17032462358474731,
      "num_tokens": 12712750.0,
      "step": 5555
    },
    {
      "entropy": 5.71341609954834,
      "epoch": 0.549624357453539,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000497703062203961,
      "loss": 5.5853,
      "mean_token_accuracy": 0.16658094078302382,
      "num_tokens": 12724464.0,
      "step": 5560
    },
    {
      "entropy": 5.749512577056885,
      "epoch": 0.5501186239620404,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004976980298053567,
      "loss": 5.6705,
      "mean_token_accuracy": 0.1623353898525238,
      "num_tokens": 12735419.0,
      "step": 5565
    },
    {
      "entropy": 5.863523006439209,
      "epoch": 0.5506128904705417,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004976929919283465,
      "loss": 5.6631,
      "mean_token_accuracy": 0.16216106861829757,
      "num_tokens": 12746279.0,
      "step": 5570
    },
    {
      "entropy": 5.748651504516602,
      "epoch": 0.5511071569790431,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004976879485730541,
      "loss": 5.7698,
      "mean_token_accuracy": 0.15983541011810304,
      "num_tokens": 12758107.0,
      "step": 5575
    },
    {
      "entropy": 5.745821332931518,
      "epoch": 0.5516014234875445,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004976828997396036,
      "loss": 5.6676,
      "mean_token_accuracy": 0.17020906507968903,
      "num_tokens": 12769354.0,
      "step": 5580
    },
    {
      "entropy": 5.821145534515381,
      "epoch": 0.5520956899960459,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004976778454281193,
      "loss": 5.6384,
      "mean_token_accuracy": 0.1623901054263115,
      "num_tokens": 12780926.0,
      "step": 5585
    },
    {
      "entropy": 5.810956811904907,
      "epoch": 0.5525899565045472,
      "grad_norm": 1.5,
      "learning_rate": 0.0004976727856387253,
      "loss": 5.6478,
      "mean_token_accuracy": 0.15976598411798476,
      "num_tokens": 12792423.0,
      "step": 5590
    },
    {
      "entropy": 5.711795043945313,
      "epoch": 0.5530842230130486,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004976677203715464,
      "loss": 5.5853,
      "mean_token_accuracy": 0.17233408838510514,
      "num_tokens": 12803688.0,
      "step": 5595
    },
    {
      "entropy": 5.709937620162964,
      "epoch": 0.55357848952155,
      "grad_norm": 1.5546875,
      "learning_rate": 0.000497662649626707,
      "loss": 5.5821,
      "mean_token_accuracy": 0.16821181625127793,
      "num_tokens": 12814617.0,
      "step": 5600
    },
    {
      "entropy": 5.731007671356201,
      "epoch": 0.5540727560300513,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004976575734043318,
      "loss": 5.6118,
      "mean_token_accuracy": 0.16292986273765564,
      "num_tokens": 12826027.0,
      "step": 5605
    },
    {
      "entropy": 5.753634214401245,
      "epoch": 0.5545670225385528,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004976524917045457,
      "loss": 5.7005,
      "mean_token_accuracy": 0.15490526854991912,
      "num_tokens": 12837454.0,
      "step": 5610
    },
    {
      "entropy": 5.759040403366089,
      "epoch": 0.5550612890470542,
      "grad_norm": 1.640625,
      "learning_rate": 0.000497647404527474,
      "loss": 5.597,
      "mean_token_accuracy": 0.16277759820222854,
      "num_tokens": 12848550.0,
      "step": 5615
    },
    {
      "entropy": 5.775821399688721,
      "epoch": 0.5555555555555556,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004976423118732413,
      "loss": 5.6663,
      "mean_token_accuracy": 0.16614630222320556,
      "num_tokens": 12858270.0,
      "step": 5620
    },
    {
      "entropy": 5.715994215011596,
      "epoch": 0.556049822064057,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004976372137419733,
      "loss": 5.6312,
      "mean_token_accuracy": 0.16284454464912415,
      "num_tokens": 12869507.0,
      "step": 5625
    },
    {
      "entropy": 5.775624847412109,
      "epoch": 0.5565440885725583,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004976321101337954,
      "loss": 5.5959,
      "mean_token_accuracy": 0.16878042072057725,
      "num_tokens": 12880097.0,
      "step": 5630
    },
    {
      "entropy": 5.714468574523925,
      "epoch": 0.5570383550810597,
      "grad_norm": 1.390625,
      "learning_rate": 0.000497627001048833,
      "loss": 5.5886,
      "mean_token_accuracy": 0.1684863582253456,
      "num_tokens": 12891302.0,
      "step": 5635
    },
    {
      "entropy": 5.662768936157226,
      "epoch": 0.5575326215895611,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004976218864872118,
      "loss": 5.6089,
      "mean_token_accuracy": 0.17541296780109406,
      "num_tokens": 12903950.0,
      "step": 5640
    },
    {
      "entropy": 5.763148927688599,
      "epoch": 0.5580268880980624,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004976167664490576,
      "loss": 5.6523,
      "mean_token_accuracy": 0.16255742013454438,
      "num_tokens": 12915713.0,
      "step": 5645
    },
    {
      "entropy": 5.76856484413147,
      "epoch": 0.5585211546065638,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004976116409344966,
      "loss": 5.6154,
      "mean_token_accuracy": 0.16408945620059967,
      "num_tokens": 12926655.0,
      "step": 5650
    },
    {
      "entropy": 5.692210292816162,
      "epoch": 0.5590154211150652,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004976065099436546,
      "loss": 5.6092,
      "mean_token_accuracy": 0.16390977948904037,
      "num_tokens": 12938144.0,
      "step": 5655
    },
    {
      "entropy": 5.833284473419189,
      "epoch": 0.5595096876235667,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004976013734766579,
      "loss": 5.7924,
      "mean_token_accuracy": 0.15775922387838365,
      "num_tokens": 12950486.0,
      "step": 5660
    },
    {
      "entropy": 5.755685710906983,
      "epoch": 0.560003954132068,
      "grad_norm": 1.71875,
      "learning_rate": 0.000497596231533633,
      "loss": 5.5279,
      "mean_token_accuracy": 0.1694363996386528,
      "num_tokens": 12960966.0,
      "step": 5665
    },
    {
      "entropy": 5.704496192932129,
      "epoch": 0.5604982206405694,
      "grad_norm": 1.65625,
      "learning_rate": 0.000497591084114706,
      "loss": 5.6336,
      "mean_token_accuracy": 0.16190723776817323,
      "num_tokens": 12971960.0,
      "step": 5670
    },
    {
      "entropy": 5.789654350280761,
      "epoch": 0.5609924871490708,
      "grad_norm": 1.6875,
      "learning_rate": 0.000497585931220004,
      "loss": 5.7266,
      "mean_token_accuracy": 0.16142417639493942,
      "num_tokens": 12984186.0,
      "step": 5675
    },
    {
      "entropy": 5.7306139945983885,
      "epoch": 0.5614867536575722,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004975807728496535,
      "loss": 5.562,
      "mean_token_accuracy": 0.17325062155723572,
      "num_tokens": 12994205.0,
      "step": 5680
    },
    {
      "entropy": 5.799353694915771,
      "epoch": 0.5619810201660735,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004975756090037815,
      "loss": 5.6845,
      "mean_token_accuracy": 0.16368188560009003,
      "num_tokens": 13006143.0,
      "step": 5685
    },
    {
      "entropy": 5.88020429611206,
      "epoch": 0.5624752866745749,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000497570439682515,
      "loss": 5.7204,
      "mean_token_accuracy": 0.16078546196222304,
      "num_tokens": 13017849.0,
      "step": 5690
    },
    {
      "entropy": 5.665234613418579,
      "epoch": 0.5629695531830763,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000497565264885981,
      "loss": 5.5975,
      "mean_token_accuracy": 0.1682819738984108,
      "num_tokens": 13030301.0,
      "step": 5695
    },
    {
      "entropy": 5.71648907661438,
      "epoch": 0.5634638196915777,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004975600846143072,
      "loss": 5.6572,
      "mean_token_accuracy": 0.16545034497976302,
      "num_tokens": 13040229.0,
      "step": 5700
    },
    {
      "entropy": 5.736776065826416,
      "epoch": 0.563958086200079,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004975548988676206,
      "loss": 5.592,
      "mean_token_accuracy": 0.16608160734176636,
      "num_tokens": 13050752.0,
      "step": 5705
    },
    {
      "entropy": 5.809312772750855,
      "epoch": 0.5644523527085805,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004975497076460491,
      "loss": 5.7965,
      "mean_token_accuracy": 0.15191729068756105,
      "num_tokens": 13062003.0,
      "step": 5710
    },
    {
      "entropy": 5.793465518951416,
      "epoch": 0.5649466192170819,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004975445109497202,
      "loss": 5.708,
      "mean_token_accuracy": 0.1548692211508751,
      "num_tokens": 13073746.0,
      "step": 5715
    },
    {
      "entropy": 5.859099817276001,
      "epoch": 0.5654408857255833,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004975393087787617,
      "loss": 5.7765,
      "mean_token_accuracy": 0.15511635392904283,
      "num_tokens": 13086068.0,
      "step": 5720
    },
    {
      "entropy": 5.81207914352417,
      "epoch": 0.5659351522340846,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004975341011333019,
      "loss": 5.6729,
      "mean_token_accuracy": 0.16146194487810134,
      "num_tokens": 13097078.0,
      "step": 5725
    },
    {
      "entropy": 5.746955919265747,
      "epoch": 0.566429418742586,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004975288880134685,
      "loss": 5.6566,
      "mean_token_accuracy": 0.16354386880993843,
      "num_tokens": 13108450.0,
      "step": 5730
    },
    {
      "entropy": 5.790970754623413,
      "epoch": 0.5669236852510874,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004975236694193901,
      "loss": 5.7057,
      "mean_token_accuracy": 0.15594887360930443,
      "num_tokens": 13120216.0,
      "step": 5735
    },
    {
      "entropy": 5.838050889968872,
      "epoch": 0.5674179517595888,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004975184453511948,
      "loss": 5.6929,
      "mean_token_accuracy": 0.15619661808013915,
      "num_tokens": 13131214.0,
      "step": 5740
    },
    {
      "entropy": 5.817377376556396,
      "epoch": 0.5679122182680901,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004975132158090112,
      "loss": 5.7391,
      "mean_token_accuracy": 0.15629347264766694,
      "num_tokens": 13143792.0,
      "step": 5745
    },
    {
      "entropy": 5.72936897277832,
      "epoch": 0.5684064847765915,
      "grad_norm": 2.78125,
      "learning_rate": 0.000497507980792968,
      "loss": 5.7115,
      "mean_token_accuracy": 0.15781604796648024,
      "num_tokens": 13156022.0,
      "step": 5750
    },
    {
      "entropy": 5.779087352752685,
      "epoch": 0.5689007512850929,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000497502740303194,
      "loss": 5.657,
      "mean_token_accuracy": 0.16479096114635466,
      "num_tokens": 13167374.0,
      "step": 5755
    },
    {
      "entropy": 5.718874120712281,
      "epoch": 0.5693950177935944,
      "grad_norm": 1.6640625,
      "learning_rate": 0.000497497494339818,
      "loss": 5.5917,
      "mean_token_accuracy": 0.17131200879812242,
      "num_tokens": 13177662.0,
      "step": 5760
    },
    {
      "entropy": 5.713671731948852,
      "epoch": 0.5698892843020957,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004974922429029692,
      "loss": 5.6569,
      "mean_token_accuracy": 0.16335739940404892,
      "num_tokens": 13188375.0,
      "step": 5765
    },
    {
      "entropy": 5.797548484802246,
      "epoch": 0.5703835508105971,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004974869859927767,
      "loss": 5.5823,
      "mean_token_accuracy": 0.16842106431722642,
      "num_tokens": 13198988.0,
      "step": 5770
    },
    {
      "entropy": 5.837228012084961,
      "epoch": 0.5708778173190985,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004974817236093697,
      "loss": 5.6669,
      "mean_token_accuracy": 0.15082263350486755,
      "num_tokens": 13210699.0,
      "step": 5775
    },
    {
      "entropy": 5.689919376373291,
      "epoch": 0.5713720838275999,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004974764557528779,
      "loss": 5.5902,
      "mean_token_accuracy": 0.16678860187530517,
      "num_tokens": 13221925.0,
      "step": 5780
    },
    {
      "entropy": 5.740795373916626,
      "epoch": 0.5718663503361012,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004974711824234308,
      "loss": 5.7047,
      "mean_token_accuracy": 0.15749675184488296,
      "num_tokens": 13233278.0,
      "step": 5785
    },
    {
      "entropy": 5.813934755325318,
      "epoch": 0.5723606168446026,
      "grad_norm": 1.421875,
      "learning_rate": 0.000497465903621158,
      "loss": 5.6713,
      "mean_token_accuracy": 0.16290869414806367,
      "num_tokens": 13245423.0,
      "step": 5790
    },
    {
      "entropy": 5.772157192230225,
      "epoch": 0.572854883353104,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004974606193461895,
      "loss": 5.7195,
      "mean_token_accuracy": 0.16523960530757903,
      "num_tokens": 13256312.0,
      "step": 5795
    },
    {
      "entropy": 5.7896302223205565,
      "epoch": 0.5733491498616053,
      "grad_norm": 1.625,
      "learning_rate": 0.0004974553295986553,
      "loss": 5.671,
      "mean_token_accuracy": 0.16062716990709305,
      "num_tokens": 13268158.0,
      "step": 5800
    },
    {
      "entropy": 5.740266132354736,
      "epoch": 0.5738434163701067,
      "grad_norm": 4.375,
      "learning_rate": 0.0004974500343786854,
      "loss": 5.6451,
      "mean_token_accuracy": 0.16627271771430968,
      "num_tokens": 13279625.0,
      "step": 5805
    },
    {
      "entropy": 5.76525650024414,
      "epoch": 0.5743376828786081,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004974447336864102,
      "loss": 5.666,
      "mean_token_accuracy": 0.15998833552002906,
      "num_tokens": 13291967.0,
      "step": 5810
    },
    {
      "entropy": 5.686205053329468,
      "epoch": 0.5748319493871096,
      "grad_norm": 1.609375,
      "learning_rate": 0.00049743942752196,
      "loss": 5.553,
      "mean_token_accuracy": 0.1758944123983383,
      "num_tokens": 13303641.0,
      "step": 5815
    },
    {
      "entropy": 5.69308819770813,
      "epoch": 0.575326215895611,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004974341158854654,
      "loss": 5.6128,
      "mean_token_accuracy": 0.1627964958548546,
      "num_tokens": 13314754.0,
      "step": 5820
    },
    {
      "entropy": 5.784170627593994,
      "epoch": 0.5758204824041123,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004974287987770571,
      "loss": 5.7281,
      "mean_token_accuracy": 0.1593188002705574,
      "num_tokens": 13325574.0,
      "step": 5825
    },
    {
      "entropy": 5.7553788185119625,
      "epoch": 0.5763147489126137,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004974234761968658,
      "loss": 5.6266,
      "mean_token_accuracy": 0.17157164961099625,
      "num_tokens": 13335946.0,
      "step": 5830
    },
    {
      "entropy": 5.772668933868408,
      "epoch": 0.5768090154211151,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004974181481450224,
      "loss": 5.6394,
      "mean_token_accuracy": 0.16848790645599365,
      "num_tokens": 13347812.0,
      "step": 5835
    },
    {
      "entropy": 5.708554220199585,
      "epoch": 0.5773032819296164,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004974128146216581,
      "loss": 5.5988,
      "mean_token_accuracy": 0.16473570913076402,
      "num_tokens": 13358492.0,
      "step": 5840
    },
    {
      "entropy": 5.8079423904418945,
      "epoch": 0.5777975484381178,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004974074756269042,
      "loss": 5.7656,
      "mean_token_accuracy": 0.15715205073356628,
      "num_tokens": 13370465.0,
      "step": 5845
    },
    {
      "entropy": 5.811331558227539,
      "epoch": 0.5782918149466192,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004974021311608917,
      "loss": 5.6226,
      "mean_token_accuracy": 0.16963550597429275,
      "num_tokens": 13382328.0,
      "step": 5850
    },
    {
      "entropy": 5.728660964965821,
      "epoch": 0.5787860814551206,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004973967812237523,
      "loss": 5.7362,
      "mean_token_accuracy": 0.1589743450284004,
      "num_tokens": 13393439.0,
      "step": 5855
    },
    {
      "entropy": 5.764760494232178,
      "epoch": 0.5792803479636219,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004973914258156177,
      "loss": 5.66,
      "mean_token_accuracy": 0.16472170054912566,
      "num_tokens": 13404778.0,
      "step": 5860
    },
    {
      "entropy": 5.827328157424927,
      "epoch": 0.5797746144721234,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004973860649366195,
      "loss": 5.6623,
      "mean_token_accuracy": 0.16409119665622712,
      "num_tokens": 13417031.0,
      "step": 5865
    },
    {
      "entropy": 5.772861671447754,
      "epoch": 0.5802688809806248,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004973806985868895,
      "loss": 5.6507,
      "mean_token_accuracy": 0.16444518119096757,
      "num_tokens": 13428030.0,
      "step": 5870
    },
    {
      "entropy": 5.775449323654175,
      "epoch": 0.5807631474891262,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004973753267665599,
      "loss": 5.651,
      "mean_token_accuracy": 0.1662319228053093,
      "num_tokens": 13438934.0,
      "step": 5875
    },
    {
      "entropy": 5.7868541240692135,
      "epoch": 0.5812574139976275,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004973699494757627,
      "loss": 5.6287,
      "mean_token_accuracy": 0.1601264923810959,
      "num_tokens": 13450542.0,
      "step": 5880
    },
    {
      "entropy": 5.67889986038208,
      "epoch": 0.5817516805061289,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004973645667146302,
      "loss": 5.6767,
      "mean_token_accuracy": 0.1716054067015648,
      "num_tokens": 13462382.0,
      "step": 5885
    },
    {
      "entropy": 5.7738378047943115,
      "epoch": 0.5822459470146303,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004973591784832949,
      "loss": 5.6777,
      "mean_token_accuracy": 0.15822771340608596,
      "num_tokens": 13473852.0,
      "step": 5890
    },
    {
      "entropy": 5.811481475830078,
      "epoch": 0.5827402135231317,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004973537847818894,
      "loss": 5.7087,
      "mean_token_accuracy": 0.15851057916879654,
      "num_tokens": 13485763.0,
      "step": 5895
    },
    {
      "entropy": 5.740776443481446,
      "epoch": 0.583234480031633,
      "grad_norm": 1.25,
      "learning_rate": 0.0004973483856105462,
      "loss": 5.6722,
      "mean_token_accuracy": 0.16087493374943734,
      "num_tokens": 13498154.0,
      "step": 5900
    },
    {
      "entropy": 5.851483392715454,
      "epoch": 0.5837287465401344,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004973429809693983,
      "loss": 5.7124,
      "mean_token_accuracy": 0.15585886985063552,
      "num_tokens": 13510563.0,
      "step": 5905
    },
    {
      "entropy": 5.849711275100708,
      "epoch": 0.5842230130486358,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004973375708585785,
      "loss": 5.7279,
      "mean_token_accuracy": 0.16208239942789077,
      "num_tokens": 13521987.0,
      "step": 5910
    },
    {
      "entropy": 5.787547874450683,
      "epoch": 0.5847172795571373,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004973321552782201,
      "loss": 5.6743,
      "mean_token_accuracy": 0.16171410232782363,
      "num_tokens": 13534075.0,
      "step": 5915
    },
    {
      "entropy": 5.83193187713623,
      "epoch": 0.5852115460656386,
      "grad_norm": 1.46875,
      "learning_rate": 0.000497326734228456,
      "loss": 5.8267,
      "mean_token_accuracy": 0.15176279693841935,
      "num_tokens": 13546201.0,
      "step": 5920
    },
    {
      "entropy": 5.834552907943726,
      "epoch": 0.58570581257414,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004973213077094199,
      "loss": 5.6472,
      "mean_token_accuracy": 0.16397160589694976,
      "num_tokens": 13556985.0,
      "step": 5925
    },
    {
      "entropy": 5.82445592880249,
      "epoch": 0.5862000790826414,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004973158757212451,
      "loss": 5.6773,
      "mean_token_accuracy": 0.1634679839015007,
      "num_tokens": 13568617.0,
      "step": 5930
    },
    {
      "entropy": 5.668050289154053,
      "epoch": 0.5866943455911428,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004973104382640652,
      "loss": 5.6194,
      "mean_token_accuracy": 0.16359573006629943,
      "num_tokens": 13581326.0,
      "step": 5935
    },
    {
      "entropy": 5.706488561630249,
      "epoch": 0.5871886120996441,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004973049953380141,
      "loss": 5.6349,
      "mean_token_accuracy": 0.1637824833393097,
      "num_tokens": 13591776.0,
      "step": 5940
    },
    {
      "entropy": 5.85504903793335,
      "epoch": 0.5876828786081455,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004972995469432256,
      "loss": 5.7793,
      "mean_token_accuracy": 0.15763178318738938,
      "num_tokens": 13603475.0,
      "step": 5945
    },
    {
      "entropy": 5.793098735809326,
      "epoch": 0.5881771451166469,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004972940930798338,
      "loss": 5.6696,
      "mean_token_accuracy": 0.1648968607187271,
      "num_tokens": 13615348.0,
      "step": 5950
    },
    {
      "entropy": 5.720390892028808,
      "epoch": 0.5886714116251482,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004972886337479727,
      "loss": 5.5652,
      "mean_token_accuracy": 0.16782755702733992,
      "num_tokens": 13625982.0,
      "step": 5955
    },
    {
      "entropy": 5.809796619415283,
      "epoch": 0.5891656781336496,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004972831689477768,
      "loss": 5.6988,
      "mean_token_accuracy": 0.15440213531255723,
      "num_tokens": 13637203.0,
      "step": 5960
    },
    {
      "entropy": 5.763940525054932,
      "epoch": 0.5896599446421511,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004972776986793806,
      "loss": 5.6135,
      "mean_token_accuracy": 0.1651282861828804,
      "num_tokens": 13648880.0,
      "step": 5965
    },
    {
      "entropy": 5.771618938446045,
      "epoch": 0.5901542111506525,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004972722229429184,
      "loss": 5.697,
      "mean_token_accuracy": 0.16329590529203414,
      "num_tokens": 13660266.0,
      "step": 5970
    },
    {
      "entropy": 5.889302015304565,
      "epoch": 0.5906484776591538,
      "grad_norm": 1.53125,
      "learning_rate": 0.000497266741738525,
      "loss": 5.7748,
      "mean_token_accuracy": 0.14944686591625214,
      "num_tokens": 13672063.0,
      "step": 5975
    },
    {
      "entropy": 5.85100908279419,
      "epoch": 0.5911427441676552,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004972612550663353,
      "loss": 5.6693,
      "mean_token_accuracy": 0.16710522919893264,
      "num_tokens": 13683379.0,
      "step": 5980
    },
    {
      "entropy": 5.740947961807251,
      "epoch": 0.5916370106761566,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004972557629264842,
      "loss": 5.6176,
      "mean_token_accuracy": 0.16787842363119126,
      "num_tokens": 13694512.0,
      "step": 5985
    },
    {
      "entropy": 5.738776826858521,
      "epoch": 0.592131277184658,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004972502653191069,
      "loss": 5.6258,
      "mean_token_accuracy": 0.15922403559088708,
      "num_tokens": 13705642.0,
      "step": 5990
    },
    {
      "entropy": 5.705231809616089,
      "epoch": 0.5926255436931593,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004972447622443387,
      "loss": 5.5422,
      "mean_token_accuracy": 0.1716584861278534,
      "num_tokens": 13717367.0,
      "step": 5995
    },
    {
      "entropy": 5.697733211517334,
      "epoch": 0.5931198102016607,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004972392537023147,
      "loss": 5.6232,
      "mean_token_accuracy": 0.16564393490552903,
      "num_tokens": 13728577.0,
      "step": 6000
    },
    {
      "epoch": 0.5931198102016607,
      "eval_entropy": 5.626806823852464,
      "eval_loss": 5.705826759338379,
      "eval_mean_token_accuracy": 0.16681483873288716,
      "eval_num_tokens": 13728577.0,
      "eval_runtime": 26.5814,
      "eval_samples_per_second": 1223.15,
      "eval_steps_per_second": 152.927,
      "step": 6000
    },
    {
      "entropy": 5.774589395523071,
      "epoch": 0.5936140767101621,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004972337396931707,
      "loss": 5.6272,
      "mean_token_accuracy": 0.16147202998399734,
      "num_tokens": 13739016.0,
      "step": 6005
    },
    {
      "entropy": 5.8219915390014645,
      "epoch": 0.5941083432186635,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004972282202170421,
      "loss": 5.6713,
      "mean_token_accuracy": 0.15762737691402434,
      "num_tokens": 13749971.0,
      "step": 6010
    },
    {
      "entropy": 5.744158411026001,
      "epoch": 0.5946026097271648,
      "grad_norm": 1.484375,
      "learning_rate": 0.000497222695274065,
      "loss": 5.7037,
      "mean_token_accuracy": 0.1618531882762909,
      "num_tokens": 13760927.0,
      "step": 6015
    },
    {
      "entropy": 5.762528657913208,
      "epoch": 0.5950968762356663,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004972171648643749,
      "loss": 5.6606,
      "mean_token_accuracy": 0.16211145669221877,
      "num_tokens": 13772725.0,
      "step": 6020
    },
    {
      "entropy": 5.793417167663574,
      "epoch": 0.5955911427441677,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004972116289881082,
      "loss": 5.6606,
      "mean_token_accuracy": 0.16376404762268065,
      "num_tokens": 13783427.0,
      "step": 6025
    },
    {
      "entropy": 5.758443784713745,
      "epoch": 0.5960854092526691,
      "grad_norm": 1.375,
      "learning_rate": 0.000497206087645401,
      "loss": 5.5688,
      "mean_token_accuracy": 0.1702422946691513,
      "num_tokens": 13794045.0,
      "step": 6030
    },
    {
      "entropy": 5.793889379501342,
      "epoch": 0.5965796757611704,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004972005408363894,
      "loss": 5.6123,
      "mean_token_accuracy": 0.16738815009593963,
      "num_tokens": 13805629.0,
      "step": 6035
    },
    {
      "entropy": 5.729014110565186,
      "epoch": 0.5970739422696718,
      "grad_norm": 1.375,
      "learning_rate": 0.0004971949885612102,
      "loss": 5.725,
      "mean_token_accuracy": 0.15540962368249894,
      "num_tokens": 13817526.0,
      "step": 6040
    },
    {
      "entropy": 5.79081506729126,
      "epoch": 0.5975682087781732,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004971894308199997,
      "loss": 5.6337,
      "mean_token_accuracy": 0.16940680146217346,
      "num_tokens": 13828553.0,
      "step": 6045
    },
    {
      "entropy": 5.802762985229492,
      "epoch": 0.5980624752866746,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004971838676128947,
      "loss": 5.6454,
      "mean_token_accuracy": 0.16584882736206055,
      "num_tokens": 13840323.0,
      "step": 6050
    },
    {
      "entropy": 5.723171710968018,
      "epoch": 0.5985567417951759,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004971782989400323,
      "loss": 5.6337,
      "mean_token_accuracy": 0.16908691674470902,
      "num_tokens": 13853524.0,
      "step": 6055
    },
    {
      "entropy": 5.715406084060669,
      "epoch": 0.5990510083036773,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004971727248015491,
      "loss": 5.604,
      "mean_token_accuracy": 0.1647974133491516,
      "num_tokens": 13864462.0,
      "step": 6060
    },
    {
      "entropy": 5.7595452785491945,
      "epoch": 0.5995452748121787,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004971671451975823,
      "loss": 5.6324,
      "mean_token_accuracy": 0.16867476999759673,
      "num_tokens": 13876615.0,
      "step": 6065
    },
    {
      "entropy": 5.753623008728027,
      "epoch": 0.6000395413206802,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004971615601282695,
      "loss": 5.7443,
      "mean_token_accuracy": 0.15899960994720458,
      "num_tokens": 13887652.0,
      "step": 6070
    },
    {
      "entropy": 5.7975993156433105,
      "epoch": 0.6005338078291815,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004971559695937477,
      "loss": 5.6126,
      "mean_token_accuracy": 0.16469364762306213,
      "num_tokens": 13898961.0,
      "step": 6075
    },
    {
      "entropy": 5.841239309310913,
      "epoch": 0.6010280743376829,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004971503735941547,
      "loss": 5.687,
      "mean_token_accuracy": 0.165712633728981,
      "num_tokens": 13909904.0,
      "step": 6080
    },
    {
      "entropy": 5.8075621128082275,
      "epoch": 0.6015223408461843,
      "grad_norm": 1.390625,
      "learning_rate": 0.000497144772129628,
      "loss": 5.7012,
      "mean_token_accuracy": 0.15533842593431474,
      "num_tokens": 13921376.0,
      "step": 6085
    },
    {
      "entropy": 5.714434671401977,
      "epoch": 0.6020166073546857,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004971391652003054,
      "loss": 5.6479,
      "mean_token_accuracy": 0.1701917678117752,
      "num_tokens": 13932925.0,
      "step": 6090
    },
    {
      "entropy": 5.8097669124603275,
      "epoch": 0.602510873863187,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000497133552806325,
      "loss": 5.6328,
      "mean_token_accuracy": 0.16505085527896882,
      "num_tokens": 13945010.0,
      "step": 6095
    },
    {
      "entropy": 5.7640375137329105,
      "epoch": 0.6030051403716884,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004971279349478245,
      "loss": 5.6537,
      "mean_token_accuracy": 0.1582181602716446,
      "num_tokens": 13957710.0,
      "step": 6100
    },
    {
      "entropy": 5.662689304351806,
      "epoch": 0.6034994068801898,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004971223116249424,
      "loss": 5.5516,
      "mean_token_accuracy": 0.17787012457847595,
      "num_tokens": 13967819.0,
      "step": 6105
    },
    {
      "entropy": 5.861082077026367,
      "epoch": 0.6039936733886911,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004971166828378172,
      "loss": 5.8144,
      "mean_token_accuracy": 0.15251778960227966,
      "num_tokens": 13980342.0,
      "step": 6110
    },
    {
      "entropy": 5.833240985870361,
      "epoch": 0.6044879398971925,
      "grad_norm": 1.359375,
      "learning_rate": 0.000497111048586587,
      "loss": 5.6901,
      "mean_token_accuracy": 0.16436685621738434,
      "num_tokens": 13990370.0,
      "step": 6115
    },
    {
      "entropy": 5.7875110626220705,
      "epoch": 0.604982206405694,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004971054088713905,
      "loss": 5.6878,
      "mean_token_accuracy": 0.16689518690109253,
      "num_tokens": 14001761.0,
      "step": 6120
    },
    {
      "entropy": 5.755599641799927,
      "epoch": 0.6054764729141954,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004970997636923665,
      "loss": 5.6926,
      "mean_token_accuracy": 0.16299993991851808,
      "num_tokens": 14012980.0,
      "step": 6125
    },
    {
      "entropy": 5.748826742172241,
      "epoch": 0.6059707394226967,
      "grad_norm": 1.484375,
      "learning_rate": 0.000497094113049654,
      "loss": 5.5902,
      "mean_token_accuracy": 0.16267894804477692,
      "num_tokens": 14023609.0,
      "step": 6130
    },
    {
      "entropy": 5.838775253295898,
      "epoch": 0.6064650059311981,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004970884569433917,
      "loss": 5.7649,
      "mean_token_accuracy": 0.15493567585945128,
      "num_tokens": 14036052.0,
      "step": 6135
    },
    {
      "entropy": 5.775563335418701,
      "epoch": 0.6069592724396995,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000497082795373719,
      "loss": 5.5995,
      "mean_token_accuracy": 0.1711900442838669,
      "num_tokens": 14047932.0,
      "step": 6140
    },
    {
      "entropy": 5.798108720779419,
      "epoch": 0.6074535389482009,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004970771283407749,
      "loss": 5.6819,
      "mean_token_accuracy": 0.1644543319940567,
      "num_tokens": 14059674.0,
      "step": 6145
    },
    {
      "entropy": 5.764031267166137,
      "epoch": 0.6079478054567022,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004970714558446992,
      "loss": 5.6969,
      "mean_token_accuracy": 0.16591319292783738,
      "num_tokens": 14070753.0,
      "step": 6150
    },
    {
      "entropy": 5.776700687408447,
      "epoch": 0.6084420719652036,
      "grad_norm": 3.609375,
      "learning_rate": 0.0004970657778856312,
      "loss": 5.6721,
      "mean_token_accuracy": 0.16524501740932465,
      "num_tokens": 14083565.0,
      "step": 6155
    },
    {
      "entropy": 5.9135137557983395,
      "epoch": 0.608936338473705,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004970600944637106,
      "loss": 5.7546,
      "mean_token_accuracy": 0.15822430551052094,
      "num_tokens": 14095008.0,
      "step": 6160
    },
    {
      "entropy": 5.671637296676636,
      "epoch": 0.6094306049822064,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004970544055790772,
      "loss": 5.5445,
      "mean_token_accuracy": 0.16504374593496324,
      "num_tokens": 14106586.0,
      "step": 6165
    },
    {
      "entropy": 5.657175731658936,
      "epoch": 0.6099248714907078,
      "grad_norm": 1.46875,
      "learning_rate": 0.000497048711231871,
      "loss": 5.5317,
      "mean_token_accuracy": 0.17381640821695327,
      "num_tokens": 14117778.0,
      "step": 6170
    },
    {
      "entropy": 5.707796049118042,
      "epoch": 0.6104191379992092,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004970430114222319,
      "loss": 5.6119,
      "mean_token_accuracy": 0.16493278443813325,
      "num_tokens": 14129644.0,
      "step": 6175
    },
    {
      "entropy": 5.6628653049469,
      "epoch": 0.6109134045077106,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004970373061503005,
      "loss": 5.5874,
      "mean_token_accuracy": 0.1685387223958969,
      "num_tokens": 14141230.0,
      "step": 6180
    },
    {
      "entropy": 5.737487840652466,
      "epoch": 0.611407671016212,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004970315954162169,
      "loss": 5.7221,
      "mean_token_accuracy": 0.16381432712078095,
      "num_tokens": 14151622.0,
      "step": 6185
    },
    {
      "entropy": 5.814573860168457,
      "epoch": 0.6119019375247133,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004970258792201214,
      "loss": 5.6591,
      "mean_token_accuracy": 0.16623520851135254,
      "num_tokens": 14163641.0,
      "step": 6190
    },
    {
      "entropy": 5.805066299438477,
      "epoch": 0.6123962040332147,
      "grad_norm": 1.328125,
      "learning_rate": 0.000497020157562155,
      "loss": 5.6116,
      "mean_token_accuracy": 0.1692408263683319,
      "num_tokens": 14175154.0,
      "step": 6195
    },
    {
      "entropy": 5.788309717178345,
      "epoch": 0.6128904705417161,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004970144304424582,
      "loss": 5.6444,
      "mean_token_accuracy": 0.1643928676843643,
      "num_tokens": 14185472.0,
      "step": 6200
    },
    {
      "entropy": 5.754764270782471,
      "epoch": 0.6133847370502175,
      "grad_norm": 2.046875,
      "learning_rate": 0.000497008697861172,
      "loss": 5.6203,
      "mean_token_accuracy": 0.16751834899187087,
      "num_tokens": 14196992.0,
      "step": 6205
    },
    {
      "entropy": 5.700463199615479,
      "epoch": 0.6138790035587188,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004970029598184375,
      "loss": 5.6121,
      "mean_token_accuracy": 0.1657865524291992,
      "num_tokens": 14207732.0,
      "step": 6210
    },
    {
      "entropy": 5.877051067352295,
      "epoch": 0.6143732700672202,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004969972163143956,
      "loss": 5.7594,
      "mean_token_accuracy": 0.15141110718250275,
      "num_tokens": 14219164.0,
      "step": 6215
    },
    {
      "entropy": 5.8117677688598635,
      "epoch": 0.6148675365757216,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004969914673491878,
      "loss": 5.7159,
      "mean_token_accuracy": 0.1557608500123024,
      "num_tokens": 14231687.0,
      "step": 6220
    },
    {
      "entropy": 5.775439739227295,
      "epoch": 0.6153618030842231,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004969857129229555,
      "loss": 5.5532,
      "mean_token_accuracy": 0.16914770603179932,
      "num_tokens": 14243099.0,
      "step": 6225
    },
    {
      "entropy": 5.804778575897217,
      "epoch": 0.6158560695927244,
      "grad_norm": 1.5,
      "learning_rate": 0.0004969799530358402,
      "loss": 5.7586,
      "mean_token_accuracy": 0.15600511133670808,
      "num_tokens": 14255835.0,
      "step": 6230
    },
    {
      "entropy": 5.7494730949401855,
      "epoch": 0.6163503361012258,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004969741876879835,
      "loss": 5.5926,
      "mean_token_accuracy": 0.16058395355939864,
      "num_tokens": 14266661.0,
      "step": 6235
    },
    {
      "entropy": 5.7279599666595455,
      "epoch": 0.6168446026097272,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004969684168795275,
      "loss": 5.6731,
      "mean_token_accuracy": 0.1628929853439331,
      "num_tokens": 14278474.0,
      "step": 6240
    },
    {
      "entropy": 5.822378301620484,
      "epoch": 0.6173388691182286,
      "grad_norm": 1.375,
      "learning_rate": 0.0004969626406106139,
      "loss": 5.6425,
      "mean_token_accuracy": 0.16124500781297685,
      "num_tokens": 14289516.0,
      "step": 6245
    },
    {
      "entropy": 5.696770429611206,
      "epoch": 0.6178331356267299,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004969568588813851,
      "loss": 5.6503,
      "mean_token_accuracy": 0.16988480165600778,
      "num_tokens": 14301312.0,
      "step": 6250
    },
    {
      "entropy": 5.776571321487427,
      "epoch": 0.6183274021352313,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004969510716919829,
      "loss": 5.6151,
      "mean_token_accuracy": 0.17066278606653212,
      "num_tokens": 14312009.0,
      "step": 6255
    },
    {
      "entropy": 5.770249891281128,
      "epoch": 0.6188216686437327,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00049694527904255,
      "loss": 5.6573,
      "mean_token_accuracy": 0.17074283510446547,
      "num_tokens": 14322754.0,
      "step": 6260
    },
    {
      "entropy": 5.6543920040130615,
      "epoch": 0.619315935152234,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004969394809332287,
      "loss": 5.5159,
      "mean_token_accuracy": 0.1669551819562912,
      "num_tokens": 14333939.0,
      "step": 6265
    },
    {
      "entropy": 5.746486043930053,
      "epoch": 0.6198102016607354,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004969336773641619,
      "loss": 5.6323,
      "mean_token_accuracy": 0.16919056326150894,
      "num_tokens": 14345987.0,
      "step": 6270
    },
    {
      "entropy": 5.81604266166687,
      "epoch": 0.6203044681692369,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004969278683354921,
      "loss": 5.6322,
      "mean_token_accuracy": 0.1565315842628479,
      "num_tokens": 14356974.0,
      "step": 6275
    },
    {
      "entropy": 5.780891370773316,
      "epoch": 0.6207987346777383,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004969220538473625,
      "loss": 5.6569,
      "mean_token_accuracy": 0.15674652606248857,
      "num_tokens": 14368175.0,
      "step": 6280
    },
    {
      "entropy": 5.7501989841461185,
      "epoch": 0.6212930011862396,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004969162338999157,
      "loss": 5.7288,
      "mean_token_accuracy": 0.16223904192447663,
      "num_tokens": 14379067.0,
      "step": 6285
    },
    {
      "entropy": 5.777649879455566,
      "epoch": 0.621787267694741,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004969104084932953,
      "loss": 5.676,
      "mean_token_accuracy": 0.1620064541697502,
      "num_tokens": 14390483.0,
      "step": 6290
    },
    {
      "entropy": 5.730044412612915,
      "epoch": 0.6222815342032424,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004969045776276443,
      "loss": 5.6918,
      "mean_token_accuracy": 0.16353509873151778,
      "num_tokens": 14402757.0,
      "step": 6295
    },
    {
      "entropy": 5.731755065917969,
      "epoch": 0.6227758007117438,
      "grad_norm": 1.75,
      "learning_rate": 0.0004968987413031063,
      "loss": 5.6327,
      "mean_token_accuracy": 0.16243165880441665,
      "num_tokens": 14413829.0,
      "step": 6300
    },
    {
      "entropy": 5.767355728149414,
      "epoch": 0.6232700672202451,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004968928995198249,
      "loss": 5.7327,
      "mean_token_accuracy": 0.1529085859656334,
      "num_tokens": 14425165.0,
      "step": 6305
    },
    {
      "entropy": 5.852034854888916,
      "epoch": 0.6237643337287465,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004968870522779437,
      "loss": 5.6647,
      "mean_token_accuracy": 0.1615695297718048,
      "num_tokens": 14436264.0,
      "step": 6310
    },
    {
      "entropy": 5.729472541809082,
      "epoch": 0.6242586002372479,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004968811995776066,
      "loss": 5.6122,
      "mean_token_accuracy": 0.1666173070669174,
      "num_tokens": 14447384.0,
      "step": 6315
    },
    {
      "entropy": 5.7991142749786375,
      "epoch": 0.6247528667457493,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004968753414189576,
      "loss": 5.7192,
      "mean_token_accuracy": 0.15756288915872574,
      "num_tokens": 14460013.0,
      "step": 6320
    },
    {
      "entropy": 5.823254251480103,
      "epoch": 0.6252471332542507,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004968694778021408,
      "loss": 5.608,
      "mean_token_accuracy": 0.16636418253183366,
      "num_tokens": 14472452.0,
      "step": 6325
    },
    {
      "entropy": 5.7357104301452635,
      "epoch": 0.6257413997627521,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004968636087273003,
      "loss": 5.651,
      "mean_token_accuracy": 0.16405002474784852,
      "num_tokens": 14484487.0,
      "step": 6330
    },
    {
      "entropy": 5.737237119674683,
      "epoch": 0.6262356662712535,
      "grad_norm": 1.5,
      "learning_rate": 0.0004968577341945806,
      "loss": 5.5763,
      "mean_token_accuracy": 0.1684148997068405,
      "num_tokens": 14496043.0,
      "step": 6335
    },
    {
      "entropy": 5.790759181976318,
      "epoch": 0.6267299327797549,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004968518542041264,
      "loss": 5.6792,
      "mean_token_accuracy": 0.16304038614034652,
      "num_tokens": 14508089.0,
      "step": 6340
    },
    {
      "entropy": 5.8391822338104244,
      "epoch": 0.6272241992882562,
      "grad_norm": 1.7109375,
      "learning_rate": 0.000496845968756082,
      "loss": 5.7369,
      "mean_token_accuracy": 0.15663580745458602,
      "num_tokens": 14519480.0,
      "step": 6345
    },
    {
      "entropy": 5.820170545578003,
      "epoch": 0.6277184657967576,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004968400778505925,
      "loss": 5.6587,
      "mean_token_accuracy": 0.17018893361091614,
      "num_tokens": 14531214.0,
      "step": 6350
    },
    {
      "entropy": 5.8212890625,
      "epoch": 0.628212732305259,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004968341814878025,
      "loss": 5.7239,
      "mean_token_accuracy": 0.16491053253412247,
      "num_tokens": 14544118.0,
      "step": 6355
    },
    {
      "entropy": 5.784654712677002,
      "epoch": 0.6287069988137604,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004968282796678573,
      "loss": 5.6231,
      "mean_token_accuracy": 0.163566854596138,
      "num_tokens": 14555545.0,
      "step": 6360
    },
    {
      "entropy": 5.740172290802002,
      "epoch": 0.6292012653222617,
      "grad_norm": 1.40625,
      "learning_rate": 0.000496822372390902,
      "loss": 5.6403,
      "mean_token_accuracy": 0.16461472660303117,
      "num_tokens": 14567157.0,
      "step": 6365
    },
    {
      "entropy": 5.730631160736084,
      "epoch": 0.6296955318307631,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000496816459657082,
      "loss": 5.6034,
      "mean_token_accuracy": 0.16728046983480455,
      "num_tokens": 14577074.0,
      "step": 6370
    },
    {
      "entropy": 5.794475650787353,
      "epoch": 0.6301897983392646,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004968105414665426,
      "loss": 5.6688,
      "mean_token_accuracy": 0.16565605998039246,
      "num_tokens": 14589133.0,
      "step": 6375
    },
    {
      "entropy": 5.771366834640503,
      "epoch": 0.630684064847766,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004968046178194295,
      "loss": 5.6355,
      "mean_token_accuracy": 0.1643053725361824,
      "num_tokens": 14600985.0,
      "step": 6380
    },
    {
      "entropy": 5.77326807975769,
      "epoch": 0.6311783313562673,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004967986887158884,
      "loss": 5.6097,
      "mean_token_accuracy": 0.16638877093791962,
      "num_tokens": 14612163.0,
      "step": 6385
    },
    {
      "entropy": 5.71264967918396,
      "epoch": 0.6316725978647687,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004967927541560651,
      "loss": 5.7184,
      "mean_token_accuracy": 0.16650869697332382,
      "num_tokens": 14623045.0,
      "step": 6390
    },
    {
      "entropy": 5.791554927825928,
      "epoch": 0.6321668643732701,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004967868141401056,
      "loss": 5.6893,
      "mean_token_accuracy": 0.16229858845472336,
      "num_tokens": 14634696.0,
      "step": 6395
    },
    {
      "entropy": 5.850910758972168,
      "epoch": 0.6326611308817714,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004967808686681562,
      "loss": 5.6875,
      "mean_token_accuracy": 0.16024835854768754,
      "num_tokens": 14645589.0,
      "step": 6400
    },
    {
      "entropy": 5.799806213378906,
      "epoch": 0.6331553973902728,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004967749177403629,
      "loss": 5.6237,
      "mean_token_accuracy": 0.16475450098514557,
      "num_tokens": 14657199.0,
      "step": 6405
    },
    {
      "entropy": 5.772099733352661,
      "epoch": 0.6336496638987742,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004967689613568723,
      "loss": 5.6801,
      "mean_token_accuracy": 0.159478160738945,
      "num_tokens": 14668767.0,
      "step": 6410
    },
    {
      "entropy": 5.73914680480957,
      "epoch": 0.6341439304072756,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004967629995178309,
      "loss": 5.5801,
      "mean_token_accuracy": 0.16919685602188111,
      "num_tokens": 14680860.0,
      "step": 6415
    },
    {
      "entropy": 5.6863480567932125,
      "epoch": 0.6346381969157769,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004967570322233852,
      "loss": 5.541,
      "mean_token_accuracy": 0.16810448914766313,
      "num_tokens": 14692134.0,
      "step": 6420
    },
    {
      "entropy": 5.723415756225586,
      "epoch": 0.6351324634242784,
      "grad_norm": 1.625,
      "learning_rate": 0.0004967510594736822,
      "loss": 5.6543,
      "mean_token_accuracy": 0.16126812547445296,
      "num_tokens": 14703323.0,
      "step": 6425
    },
    {
      "entropy": 5.722515678405761,
      "epoch": 0.6356267299327798,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004967450812688686,
      "loss": 5.5869,
      "mean_token_accuracy": 0.1672411799430847,
      "num_tokens": 14714453.0,
      "step": 6430
    },
    {
      "entropy": 5.736376619338989,
      "epoch": 0.6361209964412812,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004967390976090917,
      "loss": 5.6807,
      "mean_token_accuracy": 0.1590673267841339,
      "num_tokens": 14726562.0,
      "step": 6435
    },
    {
      "entropy": 5.687188625335693,
      "epoch": 0.6366152629497825,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004967331084944986,
      "loss": 5.5045,
      "mean_token_accuracy": 0.17426086068153382,
      "num_tokens": 14738054.0,
      "step": 6440
    },
    {
      "entropy": 5.642411756515503,
      "epoch": 0.6371095294582839,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004967271139252365,
      "loss": 5.6294,
      "mean_token_accuracy": 0.1622166484594345,
      "num_tokens": 14749037.0,
      "step": 6445
    },
    {
      "entropy": 5.775140428543091,
      "epoch": 0.6376037959667853,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004967211139014531,
      "loss": 5.5897,
      "mean_token_accuracy": 0.16244631856679917,
      "num_tokens": 14759667.0,
      "step": 6450
    },
    {
      "entropy": 5.7774248123168945,
      "epoch": 0.6380980624752867,
      "grad_norm": 1.9765625,
      "learning_rate": 0.000496715108423296,
      "loss": 5.6162,
      "mean_token_accuracy": 0.1629392221570015,
      "num_tokens": 14770866.0,
      "step": 6455
    },
    {
      "entropy": 5.754608488082885,
      "epoch": 0.638592328983788,
      "grad_norm": 1.875,
      "learning_rate": 0.0004967090974909127,
      "loss": 5.6312,
      "mean_token_accuracy": 0.16380554884672166,
      "num_tokens": 14783095.0,
      "step": 6460
    },
    {
      "entropy": 5.772108030319214,
      "epoch": 0.6390865954922894,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004967030811044512,
      "loss": 5.6992,
      "mean_token_accuracy": 0.15968350768089296,
      "num_tokens": 14794046.0,
      "step": 6465
    },
    {
      "entropy": 5.756714677810669,
      "epoch": 0.6395808620007908,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004966970592640596,
      "loss": 5.6757,
      "mean_token_accuracy": 0.16397481709718703,
      "num_tokens": 14807295.0,
      "step": 6470
    },
    {
      "entropy": 5.783436918258667,
      "epoch": 0.6400751285092922,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004966910319698859,
      "loss": 5.6379,
      "mean_token_accuracy": 0.1722213000059128,
      "num_tokens": 14819109.0,
      "step": 6475
    },
    {
      "entropy": 5.755461025238037,
      "epoch": 0.6405693950177936,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004966849992220784,
      "loss": 5.6022,
      "mean_token_accuracy": 0.16542187631130217,
      "num_tokens": 14830719.0,
      "step": 6480
    },
    {
      "entropy": 5.751820373535156,
      "epoch": 0.641063661526295,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004966789610207856,
      "loss": 5.6848,
      "mean_token_accuracy": 0.1639040231704712,
      "num_tokens": 14843392.0,
      "step": 6485
    },
    {
      "entropy": 5.740760231018067,
      "epoch": 0.6415579280347964,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004966729173661559,
      "loss": 5.5618,
      "mean_token_accuracy": 0.17255358248949051,
      "num_tokens": 14855745.0,
      "step": 6490
    },
    {
      "entropy": 5.813320350646973,
      "epoch": 0.6420521945432978,
      "grad_norm": 1.625,
      "learning_rate": 0.000496666868258338,
      "loss": 5.7094,
      "mean_token_accuracy": 0.15964556038379668,
      "num_tokens": 14868053.0,
      "step": 6495
    },
    {
      "entropy": 5.76065993309021,
      "epoch": 0.6425464610517991,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004966608136974809,
      "loss": 5.6266,
      "mean_token_accuracy": 0.16805339008569717,
      "num_tokens": 14880095.0,
      "step": 6500
    },
    {
      "entropy": 5.819428634643555,
      "epoch": 0.6430407275603005,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004966547536837333,
      "loss": 5.6923,
      "mean_token_accuracy": 0.15894569158554078,
      "num_tokens": 14891013.0,
      "step": 6505
    },
    {
      "entropy": 5.784639310836792,
      "epoch": 0.6435349940688019,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004966486882172444,
      "loss": 5.6024,
      "mean_token_accuracy": 0.16922138929367064,
      "num_tokens": 14901875.0,
      "step": 6510
    },
    {
      "entropy": 5.730242681503296,
      "epoch": 0.6440292605773033,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004966426172981634,
      "loss": 5.6333,
      "mean_token_accuracy": 0.16815225183963775,
      "num_tokens": 14912620.0,
      "step": 6515
    },
    {
      "entropy": 5.685377740859986,
      "epoch": 0.6445235270858046,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004966365409266396,
      "loss": 5.6164,
      "mean_token_accuracy": 0.1710914969444275,
      "num_tokens": 14923128.0,
      "step": 6520
    },
    {
      "entropy": 5.724513721466065,
      "epoch": 0.645017793594306,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004966304591028225,
      "loss": 5.6079,
      "mean_token_accuracy": 0.16572261601686478,
      "num_tokens": 14933621.0,
      "step": 6525
    },
    {
      "entropy": 5.730213737487793,
      "epoch": 0.6455120601028075,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004966243718268617,
      "loss": 5.569,
      "mean_token_accuracy": 0.16851237118244172,
      "num_tokens": 14943292.0,
      "step": 6530
    },
    {
      "entropy": 5.706056022644043,
      "epoch": 0.6460063266113089,
      "grad_norm": 1.421875,
      "learning_rate": 0.000496618279098907,
      "loss": 5.6307,
      "mean_token_accuracy": 0.16001626700162888,
      "num_tokens": 14953998.0,
      "step": 6535
    },
    {
      "entropy": 5.784326171875,
      "epoch": 0.6465005931198102,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004966121809191083,
      "loss": 5.6679,
      "mean_token_accuracy": 0.16561168134212495,
      "num_tokens": 14965430.0,
      "step": 6540
    },
    {
      "entropy": 5.74091362953186,
      "epoch": 0.6469948596283116,
      "grad_norm": 1.875,
      "learning_rate": 0.0004966060772876156,
      "loss": 5.5362,
      "mean_token_accuracy": 0.17125169187784195,
      "num_tokens": 14977124.0,
      "step": 6545
    },
    {
      "entropy": 5.742420101165772,
      "epoch": 0.647489126136813,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004965999682045788,
      "loss": 5.5547,
      "mean_token_accuracy": 0.1699445702135563,
      "num_tokens": 14987069.0,
      "step": 6550
    },
    {
      "entropy": 5.677365732192993,
      "epoch": 0.6479833926453143,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004965938536701486,
      "loss": 5.6403,
      "mean_token_accuracy": 0.17897330820560456,
      "num_tokens": 14997919.0,
      "step": 6555
    },
    {
      "entropy": 5.7578435897827145,
      "epoch": 0.6484776591538157,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004965877336844752,
      "loss": 5.6947,
      "mean_token_accuracy": 0.16872752606868743,
      "num_tokens": 15010042.0,
      "step": 6560
    },
    {
      "entropy": 5.809809398651123,
      "epoch": 0.6489719256623171,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004965816082477091,
      "loss": 5.561,
      "mean_token_accuracy": 0.16624810099601744,
      "num_tokens": 15021166.0,
      "step": 6565
    },
    {
      "entropy": 5.805764436721802,
      "epoch": 0.6494661921708185,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004965754773600011,
      "loss": 5.6452,
      "mean_token_accuracy": 0.16420420557260512,
      "num_tokens": 15032182.0,
      "step": 6570
    },
    {
      "entropy": 5.6941694736480715,
      "epoch": 0.6499604586793198,
      "grad_norm": 1.8125,
      "learning_rate": 0.000496569341021502,
      "loss": 5.5943,
      "mean_token_accuracy": 0.16626092344522475,
      "num_tokens": 15043054.0,
      "step": 6575
    },
    {
      "entropy": 5.68024435043335,
      "epoch": 0.6504547251878213,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004965631992323628,
      "loss": 5.531,
      "mean_token_accuracy": 0.1773110330104828,
      "num_tokens": 15055462.0,
      "step": 6580
    },
    {
      "entropy": 5.748437976837158,
      "epoch": 0.6509489916963227,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004965570519927345,
      "loss": 5.6462,
      "mean_token_accuracy": 0.16381797343492507,
      "num_tokens": 15066849.0,
      "step": 6585
    },
    {
      "entropy": 5.7041933059692385,
      "epoch": 0.6514432582048241,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004965508993027682,
      "loss": 5.6044,
      "mean_token_accuracy": 0.16580681204795839,
      "num_tokens": 15078388.0,
      "step": 6590
    },
    {
      "entropy": 5.816247892379761,
      "epoch": 0.6519375247133254,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004965447411626156,
      "loss": 5.6364,
      "mean_token_accuracy": 0.16688652336597443,
      "num_tokens": 15090474.0,
      "step": 6595
    },
    {
      "entropy": 5.815159559249878,
      "epoch": 0.6524317912218268,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004965385775724279,
      "loss": 5.6316,
      "mean_token_accuracy": 0.16053054630756378,
      "num_tokens": 15101104.0,
      "step": 6600
    },
    {
      "entropy": 5.749491405487061,
      "epoch": 0.6529260577303282,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004965324085323569,
      "loss": 5.6476,
      "mean_token_accuracy": 0.15748330056667328,
      "num_tokens": 15112036.0,
      "step": 6605
    },
    {
      "entropy": 5.803199338912964,
      "epoch": 0.6534203242388296,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004965262340425541,
      "loss": 5.7597,
      "mean_token_accuracy": 0.1529676854610443,
      "num_tokens": 15124913.0,
      "step": 6610
    },
    {
      "entropy": 5.871570348739624,
      "epoch": 0.6539145907473309,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004965200541031717,
      "loss": 5.6563,
      "mean_token_accuracy": 0.1651740312576294,
      "num_tokens": 15135555.0,
      "step": 6615
    },
    {
      "entropy": 5.859944820404053,
      "epoch": 0.6544088572558323,
      "grad_norm": 1.5,
      "learning_rate": 0.0004965138687143616,
      "loss": 5.7435,
      "mean_token_accuracy": 0.15558698326349257,
      "num_tokens": 15146646.0,
      "step": 6620
    },
    {
      "entropy": 5.7530012130737305,
      "epoch": 0.6549031237643337,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004965076778762759,
      "loss": 5.6074,
      "mean_token_accuracy": 0.16277364492416382,
      "num_tokens": 15157101.0,
      "step": 6625
    },
    {
      "entropy": 5.797142171859742,
      "epoch": 0.6553973902728352,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004965014815890671,
      "loss": 5.6227,
      "mean_token_accuracy": 0.16162085384130478,
      "num_tokens": 15169645.0,
      "step": 6630
    },
    {
      "entropy": 5.7598498344421385,
      "epoch": 0.6558916567813365,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004964952798528874,
      "loss": 5.5373,
      "mean_token_accuracy": 0.1664197862148285,
      "num_tokens": 15181018.0,
      "step": 6635
    },
    {
      "entropy": 5.747228479385376,
      "epoch": 0.6563859232898379,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004964890726678894,
      "loss": 5.6154,
      "mean_token_accuracy": 0.1631882131099701,
      "num_tokens": 15191880.0,
      "step": 6640
    },
    {
      "entropy": 5.683012437820435,
      "epoch": 0.6568801897983393,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004964828600342259,
      "loss": 5.5831,
      "mean_token_accuracy": 0.16633551120758056,
      "num_tokens": 15203886.0,
      "step": 6645
    },
    {
      "entropy": 5.758241844177246,
      "epoch": 0.6573744563068407,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004964766419520497,
      "loss": 5.571,
      "mean_token_accuracy": 0.1714767888188362,
      "num_tokens": 15215837.0,
      "step": 6650
    },
    {
      "entropy": 5.789406251907349,
      "epoch": 0.657868722815342,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004964704184215138,
      "loss": 5.6685,
      "mean_token_accuracy": 0.1581181839108467,
      "num_tokens": 15226833.0,
      "step": 6655
    },
    {
      "entropy": 5.738519811630249,
      "epoch": 0.6583629893238434,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004964641894427712,
      "loss": 5.7133,
      "mean_token_accuracy": 0.15806318670511246,
      "num_tokens": 15239312.0,
      "step": 6660
    },
    {
      "entropy": 5.788010501861573,
      "epoch": 0.6588572558323448,
      "grad_norm": 1.5,
      "learning_rate": 0.0004964579550159752,
      "loss": 5.5938,
      "mean_token_accuracy": 0.1595676064491272,
      "num_tokens": 15250348.0,
      "step": 6665
    },
    {
      "entropy": 5.707490110397339,
      "epoch": 0.6593515223408462,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004964517151412791,
      "loss": 5.654,
      "mean_token_accuracy": 0.16972415298223495,
      "num_tokens": 15261445.0,
      "step": 6670
    },
    {
      "entropy": 5.780685329437256,
      "epoch": 0.6598457888493475,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004964454698188364,
      "loss": 5.5768,
      "mean_token_accuracy": 0.16413053274154663,
      "num_tokens": 15271606.0,
      "step": 6675
    },
    {
      "entropy": 5.711971473693848,
      "epoch": 0.6603400553578489,
      "grad_norm": 1.5,
      "learning_rate": 0.000496439219048801,
      "loss": 5.6486,
      "mean_token_accuracy": 0.16437918245792388,
      "num_tokens": 15283690.0,
      "step": 6680
    },
    {
      "entropy": 5.715126705169678,
      "epoch": 0.6608343218663504,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004964329628313263,
      "loss": 5.6089,
      "mean_token_accuracy": 0.17137991935014724,
      "num_tokens": 15294792.0,
      "step": 6685
    },
    {
      "entropy": 5.808073616027832,
      "epoch": 0.6613285883748518,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004964267011665664,
      "loss": 5.6457,
      "mean_token_accuracy": 0.1627802923321724,
      "num_tokens": 15305636.0,
      "step": 6690
    },
    {
      "entropy": 5.7131470203399655,
      "epoch": 0.6618228548833531,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0004964204340546753,
      "loss": 5.6345,
      "mean_token_accuracy": 0.1627189040184021,
      "num_tokens": 15316495.0,
      "step": 6695
    },
    {
      "entropy": 5.721789312362671,
      "epoch": 0.6623171213918545,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004964141614958071,
      "loss": 5.6201,
      "mean_token_accuracy": 0.16983213126659394,
      "num_tokens": 15328246.0,
      "step": 6700
    },
    {
      "entropy": 5.816872072219849,
      "epoch": 0.6628113879003559,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004964078834901163,
      "loss": 5.5891,
      "mean_token_accuracy": 0.16585003212094307,
      "num_tokens": 15339229.0,
      "step": 6705
    },
    {
      "entropy": 5.7035352230072025,
      "epoch": 0.6633056544088572,
      "grad_norm": 1.640625,
      "learning_rate": 0.000496401600037757,
      "loss": 5.6099,
      "mean_token_accuracy": 0.164272303879261,
      "num_tokens": 15350591.0,
      "step": 6710
    },
    {
      "entropy": 5.812136697769165,
      "epoch": 0.6637999209173586,
      "grad_norm": 1.625,
      "learning_rate": 0.0004963953111388842,
      "loss": 5.7277,
      "mean_token_accuracy": 0.1567153736948967,
      "num_tokens": 15363726.0,
      "step": 6715
    },
    {
      "entropy": 5.7724522113800045,
      "epoch": 0.66429418742586,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004963890167936524,
      "loss": 5.6112,
      "mean_token_accuracy": 0.16536572575569153,
      "num_tokens": 15375569.0,
      "step": 6720
    },
    {
      "entropy": 5.748994970321656,
      "epoch": 0.6647884539343614,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004963827170022164,
      "loss": 5.6863,
      "mean_token_accuracy": 0.1543663889169693,
      "num_tokens": 15387323.0,
      "step": 6725
    },
    {
      "entropy": 5.7563999652862545,
      "epoch": 0.6652827204428627,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004963764117647311,
      "loss": 5.5605,
      "mean_token_accuracy": 0.15932124704122544,
      "num_tokens": 15397912.0,
      "step": 6730
    },
    {
      "entropy": 5.691759538650513,
      "epoch": 0.6657769869513642,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004963701010813519,
      "loss": 5.4786,
      "mean_token_accuracy": 0.1785081997513771,
      "num_tokens": 15409469.0,
      "step": 6735
    },
    {
      "entropy": 5.6755051612854,
      "epoch": 0.6662712534598656,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004963637849522338,
      "loss": 5.5229,
      "mean_token_accuracy": 0.16746993511915206,
      "num_tokens": 15419681.0,
      "step": 6740
    },
    {
      "entropy": 5.778180408477783,
      "epoch": 0.666765519968367,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004963574633775323,
      "loss": 5.7094,
      "mean_token_accuracy": 0.1608496442437172,
      "num_tokens": 15431237.0,
      "step": 6745
    },
    {
      "entropy": 5.766178703308105,
      "epoch": 0.6672597864768683,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004963511363574027,
      "loss": 5.6216,
      "mean_token_accuracy": 0.16089110523462297,
      "num_tokens": 15442366.0,
      "step": 6750
    },
    {
      "entropy": 5.740045976638794,
      "epoch": 0.6677540529853697,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004963448038920009,
      "loss": 5.6081,
      "mean_token_accuracy": 0.15921292304992676,
      "num_tokens": 15453121.0,
      "step": 6755
    },
    {
      "entropy": 5.734017372131348,
      "epoch": 0.6682483194938711,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004963384659814827,
      "loss": 5.5849,
      "mean_token_accuracy": 0.17104532420635224,
      "num_tokens": 15463489.0,
      "step": 6760
    },
    {
      "entropy": 5.7487220287323,
      "epoch": 0.6687425860023725,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004963321226260038,
      "loss": 5.6448,
      "mean_token_accuracy": 0.16038959920406343,
      "num_tokens": 15475217.0,
      "step": 6765
    },
    {
      "entropy": 5.736399412155151,
      "epoch": 0.6692368525108738,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004963257738257203,
      "loss": 5.6264,
      "mean_token_accuracy": 0.16269309297204018,
      "num_tokens": 15485844.0,
      "step": 6770
    },
    {
      "entropy": 5.7635212421417235,
      "epoch": 0.6697311190193752,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004963194195807886,
      "loss": 5.6147,
      "mean_token_accuracy": 0.16412168890237808,
      "num_tokens": 15496835.0,
      "step": 6775
    },
    {
      "entropy": 5.7644247055053714,
      "epoch": 0.6702253855278766,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004963130598913646,
      "loss": 5.6523,
      "mean_token_accuracy": 0.1657743513584137,
      "num_tokens": 15508087.0,
      "step": 6780
    },
    {
      "entropy": 5.7958824157714846,
      "epoch": 0.6707196520363781,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004963066947576051,
      "loss": 5.6557,
      "mean_token_accuracy": 0.15826882794499397,
      "num_tokens": 15519596.0,
      "step": 6785
    },
    {
      "entropy": 5.780778408050537,
      "epoch": 0.6712139185448794,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004963003241796666,
      "loss": 5.6412,
      "mean_token_accuracy": 0.1555856592953205,
      "num_tokens": 15530830.0,
      "step": 6790
    },
    {
      "entropy": 5.706460666656494,
      "epoch": 0.6717081850533808,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004962939481577059,
      "loss": 5.5405,
      "mean_token_accuracy": 0.17752100974321366,
      "num_tokens": 15543195.0,
      "step": 6795
    },
    {
      "entropy": 5.749172973632812,
      "epoch": 0.6722024515618822,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004962875666918796,
      "loss": 5.6553,
      "mean_token_accuracy": 0.1612786501646042,
      "num_tokens": 15554465.0,
      "step": 6800
    },
    {
      "entropy": 5.6718591213226315,
      "epoch": 0.6726967180703836,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004962811797823448,
      "loss": 5.6188,
      "mean_token_accuracy": 0.1627206638455391,
      "num_tokens": 15566016.0,
      "step": 6805
    },
    {
      "entropy": 5.799967336654663,
      "epoch": 0.6731909845788849,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004962747874292588,
      "loss": 5.7041,
      "mean_token_accuracy": 0.15960169285535813,
      "num_tokens": 15576843.0,
      "step": 6810
    },
    {
      "entropy": 5.840959692001343,
      "epoch": 0.6736852510873863,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004962683896327786,
      "loss": 5.631,
      "mean_token_accuracy": 0.1652127519249916,
      "num_tokens": 15588390.0,
      "step": 6815
    },
    {
      "entropy": 5.829801368713379,
      "epoch": 0.6741795175958877,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004962619863930616,
      "loss": 5.7312,
      "mean_token_accuracy": 0.1608808308839798,
      "num_tokens": 15600270.0,
      "step": 6820
    },
    {
      "entropy": 5.735422515869141,
      "epoch": 0.674673784104389,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004962555777102656,
      "loss": 5.5721,
      "mean_token_accuracy": 0.16506532430648804,
      "num_tokens": 15611617.0,
      "step": 6825
    },
    {
      "entropy": 5.707896947860718,
      "epoch": 0.6751680506128904,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000496249163584548,
      "loss": 5.6483,
      "mean_token_accuracy": 0.16650721281766892,
      "num_tokens": 15622128.0,
      "step": 6830
    },
    {
      "entropy": 5.843118000030517,
      "epoch": 0.6756623171213919,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004962427440160665,
      "loss": 5.658,
      "mean_token_accuracy": 0.15920696258544922,
      "num_tokens": 15635250.0,
      "step": 6835
    },
    {
      "entropy": 5.830879068374633,
      "epoch": 0.6761565836298933,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004962363190049794,
      "loss": 5.7233,
      "mean_token_accuracy": 0.15867377519607545,
      "num_tokens": 15646535.0,
      "step": 6840
    },
    {
      "entropy": 5.755466175079346,
      "epoch": 0.6766508501383947,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004962298885514444,
      "loss": 5.6739,
      "mean_token_accuracy": 0.16584515273571016,
      "num_tokens": 15658399.0,
      "step": 6845
    },
    {
      "entropy": 5.796150159835816,
      "epoch": 0.677145116646896,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004962234526556199,
      "loss": 5.6215,
      "mean_token_accuracy": 0.1704525589942932,
      "num_tokens": 15669321.0,
      "step": 6850
    },
    {
      "entropy": 5.691433572769165,
      "epoch": 0.6776393831553974,
      "grad_norm": 1.7421875,
      "learning_rate": 0.000496217011317664,
      "loss": 5.5963,
      "mean_token_accuracy": 0.16430669128894806,
      "num_tokens": 15681666.0,
      "step": 6855
    },
    {
      "entropy": 5.666491460800171,
      "epoch": 0.6781336496638988,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004962105645377354,
      "loss": 5.6065,
      "mean_token_accuracy": 0.16990689039230347,
      "num_tokens": 15693432.0,
      "step": 6860
    },
    {
      "entropy": 5.745400285720825,
      "epoch": 0.6786279161724001,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004962041123159926,
      "loss": 5.5414,
      "mean_token_accuracy": 0.16931298077106477,
      "num_tokens": 15705151.0,
      "step": 6865
    },
    {
      "entropy": 5.697288370132446,
      "epoch": 0.6791221826809015,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004961976546525943,
      "loss": 5.55,
      "mean_token_accuracy": 0.16901147365570068,
      "num_tokens": 15716549.0,
      "step": 6870
    },
    {
      "entropy": 5.756910037994385,
      "epoch": 0.6796164491894029,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004961911915476994,
      "loss": 5.6921,
      "mean_token_accuracy": 0.16362438648939132,
      "num_tokens": 15727499.0,
      "step": 6875
    },
    {
      "entropy": 5.732259511947632,
      "epoch": 0.6801107156979043,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004961847230014668,
      "loss": 5.5914,
      "mean_token_accuracy": 0.16787823289632797,
      "num_tokens": 15739196.0,
      "step": 6880
    },
    {
      "entropy": 5.740002965927124,
      "epoch": 0.6806049822064056,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004961782490140558,
      "loss": 5.5897,
      "mean_token_accuracy": 0.17234150767326356,
      "num_tokens": 15750890.0,
      "step": 6885
    },
    {
      "entropy": 5.747622394561768,
      "epoch": 0.6810992487149071,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004961717695856256,
      "loss": 5.6336,
      "mean_token_accuracy": 0.17156925797462463,
      "num_tokens": 15762366.0,
      "step": 6890
    },
    {
      "entropy": 5.716518783569336,
      "epoch": 0.6815935152234085,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004961652847163354,
      "loss": 5.6004,
      "mean_token_accuracy": 0.1655272826552391,
      "num_tokens": 15773962.0,
      "step": 6895
    },
    {
      "entropy": 5.626062822341919,
      "epoch": 0.6820877817319099,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000496158794406345,
      "loss": 5.4643,
      "mean_token_accuracy": 0.1829729974269867,
      "num_tokens": 15784563.0,
      "step": 6900
    },
    {
      "entropy": 5.7201660633087155,
      "epoch": 0.6825820482404112,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004961522986558139,
      "loss": 5.603,
      "mean_token_accuracy": 0.1636321648955345,
      "num_tokens": 15796461.0,
      "step": 6905
    },
    {
      "entropy": 5.754406833648682,
      "epoch": 0.6830763147489126,
      "grad_norm": 1.46875,
      "learning_rate": 0.000496145797464902,
      "loss": 5.6178,
      "mean_token_accuracy": 0.16484073251485826,
      "num_tokens": 15807303.0,
      "step": 6910
    },
    {
      "entropy": 5.676671648025513,
      "epoch": 0.683570581257414,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004961392908337691,
      "loss": 5.5537,
      "mean_token_accuracy": 0.1768821358680725,
      "num_tokens": 15818964.0,
      "step": 6915
    },
    {
      "entropy": 5.7280134677886965,
      "epoch": 0.6840648477659154,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004961327787625754,
      "loss": 5.7026,
      "mean_token_accuracy": 0.15737905353307724,
      "num_tokens": 15831181.0,
      "step": 6920
    },
    {
      "entropy": 5.7252861022949215,
      "epoch": 0.6845591142744167,
      "grad_norm": 1.40625,
      "learning_rate": 0.000496126261251481,
      "loss": 5.5628,
      "mean_token_accuracy": 0.16701388210058213,
      "num_tokens": 15842442.0,
      "step": 6925
    },
    {
      "entropy": 5.780656528472901,
      "epoch": 0.6850533807829181,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004961197383006463,
      "loss": 5.659,
      "mean_token_accuracy": 0.16329487413167953,
      "num_tokens": 15854755.0,
      "step": 6930
    },
    {
      "entropy": 5.642618083953858,
      "epoch": 0.6855476472914195,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004961132099102316,
      "loss": 5.539,
      "mean_token_accuracy": 0.16859302073717117,
      "num_tokens": 15866300.0,
      "step": 6935
    },
    {
      "entropy": 5.686197423934937,
      "epoch": 0.686041913799921,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004961066760803978,
      "loss": 5.5276,
      "mean_token_accuracy": 0.17215495705604553,
      "num_tokens": 15877165.0,
      "step": 6940
    },
    {
      "entropy": 5.787747859954834,
      "epoch": 0.6865361803084223,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004961001368113054,
      "loss": 5.6158,
      "mean_token_accuracy": 0.16086900383234023,
      "num_tokens": 15888893.0,
      "step": 6945
    },
    {
      "entropy": 5.739814615249633,
      "epoch": 0.6870304468169237,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004960935921031153,
      "loss": 5.6375,
      "mean_token_accuracy": 0.1608149915933609,
      "num_tokens": 15899994.0,
      "step": 6950
    },
    {
      "entropy": 5.686029529571533,
      "epoch": 0.6875247133254251,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004960870419559887,
      "loss": 5.538,
      "mean_token_accuracy": 0.1723516635596752,
      "num_tokens": 15911330.0,
      "step": 6955
    },
    {
      "entropy": 5.722923231124878,
      "epoch": 0.6880189798339265,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004960804863700864,
      "loss": 5.6281,
      "mean_token_accuracy": 0.16891862601041793,
      "num_tokens": 15922397.0,
      "step": 6960
    },
    {
      "entropy": 5.720265007019043,
      "epoch": 0.6885132463424278,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004960739253455699,
      "loss": 5.6056,
      "mean_token_accuracy": 0.1666198566555977,
      "num_tokens": 15934628.0,
      "step": 6965
    },
    {
      "entropy": 5.689791822433472,
      "epoch": 0.6890075128509292,
      "grad_norm": 1.375,
      "learning_rate": 0.0004960673588826005,
      "loss": 5.5806,
      "mean_token_accuracy": 0.16941634565591812,
      "num_tokens": 15946599.0,
      "step": 6970
    },
    {
      "entropy": 5.775901412963867,
      "epoch": 0.6895017793594306,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004960607869813398,
      "loss": 5.5979,
      "mean_token_accuracy": 0.1695634126663208,
      "num_tokens": 15957402.0,
      "step": 6975
    },
    {
      "entropy": 5.756583118438721,
      "epoch": 0.689996045867932,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004960542096419494,
      "loss": 5.6477,
      "mean_token_accuracy": 0.16816511303186416,
      "num_tokens": 15968091.0,
      "step": 6980
    },
    {
      "entropy": 5.680802536010742,
      "epoch": 0.6904903123764333,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004960476268645912,
      "loss": 5.6518,
      "mean_token_accuracy": 0.15996556580066681,
      "num_tokens": 15980809.0,
      "step": 6985
    },
    {
      "entropy": 5.726210641860962,
      "epoch": 0.6909845788849348,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004960410386494271,
      "loss": 5.6232,
      "mean_token_accuracy": 0.1655051276087761,
      "num_tokens": 15992701.0,
      "step": 6990
    },
    {
      "entropy": 5.869408893585205,
      "epoch": 0.6914788453934362,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004960344449966191,
      "loss": 5.706,
      "mean_token_accuracy": 0.1593773052096367,
      "num_tokens": 16005871.0,
      "step": 6995
    },
    {
      "entropy": 5.792735528945923,
      "epoch": 0.6919731119019376,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004960278459063296,
      "loss": 5.7126,
      "mean_token_accuracy": 0.1535090833902359,
      "num_tokens": 16018126.0,
      "step": 7000
    },
    {
      "entropy": 5.701186609268189,
      "epoch": 0.6924673784104389,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004960212413787206,
      "loss": 5.5467,
      "mean_token_accuracy": 0.16927996426820754,
      "num_tokens": 16028694.0,
      "step": 7005
    },
    {
      "entropy": 5.729494571685791,
      "epoch": 0.6929616449189403,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000496014631413955,
      "loss": 5.654,
      "mean_token_accuracy": 0.16199876368045807,
      "num_tokens": 16040354.0,
      "step": 7010
    },
    {
      "entropy": 5.756959724426269,
      "epoch": 0.6934559114274417,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004960080160121951,
      "loss": 5.6055,
      "mean_token_accuracy": 0.1653362527489662,
      "num_tokens": 16051713.0,
      "step": 7015
    },
    {
      "entropy": 5.799495077133178,
      "epoch": 0.693950177935943,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004960013951736036,
      "loss": 5.6939,
      "mean_token_accuracy": 0.1569141462445259,
      "num_tokens": 16064737.0,
      "step": 7020
    },
    {
      "entropy": 5.724238586425781,
      "epoch": 0.6944444444444444,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004959947688983436,
      "loss": 5.5548,
      "mean_token_accuracy": 0.17220764309167863,
      "num_tokens": 16075515.0,
      "step": 7025
    },
    {
      "entropy": 5.769337606430054,
      "epoch": 0.6949387109529458,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004959881371865779,
      "loss": 5.6266,
      "mean_token_accuracy": 0.1644908979535103,
      "num_tokens": 16088702.0,
      "step": 7030
    },
    {
      "entropy": 5.683289098739624,
      "epoch": 0.6954329774614472,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004959815000384699,
      "loss": 5.5556,
      "mean_token_accuracy": 0.16567983478307724,
      "num_tokens": 16101122.0,
      "step": 7035
    },
    {
      "entropy": 5.749926900863647,
      "epoch": 0.6959272439699487,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004959748574541826,
      "loss": 5.6294,
      "mean_token_accuracy": 0.16365983486175537,
      "num_tokens": 16112817.0,
      "step": 7040
    },
    {
      "entropy": 5.845486259460449,
      "epoch": 0.69642151047845,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004959682094338795,
      "loss": 5.7138,
      "mean_token_accuracy": 0.1570294588804245,
      "num_tokens": 16124827.0,
      "step": 7045
    },
    {
      "entropy": 5.6996053695678714,
      "epoch": 0.6969157769869514,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004959615559777243,
      "loss": 5.5905,
      "mean_token_accuracy": 0.1688149690628052,
      "num_tokens": 16136940.0,
      "step": 7050
    },
    {
      "entropy": 5.72951307296753,
      "epoch": 0.6974100434954528,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004959548970858805,
      "loss": 5.5936,
      "mean_token_accuracy": 0.17095874845981598,
      "num_tokens": 16147541.0,
      "step": 7055
    },
    {
      "entropy": 5.721037769317627,
      "epoch": 0.6979043100039541,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004959482327585119,
      "loss": 5.6199,
      "mean_token_accuracy": 0.16173220723867415,
      "num_tokens": 16157626.0,
      "step": 7060
    },
    {
      "entropy": 5.744309997558593,
      "epoch": 0.6983985765124555,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004959415629957824,
      "loss": 5.5777,
      "mean_token_accuracy": 0.16985894292593,
      "num_tokens": 16168494.0,
      "step": 7065
    },
    {
      "entropy": 5.782959604263306,
      "epoch": 0.6988928430209569,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004959348877978563,
      "loss": 5.6841,
      "mean_token_accuracy": 0.165982449054718,
      "num_tokens": 16180749.0,
      "step": 7070
    },
    {
      "entropy": 5.719228601455688,
      "epoch": 0.6993871095294583,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004959282071648977,
      "loss": 5.6362,
      "mean_token_accuracy": 0.1687002196907997,
      "num_tokens": 16193204.0,
      "step": 7075
    },
    {
      "entropy": 5.739464950561524,
      "epoch": 0.6998813760379596,
      "grad_norm": 1.40625,
      "learning_rate": 0.000495921521097071,
      "loss": 5.5934,
      "mean_token_accuracy": 0.1826246902346611,
      "num_tokens": 16204303.0,
      "step": 7080
    },
    {
      "entropy": 5.705110502243042,
      "epoch": 0.700375642546461,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004959148295945405,
      "loss": 5.5489,
      "mean_token_accuracy": 0.1684345468878746,
      "num_tokens": 16215261.0,
      "step": 7085
    },
    {
      "entropy": 5.693350887298584,
      "epoch": 0.7008699090549624,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004959081326574708,
      "loss": 5.5561,
      "mean_token_accuracy": 0.1730835407972336,
      "num_tokens": 16226973.0,
      "step": 7090
    },
    {
      "entropy": 5.602511167526245,
      "epoch": 0.7013641755634639,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004959014302860271,
      "loss": 5.4411,
      "mean_token_accuracy": 0.1794707089662552,
      "num_tokens": 16239152.0,
      "step": 7095
    },
    {
      "entropy": 5.743041944503784,
      "epoch": 0.7018584420719652,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004958947224803737,
      "loss": 5.5775,
      "mean_token_accuracy": 0.17361103594303132,
      "num_tokens": 16251884.0,
      "step": 7100
    },
    {
      "entropy": 5.721180820465088,
      "epoch": 0.7023527085804666,
      "grad_norm": 1.625,
      "learning_rate": 0.0004958880092406759,
      "loss": 5.6523,
      "mean_token_accuracy": 0.15597390979528428,
      "num_tokens": 16262497.0,
      "step": 7105
    },
    {
      "entropy": 5.696573257446289,
      "epoch": 0.702846975088968,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004958812905670988,
      "loss": 5.6,
      "mean_token_accuracy": 0.16918797194957733,
      "num_tokens": 16274624.0,
      "step": 7110
    },
    {
      "entropy": 5.794445562362671,
      "epoch": 0.7033412415974694,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004958745664598077,
      "loss": 5.6083,
      "mean_token_accuracy": 0.1654810830950737,
      "num_tokens": 16286096.0,
      "step": 7115
    },
    {
      "entropy": 5.691314268112182,
      "epoch": 0.7038355081059707,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004958678369189679,
      "loss": 5.5325,
      "mean_token_accuracy": 0.17520693093538284,
      "num_tokens": 16297853.0,
      "step": 7120
    },
    {
      "entropy": 5.690968465805054,
      "epoch": 0.7043297746144721,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004958611019447452,
      "loss": 5.6683,
      "mean_token_accuracy": 0.16415596902370452,
      "num_tokens": 16310708.0,
      "step": 7125
    },
    {
      "entropy": 5.79909029006958,
      "epoch": 0.7048240411229735,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004958543615373051,
      "loss": 5.6072,
      "mean_token_accuracy": 0.1694961205124855,
      "num_tokens": 16321216.0,
      "step": 7130
    },
    {
      "entropy": 5.719899320602417,
      "epoch": 0.7053183076314749,
      "grad_norm": 3.0625,
      "learning_rate": 0.0004958476156968134,
      "loss": 5.5688,
      "mean_token_accuracy": 0.17243465632200242,
      "num_tokens": 16331982.0,
      "step": 7135
    },
    {
      "entropy": 5.701531553268433,
      "epoch": 0.7058125741399762,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000495840864423436,
      "loss": 5.6554,
      "mean_token_accuracy": 0.16359290480613708,
      "num_tokens": 16343001.0,
      "step": 7140
    },
    {
      "entropy": 5.845033454895019,
      "epoch": 0.7063068406484777,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004958341077173392,
      "loss": 5.6023,
      "mean_token_accuracy": 0.16307328194379805,
      "num_tokens": 16355519.0,
      "step": 7145
    },
    {
      "entropy": 5.748544406890869,
      "epoch": 0.7068011071569791,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004958273455786892,
      "loss": 5.5802,
      "mean_token_accuracy": 0.16546572297811507,
      "num_tokens": 16366342.0,
      "step": 7150
    },
    {
      "entropy": 5.7766571044921875,
      "epoch": 0.7072953736654805,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004958205780076521,
      "loss": 5.6721,
      "mean_token_accuracy": 0.15858823508024217,
      "num_tokens": 16378395.0,
      "step": 7155
    },
    {
      "entropy": 5.728197002410889,
      "epoch": 0.7077896401739818,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004958138050043945,
      "loss": 5.5582,
      "mean_token_accuracy": 0.16583077907562255,
      "num_tokens": 16388946.0,
      "step": 7160
    },
    {
      "entropy": 5.657739543914795,
      "epoch": 0.7082839066824832,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004958070265690831,
      "loss": 5.6317,
      "mean_token_accuracy": 0.16495919078588486,
      "num_tokens": 16399824.0,
      "step": 7165
    },
    {
      "entropy": 5.7876451969146725,
      "epoch": 0.7087781731909846,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004958002427018845,
      "loss": 5.6197,
      "mean_token_accuracy": 0.1602352723479271,
      "num_tokens": 16411932.0,
      "step": 7170
    },
    {
      "entropy": 5.815198183059692,
      "epoch": 0.709272439699486,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004957934534029658,
      "loss": 5.6052,
      "mean_token_accuracy": 0.163335819542408,
      "num_tokens": 16423634.0,
      "step": 7175
    },
    {
      "entropy": 5.7700950622558596,
      "epoch": 0.7097667062079873,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004957866586724938,
      "loss": 5.6627,
      "mean_token_accuracy": 0.1671762764453888,
      "num_tokens": 16434700.0,
      "step": 7180
    },
    {
      "entropy": 5.7832824230194095,
      "epoch": 0.7102609727164887,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004957798585106359,
      "loss": 5.6468,
      "mean_token_accuracy": 0.16163350492715836,
      "num_tokens": 16445829.0,
      "step": 7185
    },
    {
      "entropy": 5.68034815788269,
      "epoch": 0.7107552392249901,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000495773052917559,
      "loss": 5.5141,
      "mean_token_accuracy": 0.17704764306545256,
      "num_tokens": 16456352.0,
      "step": 7190
    },
    {
      "entropy": 5.6992011070251465,
      "epoch": 0.7112495057334915,
      "grad_norm": 1.5,
      "learning_rate": 0.0004957662418934309,
      "loss": 5.5365,
      "mean_token_accuracy": 0.1718830108642578,
      "num_tokens": 16467542.0,
      "step": 7195
    },
    {
      "entropy": 5.620245409011841,
      "epoch": 0.7117437722419929,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004957594254384189,
      "loss": 5.4978,
      "mean_token_accuracy": 0.17775754928588866,
      "num_tokens": 16478614.0,
      "step": 7200
    },
    {
      "entropy": 5.7373486995697025,
      "epoch": 0.7122380387504943,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004957526035526908,
      "loss": 5.5665,
      "mean_token_accuracy": 0.16869405061006545,
      "num_tokens": 16489929.0,
      "step": 7205
    },
    {
      "entropy": 5.8183043003082275,
      "epoch": 0.7127323052589957,
      "grad_norm": 2.40625,
      "learning_rate": 0.0004957457762364146,
      "loss": 5.6891,
      "mean_token_accuracy": 0.15675144046545028,
      "num_tokens": 16502532.0,
      "step": 7210
    },
    {
      "entropy": 5.806992864608764,
      "epoch": 0.713226571767497,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004957389434897578,
      "loss": 5.6969,
      "mean_token_accuracy": 0.16106098294258117,
      "num_tokens": 16513805.0,
      "step": 7215
    },
    {
      "entropy": 5.771140623092651,
      "epoch": 0.7137208382759984,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004957321053128889,
      "loss": 5.6764,
      "mean_token_accuracy": 0.16268240809440612,
      "num_tokens": 16525189.0,
      "step": 7220
    },
    {
      "entropy": 5.7535521507263185,
      "epoch": 0.7142151047844998,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004957252617059759,
      "loss": 5.6751,
      "mean_token_accuracy": 0.16041869074106216,
      "num_tokens": 16538246.0,
      "step": 7225
    },
    {
      "entropy": 5.8260119438171385,
      "epoch": 0.7147093712930012,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0004957184126691873,
      "loss": 5.6609,
      "mean_token_accuracy": 0.16416329145431519,
      "num_tokens": 16550036.0,
      "step": 7230
    },
    {
      "entropy": 5.74976224899292,
      "epoch": 0.7152036378015025,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004957115582026915,
      "loss": 5.6072,
      "mean_token_accuracy": 0.16773174107074737,
      "num_tokens": 16562262.0,
      "step": 7235
    },
    {
      "entropy": 5.707104778289795,
      "epoch": 0.7156979043100039,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000495704698306657,
      "loss": 5.5791,
      "mean_token_accuracy": 0.1737347051501274,
      "num_tokens": 16574213.0,
      "step": 7240
    },
    {
      "entropy": 5.802697229385376,
      "epoch": 0.7161921708185054,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004956978329812528,
      "loss": 5.6479,
      "mean_token_accuracy": 0.16947502195835112,
      "num_tokens": 16586063.0,
      "step": 7245
    },
    {
      "entropy": 5.790889883041382,
      "epoch": 0.7166864373270068,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004956909622266476,
      "loss": 5.6501,
      "mean_token_accuracy": 0.16264750808477402,
      "num_tokens": 16598014.0,
      "step": 7250
    },
    {
      "entropy": 5.731391477584839,
      "epoch": 0.7171807038355081,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004956840860430104,
      "loss": 5.6127,
      "mean_token_accuracy": 0.16996590346097945,
      "num_tokens": 16609128.0,
      "step": 7255
    },
    {
      "entropy": 5.741106462478638,
      "epoch": 0.7176749703440095,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004956772044305106,
      "loss": 5.5423,
      "mean_token_accuracy": 0.1706029072403908,
      "num_tokens": 16620095.0,
      "step": 7260
    },
    {
      "entropy": 5.748459672927856,
      "epoch": 0.7181692368525109,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004956703173893173,
      "loss": 5.5585,
      "mean_token_accuracy": 0.16713149696588517,
      "num_tokens": 16630919.0,
      "step": 7265
    },
    {
      "entropy": 5.7497953414917,
      "epoch": 0.7186635033610123,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004956634249196001,
      "loss": 5.6744,
      "mean_token_accuracy": 0.1614256903529167,
      "num_tokens": 16642111.0,
      "step": 7270
    },
    {
      "entropy": 5.7414665699005125,
      "epoch": 0.7191577698695136,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004956565270215282,
      "loss": 5.652,
      "mean_token_accuracy": 0.16246819943189622,
      "num_tokens": 16654805.0,
      "step": 7275
    },
    {
      "entropy": 5.853487873077393,
      "epoch": 0.719652036378015,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004956496236952716,
      "loss": 5.6905,
      "mean_token_accuracy": 0.16142804622650148,
      "num_tokens": 16664735.0,
      "step": 7280
    },
    {
      "entropy": 5.802929544448853,
      "epoch": 0.7201463028865164,
      "grad_norm": 1.5,
      "learning_rate": 0.000495642714941,
      "loss": 5.5447,
      "mean_token_accuracy": 0.16482402384281158,
      "num_tokens": 16676163.0,
      "step": 7285
    },
    {
      "entropy": 5.81269097328186,
      "epoch": 0.7206405693950177,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004956358007588834,
      "loss": 5.7549,
      "mean_token_accuracy": 0.15713878720998764,
      "num_tokens": 16689683.0,
      "step": 7290
    },
    {
      "entropy": 5.856641530990601,
      "epoch": 0.7211348359035191,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004956288811490919,
      "loss": 5.682,
      "mean_token_accuracy": 0.1596776008605957,
      "num_tokens": 16702193.0,
      "step": 7295
    },
    {
      "entropy": 5.684182596206665,
      "epoch": 0.7216291024120206,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004956219561117955,
      "loss": 5.501,
      "mean_token_accuracy": 0.17966277301311492,
      "num_tokens": 16713489.0,
      "step": 7300
    },
    {
      "entropy": 5.68188705444336,
      "epoch": 0.722123368920522,
      "grad_norm": 1.625,
      "learning_rate": 0.000495615025647165,
      "loss": 5.6197,
      "mean_token_accuracy": 0.16612053662538528,
      "num_tokens": 16725103.0,
      "step": 7305
    },
    {
      "entropy": 5.735351848602295,
      "epoch": 0.7226176354290234,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004956080897553705,
      "loss": 5.5905,
      "mean_token_accuracy": 0.16317443400621415,
      "num_tokens": 16737056.0,
      "step": 7310
    },
    {
      "entropy": 5.698517084121704,
      "epoch": 0.7231119019375247,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004956011484365829,
      "loss": 5.5534,
      "mean_token_accuracy": 0.1738612711429596,
      "num_tokens": 16747347.0,
      "step": 7315
    },
    {
      "entropy": 5.692885541915894,
      "epoch": 0.7236061684460261,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004955942016909728,
      "loss": 5.571,
      "mean_token_accuracy": 0.1722766116261482,
      "num_tokens": 16758738.0,
      "step": 7320
    },
    {
      "entropy": 5.711320209503174,
      "epoch": 0.7241004349545275,
      "grad_norm": 1.4375,
      "learning_rate": 0.000495587249518711,
      "loss": 5.6174,
      "mean_token_accuracy": 0.163837032020092,
      "num_tokens": 16770867.0,
      "step": 7325
    },
    {
      "entropy": 5.768688583374024,
      "epoch": 0.7245947014630288,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004955802919199687,
      "loss": 5.5512,
      "mean_token_accuracy": 0.17396171689033507,
      "num_tokens": 16781537.0,
      "step": 7330
    },
    {
      "entropy": 5.738880825042725,
      "epoch": 0.7250889679715302,
      "grad_norm": 1.484375,
      "learning_rate": 0.000495573328894917,
      "loss": 5.588,
      "mean_token_accuracy": 0.16630553752183913,
      "num_tokens": 16792250.0,
      "step": 7335
    },
    {
      "entropy": 5.720074129104614,
      "epoch": 0.7255832344800316,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004955663604437272,
      "loss": 5.5543,
      "mean_token_accuracy": 0.16450802534818648,
      "num_tokens": 16803673.0,
      "step": 7340
    },
    {
      "entropy": 5.7297827243804935,
      "epoch": 0.726077500988533,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004955593865665707,
      "loss": 5.6732,
      "mean_token_accuracy": 0.16134673357009888,
      "num_tokens": 16815484.0,
      "step": 7345
    },
    {
      "entropy": 5.758619976043701,
      "epoch": 0.7265717674970344,
      "grad_norm": 1.6875,
      "learning_rate": 0.000495552407263619,
      "loss": 5.6215,
      "mean_token_accuracy": 0.16498324126005173,
      "num_tokens": 16827097.0,
      "step": 7350
    },
    {
      "entropy": 5.828380537033081,
      "epoch": 0.7270660340055358,
      "grad_norm": 1.75,
      "learning_rate": 0.0004955454225350441,
      "loss": 5.581,
      "mean_token_accuracy": 0.16659811586141587,
      "num_tokens": 16838235.0,
      "step": 7355
    },
    {
      "entropy": 5.756051635742187,
      "epoch": 0.7275603005140372,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004955384323810174,
      "loss": 5.5903,
      "mean_token_accuracy": 0.16529640108346938,
      "num_tokens": 16847831.0,
      "step": 7360
    },
    {
      "entropy": 5.743377733230591,
      "epoch": 0.7280545670225386,
      "grad_norm": 2.6875,
      "learning_rate": 0.000495531436801711,
      "loss": 5.6226,
      "mean_token_accuracy": 0.16450040489435197,
      "num_tokens": 16859607.0,
      "step": 7365
    },
    {
      "entropy": 5.863069009780884,
      "epoch": 0.7285488335310399,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004955244357972972,
      "loss": 5.6743,
      "mean_token_accuracy": 0.16126021295785903,
      "num_tokens": 16869787.0,
      "step": 7370
    },
    {
      "entropy": 5.708996438980103,
      "epoch": 0.7290431000395413,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000495517429367948,
      "loss": 5.5496,
      "mean_token_accuracy": 0.16942000091075898,
      "num_tokens": 16880819.0,
      "step": 7375
    },
    {
      "entropy": 5.769755458831787,
      "epoch": 0.7295373665480427,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004955104175138358,
      "loss": 5.6697,
      "mean_token_accuracy": 0.16085806638002395,
      "num_tokens": 16893247.0,
      "step": 7380
    },
    {
      "entropy": 5.781706809997559,
      "epoch": 0.7300316330565441,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004955034002351332,
      "loss": 5.5792,
      "mean_token_accuracy": 0.16594864279031754,
      "num_tokens": 16904469.0,
      "step": 7385
    },
    {
      "entropy": 5.7706982612609865,
      "epoch": 0.7305258995650454,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004954963775320127,
      "loss": 5.7237,
      "mean_token_accuracy": 0.16324723958969117,
      "num_tokens": 16918433.0,
      "step": 7390
    },
    {
      "entropy": 5.801693916320801,
      "epoch": 0.7310201660735468,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000495489349404647,
      "loss": 5.6518,
      "mean_token_accuracy": 0.1594475969672203,
      "num_tokens": 16930905.0,
      "step": 7395
    },
    {
      "entropy": 5.820454740524292,
      "epoch": 0.7315144325820483,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004954823158532093,
      "loss": 5.6792,
      "mean_token_accuracy": 0.15785257816314696,
      "num_tokens": 16943461.0,
      "step": 7400
    },
    {
      "entropy": 5.702722454071045,
      "epoch": 0.7320086990905497,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004954752768778723,
      "loss": 5.6351,
      "mean_token_accuracy": 0.16629653424024582,
      "num_tokens": 16955539.0,
      "step": 7405
    },
    {
      "entropy": 5.735093688964843,
      "epoch": 0.732502965599051,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004954682324788093,
      "loss": 5.5956,
      "mean_token_accuracy": 0.16030968278646468,
      "num_tokens": 16967508.0,
      "step": 7410
    },
    {
      "entropy": 5.771734523773193,
      "epoch": 0.7329972321075524,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004954611826561935,
      "loss": 5.5661,
      "mean_token_accuracy": 0.1696939140558243,
      "num_tokens": 16979033.0,
      "step": 7415
    },
    {
      "entropy": 5.715194416046143,
      "epoch": 0.7334914986160538,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004954541274101985,
      "loss": 5.5482,
      "mean_token_accuracy": 0.17350004613399506,
      "num_tokens": 16991874.0,
      "step": 7420
    },
    {
      "entropy": 5.8296373844146725,
      "epoch": 0.7339857651245552,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004954470667409978,
      "loss": 5.7261,
      "mean_token_accuracy": 0.15965375751256944,
      "num_tokens": 17003160.0,
      "step": 7425
    },
    {
      "entropy": 5.749544095993042,
      "epoch": 0.7344800316330565,
      "grad_norm": 1.6015625,
      "learning_rate": 0.000495440000648765,
      "loss": 5.5817,
      "mean_token_accuracy": 0.16631619036197662,
      "num_tokens": 17014043.0,
      "step": 7430
    },
    {
      "entropy": 5.759813642501831,
      "epoch": 0.7349742981415579,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004954329291336739,
      "loss": 5.5436,
      "mean_token_accuracy": 0.16683925092220306,
      "num_tokens": 17025117.0,
      "step": 7435
    },
    {
      "entropy": 5.66251392364502,
      "epoch": 0.7354685646500593,
      "grad_norm": 2.25,
      "learning_rate": 0.0004954258521958988,
      "loss": 5.611,
      "mean_token_accuracy": 0.1650553524494171,
      "num_tokens": 17037366.0,
      "step": 7440
    },
    {
      "entropy": 5.764198064804077,
      "epoch": 0.7359628311585606,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004954187698356133,
      "loss": 5.6507,
      "mean_token_accuracy": 0.16029622852802278,
      "num_tokens": 17050683.0,
      "step": 7445
    },
    {
      "entropy": 5.708848094940185,
      "epoch": 0.7364570976670621,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004954116820529921,
      "loss": 5.5338,
      "mean_token_accuracy": 0.1740710735321045,
      "num_tokens": 17062403.0,
      "step": 7450
    },
    {
      "entropy": 5.714789724349975,
      "epoch": 0.7369513641755635,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004954045888482091,
      "loss": 5.5412,
      "mean_token_accuracy": 0.1723011016845703,
      "num_tokens": 17072688.0,
      "step": 7455
    },
    {
      "entropy": 5.749787712097168,
      "epoch": 0.7374456306840649,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004953974902214391,
      "loss": 5.6045,
      "mean_token_accuracy": 0.16877943873405457,
      "num_tokens": 17084118.0,
      "step": 7460
    },
    {
      "entropy": 5.808379793167115,
      "epoch": 0.7379398971925663,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004953903861728567,
      "loss": 5.6354,
      "mean_token_accuracy": 0.16455026119947433,
      "num_tokens": 17095600.0,
      "step": 7465
    },
    {
      "entropy": 5.8653984546661375,
      "epoch": 0.7384341637010676,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004953832767026365,
      "loss": 5.7431,
      "mean_token_accuracy": 0.15749269872903823,
      "num_tokens": 17107934.0,
      "step": 7470
    },
    {
      "entropy": 5.745365571975708,
      "epoch": 0.738928430209569,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004953761618109535,
      "loss": 5.6656,
      "mean_token_accuracy": 0.15862454771995543,
      "num_tokens": 17119578.0,
      "step": 7475
    },
    {
      "entropy": 5.762130832672119,
      "epoch": 0.7394226967180704,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004953690414979828,
      "loss": 5.6451,
      "mean_token_accuracy": 0.16063157021999358,
      "num_tokens": 17129928.0,
      "step": 7480
    },
    {
      "entropy": 5.7570428371429445,
      "epoch": 0.7399169632265717,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004953619157638995,
      "loss": 5.6536,
      "mean_token_accuracy": 0.16045423299074174,
      "num_tokens": 17141149.0,
      "step": 7485
    },
    {
      "entropy": 5.7265997409820555,
      "epoch": 0.7404112297350731,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004953547846088789,
      "loss": 5.5467,
      "mean_token_accuracy": 0.16523138135671617,
      "num_tokens": 17152577.0,
      "step": 7490
    },
    {
      "entropy": 5.703083562850952,
      "epoch": 0.7409054962435745,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004953476480330964,
      "loss": 5.5679,
      "mean_token_accuracy": 0.16664033830165864,
      "num_tokens": 17163997.0,
      "step": 7495
    },
    {
      "entropy": 5.722478723526001,
      "epoch": 0.741399762752076,
      "grad_norm": 1.375,
      "learning_rate": 0.0004953405060367276,
      "loss": 5.5299,
      "mean_token_accuracy": 0.16978507190942765,
      "num_tokens": 17175317.0,
      "step": 7500
    },
    {
      "entropy": 5.63282732963562,
      "epoch": 0.7418940292605773,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004953333586199481,
      "loss": 5.5892,
      "mean_token_accuracy": 0.17345013469457626,
      "num_tokens": 17185615.0,
      "step": 7505
    },
    {
      "entropy": 5.815655708312988,
      "epoch": 0.7423882957690787,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004953262057829338,
      "loss": 5.6085,
      "mean_token_accuracy": 0.16249293237924575,
      "num_tokens": 17198160.0,
      "step": 7510
    },
    {
      "entropy": 5.761313104629517,
      "epoch": 0.7428825622775801,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004953190475258607,
      "loss": 5.5911,
      "mean_token_accuracy": 0.1688827559351921,
      "num_tokens": 17209652.0,
      "step": 7515
    },
    {
      "entropy": 5.756549215316772,
      "epoch": 0.7433768287860815,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004953118838489049,
      "loss": 5.5966,
      "mean_token_accuracy": 0.1644274726510048,
      "num_tokens": 17220832.0,
      "step": 7520
    },
    {
      "entropy": 5.7113487243652346,
      "epoch": 0.7438710952945828,
      "grad_norm": 2.125,
      "learning_rate": 0.0004953047147522424,
      "loss": 5.5467,
      "mean_token_accuracy": 0.16912853568792344,
      "num_tokens": 17231952.0,
      "step": 7525
    },
    {
      "entropy": 5.634316873550415,
      "epoch": 0.7443653618030842,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004952975402360499,
      "loss": 5.5471,
      "mean_token_accuracy": 0.17110426276922225,
      "num_tokens": 17242705.0,
      "step": 7530
    },
    {
      "entropy": 5.7730433464050295,
      "epoch": 0.7448596283115856,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004952903603005037,
      "loss": 5.6208,
      "mean_token_accuracy": 0.1608292430639267,
      "num_tokens": 17253946.0,
      "step": 7535
    },
    {
      "entropy": 5.856173133850097,
      "epoch": 0.745353894820087,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004952831749457804,
      "loss": 5.6546,
      "mean_token_accuracy": 0.17365415692329406,
      "num_tokens": 17265732.0,
      "step": 7540
    },
    {
      "entropy": 5.735471105575561,
      "epoch": 0.7458481613285883,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004952759841720569,
      "loss": 5.5842,
      "mean_token_accuracy": 0.1696339651942253,
      "num_tokens": 17277154.0,
      "step": 7545
    },
    {
      "entropy": 5.73749475479126,
      "epoch": 0.7463424278370897,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004952687879795101,
      "loss": 5.6365,
      "mean_token_accuracy": 0.16797012239694595,
      "num_tokens": 17288179.0,
      "step": 7550
    },
    {
      "entropy": 5.746026420593262,
      "epoch": 0.7468366943455912,
      "grad_norm": 3.375,
      "learning_rate": 0.0004952615863683167,
      "loss": 5.577,
      "mean_token_accuracy": 0.17081994265317918,
      "num_tokens": 17299435.0,
      "step": 7555
    },
    {
      "entropy": 5.6489966869354244,
      "epoch": 0.7473309608540926,
      "grad_norm": 2.875,
      "learning_rate": 0.0004952543793386543,
      "loss": 5.4363,
      "mean_token_accuracy": 0.18596286624670028,
      "num_tokens": 17309442.0,
      "step": 7560
    },
    {
      "entropy": 5.76450023651123,
      "epoch": 0.7478252273625939,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004952471668906999,
      "loss": 5.6504,
      "mean_token_accuracy": 0.15946330577135087,
      "num_tokens": 17322072.0,
      "step": 7565
    },
    {
      "entropy": 5.770624351501465,
      "epoch": 0.7483194938710953,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004952399490246311,
      "loss": 5.5672,
      "mean_token_accuracy": 0.1691721722483635,
      "num_tokens": 17333458.0,
      "step": 7570
    },
    {
      "entropy": 5.647971057891846,
      "epoch": 0.7488137603795967,
      "grad_norm": 2.265625,
      "learning_rate": 0.0004952327257406253,
      "loss": 5.5297,
      "mean_token_accuracy": 0.1690526247024536,
      "num_tokens": 17344319.0,
      "step": 7575
    },
    {
      "entropy": 5.714942932128906,
      "epoch": 0.749308026888098,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004952254970388603,
      "loss": 5.6395,
      "mean_token_accuracy": 0.16880596280097962,
      "num_tokens": 17357016.0,
      "step": 7580
    },
    {
      "entropy": 5.754970455169678,
      "epoch": 0.7498022933965994,
      "grad_norm": 1.453125,
      "learning_rate": 0.000495218262919514,
      "loss": 5.5313,
      "mean_token_accuracy": 0.1711251214146614,
      "num_tokens": 17368574.0,
      "step": 7585
    },
    {
      "entropy": 5.609649229049682,
      "epoch": 0.7502965599051008,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004952110233827642,
      "loss": 5.4966,
      "mean_token_accuracy": 0.1723710983991623,
      "num_tokens": 17379858.0,
      "step": 7590
    },
    {
      "entropy": 5.747307205200196,
      "epoch": 0.7507908264136022,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004952037784287891,
      "loss": 5.6442,
      "mean_token_accuracy": 0.1659706175327301,
      "num_tokens": 17391543.0,
      "step": 7595
    },
    {
      "entropy": 5.772963285446167,
      "epoch": 0.7512850929221035,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004951965280577668,
      "loss": 5.5947,
      "mean_token_accuracy": 0.17050851732492447,
      "num_tokens": 17403193.0,
      "step": 7600
    },
    {
      "entropy": 5.756211566925049,
      "epoch": 0.751779359430605,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004951892722698759,
      "loss": 5.6992,
      "mean_token_accuracy": 0.16200622618198396,
      "num_tokens": 17415866.0,
      "step": 7605
    },
    {
      "entropy": 5.665574550628662,
      "epoch": 0.7522736259391064,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004951820110652946,
      "loss": 5.5441,
      "mean_token_accuracy": 0.16659905314445494,
      "num_tokens": 17427880.0,
      "step": 7610
    },
    {
      "entropy": 5.76431713104248,
      "epoch": 0.7527678924476078,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004951747444442017,
      "loss": 5.5778,
      "mean_token_accuracy": 0.16731993854045868,
      "num_tokens": 17440295.0,
      "step": 7615
    },
    {
      "entropy": 5.792757797241211,
      "epoch": 0.7532621589561092,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004951674724067759,
      "loss": 5.6111,
      "mean_token_accuracy": 0.17040710896253586,
      "num_tokens": 17451636.0,
      "step": 7620
    },
    {
      "entropy": 5.778954076766968,
      "epoch": 0.7537564254646105,
      "grad_norm": 2.0,
      "learning_rate": 0.0004951601949531962,
      "loss": 5.6619,
      "mean_token_accuracy": 0.16695151776075362,
      "num_tokens": 17464376.0,
      "step": 7625
    },
    {
      "entropy": 5.734000110626221,
      "epoch": 0.7542506919731119,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004951529120836416,
      "loss": 5.5152,
      "mean_token_accuracy": 0.1681979015469551,
      "num_tokens": 17475098.0,
      "step": 7630
    },
    {
      "entropy": 5.651544904708862,
      "epoch": 0.7547449584816133,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004951456237982911,
      "loss": 5.5267,
      "mean_token_accuracy": 0.16398171633481978,
      "num_tokens": 17486176.0,
      "step": 7635
    },
    {
      "entropy": 5.710281324386597,
      "epoch": 0.7552392249901146,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004951383300973242,
      "loss": 5.5864,
      "mean_token_accuracy": 0.16826897412538527,
      "num_tokens": 17498386.0,
      "step": 7640
    },
    {
      "entropy": 5.766921281814575,
      "epoch": 0.755733491498616,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004951310309809201,
      "loss": 5.6293,
      "mean_token_accuracy": 0.16339175999164582,
      "num_tokens": 17510409.0,
      "step": 7645
    },
    {
      "entropy": 5.711449527740479,
      "epoch": 0.7562277580071174,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004951237264492585,
      "loss": 5.566,
      "mean_token_accuracy": 0.1698060467839241,
      "num_tokens": 17521803.0,
      "step": 7650
    },
    {
      "entropy": 5.692463731765747,
      "epoch": 0.7567220245156189,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004951164165025192,
      "loss": 5.6248,
      "mean_token_accuracy": 0.16975642889738082,
      "num_tokens": 17532974.0,
      "step": 7655
    },
    {
      "entropy": 5.783552932739258,
      "epoch": 0.7572162910241202,
      "grad_norm": 2.125,
      "learning_rate": 0.0004951091011408818,
      "loss": 5.5994,
      "mean_token_accuracy": 0.1688157469034195,
      "num_tokens": 17544309.0,
      "step": 7660
    },
    {
      "entropy": 5.709373950958252,
      "epoch": 0.7577105575326216,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004951017803645264,
      "loss": 5.5425,
      "mean_token_accuracy": 0.16980530321598053,
      "num_tokens": 17555395.0,
      "step": 7665
    },
    {
      "entropy": 5.680856132507325,
      "epoch": 0.758204824041123,
      "grad_norm": 1.7109375,
      "learning_rate": 0.000495094454173633,
      "loss": 5.5023,
      "mean_token_accuracy": 0.17162813395261764,
      "num_tokens": 17567434.0,
      "step": 7670
    },
    {
      "entropy": 5.778611326217652,
      "epoch": 0.7586990905496244,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004950871225683819,
      "loss": 5.6471,
      "mean_token_accuracy": 0.16490187793970107,
      "num_tokens": 17578381.0,
      "step": 7675
    },
    {
      "entropy": 5.737913703918457,
      "epoch": 0.7591933570581257,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004950797855489535,
      "loss": 5.5999,
      "mean_token_accuracy": 0.16983857303857802,
      "num_tokens": 17589732.0,
      "step": 7680
    },
    {
      "entropy": 5.79100489616394,
      "epoch": 0.7596876235666271,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004950724431155283,
      "loss": 5.5374,
      "mean_token_accuracy": 0.16733123809099198,
      "num_tokens": 17601092.0,
      "step": 7685
    },
    {
      "entropy": 5.717710208892822,
      "epoch": 0.7601818900751285,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004950650952682867,
      "loss": 5.6421,
      "mean_token_accuracy": 0.16308820098638535,
      "num_tokens": 17613268.0,
      "step": 7690
    },
    {
      "entropy": 5.780374574661255,
      "epoch": 0.7606761565836299,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004950577420074098,
      "loss": 5.536,
      "mean_token_accuracy": 0.16852555721998214,
      "num_tokens": 17623500.0,
      "step": 7695
    },
    {
      "entropy": 5.770693397521972,
      "epoch": 0.7611704230921312,
      "grad_norm": 1.5,
      "learning_rate": 0.0004950503833330781,
      "loss": 5.7473,
      "mean_token_accuracy": 0.1554810255765915,
      "num_tokens": 17636151.0,
      "step": 7700
    },
    {
      "entropy": 5.761449670791626,
      "epoch": 0.7616646896006327,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004950430192454729,
      "loss": 5.6112,
      "mean_token_accuracy": 0.16518060564994813,
      "num_tokens": 17648392.0,
      "step": 7705
    },
    {
      "entropy": 5.72187991142273,
      "epoch": 0.7621589561091341,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004950356497447754,
      "loss": 5.5495,
      "mean_token_accuracy": 0.16627350598573684,
      "num_tokens": 17659777.0,
      "step": 7710
    },
    {
      "entropy": 5.785965776443481,
      "epoch": 0.7626532226176355,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004950282748311666,
      "loss": 5.6516,
      "mean_token_accuracy": 0.16374578326940536,
      "num_tokens": 17671986.0,
      "step": 7715
    },
    {
      "entropy": 5.734409761428833,
      "epoch": 0.7631474891261368,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004950208945048283,
      "loss": 5.576,
      "mean_token_accuracy": 0.16800594031810762,
      "num_tokens": 17683740.0,
      "step": 7720
    },
    {
      "entropy": 5.712638425827026,
      "epoch": 0.7636417556346382,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004950135087659417,
      "loss": 5.6567,
      "mean_token_accuracy": 0.16560078561306,
      "num_tokens": 17696034.0,
      "step": 7725
    },
    {
      "entropy": 5.751904010772705,
      "epoch": 0.7641360221431396,
      "grad_norm": 2.5625,
      "learning_rate": 0.0004950061176146887,
      "loss": 5.5864,
      "mean_token_accuracy": 0.16943542659282684,
      "num_tokens": 17707783.0,
      "step": 7730
    },
    {
      "entropy": 5.737476634979248,
      "epoch": 0.764630288651641,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004949987210512511,
      "loss": 5.5474,
      "mean_token_accuracy": 0.17262809574604035,
      "num_tokens": 17718127.0,
      "step": 7735
    },
    {
      "entropy": 5.739609050750732,
      "epoch": 0.7651245551601423,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004949913190758109,
      "loss": 5.6278,
      "mean_token_accuracy": 0.1614113688468933,
      "num_tokens": 17729798.0,
      "step": 7740
    },
    {
      "entropy": 5.787795925140381,
      "epoch": 0.7656188216686437,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004949839116885499,
      "loss": 5.5846,
      "mean_token_accuracy": 0.16649723351001738,
      "num_tokens": 17741989.0,
      "step": 7745
    },
    {
      "entropy": 5.696823358535767,
      "epoch": 0.7661130881771451,
      "grad_norm": 1.9375,
      "learning_rate": 0.0004949764988896508,
      "loss": 5.6303,
      "mean_token_accuracy": 0.16772397458553315,
      "num_tokens": 17753737.0,
      "step": 7750
    },
    {
      "entropy": 5.672492122650146,
      "epoch": 0.7666073546856464,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004949690806792955,
      "loss": 5.4977,
      "mean_token_accuracy": 0.1778843492269516,
      "num_tokens": 17764158.0,
      "step": 7755
    },
    {
      "entropy": 5.745405960083008,
      "epoch": 0.7671016211941479,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004949616570576668,
      "loss": 5.6329,
      "mean_token_accuracy": 0.16734983325004577,
      "num_tokens": 17775878.0,
      "step": 7760
    },
    {
      "entropy": 5.7288330078125,
      "epoch": 0.7675958877026493,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004949542280249473,
      "loss": 5.5491,
      "mean_token_accuracy": 0.17019440233707428,
      "num_tokens": 17786107.0,
      "step": 7765
    },
    {
      "entropy": 5.716768264770508,
      "epoch": 0.7680901542111507,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004949467935813196,
      "loss": 5.6092,
      "mean_token_accuracy": 0.16628210842609406,
      "num_tokens": 17798470.0,
      "step": 7770
    },
    {
      "entropy": 5.762594699859619,
      "epoch": 0.768584420719652,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004949393537269666,
      "loss": 5.6083,
      "mean_token_accuracy": 0.16674718111753464,
      "num_tokens": 17811125.0,
      "step": 7775
    },
    {
      "entropy": 5.714617395401001,
      "epoch": 0.7690786872281534,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004949319084620714,
      "loss": 5.4803,
      "mean_token_accuracy": 0.17195260524749756,
      "num_tokens": 17821535.0,
      "step": 7780
    },
    {
      "entropy": 5.794731998443604,
      "epoch": 0.7695729537366548,
      "grad_norm": 1.390625,
      "learning_rate": 0.000494924457786817,
      "loss": 5.7081,
      "mean_token_accuracy": 0.15793583393096924,
      "num_tokens": 17833493.0,
      "step": 7785
    },
    {
      "entropy": 5.854009294509888,
      "epoch": 0.7700672202451562,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004949170017013871,
      "loss": 5.6708,
      "mean_token_accuracy": 0.16421451568603515,
      "num_tokens": 17844705.0,
      "step": 7790
    },
    {
      "entropy": 5.773963212966919,
      "epoch": 0.7705614867536575,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004949095402059646,
      "loss": 5.6724,
      "mean_token_accuracy": 0.15974969714879989,
      "num_tokens": 17856088.0,
      "step": 7795
    },
    {
      "entropy": 5.8078529834747314,
      "epoch": 0.7710557532621589,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004949020733007334,
      "loss": 5.6566,
      "mean_token_accuracy": 0.16017177999019622,
      "num_tokens": 17867602.0,
      "step": 7800
    },
    {
      "entropy": 5.803476333618164,
      "epoch": 0.7715500197706603,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000494894600985877,
      "loss": 5.6991,
      "mean_token_accuracy": 0.1625214010477066,
      "num_tokens": 17880487.0,
      "step": 7805
    },
    {
      "entropy": 5.83843674659729,
      "epoch": 0.7720442862791618,
      "grad_norm": 2.0,
      "learning_rate": 0.0004948871232615794,
      "loss": 5.6192,
      "mean_token_accuracy": 0.16537441909313202,
      "num_tokens": 17891690.0,
      "step": 7810
    },
    {
      "entropy": 5.819495391845703,
      "epoch": 0.7725385527876631,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004948796401280244,
      "loss": 5.667,
      "mean_token_accuracy": 0.16117413341999054,
      "num_tokens": 17903165.0,
      "step": 7815
    },
    {
      "entropy": 5.731281900405884,
      "epoch": 0.7730328192961645,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004948721515853962,
      "loss": 5.5061,
      "mean_token_accuracy": 0.17569276541471482,
      "num_tokens": 17914492.0,
      "step": 7820
    },
    {
      "entropy": 5.739938879013062,
      "epoch": 0.7735270858046659,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004948646576338789,
      "loss": 5.5912,
      "mean_token_accuracy": 0.1642355740070343,
      "num_tokens": 17926231.0,
      "step": 7825
    },
    {
      "entropy": 5.72123498916626,
      "epoch": 0.7740213523131673,
      "grad_norm": 1.7421875,
      "learning_rate": 0.000494857158273657,
      "loss": 5.5679,
      "mean_token_accuracy": 0.1720128521323204,
      "num_tokens": 17937167.0,
      "step": 7830
    },
    {
      "entropy": 5.773499441146851,
      "epoch": 0.7745156188216686,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004948496535049148,
      "loss": 5.6419,
      "mean_token_accuracy": 0.16745395958423615,
      "num_tokens": 17948939.0,
      "step": 7835
    },
    {
      "entropy": 5.710143995285034,
      "epoch": 0.77500988533017,
      "grad_norm": 1.78125,
      "learning_rate": 0.000494842143327837,
      "loss": 5.431,
      "mean_token_accuracy": 0.18013058453798295,
      "num_tokens": 17959250.0,
      "step": 7840
    },
    {
      "entropy": 5.673285865783692,
      "epoch": 0.7755041518386714,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004948346277426084,
      "loss": 5.6599,
      "mean_token_accuracy": 0.16390762478113174,
      "num_tokens": 17970191.0,
      "step": 7845
    },
    {
      "entropy": 5.7622730255126955,
      "epoch": 0.7759984183471728,
      "grad_norm": 2.0,
      "learning_rate": 0.000494827106749414,
      "loss": 5.5742,
      "mean_token_accuracy": 0.17012883871793746,
      "num_tokens": 17982114.0,
      "step": 7850
    },
    {
      "entropy": 5.843960189819336,
      "epoch": 0.7764926848556741,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004948195803484385,
      "loss": 5.6825,
      "mean_token_accuracy": 0.1603810116648674,
      "num_tokens": 17994066.0,
      "step": 7855
    },
    {
      "entropy": 5.718367099761963,
      "epoch": 0.7769869513641756,
      "grad_norm": 3.203125,
      "learning_rate": 0.0004948120485398673,
      "loss": 5.5979,
      "mean_token_accuracy": 0.16301825344562532,
      "num_tokens": 18005406.0,
      "step": 7860
    },
    {
      "entropy": 5.71928973197937,
      "epoch": 0.777481217872677,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004948045113238856,
      "loss": 5.614,
      "mean_token_accuracy": 0.16130679994821548,
      "num_tokens": 18016773.0,
      "step": 7865
    },
    {
      "entropy": 5.8236809253692625,
      "epoch": 0.7779754843811784,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004947969687006787,
      "loss": 5.6569,
      "mean_token_accuracy": 0.16104890108108522,
      "num_tokens": 18028682.0,
      "step": 7870
    },
    {
      "entropy": 5.822891807556152,
      "epoch": 0.7784697508896797,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004947894206704324,
      "loss": 5.6201,
      "mean_token_accuracy": 0.16524371057748793,
      "num_tokens": 18039525.0,
      "step": 7875
    },
    {
      "entropy": 5.726070070266724,
      "epoch": 0.7789640173981811,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004947818672333323,
      "loss": 5.5996,
      "mean_token_accuracy": 0.163727305829525,
      "num_tokens": 18049918.0,
      "step": 7880
    },
    {
      "entropy": 5.791027450561524,
      "epoch": 0.7794582839066825,
      "grad_norm": 2.046875,
      "learning_rate": 0.000494774308389564,
      "loss": 5.643,
      "mean_token_accuracy": 0.16190916895866395,
      "num_tokens": 18061019.0,
      "step": 7885
    },
    {
      "entropy": 5.775240516662597,
      "epoch": 0.7799525504151839,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004947667441393138,
      "loss": 5.5796,
      "mean_token_accuracy": 0.16727473735809326,
      "num_tokens": 18072123.0,
      "step": 7890
    },
    {
      "entropy": 5.710614395141602,
      "epoch": 0.7804468169236852,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004947591744827674,
      "loss": 5.6008,
      "mean_token_accuracy": 0.16120128184556962,
      "num_tokens": 18083354.0,
      "step": 7895
    },
    {
      "entropy": 5.802106952667236,
      "epoch": 0.7809410834321866,
      "grad_norm": 2.1875,
      "learning_rate": 0.0004947515994201113,
      "loss": 5.6914,
      "mean_token_accuracy": 0.15860090479254724,
      "num_tokens": 18094188.0,
      "step": 7900
    },
    {
      "entropy": 5.792725992202759,
      "epoch": 0.781435349940688,
      "grad_norm": 2.359375,
      "learning_rate": 0.0004947440189515318,
      "loss": 5.6032,
      "mean_token_accuracy": 0.16835448443889617,
      "num_tokens": 18106198.0,
      "step": 7905
    },
    {
      "entropy": 5.766251420974731,
      "epoch": 0.7819296164491895,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004947364330772154,
      "loss": 5.5687,
      "mean_token_accuracy": 0.1729208290576935,
      "num_tokens": 18117137.0,
      "step": 7910
    },
    {
      "entropy": 5.698108386993408,
      "epoch": 0.7824238829576908,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004947288417973486,
      "loss": 5.6236,
      "mean_token_accuracy": 0.15833986103534697,
      "num_tokens": 18127341.0,
      "step": 7915
    },
    {
      "entropy": 5.735625219345093,
      "epoch": 0.7829181494661922,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004947212451121182,
      "loss": 5.5679,
      "mean_token_accuracy": 0.16582157760858535,
      "num_tokens": 18138333.0,
      "step": 7920
    },
    {
      "entropy": 5.854029607772827,
      "epoch": 0.7834124159746936,
      "grad_norm": 1.8125,
      "learning_rate": 0.000494713643021711,
      "loss": 5.8224,
      "mean_token_accuracy": 0.15139006227254867,
      "num_tokens": 18150961.0,
      "step": 7925
    },
    {
      "entropy": 5.845320081710815,
      "epoch": 0.783906682483195,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004947060355263142,
      "loss": 5.6045,
      "mean_token_accuracy": 0.1659821465611458,
      "num_tokens": 18162604.0,
      "step": 7930
    },
    {
      "entropy": 5.793997812271118,
      "epoch": 0.7844009489916963,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004946984226261149,
      "loss": 5.6423,
      "mean_token_accuracy": 0.16274237185716628,
      "num_tokens": 18174202.0,
      "step": 7935
    },
    {
      "entropy": 5.815287303924561,
      "epoch": 0.7848952155001977,
      "grad_norm": 2.765625,
      "learning_rate": 0.0004946908043213001,
      "loss": 5.649,
      "mean_token_accuracy": 0.16121193915605544,
      "num_tokens": 18185942.0,
      "step": 7940
    },
    {
      "entropy": 5.766228580474854,
      "epoch": 0.7853894820086991,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004946831806120576,
      "loss": 5.5758,
      "mean_token_accuracy": 0.16358782947063447,
      "num_tokens": 18198897.0,
      "step": 7945
    },
    {
      "entropy": 5.713035774230957,
      "epoch": 0.7858837485172004,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004946755514985748,
      "loss": 5.494,
      "mean_token_accuracy": 0.17339199781417847,
      "num_tokens": 18211439.0,
      "step": 7950
    },
    {
      "entropy": 5.67184591293335,
      "epoch": 0.7863780150257018,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004946679169810392,
      "loss": 5.5178,
      "mean_token_accuracy": 0.1753073200583458,
      "num_tokens": 18221864.0,
      "step": 7955
    },
    {
      "entropy": 5.699005889892578,
      "epoch": 0.7868722815342032,
      "grad_norm": 2.453125,
      "learning_rate": 0.000494660277059639,
      "loss": 5.6852,
      "mean_token_accuracy": 0.1599324032664299,
      "num_tokens": 18234005.0,
      "step": 7960
    },
    {
      "entropy": 5.764883041381836,
      "epoch": 0.7873665480427047,
      "grad_norm": 2.125,
      "learning_rate": 0.0004946526317345617,
      "loss": 5.5763,
      "mean_token_accuracy": 0.17003808170557022,
      "num_tokens": 18246804.0,
      "step": 7965
    },
    {
      "entropy": 5.78763518333435,
      "epoch": 0.787860814551206,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004946449810059957,
      "loss": 5.5721,
      "mean_token_accuracy": 0.16835487633943558,
      "num_tokens": 18259192.0,
      "step": 7970
    },
    {
      "entropy": 5.7099353790283205,
      "epoch": 0.7883550810597074,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004946373248741289,
      "loss": 5.6096,
      "mean_token_accuracy": 0.16795953810214997,
      "num_tokens": 18271073.0,
      "step": 7975
    },
    {
      "entropy": 5.746679973602295,
      "epoch": 0.7888493475682088,
      "grad_norm": 1.8046875,
      "learning_rate": 0.00049462966333915,
      "loss": 5.607,
      "mean_token_accuracy": 0.16693271249532698,
      "num_tokens": 18281650.0,
      "step": 7980
    },
    {
      "entropy": 5.821167373657227,
      "epoch": 0.7893436140767102,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004946219964012473,
      "loss": 5.5497,
      "mean_token_accuracy": 0.16873108595609665,
      "num_tokens": 18292217.0,
      "step": 7985
    },
    {
      "entropy": 5.650928974151611,
      "epoch": 0.7898378805852115,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004946143240606095,
      "loss": 5.5557,
      "mean_token_accuracy": 0.16773455291986467,
      "num_tokens": 18305210.0,
      "step": 7990
    },
    {
      "entropy": 5.711899185180664,
      "epoch": 0.7903321470937129,
      "grad_norm": 2.0,
      "learning_rate": 0.0004946066463174251,
      "loss": 5.5226,
      "mean_token_accuracy": 0.172981558740139,
      "num_tokens": 18316418.0,
      "step": 7995
    },
    {
      "entropy": 5.6965254783630375,
      "epoch": 0.7908264136022143,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004945989631718831,
      "loss": 5.6019,
      "mean_token_accuracy": 0.16918894946575164,
      "num_tokens": 18328637.0,
      "step": 8000
    },
    {
      "entropy": 5.708858489990234,
      "epoch": 0.7913206801107157,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004945912746241727,
      "loss": 5.4984,
      "mean_token_accuracy": 0.17455985844135286,
      "num_tokens": 18340292.0,
      "step": 8005
    },
    {
      "entropy": 5.734827184677124,
      "epoch": 0.791814946619217,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004945835806744828,
      "loss": 5.5826,
      "mean_token_accuracy": 0.16928263306617736,
      "num_tokens": 18350784.0,
      "step": 8010
    },
    {
      "entropy": 5.804034471511841,
      "epoch": 0.7923092131277185,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004945758813230028,
      "loss": 5.6269,
      "mean_token_accuracy": 0.16163259893655776,
      "num_tokens": 18363147.0,
      "step": 8015
    },
    {
      "entropy": 5.79567584991455,
      "epoch": 0.7928034796362199,
      "grad_norm": 2.25,
      "learning_rate": 0.000494568176569922,
      "loss": 5.5995,
      "mean_token_accuracy": 0.1675742194056511,
      "num_tokens": 18374152.0,
      "step": 8020
    },
    {
      "entropy": 5.745318222045898,
      "epoch": 0.7932977461447213,
      "grad_norm": 1.78125,
      "learning_rate": 0.00049456046641543,
      "loss": 5.4946,
      "mean_token_accuracy": 0.16881230026483535,
      "num_tokens": 18385321.0,
      "step": 8025
    },
    {
      "entropy": 5.675108766555786,
      "epoch": 0.7937920126532226,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0004945527508597165,
      "loss": 5.5988,
      "mean_token_accuracy": 0.1631831556558609,
      "num_tokens": 18397072.0,
      "step": 8030
    },
    {
      "entropy": 5.730170154571534,
      "epoch": 0.794286279161724,
      "grad_norm": 1.625,
      "learning_rate": 0.0004945450299029713,
      "loss": 5.5449,
      "mean_token_accuracy": 0.16746590435504913,
      "num_tokens": 18408172.0,
      "step": 8035
    },
    {
      "entropy": 5.788347434997559,
      "epoch": 0.7947805456702254,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004945373035453842,
      "loss": 5.5505,
      "mean_token_accuracy": 0.1692645326256752,
      "num_tokens": 18419350.0,
      "step": 8040
    },
    {
      "entropy": 5.686052751541138,
      "epoch": 0.7952748121787268,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004945295717871454,
      "loss": 5.5926,
      "mean_token_accuracy": 0.16614610105752944,
      "num_tokens": 18430907.0,
      "step": 8045
    },
    {
      "entropy": 5.753582334518432,
      "epoch": 0.7957690786872281,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004945218346284451,
      "loss": 5.5776,
      "mean_token_accuracy": 0.16936265379190446,
      "num_tokens": 18441167.0,
      "step": 8050
    },
    {
      "entropy": 5.731485366821289,
      "epoch": 0.7962633451957295,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004945140920694736,
      "loss": 5.4865,
      "mean_token_accuracy": 0.17925694435834885,
      "num_tokens": 18452025.0,
      "step": 8055
    },
    {
      "entropy": 5.770360326766967,
      "epoch": 0.7967576117042309,
      "grad_norm": 2.484375,
      "learning_rate": 0.0004945063441104214,
      "loss": 5.7099,
      "mean_token_accuracy": 0.15428086891770362,
      "num_tokens": 18464305.0,
      "step": 8060
    },
    {
      "entropy": 5.755635833740234,
      "epoch": 0.7972518782127324,
      "grad_norm": 2.359375,
      "learning_rate": 0.0004944985907514791,
      "loss": 5.6277,
      "mean_token_accuracy": 0.16804383099079132,
      "num_tokens": 18476217.0,
      "step": 8065
    },
    {
      "entropy": 5.840139961242675,
      "epoch": 0.7977461447212337,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004944908319928372,
      "loss": 5.6014,
      "mean_token_accuracy": 0.16283569633960723,
      "num_tokens": 18486967.0,
      "step": 8070
    },
    {
      "entropy": 5.804177236557007,
      "epoch": 0.7982404112297351,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0004944830678346871,
      "loss": 5.5772,
      "mean_token_accuracy": 0.16123078316450118,
      "num_tokens": 18498918.0,
      "step": 8075
    },
    {
      "entropy": 5.735114192962646,
      "epoch": 0.7987346777382365,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004944752982772194,
      "loss": 5.5767,
      "mean_token_accuracy": 0.1721997916698456,
      "num_tokens": 18509424.0,
      "step": 8080
    },
    {
      "entropy": 5.687436103820801,
      "epoch": 0.7992289442467378,
      "grad_norm": 2.1875,
      "learning_rate": 0.0004944675233206252,
      "loss": 5.485,
      "mean_token_accuracy": 0.17480947375297545,
      "num_tokens": 18520444.0,
      "step": 8085
    },
    {
      "entropy": 5.692551803588867,
      "epoch": 0.7997232107552392,
      "grad_norm": 3.265625,
      "learning_rate": 0.000494459742965096,
      "loss": 5.5625,
      "mean_token_accuracy": 0.17156413346529006,
      "num_tokens": 18531874.0,
      "step": 8090
    },
    {
      "entropy": 5.79705581665039,
      "epoch": 0.8002174772637406,
      "grad_norm": 2.140625,
      "learning_rate": 0.000494451957210823,
      "loss": 5.6377,
      "mean_token_accuracy": 0.16214412003755568,
      "num_tokens": 18542492.0,
      "step": 8095
    },
    {
      "entropy": 5.810577726364135,
      "epoch": 0.800711743772242,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004944441660579979,
      "loss": 5.6892,
      "mean_token_accuracy": 0.15392046570777893,
      "num_tokens": 18554477.0,
      "step": 8100
    },
    {
      "entropy": 5.748530530929566,
      "epoch": 0.8012060102807433,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004944363695068121,
      "loss": 5.5564,
      "mean_token_accuracy": 0.17218154817819595,
      "num_tokens": 18565416.0,
      "step": 8105
    },
    {
      "entropy": 5.733624601364136,
      "epoch": 0.8017002767892447,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004944285675574576,
      "loss": 5.5239,
      "mean_token_accuracy": 0.17163797467947006,
      "num_tokens": 18575949.0,
      "step": 8110
    },
    {
      "entropy": 5.705946159362793,
      "epoch": 0.8021945432977462,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0004944207602101265,
      "loss": 5.6401,
      "mean_token_accuracy": 0.16614215821027756,
      "num_tokens": 18588177.0,
      "step": 8115
    },
    {
      "entropy": 5.770335531234741,
      "epoch": 0.8026888098062476,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004944129474650104,
      "loss": 5.4932,
      "mean_token_accuracy": 0.16864445805549622,
      "num_tokens": 18599112.0,
      "step": 8120
    },
    {
      "entropy": 5.58718318939209,
      "epoch": 0.803183076314749,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000494405129322302,
      "loss": 5.4515,
      "mean_token_accuracy": 0.17800434678792953,
      "num_tokens": 18610432.0,
      "step": 8125
    },
    {
      "entropy": 5.708334064483642,
      "epoch": 0.8036773428232503,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004943973057821931,
      "loss": 5.5363,
      "mean_token_accuracy": 0.16058900356292724,
      "num_tokens": 18621739.0,
      "step": 8130
    },
    {
      "entropy": 5.75359206199646,
      "epoch": 0.8041716093317517,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004943894768448768,
      "loss": 5.5513,
      "mean_token_accuracy": 0.16682366132736207,
      "num_tokens": 18631983.0,
      "step": 8135
    },
    {
      "entropy": 5.758172178268433,
      "epoch": 0.8046658758402531,
      "grad_norm": 1.78125,
      "learning_rate": 0.000494381642510545,
      "loss": 5.6501,
      "mean_token_accuracy": 0.1592622384428978,
      "num_tokens": 18643507.0,
      "step": 8140
    },
    {
      "entropy": 5.856941986083984,
      "epoch": 0.8051601423487544,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004943738027793909,
      "loss": 5.6847,
      "mean_token_accuracy": 0.16642322838306428,
      "num_tokens": 18654516.0,
      "step": 8145
    },
    {
      "entropy": 5.860350036621094,
      "epoch": 0.8056544088572558,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004943659576516072,
      "loss": 5.6642,
      "mean_token_accuracy": 0.16003682017326354,
      "num_tokens": 18666855.0,
      "step": 8150
    },
    {
      "entropy": 5.720005226135254,
      "epoch": 0.8061486753657572,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004943581071273869,
      "loss": 5.6396,
      "mean_token_accuracy": 0.16216532737016678,
      "num_tokens": 18679105.0,
      "step": 8155
    },
    {
      "entropy": 5.747036457061768,
      "epoch": 0.8066429418742586,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004943502512069232,
      "loss": 5.7048,
      "mean_token_accuracy": 0.16681862473487855,
      "num_tokens": 18691349.0,
      "step": 8160
    },
    {
      "entropy": 5.866746234893799,
      "epoch": 0.8071372083827599,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0004943423898904092,
      "loss": 5.7216,
      "mean_token_accuracy": 0.16013684570789338,
      "num_tokens": 18703773.0,
      "step": 8165
    },
    {
      "entropy": 5.770022296905518,
      "epoch": 0.8076314748912614,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004943345231780385,
      "loss": 5.5964,
      "mean_token_accuracy": 0.16229512095451354,
      "num_tokens": 18715281.0,
      "step": 8170
    },
    {
      "entropy": 5.730436706542969,
      "epoch": 0.8081257413997628,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004943266510700043,
      "loss": 5.5626,
      "mean_token_accuracy": 0.16943300813436507,
      "num_tokens": 18726446.0,
      "step": 8175
    },
    {
      "entropy": 5.711192083358765,
      "epoch": 0.8086200079082642,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004943187735665007,
      "loss": 5.4443,
      "mean_token_accuracy": 0.17946724593639374,
      "num_tokens": 18737348.0,
      "step": 8180
    },
    {
      "entropy": 5.722460412979126,
      "epoch": 0.8091142744167655,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004943108906677211,
      "loss": 5.5377,
      "mean_token_accuracy": 0.16783976107835769,
      "num_tokens": 18748921.0,
      "step": 8185
    },
    {
      "entropy": 5.6996664047241214,
      "epoch": 0.8096085409252669,
      "grad_norm": 2.46875,
      "learning_rate": 0.0004943030023738595,
      "loss": 5.5646,
      "mean_token_accuracy": 0.17357683777809144,
      "num_tokens": 18761252.0,
      "step": 8190
    },
    {
      "entropy": 5.712693071365356,
      "epoch": 0.8101028074337683,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004942951086851102,
      "loss": 5.5342,
      "mean_token_accuracy": 0.1661703333258629,
      "num_tokens": 18772915.0,
      "step": 8195
    },
    {
      "entropy": 5.715088748931885,
      "epoch": 0.8105970739422697,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004942872096016671,
      "loss": 5.5609,
      "mean_token_accuracy": 0.17356885224580765,
      "num_tokens": 18783533.0,
      "step": 8200
    },
    {
      "entropy": 5.696490907669068,
      "epoch": 0.811091340450771,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004942793051237247,
      "loss": 5.5334,
      "mean_token_accuracy": 0.16562338173389435,
      "num_tokens": 18795068.0,
      "step": 8205
    },
    {
      "entropy": 5.799958515167236,
      "epoch": 0.8115856069592724,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004942713952514773,
      "loss": 5.718,
      "mean_token_accuracy": 0.15598101913928986,
      "num_tokens": 18808105.0,
      "step": 8210
    },
    {
      "entropy": 5.74779863357544,
      "epoch": 0.8120798734677738,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0004942634799851196,
      "loss": 5.5751,
      "mean_token_accuracy": 0.16392529457807542,
      "num_tokens": 18819275.0,
      "step": 8215
    },
    {
      "entropy": 5.709667634963989,
      "epoch": 0.8125741399762753,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004942555593248463,
      "loss": 5.5199,
      "mean_token_accuracy": 0.17542681843042374,
      "num_tokens": 18828535.0,
      "step": 8220
    },
    {
      "entropy": 5.7056725978851315,
      "epoch": 0.8130684064847766,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0004942476332708522,
      "loss": 5.5377,
      "mean_token_accuracy": 0.17447522431612014,
      "num_tokens": 18839983.0,
      "step": 8225
    },
    {
      "entropy": 5.779212856292725,
      "epoch": 0.813562672993278,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004942397018233322,
      "loss": 5.6074,
      "mean_token_accuracy": 0.16651627272367478,
      "num_tokens": 18851875.0,
      "step": 8230
    },
    {
      "entropy": 5.742778253555298,
      "epoch": 0.8140569395017794,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004942317649824816,
      "loss": 5.6173,
      "mean_token_accuracy": 0.1662682831287384,
      "num_tokens": 18863629.0,
      "step": 8235
    },
    {
      "entropy": 5.688341474533081,
      "epoch": 0.8145512060102807,
      "grad_norm": 1.625,
      "learning_rate": 0.0004942238227484956,
      "loss": 5.5326,
      "mean_token_accuracy": 0.17284415364265443,
      "num_tokens": 18875660.0,
      "step": 8240
    },
    {
      "entropy": 5.7825586795806885,
      "epoch": 0.8150454725187821,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004942158751215696,
      "loss": 5.629,
      "mean_token_accuracy": 0.16246573776006698,
      "num_tokens": 18887140.0,
      "step": 8245
    },
    {
      "entropy": 5.751757287979126,
      "epoch": 0.8155397390272835,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004942079221018989,
      "loss": 5.6324,
      "mean_token_accuracy": 0.1648348778486252,
      "num_tokens": 18899386.0,
      "step": 8250
    },
    {
      "entropy": 5.828459548950195,
      "epoch": 0.8160340055357849,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004941999636896795,
      "loss": 5.6788,
      "mean_token_accuracy": 0.15502452254295349,
      "num_tokens": 18911187.0,
      "step": 8255
    },
    {
      "entropy": 5.741218852996826,
      "epoch": 0.8165282720442862,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004941919998851068,
      "loss": 5.543,
      "mean_token_accuracy": 0.1658144935965538,
      "num_tokens": 18922522.0,
      "step": 8260
    },
    {
      "entropy": 5.624863624572754,
      "epoch": 0.8170225385527876,
      "grad_norm": 2.265625,
      "learning_rate": 0.0004941840306883771,
      "loss": 5.4739,
      "mean_token_accuracy": 0.17052083611488342,
      "num_tokens": 18932689.0,
      "step": 8265
    },
    {
      "entropy": 5.811828184127807,
      "epoch": 0.8175168050612891,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004941760560996861,
      "loss": 5.7242,
      "mean_token_accuracy": 0.16082705110311507,
      "num_tokens": 18944296.0,
      "step": 8270
    },
    {
      "entropy": 5.806406307220459,
      "epoch": 0.8180110715697905,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004941680761192302,
      "loss": 5.6438,
      "mean_token_accuracy": 0.1650134488940239,
      "num_tokens": 18955449.0,
      "step": 8275
    },
    {
      "entropy": 5.840929269790649,
      "epoch": 0.8185053380782918,
      "grad_norm": 1.625,
      "learning_rate": 0.0004941600907472055,
      "loss": 5.6476,
      "mean_token_accuracy": 0.16064279079437255,
      "num_tokens": 18969065.0,
      "step": 8280
    },
    {
      "entropy": 5.764996004104614,
      "epoch": 0.8189996045867932,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004941520999838088,
      "loss": 5.6348,
      "mean_token_accuracy": 0.17293646186590195,
      "num_tokens": 18979801.0,
      "step": 8285
    },
    {
      "entropy": 5.722033262252808,
      "epoch": 0.8194938710952946,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004941441038292363,
      "loss": 5.6818,
      "mean_token_accuracy": 0.16279126033186914,
      "num_tokens": 18990795.0,
      "step": 8290
    },
    {
      "entropy": 5.803367376327515,
      "epoch": 0.819988137603796,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004941361022836847,
      "loss": 5.6322,
      "mean_token_accuracy": 0.16177277714014054,
      "num_tokens": 19002318.0,
      "step": 8295
    },
    {
      "entropy": 5.774603319168091,
      "epoch": 0.8204824041122973,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004941280953473513,
      "loss": 5.6415,
      "mean_token_accuracy": 0.1721104621887207,
      "num_tokens": 19014977.0,
      "step": 8300
    },
    {
      "entropy": 5.795838165283203,
      "epoch": 0.8209766706207987,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004941200830204325,
      "loss": 5.6524,
      "mean_token_accuracy": 0.16473191753029823,
      "num_tokens": 19026081.0,
      "step": 8305
    },
    {
      "entropy": 5.792309236526489,
      "epoch": 0.8214709371293001,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004941120653031256,
      "loss": 5.6311,
      "mean_token_accuracy": 0.1613746091723442,
      "num_tokens": 19037694.0,
      "step": 8310
    },
    {
      "entropy": 5.711976194381714,
      "epoch": 0.8219652036378015,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004941040421956282,
      "loss": 5.5316,
      "mean_token_accuracy": 0.17463389784097672,
      "num_tokens": 19048344.0,
      "step": 8315
    },
    {
      "entropy": 5.784969282150269,
      "epoch": 0.8224594701463029,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004940960136981372,
      "loss": 5.6022,
      "mean_token_accuracy": 0.16066683381795882,
      "num_tokens": 19060213.0,
      "step": 8320
    },
    {
      "entropy": 5.7101997375488285,
      "epoch": 0.8229537366548043,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0004940879798108503,
      "loss": 5.4935,
      "mean_token_accuracy": 0.17741834223270417,
      "num_tokens": 19070881.0,
      "step": 8325
    },
    {
      "entropy": 5.726340198516846,
      "epoch": 0.8234480031633057,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004940799405339651,
      "loss": 5.5341,
      "mean_token_accuracy": 0.16876033395528794,
      "num_tokens": 19082206.0,
      "step": 8330
    },
    {
      "entropy": 5.667361116409301,
      "epoch": 0.8239422696718071,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004940718958676794,
      "loss": 5.5631,
      "mean_token_accuracy": 0.1733643665909767,
      "num_tokens": 19094014.0,
      "step": 8335
    },
    {
      "entropy": 5.789534711837769,
      "epoch": 0.8244365361803084,
      "grad_norm": 2.28125,
      "learning_rate": 0.000494063845812191,
      "loss": 5.5522,
      "mean_token_accuracy": 0.17264978140592574,
      "num_tokens": 19106859.0,
      "step": 8340
    },
    {
      "entropy": 5.654939222335815,
      "epoch": 0.8249308026888098,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004940557903676981,
      "loss": 5.5968,
      "mean_token_accuracy": 0.1693487599492073,
      "num_tokens": 19118296.0,
      "step": 8345
    },
    {
      "entropy": 5.741930246353149,
      "epoch": 0.8254250691973112,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004940477295343988,
      "loss": 5.544,
      "mean_token_accuracy": 0.1718739315867424,
      "num_tokens": 19130212.0,
      "step": 8350
    },
    {
      "entropy": 5.71384630203247,
      "epoch": 0.8259193357058126,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004940396633124912,
      "loss": 5.5218,
      "mean_token_accuracy": 0.1781422570347786,
      "num_tokens": 19140798.0,
      "step": 8355
    },
    {
      "entropy": 5.773586368560791,
      "epoch": 0.8264136022143139,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004940315917021742,
      "loss": 5.6263,
      "mean_token_accuracy": 0.16828884929418564,
      "num_tokens": 19151875.0,
      "step": 8360
    },
    {
      "entropy": 5.8171134948730465,
      "epoch": 0.8269078687228153,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004940235147036457,
      "loss": 5.6904,
      "mean_token_accuracy": 0.15664152204990386,
      "num_tokens": 19164532.0,
      "step": 8365
    },
    {
      "entropy": 5.73322434425354,
      "epoch": 0.8274021352313167,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000494015432317105,
      "loss": 5.6274,
      "mean_token_accuracy": 0.165276001393795,
      "num_tokens": 19176828.0,
      "step": 8370
    },
    {
      "entropy": 5.6323466300964355,
      "epoch": 0.8278964017398182,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004940073445427506,
      "loss": 5.4644,
      "mean_token_accuracy": 0.1765448972582817,
      "num_tokens": 19188118.0,
      "step": 8375
    },
    {
      "entropy": 5.726210451126098,
      "epoch": 0.8283906682483195,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004939992513807816,
      "loss": 5.6424,
      "mean_token_accuracy": 0.1584145173430443,
      "num_tokens": 19201009.0,
      "step": 8380
    },
    {
      "entropy": 5.74536657333374,
      "epoch": 0.8288849347568209,
      "grad_norm": 1.625,
      "learning_rate": 0.0004939911528313971,
      "loss": 5.543,
      "mean_token_accuracy": 0.16552556604146956,
      "num_tokens": 19212747.0,
      "step": 8385
    },
    {
      "entropy": 5.7278941631317135,
      "epoch": 0.8293792012653223,
      "grad_norm": 1.5390625,
      "learning_rate": 0.000493983048894796,
      "loss": 5.6061,
      "mean_token_accuracy": 0.16094849705696107,
      "num_tokens": 19224043.0,
      "step": 8390
    },
    {
      "entropy": 5.742194604873657,
      "epoch": 0.8298734677738236,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004939749395711782,
      "loss": 5.5166,
      "mean_token_accuracy": 0.16918471604585647,
      "num_tokens": 19235475.0,
      "step": 8395
    },
    {
      "entropy": 5.848622798919678,
      "epoch": 0.830367734282325,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004939668248607428,
      "loss": 5.6148,
      "mean_token_accuracy": 0.16481879055500032,
      "num_tokens": 19247208.0,
      "step": 8400
    },
    {
      "entropy": 5.7011463165283205,
      "epoch": 0.8308620007908264,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004939587047636896,
      "loss": 5.5725,
      "mean_token_accuracy": 0.16622405797243117,
      "num_tokens": 19260172.0,
      "step": 8405
    },
    {
      "entropy": 5.751612997055053,
      "epoch": 0.8313562672993278,
      "grad_norm": 1.625,
      "learning_rate": 0.0004939505792802182,
      "loss": 5.6284,
      "mean_token_accuracy": 0.1683248460292816,
      "num_tokens": 19271173.0,
      "step": 8410
    },
    {
      "entropy": 5.7848186016082765,
      "epoch": 0.8318505338078291,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004939424484105287,
      "loss": 5.5965,
      "mean_token_accuracy": 0.17174357622861863,
      "num_tokens": 19282693.0,
      "step": 8415
    },
    {
      "entropy": 5.73541841506958,
      "epoch": 0.8323448003163305,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004939343121548209,
      "loss": 5.6297,
      "mean_token_accuracy": 0.15996363162994384,
      "num_tokens": 19293883.0,
      "step": 8420
    },
    {
      "entropy": 5.6662674903869625,
      "epoch": 0.832839066824832,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004939261705132951,
      "loss": 5.474,
      "mean_token_accuracy": 0.1741936683654785,
      "num_tokens": 19304530.0,
      "step": 8425
    },
    {
      "entropy": 5.645070791244507,
      "epoch": 0.8333333333333334,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004939180234861516,
      "loss": 5.5249,
      "mean_token_accuracy": 0.17261814922094346,
      "num_tokens": 19315765.0,
      "step": 8430
    },
    {
      "entropy": 5.750543737411499,
      "epoch": 0.8338275998418347,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004939098710735906,
      "loss": 5.5881,
      "mean_token_accuracy": 0.16688463538885118,
      "num_tokens": 19328484.0,
      "step": 8435
    },
    {
      "entropy": 5.6985454082489015,
      "epoch": 0.8343218663503361,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004939017132758129,
      "loss": 5.5335,
      "mean_token_accuracy": 0.17178886979818345,
      "num_tokens": 19340594.0,
      "step": 8440
    },
    {
      "entropy": 5.813464736938476,
      "epoch": 0.8348161328588375,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004938935500930192,
      "loss": 5.6238,
      "mean_token_accuracy": 0.16470698714256288,
      "num_tokens": 19351105.0,
      "step": 8445
    },
    {
      "entropy": 5.729646682739258,
      "epoch": 0.8353103993673389,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004938853815254101,
      "loss": 5.5312,
      "mean_token_accuracy": 0.17212710082530974,
      "num_tokens": 19361619.0,
      "step": 8450
    },
    {
      "entropy": 5.739841556549072,
      "epoch": 0.8358046658758402,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004938772075731867,
      "loss": 5.6958,
      "mean_token_accuracy": 0.16396330744028093,
      "num_tokens": 19374251.0,
      "step": 8455
    },
    {
      "entropy": 5.80084490776062,
      "epoch": 0.8362989323843416,
      "grad_norm": 2.625,
      "learning_rate": 0.0004938690282365499,
      "loss": 5.6386,
      "mean_token_accuracy": 0.1563650041818619,
      "num_tokens": 19386300.0,
      "step": 8460
    },
    {
      "entropy": 5.709263563156128,
      "epoch": 0.836793198892843,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004938608435157012,
      "loss": 5.5593,
      "mean_token_accuracy": 0.1674885407090187,
      "num_tokens": 19397082.0,
      "step": 8465
    },
    {
      "entropy": 5.699165153503418,
      "epoch": 0.8372874654013444,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004938526534108417,
      "loss": 5.6282,
      "mean_token_accuracy": 0.16458649039268494,
      "num_tokens": 19409123.0,
      "step": 8470
    },
    {
      "entropy": 5.634624910354614,
      "epoch": 0.8377817319098458,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004938444579221732,
      "loss": 5.3999,
      "mean_token_accuracy": 0.1866477444767952,
      "num_tokens": 19419854.0,
      "step": 8475
    },
    {
      "entropy": 5.757629108428955,
      "epoch": 0.8382759984183472,
      "grad_norm": 1.5,
      "learning_rate": 0.0004938362570498969,
      "loss": 5.6005,
      "mean_token_accuracy": 0.16824939399957656,
      "num_tokens": 19430825.0,
      "step": 8480
    },
    {
      "entropy": 5.708920335769653,
      "epoch": 0.8387702649268486,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004938280507942146,
      "loss": 5.6112,
      "mean_token_accuracy": 0.16124509274959564,
      "num_tokens": 19442617.0,
      "step": 8485
    },
    {
      "entropy": 5.83821210861206,
      "epoch": 0.83926453143535,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004938198391553285,
      "loss": 5.6299,
      "mean_token_accuracy": 0.1635190352797508,
      "num_tokens": 19453911.0,
      "step": 8490
    },
    {
      "entropy": 5.6762285232543945,
      "epoch": 0.8397587979438513,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004938116221334404,
      "loss": 5.5267,
      "mean_token_accuracy": 0.17146094739437104,
      "num_tokens": 19464837.0,
      "step": 8495
    },
    {
      "entropy": 5.660635662078858,
      "epoch": 0.8402530644523527,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004938033997287524,
      "loss": 5.5105,
      "mean_token_accuracy": 0.17637209445238114,
      "num_tokens": 19477067.0,
      "step": 8500
    },
    {
      "entropy": 5.66599178314209,
      "epoch": 0.8407473309608541,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004937951719414668,
      "loss": 5.5559,
      "mean_token_accuracy": 0.1712522029876709,
      "num_tokens": 19489442.0,
      "step": 8505
    },
    {
      "entropy": 5.6885395526885985,
      "epoch": 0.8412415974693555,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004937869387717861,
      "loss": 5.525,
      "mean_token_accuracy": 0.16600052267313004,
      "num_tokens": 19501197.0,
      "step": 8510
    },
    {
      "entropy": 5.7533753395080565,
      "epoch": 0.8417358639778568,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004937787002199127,
      "loss": 5.551,
      "mean_token_accuracy": 0.1710121750831604,
      "num_tokens": 19513311.0,
      "step": 8515
    },
    {
      "entropy": 5.8224297046661375,
      "epoch": 0.8422301304863582,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004937704562860493,
      "loss": 5.6781,
      "mean_token_accuracy": 0.16092755049467086,
      "num_tokens": 19525643.0,
      "step": 8520
    },
    {
      "entropy": 5.6762122631073,
      "epoch": 0.8427243969948597,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004937622069703988,
      "loss": 5.485,
      "mean_token_accuracy": 0.1767942041158676,
      "num_tokens": 19537478.0,
      "step": 8525
    },
    {
      "entropy": 5.732364177703857,
      "epoch": 0.843218663503361,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004937539522731641,
      "loss": 5.5552,
      "mean_token_accuracy": 0.17082753777503967,
      "num_tokens": 19549370.0,
      "step": 8530
    },
    {
      "entropy": 5.772033739089966,
      "epoch": 0.8437129300118624,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004937456921945482,
      "loss": 5.5,
      "mean_token_accuracy": 0.17610560804605485,
      "num_tokens": 19561793.0,
      "step": 8535
    },
    {
      "entropy": 5.596465587615967,
      "epoch": 0.8442071965203638,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004937374267347543,
      "loss": 5.4331,
      "mean_token_accuracy": 0.18058519512414933,
      "num_tokens": 19573603.0,
      "step": 8540
    },
    {
      "entropy": 5.679430818557739,
      "epoch": 0.8447014630288652,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004937291558939858,
      "loss": 5.6086,
      "mean_token_accuracy": 0.1616741880774498,
      "num_tokens": 19584848.0,
      "step": 8545
    },
    {
      "entropy": 5.734065532684326,
      "epoch": 0.8451957295373665,
      "grad_norm": 1.5,
      "learning_rate": 0.0004937208796724461,
      "loss": 5.5647,
      "mean_token_accuracy": 0.16682935655117034,
      "num_tokens": 19594965.0,
      "step": 8550
    },
    {
      "entropy": 5.741935729980469,
      "epoch": 0.8456899960458679,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004937125980703389,
      "loss": 5.4621,
      "mean_token_accuracy": 0.17768049091100693,
      "num_tokens": 19607347.0,
      "step": 8555
    },
    {
      "entropy": 5.737591218948364,
      "epoch": 0.8461842625543693,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004937043110878678,
      "loss": 5.5231,
      "mean_token_accuracy": 0.17151321321725846,
      "num_tokens": 19618614.0,
      "step": 8560
    },
    {
      "entropy": 5.729369211196899,
      "epoch": 0.8466785290628707,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004936960187252367,
      "loss": 5.5566,
      "mean_token_accuracy": 0.16908628940582277,
      "num_tokens": 19630409.0,
      "step": 8565
    },
    {
      "entropy": 5.691347980499268,
      "epoch": 0.847172795571372,
      "grad_norm": 2.125,
      "learning_rate": 0.0004936877209826497,
      "loss": 5.5892,
      "mean_token_accuracy": 0.16252141147851945,
      "num_tokens": 19642183.0,
      "step": 8570
    },
    {
      "entropy": 5.7433980941772464,
      "epoch": 0.8476670620798734,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004936794178603108,
      "loss": 5.585,
      "mean_token_accuracy": 0.17702843695878984,
      "num_tokens": 19653846.0,
      "step": 8575
    },
    {
      "entropy": 5.765725612640381,
      "epoch": 0.8481613285883749,
      "grad_norm": 2.234375,
      "learning_rate": 0.0004936711093584242,
      "loss": 5.5551,
      "mean_token_accuracy": 0.1723070040345192,
      "num_tokens": 19664183.0,
      "step": 8580
    },
    {
      "entropy": 5.719138860702515,
      "epoch": 0.8486555950968763,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004936627954771945,
      "loss": 5.552,
      "mean_token_accuracy": 0.1693572610616684,
      "num_tokens": 19675863.0,
      "step": 8585
    },
    {
      "entropy": 5.646349763870239,
      "epoch": 0.8491498616053776,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004936544762168261,
      "loss": 5.571,
      "mean_token_accuracy": 0.17368430644273758,
      "num_tokens": 19688410.0,
      "step": 8590
    },
    {
      "entropy": 5.736141061782837,
      "epoch": 0.849644128113879,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004936461515775237,
      "loss": 5.5386,
      "mean_token_accuracy": 0.16690851300954818,
      "num_tokens": 19698849.0,
      "step": 8595
    },
    {
      "entropy": 5.842454051971435,
      "epoch": 0.8501383946223804,
      "grad_norm": 1.75,
      "learning_rate": 0.000493637821559492,
      "loss": 5.7097,
      "mean_token_accuracy": 0.15919574201107026,
      "num_tokens": 19710761.0,
      "step": 8600
    },
    {
      "entropy": 5.764397382736206,
      "epoch": 0.8506326611308818,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004936294861629358,
      "loss": 5.5974,
      "mean_token_accuracy": 0.1675185576081276,
      "num_tokens": 19722530.0,
      "step": 8605
    },
    {
      "entropy": 5.670903205871582,
      "epoch": 0.8511269276393831,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004936211453880606,
      "loss": 5.512,
      "mean_token_accuracy": 0.17287883907556534,
      "num_tokens": 19733352.0,
      "step": 8610
    },
    {
      "entropy": 5.7012639999389645,
      "epoch": 0.8516211941478845,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004936127992350712,
      "loss": 5.5806,
      "mean_token_accuracy": 0.17090388685464858,
      "num_tokens": 19745071.0,
      "step": 8615
    },
    {
      "entropy": 5.742192649841309,
      "epoch": 0.8521154606563859,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004936044477041729,
      "loss": 5.5244,
      "mean_token_accuracy": 0.17320816963911057,
      "num_tokens": 19756039.0,
      "step": 8620
    },
    {
      "entropy": 5.722411012649536,
      "epoch": 0.8526097271648873,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004935960907955715,
      "loss": 5.619,
      "mean_token_accuracy": 0.16041606664657593,
      "num_tokens": 19768728.0,
      "step": 8625
    },
    {
      "entropy": 5.718424987792969,
      "epoch": 0.8531039936733887,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004935877285094722,
      "loss": 5.5978,
      "mean_token_accuracy": 0.1654672145843506,
      "num_tokens": 19780010.0,
      "step": 8630
    },
    {
      "entropy": 5.7518528461456295,
      "epoch": 0.8535982601818901,
      "grad_norm": 2.53125,
      "learning_rate": 0.000493579360846081,
      "loss": 5.5281,
      "mean_token_accuracy": 0.17875559329986573,
      "num_tokens": 19790082.0,
      "step": 8635
    },
    {
      "entropy": 5.7833168506622314,
      "epoch": 0.8540925266903915,
      "grad_norm": 2.40625,
      "learning_rate": 0.0004935709878056034,
      "loss": 5.6247,
      "mean_token_accuracy": 0.16765579134225844,
      "num_tokens": 19801702.0,
      "step": 8640
    },
    {
      "entropy": 5.710736656188965,
      "epoch": 0.8545867931988929,
      "grad_norm": 2.3125,
      "learning_rate": 0.0004935626093882457,
      "loss": 5.6742,
      "mean_token_accuracy": 0.16747738271951676,
      "num_tokens": 19812173.0,
      "step": 8645
    },
    {
      "entropy": 5.680256938934326,
      "epoch": 0.8550810597073942,
      "grad_norm": 2.265625,
      "learning_rate": 0.000493554225594214,
      "loss": 5.5398,
      "mean_token_accuracy": 0.17358147948980332,
      "num_tokens": 19823286.0,
      "step": 8650
    },
    {
      "entropy": 5.7853436946868895,
      "epoch": 0.8555753262158956,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004935458364237143,
      "loss": 5.5524,
      "mean_token_accuracy": 0.17029246538877488,
      "num_tokens": 19834472.0,
      "step": 8655
    },
    {
      "entropy": 5.784717130661011,
      "epoch": 0.856069592724397,
      "grad_norm": 2.28125,
      "learning_rate": 0.0004935374418769532,
      "loss": 5.5765,
      "mean_token_accuracy": 0.16844407021999358,
      "num_tokens": 19845073.0,
      "step": 8660
    },
    {
      "entropy": 5.765900182723999,
      "epoch": 0.8565638592328984,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004935290419541371,
      "loss": 5.6283,
      "mean_token_accuracy": 0.1618078276515007,
      "num_tokens": 19856762.0,
      "step": 8665
    },
    {
      "entropy": 5.711626052856445,
      "epoch": 0.8570581257413997,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004935206366554726,
      "loss": 5.5423,
      "mean_token_accuracy": 0.17094463855028152,
      "num_tokens": 19867358.0,
      "step": 8670
    },
    {
      "entropy": 5.702182626724243,
      "epoch": 0.8575523922499011,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004935122259811667,
      "loss": 5.5829,
      "mean_token_accuracy": 0.1757657617330551,
      "num_tokens": 19878527.0,
      "step": 8675
    },
    {
      "entropy": 5.77548394203186,
      "epoch": 0.8580466587584026,
      "grad_norm": 2.109375,
      "learning_rate": 0.000493503809931426,
      "loss": 5.6172,
      "mean_token_accuracy": 0.1653773806989193,
      "num_tokens": 19889680.0,
      "step": 8680
    },
    {
      "entropy": 5.751953983306885,
      "epoch": 0.858540925266904,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004934953885064577,
      "loss": 5.5199,
      "mean_token_accuracy": 0.17434027343988417,
      "num_tokens": 19900569.0,
      "step": 8685
    },
    {
      "entropy": 5.661594724655151,
      "epoch": 0.8590351917754053,
      "grad_norm": 1.640625,
      "learning_rate": 0.000493486961706469,
      "loss": 5.4962,
      "mean_token_accuracy": 0.17351965010166168,
      "num_tokens": 19912537.0,
      "step": 8690
    },
    {
      "entropy": 5.705906391143799,
      "epoch": 0.8595294582839067,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004934785295316672,
      "loss": 5.5611,
      "mean_token_accuracy": 0.16770893782377244,
      "num_tokens": 19923931.0,
      "step": 8695
    },
    {
      "entropy": 5.814125823974609,
      "epoch": 0.8600237247924081,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0004934700919822596,
      "loss": 5.6378,
      "mean_token_accuracy": 0.16237291246652602,
      "num_tokens": 19933956.0,
      "step": 8700
    },
    {
      "entropy": 5.696491146087647,
      "epoch": 0.8605179913009094,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004934616490584538,
      "loss": 5.5152,
      "mean_token_accuracy": 0.1765739381313324,
      "num_tokens": 19945603.0,
      "step": 8705
    },
    {
      "entropy": 5.76496958732605,
      "epoch": 0.8610122578094108,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004934532007604577,
      "loss": 5.5922,
      "mean_token_accuracy": 0.167815600335598,
      "num_tokens": 19956358.0,
      "step": 8710
    },
    {
      "entropy": 5.721701908111572,
      "epoch": 0.8615065243179122,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004934447470884788,
      "loss": 5.5212,
      "mean_token_accuracy": 0.17156184166669847,
      "num_tokens": 19967856.0,
      "step": 8715
    },
    {
      "entropy": 5.672111463546753,
      "epoch": 0.8620007908264136,
      "grad_norm": 2.328125,
      "learning_rate": 0.0004934362880427253,
      "loss": 5.5499,
      "mean_token_accuracy": 0.1714496999979019,
      "num_tokens": 19979323.0,
      "step": 8720
    },
    {
      "entropy": 5.793496084213257,
      "epoch": 0.8624950573349149,
      "grad_norm": 2.3125,
      "learning_rate": 0.0004934278236234052,
      "loss": 5.6038,
      "mean_token_accuracy": 0.1626463383436203,
      "num_tokens": 19991543.0,
      "step": 8725
    },
    {
      "entropy": 5.796078205108643,
      "epoch": 0.8629893238434164,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004934193538307268,
      "loss": 5.5933,
      "mean_token_accuracy": 0.16286921501159668,
      "num_tokens": 20003621.0,
      "step": 8730
    },
    {
      "entropy": 5.684550333023071,
      "epoch": 0.8634835903519178,
      "grad_norm": 2.515625,
      "learning_rate": 0.0004934108786648983,
      "loss": 5.5786,
      "mean_token_accuracy": 0.16030988246202468,
      "num_tokens": 20015183.0,
      "step": 8735
    },
    {
      "entropy": 5.740259408950806,
      "epoch": 0.8639778568604192,
      "grad_norm": 1.5,
      "learning_rate": 0.0004934023981261285,
      "loss": 5.5143,
      "mean_token_accuracy": 0.17002393901348115,
      "num_tokens": 20026735.0,
      "step": 8740
    },
    {
      "entropy": 5.755583667755127,
      "epoch": 0.8644721233689205,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004933939122146256,
      "loss": 5.493,
      "mean_token_accuracy": 0.16880759298801423,
      "num_tokens": 20037410.0,
      "step": 8745
    },
    {
      "entropy": 5.5991795539855955,
      "epoch": 0.8649663898774219,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004933854209305987,
      "loss": 5.4981,
      "mean_token_accuracy": 0.17080789059400558,
      "num_tokens": 20048972.0,
      "step": 8750
    },
    {
      "entropy": 5.707715702056885,
      "epoch": 0.8654606563859233,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0004933769242742566,
      "loss": 5.5287,
      "mean_token_accuracy": 0.17359218299388884,
      "num_tokens": 20059667.0,
      "step": 8755
    },
    {
      "entropy": 5.707452774047852,
      "epoch": 0.8659549228944247,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004933684222458082,
      "loss": 5.4851,
      "mean_token_accuracy": 0.17378606051206588,
      "num_tokens": 20070612.0,
      "step": 8760
    },
    {
      "entropy": 5.647929525375366,
      "epoch": 0.866449189402926,
      "grad_norm": 3.265625,
      "learning_rate": 0.0004933599148454628,
      "loss": 5.5018,
      "mean_token_accuracy": 0.17339118272066117,
      "num_tokens": 20081445.0,
      "step": 8765
    },
    {
      "entropy": 5.774861288070679,
      "epoch": 0.8669434559114274,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004933514020734296,
      "loss": 5.6464,
      "mean_token_accuracy": 0.16188667863607406,
      "num_tokens": 20094414.0,
      "step": 8770
    },
    {
      "entropy": 5.723762559890747,
      "epoch": 0.8674377224199288,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004933428839299181,
      "loss": 5.4601,
      "mean_token_accuracy": 0.1835716888308525,
      "num_tokens": 20106793.0,
      "step": 8775
    },
    {
      "entropy": 5.698265886306762,
      "epoch": 0.8679319889284303,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004933343604151376,
      "loss": 5.5529,
      "mean_token_accuracy": 0.17220603823661804,
      "num_tokens": 20118582.0,
      "step": 8780
    },
    {
      "entropy": 5.741739892959595,
      "epoch": 0.8684262554369316,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004933258315292982,
      "loss": 5.5108,
      "mean_token_accuracy": 0.1700451508164406,
      "num_tokens": 20130548.0,
      "step": 8785
    },
    {
      "entropy": 5.772466707229614,
      "epoch": 0.868920521945433,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004933172972726094,
      "loss": 5.6551,
      "mean_token_accuracy": 0.16548369973897933,
      "num_tokens": 20141821.0,
      "step": 8790
    },
    {
      "entropy": 5.714755010604859,
      "epoch": 0.8694147884539344,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004933087576452811,
      "loss": 5.5783,
      "mean_token_accuracy": 0.17281537801027297,
      "num_tokens": 20152071.0,
      "step": 8795
    },
    {
      "entropy": 5.786811208724975,
      "epoch": 0.8699090549624358,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004933002126475235,
      "loss": 5.5795,
      "mean_token_accuracy": 0.16612692028284073,
      "num_tokens": 20163780.0,
      "step": 8800
    },
    {
      "entropy": 5.738742542266846,
      "epoch": 0.8704033214709371,
      "grad_norm": 2.328125,
      "learning_rate": 0.0004932916622795468,
      "loss": 5.6145,
      "mean_token_accuracy": 0.16578862071037292,
      "num_tokens": 20175477.0,
      "step": 8805
    },
    {
      "entropy": 5.624256181716919,
      "epoch": 0.8708975879794385,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004932831065415614,
      "loss": 5.5104,
      "mean_token_accuracy": 0.17047771364450454,
      "num_tokens": 20189452.0,
      "step": 8810
    },
    {
      "entropy": 5.768142175674439,
      "epoch": 0.8713918544879399,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004932745454337778,
      "loss": 5.6011,
      "mean_token_accuracy": 0.16280566155910492,
      "num_tokens": 20201250.0,
      "step": 8815
    },
    {
      "entropy": 5.7166344165802006,
      "epoch": 0.8718861209964412,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004932659789564065,
      "loss": 5.5081,
      "mean_token_accuracy": 0.17887499630451204,
      "num_tokens": 20212394.0,
      "step": 8820
    },
    {
      "entropy": 5.729531717300415,
      "epoch": 0.8723803875049426,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004932574071096581,
      "loss": 5.5072,
      "mean_token_accuracy": 0.17219230085611342,
      "num_tokens": 20223262.0,
      "step": 8825
    },
    {
      "entropy": 5.698611259460449,
      "epoch": 0.872874654013444,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004932488298937437,
      "loss": 5.6138,
      "mean_token_accuracy": 0.1673371374607086,
      "num_tokens": 20234088.0,
      "step": 8830
    },
    {
      "entropy": 5.660132932662964,
      "epoch": 0.8733689205219455,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004932402473088743,
      "loss": 5.4931,
      "mean_token_accuracy": 0.17071036249399185,
      "num_tokens": 20245373.0,
      "step": 8835
    },
    {
      "entropy": 5.775502157211304,
      "epoch": 0.8738631870304469,
      "grad_norm": 2.984375,
      "learning_rate": 0.0004932316593552609,
      "loss": 5.6064,
      "mean_token_accuracy": 0.17334347367286682,
      "num_tokens": 20255824.0,
      "step": 8840
    },
    {
      "entropy": 5.661215305328369,
      "epoch": 0.8743574535389482,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004932230660331148,
      "loss": 5.4655,
      "mean_token_accuracy": 0.17837783545255662,
      "num_tokens": 20266576.0,
      "step": 8845
    },
    {
      "entropy": 5.790356683731079,
      "epoch": 0.8748517200474496,
      "grad_norm": 2.3125,
      "learning_rate": 0.0004932144673426475,
      "loss": 5.5982,
      "mean_token_accuracy": 0.16162995994091034,
      "num_tokens": 20277874.0,
      "step": 8850
    },
    {
      "entropy": 5.695214176177979,
      "epoch": 0.875345986555951,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004932058632840705,
      "loss": 5.5188,
      "mean_token_accuracy": 0.172012996673584,
      "num_tokens": 20288403.0,
      "step": 8855
    },
    {
      "entropy": 5.764313507080078,
      "epoch": 0.8758402530644523,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004931972538575952,
      "loss": 5.5804,
      "mean_token_accuracy": 0.16243908554315567,
      "num_tokens": 20299564.0,
      "step": 8860
    },
    {
      "entropy": 5.680908489227295,
      "epoch": 0.8763345195729537,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004931886390634338,
      "loss": 5.4998,
      "mean_token_accuracy": 0.17179912477731704,
      "num_tokens": 20312208.0,
      "step": 8865
    },
    {
      "entropy": 5.705981826782226,
      "epoch": 0.8768287860814551,
      "grad_norm": 1.828125,
      "learning_rate": 0.000493180018901798,
      "loss": 5.6039,
      "mean_token_accuracy": 0.16613852381706237,
      "num_tokens": 20323486.0,
      "step": 8870
    },
    {
      "entropy": 5.804632759094238,
      "epoch": 0.8773230525899565,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004931713933728997,
      "loss": 5.6091,
      "mean_token_accuracy": 0.16351276636123657,
      "num_tokens": 20335806.0,
      "step": 8875
    },
    {
      "entropy": 5.742422723770142,
      "epoch": 0.8778173190984578,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004931627624769515,
      "loss": 5.4946,
      "mean_token_accuracy": 0.1730387896299362,
      "num_tokens": 20347216.0,
      "step": 8880
    },
    {
      "entropy": 5.6844422817230225,
      "epoch": 0.8783115856069593,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004931541262141653,
      "loss": 5.5852,
      "mean_token_accuracy": 0.1655227795243263,
      "num_tokens": 20358938.0,
      "step": 8885
    },
    {
      "entropy": 5.708975553512573,
      "epoch": 0.8788058521154607,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004931454845847539,
      "loss": 5.5966,
      "mean_token_accuracy": 0.16430084109306337,
      "num_tokens": 20370742.0,
      "step": 8890
    },
    {
      "entropy": 5.653120803833008,
      "epoch": 0.8793001186239621,
      "grad_norm": 2.578125,
      "learning_rate": 0.0004931368375889298,
      "loss": 5.4753,
      "mean_token_accuracy": 0.17863573729991913,
      "num_tokens": 20382100.0,
      "step": 8895
    },
    {
      "entropy": 5.774899673461914,
      "epoch": 0.8797943851324634,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004931281852269054,
      "loss": 5.6282,
      "mean_token_accuracy": 0.16720043271780013,
      "num_tokens": 20393348.0,
      "step": 8900
    },
    {
      "entropy": 5.726962852478027,
      "epoch": 0.8802886516409648,
      "grad_norm": 1.8359375,
      "learning_rate": 0.000493119527498894,
      "loss": 5.4671,
      "mean_token_accuracy": 0.17663811445236205,
      "num_tokens": 20403471.0,
      "step": 8905
    },
    {
      "entropy": 5.726318407058716,
      "epoch": 0.8807829181494662,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004931108644051084,
      "loss": 5.5251,
      "mean_token_accuracy": 0.1799784854054451,
      "num_tokens": 20415146.0,
      "step": 8910
    },
    {
      "entropy": 5.739230489730835,
      "epoch": 0.8812771846579676,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004931021959457617,
      "loss": 5.5466,
      "mean_token_accuracy": 0.1642392247915268,
      "num_tokens": 20425857.0,
      "step": 8915
    },
    {
      "entropy": 5.7194342613220215,
      "epoch": 0.8817714511664689,
      "grad_norm": 1.8046875,
      "learning_rate": 0.000493093522121067,
      "loss": 5.659,
      "mean_token_accuracy": 0.16173420399427413,
      "num_tokens": 20437987.0,
      "step": 8920
    },
    {
      "entropy": 5.692117977142334,
      "epoch": 0.8822657176749703,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004930848429312379,
      "loss": 5.4962,
      "mean_token_accuracy": 0.17015793770551682,
      "num_tokens": 20449079.0,
      "step": 8925
    },
    {
      "entropy": 5.791934490203857,
      "epoch": 0.8827599841834717,
      "grad_norm": 2.203125,
      "learning_rate": 0.0004930761583764879,
      "loss": 5.6569,
      "mean_token_accuracy": 0.16063505560159683,
      "num_tokens": 20461032.0,
      "step": 8930
    },
    {
      "entropy": 5.79985876083374,
      "epoch": 0.8832542506919732,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004930674684570306,
      "loss": 5.6377,
      "mean_token_accuracy": 0.16756799966096877,
      "num_tokens": 20471825.0,
      "step": 8935
    },
    {
      "entropy": 5.73646650314331,
      "epoch": 0.8837485172004745,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004930587731730796,
      "loss": 5.5765,
      "mean_token_accuracy": 0.1716840773820877,
      "num_tokens": 20483329.0,
      "step": 8940
    },
    {
      "entropy": 5.716999340057373,
      "epoch": 0.8842427837089759,
      "grad_norm": 1.609375,
      "learning_rate": 0.000493050072524849,
      "loss": 5.5456,
      "mean_token_accuracy": 0.16333414018154144,
      "num_tokens": 20494537.0,
      "step": 8945
    },
    {
      "entropy": 5.679772901535034,
      "epoch": 0.8847370502174773,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004930413665125528,
      "loss": 5.5126,
      "mean_token_accuracy": 0.1739257425069809,
      "num_tokens": 20506205.0,
      "step": 8950
    },
    {
      "entropy": 5.771314525604248,
      "epoch": 0.8852313167259787,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004930326551364052,
      "loss": 5.5617,
      "mean_token_accuracy": 0.16719167828559875,
      "num_tokens": 20517447.0,
      "step": 8955
    },
    {
      "entropy": 5.727482795715332,
      "epoch": 0.88572558323448,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004930239383966204,
      "loss": 5.6009,
      "mean_token_accuracy": 0.169428226351738,
      "num_tokens": 20529435.0,
      "step": 8960
    },
    {
      "entropy": 5.687300539016723,
      "epoch": 0.8862198497429814,
      "grad_norm": 2.265625,
      "learning_rate": 0.0004930152162934128,
      "loss": 5.4441,
      "mean_token_accuracy": 0.1800736054778099,
      "num_tokens": 20540151.0,
      "step": 8965
    },
    {
      "entropy": 5.684276580810547,
      "epoch": 0.8867141162514828,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004930064888269973,
      "loss": 5.5866,
      "mean_token_accuracy": 0.1656283736228943,
      "num_tokens": 20552355.0,
      "step": 8970
    },
    {
      "entropy": 5.818031358718872,
      "epoch": 0.8872083827599841,
      "grad_norm": 2.421875,
      "learning_rate": 0.0004929977559975881,
      "loss": 5.5924,
      "mean_token_accuracy": 0.16680291742086412,
      "num_tokens": 20562129.0,
      "step": 8975
    },
    {
      "entropy": 5.74086275100708,
      "epoch": 0.8877026492684855,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004929890178054004,
      "loss": 5.5582,
      "mean_token_accuracy": 0.16622466892004012,
      "num_tokens": 20573232.0,
      "step": 8980
    },
    {
      "entropy": 5.725743961334229,
      "epoch": 0.888196915776987,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004929802742506491,
      "loss": 5.476,
      "mean_token_accuracy": 0.17476498931646348,
      "num_tokens": 20582986.0,
      "step": 8985
    },
    {
      "entropy": 5.77103328704834,
      "epoch": 0.8886911822854884,
      "grad_norm": 2.75,
      "learning_rate": 0.0004929715253335492,
      "loss": 5.5539,
      "mean_token_accuracy": 0.17096674740314483,
      "num_tokens": 20593619.0,
      "step": 8990
    },
    {
      "entropy": 5.672032690048217,
      "epoch": 0.8891854487939898,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0004929627710543159,
      "loss": 5.5082,
      "mean_token_accuracy": 0.17296775728464125,
      "num_tokens": 20604748.0,
      "step": 8995
    },
    {
      "entropy": 5.771025037765503,
      "epoch": 0.8896797153024911,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004929540114131647,
      "loss": 5.6207,
      "mean_token_accuracy": 0.16361732929944992,
      "num_tokens": 20615915.0,
      "step": 9000
    },
    {
      "epoch": 0.8896797153024911,
      "eval_entropy": 5.632198041655481,
      "eval_loss": 5.590260028839111,
      "eval_mean_token_accuracy": 0.17387622830081104,
      "eval_num_tokens": 20615915.0,
      "eval_runtime": 26.5769,
      "eval_samples_per_second": 1223.354,
      "eval_steps_per_second": 152.952,
      "step": 9000
    },
    {
      "entropy": 5.720098304748535,
      "epoch": 0.8901739818109925,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004929452464103109,
      "loss": 5.4836,
      "mean_token_accuracy": 0.176371668279171,
      "num_tokens": 20627127.0,
      "step": 9005
    },
    {
      "entropy": 5.78907413482666,
      "epoch": 0.8906682483194939,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004929364760459704,
      "loss": 5.5641,
      "mean_token_accuracy": 0.16439370661973954,
      "num_tokens": 20638568.0,
      "step": 9010
    },
    {
      "entropy": 5.714658117294311,
      "epoch": 0.8911625148279952,
      "grad_norm": 2.4375,
      "learning_rate": 0.0004929277003203586,
      "loss": 5.4954,
      "mean_token_accuracy": 0.17102617770433426,
      "num_tokens": 20650803.0,
      "step": 9015
    },
    {
      "entropy": 5.7387138366699215,
      "epoch": 0.8916567813364966,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004929189192336916,
      "loss": 5.6561,
      "mean_token_accuracy": 0.15735799074172974,
      "num_tokens": 20663021.0,
      "step": 9020
    },
    {
      "entropy": 5.742475605010986,
      "epoch": 0.892151047844998,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004929101327861854,
      "loss": 5.5317,
      "mean_token_accuracy": 0.1722045734524727,
      "num_tokens": 20675144.0,
      "step": 9025
    },
    {
      "entropy": 5.692175197601318,
      "epoch": 0.8926453143534994,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004929013409780562,
      "loss": 5.526,
      "mean_token_accuracy": 0.1707400217652321,
      "num_tokens": 20685941.0,
      "step": 9030
    },
    {
      "entropy": 5.626820421218872,
      "epoch": 0.8931395808620007,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004928925438095202,
      "loss": 5.5671,
      "mean_token_accuracy": 0.16542149782180787,
      "num_tokens": 20696996.0,
      "step": 9035
    },
    {
      "entropy": 5.73852801322937,
      "epoch": 0.8936338473705022,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004928837412807938,
      "loss": 5.5197,
      "mean_token_accuracy": 0.16998820304870604,
      "num_tokens": 20708571.0,
      "step": 9040
    },
    {
      "entropy": 5.754188203811646,
      "epoch": 0.8941281138790036,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004928749333920935,
      "loss": 5.5551,
      "mean_token_accuracy": 0.17240752577781676,
      "num_tokens": 20719816.0,
      "step": 9045
    },
    {
      "entropy": 5.687234401702881,
      "epoch": 0.894622380387505,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000492866120143636,
      "loss": 5.5359,
      "mean_token_accuracy": 0.16947951912879944,
      "num_tokens": 20731540.0,
      "step": 9050
    },
    {
      "entropy": 5.726993370056152,
      "epoch": 0.8951166468960063,
      "grad_norm": 2.265625,
      "learning_rate": 0.0004928573015356382,
      "loss": 5.5456,
      "mean_token_accuracy": 0.17137984484434127,
      "num_tokens": 20742949.0,
      "step": 9055
    },
    {
      "entropy": 5.729455184936524,
      "epoch": 0.8956109134045077,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004928484775683171,
      "loss": 5.5549,
      "mean_token_accuracy": 0.17138264775276185,
      "num_tokens": 20752840.0,
      "step": 9060
    },
    {
      "entropy": 5.687931203842163,
      "epoch": 0.8961051799130091,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004928396482418895,
      "loss": 5.5632,
      "mean_token_accuracy": 0.16409212499856948,
      "num_tokens": 20763772.0,
      "step": 9065
    },
    {
      "entropy": 5.720460271835327,
      "epoch": 0.8965994464215105,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004928308135565728,
      "loss": 5.5444,
      "mean_token_accuracy": 0.17190071940422058,
      "num_tokens": 20775815.0,
      "step": 9070
    },
    {
      "entropy": 5.716368055343628,
      "epoch": 0.8970937129300118,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004928219735125843,
      "loss": 5.5737,
      "mean_token_accuracy": 0.167434960603714,
      "num_tokens": 20787307.0,
      "step": 9075
    },
    {
      "entropy": 5.737261772155762,
      "epoch": 0.8975879794385132,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004928131281101415,
      "loss": 5.6537,
      "mean_token_accuracy": 0.16068975925445556,
      "num_tokens": 20798910.0,
      "step": 9080
    },
    {
      "entropy": 5.870688104629517,
      "epoch": 0.8980822459470146,
      "grad_norm": 2.5,
      "learning_rate": 0.000492804277349462,
      "loss": 5.5793,
      "mean_token_accuracy": 0.1611576959490776,
      "num_tokens": 20808628.0,
      "step": 9085
    },
    {
      "entropy": 5.729201316833496,
      "epoch": 0.8985765124555161,
      "grad_norm": 2.234375,
      "learning_rate": 0.0004927954212307634,
      "loss": 5.494,
      "mean_token_accuracy": 0.1680928349494934,
      "num_tokens": 20820117.0,
      "step": 9090
    },
    {
      "entropy": 5.589667654037475,
      "epoch": 0.8990707789640174,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004927865597542637,
      "loss": 5.4608,
      "mean_token_accuracy": 0.17886472791433333,
      "num_tokens": 20830859.0,
      "step": 9095
    },
    {
      "entropy": 5.716715574264526,
      "epoch": 0.8995650454725188,
      "grad_norm": 2.59375,
      "learning_rate": 0.0004927776929201807,
      "loss": 5.4836,
      "mean_token_accuracy": 0.17389875203371047,
      "num_tokens": 20843016.0,
      "step": 9100
    },
    {
      "entropy": 5.700129270553589,
      "epoch": 0.9000593119810202,
      "grad_norm": 2.203125,
      "learning_rate": 0.0004927688207287328,
      "loss": 5.4976,
      "mean_token_accuracy": 0.16986022442579268,
      "num_tokens": 20854873.0,
      "step": 9105
    },
    {
      "entropy": 5.771597194671631,
      "epoch": 0.9005535784895216,
      "grad_norm": 2.0,
      "learning_rate": 0.0004927599431801381,
      "loss": 5.6897,
      "mean_token_accuracy": 0.1707574725151062,
      "num_tokens": 20866756.0,
      "step": 9110
    },
    {
      "entropy": 5.737335109710694,
      "epoch": 0.9010478449980229,
      "grad_norm": 2.703125,
      "learning_rate": 0.000492751060274615,
      "loss": 5.4831,
      "mean_token_accuracy": 0.17735593020915985,
      "num_tokens": 20878859.0,
      "step": 9115
    },
    {
      "entropy": 5.673659038543701,
      "epoch": 0.9015421115065243,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004927421720123821,
      "loss": 5.5392,
      "mean_token_accuracy": 0.1698540985584259,
      "num_tokens": 20889188.0,
      "step": 9120
    },
    {
      "entropy": 5.691730403900147,
      "epoch": 0.9020363780150257,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0004927332783936579,
      "loss": 5.4922,
      "mean_token_accuracy": 0.17076885402202607,
      "num_tokens": 20900318.0,
      "step": 9125
    },
    {
      "entropy": 5.729398012161255,
      "epoch": 0.902530644523527,
      "grad_norm": 2.40625,
      "learning_rate": 0.0004927243794186614,
      "loss": 5.5728,
      "mean_token_accuracy": 0.16684664338827132,
      "num_tokens": 20912384.0,
      "step": 9130
    },
    {
      "entropy": 5.691777324676513,
      "epoch": 0.9030249110320284,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004927154750876112,
      "loss": 5.5263,
      "mean_token_accuracy": 0.16683476269245148,
      "num_tokens": 20923002.0,
      "step": 9135
    },
    {
      "entropy": 5.720336961746216,
      "epoch": 0.9035191775405299,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004927065654007268,
      "loss": 5.6612,
      "mean_token_accuracy": 0.16474491804838182,
      "num_tokens": 20934235.0,
      "step": 9140
    },
    {
      "entropy": 5.681716823577881,
      "epoch": 0.9040134440490313,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004926976503582269,
      "loss": 5.5106,
      "mean_token_accuracy": 0.16745854318141937,
      "num_tokens": 20946788.0,
      "step": 9145
    },
    {
      "entropy": 5.7602071285247805,
      "epoch": 0.9045077105575327,
      "grad_norm": 1.734375,
      "learning_rate": 0.000492688729960331,
      "loss": 5.548,
      "mean_token_accuracy": 0.17626707553863524,
      "num_tokens": 20958963.0,
      "step": 9150
    },
    {
      "entropy": 5.7697436809539795,
      "epoch": 0.905001977066034,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004926798042072587,
      "loss": 5.5258,
      "mean_token_accuracy": 0.17248127460479737,
      "num_tokens": 20969630.0,
      "step": 9155
    },
    {
      "entropy": 5.671909046173096,
      "epoch": 0.9054962435745354,
      "grad_norm": 1.625,
      "learning_rate": 0.0004926708730992294,
      "loss": 5.503,
      "mean_token_accuracy": 0.17446861267089844,
      "num_tokens": 20980489.0,
      "step": 9160
    },
    {
      "entropy": 5.77673864364624,
      "epoch": 0.9059905100830368,
      "grad_norm": 1.5,
      "learning_rate": 0.0004926619366364629,
      "loss": 5.5565,
      "mean_token_accuracy": 0.1746782347559929,
      "num_tokens": 20992199.0,
      "step": 9165
    },
    {
      "entropy": 5.753552103042603,
      "epoch": 0.9064847765915381,
      "grad_norm": 1.9765625,
      "learning_rate": 0.000492652994819179,
      "loss": 5.5754,
      "mean_token_accuracy": 0.166583052277565,
      "num_tokens": 21003680.0,
      "step": 9170
    },
    {
      "entropy": 5.733190870285034,
      "epoch": 0.9069790431000395,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004926440476475975,
      "loss": 5.484,
      "mean_token_accuracy": 0.1758432000875473,
      "num_tokens": 21014191.0,
      "step": 9175
    },
    {
      "entropy": 5.724966335296631,
      "epoch": 0.9074733096085409,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004926350951219388,
      "loss": 5.4859,
      "mean_token_accuracy": 0.17352897375822068,
      "num_tokens": 21025692.0,
      "step": 9180
    },
    {
      "entropy": 5.6547223091125485,
      "epoch": 0.9079675761170423,
      "grad_norm": 2.390625,
      "learning_rate": 0.000492626137242423,
      "loss": 5.5037,
      "mean_token_accuracy": 0.1735650509595871,
      "num_tokens": 21036643.0,
      "step": 9185
    },
    {
      "entropy": 5.712827491760254,
      "epoch": 0.9084618426255437,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004926171740092704,
      "loss": 5.5935,
      "mean_token_accuracy": 0.16322490721940994,
      "num_tokens": 21047509.0,
      "step": 9190
    },
    {
      "entropy": 5.721576261520386,
      "epoch": 0.9089561091340451,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004926082054227016,
      "loss": 5.6401,
      "mean_token_accuracy": 0.17004420459270478,
      "num_tokens": 21059854.0,
      "step": 9195
    },
    {
      "entropy": 5.7053791046142575,
      "epoch": 0.9094503756425465,
      "grad_norm": 2.1875,
      "learning_rate": 0.0004925992314829372,
      "loss": 5.4607,
      "mean_token_accuracy": 0.17410786002874373,
      "num_tokens": 21069926.0,
      "step": 9200
    },
    {
      "entropy": 5.750688266754151,
      "epoch": 0.9099446421510479,
      "grad_norm": 1.6171875,
      "learning_rate": 0.000492590252190198,
      "loss": 5.5647,
      "mean_token_accuracy": 0.1693345457315445,
      "num_tokens": 21081580.0,
      "step": 9205
    },
    {
      "entropy": 5.710610485076904,
      "epoch": 0.9104389086595492,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004925812675447047,
      "loss": 5.4592,
      "mean_token_accuracy": 0.17503467500209807,
      "num_tokens": 21092510.0,
      "step": 9210
    },
    {
      "entropy": 5.640002012252808,
      "epoch": 0.9109331751680506,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004925722775466785,
      "loss": 5.4806,
      "mean_token_accuracy": 0.17165618687868117,
      "num_tokens": 21103995.0,
      "step": 9215
    },
    {
      "entropy": 5.743108701705933,
      "epoch": 0.911427441676552,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004925632821963406,
      "loss": 5.5564,
      "mean_token_accuracy": 0.1709841176867485,
      "num_tokens": 21114949.0,
      "step": 9220
    },
    {
      "entropy": 5.728981018066406,
      "epoch": 0.9119217081850534,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004925542814939121,
      "loss": 5.4762,
      "mean_token_accuracy": 0.1855828732252121,
      "num_tokens": 21125810.0,
      "step": 9225
    },
    {
      "entropy": 5.658492136001587,
      "epoch": 0.9124159746935547,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004925452754396146,
      "loss": 5.5463,
      "mean_token_accuracy": 0.16577664613723755,
      "num_tokens": 21137728.0,
      "step": 9230
    },
    {
      "entropy": 5.7361180782318115,
      "epoch": 0.9129102412020561,
      "grad_norm": 1.625,
      "learning_rate": 0.0004925362640336696,
      "loss": 5.5207,
      "mean_token_accuracy": 0.17037809193134307,
      "num_tokens": 21149249.0,
      "step": 9235
    },
    {
      "entropy": 5.74154806137085,
      "epoch": 0.9134045077105575,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004925272472762987,
      "loss": 5.6152,
      "mean_token_accuracy": 0.16150472760200502,
      "num_tokens": 21162338.0,
      "step": 9240
    },
    {
      "entropy": 5.618438482284546,
      "epoch": 0.913898774219059,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004925182251677238,
      "loss": 5.3848,
      "mean_token_accuracy": 0.1800136461853981,
      "num_tokens": 21172953.0,
      "step": 9245
    },
    {
      "entropy": 5.699621725082397,
      "epoch": 0.9143930407275603,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004925091977081668,
      "loss": 5.5477,
      "mean_token_accuracy": 0.17176782488822936,
      "num_tokens": 21183835.0,
      "step": 9250
    },
    {
      "entropy": 5.699409437179566,
      "epoch": 0.9148873072360617,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004925001648978498,
      "loss": 5.5087,
      "mean_token_accuracy": 0.17691364139318466,
      "num_tokens": 21195392.0,
      "step": 9255
    },
    {
      "entropy": 5.810751867294312,
      "epoch": 0.9153815737445631,
      "grad_norm": 2.09375,
      "learning_rate": 0.000492491126736995,
      "loss": 5.6633,
      "mean_token_accuracy": 0.15760257691144944,
      "num_tokens": 21207213.0,
      "step": 9260
    },
    {
      "entropy": 5.801012849807739,
      "epoch": 0.9158758402530645,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004924820832258248,
      "loss": 5.6224,
      "mean_token_accuracy": 0.1684015840291977,
      "num_tokens": 21218843.0,
      "step": 9265
    },
    {
      "entropy": 5.700365972518921,
      "epoch": 0.9163701067615658,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004924730343645614,
      "loss": 5.5762,
      "mean_token_accuracy": 0.1689996376633644,
      "num_tokens": 21229596.0,
      "step": 9270
    },
    {
      "entropy": 5.770077037811279,
      "epoch": 0.9168643732700672,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004924639801534278,
      "loss": 5.5118,
      "mean_token_accuracy": 0.17067157924175264,
      "num_tokens": 21240040.0,
      "step": 9275
    },
    {
      "entropy": 5.7615241527557375,
      "epoch": 0.9173586397785686,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004924549205926465,
      "loss": 5.6402,
      "mean_token_accuracy": 0.1597430318593979,
      "num_tokens": 21252046.0,
      "step": 9280
    },
    {
      "entropy": 5.75082573890686,
      "epoch": 0.91785290628707,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004924458556824405,
      "loss": 5.5675,
      "mean_token_accuracy": 0.16260437667369843,
      "num_tokens": 21263720.0,
      "step": 9285
    },
    {
      "entropy": 5.767461442947388,
      "epoch": 0.9183471727955713,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004924367854230326,
      "loss": 5.5755,
      "mean_token_accuracy": 0.16299645304679872,
      "num_tokens": 21275751.0,
      "step": 9290
    },
    {
      "entropy": 5.7449291229248045,
      "epoch": 0.9188414393040728,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004924277098146461,
      "loss": 5.5632,
      "mean_token_accuracy": 0.17114244848489762,
      "num_tokens": 21287081.0,
      "step": 9295
    },
    {
      "entropy": 5.649151706695557,
      "epoch": 0.9193357058125742,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004924186288575043,
      "loss": 5.5221,
      "mean_token_accuracy": 0.17164077907800673,
      "num_tokens": 21298007.0,
      "step": 9300
    },
    {
      "entropy": 5.747290325164795,
      "epoch": 0.9198299723210756,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004924095425518304,
      "loss": 5.5545,
      "mean_token_accuracy": 0.167376746237278,
      "num_tokens": 21309416.0,
      "step": 9305
    },
    {
      "entropy": 5.664888668060303,
      "epoch": 0.9203242388295769,
      "grad_norm": 1.671875,
      "learning_rate": 0.000492400450897848,
      "loss": 5.5199,
      "mean_token_accuracy": 0.17147454023361205,
      "num_tokens": 21321129.0,
      "step": 9310
    },
    {
      "entropy": 5.7010034084320065,
      "epoch": 0.9208185053380783,
      "grad_norm": 1.75,
      "learning_rate": 0.0004923913538957809,
      "loss": 5.5081,
      "mean_token_accuracy": 0.17650321274995803,
      "num_tokens": 21331706.0,
      "step": 9315
    },
    {
      "entropy": 5.72475118637085,
      "epoch": 0.9213127718465797,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004923822515458526,
      "loss": 5.5453,
      "mean_token_accuracy": 0.1709713250398636,
      "num_tokens": 21343738.0,
      "step": 9320
    },
    {
      "entropy": 5.730406284332275,
      "epoch": 0.921807038355081,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004923731438482873,
      "loss": 5.5361,
      "mean_token_accuracy": 0.17120832204818726,
      "num_tokens": 21354714.0,
      "step": 9325
    },
    {
      "entropy": 5.795811319351197,
      "epoch": 0.9223013048635824,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004923640308033088,
      "loss": 5.5813,
      "mean_token_accuracy": 0.1637384533882141,
      "num_tokens": 21365947.0,
      "step": 9330
    },
    {
      "entropy": 5.835444068908691,
      "epoch": 0.9227955713720838,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004923549124111416,
      "loss": 5.708,
      "mean_token_accuracy": 0.15868209153413773,
      "num_tokens": 21376729.0,
      "step": 9335
    },
    {
      "entropy": 5.694757604598999,
      "epoch": 0.9232898378805852,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004923457886720098,
      "loss": 5.5016,
      "mean_token_accuracy": 0.1803483709692955,
      "num_tokens": 21388017.0,
      "step": 9340
    },
    {
      "entropy": 5.630907011032105,
      "epoch": 0.9237841043890866,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004923366595861378,
      "loss": 5.4776,
      "mean_token_accuracy": 0.18337951153516768,
      "num_tokens": 21399566.0,
      "step": 9345
    },
    {
      "entropy": 5.760402250289917,
      "epoch": 0.924278370897588,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004923275251537503,
      "loss": 5.5908,
      "mean_token_accuracy": 0.16621706634759903,
      "num_tokens": 21410276.0,
      "step": 9350
    },
    {
      "entropy": 5.716007423400879,
      "epoch": 0.9247726374060894,
      "grad_norm": 1.890625,
      "learning_rate": 0.000492318385375072,
      "loss": 5.5176,
      "mean_token_accuracy": 0.17807159721851348,
      "num_tokens": 21421090.0,
      "step": 9355
    },
    {
      "entropy": 5.69927749633789,
      "epoch": 0.9252669039145908,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004923092402503277,
      "loss": 5.4996,
      "mean_token_accuracy": 0.17431414425373076,
      "num_tokens": 21431057.0,
      "step": 9360
    },
    {
      "entropy": 5.734055471420288,
      "epoch": 0.9257611704230921,
      "grad_norm": 1.75,
      "learning_rate": 0.0004923000897797423,
      "loss": 5.5753,
      "mean_token_accuracy": 0.16445844024419784,
      "num_tokens": 21443849.0,
      "step": 9365
    },
    {
      "entropy": 5.802360677719117,
      "epoch": 0.9262554369315935,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004922909339635411,
      "loss": 5.6275,
      "mean_token_accuracy": 0.1604536235332489,
      "num_tokens": 21454824.0,
      "step": 9370
    },
    {
      "entropy": 5.740575313568115,
      "epoch": 0.9267497034400949,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004922817728019492,
      "loss": 5.5393,
      "mean_token_accuracy": 0.16994927674531937,
      "num_tokens": 21466082.0,
      "step": 9375
    },
    {
      "entropy": 5.7078859329223635,
      "epoch": 0.9272439699485963,
      "grad_norm": 2.375,
      "learning_rate": 0.0004922726062951918,
      "loss": 5.5113,
      "mean_token_accuracy": 0.17064426243305206,
      "num_tokens": 21477805.0,
      "step": 9380
    },
    {
      "entropy": 5.744085168838501,
      "epoch": 0.9277382364570976,
      "grad_norm": 1.5,
      "learning_rate": 0.0004922634344434948,
      "loss": 5.5207,
      "mean_token_accuracy": 0.16478744000196457,
      "num_tokens": 21490014.0,
      "step": 9385
    },
    {
      "entropy": 5.759122610092163,
      "epoch": 0.928232502965599,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004922542572470835,
      "loss": 5.5924,
      "mean_token_accuracy": 0.16787174344062805,
      "num_tokens": 21501713.0,
      "step": 9390
    },
    {
      "entropy": 5.623011350631714,
      "epoch": 0.9287267694741005,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004922450747061837,
      "loss": 5.4546,
      "mean_token_accuracy": 0.1778919905424118,
      "num_tokens": 21513258.0,
      "step": 9395
    },
    {
      "entropy": 5.682714223861694,
      "epoch": 0.9292210359826019,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004922358868210214,
      "loss": 5.5475,
      "mean_token_accuracy": 0.17112779170274733,
      "num_tokens": 21526259.0,
      "step": 9400
    },
    {
      "entropy": 5.695071601867676,
      "epoch": 0.9297153024911032,
      "grad_norm": 2.46875,
      "learning_rate": 0.0004922266935918227,
      "loss": 5.5149,
      "mean_token_accuracy": 0.16989525109529496,
      "num_tokens": 21539144.0,
      "step": 9405
    },
    {
      "entropy": 5.778148746490478,
      "epoch": 0.9302095689996046,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004922174950188136,
      "loss": 5.5996,
      "mean_token_accuracy": 0.16327397972345353,
      "num_tokens": 21550550.0,
      "step": 9410
    },
    {
      "entropy": 5.7194273471832275,
      "epoch": 0.930703835508106,
      "grad_norm": 2.5,
      "learning_rate": 0.0004922082911022204,
      "loss": 5.5015,
      "mean_token_accuracy": 0.16759830862283706,
      "num_tokens": 21562491.0,
      "step": 9415
    },
    {
      "entropy": 5.68703031539917,
      "epoch": 0.9311981020166074,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004921990818422694,
      "loss": 5.5711,
      "mean_token_accuracy": 0.16791069358587266,
      "num_tokens": 21573679.0,
      "step": 9420
    },
    {
      "entropy": 5.736935138702393,
      "epoch": 0.9316923685251087,
      "grad_norm": 2.25,
      "learning_rate": 0.0004921898672391874,
      "loss": 5.487,
      "mean_token_accuracy": 0.17206409722566604,
      "num_tokens": 21585491.0,
      "step": 9425
    },
    {
      "entropy": 5.795855474472046,
      "epoch": 0.9321866350336101,
      "grad_norm": 2.421875,
      "learning_rate": 0.0004921806472932011,
      "loss": 5.5667,
      "mean_token_accuracy": 0.17183364033699036,
      "num_tokens": 21598302.0,
      "step": 9430
    },
    {
      "entropy": 5.709362936019898,
      "epoch": 0.9326809015421115,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004921714220045371,
      "loss": 5.5457,
      "mean_token_accuracy": 0.17069749385118485,
      "num_tokens": 21608442.0,
      "step": 9435
    },
    {
      "entropy": 5.72504825592041,
      "epoch": 0.9331751680506128,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004921621913734223,
      "loss": 5.5527,
      "mean_token_accuracy": 0.1707771122455597,
      "num_tokens": 21619594.0,
      "step": 9440
    },
    {
      "entropy": 5.791470766067505,
      "epoch": 0.9336694345591142,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004921529554000841,
      "loss": 5.5785,
      "mean_token_accuracy": 0.16914348304271698,
      "num_tokens": 21631065.0,
      "step": 9445
    },
    {
      "entropy": 5.712844276428223,
      "epoch": 0.9341637010676157,
      "grad_norm": 5.28125,
      "learning_rate": 0.0004921437140847493,
      "loss": 5.4894,
      "mean_token_accuracy": 0.17622862309217452,
      "num_tokens": 21640949.0,
      "step": 9450
    },
    {
      "entropy": 5.776348352432251,
      "epoch": 0.9346579675761171,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004921344674276456,
      "loss": 5.6225,
      "mean_token_accuracy": 0.1652636095881462,
      "num_tokens": 21651189.0,
      "step": 9455
    },
    {
      "entropy": 5.8496729850769045,
      "epoch": 0.9351522340846185,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004921252154290003,
      "loss": 5.6029,
      "mean_token_accuracy": 0.1607473149895668,
      "num_tokens": 21662790.0,
      "step": 9460
    },
    {
      "entropy": 5.7654030323028564,
      "epoch": 0.9356465005931198,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000492115958089041,
      "loss": 5.6467,
      "mean_token_accuracy": 0.16289750188589097,
      "num_tokens": 21674980.0,
      "step": 9465
    },
    {
      "entropy": 5.699536752700806,
      "epoch": 0.9361407671016212,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004921066954079955,
      "loss": 5.5328,
      "mean_token_accuracy": 0.1692644625902176,
      "num_tokens": 21687721.0,
      "step": 9470
    },
    {
      "entropy": 5.763693237304688,
      "epoch": 0.9366350336101226,
      "grad_norm": 2.328125,
      "learning_rate": 0.0004920974273860915,
      "loss": 5.6703,
      "mean_token_accuracy": 0.15921183824539184,
      "num_tokens": 21699522.0,
      "step": 9475
    },
    {
      "entropy": 5.782375383377075,
      "epoch": 0.9371293001186239,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004920881540235572,
      "loss": 5.4274,
      "mean_token_accuracy": 0.17436766028404235,
      "num_tokens": 21709972.0,
      "step": 9480
    },
    {
      "entropy": 5.62567663192749,
      "epoch": 0.9376235666271253,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004920788753206205,
      "loss": 5.5437,
      "mean_token_accuracy": 0.17245012819766997,
      "num_tokens": 21722172.0,
      "step": 9485
    },
    {
      "entropy": 5.6970696449279785,
      "epoch": 0.9381178331356267,
      "grad_norm": 1.9296875,
      "learning_rate": 0.00049206959127751,
      "loss": 5.5502,
      "mean_token_accuracy": 0.16532090157270432,
      "num_tokens": 21734314.0,
      "step": 9490
    },
    {
      "entropy": 5.741184091567993,
      "epoch": 0.9386120996441281,
      "grad_norm": 2.765625,
      "learning_rate": 0.0004920603018944537,
      "loss": 5.5199,
      "mean_token_accuracy": 0.17438743263483047,
      "num_tokens": 21744031.0,
      "step": 9495
    },
    {
      "entropy": 5.715707159042358,
      "epoch": 0.9391063661526295,
      "grad_norm": 2.390625,
      "learning_rate": 0.0004920510071716803,
      "loss": 5.4969,
      "mean_token_accuracy": 0.1746805191040039,
      "num_tokens": 21755653.0,
      "step": 9500
    },
    {
      "entropy": 5.816602802276611,
      "epoch": 0.9396006326611309,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004920417071094186,
      "loss": 5.594,
      "mean_token_accuracy": 0.1655436113476753,
      "num_tokens": 21768513.0,
      "step": 9505
    },
    {
      "entropy": 5.744887208938598,
      "epoch": 0.9400948991696323,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004920324017078971,
      "loss": 5.5205,
      "mean_token_accuracy": 0.16552363485097885,
      "num_tokens": 21780857.0,
      "step": 9510
    },
    {
      "entropy": 5.665854024887085,
      "epoch": 0.9405891656781337,
      "grad_norm": 1.90625,
      "learning_rate": 0.000492023090967345,
      "loss": 5.5126,
      "mean_token_accuracy": 0.17808787375688553,
      "num_tokens": 21792514.0,
      "step": 9515
    },
    {
      "entropy": 5.8031830310821535,
      "epoch": 0.941083432186635,
      "grad_norm": 1.6953125,
      "learning_rate": 0.000492013774887991,
      "loss": 5.5716,
      "mean_token_accuracy": 0.1701519325375557,
      "num_tokens": 21803154.0,
      "step": 9520
    },
    {
      "entropy": 5.694848203659058,
      "epoch": 0.9415776986951364,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004920044534700645,
      "loss": 5.5582,
      "mean_token_accuracy": 0.16919979006052016,
      "num_tokens": 21814439.0,
      "step": 9525
    },
    {
      "entropy": 5.6712646484375,
      "epoch": 0.9420719652036378,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004919951267137949,
      "loss": 5.5032,
      "mean_token_accuracy": 0.17841036319732667,
      "num_tokens": 21826870.0,
      "step": 9530
    },
    {
      "entropy": 5.803434705734253,
      "epoch": 0.9425662317121392,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004919857946194114,
      "loss": 5.4939,
      "mean_token_accuracy": 0.17100831866264343,
      "num_tokens": 21837627.0,
      "step": 9535
    },
    {
      "entropy": 5.663749742507934,
      "epoch": 0.9430604982206405,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004919764571871438,
      "loss": 5.6254,
      "mean_token_accuracy": 0.16555337458848954,
      "num_tokens": 21848891.0,
      "step": 9540
    },
    {
      "entropy": 5.804639577865601,
      "epoch": 0.9435547647291419,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004919671144172216,
      "loss": 5.5292,
      "mean_token_accuracy": 0.16971494555473327,
      "num_tokens": 21860094.0,
      "step": 9545
    },
    {
      "entropy": 5.770897102355957,
      "epoch": 0.9440490312376434,
      "grad_norm": 1.5,
      "learning_rate": 0.0004919577663098748,
      "loss": 5.5062,
      "mean_token_accuracy": 0.17213414311408998,
      "num_tokens": 21870881.0,
      "step": 9550
    },
    {
      "entropy": 5.708716773986817,
      "epoch": 0.9445432977461448,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004919484128653333,
      "loss": 5.5525,
      "mean_token_accuracy": 0.16999221891164779,
      "num_tokens": 21882825.0,
      "step": 9555
    },
    {
      "entropy": 5.736716032028198,
      "epoch": 0.9450375642546461,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004919390540838271,
      "loss": 5.6146,
      "mean_token_accuracy": 0.16759835928678513,
      "num_tokens": 21894952.0,
      "step": 9560
    },
    {
      "entropy": 5.760945129394531,
      "epoch": 0.9455318307631475,
      "grad_norm": 1.5,
      "learning_rate": 0.0004919296899655866,
      "loss": 5.5588,
      "mean_token_accuracy": 0.16439727395772935,
      "num_tokens": 21906893.0,
      "step": 9565
    },
    {
      "entropy": 5.770154666900635,
      "epoch": 0.9460260972716489,
      "grad_norm": 1.8203125,
      "learning_rate": 0.000491920320510842,
      "loss": 5.619,
      "mean_token_accuracy": 0.16570377796888353,
      "num_tokens": 21918571.0,
      "step": 9570
    },
    {
      "entropy": 5.63245735168457,
      "epoch": 0.9465203637801503,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004919109457198239,
      "loss": 5.4869,
      "mean_token_accuracy": 0.17033022046089172,
      "num_tokens": 21929488.0,
      "step": 9575
    },
    {
      "entropy": 5.692418336868286,
      "epoch": 0.9470146302886516,
      "grad_norm": 2.625,
      "learning_rate": 0.0004919015655927629,
      "loss": 5.4875,
      "mean_token_accuracy": 0.16792791038751603,
      "num_tokens": 21940368.0,
      "step": 9580
    },
    {
      "entropy": 5.777459001541137,
      "epoch": 0.947508896797153,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004918921801298897,
      "loss": 5.6152,
      "mean_token_accuracy": 0.16874023675918579,
      "num_tokens": 21951171.0,
      "step": 9585
    },
    {
      "entropy": 5.767289924621582,
      "epoch": 0.9480031633056544,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004918827893314353,
      "loss": 5.5649,
      "mean_token_accuracy": 0.1750333920121193,
      "num_tokens": 21962312.0,
      "step": 9590
    },
    {
      "entropy": 5.7734699726104735,
      "epoch": 0.9484974298141557,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004918733931976305,
      "loss": 5.4953,
      "mean_token_accuracy": 0.17125972658395766,
      "num_tokens": 21971994.0,
      "step": 9595
    },
    {
      "entropy": 5.778622150421143,
      "epoch": 0.9489916963226572,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004918639917287067,
      "loss": 5.593,
      "mean_token_accuracy": 0.16450224071741104,
      "num_tokens": 21983084.0,
      "step": 9600
    },
    {
      "entropy": 5.701753854751587,
      "epoch": 0.9494859628311586,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0004918545849248951,
      "loss": 5.6045,
      "mean_token_accuracy": 0.16481069773435592,
      "num_tokens": 21995005.0,
      "step": 9605
    },
    {
      "entropy": 5.677874517440796,
      "epoch": 0.94998022933966,
      "grad_norm": 1.75,
      "learning_rate": 0.000491845172786427,
      "loss": 5.5477,
      "mean_token_accuracy": 0.17119743973016738,
      "num_tokens": 22006377.0,
      "step": 9610
    },
    {
      "entropy": 5.747830867767334,
      "epoch": 0.9504744958481613,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004918357553135339,
      "loss": 5.6163,
      "mean_token_accuracy": 0.167357137799263,
      "num_tokens": 22019663.0,
      "step": 9615
    },
    {
      "entropy": 5.847278070449829,
      "epoch": 0.9509687623566627,
      "grad_norm": 1.9375,
      "learning_rate": 0.0004918263325064476,
      "loss": 5.6026,
      "mean_token_accuracy": 0.16540860384702682,
      "num_tokens": 22030646.0,
      "step": 9620
    },
    {
      "entropy": 5.759588670730591,
      "epoch": 0.9514630288651641,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004918169043654,
      "loss": 5.5325,
      "mean_token_accuracy": 0.17217915654182434,
      "num_tokens": 22043563.0,
      "step": 9625
    },
    {
      "entropy": 5.673029232025146,
      "epoch": 0.9519572953736655,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004918074708906229,
      "loss": 5.506,
      "mean_token_accuracy": 0.17214220017194748,
      "num_tokens": 22055224.0,
      "step": 9630
    },
    {
      "entropy": 5.713845300674438,
      "epoch": 0.9524515618821668,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004917980320823483,
      "loss": 5.5903,
      "mean_token_accuracy": 0.16551012098789214,
      "num_tokens": 22067445.0,
      "step": 9635
    },
    {
      "entropy": 5.742143154144287,
      "epoch": 0.9529458283906682,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004917885879408085,
      "loss": 5.502,
      "mean_token_accuracy": 0.17228208035230635,
      "num_tokens": 22078565.0,
      "step": 9640
    },
    {
      "entropy": 5.605641555786133,
      "epoch": 0.9534400948991696,
      "grad_norm": 1.875,
      "learning_rate": 0.0004917791384662359,
      "loss": 5.4062,
      "mean_token_accuracy": 0.1809736594557762,
      "num_tokens": 22088658.0,
      "step": 9645
    },
    {
      "entropy": 5.66534013748169,
      "epoch": 0.953934361407671,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004917696836588628,
      "loss": 5.5396,
      "mean_token_accuracy": 0.16359647661447524,
      "num_tokens": 22100747.0,
      "step": 9650
    },
    {
      "entropy": 5.658927488327026,
      "epoch": 0.9544286279161724,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004917602235189218,
      "loss": 5.5572,
      "mean_token_accuracy": 0.16690735518932343,
      "num_tokens": 22113559.0,
      "step": 9655
    },
    {
      "entropy": 5.807874345779419,
      "epoch": 0.9549228944246738,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004917507580466457,
      "loss": 5.558,
      "mean_token_accuracy": 0.16766222715377807,
      "num_tokens": 22124678.0,
      "step": 9660
    },
    {
      "entropy": 5.729693412780762,
      "epoch": 0.9554171609331752,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004917412872422674,
      "loss": 5.5451,
      "mean_token_accuracy": 0.17450609058141708,
      "num_tokens": 22136462.0,
      "step": 9665
    },
    {
      "entropy": 5.695987272262573,
      "epoch": 0.9559114274416766,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004917318111060198,
      "loss": 5.5553,
      "mean_token_accuracy": 0.17311072945594788,
      "num_tokens": 22149101.0,
      "step": 9670
    },
    {
      "entropy": 5.716809606552124,
      "epoch": 0.9564056939501779,
      "grad_norm": 2.0625,
      "learning_rate": 0.000491722329638136,
      "loss": 5.5018,
      "mean_token_accuracy": 0.17296547144651414,
      "num_tokens": 22159574.0,
      "step": 9675
    },
    {
      "entropy": 5.746746778488159,
      "epoch": 0.9568999604586793,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004917128428388492,
      "loss": 5.5792,
      "mean_token_accuracy": 0.17071614861488343,
      "num_tokens": 22172308.0,
      "step": 9680
    },
    {
      "entropy": 5.729102516174317,
      "epoch": 0.9573942269671807,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004917033507083929,
      "loss": 5.5644,
      "mean_token_accuracy": 0.16269811391830444,
      "num_tokens": 22183714.0,
      "step": 9685
    },
    {
      "entropy": 5.662445402145385,
      "epoch": 0.9578884934756821,
      "grad_norm": 3.765625,
      "learning_rate": 0.0004916938532470006,
      "loss": 5.4123,
      "mean_token_accuracy": 0.17422082275152206,
      "num_tokens": 22194785.0,
      "step": 9690
    },
    {
      "entropy": 5.710245752334595,
      "epoch": 0.9583827599841834,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004916843504549058,
      "loss": 5.5693,
      "mean_token_accuracy": 0.17288799285888673,
      "num_tokens": 22205998.0,
      "step": 9695
    },
    {
      "entropy": 5.723492574691773,
      "epoch": 0.9588770264926848,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004916748423323424,
      "loss": 5.5291,
      "mean_token_accuracy": 0.16438755691051482,
      "num_tokens": 22216114.0,
      "step": 9700
    },
    {
      "entropy": 5.6744773387908936,
      "epoch": 0.9593712930011863,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004916653288795443,
      "loss": 5.4928,
      "mean_token_accuracy": 0.17621395885944366,
      "num_tokens": 22226793.0,
      "step": 9705
    },
    {
      "entropy": 5.718301105499267,
      "epoch": 0.9598655595096877,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004916558100967455,
      "loss": 5.6036,
      "mean_token_accuracy": 0.17109611630439758,
      "num_tokens": 22237444.0,
      "step": 9710
    },
    {
      "entropy": 5.724616670608521,
      "epoch": 0.960359826018189,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004916462859841801,
      "loss": 5.5622,
      "mean_token_accuracy": 0.16795298755168914,
      "num_tokens": 22250704.0,
      "step": 9715
    },
    {
      "entropy": 5.742507457733154,
      "epoch": 0.9608540925266904,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004916367565420826,
      "loss": 5.5823,
      "mean_token_accuracy": 0.17305338829755784,
      "num_tokens": 22264533.0,
      "step": 9720
    },
    {
      "entropy": 5.818283224105835,
      "epoch": 0.9613483590351918,
      "grad_norm": 1.75,
      "learning_rate": 0.0004916272217706871,
      "loss": 5.5864,
      "mean_token_accuracy": 0.16755596548318863,
      "num_tokens": 22276510.0,
      "step": 9725
    },
    {
      "entropy": 5.765133094787598,
      "epoch": 0.9618426255436932,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0004916176816702284,
      "loss": 5.6159,
      "mean_token_accuracy": 0.16529442965984345,
      "num_tokens": 22288634.0,
      "step": 9730
    },
    {
      "entropy": 5.789472341537476,
      "epoch": 0.9623368920521945,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004916081362409411,
      "loss": 5.5959,
      "mean_token_accuracy": 0.16372036784887314,
      "num_tokens": 22299781.0,
      "step": 9735
    },
    {
      "entropy": 5.670352220535278,
      "epoch": 0.9628311585606959,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00049159858548306,
      "loss": 5.4344,
      "mean_token_accuracy": 0.1774035409092903,
      "num_tokens": 22310666.0,
      "step": 9740
    },
    {
      "entropy": 5.744387197494507,
      "epoch": 0.9633254250691973,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004915890293968202,
      "loss": 5.6118,
      "mean_token_accuracy": 0.16391652673482895,
      "num_tokens": 22322811.0,
      "step": 9745
    },
    {
      "entropy": 5.748639297485352,
      "epoch": 0.9638196915776986,
      "grad_norm": 2.28125,
      "learning_rate": 0.0004915794679824567,
      "loss": 5.5789,
      "mean_token_accuracy": 0.16041059792041779,
      "num_tokens": 22335017.0,
      "step": 9750
    },
    {
      "entropy": 5.667694616317749,
      "epoch": 0.9643139580862001,
      "grad_norm": 2.453125,
      "learning_rate": 0.0004915699012402047,
      "loss": 5.4992,
      "mean_token_accuracy": 0.1760917529463768,
      "num_tokens": 22346250.0,
      "step": 9755
    },
    {
      "entropy": 5.755313062667847,
      "epoch": 0.9648082245947015,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004915603291702995,
      "loss": 5.5462,
      "mean_token_accuracy": 0.17052748203277587,
      "num_tokens": 22357436.0,
      "step": 9760
    },
    {
      "entropy": 5.755286169052124,
      "epoch": 0.9653024911032029,
      "grad_norm": 2.546875,
      "learning_rate": 0.0004915507517729765,
      "loss": 5.4888,
      "mean_token_accuracy": 0.174141626060009,
      "num_tokens": 22368312.0,
      "step": 9765
    },
    {
      "entropy": 5.719630479812622,
      "epoch": 0.9657967576117042,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004915411690484715,
      "loss": 5.5166,
      "mean_token_accuracy": 0.1709474354982376,
      "num_tokens": 22378659.0,
      "step": 9770
    },
    {
      "entropy": 5.720451307296753,
      "epoch": 0.9662910241202056,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004915315809970202,
      "loss": 5.5186,
      "mean_token_accuracy": 0.17279943972826003,
      "num_tokens": 22390642.0,
      "step": 9775
    },
    {
      "entropy": 5.779125785827636,
      "epoch": 0.966785290628707,
      "grad_norm": 2.328125,
      "learning_rate": 0.0004915219876188583,
      "loss": 5.6254,
      "mean_token_accuracy": 0.16319596916437148,
      "num_tokens": 22402484.0,
      "step": 9780
    },
    {
      "entropy": 5.709197664260865,
      "epoch": 0.9672795571372084,
      "grad_norm": 2.34375,
      "learning_rate": 0.0004915123889142221,
      "loss": 5.5234,
      "mean_token_accuracy": 0.17170323580503463,
      "num_tokens": 22413315.0,
      "step": 9785
    },
    {
      "entropy": 5.778808069229126,
      "epoch": 0.9677738236457097,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004915027848833472,
      "loss": 5.5834,
      "mean_token_accuracy": 0.16111094057559966,
      "num_tokens": 22425123.0,
      "step": 9790
    },
    {
      "entropy": 5.730625629425049,
      "epoch": 0.9682680901542111,
      "grad_norm": 2.359375,
      "learning_rate": 0.0004914931755264705,
      "loss": 5.4801,
      "mean_token_accuracy": 0.17246174812316895,
      "num_tokens": 22436437.0,
      "step": 9795
    },
    {
      "entropy": 5.721488285064697,
      "epoch": 0.9687623566627125,
      "grad_norm": 1.671875,
      "learning_rate": 0.000491483560843828,
      "loss": 5.4722,
      "mean_token_accuracy": 0.17317120879888534,
      "num_tokens": 22446677.0,
      "step": 9800
    },
    {
      "entropy": 5.724382066726685,
      "epoch": 0.969256623171214,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004914739408356563,
      "loss": 5.5369,
      "mean_token_accuracy": 0.17374182641506195,
      "num_tokens": 22457597.0,
      "step": 9805
    },
    {
      "entropy": 5.784380340576172,
      "epoch": 0.9697508896797153,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004914643155021921,
      "loss": 5.6127,
      "mean_token_accuracy": 0.17223210036754608,
      "num_tokens": 22469715.0,
      "step": 9810
    },
    {
      "entropy": 5.738711214065551,
      "epoch": 0.9702451561882167,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004914546848436722,
      "loss": 5.5637,
      "mean_token_accuracy": 0.16941601932048797,
      "num_tokens": 22479840.0,
      "step": 9815
    },
    {
      "entropy": 5.689458894729614,
      "epoch": 0.9707394226967181,
      "grad_norm": 7.0625,
      "learning_rate": 0.0004914450488603334,
      "loss": 5.4809,
      "mean_token_accuracy": 0.17289411127567292,
      "num_tokens": 22491017.0,
      "step": 9820
    },
    {
      "entropy": 5.727369022369385,
      "epoch": 0.9712336892052195,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004914354075524129,
      "loss": 5.4819,
      "mean_token_accuracy": 0.17074205428361894,
      "num_tokens": 22502308.0,
      "step": 9825
    },
    {
      "entropy": 5.597308731079101,
      "epoch": 0.9717279557137208,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004914257609201477,
      "loss": 5.4226,
      "mean_token_accuracy": 0.18115499317646028,
      "num_tokens": 22513090.0,
      "step": 9830
    },
    {
      "entropy": 5.680245494842529,
      "epoch": 0.9722222222222222,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004914161089637752,
      "loss": 5.4995,
      "mean_token_accuracy": 0.17669508904218673,
      "num_tokens": 22523391.0,
      "step": 9835
    },
    {
      "entropy": 5.68854022026062,
      "epoch": 0.9727164887307236,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004914064516835329,
      "loss": 5.4841,
      "mean_token_accuracy": 0.1739958181977272,
      "num_tokens": 22536038.0,
      "step": 9840
    },
    {
      "entropy": 5.7563879013061525,
      "epoch": 0.973210755239225,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004913967890796583,
      "loss": 5.5307,
      "mean_token_accuracy": 0.16982893347740174,
      "num_tokens": 22546897.0,
      "step": 9845
    },
    {
      "entropy": 5.715967416763306,
      "epoch": 0.9737050217477263,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004913871211523891,
      "loss": 5.5775,
      "mean_token_accuracy": 0.16490885317325593,
      "num_tokens": 22558240.0,
      "step": 9850
    },
    {
      "entropy": 5.6967414855957035,
      "epoch": 0.9741992882562278,
      "grad_norm": 2.4375,
      "learning_rate": 0.0004913774479019633,
      "loss": 5.5312,
      "mean_token_accuracy": 0.17213609218597412,
      "num_tokens": 22570540.0,
      "step": 9855
    },
    {
      "entropy": 5.81184492111206,
      "epoch": 0.9746935547647292,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004913677693286185,
      "loss": 5.6159,
      "mean_token_accuracy": 0.16770697087049485,
      "num_tokens": 22582657.0,
      "step": 9860
    },
    {
      "entropy": 5.81785569190979,
      "epoch": 0.9751878212732306,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004913580854325932,
      "loss": 5.6541,
      "mean_token_accuracy": 0.16404767334461212,
      "num_tokens": 22593445.0,
      "step": 9865
    },
    {
      "entropy": 5.690832710266113,
      "epoch": 0.9756820877817319,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004913483962141253,
      "loss": 5.501,
      "mean_token_accuracy": 0.16884115636348723,
      "num_tokens": 22604631.0,
      "step": 9870
    },
    {
      "entropy": 5.682158422470093,
      "epoch": 0.9761763542902333,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004913387016734533,
      "loss": 5.5136,
      "mean_token_accuracy": 0.17531213015317917,
      "num_tokens": 22615130.0,
      "step": 9875
    },
    {
      "entropy": 5.705617094039917,
      "epoch": 0.9766706207987347,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004913290018108157,
      "loss": 5.5195,
      "mean_token_accuracy": 0.1676092565059662,
      "num_tokens": 22627372.0,
      "step": 9880
    },
    {
      "entropy": 5.794613313674927,
      "epoch": 0.977164887307236,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004913192966264511,
      "loss": 5.6028,
      "mean_token_accuracy": 0.16676058620214462,
      "num_tokens": 22640354.0,
      "step": 9885
    },
    {
      "entropy": 5.773149681091309,
      "epoch": 0.9776591538157374,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004913095861205982,
      "loss": 5.603,
      "mean_token_accuracy": 0.1643069788813591,
      "num_tokens": 22651717.0,
      "step": 9890
    },
    {
      "entropy": 5.84906964302063,
      "epoch": 0.9781534203242388,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004912998702934961,
      "loss": 5.6459,
      "mean_token_accuracy": 0.16125912070274354,
      "num_tokens": 22663334.0,
      "step": 9895
    },
    {
      "entropy": 5.792688226699829,
      "epoch": 0.9786476868327402,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004912901491453835,
      "loss": 5.545,
      "mean_token_accuracy": 0.17154308557510375,
      "num_tokens": 22675894.0,
      "step": 9900
    },
    {
      "entropy": 5.651049518585205,
      "epoch": 0.9791419533412415,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004912804226764997,
      "loss": 5.4765,
      "mean_token_accuracy": 0.18328205943107606,
      "num_tokens": 22687861.0,
      "step": 9905
    },
    {
      "entropy": 5.710421419143676,
      "epoch": 0.979636219849743,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004912706908870838,
      "loss": 5.4763,
      "mean_token_accuracy": 0.1715717598795891,
      "num_tokens": 22698346.0,
      "step": 9910
    },
    {
      "entropy": 5.703184700012207,
      "epoch": 0.9801304863582444,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004912609537773755,
      "loss": 5.5988,
      "mean_token_accuracy": 0.16239245235919952,
      "num_tokens": 22710665.0,
      "step": 9915
    },
    {
      "entropy": 5.711629629135132,
      "epoch": 0.9806247528667458,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004912512113476142,
      "loss": 5.5132,
      "mean_token_accuracy": 0.17588357776403427,
      "num_tokens": 22720989.0,
      "step": 9920
    },
    {
      "entropy": 5.689323139190674,
      "epoch": 0.9811190193752471,
      "grad_norm": 2.59375,
      "learning_rate": 0.0004912414635980395,
      "loss": 5.5147,
      "mean_token_accuracy": 0.17529048919677734,
      "num_tokens": 22732265.0,
      "step": 9925
    },
    {
      "entropy": 5.712391090393067,
      "epoch": 0.9816132858837485,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004912317105288912,
      "loss": 5.5162,
      "mean_token_accuracy": 0.16600595265626908,
      "num_tokens": 22742435.0,
      "step": 9930
    },
    {
      "entropy": 5.696404027938843,
      "epoch": 0.9821075523922499,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004912219521404094,
      "loss": 5.4883,
      "mean_token_accuracy": 0.17258599251508713,
      "num_tokens": 22753750.0,
      "step": 9935
    },
    {
      "entropy": 5.778047513961792,
      "epoch": 0.9826018189007513,
      "grad_norm": 4.46875,
      "learning_rate": 0.000491212188432834,
      "loss": 5.5516,
      "mean_token_accuracy": 0.16767684519290924,
      "num_tokens": 22765549.0,
      "step": 9940
    },
    {
      "entropy": 5.719587564468384,
      "epoch": 0.9830960854092526,
      "grad_norm": 2.453125,
      "learning_rate": 0.0004912024194064052,
      "loss": 5.611,
      "mean_token_accuracy": 0.16390867978334428,
      "num_tokens": 22777882.0,
      "step": 9945
    },
    {
      "entropy": 5.779613828659057,
      "epoch": 0.983590351917754,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004911926450613633,
      "loss": 5.6056,
      "mean_token_accuracy": 0.16340621560811996,
      "num_tokens": 22789671.0,
      "step": 9950
    },
    {
      "entropy": 5.783669090270996,
      "epoch": 0.9840846184262554,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004911828653979488,
      "loss": 5.5757,
      "mean_token_accuracy": 0.16210843324661256,
      "num_tokens": 22801330.0,
      "step": 9955
    },
    {
      "entropy": 5.725742197036743,
      "epoch": 0.9845788849347569,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004911730804164024,
      "loss": 5.5996,
      "mean_token_accuracy": 0.16601726561784744,
      "num_tokens": 22813998.0,
      "step": 9960
    },
    {
      "entropy": 5.732435894012451,
      "epoch": 0.9850731514432582,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004911632901169645,
      "loss": 5.5312,
      "mean_token_accuracy": 0.17036539912223816,
      "num_tokens": 22825048.0,
      "step": 9965
    },
    {
      "entropy": 5.63055191040039,
      "epoch": 0.9855674179517596,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004911534944998762,
      "loss": 5.4513,
      "mean_token_accuracy": 0.17461760640144347,
      "num_tokens": 22835450.0,
      "step": 9970
    },
    {
      "entropy": 5.666855812072754,
      "epoch": 0.986061684460261,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004911436935653785,
      "loss": 5.4842,
      "mean_token_accuracy": 0.1743703842163086,
      "num_tokens": 22847320.0,
      "step": 9975
    },
    {
      "entropy": 5.726233386993409,
      "epoch": 0.9865559509687624,
      "grad_norm": 2.1875,
      "learning_rate": 0.0004911338873137122,
      "loss": 5.4607,
      "mean_token_accuracy": 0.17830847352743148,
      "num_tokens": 22857676.0,
      "step": 9980
    },
    {
      "entropy": 5.659705543518067,
      "epoch": 0.9870502174772637,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000491124075745119,
      "loss": 5.4851,
      "mean_token_accuracy": 0.17840947955846786,
      "num_tokens": 22869767.0,
      "step": 9985
    },
    {
      "entropy": 5.7767034530639645,
      "epoch": 0.9875444839857651,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004911142588598399,
      "loss": 5.5869,
      "mean_token_accuracy": 0.17025674283504486,
      "num_tokens": 22880096.0,
      "step": 9990
    },
    {
      "entropy": 5.801022672653199,
      "epoch": 0.9880387504942665,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004911044366581165,
      "loss": 5.4958,
      "mean_token_accuracy": 0.17515075951814651,
      "num_tokens": 22892436.0,
      "step": 9995
    },
    {
      "entropy": 5.759911441802979,
      "epoch": 0.9885330170027679,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004910946091401903,
      "loss": 5.5384,
      "mean_token_accuracy": 0.16783835142850875,
      "num_tokens": 22904641.0,
      "step": 10000
    },
    {
      "entropy": 5.6862218379974365,
      "epoch": 0.9890272835112692,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004910847763063033,
      "loss": 5.3948,
      "mean_token_accuracy": 0.17469682544469833,
      "num_tokens": 22915372.0,
      "step": 10005
    },
    {
      "entropy": 5.756756591796875,
      "epoch": 0.9895215500197707,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004910749381566973,
      "loss": 5.6249,
      "mean_token_accuracy": 0.15831247717142105,
      "num_tokens": 22926583.0,
      "step": 10010
    },
    {
      "entropy": 5.726555013656617,
      "epoch": 0.9900158165282721,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004910650946916143,
      "loss": 5.4904,
      "mean_token_accuracy": 0.17097898572683334,
      "num_tokens": 22938103.0,
      "step": 10015
    },
    {
      "entropy": 5.692210149765015,
      "epoch": 0.9905100830367735,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004910552459112964,
      "loss": 5.5445,
      "mean_token_accuracy": 0.17311066538095474,
      "num_tokens": 22949475.0,
      "step": 10020
    },
    {
      "entropy": 5.73862590789795,
      "epoch": 0.9910043495452748,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004910453918159861,
      "loss": 5.5659,
      "mean_token_accuracy": 0.16979775428771973,
      "num_tokens": 22961404.0,
      "step": 10025
    },
    {
      "entropy": 5.722372436523438,
      "epoch": 0.9914986160537762,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004910355324059255,
      "loss": 5.5298,
      "mean_token_accuracy": 0.17193541228771209,
      "num_tokens": 22973009.0,
      "step": 10030
    },
    {
      "entropy": 5.701122713088989,
      "epoch": 0.9919928825622776,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004910256676813572,
      "loss": 5.576,
      "mean_token_accuracy": 0.17126102447509767,
      "num_tokens": 22985104.0,
      "step": 10035
    },
    {
      "entropy": 5.713639116287231,
      "epoch": 0.992487149070779,
      "grad_norm": 2.421875,
      "learning_rate": 0.0004910157976425241,
      "loss": 5.5728,
      "mean_token_accuracy": 0.1672808587551117,
      "num_tokens": 22997133.0,
      "step": 10040
    },
    {
      "entropy": 5.697215032577515,
      "epoch": 0.9929814155792803,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004910059222896689,
      "loss": 5.4149,
      "mean_token_accuracy": 0.17880269140005112,
      "num_tokens": 23007621.0,
      "step": 10045
    },
    {
      "entropy": 5.758591985702514,
      "epoch": 0.9934756820877817,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004909960416230344,
      "loss": 5.5798,
      "mean_token_accuracy": 0.1678140103816986,
      "num_tokens": 23018254.0,
      "step": 10050
    },
    {
      "entropy": 5.6953295230865475,
      "epoch": 0.9939699485962831,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004909861556428638,
      "loss": 5.4975,
      "mean_token_accuracy": 0.17135821878910065,
      "num_tokens": 23030733.0,
      "step": 10055
    },
    {
      "entropy": 5.6859265804290775,
      "epoch": 0.9944642151047846,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004909762643494003,
      "loss": 5.4804,
      "mean_token_accuracy": 0.16995348781347275,
      "num_tokens": 23043086.0,
      "step": 10060
    },
    {
      "entropy": 5.784011268615723,
      "epoch": 0.9949584816132859,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004909663677428872,
      "loss": 5.6515,
      "mean_token_accuracy": 0.16197943538427353,
      "num_tokens": 23055281.0,
      "step": 10065
    },
    {
      "entropy": 5.6577071189880375,
      "epoch": 0.9954527481217873,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004909564658235678,
      "loss": 5.531,
      "mean_token_accuracy": 0.17189735770225525,
      "num_tokens": 23067332.0,
      "step": 10070
    },
    {
      "entropy": 5.707271242141724,
      "epoch": 0.9959470146302887,
      "grad_norm": 1.75,
      "learning_rate": 0.000490946558591686,
      "loss": 5.5062,
      "mean_token_accuracy": 0.17119333148002625,
      "num_tokens": 23079278.0,
      "step": 10075
    },
    {
      "entropy": 5.704014110565185,
      "epoch": 0.99644128113879,
      "grad_norm": 1.75,
      "learning_rate": 0.0004909366460474852,
      "loss": 5.5154,
      "mean_token_accuracy": 0.1689569190144539,
      "num_tokens": 23092177.0,
      "step": 10080
    },
    {
      "entropy": 5.638453149795533,
      "epoch": 0.9969355476472914,
      "grad_norm": 1.625,
      "learning_rate": 0.0004909267281912095,
      "loss": 5.5239,
      "mean_token_accuracy": 0.17321101874113082,
      "num_tokens": 23103249.0,
      "step": 10085
    },
    {
      "entropy": 5.669422817230225,
      "epoch": 0.9974298141557928,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004909168050231029,
      "loss": 5.5578,
      "mean_token_accuracy": 0.1760734960436821,
      "num_tokens": 23113821.0,
      "step": 10090
    },
    {
      "entropy": 5.777122926712036,
      "epoch": 0.9979240806642942,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004909068765434093,
      "loss": 5.5645,
      "mean_token_accuracy": 0.16094685047864915,
      "num_tokens": 23124819.0,
      "step": 10095
    },
    {
      "entropy": 5.758470964431763,
      "epoch": 0.9984183471727955,
      "grad_norm": 1.7421875,
      "learning_rate": 0.000490896942752373,
      "loss": 5.5615,
      "mean_token_accuracy": 0.16750091016292573,
      "num_tokens": 23135590.0,
      "step": 10100
    },
    {
      "entropy": 5.721115493774414,
      "epoch": 0.9989126136812969,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004908870036502384,
      "loss": 5.5313,
      "mean_token_accuracy": 0.1731726974248886,
      "num_tokens": 23147290.0,
      "step": 10105
    },
    {
      "entropy": 5.744794416427612,
      "epoch": 0.9994068801897983,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004908770592372501,
      "loss": 5.5257,
      "mean_token_accuracy": 0.17112929821014405,
      "num_tokens": 23158276.0,
      "step": 10110
    },
    {
      "entropy": 5.714811325073242,
      "epoch": 0.9999011466982998,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004908671095136526,
      "loss": 5.5044,
      "mean_token_accuracy": 0.1722613126039505,
      "num_tokens": 23169174.0,
      "step": 10115
    },
    {
      "entropy": 5.721333026885986,
      "epoch": 1.0003954132068011,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004908571544796908,
      "loss": 5.5284,
      "mean_token_accuracy": 0.16843016892671586,
      "num_tokens": 23180377.0,
      "step": 10120
    },
    {
      "entropy": 5.708949089050293,
      "epoch": 1.0008896797153024,
      "grad_norm": 2.375,
      "learning_rate": 0.0004908471941356094,
      "loss": 5.4766,
      "mean_token_accuracy": 0.17432643324136735,
      "num_tokens": 23191158.0,
      "step": 10125
    },
    {
      "entropy": 5.743306827545166,
      "epoch": 1.001383946223804,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004908372284816537,
      "loss": 5.4925,
      "mean_token_accuracy": 0.1740867018699646,
      "num_tokens": 23201956.0,
      "step": 10130
    },
    {
      "entropy": 5.684691715240478,
      "epoch": 1.0018782127323052,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004908272575180685,
      "loss": 5.4276,
      "mean_token_accuracy": 0.1736741080880165,
      "num_tokens": 23213449.0,
      "step": 10135
    },
    {
      "entropy": 5.775772142410278,
      "epoch": 1.0023724792408066,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004908172812450993,
      "loss": 5.5231,
      "mean_token_accuracy": 0.15992673933506013,
      "num_tokens": 23226514.0,
      "step": 10140
    },
    {
      "entropy": 5.662961006164551,
      "epoch": 1.0028667457493081,
      "grad_norm": 1.5,
      "learning_rate": 0.0004908072996629917,
      "loss": 5.445,
      "mean_token_accuracy": 0.16994922310113908,
      "num_tokens": 23239322.0,
      "step": 10145
    },
    {
      "entropy": 5.707709884643554,
      "epoch": 1.0033610122578094,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004907973127719909,
      "loss": 5.4795,
      "mean_token_accuracy": 0.17108853608369828,
      "num_tokens": 23250990.0,
      "step": 10150
    },
    {
      "entropy": 5.72198257446289,
      "epoch": 1.0038552787663109,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004907873205723427,
      "loss": 5.5102,
      "mean_token_accuracy": 0.16593290269374847,
      "num_tokens": 23261630.0,
      "step": 10155
    },
    {
      "entropy": 5.7621190547943115,
      "epoch": 1.0043495452748121,
      "grad_norm": 1.78125,
      "learning_rate": 0.000490777323064293,
      "loss": 5.4764,
      "mean_token_accuracy": 0.16607781499624252,
      "num_tokens": 23272823.0,
      "step": 10160
    },
    {
      "entropy": 5.625604629516602,
      "epoch": 1.0048438117833136,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004907673202480876,
      "loss": 5.2859,
      "mean_token_accuracy": 0.19299798607826232,
      "num_tokens": 23282827.0,
      "step": 10165
    },
    {
      "entropy": 5.698440217971802,
      "epoch": 1.0053380782918149,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004907573121239727,
      "loss": 5.4919,
      "mean_token_accuracy": 0.17170169204473495,
      "num_tokens": 23295359.0,
      "step": 10170
    },
    {
      "entropy": 5.697689151763916,
      "epoch": 1.0058323448003164,
      "grad_norm": 2.328125,
      "learning_rate": 0.0004907472986921945,
      "loss": 5.4573,
      "mean_token_accuracy": 0.17231786549091338,
      "num_tokens": 23305932.0,
      "step": 10175
    },
    {
      "entropy": 5.676281976699829,
      "epoch": 1.0063266113088176,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004907372799529992,
      "loss": 5.4342,
      "mean_token_accuracy": 0.1741688385605812,
      "num_tokens": 23317856.0,
      "step": 10180
    },
    {
      "entropy": 5.693381118774414,
      "epoch": 1.006820877817319,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004907272559066334,
      "loss": 5.3396,
      "mean_token_accuracy": 0.18660397678613663,
      "num_tokens": 23328071.0,
      "step": 10185
    },
    {
      "entropy": 5.651421022415161,
      "epoch": 1.0073151443258206,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004907172265533436,
      "loss": 5.5209,
      "mean_token_accuracy": 0.1700335383415222,
      "num_tokens": 23339601.0,
      "step": 10190
    },
    {
      "entropy": 5.740713691711425,
      "epoch": 1.0078094108343219,
      "grad_norm": 1.75,
      "learning_rate": 0.0004907071918933766,
      "loss": 5.4867,
      "mean_token_accuracy": 0.17361918538808824,
      "num_tokens": 23351672.0,
      "step": 10195
    },
    {
      "entropy": 5.7681866645812985,
      "epoch": 1.0083036773428233,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004906971519269793,
      "loss": 5.5063,
      "mean_token_accuracy": 0.17053833603858948,
      "num_tokens": 23361684.0,
      "step": 10200
    },
    {
      "entropy": 5.615524816513061,
      "epoch": 1.0087979438513246,
      "grad_norm": 1.625,
      "learning_rate": 0.0004906871066543984,
      "loss": 5.4756,
      "mean_token_accuracy": 0.17541383057832718,
      "num_tokens": 23373969.0,
      "step": 10205
    },
    {
      "entropy": 5.7493010520935055,
      "epoch": 1.009292210359826,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004906770560758813,
      "loss": 5.5105,
      "mean_token_accuracy": 0.16009564846754074,
      "num_tokens": 23384747.0,
      "step": 10210
    },
    {
      "entropy": 5.768032264709473,
      "epoch": 1.0097864768683273,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004906670001916752,
      "loss": 5.4713,
      "mean_token_accuracy": 0.1735076993703842,
      "num_tokens": 23395900.0,
      "step": 10215
    },
    {
      "entropy": 5.717750883102417,
      "epoch": 1.0102807433768288,
      "grad_norm": 3.453125,
      "learning_rate": 0.0004906569390020274,
      "loss": 5.4877,
      "mean_token_accuracy": 0.17244392186403273,
      "num_tokens": 23408912.0,
      "step": 10220
    },
    {
      "entropy": 5.712572050094605,
      "epoch": 1.01077500988533,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004906468725071855,
      "loss": 5.4131,
      "mean_token_accuracy": 0.17759112566709517,
      "num_tokens": 23419042.0,
      "step": 10225
    },
    {
      "entropy": 5.693017435073853,
      "epoch": 1.0112692763938316,
      "grad_norm": 1.609375,
      "learning_rate": 0.000490636800707397,
      "loss": 5.4527,
      "mean_token_accuracy": 0.17689252644777298,
      "num_tokens": 23430861.0,
      "step": 10230
    },
    {
      "entropy": 5.77862868309021,
      "epoch": 1.0117635429023328,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004906267236029099,
      "loss": 5.5539,
      "mean_token_accuracy": 0.1663425713777542,
      "num_tokens": 23443489.0,
      "step": 10235
    },
    {
      "entropy": 5.709869241714477,
      "epoch": 1.0122578094108343,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004906166411939718,
      "loss": 5.476,
      "mean_token_accuracy": 0.17230155020952226,
      "num_tokens": 23454327.0,
      "step": 10240
    },
    {
      "entropy": 5.6665137767791744,
      "epoch": 1.0127520759193358,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004906065534808309,
      "loss": 5.38,
      "mean_token_accuracy": 0.17807909101247787,
      "num_tokens": 23466684.0,
      "step": 10245
    },
    {
      "entropy": 5.678644275665283,
      "epoch": 1.013246342427837,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004905964604637353,
      "loss": 5.477,
      "mean_token_accuracy": 0.17528004497289656,
      "num_tokens": 23478347.0,
      "step": 10250
    },
    {
      "entropy": 5.622536945343017,
      "epoch": 1.0137406089363385,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004905863621429334,
      "loss": 5.3539,
      "mean_token_accuracy": 0.17825525999069214,
      "num_tokens": 23489588.0,
      "step": 10255
    },
    {
      "entropy": 5.642977619171143,
      "epoch": 1.0142348754448398,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004905762585186735,
      "loss": 5.4355,
      "mean_token_accuracy": 0.17546121925115585,
      "num_tokens": 23501347.0,
      "step": 10260
    },
    {
      "entropy": 5.768431997299194,
      "epoch": 1.0147291419533413,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004905661495912042,
      "loss": 5.5025,
      "mean_token_accuracy": 0.16941647082567216,
      "num_tokens": 23513036.0,
      "step": 10265
    },
    {
      "entropy": 5.641936492919922,
      "epoch": 1.0152234084618426,
      "grad_norm": 3.3125,
      "learning_rate": 0.0004905560353607742,
      "loss": 5.385,
      "mean_token_accuracy": 0.17548509687185287,
      "num_tokens": 23524166.0,
      "step": 10270
    },
    {
      "entropy": 5.633813858032227,
      "epoch": 1.015717674970344,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004905459158276323,
      "loss": 5.4743,
      "mean_token_accuracy": 0.17159906774759293,
      "num_tokens": 23536638.0,
      "step": 10275
    },
    {
      "entropy": 5.7214484214782715,
      "epoch": 1.0162119414788453,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004905357909920275,
      "loss": 5.4902,
      "mean_token_accuracy": 0.17271796464920045,
      "num_tokens": 23548567.0,
      "step": 10280
    },
    {
      "entropy": 5.696207189559937,
      "epoch": 1.0167062079873468,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004905256608542086,
      "loss": 5.4225,
      "mean_token_accuracy": 0.1777705207467079,
      "num_tokens": 23560108.0,
      "step": 10285
    },
    {
      "entropy": 5.632985210418701,
      "epoch": 1.0172004744958483,
      "grad_norm": 3.0625,
      "learning_rate": 0.0004905155254144251,
      "loss": 5.4225,
      "mean_token_accuracy": 0.17967845052480697,
      "num_tokens": 23571542.0,
      "step": 10290
    },
    {
      "entropy": 5.601810550689697,
      "epoch": 1.0176947410043495,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004905053846729263,
      "loss": 5.4001,
      "mean_token_accuracy": 0.17619881480932237,
      "num_tokens": 23582200.0,
      "step": 10295
    },
    {
      "entropy": 5.703922843933105,
      "epoch": 1.018189007512851,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004904952386299616,
      "loss": 5.4895,
      "mean_token_accuracy": 0.16764492392539979,
      "num_tokens": 23594530.0,
      "step": 10300
    },
    {
      "entropy": 5.69450831413269,
      "epoch": 1.0186832740213523,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004904850872857805,
      "loss": 5.3444,
      "mean_token_accuracy": 0.19017274230718612,
      "num_tokens": 23605697.0,
      "step": 10305
    },
    {
      "entropy": 5.652023887634277,
      "epoch": 1.0191775405298538,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000490474930640633,
      "loss": 5.5008,
      "mean_token_accuracy": 0.17088642716407776,
      "num_tokens": 23615730.0,
      "step": 10310
    },
    {
      "entropy": 5.571003818511963,
      "epoch": 1.019671807038355,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004904647686947687,
      "loss": 5.4094,
      "mean_token_accuracy": 0.182180692255497,
      "num_tokens": 23627242.0,
      "step": 10315
    },
    {
      "entropy": 5.624724102020264,
      "epoch": 1.0201660735468565,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004904546014484376,
      "loss": 5.3522,
      "mean_token_accuracy": 0.18565500676631927,
      "num_tokens": 23639450.0,
      "step": 10320
    },
    {
      "entropy": 5.730679082870483,
      "epoch": 1.0206603400553578,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004904444289018898,
      "loss": 5.4622,
      "mean_token_accuracy": 0.17341006845235823,
      "num_tokens": 23651154.0,
      "step": 10325
    },
    {
      "entropy": 5.69901762008667,
      "epoch": 1.0211546065638593,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004904342510553758,
      "loss": 5.5191,
      "mean_token_accuracy": 0.16899913549423218,
      "num_tokens": 23663996.0,
      "step": 10330
    },
    {
      "entropy": 5.6800093173980715,
      "epoch": 1.0216488730723605,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004904240679091458,
      "loss": 5.4458,
      "mean_token_accuracy": 0.172891241312027,
      "num_tokens": 23676270.0,
      "step": 10335
    },
    {
      "entropy": 5.746737003326416,
      "epoch": 1.022143139580862,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004904138794634502,
      "loss": 5.4254,
      "mean_token_accuracy": 0.17328813523054123,
      "num_tokens": 23687483.0,
      "step": 10340
    },
    {
      "entropy": 5.73184905052185,
      "epoch": 1.0226374060893635,
      "grad_norm": 5.375,
      "learning_rate": 0.0004904036857185398,
      "loss": 5.5057,
      "mean_token_accuracy": 0.17040742337703704,
      "num_tokens": 23697956.0,
      "step": 10345
    },
    {
      "entropy": 5.670113945007325,
      "epoch": 1.0231316725978647,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004903934866746653,
      "loss": 5.4599,
      "mean_token_accuracy": 0.17762310802936554,
      "num_tokens": 23709029.0,
      "step": 10350
    },
    {
      "entropy": 5.65038914680481,
      "epoch": 1.0236259391063662,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004903832823320777,
      "loss": 5.3957,
      "mean_token_accuracy": 0.1827886775135994,
      "num_tokens": 23719792.0,
      "step": 10355
    },
    {
      "entropy": 5.703197288513183,
      "epoch": 1.0241202056148675,
      "grad_norm": 2.625,
      "learning_rate": 0.0004903730726910278,
      "loss": 5.4657,
      "mean_token_accuracy": 0.1736227661371231,
      "num_tokens": 23730201.0,
      "step": 10360
    },
    {
      "entropy": 5.679931545257569,
      "epoch": 1.024614472123369,
      "grad_norm": 1.671875,
      "learning_rate": 0.000490362857751767,
      "loss": 5.414,
      "mean_token_accuracy": 0.17664798051118852,
      "num_tokens": 23741743.0,
      "step": 10365
    },
    {
      "entropy": 5.661805295944214,
      "epoch": 1.0251087386318702,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004903526375145465,
      "loss": 5.3697,
      "mean_token_accuracy": 0.17597247511148453,
      "num_tokens": 23753879.0,
      "step": 10370
    },
    {
      "entropy": 5.698982429504395,
      "epoch": 1.0256030051403717,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004903424119796176,
      "loss": 5.5251,
      "mean_token_accuracy": 0.16546280086040496,
      "num_tokens": 23765062.0,
      "step": 10375
    },
    {
      "entropy": 5.687112331390381,
      "epoch": 1.026097271648873,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004903321811472321,
      "loss": 5.4873,
      "mean_token_accuracy": 0.17693398892879486,
      "num_tokens": 23777261.0,
      "step": 10380
    },
    {
      "entropy": 5.698892879486084,
      "epoch": 1.0265915381573745,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004903219450176413,
      "loss": 5.48,
      "mean_token_accuracy": 0.17271430939435958,
      "num_tokens": 23789247.0,
      "step": 10385
    },
    {
      "entropy": 5.683625364303589,
      "epoch": 1.0270858046658757,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004903117035910975,
      "loss": 5.3765,
      "mean_token_accuracy": 0.17468706369400025,
      "num_tokens": 23798981.0,
      "step": 10390
    },
    {
      "entropy": 5.710557699203491,
      "epoch": 1.0275800711743772,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004903014568678523,
      "loss": 5.4939,
      "mean_token_accuracy": 0.17281184941530228,
      "num_tokens": 23810856.0,
      "step": 10395
    },
    {
      "entropy": 5.70170259475708,
      "epoch": 1.0280743376828787,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004902912048481579,
      "loss": 5.4876,
      "mean_token_accuracy": 0.17184120863676072,
      "num_tokens": 23821544.0,
      "step": 10400
    },
    {
      "entropy": 5.665974378585815,
      "epoch": 1.02856860419138,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004902809475322664,
      "loss": 5.423,
      "mean_token_accuracy": 0.1731875240802765,
      "num_tokens": 23831972.0,
      "step": 10405
    },
    {
      "entropy": 5.758858346939087,
      "epoch": 1.0290628706998814,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004902706849204302,
      "loss": 5.4198,
      "mean_token_accuracy": 0.17724981009960175,
      "num_tokens": 23843451.0,
      "step": 10410
    },
    {
      "entropy": 5.674815464019775,
      "epoch": 1.0295571372083827,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004902604170129017,
      "loss": 5.4488,
      "mean_token_accuracy": 0.17553705275058745,
      "num_tokens": 23854060.0,
      "step": 10415
    },
    {
      "entropy": 5.753055143356323,
      "epoch": 1.0300514037168842,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004902501438099337,
      "loss": 5.5603,
      "mean_token_accuracy": 0.1612450122833252,
      "num_tokens": 23865722.0,
      "step": 10420
    },
    {
      "entropy": 5.7256828308105465,
      "epoch": 1.0305456702253855,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004902398653117787,
      "loss": 5.4783,
      "mean_token_accuracy": 0.1738601580262184,
      "num_tokens": 23875888.0,
      "step": 10425
    },
    {
      "entropy": 5.726684522628784,
      "epoch": 1.031039936733887,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004902295815186896,
      "loss": 5.4912,
      "mean_token_accuracy": 0.16859359592199324,
      "num_tokens": 23887816.0,
      "step": 10430
    },
    {
      "entropy": 5.703883075714112,
      "epoch": 1.0315342032423882,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004902192924309193,
      "loss": 5.4453,
      "mean_token_accuracy": 0.1753258228302002,
      "num_tokens": 23900468.0,
      "step": 10435
    },
    {
      "entropy": 5.7410656929016115,
      "epoch": 1.0320284697508897,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004902089980487211,
      "loss": 5.4022,
      "mean_token_accuracy": 0.17207707464694977,
      "num_tokens": 23912086.0,
      "step": 10440
    },
    {
      "entropy": 5.670900201797485,
      "epoch": 1.032522736259391,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004901986983723483,
      "loss": 5.448,
      "mean_token_accuracy": 0.17552540749311446,
      "num_tokens": 23923262.0,
      "step": 10445
    },
    {
      "entropy": 5.600853347778321,
      "epoch": 1.0330170027678924,
      "grad_norm": 3.34375,
      "learning_rate": 0.000490188393402054,
      "loss": 5.4123,
      "mean_token_accuracy": 0.172198124229908,
      "num_tokens": 23934660.0,
      "step": 10450
    },
    {
      "entropy": 5.687058019638061,
      "epoch": 1.033511269276394,
      "grad_norm": 2.40625,
      "learning_rate": 0.0004901780831380919,
      "loss": 5.4066,
      "mean_token_accuracy": 0.18070681244134904,
      "num_tokens": 23945822.0,
      "step": 10455
    },
    {
      "entropy": 5.681888008117676,
      "epoch": 1.0340055357848952,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004901677675807155,
      "loss": 5.4855,
      "mean_token_accuracy": 0.17444808334112166,
      "num_tokens": 23957312.0,
      "step": 10460
    },
    {
      "entropy": 5.696739625930786,
      "epoch": 1.0344998022933967,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004901574467301786,
      "loss": 5.4576,
      "mean_token_accuracy": 0.1726560115814209,
      "num_tokens": 23968477.0,
      "step": 10465
    },
    {
      "entropy": 5.620746803283692,
      "epoch": 1.034994068801898,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0004901471205867353,
      "loss": 5.382,
      "mean_token_accuracy": 0.17293237149715424,
      "num_tokens": 23979041.0,
      "step": 10470
    },
    {
      "entropy": 5.739405393600464,
      "epoch": 1.0354883353103994,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004901367891506393,
      "loss": 5.4879,
      "mean_token_accuracy": 0.1751044735312462,
      "num_tokens": 23990947.0,
      "step": 10475
    },
    {
      "entropy": 5.715644311904907,
      "epoch": 1.0359826018189007,
      "grad_norm": 1.640625,
      "learning_rate": 0.000490126452422145,
      "loss": 5.3961,
      "mean_token_accuracy": 0.18013232350349426,
      "num_tokens": 24001652.0,
      "step": 10480
    },
    {
      "entropy": 5.732950448989868,
      "epoch": 1.0364768683274022,
      "grad_norm": 2.4375,
      "learning_rate": 0.0004901161104015066,
      "loss": 5.4818,
      "mean_token_accuracy": 0.1697138473391533,
      "num_tokens": 24012707.0,
      "step": 10485
    },
    {
      "entropy": 5.5767247676849365,
      "epoch": 1.0369711348359034,
      "grad_norm": 1.9375,
      "learning_rate": 0.0004901057630889784,
      "loss": 5.3276,
      "mean_token_accuracy": 0.1830436870455742,
      "num_tokens": 24024136.0,
      "step": 10490
    },
    {
      "entropy": 5.692571210861206,
      "epoch": 1.037465401344405,
      "grad_norm": 2.296875,
      "learning_rate": 0.0004900954104848152,
      "loss": 5.4736,
      "mean_token_accuracy": 0.17764198333024978,
      "num_tokens": 24035586.0,
      "step": 10495
    },
    {
      "entropy": 5.698720121383667,
      "epoch": 1.0379596678529064,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004900850525892714,
      "loss": 5.4103,
      "mean_token_accuracy": 0.17735551595687865,
      "num_tokens": 24045338.0,
      "step": 10500
    },
    {
      "entropy": 5.730284309387207,
      "epoch": 1.0384539343614076,
      "grad_norm": 1.984375,
      "learning_rate": 0.000490074689402602,
      "loss": 5.5,
      "mean_token_accuracy": 0.16347283273935317,
      "num_tokens": 24056898.0,
      "step": 10505
    },
    {
      "entropy": 5.6666100978851315,
      "epoch": 1.0389482008699091,
      "grad_norm": 1.5,
      "learning_rate": 0.0004900643209250619,
      "loss": 5.3773,
      "mean_token_accuracy": 0.18072302937507628,
      "num_tokens": 24068998.0,
      "step": 10510
    },
    {
      "entropy": 5.636692190170288,
      "epoch": 1.0394424673784104,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0004900539471569061,
      "loss": 5.4409,
      "mean_token_accuracy": 0.1745138868689537,
      "num_tokens": 24080301.0,
      "step": 10515
    },
    {
      "entropy": 5.657697582244873,
      "epoch": 1.0399367338869119,
      "grad_norm": 1.9140625,
      "learning_rate": 0.00049004356809839,
      "loss": 5.4701,
      "mean_token_accuracy": 0.17101382613182067,
      "num_tokens": 24092601.0,
      "step": 10520
    },
    {
      "entropy": 5.801998853683472,
      "epoch": 1.0404310003954131,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004900331837497686,
      "loss": 5.5748,
      "mean_token_accuracy": 0.16291743665933608,
      "num_tokens": 24105621.0,
      "step": 10525
    },
    {
      "entropy": 5.733995389938355,
      "epoch": 1.0409252669039146,
      "grad_norm": 2.46875,
      "learning_rate": 0.0004900227941112976,
      "loss": 5.446,
      "mean_token_accuracy": 0.1833525836467743,
      "num_tokens": 24117926.0,
      "step": 10530
    },
    {
      "entropy": 5.664119672775269,
      "epoch": 1.0414195334124159,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004900123991832323,
      "loss": 5.4049,
      "mean_token_accuracy": 0.17392261922359467,
      "num_tokens": 24129119.0,
      "step": 10535
    },
    {
      "entropy": 5.604949569702148,
      "epoch": 1.0419137999209174,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004900019989658288,
      "loss": 5.4643,
      "mean_token_accuracy": 0.16780135184526443,
      "num_tokens": 24141726.0,
      "step": 10540
    },
    {
      "entropy": 5.795202016830444,
      "epoch": 1.0424080664294189,
      "grad_norm": 2.234375,
      "learning_rate": 0.0004899915934593427,
      "loss": 5.4902,
      "mean_token_accuracy": 0.17176624685525893,
      "num_tokens": 24152912.0,
      "step": 10545
    },
    {
      "entropy": 5.802124071121216,
      "epoch": 1.0429023329379201,
      "grad_norm": 2.59375,
      "learning_rate": 0.0004899811826640302,
      "loss": 5.5042,
      "mean_token_accuracy": 0.16834273487329482,
      "num_tokens": 24163602.0,
      "step": 10550
    },
    {
      "entropy": 5.677811765670777,
      "epoch": 1.0433965994464216,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004899707665801471,
      "loss": 5.499,
      "mean_token_accuracy": 0.17468012869358063,
      "num_tokens": 24175904.0,
      "step": 10555
    },
    {
      "entropy": 5.661863040924072,
      "epoch": 1.0438908659549229,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004899603452079499,
      "loss": 5.4212,
      "mean_token_accuracy": 0.1718466505408287,
      "num_tokens": 24186438.0,
      "step": 10560
    },
    {
      "entropy": 5.694998502731323,
      "epoch": 1.0443851324634243,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004899499185476948,
      "loss": 5.545,
      "mean_token_accuracy": 0.1677091032266617,
      "num_tokens": 24198357.0,
      "step": 10565
    },
    {
      "entropy": 5.8026610851287845,
      "epoch": 1.0448793989719256,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004899394865996383,
      "loss": 5.5847,
      "mean_token_accuracy": 0.1644972965121269,
      "num_tokens": 24211257.0,
      "step": 10570
    },
    {
      "entropy": 5.787684726715088,
      "epoch": 1.045373665480427,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004899290493640371,
      "loss": 5.5521,
      "mean_token_accuracy": 0.16253542751073838,
      "num_tokens": 24223189.0,
      "step": 10575
    },
    {
      "entropy": 5.669928932189942,
      "epoch": 1.0458679319889284,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004899186068411481,
      "loss": 5.4573,
      "mean_token_accuracy": 0.17435319125652313,
      "num_tokens": 24234701.0,
      "step": 10580
    },
    {
      "entropy": 5.6718775749206545,
      "epoch": 1.0463621984974298,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004899081590312278,
      "loss": 5.4448,
      "mean_token_accuracy": 0.17705156952142714,
      "num_tokens": 24245562.0,
      "step": 10585
    },
    {
      "entropy": 5.695548248291016,
      "epoch": 1.046856465005931,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004898977059345336,
      "loss": 5.4456,
      "mean_token_accuracy": 0.16797035187482834,
      "num_tokens": 24256479.0,
      "step": 10590
    },
    {
      "entropy": 5.726232385635376,
      "epoch": 1.0473507315144326,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004898872475513225,
      "loss": 5.5574,
      "mean_token_accuracy": 0.16119212806224822,
      "num_tokens": 24266900.0,
      "step": 10595
    },
    {
      "entropy": 5.801811981201172,
      "epoch": 1.047844998022934,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004898767838818517,
      "loss": 5.5412,
      "mean_token_accuracy": 0.1650843694806099,
      "num_tokens": 24278915.0,
      "step": 10600
    },
    {
      "entropy": 5.725538301467895,
      "epoch": 1.0483392645314353,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004898663149263787,
      "loss": 5.3705,
      "mean_token_accuracy": 0.18136925846338273,
      "num_tokens": 24289584.0,
      "step": 10605
    },
    {
      "entropy": 5.665912246704101,
      "epoch": 1.0488335310399368,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004898558406851611,
      "loss": 5.4032,
      "mean_token_accuracy": 0.17417797446250916,
      "num_tokens": 24300922.0,
      "step": 10610
    },
    {
      "entropy": 5.6453601837158205,
      "epoch": 1.049327797548438,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004898453611584564,
      "loss": 5.4113,
      "mean_token_accuracy": 0.17546495646238328,
      "num_tokens": 24311238.0,
      "step": 10615
    },
    {
      "entropy": 5.6730499267578125,
      "epoch": 1.0498220640569396,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004898348763465226,
      "loss": 5.3961,
      "mean_token_accuracy": 0.17771839201450348,
      "num_tokens": 24322417.0,
      "step": 10620
    },
    {
      "entropy": 5.665038681030273,
      "epoch": 1.0503163305654408,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004898243862496173,
      "loss": 5.3801,
      "mean_token_accuracy": 0.182219797372818,
      "num_tokens": 24332751.0,
      "step": 10625
    },
    {
      "entropy": 5.717585182189941,
      "epoch": 1.0508105970739423,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004898138908679989,
      "loss": 5.5427,
      "mean_token_accuracy": 0.1624419942498207,
      "num_tokens": 24345082.0,
      "step": 10630
    },
    {
      "entropy": 5.710233116149903,
      "epoch": 1.0513048635824436,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004898033902019255,
      "loss": 5.4797,
      "mean_token_accuracy": 0.1731332704424858,
      "num_tokens": 24355863.0,
      "step": 10635
    },
    {
      "entropy": 5.840935707092285,
      "epoch": 1.051799130090945,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004897928842516553,
      "loss": 5.5313,
      "mean_token_accuracy": 0.16327820122241973,
      "num_tokens": 24367763.0,
      "step": 10640
    },
    {
      "entropy": 5.68679575920105,
      "epoch": 1.0522933965994463,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004897823730174468,
      "loss": 5.4751,
      "mean_token_accuracy": 0.16723699867725372,
      "num_tokens": 24378943.0,
      "step": 10645
    },
    {
      "entropy": 5.6792675971984865,
      "epoch": 1.0527876631079478,
      "grad_norm": 2.296875,
      "learning_rate": 0.0004897718564995587,
      "loss": 5.4444,
      "mean_token_accuracy": 0.1703573137521744,
      "num_tokens": 24389508.0,
      "step": 10650
    },
    {
      "entropy": 5.681413888931274,
      "epoch": 1.0532819296164493,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004897613346982495,
      "loss": 5.358,
      "mean_token_accuracy": 0.17905720472335815,
      "num_tokens": 24400710.0,
      "step": 10655
    },
    {
      "entropy": 5.6566481590271,
      "epoch": 1.0537761961249505,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004897508076137783,
      "loss": 5.4602,
      "mean_token_accuracy": 0.17372193336486816,
      "num_tokens": 24412854.0,
      "step": 10660
    },
    {
      "entropy": 5.7195087432861325,
      "epoch": 1.054270462633452,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004897402752464037,
      "loss": 5.4846,
      "mean_token_accuracy": 0.17400692403316498,
      "num_tokens": 24423212.0,
      "step": 10665
    },
    {
      "entropy": 5.694505739212036,
      "epoch": 1.0547647291419533,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004897297375963852,
      "loss": 5.4786,
      "mean_token_accuracy": 0.17213946282863618,
      "num_tokens": 24436091.0,
      "step": 10670
    },
    {
      "entropy": 5.61199369430542,
      "epoch": 1.0552589956504548,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004897191946639819,
      "loss": 5.3749,
      "mean_token_accuracy": 0.17962474077939988,
      "num_tokens": 24448499.0,
      "step": 10675
    },
    {
      "entropy": 5.767547607421875,
      "epoch": 1.055753262158956,
      "grad_norm": 2.296875,
      "learning_rate": 0.0004897086464494529,
      "loss": 5.5765,
      "mean_token_accuracy": 0.16218212693929673,
      "num_tokens": 24460732.0,
      "step": 10680
    },
    {
      "entropy": 5.746901226043701,
      "epoch": 1.0562475286674575,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004896980929530579,
      "loss": 5.4556,
      "mean_token_accuracy": 0.1791859969496727,
      "num_tokens": 24472460.0,
      "step": 10685
    },
    {
      "entropy": 5.6408806324005125,
      "epoch": 1.0567417951759588,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004896875341750567,
      "loss": 5.4363,
      "mean_token_accuracy": 0.17761030346155166,
      "num_tokens": 24483820.0,
      "step": 10690
    },
    {
      "entropy": 5.626603603363037,
      "epoch": 1.0572360616844603,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004896769701157087,
      "loss": 5.3462,
      "mean_token_accuracy": 0.18297673314809798,
      "num_tokens": 24495775.0,
      "step": 10695
    },
    {
      "entropy": 5.670195913314819,
      "epoch": 1.0577303281929615,
      "grad_norm": 1.671875,
      "learning_rate": 0.000489666400775274,
      "loss": 5.3809,
      "mean_token_accuracy": 0.18010785877704621,
      "num_tokens": 24505589.0,
      "step": 10700
    },
    {
      "entropy": 5.646228551864624,
      "epoch": 1.058224594701463,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004896558261540125,
      "loss": 5.4354,
      "mean_token_accuracy": 0.17763150334358216,
      "num_tokens": 24515988.0,
      "step": 10705
    },
    {
      "entropy": 5.576885223388672,
      "epoch": 1.0587188612099645,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004896452462521844,
      "loss": 5.4076,
      "mean_token_accuracy": 0.1800934761762619,
      "num_tokens": 24528273.0,
      "step": 10710
    },
    {
      "entropy": 5.745494604110718,
      "epoch": 1.0592131277184658,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004896346610700498,
      "loss": 5.4449,
      "mean_token_accuracy": 0.1785084530711174,
      "num_tokens": 24539014.0,
      "step": 10715
    },
    {
      "entropy": 5.642323732376099,
      "epoch": 1.0597073942269672,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004896240706078694,
      "loss": 5.4184,
      "mean_token_accuracy": 0.17620733231306077,
      "num_tokens": 24550210.0,
      "step": 10720
    },
    {
      "entropy": 5.64607834815979,
      "epoch": 1.0602016607354685,
      "grad_norm": 2.640625,
      "learning_rate": 0.0004896134748659036,
      "loss": 5.376,
      "mean_token_accuracy": 0.17596860975027084,
      "num_tokens": 24560934.0,
      "step": 10725
    },
    {
      "entropy": 5.6721272468566895,
      "epoch": 1.06069592724397,
      "grad_norm": 2.03125,
      "learning_rate": 0.000489602873844413,
      "loss": 5.4692,
      "mean_token_accuracy": 0.16910083293914796,
      "num_tokens": 24572274.0,
      "step": 10730
    },
    {
      "entropy": 5.738881397247314,
      "epoch": 1.0611901937524713,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004895922675436583,
      "loss": 5.4737,
      "mean_token_accuracy": 0.17420842200517656,
      "num_tokens": 24583950.0,
      "step": 10735
    },
    {
      "entropy": 5.706207895278931,
      "epoch": 1.0616844602609727,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004895816559639006,
      "loss": 5.4099,
      "mean_token_accuracy": 0.1797514706850052,
      "num_tokens": 24597058.0,
      "step": 10740
    },
    {
      "entropy": 5.60955810546875,
      "epoch": 1.062178726769474,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004895710391054008,
      "loss": 5.2857,
      "mean_token_accuracy": 0.18629123270511627,
      "num_tokens": 24608604.0,
      "step": 10745
    },
    {
      "entropy": 5.703134679794312,
      "epoch": 1.0626729932779755,
      "grad_norm": 2.40625,
      "learning_rate": 0.0004895604169684203,
      "loss": 5.5008,
      "mean_token_accuracy": 0.17173027396202087,
      "num_tokens": 24619707.0,
      "step": 10750
    },
    {
      "entropy": 5.729007291793823,
      "epoch": 1.063167259786477,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004895497895532202,
      "loss": 5.4317,
      "mean_token_accuracy": 0.1732850044965744,
      "num_tokens": 24629579.0,
      "step": 10755
    },
    {
      "entropy": 5.62091293334961,
      "epoch": 1.0636615262949782,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004895391568600619,
      "loss": 5.4496,
      "mean_token_accuracy": 0.1655258521437645,
      "num_tokens": 24641358.0,
      "step": 10760
    },
    {
      "entropy": 5.623490047454834,
      "epoch": 1.0641557928034797,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004895285188892071,
      "loss": 5.4152,
      "mean_token_accuracy": 0.17440328747034073,
      "num_tokens": 24652293.0,
      "step": 10765
    },
    {
      "entropy": 5.7239823818206785,
      "epoch": 1.064650059311981,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004895178756409175,
      "loss": 5.5595,
      "mean_token_accuracy": 0.17060342133045198,
      "num_tokens": 24663370.0,
      "step": 10770
    },
    {
      "entropy": 5.7556455612182615,
      "epoch": 1.0651443258204825,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004895072271154548,
      "loss": 5.5119,
      "mean_token_accuracy": 0.17038141191005707,
      "num_tokens": 24675037.0,
      "step": 10775
    },
    {
      "entropy": 5.7060566425323485,
      "epoch": 1.0656385923289837,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004894965733130812,
      "loss": 5.4312,
      "mean_token_accuracy": 0.17229524701833726,
      "num_tokens": 24685680.0,
      "step": 10780
    },
    {
      "entropy": 5.623900699615478,
      "epoch": 1.0661328588374852,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004894859142340585,
      "loss": 5.3981,
      "mean_token_accuracy": 0.1753751665353775,
      "num_tokens": 24697831.0,
      "step": 10785
    },
    {
      "entropy": 5.686001968383789,
      "epoch": 1.0666271253459865,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004894752498786491,
      "loss": 5.4301,
      "mean_token_accuracy": 0.17435342669487,
      "num_tokens": 24710226.0,
      "step": 10790
    },
    {
      "entropy": 5.684980154037476,
      "epoch": 1.067121391854488,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004894645802471152,
      "loss": 5.4577,
      "mean_token_accuracy": 0.1775508314371109,
      "num_tokens": 24721743.0,
      "step": 10795
    },
    {
      "entropy": 5.685684299468994,
      "epoch": 1.0676156583629894,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004894539053397194,
      "loss": 5.4477,
      "mean_token_accuracy": 0.16322674304246904,
      "num_tokens": 24733804.0,
      "step": 10800
    },
    {
      "entropy": 5.712589454650879,
      "epoch": 1.0681099248714907,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004894432251567242,
      "loss": 5.3965,
      "mean_token_accuracy": 0.1844434380531311,
      "num_tokens": 24745285.0,
      "step": 10805
    },
    {
      "entropy": 5.678305912017822,
      "epoch": 1.0686041913799922,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004894325396983925,
      "loss": 5.4555,
      "mean_token_accuracy": 0.1691162258386612,
      "num_tokens": 24756069.0,
      "step": 10810
    },
    {
      "entropy": 5.610199117660523,
      "epoch": 1.0690984578884934,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004894218489649869,
      "loss": 5.3803,
      "mean_token_accuracy": 0.17897149920463562,
      "num_tokens": 24766935.0,
      "step": 10815
    },
    {
      "entropy": 5.7185900688171385,
      "epoch": 1.069592724396995,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004894111529567707,
      "loss": 5.4289,
      "mean_token_accuracy": 0.17824586480855942,
      "num_tokens": 24777642.0,
      "step": 10820
    },
    {
      "entropy": 5.651887607574463,
      "epoch": 1.0700869909054962,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004894004516740069,
      "loss": 5.377,
      "mean_token_accuracy": 0.1777845025062561,
      "num_tokens": 24788110.0,
      "step": 10825
    },
    {
      "entropy": 5.710649204254151,
      "epoch": 1.0705812574139977,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004893897451169586,
      "loss": 5.5152,
      "mean_token_accuracy": 0.16544357389211656,
      "num_tokens": 24799976.0,
      "step": 10830
    },
    {
      "entropy": 5.757799196243286,
      "epoch": 1.071075523922499,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004893790332858893,
      "loss": 5.4607,
      "mean_token_accuracy": 0.17245256751775742,
      "num_tokens": 24811131.0,
      "step": 10835
    },
    {
      "entropy": 5.70740327835083,
      "epoch": 1.0715697904310004,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004893683161810626,
      "loss": 5.4196,
      "mean_token_accuracy": 0.17343536615371705,
      "num_tokens": 24822137.0,
      "step": 10840
    },
    {
      "entropy": 5.5790143489837645,
      "epoch": 1.0720640569395017,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004893575938027421,
      "loss": 5.4147,
      "mean_token_accuracy": 0.17936326563358307,
      "num_tokens": 24832369.0,
      "step": 10845
    },
    {
      "entropy": 5.697907638549805,
      "epoch": 1.0725583234480032,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004893468661511914,
      "loss": 5.474,
      "mean_token_accuracy": 0.1761583521962166,
      "num_tokens": 24843351.0,
      "step": 10850
    },
    {
      "entropy": 5.8073609352111815,
      "epoch": 1.0730525899565047,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004893361332266746,
      "loss": 5.5499,
      "mean_token_accuracy": 0.16175124049186707,
      "num_tokens": 24855019.0,
      "step": 10855
    },
    {
      "entropy": 5.697947549819946,
      "epoch": 1.073546856465006,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004893253950294557,
      "loss": 5.4525,
      "mean_token_accuracy": 0.174090975522995,
      "num_tokens": 24866224.0,
      "step": 10860
    },
    {
      "entropy": 5.64131121635437,
      "epoch": 1.0740411229735074,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004893146515597989,
      "loss": 5.3759,
      "mean_token_accuracy": 0.1829046055674553,
      "num_tokens": 24876995.0,
      "step": 10865
    },
    {
      "entropy": 5.67849440574646,
      "epoch": 1.0745353894820087,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004893039028179683,
      "loss": 5.4737,
      "mean_token_accuracy": 0.17300374507904054,
      "num_tokens": 24887846.0,
      "step": 10870
    },
    {
      "entropy": 5.741180038452148,
      "epoch": 1.0750296559905101,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004892931488042285,
      "loss": 5.5989,
      "mean_token_accuracy": 0.16268961429595946,
      "num_tokens": 24899221.0,
      "step": 10875
    },
    {
      "entropy": 5.758380556106568,
      "epoch": 1.0755239224990114,
      "grad_norm": 1.6484375,
      "learning_rate": 0.000489282389518844,
      "loss": 5.5469,
      "mean_token_accuracy": 0.1690807208418846,
      "num_tokens": 24912103.0,
      "step": 10880
    },
    {
      "entropy": 5.792465305328369,
      "epoch": 1.076018189007513,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004892716249620795,
      "loss": 5.5274,
      "mean_token_accuracy": 0.16700552254915238,
      "num_tokens": 24925465.0,
      "step": 10885
    },
    {
      "entropy": 5.712474822998047,
      "epoch": 1.0765124555160142,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004892608551341997,
      "loss": 5.3236,
      "mean_token_accuracy": 0.17507287859916687,
      "num_tokens": 24935136.0,
      "step": 10890
    },
    {
      "entropy": 5.653912019729614,
      "epoch": 1.0770067220245156,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004892500800354697,
      "loss": 5.4631,
      "mean_token_accuracy": 0.16899198442697524,
      "num_tokens": 24946845.0,
      "step": 10895
    },
    {
      "entropy": 5.616721534729004,
      "epoch": 1.077500988533017,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004892392996661544,
      "loss": 5.3935,
      "mean_token_accuracy": 0.17076674848794937,
      "num_tokens": 24958337.0,
      "step": 10900
    },
    {
      "entropy": 5.721500301361084,
      "epoch": 1.0779952550415184,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004892285140265192,
      "loss": 5.5059,
      "mean_token_accuracy": 0.16908132582902907,
      "num_tokens": 24969505.0,
      "step": 10905
    },
    {
      "entropy": 5.653558921813965,
      "epoch": 1.0784895215500199,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004892177231168292,
      "loss": 5.4578,
      "mean_token_accuracy": 0.17113558501005172,
      "num_tokens": 24981577.0,
      "step": 10910
    },
    {
      "entropy": 5.753542566299439,
      "epoch": 1.0789837880585211,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004892069269373502,
      "loss": 5.509,
      "mean_token_accuracy": 0.17284602224826812,
      "num_tokens": 24992758.0,
      "step": 10915
    },
    {
      "entropy": 5.739348220825195,
      "epoch": 1.0794780545670226,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004891961254883474,
      "loss": 5.4784,
      "mean_token_accuracy": 0.17648012042045594,
      "num_tokens": 25003057.0,
      "step": 10920
    },
    {
      "entropy": 5.653453207015991,
      "epoch": 1.0799723210755239,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004891853187700867,
      "loss": 5.4636,
      "mean_token_accuracy": 0.17451973110437394,
      "num_tokens": 25014900.0,
      "step": 10925
    },
    {
      "entropy": 5.64291934967041,
      "epoch": 1.0804665875840254,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004891745067828339,
      "loss": 5.3834,
      "mean_token_accuracy": 0.18131772428750992,
      "num_tokens": 25026313.0,
      "step": 10930
    },
    {
      "entropy": 5.708517026901245,
      "epoch": 1.0809608540925266,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004891636895268551,
      "loss": 5.4679,
      "mean_token_accuracy": 0.17502452582120895,
      "num_tokens": 25037014.0,
      "step": 10935
    },
    {
      "entropy": 5.752079343795776,
      "epoch": 1.081455120601028,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004891528670024162,
      "loss": 5.5584,
      "mean_token_accuracy": 0.17423778027296066,
      "num_tokens": 25048051.0,
      "step": 10940
    },
    {
      "entropy": 5.632770633697509,
      "epoch": 1.0819493871095294,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004891420392097837,
      "loss": 5.3781,
      "mean_token_accuracy": 0.18239771574735641,
      "num_tokens": 25057825.0,
      "step": 10945
    },
    {
      "entropy": 5.616934061050415,
      "epoch": 1.0824436536180309,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004891312061492236,
      "loss": 5.4173,
      "mean_token_accuracy": 0.17249033600091934,
      "num_tokens": 25067702.0,
      "step": 10950
    },
    {
      "entropy": 5.694773626327515,
      "epoch": 1.0829379201265321,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004891203678210027,
      "loss": 5.403,
      "mean_token_accuracy": 0.1786690905690193,
      "num_tokens": 25078246.0,
      "step": 10955
    },
    {
      "entropy": 5.680955410003662,
      "epoch": 1.0834321866350336,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004891095242253876,
      "loss": 5.4173,
      "mean_token_accuracy": 0.17662651240825653,
      "num_tokens": 25089267.0,
      "step": 10960
    },
    {
      "entropy": 5.7063781261444095,
      "epoch": 1.083926453143535,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004890986753626449,
      "loss": 5.509,
      "mean_token_accuracy": 0.16996352672576903,
      "num_tokens": 25101392.0,
      "step": 10965
    },
    {
      "entropy": 5.698831224441529,
      "epoch": 1.0844207196520363,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004890878212330418,
      "loss": 5.5032,
      "mean_token_accuracy": 0.17343595176935195,
      "num_tokens": 25113309.0,
      "step": 10970
    },
    {
      "entropy": 5.668400239944458,
      "epoch": 1.0849149861605378,
      "grad_norm": 1.5,
      "learning_rate": 0.0004890769618368448,
      "loss": 5.3325,
      "mean_token_accuracy": 0.1874750480055809,
      "num_tokens": 25124803.0,
      "step": 10975
    },
    {
      "entropy": 5.680614185333252,
      "epoch": 1.085409252669039,
      "grad_norm": 5.25,
      "learning_rate": 0.0004890660971743215,
      "loss": 5.4266,
      "mean_token_accuracy": 0.1783684939146042,
      "num_tokens": 25136443.0,
      "step": 10980
    },
    {
      "entropy": 5.707544231414795,
      "epoch": 1.0859035191775406,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004890552272457389,
      "loss": 5.4362,
      "mean_token_accuracy": 0.1714188352227211,
      "num_tokens": 25147128.0,
      "step": 10985
    },
    {
      "entropy": 5.620350742340088,
      "epoch": 1.0863977856860418,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004890443520513645,
      "loss": 5.4871,
      "mean_token_accuracy": 0.17620726972818374,
      "num_tokens": 25159420.0,
      "step": 10990
    },
    {
      "entropy": 5.637927675247193,
      "epoch": 1.0868920521945433,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004890334715914659,
      "loss": 5.4119,
      "mean_token_accuracy": 0.17815214842557908,
      "num_tokens": 25171624.0,
      "step": 10995
    },
    {
      "entropy": 5.660104274749756,
      "epoch": 1.0873863187030446,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004890225858663106,
      "loss": 5.3573,
      "mean_token_accuracy": 0.18545863330364226,
      "num_tokens": 25181610.0,
      "step": 11000
    },
    {
      "entropy": 5.703441953659057,
      "epoch": 1.087880585211546,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004890116948761666,
      "loss": 5.3451,
      "mean_token_accuracy": 0.17712744921445847,
      "num_tokens": 25191852.0,
      "step": 11005
    },
    {
      "entropy": 5.696305465698242,
      "epoch": 1.0883748517200476,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004890007986213018,
      "loss": 5.5346,
      "mean_token_accuracy": 0.166000697016716,
      "num_tokens": 25204240.0,
      "step": 11010
    },
    {
      "entropy": 5.556787633895874,
      "epoch": 1.0888691182285488,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004889898971019838,
      "loss": 5.3813,
      "mean_token_accuracy": 0.1833414316177368,
      "num_tokens": 25215512.0,
      "step": 11015
    },
    {
      "entropy": 5.690842151641846,
      "epoch": 1.0893633847370503,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004889789903184815,
      "loss": 5.4802,
      "mean_token_accuracy": 0.16919929832220076,
      "num_tokens": 25227903.0,
      "step": 11020
    },
    {
      "entropy": 5.730541086196899,
      "epoch": 1.0898576512455516,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004889680782710626,
      "loss": 5.4147,
      "mean_token_accuracy": 0.17766868472099304,
      "num_tokens": 25240204.0,
      "step": 11025
    },
    {
      "entropy": 5.620070505142212,
      "epoch": 1.090351917754053,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004889571609599958,
      "loss": 5.3853,
      "mean_token_accuracy": 0.17708191126585007,
      "num_tokens": 25251978.0,
      "step": 11030
    },
    {
      "entropy": 5.670386457443238,
      "epoch": 1.0908461842625543,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004889462383855498,
      "loss": 5.4195,
      "mean_token_accuracy": 0.1792678251862526,
      "num_tokens": 25262496.0,
      "step": 11035
    },
    {
      "entropy": 5.691803979873657,
      "epoch": 1.0913404507710558,
      "grad_norm": 1.5859375,
      "learning_rate": 0.000488935310547993,
      "loss": 5.4608,
      "mean_token_accuracy": 0.17530543655157088,
      "num_tokens": 25272814.0,
      "step": 11040
    },
    {
      "entropy": 5.724419021606446,
      "epoch": 1.091834717279557,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004889243774475945,
      "loss": 5.3906,
      "mean_token_accuracy": 0.17630032896995546,
      "num_tokens": 25283749.0,
      "step": 11045
    },
    {
      "entropy": 5.640465211868286,
      "epoch": 1.0923289837880585,
      "grad_norm": 1.8125,
      "learning_rate": 0.000488913439084623,
      "loss": 5.4399,
      "mean_token_accuracy": 0.17799849957227706,
      "num_tokens": 25295993.0,
      "step": 11050
    },
    {
      "entropy": 5.632370233535767,
      "epoch": 1.09282325029656,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004889024954593477,
      "loss": 5.474,
      "mean_token_accuracy": 0.17380830645561218,
      "num_tokens": 25309011.0,
      "step": 11055
    },
    {
      "entropy": 5.681922769546508,
      "epoch": 1.0933175168050613,
      "grad_norm": 1.25,
      "learning_rate": 0.000488891546572038,
      "loss": 5.4085,
      "mean_token_accuracy": 0.17508513629436492,
      "num_tokens": 25320901.0,
      "step": 11060
    },
    {
      "entropy": 5.699883890151978,
      "epoch": 1.0938117833135628,
      "grad_norm": 1.9765625,
      "learning_rate": 0.000488880592422963,
      "loss": 5.389,
      "mean_token_accuracy": 0.1825525000691414,
      "num_tokens": 25333042.0,
      "step": 11065
    },
    {
      "entropy": 5.62830171585083,
      "epoch": 1.094306049822064,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004888696330123921,
      "loss": 5.4603,
      "mean_token_accuracy": 0.17420888245105742,
      "num_tokens": 25346413.0,
      "step": 11070
    },
    {
      "entropy": 5.667964601516724,
      "epoch": 1.0948003163305655,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004888586683405952,
      "loss": 5.4759,
      "mean_token_accuracy": 0.1709025666117668,
      "num_tokens": 25358104.0,
      "step": 11075
    },
    {
      "entropy": 5.679701471328736,
      "epoch": 1.0952945828390668,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004888476984078419,
      "loss": 5.4473,
      "mean_token_accuracy": 0.17665988206863403,
      "num_tokens": 25369485.0,
      "step": 11080
    },
    {
      "entropy": 5.7206920146942135,
      "epoch": 1.0957888493475683,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000488836723214402,
      "loss": 5.5008,
      "mean_token_accuracy": 0.17268404215574265,
      "num_tokens": 25380349.0,
      "step": 11085
    },
    {
      "entropy": 5.721868324279785,
      "epoch": 1.0962831158560695,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004888257427605456,
      "loss": 5.5257,
      "mean_token_accuracy": 0.1666341729462147,
      "num_tokens": 25393042.0,
      "step": 11090
    },
    {
      "entropy": 5.761946821212769,
      "epoch": 1.096777382364571,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004888147570465426,
      "loss": 5.524,
      "mean_token_accuracy": 0.17503573894500732,
      "num_tokens": 25404470.0,
      "step": 11095
    },
    {
      "entropy": 5.627283143997192,
      "epoch": 1.0972716488730723,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004888037660726635,
      "loss": 5.3685,
      "mean_token_accuracy": 0.18558280915021896,
      "num_tokens": 25415397.0,
      "step": 11100
    },
    {
      "entropy": 5.679556846618652,
      "epoch": 1.0977659153815738,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004887927698391786,
      "loss": 5.5463,
      "mean_token_accuracy": 0.1637173056602478,
      "num_tokens": 25427409.0,
      "step": 11105
    },
    {
      "entropy": 5.718594217300415,
      "epoch": 1.0982601818900752,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004887817683463583,
      "loss": 5.399,
      "mean_token_accuracy": 0.17618545591831208,
      "num_tokens": 25438063.0,
      "step": 11110
    },
    {
      "entropy": 5.614250230789184,
      "epoch": 1.0987544483985765,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004887707615944734,
      "loss": 5.3631,
      "mean_token_accuracy": 0.1788792833685875,
      "num_tokens": 25449242.0,
      "step": 11115
    },
    {
      "entropy": 5.693329954147339,
      "epoch": 1.099248714907078,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004887597495837946,
      "loss": 5.486,
      "mean_token_accuracy": 0.17618035823106765,
      "num_tokens": 25461434.0,
      "step": 11120
    },
    {
      "entropy": 5.659618759155274,
      "epoch": 1.0997429814155792,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004887487323145929,
      "loss": 5.4355,
      "mean_token_accuracy": 0.17905406504869462,
      "num_tokens": 25472035.0,
      "step": 11125
    },
    {
      "entropy": 5.656661128997802,
      "epoch": 1.1002372479240807,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000488737709787139,
      "loss": 5.4626,
      "mean_token_accuracy": 0.17480503022670746,
      "num_tokens": 25483143.0,
      "step": 11130
    },
    {
      "entropy": 5.6812938213348385,
      "epoch": 1.100731514432582,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004887266820017044,
      "loss": 5.6167,
      "mean_token_accuracy": 0.16335411220788956,
      "num_tokens": 25496873.0,
      "step": 11135
    },
    {
      "entropy": 5.689480924606324,
      "epoch": 1.1012257809410835,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004887156489585603,
      "loss": 5.4049,
      "mean_token_accuracy": 0.17622639536857604,
      "num_tokens": 25508661.0,
      "step": 11140
    },
    {
      "entropy": 5.794542360305786,
      "epoch": 1.1017200474495847,
      "grad_norm": 1.296875,
      "learning_rate": 0.000488704610657978,
      "loss": 5.5578,
      "mean_token_accuracy": 0.16312139481306076,
      "num_tokens": 25520789.0,
      "step": 11145
    },
    {
      "entropy": 5.568453454971314,
      "epoch": 1.1022143139580862,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004886935671002293,
      "loss": 5.4133,
      "mean_token_accuracy": 0.17906275987625123,
      "num_tokens": 25531131.0,
      "step": 11150
    },
    {
      "entropy": 5.756692600250244,
      "epoch": 1.1027085804665875,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004886825182855855,
      "loss": 5.4954,
      "mean_token_accuracy": 0.16403940469026565,
      "num_tokens": 25543803.0,
      "step": 11155
    },
    {
      "entropy": 5.766486978530883,
      "epoch": 1.103202846975089,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004886714642143188,
      "loss": 5.4045,
      "mean_token_accuracy": 0.18338447213172912,
      "num_tokens": 25553588.0,
      "step": 11160
    },
    {
      "entropy": 5.64628849029541,
      "epoch": 1.1036971134835905,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004886604048867007,
      "loss": 5.4601,
      "mean_token_accuracy": 0.17688646167516708,
      "num_tokens": 25566175.0,
      "step": 11165
    },
    {
      "entropy": 5.737870454788208,
      "epoch": 1.1041913799920917,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004886493403030035,
      "loss": 5.5392,
      "mean_token_accuracy": 0.1695457637310028,
      "num_tokens": 25578033.0,
      "step": 11170
    },
    {
      "entropy": 5.698987054824829,
      "epoch": 1.1046856465005932,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004886382704634996,
      "loss": 5.4029,
      "mean_token_accuracy": 0.1759108915925026,
      "num_tokens": 25588759.0,
      "step": 11175
    },
    {
      "entropy": 5.61225323677063,
      "epoch": 1.1051799130090945,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004886271953684609,
      "loss": 5.315,
      "mean_token_accuracy": 0.1810745894908905,
      "num_tokens": 25599200.0,
      "step": 11180
    },
    {
      "entropy": 5.617602348327637,
      "epoch": 1.105674179517596,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004886161150181602,
      "loss": 5.4146,
      "mean_token_accuracy": 0.17927956581115723,
      "num_tokens": 25610279.0,
      "step": 11185
    },
    {
      "entropy": 5.6308276653289795,
      "epoch": 1.1061684460260972,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004886050294128698,
      "loss": 5.3885,
      "mean_token_accuracy": 0.1758725807070732,
      "num_tokens": 25622062.0,
      "step": 11190
    },
    {
      "entropy": 5.702178955078125,
      "epoch": 1.1066627125345987,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004885939385528628,
      "loss": 5.4409,
      "mean_token_accuracy": 0.17086283713579178,
      "num_tokens": 25633451.0,
      "step": 11195
    },
    {
      "entropy": 5.633413314819336,
      "epoch": 1.1071569790431,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004885828424384114,
      "loss": 5.3822,
      "mean_token_accuracy": 0.1800318941473961,
      "num_tokens": 25644537.0,
      "step": 11200
    },
    {
      "entropy": 5.624868011474609,
      "epoch": 1.1076512455516014,
      "grad_norm": 1.625,
      "learning_rate": 0.0004885717410697892,
      "loss": 5.418,
      "mean_token_accuracy": 0.1765780344605446,
      "num_tokens": 25655570.0,
      "step": 11205
    },
    {
      "entropy": 5.767362356185913,
      "epoch": 1.1081455120601027,
      "grad_norm": 1.546875,
      "learning_rate": 0.000488560634447269,
      "loss": 5.5663,
      "mean_token_accuracy": 0.16615985780954362,
      "num_tokens": 25669470.0,
      "step": 11210
    },
    {
      "entropy": 5.728941822052002,
      "epoch": 1.1086397785686042,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000488549522571124,
      "loss": 5.4746,
      "mean_token_accuracy": 0.17191808223724364,
      "num_tokens": 25682347.0,
      "step": 11215
    },
    {
      "entropy": 5.7414826393127445,
      "epoch": 1.1091340450771057,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004885384054416276,
      "loss": 5.541,
      "mean_token_accuracy": 0.16353462636470795,
      "num_tokens": 25694038.0,
      "step": 11220
    },
    {
      "entropy": 5.668048620223999,
      "epoch": 1.109628311585607,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004885272830590534,
      "loss": 5.3421,
      "mean_token_accuracy": 0.17849619090557098,
      "num_tokens": 25705850.0,
      "step": 11225
    },
    {
      "entropy": 5.656874418258667,
      "epoch": 1.1101225780941084,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004885161554236748,
      "loss": 5.4117,
      "mean_token_accuracy": 0.17788560837507247,
      "num_tokens": 25716294.0,
      "step": 11230
    },
    {
      "entropy": 5.583490800857544,
      "epoch": 1.1106168446026097,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004885050225357657,
      "loss": 5.4142,
      "mean_token_accuracy": 0.18146836310625075,
      "num_tokens": 25727367.0,
      "step": 11235
    },
    {
      "entropy": 5.737776374816894,
      "epoch": 1.1111111111111112,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004884938843955999,
      "loss": 5.5347,
      "mean_token_accuracy": 0.16589053124189376,
      "num_tokens": 25738352.0,
      "step": 11240
    },
    {
      "entropy": 5.663060474395752,
      "epoch": 1.1116053776196124,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004884827410034515,
      "loss": 5.3529,
      "mean_token_accuracy": 0.18134012520313264,
      "num_tokens": 25749940.0,
      "step": 11245
    },
    {
      "entropy": 5.630949878692627,
      "epoch": 1.112099644128114,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004884715923595945,
      "loss": 5.4435,
      "mean_token_accuracy": 0.16812354773283006,
      "num_tokens": 25761251.0,
      "step": 11250
    },
    {
      "entropy": 5.664326858520508,
      "epoch": 1.1125939106366152,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004884604384643032,
      "loss": 5.4801,
      "mean_token_accuracy": 0.17426120340824128,
      "num_tokens": 25772382.0,
      "step": 11255
    },
    {
      "entropy": 5.637487506866455,
      "epoch": 1.1130881771451167,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000488449279317852,
      "loss": 5.4089,
      "mean_token_accuracy": 0.17870349436998367,
      "num_tokens": 25784960.0,
      "step": 11260
    },
    {
      "entropy": 5.694792127609253,
      "epoch": 1.113582443653618,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004884381149205154,
      "loss": 5.3988,
      "mean_token_accuracy": 0.17258140295743943,
      "num_tokens": 25796644.0,
      "step": 11265
    },
    {
      "entropy": 5.656617069244385,
      "epoch": 1.1140767101621194,
      "grad_norm": 1.375,
      "learning_rate": 0.0004884269452725681,
      "loss": 5.4953,
      "mean_token_accuracy": 0.17754077017307282,
      "num_tokens": 25807712.0,
      "step": 11270
    },
    {
      "entropy": 5.650223445892334,
      "epoch": 1.1145709766706209,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004884157703742848,
      "loss": 5.3863,
      "mean_token_accuracy": 0.17846150696277618,
      "num_tokens": 25817944.0,
      "step": 11275
    },
    {
      "entropy": 5.60605902671814,
      "epoch": 1.1150652431791221,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004884045902259405,
      "loss": 5.4157,
      "mean_token_accuracy": 0.17873818725347518,
      "num_tokens": 25830801.0,
      "step": 11280
    },
    {
      "entropy": 5.6828076362609865,
      "epoch": 1.1155595096876236,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004883934048278102,
      "loss": 5.4188,
      "mean_token_accuracy": 0.17201802283525466,
      "num_tokens": 25843877.0,
      "step": 11285
    },
    {
      "entropy": 5.667689371109009,
      "epoch": 1.116053776196125,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004883822141801689,
      "loss": 5.3798,
      "mean_token_accuracy": 0.1794407308101654,
      "num_tokens": 25853763.0,
      "step": 11290
    },
    {
      "entropy": 5.689780902862549,
      "epoch": 1.1165480427046264,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004883710182832921,
      "loss": 5.5016,
      "mean_token_accuracy": 0.16742501109838487,
      "num_tokens": 25864759.0,
      "step": 11295
    },
    {
      "entropy": 5.770525407791138,
      "epoch": 1.1170423092131276,
      "grad_norm": 1.75,
      "learning_rate": 0.0004883598171374552,
      "loss": 5.5681,
      "mean_token_accuracy": 0.16486529558897017,
      "num_tokens": 25877325.0,
      "step": 11300
    },
    {
      "entropy": 5.675923871994018,
      "epoch": 1.1175365757216291,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004883486107429337,
      "loss": 5.4036,
      "mean_token_accuracy": 0.17564295828342438,
      "num_tokens": 25889308.0,
      "step": 11305
    },
    {
      "entropy": 5.591322946548462,
      "epoch": 1.1180308422301304,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004883373991000031,
      "loss": 5.4685,
      "mean_token_accuracy": 0.17322899848222734,
      "num_tokens": 25901721.0,
      "step": 11310
    },
    {
      "entropy": 5.6956439971923825,
      "epoch": 1.1185251087386319,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004883261822089395,
      "loss": 5.4546,
      "mean_token_accuracy": 0.17071697115898132,
      "num_tokens": 25912399.0,
      "step": 11315
    },
    {
      "entropy": 5.630310440063477,
      "epoch": 1.1190193752471334,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004883149600700188,
      "loss": 5.3662,
      "mean_token_accuracy": 0.1815431609749794,
      "num_tokens": 25923560.0,
      "step": 11320
    },
    {
      "entropy": 5.673865890502929,
      "epoch": 1.1195136417556346,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004883037326835169,
      "loss": 5.4309,
      "mean_token_accuracy": 0.17131112664937972,
      "num_tokens": 25933408.0,
      "step": 11325
    },
    {
      "entropy": 5.633373880386353,
      "epoch": 1.120007908264136,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004882925000497101,
      "loss": 5.4269,
      "mean_token_accuracy": 0.17251218110322952,
      "num_tokens": 25944492.0,
      "step": 11330
    },
    {
      "entropy": 5.693177509307861,
      "epoch": 1.1205021747726374,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004882812621688747,
      "loss": 5.4292,
      "mean_token_accuracy": 0.18335856199264527,
      "num_tokens": 25956439.0,
      "step": 11335
    },
    {
      "entropy": 5.681436491012573,
      "epoch": 1.1209964412811388,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004882700190412873,
      "loss": 5.4592,
      "mean_token_accuracy": 0.16837025731801986,
      "num_tokens": 25968765.0,
      "step": 11340
    },
    {
      "entropy": 5.636088705062866,
      "epoch": 1.12149070778964,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004882587706672242,
      "loss": 5.4631,
      "mean_token_accuracy": 0.1741855710744858,
      "num_tokens": 25981300.0,
      "step": 11345
    },
    {
      "entropy": 5.665346479415893,
      "epoch": 1.1219849742981416,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004882475170469622,
      "loss": 5.4121,
      "mean_token_accuracy": 0.17627294659614562,
      "num_tokens": 25991655.0,
      "step": 11350
    },
    {
      "entropy": 5.678512907028198,
      "epoch": 1.1224792408066429,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004882362581807783,
      "loss": 5.4,
      "mean_token_accuracy": 0.17776245772838592,
      "num_tokens": 26002989.0,
      "step": 11355
    },
    {
      "entropy": 5.621101999282837,
      "epoch": 1.1229735073151443,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004882249940689494,
      "loss": 5.3993,
      "mean_token_accuracy": 0.1808344915509224,
      "num_tokens": 26013807.0,
      "step": 11360
    },
    {
      "entropy": 5.691890525817871,
      "epoch": 1.1234677738236458,
      "grad_norm": 1.5,
      "learning_rate": 0.00048821372471175245,
      "loss": 5.53,
      "mean_token_accuracy": 0.16830582469701766,
      "num_tokens": 26027099.0,
      "step": 11365
    },
    {
      "entropy": 5.722969627380371,
      "epoch": 1.123962040332147,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048820245010946486,
      "loss": 5.4276,
      "mean_token_accuracy": 0.17034331411123277,
      "num_tokens": 26039046.0,
      "step": 11370
    },
    {
      "entropy": 5.588412952423096,
      "epoch": 1.1244563068406486,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004881911702623639,
      "loss": 5.2913,
      "mean_token_accuracy": 0.18566345870494844,
      "num_tokens": 26049763.0,
      "step": 11375
    },
    {
      "entropy": 5.5760393142700195,
      "epoch": 1.1249505733491498,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004881798851707271,
      "loss": 5.4488,
      "mean_token_accuracy": 0.17723255157470702,
      "num_tokens": 26062227.0,
      "step": 11380
    },
    {
      "entropy": 5.650962877273559,
      "epoch": 1.1254448398576513,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048816859483483214,
      "loss": 5.4271,
      "mean_token_accuracy": 0.1804071545600891,
      "num_tokens": 26074019.0,
      "step": 11385
    },
    {
      "entropy": 5.723393440246582,
      "epoch": 1.1259391063661526,
      "grad_norm": 1.8125,
      "learning_rate": 0.00048815729925495656,
      "loss": 5.5132,
      "mean_token_accuracy": 0.17423951178789138,
      "num_tokens": 26086169.0,
      "step": 11390
    },
    {
      "entropy": 5.620085525512695,
      "epoch": 1.126433372874654,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048814599843137837,
      "loss": 5.428,
      "mean_token_accuracy": 0.16667740494012834,
      "num_tokens": 26097677.0,
      "step": 11395
    },
    {
      "entropy": 5.610358047485351,
      "epoch": 1.1269276393831553,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004881346923643755,
      "loss": 5.339,
      "mean_token_accuracy": 0.18491210490465165,
      "num_tokens": 26108825.0,
      "step": 11400
    },
    {
      "entropy": 5.752606773376465,
      "epoch": 1.1274219058916568,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004881233810542263,
      "loss": 5.4979,
      "mean_token_accuracy": 0.16972808837890624,
      "num_tokens": 26120274.0,
      "step": 11405
    },
    {
      "entropy": 5.609503364562988,
      "epoch": 1.127916172400158,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00048811206450120875,
      "loss": 5.3985,
      "mean_token_accuracy": 0.17722068279981612,
      "num_tokens": 26131964.0,
      "step": 11410
    },
    {
      "entropy": 5.661878633499145,
      "epoch": 1.1284104389086596,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004881007427056013,
      "loss": 5.4566,
      "mean_token_accuracy": 0.1762293979525566,
      "num_tokens": 26143308.0,
      "step": 11415
    },
    {
      "entropy": 5.70899338722229,
      "epoch": 1.128904705417161,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004880894156676826,
      "loss": 5.4904,
      "mean_token_accuracy": 0.16551888436079026,
      "num_tokens": 26153951.0,
      "step": 11420
    },
    {
      "entropy": 5.663500785827637,
      "epoch": 1.1293989719256623,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004880780833877312,
      "loss": 5.3573,
      "mean_token_accuracy": 0.17926861345767975,
      "num_tokens": 26165951.0,
      "step": 11425
    },
    {
      "entropy": 5.631722688674927,
      "epoch": 1.1298932384341638,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004880667458660259,
      "loss": 5.4365,
      "mean_token_accuracy": 0.16626278012990953,
      "num_tokens": 26176751.0,
      "step": 11430
    },
    {
      "entropy": 5.715836238861084,
      "epoch": 1.130387504942665,
      "grad_norm": 1.875,
      "learning_rate": 0.0004880554031028456,
      "loss": 5.5231,
      "mean_token_accuracy": 0.17225386649370195,
      "num_tokens": 26187682.0,
      "step": 11435
    },
    {
      "entropy": 5.781738328933716,
      "epoch": 1.1308817714511665,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00048804405509846933,
      "loss": 5.4891,
      "mean_token_accuracy": 0.1706399753689766,
      "num_tokens": 26198960.0,
      "step": 11440
    },
    {
      "entropy": 5.636456251144409,
      "epoch": 1.1313760379596678,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004880327018531762,
      "loss": 5.4698,
      "mean_token_accuracy": 0.17297246754169465,
      "num_tokens": 26210198.0,
      "step": 11445
    },
    {
      "entropy": 5.605314636230469,
      "epoch": 1.1318703044681693,
      "grad_norm": 1.5,
      "learning_rate": 0.00048802134336724566,
      "loss": 5.36,
      "mean_token_accuracy": 0.17575148046016692,
      "num_tokens": 26220817.0,
      "step": 11450
    },
    {
      "entropy": 5.647216749191284,
      "epoch": 1.1323645709766705,
      "grad_norm": 1.5625,
      "learning_rate": 0.00048800997964095687,
      "loss": 5.3927,
      "mean_token_accuracy": 0.18376834839582443,
      "num_tokens": 26233657.0,
      "step": 11455
    },
    {
      "entropy": 5.636550951004028,
      "epoch": 1.132858837485172,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004879986106745897,
      "loss": 5.434,
      "mean_token_accuracy": 0.17422324866056443,
      "num_tokens": 26245566.0,
      "step": 11460
    },
    {
      "entropy": 5.666111087799072,
      "epoch": 1.1333531039936733,
      "grad_norm": 1.5625,
      "learning_rate": 0.00048798723646842357,
      "loss": 5.3695,
      "mean_token_accuracy": 0.18181964755058289,
      "num_tokens": 26256717.0,
      "step": 11465
    },
    {
      "entropy": 5.6364631175994875,
      "epoch": 1.1338473705021748,
      "grad_norm": 2.328125,
      "learning_rate": 0.0004879758570227384,
      "loss": 5.4569,
      "mean_token_accuracy": 0.17115849405527114,
      "num_tokens": 26267926.0,
      "step": 11470
    },
    {
      "entropy": 5.620614242553711,
      "epoch": 1.1343416370106763,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048796447233781406,
      "loss": 5.4271,
      "mean_token_accuracy": 0.17267002314329147,
      "num_tokens": 26278666.0,
      "step": 11475
    },
    {
      "entropy": 5.667811727523803,
      "epoch": 1.1348359035191775,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004879530824139308,
      "loss": 5.3947,
      "mean_token_accuracy": 0.1804790809750557,
      "num_tokens": 26289961.0,
      "step": 11480
    },
    {
      "entropy": 5.627564764022827,
      "epoch": 1.135330170027679,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048794168725136854,
      "loss": 5.4312,
      "mean_token_accuracy": 0.17107535451650618,
      "num_tokens": 26301245.0,
      "step": 11485
    },
    {
      "entropy": 5.687054109573364,
      "epoch": 1.1358244365361803,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004879302868504078,
      "loss": 5.4727,
      "mean_token_accuracy": 0.17107676416635514,
      "num_tokens": 26312598.0,
      "step": 11490
    },
    {
      "entropy": 5.755224800109863,
      "epoch": 1.1363187030446817,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004879188812113289,
      "loss": 5.4642,
      "mean_token_accuracy": 0.17363066375255584,
      "num_tokens": 26324300.0,
      "step": 11495
    },
    {
      "entropy": 5.564968395233154,
      "epoch": 1.136812969553183,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004879074703344125,
      "loss": 5.438,
      "mean_token_accuracy": 0.17608712017536163,
      "num_tokens": 26335789.0,
      "step": 11500
    },
    {
      "entropy": 5.714873218536377,
      "epoch": 1.1373072360616845,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004878960542199394,
      "loss": 5.4235,
      "mean_token_accuracy": 0.1759517163038254,
      "num_tokens": 26347336.0,
      "step": 11505
    },
    {
      "entropy": 5.72796597480774,
      "epoch": 1.1378015025701858,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048788463286819037,
      "loss": 5.4128,
      "mean_token_accuracy": 0.17386628687381744,
      "num_tokens": 26358176.0,
      "step": 11510
    },
    {
      "entropy": 5.591306924819946,
      "epoch": 1.1382957690786872,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004878732062794462,
      "loss": 5.3957,
      "mean_token_accuracy": 0.17131773829460145,
      "num_tokens": 26368915.0,
      "step": 11515
    },
    {
      "entropy": 5.653985548019409,
      "epoch": 1.1387900355871885,
      "grad_norm": 1.53125,
      "learning_rate": 0.00048786177445398816,
      "loss": 5.4692,
      "mean_token_accuracy": 0.17206621021032334,
      "num_tokens": 26381687.0,
      "step": 11520
    },
    {
      "entropy": 5.693347501754761,
      "epoch": 1.13928430209569,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004878503373920975,
      "loss": 5.5225,
      "mean_token_accuracy": 0.1717299461364746,
      "num_tokens": 26394289.0,
      "step": 11525
    },
    {
      "entropy": 5.627802801132202,
      "epoch": 1.1397785686041915,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004878388950940555,
      "loss": 5.3886,
      "mean_token_accuracy": 0.18488000631332396,
      "num_tokens": 26405743.0,
      "step": 11530
    },
    {
      "entropy": 5.742920541763306,
      "epoch": 1.1402728351126927,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048782744756014365,
      "loss": 5.5207,
      "mean_token_accuracy": 0.16564180701971054,
      "num_tokens": 26417255.0,
      "step": 11535
    },
    {
      "entropy": 5.746685361862182,
      "epoch": 1.1407671016211942,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004878159947906436,
      "loss": 5.5333,
      "mean_token_accuracy": 0.17083633840084075,
      "num_tokens": 26428992.0,
      "step": 11540
    },
    {
      "entropy": 5.619696760177613,
      "epoch": 1.1412613681296955,
      "grad_norm": 1.40625,
      "learning_rate": 0.000487804536785837,
      "loss": 5.4275,
      "mean_token_accuracy": 0.18058303594589234,
      "num_tokens": 26440375.0,
      "step": 11545
    },
    {
      "entropy": 5.680644845962524,
      "epoch": 1.141755634638197,
      "grad_norm": 1.25,
      "learning_rate": 0.00048779307354600586,
      "loss": 5.4876,
      "mean_token_accuracy": 0.16940272301435472,
      "num_tokens": 26452253.0,
      "step": 11550
    },
    {
      "entropy": 5.740002298355103,
      "epoch": 1.1422499011466982,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00048778160507143204,
      "loss": 5.5166,
      "mean_token_accuracy": 0.16702307760715485,
      "num_tokens": 26463766.0,
      "step": 11555
    },
    {
      "entropy": 5.634490919113159,
      "epoch": 1.1427441676551997,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004877701313623978,
      "loss": 5.4744,
      "mean_token_accuracy": 0.1724281281232834,
      "num_tokens": 26475869.0,
      "step": 11560
    },
    {
      "entropy": 5.70414628982544,
      "epoch": 1.1432384341637012,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004877586524191852,
      "loss": 5.4498,
      "mean_token_accuracy": 0.17543199509382248,
      "num_tokens": 26487142.0,
      "step": 11565
    },
    {
      "entropy": 5.661681413650513,
      "epoch": 1.1437327006722025,
      "grad_norm": 1.375,
      "learning_rate": 0.0004877471682420768,
      "loss": 5.2898,
      "mean_token_accuracy": 0.18153191357851028,
      "num_tokens": 26499019.0,
      "step": 11570
    },
    {
      "entropy": 5.596389675140381,
      "epoch": 1.1442269671807037,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004877356788313551,
      "loss": 5.3995,
      "mean_token_accuracy": 0.18449472784996032,
      "num_tokens": 26510060.0,
      "step": 11575
    },
    {
      "entropy": 5.596276760101318,
      "epoch": 1.1447212336892052,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004877241841873027,
      "loss": 5.3307,
      "mean_token_accuracy": 0.18034512549638748,
      "num_tokens": 26520723.0,
      "step": 11580
    },
    {
      "entropy": 5.603102827072144,
      "epoch": 1.1452155001977067,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004877126843102023,
      "loss": 5.4184,
      "mean_token_accuracy": 0.17794054597616196,
      "num_tokens": 26532545.0,
      "step": 11585
    },
    {
      "entropy": 5.695872735977173,
      "epoch": 1.145709766706208,
      "grad_norm": 1.7890625,
      "learning_rate": 0.00048770117920033697,
      "loss": 5.4048,
      "mean_token_accuracy": 0.18506992906332015,
      "num_tokens": 26542993.0,
      "step": 11590
    },
    {
      "entropy": 5.682855701446533,
      "epoch": 1.1462040332147094,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004876896688579895,
      "loss": 5.4174,
      "mean_token_accuracy": 0.1772419184446335,
      "num_tokens": 26554658.0,
      "step": 11595
    },
    {
      "entropy": 5.653793478012085,
      "epoch": 1.1466982997232107,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004876781532834433,
      "loss": 5.4349,
      "mean_token_accuracy": 0.18008334040641785,
      "num_tokens": 26565925.0,
      "step": 11600
    },
    {
      "entropy": 5.606824779510498,
      "epoch": 1.1471925662317122,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004876666324769815,
      "loss": 5.4022,
      "mean_token_accuracy": 0.1800542414188385,
      "num_tokens": 26577045.0,
      "step": 11605
    },
    {
      "entropy": 5.635615015029908,
      "epoch": 1.1476868327402134,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004876551064388875,
      "loss": 5.4863,
      "mean_token_accuracy": 0.1699806660413742,
      "num_tokens": 26587773.0,
      "step": 11610
    },
    {
      "entropy": 5.675181722640991,
      "epoch": 1.148181099248715,
      "grad_norm": 1.65625,
      "learning_rate": 0.00048764357516944485,
      "loss": 5.3888,
      "mean_token_accuracy": 0.17851745188236237,
      "num_tokens": 26597861.0,
      "step": 11615
    },
    {
      "entropy": 5.737946939468384,
      "epoch": 1.1486753657572164,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00048763203866893735,
      "loss": 5.4359,
      "mean_token_accuracy": 0.17790672332048416,
      "num_tokens": 26607695.0,
      "step": 11620
    },
    {
      "entropy": 5.650989007949829,
      "epoch": 1.1491696322657177,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00048762049693764866,
      "loss": 5.4952,
      "mean_token_accuracy": 0.1648597463965416,
      "num_tokens": 26619845.0,
      "step": 11625
    },
    {
      "entropy": 5.743856191635132,
      "epoch": 1.1496638987742192,
      "grad_norm": 2.390625,
      "learning_rate": 0.00048760894997586284,
      "loss": 5.3941,
      "mean_token_accuracy": 0.17170243710279465,
      "num_tokens": 26630279.0,
      "step": 11630
    },
    {
      "entropy": 5.61414909362793,
      "epoch": 1.1501581652827204,
      "grad_norm": 1.75,
      "learning_rate": 0.0004875973977838639,
      "loss": 5.4093,
      "mean_token_accuracy": 0.1825347810983658,
      "num_tokens": 26640996.0,
      "step": 11635
    },
    {
      "entropy": 5.552602672576905,
      "epoch": 1.150652431791222,
      "grad_norm": 1.59375,
      "learning_rate": 0.00048758584036193587,
      "loss": 5.3398,
      "mean_token_accuracy": 0.17941114753484727,
      "num_tokens": 26652416.0,
      "step": 11640
    },
    {
      "entropy": 5.676434087753296,
      "epoch": 1.1511466982997232,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004875742777103632,
      "loss": 5.4468,
      "mean_token_accuracy": 0.17161732017993928,
      "num_tokens": 26663649.0,
      "step": 11645
    },
    {
      "entropy": 5.709451341629029,
      "epoch": 1.1516409648082246,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048756270982943035,
      "loss": 5.5184,
      "mean_token_accuracy": 0.16836084872484208,
      "num_tokens": 26676138.0,
      "step": 11650
    },
    {
      "entropy": 5.7523651123046875,
      "epoch": 1.152135231316726,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004875511367194219,
      "loss": 5.4705,
      "mean_token_accuracy": 0.1685570076107979,
      "num_tokens": 26687364.0,
      "step": 11655
    },
    {
      "entropy": 5.661980247497558,
      "epoch": 1.1526294978252274,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004875395583806225,
      "loss": 5.4327,
      "mean_token_accuracy": 0.1747313290834427,
      "num_tokens": 26698135.0,
      "step": 11660
    },
    {
      "entropy": 5.6518364429473875,
      "epoch": 1.1531237643337287,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004875279748133169,
      "loss": 5.3871,
      "mean_token_accuracy": 0.17886864840984346,
      "num_tokens": 26710495.0,
      "step": 11665
    },
    {
      "entropy": 5.7325492858886715,
      "epoch": 1.1536180308422301,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004875163860177902,
      "loss": 5.5527,
      "mean_token_accuracy": 0.1664167106151581,
      "num_tokens": 26722876.0,
      "step": 11670
    },
    {
      "entropy": 5.665588855743408,
      "epoch": 1.1541122973507316,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00048750479199432753,
      "loss": 5.4165,
      "mean_token_accuracy": 0.17477668076753616,
      "num_tokens": 26733548.0,
      "step": 11675
    },
    {
      "entropy": 5.550989198684692,
      "epoch": 1.1546065638592329,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004874931927432139,
      "loss": 5.3427,
      "mean_token_accuracy": 0.18304872363805771,
      "num_tokens": 26744754.0,
      "step": 11680
    },
    {
      "entropy": 5.619116973876953,
      "epoch": 1.1551008303677344,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004874815882647348,
      "loss": 5.4721,
      "mean_token_accuracy": 0.17246431410312651,
      "num_tokens": 26757416.0,
      "step": 11685
    },
    {
      "entropy": 5.643358421325684,
      "epoch": 1.1555950968762356,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004874699785591756,
      "loss": 5.3776,
      "mean_token_accuracy": 0.17925926446914672,
      "num_tokens": 26768572.0,
      "step": 11690
    },
    {
      "entropy": 5.593602085113526,
      "epoch": 1.1560893633847371,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00048745836362682203,
      "loss": 5.3712,
      "mean_token_accuracy": 0.18403592556715012,
      "num_tokens": 26779309.0,
      "step": 11695
    },
    {
      "entropy": 5.7210639953613285,
      "epoch": 1.1565836298932384,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004874467434679598,
      "loss": 5.4628,
      "mean_token_accuracy": 0.17899265587329866,
      "num_tokens": 26791107.0,
      "step": 11700
    },
    {
      "entropy": 5.63674488067627,
      "epoch": 1.1570778964017399,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004874351180828746,
      "loss": 5.3522,
      "mean_token_accuracy": 0.18383544087409973,
      "num_tokens": 26801334.0,
      "step": 11705
    },
    {
      "entropy": 5.582541036605835,
      "epoch": 1.1575721629102411,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004874234874718526,
      "loss": 5.3867,
      "mean_token_accuracy": 0.17482873648405076,
      "num_tokens": 26812972.0,
      "step": 11710
    },
    {
      "entropy": 5.675421714782715,
      "epoch": 1.1580664294187426,
      "grad_norm": 2.453125,
      "learning_rate": 0.00048741185163517984,
      "loss": 5.4429,
      "mean_token_accuracy": 0.17699123919010162,
      "num_tokens": 26823683.0,
      "step": 11715
    },
    {
      "entropy": 5.6697557926177975,
      "epoch": 1.1585606959272439,
      "grad_norm": 1.5625,
      "learning_rate": 0.00048740021057314267,
      "loss": 5.4169,
      "mean_token_accuracy": 0.18200384974479675,
      "num_tokens": 26835898.0,
      "step": 11720
    },
    {
      "entropy": 5.661006021499634,
      "epoch": 1.1590549624357454,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048738856428602736,
      "loss": 5.3759,
      "mean_token_accuracy": 0.18145013302564622,
      "num_tokens": 26847363.0,
      "step": 11725
    },
    {
      "entropy": 5.6185790538787845,
      "epoch": 1.1595492289442468,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004873769127741203,
      "loss": 5.3423,
      "mean_token_accuracy": 0.17896101325750352,
      "num_tokens": 26859081.0,
      "step": 11730
    },
    {
      "entropy": 5.640149259567261,
      "epoch": 1.160043495452748,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004873652560377083,
      "loss": 5.3649,
      "mean_token_accuracy": 0.1796891063451767,
      "num_tokens": 26869966.0,
      "step": 11735
    },
    {
      "entropy": 5.630725002288818,
      "epoch": 1.1605377619612496,
      "grad_norm": 1.6171875,
      "learning_rate": 0.000487353594077078,
      "loss": 5.3721,
      "mean_token_accuracy": 0.1796877056360245,
      "num_tokens": 26880673.0,
      "step": 11740
    },
    {
      "entropy": 5.648007583618164,
      "epoch": 1.1610320284697508,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00048734192689251643,
      "loss": 5.4095,
      "mean_token_accuracy": 0.1808043971657753,
      "num_tokens": 26892306.0,
      "step": 11745
    },
    {
      "entropy": 5.743666696548462,
      "epoch": 1.1615262949782523,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004873302544843104,
      "loss": 5.4672,
      "mean_token_accuracy": 0.17247305661439896,
      "num_tokens": 26904655.0,
      "step": 11750
    },
    {
      "entropy": 5.609138011932373,
      "epoch": 1.1620205614867536,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004873185768527473,
      "loss": 5.4781,
      "mean_token_accuracy": 0.17411849796772003,
      "num_tokens": 26916580.0,
      "step": 11755
    },
    {
      "entropy": 5.674717092514038,
      "epoch": 1.162514827995255,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048730689399811427,
      "loss": 5.4097,
      "mean_token_accuracy": 0.17701719105243682,
      "num_tokens": 26927033.0,
      "step": 11760
    },
    {
      "entropy": 5.680281448364258,
      "epoch": 1.1630090945037563,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004872952059206986,
      "loss": 5.4916,
      "mean_token_accuracy": 0.17104896754026414,
      "num_tokens": 26938594.0,
      "step": 11765
    },
    {
      "entropy": 5.623374891281128,
      "epoch": 1.1635033610122578,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004872835126207881,
      "loss": 5.3698,
      "mean_token_accuracy": 0.18213477730751038,
      "num_tokens": 26949595.0,
      "step": 11770
    },
    {
      "entropy": 5.711515045166015,
      "epoch": 1.163997627520759,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004872718140986701,
      "loss": 5.5018,
      "mean_token_accuracy": 0.17131920754909516,
      "num_tokens": 26961692.0,
      "step": 11775
    },
    {
      "entropy": 5.654244089126587,
      "epoch": 1.1644918940292606,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00048726011035463265,
      "loss": 5.343,
      "mean_token_accuracy": 0.17828375101089478,
      "num_tokens": 26972062.0,
      "step": 11780
    },
    {
      "entropy": 5.701637077331543,
      "epoch": 1.164986160537762,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004872484013889635,
      "loss": 5.3837,
      "mean_token_accuracy": 0.17885130345821382,
      "num_tokens": 26983353.0,
      "step": 11785
    },
    {
      "entropy": 5.711111164093017,
      "epoch": 1.1654804270462633,
      "grad_norm": 1.5625,
      "learning_rate": 0.00048723668720195067,
      "loss": 5.5137,
      "mean_token_accuracy": 0.16809093058109284,
      "num_tokens": 26994828.0,
      "step": 11790
    },
    {
      "entropy": 5.595471048355103,
      "epoch": 1.1659746935547648,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004872249677938825,
      "loss": 5.4147,
      "mean_token_accuracy": 0.17792768776416779,
      "num_tokens": 27008311.0,
      "step": 11795
    },
    {
      "entropy": 5.592234420776367,
      "epoch": 1.166468960063266,
      "grad_norm": 2.125,
      "learning_rate": 0.0004872132431650472,
      "loss": 5.3809,
      "mean_token_accuracy": 0.18287635445594788,
      "num_tokens": 27018984.0,
      "step": 11800
    },
    {
      "entropy": 5.6987982273101805,
      "epoch": 1.1669632265717675,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00048720151331573317,
      "loss": 5.4973,
      "mean_token_accuracy": 0.1673872858285904,
      "num_tokens": 27029183.0,
      "step": 11805
    },
    {
      "entropy": 5.595653820037842,
      "epoch": 1.1674574930802688,
      "grad_norm": 1.6328125,
      "learning_rate": 0.000487189778246229,
      "loss": 5.344,
      "mean_token_accuracy": 0.18547672182321548,
      "num_tokens": 27039661.0,
      "step": 11810
    },
    {
      "entropy": 5.69262924194336,
      "epoch": 1.1679517595887703,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00048717803795682334,
      "loss": 5.5111,
      "mean_token_accuracy": 0.16858460158109664,
      "num_tokens": 27051685.0,
      "step": 11815
    },
    {
      "entropy": 5.751629972457886,
      "epoch": 1.1684460260972718,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004871662924478051,
      "loss": 5.4691,
      "mean_token_accuracy": 0.1781709909439087,
      "num_tokens": 27062096.0,
      "step": 11820
    },
    {
      "entropy": 5.636955308914184,
      "epoch": 1.168940292605773,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004871545417194631,
      "loss": 5.367,
      "mean_token_accuracy": 0.17909791320562363,
      "num_tokens": 27073059.0,
      "step": 11825
    },
    {
      "entropy": 5.708993577957154,
      "epoch": 1.1694345591142743,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00048714278577208643,
      "loss": 5.4533,
      "mean_token_accuracy": 0.1752929374575615,
      "num_tokens": 27085272.0,
      "step": 11830
    },
    {
      "entropy": 5.700922155380249,
      "epoch": 1.1699288256227758,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004871310246059644,
      "loss": 5.4374,
      "mean_token_accuracy": 0.16876550614833832,
      "num_tokens": 27096651.0,
      "step": 11835
    },
    {
      "entropy": 5.694065380096435,
      "epoch": 1.1704230921312773,
      "grad_norm": 1.515625,
      "learning_rate": 0.00048711925822138627,
      "loss": 5.3697,
      "mean_token_accuracy": 0.17503898441791535,
      "num_tokens": 27108180.0,
      "step": 11840
    },
    {
      "entropy": 5.717705345153808,
      "epoch": 1.1709173586397785,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00048710748661864135,
      "loss": 5.4577,
      "mean_token_accuracy": 0.1774182364344597,
      "num_tokens": 27118902.0,
      "step": 11845
    },
    {
      "entropy": 5.639253377914429,
      "epoch": 1.17141162514828,
      "grad_norm": 1.375,
      "learning_rate": 0.00048709570979801946,
      "loss": 5.4198,
      "mean_token_accuracy": 0.17686117738485335,
      "num_tokens": 27131279.0,
      "step": 11850
    },
    {
      "entropy": 5.621879291534424,
      "epoch": 1.1719058916567813,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00048708392775981014,
      "loss": 5.393,
      "mean_token_accuracy": 0.18451005816459656,
      "num_tokens": 27142983.0,
      "step": 11855
    },
    {
      "entropy": 5.660607719421387,
      "epoch": 1.1724001581652828,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004870721405043033,
      "loss": 5.3776,
      "mean_token_accuracy": 0.1748368352651596,
      "num_tokens": 27154516.0,
      "step": 11860
    },
    {
      "entropy": 5.676554727554321,
      "epoch": 1.172894424673784,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004870603480317889,
      "loss": 5.4369,
      "mean_token_accuracy": 0.17714077830314637,
      "num_tokens": 27164717.0,
      "step": 11865
    },
    {
      "entropy": 5.712461853027344,
      "epoch": 1.1733886911822855,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048704855034255713,
      "loss": 5.4437,
      "mean_token_accuracy": 0.18074197918176652,
      "num_tokens": 27174830.0,
      "step": 11870
    },
    {
      "entropy": 5.637642765045166,
      "epoch": 1.173882957690787,
      "grad_norm": 1.5625,
      "learning_rate": 0.00048703674743689806,
      "loss": 5.4109,
      "mean_token_accuracy": 0.1771542340517044,
      "num_tokens": 27186724.0,
      "step": 11875
    },
    {
      "entropy": 5.574132919311523,
      "epoch": 1.1743772241992882,
      "grad_norm": 1.5,
      "learning_rate": 0.0004870249393151021,
      "loss": 5.3843,
      "mean_token_accuracy": 0.17653077244758605,
      "num_tokens": 27198416.0,
      "step": 11880
    },
    {
      "entropy": 5.641961002349854,
      "epoch": 1.1748714907077897,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00048701312597745986,
      "loss": 5.307,
      "mean_token_accuracy": 0.18012466728687287,
      "num_tokens": 27209347.0,
      "step": 11885
    },
    {
      "entropy": 5.6432661533355715,
      "epoch": 1.175365757216291,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00048700130742426165,
      "loss": 5.3988,
      "mean_token_accuracy": 0.1704787015914917,
      "num_tokens": 27221061.0,
      "step": 11890
    },
    {
      "entropy": 5.629746580123902,
      "epoch": 1.1758600237247925,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004869894836557985,
      "loss": 5.3737,
      "mean_token_accuracy": 0.18714386522769927,
      "num_tokens": 27233436.0,
      "step": 11895
    },
    {
      "entropy": 5.657157850265503,
      "epoch": 1.1763542902332937,
      "grad_norm": 1.546875,
      "learning_rate": 0.00048697765467236117,
      "loss": 5.3814,
      "mean_token_accuracy": 0.18612172752618789,
      "num_tokens": 27246308.0,
      "step": 11900
    },
    {
      "entropy": 5.651254463195801,
      "epoch": 1.1768485567417952,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048696582047424053,
      "loss": 5.4382,
      "mean_token_accuracy": 0.17277199029922485,
      "num_tokens": 27258187.0,
      "step": 11905
    },
    {
      "entropy": 5.628514671325684,
      "epoch": 1.1773428232502965,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00048695398106172793,
      "loss": 5.4002,
      "mean_token_accuracy": 0.18142856806516647,
      "num_tokens": 27269411.0,
      "step": 11910
    },
    {
      "entropy": 5.630224800109863,
      "epoch": 1.177837089758798,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004869421364351145,
      "loss": 5.4196,
      "mean_token_accuracy": 0.1757066622376442,
      "num_tokens": 27280592.0,
      "step": 11915
    },
    {
      "entropy": 5.749567890167237,
      "epoch": 1.1783313562672992,
      "grad_norm": 1.53125,
      "learning_rate": 0.00048693028659469156,
      "loss": 5.5001,
      "mean_token_accuracy": 0.16359154284000396,
      "num_tokens": 27292151.0,
      "step": 11920
    },
    {
      "entropy": 5.660740852355957,
      "epoch": 1.1788256227758007,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00048691843154075075,
      "loss": 5.3616,
      "mean_token_accuracy": 0.17889530807733536,
      "num_tokens": 27305371.0,
      "step": 11925
    },
    {
      "entropy": 5.623413515090943,
      "epoch": 1.1793198892843022,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048690657127358367,
      "loss": 5.4557,
      "mean_token_accuracy": 0.1730481877923012,
      "num_tokens": 27316810.0,
      "step": 11930
    },
    {
      "entropy": 5.645069646835327,
      "epoch": 1.1798141557928035,
      "grad_norm": 1.53125,
      "learning_rate": 0.00048689470579348207,
      "loss": 5.4366,
      "mean_token_accuracy": 0.17546434402465821,
      "num_tokens": 27328412.0,
      "step": 11935
    },
    {
      "entropy": 5.707151222229004,
      "epoch": 1.180308422301305,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00048688283510073775,
      "loss": 5.4833,
      "mean_token_accuracy": 0.17414152175188063,
      "num_tokens": 27340216.0,
      "step": 11940
    },
    {
      "entropy": 5.622820711135864,
      "epoch": 1.1808026888098062,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004868709591956429,
      "loss": 5.3127,
      "mean_token_accuracy": 0.17851535528898238,
      "num_tokens": 27350290.0,
      "step": 11945
    },
    {
      "entropy": 5.634811019897461,
      "epoch": 1.1812969553183077,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00048685907807848955,
      "loss": 5.486,
      "mean_token_accuracy": 0.1703660801053047,
      "num_tokens": 27360639.0,
      "step": 11950
    },
    {
      "entropy": 5.62358283996582,
      "epoch": 1.181791221826809,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048684719174957005,
      "loss": 5.3517,
      "mean_token_accuracy": 0.1807660326361656,
      "num_tokens": 27372059.0,
      "step": 11955
    },
    {
      "entropy": 5.663116455078125,
      "epoch": 1.1822854883353104,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004868353002091767,
      "loss": 5.4695,
      "mean_token_accuracy": 0.16530514657497405,
      "num_tokens": 27384308.0,
      "step": 11960
    },
    {
      "entropy": 5.575397205352783,
      "epoch": 1.1827797548438117,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004868234034576021,
      "loss": 5.2754,
      "mean_token_accuracy": 0.1968613848090172,
      "num_tokens": 27395510.0,
      "step": 11965
    },
    {
      "entropy": 5.610069465637207,
      "epoch": 1.1832740213523132,
      "grad_norm": 1.34375,
      "learning_rate": 0.000486811501495139,
      "loss": 5.404,
      "mean_token_accuracy": 0.17248845845460892,
      "num_tokens": 27406697.0,
      "step": 11970
    },
    {
      "entropy": 5.638059663772583,
      "epoch": 1.1837682878608144,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004867995943220801,
      "loss": 5.4212,
      "mean_token_accuracy": 0.18095448911190032,
      "num_tokens": 27418047.0,
      "step": 11975
    },
    {
      "entropy": 5.706262063980103,
      "epoch": 1.184262554369316,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00048678768193871825,
      "loss": 5.5287,
      "mean_token_accuracy": 0.16767776757478714,
      "num_tokens": 27430044.0,
      "step": 11980
    },
    {
      "entropy": 5.656655359268188,
      "epoch": 1.1847568208778174,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004867757643453466,
      "loss": 5.4489,
      "mean_token_accuracy": 0.17198732495307922,
      "num_tokens": 27441392.0,
      "step": 11985
    },
    {
      "entropy": 5.703206872940063,
      "epoch": 1.1852510873863187,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048676384154225827,
      "loss": 5.5398,
      "mean_token_accuracy": 0.16984956264495848,
      "num_tokens": 27451356.0,
      "step": 11990
    },
    {
      "entropy": 5.602766323089599,
      "epoch": 1.1857453538948202,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004867519135297466,
      "loss": 5.3635,
      "mean_token_accuracy": 0.1859638586640358,
      "num_tokens": 27462358.0,
      "step": 11995
    },
    {
      "entropy": 5.5955119132995605,
      "epoch": 1.1862396204033214,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048673998030810507,
      "loss": 5.3555,
      "mean_token_accuracy": 0.17727247327566148,
      "num_tokens": 27474144.0,
      "step": 12000
    },
    {
      "epoch": 1.1862396204033214,
      "eval_entropy": 5.5140220208068325,
      "eval_loss": 5.510257244110107,
      "eval_mean_token_accuracy": 0.18012088329369733,
      "eval_num_tokens": 27474144.0,
      "eval_runtime": 26.4017,
      "eval_samples_per_second": 1231.475,
      "eval_steps_per_second": 153.967,
      "step": 12000
    },
    {
      "entropy": 5.739811515808105,
      "epoch": 1.186733886911823,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048672804187762704,
      "loss": 5.5097,
      "mean_token_accuracy": 0.16732267439365386,
      "num_tokens": 27486337.0,
      "step": 12005
    },
    {
      "entropy": 5.657247114181518,
      "epoch": 1.1872281534203242,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004867160982386065,
      "loss": 5.4336,
      "mean_token_accuracy": 0.18076617270708084,
      "num_tokens": 27498686.0,
      "step": 12010
    },
    {
      "entropy": 5.7094970703125,
      "epoch": 1.1877224199288257,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00048670414939133694,
      "loss": 5.4611,
      "mean_token_accuracy": 0.17435312420129775,
      "num_tokens": 27509850.0,
      "step": 12015
    },
    {
      "entropy": 5.778923892974854,
      "epoch": 1.188216686437327,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004866921953361125,
      "loss": 5.558,
      "mean_token_accuracy": 0.16622319221496581,
      "num_tokens": 27521994.0,
      "step": 12020
    },
    {
      "entropy": 5.635497379302978,
      "epoch": 1.1887109529458284,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004866802360732273,
      "loss": 5.4934,
      "mean_token_accuracy": 0.17442448139190675,
      "num_tokens": 27534980.0,
      "step": 12025
    },
    {
      "entropy": 5.672856235504151,
      "epoch": 1.1892052194543297,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004866682716029754,
      "loss": 5.3913,
      "mean_token_accuracy": 0.17617794126272202,
      "num_tokens": 27546461.0,
      "step": 12030
    },
    {
      "entropy": 5.665174865722657,
      "epoch": 1.1896994859628311,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048665630192565117,
      "loss": 5.3281,
      "mean_token_accuracy": 0.18198878914117814,
      "num_tokens": 27557076.0,
      "step": 12035
    },
    {
      "entropy": 5.563039493560791,
      "epoch": 1.1901937524713326,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004866443270415491,
      "loss": 5.4388,
      "mean_token_accuracy": 0.18048588037490845,
      "num_tokens": 27569807.0,
      "step": 12040
    },
    {
      "entropy": 5.675030899047852,
      "epoch": 1.190688018979834,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004866323469509637,
      "loss": 5.3448,
      "mean_token_accuracy": 0.18154726922512054,
      "num_tokens": 27579858.0,
      "step": 12045
    },
    {
      "entropy": 5.646805286407471,
      "epoch": 1.1911822854883354,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004866203616541898,
      "loss": 5.4392,
      "mean_token_accuracy": 0.17135237157344818,
      "num_tokens": 27590192.0,
      "step": 12050
    },
    {
      "entropy": 5.616262292861938,
      "epoch": 1.1916765519968366,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048660837115152214,
      "loss": 5.4372,
      "mean_token_accuracy": 0.18135599046945572,
      "num_tokens": 27601756.0,
      "step": 12055
    },
    {
      "entropy": 5.627850341796875,
      "epoch": 1.1921708185053381,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004865963754432557,
      "loss": 5.3573,
      "mean_token_accuracy": 0.18076136261224746,
      "num_tokens": 27614019.0,
      "step": 12060
    },
    {
      "entropy": 5.588666105270386,
      "epoch": 1.1926650850138394,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004865843745296856,
      "loss": 5.4031,
      "mean_token_accuracy": 0.17696809470653535,
      "num_tokens": 27626900.0,
      "step": 12065
    },
    {
      "entropy": 5.618898963928222,
      "epoch": 1.1931593515223409,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048657236841110707,
      "loss": 5.3423,
      "mean_token_accuracy": 0.18124785125255585,
      "num_tokens": 27637897.0,
      "step": 12070
    },
    {
      "entropy": 5.67648663520813,
      "epoch": 1.1936536180308421,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004865603570878155,
      "loss": 5.4485,
      "mean_token_accuracy": 0.1725422888994217,
      "num_tokens": 27649523.0,
      "step": 12075
    },
    {
      "entropy": 5.703610610961914,
      "epoch": 1.1941478845393436,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004865483405601063,
      "loss": 5.4665,
      "mean_token_accuracy": 0.17718684077262878,
      "num_tokens": 27659406.0,
      "step": 12080
    },
    {
      "entropy": 5.589076232910156,
      "epoch": 1.1946421510478449,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004865363188282751,
      "loss": 5.3647,
      "mean_token_accuracy": 0.1768591359257698,
      "num_tokens": 27670912.0,
      "step": 12085
    },
    {
      "entropy": 5.655852746963501,
      "epoch": 1.1951364175563464,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048652429189261767,
      "loss": 5.4101,
      "mean_token_accuracy": 0.18040030747652053,
      "num_tokens": 27681667.0,
      "step": 12090
    },
    {
      "entropy": 5.652654838562012,
      "epoch": 1.1956306840648478,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004865122597534298,
      "loss": 5.4287,
      "mean_token_accuracy": 0.17895926535129547,
      "num_tokens": 27692170.0,
      "step": 12095
    },
    {
      "entropy": 5.5692291259765625,
      "epoch": 1.196124950573349,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004865002224110075,
      "loss": 5.2976,
      "mean_token_accuracy": 0.18794215023517608,
      "num_tokens": 27702533.0,
      "step": 12100
    },
    {
      "entropy": 5.615765953063965,
      "epoch": 1.1966192170818506,
      "grad_norm": 1.5,
      "learning_rate": 0.000486488179865647,
      "loss": 5.3229,
      "mean_token_accuracy": 0.18321851789951324,
      "num_tokens": 27713795.0,
      "step": 12105
    },
    {
      "entropy": 5.655087471008301,
      "epoch": 1.1971134835903519,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004864761321176444,
      "loss": 5.4615,
      "mean_token_accuracy": 0.1752711534500122,
      "num_tokens": 27725404.0,
      "step": 12110
    },
    {
      "entropy": 5.595058250427246,
      "epoch": 1.1976077500988533,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004864640791672962,
      "loss": 5.3531,
      "mean_token_accuracy": 0.17602431178092956,
      "num_tokens": 27737537.0,
      "step": 12115
    },
    {
      "entropy": 5.596295928955078,
      "epoch": 1.1981020166073546,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004864520210148988,
      "loss": 5.3576,
      "mean_token_accuracy": 0.18118408620357512,
      "num_tokens": 27749841.0,
      "step": 12120
    },
    {
      "entropy": 5.6556768894195555,
      "epoch": 1.198596283115856,
      "grad_norm": 1.375,
      "learning_rate": 0.0004864399576607489,
      "loss": 5.471,
      "mean_token_accuracy": 0.17475775331258775,
      "num_tokens": 27760302.0,
      "step": 12125
    },
    {
      "entropy": 5.6397792339324955,
      "epoch": 1.1990905496243576,
      "grad_norm": 2.125,
      "learning_rate": 0.00048642788910514324,
      "loss": 5.4584,
      "mean_token_accuracy": 0.17943050116300582,
      "num_tokens": 27770989.0,
      "step": 12130
    },
    {
      "entropy": 5.714544105529785,
      "epoch": 1.1995848161328588,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00048641581534837863,
      "loss": 5.4042,
      "mean_token_accuracy": 0.17871297597885133,
      "num_tokens": 27785196.0,
      "step": 12135
    },
    {
      "entropy": 5.620921564102173,
      "epoch": 1.2000790826413603,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00048640373639075225,
      "loss": 5.4314,
      "mean_token_accuracy": 0.17638997584581376,
      "num_tokens": 27795588.0,
      "step": 12140
    },
    {
      "entropy": 5.565603399276734,
      "epoch": 1.2005733491498616,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00048639165223256117,
      "loss": 5.3449,
      "mean_token_accuracy": 0.17768989205360414,
      "num_tokens": 27807204.0,
      "step": 12145
    },
    {
      "entropy": 5.617923736572266,
      "epoch": 1.201067615658363,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048637956287410263,
      "loss": 5.3388,
      "mean_token_accuracy": 0.18450913578271866,
      "num_tokens": 27817959.0,
      "step": 12150
    },
    {
      "entropy": 5.667637157440185,
      "epoch": 1.2015618821668643,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048636746831567403,
      "loss": 5.5082,
      "mean_token_accuracy": 0.170717990398407,
      "num_tokens": 27830736.0,
      "step": 12155
    },
    {
      "entropy": 5.660735654830932,
      "epoch": 1.2020561486753658,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004863553685575729,
      "loss": 5.4385,
      "mean_token_accuracy": 0.1692439377307892,
      "num_tokens": 27840992.0,
      "step": 12160
    },
    {
      "entropy": 5.676145792007446,
      "epoch": 1.202550415183867,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004863432636000969,
      "loss": 5.4075,
      "mean_token_accuracy": 0.17512346804141998,
      "num_tokens": 27852747.0,
      "step": 12165
    },
    {
      "entropy": 5.640401172637939,
      "epoch": 1.2030446816923686,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004863311534435439,
      "loss": 5.4989,
      "mean_token_accuracy": 0.17363506257534028,
      "num_tokens": 27865136.0,
      "step": 12170
    },
    {
      "entropy": 5.712698268890381,
      "epoch": 1.2035389482008698,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00048631903808821176,
      "loss": 5.3447,
      "mean_token_accuracy": 0.18180596083402634,
      "num_tokens": 27876489.0,
      "step": 12175
    },
    {
      "entropy": 5.613698101043701,
      "epoch": 1.2040332147093713,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048630691753439844,
      "loss": 5.3516,
      "mean_token_accuracy": 0.1796725422143936,
      "num_tokens": 27886478.0,
      "step": 12180
    },
    {
      "entropy": 5.571279525756836,
      "epoch": 1.2045274812178728,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004862947917824022,
      "loss": 5.374,
      "mean_token_accuracy": 0.1868493989109993,
      "num_tokens": 27897473.0,
      "step": 12185
    },
    {
      "entropy": 5.614389181137085,
      "epoch": 1.205021747726374,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048628266083252124,
      "loss": 5.3351,
      "mean_token_accuracy": 0.17487846314907074,
      "num_tokens": 27908390.0,
      "step": 12190
    },
    {
      "entropy": 5.696160984039307,
      "epoch": 1.2055160142348755,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004862705246850541,
      "loss": 5.4058,
      "mean_token_accuracy": 0.17434001415967942,
      "num_tokens": 27919372.0,
      "step": 12195
    },
    {
      "entropy": 5.5829774856567385,
      "epoch": 1.2060102807433768,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004862583833402993,
      "loss": 5.3797,
      "mean_token_accuracy": 0.17981068789958954,
      "num_tokens": 27931294.0,
      "step": 12200
    },
    {
      "entropy": 5.660366868972778,
      "epoch": 1.2065045472518783,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004862462367985554,
      "loss": 5.4295,
      "mean_token_accuracy": 0.1760703966021538,
      "num_tokens": 27941782.0,
      "step": 12205
    },
    {
      "entropy": 5.620453786849976,
      "epoch": 1.2069988137603795,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004862340850601214,
      "loss": 5.3257,
      "mean_token_accuracy": 0.19191028475761412,
      "num_tokens": 27951703.0,
      "step": 12210
    },
    {
      "entropy": 5.655619049072266,
      "epoch": 1.207493080268881,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004862219281252961,
      "loss": 5.3656,
      "mean_token_accuracy": 0.18273694962263107,
      "num_tokens": 27962049.0,
      "step": 12215
    },
    {
      "entropy": 5.675875425338745,
      "epoch": 1.2079873467773823,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004862097659943786,
      "loss": 5.4074,
      "mean_token_accuracy": 0.17820671647787095,
      "num_tokens": 27973502.0,
      "step": 12220
    },
    {
      "entropy": 5.550475168228149,
      "epoch": 1.2084816132858838,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048619759866766807,
      "loss": 5.2836,
      "mean_token_accuracy": 0.18255395740270614,
      "num_tokens": 27984653.0,
      "step": 12225
    },
    {
      "entropy": 5.638367605209351,
      "epoch": 1.208975879794385,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00048618542614546384,
      "loss": 5.4858,
      "mean_token_accuracy": 0.16729370206594468,
      "num_tokens": 27995152.0,
      "step": 12230
    },
    {
      "entropy": 5.725869369506836,
      "epoch": 1.2094701463028865,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048617324842806545,
      "loss": 5.5598,
      "mean_token_accuracy": 0.16973497122526168,
      "num_tokens": 28006928.0,
      "step": 12235
    },
    {
      "entropy": 5.67096529006958,
      "epoch": 1.209964412811388,
      "grad_norm": 1.453125,
      "learning_rate": 0.00048616106551577226,
      "loss": 5.3859,
      "mean_token_accuracy": 0.17552573084831238,
      "num_tokens": 28018004.0,
      "step": 12240
    },
    {
      "entropy": 5.682720947265625,
      "epoch": 1.2104586793198893,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048614887740888426,
      "loss": 5.4847,
      "mean_token_accuracy": 0.16631695330142976,
      "num_tokens": 28030645.0,
      "step": 12245
    },
    {
      "entropy": 5.6960653305053714,
      "epoch": 1.2109529458283907,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000486136684107701,
      "loss": 5.3983,
      "mean_token_accuracy": 0.18029665350914,
      "num_tokens": 28042591.0,
      "step": 12250
    },
    {
      "entropy": 5.589819478988647,
      "epoch": 1.211447212336892,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004861244856125226,
      "loss": 5.351,
      "mean_token_accuracy": 0.18234280049800872,
      "num_tokens": 28053285.0,
      "step": 12255
    },
    {
      "entropy": 5.518368053436279,
      "epoch": 1.2119414788453935,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004861122819236491,
      "loss": 5.335,
      "mean_token_accuracy": 0.1806777760386467,
      "num_tokens": 28065747.0,
      "step": 12260
    },
    {
      "entropy": 5.627572441101075,
      "epoch": 1.2124357453538948,
      "grad_norm": 1.375,
      "learning_rate": 0.00048610007304138067,
      "loss": 5.3808,
      "mean_token_accuracy": 0.17638372629880905,
      "num_tokens": 28077112.0,
      "step": 12265
    },
    {
      "entropy": 5.630892562866211,
      "epoch": 1.2129300118623962,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004860878589660177,
      "loss": 5.3865,
      "mean_token_accuracy": 0.18621813058853148,
      "num_tokens": 28088835.0,
      "step": 12270
    },
    {
      "entropy": 5.733489227294922,
      "epoch": 1.2134242783708975,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004860756396978607,
      "loss": 5.4748,
      "mean_token_accuracy": 0.16959445178508759,
      "num_tokens": 28100921.0,
      "step": 12275
    },
    {
      "entropy": 5.681570243835449,
      "epoch": 1.213918544879399,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004860634152372102,
      "loss": 5.3867,
      "mean_token_accuracy": 0.17884536385536193,
      "num_tokens": 28112531.0,
      "step": 12280
    },
    {
      "entropy": 5.6841654777526855,
      "epoch": 1.2144128113879002,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004860511855843669,
      "loss": 5.4332,
      "mean_token_accuracy": 0.17698246389627456,
      "num_tokens": 28123821.0,
      "step": 12285
    },
    {
      "entropy": 5.626000261306762,
      "epoch": 1.2149070778964017,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00048603895073963177,
      "loss": 5.3761,
      "mean_token_accuracy": 0.1729662999510765,
      "num_tokens": 28135046.0,
      "step": 12290
    },
    {
      "entropy": 5.6053605556488035,
      "epoch": 1.2154013444049032,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048602671070330564,
      "loss": 5.3832,
      "mean_token_accuracy": 0.18139130920171737,
      "num_tokens": 28145870.0,
      "step": 12295
    },
    {
      "entropy": 5.6928201675415036,
      "epoch": 1.2158956109134045,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00048601446547568973,
      "loss": 5.3982,
      "mean_token_accuracy": 0.17751380056142807,
      "num_tokens": 28156078.0,
      "step": 12300
    },
    {
      "entropy": 5.659221839904785,
      "epoch": 1.216389877421906,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004860022150570852,
      "loss": 5.3714,
      "mean_token_accuracy": 0.1776227831840515,
      "num_tokens": 28166658.0,
      "step": 12305
    },
    {
      "entropy": 5.5944211959838865,
      "epoch": 1.2168841439304072,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048598995944779344,
      "loss": 5.4104,
      "mean_token_accuracy": 0.17678799480199814,
      "num_tokens": 28178890.0,
      "step": 12310
    },
    {
      "entropy": 5.612985229492187,
      "epoch": 1.2173784104389087,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048597769864811604,
      "loss": 5.3727,
      "mean_token_accuracy": 0.18694625943899154,
      "num_tokens": 28191101.0,
      "step": 12315
    },
    {
      "entropy": 5.64115424156189,
      "epoch": 1.21787267694741,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004859654326583544,
      "loss": 5.4669,
      "mean_token_accuracy": 0.17831785529851912,
      "num_tokens": 28202877.0,
      "step": 12320
    },
    {
      "entropy": 5.680452442169189,
      "epoch": 1.2183669434559115,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004859531614788105,
      "loss": 5.4376,
      "mean_token_accuracy": 0.1753949835896492,
      "num_tokens": 28214762.0,
      "step": 12325
    },
    {
      "entropy": 5.682998323440552,
      "epoch": 1.2188612099644127,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000485940885109786,
      "loss": 5.3839,
      "mean_token_accuracy": 0.17759552896022796,
      "num_tokens": 28225830.0,
      "step": 12330
    },
    {
      "entropy": 5.557255125045776,
      "epoch": 1.2193554764729142,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00048592860355158306,
      "loss": 5.3635,
      "mean_token_accuracy": 0.18647830337285995,
      "num_tokens": 28237302.0,
      "step": 12335
    },
    {
      "entropy": 5.612345838546753,
      "epoch": 1.2198497429814155,
      "grad_norm": 1.5,
      "learning_rate": 0.0004859163168045038,
      "loss": 5.3754,
      "mean_token_accuracy": 0.1846173033118248,
      "num_tokens": 28247658.0,
      "step": 12340
    },
    {
      "entropy": 5.613757658004761,
      "epoch": 1.220344009489917,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004859040248688503,
      "loss": 5.4395,
      "mean_token_accuracy": 0.1818190962076187,
      "num_tokens": 28258473.0,
      "step": 12345
    },
    {
      "entropy": 5.741697692871094,
      "epoch": 1.2208382759984184,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048589172774492515,
      "loss": 5.5443,
      "mean_token_accuracy": 0.17012276500463486,
      "num_tokens": 28270167.0,
      "step": 12350
    },
    {
      "entropy": 5.60206937789917,
      "epoch": 1.2213325425069197,
      "grad_norm": 1.375,
      "learning_rate": 0.00048587942543303075,
      "loss": 5.3912,
      "mean_token_accuracy": 0.18306322246789933,
      "num_tokens": 28282047.0,
      "step": 12355
    },
    {
      "entropy": 5.640376424789428,
      "epoch": 1.2218268090154212,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004858671179334698,
      "loss": 5.3562,
      "mean_token_accuracy": 0.18394216746091843,
      "num_tokens": 28292868.0,
      "step": 12360
    },
    {
      "entropy": 5.705264711380005,
      "epoch": 1.2223210755239224,
      "grad_norm": 1.8359375,
      "learning_rate": 0.00048585480524654504,
      "loss": 5.5139,
      "mean_token_accuracy": 0.1703129768371582,
      "num_tokens": 28303816.0,
      "step": 12365
    },
    {
      "entropy": 5.7533531665802,
      "epoch": 1.222815342032424,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004858424873725593,
      "loss": 5.4959,
      "mean_token_accuracy": 0.17069889158010482,
      "num_tokens": 28315191.0,
      "step": 12370
    },
    {
      "entropy": 5.669316530227661,
      "epoch": 1.2233096085409252,
      "grad_norm": 1.375,
      "learning_rate": 0.0004858301643118157,
      "loss": 5.3583,
      "mean_token_accuracy": 0.18108568489551544,
      "num_tokens": 28325312.0,
      "step": 12375
    },
    {
      "entropy": 5.642570304870605,
      "epoch": 1.2238038750494267,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004858178360646174,
      "loss": 5.4362,
      "mean_token_accuracy": 0.1718377485871315,
      "num_tokens": 28336171.0,
      "step": 12380
    },
    {
      "entropy": 5.635008192062378,
      "epoch": 1.2242981415579282,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00048580550263126757,
      "loss": 5.412,
      "mean_token_accuracy": 0.16938796043395996,
      "num_tokens": 28347488.0,
      "step": 12385
    },
    {
      "entropy": 5.6321797370910645,
      "epoch": 1.2247924080664294,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004857931640120697,
      "loss": 5.3261,
      "mean_token_accuracy": 0.18346469849348068,
      "num_tokens": 28358653.0,
      "step": 12390
    },
    {
      "entropy": 5.619165849685669,
      "epoch": 1.2252866745749307,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048578082020732724,
      "loss": 5.3817,
      "mean_token_accuracy": 0.19016854465007782,
      "num_tokens": 28369588.0,
      "step": 12395
    },
    {
      "entropy": 5.605570983886719,
      "epoch": 1.2257809410834322,
      "grad_norm": 1.453125,
      "learning_rate": 0.000485768471217344,
      "loss": 5.3182,
      "mean_token_accuracy": 0.18049757480621337,
      "num_tokens": 28382009.0,
      "step": 12400
    },
    {
      "entropy": 5.700220441818237,
      "epoch": 1.2262752075919336,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004857561170424235,
      "loss": 5.4522,
      "mean_token_accuracy": 0.17025645822286606,
      "num_tokens": 28394376.0,
      "step": 12405
    },
    {
      "entropy": 5.66279354095459,
      "epoch": 1.226769474100435,
      "grad_norm": 1.8671875,
      "learning_rate": 0.00048574375768286996,
      "loss": 5.4606,
      "mean_token_accuracy": 0.1747723937034607,
      "num_tokens": 28406453.0,
      "step": 12410
    },
    {
      "entropy": 5.681567859649658,
      "epoch": 1.2272637406089364,
      "grad_norm": 1.703125,
      "learning_rate": 0.00048573139313898723,
      "loss": 5.4079,
      "mean_token_accuracy": 0.1819054201245308,
      "num_tokens": 28417178.0,
      "step": 12415
    },
    {
      "entropy": 5.6706850051879885,
      "epoch": 1.2277580071174377,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004857190234110795,
      "loss": 5.3876,
      "mean_token_accuracy": 0.1764465883374214,
      "num_tokens": 28428396.0,
      "step": 12420
    },
    {
      "entropy": 5.49635157585144,
      "epoch": 1.2282522736259391,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004857066484994512,
      "loss": 5.3447,
      "mean_token_accuracy": 0.18193579614162445,
      "num_tokens": 28441239.0,
      "step": 12425
    },
    {
      "entropy": 5.7104579448699955,
      "epoch": 1.2287465401344404,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004856942684044067,
      "loss": 5.5477,
      "mean_token_accuracy": 0.18025958240032197,
      "num_tokens": 28453671.0,
      "step": 12430
    },
    {
      "entropy": 5.6799072265625,
      "epoch": 1.2292408066429419,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00048568188312625037,
      "loss": 5.3981,
      "mean_token_accuracy": 0.17662502974271774,
      "num_tokens": 28465922.0,
      "step": 12435
    },
    {
      "entropy": 5.650804281234741,
      "epoch": 1.2297350731514434,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004856694926652871,
      "loss": 5.3942,
      "mean_token_accuracy": 0.17636173516511916,
      "num_tokens": 28477559.0,
      "step": 12440
    },
    {
      "entropy": 5.637345218658448,
      "epoch": 1.2302293396599446,
      "grad_norm": 1.515625,
      "learning_rate": 0.00048565709702182164,
      "loss": 5.4299,
      "mean_token_accuracy": 0.17880400121212006,
      "num_tokens": 28489276.0,
      "step": 12445
    },
    {
      "entropy": 5.6748137950897215,
      "epoch": 1.2307236061684461,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004856446961961589,
      "loss": 5.528,
      "mean_token_accuracy": 0.17531248331069946,
      "num_tokens": 28500821.0,
      "step": 12450
    },
    {
      "entropy": 5.658227396011353,
      "epoch": 1.2312178726769474,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00048563229018860394,
      "loss": 5.4392,
      "mean_token_accuracy": 0.1774931490421295,
      "num_tokens": 28513646.0,
      "step": 12455
    },
    {
      "entropy": 5.683524656295776,
      "epoch": 1.2317121391854489,
      "grad_norm": 1.5625,
      "learning_rate": 0.00048561987899946204,
      "loss": 5.4276,
      "mean_token_accuracy": 0.1752988278865814,
      "num_tokens": 28525508.0,
      "step": 12460
    },
    {
      "entropy": 5.7381178855896,
      "epoch": 1.2322064056939501,
      "grad_norm": 2.03125,
      "learning_rate": 0.00048560746262903836,
      "loss": 5.4302,
      "mean_token_accuracy": 0.176610304415226,
      "num_tokens": 28537863.0,
      "step": 12465
    },
    {
      "entropy": 5.599919939041138,
      "epoch": 1.2327006722024516,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004855950410776385,
      "loss": 5.4294,
      "mean_token_accuracy": 0.17595396041870118,
      "num_tokens": 28550163.0,
      "step": 12470
    },
    {
      "entropy": 5.62755274772644,
      "epoch": 1.2331949387109529,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004855826143455678,
      "loss": 5.3184,
      "mean_token_accuracy": 0.1826096773147583,
      "num_tokens": 28560501.0,
      "step": 12475
    },
    {
      "entropy": 5.739967250823975,
      "epoch": 1.2336892052194544,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004855701824331323,
      "loss": 5.4661,
      "mean_token_accuracy": 0.16783886551856994,
      "num_tokens": 28573057.0,
      "step": 12480
    },
    {
      "entropy": 5.6555641174316404,
      "epoch": 1.2341834717279556,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048555774534063753,
      "loss": 5.4591,
      "mean_token_accuracy": 0.1719682291150093,
      "num_tokens": 28585147.0,
      "step": 12485
    },
    {
      "entropy": 5.5736206531524655,
      "epoch": 1.234677738236457,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004855453030683896,
      "loss": 5.4346,
      "mean_token_accuracy": 0.17766974866390228,
      "num_tokens": 28597774.0,
      "step": 12490
    },
    {
      "entropy": 5.685682868957519,
      "epoch": 1.2351720047449586,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004855328556166945,
      "loss": 5.3544,
      "mean_token_accuracy": 0.18068389743566513,
      "num_tokens": 28610031.0,
      "step": 12495
    },
    {
      "entropy": 5.687689638137817,
      "epoch": 1.2356662712534598,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004855204029858586,
      "loss": 5.4209,
      "mean_token_accuracy": 0.17408981323242187,
      "num_tokens": 28621485.0,
      "step": 12500
    },
    {
      "entropy": 5.578461313247681,
      "epoch": 1.2361605377619613,
      "grad_norm": 1.5,
      "learning_rate": 0.0004855079451761881,
      "loss": 5.3359,
      "mean_token_accuracy": 0.1862749397754669,
      "num_tokens": 28632213.0,
      "step": 12505
    },
    {
      "entropy": 5.695471096038818,
      "epoch": 1.2366548042704626,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004854954821879894,
      "loss": 5.5189,
      "mean_token_accuracy": 0.1741614282131195,
      "num_tokens": 28643994.0,
      "step": 12510
    },
    {
      "entropy": 5.770418787002564,
      "epoch": 1.237149070778964,
      "grad_norm": 1.53125,
      "learning_rate": 0.00048548301402156927,
      "loss": 5.4386,
      "mean_token_accuracy": 0.17961105108261108,
      "num_tokens": 28656556.0,
      "step": 12515
    },
    {
      "entropy": 5.640476369857788,
      "epoch": 1.2376433372874653,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004854705406772343,
      "loss": 5.426,
      "mean_token_accuracy": 0.18069577664136888,
      "num_tokens": 28668022.0,
      "step": 12520
    },
    {
      "entropy": 5.609474754333496,
      "epoch": 1.2381376037959668,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048545806215529136,
      "loss": 5.4312,
      "mean_token_accuracy": 0.1721807077527046,
      "num_tokens": 28679074.0,
      "step": 12525
    },
    {
      "entropy": 5.630985260009766,
      "epoch": 1.238631870304468,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048544557845604753,
      "loss": 5.3557,
      "mean_token_accuracy": 0.18284886479377746,
      "num_tokens": 28691872.0,
      "step": 12530
    },
    {
      "entropy": 5.736825895309448,
      "epoch": 1.2391261368129696,
      "grad_norm": 1.625,
      "learning_rate": 0.00048543308957980973,
      "loss": 5.4639,
      "mean_token_accuracy": 0.17925804555416108,
      "num_tokens": 28703367.0,
      "step": 12535
    },
    {
      "entropy": 5.581083488464356,
      "epoch": 1.2396204033214708,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004854205955268853,
      "loss": 5.2938,
      "mean_token_accuracy": 0.18966979831457137,
      "num_tokens": 28713841.0,
      "step": 12540
    },
    {
      "entropy": 5.617019414901733,
      "epoch": 1.2401146698299723,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00048540809629758167,
      "loss": 5.4704,
      "mean_token_accuracy": 0.1692992240190506,
      "num_tokens": 28724924.0,
      "step": 12545
    },
    {
      "entropy": 5.6759274959564205,
      "epoch": 1.2406089363384738,
      "grad_norm": 1.25,
      "learning_rate": 0.0004853955918922061,
      "loss": 5.4277,
      "mean_token_accuracy": 0.17921543270349502,
      "num_tokens": 28736236.0,
      "step": 12550
    },
    {
      "entropy": 5.70656533241272,
      "epoch": 1.241103202846975,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004853830823110663,
      "loss": 5.466,
      "mean_token_accuracy": 0.1763219028711319,
      "num_tokens": 28748841.0,
      "step": 12555
    },
    {
      "entropy": 5.610775709152222,
      "epoch": 1.2415974693554765,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004853705675544702,
      "loss": 5.3275,
      "mean_token_accuracy": 0.18271404206752778,
      "num_tokens": 28759491.0,
      "step": 12560
    },
    {
      "entropy": 5.524050188064575,
      "epoch": 1.2420917358639778,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004853580476227254,
      "loss": 5.2976,
      "mean_token_accuracy": 0.18501685708761215,
      "num_tokens": 28771668.0,
      "step": 12565
    },
    {
      "entropy": 5.5875575065612795,
      "epoch": 1.2425860023724793,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00048534552251614,
      "loss": 5.4221,
      "mean_token_accuracy": 0.17837006449699402,
      "num_tokens": 28783102.0,
      "step": 12570
    },
    {
      "entropy": 5.6870335102081295,
      "epoch": 1.2430802688809806,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048533299223502217,
      "loss": 5.4271,
      "mean_token_accuracy": 0.17940369248390198,
      "num_tokens": 28794643.0,
      "step": 12575
    },
    {
      "entropy": 5.661948823928833,
      "epoch": 1.243574535389482,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004853204567796801,
      "loss": 5.4331,
      "mean_token_accuracy": 0.17136591374874116,
      "num_tokens": 28806597.0,
      "step": 12580
    },
    {
      "entropy": 5.632091426849366,
      "epoch": 1.2440688018979833,
      "grad_norm": 1.609375,
      "learning_rate": 0.000485307916150422,
      "loss": 5.3856,
      "mean_token_accuracy": 0.17984899282455444,
      "num_tokens": 28817379.0,
      "step": 12585
    },
    {
      "entropy": 5.621669912338257,
      "epoch": 1.2445630684064848,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004852953703475567,
      "loss": 5.3909,
      "mean_token_accuracy": 0.17360709905624389,
      "num_tokens": 28828412.0,
      "step": 12590
    },
    {
      "entropy": 5.587964200973511,
      "epoch": 1.245057334914986,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00048528281937139264,
      "loss": 5.316,
      "mean_token_accuracy": 0.1840084671974182,
      "num_tokens": 28839888.0,
      "step": 12595
    },
    {
      "entropy": 5.704300451278686,
      "epoch": 1.2455516014234875,
      "grad_norm": 1.5,
      "learning_rate": 0.00048527026322223854,
      "loss": 5.4969,
      "mean_token_accuracy": 0.1701827421784401,
      "num_tokens": 28852113.0,
      "step": 12600
    },
    {
      "entropy": 5.692036962509155,
      "epoch": 1.246045867931989,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004852577019004035,
      "loss": 5.4663,
      "mean_token_accuracy": 0.17194680720567704,
      "num_tokens": 28864715.0,
      "step": 12605
    },
    {
      "entropy": 5.597376585006714,
      "epoch": 1.2465401344404903,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004852451354061962,
      "loss": 5.2876,
      "mean_token_accuracy": 0.18463059812784194,
      "num_tokens": 28875166.0,
      "step": 12610
    },
    {
      "entropy": 5.582881498336792,
      "epoch": 1.2470344009489918,
      "grad_norm": 1.6875,
      "learning_rate": 0.00048523256373992604,
      "loss": 5.3724,
      "mean_token_accuracy": 0.18343687802553177,
      "num_tokens": 28885946.0,
      "step": 12615
    },
    {
      "entropy": 5.611480045318603,
      "epoch": 1.247528667457493,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004852199869019021,
      "loss": 5.3716,
      "mean_token_accuracy": 0.17991825491189956,
      "num_tokens": 28896654.0,
      "step": 12620
    },
    {
      "entropy": 5.6644855499267575,
      "epoch": 1.2480229339659945,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00048520740489243396,
      "loss": 5.4131,
      "mean_token_accuracy": 0.17587956190109252,
      "num_tokens": 28908306.0,
      "step": 12625
    },
    {
      "entropy": 5.675091171264649,
      "epoch": 1.2485172004744958,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048519481771183107,
      "loss": 5.4399,
      "mean_token_accuracy": 0.16886961460113525,
      "num_tokens": 28921044.0,
      "step": 12630
    },
    {
      "entropy": 5.631007289886474,
      "epoch": 1.2490114669829973,
      "grad_norm": 1.8203125,
      "learning_rate": 0.000485182225360403,
      "loss": 5.3418,
      "mean_token_accuracy": 0.18043060451745987,
      "num_tokens": 28932290.0,
      "step": 12635
    },
    {
      "entropy": 5.658507061004639,
      "epoch": 1.2495057334914987,
      "grad_norm": 1.5,
      "learning_rate": 0.0004851696278384596,
      "loss": 5.3923,
      "mean_token_accuracy": 0.18084783405065535,
      "num_tokens": 28943482.0,
      "step": 12640
    },
    {
      "entropy": 5.562141704559326,
      "epoch": 1.25,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004851570251463107,
      "loss": 5.4219,
      "mean_token_accuracy": 0.1782122403383255,
      "num_tokens": 28954749.0,
      "step": 12645
    },
    {
      "entropy": 5.720483493804932,
      "epoch": 1.2504942665085013,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048514441728426646,
      "loss": 5.4769,
      "mean_token_accuracy": 0.16636212468147277,
      "num_tokens": 28965304.0,
      "step": 12650
    },
    {
      "entropy": 5.611703300476075,
      "epoch": 1.2509885330170027,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004851318042526369,
      "loss": 5.3448,
      "mean_token_accuracy": 0.1800813630223274,
      "num_tokens": 28976267.0,
      "step": 12655
    },
    {
      "entropy": 5.700345659255982,
      "epoch": 1.2514827995255042,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004851191860517324,
      "loss": 5.5044,
      "mean_token_accuracy": 0.17805597186088562,
      "num_tokens": 28986492.0,
      "step": 12660
    },
    {
      "entropy": 5.684275484085083,
      "epoch": 1.2519770660340055,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00048510656268186325,
      "loss": 5.3714,
      "mean_token_accuracy": 0.18066300004720687,
      "num_tokens": 28996693.0,
      "step": 12665
    },
    {
      "entropy": 5.671869564056396,
      "epoch": 1.252471332542507,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004850939341433401,
      "loss": 5.4352,
      "mean_token_accuracy": 0.1819142982363701,
      "num_tokens": 29008498.0,
      "step": 12670
    },
    {
      "entropy": 5.69298095703125,
      "epoch": 1.2529655990510082,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00048508130043647366,
      "loss": 5.511,
      "mean_token_accuracy": 0.17398120164871217,
      "num_tokens": 29018563.0,
      "step": 12675
    },
    {
      "entropy": 5.645291996002197,
      "epoch": 1.2534598655595097,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004850686615615745,
      "loss": 5.3661,
      "mean_token_accuracy": 0.17985970824956893,
      "num_tokens": 29029204.0,
      "step": 12680
    },
    {
      "entropy": 5.583032321929932,
      "epoch": 1.253954132068011,
      "grad_norm": 1.953125,
      "learning_rate": 0.00048505601751895383,
      "loss": 5.3941,
      "mean_token_accuracy": 0.17486785352230072,
      "num_tokens": 29041112.0,
      "step": 12685
    },
    {
      "entropy": 5.741714668273926,
      "epoch": 1.2544483985765125,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004850433683089224,
      "loss": 5.4221,
      "mean_token_accuracy": 0.17500852644443513,
      "num_tokens": 29052791.0,
      "step": 12690
    },
    {
      "entropy": 5.679446697235107,
      "epoch": 1.254942665085014,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004850307139317916,
      "loss": 5.4684,
      "mean_token_accuracy": 0.17206247001886368,
      "num_tokens": 29064525.0,
      "step": 12695
    },
    {
      "entropy": 5.48385853767395,
      "epoch": 1.2554369315935152,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004850180543878726,
      "loss": 5.3237,
      "mean_token_accuracy": 0.18715211004018784,
      "num_tokens": 29077265.0,
      "step": 12700
    },
    {
      "entropy": 5.634773635864258,
      "epoch": 1.2559311981020165,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004850053896774769,
      "loss": 5.3379,
      "mean_token_accuracy": 0.18502485603094102,
      "num_tokens": 29088022.0,
      "step": 12705
    },
    {
      "entropy": 5.730953121185303,
      "epoch": 1.256425464610518,
      "grad_norm": 2.328125,
      "learning_rate": 0.0004849927198009161,
      "loss": 5.3939,
      "mean_token_accuracy": 0.17817443013191223,
      "num_tokens": 29099901.0,
      "step": 12710
    },
    {
      "entropy": 5.64830379486084,
      "epoch": 1.2569197311190194,
      "grad_norm": 2.140625,
      "learning_rate": 0.00048498004475850174,
      "loss": 5.3882,
      "mean_token_accuracy": 0.17698948979377746,
      "num_tokens": 29111249.0,
      "step": 12715
    },
    {
      "entropy": 5.507478952407837,
      "epoch": 1.2574139976275207,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00048496736455054567,
      "loss": 5.3591,
      "mean_token_accuracy": 0.18044111728668213,
      "num_tokens": 29121261.0,
      "step": 12720
    },
    {
      "entropy": 5.656966972351074,
      "epoch": 1.2579082641360222,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00048495467917736003,
      "loss": 5.5315,
      "mean_token_accuracy": 0.17606231123208999,
      "num_tokens": 29133987.0,
      "step": 12725
    },
    {
      "entropy": 5.772422790527344,
      "epoch": 1.2584025306445235,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0004849419886392565,
      "loss": 5.4321,
      "mean_token_accuracy": 0.18040544241666795,
      "num_tokens": 29146499.0,
      "step": 12730
    },
    {
      "entropy": 5.609433269500732,
      "epoch": 1.258896797153025,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004849292929365476,
      "loss": 5.3103,
      "mean_token_accuracy": 0.18331663608551024,
      "num_tokens": 29157892.0,
      "step": 12735
    },
    {
      "entropy": 5.621004962921143,
      "epoch": 1.2593910636615262,
      "grad_norm": 3.875,
      "learning_rate": 0.0004849165920695455,
      "loss": 5.4279,
      "mean_token_accuracy": 0.17835961282253265,
      "num_tokens": 29169823.0,
      "step": 12740
    },
    {
      "entropy": 5.75237922668457,
      "epoch": 1.2598853301700277,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004849038860385627,
      "loss": 5.462,
      "mean_token_accuracy": 0.1735948607325554,
      "num_tokens": 29181065.0,
      "step": 12745
    },
    {
      "entropy": 5.64034070968628,
      "epoch": 1.2603795966785292,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004848911748439117,
      "loss": 5.3708,
      "mean_token_accuracy": 0.17959531843662263,
      "num_tokens": 29192711.0,
      "step": 12750
    },
    {
      "entropy": 5.552733993530273,
      "epoch": 1.2608738631870304,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004848784584859052,
      "loss": 5.3191,
      "mean_token_accuracy": 0.18066852986812593,
      "num_tokens": 29204688.0,
      "step": 12755
    },
    {
      "entropy": 5.566748905181885,
      "epoch": 1.2613681296955317,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004848657369648561,
      "loss": 5.366,
      "mean_token_accuracy": 0.18262538462877273,
      "num_tokens": 29216326.0,
      "step": 12760
    },
    {
      "entropy": 5.606880521774292,
      "epoch": 1.2618623962040332,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00048485301028107736,
      "loss": 5.3765,
      "mean_token_accuracy": 0.184257772564888,
      "num_tokens": 29226805.0,
      "step": 12765
    },
    {
      "entropy": 5.584829711914063,
      "epoch": 1.2623566627125347,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000484840278434882,
      "loss": 5.365,
      "mean_token_accuracy": 0.1775599479675293,
      "num_tokens": 29239334.0,
      "step": 12770
    },
    {
      "entropy": 5.6800188541412355,
      "epoch": 1.262850929221036,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004848275414265832,
      "loss": 5.3789,
      "mean_token_accuracy": 0.1781705290079117,
      "num_tokens": 29250443.0,
      "step": 12775
    },
    {
      "entropy": 5.594029474258423,
      "epoch": 1.2633451957295374,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048481479925649435,
      "loss": 5.3356,
      "mean_token_accuracy": 0.17795343548059464,
      "num_tokens": 29261209.0,
      "step": 12780
    },
    {
      "entropy": 5.6312315464019775,
      "epoch": 1.2638394622380387,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00048480205192492887,
      "loss": 5.4764,
      "mean_token_accuracy": 0.16658923476934434,
      "num_tokens": 29273761.0,
      "step": 12785
    },
    {
      "entropy": 5.719128370285034,
      "epoch": 1.2643337287465402,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004847892994322004,
      "loss": 5.4512,
      "mean_token_accuracy": 0.17355400174856186,
      "num_tokens": 29285157.0,
      "step": 12790
    },
    {
      "entropy": 5.709945917129517,
      "epoch": 1.2648279952550414,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004847765417786226,
      "loss": 5.4174,
      "mean_token_accuracy": 0.177736359834671,
      "num_tokens": 29295903.0,
      "step": 12795
    },
    {
      "entropy": 5.705383539199829,
      "epoch": 1.265322261763543,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004847637789645093,
      "loss": 5.45,
      "mean_token_accuracy": 0.17303601056337356,
      "num_tokens": 29307865.0,
      "step": 12800
    },
    {
      "entropy": 5.752659273147583,
      "epoch": 1.2658165282720444,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004847510109901745,
      "loss": 5.5215,
      "mean_token_accuracy": 0.16498949825763704,
      "num_tokens": 29319614.0,
      "step": 12805
    },
    {
      "entropy": 5.637752914428711,
      "epoch": 1.2663107947805456,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004847382378559323,
      "loss": 5.4065,
      "mean_token_accuracy": 0.1753526359796524,
      "num_tokens": 29330306.0,
      "step": 12810
    },
    {
      "entropy": 5.641529321670532,
      "epoch": 1.2668050612890471,
      "grad_norm": 1.8515625,
      "learning_rate": 0.00048472545956209696,
      "loss": 5.3708,
      "mean_token_accuracy": 0.18000748455524446,
      "num_tokens": 29341611.0,
      "step": 12815
    },
    {
      "entropy": 5.605336093902588,
      "epoch": 1.2672993277975484,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004847126761089827,
      "loss": 5.3085,
      "mean_token_accuracy": 0.18791087716817856,
      "num_tokens": 29353481.0,
      "step": 12820
    },
    {
      "entropy": 5.60109224319458,
      "epoch": 1.2677935943060499,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004846998874969041,
      "loss": 5.2939,
      "mean_token_accuracy": 0.18456100970506667,
      "num_tokens": 29365388.0,
      "step": 12825
    },
    {
      "entropy": 5.681017684936523,
      "epoch": 1.2682878608145511,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004846870937261758,
      "loss": 5.4539,
      "mean_token_accuracy": 0.16939304769039154,
      "num_tokens": 29377805.0,
      "step": 12830
    },
    {
      "entropy": 5.595469951629639,
      "epoch": 1.2687821273230526,
      "grad_norm": 1.546875,
      "learning_rate": 0.00048467429479711233,
      "loss": 5.3411,
      "mean_token_accuracy": 0.18266978710889817,
      "num_tokens": 29389736.0,
      "step": 12835
    },
    {
      "entropy": 5.592105388641357,
      "epoch": 1.269276393831554,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00048466149071002876,
      "loss": 5.3381,
      "mean_token_accuracy": 0.1925693854689598,
      "num_tokens": 29401256.0,
      "step": 12840
    },
    {
      "entropy": 5.6697304248809814,
      "epoch": 1.2697706603400554,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048464868146524,
      "loss": 5.3749,
      "mean_token_accuracy": 0.17923028022050858,
      "num_tokens": 29413287.0,
      "step": 12845
    },
    {
      "entropy": 5.653916311264038,
      "epoch": 1.2702649268485566,
      "grad_norm": 1.7890625,
      "learning_rate": 0.00048463586706306113,
      "loss": 5.4282,
      "mean_token_accuracy": 0.17880818396806716,
      "num_tokens": 29424151.0,
      "step": 12850
    },
    {
      "entropy": 5.658403825759888,
      "epoch": 1.2707591933570581,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004846230475038074,
      "loss": 5.3897,
      "mean_token_accuracy": 0.17668010145425797,
      "num_tokens": 29435006.0,
      "step": 12855
    },
    {
      "entropy": 5.588736152648925,
      "epoch": 1.2712534598655596,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004846102227877942,
      "loss": 5.39,
      "mean_token_accuracy": 0.18401414453983306,
      "num_tokens": 29446453.0,
      "step": 12860
    },
    {
      "entropy": 5.690087890625,
      "epoch": 1.2717477263740609,
      "grad_norm": 1.734375,
      "learning_rate": 0.00048459739291533706,
      "loss": 5.4515,
      "mean_token_accuracy": 0.17386644184589387,
      "num_tokens": 29458468.0,
      "step": 12865
    },
    {
      "entropy": 5.7500592231750485,
      "epoch": 1.2722419928825623,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004845845578867514,
      "loss": 5.4968,
      "mean_token_accuracy": 0.1727755606174469,
      "num_tokens": 29471167.0,
      "step": 12870
    },
    {
      "entropy": 5.658725786209106,
      "epoch": 1.2727362593910636,
      "grad_norm": 1.640625,
      "learning_rate": 0.00048457171770235327,
      "loss": 5.4108,
      "mean_token_accuracy": 0.1813260719180107,
      "num_tokens": 29482668.0,
      "step": 12875
    },
    {
      "entropy": 5.689535808563233,
      "epoch": 1.273230525899565,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004845588723624582,
      "loss": 5.4646,
      "mean_token_accuracy": 0.17666347622871398,
      "num_tokens": 29495424.0,
      "step": 12880
    },
    {
      "entropy": 5.6919396877288815,
      "epoch": 1.2737247924080664,
      "grad_norm": 2.546875,
      "learning_rate": 0.0004845460218673825,
      "loss": 5.461,
      "mean_token_accuracy": 0.17963027060031891,
      "num_tokens": 29508388.0,
      "step": 12885
    },
    {
      "entropy": 5.630347633361817,
      "epoch": 1.2742190589165678,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004845331662174421,
      "loss": 5.3609,
      "mean_token_accuracy": 0.18002698421478272,
      "num_tokens": 29519884.0,
      "step": 12890
    },
    {
      "entropy": 5.687825918197632,
      "epoch": 1.2747133254250693,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004845203054129533,
      "loss": 5.5531,
      "mean_token_accuracy": 0.16874832510948182,
      "num_tokens": 29532398.0,
      "step": 12895
    },
    {
      "entropy": 5.671998405456543,
      "epoch": 1.2752075919335706,
      "grad_norm": 1.7890625,
      "learning_rate": 0.00048450743945423246,
      "loss": 5.4333,
      "mean_token_accuracy": 0.18099327236413956,
      "num_tokens": 29544394.0,
      "step": 12900
    },
    {
      "entropy": 5.70644359588623,
      "epoch": 1.2757018584420718,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004844945683415961,
      "loss": 5.4868,
      "mean_token_accuracy": 0.17671926468610763,
      "num_tokens": 29555027.0,
      "step": 12905
    },
    {
      "entropy": 5.665847635269165,
      "epoch": 1.2761961249505733,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000484481692075361,
      "loss": 5.4197,
      "mean_token_accuracy": 0.17745458632707595,
      "num_tokens": 29567386.0,
      "step": 12910
    },
    {
      "entropy": 5.733074712753296,
      "epoch": 1.2766903914590748,
      "grad_norm": 1.515625,
      "learning_rate": 0.00048446881065584367,
      "loss": 5.5363,
      "mean_token_accuracy": 0.1666639506816864,
      "num_tokens": 29577832.0,
      "step": 12915
    },
    {
      "entropy": 5.601913690567017,
      "epoch": 1.277184657967576,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00048445592408336106,
      "loss": 5.3305,
      "mean_token_accuracy": 0.1860379919409752,
      "num_tokens": 29590163.0,
      "step": 12920
    },
    {
      "entropy": 5.672854566574097,
      "epoch": 1.2776789244760776,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048444303235823023,
      "loss": 5.3663,
      "mean_token_accuracy": 0.1835150256752968,
      "num_tokens": 29601531.0,
      "step": 12925
    },
    {
      "entropy": 5.674741649627686,
      "epoch": 1.2781731909845788,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004844301354807683,
      "loss": 5.4555,
      "mean_token_accuracy": 0.17768363654613495,
      "num_tokens": 29612417.0,
      "step": 12930
    },
    {
      "entropy": 5.703456687927246,
      "epoch": 1.2786674574930803,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048441723345129265,
      "loss": 5.4721,
      "mean_token_accuracy": 0.17694191187620162,
      "num_tokens": 29622087.0,
      "step": 12935
    },
    {
      "entropy": 5.780738687515258,
      "epoch": 1.2791617240015816,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004844043262701205,
      "loss": 5.5177,
      "mean_token_accuracy": 0.17039716243743896,
      "num_tokens": 29633564.0,
      "step": 12940
    },
    {
      "entropy": 5.580569744110107,
      "epoch": 1.279655990510083,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004843914139375695,
      "loss": 5.3708,
      "mean_token_accuracy": 0.18296968638896943,
      "num_tokens": 29644626.0,
      "step": 12945
    },
    {
      "entropy": 5.5537426471710205,
      "epoch": 1.2801502570185845,
      "grad_norm": 1.484375,
      "learning_rate": 0.00048437849645395714,
      "loss": 5.3422,
      "mean_token_accuracy": 0.18308935314416885,
      "num_tokens": 29655587.0,
      "step": 12950
    },
    {
      "entropy": 5.691879844665527,
      "epoch": 1.2806445235270858,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004843655738196013,
      "loss": 5.3843,
      "mean_token_accuracy": 0.1844848096370697,
      "num_tokens": 29665067.0,
      "step": 12955
    },
    {
      "entropy": 5.70751805305481,
      "epoch": 1.281138790035587,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004843526460348199,
      "loss": 5.373,
      "mean_token_accuracy": 0.1804453819990158,
      "num_tokens": 29676663.0,
      "step": 12960
    },
    {
      "entropy": 5.522183752059936,
      "epoch": 1.2816330565440885,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004843397130999309,
      "loss": 5.3273,
      "mean_token_accuracy": 0.18253915756940842,
      "num_tokens": 29688242.0,
      "step": 12965
    },
    {
      "entropy": 5.678976202011109,
      "epoch": 1.28212732305259,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048432677501525253,
      "loss": 5.4461,
      "mean_token_accuracy": 0.16873905509710313,
      "num_tokens": 29699184.0,
      "step": 12970
    },
    {
      "entropy": 5.613884782791137,
      "epoch": 1.2826215895610913,
      "grad_norm": 1.703125,
      "learning_rate": 0.000484313831781103,
      "loss": 5.3743,
      "mean_token_accuracy": 0.17882349491119384,
      "num_tokens": 29709610.0,
      "step": 12975
    },
    {
      "entropy": 5.573097181320191,
      "epoch": 1.2831158560695928,
      "grad_norm": 1.5,
      "learning_rate": 0.0004843008833978007,
      "loss": 5.3236,
      "mean_token_accuracy": 0.1840164691209793,
      "num_tokens": 29720359.0,
      "step": 12980
    },
    {
      "entropy": 5.6951256275177,
      "epoch": 1.283610122578094,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048428792986566425,
      "loss": 5.5031,
      "mean_token_accuracy": 0.1707326963543892,
      "num_tokens": 29732411.0,
      "step": 12985
    },
    {
      "entropy": 5.68478479385376,
      "epoch": 1.2841043890865955,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004842749711850122,
      "loss": 5.2516,
      "mean_token_accuracy": 0.19107532799243926,
      "num_tokens": 29743394.0,
      "step": 12990
    },
    {
      "entropy": 5.639246797561645,
      "epoch": 1.2845986555950968,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00048426200735616343,
      "loss": 5.4283,
      "mean_token_accuracy": 0.17719055712223053,
      "num_tokens": 29754798.0,
      "step": 12995
    },
    {
      "entropy": 5.648987913131714,
      "epoch": 1.2850929221035983,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004842490383794368,
      "loss": 5.4049,
      "mean_token_accuracy": 0.18132104277610778,
      "num_tokens": 29767891.0,
      "step": 13000
    },
    {
      "entropy": 5.588558721542358,
      "epoch": 1.2855871886120998,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048423606425515136,
      "loss": 5.2791,
      "mean_token_accuracy": 0.18616220355033875,
      "num_tokens": 29780264.0,
      "step": 13005
    },
    {
      "entropy": 5.65120644569397,
      "epoch": 1.286081455120601,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048422308498362635,
      "loss": 5.4696,
      "mean_token_accuracy": 0.17653971463441848,
      "num_tokens": 29792184.0,
      "step": 13010
    },
    {
      "entropy": 5.647147989273071,
      "epoch": 1.2865757216291023,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004842101005651809,
      "loss": 5.3701,
      "mean_token_accuracy": 0.17719782441854476,
      "num_tokens": 29802651.0,
      "step": 13015
    },
    {
      "entropy": 5.607454633712768,
      "epoch": 1.2870699881376038,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004841971110001346,
      "loss": 5.3772,
      "mean_token_accuracy": 0.1856875702738762,
      "num_tokens": 29814172.0,
      "step": 13020
    },
    {
      "entropy": 5.711566686630249,
      "epoch": 1.2875642546461052,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048418411628880684,
      "loss": 5.4764,
      "mean_token_accuracy": 0.171826508641243,
      "num_tokens": 29826636.0,
      "step": 13025
    },
    {
      "entropy": 5.603142166137696,
      "epoch": 1.2880585211546065,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004841711164315174,
      "loss": 5.3736,
      "mean_token_accuracy": 0.17839546054601668,
      "num_tokens": 29838532.0,
      "step": 13030
    },
    {
      "entropy": 5.600117206573486,
      "epoch": 1.288552787663108,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000484158111428586,
      "loss": 5.3632,
      "mean_token_accuracy": 0.18144417703151702,
      "num_tokens": 29849889.0,
      "step": 13035
    },
    {
      "entropy": 5.745006275177002,
      "epoch": 1.2890470541716093,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048414510128033265,
      "loss": 5.4663,
      "mean_token_accuracy": 0.1679910659790039,
      "num_tokens": 29860856.0,
      "step": 13040
    },
    {
      "entropy": 5.638733720779419,
      "epoch": 1.2895413206801107,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004841320859870774,
      "loss": 5.3746,
      "mean_token_accuracy": 0.18261008709669113,
      "num_tokens": 29871921.0,
      "step": 13045
    },
    {
      "entropy": 5.5799013614654545,
      "epoch": 1.290035587188612,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048411906554914037,
      "loss": 5.3832,
      "mean_token_accuracy": 0.1773080348968506,
      "num_tokens": 29883410.0,
      "step": 13050
    },
    {
      "entropy": 5.750530576705932,
      "epoch": 1.2905298536971135,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048410603996684184,
      "loss": 5.5035,
      "mean_token_accuracy": 0.17208611369132995,
      "num_tokens": 29895640.0,
      "step": 13055
    },
    {
      "entropy": 5.629776525497436,
      "epoch": 1.291024120205615,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004840930092405023,
      "loss": 5.3102,
      "mean_token_accuracy": 0.18957563042640685,
      "num_tokens": 29907797.0,
      "step": 13060
    },
    {
      "entropy": 5.530115270614624,
      "epoch": 1.2915183867141162,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048407997337044235,
      "loss": 5.3943,
      "mean_token_accuracy": 0.18360997438430787,
      "num_tokens": 29919085.0,
      "step": 13065
    },
    {
      "entropy": 5.658555698394776,
      "epoch": 1.2920126532226177,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004840669323569826,
      "loss": 5.4169,
      "mean_token_accuracy": 0.17061354964971542,
      "num_tokens": 29930658.0,
      "step": 13070
    },
    {
      "entropy": 5.692487525939941,
      "epoch": 1.292506919731119,
      "grad_norm": 1.5625,
      "learning_rate": 0.00048405388620044383,
      "loss": 5.3639,
      "mean_token_accuracy": 0.1829603925347328,
      "num_tokens": 29942550.0,
      "step": 13075
    },
    {
      "entropy": 5.569294452667236,
      "epoch": 1.2930011862396205,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004840408349011471,
      "loss": 5.3512,
      "mean_token_accuracy": 0.1800552323460579,
      "num_tokens": 29953600.0,
      "step": 13080
    },
    {
      "entropy": 5.6464011669158936,
      "epoch": 1.2934954527481217,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004840277784594133,
      "loss": 5.368,
      "mean_token_accuracy": 0.18142091184854509,
      "num_tokens": 29965694.0,
      "step": 13085
    },
    {
      "entropy": 5.645879220962525,
      "epoch": 1.2939897192566232,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004840147168755638,
      "loss": 5.3797,
      "mean_token_accuracy": 0.18047715574502946,
      "num_tokens": 29977216.0,
      "step": 13090
    },
    {
      "entropy": 5.55439863204956,
      "epoch": 1.2944839857651247,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004840016501499198,
      "loss": 5.2762,
      "mean_token_accuracy": 0.18975744545459747,
      "num_tokens": 29988036.0,
      "step": 13095
    },
    {
      "entropy": 5.681125593185425,
      "epoch": 1.294978252273626,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048398857828280286,
      "loss": 5.4432,
      "mean_token_accuracy": 0.1805913582444191,
      "num_tokens": 30000711.0,
      "step": 13100
    },
    {
      "entropy": 5.695340204238891,
      "epoch": 1.2954725187821272,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048397550127453445,
      "loss": 5.4626,
      "mean_token_accuracy": 0.17566647827625276,
      "num_tokens": 30013607.0,
      "step": 13105
    },
    {
      "entropy": 5.6256498336792,
      "epoch": 1.2959667852906287,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004839624191254362,
      "loss": 5.3773,
      "mean_token_accuracy": 0.178213170170784,
      "num_tokens": 30024375.0,
      "step": 13110
    },
    {
      "entropy": 5.620825958251953,
      "epoch": 1.2964610517991302,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048394933183583004,
      "loss": 5.424,
      "mean_token_accuracy": 0.1755230352282524,
      "num_tokens": 30035476.0,
      "step": 13115
    },
    {
      "entropy": 5.6692657470703125,
      "epoch": 1.2969553183076314,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000483936239406038,
      "loss": 5.4222,
      "mean_token_accuracy": 0.17723745703697205,
      "num_tokens": 30047193.0,
      "step": 13120
    },
    {
      "entropy": 5.735631275177002,
      "epoch": 1.297449584816133,
      "grad_norm": 2.921875,
      "learning_rate": 0.000483923141836382,
      "loss": 5.4347,
      "mean_token_accuracy": 0.17823024243116378,
      "num_tokens": 30059873.0,
      "step": 13125
    },
    {
      "entropy": 5.679801034927368,
      "epoch": 1.2979438513246342,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004839100391271843,
      "loss": 5.4182,
      "mean_token_accuracy": 0.17939580380916595,
      "num_tokens": 30071318.0,
      "step": 13130
    },
    {
      "entropy": 5.5987104892730715,
      "epoch": 1.2984381178331357,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004838969312787672,
      "loss": 5.3939,
      "mean_token_accuracy": 0.18115571439266204,
      "num_tokens": 30082750.0,
      "step": 13135
    },
    {
      "entropy": 5.632969570159912,
      "epoch": 1.298932384341637,
      "grad_norm": 1.5,
      "learning_rate": 0.0004838838182914532,
      "loss": 5.4,
      "mean_token_accuracy": 0.17292409688234328,
      "num_tokens": 30093730.0,
      "step": 13140
    },
    {
      "entropy": 5.715986251831055,
      "epoch": 1.2994266508501384,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004838707001655649,
      "loss": 5.4078,
      "mean_token_accuracy": 0.18056964725255967,
      "num_tokens": 30104776.0,
      "step": 13145
    },
    {
      "entropy": 5.558435106277466,
      "epoch": 1.29992091735864,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004838575769014249,
      "loss": 5.2964,
      "mean_token_accuracy": 0.18821469694375992,
      "num_tokens": 30115110.0,
      "step": 13150
    },
    {
      "entropy": 5.619421100616455,
      "epoch": 1.3004151838671412,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004838444484993562,
      "loss": 5.3691,
      "mean_token_accuracy": 0.18099021315574645,
      "num_tokens": 30126678.0,
      "step": 13155
    },
    {
      "entropy": 5.6659797668457035,
      "epoch": 1.3009094503756424,
      "grad_norm": 1.5,
      "learning_rate": 0.00048383131495968165,
      "loss": 5.3447,
      "mean_token_accuracy": 0.17956814169883728,
      "num_tokens": 30137703.0,
      "step": 13160
    },
    {
      "entropy": 5.579706239700317,
      "epoch": 1.301403716884144,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00048381817628272427,
      "loss": 5.4064,
      "mean_token_accuracy": 0.17738770693540573,
      "num_tokens": 30147624.0,
      "step": 13165
    },
    {
      "entropy": 5.613440656661988,
      "epoch": 1.3018979833926454,
      "grad_norm": 1.5,
      "learning_rate": 0.0004838050324688074,
      "loss": 5.3454,
      "mean_token_accuracy": 0.18368444889783858,
      "num_tokens": 30158863.0,
      "step": 13170
    },
    {
      "entropy": 5.64213981628418,
      "epoch": 1.3023922499011467,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004837918835182544,
      "loss": 5.3858,
      "mean_token_accuracy": 0.1810169920325279,
      "num_tokens": 30170876.0,
      "step": 13175
    },
    {
      "entropy": 5.606496524810791,
      "epoch": 1.3028865164096481,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004837787294313886,
      "loss": 5.3255,
      "mean_token_accuracy": 0.1836916372179985,
      "num_tokens": 30182267.0,
      "step": 13180
    },
    {
      "entropy": 5.632263326644898,
      "epoch": 1.3033807829181494,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004837655702085337,
      "loss": 5.4063,
      "mean_token_accuracy": 0.1785965159535408,
      "num_tokens": 30193572.0,
      "step": 13185
    },
    {
      "entropy": 5.615349769592285,
      "epoch": 1.303875049426651,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004837524058500134,
      "loss": 5.3248,
      "mean_token_accuracy": 0.18587929308414458,
      "num_tokens": 30206341.0,
      "step": 13190
    },
    {
      "entropy": 5.663844728469849,
      "epoch": 1.3043693159351522,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048373923635615146,
      "loss": 5.4009,
      "mean_token_accuracy": 0.17908063381910325,
      "num_tokens": 30216067.0,
      "step": 13195
    },
    {
      "entropy": 5.630838346481323,
      "epoch": 1.3048635824436536,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048372606172727197,
      "loss": 5.4723,
      "mean_token_accuracy": 0.17101092785596847,
      "num_tokens": 30227033.0,
      "step": 13200
    },
    {
      "entropy": 5.700982999801636,
      "epoch": 1.3053578489521551,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000483712881963699,
      "loss": 5.4258,
      "mean_token_accuracy": 0.1803620934486389,
      "num_tokens": 30238766.0,
      "step": 13205
    },
    {
      "entropy": 5.660584402084351,
      "epoch": 1.3058521154606564,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048369969706575666,
      "loss": 5.4746,
      "mean_token_accuracy": 0.17368588596582413,
      "num_tokens": 30250180.0,
      "step": 13210
    },
    {
      "entropy": 5.590638971328735,
      "epoch": 1.3063463819691576,
      "grad_norm": 1.25,
      "learning_rate": 0.00048368650703376956,
      "loss": 5.274,
      "mean_token_accuracy": 0.19140132665634155,
      "num_tokens": 30260704.0,
      "step": 13215
    },
    {
      "entropy": 5.59091215133667,
      "epoch": 1.3068406484776591,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004836733118680619,
      "loss": 5.3594,
      "mean_token_accuracy": 0.18530572801828385,
      "num_tokens": 30273547.0,
      "step": 13220
    },
    {
      "entropy": 5.605325078964233,
      "epoch": 1.3073349149861606,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004836601115689584,
      "loss": 5.3659,
      "mean_token_accuracy": 0.17476717978715897,
      "num_tokens": 30284872.0,
      "step": 13225
    },
    {
      "entropy": 5.570671653747558,
      "epoch": 1.3078291814946619,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004836469061367838,
      "loss": 5.399,
      "mean_token_accuracy": 0.17890868335962296,
      "num_tokens": 30296768.0,
      "step": 13230
    },
    {
      "entropy": 5.591505002975464,
      "epoch": 1.3083234480031634,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048363369557186293,
      "loss": 5.3012,
      "mean_token_accuracy": 0.18937211483716965,
      "num_tokens": 30308307.0,
      "step": 13235
    },
    {
      "entropy": 5.722180032730103,
      "epoch": 1.3088177145116646,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048362047987452085,
      "loss": 5.3935,
      "mean_token_accuracy": 0.17431413680315017,
      "num_tokens": 30318987.0,
      "step": 13240
    },
    {
      "entropy": 5.643228530883789,
      "epoch": 1.309311981020166,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004836072590450825,
      "loss": 5.4181,
      "mean_token_accuracy": 0.1802129030227661,
      "num_tokens": 30330401.0,
      "step": 13245
    },
    {
      "entropy": 5.645843267440796,
      "epoch": 1.3098062475286674,
      "grad_norm": 1.40625,
      "learning_rate": 0.00048359403308387327,
      "loss": 5.3463,
      "mean_token_accuracy": 0.1787939503788948,
      "num_tokens": 30341485.0,
      "step": 13250
    },
    {
      "entropy": 5.693995666503906,
      "epoch": 1.3103005140371689,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00048358080199121845,
      "loss": 5.4316,
      "mean_token_accuracy": 0.1705599918961525,
      "num_tokens": 30352854.0,
      "step": 13255
    },
    {
      "entropy": 5.679081201553345,
      "epoch": 1.3107947805456703,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048356756576744354,
      "loss": 5.4305,
      "mean_token_accuracy": 0.178836627304554,
      "num_tokens": 30363944.0,
      "step": 13260
    },
    {
      "entropy": 5.6488221168518065,
      "epoch": 1.3112890470541716,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004835543244128742,
      "loss": 5.4293,
      "mean_token_accuracy": 0.1715848907828331,
      "num_tokens": 30375678.0,
      "step": 13265
    },
    {
      "entropy": 5.635829257965088,
      "epoch": 1.3117833135626729,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004835410779278361,
      "loss": 5.407,
      "mean_token_accuracy": 0.18080176115036012,
      "num_tokens": 30387206.0,
      "step": 13270
    },
    {
      "entropy": 5.642653369903565,
      "epoch": 1.3122775800711743,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004835278263126551,
      "loss": 5.4147,
      "mean_token_accuracy": 0.1762957602739334,
      "num_tokens": 30398769.0,
      "step": 13275
    },
    {
      "entropy": 5.664876508712768,
      "epoch": 1.3127718465796758,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004835145695676572,
      "loss": 5.3668,
      "mean_token_accuracy": 0.18843472748994827,
      "num_tokens": 30409626.0,
      "step": 13280
    },
    {
      "entropy": 5.629417657852173,
      "epoch": 1.313266113088177,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004835013076931686,
      "loss": 5.3782,
      "mean_token_accuracy": 0.17804671674966813,
      "num_tokens": 30421142.0,
      "step": 13285
    },
    {
      "entropy": 5.59480242729187,
      "epoch": 1.3137603795966786,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048348804068951547,
      "loss": 5.2425,
      "mean_token_accuracy": 0.1974676877260208,
      "num_tokens": 30430438.0,
      "step": 13290
    },
    {
      "entropy": 5.644433116912841,
      "epoch": 1.3142546461051798,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048347476855702413,
      "loss": 5.4122,
      "mean_token_accuracy": 0.17402077913284303,
      "num_tokens": 30442351.0,
      "step": 13295
    },
    {
      "entropy": 5.6162878513336185,
      "epoch": 1.3147489126136813,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048346149129602125,
      "loss": 5.444,
      "mean_token_accuracy": 0.1723179340362549,
      "num_tokens": 30454987.0,
      "step": 13300
    },
    {
      "entropy": 5.59311089515686,
      "epoch": 1.3152431791221826,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00048344820890683333,
      "loss": 5.3345,
      "mean_token_accuracy": 0.18361980020999907,
      "num_tokens": 30466186.0,
      "step": 13305
    },
    {
      "entropy": 5.6795103549957275,
      "epoch": 1.315737445630684,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048343492138978704,
      "loss": 5.4492,
      "mean_token_accuracy": 0.17615232169628142,
      "num_tokens": 30477944.0,
      "step": 13310
    },
    {
      "entropy": 5.65897126197815,
      "epoch": 1.3162317121391856,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004834216287452094,
      "loss": 5.3699,
      "mean_token_accuracy": 0.18468839526176453,
      "num_tokens": 30489922.0,
      "step": 13315
    },
    {
      "entropy": 5.608570003509522,
      "epoch": 1.3167259786476868,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004834083309734274,
      "loss": 5.3909,
      "mean_token_accuracy": 0.18573866486549379,
      "num_tokens": 30501739.0,
      "step": 13320
    },
    {
      "entropy": 5.702604866027832,
      "epoch": 1.3172202451561883,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004833950280747681,
      "loss": 5.4557,
      "mean_token_accuracy": 0.1728788986802101,
      "num_tokens": 30512213.0,
      "step": 13325
    },
    {
      "entropy": 5.69100136756897,
      "epoch": 1.3177145116646896,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004833817200495588,
      "loss": 5.3906,
      "mean_token_accuracy": 0.17468763142824173,
      "num_tokens": 30524300.0,
      "step": 13330
    },
    {
      "entropy": 5.623618125915527,
      "epoch": 1.318208778173191,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048336840689812676,
      "loss": 5.4309,
      "mean_token_accuracy": 0.17311019748449324,
      "num_tokens": 30536137.0,
      "step": 13335
    },
    {
      "entropy": 5.611327219009399,
      "epoch": 1.3187030446816923,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048335508862079975,
      "loss": 5.3965,
      "mean_token_accuracy": 0.1716880276799202,
      "num_tokens": 30548657.0,
      "step": 13340
    },
    {
      "entropy": 5.6862232208251955,
      "epoch": 1.3191973111901938,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048334176521790513,
      "loss": 5.4205,
      "mean_token_accuracy": 0.17718607038259507,
      "num_tokens": 30560274.0,
      "step": 13345
    },
    {
      "entropy": 5.574866247177124,
      "epoch": 1.3196915776986953,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004833284366897707,
      "loss": 5.2735,
      "mean_token_accuracy": 0.18969914615154265,
      "num_tokens": 30571026.0,
      "step": 13350
    },
    {
      "entropy": 5.578272104263306,
      "epoch": 1.3201858442071965,
      "grad_norm": 1.578125,
      "learning_rate": 0.00048331510303672456,
      "loss": 5.2936,
      "mean_token_accuracy": 0.18864040672779084,
      "num_tokens": 30582210.0,
      "step": 13355
    },
    {
      "entropy": 5.644514846801758,
      "epoch": 1.3206801107156978,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004833017642590945,
      "loss": 5.4444,
      "mean_token_accuracy": 0.17006874233484268,
      "num_tokens": 30594049.0,
      "step": 13360
    },
    {
      "entropy": 5.675704860687256,
      "epoch": 1.3211743772241993,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00048328842035720863,
      "loss": 5.398,
      "mean_token_accuracy": 0.17326983362436293,
      "num_tokens": 30604713.0,
      "step": 13365
    },
    {
      "entropy": 5.597809505462647,
      "epoch": 1.3216686437327008,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004832750713313954,
      "loss": 5.3361,
      "mean_token_accuracy": 0.17634675949811934,
      "num_tokens": 30615906.0,
      "step": 13370
    },
    {
      "entropy": 5.643567705154419,
      "epoch": 1.322162910241202,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048326171718198315,
      "loss": 5.4583,
      "mean_token_accuracy": 0.17781806215643883,
      "num_tokens": 30628191.0,
      "step": 13375
    },
    {
      "entropy": 5.679811906814575,
      "epoch": 1.3226571767497035,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048324835790930027,
      "loss": 5.4629,
      "mean_token_accuracy": 0.17555269449949265,
      "num_tokens": 30640667.0,
      "step": 13380
    },
    {
      "entropy": 5.63090705871582,
      "epoch": 1.3231514432582048,
      "grad_norm": 2.84375,
      "learning_rate": 0.0004832349935136755,
      "loss": 5.4278,
      "mean_token_accuracy": 0.17625583261251448,
      "num_tokens": 30651779.0,
      "step": 13385
    },
    {
      "entropy": 5.5488059520721436,
      "epoch": 1.3236457097667063,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004832216239954376,
      "loss": 5.2572,
      "mean_token_accuracy": 0.19061069041490555,
      "num_tokens": 30662396.0,
      "step": 13390
    },
    {
      "entropy": 5.62912073135376,
      "epoch": 1.3241399762752075,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004832082493549154,
      "loss": 5.3982,
      "mean_token_accuracy": 0.1802899122238159,
      "num_tokens": 30674476.0,
      "step": 13395
    },
    {
      "entropy": 5.6277824401855465,
      "epoch": 1.324634242783709,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004831948695924381,
      "loss": 5.3869,
      "mean_token_accuracy": 0.18105309307575226,
      "num_tokens": 30687417.0,
      "step": 13400
    },
    {
      "entropy": 5.647561931610108,
      "epoch": 1.3251285092922105,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004831814847083346,
      "loss": 5.4238,
      "mean_token_accuracy": 0.17972079813480377,
      "num_tokens": 30698925.0,
      "step": 13405
    },
    {
      "entropy": 5.659072542190552,
      "epoch": 1.3256227758007118,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00048316809470293426,
      "loss": 5.4237,
      "mean_token_accuracy": 0.17830995619297027,
      "num_tokens": 30711237.0,
      "step": 13410
    },
    {
      "entropy": 5.6586585521698,
      "epoch": 1.326117042309213,
      "grad_norm": 1.625,
      "learning_rate": 0.00048315469957656656,
      "loss": 5.4432,
      "mean_token_accuracy": 0.17360139042139053,
      "num_tokens": 30721900.0,
      "step": 13415
    },
    {
      "entropy": 5.680593538284302,
      "epoch": 1.3266113088177145,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004831412993295609,
      "loss": 5.3856,
      "mean_token_accuracy": 0.17664714008569718,
      "num_tokens": 30734738.0,
      "step": 13420
    },
    {
      "entropy": 5.645633602142334,
      "epoch": 1.327105575326216,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048312789396224697,
      "loss": 5.4344,
      "mean_token_accuracy": 0.17951468527317047,
      "num_tokens": 30745849.0,
      "step": 13425
    },
    {
      "entropy": 5.661584997177124,
      "epoch": 1.3275998418347172,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048311448347495466,
      "loss": 5.3987,
      "mean_token_accuracy": 0.1778587281703949,
      "num_tokens": 30757118.0,
      "step": 13430
    },
    {
      "entropy": 5.692550134658814,
      "epoch": 1.3280941083432187,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004831010678680137,
      "loss": 5.4904,
      "mean_token_accuracy": 0.17638929784297944,
      "num_tokens": 30768692.0,
      "step": 13435
    },
    {
      "entropy": 5.683389329910279,
      "epoch": 1.32858837485172,
      "grad_norm": 2.875,
      "learning_rate": 0.0004830876471417542,
      "loss": 5.4344,
      "mean_token_accuracy": 0.17975543588399887,
      "num_tokens": 30780241.0,
      "step": 13440
    },
    {
      "entropy": 5.606634998321534,
      "epoch": 1.3290826413602215,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004830742212965063,
      "loss": 5.4172,
      "mean_token_accuracy": 0.17563898265361785,
      "num_tokens": 30792733.0,
      "step": 13445
    },
    {
      "entropy": 5.620973730087281,
      "epoch": 1.3295769078687227,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004830607903326003,
      "loss": 5.3679,
      "mean_token_accuracy": 0.1809845268726349,
      "num_tokens": 30803997.0,
      "step": 13450
    },
    {
      "entropy": 5.593581581115723,
      "epoch": 1.3300711743772242,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004830473542503665,
      "loss": 5.3356,
      "mean_token_accuracy": 0.18298769295215606,
      "num_tokens": 30815322.0,
      "step": 13455
    },
    {
      "entropy": 5.6125345706939695,
      "epoch": 1.3305654408857257,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048303391305013556,
      "loss": 5.3639,
      "mean_token_accuracy": 0.17978909313678743,
      "num_tokens": 30825572.0,
      "step": 13460
    },
    {
      "entropy": 5.599508953094483,
      "epoch": 1.331059707394227,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048302046673223816,
      "loss": 5.2795,
      "mean_token_accuracy": 0.18489835411310196,
      "num_tokens": 30837451.0,
      "step": 13465
    },
    {
      "entropy": 5.587284898757934,
      "epoch": 1.3315539739027282,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004830070152970049,
      "loss": 5.3613,
      "mean_token_accuracy": 0.1831515297293663,
      "num_tokens": 30849418.0,
      "step": 13470
    },
    {
      "entropy": 5.75219316482544,
      "epoch": 1.3320482404112297,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048299355874476685,
      "loss": 5.4682,
      "mean_token_accuracy": 0.1709510028362274,
      "num_tokens": 30860387.0,
      "step": 13475
    },
    {
      "entropy": 5.5926683902740475,
      "epoch": 1.3325425069197312,
      "grad_norm": 1.328125,
      "learning_rate": 0.000482980097075855,
      "loss": 5.3502,
      "mean_token_accuracy": 0.18210737854242326,
      "num_tokens": 30871599.0,
      "step": 13480
    },
    {
      "entropy": 5.561595392227173,
      "epoch": 1.3330367734282325,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048296663029060046,
      "loss": 5.2984,
      "mean_token_accuracy": 0.18697010576725007,
      "num_tokens": 30882298.0,
      "step": 13485
    },
    {
      "entropy": 5.571579360961914,
      "epoch": 1.333531039936734,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004829531583893345,
      "loss": 5.3999,
      "mean_token_accuracy": 0.18073733001947404,
      "num_tokens": 30893589.0,
      "step": 13490
    },
    {
      "entropy": 5.619697856903076,
      "epoch": 1.3340253064452352,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004829396813723886,
      "loss": 5.3424,
      "mean_token_accuracy": 0.17959402352571488,
      "num_tokens": 30905041.0,
      "step": 13495
    },
    {
      "entropy": 5.674782562255859,
      "epoch": 1.3345195729537367,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004829261992400944,
      "loss": 5.4271,
      "mean_token_accuracy": 0.17518270611763,
      "num_tokens": 30916244.0,
      "step": 13500
    },
    {
      "entropy": 5.638361072540283,
      "epoch": 1.335013839462238,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004829127119927833,
      "loss": 5.3625,
      "mean_token_accuracy": 0.18202189803123475,
      "num_tokens": 30927444.0,
      "step": 13505
    },
    {
      "entropy": 5.583319044113159,
      "epoch": 1.3355081059707394,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004828992196307873,
      "loss": 5.3668,
      "mean_token_accuracy": 0.18954140990972518,
      "num_tokens": 30939402.0,
      "step": 13510
    },
    {
      "entropy": 5.689163446426392,
      "epoch": 1.336002372479241,
      "grad_norm": 1.59375,
      "learning_rate": 0.00048288572215443825,
      "loss": 5.4479,
      "mean_token_accuracy": 0.17430518418550492,
      "num_tokens": 30951598.0,
      "step": 13515
    },
    {
      "entropy": 5.66749677658081,
      "epoch": 1.3364966389877422,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004828722195640681,
      "loss": 5.4196,
      "mean_token_accuracy": 0.17818957567214966,
      "num_tokens": 30963897.0,
      "step": 13520
    },
    {
      "entropy": 5.6022481441497805,
      "epoch": 1.3369909054962434,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00048285871186000916,
      "loss": 5.3421,
      "mean_token_accuracy": 0.17971886992454528,
      "num_tokens": 30975143.0,
      "step": 13525
    },
    {
      "entropy": 5.644520902633667,
      "epoch": 1.337485172004745,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048284519904259374,
      "loss": 5.3607,
      "mean_token_accuracy": 0.17373865097761154,
      "num_tokens": 30986656.0,
      "step": 13530
    },
    {
      "entropy": 5.713407182693482,
      "epoch": 1.3379794385132464,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004828316811121541,
      "loss": 5.4979,
      "mean_token_accuracy": 0.1779793843626976,
      "num_tokens": 30999904.0,
      "step": 13535
    },
    {
      "entropy": 5.618749570846558,
      "epoch": 1.3384737050217477,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048281815806902286,
      "loss": 5.2598,
      "mean_token_accuracy": 0.18969714790582656,
      "num_tokens": 31012287.0,
      "step": 13540
    },
    {
      "entropy": 5.639257431030273,
      "epoch": 1.3389679715302492,
      "grad_norm": 1.375,
      "learning_rate": 0.00048280462991353267,
      "loss": 5.3672,
      "mean_token_accuracy": 0.1851990595459938,
      "num_tokens": 31023954.0,
      "step": 13545
    },
    {
      "entropy": 5.573872852325439,
      "epoch": 1.3394622380387504,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004827910966460164,
      "loss": 5.3272,
      "mean_token_accuracy": 0.18471778333187103,
      "num_tokens": 31035365.0,
      "step": 13550
    },
    {
      "entropy": 5.612487077713013,
      "epoch": 1.339956504547252,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004827775582668069,
      "loss": 5.3592,
      "mean_token_accuracy": 0.1845427244901657,
      "num_tokens": 31046814.0,
      "step": 13555
    },
    {
      "entropy": 5.572548675537109,
      "epoch": 1.3404507710557532,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004827640147762372,
      "loss": 5.3232,
      "mean_token_accuracy": 0.18366092592477798,
      "num_tokens": 31058112.0,
      "step": 13560
    },
    {
      "entropy": 5.55669584274292,
      "epoch": 1.3409450375642546,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00048275046617464055,
      "loss": 5.3288,
      "mean_token_accuracy": 0.18401081562042237,
      "num_tokens": 31068703.0,
      "step": 13565
    },
    {
      "entropy": 5.645600700378418,
      "epoch": 1.3414393040727561,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004827369124623502,
      "loss": 5.364,
      "mean_token_accuracy": 0.18467544466257096,
      "num_tokens": 31080018.0,
      "step": 13570
    },
    {
      "entropy": 5.678955984115601,
      "epoch": 1.3419335705812574,
      "grad_norm": 2.0,
      "learning_rate": 0.00048272335363969956,
      "loss": 5.4271,
      "mean_token_accuracy": 0.17893292903900146,
      "num_tokens": 31091103.0,
      "step": 13575
    },
    {
      "entropy": 5.706071186065674,
      "epoch": 1.3424278370897589,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004827097897070222,
      "loss": 5.4285,
      "mean_token_accuracy": 0.17564288526773453,
      "num_tokens": 31102731.0,
      "step": 13580
    },
    {
      "entropy": 5.541435718536377,
      "epoch": 1.3429221035982601,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004826962206646518,
      "loss": 5.2902,
      "mean_token_accuracy": 0.1952822133898735,
      "num_tokens": 31114080.0,
      "step": 13585
    },
    {
      "entropy": 5.575323390960693,
      "epoch": 1.3434163701067616,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048268264651292215,
      "loss": 5.3583,
      "mean_token_accuracy": 0.17633841186761856,
      "num_tokens": 31125875.0,
      "step": 13590
    },
    {
      "entropy": 5.661813354492187,
      "epoch": 1.3439106366152629,
      "grad_norm": 2.234375,
      "learning_rate": 0.00048266906725216713,
      "loss": 5.356,
      "mean_token_accuracy": 0.18115421682596206,
      "num_tokens": 31139246.0,
      "step": 13595
    },
    {
      "entropy": 5.604208612442017,
      "epoch": 1.3444049031237644,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00048265548288272086,
      "loss": 5.2787,
      "mean_token_accuracy": 0.1843094438314438,
      "num_tokens": 31150606.0,
      "step": 13600
    },
    {
      "entropy": 5.570496940612793,
      "epoch": 1.3448991696322659,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048264189340491755,
      "loss": 5.4089,
      "mean_token_accuracy": 0.180218605697155,
      "num_tokens": 31162413.0,
      "step": 13605
    },
    {
      "entropy": 5.655467796325683,
      "epoch": 1.3453934361407671,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048262829881909135,
      "loss": 5.4031,
      "mean_token_accuracy": 0.1763067677617073,
      "num_tokens": 31172832.0,
      "step": 13610
    },
    {
      "entropy": 5.650714921951294,
      "epoch": 1.3458877026492684,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048261469912557685,
      "loss": 5.4257,
      "mean_token_accuracy": 0.17724828869104386,
      "num_tokens": 31186172.0,
      "step": 13615
    },
    {
      "entropy": 5.635502576828003,
      "epoch": 1.3463819691577699,
      "grad_norm": 1.640625,
      "learning_rate": 0.00048260109432470856,
      "loss": 5.2927,
      "mean_token_accuracy": 0.1912631407380104,
      "num_tokens": 31197580.0,
      "step": 13620
    },
    {
      "entropy": 5.613646411895752,
      "epoch": 1.3468762356662713,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048258748441682114,
      "loss": 5.3589,
      "mean_token_accuracy": 0.18065993934869767,
      "num_tokens": 31209262.0,
      "step": 13625
    },
    {
      "entropy": 5.5887237071990965,
      "epoch": 1.3473705021747726,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004825738694022494,
      "loss": 5.3555,
      "mean_token_accuracy": 0.18029915988445283,
      "num_tokens": 31221722.0,
      "step": 13630
    },
    {
      "entropy": 5.602186250686645,
      "epoch": 1.347864768683274,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048256024928132826,
      "loss": 5.3071,
      "mean_token_accuracy": 0.1955119326710701,
      "num_tokens": 31232871.0,
      "step": 13635
    },
    {
      "entropy": 5.613846206665039,
      "epoch": 1.3483590351917754,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004825466240543928,
      "loss": 5.3822,
      "mean_token_accuracy": 0.1794242352247238,
      "num_tokens": 31244010.0,
      "step": 13640
    },
    {
      "entropy": 5.56691312789917,
      "epoch": 1.3488533017002768,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004825329937217782,
      "loss": 5.3769,
      "mean_token_accuracy": 0.17975466847419738,
      "num_tokens": 31255643.0,
      "step": 13645
    },
    {
      "entropy": 5.712325811386108,
      "epoch": 1.349347568208778,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004825193582838196,
      "loss": 5.476,
      "mean_token_accuracy": 0.17626722902059555,
      "num_tokens": 31266764.0,
      "step": 13650
    },
    {
      "entropy": 5.691398477554321,
      "epoch": 1.3498418347172796,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004825057177408528,
      "loss": 5.3355,
      "mean_token_accuracy": 0.1884911060333252,
      "num_tokens": 31277533.0,
      "step": 13655
    },
    {
      "entropy": 5.59542818069458,
      "epoch": 1.350336101225781,
      "grad_norm": 1.34375,
      "learning_rate": 0.000482492072093213,
      "loss": 5.358,
      "mean_token_accuracy": 0.17958372235298156,
      "num_tokens": 31289509.0,
      "step": 13660
    },
    {
      "entropy": 5.599795770645142,
      "epoch": 1.3508303677342823,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048247842134123615,
      "loss": 5.4155,
      "mean_token_accuracy": 0.1811332657933235,
      "num_tokens": 31300909.0,
      "step": 13665
    },
    {
      "entropy": 5.663826608657837,
      "epoch": 1.3513246342427836,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004824647654852579,
      "loss": 5.4295,
      "mean_token_accuracy": 0.1764205664396286,
      "num_tokens": 31312148.0,
      "step": 13670
    },
    {
      "entropy": 5.603689527511596,
      "epoch": 1.351818900751285,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004824511045256142,
      "loss": 5.2806,
      "mean_token_accuracy": 0.1938604086637497,
      "num_tokens": 31323388.0,
      "step": 13675
    },
    {
      "entropy": 5.589712715148925,
      "epoch": 1.3523131672597866,
      "grad_norm": 1.796875,
      "learning_rate": 0.00048243743846264125,
      "loss": 5.2603,
      "mean_token_accuracy": 0.18790659308433533,
      "num_tokens": 31334103.0,
      "step": 13680
    },
    {
      "entropy": 5.608765602111816,
      "epoch": 1.3528074337682878,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004824237672966751,
      "loss": 5.3598,
      "mean_token_accuracy": 0.1805233985185623,
      "num_tokens": 31345456.0,
      "step": 13685
    },
    {
      "entropy": 5.521354341506958,
      "epoch": 1.3533017002767893,
      "grad_norm": 1.5625,
      "learning_rate": 0.000482410091028052,
      "loss": 5.2303,
      "mean_token_accuracy": 0.19384558498859406,
      "num_tokens": 31356482.0,
      "step": 13690
    },
    {
      "entropy": 5.634257793426514,
      "epoch": 1.3537959667852906,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00048239640965710854,
      "loss": 5.4417,
      "mean_token_accuracy": 0.1719844087958336,
      "num_tokens": 31367103.0,
      "step": 13695
    },
    {
      "entropy": 5.691772508621216,
      "epoch": 1.354290233293792,
      "grad_norm": 1.828125,
      "learning_rate": 0.00048238272318418127,
      "loss": 5.4011,
      "mean_token_accuracy": 0.17740811556577682,
      "num_tokens": 31378564.0,
      "step": 13700
    },
    {
      "entropy": 5.645037221908569,
      "epoch": 1.3547844998022933,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048236903160960677,
      "loss": 5.3894,
      "mean_token_accuracy": 0.18022141307592393,
      "num_tokens": 31391435.0,
      "step": 13705
    },
    {
      "entropy": 5.6197977542877195,
      "epoch": 1.3552787663107948,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000482355334933722,
      "loss": 5.4029,
      "mean_token_accuracy": 0.1759616881608963,
      "num_tokens": 31403422.0,
      "step": 13710
    },
    {
      "entropy": 5.604930877685547,
      "epoch": 1.3557730328192963,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048234163315686375,
      "loss": 5.2503,
      "mean_token_accuracy": 0.19544390738010406,
      "num_tokens": 31414940.0,
      "step": 13715
    },
    {
      "entropy": 5.6139758110046385,
      "epoch": 1.3562672993277975,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004823279262793692,
      "loss": 5.4644,
      "mean_token_accuracy": 0.17626293748617172,
      "num_tokens": 31426369.0,
      "step": 13720
    },
    {
      "entropy": 5.577859354019165,
      "epoch": 1.3567615658362988,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048231421430157547,
      "loss": 5.3865,
      "mean_token_accuracy": 0.18294183164834976,
      "num_tokens": 31438426.0,
      "step": 13725
    },
    {
      "entropy": 5.561729478836059,
      "epoch": 1.3572558323448003,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004823004972238199,
      "loss": 5.3096,
      "mean_token_accuracy": 0.1826370596885681,
      "num_tokens": 31450230.0,
      "step": 13730
    },
    {
      "entropy": 5.656190872192383,
      "epoch": 1.3577500988533018,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048228677504643996,
      "loss": 5.4813,
      "mean_token_accuracy": 0.17086569517850875,
      "num_tokens": 31462564.0,
      "step": 13735
    },
    {
      "entropy": 5.6548137187957765,
      "epoch": 1.358244365361803,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004822730477697732,
      "loss": 5.328,
      "mean_token_accuracy": 0.18942943811416627,
      "num_tokens": 31475053.0,
      "step": 13740
    },
    {
      "entropy": 5.570001649856567,
      "epoch": 1.3587386318703045,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004822593153941573,
      "loss": 5.3696,
      "mean_token_accuracy": 0.18805706948041917,
      "num_tokens": 31487187.0,
      "step": 13745
    },
    {
      "entropy": 5.668126106262207,
      "epoch": 1.3592328983788058,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048224557791993013,
      "loss": 5.4042,
      "mean_token_accuracy": 0.18166119307279588,
      "num_tokens": 31499288.0,
      "step": 13750
    },
    {
      "entropy": 5.618986558914185,
      "epoch": 1.3597271648873073,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048223183534742953,
      "loss": 5.3463,
      "mean_token_accuracy": 0.18207521587610245,
      "num_tokens": 31510113.0,
      "step": 13755
    },
    {
      "entropy": 5.575040054321289,
      "epoch": 1.3602214313958085,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048221808767699373,
      "loss": 5.3085,
      "mean_token_accuracy": 0.1831543982028961,
      "num_tokens": 31520900.0,
      "step": 13760
    },
    {
      "entropy": 5.651335525512695,
      "epoch": 1.36071569790431,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004822043349089607,
      "loss": 5.3724,
      "mean_token_accuracy": 0.1856143057346344,
      "num_tokens": 31533085.0,
      "step": 13765
    },
    {
      "entropy": 5.6594603061676025,
      "epoch": 1.3612099644128115,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004821905770436689,
      "loss": 5.3707,
      "mean_token_accuracy": 0.17452322691679,
      "num_tokens": 31544287.0,
      "step": 13770
    },
    {
      "entropy": 5.567810916900635,
      "epoch": 1.3617042309213128,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004821768140814568,
      "loss": 5.3368,
      "mean_token_accuracy": 0.18782113939523698,
      "num_tokens": 31555233.0,
      "step": 13775
    },
    {
      "entropy": 5.607983779907227,
      "epoch": 1.362198497429814,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004821630460226629,
      "loss": 5.2746,
      "mean_token_accuracy": 0.18981120437383653,
      "num_tokens": 31565421.0,
      "step": 13780
    },
    {
      "entropy": 5.626432085037232,
      "epoch": 1.3626927639383155,
      "grad_norm": 1.375,
      "learning_rate": 0.000482149272867626,
      "loss": 5.3883,
      "mean_token_accuracy": 0.1856262981891632,
      "num_tokens": 31576787.0,
      "step": 13785
    },
    {
      "entropy": 5.64390196800232,
      "epoch": 1.363187030446817,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048213549461668476,
      "loss": 5.4412,
      "mean_token_accuracy": 0.17790410965681075,
      "num_tokens": 31587563.0,
      "step": 13790
    },
    {
      "entropy": 5.643861103057861,
      "epoch": 1.3636812969553183,
      "grad_norm": 1.765625,
      "learning_rate": 0.00048212171127017835,
      "loss": 5.3339,
      "mean_token_accuracy": 0.18473720997571946,
      "num_tokens": 31598186.0,
      "step": 13795
    },
    {
      "entropy": 5.666094207763672,
      "epoch": 1.3641755634638197,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004821079228284456,
      "loss": 5.3588,
      "mean_token_accuracy": 0.1768627107143402,
      "num_tokens": 31609458.0,
      "step": 13800
    },
    {
      "entropy": 5.628021001815796,
      "epoch": 1.364669829972321,
      "grad_norm": 1.53125,
      "learning_rate": 0.00048209412929182586,
      "loss": 5.2808,
      "mean_token_accuracy": 0.18881404101848603,
      "num_tokens": 31619497.0,
      "step": 13805
    },
    {
      "entropy": 5.6065410614013675,
      "epoch": 1.3651640964808225,
      "grad_norm": 1.5,
      "learning_rate": 0.0004820803306606584,
      "loss": 5.3858,
      "mean_token_accuracy": 0.17492412477731706,
      "num_tokens": 31630838.0,
      "step": 13810
    },
    {
      "entropy": 5.652844095230103,
      "epoch": 1.3656583629893237,
      "grad_norm": 2.359375,
      "learning_rate": 0.0004820665269352827,
      "loss": 5.4446,
      "mean_token_accuracy": 0.1760430321097374,
      "num_tokens": 31642234.0,
      "step": 13815
    },
    {
      "entropy": 5.6957813739776615,
      "epoch": 1.3661526294978252,
      "grad_norm": 1.703125,
      "learning_rate": 0.00048205271811603827,
      "loss": 5.4052,
      "mean_token_accuracy": 0.17412530183792113,
      "num_tokens": 31653283.0,
      "step": 13820
    },
    {
      "entropy": 5.693255805969239,
      "epoch": 1.3666468960063267,
      "grad_norm": 2.1875,
      "learning_rate": 0.0004820389042032649,
      "loss": 5.3407,
      "mean_token_accuracy": 0.18445192277431488,
      "num_tokens": 31663284.0,
      "step": 13825
    },
    {
      "entropy": 5.555064392089844,
      "epoch": 1.367141162514828,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004820250851973023,
      "loss": 5.3024,
      "mean_token_accuracy": 0.18553722500801087,
      "num_tokens": 31674379.0,
      "step": 13830
    },
    {
      "entropy": 5.6579591751098635,
      "epoch": 1.3676354290233292,
      "grad_norm": 1.5,
      "learning_rate": 0.0004820112610984906,
      "loss": 5.4597,
      "mean_token_accuracy": 0.17677143663167955,
      "num_tokens": 31686420.0,
      "step": 13835
    },
    {
      "entropy": 5.556261730194092,
      "epoch": 1.3681296955318307,
      "grad_norm": 2.171875,
      "learning_rate": 0.00048199743190716966,
      "loss": 5.2543,
      "mean_token_accuracy": 0.19362638145685196,
      "num_tokens": 31696919.0,
      "step": 13840
    },
    {
      "entropy": 5.591260147094727,
      "epoch": 1.3686239620403322,
      "grad_norm": 1.8359375,
      "learning_rate": 0.00048198359762367983,
      "loss": 5.3678,
      "mean_token_accuracy": 0.17782708555459975,
      "num_tokens": 31708878.0,
      "step": 13845
    },
    {
      "entropy": 5.6437853336334225,
      "epoch": 1.3691182285488335,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00048196975824836135,
      "loss": 5.388,
      "mean_token_accuracy": 0.1815121129155159,
      "num_tokens": 31719348.0,
      "step": 13850
    },
    {
      "entropy": 5.621160697937012,
      "epoch": 1.369612495057335,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004819559137815547,
      "loss": 5.3446,
      "mean_token_accuracy": 0.1858145609498024,
      "num_tokens": 31729349.0,
      "step": 13855
    },
    {
      "entropy": 5.674253654479981,
      "epoch": 1.3701067615658362,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004819420642236005,
      "loss": 5.3615,
      "mean_token_accuracy": 0.18043985068798066,
      "num_tokens": 31740345.0,
      "step": 13860
    },
    {
      "entropy": 5.676819467544556,
      "epoch": 1.3706010280743377,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004819282095748395,
      "loss": 5.3866,
      "mean_token_accuracy": 0.17638078480958938,
      "num_tokens": 31752231.0,
      "step": 13865
    },
    {
      "entropy": 5.654842710494995,
      "epoch": 1.371095294582839,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004819143498356123,
      "loss": 5.3926,
      "mean_token_accuracy": 0.1823098376393318,
      "num_tokens": 31764248.0,
      "step": 13870
    },
    {
      "entropy": 5.551671409606934,
      "epoch": 1.3715895610913404,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004819004850062601,
      "loss": 5.2235,
      "mean_token_accuracy": 0.18993889391422272,
      "num_tokens": 31774173.0,
      "step": 13875
    },
    {
      "entropy": 5.609506845474243,
      "epoch": 1.372083827599842,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048188661508712376,
      "loss": 5.3902,
      "mean_token_accuracy": 0.17847636342048645,
      "num_tokens": 31784860.0,
      "step": 13880
    },
    {
      "entropy": 5.618414068222046,
      "epoch": 1.3725780941083432,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004818727400785447,
      "loss": 5.4058,
      "mean_token_accuracy": 0.1797900229692459,
      "num_tokens": 31797012.0,
      "step": 13885
    },
    {
      "entropy": 5.597910213470459,
      "epoch": 1.3730723606168447,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004818588599808641,
      "loss": 5.3025,
      "mean_token_accuracy": 0.1837613582611084,
      "num_tokens": 31808216.0,
      "step": 13890
    },
    {
      "entropy": 5.606488990783691,
      "epoch": 1.373566627125346,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048184497479442345,
      "loss": 5.4182,
      "mean_token_accuracy": 0.17232680320739746,
      "num_tokens": 31820097.0,
      "step": 13895
    },
    {
      "entropy": 5.6867293357849125,
      "epoch": 1.3740608936338474,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004818310845195643,
      "loss": 5.4774,
      "mean_token_accuracy": 0.17236514538526534,
      "num_tokens": 31832252.0,
      "step": 13900
    },
    {
      "entropy": 5.651212930679321,
      "epoch": 1.3745551601423487,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004818171891566284,
      "loss": 5.4432,
      "mean_token_accuracy": 0.17456165552139283,
      "num_tokens": 31842928.0,
      "step": 13905
    },
    {
      "entropy": 5.629182624816894,
      "epoch": 1.3750494266508502,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004818032887059576,
      "loss": 5.2923,
      "mean_token_accuracy": 0.19028321355581285,
      "num_tokens": 31854072.0,
      "step": 13910
    },
    {
      "entropy": 5.6791047096252445,
      "epoch": 1.3755436931593517,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004817893831678938,
      "loss": 5.4136,
      "mean_token_accuracy": 0.1805824011564255,
      "num_tokens": 31865163.0,
      "step": 13915
    },
    {
      "entropy": 5.526577377319336,
      "epoch": 1.376037959667853,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00048177547254277904,
      "loss": 5.287,
      "mean_token_accuracy": 0.18845449686050414,
      "num_tokens": 31876583.0,
      "step": 13920
    },
    {
      "entropy": 5.634654188156128,
      "epoch": 1.3765322261763542,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004817615568309557,
      "loss": 5.3403,
      "mean_token_accuracy": 0.18297360688447953,
      "num_tokens": 31886794.0,
      "step": 13925
    },
    {
      "entropy": 5.69839129447937,
      "epoch": 1.3770264926848557,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004817476360327658,
      "loss": 5.4266,
      "mean_token_accuracy": 0.17962846904993057,
      "num_tokens": 31899980.0,
      "step": 13930
    },
    {
      "entropy": 5.559878158569336,
      "epoch": 1.3775207591933571,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004817337101485521,
      "loss": 5.2333,
      "mean_token_accuracy": 0.19485677629709244,
      "num_tokens": 31909906.0,
      "step": 13935
    },
    {
      "entropy": 5.531175947189331,
      "epoch": 1.3780150257018584,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00048171977917865706,
      "loss": 5.3175,
      "mean_token_accuracy": 0.18942902237176895,
      "num_tokens": 31920499.0,
      "step": 13940
    },
    {
      "entropy": 5.533161878585815,
      "epoch": 1.37850929221036,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00048170584312342337,
      "loss": 5.3218,
      "mean_token_accuracy": 0.19081737995147705,
      "num_tokens": 31931031.0,
      "step": 13945
    },
    {
      "entropy": 5.654254961013794,
      "epoch": 1.3790035587188612,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004816919019831939,
      "loss": 5.3158,
      "mean_token_accuracy": 0.18491162210702897,
      "num_tokens": 31941887.0,
      "step": 13950
    },
    {
      "entropy": 5.6600535869598385,
      "epoch": 1.3794978252273626,
      "grad_norm": 1.8828125,
      "learning_rate": 0.00048167795575831154,
      "loss": 5.4021,
      "mean_token_accuracy": 0.17590015530586242,
      "num_tokens": 31953727.0,
      "step": 13955
    },
    {
      "entropy": 5.611099529266357,
      "epoch": 1.379992091735864,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004816640044491193,
      "loss": 5.3638,
      "mean_token_accuracy": 0.18339920341968535,
      "num_tokens": 31965418.0,
      "step": 13960
    },
    {
      "entropy": 5.619025993347168,
      "epoch": 1.3804863582443654,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004816500480559607,
      "loss": 5.3919,
      "mean_token_accuracy": 0.18112977892160415,
      "num_tokens": 31976528.0,
      "step": 13965
    },
    {
      "entropy": 5.642622089385986,
      "epoch": 1.3809806247528669,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048163608657917874,
      "loss": 5.38,
      "mean_token_accuracy": 0.17919955253601075,
      "num_tokens": 31987921.0,
      "step": 13970
    },
    {
      "entropy": 5.587106704711914,
      "epoch": 1.3814748912613681,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000481622120019117,
      "loss": 5.2548,
      "mean_token_accuracy": 0.18843722045421601,
      "num_tokens": 31998824.0,
      "step": 13975
    },
    {
      "entropy": 5.617704248428344,
      "epoch": 1.3819691577698694,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004816081483761191,
      "loss": 5.4425,
      "mean_token_accuracy": 0.17394677996635438,
      "num_tokens": 32011690.0,
      "step": 13980
    },
    {
      "entropy": 5.658710956573486,
      "epoch": 1.3824634242783709,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004815941716505286,
      "loss": 5.4222,
      "mean_token_accuracy": 0.18356541395187378,
      "num_tokens": 32022946.0,
      "step": 13985
    },
    {
      "entropy": 5.655589818954468,
      "epoch": 1.3829576907868724,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048158018984268954,
      "loss": 5.3743,
      "mean_token_accuracy": 0.17571795731782913,
      "num_tokens": 32033683.0,
      "step": 13990
    },
    {
      "entropy": 5.6693085670471195,
      "epoch": 1.3834519572953736,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004815662029529457,
      "loss": 5.4298,
      "mean_token_accuracy": 0.1724245861172676,
      "num_tokens": 32045334.0,
      "step": 13995
    },
    {
      "entropy": 5.589751243591309,
      "epoch": 1.383946223803875,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048155221098164123,
      "loss": 5.316,
      "mean_token_accuracy": 0.18133876323699952,
      "num_tokens": 32056425.0,
      "step": 14000
    },
    {
      "entropy": 5.628468751907349,
      "epoch": 1.3844404903123764,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048153821392912035,
      "loss": 5.3674,
      "mean_token_accuracy": 0.18083936721086502,
      "num_tokens": 32068053.0,
      "step": 14005
    },
    {
      "entropy": 5.663592529296875,
      "epoch": 1.3849347568208779,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004815242117957273,
      "loss": 5.3751,
      "mean_token_accuracy": 0.18200017362833024,
      "num_tokens": 32079462.0,
      "step": 14010
    },
    {
      "entropy": 5.54339189529419,
      "epoch": 1.3854290233293791,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048151020458180667,
      "loss": 5.3512,
      "mean_token_accuracy": 0.18686241209506987,
      "num_tokens": 32090566.0,
      "step": 14015
    },
    {
      "entropy": 5.722955894470215,
      "epoch": 1.3859232898378806,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000481496192287703,
      "loss": 5.43,
      "mean_token_accuracy": 0.17913904488086702,
      "num_tokens": 32101535.0,
      "step": 14020
    },
    {
      "entropy": 5.600596332550049,
      "epoch": 1.386417556346382,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004814821749137609,
      "loss": 5.2919,
      "mean_token_accuracy": 0.18575020730495453,
      "num_tokens": 32112450.0,
      "step": 14025
    },
    {
      "entropy": 5.551318883895874,
      "epoch": 1.3869118228548833,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004814681524603252,
      "loss": 5.4073,
      "mean_token_accuracy": 0.17307053953409196,
      "num_tokens": 32123675.0,
      "step": 14030
    },
    {
      "entropy": 5.6738694190979,
      "epoch": 1.3874060893633846,
      "grad_norm": 1.390625,
      "learning_rate": 0.000481454124927741,
      "loss": 5.4015,
      "mean_token_accuracy": 0.17737227082252502,
      "num_tokens": 32135283.0,
      "step": 14035
    },
    {
      "entropy": 5.607222080230713,
      "epoch": 1.387900355871886,
      "grad_norm": 1.546875,
      "learning_rate": 0.00048144009231635326,
      "loss": 5.3128,
      "mean_token_accuracy": 0.19094865769147873,
      "num_tokens": 32146715.0,
      "step": 14040
    },
    {
      "entropy": 5.5505311489105225,
      "epoch": 1.3883946223803876,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004814260546265073,
      "loss": 5.3444,
      "mean_token_accuracy": 0.1852043941617012,
      "num_tokens": 32158323.0,
      "step": 14045
    },
    {
      "entropy": 5.629556179046631,
      "epoch": 1.3888888888888888,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004814120118585483,
      "loss": 5.4083,
      "mean_token_accuracy": 0.1811718225479126,
      "num_tokens": 32170453.0,
      "step": 14050
    },
    {
      "entropy": 5.740184307098389,
      "epoch": 1.3893831553973903,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004813979640128218,
      "loss": 5.4245,
      "mean_token_accuracy": 0.17536792308092117,
      "num_tokens": 32181596.0,
      "step": 14055
    },
    {
      "entropy": 5.71956353187561,
      "epoch": 1.3898774219058916,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004813839110896734,
      "loss": 5.3981,
      "mean_token_accuracy": 0.18018420189619064,
      "num_tokens": 32194603.0,
      "step": 14060
    },
    {
      "entropy": 5.5669067859649655,
      "epoch": 1.390371688414393,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00048136985308944874,
      "loss": 5.3001,
      "mean_token_accuracy": 0.18715957105159758,
      "num_tokens": 32206064.0,
      "step": 14065
    },
    {
      "entropy": 5.618303680419922,
      "epoch": 1.3908659549228943,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048135579001249367,
      "loss": 5.3729,
      "mean_token_accuracy": 0.1751048281788826,
      "num_tokens": 32217254.0,
      "step": 14070
    },
    {
      "entropy": 5.764079332351685,
      "epoch": 1.3913602214313958,
      "grad_norm": 2.859375,
      "learning_rate": 0.0004813417218591542,
      "loss": 5.5131,
      "mean_token_accuracy": 0.16822995394468307,
      "num_tokens": 32228821.0,
      "step": 14075
    },
    {
      "entropy": 5.6994609355926515,
      "epoch": 1.3918544879398973,
      "grad_norm": 2.5625,
      "learning_rate": 0.0004813276486297763,
      "loss": 5.3707,
      "mean_token_accuracy": 0.17629912197589875,
      "num_tokens": 32239476.0,
      "step": 14080
    },
    {
      "entropy": 5.600485897064209,
      "epoch": 1.3923487544483986,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004813135703247062,
      "loss": 5.3305,
      "mean_token_accuracy": 0.184663724899292,
      "num_tokens": 32250975.0,
      "step": 14085
    },
    {
      "entropy": 5.651822423934936,
      "epoch": 1.3928430209568998,
      "grad_norm": 1.8125,
      "learning_rate": 0.00048129948694429036,
      "loss": 5.3117,
      "mean_token_accuracy": 0.19396121203899383,
      "num_tokens": 32261675.0,
      "step": 14090
    },
    {
      "entropy": 5.5526707649230955,
      "epoch": 1.3933372874654013,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048128539848887517,
      "loss": 5.2218,
      "mean_token_accuracy": 0.1985325649380684,
      "num_tokens": 32272523.0,
      "step": 14095
    },
    {
      "entropy": 5.589223480224609,
      "epoch": 1.3938315539739028,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00048127130495880703,
      "loss": 5.3903,
      "mean_token_accuracy": 0.17744339108467103,
      "num_tokens": 32284140.0,
      "step": 14100
    },
    {
      "entropy": 5.598843765258789,
      "epoch": 1.394325820482404,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004812572063544329,
      "loss": 5.3217,
      "mean_token_accuracy": 0.17905248701572418,
      "num_tokens": 32296286.0,
      "step": 14105
    },
    {
      "entropy": 5.6526764869689945,
      "epoch": 1.3948200869909055,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00048124310267609946,
      "loss": 5.3731,
      "mean_token_accuracy": 0.18060370683670043,
      "num_tokens": 32307394.0,
      "step": 14110
    },
    {
      "entropy": 5.681780576705933,
      "epoch": 1.3953143534994068,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004812289939241537,
      "loss": 5.4591,
      "mean_token_accuracy": 0.17379377037286758,
      "num_tokens": 32318788.0,
      "step": 14115
    },
    {
      "entropy": 5.663927268981934,
      "epoch": 1.3958086200079083,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004812148800989427,
      "loss": 5.4546,
      "mean_token_accuracy": 0.17521916776895524,
      "num_tokens": 32332395.0,
      "step": 14120
    },
    {
      "entropy": 5.688994550704956,
      "epoch": 1.3963028865164095,
      "grad_norm": 1.609375,
      "learning_rate": 0.00048120076120081363,
      "loss": 5.4187,
      "mean_token_accuracy": 0.17732064425945282,
      "num_tokens": 32344226.0,
      "step": 14125
    },
    {
      "entropy": 5.667909383773804,
      "epoch": 1.396797153024911,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048118663723011393,
      "loss": 5.3973,
      "mean_token_accuracy": 0.17809137254953383,
      "num_tokens": 32354677.0,
      "step": 14130
    },
    {
      "entropy": 5.627816009521484,
      "epoch": 1.3972914195334125,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004811725081871909,
      "loss": 5.3023,
      "mean_token_accuracy": 0.18046956956386567,
      "num_tokens": 32365054.0,
      "step": 14135
    },
    {
      "entropy": 5.55205659866333,
      "epoch": 1.3977856860419138,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004811583740723922,
      "loss": 5.293,
      "mean_token_accuracy": 0.19918642193078995,
      "num_tokens": 32376215.0,
      "step": 14140
    },
    {
      "entropy": 5.717932224273682,
      "epoch": 1.3982799525504153,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048114423488606543,
      "loss": 5.3989,
      "mean_token_accuracy": 0.17818801999092101,
      "num_tokens": 32386880.0,
      "step": 14145
    },
    {
      "entropy": 5.639126348495483,
      "epoch": 1.3987742190589165,
      "grad_norm": 1.484375,
      "learning_rate": 0.00048113009062855866,
      "loss": 5.4097,
      "mean_token_accuracy": 0.18310538828372955,
      "num_tokens": 32398227.0,
      "step": 14150
    },
    {
      "entropy": 5.680119609832763,
      "epoch": 1.399268485567418,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004811159413002196,
      "loss": 5.4268,
      "mean_token_accuracy": 0.177865169942379,
      "num_tokens": 32410091.0,
      "step": 14155
    },
    {
      "entropy": 5.661602926254273,
      "epoch": 1.3997627520759193,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004811017869013963,
      "loss": 5.3683,
      "mean_token_accuracy": 0.17924144864082336,
      "num_tokens": 32422710.0,
      "step": 14160
    },
    {
      "entropy": 5.65218677520752,
      "epoch": 1.4002570185844208,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004810876274324372,
      "loss": 5.405,
      "mean_token_accuracy": 0.1841844215989113,
      "num_tokens": 32433691.0,
      "step": 14165
    },
    {
      "entropy": 5.608447647094726,
      "epoch": 1.4007512850929222,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004810734628936905,
      "loss": 5.3363,
      "mean_token_accuracy": 0.18748170435428618,
      "num_tokens": 32445370.0,
      "step": 14170
    },
    {
      "entropy": 5.606389427185059,
      "epoch": 1.4012455516014235,
      "grad_norm": 1.453125,
      "learning_rate": 0.00048105929328550465,
      "loss": 5.307,
      "mean_token_accuracy": 0.18583703488111497,
      "num_tokens": 32456737.0,
      "step": 14175
    },
    {
      "entropy": 5.5614348411560055,
      "epoch": 1.4017398181099248,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004810451186082282,
      "loss": 5.3748,
      "mean_token_accuracy": 0.17483311295509338,
      "num_tokens": 32467957.0,
      "step": 14180
    },
    {
      "entropy": 5.632024574279785,
      "epoch": 1.4022340846184262,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048103093886220984,
      "loss": 5.4392,
      "mean_token_accuracy": 0.1730569750070572,
      "num_tokens": 32481697.0,
      "step": 14185
    },
    {
      "entropy": 5.6059290885925295,
      "epoch": 1.4027283511269277,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048101675404779834,
      "loss": 5.2545,
      "mean_token_accuracy": 0.1884591683745384,
      "num_tokens": 32492828.0,
      "step": 14190
    },
    {
      "entropy": 5.736994743347168,
      "epoch": 1.403222617635429,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004810025641653428,
      "loss": 5.506,
      "mean_token_accuracy": 0.17533777952194213,
      "num_tokens": 32505120.0,
      "step": 14195
    },
    {
      "entropy": 5.582390546798706,
      "epoch": 1.4037168841439305,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004809883692151923,
      "loss": 5.3188,
      "mean_token_accuracy": 0.18648334741592407,
      "num_tokens": 32516837.0,
      "step": 14200
    },
    {
      "entropy": 5.57723822593689,
      "epoch": 1.4042111506524317,
      "grad_norm": 1.6875,
      "learning_rate": 0.00048097416919769595,
      "loss": 5.2735,
      "mean_token_accuracy": 0.19068894982337953,
      "num_tokens": 32526906.0,
      "step": 14205
    },
    {
      "entropy": 5.657097005844117,
      "epoch": 1.4047054171609332,
      "grad_norm": 1.53125,
      "learning_rate": 0.00048095996411320294,
      "loss": 5.425,
      "mean_token_accuracy": 0.17666495591402054,
      "num_tokens": 32537929.0,
      "step": 14210
    },
    {
      "entropy": 5.59872350692749,
      "epoch": 1.4051996836694345,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00048094575396206296,
      "loss": 5.4146,
      "mean_token_accuracy": 0.18222443759441376,
      "num_tokens": 32549458.0,
      "step": 14215
    },
    {
      "entropy": 5.615854358673095,
      "epoch": 1.405693950177936,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004809315387446255,
      "loss": 5.3442,
      "mean_token_accuracy": 0.18087542057037354,
      "num_tokens": 32560244.0,
      "step": 14220
    },
    {
      "entropy": 5.6728517532348635,
      "epoch": 1.4061882166864375,
      "grad_norm": 1.375,
      "learning_rate": 0.0004809173184612402,
      "loss": 5.4353,
      "mean_token_accuracy": 0.17692623138427735,
      "num_tokens": 32571337.0,
      "step": 14225
    },
    {
      "entropy": 5.589094638824463,
      "epoch": 1.4066824831949387,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004809030931122569,
      "loss": 5.3429,
      "mean_token_accuracy": 0.1807697132229805,
      "num_tokens": 32582325.0,
      "step": 14230
    },
    {
      "entropy": 5.706233787536621,
      "epoch": 1.40717674970344,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048088886269802565,
      "loss": 5.4044,
      "mean_token_accuracy": 0.1818245232105255,
      "num_tokens": 32592953.0,
      "step": 14235
    },
    {
      "entropy": 5.615331029891967,
      "epoch": 1.4076710162119415,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004808746272188964,
      "loss": 5.3158,
      "mean_token_accuracy": 0.18511224091053008,
      "num_tokens": 32604190.0,
      "step": 14240
    },
    {
      "entropy": 5.6900841236114506,
      "epoch": 1.408165282720443,
      "grad_norm": 1.671875,
      "learning_rate": 0.00048086038667521935,
      "loss": 5.3138,
      "mean_token_accuracy": 0.18650416880846024,
      "num_tokens": 32614467.0,
      "step": 14245
    },
    {
      "entropy": 5.555642032623291,
      "epoch": 1.4086595492289442,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004808461410673449,
      "loss": 5.3312,
      "mean_token_accuracy": 0.1829972445964813,
      "num_tokens": 32624022.0,
      "step": 14250
    },
    {
      "entropy": 5.52761435508728,
      "epoch": 1.4091538157374457,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004808318903956234,
      "loss": 5.3392,
      "mean_token_accuracy": 0.18747462332248688,
      "num_tokens": 32634363.0,
      "step": 14255
    },
    {
      "entropy": 5.600794696807862,
      "epoch": 1.409648082245947,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004808176346604055,
      "loss": 5.274,
      "mean_token_accuracy": 0.1910204216837883,
      "num_tokens": 32646143.0,
      "step": 14260
    },
    {
      "entropy": 5.624323749542237,
      "epoch": 1.4101423487544484,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004808033738620419,
      "loss": 5.3624,
      "mean_token_accuracy": 0.17982067912817,
      "num_tokens": 32658735.0,
      "step": 14265
    },
    {
      "entropy": 5.684943056106567,
      "epoch": 1.4106366152629497,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00048078910800088327,
      "loss": 5.4734,
      "mean_token_accuracy": 0.17568115144968033,
      "num_tokens": 32669337.0,
      "step": 14270
    },
    {
      "entropy": 5.5618006706237795,
      "epoch": 1.4111308817714512,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004807748370772807,
      "loss": 5.2686,
      "mean_token_accuracy": 0.1903432548046112,
      "num_tokens": 32680072.0,
      "step": 14275
    },
    {
      "entropy": 5.59342131614685,
      "epoch": 1.4116251482799527,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048076056109158525,
      "loss": 5.3226,
      "mean_token_accuracy": 0.18741660118103026,
      "num_tokens": 32691142.0,
      "step": 14280
    },
    {
      "entropy": 5.68325514793396,
      "epoch": 1.412119414788454,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004807462800441481,
      "loss": 5.4168,
      "mean_token_accuracy": 0.17906888872385024,
      "num_tokens": 32702482.0,
      "step": 14285
    },
    {
      "entropy": 5.658294582366944,
      "epoch": 1.4126136812969552,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004807319939353204,
      "loss": 5.4193,
      "mean_token_accuracy": 0.17739390581846237,
      "num_tokens": 32714456.0,
      "step": 14290
    },
    {
      "entropy": 5.650444650650025,
      "epoch": 1.4131079478054567,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048071770276545377,
      "loss": 5.3477,
      "mean_token_accuracy": 0.177822445333004,
      "num_tokens": 32726305.0,
      "step": 14295
    },
    {
      "entropy": 5.518589162826538,
      "epoch": 1.4136022143139582,
      "grad_norm": 1.453125,
      "learning_rate": 0.00048070340653489984,
      "loss": 5.2948,
      "mean_token_accuracy": 0.19069047570228576,
      "num_tokens": 32736446.0,
      "step": 14300
    },
    {
      "entropy": 5.62140154838562,
      "epoch": 1.4140964808224594,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048068910524401014,
      "loss": 5.3666,
      "mean_token_accuracy": 0.17754603624343873,
      "num_tokens": 32746798.0,
      "step": 14305
    },
    {
      "entropy": 5.5328234195709225,
      "epoch": 1.414590747330961,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00048067479889313654,
      "loss": 5.2784,
      "mean_token_accuracy": 0.1890805631875992,
      "num_tokens": 32758647.0,
      "step": 14310
    },
    {
      "entropy": 5.580541706085205,
      "epoch": 1.4150850138394622,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048066048748263097,
      "loss": 5.357,
      "mean_token_accuracy": 0.1755592092871666,
      "num_tokens": 32770759.0,
      "step": 14315
    },
    {
      "entropy": 5.640110445022583,
      "epoch": 1.4155792803479637,
      "grad_norm": 1.8203125,
      "learning_rate": 0.00048064617101284553,
      "loss": 5.3578,
      "mean_token_accuracy": 0.18834201991558075,
      "num_tokens": 32783398.0,
      "step": 14320
    },
    {
      "entropy": 5.665228271484375,
      "epoch": 1.416073546856465,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048063184948413237,
      "loss": 5.3701,
      "mean_token_accuracy": 0.17614586055278778,
      "num_tokens": 32795995.0,
      "step": 14325
    },
    {
      "entropy": 5.674857711791992,
      "epoch": 1.4165678133649664,
      "grad_norm": 1.65625,
      "learning_rate": 0.00048061752289684385,
      "loss": 5.4351,
      "mean_token_accuracy": 0.17878583669662476,
      "num_tokens": 32808700.0,
      "step": 14330
    },
    {
      "entropy": 5.661518812179565,
      "epoch": 1.4170620798734679,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004806031912513323,
      "loss": 5.5123,
      "mean_token_accuracy": 0.1630460724234581,
      "num_tokens": 32820815.0,
      "step": 14335
    },
    {
      "entropy": 5.639752626419067,
      "epoch": 1.4175563463819691,
      "grad_norm": 1.5625,
      "learning_rate": 0.00048058885454795036,
      "loss": 5.3892,
      "mean_token_accuracy": 0.18315636217594147,
      "num_tokens": 32832216.0,
      "step": 14340
    },
    {
      "entropy": 5.700220727920533,
      "epoch": 1.4180506128904704,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004805745127870508,
      "loss": 5.4666,
      "mean_token_accuracy": 0.17161145955324172,
      "num_tokens": 32845118.0,
      "step": 14345
    },
    {
      "entropy": 5.6551275730133055,
      "epoch": 1.418544879398972,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004805601659689862,
      "loss": 5.3048,
      "mean_token_accuracy": 0.1845885768532753,
      "num_tokens": 32855662.0,
      "step": 14350
    },
    {
      "entropy": 5.639504289627075,
      "epoch": 1.4190391459074734,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004805458140941097,
      "loss": 5.4253,
      "mean_token_accuracy": 0.18153348565101624,
      "num_tokens": 32865530.0,
      "step": 14355
    },
    {
      "entropy": 5.6355565071105955,
      "epoch": 1.4195334124159746,
      "grad_norm": 2.03125,
      "learning_rate": 0.00048053145716277424,
      "loss": 5.4103,
      "mean_token_accuracy": 0.17455045580863954,
      "num_tokens": 32877542.0,
      "step": 14360
    },
    {
      "entropy": 5.604001712799072,
      "epoch": 1.4200276789244761,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004805170951753331,
      "loss": 5.353,
      "mean_token_accuracy": 0.18324343115091324,
      "num_tokens": 32888197.0,
      "step": 14365
    },
    {
      "entropy": 5.632210922241211,
      "epoch": 1.4205219454329774,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004805027281321395,
      "loss": 5.3738,
      "mean_token_accuracy": 0.17722806483507156,
      "num_tokens": 32899633.0,
      "step": 14370
    },
    {
      "entropy": 5.548191928863526,
      "epoch": 1.4210162119414789,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004804883560335469,
      "loss": 5.3305,
      "mean_token_accuracy": 0.18591197431087494,
      "num_tokens": 32910334.0,
      "step": 14375
    },
    {
      "entropy": 5.695159244537353,
      "epoch": 1.4215104784499801,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004804739788799089,
      "loss": 5.4091,
      "mean_token_accuracy": 0.18576927334070206,
      "num_tokens": 32920071.0,
      "step": 14380
    },
    {
      "entropy": 5.610476636886597,
      "epoch": 1.4220047449584816,
      "grad_norm": 1.375,
      "learning_rate": 0.00048045959667157914,
      "loss": 5.38,
      "mean_token_accuracy": 0.18382232934236525,
      "num_tokens": 32931391.0,
      "step": 14385
    },
    {
      "entropy": 5.6042783737182615,
      "epoch": 1.422499011466983,
      "grad_norm": 1.375,
      "learning_rate": 0.0004804452094089114,
      "loss": 5.3625,
      "mean_token_accuracy": 0.1807640716433525,
      "num_tokens": 32941831.0,
      "step": 14390
    },
    {
      "entropy": 5.620062637329101,
      "epoch": 1.4229932779754844,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004804308170922597,
      "loss": 5.3404,
      "mean_token_accuracy": 0.18507003784179688,
      "num_tokens": 32952949.0,
      "step": 14395
    },
    {
      "entropy": 5.6729881286621096,
      "epoch": 1.4234875444839858,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004804164197219779,
      "loss": 5.4439,
      "mean_token_accuracy": 0.17989460229873658,
      "num_tokens": 32964411.0,
      "step": 14400
    },
    {
      "entropy": 5.682026958465576,
      "epoch": 1.423981810992487,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004804020172984204,
      "loss": 5.3853,
      "mean_token_accuracy": 0.17940158098936082,
      "num_tokens": 32975484.0,
      "step": 14405
    },
    {
      "entropy": 5.593816804885864,
      "epoch": 1.4244760775009886,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048038760982194135,
      "loss": 5.2949,
      "mean_token_accuracy": 0.18834904432296753,
      "num_tokens": 32986691.0,
      "step": 14410
    },
    {
      "entropy": 5.609579086303711,
      "epoch": 1.4249703440094899,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004803731972928954,
      "loss": 5.3163,
      "mean_token_accuracy": 0.18441108167171477,
      "num_tokens": 32998357.0,
      "step": 14415
    },
    {
      "entropy": 5.66904764175415,
      "epoch": 1.4254646105179913,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048035877971163674,
      "loss": 5.3929,
      "mean_token_accuracy": 0.18656598627567292,
      "num_tokens": 33009451.0,
      "step": 14420
    },
    {
      "entropy": 5.507620096206665,
      "epoch": 1.4259588770264928,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004803443570785203,
      "loss": 5.2737,
      "mean_token_accuracy": 0.18558445572853088,
      "num_tokens": 33020143.0,
      "step": 14425
    },
    {
      "entropy": 5.563524532318115,
      "epoch": 1.426453143534994,
      "grad_norm": 1.5625,
      "learning_rate": 0.00048032992939390086,
      "loss": 5.2718,
      "mean_token_accuracy": 0.19021858423948287,
      "num_tokens": 33031448.0,
      "step": 14430
    },
    {
      "entropy": 5.684447813034057,
      "epoch": 1.4269474100434953,
      "grad_norm": 1.375,
      "learning_rate": 0.0004803154966581333,
      "loss": 5.3877,
      "mean_token_accuracy": 0.17983870953321457,
      "num_tokens": 33041545.0,
      "step": 14435
    },
    {
      "entropy": 5.606578302383423,
      "epoch": 1.4274416765519968,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048030105887157253,
      "loss": 5.3479,
      "mean_token_accuracy": 0.18550915122032166,
      "num_tokens": 33051681.0,
      "step": 14440
    },
    {
      "entropy": 5.697551631927491,
      "epoch": 1.4279359430604983,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048028661603457406,
      "loss": 5.3928,
      "mean_token_accuracy": 0.1820794641971588,
      "num_tokens": 33062735.0,
      "step": 14445
    },
    {
      "entropy": 5.568870639801025,
      "epoch": 1.4284302095689996,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004802721681474929,
      "loss": 5.2552,
      "mean_token_accuracy": 0.19466136991977692,
      "num_tokens": 33073975.0,
      "step": 14450
    },
    {
      "entropy": 5.599568557739258,
      "epoch": 1.428924476077501,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048025771521068456,
      "loss": 5.3619,
      "mean_token_accuracy": 0.18302928954362868,
      "num_tokens": 33086928.0,
      "step": 14455
    },
    {
      "entropy": 5.6740195751190186,
      "epoch": 1.4294187425860023,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048024325722450454,
      "loss": 5.5036,
      "mean_token_accuracy": 0.17859254628419877,
      "num_tokens": 33099073.0,
      "step": 14460
    },
    {
      "entropy": 5.719235181808472,
      "epoch": 1.4299130090945038,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004802287941893087,
      "loss": 5.3685,
      "mean_token_accuracy": 0.17799197733402253,
      "num_tokens": 33110245.0,
      "step": 14465
    },
    {
      "entropy": 5.6248149394989015,
      "epoch": 1.430407275603005,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004802143261054525,
      "loss": 5.3428,
      "mean_token_accuracy": 0.18073080480098724,
      "num_tokens": 33121659.0,
      "step": 14470
    },
    {
      "entropy": 5.596703433990479,
      "epoch": 1.4309015421115066,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048019985297329215,
      "loss": 5.415,
      "mean_token_accuracy": 0.1760291948914528,
      "num_tokens": 33133296.0,
      "step": 14475
    },
    {
      "entropy": 5.616138505935669,
      "epoch": 1.431395808620008,
      "grad_norm": 1.25,
      "learning_rate": 0.0004801853747931836,
      "loss": 5.4193,
      "mean_token_accuracy": 0.1771381080150604,
      "num_tokens": 33145069.0,
      "step": 14480
    },
    {
      "entropy": 5.683020067214966,
      "epoch": 1.4318900751285093,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048017089156548293,
      "loss": 5.4134,
      "mean_token_accuracy": 0.1787265807390213,
      "num_tokens": 33156089.0,
      "step": 14485
    },
    {
      "entropy": 5.733552122116089,
      "epoch": 1.4323843416370106,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048015640329054655,
      "loss": 5.3829,
      "mean_token_accuracy": 0.1780915677547455,
      "num_tokens": 33168300.0,
      "step": 14490
    },
    {
      "entropy": 5.55993766784668,
      "epoch": 1.432878608145512,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048014190996873075,
      "loss": 5.272,
      "mean_token_accuracy": 0.19346413612365723,
      "num_tokens": 33179072.0,
      "step": 14495
    },
    {
      "entropy": 5.580727434158325,
      "epoch": 1.4333728746540135,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048012741160039223,
      "loss": 5.3292,
      "mean_token_accuracy": 0.18078672736883164,
      "num_tokens": 33190888.0,
      "step": 14500
    },
    {
      "entropy": 5.663186931610108,
      "epoch": 1.4338671411625148,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004801129081858875,
      "loss": 5.4342,
      "mean_token_accuracy": 0.17907172739505767,
      "num_tokens": 33202747.0,
      "step": 14505
    },
    {
      "entropy": 5.684048700332641,
      "epoch": 1.4343614076710163,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004800983997255735,
      "loss": 5.3817,
      "mean_token_accuracy": 0.18273817002773285,
      "num_tokens": 33213917.0,
      "step": 14510
    },
    {
      "entropy": 5.59083662033081,
      "epoch": 1.4348556741795175,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004800838862198069,
      "loss": 5.3334,
      "mean_token_accuracy": 0.17874141037464142,
      "num_tokens": 33223924.0,
      "step": 14515
    },
    {
      "entropy": 5.618529605865478,
      "epoch": 1.435349940688019,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000480069367668945,
      "loss": 5.2909,
      "mean_token_accuracy": 0.18572077304124832,
      "num_tokens": 33234487.0,
      "step": 14520
    },
    {
      "entropy": 5.598186922073364,
      "epoch": 1.4358442071965203,
      "grad_norm": 1.734375,
      "learning_rate": 0.00048005484407334474,
      "loss": 5.386,
      "mean_token_accuracy": 0.18556090593338012,
      "num_tokens": 33245585.0,
      "step": 14525
    },
    {
      "entropy": 5.606752204895019,
      "epoch": 1.4363384737050218,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004800403154333636,
      "loss": 5.2933,
      "mean_token_accuracy": 0.18794708997011184,
      "num_tokens": 33256545.0,
      "step": 14530
    },
    {
      "entropy": 5.580558729171753,
      "epoch": 1.4368327402135233,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004800257817493589,
      "loss": 5.2852,
      "mean_token_accuracy": 0.1874091848731041,
      "num_tokens": 33267946.0,
      "step": 14535
    },
    {
      "entropy": 5.604639720916748,
      "epoch": 1.4373270067220245,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00048001124302168814,
      "loss": 5.3185,
      "mean_token_accuracy": 0.1865818217396736,
      "num_tokens": 33279099.0,
      "step": 14540
    },
    {
      "entropy": 5.683906126022339,
      "epoch": 1.4378212732305258,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004799966992507089,
      "loss": 5.4147,
      "mean_token_accuracy": 0.1787988394498825,
      "num_tokens": 33290544.0,
      "step": 14545
    },
    {
      "entropy": 5.591766595840454,
      "epoch": 1.4383155397390273,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047998215043677914,
      "loss": 5.268,
      "mean_token_accuracy": 0.18995272368192673,
      "num_tokens": 33302317.0,
      "step": 14550
    },
    {
      "entropy": 5.580928659439087,
      "epoch": 1.4388098062475287,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004799675965802567,
      "loss": 5.3077,
      "mean_token_accuracy": 0.1868557170033455,
      "num_tokens": 33313553.0,
      "step": 14555
    },
    {
      "entropy": 5.645744895935058,
      "epoch": 1.43930407275603,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047995303768149955,
      "loss": 5.3678,
      "mean_token_accuracy": 0.1820383131504059,
      "num_tokens": 33324610.0,
      "step": 14560
    },
    {
      "entropy": 5.574640274047852,
      "epoch": 1.4397983392645315,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004799384737408659,
      "loss": 5.3641,
      "mean_token_accuracy": 0.18238909542560577,
      "num_tokens": 33335562.0,
      "step": 14565
    },
    {
      "entropy": 5.696072626113891,
      "epoch": 1.4402926057730328,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047992390475871396,
      "loss": 5.4947,
      "mean_token_accuracy": 0.17222489416599274,
      "num_tokens": 33346778.0,
      "step": 14570
    },
    {
      "entropy": 5.667906379699707,
      "epoch": 1.4407868722815342,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004799093307354022,
      "loss": 5.3618,
      "mean_token_accuracy": 0.1750493198633194,
      "num_tokens": 33358712.0,
      "step": 14575
    },
    {
      "entropy": 5.6158068656921385,
      "epoch": 1.4412811387900355,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004798947516712891,
      "loss": 5.3751,
      "mean_token_accuracy": 0.18001998364925384,
      "num_tokens": 33369293.0,
      "step": 14580
    },
    {
      "entropy": 5.661997842788696,
      "epoch": 1.441775405298537,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047988016756673335,
      "loss": 5.4379,
      "mean_token_accuracy": 0.17685143798589706,
      "num_tokens": 33383119.0,
      "step": 14585
    },
    {
      "entropy": 5.667682361602783,
      "epoch": 1.4422696718070385,
      "grad_norm": 1.25,
      "learning_rate": 0.0004798655784220937,
      "loss": 5.3517,
      "mean_token_accuracy": 0.18535353243350983,
      "num_tokens": 33395334.0,
      "step": 14590
    },
    {
      "entropy": 5.51068754196167,
      "epoch": 1.4427639383155397,
      "grad_norm": 1.46875,
      "learning_rate": 0.000479850984237729,
      "loss": 5.2635,
      "mean_token_accuracy": 0.19356120079755784,
      "num_tokens": 33406676.0,
      "step": 14595
    },
    {
      "entropy": 5.705871438980102,
      "epoch": 1.443258204824041,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004798363850139983,
      "loss": 5.4089,
      "mean_token_accuracy": 0.17501930743455887,
      "num_tokens": 33419153.0,
      "step": 14600
    },
    {
      "entropy": 5.617748403549195,
      "epoch": 1.4437524713325425,
      "grad_norm": 1.25,
      "learning_rate": 0.0004798217807512608,
      "loss": 5.296,
      "mean_token_accuracy": 0.18446858376264572,
      "num_tokens": 33430146.0,
      "step": 14605
    },
    {
      "entropy": 5.632235765457153,
      "epoch": 1.444246737841044,
      "grad_norm": 1.5,
      "learning_rate": 0.00047980717144987574,
      "loss": 5.4431,
      "mean_token_accuracy": 0.17428609281778334,
      "num_tokens": 33441484.0,
      "step": 14610
    },
    {
      "entropy": 5.666584444046021,
      "epoch": 1.4447410043495452,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047979255711020236,
      "loss": 5.2821,
      "mean_token_accuracy": 0.18434666842222214,
      "num_tokens": 33451817.0,
      "step": 14615
    },
    {
      "entropy": 5.581913089752197,
      "epoch": 1.4452352708580467,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004797779377326004,
      "loss": 5.3709,
      "mean_token_accuracy": 0.1810087189078331,
      "num_tokens": 33463805.0,
      "step": 14620
    },
    {
      "entropy": 5.60614013671875,
      "epoch": 1.445729537366548,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004797633133174295,
      "loss": 5.4005,
      "mean_token_accuracy": 0.17057690471410752,
      "num_tokens": 33475931.0,
      "step": 14625
    },
    {
      "entropy": 5.6683330059051515,
      "epoch": 1.4462238038750495,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004797486838650491,
      "loss": 5.3275,
      "mean_token_accuracy": 0.18055067509412764,
      "num_tokens": 33487306.0,
      "step": 14630
    },
    {
      "entropy": 5.612042379379273,
      "epoch": 1.4467180703835507,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004797340493758195,
      "loss": 5.3693,
      "mean_token_accuracy": 0.17981306463479996,
      "num_tokens": 33499049.0,
      "step": 14635
    },
    {
      "entropy": 5.627777528762818,
      "epoch": 1.4472123368920522,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004797194098501004,
      "loss": 5.3484,
      "mean_token_accuracy": 0.18871730715036392,
      "num_tokens": 33509626.0,
      "step": 14640
    },
    {
      "entropy": 5.622543001174927,
      "epoch": 1.4477066034005537,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004797047652882521,
      "loss": 5.4133,
      "mean_token_accuracy": 0.17613703310489653,
      "num_tokens": 33521510.0,
      "step": 14645
    },
    {
      "entropy": 5.574356555938721,
      "epoch": 1.448200869909055,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047969011569063487,
      "loss": 5.3423,
      "mean_token_accuracy": 0.18178556263446807,
      "num_tokens": 33532622.0,
      "step": 14650
    },
    {
      "entropy": 5.5780839920043945,
      "epoch": 1.4486951364175564,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004796754610576091,
      "loss": 5.3044,
      "mean_token_accuracy": 0.1923196479678154,
      "num_tokens": 33544196.0,
      "step": 14655
    },
    {
      "entropy": 5.6785729885101315,
      "epoch": 1.4491894029260577,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00047966080138953505,
      "loss": 5.4051,
      "mean_token_accuracy": 0.1795321062207222,
      "num_tokens": 33554927.0,
      "step": 14660
    },
    {
      "entropy": 5.669644927978515,
      "epoch": 1.4496836694345592,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004796461366867736,
      "loss": 5.4302,
      "mean_token_accuracy": 0.1686223939061165,
      "num_tokens": 33566412.0,
      "step": 14665
    },
    {
      "entropy": 5.5711634159088135,
      "epoch": 1.4501779359430604,
      "grad_norm": 1.5625,
      "learning_rate": 0.00047963146694968546,
      "loss": 5.3479,
      "mean_token_accuracy": 0.18360590636730195,
      "num_tokens": 33578010.0,
      "step": 14670
    },
    {
      "entropy": 5.643803644180298,
      "epoch": 1.450672202451562,
      "grad_norm": 1.375,
      "learning_rate": 0.0004796167921786315,
      "loss": 5.4302,
      "mean_token_accuracy": 0.17864201962947845,
      "num_tokens": 33589647.0,
      "step": 14675
    },
    {
      "entropy": 5.730944204330444,
      "epoch": 1.4511664689600634,
      "grad_norm": 2.09375,
      "learning_rate": 0.00047960211237397274,
      "loss": 5.5316,
      "mean_token_accuracy": 0.17348184138536454,
      "num_tokens": 33601292.0,
      "step": 14680
    },
    {
      "entropy": 5.689969062805176,
      "epoch": 1.4516607354685647,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047958742753607015,
      "loss": 5.3746,
      "mean_token_accuracy": 0.18604425042867662,
      "num_tokens": 33612878.0,
      "step": 14685
    },
    {
      "entropy": 5.6125401020050045,
      "epoch": 1.452155001977066,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047957273766528524,
      "loss": 5.3719,
      "mean_token_accuracy": 0.18531243354082108,
      "num_tokens": 33625680.0,
      "step": 14690
    },
    {
      "entropy": 5.587348461151123,
      "epoch": 1.4526492684855674,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004795580427619791,
      "loss": 5.312,
      "mean_token_accuracy": 0.18208350688219072,
      "num_tokens": 33637066.0,
      "step": 14695
    },
    {
      "entropy": 5.6511941909790036,
      "epoch": 1.453143534994069,
      "grad_norm": 1.421875,
      "learning_rate": 0.00047954334282651357,
      "loss": 5.4694,
      "mean_token_accuracy": 0.17673402577638625,
      "num_tokens": 33648213.0,
      "step": 14700
    },
    {
      "entropy": 5.667943525314331,
      "epoch": 1.4536378015025702,
      "grad_norm": 2.390625,
      "learning_rate": 0.0004795286378592499,
      "loss": 5.3321,
      "mean_token_accuracy": 0.18379795253276826,
      "num_tokens": 33661582.0,
      "step": 14705
    },
    {
      "entropy": 5.605260801315308,
      "epoch": 1.4541320680110716,
      "grad_norm": 1.375,
      "learning_rate": 0.0004795139278605501,
      "loss": 5.3244,
      "mean_token_accuracy": 0.18694020211696624,
      "num_tokens": 33673122.0,
      "step": 14710
    },
    {
      "entropy": 5.66150803565979,
      "epoch": 1.454626334519573,
      "grad_norm": 1.734375,
      "learning_rate": 0.00047949921283077594,
      "loss": 5.3201,
      "mean_token_accuracy": 0.18457397520542146,
      "num_tokens": 33683926.0,
      "step": 14715
    },
    {
      "entropy": 5.630133533477784,
      "epoch": 1.4551206010280744,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004794844927702894,
      "loss": 5.3744,
      "mean_token_accuracy": 0.18353369385004042,
      "num_tokens": 33694529.0,
      "step": 14720
    },
    {
      "entropy": 5.634875106811523,
      "epoch": 1.4556148675365757,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004794697676794526,
      "loss": 5.3737,
      "mean_token_accuracy": 0.18258169889450074,
      "num_tokens": 33704618.0,
      "step": 14725
    },
    {
      "entropy": 5.579412269592285,
      "epoch": 1.4561091340450771,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004794550375586278,
      "loss": 5.3126,
      "mean_token_accuracy": 0.17945686131715774,
      "num_tokens": 33716354.0,
      "step": 14730
    },
    {
      "entropy": 5.59295973777771,
      "epoch": 1.4566034005535786,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004794403024081774,
      "loss": 5.2918,
      "mean_token_accuracy": 0.19096234291791916,
      "num_tokens": 33728065.0,
      "step": 14735
    },
    {
      "entropy": 5.6533135890960695,
      "epoch": 1.4570976670620799,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004794255622284638,
      "loss": 5.3909,
      "mean_token_accuracy": 0.1789734572172165,
      "num_tokens": 33740207.0,
      "step": 14740
    },
    {
      "entropy": 5.674638748168945,
      "epoch": 1.4575919335705811,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047941081701984966,
      "loss": 5.3328,
      "mean_token_accuracy": 0.1822485476732254,
      "num_tokens": 33750768.0,
      "step": 14745
    },
    {
      "entropy": 5.612480306625367,
      "epoch": 1.4580862000790826,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047939606678269774,
      "loss": 5.3852,
      "mean_token_accuracy": 0.1883511036634445,
      "num_tokens": 33762029.0,
      "step": 14750
    },
    {
      "entropy": 5.574562501907349,
      "epoch": 1.4585804665875841,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004793813115173708,
      "loss": 5.2925,
      "mean_token_accuracy": 0.19058095961809157,
      "num_tokens": 33773483.0,
      "step": 14755
    },
    {
      "entropy": 5.6138105392456055,
      "epoch": 1.4590747330960854,
      "grad_norm": 1.5390625,
      "learning_rate": 0.000479366551224232,
      "loss": 5.336,
      "mean_token_accuracy": 0.18321554660797118,
      "num_tokens": 33784404.0,
      "step": 14760
    },
    {
      "entropy": 5.57039155960083,
      "epoch": 1.4595689996045869,
      "grad_norm": 1.75,
      "learning_rate": 0.0004793517859036443,
      "loss": 5.3379,
      "mean_token_accuracy": 0.18563806414604186,
      "num_tokens": 33794625.0,
      "step": 14765
    },
    {
      "entropy": 5.6151659965515135,
      "epoch": 1.4600632661130881,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004793370155559709,
      "loss": 5.3813,
      "mean_token_accuracy": 0.18430256098508835,
      "num_tokens": 33805872.0,
      "step": 14770
    },
    {
      "entropy": 5.544369125366211,
      "epoch": 1.4605575326215896,
      "grad_norm": 1.625,
      "learning_rate": 0.0004793222401815753,
      "loss": 5.2082,
      "mean_token_accuracy": 0.19724307656288148,
      "num_tokens": 33815979.0,
      "step": 14775
    },
    {
      "entropy": 5.6172308921813965,
      "epoch": 1.4610517991300909,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004793074597808208,
      "loss": 5.4079,
      "mean_token_accuracy": 0.1793598398566246,
      "num_tokens": 33827571.0,
      "step": 14780
    },
    {
      "entropy": 5.597890138626099,
      "epoch": 1.4615460656385924,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047929267435407115,
      "loss": 5.3719,
      "mean_token_accuracy": 0.17711426466703414,
      "num_tokens": 33838749.0,
      "step": 14785
    },
    {
      "entropy": 5.662248945236206,
      "epoch": 1.4620403321470938,
      "grad_norm": 1.53125,
      "learning_rate": 0.00047927788390168997,
      "loss": 5.3564,
      "mean_token_accuracy": 0.1814606159925461,
      "num_tokens": 33849494.0,
      "step": 14790
    },
    {
      "entropy": 5.693349552154541,
      "epoch": 1.462534598655595,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047926308842404126,
      "loss": 5.4203,
      "mean_token_accuracy": 0.17738244980573653,
      "num_tokens": 33860586.0,
      "step": 14795
    },
    {
      "entropy": 5.560922241210937,
      "epoch": 1.4630288651640964,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004792482879214888,
      "loss": 5.1987,
      "mean_token_accuracy": 0.19594061523675918,
      "num_tokens": 33871261.0,
      "step": 14800
    },
    {
      "entropy": 5.586266946792603,
      "epoch": 1.4635231316725978,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004792334823943968,
      "loss": 5.339,
      "mean_token_accuracy": 0.19146328121423722,
      "num_tokens": 33883160.0,
      "step": 14805
    },
    {
      "entropy": 5.6598132133483885,
      "epoch": 1.4640173981810993,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004792186718431294,
      "loss": 5.3454,
      "mean_token_accuracy": 0.18808734118938447,
      "num_tokens": 33894250.0,
      "step": 14810
    },
    {
      "entropy": 5.581802940368652,
      "epoch": 1.4645116646896006,
      "grad_norm": 1.46875,
      "learning_rate": 0.000479203856268051,
      "loss": 5.3342,
      "mean_token_accuracy": 0.18571853786706924,
      "num_tokens": 33905076.0,
      "step": 14815
    },
    {
      "entropy": 5.518166732788086,
      "epoch": 1.465005931198102,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004791890356695261,
      "loss": 5.3344,
      "mean_token_accuracy": 0.18704134076833726,
      "num_tokens": 33916982.0,
      "step": 14820
    },
    {
      "entropy": 5.71618881225586,
      "epoch": 1.4655001977066033,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004791742100479193,
      "loss": 5.4066,
      "mean_token_accuracy": 0.17929370254278182,
      "num_tokens": 33928306.0,
      "step": 14825
    },
    {
      "entropy": 5.745555257797241,
      "epoch": 1.4659944642151048,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004791593794035952,
      "loss": 5.4343,
      "mean_token_accuracy": 0.17960487604141234,
      "num_tokens": 33940371.0,
      "step": 14830
    },
    {
      "entropy": 5.645412349700928,
      "epoch": 1.466488730723606,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004791445437369187,
      "loss": 5.3945,
      "mean_token_accuracy": 0.1763386607170105,
      "num_tokens": 33952250.0,
      "step": 14835
    },
    {
      "entropy": 5.656629753112793,
      "epoch": 1.4669829972321076,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004791297030482547,
      "loss": 5.3727,
      "mean_token_accuracy": 0.18149725198745728,
      "num_tokens": 33963426.0,
      "step": 14840
    },
    {
      "entropy": 5.589711332321167,
      "epoch": 1.467477263740609,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004791148573379684,
      "loss": 5.3499,
      "mean_token_accuracy": 0.1780073583126068,
      "num_tokens": 33974982.0,
      "step": 14845
    },
    {
      "entropy": 5.5783624172210695,
      "epoch": 1.4679715302491103,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047910000660642487,
      "loss": 5.3534,
      "mean_token_accuracy": 0.18175674527883529,
      "num_tokens": 33986413.0,
      "step": 14850
    },
    {
      "entropy": 5.66426191329956,
      "epoch": 1.4684657967576116,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004790851508539896,
      "loss": 5.297,
      "mean_token_accuracy": 0.18410698622465133,
      "num_tokens": 33996821.0,
      "step": 14855
    },
    {
      "entropy": 5.592700386047364,
      "epoch": 1.468960063266113,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004790702900810279,
      "loss": 5.3648,
      "mean_token_accuracy": 0.17952213436365128,
      "num_tokens": 34008817.0,
      "step": 14860
    },
    {
      "entropy": 5.567141056060791,
      "epoch": 1.4694543297746145,
      "grad_norm": 1.8359375,
      "learning_rate": 0.00047905542428790536,
      "loss": 5.358,
      "mean_token_accuracy": 0.18273835480213166,
      "num_tokens": 34020051.0,
      "step": 14865
    },
    {
      "entropy": 5.599887275695801,
      "epoch": 1.4699485962831158,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004790405534749878,
      "loss": 5.3029,
      "mean_token_accuracy": 0.1902901217341423,
      "num_tokens": 34032764.0,
      "step": 14870
    },
    {
      "entropy": 5.639361333847046,
      "epoch": 1.4704428627916173,
      "grad_norm": 1.5,
      "learning_rate": 0.000479025677642641,
      "loss": 5.3703,
      "mean_token_accuracy": 0.17783863842487335,
      "num_tokens": 34042868.0,
      "step": 14875
    },
    {
      "entropy": 5.629166746139527,
      "epoch": 1.4709371293001186,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004790107967912309,
      "loss": 5.3929,
      "mean_token_accuracy": 0.18257658332586288,
      "num_tokens": 34054807.0,
      "step": 14880
    },
    {
      "entropy": 5.662144660949707,
      "epoch": 1.47143139580862,
      "grad_norm": 1.859375,
      "learning_rate": 0.00047899591092112343,
      "loss": 5.4128,
      "mean_token_accuracy": 0.1783519834280014,
      "num_tokens": 34066378.0,
      "step": 14885
    },
    {
      "entropy": 5.615582084655761,
      "epoch": 1.4719256623171213,
      "grad_norm": 1.6484375,
      "learning_rate": 0.000478981020032685,
      "loss": 5.3865,
      "mean_token_accuracy": 0.17028851956129074,
      "num_tokens": 34078267.0,
      "step": 14890
    },
    {
      "entropy": 5.645502710342408,
      "epoch": 1.4724199288256228,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047896612412628184,
      "loss": 5.4083,
      "mean_token_accuracy": 0.17240892648696898,
      "num_tokens": 34090356.0,
      "step": 14895
    },
    {
      "entropy": 5.6976823806762695,
      "epoch": 1.4729141953341243,
      "grad_norm": 1.65625,
      "learning_rate": 0.00047895122320228026,
      "loss": 5.3396,
      "mean_token_accuracy": 0.1850088730454445,
      "num_tokens": 34101143.0,
      "step": 14900
    },
    {
      "entropy": 5.593738746643067,
      "epoch": 1.4734084618426255,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047893631726104703,
      "loss": 5.3688,
      "mean_token_accuracy": 0.18088444620370864,
      "num_tokens": 34112655.0,
      "step": 14905
    },
    {
      "entropy": 5.615711688995361,
      "epoch": 1.4739027283511268,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004789214063029488,
      "loss": 5.3873,
      "mean_token_accuracy": 0.18227738440036773,
      "num_tokens": 34123825.0,
      "step": 14910
    },
    {
      "entropy": 5.62659273147583,
      "epoch": 1.4743969948596283,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047890649032835233,
      "loss": 5.3672,
      "mean_token_accuracy": 0.18306371867656707,
      "num_tokens": 34135416.0,
      "step": 14915
    },
    {
      "entropy": 5.672122526168823,
      "epoch": 1.4748912613681298,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004788915693376245,
      "loss": 5.402,
      "mean_token_accuracy": 0.180454520881176,
      "num_tokens": 34146797.0,
      "step": 14920
    },
    {
      "entropy": 5.705274295806885,
      "epoch": 1.475385527876631,
      "grad_norm": 1.9609375,
      "learning_rate": 0.00047887664333113244,
      "loss": 5.4239,
      "mean_token_accuracy": 0.1766470640897751,
      "num_tokens": 34158749.0,
      "step": 14925
    },
    {
      "entropy": 5.603371286392212,
      "epoch": 1.4758797943851325,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004788617123092434,
      "loss": 5.2922,
      "mean_token_accuracy": 0.1885598421096802,
      "num_tokens": 34169694.0,
      "step": 14930
    },
    {
      "entropy": 5.615245246887207,
      "epoch": 1.4763740608936338,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004788467762723245,
      "loss": 5.4114,
      "mean_token_accuracy": 0.1805598571896553,
      "num_tokens": 34182166.0,
      "step": 14935
    },
    {
      "entropy": 5.634717798233032,
      "epoch": 1.4768683274021353,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047883183522074335,
      "loss": 5.3035,
      "mean_token_accuracy": 0.18629665821790695,
      "num_tokens": 34193651.0,
      "step": 14940
    },
    {
      "entropy": 5.6833878517150875,
      "epoch": 1.4773625939106365,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047881688915486743,
      "loss": 5.4213,
      "mean_token_accuracy": 0.17869849354028702,
      "num_tokens": 34204173.0,
      "step": 14945
    },
    {
      "entropy": 5.597486066818237,
      "epoch": 1.477856860419138,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00047880193807506446,
      "loss": 5.3334,
      "mean_token_accuracy": 0.19496313184499742,
      "num_tokens": 34214872.0,
      "step": 14950
    },
    {
      "entropy": 5.629118156433106,
      "epoch": 1.4783511269276395,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004787869819817021,
      "loss": 5.3415,
      "mean_token_accuracy": 0.18141088783740997,
      "num_tokens": 34226433.0,
      "step": 14955
    },
    {
      "entropy": 5.576784658432007,
      "epoch": 1.4788453934361407,
      "grad_norm": 1.9296875,
      "learning_rate": 0.00047877202087514843,
      "loss": 5.3673,
      "mean_token_accuracy": 0.18549707531929016,
      "num_tokens": 34238047.0,
      "step": 14960
    },
    {
      "entropy": 5.634697961807251,
      "epoch": 1.4793396599446422,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004787570547557715,
      "loss": 5.3913,
      "mean_token_accuracy": 0.18384256958961487,
      "num_tokens": 34250445.0,
      "step": 14965
    },
    {
      "entropy": 5.6461282730102536,
      "epoch": 1.4798339264531435,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00047874208362393925,
      "loss": 5.2671,
      "mean_token_accuracy": 0.18621405214071274,
      "num_tokens": 34261324.0,
      "step": 14970
    },
    {
      "entropy": 5.609235572814941,
      "epoch": 1.480328192961645,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047872710748002024,
      "loss": 5.3665,
      "mean_token_accuracy": 0.18379658907651902,
      "num_tokens": 34272074.0,
      "step": 14975
    },
    {
      "entropy": 5.549933052062988,
      "epoch": 1.4808224594701462,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004787121263243827,
      "loss": 5.3333,
      "mean_token_accuracy": 0.18016744703054427,
      "num_tokens": 34282958.0,
      "step": 14980
    },
    {
      "entropy": 5.621416664123535,
      "epoch": 1.4813167259786477,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047869714015739533,
      "loss": 5.2779,
      "mean_token_accuracy": 0.18833871483802794,
      "num_tokens": 34294999.0,
      "step": 14985
    },
    {
      "entropy": 5.649327230453491,
      "epoch": 1.4818109924871492,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00047868214897942675,
      "loss": 5.3175,
      "mean_token_accuracy": 0.18065880984067917,
      "num_tokens": 34305363.0,
      "step": 14990
    },
    {
      "entropy": 5.6238977909088135,
      "epoch": 1.4823052589956505,
      "grad_norm": 1.375,
      "learning_rate": 0.0004786671527908456,
      "loss": 5.4132,
      "mean_token_accuracy": 0.1785098761320114,
      "num_tokens": 34317467.0,
      "step": 14995
    },
    {
      "entropy": 5.6816877841949465,
      "epoch": 1.4827995255041517,
      "grad_norm": 1.65625,
      "learning_rate": 0.000478652151592021,
      "loss": 5.4319,
      "mean_token_accuracy": 0.17484955340623856,
      "num_tokens": 34331058.0,
      "step": 15000
    },
    {
      "epoch": 1.4827995255041517,
      "eval_entropy": 5.423371398874168,
      "eval_loss": 5.422661304473877,
      "eval_mean_token_accuracy": 0.1868826922308768,
      "eval_num_tokens": 34331058.0,
      "eval_runtime": 26.577,
      "eval_samples_per_second": 1223.349,
      "eval_steps_per_second": 152.952,
      "step": 15000
    },
    {
      "entropy": 5.617215776443482,
      "epoch": 1.4832937920126532,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004786371453833217,
      "loss": 5.4065,
      "mean_token_accuracy": 0.18247174173593522,
      "num_tokens": 34342593.0,
      "step": 15005
    },
    {
      "entropy": 5.674700450897217,
      "epoch": 1.4837880585211547,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00047862213416511717,
      "loss": 5.3839,
      "mean_token_accuracy": 0.17913443297147752,
      "num_tokens": 34353507.0,
      "step": 15010
    },
    {
      "entropy": 5.575383567810059,
      "epoch": 1.484282325029656,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004786071179377766,
      "loss": 5.2837,
      "mean_token_accuracy": 0.1940351888537407,
      "num_tokens": 34365606.0,
      "step": 15015
    },
    {
      "entropy": 5.60756778717041,
      "epoch": 1.4847765915381574,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004785920967016692,
      "loss": 5.3131,
      "mean_token_accuracy": 0.18299914747476578,
      "num_tokens": 34375925.0,
      "step": 15020
    },
    {
      "entropy": 5.669091081619262,
      "epoch": 1.4852708580466587,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047857707045716473,
      "loss": 5.4123,
      "mean_token_accuracy": 0.18027361631393432,
      "num_tokens": 34387806.0,
      "step": 15025
    },
    {
      "entropy": 5.611959981918335,
      "epoch": 1.4857651245551602,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047856203920463273,
      "loss": 5.2747,
      "mean_token_accuracy": 0.1909378230571747,
      "num_tokens": 34400367.0,
      "step": 15030
    },
    {
      "entropy": 5.576943254470825,
      "epoch": 1.4862593910636615,
      "grad_norm": 1.8671875,
      "learning_rate": 0.000478547002944443,
      "loss": 5.3497,
      "mean_token_accuracy": 0.18589189946651458,
      "num_tokens": 34412082.0,
      "step": 15035
    },
    {
      "entropy": 5.598098707199097,
      "epoch": 1.486753657572163,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004785319616769654,
      "loss": 5.3018,
      "mean_token_accuracy": 0.18902516961097718,
      "num_tokens": 34423315.0,
      "step": 15040
    },
    {
      "entropy": 5.659210872650147,
      "epoch": 1.4872479240806644,
      "grad_norm": 1.625,
      "learning_rate": 0.00047851691540256995,
      "loss": 5.3467,
      "mean_token_accuracy": 0.17677914798259736,
      "num_tokens": 34433748.0,
      "step": 15045
    },
    {
      "entropy": 5.602761507034302,
      "epoch": 1.4877421905891657,
      "grad_norm": 1.671875,
      "learning_rate": 0.00047850186412162684,
      "loss": 5.3128,
      "mean_token_accuracy": 0.1851986452937126,
      "num_tokens": 34444425.0,
      "step": 15050
    },
    {
      "entropy": 5.604499626159668,
      "epoch": 1.488236457097667,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004784868078345063,
      "loss": 5.371,
      "mean_token_accuracy": 0.17852768152952195,
      "num_tokens": 34456350.0,
      "step": 15055
    },
    {
      "entropy": 5.699484825134277,
      "epoch": 1.4887307236061684,
      "grad_norm": 2.234375,
      "learning_rate": 0.0004784717465415787,
      "loss": 5.4423,
      "mean_token_accuracy": 0.17603701055049897,
      "num_tokens": 34466606.0,
      "step": 15060
    },
    {
      "entropy": 5.644379138946533,
      "epoch": 1.48922499011467,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047845668024321455,
      "loss": 5.3416,
      "mean_token_accuracy": 0.17729568034410476,
      "num_tokens": 34478085.0,
      "step": 15065
    },
    {
      "entropy": 5.585989332199096,
      "epoch": 1.4897192566231712,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004784416089397846,
      "loss": 5.4402,
      "mean_token_accuracy": 0.17838693410158157,
      "num_tokens": 34488437.0,
      "step": 15070
    },
    {
      "entropy": 5.661489820480346,
      "epoch": 1.4902135231316727,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004784265326316594,
      "loss": 5.3424,
      "mean_token_accuracy": 0.1784212201833725,
      "num_tokens": 34499999.0,
      "step": 15075
    },
    {
      "entropy": 5.649656057357788,
      "epoch": 1.490707789640174,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00047841145131921,
      "loss": 5.3981,
      "mean_token_accuracy": 0.18453755974769592,
      "num_tokens": 34510509.0,
      "step": 15080
    },
    {
      "entropy": 5.631809711456299,
      "epoch": 1.4912020561486754,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047839636500280724,
      "loss": 5.3683,
      "mean_token_accuracy": 0.18088580518960953,
      "num_tokens": 34521192.0,
      "step": 15085
    },
    {
      "entropy": 5.712775325775146,
      "epoch": 1.4916963226571767,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004783812736828225,
      "loss": 5.3719,
      "mean_token_accuracy": 0.18623314797878265,
      "num_tokens": 34532401.0,
      "step": 15090
    },
    {
      "entropy": 5.645899915695191,
      "epoch": 1.4921905891656781,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00047836617735962675,
      "loss": 5.3545,
      "mean_token_accuracy": 0.1802325502038002,
      "num_tokens": 34544306.0,
      "step": 15095
    },
    {
      "entropy": 5.545193767547607,
      "epoch": 1.4926848556741796,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004783510760335915,
      "loss": 5.3018,
      "mean_token_accuracy": 0.18025859892368318,
      "num_tokens": 34557122.0,
      "step": 15100
    },
    {
      "entropy": 5.589693403244018,
      "epoch": 1.493179122182681,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004783359697050883,
      "loss": 5.3085,
      "mean_token_accuracy": 0.18609167486429215,
      "num_tokens": 34568302.0,
      "step": 15105
    },
    {
      "entropy": 5.639269113540649,
      "epoch": 1.4936733886911822,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047832085837448864,
      "loss": 5.2383,
      "mean_token_accuracy": 0.19049194008111953,
      "num_tokens": 34578536.0,
      "step": 15110
    },
    {
      "entropy": 5.596101903915406,
      "epoch": 1.4941676551996836,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00047830574204216436,
      "loss": 5.3497,
      "mean_token_accuracy": 0.17946695685386657,
      "num_tokens": 34590333.0,
      "step": 15115
    },
    {
      "entropy": 5.646733188629151,
      "epoch": 1.4946619217081851,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004782906207084873,
      "loss": 5.3667,
      "mean_token_accuracy": 0.17681428641080857,
      "num_tokens": 34601447.0,
      "step": 15120
    },
    {
      "entropy": 5.644145345687866,
      "epoch": 1.4951561882166864,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004782754943738294,
      "loss": 5.3793,
      "mean_token_accuracy": 0.17728787213563918,
      "num_tokens": 34613559.0,
      "step": 15125
    },
    {
      "entropy": 5.623290920257569,
      "epoch": 1.4956504547251879,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004782603630385628,
      "loss": 5.3809,
      "mean_token_accuracy": 0.18217939585447313,
      "num_tokens": 34624354.0,
      "step": 15130
    },
    {
      "entropy": 5.635566759109497,
      "epoch": 1.4961447212336891,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004782452267030598,
      "loss": 5.4486,
      "mean_token_accuracy": 0.1808988034725189,
      "num_tokens": 34636093.0,
      "step": 15135
    },
    {
      "entropy": 5.705580520629883,
      "epoch": 1.4966389877421906,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004782300853676926,
      "loss": 5.2759,
      "mean_token_accuracy": 0.19083803296089172,
      "num_tokens": 34647503.0,
      "step": 15140
    },
    {
      "entropy": 5.601016902923584,
      "epoch": 1.4971332542506919,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047821493903283383,
      "loss": 5.4302,
      "mean_token_accuracy": 0.1819182351231575,
      "num_tokens": 34660258.0,
      "step": 15145
    },
    {
      "entropy": 5.619529151916504,
      "epoch": 1.4976275207591934,
      "grad_norm": 1.671875,
      "learning_rate": 0.000478199787698856,
      "loss": 5.3236,
      "mean_token_accuracy": 0.18092080652713777,
      "num_tokens": 34672199.0,
      "step": 15150
    },
    {
      "entropy": 5.632478189468384,
      "epoch": 1.4981217872676948,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047818463136613184,
      "loss": 5.3093,
      "mean_token_accuracy": 0.18756650537252426,
      "num_tokens": 34684419.0,
      "step": 15155
    },
    {
      "entropy": 5.436056900024414,
      "epoch": 1.498616053776196,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00047816947003503425,
      "loss": 5.1478,
      "mean_token_accuracy": 0.2000817134976387,
      "num_tokens": 34695649.0,
      "step": 15160
    },
    {
      "entropy": 5.639167928695679,
      "epoch": 1.4991103202846974,
      "grad_norm": 1.8203125,
      "learning_rate": 0.00047815430370593627,
      "loss": 5.3857,
      "mean_token_accuracy": 0.18156712800264357,
      "num_tokens": 34707019.0,
      "step": 15165
    },
    {
      "entropy": 5.627761268615723,
      "epoch": 1.4996045867931989,
      "grad_norm": 1.5,
      "learning_rate": 0.00047813913237921077,
      "loss": 5.296,
      "mean_token_accuracy": 0.19009427428245546,
      "num_tokens": 34717609.0,
      "step": 15170
    },
    {
      "entropy": 5.643815708160401,
      "epoch": 1.5000988533017003,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004781239560552312,
      "loss": 5.3758,
      "mean_token_accuracy": 0.18448631167411805,
      "num_tokens": 34728607.0,
      "step": 15175
    },
    {
      "entropy": 5.513981771469116,
      "epoch": 1.5005931198102016,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047810877473437077,
      "loss": 5.2648,
      "mean_token_accuracy": 0.18595848977565765,
      "num_tokens": 34739537.0,
      "step": 15180
    },
    {
      "entropy": 5.592280387878418,
      "epoch": 1.501087386318703,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047809358841700297,
      "loss": 5.3778,
      "mean_token_accuracy": 0.18015966415405274,
      "num_tokens": 34751814.0,
      "step": 15185
    },
    {
      "entropy": 5.681809139251709,
      "epoch": 1.5015816528272046,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004780783971035014,
      "loss": 5.298,
      "mean_token_accuracy": 0.18646501898765563,
      "num_tokens": 34762647.0,
      "step": 15190
    },
    {
      "entropy": 5.6360191822052,
      "epoch": 1.5020759193357058,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004780632007942397,
      "loss": 5.3899,
      "mean_token_accuracy": 0.17677464336156845,
      "num_tokens": 34774157.0,
      "step": 15195
    },
    {
      "entropy": 5.596209096908569,
      "epoch": 1.502570185844207,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004780479994895918,
      "loss": 5.2364,
      "mean_token_accuracy": 0.18311847001314163,
      "num_tokens": 34785244.0,
      "step": 15200
    },
    {
      "entropy": 5.628902387619019,
      "epoch": 1.5030644523527086,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004780327931899316,
      "loss": 5.3163,
      "mean_token_accuracy": 0.18666891753673553,
      "num_tokens": 34797090.0,
      "step": 15205
    },
    {
      "entropy": 5.581091642379761,
      "epoch": 1.50355871886121,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047801758189563315,
      "loss": 5.333,
      "mean_token_accuracy": 0.18319434970617293,
      "num_tokens": 34808598.0,
      "step": 15210
    },
    {
      "entropy": 5.613069725036621,
      "epoch": 1.5040529853697113,
      "grad_norm": 2.234375,
      "learning_rate": 0.00047800236560707076,
      "loss": 5.3919,
      "mean_token_accuracy": 0.1822893887758255,
      "num_tokens": 34820619.0,
      "step": 15215
    },
    {
      "entropy": 5.636681747436524,
      "epoch": 1.5045472518782126,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047798714432461864,
      "loss": 5.384,
      "mean_token_accuracy": 0.18434231877326965,
      "num_tokens": 34832463.0,
      "step": 15220
    },
    {
      "entropy": 5.657518005371093,
      "epoch": 1.505041518386714,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0004779719180486513,
      "loss": 5.3848,
      "mean_token_accuracy": 0.17879601269960405,
      "num_tokens": 34844255.0,
      "step": 15225
    },
    {
      "entropy": 5.565562343597412,
      "epoch": 1.5055357848952156,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004779566867795433,
      "loss": 5.3796,
      "mean_token_accuracy": 0.1832110434770584,
      "num_tokens": 34857215.0,
      "step": 15230
    },
    {
      "entropy": 5.569212818145752,
      "epoch": 1.5060300514037168,
      "grad_norm": 1.546875,
      "learning_rate": 0.00047794145051766935,
      "loss": 5.2977,
      "mean_token_accuracy": 0.18866186439990998,
      "num_tokens": 34868831.0,
      "step": 15235
    },
    {
      "entropy": 5.690900421142578,
      "epoch": 1.5065243179122183,
      "grad_norm": 1.25,
      "learning_rate": 0.0004779262092634042,
      "loss": 5.2405,
      "mean_token_accuracy": 0.1928907588124275,
      "num_tokens": 34879258.0,
      "step": 15240
    },
    {
      "entropy": 5.61680793762207,
      "epoch": 1.5070185844207198,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047791096301712276,
      "loss": 5.3827,
      "mean_token_accuracy": 0.17970270216464995,
      "num_tokens": 34891492.0,
      "step": 15245
    },
    {
      "entropy": 5.550572633743286,
      "epoch": 1.507512850929221,
      "grad_norm": 1.609375,
      "learning_rate": 0.00047789571177920024,
      "loss": 5.2608,
      "mean_token_accuracy": 0.1808638244867325,
      "num_tokens": 34901836.0,
      "step": 15250
    },
    {
      "entropy": 5.590566205978393,
      "epoch": 1.5080071174377223,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004778804555500117,
      "loss": 5.3904,
      "mean_token_accuracy": 0.18146324008703232,
      "num_tokens": 34913892.0,
      "step": 15255
    },
    {
      "entropy": 5.681615972518921,
      "epoch": 1.5085013839462238,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004778651943299324,
      "loss": 5.4219,
      "mean_token_accuracy": 0.17938367873430253,
      "num_tokens": 34925657.0,
      "step": 15260
    },
    {
      "entropy": 5.6409605026245115,
      "epoch": 1.5089956504547253,
      "grad_norm": 1.296875,
      "learning_rate": 0.000477849928119338,
      "loss": 5.3731,
      "mean_token_accuracy": 0.17458854913711547,
      "num_tokens": 34937921.0,
      "step": 15265
    },
    {
      "entropy": 5.601303052902222,
      "epoch": 1.5094899169632265,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004778346569186038,
      "loss": 5.3644,
      "mean_token_accuracy": 0.18017712980508804,
      "num_tokens": 34948742.0,
      "step": 15270
    },
    {
      "entropy": 5.589604139328003,
      "epoch": 1.5099841834717278,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047781938072810555,
      "loss": 5.3525,
      "mean_token_accuracy": 0.18522555977106095,
      "num_tokens": 34960912.0,
      "step": 15275
    },
    {
      "entropy": 5.65964641571045,
      "epoch": 1.5104784499802293,
      "grad_norm": 3.40625,
      "learning_rate": 0.00047780409954821914,
      "loss": 5.3997,
      "mean_token_accuracy": 0.17724638432264328,
      "num_tokens": 34971188.0,
      "step": 15280
    },
    {
      "entropy": 5.617594957351685,
      "epoch": 1.5109727164887308,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047778881337932037,
      "loss": 5.3253,
      "mean_token_accuracy": 0.17985613346099855,
      "num_tokens": 34983316.0,
      "step": 15285
    },
    {
      "entropy": 5.699150085449219,
      "epoch": 1.511466982997232,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00047777352222178537,
      "loss": 5.4325,
      "mean_token_accuracy": 0.1703278660774231,
      "num_tokens": 34995698.0,
      "step": 15290
    },
    {
      "entropy": 5.606994342803955,
      "epoch": 1.5119612495057335,
      "grad_norm": 1.40625,
      "learning_rate": 0.00047775822607599023,
      "loss": 5.3422,
      "mean_token_accuracy": 0.1865560382604599,
      "num_tokens": 35007296.0,
      "step": 15295
    },
    {
      "entropy": 5.688381290435791,
      "epoch": 1.512455516014235,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004777429249423113,
      "loss": 5.3922,
      "mean_token_accuracy": 0.1801751211285591,
      "num_tokens": 35019126.0,
      "step": 15300
    },
    {
      "entropy": 5.591096115112305,
      "epoch": 1.5129497825227363,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00047772761882112493,
      "loss": 5.304,
      "mean_token_accuracy": 0.1849311724305153,
      "num_tokens": 35030213.0,
      "step": 15305
    },
    {
      "entropy": 5.54937252998352,
      "epoch": 1.5134440490312375,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004777123077128077,
      "loss": 5.2888,
      "mean_token_accuracy": 0.1802014023065567,
      "num_tokens": 35043134.0,
      "step": 15310
    },
    {
      "entropy": 5.677251100540161,
      "epoch": 1.513938315539739,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004776969916177363,
      "loss": 5.3966,
      "mean_token_accuracy": 0.1778102621436119,
      "num_tokens": 35055484.0,
      "step": 15315
    },
    {
      "entropy": 5.667330169677735,
      "epoch": 1.5144325820482405,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004776816705362874,
      "loss": 5.3822,
      "mean_token_accuracy": 0.18275257647037507,
      "num_tokens": 35067037.0,
      "step": 15320
    },
    {
      "entropy": 5.573491811752319,
      "epoch": 1.5149268485567418,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004776663444688381,
      "loss": 5.3004,
      "mean_token_accuracy": 0.18107000291347503,
      "num_tokens": 35078340.0,
      "step": 15325
    },
    {
      "entropy": 5.604687261581421,
      "epoch": 1.515421115065243,
      "grad_norm": 1.375,
      "learning_rate": 0.0004776510134157651,
      "loss": 5.3984,
      "mean_token_accuracy": 0.1840140089392662,
      "num_tokens": 35090526.0,
      "step": 15330
    },
    {
      "entropy": 5.63254017829895,
      "epoch": 1.5159153815737447,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047763567737744576,
      "loss": 5.425,
      "mean_token_accuracy": 0.1753411650657654,
      "num_tokens": 35103757.0,
      "step": 15335
    },
    {
      "entropy": 5.624457502365113,
      "epoch": 1.516409648082246,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004776203363542574,
      "loss": 5.2841,
      "mean_token_accuracy": 0.18666779547929763,
      "num_tokens": 35115694.0,
      "step": 15340
    },
    {
      "entropy": 5.590429639816284,
      "epoch": 1.5169039145907472,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004776049903465773,
      "loss": 5.288,
      "mean_token_accuracy": 0.18901096284389496,
      "num_tokens": 35126691.0,
      "step": 15345
    },
    {
      "entropy": 5.584354209899902,
      "epoch": 1.5173981810992487,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000477589639354783,
      "loss": 5.3153,
      "mean_token_accuracy": 0.18446731865406035,
      "num_tokens": 35137851.0,
      "step": 15350
    },
    {
      "entropy": 5.603511905670166,
      "epoch": 1.5178924476077502,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004775742833792521,
      "loss": 5.342,
      "mean_token_accuracy": 0.18033986687660217,
      "num_tokens": 35148583.0,
      "step": 15355
    },
    {
      "entropy": 5.627356958389282,
      "epoch": 1.5183867141162515,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004775589224203625,
      "loss": 5.3673,
      "mean_token_accuracy": 0.17984215170145035,
      "num_tokens": 35161469.0,
      "step": 15360
    },
    {
      "entropy": 5.60552659034729,
      "epoch": 1.5188809806247527,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047754355647849185,
      "loss": 5.2445,
      "mean_token_accuracy": 0.19505372643470764,
      "num_tokens": 35172349.0,
      "step": 15365
    },
    {
      "entropy": 5.565273380279541,
      "epoch": 1.5193752471332542,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047752818555401845,
      "loss": 5.2737,
      "mean_token_accuracy": 0.18398473858833314,
      "num_tokens": 35183118.0,
      "step": 15370
    },
    {
      "entropy": 5.685931777954101,
      "epoch": 1.5198695136417557,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047751280964732014,
      "loss": 5.329,
      "mean_token_accuracy": 0.18145417124032975,
      "num_tokens": 35195702.0,
      "step": 15375
    },
    {
      "entropy": 5.727619886398315,
      "epoch": 1.520363780150257,
      "grad_norm": 1.375,
      "learning_rate": 0.00047749742875877535,
      "loss": 5.4801,
      "mean_token_accuracy": 0.17211327850818633,
      "num_tokens": 35207356.0,
      "step": 15380
    },
    {
      "entropy": 5.616939640045166,
      "epoch": 1.5208580466587585,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004774820428887624,
      "loss": 5.3531,
      "mean_token_accuracy": 0.1812470391392708,
      "num_tokens": 35219207.0,
      "step": 15385
    },
    {
      "entropy": 5.580738639831543,
      "epoch": 1.52135231316726,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004774666520376597,
      "loss": 5.3034,
      "mean_token_accuracy": 0.18335646986961365,
      "num_tokens": 35230396.0,
      "step": 15390
    },
    {
      "entropy": 5.595555925369263,
      "epoch": 1.5218465796757612,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004774512562058461,
      "loss": 5.3771,
      "mean_token_accuracy": 0.18237807005643844,
      "num_tokens": 35242320.0,
      "step": 15395
    },
    {
      "entropy": 5.710798025131226,
      "epoch": 1.5223408461842625,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004774358553937001,
      "loss": 5.4071,
      "mean_token_accuracy": 0.17716021537780763,
      "num_tokens": 35252926.0,
      "step": 15400
    },
    {
      "entropy": 5.7017210006713865,
      "epoch": 1.522835112692764,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004774204496016006,
      "loss": 5.4309,
      "mean_token_accuracy": 0.17283237874507903,
      "num_tokens": 35263833.0,
      "step": 15405
    },
    {
      "entropy": 5.605069303512574,
      "epoch": 1.5233293792012654,
      "grad_norm": 1.578125,
      "learning_rate": 0.00047740503882992677,
      "loss": 5.3906,
      "mean_token_accuracy": 0.1789207637310028,
      "num_tokens": 35274951.0,
      "step": 15410
    },
    {
      "entropy": 5.672837018966675,
      "epoch": 1.5238236457097667,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004773896230790575,
      "loss": 5.4184,
      "mean_token_accuracy": 0.1774936020374298,
      "num_tokens": 35287940.0,
      "step": 15415
    },
    {
      "entropy": 5.564551210403442,
      "epoch": 1.524317912218268,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047737420234937215,
      "loss": 5.3367,
      "mean_token_accuracy": 0.18867705762386322,
      "num_tokens": 35299321.0,
      "step": 15420
    },
    {
      "entropy": 5.593753147125244,
      "epoch": 1.5248121787267694,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047735877664125,
      "loss": 5.3143,
      "mean_token_accuracy": 0.1857372283935547,
      "num_tokens": 35310181.0,
      "step": 15425
    },
    {
      "entropy": 5.668422603607178,
      "epoch": 1.525306445235271,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00047734334595507045,
      "loss": 5.3434,
      "mean_token_accuracy": 0.18003177791833877,
      "num_tokens": 35320435.0,
      "step": 15430
    },
    {
      "entropy": 5.6462079048156735,
      "epoch": 1.5258007117437722,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004773279102912134,
      "loss": 5.4235,
      "mean_token_accuracy": 0.1774923861026764,
      "num_tokens": 35332533.0,
      "step": 15435
    },
    {
      "entropy": 5.571061515808106,
      "epoch": 1.5262949782522737,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004773124696500582,
      "loss": 5.307,
      "mean_token_accuracy": 0.1932358369231224,
      "num_tokens": 35344289.0,
      "step": 15440
    },
    {
      "entropy": 5.596386289596557,
      "epoch": 1.5267892447607752,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004772970240319849,
      "loss": 5.2816,
      "mean_token_accuracy": 0.191512893140316,
      "num_tokens": 35355679.0,
      "step": 15445
    },
    {
      "entropy": 5.571349477767944,
      "epoch": 1.5272835112692764,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047728157343737346,
      "loss": 5.3003,
      "mean_token_accuracy": 0.18744294792413713,
      "num_tokens": 35367489.0,
      "step": 15450
    },
    {
      "entropy": 5.565728282928466,
      "epoch": 1.5277777777777777,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047726611786660396,
      "loss": 5.2218,
      "mean_token_accuracy": 0.18942863643169403,
      "num_tokens": 35378349.0,
      "step": 15455
    },
    {
      "entropy": 5.6377938747406,
      "epoch": 1.5282720442862792,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047725065732005654,
      "loss": 5.3806,
      "mean_token_accuracy": 0.18526216149330138,
      "num_tokens": 35388927.0,
      "step": 15460
    },
    {
      "entropy": 5.587651300430298,
      "epoch": 1.5287663107947806,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004772351917981115,
      "loss": 5.3059,
      "mean_token_accuracy": 0.1870112270116806,
      "num_tokens": 35401859.0,
      "step": 15465
    },
    {
      "entropy": 5.659857892990113,
      "epoch": 1.529260577303282,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004772197213011494,
      "loss": 5.3437,
      "mean_token_accuracy": 0.18049770146608352,
      "num_tokens": 35412735.0,
      "step": 15470
    },
    {
      "entropy": 5.573085069656372,
      "epoch": 1.5297548438117832,
      "grad_norm": 2.234375,
      "learning_rate": 0.0004772042458295509,
      "loss": 5.3084,
      "mean_token_accuracy": 0.18320072442293167,
      "num_tokens": 35422960.0,
      "step": 15475
    },
    {
      "entropy": 5.676775074005127,
      "epoch": 1.5302491103202847,
      "grad_norm": 1.703125,
      "learning_rate": 0.00047718876538369655,
      "loss": 5.4291,
      "mean_token_accuracy": 0.1745563790202141,
      "num_tokens": 35435054.0,
      "step": 15480
    },
    {
      "entropy": 5.651961374282837,
      "epoch": 1.5307433768287861,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00047717327996396716,
      "loss": 5.3786,
      "mean_token_accuracy": 0.1803923726081848,
      "num_tokens": 35447388.0,
      "step": 15485
    },
    {
      "entropy": 5.688505458831787,
      "epoch": 1.5312376433372874,
      "grad_norm": 2.28125,
      "learning_rate": 0.0004771577895707437,
      "loss": 5.3644,
      "mean_token_accuracy": 0.18185846358537675,
      "num_tokens": 35459181.0,
      "step": 15490
    },
    {
      "entropy": 5.530900859832764,
      "epoch": 1.5317319098457889,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004771422942044073,
      "loss": 5.2051,
      "mean_token_accuracy": 0.18724495470523833,
      "num_tokens": 35469978.0,
      "step": 15495
    },
    {
      "entropy": 5.540166616439819,
      "epoch": 1.5322261763542904,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004771267938653391,
      "loss": 5.3127,
      "mean_token_accuracy": 0.1894787386059761,
      "num_tokens": 35482196.0,
      "step": 15500
    },
    {
      "entropy": 5.575621795654297,
      "epoch": 1.5327204428627916,
      "grad_norm": 1.5,
      "learning_rate": 0.0004771112885539204,
      "loss": 5.2472,
      "mean_token_accuracy": 0.1961321771144867,
      "num_tokens": 35493025.0,
      "step": 15505
    },
    {
      "entropy": 5.54494423866272,
      "epoch": 1.533214709371293,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047709577827053257,
      "loss": 5.3411,
      "mean_token_accuracy": 0.18795692324638366,
      "num_tokens": 35504379.0,
      "step": 15510
    },
    {
      "entropy": 5.630211114883423,
      "epoch": 1.5337089758797944,
      "grad_norm": 1.65625,
      "learning_rate": 0.00047708026301555736,
      "loss": 5.2935,
      "mean_token_accuracy": 0.1893584430217743,
      "num_tokens": 35515401.0,
      "step": 15515
    },
    {
      "entropy": 5.6996245861053465,
      "epoch": 1.5342032423882959,
      "grad_norm": 1.25,
      "learning_rate": 0.00047706474278937626,
      "loss": 5.3724,
      "mean_token_accuracy": 0.18451106250286103,
      "num_tokens": 35527625.0,
      "step": 15520
    },
    {
      "entropy": 5.608449029922485,
      "epoch": 1.5346975088967971,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047704921759237116,
      "loss": 5.3286,
      "mean_token_accuracy": 0.181321482360363,
      "num_tokens": 35539484.0,
      "step": 15525
    },
    {
      "entropy": 5.678898477554322,
      "epoch": 1.5351917754052984,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00047703368742492403,
      "loss": 5.3431,
      "mean_token_accuracy": 0.17915442436933518,
      "num_tokens": 35550362.0,
      "step": 15530
    },
    {
      "entropy": 5.677475261688232,
      "epoch": 1.5356860419137999,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004770181522874167,
      "loss": 5.3221,
      "mean_token_accuracy": 0.18542334884405137,
      "num_tokens": 35561564.0,
      "step": 15535
    },
    {
      "entropy": 5.6013877391815186,
      "epoch": 1.5361803084223014,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004770026121802316,
      "loss": 5.2701,
      "mean_token_accuracy": 0.18590582013130189,
      "num_tokens": 35572216.0,
      "step": 15540
    },
    {
      "entropy": 5.5432319164276125,
      "epoch": 1.5366745749308026,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004769870671037509,
      "loss": 5.2353,
      "mean_token_accuracy": 0.19741733223199845,
      "num_tokens": 35584471.0,
      "step": 15545
    },
    {
      "entropy": 5.587041139602661,
      "epoch": 1.537168841439304,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047697151705835693,
      "loss": 5.3102,
      "mean_token_accuracy": 0.1844884932041168,
      "num_tokens": 35596551.0,
      "step": 15550
    },
    {
      "entropy": 5.627885103225708,
      "epoch": 1.5376631079478056,
      "grad_norm": 1.375,
      "learning_rate": 0.0004769559620444324,
      "loss": 5.3237,
      "mean_token_accuracy": 0.1878990888595581,
      "num_tokens": 35607425.0,
      "step": 15555
    },
    {
      "entropy": 5.5082183361053465,
      "epoch": 1.5381573744563068,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047694040206235983,
      "loss": 5.2775,
      "mean_token_accuracy": 0.1897694483399391,
      "num_tokens": 35618900.0,
      "step": 15560
    },
    {
      "entropy": 5.622366571426392,
      "epoch": 1.538651640964808,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000476924837112522,
      "loss": 5.2953,
      "mean_token_accuracy": 0.18988023847341537,
      "num_tokens": 35630126.0,
      "step": 15565
    },
    {
      "entropy": 5.60708441734314,
      "epoch": 1.5391459074733096,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00047690926719530193,
      "loss": 5.3599,
      "mean_token_accuracy": 0.18527670949697495,
      "num_tokens": 35642406.0,
      "step": 15570
    },
    {
      "entropy": 5.5634284019470215,
      "epoch": 1.539640173981811,
      "grad_norm": 1.9765625,
      "learning_rate": 0.00047689369231108253,
      "loss": 5.357,
      "mean_token_accuracy": 0.18218495100736617,
      "num_tokens": 35654701.0,
      "step": 15575
    },
    {
      "entropy": 5.7010178565979,
      "epoch": 1.5401344404903123,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000476878112460247,
      "loss": 5.4242,
      "mean_token_accuracy": 0.17453216910362243,
      "num_tokens": 35665668.0,
      "step": 15580
    },
    {
      "entropy": 5.609474086761475,
      "epoch": 1.5406287069988136,
      "grad_norm": 1.65625,
      "learning_rate": 0.00047686252764317854,
      "loss": 5.3296,
      "mean_token_accuracy": 0.18516783863306047,
      "num_tokens": 35676940.0,
      "step": 15585
    },
    {
      "entropy": 5.6635418891906735,
      "epoch": 1.5411229735073153,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047684693786026054,
      "loss": 5.4186,
      "mean_token_accuracy": 0.17526114284992217,
      "num_tokens": 35687860.0,
      "step": 15590
    },
    {
      "entropy": 5.582547950744629,
      "epoch": 1.5416172400158166,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047683134311187663,
      "loss": 5.2703,
      "mean_token_accuracy": 0.1913952797651291,
      "num_tokens": 35698349.0,
      "step": 15595
    },
    {
      "entropy": 5.566019010543823,
      "epoch": 1.5421115065243178,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047681574339841044,
      "loss": 5.3376,
      "mean_token_accuracy": 0.18827067464590072,
      "num_tokens": 35709649.0,
      "step": 15600
    },
    {
      "entropy": 5.566201066970825,
      "epoch": 1.5426057730328193,
      "grad_norm": 1.453125,
      "learning_rate": 0.00047680013872024546,
      "loss": 5.2468,
      "mean_token_accuracy": 0.18854983896017075,
      "num_tokens": 35720806.0,
      "step": 15605
    },
    {
      "entropy": 5.624632406234741,
      "epoch": 1.5431000395413208,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004767845290777659,
      "loss": 5.2942,
      "mean_token_accuracy": 0.1844210222363472,
      "num_tokens": 35732044.0,
      "step": 15610
    },
    {
      "entropy": 5.577399349212646,
      "epoch": 1.543594306049822,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004767689144713556,
      "loss": 5.3506,
      "mean_token_accuracy": 0.18211054801940918,
      "num_tokens": 35743758.0,
      "step": 15615
    },
    {
      "entropy": 5.645985889434814,
      "epoch": 1.5440885725583233,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047675329490139866,
      "loss": 5.2863,
      "mean_token_accuracy": 0.1800716131925583,
      "num_tokens": 35756184.0,
      "step": 15620
    },
    {
      "entropy": 5.590408897399902,
      "epoch": 1.5445828390668248,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00047673767036827947,
      "loss": 5.3153,
      "mean_token_accuracy": 0.18041884899139404,
      "num_tokens": 35768255.0,
      "step": 15625
    },
    {
      "entropy": 5.592480325698853,
      "epoch": 1.5450771055753263,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004767220408723822,
      "loss": 5.3666,
      "mean_token_accuracy": 0.17498636841773987,
      "num_tokens": 35779111.0,
      "step": 15630
    },
    {
      "entropy": 5.657261562347412,
      "epoch": 1.5455713720838276,
      "grad_norm": 2.671875,
      "learning_rate": 0.0004767064064140915,
      "loss": 5.3822,
      "mean_token_accuracy": 0.18287851214408873,
      "num_tokens": 35791676.0,
      "step": 15635
    },
    {
      "entropy": 5.601535701751709,
      "epoch": 1.546065638592329,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004766907669937919,
      "loss": 5.2741,
      "mean_token_accuracy": 0.19282286912202834,
      "num_tokens": 35801868.0,
      "step": 15640
    },
    {
      "entropy": 5.667299222946167,
      "epoch": 1.5465599051008305,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000476675122611868,
      "loss": 5.4693,
      "mean_token_accuracy": 0.17500021457672119,
      "num_tokens": 35812428.0,
      "step": 15645
    },
    {
      "entropy": 5.625846576690674,
      "epoch": 1.5470541716093318,
      "grad_norm": 1.484375,
      "learning_rate": 0.000476659473268705,
      "loss": 5.3086,
      "mean_token_accuracy": 0.1865147680044174,
      "num_tokens": 35823600.0,
      "step": 15650
    },
    {
      "entropy": 5.625600957870484,
      "epoch": 1.547548438117833,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004766438189646876,
      "loss": 5.34,
      "mean_token_accuracy": 0.18703303933143617,
      "num_tokens": 35834988.0,
      "step": 15655
    },
    {
      "entropy": 5.617691516876221,
      "epoch": 1.5480427046263345,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00047662815970020086,
      "loss": 5.3453,
      "mean_token_accuracy": 0.1862446442246437,
      "num_tokens": 35846110.0,
      "step": 15660
    },
    {
      "entropy": 5.529103136062622,
      "epoch": 1.548536971134836,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004766124954756302,
      "loss": 5.3164,
      "mean_token_accuracy": 0.18625538796186447,
      "num_tokens": 35858276.0,
      "step": 15665
    },
    {
      "entropy": 5.711389350891113,
      "epoch": 1.5490312376433373,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047659682629136094,
      "loss": 5.3977,
      "mean_token_accuracy": 0.1807095542550087,
      "num_tokens": 35870705.0,
      "step": 15670
    },
    {
      "entropy": 5.649735975265503,
      "epoch": 1.5495255041518385,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004765811521477784,
      "loss": 5.3286,
      "mean_token_accuracy": 0.18542488515377045,
      "num_tokens": 35882298.0,
      "step": 15675
    },
    {
      "entropy": 5.538140392303466,
      "epoch": 1.55001977066034,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004765654730452682,
      "loss": 5.2647,
      "mean_token_accuracy": 0.19012416154146194,
      "num_tokens": 35894665.0,
      "step": 15680
    },
    {
      "entropy": 5.627080726623535,
      "epoch": 1.5505140371688415,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004765497889842161,
      "loss": 5.3343,
      "mean_token_accuracy": 0.18909312933683395,
      "num_tokens": 35905893.0,
      "step": 15685
    },
    {
      "entropy": 5.624281024932861,
      "epoch": 1.5510083036773428,
      "grad_norm": 2.359375,
      "learning_rate": 0.00047653409996500803,
      "loss": 5.2838,
      "mean_token_accuracy": 0.1896094098687172,
      "num_tokens": 35917883.0,
      "step": 15690
    },
    {
      "entropy": 5.620629692077637,
      "epoch": 1.5515025701858443,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047651840598802976,
      "loss": 5.3608,
      "mean_token_accuracy": 0.18712644875049592,
      "num_tokens": 35929863.0,
      "step": 15695
    },
    {
      "entropy": 5.6553479671478275,
      "epoch": 1.5519968366943457,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004765027070536674,
      "loss": 5.3804,
      "mean_token_accuracy": 0.18213234543800355,
      "num_tokens": 35940843.0,
      "step": 15700
    },
    {
      "entropy": 5.579813432693482,
      "epoch": 1.552491103202847,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004764870031623073,
      "loss": 5.2457,
      "mean_token_accuracy": 0.18896625339984893,
      "num_tokens": 35953516.0,
      "step": 15705
    },
    {
      "entropy": 5.545122337341309,
      "epoch": 1.5529853697113483,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00047647129431433566,
      "loss": 5.3128,
      "mean_token_accuracy": 0.18391638547182082,
      "num_tokens": 35965481.0,
      "step": 15710
    },
    {
      "entropy": 5.627437734603882,
      "epoch": 1.5534796362198497,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004764555805101388,
      "loss": 5.3941,
      "mean_token_accuracy": 0.18486918061971663,
      "num_tokens": 35976434.0,
      "step": 15715
    },
    {
      "entropy": 5.619044017791748,
      "epoch": 1.5539739027283512,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004764398617501035,
      "loss": 5.28,
      "mean_token_accuracy": 0.18788978457450867,
      "num_tokens": 35987644.0,
      "step": 15720
    },
    {
      "entropy": 5.628942060470581,
      "epoch": 1.5544681692368525,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004764241380346163,
      "loss": 5.3736,
      "mean_token_accuracy": 0.18167052417993546,
      "num_tokens": 35999630.0,
      "step": 15725
    },
    {
      "entropy": 5.698288631439209,
      "epoch": 1.5549624357453538,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004764084093640641,
      "loss": 5.4317,
      "mean_token_accuracy": 0.18579235970973967,
      "num_tokens": 36011889.0,
      "step": 15730
    },
    {
      "entropy": 5.65256633758545,
      "epoch": 1.5554567022538552,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004763926757388338,
      "loss": 5.4058,
      "mean_token_accuracy": 0.17793132364749908,
      "num_tokens": 36024131.0,
      "step": 15735
    },
    {
      "entropy": 5.6088632106781,
      "epoch": 1.5559509687623567,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004763769371593124,
      "loss": 5.3134,
      "mean_token_accuracy": 0.19488660544157027,
      "num_tokens": 36035759.0,
      "step": 15740
    },
    {
      "entropy": 5.632501840591431,
      "epoch": 1.556445235270858,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004763611936258871,
      "loss": 5.3827,
      "mean_token_accuracy": 0.1786652162671089,
      "num_tokens": 36047882.0,
      "step": 15745
    },
    {
      "entropy": 5.5827916145324705,
      "epoch": 1.5569395017793595,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004763454451389452,
      "loss": 5.3428,
      "mean_token_accuracy": 0.181227470934391,
      "num_tokens": 36058268.0,
      "step": 15750
    },
    {
      "entropy": 5.680436992645264,
      "epoch": 1.557433768287861,
      "grad_norm": 1.4375,
      "learning_rate": 0.00047632969169887407,
      "loss": 5.3659,
      "mean_token_accuracy": 0.18153930008411406,
      "num_tokens": 36070797.0,
      "step": 15755
    },
    {
      "entropy": 5.607543659210205,
      "epoch": 1.5579280347963622,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004763139333060613,
      "loss": 5.3827,
      "mean_token_accuracy": 0.1859066367149353,
      "num_tokens": 36082858.0,
      "step": 15760
    },
    {
      "entropy": 5.621573448181152,
      "epoch": 1.5584223013048635,
      "grad_norm": 1.5625,
      "learning_rate": 0.00047629816996089464,
      "loss": 5.3434,
      "mean_token_accuracy": 0.1824665665626526,
      "num_tokens": 36094825.0,
      "step": 15765
    },
    {
      "entropy": 5.663739156723023,
      "epoch": 1.558916567813365,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047628240166376173,
      "loss": 5.3562,
      "mean_token_accuracy": 0.1769666850566864,
      "num_tokens": 36106379.0,
      "step": 15770
    },
    {
      "entropy": 5.542728328704834,
      "epoch": 1.5594108343218664,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047626662841505045,
      "loss": 5.2548,
      "mean_token_accuracy": 0.18893215656280518,
      "num_tokens": 36117137.0,
      "step": 15775
    },
    {
      "entropy": 5.597761726379394,
      "epoch": 1.5599051008303677,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000476250850215149,
      "loss": 5.3735,
      "mean_token_accuracy": 0.1764802649617195,
      "num_tokens": 36128420.0,
      "step": 15780
    },
    {
      "entropy": 5.660768747329712,
      "epoch": 1.560399367338869,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004762350670644453,
      "loss": 5.294,
      "mean_token_accuracy": 0.18913516253232956,
      "num_tokens": 36139017.0,
      "step": 15785
    },
    {
      "entropy": 5.613294553756714,
      "epoch": 1.5608936338473705,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004762192789633278,
      "loss": 5.3995,
      "mean_token_accuracy": 0.1834932953119278,
      "num_tokens": 36151748.0,
      "step": 15790
    },
    {
      "entropy": 5.630767488479615,
      "epoch": 1.561387900355872,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047620348591218475,
      "loss": 5.3672,
      "mean_token_accuracy": 0.18219092190265657,
      "num_tokens": 36163114.0,
      "step": 15795
    },
    {
      "entropy": 5.632153034210205,
      "epoch": 1.5618821668643732,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004761876879114048,
      "loss": 5.3777,
      "mean_token_accuracy": 0.18503647148609162,
      "num_tokens": 36174601.0,
      "step": 15800
    },
    {
      "entropy": 5.632711267471313,
      "epoch": 1.5623764333728747,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004761718849613765,
      "loss": 5.305,
      "mean_token_accuracy": 0.187123903632164,
      "num_tokens": 36185081.0,
      "step": 15805
    },
    {
      "entropy": 5.618956899642944,
      "epoch": 1.5628706998813762,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004761560770624886,
      "loss": 5.3568,
      "mean_token_accuracy": 0.18568941950798035,
      "num_tokens": 36196547.0,
      "step": 15810
    },
    {
      "entropy": 5.615840435028076,
      "epoch": 1.5633649663898774,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004761402642151301,
      "loss": 5.2683,
      "mean_token_accuracy": 0.19079804569482803,
      "num_tokens": 36208115.0,
      "step": 15815
    },
    {
      "entropy": 5.630123519897461,
      "epoch": 1.5638592328983787,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004761244464196897,
      "loss": 5.33,
      "mean_token_accuracy": 0.1871498331427574,
      "num_tokens": 36219571.0,
      "step": 15820
    },
    {
      "entropy": 5.619361352920532,
      "epoch": 1.5643534994068802,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004761086236765569,
      "loss": 5.3559,
      "mean_token_accuracy": 0.18395351320505143,
      "num_tokens": 36232219.0,
      "step": 15825
    },
    {
      "entropy": 5.60046420097351,
      "epoch": 1.5648477659153817,
      "grad_norm": 1.5,
      "learning_rate": 0.0004760927959861206,
      "loss": 5.3412,
      "mean_token_accuracy": 0.18310203701257705,
      "num_tokens": 36243761.0,
      "step": 15830
    },
    {
      "entropy": 5.575168752670288,
      "epoch": 1.565342032423883,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004760769633487704,
      "loss": 5.3183,
      "mean_token_accuracy": 0.18504491746425628,
      "num_tokens": 36255466.0,
      "step": 15835
    },
    {
      "entropy": 5.591471767425537,
      "epoch": 1.5658362989323842,
      "grad_norm": 1.5,
      "learning_rate": 0.00047606112576489574,
      "loss": 5.3694,
      "mean_token_accuracy": 0.17849458903074264,
      "num_tokens": 36267252.0,
      "step": 15840
    },
    {
      "entropy": 5.676768827438354,
      "epoch": 1.566330565440886,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004760452832348862,
      "loss": 5.3434,
      "mean_token_accuracy": 0.19355549216270446,
      "num_tokens": 36278149.0,
      "step": 15845
    },
    {
      "entropy": 5.784940767288208,
      "epoch": 1.5668248319493872,
      "grad_norm": 1.6875,
      "learning_rate": 0.00047602943575913146,
      "loss": 5.3878,
      "mean_token_accuracy": 0.1894302859902382,
      "num_tokens": 36290583.0,
      "step": 15850
    },
    {
      "entropy": 5.626826572418213,
      "epoch": 1.5673190984578884,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004760135833380214,
      "loss": 5.3527,
      "mean_token_accuracy": 0.17807947248220443,
      "num_tokens": 36301508.0,
      "step": 15855
    },
    {
      "entropy": 5.574789237976074,
      "epoch": 1.56781336496639,
      "grad_norm": 1.421875,
      "learning_rate": 0.000475997725971946,
      "loss": 5.2702,
      "mean_token_accuracy": 0.19079479277133943,
      "num_tokens": 36312488.0,
      "step": 15860
    },
    {
      "entropy": 5.63427586555481,
      "epoch": 1.5683076314748914,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004759818636612954,
      "loss": 5.2799,
      "mean_token_accuracy": 0.18527939766645432,
      "num_tokens": 36323263.0,
      "step": 15865
    },
    {
      "entropy": 5.601292610168457,
      "epoch": 1.5688018979833926,
      "grad_norm": 1.671875,
      "learning_rate": 0.00047596599640645974,
      "loss": 5.3287,
      "mean_token_accuracy": 0.17838988304138184,
      "num_tokens": 36336095.0,
      "step": 15870
    },
    {
      "entropy": 5.574136352539062,
      "epoch": 1.569296164491894,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00047595012420782946,
      "loss": 5.3711,
      "mean_token_accuracy": 0.18640863448381423,
      "num_tokens": 36346653.0,
      "step": 15875
    },
    {
      "entropy": 5.610943984985352,
      "epoch": 1.5697904310003954,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004759342470657949,
      "loss": 5.3091,
      "mean_token_accuracy": 0.18667134642601013,
      "num_tokens": 36358190.0,
      "step": 15880
    },
    {
      "entropy": 5.63078088760376,
      "epoch": 1.5702846975088969,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004759183649807467,
      "loss": 5.3503,
      "mean_token_accuracy": 0.1857827052474022,
      "num_tokens": 36369037.0,
      "step": 15885
    },
    {
      "entropy": 5.651617813110351,
      "epoch": 1.5707789640173981,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004759024779530755,
      "loss": 5.3171,
      "mean_token_accuracy": 0.18673258870840073,
      "num_tokens": 36379881.0,
      "step": 15890
    },
    {
      "entropy": 5.641597414016724,
      "epoch": 1.5712732305258994,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047588658598317227,
      "loss": 5.2663,
      "mean_token_accuracy": 0.19190889745950698,
      "num_tokens": 36390927.0,
      "step": 15895
    },
    {
      "entropy": 5.6638843536376955,
      "epoch": 1.571767497034401,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047587068907142783,
      "loss": 5.4153,
      "mean_token_accuracy": 0.18489917516708373,
      "num_tokens": 36402398.0,
      "step": 15900
    },
    {
      "entropy": 5.55378794670105,
      "epoch": 1.5722617635429024,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004758547872182332,
      "loss": 5.2635,
      "mean_token_accuracy": 0.18673126250505448,
      "num_tokens": 36413991.0,
      "step": 15905
    },
    {
      "entropy": 5.541579818725586,
      "epoch": 1.5727560300514036,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004758388804239797,
      "loss": 5.3296,
      "mean_token_accuracy": 0.1853797420859337,
      "num_tokens": 36425807.0,
      "step": 15910
    },
    {
      "entropy": 5.701465177536011,
      "epoch": 1.5732502965599051,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004758229686890586,
      "loss": 5.3825,
      "mean_token_accuracy": 0.18051216751337051,
      "num_tokens": 36437968.0,
      "step": 15915
    },
    {
      "entropy": 5.669748830795288,
      "epoch": 1.5737445630684066,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047580705201386134,
      "loss": 5.3541,
      "mean_token_accuracy": 0.17972855120897294,
      "num_tokens": 36449803.0,
      "step": 15920
    },
    {
      "entropy": 5.6376667499542235,
      "epoch": 1.5742388295769079,
      "grad_norm": 1.5,
      "learning_rate": 0.0004757911303987794,
      "loss": 5.427,
      "mean_token_accuracy": 0.17511660605669022,
      "num_tokens": 36461849.0,
      "step": 15925
    },
    {
      "entropy": 5.622438859939575,
      "epoch": 1.5747330960854091,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004757752038442045,
      "loss": 5.2888,
      "mean_token_accuracy": 0.1888489231467247,
      "num_tokens": 36473324.0,
      "step": 15930
    },
    {
      "entropy": 5.675995826721191,
      "epoch": 1.5752273625939106,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004757592723505284,
      "loss": 5.3378,
      "mean_token_accuracy": 0.1818197563290596,
      "num_tokens": 36483949.0,
      "step": 15935
    },
    {
      "entropy": 5.529774618148804,
      "epoch": 1.575721629102412,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004757433359181431,
      "loss": 5.2579,
      "mean_token_accuracy": 0.1928209513425827,
      "num_tokens": 36495308.0,
      "step": 15940
    },
    {
      "entropy": 5.525200986862183,
      "epoch": 1.5762158956109134,
      "grad_norm": 1.625,
      "learning_rate": 0.00047572739454744057,
      "loss": 5.3082,
      "mean_token_accuracy": 0.1916145220398903,
      "num_tokens": 36506052.0,
      "step": 15945
    },
    {
      "entropy": 5.522533226013183,
      "epoch": 1.5767101621194148,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000475711448238813,
      "loss": 5.2459,
      "mean_token_accuracy": 0.19694708287715912,
      "num_tokens": 36519175.0,
      "step": 15950
    },
    {
      "entropy": 5.659843587875367,
      "epoch": 1.5772044286279163,
      "grad_norm": 1.4375,
      "learning_rate": 0.00047569549699265275,
      "loss": 5.3167,
      "mean_token_accuracy": 0.1867457538843155,
      "num_tokens": 36529736.0,
      "step": 15955
    },
    {
      "entropy": 5.558991813659668,
      "epoch": 1.5776986951364176,
      "grad_norm": 1.546875,
      "learning_rate": 0.00047567954080935206,
      "loss": 5.2563,
      "mean_token_accuracy": 0.1849062830209732,
      "num_tokens": 36541715.0,
      "step": 15960
    },
    {
      "entropy": 5.637162828445435,
      "epoch": 1.5781929616449188,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004756635796893036,
      "loss": 5.3477,
      "mean_token_accuracy": 0.1868065819144249,
      "num_tokens": 36553807.0,
      "step": 15965
    },
    {
      "entropy": 5.679484033584595,
      "epoch": 1.5786872281534203,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004756476136328998,
      "loss": 5.3842,
      "mean_token_accuracy": 0.18059302419424056,
      "num_tokens": 36566053.0,
      "step": 15970
    },
    {
      "entropy": 5.6261743068695065,
      "epoch": 1.5791814946619218,
      "grad_norm": 1.4375,
      "learning_rate": 0.00047563164264053374,
      "loss": 5.378,
      "mean_token_accuracy": 0.1833268404006958,
      "num_tokens": 36577637.0,
      "step": 15975
    },
    {
      "entropy": 5.624777889251709,
      "epoch": 1.579675761170423,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047561566671259803,
      "loss": 5.3129,
      "mean_token_accuracy": 0.18292995393276215,
      "num_tokens": 36587767.0,
      "step": 15980
    },
    {
      "entropy": 5.534054374694824,
      "epoch": 1.5801700276789243,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047559968584948587,
      "loss": 5.2268,
      "mean_token_accuracy": 0.19400970488786698,
      "num_tokens": 36599163.0,
      "step": 15985
    },
    {
      "entropy": 5.61691107749939,
      "epoch": 1.5806642941874258,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004755837000515903,
      "loss": 5.3975,
      "mean_token_accuracy": 0.17972329407930374,
      "num_tokens": 36611170.0,
      "step": 15990
    },
    {
      "entropy": 5.720366668701172,
      "epoch": 1.5811585606959273,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004755677093193046,
      "loss": 5.4726,
      "mean_token_accuracy": 0.17461608797311784,
      "num_tokens": 36622907.0,
      "step": 15995
    },
    {
      "entropy": 5.636205244064331,
      "epoch": 1.5816528272044286,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004755517136530221,
      "loss": 5.3581,
      "mean_token_accuracy": 0.18401266932487487,
      "num_tokens": 36635197.0,
      "step": 16000
    },
    {
      "entropy": 5.6060850620269775,
      "epoch": 1.58214709371293,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047553571305313636,
      "loss": 5.2775,
      "mean_token_accuracy": 0.18767874985933303,
      "num_tokens": 36646965.0,
      "step": 16005
    },
    {
      "entropy": 5.530509090423584,
      "epoch": 1.5826413602214315,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047551970752004097,
      "loss": 5.2665,
      "mean_token_accuracy": 0.20047152787446976,
      "num_tokens": 36657344.0,
      "step": 16010
    },
    {
      "entropy": 5.524829912185669,
      "epoch": 1.5831356267299328,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047550369705412964,
      "loss": 5.2687,
      "mean_token_accuracy": 0.19154494553804396,
      "num_tokens": 36668833.0,
      "step": 16015
    },
    {
      "entropy": 5.599942874908447,
      "epoch": 1.583629893238434,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004754876816557963,
      "loss": 5.261,
      "mean_token_accuracy": 0.1909753382205963,
      "num_tokens": 36680902.0,
      "step": 16020
    },
    {
      "entropy": 5.574006414413452,
      "epoch": 1.5841241597469355,
      "grad_norm": 1.5,
      "learning_rate": 0.00047547166132543483,
      "loss": 5.2627,
      "mean_token_accuracy": 0.19489651471376418,
      "num_tokens": 36691312.0,
      "step": 16025
    },
    {
      "entropy": 5.588957500457764,
      "epoch": 1.584618426255437,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047545563606343945,
      "loss": 5.3141,
      "mean_token_accuracy": 0.1848904609680176,
      "num_tokens": 36701091.0,
      "step": 16030
    },
    {
      "entropy": 5.630921983718872,
      "epoch": 1.5851126927639383,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004754396058702043,
      "loss": 5.355,
      "mean_token_accuracy": 0.1891075000166893,
      "num_tokens": 36712524.0,
      "step": 16035
    },
    {
      "entropy": 5.652781963348389,
      "epoch": 1.5856069592724396,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047542357074612375,
      "loss": 5.3693,
      "mean_token_accuracy": 0.1796592429280281,
      "num_tokens": 36724181.0,
      "step": 16040
    },
    {
      "entropy": 5.607776737213134,
      "epoch": 1.586101225780941,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047540753069159224,
      "loss": 5.271,
      "mean_token_accuracy": 0.19027665108442307,
      "num_tokens": 36734777.0,
      "step": 16045
    },
    {
      "entropy": 5.624070930480957,
      "epoch": 1.5865954922894425,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047539148570700443,
      "loss": 5.3444,
      "mean_token_accuracy": 0.1836364358663559,
      "num_tokens": 36746437.0,
      "step": 16050
    },
    {
      "entropy": 5.701212692260742,
      "epoch": 1.5870897587979438,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047537543579275496,
      "loss": 5.479,
      "mean_token_accuracy": 0.177655591070652,
      "num_tokens": 36759615.0,
      "step": 16055
    },
    {
      "entropy": 5.629216146469116,
      "epoch": 1.5875840253064453,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004753593809492387,
      "loss": 5.3266,
      "mean_token_accuracy": 0.18154437094926834,
      "num_tokens": 36770762.0,
      "step": 16060
    },
    {
      "entropy": 5.618433475494385,
      "epoch": 1.5880782918149468,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004753433211768506,
      "loss": 5.3087,
      "mean_token_accuracy": 0.18189213126897813,
      "num_tokens": 36781410.0,
      "step": 16065
    },
    {
      "entropy": 5.6102643489837645,
      "epoch": 1.588572558323448,
      "grad_norm": 2.34375,
      "learning_rate": 0.00047532725647598563,
      "loss": 5.333,
      "mean_token_accuracy": 0.18333612531423568,
      "num_tokens": 36792737.0,
      "step": 16070
    },
    {
      "entropy": 5.574700546264649,
      "epoch": 1.5890668248319493,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00047531118684703916,
      "loss": 5.2375,
      "mean_token_accuracy": 0.18918032348155975,
      "num_tokens": 36804626.0,
      "step": 16075
    },
    {
      "entropy": 5.582992458343506,
      "epoch": 1.5895610913404508,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004752951122904064,
      "loss": 5.3336,
      "mean_token_accuracy": 0.18058971762657167,
      "num_tokens": 36815964.0,
      "step": 16080
    },
    {
      "entropy": 5.6143965244293215,
      "epoch": 1.5900553578489522,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004752790328064828,
      "loss": 5.2576,
      "mean_token_accuracy": 0.19323818981647492,
      "num_tokens": 36828180.0,
      "step": 16085
    },
    {
      "entropy": 5.568392133712768,
      "epoch": 1.5905496243574535,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047526294839566393,
      "loss": 5.2713,
      "mean_token_accuracy": 0.1873632401227951,
      "num_tokens": 36839589.0,
      "step": 16090
    },
    {
      "entropy": 5.591415548324585,
      "epoch": 1.5910438908659548,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004752468590583454,
      "loss": 5.2573,
      "mean_token_accuracy": 0.1895088165998459,
      "num_tokens": 36850392.0,
      "step": 16095
    },
    {
      "entropy": 5.534626626968384,
      "epoch": 1.5915381573744563,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004752307647949232,
      "loss": 5.2631,
      "mean_token_accuracy": 0.19192342609167098,
      "num_tokens": 36862334.0,
      "step": 16100
    },
    {
      "entropy": 5.5994960308074955,
      "epoch": 1.5920324238829577,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000475214665605793,
      "loss": 5.2902,
      "mean_token_accuracy": 0.18753493577241898,
      "num_tokens": 36874837.0,
      "step": 16105
    },
    {
      "entropy": 5.619603300094605,
      "epoch": 1.592526690391459,
      "grad_norm": 1.46875,
      "learning_rate": 0.000475198561491351,
      "loss": 5.3064,
      "mean_token_accuracy": 0.1915009081363678,
      "num_tokens": 36886375.0,
      "step": 16110
    },
    {
      "entropy": 5.618808841705322,
      "epoch": 1.5930209568999605,
      "grad_norm": 2.234375,
      "learning_rate": 0.0004751824524519934,
      "loss": 5.3447,
      "mean_token_accuracy": 0.18166275769472123,
      "num_tokens": 36897731.0,
      "step": 16115
    },
    {
      "entropy": 5.592964029312133,
      "epoch": 1.593515223408462,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004751663384881163,
      "loss": 5.3723,
      "mean_token_accuracy": 0.1810892716050148,
      "num_tokens": 36909647.0,
      "step": 16120
    },
    {
      "entropy": 5.582522249221801,
      "epoch": 1.5940094899169632,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004751502196001163,
      "loss": 5.3169,
      "mean_token_accuracy": 0.18631683588027953,
      "num_tokens": 36919960.0,
      "step": 16125
    },
    {
      "entropy": 5.639509057998657,
      "epoch": 1.5945037564254645,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004751340957883898,
      "loss": 5.34,
      "mean_token_accuracy": 0.18041442185640336,
      "num_tokens": 36930520.0,
      "step": 16130
    },
    {
      "entropy": 5.668814468383789,
      "epoch": 1.594998022933966,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004751179670533335,
      "loss": 5.3237,
      "mean_token_accuracy": 0.18239227086305618,
      "num_tokens": 36942627.0,
      "step": 16135
    },
    {
      "entropy": 5.598042011260986,
      "epoch": 1.5954922894424675,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004751018333953442,
      "loss": 5.3671,
      "mean_token_accuracy": 0.18338391780853272,
      "num_tokens": 36954276.0,
      "step": 16140
    },
    {
      "entropy": 5.5386909484863285,
      "epoch": 1.5959865559509687,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004750856948148187,
      "loss": 5.3036,
      "mean_token_accuracy": 0.19186697751283646,
      "num_tokens": 36965840.0,
      "step": 16145
    },
    {
      "entropy": 5.584394693374634,
      "epoch": 1.59648082245947,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004750695513121541,
      "loss": 5.2215,
      "mean_token_accuracy": 0.1885249361395836,
      "num_tokens": 36977133.0,
      "step": 16150
    },
    {
      "entropy": 5.63659405708313,
      "epoch": 1.5969750889679717,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004750534028877475,
      "loss": 5.3085,
      "mean_token_accuracy": 0.18535743057727813,
      "num_tokens": 36989013.0,
      "step": 16155
    },
    {
      "entropy": 5.546811771392822,
      "epoch": 1.597469355476473,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004750372495419962,
      "loss": 5.2693,
      "mean_token_accuracy": 0.18440254777669907,
      "num_tokens": 37000324.0,
      "step": 16160
    },
    {
      "entropy": 5.623281574249267,
      "epoch": 1.5979636219849742,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004750210912752975,
      "loss": 5.3469,
      "mean_token_accuracy": 0.18501806557178496,
      "num_tokens": 37010551.0,
      "step": 16165
    },
    {
      "entropy": 5.712255382537842,
      "epoch": 1.5984578884934757,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004750049280880489,
      "loss": 5.3869,
      "mean_token_accuracy": 0.17981976121664048,
      "num_tokens": 37022732.0,
      "step": 16170
    },
    {
      "entropy": 5.61658091545105,
      "epoch": 1.5989521550019772,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047498875998064805,
      "loss": 5.3244,
      "mean_token_accuracy": 0.1808415398001671,
      "num_tokens": 37033230.0,
      "step": 16175
    },
    {
      "entropy": 5.555230951309204,
      "epoch": 1.5994464215104784,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004749725869534926,
      "loss": 5.2248,
      "mean_token_accuracy": 0.1938748687505722,
      "num_tokens": 37043854.0,
      "step": 16180
    },
    {
      "entropy": 5.626410484313965,
      "epoch": 1.5999406880189797,
      "grad_norm": 1.46875,
      "learning_rate": 0.00047495640900698064,
      "loss": 5.3321,
      "mean_token_accuracy": 0.18536791801452637,
      "num_tokens": 37055208.0,
      "step": 16185
    },
    {
      "entropy": 5.60192174911499,
      "epoch": 1.6004349545274812,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047494022614150995,
      "loss": 5.4013,
      "mean_token_accuracy": 0.17959013283252717,
      "num_tokens": 37065938.0,
      "step": 16190
    },
    {
      "entropy": 5.597919464111328,
      "epoch": 1.6009292210359827,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004749240383574786,
      "loss": 5.2311,
      "mean_token_accuracy": 0.18632403314113616,
      "num_tokens": 37077252.0,
      "step": 16195
    },
    {
      "entropy": 5.576686239242553,
      "epoch": 1.601423487544484,
      "grad_norm": 1.453125,
      "learning_rate": 0.000474907845655285,
      "loss": 5.2873,
      "mean_token_accuracy": 0.19015153646469116,
      "num_tokens": 37087443.0,
      "step": 16200
    },
    {
      "entropy": 5.550740718841553,
      "epoch": 1.6019177540529854,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004748916480353273,
      "loss": 5.1721,
      "mean_token_accuracy": 0.20141243636608125,
      "num_tokens": 37097621.0,
      "step": 16205
    },
    {
      "entropy": 5.580372714996338,
      "epoch": 1.602412020561487,
      "grad_norm": 1.4375,
      "learning_rate": 0.00047487544549800403,
      "loss": 5.3682,
      "mean_token_accuracy": 0.18480129688978195,
      "num_tokens": 37110026.0,
      "step": 16210
    },
    {
      "entropy": 5.553441381454467,
      "epoch": 1.6029062870699882,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004748592380437138,
      "loss": 5.2897,
      "mean_token_accuracy": 0.18632829785346985,
      "num_tokens": 37121447.0,
      "step": 16215
    },
    {
      "entropy": 5.615175247192383,
      "epoch": 1.6034005535784894,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004748430256728554,
      "loss": 5.3293,
      "mean_token_accuracy": 0.18522191047668457,
      "num_tokens": 37132245.0,
      "step": 16220
    },
    {
      "entropy": 5.720344877243042,
      "epoch": 1.603894820086991,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047482680838582754,
      "loss": 5.3626,
      "mean_token_accuracy": 0.18131855726242066,
      "num_tokens": 37144135.0,
      "step": 16225
    },
    {
      "entropy": 5.659983491897583,
      "epoch": 1.6043890865954924,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004748105861830292,
      "loss": 5.3905,
      "mean_token_accuracy": 0.1781242623925209,
      "num_tokens": 37155716.0,
      "step": 16230
    },
    {
      "entropy": 5.665464925765991,
      "epoch": 1.6048833531039937,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004747943590648594,
      "loss": 5.413,
      "mean_token_accuracy": 0.1751346319913864,
      "num_tokens": 37166701.0,
      "step": 16235
    },
    {
      "entropy": 5.645764589309692,
      "epoch": 1.605377619612495,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047477812703171737,
      "loss": 5.3582,
      "mean_token_accuracy": 0.17787411212921142,
      "num_tokens": 37177655.0,
      "step": 16240
    },
    {
      "entropy": 5.611736822128296,
      "epoch": 1.6058718861209964,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047476189008400246,
      "loss": 5.3171,
      "mean_token_accuracy": 0.1821049377322197,
      "num_tokens": 37189925.0,
      "step": 16245
    },
    {
      "entropy": 5.587080430984497,
      "epoch": 1.606366152629498,
      "grad_norm": 1.296875,
      "learning_rate": 0.000474745648222114,
      "loss": 5.288,
      "mean_token_accuracy": 0.1887748658657074,
      "num_tokens": 37200756.0,
      "step": 16250
    },
    {
      "entropy": 5.621827220916748,
      "epoch": 1.6068604191379992,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004747294014464516,
      "loss": 5.3373,
      "mean_token_accuracy": 0.18284670561552047,
      "num_tokens": 37213252.0,
      "step": 16255
    },
    {
      "entropy": 5.698238182067871,
      "epoch": 1.6073546856465006,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000474713149757415,
      "loss": 5.3153,
      "mean_token_accuracy": 0.18241272121667862,
      "num_tokens": 37223754.0,
      "step": 16260
    },
    {
      "entropy": 5.612072324752807,
      "epoch": 1.6078489521550021,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047469689315540395,
      "loss": 5.4125,
      "mean_token_accuracy": 0.17724613547325135,
      "num_tokens": 37235930.0,
      "step": 16265
    },
    {
      "entropy": 5.559077358245849,
      "epoch": 1.6083432186635034,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047468063164081826,
      "loss": 5.1992,
      "mean_token_accuracy": 0.19498272836208344,
      "num_tokens": 37246794.0,
      "step": 16270
    },
    {
      "entropy": 5.570011711120605,
      "epoch": 1.6088374851720046,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004746643652140581,
      "loss": 5.2809,
      "mean_token_accuracy": 0.18514157086610794,
      "num_tokens": 37257523.0,
      "step": 16275
    },
    {
      "entropy": 5.5720607280731205,
      "epoch": 1.6093317516805061,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047464809387552366,
      "loss": 5.3975,
      "mean_token_accuracy": 0.1843038469552994,
      "num_tokens": 37269505.0,
      "step": 16280
    },
    {
      "entropy": 5.665265703201294,
      "epoch": 1.6098260181890076,
      "grad_norm": 1.1875,
      "learning_rate": 0.000474631817625615,
      "loss": 5.2979,
      "mean_token_accuracy": 0.18875951766967775,
      "num_tokens": 37281139.0,
      "step": 16285
    },
    {
      "entropy": 5.5413368225097654,
      "epoch": 1.6103202846975089,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004746155364647327,
      "loss": 5.2781,
      "mean_token_accuracy": 0.18880391567945481,
      "num_tokens": 37291784.0,
      "step": 16290
    },
    {
      "entropy": 5.616222238540649,
      "epoch": 1.6108145512060101,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047459925039327727,
      "loss": 5.3411,
      "mean_token_accuracy": 0.18243536949157715,
      "num_tokens": 37302754.0,
      "step": 16295
    },
    {
      "entropy": 5.56536831855774,
      "epoch": 1.6113088177145116,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047458295941164933,
      "loss": 5.2721,
      "mean_token_accuracy": 0.18999520540237427,
      "num_tokens": 37313478.0,
      "step": 16300
    },
    {
      "entropy": 5.568728113174439,
      "epoch": 1.611803084223013,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004745666635202496,
      "loss": 5.301,
      "mean_token_accuracy": 0.18519092500209808,
      "num_tokens": 37325223.0,
      "step": 16305
    },
    {
      "entropy": 5.628871822357178,
      "epoch": 1.6122973507315144,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000474550362719479,
      "loss": 5.3034,
      "mean_token_accuracy": 0.18422407805919647,
      "num_tokens": 37336794.0,
      "step": 16310
    },
    {
      "entropy": 5.72665753364563,
      "epoch": 1.6127916172400159,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047453405700973844,
      "loss": 5.4531,
      "mean_token_accuracy": 0.17323654890060425,
      "num_tokens": 37349171.0,
      "step": 16315
    },
    {
      "entropy": 5.633377599716186,
      "epoch": 1.6132858837485173,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004745177463914292,
      "loss": 5.4265,
      "mean_token_accuracy": 0.17821444123983382,
      "num_tokens": 37362201.0,
      "step": 16320
    },
    {
      "entropy": 5.654541254043579,
      "epoch": 1.6137801502570186,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047450143086495244,
      "loss": 5.3534,
      "mean_token_accuracy": 0.1801922008395195,
      "num_tokens": 37373858.0,
      "step": 16325
    },
    {
      "entropy": 5.64607801437378,
      "epoch": 1.6142744167655199,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004744851104307095,
      "loss": 5.358,
      "mean_token_accuracy": 0.1817267581820488,
      "num_tokens": 37385973.0,
      "step": 16330
    },
    {
      "entropy": 5.644533395767212,
      "epoch": 1.6147686832740213,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047446878508910194,
      "loss": 5.3912,
      "mean_token_accuracy": 0.18212875872850418,
      "num_tokens": 37398054.0,
      "step": 16335
    },
    {
      "entropy": 5.673584794998169,
      "epoch": 1.6152629497825228,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047445245484053133,
      "loss": 5.3355,
      "mean_token_accuracy": 0.18101910054683684,
      "num_tokens": 37408656.0,
      "step": 16340
    },
    {
      "entropy": 5.547592878341675,
      "epoch": 1.615757216291024,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004744361196853994,
      "loss": 5.2153,
      "mean_token_accuracy": 0.19336293935775756,
      "num_tokens": 37420454.0,
      "step": 16345
    },
    {
      "entropy": 5.612527751922608,
      "epoch": 1.6162514827995254,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004744197796241079,
      "loss": 5.3322,
      "mean_token_accuracy": 0.1863848313689232,
      "num_tokens": 37431879.0,
      "step": 16350
    },
    {
      "entropy": 5.631636095046997,
      "epoch": 1.6167457493080268,
      "grad_norm": 1.375,
      "learning_rate": 0.0004744034346570589,
      "loss": 5.3597,
      "mean_token_accuracy": 0.18099877387285232,
      "num_tokens": 37443487.0,
      "step": 16355
    },
    {
      "entropy": 5.544313859939575,
      "epoch": 1.6172400158165283,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004743870847846546,
      "loss": 5.2609,
      "mean_token_accuracy": 0.18936059176921843,
      "num_tokens": 37455726.0,
      "step": 16360
    },
    {
      "entropy": 5.625551223754883,
      "epoch": 1.6177342823250296,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004743707300072969,
      "loss": 5.349,
      "mean_token_accuracy": 0.1823552057147026,
      "num_tokens": 37466735.0,
      "step": 16365
    },
    {
      "entropy": 5.668590641021728,
      "epoch": 1.618228548833531,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047435437032538843,
      "loss": 5.4241,
      "mean_token_accuracy": 0.18433574587106705,
      "num_tokens": 37479599.0,
      "step": 16370
    },
    {
      "entropy": 5.549736022949219,
      "epoch": 1.6187228153420326,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00047433800573933147,
      "loss": 5.3036,
      "mean_token_accuracy": 0.1854292258620262,
      "num_tokens": 37490896.0,
      "step": 16375
    },
    {
      "entropy": 5.54924168586731,
      "epoch": 1.6192170818505338,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047432163624952874,
      "loss": 5.3313,
      "mean_token_accuracy": 0.1861521676182747,
      "num_tokens": 37501474.0,
      "step": 16380
    },
    {
      "entropy": 5.5916643142700195,
      "epoch": 1.619711348359035,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004743052618563827,
      "loss": 5.3492,
      "mean_token_accuracy": 0.18559490293264388,
      "num_tokens": 37513410.0,
      "step": 16385
    },
    {
      "entropy": 5.5729100704193115,
      "epoch": 1.6202056148675366,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004742888825602964,
      "loss": 5.2347,
      "mean_token_accuracy": 0.1972057580947876,
      "num_tokens": 37522612.0,
      "step": 16390
    },
    {
      "entropy": 5.712789106369018,
      "epoch": 1.620699881376038,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004742724983616726,
      "loss": 5.3924,
      "mean_token_accuracy": 0.17707159370183945,
      "num_tokens": 37534320.0,
      "step": 16395
    },
    {
      "entropy": 5.630206823348999,
      "epoch": 1.6211941478845393,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00047425610926091444,
      "loss": 5.3685,
      "mean_token_accuracy": 0.17957195192575454,
      "num_tokens": 37545599.0,
      "step": 16400
    },
    {
      "entropy": 5.619510316848755,
      "epoch": 1.6216884143930406,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004742397152584251,
      "loss": 5.4126,
      "mean_token_accuracy": 0.190477953851223,
      "num_tokens": 37558002.0,
      "step": 16405
    },
    {
      "entropy": 5.5778505325317385,
      "epoch": 1.6221826809015423,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047422331635460784,
      "loss": 5.2928,
      "mean_token_accuracy": 0.18724770545959474,
      "num_tokens": 37570808.0,
      "step": 16410
    },
    {
      "entropy": 5.650541353225708,
      "epoch": 1.6226769474100435,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00047420691254986615,
      "loss": 5.3349,
      "mean_token_accuracy": 0.18815899789333343,
      "num_tokens": 37581065.0,
      "step": 16415
    },
    {
      "entropy": 5.616580963134766,
      "epoch": 1.6231712139185448,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047419050384460345,
      "loss": 5.3149,
      "mean_token_accuracy": 0.1897153675556183,
      "num_tokens": 37592472.0,
      "step": 16420
    },
    {
      "entropy": 5.612893772125244,
      "epoch": 1.6236654804270463,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004741740902392235,
      "loss": 5.3053,
      "mean_token_accuracy": 0.1877980872988701,
      "num_tokens": 37604601.0,
      "step": 16425
    },
    {
      "entropy": 5.626508188247681,
      "epoch": 1.6241597469355478,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004741576717341299,
      "loss": 5.2873,
      "mean_token_accuracy": 0.187255597114563,
      "num_tokens": 37615978.0,
      "step": 16430
    },
    {
      "entropy": 5.523616456985474,
      "epoch": 1.624654013444049,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004741412483297268,
      "loss": 5.2555,
      "mean_token_accuracy": 0.1919064223766327,
      "num_tokens": 37627532.0,
      "step": 16435
    },
    {
      "entropy": 5.601925039291382,
      "epoch": 1.6251482799525503,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004741248200264181,
      "loss": 5.3752,
      "mean_token_accuracy": 0.1803853392601013,
      "num_tokens": 37640508.0,
      "step": 16440
    },
    {
      "entropy": 5.604682588577271,
      "epoch": 1.6256425464610518,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004741083868246079,
      "loss": 5.2514,
      "mean_token_accuracy": 0.1836131200194359,
      "num_tokens": 37652095.0,
      "step": 16445
    },
    {
      "entropy": 5.555622911453247,
      "epoch": 1.6261368129695533,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004740919487247005,
      "loss": 5.2743,
      "mean_token_accuracy": 0.19207785427570342,
      "num_tokens": 37663481.0,
      "step": 16450
    },
    {
      "entropy": 5.618346166610718,
      "epoch": 1.6266310794780545,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004740755057271003,
      "loss": 5.2936,
      "mean_token_accuracy": 0.18932325541973113,
      "num_tokens": 37675259.0,
      "step": 16455
    },
    {
      "entropy": 5.566410541534424,
      "epoch": 1.627125345986556,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047405905783221177,
      "loss": 5.2691,
      "mean_token_accuracy": 0.18333564102649688,
      "num_tokens": 37686006.0,
      "step": 16460
    },
    {
      "entropy": 5.579566097259521,
      "epoch": 1.6276196124950575,
      "grad_norm": 1.578125,
      "learning_rate": 0.00047404260504043945,
      "loss": 5.3272,
      "mean_token_accuracy": 0.18431832492351533,
      "num_tokens": 37697642.0,
      "step": 16465
    },
    {
      "entropy": 5.660151815414428,
      "epoch": 1.6281138790035588,
      "grad_norm": 1.40625,
      "learning_rate": 0.00047402614735218824,
      "loss": 5.3415,
      "mean_token_accuracy": 0.18479539901018144,
      "num_tokens": 37707714.0,
      "step": 16470
    },
    {
      "entropy": 5.603965520858765,
      "epoch": 1.62860814551206,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004740096847678629,
      "loss": 5.3334,
      "mean_token_accuracy": 0.18693797290325165,
      "num_tokens": 37719562.0,
      "step": 16475
    },
    {
      "entropy": 5.567446994781494,
      "epoch": 1.6291024120205615,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004739932172878685,
      "loss": 5.2968,
      "mean_token_accuracy": 0.18708339482545852,
      "num_tokens": 37731521.0,
      "step": 16480
    },
    {
      "entropy": 5.675037574768067,
      "epoch": 1.629596678529063,
      "grad_norm": 1.8515625,
      "learning_rate": 0.00047397674491261003,
      "loss": 5.3605,
      "mean_token_accuracy": 0.18157930076122283,
      "num_tokens": 37743015.0,
      "step": 16485
    },
    {
      "entropy": 5.62669267654419,
      "epoch": 1.6300909450375642,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047396026764249285,
      "loss": 5.3398,
      "mean_token_accuracy": 0.17811971306800842,
      "num_tokens": 37754219.0,
      "step": 16490
    },
    {
      "entropy": 5.547645044326782,
      "epoch": 1.6305852115460655,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004739437854779222,
      "loss": 5.3645,
      "mean_token_accuracy": 0.1810488909482956,
      "num_tokens": 37765738.0,
      "step": 16495
    },
    {
      "entropy": 5.436764430999756,
      "epoch": 1.631079478054567,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047392729841930345,
      "loss": 5.1029,
      "mean_token_accuracy": 0.20867561250925065,
      "num_tokens": 37778647.0,
      "step": 16500
    },
    {
      "entropy": 5.607191371917724,
      "epoch": 1.6315737445630685,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047391080646704237,
      "loss": 5.311,
      "mean_token_accuracy": 0.18862325698137283,
      "num_tokens": 37790380.0,
      "step": 16505
    },
    {
      "entropy": 5.565876245498657,
      "epoch": 1.6320680110715697,
      "grad_norm": 1.375,
      "learning_rate": 0.00047389430962154464,
      "loss": 5.3182,
      "mean_token_accuracy": 0.17928051203489304,
      "num_tokens": 37802474.0,
      "step": 16510
    },
    {
      "entropy": 5.5442015647888185,
      "epoch": 1.6325622775800712,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00047387780788321594,
      "loss": 5.3816,
      "mean_token_accuracy": 0.17658133208751678,
      "num_tokens": 37814154.0,
      "step": 16515
    },
    {
      "entropy": 5.692914152145386,
      "epoch": 1.6330565440885727,
      "grad_norm": 1.5,
      "learning_rate": 0.0004738613012524624,
      "loss": 5.3094,
      "mean_token_accuracy": 0.18607857823371887,
      "num_tokens": 37825371.0,
      "step": 16520
    },
    {
      "entropy": 5.652116012573242,
      "epoch": 1.633550810597074,
      "grad_norm": 1.625,
      "learning_rate": 0.00047384478972968997,
      "loss": 5.3403,
      "mean_token_accuracy": 0.18100840896368026,
      "num_tokens": 37836475.0,
      "step": 16525
    },
    {
      "entropy": 5.636460876464843,
      "epoch": 1.6340450771055752,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047382827331530483,
      "loss": 5.3688,
      "mean_token_accuracy": 0.1857140064239502,
      "num_tokens": 37848413.0,
      "step": 16530
    },
    {
      "entropy": 5.556559610366821,
      "epoch": 1.6345393436140767,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004738117520097133,
      "loss": 5.2299,
      "mean_token_accuracy": 0.1951899230480194,
      "num_tokens": 37858174.0,
      "step": 16535
    },
    {
      "entropy": 5.601485538482666,
      "epoch": 1.6350336101225782,
      "grad_norm": 1.6015625,
      "learning_rate": 0.000473795225813322,
      "loss": 5.364,
      "mean_token_accuracy": 0.18491999059915543,
      "num_tokens": 37869987.0,
      "step": 16540
    },
    {
      "entropy": 5.5809314250946045,
      "epoch": 1.6355278766310795,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047377869472653704,
      "loss": 5.2519,
      "mean_token_accuracy": 0.19710858911275864,
      "num_tokens": 37882050.0,
      "step": 16545
    },
    {
      "entropy": 5.611066007614136,
      "epoch": 1.6360221431395807,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004737621587497655,
      "loss": 5.2973,
      "mean_token_accuracy": 0.1831185296177864,
      "num_tokens": 37894537.0,
      "step": 16550
    },
    {
      "entropy": 5.640512275695801,
      "epoch": 1.6365164096480822,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000473745617883414,
      "loss": 5.2467,
      "mean_token_accuracy": 0.1942311331629753,
      "num_tokens": 37905795.0,
      "step": 16555
    },
    {
      "entropy": 5.5917558670043945,
      "epoch": 1.6370106761565837,
      "grad_norm": 1.484375,
      "learning_rate": 0.00047372907212788953,
      "loss": 5.3249,
      "mean_token_accuracy": 0.1859632760286331,
      "num_tokens": 37918778.0,
      "step": 16560
    },
    {
      "entropy": 5.649238777160645,
      "epoch": 1.637504942665085,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00047371252148359897,
      "loss": 5.3954,
      "mean_token_accuracy": 0.1771112322807312,
      "num_tokens": 37930963.0,
      "step": 16565
    },
    {
      "entropy": 5.634269762039184,
      "epoch": 1.6379992091735864,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004736959659509496,
      "loss": 5.3853,
      "mean_token_accuracy": 0.1860823005437851,
      "num_tokens": 37943067.0,
      "step": 16570
    },
    {
      "entropy": 5.605807065963745,
      "epoch": 1.638493475682088,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004736794055303487,
      "loss": 5.3529,
      "mean_token_accuracy": 0.18497975617647172,
      "num_tokens": 37953425.0,
      "step": 16575
    },
    {
      "entropy": 5.5283989906311035,
      "epoch": 1.6389877421905892,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047366284022220346,
      "loss": 5.2605,
      "mean_token_accuracy": 0.18783415257930755,
      "num_tokens": 37965801.0,
      "step": 16580
    },
    {
      "entropy": 5.683371210098267,
      "epoch": 1.6394820086990904,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004736462700269216,
      "loss": 5.3586,
      "mean_token_accuracy": 0.1798913896083832,
      "num_tokens": 37978169.0,
      "step": 16585
    },
    {
      "entropy": 5.63033618927002,
      "epoch": 1.639976275207592,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047362969494491063,
      "loss": 5.3284,
      "mean_token_accuracy": 0.18879836350679396,
      "num_tokens": 37990154.0,
      "step": 16590
    },
    {
      "entropy": 5.617295026779175,
      "epoch": 1.6404705417160934,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047361311497657845,
      "loss": 5.2305,
      "mean_token_accuracy": 0.19250458329916,
      "num_tokens": 38001249.0,
      "step": 16595
    },
    {
      "entropy": 5.53756685256958,
      "epoch": 1.6409648082245947,
      "grad_norm": 1.375,
      "learning_rate": 0.0004735965301223328,
      "loss": 5.283,
      "mean_token_accuracy": 0.19124522656202317,
      "num_tokens": 38013453.0,
      "step": 16600
    },
    {
      "entropy": 5.620217370986938,
      "epoch": 1.641459074733096,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047357994038258166,
      "loss": 5.4104,
      "mean_token_accuracy": 0.17685119509696962,
      "num_tokens": 38024463.0,
      "step": 16605
    },
    {
      "entropy": 5.6609179973602295,
      "epoch": 1.6419533412415974,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00047356334575773324,
      "loss": 5.3709,
      "mean_token_accuracy": 0.18480053395032883,
      "num_tokens": 38036131.0,
      "step": 16610
    },
    {
      "entropy": 5.5884881019592285,
      "epoch": 1.642447607750099,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004735467462481956,
      "loss": 5.2971,
      "mean_token_accuracy": 0.18905212134122848,
      "num_tokens": 38047428.0,
      "step": 16615
    },
    {
      "entropy": 5.5827028274536135,
      "epoch": 1.6429418742586002,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004735301418543773,
      "loss": 5.3279,
      "mean_token_accuracy": 0.18912371695041658,
      "num_tokens": 38059864.0,
      "step": 16620
    },
    {
      "entropy": 5.6316742420196535,
      "epoch": 1.6434361407671016,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047351353257668675,
      "loss": 5.3442,
      "mean_token_accuracy": 0.17932978868484498,
      "num_tokens": 38071356.0,
      "step": 16625
    },
    {
      "entropy": 5.6271758556365965,
      "epoch": 1.6439304072756031,
      "grad_norm": 1.5,
      "learning_rate": 0.0004734969184155324,
      "loss": 5.3443,
      "mean_token_accuracy": 0.17520896345376968,
      "num_tokens": 38082124.0,
      "step": 16630
    },
    {
      "entropy": 5.642376804351807,
      "epoch": 1.6444246737841044,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00047348029937132306,
      "loss": 5.3438,
      "mean_token_accuracy": 0.1800471603870392,
      "num_tokens": 38093451.0,
      "step": 16635
    },
    {
      "entropy": 5.5985941886901855,
      "epoch": 1.6449189402926057,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047346367544446753,
      "loss": 5.3464,
      "mean_token_accuracy": 0.17877690345048905,
      "num_tokens": 38105078.0,
      "step": 16640
    },
    {
      "entropy": 5.5512065410614015,
      "epoch": 1.6454132068011071,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004734470466353749,
      "loss": 5.1954,
      "mean_token_accuracy": 0.19351691603660584,
      "num_tokens": 38114953.0,
      "step": 16645
    },
    {
      "entropy": 5.576357316970825,
      "epoch": 1.6459074733096086,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004734304129444542,
      "loss": 5.2643,
      "mean_token_accuracy": 0.18573001325130462,
      "num_tokens": 38126465.0,
      "step": 16650
    },
    {
      "entropy": 5.688961219787598,
      "epoch": 1.6464017398181099,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004734137743721144,
      "loss": 5.3989,
      "mean_token_accuracy": 0.18103523552417755,
      "num_tokens": 38139223.0,
      "step": 16655
    },
    {
      "entropy": 5.517856693267822,
      "epoch": 1.6468960063266112,
      "grad_norm": 1.484375,
      "learning_rate": 0.000473397130918765,
      "loss": 5.278,
      "mean_token_accuracy": 0.19320471286773683,
      "num_tokens": 38151676.0,
      "step": 16660
    },
    {
      "entropy": 5.570414400100708,
      "epoch": 1.6473902728351129,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047338048258481545,
      "loss": 5.288,
      "mean_token_accuracy": 0.1813721239566803,
      "num_tokens": 38162759.0,
      "step": 16665
    },
    {
      "entropy": 5.579239845275879,
      "epoch": 1.6478845393436141,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004733638293706752,
      "loss": 5.2953,
      "mean_token_accuracy": 0.1891671448945999,
      "num_tokens": 38174820.0,
      "step": 16670
    },
    {
      "entropy": 5.528016471862793,
      "epoch": 1.6483788058521154,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000473347171276754,
      "loss": 5.1845,
      "mean_token_accuracy": 0.19763631820678712,
      "num_tokens": 38185558.0,
      "step": 16675
    },
    {
      "entropy": 5.577135324478149,
      "epoch": 1.6488730723606169,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004733305083034617,
      "loss": 5.3602,
      "mean_token_accuracy": 0.1821441262960434,
      "num_tokens": 38198231.0,
      "step": 16680
    },
    {
      "entropy": 5.645250463485718,
      "epoch": 1.6493673388691183,
      "grad_norm": 1.375,
      "learning_rate": 0.00047331384045120803,
      "loss": 5.3279,
      "mean_token_accuracy": 0.18815534561872482,
      "num_tokens": 38209539.0,
      "step": 16685
    },
    {
      "entropy": 5.629117488861084,
      "epoch": 1.6498616053776196,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004732971677204031,
      "loss": 5.3586,
      "mean_token_accuracy": 0.17826489508152008,
      "num_tokens": 38220873.0,
      "step": 16690
    },
    {
      "entropy": 5.61576828956604,
      "epoch": 1.6503558718861209,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047328049011145714,
      "loss": 5.2584,
      "mean_token_accuracy": 0.19291307777166367,
      "num_tokens": 38231186.0,
      "step": 16695
    },
    {
      "entropy": 5.634268760681152,
      "epoch": 1.6508501383946224,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004732638076247804,
      "loss": 5.384,
      "mean_token_accuracy": 0.17331185787916184,
      "num_tokens": 38242152.0,
      "step": 16700
    },
    {
      "entropy": 5.664009284973145,
      "epoch": 1.6513444049031238,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047324712026078315,
      "loss": 5.3286,
      "mean_token_accuracy": 0.18384687900543212,
      "num_tokens": 38254281.0,
      "step": 16705
    },
    {
      "entropy": 5.559139442443848,
      "epoch": 1.651838671411625,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000473230428019876,
      "loss": 5.1981,
      "mean_token_accuracy": 0.1868847668170929,
      "num_tokens": 38265120.0,
      "step": 16710
    },
    {
      "entropy": 5.592050409317016,
      "epoch": 1.6523329379201266,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047321373090246957,
      "loss": 5.2522,
      "mean_token_accuracy": 0.19010176956653596,
      "num_tokens": 38275509.0,
      "step": 16715
    },
    {
      "entropy": 5.543418598175049,
      "epoch": 1.652827204428628,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047319702890897457,
      "loss": 5.2443,
      "mean_token_accuracy": 0.19703299701213836,
      "num_tokens": 38285586.0,
      "step": 16720
    },
    {
      "entropy": 5.586500930786133,
      "epoch": 1.6533214709371293,
      "grad_norm": 1.7734375,
      "learning_rate": 0.000473180322039802,
      "loss": 5.2679,
      "mean_token_accuracy": 0.1873980015516281,
      "num_tokens": 38296091.0,
      "step": 16725
    },
    {
      "entropy": 5.597022199630738,
      "epoch": 1.6538157374456306,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004731636102953627,
      "loss": 5.3475,
      "mean_token_accuracy": 0.18359087109565736,
      "num_tokens": 38308520.0,
      "step": 16730
    },
    {
      "entropy": 5.635043001174926,
      "epoch": 1.654310003954132,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004731468936760679,
      "loss": 5.3257,
      "mean_token_accuracy": 0.1828442320227623,
      "num_tokens": 38320584.0,
      "step": 16735
    },
    {
      "entropy": 5.59954867362976,
      "epoch": 1.6548042704626336,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004731301721823287,
      "loss": 5.2892,
      "mean_token_accuracy": 0.18667288422584533,
      "num_tokens": 38332646.0,
      "step": 16740
    },
    {
      "entropy": 5.630438184738159,
      "epoch": 1.6552985369711348,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004731134458145566,
      "loss": 5.3275,
      "mean_token_accuracy": 0.18137831836938859,
      "num_tokens": 38343382.0,
      "step": 16745
    },
    {
      "entropy": 5.69502820968628,
      "epoch": 1.655792803479636,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047309671457316295,
      "loss": 5.3761,
      "mean_token_accuracy": 0.17543837130069734,
      "num_tokens": 38354809.0,
      "step": 16750
    },
    {
      "entropy": 5.583798122406006,
      "epoch": 1.6562870699881376,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047307997845855946,
      "loss": 5.3126,
      "mean_token_accuracy": 0.1855696126818657,
      "num_tokens": 38366524.0,
      "step": 16755
    },
    {
      "entropy": 5.5409763813018795,
      "epoch": 1.656781336496639,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004730632374711578,
      "loss": 5.2419,
      "mean_token_accuracy": 0.19076735526323318,
      "num_tokens": 38377645.0,
      "step": 16760
    },
    {
      "entropy": 5.6520342350006105,
      "epoch": 1.6572756030051403,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004730464916113696,
      "loss": 5.3838,
      "mean_token_accuracy": 0.1905674546957016,
      "num_tokens": 38389633.0,
      "step": 16765
    },
    {
      "entropy": 5.6495801448822025,
      "epoch": 1.6577698695136418,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00047302974087960704,
      "loss": 5.3665,
      "mean_token_accuracy": 0.18001354038715361,
      "num_tokens": 38401255.0,
      "step": 16770
    },
    {
      "entropy": 5.6040256977081295,
      "epoch": 1.6582641360221433,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004730129852762822,
      "loss": 5.3101,
      "mean_token_accuracy": 0.18237418830394744,
      "num_tokens": 38413160.0,
      "step": 16775
    },
    {
      "entropy": 5.593385553359985,
      "epoch": 1.6587584025306445,
      "grad_norm": 1.53125,
      "learning_rate": 0.00047299622480180716,
      "loss": 5.3571,
      "mean_token_accuracy": 0.1793212816119194,
      "num_tokens": 38423789.0,
      "step": 16780
    },
    {
      "entropy": 5.61279182434082,
      "epoch": 1.6592526690391458,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047297945945659424,
      "loss": 5.3137,
      "mean_token_accuracy": 0.18210093080997466,
      "num_tokens": 38435031.0,
      "step": 16785
    },
    {
      "entropy": 5.595762157440186,
      "epoch": 1.6597469355476473,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047296268924105594,
      "loss": 5.2917,
      "mean_token_accuracy": 0.19297283589839936,
      "num_tokens": 38445146.0,
      "step": 16790
    },
    {
      "entropy": 5.646050357818604,
      "epoch": 1.6602412020561488,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047294591415560484,
      "loss": 5.3501,
      "mean_token_accuracy": 0.1796015441417694,
      "num_tokens": 38457429.0,
      "step": 16795
    },
    {
      "entropy": 5.633969354629516,
      "epoch": 1.66073546856465,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004729291342006535,
      "loss": 5.3239,
      "mean_token_accuracy": 0.188394233584404,
      "num_tokens": 38468218.0,
      "step": 16800
    },
    {
      "entropy": 5.651690578460693,
      "epoch": 1.6612297350731513,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004729123493766147,
      "loss": 5.3682,
      "mean_token_accuracy": 0.18079716563224793,
      "num_tokens": 38480841.0,
      "step": 16805
    },
    {
      "entropy": 5.5538781642913815,
      "epoch": 1.6617240015816528,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004728955596839014,
      "loss": 5.2478,
      "mean_token_accuracy": 0.184818834066391,
      "num_tokens": 38493115.0,
      "step": 16810
    },
    {
      "entropy": 5.6409104347229,
      "epoch": 1.6622182680901543,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004728787651229267,
      "loss": 5.2922,
      "mean_token_accuracy": 0.18741248697042465,
      "num_tokens": 38505187.0,
      "step": 16815
    },
    {
      "entropy": 5.62151689529419,
      "epoch": 1.6627125345986555,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047286196569410366,
      "loss": 5.4234,
      "mean_token_accuracy": 0.17665693461894988,
      "num_tokens": 38518736.0,
      "step": 16820
    },
    {
      "entropy": 5.54524941444397,
      "epoch": 1.663206801107157,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047284516139784556,
      "loss": 5.2484,
      "mean_token_accuracy": 0.19550250321626664,
      "num_tokens": 38529703.0,
      "step": 16825
    },
    {
      "entropy": 5.582247686386109,
      "epoch": 1.6637010676156585,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004728283522345658,
      "loss": 5.3141,
      "mean_token_accuracy": 0.178050996363163,
      "num_tokens": 38542200.0,
      "step": 16830
    },
    {
      "entropy": 5.633548259735107,
      "epoch": 1.6641953341241598,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004728115382046779,
      "loss": 5.3515,
      "mean_token_accuracy": 0.18050352483987808,
      "num_tokens": 38552921.0,
      "step": 16835
    },
    {
      "entropy": 5.61789493560791,
      "epoch": 1.664689600632661,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004727947193085954,
      "loss": 5.3661,
      "mean_token_accuracy": 0.18265555948019027,
      "num_tokens": 38564837.0,
      "step": 16840
    },
    {
      "entropy": 5.625684118270874,
      "epoch": 1.6651838671411625,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047277789554673217,
      "loss": 5.319,
      "mean_token_accuracy": 0.18435919731855394,
      "num_tokens": 38575786.0,
      "step": 16845
    },
    {
      "entropy": 5.649453401565552,
      "epoch": 1.665678133649664,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047276106691950204,
      "loss": 5.357,
      "mean_token_accuracy": 0.17883790731430055,
      "num_tokens": 38588232.0,
      "step": 16850
    },
    {
      "entropy": 5.615127325057983,
      "epoch": 1.6661724001581653,
      "grad_norm": 1.484375,
      "learning_rate": 0.00047274423342731896,
      "loss": 5.2989,
      "mean_token_accuracy": 0.182313534617424,
      "num_tokens": 38600276.0,
      "step": 16855
    },
    {
      "entropy": 5.621436691284179,
      "epoch": 1.6666666666666665,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004727273950705971,
      "loss": 5.3232,
      "mean_token_accuracy": 0.18429888784885406,
      "num_tokens": 38611879.0,
      "step": 16860
    },
    {
      "entropy": 5.612408685684204,
      "epoch": 1.667160933175168,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047271055184975057,
      "loss": 5.3806,
      "mean_token_accuracy": 0.17519059628248215,
      "num_tokens": 38623032.0,
      "step": 16865
    },
    {
      "entropy": 5.678209400177002,
      "epoch": 1.6676551996836695,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004726937037651939,
      "loss": 5.3433,
      "mean_token_accuracy": 0.18555716872215272,
      "num_tokens": 38633864.0,
      "step": 16870
    },
    {
      "entropy": 5.558736944198609,
      "epoch": 1.6681494661921707,
      "grad_norm": 1.484375,
      "learning_rate": 0.00047267685081734133,
      "loss": 5.2968,
      "mean_token_accuracy": 0.1855804815888405,
      "num_tokens": 38645208.0,
      "step": 16875
    },
    {
      "entropy": 5.527917814254761,
      "epoch": 1.6686437327006722,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004726599930066076,
      "loss": 5.2524,
      "mean_token_accuracy": 0.18979646563529967,
      "num_tokens": 38655181.0,
      "step": 16880
    },
    {
      "entropy": 5.700963878631592,
      "epoch": 1.6691379992091737,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004726431303334074,
      "loss": 5.3787,
      "mean_token_accuracy": 0.18114294111728668,
      "num_tokens": 38667381.0,
      "step": 16885
    },
    {
      "entropy": 5.640314674377441,
      "epoch": 1.669632265717675,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047262626279815555,
      "loss": 5.337,
      "mean_token_accuracy": 0.18546753078699113,
      "num_tokens": 38680479.0,
      "step": 16890
    },
    {
      "entropy": 5.619797801971435,
      "epoch": 1.6701265322261762,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000472609390401267,
      "loss": 5.3317,
      "mean_token_accuracy": 0.18610499650239945,
      "num_tokens": 38691737.0,
      "step": 16895
    },
    {
      "entropy": 5.60442419052124,
      "epoch": 1.6706207987346777,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004725925131431567,
      "loss": 5.2764,
      "mean_token_accuracy": 0.19037256091833116,
      "num_tokens": 38703306.0,
      "step": 16900
    },
    {
      "entropy": 5.617859268188477,
      "epoch": 1.6711150652431792,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00047257563102424,
      "loss": 5.313,
      "mean_token_accuracy": 0.18335749804973603,
      "num_tokens": 38714683.0,
      "step": 16905
    },
    {
      "entropy": 5.633234119415283,
      "epoch": 1.6716093317516805,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004725587440449321,
      "loss": 5.3991,
      "mean_token_accuracy": 0.1770082876086235,
      "num_tokens": 38727705.0,
      "step": 16910
    },
    {
      "entropy": 5.6590173721313475,
      "epoch": 1.6721035982601817,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047254185220564843,
      "loss": 5.3825,
      "mean_token_accuracy": 0.1855751559138298,
      "num_tokens": 38738687.0,
      "step": 16915
    },
    {
      "entropy": 5.5989988327026365,
      "epoch": 1.6725978647686834,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047252495550680453,
      "loss": 5.2853,
      "mean_token_accuracy": 0.18258834034204482,
      "num_tokens": 38749437.0,
      "step": 16920
    },
    {
      "entropy": 5.50099630355835,
      "epoch": 1.6730921312771847,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004725080539488162,
      "loss": 5.1821,
      "mean_token_accuracy": 0.20848096907138824,
      "num_tokens": 38761342.0,
      "step": 16925
    },
    {
      "entropy": 5.627041053771973,
      "epoch": 1.673586397785686,
      "grad_norm": 2.296875,
      "learning_rate": 0.00047249114753209896,
      "loss": 5.3708,
      "mean_token_accuracy": 0.18091026842594146,
      "num_tokens": 38773019.0,
      "step": 16930
    },
    {
      "entropy": 5.709657430648804,
      "epoch": 1.6740806642941874,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047247423625706894,
      "loss": 5.3533,
      "mean_token_accuracy": 0.18081678599119186,
      "num_tokens": 38785291.0,
      "step": 16935
    },
    {
      "entropy": 5.614141654968262,
      "epoch": 1.674574930802689,
      "grad_norm": 1.453125,
      "learning_rate": 0.000472457320124142,
      "loss": 5.3245,
      "mean_token_accuracy": 0.18269609063863754,
      "num_tokens": 38796926.0,
      "step": 16940
    },
    {
      "entropy": 5.640139865875244,
      "epoch": 1.6750691973111902,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047244039913373444,
      "loss": 5.3089,
      "mean_token_accuracy": 0.18822263777256013,
      "num_tokens": 38809839.0,
      "step": 16945
    },
    {
      "entropy": 5.665899753570557,
      "epoch": 1.6755634638196915,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004724234732862624,
      "loss": 5.4078,
      "mean_token_accuracy": 0.174242602288723,
      "num_tokens": 38823059.0,
      "step": 16950
    },
    {
      "entropy": 5.56723051071167,
      "epoch": 1.676057730328193,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004724065425821422,
      "loss": 5.2031,
      "mean_token_accuracy": 0.19865086078643798,
      "num_tokens": 38834275.0,
      "step": 16955
    },
    {
      "entropy": 5.579958963394165,
      "epoch": 1.6765519968366944,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004723896070217905,
      "loss": 5.2993,
      "mean_token_accuracy": 0.18667323440313338,
      "num_tokens": 38845992.0,
      "step": 16960
    },
    {
      "entropy": 5.631908941268921,
      "epoch": 1.6770462633451957,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004723726666056239,
      "loss": 5.3728,
      "mean_token_accuracy": 0.18548421561717987,
      "num_tokens": 38857269.0,
      "step": 16965
    },
    {
      "entropy": 5.625741910934448,
      "epoch": 1.677540529853697,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00047235572133405904,
      "loss": 5.3806,
      "mean_token_accuracy": 0.18023659884929658,
      "num_tokens": 38869393.0,
      "step": 16970
    },
    {
      "entropy": 5.69693169593811,
      "epoch": 1.6780347963621987,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047233877120751277,
      "loss": 5.345,
      "mean_token_accuracy": 0.18244201838970184,
      "num_tokens": 38881625.0,
      "step": 16975
    },
    {
      "entropy": 5.61612720489502,
      "epoch": 1.6785290628707,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047232181622640216,
      "loss": 5.3105,
      "mean_token_accuracy": 0.1850807934999466,
      "num_tokens": 38892452.0,
      "step": 16980
    },
    {
      "entropy": 5.599652290344238,
      "epoch": 1.6790233293792012,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004723048563911443,
      "loss": 5.1921,
      "mean_token_accuracy": 0.19920271188020705,
      "num_tokens": 38902415.0,
      "step": 16985
    },
    {
      "entropy": 5.559013795852661,
      "epoch": 1.6795175958877027,
      "grad_norm": 1.453125,
      "learning_rate": 0.00047228789170215623,
      "loss": 5.2564,
      "mean_token_accuracy": 0.1962148904800415,
      "num_tokens": 38912655.0,
      "step": 16990
    },
    {
      "entropy": 5.604779148101807,
      "epoch": 1.6800118623962041,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00047227092215985553,
      "loss": 5.3137,
      "mean_token_accuracy": 0.1834467902779579,
      "num_tokens": 38923295.0,
      "step": 16995
    },
    {
      "entropy": 5.628239107131958,
      "epoch": 1.6805061289047054,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004722539477646595,
      "loss": 5.3356,
      "mean_token_accuracy": 0.1802532345056534,
      "num_tokens": 38934879.0,
      "step": 17000
    },
    {
      "entropy": 5.684195280075073,
      "epoch": 1.6810003954132067,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047223696851698575,
      "loss": 5.4377,
      "mean_token_accuracy": 0.17668186277151107,
      "num_tokens": 38946574.0,
      "step": 17005
    },
    {
      "entropy": 5.59917459487915,
      "epoch": 1.6814946619217082,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000472219984417252,
      "loss": 5.3444,
      "mean_token_accuracy": 0.18711143136024475,
      "num_tokens": 38958136.0,
      "step": 17010
    },
    {
      "entropy": 5.639297246932983,
      "epoch": 1.6819889284302096,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004722029954658761,
      "loss": 5.2896,
      "mean_token_accuracy": 0.19137297570705414,
      "num_tokens": 38969353.0,
      "step": 17015
    },
    {
      "entropy": 5.525437068939209,
      "epoch": 1.682483194938711,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004721860016632758,
      "loss": 5.2684,
      "mean_token_accuracy": 0.19085678160190583,
      "num_tokens": 38979878.0,
      "step": 17020
    },
    {
      "entropy": 5.611176776885986,
      "epoch": 1.6829774614472124,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004721690030098693,
      "loss": 5.4148,
      "mean_token_accuracy": 0.18562639951705934,
      "num_tokens": 38992847.0,
      "step": 17025
    },
    {
      "entropy": 5.622990417480469,
      "epoch": 1.6834717279557139,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004721519995060748,
      "loss": 5.2991,
      "mean_token_accuracy": 0.18584561347961426,
      "num_tokens": 39003874.0,
      "step": 17030
    },
    {
      "entropy": 5.617108631134033,
      "epoch": 1.6839659944642151,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047213499115231043,
      "loss": 5.2694,
      "mean_token_accuracy": 0.18640926629304885,
      "num_tokens": 39014327.0,
      "step": 17035
    },
    {
      "entropy": 5.605392789840698,
      "epoch": 1.6844602609727164,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004721179779489948,
      "loss": 5.2492,
      "mean_token_accuracy": 0.18275727033615113,
      "num_tokens": 39024507.0,
      "step": 17040
    },
    {
      "entropy": 5.529019927978515,
      "epoch": 1.6849545274812179,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047210095989654624,
      "loss": 5.2665,
      "mean_token_accuracy": 0.18543069660663605,
      "num_tokens": 39036184.0,
      "step": 17045
    },
    {
      "entropy": 5.532121562957764,
      "epoch": 1.6854487939897194,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047208393699538356,
      "loss": 5.233,
      "mean_token_accuracy": 0.19430117756128312,
      "num_tokens": 39047649.0,
      "step": 17050
    },
    {
      "entropy": 5.548897743225098,
      "epoch": 1.6859430604982206,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047206690924592545,
      "loss": 5.2543,
      "mean_token_accuracy": 0.18739078044891358,
      "num_tokens": 39058167.0,
      "step": 17055
    },
    {
      "entropy": 5.60099229812622,
      "epoch": 1.6864373270067219,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004720498766485907,
      "loss": 5.301,
      "mean_token_accuracy": 0.1867012232542038,
      "num_tokens": 39070099.0,
      "step": 17060
    },
    {
      "entropy": 5.612334156036377,
      "epoch": 1.6869315935152234,
      "grad_norm": 1.375,
      "learning_rate": 0.00047203283920379846,
      "loss": 5.2228,
      "mean_token_accuracy": 0.19435063153505325,
      "num_tokens": 39080337.0,
      "step": 17065
    },
    {
      "entropy": 5.710967493057251,
      "epoch": 1.6874258600237249,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047201579691196784,
      "loss": 5.4926,
      "mean_token_accuracy": 0.17481014579534532,
      "num_tokens": 39093455.0,
      "step": 17070
    },
    {
      "entropy": 5.619350147247315,
      "epoch": 1.6879201265322261,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047199874977351807,
      "loss": 5.3148,
      "mean_token_accuracy": 0.18065599948167801,
      "num_tokens": 39105424.0,
      "step": 17075
    },
    {
      "entropy": 5.576512002944947,
      "epoch": 1.6884143930407276,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004719816977888685,
      "loss": 5.253,
      "mean_token_accuracy": 0.19569894522428513,
      "num_tokens": 39117441.0,
      "step": 17080
    },
    {
      "entropy": 5.5913372993469235,
      "epoch": 1.688908659549229,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047196464095843857,
      "loss": 5.3498,
      "mean_token_accuracy": 0.18002158999443055,
      "num_tokens": 39129853.0,
      "step": 17085
    },
    {
      "entropy": 5.618453884124756,
      "epoch": 1.6894029260577303,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004719475792826479,
      "loss": 5.2688,
      "mean_token_accuracy": 0.19096520990133287,
      "num_tokens": 39141132.0,
      "step": 17090
    },
    {
      "entropy": 5.5063755989074705,
      "epoch": 1.6898971925662316,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004719305127619162,
      "loss": 5.2697,
      "mean_token_accuracy": 0.1976846918463707,
      "num_tokens": 39153892.0,
      "step": 17095
    },
    {
      "entropy": 5.623325443267822,
      "epoch": 1.690391459074733,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004719134413966635,
      "loss": 5.2339,
      "mean_token_accuracy": 0.191673544049263,
      "num_tokens": 39164410.0,
      "step": 17100
    },
    {
      "entropy": 5.5699601650238035,
      "epoch": 1.6908857255832346,
      "grad_norm": 1.40625,
      "learning_rate": 0.00047189636518730945,
      "loss": 5.2189,
      "mean_token_accuracy": 0.19667695760726928,
      "num_tokens": 39174612.0,
      "step": 17105
    },
    {
      "entropy": 5.593900680541992,
      "epoch": 1.6913799920917358,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047187928413427424,
      "loss": 5.2981,
      "mean_token_accuracy": 0.1875629559159279,
      "num_tokens": 39187355.0,
      "step": 17110
    },
    {
      "entropy": 5.61171407699585,
      "epoch": 1.691874258600237,
      "grad_norm": 1.8046875,
      "learning_rate": 0.00047186219823797815,
      "loss": 5.3324,
      "mean_token_accuracy": 0.18161751180887223,
      "num_tokens": 39199145.0,
      "step": 17115
    },
    {
      "entropy": 5.686924171447754,
      "epoch": 1.6923685251087386,
      "grad_norm": 1.484375,
      "learning_rate": 0.00047184510749884146,
      "loss": 5.3753,
      "mean_token_accuracy": 0.18694525361061096,
      "num_tokens": 39210589.0,
      "step": 17120
    },
    {
      "entropy": 5.5554056644439695,
      "epoch": 1.69286279161724,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004718280119172846,
      "loss": 5.2507,
      "mean_token_accuracy": 0.18905833512544631,
      "num_tokens": 39222657.0,
      "step": 17125
    },
    {
      "entropy": 5.543187284469605,
      "epoch": 1.6933570581257413,
      "grad_norm": 1.640625,
      "learning_rate": 0.00047181091149372807,
      "loss": 5.2296,
      "mean_token_accuracy": 0.20157125890254973,
      "num_tokens": 39233439.0,
      "step": 17130
    },
    {
      "entropy": 5.5260087013244625,
      "epoch": 1.6938513246342428,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004717938062285926,
      "loss": 5.2488,
      "mean_token_accuracy": 0.1965789318084717,
      "num_tokens": 39243787.0,
      "step": 17135
    },
    {
      "entropy": 5.625337028503418,
      "epoch": 1.6943455911427443,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047177669612229894,
      "loss": 5.3315,
      "mean_token_accuracy": 0.18010910451412201,
      "num_tokens": 39255658.0,
      "step": 17140
    },
    {
      "entropy": 5.640790319442749,
      "epoch": 1.6948398576512456,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047175958117526805,
      "loss": 5.3526,
      "mean_token_accuracy": 0.1809498891234398,
      "num_tokens": 39269480.0,
      "step": 17145
    },
    {
      "entropy": 5.632169532775879,
      "epoch": 1.6953341241597468,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047174246138792096,
      "loss": 5.3408,
      "mean_token_accuracy": 0.18934381157159805,
      "num_tokens": 39281063.0,
      "step": 17150
    },
    {
      "entropy": 5.624203538894653,
      "epoch": 1.6958283906682483,
      "grad_norm": 1.421875,
      "learning_rate": 0.00047172533676067874,
      "loss": 5.2909,
      "mean_token_accuracy": 0.18480647206306458,
      "num_tokens": 39292188.0,
      "step": 17155
    },
    {
      "entropy": 5.606174564361572,
      "epoch": 1.6963226571767498,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004717082072939628,
      "loss": 5.3258,
      "mean_token_accuracy": 0.183785642683506,
      "num_tokens": 39303661.0,
      "step": 17160
    },
    {
      "entropy": 5.593181324005127,
      "epoch": 1.696816923685251,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004716910729881944,
      "loss": 5.3035,
      "mean_token_accuracy": 0.1838001474738121,
      "num_tokens": 39315028.0,
      "step": 17165
    },
    {
      "entropy": 5.4845743656158445,
      "epoch": 1.6973111901937523,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047167393384379513,
      "loss": 5.222,
      "mean_token_accuracy": 0.19210754185914994,
      "num_tokens": 39326629.0,
      "step": 17170
    },
    {
      "entropy": 5.49462661743164,
      "epoch": 1.6978054567022538,
      "grad_norm": 1.515625,
      "learning_rate": 0.00047165678986118653,
      "loss": 5.2112,
      "mean_token_accuracy": 0.19728889763355256,
      "num_tokens": 39338821.0,
      "step": 17175
    },
    {
      "entropy": 5.590615892410279,
      "epoch": 1.6982997232107553,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004716396410407904,
      "loss": 5.2055,
      "mean_token_accuracy": 0.19036183804273604,
      "num_tokens": 39350093.0,
      "step": 17180
    },
    {
      "entropy": 5.522876310348511,
      "epoch": 1.6987939897192565,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004716224873830286,
      "loss": 5.2128,
      "mean_token_accuracy": 0.19431255161762237,
      "num_tokens": 39361120.0,
      "step": 17185
    },
    {
      "entropy": 5.584937524795532,
      "epoch": 1.699288256227758,
      "grad_norm": 1.703125,
      "learning_rate": 0.00047160532888832315,
      "loss": 5.3103,
      "mean_token_accuracy": 0.18529399782419204,
      "num_tokens": 39373040.0,
      "step": 17190
    },
    {
      "entropy": 5.607525396347046,
      "epoch": 1.6997825227362595,
      "grad_norm": 1.421875,
      "learning_rate": 0.00047158816555709597,
      "loss": 5.2519,
      "mean_token_accuracy": 0.18857672810554504,
      "num_tokens": 39384576.0,
      "step": 17195
    },
    {
      "entropy": 5.600713062286377,
      "epoch": 1.7002767892447608,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004715709973897696,
      "loss": 5.2405,
      "mean_token_accuracy": 0.19286413490772247,
      "num_tokens": 39397550.0,
      "step": 17200
    },
    {
      "entropy": 5.60021390914917,
      "epoch": 1.700771055753262,
      "grad_norm": 1.46875,
      "learning_rate": 0.00047155382438676604,
      "loss": 5.3157,
      "mean_token_accuracy": 0.1889979660511017,
      "num_tokens": 39407474.0,
      "step": 17205
    },
    {
      "entropy": 5.577948570251465,
      "epoch": 1.7012653222617635,
      "grad_norm": 1.59375,
      "learning_rate": 0.00047153664654850803,
      "loss": 5.2856,
      "mean_token_accuracy": 0.18776836693286897,
      "num_tokens": 39420104.0,
      "step": 17210
    },
    {
      "entropy": 5.661645793914795,
      "epoch": 1.701759588770265,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047151946387541795,
      "loss": 5.4216,
      "mean_token_accuracy": 0.17575209438800812,
      "num_tokens": 39431887.0,
      "step": 17215
    },
    {
      "entropy": 5.553757047653198,
      "epoch": 1.7022538552787663,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004715022763679185,
      "loss": 5.2569,
      "mean_token_accuracy": 0.18703849911689757,
      "num_tokens": 39442772.0,
      "step": 17220
    },
    {
      "entropy": 5.6448485374450685,
      "epoch": 1.7027481217872675,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047148508402643273,
      "loss": 5.3712,
      "mean_token_accuracy": 0.18131782710552216,
      "num_tokens": 39455067.0,
      "step": 17225
    },
    {
      "entropy": 5.580223608016968,
      "epoch": 1.7032423882957692,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004714678868513834,
      "loss": 5.3584,
      "mean_token_accuracy": 0.1831586554646492,
      "num_tokens": 39466700.0,
      "step": 17230
    },
    {
      "entropy": 5.6199592590332035,
      "epoch": 1.7037366548042705,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004714506848431935,
      "loss": 5.3319,
      "mean_token_accuracy": 0.19206959754228592,
      "num_tokens": 39477591.0,
      "step": 17235
    },
    {
      "entropy": 5.5991654872894285,
      "epoch": 1.7042309213127718,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047143347800228624,
      "loss": 5.3127,
      "mean_token_accuracy": 0.18071477562189103,
      "num_tokens": 39488615.0,
      "step": 17240
    },
    {
      "entropy": 5.656261301040649,
      "epoch": 1.7047251878212732,
      "grad_norm": 1.546875,
      "learning_rate": 0.000471416266329085,
      "loss": 5.3277,
      "mean_token_accuracy": 0.1849161058664322,
      "num_tokens": 39499610.0,
      "step": 17245
    },
    {
      "entropy": 5.652618741989135,
      "epoch": 1.7052194543297747,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004713990498240131,
      "loss": 5.3008,
      "mean_token_accuracy": 0.1860312044620514,
      "num_tokens": 39511180.0,
      "step": 17250
    },
    {
      "entropy": 5.580204868316651,
      "epoch": 1.705713720838276,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047138182848749415,
      "loss": 5.3333,
      "mean_token_accuracy": 0.19335238486528397,
      "num_tokens": 39523271.0,
      "step": 17255
    },
    {
      "entropy": 5.5935310363769535,
      "epoch": 1.7062079873467773,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004713646023199518,
      "loss": 5.338,
      "mean_token_accuracy": 0.19217924773693085,
      "num_tokens": 39534334.0,
      "step": 17260
    },
    {
      "entropy": 5.581711387634277,
      "epoch": 1.7067022538552787,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004713473713218097,
      "loss": 5.2431,
      "mean_token_accuracy": 0.19212531447410583,
      "num_tokens": 39545937.0,
      "step": 17265
    },
    {
      "entropy": 5.646730279922485,
      "epoch": 1.7071965203637802,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004713301354934918,
      "loss": 5.3665,
      "mean_token_accuracy": 0.18449095785617828,
      "num_tokens": 39558129.0,
      "step": 17270
    },
    {
      "entropy": 5.6310755729675295,
      "epoch": 1.7076907868722815,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004713128948354221,
      "loss": 5.3526,
      "mean_token_accuracy": 0.18579082638025285,
      "num_tokens": 39569649.0,
      "step": 17275
    },
    {
      "entropy": 5.620336723327637,
      "epoch": 1.708185053380783,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004712956493480248,
      "loss": 5.3703,
      "mean_token_accuracy": 0.1837715446949005,
      "num_tokens": 39581051.0,
      "step": 17280
    },
    {
      "entropy": 5.6267365455627445,
      "epoch": 1.7086793198892845,
      "grad_norm": 1.640625,
      "learning_rate": 0.000471278399031724,
      "loss": 5.3297,
      "mean_token_accuracy": 0.18477394729852675,
      "num_tokens": 39591225.0,
      "step": 17285
    },
    {
      "entropy": 5.576082992553711,
      "epoch": 1.7091735863977857,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004712611438869442,
      "loss": 5.2126,
      "mean_token_accuracy": 0.19559209942817687,
      "num_tokens": 39601270.0,
      "step": 17290
    },
    {
      "entropy": 5.540659236907959,
      "epoch": 1.709667852906287,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047124388391410975,
      "loss": 5.1512,
      "mean_token_accuracy": 0.19379169195890428,
      "num_tokens": 39611574.0,
      "step": 17295
    },
    {
      "entropy": 5.541296529769897,
      "epoch": 1.7101621194147885,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004712266191136453,
      "loss": 5.2399,
      "mean_token_accuracy": 0.19656410962343215,
      "num_tokens": 39622889.0,
      "step": 17300
    },
    {
      "entropy": 5.607184553146363,
      "epoch": 1.71065638592329,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004712093494859757,
      "loss": 5.2887,
      "mean_token_accuracy": 0.18673118352890014,
      "num_tokens": 39634286.0,
      "step": 17305
    },
    {
      "entropy": 5.590369510650635,
      "epoch": 1.7111506524317912,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004711920750315255,
      "loss": 5.369,
      "mean_token_accuracy": 0.18006061762571335,
      "num_tokens": 39646653.0,
      "step": 17310
    },
    {
      "entropy": 5.610978651046753,
      "epoch": 1.7116449189402925,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004711747957507199,
      "loss": 5.2906,
      "mean_token_accuracy": 0.1871691957116127,
      "num_tokens": 39658027.0,
      "step": 17315
    },
    {
      "entropy": 5.638408565521241,
      "epoch": 1.712139185448794,
      "grad_norm": 1.515625,
      "learning_rate": 0.00047115751164398395,
      "loss": 5.3227,
      "mean_token_accuracy": 0.18452032208442687,
      "num_tokens": 39669912.0,
      "step": 17320
    },
    {
      "entropy": 5.520593500137329,
      "epoch": 1.7126334519572954,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004711402227117427,
      "loss": 5.3015,
      "mean_token_accuracy": 0.1873263418674469,
      "num_tokens": 39680857.0,
      "step": 17325
    },
    {
      "entropy": 5.610478162765503,
      "epoch": 1.7131277184657967,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047112292895442156,
      "loss": 5.2755,
      "mean_token_accuracy": 0.19037665128707887,
      "num_tokens": 39692301.0,
      "step": 17330
    },
    {
      "entropy": 5.624378061294555,
      "epoch": 1.7136219849742982,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000471105630372446,
      "loss": 5.3229,
      "mean_token_accuracy": 0.1820497214794159,
      "num_tokens": 39703641.0,
      "step": 17335
    },
    {
      "entropy": 5.589098167419434,
      "epoch": 1.7141162514827997,
      "grad_norm": 1.625,
      "learning_rate": 0.00047108832696624145,
      "loss": 5.2684,
      "mean_token_accuracy": 0.1971456900238991,
      "num_tokens": 39714919.0,
      "step": 17340
    },
    {
      "entropy": 5.574347686767578,
      "epoch": 1.714610517991301,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004710710187362336,
      "loss": 5.2917,
      "mean_token_accuracy": 0.1871811717748642,
      "num_tokens": 39725658.0,
      "step": 17345
    },
    {
      "entropy": 5.559394073486328,
      "epoch": 1.7151047844998022,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047105370568284834,
      "loss": 5.2285,
      "mean_token_accuracy": 0.19985225200653076,
      "num_tokens": 39736899.0,
      "step": 17350
    },
    {
      "entropy": 5.624191617965698,
      "epoch": 1.7155990510083037,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004710363878065115,
      "loss": 5.3855,
      "mean_token_accuracy": 0.17339082807302475,
      "num_tokens": 39747875.0,
      "step": 17355
    },
    {
      "entropy": 5.688376426696777,
      "epoch": 1.7160933175168052,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004710190651076491,
      "loss": 5.3452,
      "mean_token_accuracy": 0.18529719859361649,
      "num_tokens": 39760253.0,
      "step": 17360
    },
    {
      "entropy": 5.62997407913208,
      "epoch": 1.7165875840253064,
      "grad_norm": 1.25,
      "learning_rate": 0.0004710017375866872,
      "loss": 5.3118,
      "mean_token_accuracy": 0.18715773969888688,
      "num_tokens": 39771031.0,
      "step": 17365
    },
    {
      "entropy": 5.48432297706604,
      "epoch": 1.7170818505338077,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047098440524405227,
      "loss": 5.2458,
      "mean_token_accuracy": 0.19431071132421493,
      "num_tokens": 39780850.0,
      "step": 17370
    },
    {
      "entropy": 5.607624244689942,
      "epoch": 1.7175761170423092,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004709670680801705,
      "loss": 5.3411,
      "mean_token_accuracy": 0.19476141929626464,
      "num_tokens": 39792975.0,
      "step": 17375
    },
    {
      "entropy": 5.631660604476929,
      "epoch": 1.7180703835508107,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004709497260954685,
      "loss": 5.3154,
      "mean_token_accuracy": 0.19129008650779725,
      "num_tokens": 39803574.0,
      "step": 17380
    },
    {
      "entropy": 5.542653369903564,
      "epoch": 1.718564650059312,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047093237929037284,
      "loss": 5.2919,
      "mean_token_accuracy": 0.18613855987787248,
      "num_tokens": 39814907.0,
      "step": 17385
    },
    {
      "entropy": 5.583449792861939,
      "epoch": 1.7190589165678134,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004709150276653101,
      "loss": 5.273,
      "mean_token_accuracy": 0.1841133266687393,
      "num_tokens": 39826344.0,
      "step": 17390
    },
    {
      "entropy": 5.594506120681762,
      "epoch": 1.7195531830763149,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047089767122070756,
      "loss": 5.307,
      "mean_token_accuracy": 0.18551286458969116,
      "num_tokens": 39838516.0,
      "step": 17395
    },
    {
      "entropy": 5.831244707107544,
      "epoch": 1.7200474495848161,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047088030995699173,
      "loss": 5.5101,
      "mean_token_accuracy": 0.17570637315511703,
      "num_tokens": 39850757.0,
      "step": 17400
    },
    {
      "entropy": 5.704502773284912,
      "epoch": 1.7205417160933174,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004708629438745899,
      "loss": 5.3871,
      "mean_token_accuracy": 0.1781456157565117,
      "num_tokens": 39861105.0,
      "step": 17405
    },
    {
      "entropy": 5.628919792175293,
      "epoch": 1.721035982601819,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004708455729739293,
      "loss": 5.3607,
      "mean_token_accuracy": 0.1796714797616005,
      "num_tokens": 39873999.0,
      "step": 17410
    },
    {
      "entropy": 5.579667377471924,
      "epoch": 1.7215302491103204,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047082819725543726,
      "loss": 5.3267,
      "mean_token_accuracy": 0.1815303936600685,
      "num_tokens": 39885768.0,
      "step": 17415
    },
    {
      "entropy": 5.73689341545105,
      "epoch": 1.7220245156188216,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004708108167195411,
      "loss": 5.3578,
      "mean_token_accuracy": 0.18284257799386977,
      "num_tokens": 39897548.0,
      "step": 17420
    },
    {
      "entropy": 5.575407457351685,
      "epoch": 1.722518782127323,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0004707934313666685,
      "loss": 5.277,
      "mean_token_accuracy": 0.19429749846458436,
      "num_tokens": 39909192.0,
      "step": 17425
    },
    {
      "entropy": 5.582723236083984,
      "epoch": 1.7230130486358244,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004707760411972471,
      "loss": 5.2786,
      "mean_token_accuracy": 0.18955908715724945,
      "num_tokens": 39921279.0,
      "step": 17430
    },
    {
      "entropy": 5.62904462814331,
      "epoch": 1.7235073151443259,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004707586462117048,
      "loss": 5.3231,
      "mean_token_accuracy": 0.18266988098621367,
      "num_tokens": 39931638.0,
      "step": 17435
    },
    {
      "entropy": 5.656722688674927,
      "epoch": 1.7240015816528271,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047074124641046934,
      "loss": 5.3128,
      "mean_token_accuracy": 0.18738672584295274,
      "num_tokens": 39944239.0,
      "step": 17440
    },
    {
      "entropy": 5.688676738739014,
      "epoch": 1.7244958481613286,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004707238417939689,
      "loss": 5.3812,
      "mean_token_accuracy": 0.1788727968931198,
      "num_tokens": 39956486.0,
      "step": 17445
    },
    {
      "entropy": 5.446270942687988,
      "epoch": 1.72499011466983,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004707064323626316,
      "loss": 5.1634,
      "mean_token_accuracy": 0.196450312435627,
      "num_tokens": 39966271.0,
      "step": 17450
    },
    {
      "entropy": 5.569846725463867,
      "epoch": 1.7254843811783314,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047068901811688564,
      "loss": 5.3496,
      "mean_token_accuracy": 0.18431257903575898,
      "num_tokens": 39978810.0,
      "step": 17455
    },
    {
      "entropy": 5.5953551769256595,
      "epoch": 1.7259786476868326,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004706715990571595,
      "loss": 5.2432,
      "mean_token_accuracy": 0.1884813740849495,
      "num_tokens": 39989331.0,
      "step": 17460
    },
    {
      "entropy": 5.5677149295806885,
      "epoch": 1.726472914195334,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004706541751838817,
      "loss": 5.2346,
      "mean_token_accuracy": 0.189345982670784,
      "num_tokens": 39999931.0,
      "step": 17465
    },
    {
      "entropy": 5.599851417541504,
      "epoch": 1.7269671807038356,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004706367464974808,
      "loss": 5.2322,
      "mean_token_accuracy": 0.18586724251508713,
      "num_tokens": 40010702.0,
      "step": 17470
    },
    {
      "entropy": 5.6359021186828615,
      "epoch": 1.7274614472123369,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047061931299838554,
      "loss": 5.3409,
      "mean_token_accuracy": 0.18703936487436296,
      "num_tokens": 40022897.0,
      "step": 17475
    },
    {
      "entropy": 5.550184345245361,
      "epoch": 1.7279557137208381,
      "grad_norm": 1.453125,
      "learning_rate": 0.000470601874687025,
      "loss": 5.2823,
      "mean_token_accuracy": 0.18598002195358276,
      "num_tokens": 40034057.0,
      "step": 17480
    },
    {
      "entropy": 5.614707040786743,
      "epoch": 1.7284499802293398,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004705844315638279,
      "loss": 5.3032,
      "mean_token_accuracy": 0.18859024047851564,
      "num_tokens": 40044595.0,
      "step": 17485
    },
    {
      "entropy": 5.629191207885742,
      "epoch": 1.728944246737841,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004705669836292234,
      "loss": 5.2644,
      "mean_token_accuracy": 0.18638714998960496,
      "num_tokens": 40055175.0,
      "step": 17490
    },
    {
      "entropy": 5.557345151901245,
      "epoch": 1.7294385132463423,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004705495308836408,
      "loss": 5.1877,
      "mean_token_accuracy": 0.1965081125497818,
      "num_tokens": 40065276.0,
      "step": 17495
    },
    {
      "entropy": 5.616838836669922,
      "epoch": 1.7299327797548438,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004705320733275094,
      "loss": 5.336,
      "mean_token_accuracy": 0.1858326554298401,
      "num_tokens": 40077230.0,
      "step": 17500
    },
    {
      "entropy": 5.531122255325317,
      "epoch": 1.7304270462633453,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004705146109612587,
      "loss": 5.1687,
      "mean_token_accuracy": 0.19542232453823088,
      "num_tokens": 40088058.0,
      "step": 17505
    },
    {
      "entropy": 5.479264450073242,
      "epoch": 1.7309213127718466,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004704971437853182,
      "loss": 5.2536,
      "mean_token_accuracy": 0.18729754239320756,
      "num_tokens": 40099618.0,
      "step": 17510
    },
    {
      "entropy": 5.636681604385376,
      "epoch": 1.7314155792803478,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004704796718001176,
      "loss": 5.3534,
      "mean_token_accuracy": 0.19434620887041093,
      "num_tokens": 40111088.0,
      "step": 17515
    },
    {
      "entropy": 5.655070877075195,
      "epoch": 1.7319098457888493,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047046219500608685,
      "loss": 5.3281,
      "mean_token_accuracy": 0.1856656551361084,
      "num_tokens": 40122191.0,
      "step": 17520
    },
    {
      "entropy": 5.606954717636109,
      "epoch": 1.7324041122973508,
      "grad_norm": 1.25,
      "learning_rate": 0.0004704447134036557,
      "loss": 5.3506,
      "mean_token_accuracy": 0.17780194580554962,
      "num_tokens": 40133764.0,
      "step": 17525
    },
    {
      "entropy": 5.582091045379639,
      "epoch": 1.732898378805852,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047042722699325426,
      "loss": 5.2376,
      "mean_token_accuracy": 0.19495914727449418,
      "num_tokens": 40144502.0,
      "step": 17530
    },
    {
      "entropy": 5.591317462921142,
      "epoch": 1.7333926453143536,
      "grad_norm": 1.375,
      "learning_rate": 0.0004704097357753128,
      "loss": 5.2876,
      "mean_token_accuracy": 0.18989259749650955,
      "num_tokens": 40155606.0,
      "step": 17535
    },
    {
      "entropy": 5.542151260375976,
      "epoch": 1.733886911822855,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004703922397502614,
      "loss": 5.2254,
      "mean_token_accuracy": 0.19372633397579192,
      "num_tokens": 40166551.0,
      "step": 17540
    },
    {
      "entropy": 5.652794981002808,
      "epoch": 1.7343811783313563,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047037473891853066,
      "loss": 5.4524,
      "mean_token_accuracy": 0.18176266849040984,
      "num_tokens": 40178804.0,
      "step": 17545
    },
    {
      "entropy": 5.646605539321899,
      "epoch": 1.7348754448398576,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047035723328055104,
      "loss": 5.3123,
      "mean_token_accuracy": 0.1861835315823555,
      "num_tokens": 40189861.0,
      "step": 17550
    },
    {
      "entropy": 5.54687819480896,
      "epoch": 1.735369711348359,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004703397228367532,
      "loss": 5.2289,
      "mean_token_accuracy": 0.1919678419828415,
      "num_tokens": 40202619.0,
      "step": 17555
    },
    {
      "entropy": 5.542559719085693,
      "epoch": 1.7358639778568605,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047032220758756786,
      "loss": 5.2547,
      "mean_token_accuracy": 0.18909469544887542,
      "num_tokens": 40213509.0,
      "step": 17560
    },
    {
      "entropy": 5.6481256008148195,
      "epoch": 1.7363582443653618,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004703046875334259,
      "loss": 5.3139,
      "mean_token_accuracy": 0.18515729606151582,
      "num_tokens": 40224960.0,
      "step": 17565
    },
    {
      "entropy": 5.597984170913696,
      "epoch": 1.736852510873863,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047028716267475827,
      "loss": 5.2978,
      "mean_token_accuracy": 0.1903235450387001,
      "num_tokens": 40238516.0,
      "step": 17570
    },
    {
      "entropy": 5.625869846343994,
      "epoch": 1.7373467773823645,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004702696330119963,
      "loss": 5.3755,
      "mean_token_accuracy": 0.1815244063735008,
      "num_tokens": 40249123.0,
      "step": 17575
    },
    {
      "entropy": 5.6585776805877686,
      "epoch": 1.737841043890866,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004702520985455709,
      "loss": 5.3214,
      "mean_token_accuracy": 0.19243215918540954,
      "num_tokens": 40260168.0,
      "step": 17580
    },
    {
      "entropy": 5.6365001678466795,
      "epoch": 1.7383353103993673,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047023455927591365,
      "loss": 5.386,
      "mean_token_accuracy": 0.17834709733724594,
      "num_tokens": 40273183.0,
      "step": 17585
    },
    {
      "entropy": 5.615384292602539,
      "epoch": 1.7388295769078688,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047021701520345594,
      "loss": 5.3319,
      "mean_token_accuracy": 0.18291830271482468,
      "num_tokens": 40284452.0,
      "step": 17590
    },
    {
      "entropy": 5.592526245117187,
      "epoch": 1.7393238434163703,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047019946632862944,
      "loss": 5.2221,
      "mean_token_accuracy": 0.19508425295352935,
      "num_tokens": 40295267.0,
      "step": 17595
    },
    {
      "entropy": 5.554788827896118,
      "epoch": 1.7398181099248715,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047018191265186566,
      "loss": 5.2266,
      "mean_token_accuracy": 0.19118803441524507,
      "num_tokens": 40306766.0,
      "step": 17600
    },
    {
      "entropy": 5.570209789276123,
      "epoch": 1.7403123764333728,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047016435417359667,
      "loss": 5.2406,
      "mean_token_accuracy": 0.1879921957850456,
      "num_tokens": 40317382.0,
      "step": 17605
    },
    {
      "entropy": 5.579920673370362,
      "epoch": 1.7408066429418743,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047014679089425423,
      "loss": 5.3162,
      "mean_token_accuracy": 0.1875707507133484,
      "num_tokens": 40328273.0,
      "step": 17610
    },
    {
      "entropy": 5.622617149353028,
      "epoch": 1.7413009094503757,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047012922281427046,
      "loss": 5.3317,
      "mean_token_accuracy": 0.18957780450582504,
      "num_tokens": 40340550.0,
      "step": 17615
    },
    {
      "entropy": 5.690776443481445,
      "epoch": 1.741795175958877,
      "grad_norm": 1.875,
      "learning_rate": 0.00047011164993407753,
      "loss": 5.366,
      "mean_token_accuracy": 0.1877758741378784,
      "num_tokens": 40351752.0,
      "step": 17620
    },
    {
      "entropy": 5.564198160171509,
      "epoch": 1.7422894424673783,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00047009407225410776,
      "loss": 5.2312,
      "mean_token_accuracy": 0.19071638584136963,
      "num_tokens": 40363330.0,
      "step": 17625
    },
    {
      "entropy": 5.507275009155274,
      "epoch": 1.7427837089758798,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004700764897747936,
      "loss": 5.1795,
      "mean_token_accuracy": 0.20213651657104492,
      "num_tokens": 40375674.0,
      "step": 17630
    },
    {
      "entropy": 5.527492570877075,
      "epoch": 1.7432779754843812,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004700589024965674,
      "loss": 5.2826,
      "mean_token_accuracy": 0.18857021033763885,
      "num_tokens": 40387120.0,
      "step": 17635
    },
    {
      "entropy": 5.586859512329101,
      "epoch": 1.7437722419928825,
      "grad_norm": 1.25,
      "learning_rate": 0.000470041310419862,
      "loss": 5.2899,
      "mean_token_accuracy": 0.19089189320802688,
      "num_tokens": 40399196.0,
      "step": 17640
    },
    {
      "entropy": 5.596062469482422,
      "epoch": 1.744266508501384,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004700237135451101,
      "loss": 5.2864,
      "mean_token_accuracy": 0.1881484404206276,
      "num_tokens": 40411693.0,
      "step": 17645
    },
    {
      "entropy": 5.50392255783081,
      "epoch": 1.7447607750098855,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004700061118727446,
      "loss": 5.1799,
      "mean_token_accuracy": 0.201791650056839,
      "num_tokens": 40422866.0,
      "step": 17650
    },
    {
      "entropy": 5.556821727752686,
      "epoch": 1.7452550415183867,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004699885054031985,
      "loss": 5.3388,
      "mean_token_accuracy": 0.19352139234542848,
      "num_tokens": 40435028.0,
      "step": 17655
    },
    {
      "entropy": 5.692240762710571,
      "epoch": 1.745749308026888,
      "grad_norm": 1.421875,
      "learning_rate": 0.000469970894136905,
      "loss": 5.4389,
      "mean_token_accuracy": 0.18137546479701996,
      "num_tokens": 40446802.0,
      "step": 17660
    },
    {
      "entropy": 5.704209566116333,
      "epoch": 1.7462435745353895,
      "grad_norm": 1.421875,
      "learning_rate": 0.00046995327807429715,
      "loss": 5.3588,
      "mean_token_accuracy": 0.1850094825029373,
      "num_tokens": 40459025.0,
      "step": 17665
    },
    {
      "entropy": 5.663491249084473,
      "epoch": 1.746737841043891,
      "grad_norm": 1.5,
      "learning_rate": 0.00046993565721580846,
      "loss": 5.3921,
      "mean_token_accuracy": 0.17885632663965226,
      "num_tokens": 40471486.0,
      "step": 17670
    },
    {
      "entropy": 5.588140773773193,
      "epoch": 1.7472321075523922,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004699180315618723,
      "loss": 5.2497,
      "mean_token_accuracy": 0.18945779651403427,
      "num_tokens": 40483163.0,
      "step": 17675
    },
    {
      "entropy": 5.590656995773315,
      "epoch": 1.7477263740608935,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004699004011129224,
      "loss": 5.2453,
      "mean_token_accuracy": 0.1831422969698906,
      "num_tokens": 40494475.0,
      "step": 17680
    },
    {
      "entropy": 5.529447078704834,
      "epoch": 1.748220640569395,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046988276586939234,
      "loss": 5.2886,
      "mean_token_accuracy": 0.18727412223815917,
      "num_tokens": 40504557.0,
      "step": 17685
    },
    {
      "entropy": 5.521021842956543,
      "epoch": 1.7487149070778965,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004698651258317161,
      "loss": 5.2393,
      "mean_token_accuracy": 0.1954387068748474,
      "num_tokens": 40516147.0,
      "step": 17690
    },
    {
      "entropy": 5.563278532028198,
      "epoch": 1.7492091735863977,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004698474810003274,
      "loss": 5.2377,
      "mean_token_accuracy": 0.19631812423467637,
      "num_tokens": 40527238.0,
      "step": 17695
    },
    {
      "entropy": 5.5441900253295895,
      "epoch": 1.7497034400948992,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004698298313756605,
      "loss": 5.2054,
      "mean_token_accuracy": 0.1930094137787819,
      "num_tokens": 40537470.0,
      "step": 17700
    },
    {
      "entropy": 5.543980407714844,
      "epoch": 1.7501977066034007,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046981217695814953,
      "loss": 5.2303,
      "mean_token_accuracy": 0.19392044693231583,
      "num_tokens": 40549865.0,
      "step": 17705
    },
    {
      "entropy": 5.655466508865357,
      "epoch": 1.750691973111902,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004697945177482287,
      "loss": 5.3484,
      "mean_token_accuracy": 0.17952707260847092,
      "num_tokens": 40561796.0,
      "step": 17710
    },
    {
      "entropy": 5.545645618438721,
      "epoch": 1.7511862396204032,
      "grad_norm": 1.296875,
      "learning_rate": 0.00046977685374633267,
      "loss": 5.2035,
      "mean_token_accuracy": 0.19220599681138992,
      "num_tokens": 40574597.0,
      "step": 17715
    },
    {
      "entropy": 5.5153602123260494,
      "epoch": 1.7516805061289047,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046975918495289576,
      "loss": 5.2882,
      "mean_token_accuracy": 0.18653754889965057,
      "num_tokens": 40586549.0,
      "step": 17720
    },
    {
      "entropy": 5.571998977661133,
      "epoch": 1.7521747726374062,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004697415113683527,
      "loss": 5.2617,
      "mean_token_accuracy": 0.19211031794548034,
      "num_tokens": 40597808.0,
      "step": 17725
    },
    {
      "entropy": 5.673518419265747,
      "epoch": 1.7526690391459074,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046972383299313813,
      "loss": 5.4096,
      "mean_token_accuracy": 0.18321943581104277,
      "num_tokens": 40608636.0,
      "step": 17730
    },
    {
      "entropy": 5.638261222839356,
      "epoch": 1.7531633056544087,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004697061498276872,
      "loss": 5.2918,
      "mean_token_accuracy": 0.1893991634249687,
      "num_tokens": 40619099.0,
      "step": 17735
    },
    {
      "entropy": 5.526123428344727,
      "epoch": 1.7536575721629104,
      "grad_norm": 1.515625,
      "learning_rate": 0.00046968846187243473,
      "loss": 5.2339,
      "mean_token_accuracy": 0.1968296140432358,
      "num_tokens": 40630436.0,
      "step": 17740
    },
    {
      "entropy": 5.561237144470215,
      "epoch": 1.7541518386714117,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046967076912781585,
      "loss": 5.3045,
      "mean_token_accuracy": 0.1918693706393242,
      "num_tokens": 40643078.0,
      "step": 17745
    },
    {
      "entropy": 5.586308002471924,
      "epoch": 1.754646105179913,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046965307159426594,
      "loss": 5.247,
      "mean_token_accuracy": 0.1936090588569641,
      "num_tokens": 40653368.0,
      "step": 17750
    },
    {
      "entropy": 5.581864833831787,
      "epoch": 1.7551403716884144,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004696353692722202,
      "loss": 5.2931,
      "mean_token_accuracy": 0.18677907586097717,
      "num_tokens": 40665377.0,
      "step": 17755
    },
    {
      "entropy": 5.620696687698365,
      "epoch": 1.755634638196916,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004696176621621143,
      "loss": 5.3719,
      "mean_token_accuracy": 0.18811169117689133,
      "num_tokens": 40676599.0,
      "step": 17760
    },
    {
      "entropy": 5.601365947723389,
      "epoch": 1.7561289047054172,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004695999502643836,
      "loss": 5.2585,
      "mean_token_accuracy": 0.197500604391098,
      "num_tokens": 40687619.0,
      "step": 17765
    },
    {
      "entropy": 5.526253795623779,
      "epoch": 1.7566231712139184,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046958223357946395,
      "loss": 5.2577,
      "mean_token_accuracy": 0.19111137539148332,
      "num_tokens": 40699410.0,
      "step": 17770
    },
    {
      "entropy": 5.60101637840271,
      "epoch": 1.75711743772242,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046956451210779113,
      "loss": 5.2971,
      "mean_token_accuracy": 0.18093891590833663,
      "num_tokens": 40711149.0,
      "step": 17775
    },
    {
      "entropy": 5.5357701778411865,
      "epoch": 1.7576117042309214,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046954678584980113,
      "loss": 5.2769,
      "mean_token_accuracy": 0.19323269724845887,
      "num_tokens": 40722997.0,
      "step": 17780
    },
    {
      "entropy": 5.644488525390625,
      "epoch": 1.7581059707394227,
      "grad_norm": 1.65625,
      "learning_rate": 0.00046952905480593014,
      "loss": 5.3274,
      "mean_token_accuracy": 0.1910620853304863,
      "num_tokens": 40733786.0,
      "step": 17785
    },
    {
      "entropy": 5.643467903137207,
      "epoch": 1.758600237247924,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00046951131897661415,
      "loss": 5.3881,
      "mean_token_accuracy": 0.18358676135540009,
      "num_tokens": 40745925.0,
      "step": 17790
    },
    {
      "entropy": 5.64861044883728,
      "epoch": 1.7590945037564256,
      "grad_norm": 1.25,
      "learning_rate": 0.00046949357836228946,
      "loss": 5.3557,
      "mean_token_accuracy": 0.18385769575834274,
      "num_tokens": 40760031.0,
      "step": 17795
    },
    {
      "entropy": 5.606079721450806,
      "epoch": 1.7595887702649269,
      "grad_norm": 1.375,
      "learning_rate": 0.00046947583296339266,
      "loss": 5.3676,
      "mean_token_accuracy": 0.18572872132062912,
      "num_tokens": 40771557.0,
      "step": 17800
    },
    {
      "entropy": 5.600269317626953,
      "epoch": 1.7600830367734281,
      "grad_norm": 1.4375,
      "learning_rate": 0.00046945808278036017,
      "loss": 5.3521,
      "mean_token_accuracy": 0.1796276643872261,
      "num_tokens": 40783557.0,
      "step": 17805
    },
    {
      "entropy": 5.607090711593628,
      "epoch": 1.7605773032819296,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004694403278136287,
      "loss": 5.1765,
      "mean_token_accuracy": 0.2001905098557472,
      "num_tokens": 40794213.0,
      "step": 17810
    },
    {
      "entropy": 5.480912780761718,
      "epoch": 1.7610715697904311,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00046942256806363497,
      "loss": 5.2025,
      "mean_token_accuracy": 0.19493167102336884,
      "num_tokens": 40805753.0,
      "step": 17815
    },
    {
      "entropy": 5.588007402420044,
      "epoch": 1.7615658362989324,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046940480353081587,
      "loss": 5.3982,
      "mean_token_accuracy": 0.18515846580266954,
      "num_tokens": 40818315.0,
      "step": 17820
    },
    {
      "entropy": 5.572773551940918,
      "epoch": 1.7620601028074336,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004693870342156085,
      "loss": 5.2151,
      "mean_token_accuracy": 0.2017627850174904,
      "num_tokens": 40828740.0,
      "step": 17825
    },
    {
      "entropy": 5.560178565979004,
      "epoch": 1.7625543693159351,
      "grad_norm": 2.203125,
      "learning_rate": 0.0004693692601184499,
      "loss": 5.3196,
      "mean_token_accuracy": 0.18602558225393295,
      "num_tokens": 40841035.0,
      "step": 17830
    },
    {
      "entropy": 5.544559144973755,
      "epoch": 1.7630486358244366,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004693514812397774,
      "loss": 5.276,
      "mean_token_accuracy": 0.1928926795721054,
      "num_tokens": 40851857.0,
      "step": 17835
    },
    {
      "entropy": 5.655282020568848,
      "epoch": 1.7635429023329379,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004693336975800283,
      "loss": 5.3279,
      "mean_token_accuracy": 0.1861262395977974,
      "num_tokens": 40863964.0,
      "step": 17840
    },
    {
      "entropy": 5.622573041915894,
      "epoch": 1.7640371688414394,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046931590913964013,
      "loss": 5.3072,
      "mean_token_accuracy": 0.18961825519800185,
      "num_tokens": 40876076.0,
      "step": 17845
    },
    {
      "entropy": 5.6621355533599855,
      "epoch": 1.7645314353499408,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046929811591905034,
      "loss": 5.391,
      "mean_token_accuracy": 0.18338590413331984,
      "num_tokens": 40887839.0,
      "step": 17850
    },
    {
      "entropy": 5.636042261123658,
      "epoch": 1.765025701858442,
      "grad_norm": 1.25,
      "learning_rate": 0.0004692803179186968,
      "loss": 5.356,
      "mean_token_accuracy": 0.18774719685316085,
      "num_tokens": 40898978.0,
      "step": 17855
    },
    {
      "entropy": 5.660222101211548,
      "epoch": 1.7655199683669434,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004692625151390173,
      "loss": 5.2867,
      "mean_token_accuracy": 0.18315183222293854,
      "num_tokens": 40910990.0,
      "step": 17860
    },
    {
      "entropy": 5.522454023361206,
      "epoch": 1.7660142348754448,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004692447075804498,
      "loss": 5.2366,
      "mean_token_accuracy": 0.19687196463346482,
      "num_tokens": 40922299.0,
      "step": 17865
    },
    {
      "entropy": 5.601914930343628,
      "epoch": 1.7665085013839463,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004692268952434324,
      "loss": 5.3086,
      "mean_token_accuracy": 0.18722081780433655,
      "num_tokens": 40933181.0,
      "step": 17870
    },
    {
      "entropy": 5.551723337173462,
      "epoch": 1.7670027678924476,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004692090781284032,
      "loss": 5.3011,
      "mean_token_accuracy": 0.18664287328720092,
      "num_tokens": 40944101.0,
      "step": 17875
    },
    {
      "entropy": 5.617815780639648,
      "epoch": 1.7674970344009489,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004691912562358006,
      "loss": 5.2959,
      "mean_token_accuracy": 0.1888287827372551,
      "num_tokens": 40956492.0,
      "step": 17880
    },
    {
      "entropy": 5.656340646743774,
      "epoch": 1.7679913009094503,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046917342956606287,
      "loss": 5.3508,
      "mean_token_accuracy": 0.18149456083774568,
      "num_tokens": 40967477.0,
      "step": 17885
    },
    {
      "entropy": 5.63769760131836,
      "epoch": 1.7684855674179518,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004691555981196287,
      "loss": 5.3563,
      "mean_token_accuracy": 0.18836382478475572,
      "num_tokens": 40978756.0,
      "step": 17890
    },
    {
      "entropy": 5.561192035675049,
      "epoch": 1.768979833926453,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004691377618969367,
      "loss": 5.309,
      "mean_token_accuracy": 0.18930543810129166,
      "num_tokens": 40990583.0,
      "step": 17895
    },
    {
      "entropy": 5.577971363067627,
      "epoch": 1.7694741004349546,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046911992089842565,
      "loss": 5.2247,
      "mean_token_accuracy": 0.1916408732533455,
      "num_tokens": 41002197.0,
      "step": 17900
    },
    {
      "entropy": 5.565328741073609,
      "epoch": 1.769968366943456,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004691020751245344,
      "loss": 5.3145,
      "mean_token_accuracy": 0.18819564282894136,
      "num_tokens": 41014038.0,
      "step": 17905
    },
    {
      "entropy": 5.556237030029297,
      "epoch": 1.7704626334519573,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000469084224575702,
      "loss": 5.2391,
      "mean_token_accuracy": 0.19091471135616303,
      "num_tokens": 41025752.0,
      "step": 17910
    },
    {
      "entropy": 5.554512023925781,
      "epoch": 1.7709568999604586,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004690663692523676,
      "loss": 5.2766,
      "mean_token_accuracy": 0.19569144695997237,
      "num_tokens": 41035974.0,
      "step": 17915
    },
    {
      "entropy": 5.561146450042725,
      "epoch": 1.77145116646896,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046904850915497037,
      "loss": 5.2441,
      "mean_token_accuracy": 0.18578066527843476,
      "num_tokens": 41047669.0,
      "step": 17920
    },
    {
      "entropy": 5.6602232456207275,
      "epoch": 1.7719454329774615,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004690306442839497,
      "loss": 5.3399,
      "mean_token_accuracy": 0.18088099211454392,
      "num_tokens": 41059123.0,
      "step": 17925
    },
    {
      "entropy": 5.528279161453247,
      "epoch": 1.7724396994859628,
      "grad_norm": 1.25,
      "learning_rate": 0.0004690127746397451,
      "loss": 5.2243,
      "mean_token_accuracy": 0.19681421369314195,
      "num_tokens": 41070382.0,
      "step": 17930
    },
    {
      "entropy": 5.582224130630493,
      "epoch": 1.772933965994464,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046899490022279613,
      "loss": 5.303,
      "mean_token_accuracy": 0.19314546436071395,
      "num_tokens": 41081672.0,
      "step": 17935
    },
    {
      "entropy": 5.68007607460022,
      "epoch": 1.7734282325029656,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046897702103354256,
      "loss": 5.3072,
      "mean_token_accuracy": 0.19346420019865035,
      "num_tokens": 41092448.0,
      "step": 17940
    },
    {
      "entropy": 5.614037275314331,
      "epoch": 1.773922499011467,
      "grad_norm": 1.578125,
      "learning_rate": 0.00046895913707242423,
      "loss": 5.2501,
      "mean_token_accuracy": 0.1935768410563469,
      "num_tokens": 41103316.0,
      "step": 17945
    },
    {
      "entropy": 5.550587177276611,
      "epoch": 1.7744167655199683,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004689412483398809,
      "loss": 5.2551,
      "mean_token_accuracy": 0.19109326750040054,
      "num_tokens": 41114991.0,
      "step": 17950
    },
    {
      "entropy": 5.57541823387146,
      "epoch": 1.7749110320284698,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046892335483635284,
      "loss": 5.2308,
      "mean_token_accuracy": 0.18891442567110062,
      "num_tokens": 41126693.0,
      "step": 17955
    },
    {
      "entropy": 5.603558301925659,
      "epoch": 1.7754052985369713,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004689054565622801,
      "loss": 5.2821,
      "mean_token_accuracy": 0.1939552679657936,
      "num_tokens": 41138344.0,
      "step": 17960
    },
    {
      "entropy": 5.5450798034667965,
      "epoch": 1.7758995650454725,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004688875535181032,
      "loss": 5.2758,
      "mean_token_accuracy": 0.1896183580160141,
      "num_tokens": 41148962.0,
      "step": 17965
    },
    {
      "entropy": 5.525328159332275,
      "epoch": 1.7763938315539738,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046886964570426226,
      "loss": 5.1638,
      "mean_token_accuracy": 0.19493757635354997,
      "num_tokens": 41159317.0,
      "step": 17970
    },
    {
      "entropy": 5.621600914001465,
      "epoch": 1.7768880980624753,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004688517331211981,
      "loss": 5.2586,
      "mean_token_accuracy": 0.19220978170633315,
      "num_tokens": 41170646.0,
      "step": 17975
    },
    {
      "entropy": 5.56419267654419,
      "epoch": 1.7773823645709768,
      "grad_norm": 1.25,
      "learning_rate": 0.0004688338157693511,
      "loss": 5.3111,
      "mean_token_accuracy": 0.1828644186258316,
      "num_tokens": 41182868.0,
      "step": 17980
    },
    {
      "entropy": 5.507730484008789,
      "epoch": 1.777876631079478,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046881589364916224,
      "loss": 5.2121,
      "mean_token_accuracy": 0.19631679654121398,
      "num_tokens": 41194556.0,
      "step": 17985
    },
    {
      "entropy": 5.652327251434326,
      "epoch": 1.7783708975879793,
      "grad_norm": 1.25,
      "learning_rate": 0.0004687979667610722,
      "loss": 5.2741,
      "mean_token_accuracy": 0.19276719391345978,
      "num_tokens": 41207082.0,
      "step": 17990
    },
    {
      "entropy": 5.597120857238769,
      "epoch": 1.778865164096481,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00046878003510552224,
      "loss": 5.2631,
      "mean_token_accuracy": 0.19122608453035356,
      "num_tokens": 41217436.0,
      "step": 17995
    },
    {
      "entropy": 5.532157897949219,
      "epoch": 1.7793594306049823,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004687620986829533,
      "loss": 5.2523,
      "mean_token_accuracy": 0.18658482879400254,
      "num_tokens": 41228750.0,
      "step": 18000
    },
    {
      "epoch": 1.7793594306049823,
      "eval_entropy": 5.330345546832677,
      "eval_loss": 5.35257625579834,
      "eval_mean_token_accuracy": 0.19242417482798976,
      "eval_num_tokens": 41228750.0,
      "eval_runtime": 26.5963,
      "eval_samples_per_second": 1222.464,
      "eval_steps_per_second": 152.841,
      "step": 18000
    },
    {
      "entropy": 5.541408967971802,
      "epoch": 1.7798536971134835,
      "grad_norm": 1.4375,
      "learning_rate": 0.00046874415749380673,
      "loss": 5.2687,
      "mean_token_accuracy": 0.18772680163383484,
      "num_tokens": 41239110.0,
      "step": 18005
    },
    {
      "entropy": 5.622101354598999,
      "epoch": 1.780347963621985,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004687262115385237,
      "loss": 5.2924,
      "mean_token_accuracy": 0.1887261003255844,
      "num_tokens": 41251239.0,
      "step": 18010
    },
    {
      "entropy": 5.636102485656738,
      "epoch": 1.7808422301304865,
      "grad_norm": 1.234375,
      "learning_rate": 0.000468708260817546,
      "loss": 5.3234,
      "mean_token_accuracy": 0.1871577426791191,
      "num_tokens": 41263996.0,
      "step": 18015
    },
    {
      "entropy": 5.680250120162964,
      "epoch": 1.7813364966389877,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00046869030533131493,
      "loss": 5.3904,
      "mean_token_accuracy": 0.18110288679599762,
      "num_tokens": 41277210.0,
      "step": 18020
    },
    {
      "entropy": 5.559522914886474,
      "epoch": 1.781830763147489,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004686723450802723,
      "loss": 5.2895,
      "mean_token_accuracy": 0.18636482954025269,
      "num_tokens": 41288383.0,
      "step": 18025
    },
    {
      "entropy": 5.544947814941406,
      "epoch": 1.7823250296559905,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046865438006486,
      "loss": 5.2633,
      "mean_token_accuracy": 0.18996478915214537,
      "num_tokens": 41300970.0,
      "step": 18030
    },
    {
      "entropy": 5.650715732574463,
      "epoch": 1.782819296164492,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046863641028551984,
      "loss": 5.3236,
      "mean_token_accuracy": 0.18326372057199478,
      "num_tokens": 41312289.0,
      "step": 18035
    },
    {
      "entropy": 5.570932102203369,
      "epoch": 1.7833135626729932,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046861843574269395,
      "loss": 5.2099,
      "mean_token_accuracy": 0.18888334333896636,
      "num_tokens": 41323562.0,
      "step": 18040
    },
    {
      "entropy": 5.579482316970825,
      "epoch": 1.7838078291814945,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004686004564368245,
      "loss": 5.2525,
      "mean_token_accuracy": 0.1889817014336586,
      "num_tokens": 41335175.0,
      "step": 18045
    },
    {
      "entropy": 5.581819725036621,
      "epoch": 1.7843020956899962,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004685824723683538,
      "loss": 5.2594,
      "mean_token_accuracy": 0.19679524302482604,
      "num_tokens": 41346160.0,
      "step": 18050
    },
    {
      "entropy": 5.6147631168365475,
      "epoch": 1.7847963621984975,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004685644835377242,
      "loss": 5.4026,
      "mean_token_accuracy": 0.17842501103878022,
      "num_tokens": 41358619.0,
      "step": 18055
    },
    {
      "entropy": 5.657603168487549,
      "epoch": 1.7852906287069987,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046854648994537835,
      "loss": 5.2553,
      "mean_token_accuracy": 0.19548789262771607,
      "num_tokens": 41369034.0,
      "step": 18060
    },
    {
      "entropy": 5.63514666557312,
      "epoch": 1.7857848952155002,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004685284915917588,
      "loss": 5.2915,
      "mean_token_accuracy": 0.18678614646196365,
      "num_tokens": 41380700.0,
      "step": 18065
    },
    {
      "entropy": 5.597020769119263,
      "epoch": 1.7862791617240017,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00046851048847730835,
      "loss": 5.3346,
      "mean_token_accuracy": 0.18356600701808928,
      "num_tokens": 41391297.0,
      "step": 18070
    },
    {
      "entropy": 5.616900491714477,
      "epoch": 1.786773428232503,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046849248060246986,
      "loss": 5.3764,
      "mean_token_accuracy": 0.1827303022146225,
      "num_tokens": 41403418.0,
      "step": 18075
    },
    {
      "entropy": 5.651558446884155,
      "epoch": 1.7872676947410042,
      "grad_norm": 1.328125,
      "learning_rate": 0.00046847446796768633,
      "loss": 5.295,
      "mean_token_accuracy": 0.18457084745168686,
      "num_tokens": 41414220.0,
      "step": 18080
    },
    {
      "entropy": 5.6176307678222654,
      "epoch": 1.7877619612495057,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004684564505734008,
      "loss": 5.319,
      "mean_token_accuracy": 0.18934998661279678,
      "num_tokens": 41425933.0,
      "step": 18085
    },
    {
      "entropy": 5.523052167892456,
      "epoch": 1.7882562277580072,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046843842842005665,
      "loss": 5.2598,
      "mean_token_accuracy": 0.18879807293415068,
      "num_tokens": 41438413.0,
      "step": 18090
    },
    {
      "entropy": 5.524044036865234,
      "epoch": 1.7887504942665085,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004684204015080971,
      "loss": 5.2038,
      "mean_token_accuracy": 0.19859365224838257,
      "num_tokens": 41449104.0,
      "step": 18095
    },
    {
      "entropy": 5.631772136688232,
      "epoch": 1.78924476077501,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004684023698379657,
      "loss": 5.4023,
      "mean_token_accuracy": 0.17777305692434311,
      "num_tokens": 41460766.0,
      "step": 18100
    },
    {
      "entropy": 5.654554224014282,
      "epoch": 1.7897390272835114,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004683843334101061,
      "loss": 5.3514,
      "mean_token_accuracy": 0.17843175083398818,
      "num_tokens": 41471872.0,
      "step": 18105
    },
    {
      "entropy": 5.747128486633301,
      "epoch": 1.7902332937920127,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046836629222496174,
      "loss": 5.3528,
      "mean_token_accuracy": 0.18410602062940598,
      "num_tokens": 41483313.0,
      "step": 18110
    },
    {
      "entropy": 5.535778570175171,
      "epoch": 1.790727560300514,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004683482462829766,
      "loss": 5.2054,
      "mean_token_accuracy": 0.19946459978818892,
      "num_tokens": 41492935.0,
      "step": 18115
    },
    {
      "entropy": 5.57140851020813,
      "epoch": 1.7912218268090154,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004683301955845947,
      "loss": 5.3799,
      "mean_token_accuracy": 0.1868179604411125,
      "num_tokens": 41504159.0,
      "step": 18120
    },
    {
      "entropy": 5.574690485000611,
      "epoch": 1.791716093317517,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046831214013025996,
      "loss": 5.2038,
      "mean_token_accuracy": 0.19764899611473083,
      "num_tokens": 41515812.0,
      "step": 18125
    },
    {
      "entropy": 5.5512162208557125,
      "epoch": 1.7922103598260182,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004682940799204165,
      "loss": 5.1252,
      "mean_token_accuracy": 0.20095156580209733,
      "num_tokens": 41527109.0,
      "step": 18130
    },
    {
      "entropy": 5.544265174865723,
      "epoch": 1.7927046263345194,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046827601495550884,
      "loss": 5.2606,
      "mean_token_accuracy": 0.18564793914556504,
      "num_tokens": 41537183.0,
      "step": 18135
    },
    {
      "entropy": 5.548152256011963,
      "epoch": 1.793198892843021,
      "grad_norm": 1.25,
      "learning_rate": 0.0004682579452359811,
      "loss": 5.2213,
      "mean_token_accuracy": 0.1960213914513588,
      "num_tokens": 41548170.0,
      "step": 18140
    },
    {
      "entropy": 5.5665185928344725,
      "epoch": 1.7936931593515224,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000468239870762278,
      "loss": 5.2995,
      "mean_token_accuracy": 0.18844516575336456,
      "num_tokens": 41558737.0,
      "step": 18145
    },
    {
      "entropy": 5.5536998271942135,
      "epoch": 1.7941874258600237,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000468221791534844,
      "loss": 5.2082,
      "mean_token_accuracy": 0.1973348006606102,
      "num_tokens": 41570096.0,
      "step": 18150
    },
    {
      "entropy": 5.51496114730835,
      "epoch": 1.7946816923685251,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000468203707554124,
      "loss": 5.2106,
      "mean_token_accuracy": 0.19616491943597794,
      "num_tokens": 41582670.0,
      "step": 18155
    },
    {
      "entropy": 5.502422952651978,
      "epoch": 1.7951759588770266,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004681856188205629,
      "loss": 5.2035,
      "mean_token_accuracy": 0.1993554025888443,
      "num_tokens": 41594300.0,
      "step": 18160
    },
    {
      "entropy": 5.604350566864014,
      "epoch": 1.795670225385528,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004681675253346055,
      "loss": 5.2849,
      "mean_token_accuracy": 0.18577527701854707,
      "num_tokens": 41606021.0,
      "step": 18165
    },
    {
      "entropy": 5.566544103622436,
      "epoch": 1.7961644918940292,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004681494270966971,
      "loss": 5.2349,
      "mean_token_accuracy": 0.19252312928438187,
      "num_tokens": 41617338.0,
      "step": 18170
    },
    {
      "entropy": 5.558285474777222,
      "epoch": 1.7966587584025306,
      "grad_norm": 1.375,
      "learning_rate": 0.0004681313241072828,
      "loss": 5.2868,
      "mean_token_accuracy": 0.18772979825735092,
      "num_tokens": 41628407.0,
      "step": 18175
    },
    {
      "entropy": 5.51392502784729,
      "epoch": 1.7971530249110321,
      "grad_norm": 1.390625,
      "learning_rate": 0.00046811321636680797,
      "loss": 5.2451,
      "mean_token_accuracy": 0.18794981241226197,
      "num_tokens": 41639715.0,
      "step": 18180
    },
    {
      "entropy": 5.622610569000244,
      "epoch": 1.7976472914195334,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046809510387571806,
      "loss": 5.3152,
      "mean_token_accuracy": 0.1852630317211151,
      "num_tokens": 41651078.0,
      "step": 18185
    },
    {
      "entropy": 5.602420282363892,
      "epoch": 1.7981415579280347,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004680769866344586,
      "loss": 5.3216,
      "mean_token_accuracy": 0.1861319899559021,
      "num_tokens": 41661848.0,
      "step": 18190
    },
    {
      "entropy": 5.615753889083862,
      "epoch": 1.7986358244365361,
      "grad_norm": 1.25,
      "learning_rate": 0.0004680588646434753,
      "loss": 5.3316,
      "mean_token_accuracy": 0.18385694473981856,
      "num_tokens": 41673452.0,
      "step": 18195
    },
    {
      "entropy": 5.664385557174683,
      "epoch": 1.7991300909450376,
      "grad_norm": 1.25,
      "learning_rate": 0.00046804073790321403,
      "loss": 5.3704,
      "mean_token_accuracy": 0.17711468786001205,
      "num_tokens": 41685312.0,
      "step": 18200
    },
    {
      "entropy": 5.582038927078247,
      "epoch": 1.7996243574535389,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00046802260641412065,
      "loss": 5.2588,
      "mean_token_accuracy": 0.1892946571111679,
      "num_tokens": 41696849.0,
      "step": 18205
    },
    {
      "entropy": 5.591065692901611,
      "epoch": 1.8001186239620404,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004680044701766413,
      "loss": 5.2621,
      "mean_token_accuracy": 0.19078711718320845,
      "num_tokens": 41707506.0,
      "step": 18210
    },
    {
      "entropy": 5.597609138488769,
      "epoch": 1.8006128904705418,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004679863291912219,
      "loss": 5.257,
      "mean_token_accuracy": 0.18123914897441865,
      "num_tokens": 41718981.0,
      "step": 18215
    },
    {
      "entropy": 5.602350664138794,
      "epoch": 1.8011071569790431,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004679681834583089,
      "loss": 5.3541,
      "mean_token_accuracy": 0.18708458989858628,
      "num_tokens": 41731638.0,
      "step": 18220
    },
    {
      "entropy": 5.548145198822022,
      "epoch": 1.8016014234875444,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00046795003297834864,
      "loss": 5.2092,
      "mean_token_accuracy": 0.1994609072804451,
      "num_tokens": 41742820.0,
      "step": 18225
    },
    {
      "entropy": 5.4864260196685795,
      "epoch": 1.8020956899960459,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046793187775178764,
      "loss": 5.2635,
      "mean_token_accuracy": 0.18753628432750702,
      "num_tokens": 41754933.0,
      "step": 18230
    },
    {
      "entropy": 5.585482263565064,
      "epoch": 1.8025899565045473,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004679137177790725,
      "loss": 5.2866,
      "mean_token_accuracy": 0.19378014504909516,
      "num_tokens": 41767078.0,
      "step": 18235
    },
    {
      "entropy": 5.631111192703247,
      "epoch": 1.8030842230130486,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004678955530606501,
      "loss": 5.3253,
      "mean_token_accuracy": 0.18589173257350922,
      "num_tokens": 41778456.0,
      "step": 18240
    },
    {
      "entropy": 5.634024906158447,
      "epoch": 1.8035784895215499,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046787738359696697,
      "loss": 5.3016,
      "mean_token_accuracy": 0.18533339202404023,
      "num_tokens": 41791226.0,
      "step": 18245
    },
    {
      "entropy": 5.549139165878296,
      "epoch": 1.8040727560300513,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004678592093884704,
      "loss": 5.2701,
      "mean_token_accuracy": 0.1962038204073906,
      "num_tokens": 41803765.0,
      "step": 18250
    },
    {
      "entropy": 5.520734882354736,
      "epoch": 1.8045670225385528,
      "grad_norm": 1.71875,
      "learning_rate": 0.00046784103043560736,
      "loss": 5.3042,
      "mean_token_accuracy": 0.18670569360256195,
      "num_tokens": 41815983.0,
      "step": 18255
    },
    {
      "entropy": 5.505499267578125,
      "epoch": 1.805061289047054,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00046782284673882506,
      "loss": 5.2377,
      "mean_token_accuracy": 0.19942699521780013,
      "num_tokens": 41826517.0,
      "step": 18260
    },
    {
      "entropy": 5.677127408981323,
      "epoch": 1.8055555555555556,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004678046582985708,
      "loss": 5.3327,
      "mean_token_accuracy": 0.19110236316919327,
      "num_tokens": 41837064.0,
      "step": 18265
    },
    {
      "entropy": 5.551936531066895,
      "epoch": 1.806049822064057,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000467786465115292,
      "loss": 5.2202,
      "mean_token_accuracy": 0.1949343591928482,
      "num_tokens": 41849109.0,
      "step": 18270
    },
    {
      "entropy": 5.51414794921875,
      "epoch": 1.8065440885725583,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004677682671894364,
      "loss": 5.2323,
      "mean_token_accuracy": 0.1962486356496811,
      "num_tokens": 41859349.0,
      "step": 18275
    },
    {
      "entropy": 5.50361270904541,
      "epoch": 1.8070383550810596,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046775006452145135,
      "loss": 5.2342,
      "mean_token_accuracy": 0.19619107842445374,
      "num_tokens": 41870002.0,
      "step": 18280
    },
    {
      "entropy": 5.596113586425782,
      "epoch": 1.807532621589561,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004677318571117849,
      "loss": 5.2776,
      "mean_token_accuracy": 0.18862443566322326,
      "num_tokens": 41881219.0,
      "step": 18285
    },
    {
      "entropy": 5.616438722610473,
      "epoch": 1.8080268880980626,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00046771364496088475,
      "loss": 5.2625,
      "mean_token_accuracy": 0.19454481452703476,
      "num_tokens": 41893318.0,
      "step": 18290
    },
    {
      "entropy": 5.529006004333496,
      "epoch": 1.8085211546065638,
      "grad_norm": 1.4375,
      "learning_rate": 0.00046769542806919913,
      "loss": 5.2582,
      "mean_token_accuracy": 0.19175457060337067,
      "num_tokens": 41904816.0,
      "step": 18295
    },
    {
      "entropy": 5.583883571624756,
      "epoch": 1.809015421115065,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004676772064371761,
      "loss": 5.3018,
      "mean_token_accuracy": 0.18779680281877517,
      "num_tokens": 41916293.0,
      "step": 18300
    },
    {
      "entropy": 5.568851470947266,
      "epoch": 1.8095096876235668,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046765898006526384,
      "loss": 5.2312,
      "mean_token_accuracy": 0.19844731092453002,
      "num_tokens": 41926654.0,
      "step": 18305
    },
    {
      "entropy": 5.514491987228394,
      "epoch": 1.810003954132068,
      "grad_norm": 1.25,
      "learning_rate": 0.00046764074895391084,
      "loss": 5.1967,
      "mean_token_accuracy": 0.1925481826066971,
      "num_tokens": 41937141.0,
      "step": 18310
    },
    {
      "entropy": 5.605545520782471,
      "epoch": 1.8104982206405693,
      "grad_norm": 1.375,
      "learning_rate": 0.00046762251310356544,
      "loss": 5.3676,
      "mean_token_accuracy": 0.18483726680278778,
      "num_tokens": 41947793.0,
      "step": 18315
    },
    {
      "entropy": 5.667014932632446,
      "epoch": 1.8109924871490708,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004676042725146764,
      "loss": 5.2242,
      "mean_token_accuracy": 0.19247771799564362,
      "num_tokens": 41958524.0,
      "step": 18320
    },
    {
      "entropy": 5.6089780807495115,
      "epoch": 1.8114867536575723,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046758602718769245,
      "loss": 5.2966,
      "mean_token_accuracy": 0.18608249723911285,
      "num_tokens": 41969654.0,
      "step": 18325
    },
    {
      "entropy": 5.482597970962525,
      "epoch": 1.8119810201660735,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00046756777712306225,
      "loss": 5.2454,
      "mean_token_accuracy": 0.19596764594316482,
      "num_tokens": 41980881.0,
      "step": 18330
    },
    {
      "entropy": 5.67239933013916,
      "epoch": 1.8124752866745748,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004675495223212349,
      "loss": 5.298,
      "mean_token_accuracy": 0.18897332847118378,
      "num_tokens": 41992490.0,
      "step": 18335
    },
    {
      "entropy": 5.516888236999511,
      "epoch": 1.8129695531830763,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004675312627826594,
      "loss": 5.1787,
      "mean_token_accuracy": 0.2016163647174835,
      "num_tokens": 42001765.0,
      "step": 18340
    },
    {
      "entropy": 5.507339382171631,
      "epoch": 1.8134638196915778,
      "grad_norm": 1.328125,
      "learning_rate": 0.000467512998507785,
      "loss": 5.2147,
      "mean_token_accuracy": 0.1981933295726776,
      "num_tokens": 42013122.0,
      "step": 18345
    },
    {
      "entropy": 5.679361009597779,
      "epoch": 1.813958086200079,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00046749472949706095,
      "loss": 5.3879,
      "mean_token_accuracy": 0.183884696662426,
      "num_tokens": 42023369.0,
      "step": 18350
    },
    {
      "entropy": 5.660796403884888,
      "epoch": 1.8144523527085805,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004674764557509367,
      "loss": 5.416,
      "mean_token_accuracy": 0.17849403172731398,
      "num_tokens": 42036054.0,
      "step": 18355
    },
    {
      "entropy": 5.584282064437867,
      "epoch": 1.814946619217082,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004674581772698618,
      "loss": 5.245,
      "mean_token_accuracy": 0.18998811542987823,
      "num_tokens": 42048782.0,
      "step": 18360
    },
    {
      "entropy": 5.6189008235931395,
      "epoch": 1.8154408857255833,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046743989405428584,
      "loss": 5.276,
      "mean_token_accuracy": 0.18751917034387589,
      "num_tokens": 42059438.0,
      "step": 18365
    },
    {
      "entropy": 5.626202964782715,
      "epoch": 1.8159351522340845,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004674216061046587,
      "loss": 5.304,
      "mean_token_accuracy": 0.18751108646392822,
      "num_tokens": 42070767.0,
      "step": 18370
    },
    {
      "entropy": 5.562888813018799,
      "epoch": 1.816429418742586,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004674033134214301,
      "loss": 5.2734,
      "mean_token_accuracy": 0.19426854848861694,
      "num_tokens": 42081782.0,
      "step": 18375
    },
    {
      "entropy": 5.5661766052246096,
      "epoch": 1.8169236852510875,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046738501600505016,
      "loss": 5.2641,
      "mean_token_accuracy": 0.18902918994426726,
      "num_tokens": 42094402.0,
      "step": 18380
    },
    {
      "entropy": 5.648767709732056,
      "epoch": 1.8174179517595888,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000467366713855969,
      "loss": 5.3283,
      "mean_token_accuracy": 0.18363415598869323,
      "num_tokens": 42106772.0,
      "step": 18385
    },
    {
      "entropy": 5.590387916564941,
      "epoch": 1.81791221826809,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046734840697463685,
      "loss": 5.2532,
      "mean_token_accuracy": 0.19400646090507506,
      "num_tokens": 42118093.0,
      "step": 18390
    },
    {
      "entropy": 5.568007087707519,
      "epoch": 1.8184064847765915,
      "grad_norm": 1.4375,
      "learning_rate": 0.00046733009536150413,
      "loss": 5.277,
      "mean_token_accuracy": 0.18508214205503465,
      "num_tokens": 42130319.0,
      "step": 18395
    },
    {
      "entropy": 5.649022960662842,
      "epoch": 1.818900751285093,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004673117790170211,
      "loss": 5.271,
      "mean_token_accuracy": 0.19134919643402098,
      "num_tokens": 42140867.0,
      "step": 18400
    },
    {
      "entropy": 5.558538341522217,
      "epoch": 1.8193950177935942,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046729345794163847,
      "loss": 5.2922,
      "mean_token_accuracy": 0.1920735239982605,
      "num_tokens": 42151425.0,
      "step": 18405
    },
    {
      "entropy": 5.669665861129761,
      "epoch": 1.8198892843020957,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046727513213580703,
      "loss": 5.4033,
      "mean_token_accuracy": 0.17599879801273347,
      "num_tokens": 42163553.0,
      "step": 18410
    },
    {
      "entropy": 5.610687971115112,
      "epoch": 1.8203835508105972,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004672568015999774,
      "loss": 5.3269,
      "mean_token_accuracy": 0.17906906753778457,
      "num_tokens": 42174968.0,
      "step": 18415
    },
    {
      "entropy": 5.6283410549163815,
      "epoch": 1.8208778173190985,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046723846633460074,
      "loss": 5.3065,
      "mean_token_accuracy": 0.18652440905570983,
      "num_tokens": 42186944.0,
      "step": 18420
    },
    {
      "entropy": 5.6403772830963135,
      "epoch": 1.8213720838275997,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046722012634012794,
      "loss": 5.2676,
      "mean_token_accuracy": 0.18944794088602065,
      "num_tokens": 42198096.0,
      "step": 18425
    },
    {
      "entropy": 5.581000709533692,
      "epoch": 1.8218663503361012,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00046720178161701014,
      "loss": 5.2835,
      "mean_token_accuracy": 0.19017768800258636,
      "num_tokens": 42208782.0,
      "step": 18430
    },
    {
      "entropy": 5.669473934173584,
      "epoch": 1.8223606168446027,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004671834321656988,
      "loss": 5.3425,
      "mean_token_accuracy": 0.1841800943017006,
      "num_tokens": 42219925.0,
      "step": 18435
    },
    {
      "entropy": 5.618187761306762,
      "epoch": 1.822854883353104,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046716507798664515,
      "loss": 5.2487,
      "mean_token_accuracy": 0.193385711312294,
      "num_tokens": 42231140.0,
      "step": 18440
    },
    {
      "entropy": 5.596238994598389,
      "epoch": 1.8233491498616052,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004671467190803008,
      "loss": 5.3435,
      "mean_token_accuracy": 0.18742116242647172,
      "num_tokens": 42245025.0,
      "step": 18445
    },
    {
      "entropy": 5.581628847122192,
      "epoch": 1.8238434163701067,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004671283554471174,
      "loss": 5.265,
      "mean_token_accuracy": 0.19457059502601623,
      "num_tokens": 42255776.0,
      "step": 18450
    },
    {
      "entropy": 5.585838747024536,
      "epoch": 1.8243376828786082,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004671099870875465,
      "loss": 5.2988,
      "mean_token_accuracy": 0.18524395674467087,
      "num_tokens": 42268016.0,
      "step": 18455
    },
    {
      "entropy": 5.596515226364136,
      "epoch": 1.8248319493871095,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046709161400204034,
      "loss": 5.2664,
      "mean_token_accuracy": 0.19261734038591385,
      "num_tokens": 42279521.0,
      "step": 18460
    },
    {
      "entropy": 5.578329133987427,
      "epoch": 1.825326215895611,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00046707323619105053,
      "loss": 5.2403,
      "mean_token_accuracy": 0.1931838795542717,
      "num_tokens": 42290537.0,
      "step": 18465
    },
    {
      "entropy": 5.614188957214355,
      "epoch": 1.8258204824041124,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046705485365502926,
      "loss": 5.2462,
      "mean_token_accuracy": 0.19025925695896148,
      "num_tokens": 42303119.0,
      "step": 18470
    },
    {
      "entropy": 5.641482257843018,
      "epoch": 1.8263147489126137,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046703646639442893,
      "loss": 5.2776,
      "mean_token_accuracy": 0.19422879368066787,
      "num_tokens": 42314408.0,
      "step": 18475
    },
    {
      "entropy": 5.491457939147949,
      "epoch": 1.826809015421115,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004670180744097017,
      "loss": 5.1801,
      "mean_token_accuracy": 0.19038643985986708,
      "num_tokens": 42325848.0,
      "step": 18480
    },
    {
      "entropy": 5.610499095916748,
      "epoch": 1.8273032819296164,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004669996777013,
      "loss": 5.2892,
      "mean_token_accuracy": 0.18619169294834137,
      "num_tokens": 42337737.0,
      "step": 18485
    },
    {
      "entropy": 5.537673044204712,
      "epoch": 1.827797548438118,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046698127626967645,
      "loss": 5.2006,
      "mean_token_accuracy": 0.19505991637706757,
      "num_tokens": 42348418.0,
      "step": 18490
    },
    {
      "entropy": 5.578427410125732,
      "epoch": 1.8282918149466192,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004669628701152837,
      "loss": 5.2924,
      "mean_token_accuracy": 0.18723042607307433,
      "num_tokens": 42358665.0,
      "step": 18495
    },
    {
      "entropy": 5.639978218078613,
      "epoch": 1.8287860814551204,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004669444592385746,
      "loss": 5.2455,
      "mean_token_accuracy": 0.2033316284418106,
      "num_tokens": 42369030.0,
      "step": 18500
    },
    {
      "entropy": 5.573538446426392,
      "epoch": 1.829280347963622,
      "grad_norm": 1.484375,
      "learning_rate": 0.00046692604364000203,
      "loss": 5.2671,
      "mean_token_accuracy": 0.18693264126777648,
      "num_tokens": 42380751.0,
      "step": 18505
    },
    {
      "entropy": 5.532770872116089,
      "epoch": 1.8297746144721234,
      "grad_norm": 1.453125,
      "learning_rate": 0.000466907623320019,
      "loss": 5.2881,
      "mean_token_accuracy": 0.18852583914995194,
      "num_tokens": 42391945.0,
      "step": 18510
    },
    {
      "entropy": 5.622081851959228,
      "epoch": 1.8302688809806247,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046688919827907866,
      "loss": 5.2771,
      "mean_token_accuracy": 0.185960952937603,
      "num_tokens": 42403346.0,
      "step": 18515
    },
    {
      "entropy": 5.570225095748901,
      "epoch": 1.8307631474891262,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00046687076851763433,
      "loss": 5.3272,
      "mean_token_accuracy": 0.18314879685640334,
      "num_tokens": 42414895.0,
      "step": 18520
    },
    {
      "entropy": 5.502289581298828,
      "epoch": 1.8312574139976276,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046685233403613927,
      "loss": 5.219,
      "mean_token_accuracy": 0.19286630302667618,
      "num_tokens": 42426951.0,
      "step": 18525
    },
    {
      "entropy": 5.600601100921631,
      "epoch": 1.831751680506129,
      "grad_norm": 1.46875,
      "learning_rate": 0.00046683389483504704,
      "loss": 5.2678,
      "mean_token_accuracy": 0.1882234901189804,
      "num_tokens": 42438313.0,
      "step": 18530
    },
    {
      "entropy": 5.558518648147583,
      "epoch": 1.8322459470146302,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004668154509148112,
      "loss": 5.265,
      "mean_token_accuracy": 0.1879659563302994,
      "num_tokens": 42448592.0,
      "step": 18535
    },
    {
      "entropy": 5.5985595703125,
      "epoch": 1.8327402135231317,
      "grad_norm": 1.359375,
      "learning_rate": 0.00046679700227588556,
      "loss": 5.2908,
      "mean_token_accuracy": 0.18764979392290115,
      "num_tokens": 42460732.0,
      "step": 18540
    },
    {
      "entropy": 5.670887756347656,
      "epoch": 1.8332344800316331,
      "grad_norm": 1.25,
      "learning_rate": 0.00046677854891872384,
      "loss": 5.2736,
      "mean_token_accuracy": 0.190063539147377,
      "num_tokens": 42472862.0,
      "step": 18545
    },
    {
      "entropy": 5.540297412872315,
      "epoch": 1.8337287465401344,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004667600908437802,
      "loss": 5.2134,
      "mean_token_accuracy": 0.1941196545958519,
      "num_tokens": 42483227.0,
      "step": 18550
    },
    {
      "entropy": 5.462173700332642,
      "epoch": 1.8342230130486357,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046674162805150843,
      "loss": 5.0999,
      "mean_token_accuracy": 0.20233284831047058,
      "num_tokens": 42495414.0,
      "step": 18555
    },
    {
      "entropy": 5.503639793395996,
      "epoch": 1.8347172795571374,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046672316054236293,
      "loss": 5.3033,
      "mean_token_accuracy": 0.18969578593969344,
      "num_tokens": 42507908.0,
      "step": 18560
    },
    {
      "entropy": 5.509304571151733,
      "epoch": 1.8352115460656386,
      "grad_norm": 1.703125,
      "learning_rate": 0.00046670468831679795,
      "loss": 5.1632,
      "mean_token_accuracy": 0.196702441573143,
      "num_tokens": 42519228.0,
      "step": 18565
    },
    {
      "entropy": 5.529495525360107,
      "epoch": 1.83570581257414,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004666862113752679,
      "loss": 5.2081,
      "mean_token_accuracy": 0.19709310680627823,
      "num_tokens": 42530571.0,
      "step": 18570
    },
    {
      "entropy": 5.568110704421997,
      "epoch": 1.8362000790826414,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004666677297182273,
      "loss": 5.2792,
      "mean_token_accuracy": 0.19249682277441024,
      "num_tokens": 42541781.0,
      "step": 18575
    },
    {
      "entropy": 5.59364423751831,
      "epoch": 1.8366943455911429,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004666492433461308,
      "loss": 5.3208,
      "mean_token_accuracy": 0.18383362889289856,
      "num_tokens": 42553912.0,
      "step": 18580
    },
    {
      "entropy": 5.508568954467774,
      "epoch": 1.8371886120996441,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004666307522594332,
      "loss": 5.1763,
      "mean_token_accuracy": 0.2015312060713768,
      "num_tokens": 42565812.0,
      "step": 18585
    },
    {
      "entropy": 5.565052461624146,
      "epoch": 1.8376828786081454,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004666122564585893,
      "loss": 5.2557,
      "mean_token_accuracy": 0.1809048607945442,
      "num_tokens": 42578177.0,
      "step": 18590
    },
    {
      "entropy": 5.582100915908813,
      "epoch": 1.8381771451166469,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046659375594405427,
      "loss": 5.3378,
      "mean_token_accuracy": 0.18963935673236848,
      "num_tokens": 42590336.0,
      "step": 18595
    },
    {
      "entropy": 5.642604160308838,
      "epoch": 1.8386714116251484,
      "grad_norm": 1.375,
      "learning_rate": 0.00046657525071628307,
      "loss": 5.2479,
      "mean_token_accuracy": 0.18759679794311523,
      "num_tokens": 42601438.0,
      "step": 18600
    },
    {
      "entropy": 5.527330636978149,
      "epoch": 1.8391656781336496,
      "grad_norm": 1.421875,
      "learning_rate": 0.00046655674077573104,
      "loss": 5.2364,
      "mean_token_accuracy": 0.19379695355892182,
      "num_tokens": 42612572.0,
      "step": 18605
    },
    {
      "entropy": 5.589715051651001,
      "epoch": 1.839659944642151,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046653822612285336,
      "loss": 5.3361,
      "mean_token_accuracy": 0.1899073138833046,
      "num_tokens": 42623138.0,
      "step": 18610
    },
    {
      "entropy": 5.506485176086426,
      "epoch": 1.8401542111506526,
      "grad_norm": 1.375,
      "learning_rate": 0.00046651970675810574,
      "loss": 5.2378,
      "mean_token_accuracy": 0.19500716477632524,
      "num_tokens": 42635694.0,
      "step": 18615
    },
    {
      "entropy": 5.7109466075897215,
      "epoch": 1.8406484776591538,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004665011826819435,
      "loss": 5.3667,
      "mean_token_accuracy": 0.18470097929239274,
      "num_tokens": 42648371.0,
      "step": 18620
    },
    {
      "entropy": 5.624104356765747,
      "epoch": 1.841142744167655,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046648265389482245,
      "loss": 5.2692,
      "mean_token_accuracy": 0.18536249101161956,
      "num_tokens": 42659229.0,
      "step": 18625
    },
    {
      "entropy": 5.483868980407715,
      "epoch": 1.8416370106761566,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004664641203971985,
      "loss": 5.2407,
      "mean_token_accuracy": 0.19081413596868516,
      "num_tokens": 42669991.0,
      "step": 18630
    },
    {
      "entropy": 5.513874149322509,
      "epoch": 1.842131277184658,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046644558218952745,
      "loss": 5.1632,
      "mean_token_accuracy": 0.19989305436611177,
      "num_tokens": 42681091.0,
      "step": 18635
    },
    {
      "entropy": 5.596211624145508,
      "epoch": 1.8426255436931593,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046642703927226534,
      "loss": 5.2613,
      "mean_token_accuracy": 0.18828439563512803,
      "num_tokens": 42691928.0,
      "step": 18640
    },
    {
      "entropy": 5.541527462005615,
      "epoch": 1.8431198102016606,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004664084916458684,
      "loss": 5.2494,
      "mean_token_accuracy": 0.18935770243406297,
      "num_tokens": 42704611.0,
      "step": 18645
    },
    {
      "entropy": 5.60752682685852,
      "epoch": 1.843614076710162,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004663899393107929,
      "loss": 5.2701,
      "mean_token_accuracy": 0.19085991531610488,
      "num_tokens": 42715223.0,
      "step": 18650
    },
    {
      "entropy": 5.494353199005127,
      "epoch": 1.8441083432186636,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004663713822674952,
      "loss": 5.1885,
      "mean_token_accuracy": 0.19425834268331527,
      "num_tokens": 42728177.0,
      "step": 18655
    },
    {
      "entropy": 5.6537891864776615,
      "epoch": 1.8446026097271648,
      "grad_norm": 1.484375,
      "learning_rate": 0.00046635282051643166,
      "loss": 5.3245,
      "mean_token_accuracy": 0.1818065509200096,
      "num_tokens": 42738469.0,
      "step": 18660
    },
    {
      "entropy": 5.554398918151856,
      "epoch": 1.8450968762356663,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004663342540580593,
      "loss": 5.1918,
      "mean_token_accuracy": 0.19732003659009933,
      "num_tokens": 42749915.0,
      "step": 18665
    },
    {
      "entropy": 5.6051685333251955,
      "epoch": 1.8455911427441678,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004663156828928344,
      "loss": 5.3715,
      "mean_token_accuracy": 0.1835647091269493,
      "num_tokens": 42762679.0,
      "step": 18670
    },
    {
      "entropy": 5.658066606521606,
      "epoch": 1.846085409252669,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004662971070212141,
      "loss": 5.3072,
      "mean_token_accuracy": 0.19209382832050323,
      "num_tokens": 42775351.0,
      "step": 18675
    },
    {
      "entropy": 5.551802062988282,
      "epoch": 1.8465796757611703,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046627852644365525,
      "loss": 5.2712,
      "mean_token_accuracy": 0.18978420346975328,
      "num_tokens": 42785435.0,
      "step": 18680
    },
    {
      "entropy": 5.584657335281372,
      "epoch": 1.8470739422696718,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00046625994116061505,
      "loss": 5.32,
      "mean_token_accuracy": 0.1863328516483307,
      "num_tokens": 42795737.0,
      "step": 18685
    },
    {
      "entropy": 5.61513319015503,
      "epoch": 1.8475682087781733,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004662413511725506,
      "loss": 5.3147,
      "mean_token_accuracy": 0.18654535263776778,
      "num_tokens": 42806965.0,
      "step": 18690
    },
    {
      "entropy": 5.642132377624511,
      "epoch": 1.8480624752866746,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004662227564799192,
      "loss": 5.2769,
      "mean_token_accuracy": 0.18391224443912507,
      "num_tokens": 42818346.0,
      "step": 18695
    },
    {
      "entropy": 5.493491172790527,
      "epoch": 1.8485567417951758,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046620415708317834,
      "loss": 5.2342,
      "mean_token_accuracy": 0.1959983929991722,
      "num_tokens": 42829838.0,
      "step": 18700
    },
    {
      "entropy": 5.62834062576294,
      "epoch": 1.8490510083036773,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004661855529827856,
      "loss": 5.2891,
      "mean_token_accuracy": 0.19320822954177858,
      "num_tokens": 42841212.0,
      "step": 18705
    },
    {
      "entropy": 5.536891317367553,
      "epoch": 1.8495452748121788,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00046616694417919854,
      "loss": 5.1798,
      "mean_token_accuracy": 0.19862732887268067,
      "num_tokens": 42853237.0,
      "step": 18710
    },
    {
      "entropy": 5.5685107707977295,
      "epoch": 1.85003954132068,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00046614833067287504,
      "loss": 5.2833,
      "mean_token_accuracy": 0.18617384284734725,
      "num_tokens": 42864000.0,
      "step": 18715
    },
    {
      "entropy": 5.560598230361938,
      "epoch": 1.8505338078291815,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046612971246427293,
      "loss": 5.1688,
      "mean_token_accuracy": 0.19578854739665985,
      "num_tokens": 42874491.0,
      "step": 18720
    },
    {
      "entropy": 5.579421901702881,
      "epoch": 1.851028074337683,
      "grad_norm": 1.484375,
      "learning_rate": 0.00046611108955385024,
      "loss": 5.3854,
      "mean_token_accuracy": 0.1749049425125122,
      "num_tokens": 42885947.0,
      "step": 18725
    },
    {
      "entropy": 5.542092037200928,
      "epoch": 1.8515223408461843,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004660924619420651,
      "loss": 5.2946,
      "mean_token_accuracy": 0.18822437822818755,
      "num_tokens": 42897907.0,
      "step": 18730
    },
    {
      "entropy": 5.586062812805176,
      "epoch": 1.8520166073546855,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004660738296293758,
      "loss": 5.2199,
      "mean_token_accuracy": 0.19404203891754152,
      "num_tokens": 42909332.0,
      "step": 18735
    },
    {
      "entropy": 5.642909049987793,
      "epoch": 1.852510873863187,
      "grad_norm": 2.15625,
      "learning_rate": 0.00046605519261624055,
      "loss": 5.3511,
      "mean_token_accuracy": 0.18243160843849182,
      "num_tokens": 42921877.0,
      "step": 18740
    },
    {
      "entropy": 5.587498998641967,
      "epoch": 1.8530051403716885,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00046603655090311796,
      "loss": 5.2754,
      "mean_token_accuracy": 0.18601575195789338,
      "num_tokens": 42933381.0,
      "step": 18745
    },
    {
      "entropy": 5.642437076568603,
      "epoch": 1.8534994068801898,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004660179044904666,
      "loss": 5.2904,
      "mean_token_accuracy": 0.1857299402356148,
      "num_tokens": 42944756.0,
      "step": 18750
    },
    {
      "entropy": 5.64343900680542,
      "epoch": 1.853993673388691,
      "grad_norm": 1.46875,
      "learning_rate": 0.00046599925337874516,
      "loss": 5.3174,
      "mean_token_accuracy": 0.18462852984666825,
      "num_tokens": 42955587.0,
      "step": 18755
    },
    {
      "entropy": 5.669891691207885,
      "epoch": 1.8544879398971925,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00046598059756841246,
      "loss": 5.3501,
      "mean_token_accuracy": 0.18596279621124268,
      "num_tokens": 42966150.0,
      "step": 18760
    },
    {
      "entropy": 5.612529373168945,
      "epoch": 1.854982206405694,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004659619370599274,
      "loss": 5.3032,
      "mean_token_accuracy": 0.18246298879384995,
      "num_tokens": 42978626.0,
      "step": 18765
    },
    {
      "entropy": 5.549655961990356,
      "epoch": 1.8554764729141953,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046594327185374906,
      "loss": 5.1531,
      "mean_token_accuracy": 0.1991766169667244,
      "num_tokens": 42989194.0,
      "step": 18770
    },
    {
      "entropy": 5.487170886993408,
      "epoch": 1.8559707394226967,
      "grad_norm": 1.25,
      "learning_rate": 0.00046592460195033663,
      "loss": 5.1962,
      "mean_token_accuracy": 0.19495876133441925,
      "num_tokens": 43000589.0,
      "step": 18775
    },
    {
      "entropy": 5.531355333328247,
      "epoch": 1.8564650059311982,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004659059273501494,
      "loss": 5.187,
      "mean_token_accuracy": 0.2011738732457161,
      "num_tokens": 43011417.0,
      "step": 18780
    },
    {
      "entropy": 5.615150213241577,
      "epoch": 1.8569592724396995,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004658872480536467,
      "loss": 5.2745,
      "mean_token_accuracy": 0.18746360689401625,
      "num_tokens": 43023178.0,
      "step": 18785
    },
    {
      "entropy": 5.614170503616333,
      "epoch": 1.8574535389482008,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046586856406128805,
      "loss": 5.327,
      "mean_token_accuracy": 0.18413175493478776,
      "num_tokens": 43035526.0,
      "step": 18790
    },
    {
      "entropy": 5.5767356872558596,
      "epoch": 1.8579478054567022,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004658498753735332,
      "loss": 5.2589,
      "mean_token_accuracy": 0.19100056290626527,
      "num_tokens": 43046681.0,
      "step": 18795
    },
    {
      "entropy": 5.5651201725006105,
      "epoch": 1.8584420719652037,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046583118199084177,
      "loss": 5.2506,
      "mean_token_accuracy": 0.19438367187976838,
      "num_tokens": 43058514.0,
      "step": 18800
    },
    {
      "entropy": 5.591277074813843,
      "epoch": 1.858936338473705,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004658124839136737,
      "loss": 5.3186,
      "mean_token_accuracy": 0.18576408922672272,
      "num_tokens": 43069588.0,
      "step": 18805
    },
    {
      "entropy": 5.552458763122559,
      "epoch": 1.8594306049822062,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004657937811424889,
      "loss": 5.2594,
      "mean_token_accuracy": 0.19273126423358916,
      "num_tokens": 43080811.0,
      "step": 18810
    },
    {
      "entropy": 5.677523946762085,
      "epoch": 1.859924871490708,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004657750736777474,
      "loss": 5.3091,
      "mean_token_accuracy": 0.18513265550136565,
      "num_tokens": 43093506.0,
      "step": 18815
    },
    {
      "entropy": 5.507422304153442,
      "epoch": 1.8604191379992092,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046575636151990955,
      "loss": 5.2082,
      "mean_token_accuracy": 0.1954442411661148,
      "num_tokens": 43105272.0,
      "step": 18820
    },
    {
      "entropy": 5.554544591903687,
      "epoch": 1.8609134045077105,
      "grad_norm": 1.46875,
      "learning_rate": 0.00046573764466943565,
      "loss": 5.249,
      "mean_token_accuracy": 0.18816939294338225,
      "num_tokens": 43116070.0,
      "step": 18825
    },
    {
      "entropy": 5.594967699050903,
      "epoch": 1.861407671016212,
      "grad_norm": 1.375,
      "learning_rate": 0.00046571892312678606,
      "loss": 5.2974,
      "mean_token_accuracy": 0.18564522564411162,
      "num_tokens": 43126347.0,
      "step": 18830
    },
    {
      "entropy": 5.652983951568603,
      "epoch": 1.8619019375247134,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046570019689242136,
      "loss": 5.3466,
      "mean_token_accuracy": 0.1833465501666069,
      "num_tokens": 43138752.0,
      "step": 18835
    },
    {
      "entropy": 5.643504524230957,
      "epoch": 1.8623962040332147,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004656814659668023,
      "loss": 5.4081,
      "mean_token_accuracy": 0.17944862246513366,
      "num_tokens": 43151283.0,
      "step": 18840
    },
    {
      "entropy": 5.6380236625671385,
      "epoch": 1.862890470541716,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046566273035038953,
      "loss": 5.2691,
      "mean_token_accuracy": 0.18652819991111755,
      "num_tokens": 43161643.0,
      "step": 18845
    },
    {
      "entropy": 5.612396478652954,
      "epoch": 1.8633847370502175,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046564399004364403,
      "loss": 5.2541,
      "mean_token_accuracy": 0.18873074650764465,
      "num_tokens": 43171966.0,
      "step": 18850
    },
    {
      "entropy": 5.556411170959473,
      "epoch": 1.863879003558719,
      "grad_norm": 1.328125,
      "learning_rate": 0.00046562524504702683,
      "loss": 5.2804,
      "mean_token_accuracy": 0.19142732173204421,
      "num_tokens": 43183022.0,
      "step": 18855
    },
    {
      "entropy": 5.605290365219116,
      "epoch": 1.8643732700672202,
      "grad_norm": 1.765625,
      "learning_rate": 0.00046560649536099905,
      "loss": 5.2911,
      "mean_token_accuracy": 0.1898275911808014,
      "num_tokens": 43194927.0,
      "step": 18860
    },
    {
      "entropy": 5.533089065551758,
      "epoch": 1.8648675365757215,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004655877409860218,
      "loss": 5.1955,
      "mean_token_accuracy": 0.20584086030721666,
      "num_tokens": 43205519.0,
      "step": 18865
    },
    {
      "entropy": 5.548800230026245,
      "epoch": 1.8653618030842232,
      "grad_norm": 1.390625,
      "learning_rate": 0.00046556898192255664,
      "loss": 5.3366,
      "mean_token_accuracy": 0.1795489251613617,
      "num_tokens": 43217805.0,
      "step": 18870
    },
    {
      "entropy": 5.573772764205932,
      "epoch": 1.8658560695927244,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046555021817106494,
      "loss": 5.293,
      "mean_token_accuracy": 0.18664428740739822,
      "num_tokens": 43228394.0,
      "step": 18875
    },
    {
      "entropy": 5.606089496612549,
      "epoch": 1.8663503361012257,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046553144973200835,
      "loss": 5.2537,
      "mean_token_accuracy": 0.19243418276309968,
      "num_tokens": 43240204.0,
      "step": 18880
    },
    {
      "entropy": 5.637154960632325,
      "epoch": 1.8668446026097272,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004655126766058485,
      "loss": 5.3775,
      "mean_token_accuracy": 0.18233660459518433,
      "num_tokens": 43251103.0,
      "step": 18885
    },
    {
      "entropy": 5.504503679275513,
      "epoch": 1.8673388691182287,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046549389879304724,
      "loss": 5.2004,
      "mean_token_accuracy": 0.19156041443347932,
      "num_tokens": 43263653.0,
      "step": 18890
    },
    {
      "entropy": 5.582380962371826,
      "epoch": 1.86783313562673,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004654751162940665,
      "loss": 5.2296,
      "mean_token_accuracy": 0.19206803143024445,
      "num_tokens": 43274893.0,
      "step": 18895
    },
    {
      "entropy": 5.5891376495361325,
      "epoch": 1.8683274021352312,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004654563291093684,
      "loss": 5.2327,
      "mean_token_accuracy": 0.19600287079811096,
      "num_tokens": 43287510.0,
      "step": 18900
    },
    {
      "entropy": 5.570928621292114,
      "epoch": 1.8688216686437327,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004654375372394151,
      "loss": 5.3247,
      "mean_token_accuracy": 0.1864650085568428,
      "num_tokens": 43299735.0,
      "step": 18905
    },
    {
      "entropy": 5.585889577865601,
      "epoch": 1.8693159351522342,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046541874068466876,
      "loss": 5.2192,
      "mean_token_accuracy": 0.202897310256958,
      "num_tokens": 43310401.0,
      "step": 18910
    },
    {
      "entropy": 5.557915449142456,
      "epoch": 1.8698102016607354,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046539993944559187,
      "loss": 5.2178,
      "mean_token_accuracy": 0.1919250711798668,
      "num_tokens": 43321300.0,
      "step": 18915
    },
    {
      "entropy": 5.555347490310669,
      "epoch": 1.870304468169237,
      "grad_norm": 1.5546875,
      "learning_rate": 0.000465381133522647,
      "loss": 5.2243,
      "mean_token_accuracy": 0.1899488687515259,
      "num_tokens": 43332304.0,
      "step": 18920
    },
    {
      "entropy": 5.563818407058716,
      "epoch": 1.8707987346777384,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004653623229162966,
      "loss": 5.2768,
      "mean_token_accuracy": 0.18771232813596725,
      "num_tokens": 43343524.0,
      "step": 18925
    },
    {
      "entropy": 5.618774032592773,
      "epoch": 1.8712930011862396,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046534350762700354,
      "loss": 5.2483,
      "mean_token_accuracy": 0.18217865377664566,
      "num_tokens": 43353364.0,
      "step": 18930
    },
    {
      "entropy": 5.497668838500976,
      "epoch": 1.871787267694741,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046532468765523075,
      "loss": 5.143,
      "mean_token_accuracy": 0.20881021916866302,
      "num_tokens": 43365660.0,
      "step": 18935
    },
    {
      "entropy": 5.636999845504761,
      "epoch": 1.8722815342032424,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046530586300144104,
      "loss": 5.3554,
      "mean_token_accuracy": 0.18636872321367265,
      "num_tokens": 43376882.0,
      "step": 18940
    },
    {
      "entropy": 5.628673505783081,
      "epoch": 1.8727758007117439,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046528703366609757,
      "loss": 5.2655,
      "mean_token_accuracy": 0.19539225697517396,
      "num_tokens": 43387872.0,
      "step": 18945
    },
    {
      "entropy": 5.614740705490112,
      "epoch": 1.8732700672202451,
      "grad_norm": 1.25,
      "learning_rate": 0.00046526819964966363,
      "loss": 5.3383,
      "mean_token_accuracy": 0.18545145392417908,
      "num_tokens": 43399162.0,
      "step": 18950
    },
    {
      "entropy": 5.564567518234253,
      "epoch": 1.8737643337287464,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004652493609526024,
      "loss": 5.2372,
      "mean_token_accuracy": 0.18610263466835023,
      "num_tokens": 43410163.0,
      "step": 18955
    },
    {
      "entropy": 5.598093271255493,
      "epoch": 1.8742586002372479,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004652305175753774,
      "loss": 5.2685,
      "mean_token_accuracy": 0.18761257231235504,
      "num_tokens": 43421841.0,
      "step": 18960
    },
    {
      "entropy": 5.600408363342285,
      "epoch": 1.8747528667457494,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004652116695184521,
      "loss": 5.2245,
      "mean_token_accuracy": 0.19539880007505417,
      "num_tokens": 43432301.0,
      "step": 18965
    },
    {
      "entropy": 5.621077060699463,
      "epoch": 1.8752471332542506,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004651928167822903,
      "loss": 5.3126,
      "mean_token_accuracy": 0.18520933240652085,
      "num_tokens": 43443531.0,
      "step": 18970
    },
    {
      "entropy": 5.500466680526733,
      "epoch": 1.8757413997627521,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004651739593673556,
      "loss": 5.1751,
      "mean_token_accuracy": 0.1960080534219742,
      "num_tokens": 43455230.0,
      "step": 18975
    },
    {
      "entropy": 5.578135299682617,
      "epoch": 1.8762356662712536,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004651550972741121,
      "loss": 5.3406,
      "mean_token_accuracy": 0.18186864107847214,
      "num_tokens": 43466342.0,
      "step": 18980
    },
    {
      "entropy": 5.641259622573853,
      "epoch": 1.8767299327797549,
      "grad_norm": 1.25,
      "learning_rate": 0.0004651362305030237,
      "loss": 5.3558,
      "mean_token_accuracy": 0.18394186198711396,
      "num_tokens": 43477962.0,
      "step": 18985
    },
    {
      "entropy": 5.553733062744141,
      "epoch": 1.8772241992882561,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004651173590545545,
      "loss": 5.2919,
      "mean_token_accuracy": 0.19252508878707886,
      "num_tokens": 43489252.0,
      "step": 18990
    },
    {
      "entropy": 5.650105571746826,
      "epoch": 1.8777184657967576,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046509848292916877,
      "loss": 5.2526,
      "mean_token_accuracy": 0.19828298240900039,
      "num_tokens": 43499199.0,
      "step": 18995
    },
    {
      "entropy": 5.534240865707398,
      "epoch": 1.878212732305259,
      "grad_norm": 1.40625,
      "learning_rate": 0.00046507960212733093,
      "loss": 5.241,
      "mean_token_accuracy": 0.1918190285563469,
      "num_tokens": 43509812.0,
      "step": 19000
    },
    {
      "entropy": 5.481629514694214,
      "epoch": 1.8787069988137604,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046506071664950525,
      "loss": 5.2458,
      "mean_token_accuracy": 0.19882641285657882,
      "num_tokens": 43521197.0,
      "step": 19005
    },
    {
      "entropy": 5.569634914398193,
      "epoch": 1.8792012653222616,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00046504182649615664,
      "loss": 5.275,
      "mean_token_accuracy": 0.19473995566368102,
      "num_tokens": 43532826.0,
      "step": 19010
    },
    {
      "entropy": 5.63658037185669,
      "epoch": 1.879695531830763,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046502293166774945,
      "loss": 5.2724,
      "mean_token_accuracy": 0.1822723925113678,
      "num_tokens": 43544617.0,
      "step": 19015
    },
    {
      "entropy": 5.493009328842163,
      "epoch": 1.8801897983392646,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004650040321647487,
      "loss": 5.1501,
      "mean_token_accuracy": 0.20155085921287536,
      "num_tokens": 43556006.0,
      "step": 19020
    },
    {
      "entropy": 5.587058591842651,
      "epoch": 1.8806840648477658,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004649851279876193,
      "loss": 5.2547,
      "mean_token_accuracy": 0.19055272042751312,
      "num_tokens": 43567112.0,
      "step": 19025
    },
    {
      "entropy": 5.578532886505127,
      "epoch": 1.8811783313562673,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004649662191368263,
      "loss": 5.315,
      "mean_token_accuracy": 0.19133198708295823,
      "num_tokens": 43577974.0,
      "step": 19030
    },
    {
      "entropy": 5.521011304855347,
      "epoch": 1.8816725978647688,
      "grad_norm": 1.25,
      "learning_rate": 0.0004649473056128349,
      "loss": 5.164,
      "mean_token_accuracy": 0.19877071380615235,
      "num_tokens": 43588723.0,
      "step": 19035
    },
    {
      "entropy": 5.542726278305054,
      "epoch": 1.88216686437327,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004649283874161102,
      "loss": 5.1933,
      "mean_token_accuracy": 0.1949515625834465,
      "num_tokens": 43599721.0,
      "step": 19040
    },
    {
      "entropy": 5.649376487731933,
      "epoch": 1.8826611308817713,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004649094645471177,
      "loss": 5.3023,
      "mean_token_accuracy": 0.18710735738277434,
      "num_tokens": 43610517.0,
      "step": 19045
    },
    {
      "entropy": 5.62946834564209,
      "epoch": 1.8831553973902728,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000464890537006323,
      "loss": 5.3393,
      "mean_token_accuracy": 0.18687674254179001,
      "num_tokens": 43622547.0,
      "step": 19050
    },
    {
      "entropy": 5.601474237442017,
      "epoch": 1.8836496638987743,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0004648716047941916,
      "loss": 5.1951,
      "mean_token_accuracy": 0.19289742410182953,
      "num_tokens": 43633982.0,
      "step": 19055
    },
    {
      "entropy": 5.507806730270386,
      "epoch": 1.8841439304072756,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004648526679111892,
      "loss": 5.2357,
      "mean_token_accuracy": 0.19724133461713791,
      "num_tokens": 43644888.0,
      "step": 19060
    },
    {
      "entropy": 5.590801668167114,
      "epoch": 1.8846381969157768,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004648337263577818,
      "loss": 5.2691,
      "mean_token_accuracy": 0.19030485600233077,
      "num_tokens": 43658352.0,
      "step": 19065
    },
    {
      "entropy": 5.530408668518066,
      "epoch": 1.8851324634242785,
      "grad_norm": 1.515625,
      "learning_rate": 0.00046481478013443527,
      "loss": 5.1343,
      "mean_token_accuracy": 0.19978090077638627,
      "num_tokens": 43669127.0,
      "step": 19070
    },
    {
      "entropy": 5.571334505081177,
      "epoch": 1.8856267299327798,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004647958292416157,
      "loss": 5.2541,
      "mean_token_accuracy": 0.19033319503068924,
      "num_tokens": 43681504.0,
      "step": 19075
    },
    {
      "entropy": 5.510461378097534,
      "epoch": 1.886120996441281,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004647768736797893,
      "loss": 5.1728,
      "mean_token_accuracy": 0.19345988631248473,
      "num_tokens": 43691396.0,
      "step": 19080
    },
    {
      "entropy": 5.574536609649658,
      "epoch": 1.8866152629497825,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046475791344942227,
      "loss": 5.25,
      "mean_token_accuracy": 0.19868419021368028,
      "num_tokens": 43703761.0,
      "step": 19085
    },
    {
      "entropy": 5.51921238899231,
      "epoch": 1.887109529458284,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004647389485509812,
      "loss": 5.2152,
      "mean_token_accuracy": 0.19704494178295134,
      "num_tokens": 43715041.0,
      "step": 19090
    },
    {
      "entropy": 5.600714588165284,
      "epoch": 1.8876037959667853,
      "grad_norm": 1.125,
      "learning_rate": 0.00046471997898493256,
      "loss": 5.3376,
      "mean_token_accuracy": 0.18485677540302276,
      "num_tokens": 43727060.0,
      "step": 19095
    },
    {
      "entropy": 5.628681612014771,
      "epoch": 1.8880980624752866,
      "grad_norm": 1.28125,
      "learning_rate": 0.000464701004751743,
      "loss": 5.3286,
      "mean_token_accuracy": 0.1794735386967659,
      "num_tokens": 43738996.0,
      "step": 19100
    },
    {
      "entropy": 5.541102695465088,
      "epoch": 1.888592328983788,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004646820258518793,
      "loss": 5.1586,
      "mean_token_accuracy": 0.19014810770750046,
      "num_tokens": 43749415.0,
      "step": 19105
    },
    {
      "entropy": 5.492815637588501,
      "epoch": 1.8890865954922895,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004646630422858083,
      "loss": 5.2728,
      "mean_token_accuracy": 0.19310049414634706,
      "num_tokens": 43760289.0,
      "step": 19110
    },
    {
      "entropy": 5.620342254638672,
      "epoch": 1.8895808620007908,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00046464405405399707,
      "loss": 5.269,
      "mean_token_accuracy": 0.1939537540078163,
      "num_tokens": 43771267.0,
      "step": 19115
    },
    {
      "entropy": 5.679547834396362,
      "epoch": 1.890075128509292,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046462506115691274,
      "loss": 5.2826,
      "mean_token_accuracy": 0.19144418090581894,
      "num_tokens": 43782584.0,
      "step": 19120
    },
    {
      "entropy": 5.619050025939941,
      "epoch": 1.8905693950177938,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046460606359502237,
      "loss": 5.3289,
      "mean_token_accuracy": 0.18353534787893294,
      "num_tokens": 43792738.0,
      "step": 19125
    },
    {
      "entropy": 5.537538766860962,
      "epoch": 1.891063661526295,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046458706136879354,
      "loss": 5.3112,
      "mean_token_accuracy": 0.1930339217185974,
      "num_tokens": 43804117.0,
      "step": 19130
    },
    {
      "entropy": 5.658473205566406,
      "epoch": 1.8915579280347963,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004645680544786935,
      "loss": 5.3089,
      "mean_token_accuracy": 0.18483688086271285,
      "num_tokens": 43817367.0,
      "step": 19135
    },
    {
      "entropy": 5.63974666595459,
      "epoch": 1.8920521945432978,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00046454904292518996,
      "loss": 5.2813,
      "mean_token_accuracy": 0.19192086905241013,
      "num_tokens": 43829213.0,
      "step": 19140
    },
    {
      "entropy": 5.596264791488648,
      "epoch": 1.8925464610517992,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00046453002670875055,
      "loss": 5.2566,
      "mean_token_accuracy": 0.19487277269363404,
      "num_tokens": 43839939.0,
      "step": 19145
    },
    {
      "entropy": 5.591429853439331,
      "epoch": 1.8930407275603005,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046451100582984304,
      "loss": 5.2305,
      "mean_token_accuracy": 0.19448020607233046,
      "num_tokens": 43850411.0,
      "step": 19150
    },
    {
      "entropy": 5.525946044921875,
      "epoch": 1.8935349940688018,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004644919802889354,
      "loss": 5.2163,
      "mean_token_accuracy": 0.18922030925750732,
      "num_tokens": 43861083.0,
      "step": 19155
    },
    {
      "entropy": 5.619222021102905,
      "epoch": 1.8940292605773033,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00046447295008649576,
      "loss": 5.1924,
      "mean_token_accuracy": 0.19049459844827651,
      "num_tokens": 43871712.0,
      "step": 19160
    },
    {
      "entropy": 5.640578985214233,
      "epoch": 1.8945235270858047,
      "grad_norm": 1.390625,
      "learning_rate": 0.00046445391522299204,
      "loss": 5.3008,
      "mean_token_accuracy": 0.18398344814777373,
      "num_tokens": 43884617.0,
      "step": 19165
    },
    {
      "entropy": 5.559560298919678,
      "epoch": 1.895017793594306,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046443487569889264,
      "loss": 5.2857,
      "mean_token_accuracy": 0.18895656764507293,
      "num_tokens": 43895533.0,
      "step": 19170
    },
    {
      "entropy": 5.585369825363159,
      "epoch": 1.8955120601028075,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046441583151466586,
      "loss": 5.3073,
      "mean_token_accuracy": 0.18859479576349258,
      "num_tokens": 43906616.0,
      "step": 19175
    },
    {
      "entropy": 5.582727289199829,
      "epoch": 1.896006326611309,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004643967826707803,
      "loss": 5.2478,
      "mean_token_accuracy": 0.1877148762345314,
      "num_tokens": 43917389.0,
      "step": 19180
    },
    {
      "entropy": 5.606839275360107,
      "epoch": 1.8965005931198102,
      "grad_norm": 1.5,
      "learning_rate": 0.0004643777291677045,
      "loss": 5.3168,
      "mean_token_accuracy": 0.18657988160848618,
      "num_tokens": 43929239.0,
      "step": 19185
    },
    {
      "entropy": 5.675457000732422,
      "epoch": 1.8969948596283115,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004643586710059072,
      "loss": 5.3219,
      "mean_token_accuracy": 0.1873270958662033,
      "num_tokens": 43940267.0,
      "step": 19190
    },
    {
      "entropy": 5.661898469924926,
      "epoch": 1.897489126136813,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046433960818585717,
      "loss": 5.2356,
      "mean_token_accuracy": 0.19179625064134598,
      "num_tokens": 43951505.0,
      "step": 19195
    },
    {
      "entropy": 5.537387275695801,
      "epoch": 1.8979833926453145,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00046432054070802344,
      "loss": 5.28,
      "mean_token_accuracy": 0.1934808686375618,
      "num_tokens": 43963669.0,
      "step": 19200
    },
    {
      "entropy": 5.526686668395996,
      "epoch": 1.8984776591538157,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046430146857287497,
      "loss": 5.2309,
      "mean_token_accuracy": 0.19396394044160842,
      "num_tokens": 43975814.0,
      "step": 19205
    },
    {
      "entropy": 5.576409101486206,
      "epoch": 1.898971925662317,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046428239178088106,
      "loss": 5.2433,
      "mean_token_accuracy": 0.18827661573886872,
      "num_tokens": 43987305.0,
      "step": 19210
    },
    {
      "entropy": 5.621536540985107,
      "epoch": 1.8994661921708185,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046426331033251096,
      "loss": 5.2437,
      "mean_token_accuracy": 0.1932136073708534,
      "num_tokens": 43997441.0,
      "step": 19215
    },
    {
      "entropy": 5.622320127487183,
      "epoch": 1.89996045867932,
      "grad_norm": 1.21875,
      "learning_rate": 0.000464244224228234,
      "loss": 5.2615,
      "mean_token_accuracy": 0.1928959906101227,
      "num_tokens": 44008718.0,
      "step": 19220
    },
    {
      "entropy": 5.561352062225342,
      "epoch": 1.9004547251878212,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004642251334685199,
      "loss": 5.2884,
      "mean_token_accuracy": 0.19002583175897597,
      "num_tokens": 44020785.0,
      "step": 19225
    },
    {
      "entropy": 5.527784729003907,
      "epoch": 1.9009489916963227,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00046420603805383796,
      "loss": 5.2101,
      "mean_token_accuracy": 0.1950098916888237,
      "num_tokens": 44031360.0,
      "step": 19230
    },
    {
      "entropy": 5.5717713832855225,
      "epoch": 1.9014432582048242,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004641869379846583,
      "loss": 5.1933,
      "mean_token_accuracy": 0.1987429991364479,
      "num_tokens": 44042510.0,
      "step": 19235
    },
    {
      "entropy": 5.591089200973511,
      "epoch": 1.9019375247133254,
      "grad_norm": 1.375,
      "learning_rate": 0.00046416783326145065,
      "loss": 5.2326,
      "mean_token_accuracy": 0.1879960536956787,
      "num_tokens": 44053800.0,
      "step": 19240
    },
    {
      "entropy": 5.478452539443969,
      "epoch": 1.9024317912218267,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004641487238846848,
      "loss": 5.2645,
      "mean_token_accuracy": 0.1952885791659355,
      "num_tokens": 44066727.0,
      "step": 19245
    },
    {
      "entropy": 5.646203517913818,
      "epoch": 1.9029260577303282,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00046412960985483105,
      "loss": 5.2935,
      "mean_token_accuracy": 0.1888258159160614,
      "num_tokens": 44077863.0,
      "step": 19250
    },
    {
      "entropy": 5.523417901992798,
      "epoch": 1.9034203242388297,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004641104911723596,
      "loss": 5.181,
      "mean_token_accuracy": 0.20215461254119874,
      "num_tokens": 44088218.0,
      "step": 19255
    },
    {
      "entropy": 5.661336231231689,
      "epoch": 1.903914590747331,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004640913678377407,
      "loss": 5.3853,
      "mean_token_accuracy": 0.18510138243436813,
      "num_tokens": 44101236.0,
      "step": 19260
    },
    {
      "entropy": 5.56536979675293,
      "epoch": 1.9044088572558322,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004640722398514448,
      "loss": 5.3066,
      "mean_token_accuracy": 0.18754377961158752,
      "num_tokens": 44114185.0,
      "step": 19265
    },
    {
      "entropy": 5.535950374603272,
      "epoch": 1.9049031237643337,
      "grad_norm": 1.25,
      "learning_rate": 0.00046405310721394245,
      "loss": 5.1991,
      "mean_token_accuracy": 0.19078827202320098,
      "num_tokens": 44126060.0,
      "step": 19270
    },
    {
      "entropy": 5.553313493728638,
      "epoch": 1.9053973902728352,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004640339699257044,
      "loss": 5.3101,
      "mean_token_accuracy": 0.19053225666284562,
      "num_tokens": 44137709.0,
      "step": 19275
    },
    {
      "entropy": 5.558320093154907,
      "epoch": 1.9058916567813364,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004640148279872013,
      "loss": 5.2505,
      "mean_token_accuracy": 0.19185083061456681,
      "num_tokens": 44150042.0,
      "step": 19280
    },
    {
      "entropy": 5.609388780593872,
      "epoch": 1.906385923289838,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046399568139890413,
      "loss": 5.3202,
      "mean_token_accuracy": 0.19049898386001587,
      "num_tokens": 44162926.0,
      "step": 19285
    },
    {
      "entropy": 5.575820970535278,
      "epoch": 1.9068801897983394,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046397653016128396,
      "loss": 5.1967,
      "mean_token_accuracy": 0.19688263684511184,
      "num_tokens": 44174211.0,
      "step": 19290
    },
    {
      "entropy": 5.615792608261108,
      "epoch": 1.9073744563068407,
      "grad_norm": 1.625,
      "learning_rate": 0.0004639573742748117,
      "loss": 5.3636,
      "mean_token_accuracy": 0.17644112855195998,
      "num_tokens": 44187127.0,
      "step": 19295
    },
    {
      "entropy": 5.570122337341308,
      "epoch": 1.907868722815342,
      "grad_norm": 1.328125,
      "learning_rate": 0.00046393821373995884,
      "loss": 5.2659,
      "mean_token_accuracy": 0.18965860158205033,
      "num_tokens": 44199335.0,
      "step": 19300
    },
    {
      "entropy": 5.549188899993896,
      "epoch": 1.9083629893238434,
      "grad_norm": 1.25,
      "learning_rate": 0.00046391904855719654,
      "loss": 5.1533,
      "mean_token_accuracy": 0.20120538771152496,
      "num_tokens": 44209911.0,
      "step": 19305
    },
    {
      "entropy": 5.506358337402344,
      "epoch": 1.908857255832345,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046389987872699637,
      "loss": 5.1678,
      "mean_token_accuracy": 0.195081427693367,
      "num_tokens": 44221391.0,
      "step": 19310
    },
    {
      "entropy": 5.571594524383545,
      "epoch": 1.9093515223408462,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046388070424982985,
      "loss": 5.1906,
      "mean_token_accuracy": 0.20262186825275422,
      "num_tokens": 44233138.0,
      "step": 19315
    },
    {
      "entropy": 5.564644527435303,
      "epoch": 1.9098457888493474,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004638615251261687,
      "loss": 5.2129,
      "mean_token_accuracy": 0.19723872542381288,
      "num_tokens": 44244353.0,
      "step": 19320
    },
    {
      "entropy": 5.579747581481934,
      "epoch": 1.910340055357849,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046384234135648475,
      "loss": 5.2998,
      "mean_token_accuracy": 0.1896692395210266,
      "num_tokens": 44256369.0,
      "step": 19325
    },
    {
      "entropy": 5.5785175323486325,
      "epoch": 1.9108343218663504,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046382315294124987,
      "loss": 5.2647,
      "mean_token_accuracy": 0.18858565837144853,
      "num_tokens": 44267958.0,
      "step": 19330
    },
    {
      "entropy": 5.535160303115845,
      "epoch": 1.9113285883748516,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004638039598809361,
      "loss": 5.26,
      "mean_token_accuracy": 0.19505301266908645,
      "num_tokens": 44279379.0,
      "step": 19335
    },
    {
      "entropy": 5.598449993133545,
      "epoch": 1.9118228548833531,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004637847621760158,
      "loss": 5.2252,
      "mean_token_accuracy": 0.18962267637252808,
      "num_tokens": 44289655.0,
      "step": 19340
    },
    {
      "entropy": 5.628275203704834,
      "epoch": 1.9123171213918546,
      "grad_norm": 1.546875,
      "learning_rate": 0.00046376555982696093,
      "loss": 5.321,
      "mean_token_accuracy": 0.18700777292251586,
      "num_tokens": 44301642.0,
      "step": 19345
    },
    {
      "entropy": 5.604733324050903,
      "epoch": 1.9128113879003559,
      "grad_norm": 1.25,
      "learning_rate": 0.000463746352834244,
      "loss": 5.2167,
      "mean_token_accuracy": 0.19029637724161147,
      "num_tokens": 44313260.0,
      "step": 19350
    },
    {
      "entropy": 5.610768890380859,
      "epoch": 1.9133056544088571,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004637271411983375,
      "loss": 5.2406,
      "mean_token_accuracy": 0.18816908448934555,
      "num_tokens": 44325095.0,
      "step": 19355
    },
    {
      "entropy": 5.591111850738526,
      "epoch": 1.9137999209173586,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004637079249197141,
      "loss": 5.2855,
      "mean_token_accuracy": 0.19288601726293564,
      "num_tokens": 44337381.0,
      "step": 19360
    },
    {
      "entropy": 5.538823032379151,
      "epoch": 1.91429418742586,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046368870399884644,
      "loss": 5.1961,
      "mean_token_accuracy": 0.20184044688940048,
      "num_tokens": 44348487.0,
      "step": 19365
    },
    {
      "entropy": 5.496457242965699,
      "epoch": 1.9147884539343614,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046366947843620737,
      "loss": 5.1652,
      "mean_token_accuracy": 0.19834334403276443,
      "num_tokens": 44358645.0,
      "step": 19370
    },
    {
      "entropy": 5.52705717086792,
      "epoch": 1.9152827204428626,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00046365024823226994,
      "loss": 5.2059,
      "mean_token_accuracy": 0.1986442506313324,
      "num_tokens": 44369924.0,
      "step": 19375
    },
    {
      "entropy": 5.563433265686035,
      "epoch": 1.9157769869513643,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004636310133875071,
      "loss": 5.2122,
      "mean_token_accuracy": 0.1914411172270775,
      "num_tokens": 44381684.0,
      "step": 19380
    },
    {
      "entropy": 5.684081363677978,
      "epoch": 1.9162712534598656,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046361177390239204,
      "loss": 5.3546,
      "mean_token_accuracy": 0.18746418803930281,
      "num_tokens": 44392780.0,
      "step": 19385
    },
    {
      "entropy": 5.595615720748901,
      "epoch": 1.9167655199683669,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004635925297773981,
      "loss": 5.2996,
      "mean_token_accuracy": 0.18767099529504777,
      "num_tokens": 44405330.0,
      "step": 19390
    },
    {
      "entropy": 5.633752918243408,
      "epoch": 1.9172597864768683,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004635732810129987,
      "loss": 5.3216,
      "mean_token_accuracy": 0.18701830208301545,
      "num_tokens": 44416307.0,
      "step": 19395
    },
    {
      "entropy": 5.641895198822022,
      "epoch": 1.9177540529853698,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004635540276096673,
      "loss": 5.2841,
      "mean_token_accuracy": 0.18638513535261153,
      "num_tokens": 44427481.0,
      "step": 19400
    },
    {
      "entropy": 5.508263206481933,
      "epoch": 1.918248319493871,
      "grad_norm": 1.25,
      "learning_rate": 0.00046353476956787756,
      "loss": 5.2103,
      "mean_token_accuracy": 0.19185469597578048,
      "num_tokens": 44439462.0,
      "step": 19405
    },
    {
      "entropy": 5.573173761367798,
      "epoch": 1.9187425860023724,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046351550688810317,
      "loss": 5.3033,
      "mean_token_accuracy": 0.18386206030845642,
      "num_tokens": 44452206.0,
      "step": 19410
    },
    {
      "entropy": 5.587518358230591,
      "epoch": 1.9192368525108738,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004634962395708182,
      "loss": 5.2369,
      "mean_token_accuracy": 0.19774142205715178,
      "num_tokens": 44463598.0,
      "step": 19415
    },
    {
      "entropy": 5.5582528591156,
      "epoch": 1.9197311190193753,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00046347696761649647,
      "loss": 5.2498,
      "mean_token_accuracy": 0.2010146588087082,
      "num_tokens": 44475606.0,
      "step": 19420
    },
    {
      "entropy": 5.539303159713745,
      "epoch": 1.9202253855278766,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000463457691025612,
      "loss": 5.2539,
      "mean_token_accuracy": 0.19006529748439788,
      "num_tokens": 44486948.0,
      "step": 19425
    },
    {
      "entropy": 5.555096673965454,
      "epoch": 1.920719652036378,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046343840979863924,
      "loss": 5.2116,
      "mean_token_accuracy": 0.19313639849424363,
      "num_tokens": 44497408.0,
      "step": 19430
    },
    {
      "entropy": 5.623652982711792,
      "epoch": 1.9212139185448796,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046341912393605225,
      "loss": 5.2348,
      "mean_token_accuracy": 0.19569095075130463,
      "num_tokens": 44508246.0,
      "step": 19435
    },
    {
      "entropy": 5.506088018417358,
      "epoch": 1.9217081850533808,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004633998334383256,
      "loss": 5.1685,
      "mean_token_accuracy": 0.20250879228115082,
      "num_tokens": 44520207.0,
      "step": 19440
    },
    {
      "entropy": 5.50420298576355,
      "epoch": 1.922202451561882,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046338053830593384,
      "loss": 5.1961,
      "mean_token_accuracy": 0.19127237498760224,
      "num_tokens": 44532290.0,
      "step": 19445
    },
    {
      "entropy": 5.614888095855713,
      "epoch": 1.9226967180703836,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004633612385393516,
      "loss": 5.3001,
      "mean_token_accuracy": 0.18473622500896453,
      "num_tokens": 44544305.0,
      "step": 19450
    },
    {
      "entropy": 5.586585569381714,
      "epoch": 1.923190984578885,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004633419341390537,
      "loss": 5.2952,
      "mean_token_accuracy": 0.18646225333213806,
      "num_tokens": 44556753.0,
      "step": 19455
    },
    {
      "entropy": 5.598245286941529,
      "epoch": 1.9236852510873863,
      "grad_norm": 1.140625,
      "learning_rate": 0.000463322625105515,
      "loss": 5.2421,
      "mean_token_accuracy": 0.1869811698794365,
      "num_tokens": 44567891.0,
      "step": 19460
    },
    {
      "entropy": 5.536292028427124,
      "epoch": 1.9241795175958876,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046330331143921045,
      "loss": 5.1625,
      "mean_token_accuracy": 0.20117648541927338,
      "num_tokens": 44579060.0,
      "step": 19465
    },
    {
      "entropy": 5.558172512054443,
      "epoch": 1.924673784104389,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00046328399314061525,
      "loss": 5.2799,
      "mean_token_accuracy": 0.19366924464702606,
      "num_tokens": 44591748.0,
      "step": 19470
    },
    {
      "entropy": 5.55849175453186,
      "epoch": 1.9251680506128905,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004632646702102046,
      "loss": 5.1861,
      "mean_token_accuracy": 0.19845863580703735,
      "num_tokens": 44603119.0,
      "step": 19475
    },
    {
      "entropy": 5.5738222122192385,
      "epoch": 1.9256623171213918,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004632453426484539,
      "loss": 5.2988,
      "mean_token_accuracy": 0.18786921054124833,
      "num_tokens": 44614391.0,
      "step": 19480
    },
    {
      "entropy": 5.524324750900268,
      "epoch": 1.9261565836298933,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046322601045583863,
      "loss": 5.1292,
      "mean_token_accuracy": 0.20615734159946442,
      "num_tokens": 44623950.0,
      "step": 19485
    },
    {
      "entropy": 5.514659833908081,
      "epoch": 1.9266508501383948,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004632066736328342,
      "loss": 5.1967,
      "mean_token_accuracy": 0.2011448174715042,
      "num_tokens": 44635177.0,
      "step": 19490
    },
    {
      "entropy": 5.570940351486206,
      "epoch": 1.927145116646896,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004631873321799164,
      "loss": 5.356,
      "mean_token_accuracy": 0.18559570461511612,
      "num_tokens": 44646956.0,
      "step": 19495
    },
    {
      "entropy": 5.632038164138794,
      "epoch": 1.9276393831553973,
      "grad_norm": 1.28125,
      "learning_rate": 0.000463167986097561,
      "loss": 5.2515,
      "mean_token_accuracy": 0.18865224421024324,
      "num_tokens": 44658161.0,
      "step": 19500
    },
    {
      "entropy": 5.520165205001831,
      "epoch": 1.9281336496638988,
      "grad_norm": 1.8671875,
      "learning_rate": 0.00046314863538624403,
      "loss": 5.2165,
      "mean_token_accuracy": 0.1907589241862297,
      "num_tokens": 44668823.0,
      "step": 19505
    },
    {
      "entropy": 5.529860305786133,
      "epoch": 1.9286279161724003,
      "grad_norm": 1.125,
      "learning_rate": 0.0004631292800464414,
      "loss": 5.3278,
      "mean_token_accuracy": 0.18145890831947326,
      "num_tokens": 44682285.0,
      "step": 19510
    },
    {
      "entropy": 5.661611318588257,
      "epoch": 1.9291221826809015,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046310992007862935,
      "loss": 5.3051,
      "mean_token_accuracy": 0.19078746885061265,
      "num_tokens": 44694031.0,
      "step": 19515
    },
    {
      "entropy": 5.594558238983154,
      "epoch": 1.9296164491894028,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000463090555483284,
      "loss": 5.2558,
      "mean_token_accuracy": 0.18729052990674971,
      "num_tokens": 44705636.0,
      "step": 19520
    },
    {
      "entropy": 5.582792520523071,
      "epoch": 1.9301107156979043,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046307118626088184,
      "loss": 5.2928,
      "mean_token_accuracy": 0.1860485255718231,
      "num_tokens": 44716505.0,
      "step": 19525
    },
    {
      "entropy": 5.551295423507691,
      "epoch": 1.9306049822064058,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004630518124118994,
      "loss": 5.2939,
      "mean_token_accuracy": 0.18172208666801454,
      "num_tokens": 44729298.0,
      "step": 19530
    },
    {
      "entropy": 5.572073221206665,
      "epoch": 1.931099248714907,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046303243393681307,
      "loss": 5.2346,
      "mean_token_accuracy": 0.1893751561641693,
      "num_tokens": 44740456.0,
      "step": 19535
    },
    {
      "entropy": 5.5821935653686525,
      "epoch": 1.9315935152234085,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046301305083609973,
      "loss": 5.2512,
      "mean_token_accuracy": 0.19513129144907,
      "num_tokens": 44752311.0,
      "step": 19540
    },
    {
      "entropy": 5.643353366851807,
      "epoch": 1.93208778173191,
      "grad_norm": 1.125,
      "learning_rate": 0.00046299366311023624,
      "loss": 5.292,
      "mean_token_accuracy": 0.1904216766357422,
      "num_tokens": 44763704.0,
      "step": 19545
    },
    {
      "entropy": 5.529926109313965,
      "epoch": 1.9325820482404112,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004629742707596994,
      "loss": 5.2597,
      "mean_token_accuracy": 0.19365395605564117,
      "num_tokens": 44775053.0,
      "step": 19550
    },
    {
      "entropy": 5.6153411865234375,
      "epoch": 1.9330763147489125,
      "grad_norm": 2.515625,
      "learning_rate": 0.00046295487378496643,
      "loss": 5.3611,
      "mean_token_accuracy": 0.18268836587667464,
      "num_tokens": 44786734.0,
      "step": 19555
    },
    {
      "entropy": 5.739839267730713,
      "epoch": 1.933570581257414,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046293547218651427,
      "loss": 5.406,
      "mean_token_accuracy": 0.1813743904232979,
      "num_tokens": 44798151.0,
      "step": 19560
    },
    {
      "entropy": 5.652517223358155,
      "epoch": 1.9340648477659155,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004629160659648204,
      "loss": 5.3439,
      "mean_token_accuracy": 0.18635807633399964,
      "num_tokens": 44809354.0,
      "step": 19565
    },
    {
      "entropy": 5.535543537139892,
      "epoch": 1.9345591142744167,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004628966551203622,
      "loss": 5.2561,
      "mean_token_accuracy": 0.18832801580429076,
      "num_tokens": 44820994.0,
      "step": 19570
    },
    {
      "entropy": 5.636121320724487,
      "epoch": 1.935053380782918,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046287723965361704,
      "loss": 5.3551,
      "mean_token_accuracy": 0.1882003977894783,
      "num_tokens": 44831486.0,
      "step": 19575
    },
    {
      "entropy": 5.614482831954956,
      "epoch": 1.9355476472914195,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004628578195650628,
      "loss": 5.3167,
      "mean_token_accuracy": 0.1886990875005722,
      "num_tokens": 44844078.0,
      "step": 19580
    },
    {
      "entropy": 5.519651317596436,
      "epoch": 1.936041913799921,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004628383948551769,
      "loss": 5.2105,
      "mean_token_accuracy": 0.19270310699939727,
      "num_tokens": 44854722.0,
      "step": 19585
    },
    {
      "entropy": 5.577307558059692,
      "epoch": 1.9365361803084222,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004628189655244374,
      "loss": 5.2608,
      "mean_token_accuracy": 0.19222649186849594,
      "num_tokens": 44866337.0,
      "step": 19590
    },
    {
      "entropy": 5.608915758132935,
      "epoch": 1.9370304468169237,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004627995315733222,
      "loss": 5.3141,
      "mean_token_accuracy": 0.18868790119886397,
      "num_tokens": 44878686.0,
      "step": 19595
    },
    {
      "entropy": 5.5957568168640135,
      "epoch": 1.9375247133254252,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046278009300230945,
      "loss": 5.2387,
      "mean_token_accuracy": 0.18517884612083435,
      "num_tokens": 44890339.0,
      "step": 19600
    },
    {
      "entropy": 5.669606637954712,
      "epoch": 1.9380189798339265,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004627606498118772,
      "loss": 5.3651,
      "mean_token_accuracy": 0.1811843380331993,
      "num_tokens": 44902850.0,
      "step": 19605
    },
    {
      "entropy": 5.57137770652771,
      "epoch": 1.9385132463424277,
      "grad_norm": 1.375,
      "learning_rate": 0.0004627412020025039,
      "loss": 5.2311,
      "mean_token_accuracy": 0.19209009706974028,
      "num_tokens": 44914074.0,
      "step": 19610
    },
    {
      "entropy": 5.588099431991577,
      "epoch": 1.9390075128509292,
      "grad_norm": 1.328125,
      "learning_rate": 0.00046272174957466783,
      "loss": 5.2868,
      "mean_token_accuracy": 0.18897171318531036,
      "num_tokens": 44925665.0,
      "step": 19615
    },
    {
      "entropy": 5.611680364608764,
      "epoch": 1.9395017793594307,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004627022925288476,
      "loss": 5.3356,
      "mean_token_accuracy": 0.19216190725564958,
      "num_tokens": 44936584.0,
      "step": 19620
    },
    {
      "entropy": 5.598427104949951,
      "epoch": 1.939996045867932,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004626828308655219,
      "loss": 5.2192,
      "mean_token_accuracy": 0.1929556742310524,
      "num_tokens": 44948028.0,
      "step": 19625
    },
    {
      "entropy": 5.545696115493774,
      "epoch": 1.9404903123764332,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046266336458516944,
      "loss": 5.1576,
      "mean_token_accuracy": 0.1987527310848236,
      "num_tokens": 44959504.0,
      "step": 19630
    },
    {
      "entropy": 5.580860614776611,
      "epoch": 1.940984578884935,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004626438936882691,
      "loss": 5.3012,
      "mean_token_accuracy": 0.18677317053079606,
      "num_tokens": 44971518.0,
      "step": 19635
    },
    {
      "entropy": 5.520154142379761,
      "epoch": 1.9414788453934362,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046262441817529986,
      "loss": 5.2001,
      "mean_token_accuracy": 0.19911773949861528,
      "num_tokens": 44982514.0,
      "step": 19640
    },
    {
      "entropy": 5.540999984741211,
      "epoch": 1.9419731119019374,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004626049380467408,
      "loss": 5.2016,
      "mean_token_accuracy": 0.19832732975482942,
      "num_tokens": 44995198.0,
      "step": 19645
    },
    {
      "entropy": 5.582744216918945,
      "epoch": 1.942467378410439,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046258545330307124,
      "loss": 5.2156,
      "mean_token_accuracy": 0.19470031857490538,
      "num_tokens": 45008014.0,
      "step": 19650
    },
    {
      "entropy": 5.571939516067505,
      "epoch": 1.9429616449189404,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046256596394477037,
      "loss": 5.2797,
      "mean_token_accuracy": 0.19575897455215455,
      "num_tokens": 45021169.0,
      "step": 19655
    },
    {
      "entropy": 5.535858583450318,
      "epoch": 1.9434559114274417,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004625464699723177,
      "loss": 5.1601,
      "mean_token_accuracy": 0.1978980556130409,
      "num_tokens": 45032703.0,
      "step": 19660
    },
    {
      "entropy": 5.506058692932129,
      "epoch": 1.943950177935943,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004625269713861928,
      "loss": 5.2352,
      "mean_token_accuracy": 0.19318510591983795,
      "num_tokens": 45044350.0,
      "step": 19665
    },
    {
      "entropy": 5.517756128311158,
      "epoch": 1.9444444444444444,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004625074681868753,
      "loss": 5.2197,
      "mean_token_accuracy": 0.20157890170812606,
      "num_tokens": 45055431.0,
      "step": 19670
    },
    {
      "entropy": 5.59564619064331,
      "epoch": 1.944938710952946,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004624879603748449,
      "loss": 5.2484,
      "mean_token_accuracy": 0.19636425524950027,
      "num_tokens": 45066801.0,
      "step": 19675
    },
    {
      "entropy": 5.5482871532440186,
      "epoch": 1.9454329774614472,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004624684479505817,
      "loss": 5.2189,
      "mean_token_accuracy": 0.1945919543504715,
      "num_tokens": 45078393.0,
      "step": 19680
    },
    {
      "entropy": 5.569502639770508,
      "epoch": 1.9459272439699487,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004624489309145655,
      "loss": 5.2608,
      "mean_token_accuracy": 0.191505765914917,
      "num_tokens": 45090782.0,
      "step": 19685
    },
    {
      "entropy": 5.568394994735717,
      "epoch": 1.9464215104784501,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004624294092672766,
      "loss": 5.2325,
      "mean_token_accuracy": 0.2008551374077797,
      "num_tokens": 45102079.0,
      "step": 19690
    },
    {
      "entropy": 5.561351156234741,
      "epoch": 1.9469157769869514,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004624098830091951,
      "loss": 5.2328,
      "mean_token_accuracy": 0.2017618164420128,
      "num_tokens": 45112947.0,
      "step": 19695
    },
    {
      "entropy": 5.5489341735839846,
      "epoch": 1.9474100434954527,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046239035214080146,
      "loss": 5.2599,
      "mean_token_accuracy": 0.190461228787899,
      "num_tokens": 45124577.0,
      "step": 19700
    },
    {
      "entropy": 5.572587490081787,
      "epoch": 1.9479043100039541,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004623708166625761,
      "loss": 5.2623,
      "mean_token_accuracy": 0.18926297426223754,
      "num_tokens": 45135380.0,
      "step": 19705
    },
    {
      "entropy": 5.61161732673645,
      "epoch": 1.9483985765124556,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046235127657499945,
      "loss": 5.2401,
      "mean_token_accuracy": 0.19348111897706985,
      "num_tokens": 45146926.0,
      "step": 19710
    },
    {
      "entropy": 5.5252272605896,
      "epoch": 1.9488928430209569,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004623317318785524,
      "loss": 5.2651,
      "mean_token_accuracy": 0.19426850080490113,
      "num_tokens": 45157662.0,
      "step": 19715
    },
    {
      "entropy": 5.451500654220581,
      "epoch": 1.9493871095294582,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004623121825737158,
      "loss": 5.1567,
      "mean_token_accuracy": 0.20491967350244522,
      "num_tokens": 45170239.0,
      "step": 19720
    },
    {
      "entropy": 5.568098068237305,
      "epoch": 1.9498813760379596,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046229262866097034,
      "loss": 5.2303,
      "mean_token_accuracy": 0.18728385120630264,
      "num_tokens": 45180979.0,
      "step": 19725
    },
    {
      "entropy": 5.532364797592163,
      "epoch": 1.9503756425464611,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004622730701407971,
      "loss": 5.2183,
      "mean_token_accuracy": 0.1994457483291626,
      "num_tokens": 45191160.0,
      "step": 19730
    },
    {
      "entropy": 5.560722923278808,
      "epoch": 1.9508699090549624,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004622535070136774,
      "loss": 5.236,
      "mean_token_accuracy": 0.20027394443750382,
      "num_tokens": 45202865.0,
      "step": 19735
    },
    {
      "entropy": 5.61088376045227,
      "epoch": 1.9513641755634639,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046223393928009233,
      "loss": 5.2397,
      "mean_token_accuracy": 0.19022582322359086,
      "num_tokens": 45214831.0,
      "step": 19740
    },
    {
      "entropy": 5.6594212532043455,
      "epoch": 1.9518584420719653,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046221436694052334,
      "loss": 5.359,
      "mean_token_accuracy": 0.19217007011175155,
      "num_tokens": 45226453.0,
      "step": 19745
    },
    {
      "entropy": 5.640673685073852,
      "epoch": 1.9523527085804666,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046219478999545184,
      "loss": 5.3133,
      "mean_token_accuracy": 0.1876315340399742,
      "num_tokens": 45237292.0,
      "step": 19750
    },
    {
      "entropy": 5.483549308776856,
      "epoch": 1.9528469750889679,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004621752084453595,
      "loss": 5.1399,
      "mean_token_accuracy": 0.20296119302511215,
      "num_tokens": 45248117.0,
      "step": 19755
    },
    {
      "entropy": 5.517328548431396,
      "epoch": 1.9533412415974694,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046215562229072796,
      "loss": 5.2874,
      "mean_token_accuracy": 0.17995953261852266,
      "num_tokens": 45259120.0,
      "step": 19760
    },
    {
      "entropy": 5.57343864440918,
      "epoch": 1.9538355081059708,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046213603153203906,
      "loss": 5.1818,
      "mean_token_accuracy": 0.20038791596889496,
      "num_tokens": 45269554.0,
      "step": 19765
    },
    {
      "entropy": 5.552782297134399,
      "epoch": 1.954329774614472,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004621164361697748,
      "loss": 5.2209,
      "mean_token_accuracy": 0.19018377661705016,
      "num_tokens": 45281055.0,
      "step": 19770
    },
    {
      "entropy": 5.516794729232788,
      "epoch": 1.9548240411229734,
      "grad_norm": 1.375,
      "learning_rate": 0.0004620968362044172,
      "loss": 5.1802,
      "mean_token_accuracy": 0.21001689583063127,
      "num_tokens": 45292430.0,
      "step": 19775
    },
    {
      "entropy": 5.536257076263428,
      "epoch": 1.9553183076314749,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004620772316364483,
      "loss": 5.1572,
      "mean_token_accuracy": 0.19214885234832763,
      "num_tokens": 45303417.0,
      "step": 19780
    },
    {
      "entropy": 5.577347135543823,
      "epoch": 1.9558125741399763,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004620576224663506,
      "loss": 5.3206,
      "mean_token_accuracy": 0.18164298683404922,
      "num_tokens": 45315243.0,
      "step": 19785
    },
    {
      "entropy": 5.540080928802491,
      "epoch": 1.9563068406484776,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004620380086946063,
      "loss": 5.2297,
      "mean_token_accuracy": 0.19591670483350754,
      "num_tokens": 45326746.0,
      "step": 19790
    },
    {
      "entropy": 5.538619613647461,
      "epoch": 1.956801107156979,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046201839032169794,
      "loss": 5.1899,
      "mean_token_accuracy": 0.19956814646720886,
      "num_tokens": 45338199.0,
      "step": 19795
    },
    {
      "entropy": 5.616135883331299,
      "epoch": 1.9572953736654806,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004619987673481082,
      "loss": 5.2128,
      "mean_token_accuracy": 0.1970521867275238,
      "num_tokens": 45349470.0,
      "step": 19800
    },
    {
      "entropy": 5.597288179397583,
      "epoch": 1.9577896401739818,
      "grad_norm": 1.25,
      "learning_rate": 0.00046197913977431983,
      "loss": 5.2605,
      "mean_token_accuracy": 0.18897920548915864,
      "num_tokens": 45359735.0,
      "step": 19805
    },
    {
      "entropy": 5.648198366165161,
      "epoch": 1.958283906682483,
      "grad_norm": 1.671875,
      "learning_rate": 0.00046195950760081556,
      "loss": 5.3978,
      "mean_token_accuracy": 0.18056899458169937,
      "num_tokens": 45372960.0,
      "step": 19810
    },
    {
      "entropy": 5.627838563919068,
      "epoch": 1.9587781731909846,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004619398708280784,
      "loss": 5.3193,
      "mean_token_accuracy": 0.1901567593216896,
      "num_tokens": 45385174.0,
      "step": 19815
    },
    {
      "entropy": 5.529311275482177,
      "epoch": 1.959272439699486,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004619202294565914,
      "loss": 5.2064,
      "mean_token_accuracy": 0.1971341550350189,
      "num_tokens": 45397013.0,
      "step": 19820
    },
    {
      "entropy": 5.724609422683716,
      "epoch": 1.9597667062079873,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004619005834868378,
      "loss": 5.4052,
      "mean_token_accuracy": 0.17449079304933549,
      "num_tokens": 45408863.0,
      "step": 19825
    },
    {
      "entropy": 5.578535842895508,
      "epoch": 1.9602609727164886,
      "grad_norm": 1.125,
      "learning_rate": 0.0004618809329193008,
      "loss": 5.2465,
      "mean_token_accuracy": 0.195696596801281,
      "num_tokens": 45420630.0,
      "step": 19830
    },
    {
      "entropy": 5.6050084114074705,
      "epoch": 1.96075523922499,
      "grad_norm": 1.375,
      "learning_rate": 0.0004618612777544639,
      "loss": 5.2312,
      "mean_token_accuracy": 0.1926379457116127,
      "num_tokens": 45431427.0,
      "step": 19835
    },
    {
      "entropy": 5.581892967224121,
      "epoch": 1.9612495057334915,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046184161799281057,
      "loss": 5.1469,
      "mean_token_accuracy": 0.20311880707740784,
      "num_tokens": 45442649.0,
      "step": 19840
    },
    {
      "entropy": 5.522710466384888,
      "epoch": 1.9617437722419928,
      "grad_norm": 1.40625,
      "learning_rate": 0.00046182195363482444,
      "loss": 5.2331,
      "mean_token_accuracy": 0.19719004184007644,
      "num_tokens": 45454032.0,
      "step": 19845
    },
    {
      "entropy": 5.532337713241577,
      "epoch": 1.9622380387504943,
      "grad_norm": 1.25,
      "learning_rate": 0.00046180228468098927,
      "loss": 5.2572,
      "mean_token_accuracy": 0.18732471913099288,
      "num_tokens": 45464706.0,
      "step": 19850
    },
    {
      "entropy": 5.687468719482422,
      "epoch": 1.9627323052589958,
      "grad_norm": 2.046875,
      "learning_rate": 0.00046178261113178894,
      "loss": 5.3788,
      "mean_token_accuracy": 0.17826322168111802,
      "num_tokens": 45476292.0,
      "step": 19855
    },
    {
      "entropy": 5.522291469573974,
      "epoch": 1.963226571767497,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004617629329877074,
      "loss": 5.1848,
      "mean_token_accuracy": 0.2006040558218956,
      "num_tokens": 45488189.0,
      "step": 19860
    },
    {
      "entropy": 5.610484457015991,
      "epoch": 1.9637208382759983,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004617432502492287,
      "loss": 5.2163,
      "mean_token_accuracy": 0.19530561864376067,
      "num_tokens": 45499825.0,
      "step": 19865
    },
    {
      "entropy": 5.510534763336182,
      "epoch": 1.9642151047844998,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004617235629168371,
      "loss": 5.1981,
      "mean_token_accuracy": 0.19429436922073365,
      "num_tokens": 45511164.0,
      "step": 19870
    },
    {
      "entropy": 5.567736387252808,
      "epoch": 1.9647093712930013,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004617038709910168,
      "loss": 5.3016,
      "mean_token_accuracy": 0.18901262283325196,
      "num_tokens": 45521506.0,
      "step": 19875
    },
    {
      "entropy": 5.661199903488159,
      "epoch": 1.9652036378015025,
      "grad_norm": 3.1875,
      "learning_rate": 0.0004616841744722524,
      "loss": 5.3351,
      "mean_token_accuracy": 0.18875936716794967,
      "num_tokens": 45531622.0,
      "step": 19880
    },
    {
      "entropy": 5.582238960266113,
      "epoch": 1.9656979043100038,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046166447336102836,
      "loss": 5.1977,
      "mean_token_accuracy": 0.19730779081583022,
      "num_tokens": 45541565.0,
      "step": 19885
    },
    {
      "entropy": 5.603276014328003,
      "epoch": 1.9661921708185055,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046164476765782927,
      "loss": 5.209,
      "mean_token_accuracy": 0.19174007922410966,
      "num_tokens": 45552943.0,
      "step": 19890
    },
    {
      "entropy": 5.526623392105103,
      "epoch": 1.9666864373270068,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046162505736313997,
      "loss": 5.2833,
      "mean_token_accuracy": 0.19296058565378188,
      "num_tokens": 45563914.0,
      "step": 19895
    },
    {
      "entropy": 5.498500633239746,
      "epoch": 1.967180703835508,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004616053424774453,
      "loss": 5.1767,
      "mean_token_accuracy": 0.1936721086502075,
      "num_tokens": 45575730.0,
      "step": 19900
    },
    {
      "entropy": 5.649481534957886,
      "epoch": 1.9676749703440095,
      "grad_norm": 1.40625,
      "learning_rate": 0.00046158562300123023,
      "loss": 5.3087,
      "mean_token_accuracy": 0.18902114033699036,
      "num_tokens": 45588369.0,
      "step": 19905
    },
    {
      "entropy": 5.5415239334106445,
      "epoch": 1.968169236852511,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046156589893498,
      "loss": 5.1559,
      "mean_token_accuracy": 0.1993715688586235,
      "num_tokens": 45599235.0,
      "step": 19910
    },
    {
      "entropy": 5.510575819015503,
      "epoch": 1.9686635033610123,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00046154617027917955,
      "loss": 5.1911,
      "mean_token_accuracy": 0.2003573387861252,
      "num_tokens": 45610066.0,
      "step": 19915
    },
    {
      "entropy": 5.511261463165283,
      "epoch": 1.9691577698695135,
      "grad_norm": 1.328125,
      "learning_rate": 0.00046152643703431447,
      "loss": 5.1971,
      "mean_token_accuracy": 0.19434211552143096,
      "num_tokens": 45621543.0,
      "step": 19920
    },
    {
      "entropy": 5.552016305923462,
      "epoch": 1.969652036378015,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046150669920087004,
      "loss": 5.334,
      "mean_token_accuracy": 0.180738465487957,
      "num_tokens": 45634984.0,
      "step": 19925
    },
    {
      "entropy": 5.662458992004394,
      "epoch": 1.9701463028865165,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004614869567793319,
      "loss": 5.267,
      "mean_token_accuracy": 0.1927230551838875,
      "num_tokens": 45646778.0,
      "step": 19930
    },
    {
      "entropy": 5.557553482055664,
      "epoch": 1.9706405693950177,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046146720977018567,
      "loss": 5.2277,
      "mean_token_accuracy": 0.19547680467367173,
      "num_tokens": 45658103.0,
      "step": 19935
    },
    {
      "entropy": 5.553086471557617,
      "epoch": 1.971134835903519,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004614474581739172,
      "loss": 5.1842,
      "mean_token_accuracy": 0.19174272269010545,
      "num_tokens": 45669166.0,
      "step": 19940
    },
    {
      "entropy": 5.522197628021241,
      "epoch": 1.9716291024120207,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004614277019910123,
      "loss": 5.2439,
      "mean_token_accuracy": 0.18804922252893447,
      "num_tokens": 45680667.0,
      "step": 19945
    },
    {
      "entropy": 5.582298374176025,
      "epoch": 1.972123368920522,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004614079412219571,
      "loss": 5.3445,
      "mean_token_accuracy": 0.18462157100439072,
      "num_tokens": 45692347.0,
      "step": 19950
    },
    {
      "entropy": 5.664348840713501,
      "epoch": 1.9726176354290232,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004613881758672375,
      "loss": 5.3019,
      "mean_token_accuracy": 0.19526259005069732,
      "num_tokens": 45703972.0,
      "step": 19955
    },
    {
      "entropy": 5.551305627822876,
      "epoch": 1.9731119019375247,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046136840592733995,
      "loss": 5.2061,
      "mean_token_accuracy": 0.19083081036806107,
      "num_tokens": 45715446.0,
      "step": 19960
    },
    {
      "entropy": 5.506358575820923,
      "epoch": 1.9736061684460262,
      "grad_norm": 2.109375,
      "learning_rate": 0.00046134863140275066,
      "loss": 5.2027,
      "mean_token_accuracy": 0.20202235132455826,
      "num_tokens": 45727210.0,
      "step": 19965
    },
    {
      "entropy": 5.619771718978882,
      "epoch": 1.9741004349545275,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046132885229395606,
      "loss": 5.24,
      "mean_token_accuracy": 0.19156670421361924,
      "num_tokens": 45738235.0,
      "step": 19970
    },
    {
      "entropy": 5.591773176193238,
      "epoch": 1.9745947014630287,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004613090686014429,
      "loss": 5.2398,
      "mean_token_accuracy": 0.19855031371116638,
      "num_tokens": 45750743.0,
      "step": 19975
    },
    {
      "entropy": 5.606315803527832,
      "epoch": 1.9750889679715302,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046128928032569775,
      "loss": 5.2601,
      "mean_token_accuracy": 0.1904052436351776,
      "num_tokens": 45761989.0,
      "step": 19980
    },
    {
      "entropy": 5.526347351074219,
      "epoch": 1.9755832344800317,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004612694874672073,
      "loss": 5.3045,
      "mean_token_accuracy": 0.19031201153993607,
      "num_tokens": 45775046.0,
      "step": 19985
    },
    {
      "entropy": 5.676477527618408,
      "epoch": 1.976077500988533,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004612496900264586,
      "loss": 5.3397,
      "mean_token_accuracy": 0.18965043425559996,
      "num_tokens": 45786776.0,
      "step": 19990
    },
    {
      "entropy": 5.619040203094483,
      "epoch": 1.9765717674970344,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004612298880039387,
      "loss": 5.2079,
      "mean_token_accuracy": 0.193252994120121,
      "num_tokens": 45797675.0,
      "step": 19995
    },
    {
      "entropy": 5.565063953399658,
      "epoch": 1.977066034005536,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004612100814001346,
      "loss": 5.2542,
      "mean_token_accuracy": 0.1882918357849121,
      "num_tokens": 45809897.0,
      "step": 20000
    },
    {
      "entropy": 5.6252992153167725,
      "epoch": 1.9775603005140372,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004611902702155336,
      "loss": 5.3711,
      "mean_token_accuracy": 0.18630560785531997,
      "num_tokens": 45822330.0,
      "step": 20005
    },
    {
      "entropy": 5.611008071899414,
      "epoch": 1.9780545670225385,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004611704544506232,
      "loss": 5.2279,
      "mean_token_accuracy": 0.1941098690032959,
      "num_tokens": 45834705.0,
      "step": 20010
    },
    {
      "entropy": 5.553883600234985,
      "epoch": 1.97854883353104,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004611506341058906,
      "loss": 5.2876,
      "mean_token_accuracy": 0.1864999607205391,
      "num_tokens": 45846646.0,
      "step": 20015
    },
    {
      "entropy": 5.532613134384155,
      "epoch": 1.9790431000395414,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046113080918182346,
      "loss": 5.2062,
      "mean_token_accuracy": 0.20140762329101564,
      "num_tokens": 45857252.0,
      "step": 20020
    },
    {
      "entropy": 5.5130424976348875,
      "epoch": 1.9795373665480427,
      "grad_norm": 1.71875,
      "learning_rate": 0.00046111097967890975,
      "loss": 5.171,
      "mean_token_accuracy": 0.2004624754190445,
      "num_tokens": 45867780.0,
      "step": 20025
    },
    {
      "entropy": 5.648492479324341,
      "epoch": 1.980031633056544,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004610911455976369,
      "loss": 5.2898,
      "mean_token_accuracy": 0.18554428219795227,
      "num_tokens": 45878998.0,
      "step": 20030
    },
    {
      "entropy": 5.550714015960693,
      "epoch": 1.9805258995650454,
      "grad_norm": 1.125,
      "learning_rate": 0.00046107130693849303,
      "loss": 5.1342,
      "mean_token_accuracy": 0.1965284526348114,
      "num_tokens": 45889685.0,
      "step": 20035
    },
    {
      "entropy": 5.527384567260742,
      "epoch": 1.981020166073547,
      "grad_norm": 1.25,
      "learning_rate": 0.0004610514637019662,
      "loss": 5.2493,
      "mean_token_accuracy": 0.19591145664453508,
      "num_tokens": 45900887.0,
      "step": 20040
    },
    {
      "entropy": 5.505190896987915,
      "epoch": 1.9815144325820482,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004610316158885444,
      "loss": 5.3216,
      "mean_token_accuracy": 0.18752169162034987,
      "num_tokens": 45913860.0,
      "step": 20045
    },
    {
      "entropy": 5.571532678604126,
      "epoch": 1.9820086990905497,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004610117634987161,
      "loss": 5.2101,
      "mean_token_accuracy": 0.19395423978567122,
      "num_tokens": 45924669.0,
      "step": 20050
    },
    {
      "entropy": 5.547738027572632,
      "epoch": 1.9825029655990511,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046099190653296946,
      "loss": 5.1991,
      "mean_token_accuracy": 0.18883865773677827,
      "num_tokens": 45937013.0,
      "step": 20055
    },
    {
      "entropy": 5.5249800205230715,
      "epoch": 1.9829972321075524,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004609720449917932,
      "loss": 5.1931,
      "mean_token_accuracy": 0.19548698663711547,
      "num_tokens": 45950063.0,
      "step": 20060
    },
    {
      "entropy": 5.554518175125122,
      "epoch": 1.9834914986160537,
      "grad_norm": 1.25,
      "learning_rate": 0.00046095217887567565,
      "loss": 5.2767,
      "mean_token_accuracy": 0.19285016506910324,
      "num_tokens": 45962136.0,
      "step": 20065
    },
    {
      "entropy": 5.560231637954712,
      "epoch": 1.9839857651245552,
      "grad_norm": 1.375,
      "learning_rate": 0.0004609323081851057,
      "loss": 5.1507,
      "mean_token_accuracy": 0.1959804281592369,
      "num_tokens": 45974712.0,
      "step": 20070
    },
    {
      "entropy": 5.526361465454102,
      "epoch": 1.9844800316330566,
      "grad_norm": 1.375,
      "learning_rate": 0.0004609124329205721,
      "loss": 5.2555,
      "mean_token_accuracy": 0.19503979980945588,
      "num_tokens": 45985670.0,
      "step": 20075
    },
    {
      "entropy": 5.606738519668579,
      "epoch": 1.984974298141558,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004608925530825638,
      "loss": 5.334,
      "mean_token_accuracy": 0.18032826632261276,
      "num_tokens": 45997215.0,
      "step": 20080
    },
    {
      "entropy": 5.538951253890991,
      "epoch": 1.9854685646500592,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046087266867156997,
      "loss": 5.2456,
      "mean_token_accuracy": 0.1949484020471573,
      "num_tokens": 46008479.0,
      "step": 20085
    },
    {
      "entropy": 5.596453428268433,
      "epoch": 1.9859628311585606,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046085277968807955,
      "loss": 5.2074,
      "mean_token_accuracy": 0.19372638463973998,
      "num_tokens": 46019504.0,
      "step": 20090
    },
    {
      "entropy": 5.560030174255371,
      "epoch": 1.9864570976670621,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004608328861325819,
      "loss": 5.2727,
      "mean_token_accuracy": 0.18828368782997132,
      "num_tokens": 46030218.0,
      "step": 20095
    },
    {
      "entropy": 5.595650482177734,
      "epoch": 1.9869513641755634,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004608129880055665,
      "loss": 5.2661,
      "mean_token_accuracy": 0.19005829095840454,
      "num_tokens": 46041767.0,
      "step": 20100
    },
    {
      "entropy": 5.541061592102051,
      "epoch": 1.9874456306840649,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00046079308530752274,
      "loss": 5.1328,
      "mean_token_accuracy": 0.20187895596027375,
      "num_tokens": 46052792.0,
      "step": 20105
    },
    {
      "entropy": 5.5828554153442385,
      "epoch": 1.9879398971925664,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00046077317803894024,
      "loss": 5.3241,
      "mean_token_accuracy": 0.18713027834892274,
      "num_tokens": 46064609.0,
      "step": 20110
    },
    {
      "entropy": 5.525623846054077,
      "epoch": 1.9884341637010676,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00046075326620030876,
      "loss": 5.1805,
      "mean_token_accuracy": 0.19658640623092652,
      "num_tokens": 46075624.0,
      "step": 20115
    },
    {
      "entropy": 5.547412109375,
      "epoch": 1.9889284302095689,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046073334979211813,
      "loss": 5.2219,
      "mean_token_accuracy": 0.19744790941476822,
      "num_tokens": 46086455.0,
      "step": 20120
    },
    {
      "entropy": 5.584542846679687,
      "epoch": 1.9894226967180704,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004607134288148582,
      "loss": 5.2892,
      "mean_token_accuracy": 0.18693042546510696,
      "num_tokens": 46097362.0,
      "step": 20125
    },
    {
      "entropy": 5.507166481018066,
      "epoch": 1.9899169632265719,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046069350326901915,
      "loss": 5.1803,
      "mean_token_accuracy": 0.19810830801725388,
      "num_tokens": 46107811.0,
      "step": 20130
    },
    {
      "entropy": 5.628103399276734,
      "epoch": 1.9904112297350731,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004606735731550912,
      "loss": 5.4053,
      "mean_token_accuracy": 0.1753584623336792,
      "num_tokens": 46120178.0,
      "step": 20135
    },
    {
      "entropy": 5.713874816894531,
      "epoch": 1.9909054962435744,
      "grad_norm": 1.359375,
      "learning_rate": 0.00046065363847356443,
      "loss": 5.3044,
      "mean_token_accuracy": 0.1894799068570137,
      "num_tokens": 46131416.0,
      "step": 20140
    },
    {
      "entropy": 5.628296899795532,
      "epoch": 1.991399762752076,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046063369922492943,
      "loss": 5.2631,
      "mean_token_accuracy": 0.19796158224344254,
      "num_tokens": 46142195.0,
      "step": 20145
    },
    {
      "entropy": 5.518492078781128,
      "epoch": 1.9918940292605773,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004606137554096766,
      "loss": 5.2909,
      "mean_token_accuracy": 0.18575461953878403,
      "num_tokens": 46154537.0,
      "step": 20150
    },
    {
      "entropy": 5.626855516433716,
      "epoch": 1.9923882957690786,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004605938070282965,
      "loss": 5.2422,
      "mean_token_accuracy": 0.19054343849420546,
      "num_tokens": 46166711.0,
      "step": 20155
    },
    {
      "entropy": 5.5353936672210695,
      "epoch": 1.99288256227758,
      "grad_norm": 1.25,
      "learning_rate": 0.00046057385408128013,
      "loss": 5.1891,
      "mean_token_accuracy": 0.19557373225688934,
      "num_tokens": 46178509.0,
      "step": 20160
    },
    {
      "entropy": 5.51286129951477,
      "epoch": 1.9933768287860816,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000460553896569118,
      "loss": 5.231,
      "mean_token_accuracy": 0.19280512928962706,
      "num_tokens": 46188413.0,
      "step": 20165
    },
    {
      "entropy": 5.508239459991455,
      "epoch": 1.9938710952945828,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004605339344923014,
      "loss": 5.2405,
      "mean_token_accuracy": 0.19172760993242263,
      "num_tokens": 46199708.0,
      "step": 20170
    },
    {
      "entropy": 5.539658546447754,
      "epoch": 1.994365361803084,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046051396785132113,
      "loss": 5.2213,
      "mean_token_accuracy": 0.1972035989165306,
      "num_tokens": 46210670.0,
      "step": 20175
    },
    {
      "entropy": 5.573497486114502,
      "epoch": 1.9948596283115856,
      "grad_norm": 1.25,
      "learning_rate": 0.0004604939966466684,
      "loss": 5.2151,
      "mean_token_accuracy": 0.19435738474130632,
      "num_tokens": 46220935.0,
      "step": 20180
    },
    {
      "entropy": 5.565096092224121,
      "epoch": 1.995353894820087,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004604740208788347,
      "loss": 5.2452,
      "mean_token_accuracy": 0.19673842191696167,
      "num_tokens": 46232651.0,
      "step": 20185
    },
    {
      "entropy": 5.59782395362854,
      "epoch": 1.9958481613285883,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00046045404054831124,
      "loss": 5.2138,
      "mean_token_accuracy": 0.19382842779159545,
      "num_tokens": 46244308.0,
      "step": 20190
    },
    {
      "entropy": 5.639287376403809,
      "epoch": 1.9963424278370896,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046043405565558964,
      "loss": 5.3238,
      "mean_token_accuracy": 0.18887805789709092,
      "num_tokens": 46256941.0,
      "step": 20195
    },
    {
      "entropy": 5.567622137069702,
      "epoch": 1.9968366943455913,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046041406620116144,
      "loss": 5.3042,
      "mean_token_accuracy": 0.18454636186361312,
      "num_tokens": 46270557.0,
      "step": 20200
    },
    {
      "entropy": 5.614986753463745,
      "epoch": 1.9973309608540926,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004603940721855185,
      "loss": 5.2085,
      "mean_token_accuracy": 0.19403827637434007,
      "num_tokens": 46280940.0,
      "step": 20205
    },
    {
      "entropy": 5.518739032745361,
      "epoch": 1.9978252273625938,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004603740736091527,
      "loss": 5.2957,
      "mean_token_accuracy": 0.19323249459266661,
      "num_tokens": 46292358.0,
      "step": 20210
    },
    {
      "entropy": 5.60916051864624,
      "epoch": 1.9983194938710953,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004603540704725558,
      "loss": 5.321,
      "mean_token_accuracy": 0.19341258853673934,
      "num_tokens": 46303210.0,
      "step": 20215
    },
    {
      "entropy": 5.530276966094971,
      "epoch": 1.9988137603795968,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004603340627762201,
      "loss": 5.1948,
      "mean_token_accuracy": 0.19728601574897767,
      "num_tokens": 46315152.0,
      "step": 20220
    },
    {
      "entropy": 5.531363010406494,
      "epoch": 1.999308026888098,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046031405052063756,
      "loss": 5.205,
      "mean_token_accuracy": 0.19989322274923324,
      "num_tokens": 46325776.0,
      "step": 20225
    },
    {
      "entropy": 5.550169897079468,
      "epoch": 1.9998022933965993,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046029403370630086,
      "loss": 5.2009,
      "mean_token_accuracy": 0.20072553753852845,
      "num_tokens": 46336797.0,
      "step": 20230
    },
    {
      "entropy": 5.584086227416992,
      "epoch": 2.000296559905101,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046027401233370203,
      "loss": 5.2196,
      "mean_token_accuracy": 0.19337949752807618,
      "num_tokens": 46346327.0,
      "step": 20235
    },
    {
      "entropy": 5.507830953598022,
      "epoch": 2.0007908264136023,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046025398640333373,
      "loss": 5.1679,
      "mean_token_accuracy": 0.1936369851231575,
      "num_tokens": 46358056.0,
      "step": 20240
    },
    {
      "entropy": 5.529959011077881,
      "epoch": 2.0012850929221035,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046023395591568867,
      "loss": 5.1209,
      "mean_token_accuracy": 0.1993355333805084,
      "num_tokens": 46369792.0,
      "step": 20245
    },
    {
      "entropy": 5.609037208557129,
      "epoch": 2.001779359430605,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046021392087125957,
      "loss": 5.2188,
      "mean_token_accuracy": 0.19570821076631545,
      "num_tokens": 46381224.0,
      "step": 20250
    },
    {
      "entropy": 5.621746253967285,
      "epoch": 2.0022736259391065,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004601938812705393,
      "loss": 5.2133,
      "mean_token_accuracy": 0.19333819150924683,
      "num_tokens": 46393788.0,
      "step": 20255
    },
    {
      "entropy": 5.520979309082032,
      "epoch": 2.002767892447608,
      "grad_norm": 1.25,
      "learning_rate": 0.00046017383711402076,
      "loss": 5.1367,
      "mean_token_accuracy": 0.19376198053359986,
      "num_tokens": 46405742.0,
      "step": 20260
    },
    {
      "entropy": 5.5541774272918705,
      "epoch": 2.003262158956109,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004601537884021972,
      "loss": 5.2214,
      "mean_token_accuracy": 0.19490304589271545,
      "num_tokens": 46417276.0,
      "step": 20265
    },
    {
      "entropy": 5.494325923919678,
      "epoch": 2.0037564254646103,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004601337351355617,
      "loss": 5.1115,
      "mean_token_accuracy": 0.20242203772068024,
      "num_tokens": 46427950.0,
      "step": 20270
    },
    {
      "entropy": 5.497605562210083,
      "epoch": 2.004250691973112,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004601136773146075,
      "loss": 5.1688,
      "mean_token_accuracy": 0.20465146601200104,
      "num_tokens": 46439200.0,
      "step": 20275
    },
    {
      "entropy": 5.633645677566529,
      "epoch": 2.0047449584816133,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004600936149398282,
      "loss": 5.2613,
      "mean_token_accuracy": 0.19165722131729127,
      "num_tokens": 46451176.0,
      "step": 20280
    },
    {
      "entropy": 5.562808418273926,
      "epoch": 2.0052392249901145,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00046007354801171725,
      "loss": 5.1771,
      "mean_token_accuracy": 0.198166361451149,
      "num_tokens": 46462798.0,
      "step": 20285
    },
    {
      "entropy": 5.555772829055786,
      "epoch": 2.0057334914986162,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004600534765307682,
      "loss": 5.1953,
      "mean_token_accuracy": 0.19553897380828858,
      "num_tokens": 46473914.0,
      "step": 20290
    },
    {
      "entropy": 5.646041584014893,
      "epoch": 2.0062277580071175,
      "grad_norm": 1.4375,
      "learning_rate": 0.00046003340049747513,
      "loss": 5.2398,
      "mean_token_accuracy": 0.19778451919555665,
      "num_tokens": 46485089.0,
      "step": 20295
    },
    {
      "entropy": 5.589870405197144,
      "epoch": 2.0067220245156188,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004600133199123316,
      "loss": 5.1901,
      "mean_token_accuracy": 0.19143034666776657,
      "num_tokens": 46497115.0,
      "step": 20300
    },
    {
      "entropy": 5.4115537166595455,
      "epoch": 2.00721629102412,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004599932347758316,
      "loss": 5.0796,
      "mean_token_accuracy": 0.20637826174497603,
      "num_tokens": 46508845.0,
      "step": 20305
    },
    {
      "entropy": 5.435155630111694,
      "epoch": 2.0077105575326217,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004599731450884694,
      "loss": 5.0944,
      "mean_token_accuracy": 0.20173050016164779,
      "num_tokens": 46520105.0,
      "step": 20310
    },
    {
      "entropy": 5.506902456283569,
      "epoch": 2.008204824041123,
      "grad_norm": 1.25,
      "learning_rate": 0.0004599530508507392,
      "loss": 5.12,
      "mean_token_accuracy": 0.1984819293022156,
      "num_tokens": 46530701.0,
      "step": 20315
    },
    {
      "entropy": 5.581664371490478,
      "epoch": 2.0086990905496243,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004599329520631351,
      "loss": 5.1905,
      "mean_token_accuracy": 0.19773726910352707,
      "num_tokens": 46542614.0,
      "step": 20320
    },
    {
      "entropy": 5.552907419204712,
      "epoch": 2.009193357058126,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045991284872615184,
      "loss": 5.1529,
      "mean_token_accuracy": 0.20513053387403488,
      "num_tokens": 46553624.0,
      "step": 20325
    },
    {
      "entropy": 5.506176328659057,
      "epoch": 2.0096876235666272,
      "grad_norm": 1.125,
      "learning_rate": 0.00045989274084028375,
      "loss": 5.1627,
      "mean_token_accuracy": 0.19421325176954268,
      "num_tokens": 46565599.0,
      "step": 20330
    },
    {
      "entropy": 5.5152122497558596,
      "epoch": 2.0101818900751285,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004598726284060256,
      "loss": 5.1574,
      "mean_token_accuracy": 0.20479942858219147,
      "num_tokens": 46576730.0,
      "step": 20335
    },
    {
      "entropy": 5.581531286239624,
      "epoch": 2.0106761565836297,
      "grad_norm": 1.140625,
      "learning_rate": 0.000459852511423872,
      "loss": 5.1799,
      "mean_token_accuracy": 0.19118578135967254,
      "num_tokens": 46587435.0,
      "step": 20340
    },
    {
      "entropy": 5.497976541519165,
      "epoch": 2.0111704230921315,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004598323898943181,
      "loss": 5.0724,
      "mean_token_accuracy": 0.19877846986055375,
      "num_tokens": 46598431.0,
      "step": 20345
    },
    {
      "entropy": 5.551491165161133,
      "epoch": 2.0116646896006327,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004598122638178586,
      "loss": 5.2097,
      "mean_token_accuracy": 0.18891131281852722,
      "num_tokens": 46609343.0,
      "step": 20350
    },
    {
      "entropy": 5.633590412139893,
      "epoch": 2.012158956109134,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045979213319498883,
      "loss": 5.2134,
      "mean_token_accuracy": 0.19764048457145691,
      "num_tokens": 46621844.0,
      "step": 20355
    },
    {
      "entropy": 5.499310493469238,
      "epoch": 2.0126532226176352,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004597719980262039,
      "loss": 5.1519,
      "mean_token_accuracy": 0.19989294558763504,
      "num_tokens": 46633278.0,
      "step": 20360
    },
    {
      "entropy": 5.514910316467285,
      "epoch": 2.013147489126137,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004597518583119991,
      "loss": 5.1832,
      "mean_token_accuracy": 0.1948420986533165,
      "num_tokens": 46644375.0,
      "step": 20365
    },
    {
      "entropy": 5.591037178039551,
      "epoch": 2.013641755634638,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004597317140528699,
      "loss": 5.2331,
      "mean_token_accuracy": 0.19293177723884583,
      "num_tokens": 46655631.0,
      "step": 20370
    },
    {
      "entropy": 5.474056386947632,
      "epoch": 2.0141360221431395,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045971156524931195,
      "loss": 5.0927,
      "mean_token_accuracy": 0.21048218309879302,
      "num_tokens": 46667095.0,
      "step": 20375
    },
    {
      "entropy": 5.573895978927612,
      "epoch": 2.014630288651641,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004596914119018209,
      "loss": 5.1783,
      "mean_token_accuracy": 0.19378681778907775,
      "num_tokens": 46678524.0,
      "step": 20380
    },
    {
      "entropy": 5.5605268478393555,
      "epoch": 2.0151245551601424,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045967125401089233,
      "loss": 5.1498,
      "mean_token_accuracy": 0.19463196992874146,
      "num_tokens": 46689785.0,
      "step": 20385
    },
    {
      "entropy": 5.568623876571655,
      "epoch": 2.0156188216686437,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004596510915770223,
      "loss": 5.1786,
      "mean_token_accuracy": 0.1896626442670822,
      "num_tokens": 46701109.0,
      "step": 20390
    },
    {
      "entropy": 5.53907995223999,
      "epoch": 2.016113088177145,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004596309246007068,
      "loss": 5.1325,
      "mean_token_accuracy": 0.19810078740119935,
      "num_tokens": 46711381.0,
      "step": 20395
    },
    {
      "entropy": 5.593878984451294,
      "epoch": 2.0166073546856467,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004596107530824419,
      "loss": 5.2077,
      "mean_token_accuracy": 0.19844861477613449,
      "num_tokens": 46722237.0,
      "step": 20400
    },
    {
      "entropy": 5.501253890991211,
      "epoch": 2.017101621194148,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045959057702272383,
      "loss": 5.1584,
      "mean_token_accuracy": 0.20107773393392564,
      "num_tokens": 46733679.0,
      "step": 20405
    },
    {
      "entropy": 5.591819763183594,
      "epoch": 2.017595887702649,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004595703964220489,
      "loss": 5.2617,
      "mean_token_accuracy": 0.1872195914387703,
      "num_tokens": 46746207.0,
      "step": 20410
    },
    {
      "entropy": 5.557614374160766,
      "epoch": 2.0180901542111505,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004595502112809137,
      "loss": 5.1789,
      "mean_token_accuracy": 0.1948267251253128,
      "num_tokens": 46756607.0,
      "step": 20415
    },
    {
      "entropy": 5.5202168941497805,
      "epoch": 2.018584420719652,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004595300215998145,
      "loss": 5.2106,
      "mean_token_accuracy": 0.1983547255396843,
      "num_tokens": 46768383.0,
      "step": 20420
    },
    {
      "entropy": 5.623150157928467,
      "epoch": 2.0190786872281534,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004595098273792482,
      "loss": 5.2043,
      "mean_token_accuracy": 0.1905844032764435,
      "num_tokens": 46779511.0,
      "step": 20425
    },
    {
      "entropy": 5.540560531616211,
      "epoch": 2.0195729537366547,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004594896286197116,
      "loss": 5.1727,
      "mean_token_accuracy": 0.19906250536441802,
      "num_tokens": 46791281.0,
      "step": 20430
    },
    {
      "entropy": 5.48981614112854,
      "epoch": 2.0200672202451564,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045946942532170147,
      "loss": 5.1499,
      "mean_token_accuracy": 0.20337039828300477,
      "num_tokens": 46802090.0,
      "step": 20435
    },
    {
      "entropy": 5.548567438125611,
      "epoch": 2.0205614867536577,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004594492174857149,
      "loss": 5.1877,
      "mean_token_accuracy": 0.19649834781885148,
      "num_tokens": 46814688.0,
      "step": 20440
    },
    {
      "entropy": 5.597362470626831,
      "epoch": 2.021055753262159,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045942900511224886,
      "loss": 5.2055,
      "mean_token_accuracy": 0.19544486105442047,
      "num_tokens": 46826212.0,
      "step": 20445
    },
    {
      "entropy": 5.524536418914795,
      "epoch": 2.02155001977066,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004594087882018008,
      "loss": 5.1379,
      "mean_token_accuracy": 0.20085811764001846,
      "num_tokens": 46836583.0,
      "step": 20450
    },
    {
      "entropy": 5.520880031585693,
      "epoch": 2.022044286279162,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004593885667548679,
      "loss": 5.1799,
      "mean_token_accuracy": 0.19211812317371368,
      "num_tokens": 46847973.0,
      "step": 20455
    },
    {
      "entropy": 5.4943647384643555,
      "epoch": 2.022538552787663,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045936834077194754,
      "loss": 5.1649,
      "mean_token_accuracy": 0.19420519918203355,
      "num_tokens": 46860113.0,
      "step": 20460
    },
    {
      "entropy": 5.648628282546997,
      "epoch": 2.0230328192961644,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004593481102535375,
      "loss": 5.2685,
      "mean_token_accuracy": 0.20077752619981765,
      "num_tokens": 46871578.0,
      "step": 20465
    },
    {
      "entropy": 5.544854164123535,
      "epoch": 2.0235270858046657,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045932787520013533,
      "loss": 5.1581,
      "mean_token_accuracy": 0.19711616188287734,
      "num_tokens": 46883701.0,
      "step": 20470
    },
    {
      "entropy": 5.459692001342773,
      "epoch": 2.0240213523131674,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004593076356122389,
      "loss": 5.115,
      "mean_token_accuracy": 0.1996801182627678,
      "num_tokens": 46895107.0,
      "step": 20475
    },
    {
      "entropy": 5.450007581710816,
      "epoch": 2.0245156188216686,
      "grad_norm": 1.296875,
      "learning_rate": 0.000459287391490346,
      "loss": 5.0208,
      "mean_token_accuracy": 0.20859412848949432,
      "num_tokens": 46906984.0,
      "step": 20480
    },
    {
      "entropy": 5.53294415473938,
      "epoch": 2.02500988533017,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045926714283495466,
      "loss": 5.2098,
      "mean_token_accuracy": 0.1961056888103485,
      "num_tokens": 46918582.0,
      "step": 20485
    },
    {
      "entropy": 5.573688268661499,
      "epoch": 2.0255041518386716,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00045924688964656303,
      "loss": 5.2018,
      "mean_token_accuracy": 0.19821869134902953,
      "num_tokens": 46930799.0,
      "step": 20490
    },
    {
      "entropy": 5.5752997398376465,
      "epoch": 2.025998418347173,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045922663192566926,
      "loss": 5.1261,
      "mean_token_accuracy": 0.19986872375011444,
      "num_tokens": 46941517.0,
      "step": 20495
    },
    {
      "entropy": 5.560674953460693,
      "epoch": 2.026492684855674,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045920636967277187,
      "loss": 5.234,
      "mean_token_accuracy": 0.19469612538814546,
      "num_tokens": 46953160.0,
      "step": 20500
    },
    {
      "entropy": 5.5225818157196045,
      "epoch": 2.0269869513641754,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045918610288836915,
      "loss": 5.049,
      "mean_token_accuracy": 0.20391114205121993,
      "num_tokens": 46964284.0,
      "step": 20505
    },
    {
      "entropy": 5.531648349761963,
      "epoch": 2.027481217872677,
      "grad_norm": 1.25,
      "learning_rate": 0.0004591658315729598,
      "loss": 5.2003,
      "mean_token_accuracy": 0.19108957648277283,
      "num_tokens": 46975402.0,
      "step": 20510
    },
    {
      "entropy": 5.544076728820801,
      "epoch": 2.0279754843811784,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004591455557270423,
      "loss": 5.2042,
      "mean_token_accuracy": 0.19728613495826722,
      "num_tokens": 46988629.0,
      "step": 20515
    },
    {
      "entropy": 5.557766532897949,
      "epoch": 2.0284697508896796,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045912527535111567,
      "loss": 5.2003,
      "mean_token_accuracy": 0.20109121054410933,
      "num_tokens": 47000497.0,
      "step": 20520
    },
    {
      "entropy": 5.503021335601806,
      "epoch": 2.028964017398181,
      "grad_norm": 1.328125,
      "learning_rate": 0.00045910499044567865,
      "loss": 5.1748,
      "mean_token_accuracy": 0.19138241559267044,
      "num_tokens": 47010841.0,
      "step": 20525
    },
    {
      "entropy": 5.507014369964599,
      "epoch": 2.0294582839066826,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004590847010112303,
      "loss": 5.1287,
      "mean_token_accuracy": 0.1996312364935875,
      "num_tokens": 47023131.0,
      "step": 20530
    },
    {
      "entropy": 5.610878610610962,
      "epoch": 2.029952550415184,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045906440704826973,
      "loss": 5.2343,
      "mean_token_accuracy": 0.19512814581394194,
      "num_tokens": 47034363.0,
      "step": 20535
    },
    {
      "entropy": 5.517018508911133,
      "epoch": 2.030446816923685,
      "grad_norm": 1.25,
      "learning_rate": 0.0004590441085572962,
      "loss": 5.1967,
      "mean_token_accuracy": 0.19620057344436645,
      "num_tokens": 47045584.0,
      "step": 20540
    },
    {
      "entropy": 5.522439098358154,
      "epoch": 2.030941083432187,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045902380553880913,
      "loss": 5.1675,
      "mean_token_accuracy": 0.19280802607536315,
      "num_tokens": 47056245.0,
      "step": 20545
    },
    {
      "entropy": 5.5482017517089846,
      "epoch": 2.031435349940688,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004590034979933078,
      "loss": 5.1058,
      "mean_token_accuracy": 0.19932733327150345,
      "num_tokens": 47067856.0,
      "step": 20550
    },
    {
      "entropy": 5.501329565048218,
      "epoch": 2.0319296164491893,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045898318592129197,
      "loss": 5.1196,
      "mean_token_accuracy": 0.20489915013313292,
      "num_tokens": 47079316.0,
      "step": 20555
    },
    {
      "entropy": 5.523725509643555,
      "epoch": 2.0324238829576906,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045896286932326115,
      "loss": 5.1895,
      "mean_token_accuracy": 0.18756190985441207,
      "num_tokens": 47090268.0,
      "step": 20560
    },
    {
      "entropy": 5.562956619262695,
      "epoch": 2.0329181494661923,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004589425481997152,
      "loss": 5.1572,
      "mean_token_accuracy": 0.20081699192523955,
      "num_tokens": 47100980.0,
      "step": 20565
    },
    {
      "entropy": 5.557186031341553,
      "epoch": 2.0334124159746936,
      "grad_norm": 1.15625,
      "learning_rate": 0.000458922222551154,
      "loss": 5.1091,
      "mean_token_accuracy": 0.19957223981618882,
      "num_tokens": 47112548.0,
      "step": 20570
    },
    {
      "entropy": 5.478365278244018,
      "epoch": 2.033906682483195,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004589018923780777,
      "loss": 5.2108,
      "mean_token_accuracy": 0.1920792877674103,
      "num_tokens": 47123378.0,
      "step": 20575
    },
    {
      "entropy": 5.525159454345703,
      "epoch": 2.0344009489916965,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004588815576809862,
      "loss": 5.2262,
      "mean_token_accuracy": 0.1833912506699562,
      "num_tokens": 47135279.0,
      "step": 20580
    },
    {
      "entropy": 5.644260263442993,
      "epoch": 2.034895215500198,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045886121846037996,
      "loss": 5.2084,
      "mean_token_accuracy": 0.1981610283255577,
      "num_tokens": 47146510.0,
      "step": 20585
    },
    {
      "entropy": 5.547625494003296,
      "epoch": 2.035389482008699,
      "grad_norm": 1.34375,
      "learning_rate": 0.00045884087471675917,
      "loss": 5.1133,
      "mean_token_accuracy": 0.20044450461864471,
      "num_tokens": 47158275.0,
      "step": 20590
    },
    {
      "entropy": 5.53369140625,
      "epoch": 2.0358837485172003,
      "grad_norm": 1.359375,
      "learning_rate": 0.00045882052645062433,
      "loss": 5.2101,
      "mean_token_accuracy": 0.19684501737356186,
      "num_tokens": 47170530.0,
      "step": 20595
    },
    {
      "entropy": 5.548107528686524,
      "epoch": 2.036378015025702,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000458800173662476,
      "loss": 5.1824,
      "mean_token_accuracy": 0.19195338934659958,
      "num_tokens": 47182092.0,
      "step": 20600
    },
    {
      "entropy": 5.622240734100342,
      "epoch": 2.0368722815342033,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045877981635281497,
      "loss": 5.2062,
      "mean_token_accuracy": 0.19466145038604737,
      "num_tokens": 47193184.0,
      "step": 20605
    },
    {
      "entropy": 5.564381074905396,
      "epoch": 2.0373665480427046,
      "grad_norm": 1.328125,
      "learning_rate": 0.00045875945452214186,
      "loss": 5.1809,
      "mean_token_accuracy": 0.19054548889398576,
      "num_tokens": 47205032.0,
      "step": 20610
    },
    {
      "entropy": 5.592906761169433,
      "epoch": 2.037860814551206,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00045873908817095773,
      "loss": 5.2117,
      "mean_token_accuracy": 0.19142548888921737,
      "num_tokens": 47217553.0,
      "step": 20615
    },
    {
      "entropy": 5.5924633026123045,
      "epoch": 2.0383550810597075,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004587187172997635,
      "loss": 5.2373,
      "mean_token_accuracy": 0.19821246713399887,
      "num_tokens": 47230625.0,
      "step": 20620
    },
    {
      "entropy": 5.548401880264282,
      "epoch": 2.038849347568209,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004586983419090603,
      "loss": 5.1422,
      "mean_token_accuracy": 0.19793572276830673,
      "num_tokens": 47240703.0,
      "step": 20625
    },
    {
      "entropy": 5.488960695266724,
      "epoch": 2.03934361407671,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045867796199934943,
      "loss": 5.1519,
      "mean_token_accuracy": 0.20213185101747513,
      "num_tokens": 47253768.0,
      "step": 20630
    },
    {
      "entropy": 5.562997913360595,
      "epoch": 2.0398378805852118,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004586575775711322,
      "loss": 5.121,
      "mean_token_accuracy": 0.1955211952328682,
      "num_tokens": 47264226.0,
      "step": 20635
    },
    {
      "entropy": 5.52231068611145,
      "epoch": 2.040332147093713,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00045863718862491,
      "loss": 5.1961,
      "mean_token_accuracy": 0.19778646826744078,
      "num_tokens": 47276675.0,
      "step": 20640
    },
    {
      "entropy": 5.5680882930755615,
      "epoch": 2.0408264136022143,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045861679516118447,
      "loss": 5.2459,
      "mean_token_accuracy": 0.196078522503376,
      "num_tokens": 47288876.0,
      "step": 20645
    },
    {
      "entropy": 5.520850419998169,
      "epoch": 2.0413206801107155,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004585963971804574,
      "loss": 5.1428,
      "mean_token_accuracy": 0.19315340667963027,
      "num_tokens": 47300430.0,
      "step": 20650
    },
    {
      "entropy": 5.547067546844483,
      "epoch": 2.0418149466192173,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004585759946832304,
      "loss": 5.2118,
      "mean_token_accuracy": 0.1966462329030037,
      "num_tokens": 47310822.0,
      "step": 20655
    },
    {
      "entropy": 5.550311708450318,
      "epoch": 2.0423092131277185,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045855558767000546,
      "loss": 5.2195,
      "mean_token_accuracy": 0.19339940696954727,
      "num_tokens": 47323042.0,
      "step": 20660
    },
    {
      "entropy": 5.514761114120484,
      "epoch": 2.0428034796362198,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004585351761412846,
      "loss": 5.1702,
      "mean_token_accuracy": 0.19792895317077636,
      "num_tokens": 47334003.0,
      "step": 20665
    },
    {
      "entropy": 5.52094874382019,
      "epoch": 2.043297746144721,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004585147600975699,
      "loss": 5.1514,
      "mean_token_accuracy": 0.19315087348222731,
      "num_tokens": 47346279.0,
      "step": 20670
    },
    {
      "entropy": 5.529201078414917,
      "epoch": 2.0437920126532227,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004584943395393637,
      "loss": 5.1287,
      "mean_token_accuracy": 0.194491945207119,
      "num_tokens": 47357655.0,
      "step": 20675
    },
    {
      "entropy": 5.515044546127319,
      "epoch": 2.044286279161724,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045847391446716826,
      "loss": 5.1513,
      "mean_token_accuracy": 0.1985543340444565,
      "num_tokens": 47369494.0,
      "step": 20680
    },
    {
      "entropy": 5.522799348831176,
      "epoch": 2.0447805456702253,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000458453484881486,
      "loss": 5.1918,
      "mean_token_accuracy": 0.1998739868402481,
      "num_tokens": 47380221.0,
      "step": 20685
    },
    {
      "entropy": 5.486823892593383,
      "epoch": 2.045274812178727,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004584330507828196,
      "loss": 5.0958,
      "mean_token_accuracy": 0.20503304153680801,
      "num_tokens": 47391119.0,
      "step": 20690
    },
    {
      "entropy": 5.584002876281739,
      "epoch": 2.0457690786872282,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004584126121716717,
      "loss": 5.1464,
      "mean_token_accuracy": 0.19933037757873534,
      "num_tokens": 47401828.0,
      "step": 20695
    },
    {
      "entropy": 5.4566038131713865,
      "epoch": 2.0462633451957295,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004583921690485451,
      "loss": 5.1275,
      "mean_token_accuracy": 0.20225608050823213,
      "num_tokens": 47412921.0,
      "step": 20700
    },
    {
      "entropy": 5.541837072372436,
      "epoch": 2.0467576117042308,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004583717214139427,
      "loss": 5.1991,
      "mean_token_accuracy": 0.19362544715404512,
      "num_tokens": 47424117.0,
      "step": 20705
    },
    {
      "entropy": 5.525685119628906,
      "epoch": 2.0472518782127325,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045835126926836744,
      "loss": 5.1639,
      "mean_token_accuracy": 0.19853021949529648,
      "num_tokens": 47434457.0,
      "step": 20710
    },
    {
      "entropy": 5.573009252548218,
      "epoch": 2.0477461447212337,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004583308126123225,
      "loss": 5.2282,
      "mean_token_accuracy": 0.19634588658809662,
      "num_tokens": 47446108.0,
      "step": 20715
    },
    {
      "entropy": 5.535366010665894,
      "epoch": 2.048240411229735,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00045831035144631115,
      "loss": 5.0795,
      "mean_token_accuracy": 0.20150321274995803,
      "num_tokens": 47458735.0,
      "step": 20720
    },
    {
      "entropy": 5.6063214302062985,
      "epoch": 2.0487346777382363,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045828988577083674,
      "loss": 5.1917,
      "mean_token_accuracy": 0.1986448734998703,
      "num_tokens": 47471726.0,
      "step": 20725
    },
    {
      "entropy": 5.6269289493560795,
      "epoch": 2.049228944246738,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004582694155864027,
      "loss": 5.2066,
      "mean_token_accuracy": 0.18992726504802704,
      "num_tokens": 47483940.0,
      "step": 20730
    },
    {
      "entropy": 5.440764141082764,
      "epoch": 2.0497232107552392,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045824894089351265,
      "loss": 5.1249,
      "mean_token_accuracy": 0.19728759974241256,
      "num_tokens": 47494723.0,
      "step": 20735
    },
    {
      "entropy": 5.498567533493042,
      "epoch": 2.0502174772637405,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004582284616926702,
      "loss": 5.2001,
      "mean_token_accuracy": 0.19118748903274535,
      "num_tokens": 47506454.0,
      "step": 20740
    },
    {
      "entropy": 5.593333148956299,
      "epoch": 2.050711743772242,
      "grad_norm": 1.25,
      "learning_rate": 0.0004582079779843791,
      "loss": 5.2849,
      "mean_token_accuracy": 0.19144766479730607,
      "num_tokens": 47518366.0,
      "step": 20745
    },
    {
      "entropy": 5.528135585784912,
      "epoch": 2.0512060102807435,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004581874897691434,
      "loss": 5.1853,
      "mean_token_accuracy": 0.19747180342674256,
      "num_tokens": 47530635.0,
      "step": 20750
    },
    {
      "entropy": 5.552562522888183,
      "epoch": 2.0517002767892447,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000458166997047467,
      "loss": 5.2601,
      "mean_token_accuracy": 0.1970040410757065,
      "num_tokens": 47541158.0,
      "step": 20755
    },
    {
      "entropy": 5.568430948257446,
      "epoch": 2.052194543297746,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045814649981985406,
      "loss": 5.1902,
      "mean_token_accuracy": 0.19020578265190125,
      "num_tokens": 47552392.0,
      "step": 20760
    },
    {
      "entropy": 5.6111804962158205,
      "epoch": 2.0526888098062477,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004581259980868088,
      "loss": 5.1998,
      "mean_token_accuracy": 0.1971875935792923,
      "num_tokens": 47565079.0,
      "step": 20765
    },
    {
      "entropy": 5.603356218338012,
      "epoch": 2.053183076314749,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004581054918488356,
      "loss": 5.1926,
      "mean_token_accuracy": 0.19807939678430558,
      "num_tokens": 47575529.0,
      "step": 20770
    },
    {
      "entropy": 5.448172855377197,
      "epoch": 2.05367734282325,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045808498110643886,
      "loss": 5.1086,
      "mean_token_accuracy": 0.19796554148197174,
      "num_tokens": 47587289.0,
      "step": 20775
    },
    {
      "entropy": 5.528416347503662,
      "epoch": 2.0541716093317515,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045806446586012326,
      "loss": 5.1598,
      "mean_token_accuracy": 0.19505753070116044,
      "num_tokens": 47599184.0,
      "step": 20780
    },
    {
      "entropy": 5.486471366882324,
      "epoch": 2.054665875840253,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004580439461103933,
      "loss": 5.0424,
      "mean_token_accuracy": 0.21222029477357865,
      "num_tokens": 47609651.0,
      "step": 20785
    },
    {
      "entropy": 5.578895950317383,
      "epoch": 2.0551601423487544,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004580234218577539,
      "loss": 5.1866,
      "mean_token_accuracy": 0.19902318716049194,
      "num_tokens": 47621138.0,
      "step": 20790
    },
    {
      "entropy": 5.498895883560181,
      "epoch": 2.0556544088572557,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00045800289310270995,
      "loss": 5.1802,
      "mean_token_accuracy": 0.20277549773454667,
      "num_tokens": 47631492.0,
      "step": 20795
    },
    {
      "entropy": 5.482671165466309,
      "epoch": 2.0561486753657574,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004579823598457664,
      "loss": 5.1431,
      "mean_token_accuracy": 0.20404961258172988,
      "num_tokens": 47643088.0,
      "step": 20800
    },
    {
      "entropy": 5.46958212852478,
      "epoch": 2.0566429418742587,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045796182208742846,
      "loss": 5.0631,
      "mean_token_accuracy": 0.21121376156806945,
      "num_tokens": 47653393.0,
      "step": 20805
    },
    {
      "entropy": 5.496805810928345,
      "epoch": 2.05713720838276,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004579412798282013,
      "loss": 5.1553,
      "mean_token_accuracy": 0.20107391625642776,
      "num_tokens": 47665879.0,
      "step": 20810
    },
    {
      "entropy": 5.4519476890563965,
      "epoch": 2.057631474891261,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004579207330685902,
      "loss": 5.1289,
      "mean_token_accuracy": 0.20296097844839095,
      "num_tokens": 47678339.0,
      "step": 20815
    },
    {
      "entropy": 5.5129491806030275,
      "epoch": 2.058125741399763,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004579001818091009,
      "loss": 5.1721,
      "mean_token_accuracy": 0.20384158790111542,
      "num_tokens": 47688867.0,
      "step": 20820
    },
    {
      "entropy": 5.554760456085205,
      "epoch": 2.058620007908264,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004578796260502385,
      "loss": 5.2394,
      "mean_token_accuracy": 0.19736232608556747,
      "num_tokens": 47700265.0,
      "step": 20825
    },
    {
      "entropy": 5.603539848327637,
      "epoch": 2.0591142744167654,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004578590657925091,
      "loss": 5.2519,
      "mean_token_accuracy": 0.189376100897789,
      "num_tokens": 47711539.0,
      "step": 20830
    },
    {
      "entropy": 5.581753444671631,
      "epoch": 2.059608540925267,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004578385010364182,
      "loss": 5.1808,
      "mean_token_accuracy": 0.19475315660238265,
      "num_tokens": 47722433.0,
      "step": 20835
    },
    {
      "entropy": 5.476624965667725,
      "epoch": 2.0601028074337684,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045781793178247194,
      "loss": 5.1377,
      "mean_token_accuracy": 0.20019441992044448,
      "num_tokens": 47734001.0,
      "step": 20840
    },
    {
      "entropy": 5.4575958251953125,
      "epoch": 2.0605970739422697,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004577973580311761,
      "loss": 5.1112,
      "mean_token_accuracy": 0.20995286405086516,
      "num_tokens": 47744903.0,
      "step": 20845
    },
    {
      "entropy": 5.600017356872558,
      "epoch": 2.061091340450771,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004577767797830369,
      "loss": 5.1671,
      "mean_token_accuracy": 0.20072797536849976,
      "num_tokens": 47756571.0,
      "step": 20850
    },
    {
      "entropy": 5.510620641708374,
      "epoch": 2.0615856069592726,
      "grad_norm": 1.40625,
      "learning_rate": 0.00045775619703856057,
      "loss": 5.1228,
      "mean_token_accuracy": 0.19831173568964006,
      "num_tokens": 47767555.0,
      "step": 20855
    },
    {
      "entropy": 5.638478469848633,
      "epoch": 2.062079873467774,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045773560979825347,
      "loss": 5.2054,
      "mean_token_accuracy": 0.20082833915948867,
      "num_tokens": 47779577.0,
      "step": 20860
    },
    {
      "entropy": 5.547027683258056,
      "epoch": 2.062574139976275,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045771501806262196,
      "loss": 5.2093,
      "mean_token_accuracy": 0.19598483443260192,
      "num_tokens": 47790898.0,
      "step": 20865
    },
    {
      "entropy": 5.53073558807373,
      "epoch": 2.0630684064847764,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045769442183217273,
      "loss": 5.1745,
      "mean_token_accuracy": 0.20367027819156647,
      "num_tokens": 47802008.0,
      "step": 20870
    },
    {
      "entropy": 5.51125922203064,
      "epoch": 2.063562672993278,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004576738211074123,
      "loss": 5.136,
      "mean_token_accuracy": 0.20050776302814483,
      "num_tokens": 47813544.0,
      "step": 20875
    },
    {
      "entropy": 5.562144756317139,
      "epoch": 2.0640569395017794,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004576532158888476,
      "loss": 5.1843,
      "mean_token_accuracy": 0.19587933123111725,
      "num_tokens": 47824313.0,
      "step": 20880
    },
    {
      "entropy": 5.615613603591919,
      "epoch": 2.0645512060102806,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004576326061769854,
      "loss": 5.1574,
      "mean_token_accuracy": 0.19965433776378633,
      "num_tokens": 47836653.0,
      "step": 20885
    },
    {
      "entropy": 5.561070680618286,
      "epoch": 2.065045472518782,
      "grad_norm": 1.328125,
      "learning_rate": 0.00045761199197233285,
      "loss": 5.2082,
      "mean_token_accuracy": 0.1971917361021042,
      "num_tokens": 47848237.0,
      "step": 20890
    },
    {
      "entropy": 5.4977771759033205,
      "epoch": 2.0655397390272836,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045759137327539684,
      "loss": 5.2041,
      "mean_token_accuracy": 0.19317791163921355,
      "num_tokens": 47859686.0,
      "step": 20895
    },
    {
      "entropy": 5.480271291732788,
      "epoch": 2.066034005535785,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004575707500866848,
      "loss": 5.1205,
      "mean_token_accuracy": 0.19809692353010178,
      "num_tokens": 47871730.0,
      "step": 20900
    },
    {
      "entropy": 5.616372728347779,
      "epoch": 2.066528272044286,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045755012240670395,
      "loss": 5.2169,
      "mean_token_accuracy": 0.1914566159248352,
      "num_tokens": 47883927.0,
      "step": 20905
    },
    {
      "entropy": 5.5164111137390135,
      "epoch": 2.067022538552788,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004575294902359618,
      "loss": 5.0331,
      "mean_token_accuracy": 0.20936899483203888,
      "num_tokens": 47894885.0,
      "step": 20910
    },
    {
      "entropy": 5.583211851119995,
      "epoch": 2.067516805061289,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00045750885357496583,
      "loss": 5.213,
      "mean_token_accuracy": 0.1913083329796791,
      "num_tokens": 47906183.0,
      "step": 20915
    },
    {
      "entropy": 5.467278957366943,
      "epoch": 2.0680110715697904,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004574882124242238,
      "loss": 5.1751,
      "mean_token_accuracy": 0.19382926523685456,
      "num_tokens": 47918002.0,
      "step": 20920
    },
    {
      "entropy": 5.4840000629425045,
      "epoch": 2.0685053380782916,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045746756678424343,
      "loss": 5.0848,
      "mean_token_accuracy": 0.20050887763500214,
      "num_tokens": 47928300.0,
      "step": 20925
    },
    {
      "entropy": 5.485995578765869,
      "epoch": 2.0689996045867933,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045744691665553254,
      "loss": 5.1043,
      "mean_token_accuracy": 0.20007977783679962,
      "num_tokens": 47939591.0,
      "step": 20930
    },
    {
      "entropy": 5.615813398361206,
      "epoch": 2.0694938710952946,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004574262620385993,
      "loss": 5.2231,
      "mean_token_accuracy": 0.19402552247047425,
      "num_tokens": 47953075.0,
      "step": 20935
    },
    {
      "entropy": 5.634069156646729,
      "epoch": 2.069988137603796,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045740560293395166,
      "loss": 5.262,
      "mean_token_accuracy": 0.18457178324460982,
      "num_tokens": 47964795.0,
      "step": 20940
    },
    {
      "entropy": 5.449629259109497,
      "epoch": 2.0704824041122976,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004573849393420979,
      "loss": 5.0673,
      "mean_token_accuracy": 0.21239512115716935,
      "num_tokens": 47975175.0,
      "step": 20945
    },
    {
      "entropy": 5.493186664581299,
      "epoch": 2.070976670620799,
      "grad_norm": 1.25,
      "learning_rate": 0.0004573642712635463,
      "loss": 5.2022,
      "mean_token_accuracy": 0.1958724305033684,
      "num_tokens": 47986548.0,
      "step": 20950
    },
    {
      "entropy": 5.572441673278808,
      "epoch": 2.0714709371293,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00045734359869880535,
      "loss": 5.2584,
      "mean_token_accuracy": 0.18303977847099304,
      "num_tokens": 47998759.0,
      "step": 20955
    },
    {
      "entropy": 5.527265453338623,
      "epoch": 2.0719652036378013,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004573229216483836,
      "loss": 5.1349,
      "mean_token_accuracy": 0.2040463611483574,
      "num_tokens": 48010404.0,
      "step": 20960
    },
    {
      "entropy": 5.460834789276123,
      "epoch": 2.072459470146303,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004573022401127897,
      "loss": 5.0665,
      "mean_token_accuracy": 0.20826467424631118,
      "num_tokens": 48019729.0,
      "step": 20965
    },
    {
      "entropy": 5.544267368316651,
      "epoch": 2.0729537366548043,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045728155409253236,
      "loss": 5.1807,
      "mean_token_accuracy": 0.19063322246074677,
      "num_tokens": 48031142.0,
      "step": 20970
    },
    {
      "entropy": 5.614737224578858,
      "epoch": 2.0734480031633056,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004572608635881206,
      "loss": 5.1588,
      "mean_token_accuracy": 0.20325768738985062,
      "num_tokens": 48043051.0,
      "step": 20975
    },
    {
      "entropy": 5.471118307113647,
      "epoch": 2.073942269671807,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004572401686000633,
      "loss": 5.125,
      "mean_token_accuracy": 0.20493268370628356,
      "num_tokens": 48053721.0,
      "step": 20980
    },
    {
      "entropy": 5.36272497177124,
      "epoch": 2.0744365361803085,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045721946912886957,
      "loss": 5.075,
      "mean_token_accuracy": 0.2090740829706192,
      "num_tokens": 48065208.0,
      "step": 20985
    },
    {
      "entropy": 5.587690639495849,
      "epoch": 2.07493080268881,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004571987651750486,
      "loss": 5.2475,
      "mean_token_accuracy": 0.19670464247465133,
      "num_tokens": 48076416.0,
      "step": 20990
    },
    {
      "entropy": 5.577077770233155,
      "epoch": 2.075425069197311,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004571780567391098,
      "loss": 5.1777,
      "mean_token_accuracy": 0.1944845661520958,
      "num_tokens": 48087496.0,
      "step": 20995
    },
    {
      "entropy": 5.57269434928894,
      "epoch": 2.0759193357058128,
      "grad_norm": 1.640625,
      "learning_rate": 0.00045715734382156245,
      "loss": 5.1941,
      "mean_token_accuracy": 0.19725523442029952,
      "num_tokens": 48098920.0,
      "step": 21000
    },
    {
      "epoch": 2.0759193357058128,
      "eval_entropy": 5.334154620117807,
      "eval_loss": 5.277682304382324,
      "eval_mean_token_accuracy": 0.19912817763922985,
      "eval_num_tokens": 48098920.0,
      "eval_runtime": 26.4824,
      "eval_samples_per_second": 1227.72,
      "eval_steps_per_second": 153.498,
      "step": 21000
    },
    {
      "entropy": 5.528890895843506,
      "epoch": 2.076413602214314,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004571366264229162,
      "loss": 5.2237,
      "mean_token_accuracy": 0.19388963133096696,
      "num_tokens": 48109813.0,
      "step": 21005
    },
    {
      "entropy": 5.542822360992432,
      "epoch": 2.0769078687228153,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004571159045436807,
      "loss": 5.1668,
      "mean_token_accuracy": 0.20029381662607193,
      "num_tokens": 48120736.0,
      "step": 21010
    },
    {
      "entropy": 5.488102293014526,
      "epoch": 2.0774021352313166,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004570951781843657,
      "loss": 5.1165,
      "mean_token_accuracy": 0.20595585107803344,
      "num_tokens": 48131631.0,
      "step": 21015
    },
    {
      "entropy": 5.487272691726685,
      "epoch": 2.0778964017398183,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045707444734548116,
      "loss": 5.0808,
      "mean_token_accuracy": 0.2042886048555374,
      "num_tokens": 48141662.0,
      "step": 21020
    },
    {
      "entropy": 5.462234830856323,
      "epoch": 2.0783906682483195,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004570537120275369,
      "loss": 5.1137,
      "mean_token_accuracy": 0.20591005831956863,
      "num_tokens": 48152752.0,
      "step": 21025
    },
    {
      "entropy": 5.545647811889649,
      "epoch": 2.078884934756821,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000457032972231043,
      "loss": 5.1766,
      "mean_token_accuracy": 0.1942004829645157,
      "num_tokens": 48163510.0,
      "step": 21030
    },
    {
      "entropy": 5.47556881904602,
      "epoch": 2.079379201265322,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004570122279565098,
      "loss": 5.2163,
      "mean_token_accuracy": 0.19330684691667557,
      "num_tokens": 48174241.0,
      "step": 21035
    },
    {
      "entropy": 5.593160629272461,
      "epoch": 2.0798734677738238,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045699147920444757,
      "loss": 5.2045,
      "mean_token_accuracy": 0.1897025316953659,
      "num_tokens": 48186062.0,
      "step": 21040
    },
    {
      "entropy": 5.538670587539673,
      "epoch": 2.080367734282325,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004569707259753667,
      "loss": 5.0412,
      "mean_token_accuracy": 0.20270233154296874,
      "num_tokens": 48197108.0,
      "step": 21045
    },
    {
      "entropy": 5.416890430450439,
      "epoch": 2.0808620007908263,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004569499682697778,
      "loss": 5.0998,
      "mean_token_accuracy": 0.21021183729171752,
      "num_tokens": 48208780.0,
      "step": 21050
    },
    {
      "entropy": 5.519730567932129,
      "epoch": 2.081356267299328,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004569292060881914,
      "loss": 5.1655,
      "mean_token_accuracy": 0.1987949550151825,
      "num_tokens": 48223164.0,
      "step": 21055
    },
    {
      "entropy": 5.5092614650726315,
      "epoch": 2.0818505338078293,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045690843943111824,
      "loss": 5.0778,
      "mean_token_accuracy": 0.20350499004125594,
      "num_tokens": 48234402.0,
      "step": 21060
    },
    {
      "entropy": 5.522765254974365,
      "epoch": 2.0823448003163305,
      "grad_norm": 1.484375,
      "learning_rate": 0.00045688766829906936,
      "loss": 5.144,
      "mean_token_accuracy": 0.19374675452709197,
      "num_tokens": 48245147.0,
      "step": 21065
    },
    {
      "entropy": 5.593550968170166,
      "epoch": 2.0828390668248318,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004568668926925556,
      "loss": 5.2298,
      "mean_token_accuracy": 0.1885303184390068,
      "num_tokens": 48256787.0,
      "step": 21070
    },
    {
      "entropy": 5.507479572296143,
      "epoch": 2.0833333333333335,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000456846112612088,
      "loss": 5.1447,
      "mean_token_accuracy": 0.19519422352313995,
      "num_tokens": 48267648.0,
      "step": 21075
    },
    {
      "entropy": 5.459123134613037,
      "epoch": 2.0838275998418347,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045682532805817796,
      "loss": 5.1561,
      "mean_token_accuracy": 0.19162509292364122,
      "num_tokens": 48279274.0,
      "step": 21080
    },
    {
      "entropy": 5.553543186187744,
      "epoch": 2.084321866350336,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004568045390313365,
      "loss": 5.1676,
      "mean_token_accuracy": 0.1966897279024124,
      "num_tokens": 48290849.0,
      "step": 21085
    },
    {
      "entropy": 5.527925682067871,
      "epoch": 2.0848161328588377,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045678374553207525,
      "loss": 5.2091,
      "mean_token_accuracy": 0.1925302505493164,
      "num_tokens": 48301271.0,
      "step": 21090
    },
    {
      "entropy": 5.561384582519532,
      "epoch": 2.085310399367339,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045676294756090565,
      "loss": 5.1561,
      "mean_token_accuracy": 0.19214406460523606,
      "num_tokens": 48311869.0,
      "step": 21095
    },
    {
      "entropy": 5.5914044857025145,
      "epoch": 2.0858046658758402,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004567421451183393,
      "loss": 5.1962,
      "mean_token_accuracy": 0.20460759103298187,
      "num_tokens": 48324907.0,
      "step": 21100
    },
    {
      "entropy": 5.540077161788941,
      "epoch": 2.0862989323843415,
      "grad_norm": 1.265625,
      "learning_rate": 0.000456721338204888,
      "loss": 5.1115,
      "mean_token_accuracy": 0.2043887421488762,
      "num_tokens": 48335887.0,
      "step": 21105
    },
    {
      "entropy": 5.555528974533081,
      "epoch": 2.086793198892843,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004567005268210636,
      "loss": 5.1219,
      "mean_token_accuracy": 0.19903927594423293,
      "num_tokens": 48348190.0,
      "step": 21110
    },
    {
      "entropy": 5.5656030654907225,
      "epoch": 2.0872874654013445,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00045667971096737805,
      "loss": 5.2903,
      "mean_token_accuracy": 0.18635109663009644,
      "num_tokens": 48360113.0,
      "step": 21115
    },
    {
      "entropy": 5.5635510921478275,
      "epoch": 2.0877817319098457,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004566588906443434,
      "loss": 5.2029,
      "mean_token_accuracy": 0.19759729951620103,
      "num_tokens": 48371463.0,
      "step": 21120
    },
    {
      "entropy": 5.600057697296142,
      "epoch": 2.088275998418347,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004566380658524719,
      "loss": 5.273,
      "mean_token_accuracy": 0.18369587808847426,
      "num_tokens": 48383969.0,
      "step": 21125
    },
    {
      "entropy": 5.518441915512085,
      "epoch": 2.0887702649268487,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004566172365922758,
      "loss": 5.1087,
      "mean_token_accuracy": 0.20205163061618805,
      "num_tokens": 48395551.0,
      "step": 21130
    },
    {
      "entropy": 5.518635702133179,
      "epoch": 2.08926453143535,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004565964028642675,
      "loss": 5.1333,
      "mean_token_accuracy": 0.20412324368953705,
      "num_tokens": 48408154.0,
      "step": 21135
    },
    {
      "entropy": 5.405503940582276,
      "epoch": 2.089758797943851,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004565755646689595,
      "loss": 5.1097,
      "mean_token_accuracy": 0.2058487504720688,
      "num_tokens": 48420060.0,
      "step": 21140
    },
    {
      "entropy": 5.508947467803955,
      "epoch": 2.0902530644523525,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004565547220068644,
      "loss": 5.2114,
      "mean_token_accuracy": 0.19615855813026428,
      "num_tokens": 48432347.0,
      "step": 21145
    },
    {
      "entropy": 5.661038637161255,
      "epoch": 2.090747330960854,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004565338748784951,
      "loss": 5.2518,
      "mean_token_accuracy": 0.18711526542901993,
      "num_tokens": 48443177.0,
      "step": 21150
    },
    {
      "entropy": 5.598845624923706,
      "epoch": 2.0912415974693555,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004565130232843642,
      "loss": 5.1559,
      "mean_token_accuracy": 0.19312859177589417,
      "num_tokens": 48453605.0,
      "step": 21155
    },
    {
      "entropy": 5.565904569625855,
      "epoch": 2.0917358639778567,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004564921672249848,
      "loss": 5.1892,
      "mean_token_accuracy": 0.1960737019777298,
      "num_tokens": 48465364.0,
      "step": 21160
    },
    {
      "entropy": 5.4654029369354244,
      "epoch": 2.0922301304863584,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045647130670086996,
      "loss": 5.1312,
      "mean_token_accuracy": 0.19819717556238176,
      "num_tokens": 48477255.0,
      "step": 21165
    },
    {
      "entropy": 5.50226058959961,
      "epoch": 2.0927243969948597,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004564504417125328,
      "loss": 5.1623,
      "mean_token_accuracy": 0.2010049507021904,
      "num_tokens": 48487551.0,
      "step": 21170
    },
    {
      "entropy": 5.627320575714111,
      "epoch": 2.093218663503361,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045642957226048667,
      "loss": 5.1579,
      "mean_token_accuracy": 0.19381139427423477,
      "num_tokens": 48500422.0,
      "step": 21175
    },
    {
      "entropy": 5.462879276275634,
      "epoch": 2.093712930011862,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045640869834524486,
      "loss": 5.1188,
      "mean_token_accuracy": 0.2099028706550598,
      "num_tokens": 48511588.0,
      "step": 21180
    },
    {
      "entropy": 5.531794500350952,
      "epoch": 2.094207196520364,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004563878199673209,
      "loss": 5.2162,
      "mean_token_accuracy": 0.19897359758615493,
      "num_tokens": 48522700.0,
      "step": 21185
    },
    {
      "entropy": 5.625210285186768,
      "epoch": 2.094701463028865,
      "grad_norm": 1.25,
      "learning_rate": 0.0004563669371272285,
      "loss": 5.2277,
      "mean_token_accuracy": 0.19862253963947296,
      "num_tokens": 48534253.0,
      "step": 21190
    },
    {
      "entropy": 5.48061785697937,
      "epoch": 2.0951957295373664,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045634604982548125,
      "loss": 5.1177,
      "mean_token_accuracy": 0.20375279784202577,
      "num_tokens": 48545543.0,
      "step": 21195
    },
    {
      "entropy": 5.540404462814331,
      "epoch": 2.095689996045868,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045632515806259306,
      "loss": 5.2148,
      "mean_token_accuracy": 0.19447827488183975,
      "num_tokens": 48557433.0,
      "step": 21200
    },
    {
      "entropy": 5.576367807388306,
      "epoch": 2.0961842625543694,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045630426183907786,
      "loss": 5.1964,
      "mean_token_accuracy": 0.19374802708625793,
      "num_tokens": 48570271.0,
      "step": 21205
    },
    {
      "entropy": 5.52334189414978,
      "epoch": 2.0966785290628707,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004562833611554497,
      "loss": 5.1462,
      "mean_token_accuracy": 0.19547879993915557,
      "num_tokens": 48582376.0,
      "step": 21210
    },
    {
      "entropy": 5.539679431915284,
      "epoch": 2.097172795571372,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004562624560122227,
      "loss": 5.1931,
      "mean_token_accuracy": 0.1963004246354103,
      "num_tokens": 48595188.0,
      "step": 21215
    },
    {
      "entropy": 5.515546417236328,
      "epoch": 2.0976670620798736,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004562415464099112,
      "loss": 5.0824,
      "mean_token_accuracy": 0.19943149089813234,
      "num_tokens": 48606093.0,
      "step": 21220
    },
    {
      "entropy": 5.574150323867798,
      "epoch": 2.098161328588375,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004562206323490295,
      "loss": 5.2353,
      "mean_token_accuracy": 0.19594648480415344,
      "num_tokens": 48618163.0,
      "step": 21225
    },
    {
      "entropy": 5.5406811237335205,
      "epoch": 2.098655595096876,
      "grad_norm": 1.3125,
      "learning_rate": 0.00045619971383009216,
      "loss": 5.2012,
      "mean_token_accuracy": 0.1962972328066826,
      "num_tokens": 48629771.0,
      "step": 21230
    },
    {
      "entropy": 5.521897411346435,
      "epoch": 2.0991498616053774,
      "grad_norm": 1.5,
      "learning_rate": 0.0004561787908536138,
      "loss": 5.1483,
      "mean_token_accuracy": 0.19210223257541656,
      "num_tokens": 48640409.0,
      "step": 21235
    },
    {
      "entropy": 5.489107227325439,
      "epoch": 2.099644128113879,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045615786342010896,
      "loss": 5.1487,
      "mean_token_accuracy": 0.201000614464283,
      "num_tokens": 48652223.0,
      "step": 21240
    },
    {
      "entropy": 5.554998111724854,
      "epoch": 2.1001383946223804,
      "grad_norm": 1.3125,
      "learning_rate": 0.00045613693153009264,
      "loss": 5.0645,
      "mean_token_accuracy": 0.20431340783834456,
      "num_tokens": 48663667.0,
      "step": 21245
    },
    {
      "entropy": 5.550862169265747,
      "epoch": 2.1006326611308817,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004561159951840797,
      "loss": 5.2162,
      "mean_token_accuracy": 0.19332771003246307,
      "num_tokens": 48674691.0,
      "step": 21250
    },
    {
      "entropy": 5.48462553024292,
      "epoch": 2.1011269276393834,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004560950543825852,
      "loss": 5.1772,
      "mean_token_accuracy": 0.19521435499191284,
      "num_tokens": 48686888.0,
      "step": 21255
    },
    {
      "entropy": 5.511814212799072,
      "epoch": 2.1016211941478846,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045607410912612426,
      "loss": 5.1689,
      "mean_token_accuracy": 0.19298310875892638,
      "num_tokens": 48698666.0,
      "step": 21260
    },
    {
      "entropy": 5.543023920059204,
      "epoch": 2.102115460656386,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004560531594152121,
      "loss": 5.1103,
      "mean_token_accuracy": 0.20415060371160507,
      "num_tokens": 48709793.0,
      "step": 21265
    },
    {
      "entropy": 5.403988075256348,
      "epoch": 2.102609727164887,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004560322052503642,
      "loss": 5.0596,
      "mean_token_accuracy": 0.20458496809005738,
      "num_tokens": 48720835.0,
      "step": 21270
    },
    {
      "entropy": 5.615757846832276,
      "epoch": 2.103103993673389,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004560112466320959,
      "loss": 5.3145,
      "mean_token_accuracy": 0.18941870033740998,
      "num_tokens": 48732990.0,
      "step": 21275
    },
    {
      "entropy": 5.553078365325928,
      "epoch": 2.10359826018189,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045599028356092293,
      "loss": 5.1296,
      "mean_token_accuracy": 0.1998555451631546,
      "num_tokens": 48744985.0,
      "step": 21280
    },
    {
      "entropy": 5.417953014373779,
      "epoch": 2.1040925266903914,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004559693160373608,
      "loss": 5.0637,
      "mean_token_accuracy": 0.21073752343654634,
      "num_tokens": 48755328.0,
      "step": 21285
    },
    {
      "entropy": 5.461286020278931,
      "epoch": 2.1045867931988926,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004559483440619255,
      "loss": 5.1214,
      "mean_token_accuracy": 0.20690563172101975,
      "num_tokens": 48766666.0,
      "step": 21290
    },
    {
      "entropy": 5.61240701675415,
      "epoch": 2.1050810597073943,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045592736763513285,
      "loss": 5.1694,
      "mean_token_accuracy": 0.1963336333632469,
      "num_tokens": 48779134.0,
      "step": 21295
    },
    {
      "entropy": 5.584633636474609,
      "epoch": 2.1055753262158956,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004559063867574989,
      "loss": 5.2133,
      "mean_token_accuracy": 0.18171284943819047,
      "num_tokens": 48791338.0,
      "step": 21300
    },
    {
      "entropy": 5.505694770812989,
      "epoch": 2.106069592724397,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045588540142953973,
      "loss": 5.105,
      "mean_token_accuracy": 0.20337445735931398,
      "num_tokens": 48802711.0,
      "step": 21305
    },
    {
      "entropy": 5.468663024902344,
      "epoch": 2.1065638592328986,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004558644116517717,
      "loss": 5.1363,
      "mean_token_accuracy": 0.20200977474451065,
      "num_tokens": 48815219.0,
      "step": 21310
    },
    {
      "entropy": 5.475671195983887,
      "epoch": 2.1070581257414,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045584341742471095,
      "loss": 5.1588,
      "mean_token_accuracy": 0.19961438477039337,
      "num_tokens": 48826075.0,
      "step": 21315
    },
    {
      "entropy": 5.483574295043946,
      "epoch": 2.107552392249901,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004558224187488742,
      "loss": 5.0535,
      "mean_token_accuracy": 0.19951361119747163,
      "num_tokens": 48837375.0,
      "step": 21320
    },
    {
      "entropy": 5.543083572387696,
      "epoch": 2.1080466587584024,
      "grad_norm": 1.125,
      "learning_rate": 0.00045580141562477785,
      "loss": 5.1814,
      "mean_token_accuracy": 0.1939372792840004,
      "num_tokens": 48849205.0,
      "step": 21325
    },
    {
      "entropy": 5.492909240722656,
      "epoch": 2.108540925266904,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045578040805293864,
      "loss": 5.1621,
      "mean_token_accuracy": 0.19909158498048782,
      "num_tokens": 48861700.0,
      "step": 21330
    },
    {
      "entropy": 5.553277444839478,
      "epoch": 2.1090351917754053,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045575939603387336,
      "loss": 5.2921,
      "mean_token_accuracy": 0.18944236338138581,
      "num_tokens": 48873510.0,
      "step": 21335
    },
    {
      "entropy": 5.598937034606934,
      "epoch": 2.1095294582839066,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004557383795680988,
      "loss": 5.1557,
      "mean_token_accuracy": 0.19448843598365784,
      "num_tokens": 48885011.0,
      "step": 21340
    },
    {
      "entropy": 5.522098970413208,
      "epoch": 2.1100237247924083,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045571735865613215,
      "loss": 5.1191,
      "mean_token_accuracy": 0.1973096564412117,
      "num_tokens": 48895767.0,
      "step": 21345
    },
    {
      "entropy": 5.533459138870239,
      "epoch": 2.1105179913009096,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045569633329849043,
      "loss": 5.2381,
      "mean_token_accuracy": 0.18424018770456313,
      "num_tokens": 48906616.0,
      "step": 21350
    },
    {
      "entropy": 5.543784523010254,
      "epoch": 2.111012257809411,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004556753034956909,
      "loss": 5.167,
      "mean_token_accuracy": 0.19859943240880967,
      "num_tokens": 48916841.0,
      "step": 21355
    },
    {
      "entropy": 5.471170711517334,
      "epoch": 2.111506524317912,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004556542692482508,
      "loss": 5.1391,
      "mean_token_accuracy": 0.20517609566450118,
      "num_tokens": 48928798.0,
      "step": 21360
    },
    {
      "entropy": 5.477427244186401,
      "epoch": 2.112000790826414,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004556332305566877,
      "loss": 5.1247,
      "mean_token_accuracy": 0.20221721082925798,
      "num_tokens": 48940235.0,
      "step": 21365
    },
    {
      "entropy": 5.527065801620483,
      "epoch": 2.112495057334915,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004556121874215191,
      "loss": 5.1281,
      "mean_token_accuracy": 0.20175862610340117,
      "num_tokens": 48951718.0,
      "step": 21370
    },
    {
      "entropy": 5.51894588470459,
      "epoch": 2.1129893238434163,
      "grad_norm": 1.25,
      "learning_rate": 0.0004555911398432627,
      "loss": 5.1193,
      "mean_token_accuracy": 0.2022763729095459,
      "num_tokens": 48963215.0,
      "step": 21375
    },
    {
      "entropy": 5.568140697479248,
      "epoch": 2.1134835903519176,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045557008782243626,
      "loss": 5.2373,
      "mean_token_accuracy": 0.19304773956537247,
      "num_tokens": 48975761.0,
      "step": 21380
    },
    {
      "entropy": 5.500147342681885,
      "epoch": 2.1139778568604193,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004555490313595575,
      "loss": 5.1233,
      "mean_token_accuracy": 0.20306147038936614,
      "num_tokens": 48987192.0,
      "step": 21385
    },
    {
      "entropy": 5.4811516284942625,
      "epoch": 2.1144721233689205,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004555279704551447,
      "loss": 5.0893,
      "mean_token_accuracy": 0.20535937547683716,
      "num_tokens": 48999503.0,
      "step": 21390
    },
    {
      "entropy": 5.536003065109253,
      "epoch": 2.114966389877422,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004555069051097158,
      "loss": 5.1674,
      "mean_token_accuracy": 0.1897236704826355,
      "num_tokens": 49010046.0,
      "step": 21395
    },
    {
      "entropy": 5.497274017333984,
      "epoch": 2.115460656385923,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045548583532378903,
      "loss": 5.1516,
      "mean_token_accuracy": 0.20265138447284697,
      "num_tokens": 49021028.0,
      "step": 21400
    },
    {
      "entropy": 5.572155714035034,
      "epoch": 2.1159549228944248,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004554647610978827,
      "loss": 5.267,
      "mean_token_accuracy": 0.19370944499969484,
      "num_tokens": 49034522.0,
      "step": 21405
    },
    {
      "entropy": 5.5810693264007565,
      "epoch": 2.116449189402926,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004554436824325152,
      "loss": 5.2047,
      "mean_token_accuracy": 0.19548682868480682,
      "num_tokens": 49045034.0,
      "step": 21410
    },
    {
      "entropy": 5.458191728591919,
      "epoch": 2.1169434559114273,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045542259932820515,
      "loss": 5.1583,
      "mean_token_accuracy": 0.19981651604175568,
      "num_tokens": 49056350.0,
      "step": 21415
    },
    {
      "entropy": 5.576247358322144,
      "epoch": 2.117437722419929,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004554015117854712,
      "loss": 5.189,
      "mean_token_accuracy": 0.2013806641101837,
      "num_tokens": 49067362.0,
      "step": 21420
    },
    {
      "entropy": 5.52940993309021,
      "epoch": 2.1179319889284303,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004553804198048321,
      "loss": 5.141,
      "mean_token_accuracy": 0.1986973136663437,
      "num_tokens": 49078067.0,
      "step": 21425
    },
    {
      "entropy": 5.468345785140992,
      "epoch": 2.1184262554369315,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045535932338680655,
      "loss": 5.1616,
      "mean_token_accuracy": 0.2026003271341324,
      "num_tokens": 49090205.0,
      "step": 21430
    },
    {
      "entropy": 5.464451313018799,
      "epoch": 2.118920521945433,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004553382225319138,
      "loss": 5.1459,
      "mean_token_accuracy": 0.20440313667058946,
      "num_tokens": 49101134.0,
      "step": 21435
    },
    {
      "entropy": 5.607928705215454,
      "epoch": 2.1194147884539345,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004553171172406728,
      "loss": 5.177,
      "mean_token_accuracy": 0.18961822837591172,
      "num_tokens": 49112282.0,
      "step": 21440
    },
    {
      "entropy": 5.509439611434937,
      "epoch": 2.1199090549624358,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004552960075136027,
      "loss": 5.1136,
      "mean_token_accuracy": 0.19439624547958373,
      "num_tokens": 49123228.0,
      "step": 21445
    },
    {
      "entropy": 5.470113706588745,
      "epoch": 2.120403321470937,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045527489335122284,
      "loss": 5.149,
      "mean_token_accuracy": 0.20207837224006653,
      "num_tokens": 49134769.0,
      "step": 21450
    },
    {
      "entropy": 5.448904657363892,
      "epoch": 2.1208975879794387,
      "grad_norm": 1.25,
      "learning_rate": 0.00045525377475405266,
      "loss": 5.0752,
      "mean_token_accuracy": 0.20799223184585572,
      "num_tokens": 49146096.0,
      "step": 21455
    },
    {
      "entropy": 5.5035545349121096,
      "epoch": 2.12139185448794,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045523265172261155,
      "loss": 5.159,
      "mean_token_accuracy": 0.19232870936393737,
      "num_tokens": 49156829.0,
      "step": 21460
    },
    {
      "entropy": 5.516164350509643,
      "epoch": 2.1218861209964412,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045521152425741934,
      "loss": 5.1992,
      "mean_token_accuracy": 0.1983991637825966,
      "num_tokens": 49167364.0,
      "step": 21465
    },
    {
      "entropy": 5.516932249069214,
      "epoch": 2.1223803875049425,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045519039235899573,
      "loss": 5.1927,
      "mean_token_accuracy": 0.199073189496994,
      "num_tokens": 49178283.0,
      "step": 21470
    },
    {
      "entropy": 5.629363059997559,
      "epoch": 2.122874654013444,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045516925602786043,
      "loss": 5.2529,
      "mean_token_accuracy": 0.19027069061994553,
      "num_tokens": 49190466.0,
      "step": 21475
    },
    {
      "entropy": 5.534385395050049,
      "epoch": 2.1233689205219455,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004551481152645335,
      "loss": 5.1327,
      "mean_token_accuracy": 0.19836584776639937,
      "num_tokens": 49202601.0,
      "step": 21480
    },
    {
      "entropy": 5.410324048995972,
      "epoch": 2.1238631870304467,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045512697006953504,
      "loss": 4.9985,
      "mean_token_accuracy": 0.21097538322210313,
      "num_tokens": 49214228.0,
      "step": 21485
    },
    {
      "entropy": 5.4403187274932865,
      "epoch": 2.124357453538948,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045510582044338513,
      "loss": 5.1078,
      "mean_token_accuracy": 0.19763224422931672,
      "num_tokens": 49225495.0,
      "step": 21490
    },
    {
      "entropy": 5.542464351654052,
      "epoch": 2.1248517200474497,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000455084666386604,
      "loss": 5.164,
      "mean_token_accuracy": 0.20189469158649445,
      "num_tokens": 49237348.0,
      "step": 21495
    },
    {
      "entropy": 5.511417245864868,
      "epoch": 2.125345986555951,
      "grad_norm": 1.125,
      "learning_rate": 0.00045506350789971227,
      "loss": 5.1915,
      "mean_token_accuracy": 0.19231386929750444,
      "num_tokens": 49247691.0,
      "step": 21500
    },
    {
      "entropy": 5.498218202590943,
      "epoch": 2.1258402530644522,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004550423449832302,
      "loss": 5.1303,
      "mean_token_accuracy": 0.20230920165777205,
      "num_tokens": 49259191.0,
      "step": 21505
    },
    {
      "entropy": 5.517141723632813,
      "epoch": 2.126334519572954,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004550211776376787,
      "loss": 5.1028,
      "mean_token_accuracy": 0.19951507151126863,
      "num_tokens": 49271760.0,
      "step": 21510
    },
    {
      "entropy": 5.513000011444092,
      "epoch": 2.126828786081455,
      "grad_norm": 1.125,
      "learning_rate": 0.00045500000586357807,
      "loss": 5.2236,
      "mean_token_accuracy": 0.19550841748714448,
      "num_tokens": 49283918.0,
      "step": 21515
    },
    {
      "entropy": 5.532028245925903,
      "epoch": 2.1273230525899565,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045497882966144955,
      "loss": 5.0703,
      "mean_token_accuracy": 0.20106528401374818,
      "num_tokens": 49294313.0,
      "step": 21520
    },
    {
      "entropy": 5.430132675170898,
      "epoch": 2.1278173190984577,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004549576490318138,
      "loss": 5.0849,
      "mean_token_accuracy": 0.20017001628875733,
      "num_tokens": 49305213.0,
      "step": 21525
    },
    {
      "entropy": 5.481860733032226,
      "epoch": 2.1283115856069594,
      "grad_norm": 1.140625,
      "learning_rate": 0.000454936463975192,
      "loss": 5.0905,
      "mean_token_accuracy": 0.20660116225481034,
      "num_tokens": 49317245.0,
      "step": 21530
    },
    {
      "entropy": 5.59033613204956,
      "epoch": 2.1288058521154607,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004549152744921053,
      "loss": 5.2196,
      "mean_token_accuracy": 0.19397821724414827,
      "num_tokens": 49328400.0,
      "step": 21535
    },
    {
      "entropy": 5.595189237594605,
      "epoch": 2.129300118623962,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004548940805830749,
      "loss": 5.2373,
      "mean_token_accuracy": 0.19453509896993637,
      "num_tokens": 49341094.0,
      "step": 21540
    },
    {
      "entropy": 5.469591188430786,
      "epoch": 2.129794385132463,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045487288224862224,
      "loss": 5.1287,
      "mean_token_accuracy": 0.20355660170316697,
      "num_tokens": 49352472.0,
      "step": 21545
    },
    {
      "entropy": 5.428672981262207,
      "epoch": 2.130288651640965,
      "grad_norm": 1.125,
      "learning_rate": 0.00045485167948926876,
      "loss": 5.1174,
      "mean_token_accuracy": 0.19694542437791823,
      "num_tokens": 49362432.0,
      "step": 21550
    },
    {
      "entropy": 5.472903394699097,
      "epoch": 2.130782918149466,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000454830472305536,
      "loss": 5.1363,
      "mean_token_accuracy": 0.20871778577566147,
      "num_tokens": 49373857.0,
      "step": 21555
    },
    {
      "entropy": 5.593730497360229,
      "epoch": 2.1312771846579674,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004548092606979458,
      "loss": 5.2135,
      "mean_token_accuracy": 0.19492797404527665,
      "num_tokens": 49387545.0,
      "step": 21560
    },
    {
      "entropy": 5.5043810367584225,
      "epoch": 2.131771451166469,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045478804466701984,
      "loss": 5.1438,
      "mean_token_accuracy": 0.2008009672164917,
      "num_tokens": 49398273.0,
      "step": 21565
    },
    {
      "entropy": 5.548471403121948,
      "epoch": 2.1322657176749704,
      "grad_norm": 1.125,
      "learning_rate": 0.00045476682421328015,
      "loss": 5.1496,
      "mean_token_accuracy": 0.19599671065807342,
      "num_tokens": 49408662.0,
      "step": 21570
    },
    {
      "entropy": 5.542198657989502,
      "epoch": 2.1327599841834717,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045474559933724866,
      "loss": 5.2569,
      "mean_token_accuracy": 0.19440645575523377,
      "num_tokens": 49421189.0,
      "step": 21575
    },
    {
      "entropy": 5.559695625305176,
      "epoch": 2.133254250691973,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045472437003944753,
      "loss": 5.2208,
      "mean_token_accuracy": 0.19633277952671052,
      "num_tokens": 49432878.0,
      "step": 21580
    },
    {
      "entropy": 5.55558557510376,
      "epoch": 2.1337485172004746,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004547031363203991,
      "loss": 5.2395,
      "mean_token_accuracy": 0.19586218446493148,
      "num_tokens": 49444058.0,
      "step": 21585
    },
    {
      "entropy": 5.5462547779083256,
      "epoch": 2.134242783708976,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045468189818062554,
      "loss": 5.1538,
      "mean_token_accuracy": 0.19348617941141127,
      "num_tokens": 49456602.0,
      "step": 21590
    },
    {
      "entropy": 5.514054775238037,
      "epoch": 2.134737050217477,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004546606556206494,
      "loss": 5.157,
      "mean_token_accuracy": 0.19638571590185167,
      "num_tokens": 49467287.0,
      "step": 21595
    },
    {
      "entropy": 5.544499683380127,
      "epoch": 2.135231316725979,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004546394086409933,
      "loss": 5.102,
      "mean_token_accuracy": 0.19600446075201033,
      "num_tokens": 49478662.0,
      "step": 21600
    },
    {
      "entropy": 5.468817853927613,
      "epoch": 2.13572558323448,
      "grad_norm": 1.359375,
      "learning_rate": 0.00045461815724217984,
      "loss": 5.1732,
      "mean_token_accuracy": 0.20004867464303971,
      "num_tokens": 49489365.0,
      "step": 21605
    },
    {
      "entropy": 5.518802499771118,
      "epoch": 2.1362198497429814,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045459690142473193,
      "loss": 5.208,
      "mean_token_accuracy": 0.19705337584018706,
      "num_tokens": 49501318.0,
      "step": 21610
    },
    {
      "entropy": 5.586258745193481,
      "epoch": 2.1367141162514827,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004545756411891723,
      "loss": 5.2171,
      "mean_token_accuracy": 0.19001415967941285,
      "num_tokens": 49512138.0,
      "step": 21615
    },
    {
      "entropy": 5.5695287704467775,
      "epoch": 2.1372083827599844,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004545543765360241,
      "loss": 5.1393,
      "mean_token_accuracy": 0.19415713399648665,
      "num_tokens": 49523325.0,
      "step": 21620
    },
    {
      "entropy": 5.435473775863647,
      "epoch": 2.1377026492684856,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045453310746581036,
      "loss": 5.0925,
      "mean_token_accuracy": 0.2022656336426735,
      "num_tokens": 49535382.0,
      "step": 21625
    },
    {
      "entropy": 5.486438798904419,
      "epoch": 2.138196915776987,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004545118339790543,
      "loss": 5.0742,
      "mean_token_accuracy": 0.20641831308603287,
      "num_tokens": 49546151.0,
      "step": 21630
    },
    {
      "entropy": 5.452876615524292,
      "epoch": 2.138691182285488,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004544905560762793,
      "loss": 5.1523,
      "mean_token_accuracy": 0.20317295491695403,
      "num_tokens": 49558617.0,
      "step": 21635
    },
    {
      "entropy": 5.494188022613526,
      "epoch": 2.13918544879399,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045446927375800876,
      "loss": 5.0676,
      "mean_token_accuracy": 0.2069677084684372,
      "num_tokens": 49568494.0,
      "step": 21640
    },
    {
      "entropy": 5.559956884384155,
      "epoch": 2.139679715302491,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045444798702476625,
      "loss": 5.1597,
      "mean_token_accuracy": 0.20608499348163606,
      "num_tokens": 49579763.0,
      "step": 21645
    },
    {
      "entropy": 5.473138952255249,
      "epoch": 2.1401739818109924,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004544266958770755,
      "loss": 5.0756,
      "mean_token_accuracy": 0.203498013317585,
      "num_tokens": 49591631.0,
      "step": 21650
    },
    {
      "entropy": 5.485914611816407,
      "epoch": 2.1406682483194936,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00045440540031546004,
      "loss": 5.157,
      "mean_token_accuracy": 0.19545343071222304,
      "num_tokens": 49603458.0,
      "step": 21655
    },
    {
      "entropy": 5.479412460327149,
      "epoch": 2.1411625148279954,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045438410034044397,
      "loss": 5.1098,
      "mean_token_accuracy": 0.19994595795869827,
      "num_tokens": 49614076.0,
      "step": 21660
    },
    {
      "entropy": 5.535571908950805,
      "epoch": 2.1416567813364966,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045436279595255125,
      "loss": 5.2348,
      "mean_token_accuracy": 0.19829751253128053,
      "num_tokens": 49624982.0,
      "step": 21665
    },
    {
      "entropy": 5.504553031921387,
      "epoch": 2.142151047844998,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004543414871523059,
      "loss": 5.104,
      "mean_token_accuracy": 0.20362628549337386,
      "num_tokens": 49635897.0,
      "step": 21670
    },
    {
      "entropy": 5.470793914794922,
      "epoch": 2.1426453143534996,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004543201739402321,
      "loss": 5.1085,
      "mean_token_accuracy": 0.2038417413830757,
      "num_tokens": 49647023.0,
      "step": 21675
    },
    {
      "entropy": 5.541865301132202,
      "epoch": 2.143139580862001,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045429885631685424,
      "loss": 5.209,
      "mean_token_accuracy": 0.18845810145139694,
      "num_tokens": 49658578.0,
      "step": 21680
    },
    {
      "entropy": 5.552652645111084,
      "epoch": 2.143633847370502,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004542775342826967,
      "loss": 5.14,
      "mean_token_accuracy": 0.20193170011043549,
      "num_tokens": 49670978.0,
      "step": 21685
    },
    {
      "entropy": 5.424201250076294,
      "epoch": 2.1441281138790034,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00045425620783828396,
      "loss": 5.1407,
      "mean_token_accuracy": 0.2037213549017906,
      "num_tokens": 49682553.0,
      "step": 21690
    },
    {
      "entropy": 5.517558479309082,
      "epoch": 2.144622380387505,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004542348769841407,
      "loss": 5.1459,
      "mean_token_accuracy": 0.19859371036291124,
      "num_tokens": 49694589.0,
      "step": 21695
    },
    {
      "entropy": 5.533109378814697,
      "epoch": 2.1451166468960063,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045421354172079173,
      "loss": 5.1679,
      "mean_token_accuracy": 0.20135858058929443,
      "num_tokens": 49706530.0,
      "step": 21700
    },
    {
      "entropy": 5.475673723220825,
      "epoch": 2.1456109134045076,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045419220204876175,
      "loss": 5.1499,
      "mean_token_accuracy": 0.20309580713510514,
      "num_tokens": 49718524.0,
      "step": 21705
    },
    {
      "entropy": 5.5505389213562015,
      "epoch": 2.1461051799130093,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045417085796857596,
      "loss": 5.1651,
      "mean_token_accuracy": 0.19750214070081712,
      "num_tokens": 49729555.0,
      "step": 21710
    },
    {
      "entropy": 5.5415137767791744,
      "epoch": 2.1465994464215106,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045414950948075906,
      "loss": 5.2184,
      "mean_token_accuracy": 0.19754463583230972,
      "num_tokens": 49739708.0,
      "step": 21715
    },
    {
      "entropy": 5.568691921234131,
      "epoch": 2.147093712930012,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045412815658583657,
      "loss": 5.1886,
      "mean_token_accuracy": 0.1967131555080414,
      "num_tokens": 49750997.0,
      "step": 21720
    },
    {
      "entropy": 5.493296432495117,
      "epoch": 2.147587979438513,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004541067992843336,
      "loss": 5.1439,
      "mean_token_accuracy": 0.2032516360282898,
      "num_tokens": 49761508.0,
      "step": 21725
    },
    {
      "entropy": 5.4815867900848385,
      "epoch": 2.148082245947015,
      "grad_norm": 1.46875,
      "learning_rate": 0.00045408543757677555,
      "loss": 5.0441,
      "mean_token_accuracy": 0.20154463797807692,
      "num_tokens": 49772821.0,
      "step": 21730
    },
    {
      "entropy": 5.428773832321167,
      "epoch": 2.148576512455516,
      "grad_norm": 1.234375,
      "learning_rate": 0.000454064071463688,
      "loss": 5.0733,
      "mean_token_accuracy": 0.2097561851143837,
      "num_tokens": 49783932.0,
      "step": 21735
    },
    {
      "entropy": 5.495036220550537,
      "epoch": 2.1490707789640173,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004540427009455965,
      "loss": 5.1224,
      "mean_token_accuracy": 0.19908437877893448,
      "num_tokens": 49795126.0,
      "step": 21740
    },
    {
      "entropy": 5.501727437973022,
      "epoch": 2.1495650454725186,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004540213260230267,
      "loss": 5.1596,
      "mean_token_accuracy": 0.19532833695411683,
      "num_tokens": 49806110.0,
      "step": 21745
    },
    {
      "entropy": 5.616278839111328,
      "epoch": 2.1500593119810203,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004539999466965046,
      "loss": 5.2958,
      "mean_token_accuracy": 0.1889016091823578,
      "num_tokens": 49818879.0,
      "step": 21750
    },
    {
      "entropy": 5.524253177642822,
      "epoch": 2.1505535784895216,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045397856296655603,
      "loss": 5.1064,
      "mean_token_accuracy": 0.1991901069879532,
      "num_tokens": 49828986.0,
      "step": 21755
    },
    {
      "entropy": 5.518258571624756,
      "epoch": 2.151047844998023,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045395717483370703,
      "loss": 5.187,
      "mean_token_accuracy": 0.1985077053308487,
      "num_tokens": 49841987.0,
      "step": 21760
    },
    {
      "entropy": 5.504703712463379,
      "epoch": 2.1515421115065245,
      "grad_norm": 1.359375,
      "learning_rate": 0.00045393578229848373,
      "loss": 5.1672,
      "mean_token_accuracy": 0.1920861229300499,
      "num_tokens": 49854151.0,
      "step": 21765
    },
    {
      "entropy": 5.524015855789185,
      "epoch": 2.152036378015026,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045391438536141257,
      "loss": 5.1838,
      "mean_token_accuracy": 0.2024367183446884,
      "num_tokens": 49866531.0,
      "step": 21770
    },
    {
      "entropy": 5.503477001190186,
      "epoch": 2.152530644523527,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045389298402301965,
      "loss": 5.1038,
      "mean_token_accuracy": 0.20770833343267442,
      "num_tokens": 49877737.0,
      "step": 21775
    },
    {
      "entropy": 5.616457605361939,
      "epoch": 2.1530249110320283,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004538715782838316,
      "loss": 5.1967,
      "mean_token_accuracy": 0.19559425264596939,
      "num_tokens": 49889214.0,
      "step": 21780
    },
    {
      "entropy": 5.513102626800537,
      "epoch": 2.15351917754053,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045385016814437494,
      "loss": 5.1657,
      "mean_token_accuracy": 0.20539307594299316,
      "num_tokens": 49900293.0,
      "step": 21785
    },
    {
      "entropy": 5.570395803451538,
      "epoch": 2.1540134440490313,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045382875360517645,
      "loss": 5.24,
      "mean_token_accuracy": 0.18594724535942078,
      "num_tokens": 49912193.0,
      "step": 21790
    },
    {
      "entropy": 5.564964246749878,
      "epoch": 2.1545077105575325,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004538073346667628,
      "loss": 5.1221,
      "mean_token_accuracy": 0.20388921201229096,
      "num_tokens": 49922780.0,
      "step": 21795
    },
    {
      "entropy": 5.469277191162109,
      "epoch": 2.155001977066034,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045378591132966105,
      "loss": 5.0884,
      "mean_token_accuracy": 0.2023615777492523,
      "num_tokens": 49934551.0,
      "step": 21800
    },
    {
      "entropy": 5.51720085144043,
      "epoch": 2.1554962435745355,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004537644835943982,
      "loss": 5.1404,
      "mean_token_accuracy": 0.2002007856965065,
      "num_tokens": 49945779.0,
      "step": 21805
    },
    {
      "entropy": 5.473114347457885,
      "epoch": 2.1559905100830368,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004537430514615012,
      "loss": 5.0267,
      "mean_token_accuracy": 0.20844293385744095,
      "num_tokens": 49957127.0,
      "step": 21810
    },
    {
      "entropy": 5.497587776184082,
      "epoch": 2.156484776591538,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045372161493149746,
      "loss": 5.1075,
      "mean_token_accuracy": 0.20073380023241044,
      "num_tokens": 49968679.0,
      "step": 21815
    },
    {
      "entropy": 5.512000274658203,
      "epoch": 2.1569790431000397,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004537001740049142,
      "loss": 5.1339,
      "mean_token_accuracy": 0.20407009720802308,
      "num_tokens": 49979924.0,
      "step": 21820
    },
    {
      "entropy": 5.5685670375823975,
      "epoch": 2.157473309608541,
      "grad_norm": 1.328125,
      "learning_rate": 0.000453678728682279,
      "loss": 5.1637,
      "mean_token_accuracy": 0.19451797902584075,
      "num_tokens": 49991318.0,
      "step": 21825
    },
    {
      "entropy": 5.411525678634644,
      "epoch": 2.1579675761170423,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004536572789641193,
      "loss": 5.0614,
      "mean_token_accuracy": 0.2094429239630699,
      "num_tokens": 50003096.0,
      "step": 21830
    },
    {
      "entropy": 5.4729186534881595,
      "epoch": 2.1584618426255435,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045363582485096295,
      "loss": 5.1098,
      "mean_token_accuracy": 0.20159204304218292,
      "num_tokens": 50013953.0,
      "step": 21835
    },
    {
      "entropy": 5.529362201690674,
      "epoch": 2.1589561091340452,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004536143663433375,
      "loss": 5.1217,
      "mean_token_accuracy": 0.19942850917577742,
      "num_tokens": 50025696.0,
      "step": 21840
    },
    {
      "entropy": 5.517219161987304,
      "epoch": 2.1594503756425465,
      "grad_norm": 1.125,
      "learning_rate": 0.00045359290344177094,
      "loss": 5.1176,
      "mean_token_accuracy": 0.20553035736083985,
      "num_tokens": 50036880.0,
      "step": 21845
    },
    {
      "entropy": 5.46474175453186,
      "epoch": 2.1599446421510478,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004535714361467913,
      "loss": 5.1819,
      "mean_token_accuracy": 0.19759765714406968,
      "num_tokens": 50048746.0,
      "step": 21850
    },
    {
      "entropy": 5.488521242141724,
      "epoch": 2.1604389086595495,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004535499644589266,
      "loss": 5.1115,
      "mean_token_accuracy": 0.20375827699899673,
      "num_tokens": 50060711.0,
      "step": 21855
    },
    {
      "entropy": 5.525929641723633,
      "epoch": 2.1609331751680507,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045352848837870496,
      "loss": 5.1121,
      "mean_token_accuracy": 0.20056418627500533,
      "num_tokens": 50071878.0,
      "step": 21860
    },
    {
      "entropy": 5.5766599655151365,
      "epoch": 2.161427441676552,
      "grad_norm": 1.265625,
      "learning_rate": 0.000453507007906655,
      "loss": 5.1843,
      "mean_token_accuracy": 0.19560859203338624,
      "num_tokens": 50084600.0,
      "step": 21865
    },
    {
      "entropy": 5.427535343170166,
      "epoch": 2.1619217081850532,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004534855230433048,
      "loss": 5.0292,
      "mean_token_accuracy": 0.2078049898147583,
      "num_tokens": 50095814.0,
      "step": 21870
    },
    {
      "entropy": 5.470051288604736,
      "epoch": 2.162415974693555,
      "grad_norm": 2.015625,
      "learning_rate": 0.00045346403378918314,
      "loss": 5.1804,
      "mean_token_accuracy": 0.19895249158143996,
      "num_tokens": 50108390.0,
      "step": 21875
    },
    {
      "entropy": 5.52471776008606,
      "epoch": 2.162910241202056,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004534425401448185,
      "loss": 5.2202,
      "mean_token_accuracy": 0.2009907454252243,
      "num_tokens": 50118939.0,
      "step": 21880
    },
    {
      "entropy": 5.466049480438232,
      "epoch": 2.1634045077105575,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004534210421107397,
      "loss": 5.0865,
      "mean_token_accuracy": 0.21050201058387757,
      "num_tokens": 50129510.0,
      "step": 21885
    },
    {
      "entropy": 5.485499715805053,
      "epoch": 2.1638987742190587,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004533995396874756,
      "loss": 5.1496,
      "mean_token_accuracy": 0.20328801423311232,
      "num_tokens": 50140611.0,
      "step": 21890
    },
    {
      "entropy": 5.525869512557984,
      "epoch": 2.1643930407275604,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045337803287555525,
      "loss": 5.1817,
      "mean_token_accuracy": 0.19648308008909227,
      "num_tokens": 50152214.0,
      "step": 21895
    },
    {
      "entropy": 5.619536066055298,
      "epoch": 2.1648873072360617,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004533565216755075,
      "loss": 5.2606,
      "mean_token_accuracy": 0.18822956085205078,
      "num_tokens": 50164201.0,
      "step": 21900
    },
    {
      "entropy": 5.503797435760498,
      "epoch": 2.165381573744563,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004533350060878615,
      "loss": 5.1109,
      "mean_token_accuracy": 0.20617415457963945,
      "num_tokens": 50176515.0,
      "step": 21905
    },
    {
      "entropy": 5.535329151153564,
      "epoch": 2.1658758402530642,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004533134861131469,
      "loss": 5.1539,
      "mean_token_accuracy": 0.19906970709562302,
      "num_tokens": 50187624.0,
      "step": 21910
    },
    {
      "entropy": 5.3744159698486325,
      "epoch": 2.166370106761566,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004532919617518929,
      "loss": 4.9829,
      "mean_token_accuracy": 0.21052872240543366,
      "num_tokens": 50199549.0,
      "step": 21915
    },
    {
      "entropy": 5.500524520874023,
      "epoch": 2.166864373270067,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045327043300462887,
      "loss": 5.1831,
      "mean_token_accuracy": 0.19515424221754074,
      "num_tokens": 50210920.0,
      "step": 21920
    },
    {
      "entropy": 5.524771451950073,
      "epoch": 2.1673586397785685,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004532488998718846,
      "loss": 5.1081,
      "mean_token_accuracy": 0.2101850390434265,
      "num_tokens": 50221335.0,
      "step": 21925
    },
    {
      "entropy": 5.548227214813233,
      "epoch": 2.16785290628707,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00045322736235418973,
      "loss": 5.207,
      "mean_token_accuracy": 0.19192128479480744,
      "num_tokens": 50235138.0,
      "step": 21930
    },
    {
      "entropy": 5.502050495147705,
      "epoch": 2.1683471727955714,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045320582045207405,
      "loss": 5.1108,
      "mean_token_accuracy": 0.20652009546756744,
      "num_tokens": 50246454.0,
      "step": 21935
    },
    {
      "entropy": 5.497928428649902,
      "epoch": 2.1688414393040727,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004531842741660675,
      "loss": 5.1537,
      "mean_token_accuracy": 0.1955179199576378,
      "num_tokens": 50257591.0,
      "step": 21940
    },
    {
      "entropy": 5.530389356613159,
      "epoch": 2.169335705812574,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004531627234967003,
      "loss": 5.1275,
      "mean_token_accuracy": 0.2016258120536804,
      "num_tokens": 50268557.0,
      "step": 21945
    },
    {
      "entropy": 5.583495569229126,
      "epoch": 2.1698299723210757,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004531411684445024,
      "loss": 5.2668,
      "mean_token_accuracy": 0.1932716965675354,
      "num_tokens": 50280617.0,
      "step": 21950
    },
    {
      "entropy": 5.485199451446533,
      "epoch": 2.170324238829577,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045311960901000403,
      "loss": 5.1042,
      "mean_token_accuracy": 0.19730470329523087,
      "num_tokens": 50292071.0,
      "step": 21955
    },
    {
      "entropy": 5.487246084213257,
      "epoch": 2.170818505338078,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004530980451937358,
      "loss": 5.1204,
      "mean_token_accuracy": 0.19923310726881027,
      "num_tokens": 50303389.0,
      "step": 21960
    },
    {
      "entropy": 5.498326826095581,
      "epoch": 2.17131277184658,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004530764769962279,
      "loss": 5.1802,
      "mean_token_accuracy": 0.20463499724864959,
      "num_tokens": 50315088.0,
      "step": 21965
    },
    {
      "entropy": 5.506351566314697,
      "epoch": 2.171807038355081,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004530549044180111,
      "loss": 5.1188,
      "mean_token_accuracy": 0.20129501968622207,
      "num_tokens": 50326177.0,
      "step": 21970
    },
    {
      "entropy": 5.532204484939575,
      "epoch": 2.1723013048635824,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045303332745961605,
      "loss": 5.1246,
      "mean_token_accuracy": 0.20290366113185881,
      "num_tokens": 50337333.0,
      "step": 21975
    },
    {
      "entropy": 5.550925159454346,
      "epoch": 2.1727955713720837,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045301174612157345,
      "loss": 5.1486,
      "mean_token_accuracy": 0.20356634110212327,
      "num_tokens": 50348406.0,
      "step": 21980
    },
    {
      "entropy": 5.458467435836792,
      "epoch": 2.1732898378805854,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004529901604044143,
      "loss": 5.0808,
      "mean_token_accuracy": 0.20767980813980103,
      "num_tokens": 50360057.0,
      "step": 21985
    },
    {
      "entropy": 5.5037603855133055,
      "epoch": 2.1737841043890866,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004529685703086697,
      "loss": 5.1361,
      "mean_token_accuracy": 0.20416694432497023,
      "num_tokens": 50370355.0,
      "step": 21990
    },
    {
      "entropy": 5.534438848495483,
      "epoch": 2.174278370897588,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045294697583487056,
      "loss": 5.1543,
      "mean_token_accuracy": 0.2005469873547554,
      "num_tokens": 50382193.0,
      "step": 21995
    },
    {
      "entropy": 5.465809488296509,
      "epoch": 2.174772637406089,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045292537698354824,
      "loss": 5.0979,
      "mean_token_accuracy": 0.20017559677362443,
      "num_tokens": 50394383.0,
      "step": 22000
    },
    {
      "entropy": 5.49526572227478,
      "epoch": 2.175266903914591,
      "grad_norm": 1.6015625,
      "learning_rate": 0.000452903773755234,
      "loss": 5.061,
      "mean_token_accuracy": 0.20688024759292603,
      "num_tokens": 50405006.0,
      "step": 22005
    },
    {
      "entropy": 5.586110162734985,
      "epoch": 2.175761170423092,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004528821661504594,
      "loss": 5.2197,
      "mean_token_accuracy": 0.1991715431213379,
      "num_tokens": 50416940.0,
      "step": 22010
    },
    {
      "entropy": 5.469919347763062,
      "epoch": 2.1762554369315934,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004528605541697558,
      "loss": 5.144,
      "mean_token_accuracy": 0.20250431448221207,
      "num_tokens": 50428506.0,
      "step": 22015
    },
    {
      "entropy": 5.571971225738525,
      "epoch": 2.176749703440095,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045283893781365513,
      "loss": 5.1986,
      "mean_token_accuracy": 0.19555224925279618,
      "num_tokens": 50440421.0,
      "step": 22020
    },
    {
      "entropy": 5.550393438339233,
      "epoch": 2.1772439699485964,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004528173170826889,
      "loss": 5.1978,
      "mean_token_accuracy": 0.1958077847957611,
      "num_tokens": 50451819.0,
      "step": 22025
    },
    {
      "entropy": 5.615447998046875,
      "epoch": 2.1777382364570976,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004527956919773891,
      "loss": 5.2848,
      "mean_token_accuracy": 0.18724677860736846,
      "num_tokens": 50464382.0,
      "step": 22030
    },
    {
      "entropy": 5.556777286529541,
      "epoch": 2.178232502965599,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004527740624982877,
      "loss": 5.1492,
      "mean_token_accuracy": 0.20277251452207565,
      "num_tokens": 50474773.0,
      "step": 22035
    },
    {
      "entropy": 5.494979286193848,
      "epoch": 2.1787267694741006,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045275242864591683,
      "loss": 5.08,
      "mean_token_accuracy": 0.2024929478764534,
      "num_tokens": 50485626.0,
      "step": 22040
    },
    {
      "entropy": 5.479667377471924,
      "epoch": 2.179221035982602,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045273079042080854,
      "loss": 5.0967,
      "mean_token_accuracy": 0.20122305303812027,
      "num_tokens": 50496130.0,
      "step": 22045
    },
    {
      "entropy": 5.480082607269287,
      "epoch": 2.179715302491103,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045270914782349534,
      "loss": 5.1887,
      "mean_token_accuracy": 0.2030102089047432,
      "num_tokens": 50506578.0,
      "step": 22050
    },
    {
      "entropy": 5.520930814743042,
      "epoch": 2.1802095689996044,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045268750085450946,
      "loss": 5.2351,
      "mean_token_accuracy": 0.18985148519277573,
      "num_tokens": 50517398.0,
      "step": 22055
    },
    {
      "entropy": 5.456039094924927,
      "epoch": 2.180703835508106,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004526658495143835,
      "loss": 5.0484,
      "mean_token_accuracy": 0.21241284161806107,
      "num_tokens": 50528577.0,
      "step": 22060
    },
    {
      "entropy": 5.5150164604187015,
      "epoch": 2.1811981020166074,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004526441938036501,
      "loss": 5.1234,
      "mean_token_accuracy": 0.20830536335706712,
      "num_tokens": 50539118.0,
      "step": 22065
    },
    {
      "entropy": 5.462718534469604,
      "epoch": 2.1816923685251086,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00045262253372284193,
      "loss": 5.1802,
      "mean_token_accuracy": 0.19968547224998473,
      "num_tokens": 50549934.0,
      "step": 22070
    },
    {
      "entropy": 5.5227500915527346,
      "epoch": 2.1821866350336103,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045260086927249193,
      "loss": 5.1329,
      "mean_token_accuracy": 0.20440291613340378,
      "num_tokens": 50560661.0,
      "step": 22075
    },
    {
      "entropy": 5.557682609558105,
      "epoch": 2.1826809015421116,
      "grad_norm": 1.125,
      "learning_rate": 0.000452579200453133,
      "loss": 5.196,
      "mean_token_accuracy": 0.19065165966749192,
      "num_tokens": 50573304.0,
      "step": 22080
    },
    {
      "entropy": 5.5410960674285885,
      "epoch": 2.183175168050613,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004525575272652981,
      "loss": 5.1909,
      "mean_token_accuracy": 0.19336317777633666,
      "num_tokens": 50584166.0,
      "step": 22085
    },
    {
      "entropy": 5.614899730682373,
      "epoch": 2.183669434559114,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004525358497095206,
      "loss": 5.1677,
      "mean_token_accuracy": 0.192666257917881,
      "num_tokens": 50596425.0,
      "step": 22090
    },
    {
      "entropy": 5.468387079238892,
      "epoch": 2.184163701067616,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045251416778633354,
      "loss": 5.0842,
      "mean_token_accuracy": 0.207627472281456,
      "num_tokens": 50607483.0,
      "step": 22095
    },
    {
      "entropy": 5.465862512588501,
      "epoch": 2.184657967576117,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045249248149627046,
      "loss": 5.0862,
      "mean_token_accuracy": 0.20716465413570403,
      "num_tokens": 50619655.0,
      "step": 22100
    },
    {
      "entropy": 5.51420407295227,
      "epoch": 2.1851522340846183,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045247079083986485,
      "loss": 5.1804,
      "mean_token_accuracy": 0.19372120201587678,
      "num_tokens": 50631323.0,
      "step": 22105
    },
    {
      "entropy": 5.511419153213501,
      "epoch": 2.18564650059312,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004524490958176502,
      "loss": 5.0746,
      "mean_token_accuracy": 0.2087537795305252,
      "num_tokens": 50642931.0,
      "step": 22110
    },
    {
      "entropy": 5.442334318161011,
      "epoch": 2.1861407671016213,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004524273964301603,
      "loss": 5.0537,
      "mean_token_accuracy": 0.2096376121044159,
      "num_tokens": 50655396.0,
      "step": 22115
    },
    {
      "entropy": 5.4266410827636715,
      "epoch": 2.1866350336101226,
      "grad_norm": 1.359375,
      "learning_rate": 0.00045240569267792887,
      "loss": 5.1726,
      "mean_token_accuracy": 0.20235062539577484,
      "num_tokens": 50667264.0,
      "step": 22120
    },
    {
      "entropy": 5.413727807998657,
      "epoch": 2.187129300118624,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045238398456148995,
      "loss": 5.0766,
      "mean_token_accuracy": 0.20579549074172973,
      "num_tokens": 50679073.0,
      "step": 22125
    },
    {
      "entropy": 5.568247842788696,
      "epoch": 2.1876235666271255,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045236227208137744,
      "loss": 5.1456,
      "mean_token_accuracy": 0.19937388300895692,
      "num_tokens": 50689554.0,
      "step": 22130
    },
    {
      "entropy": 5.555780267715454,
      "epoch": 2.188117833135627,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004523405552381256,
      "loss": 5.1793,
      "mean_token_accuracy": 0.19624852985143662,
      "num_tokens": 50702419.0,
      "step": 22135
    },
    {
      "entropy": 5.506539630889892,
      "epoch": 2.188612099644128,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004523188340322685,
      "loss": 5.0861,
      "mean_token_accuracy": 0.20108540058135987,
      "num_tokens": 50713016.0,
      "step": 22140
    },
    {
      "entropy": 5.541036796569824,
      "epoch": 2.1891063661526293,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004522971084643406,
      "loss": 5.1759,
      "mean_token_accuracy": 0.19601734727621078,
      "num_tokens": 50723936.0,
      "step": 22145
    },
    {
      "entropy": 5.5003172874450685,
      "epoch": 2.189600632661131,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045227537853487627,
      "loss": 5.1047,
      "mean_token_accuracy": 0.19951000660657883,
      "num_tokens": 50734339.0,
      "step": 22150
    },
    {
      "entropy": 5.555885648727417,
      "epoch": 2.1900948991696323,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045225364424441016,
      "loss": 5.2213,
      "mean_token_accuracy": 0.19235749989748002,
      "num_tokens": 50746262.0,
      "step": 22155
    },
    {
      "entropy": 5.538311958312988,
      "epoch": 2.1905891656781336,
      "grad_norm": 1.078125,
      "learning_rate": 0.000452231905593477,
      "loss": 5.1252,
      "mean_token_accuracy": 0.2023879736661911,
      "num_tokens": 50758218.0,
      "step": 22160
    },
    {
      "entropy": 5.561844730377198,
      "epoch": 2.191083432186635,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004522101625826113,
      "loss": 5.2138,
      "mean_token_accuracy": 0.19490393698215486,
      "num_tokens": 50769263.0,
      "step": 22165
    },
    {
      "entropy": 5.3979555606842045,
      "epoch": 2.1915776986951365,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004521884152123482,
      "loss": 5.0137,
      "mean_token_accuracy": 0.21521006673574447,
      "num_tokens": 50780049.0,
      "step": 22170
    },
    {
      "entropy": 5.488247966766357,
      "epoch": 2.192071965203638,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004521666634832227,
      "loss": 5.1471,
      "mean_token_accuracy": 0.19518774896860122,
      "num_tokens": 50791817.0,
      "step": 22175
    },
    {
      "entropy": 5.51728367805481,
      "epoch": 2.192566231712139,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004521449073957697,
      "loss": 5.1347,
      "mean_token_accuracy": 0.20143390446901321,
      "num_tokens": 50802669.0,
      "step": 22180
    },
    {
      "entropy": 5.518154668807983,
      "epoch": 2.1930604982206408,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045212314695052447,
      "loss": 5.1208,
      "mean_token_accuracy": 0.20137811899185182,
      "num_tokens": 50813920.0,
      "step": 22185
    },
    {
      "entropy": 5.5540753364562985,
      "epoch": 2.193554764729142,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045210138214802227,
      "loss": 5.1983,
      "mean_token_accuracy": 0.19635826349258423,
      "num_tokens": 50825828.0,
      "step": 22190
    },
    {
      "entropy": 5.5389691352844235,
      "epoch": 2.1940490312376433,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045207961298879875,
      "loss": 5.2044,
      "mean_token_accuracy": 0.19437855631113052,
      "num_tokens": 50837425.0,
      "step": 22195
    },
    {
      "entropy": 5.55733208656311,
      "epoch": 2.1945432977461445,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004520578394733892,
      "loss": 5.1613,
      "mean_token_accuracy": 0.202816841006279,
      "num_tokens": 50848720.0,
      "step": 22200
    },
    {
      "entropy": 5.482304906845092,
      "epoch": 2.1950375642546462,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045203606160232936,
      "loss": 5.1109,
      "mean_token_accuracy": 0.19998920559883118,
      "num_tokens": 50860304.0,
      "step": 22205
    },
    {
      "entropy": 5.464160203933716,
      "epoch": 2.1955318307631475,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004520142793761549,
      "loss": 5.1447,
      "mean_token_accuracy": 0.19810025691986083,
      "num_tokens": 50872099.0,
      "step": 22210
    },
    {
      "entropy": 5.612827968597412,
      "epoch": 2.1960260972716488,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004519924927954017,
      "loss": 5.2604,
      "mean_token_accuracy": 0.1948183462023735,
      "num_tokens": 50884116.0,
      "step": 22215
    },
    {
      "entropy": 5.519051885604858,
      "epoch": 2.1965203637801505,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004519707018606059,
      "loss": 5.0711,
      "mean_token_accuracy": 0.20548426806926728,
      "num_tokens": 50896293.0,
      "step": 22220
    },
    {
      "entropy": 5.4370691776275635,
      "epoch": 2.1970146302886517,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045194890657230315,
      "loss": 5.0511,
      "mean_token_accuracy": 0.20947284549474715,
      "num_tokens": 50908633.0,
      "step": 22225
    },
    {
      "entropy": 5.494141864776611,
      "epoch": 2.197508896797153,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045192710693103,
      "loss": 5.1052,
      "mean_token_accuracy": 0.20547571033239365,
      "num_tokens": 50919524.0,
      "step": 22230
    },
    {
      "entropy": 5.4792809009552,
      "epoch": 2.1980031633056543,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004519053029373224,
      "loss": 5.1585,
      "mean_token_accuracy": 0.197202630341053,
      "num_tokens": 50930157.0,
      "step": 22235
    },
    {
      "entropy": 5.625102424621582,
      "epoch": 2.198497429814156,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000451883494591717,
      "loss": 5.2223,
      "mean_token_accuracy": 0.1938192754983902,
      "num_tokens": 50942474.0,
      "step": 22240
    },
    {
      "entropy": 5.57127423286438,
      "epoch": 2.1989916963226572,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045186168189475026,
      "loss": 5.2041,
      "mean_token_accuracy": 0.1902976557612419,
      "num_tokens": 50954847.0,
      "step": 22245
    },
    {
      "entropy": 5.490539026260376,
      "epoch": 2.1994859628311585,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004518398648469586,
      "loss": 5.1814,
      "mean_token_accuracy": 0.20104799121618272,
      "num_tokens": 50966446.0,
      "step": 22250
    },
    {
      "entropy": 5.464949321746826,
      "epoch": 2.1999802293396598,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045181804344887887,
      "loss": 5.1314,
      "mean_token_accuracy": 0.20553895682096482,
      "num_tokens": 50977949.0,
      "step": 22255
    },
    {
      "entropy": 5.51120343208313,
      "epoch": 2.2004744958481615,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045179621770104786,
      "loss": 5.1674,
      "mean_token_accuracy": 0.20135581493377686,
      "num_tokens": 50989297.0,
      "step": 22260
    },
    {
      "entropy": 5.575305509567261,
      "epoch": 2.2009687623566627,
      "grad_norm": 1.375,
      "learning_rate": 0.00045177438760400244,
      "loss": 5.1951,
      "mean_token_accuracy": 0.19927078485488892,
      "num_tokens": 51001616.0,
      "step": 22265
    },
    {
      "entropy": 5.547829914093017,
      "epoch": 2.201463028865164,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004517525531582797,
      "loss": 5.1749,
      "mean_token_accuracy": 0.1956998363137245,
      "num_tokens": 51012870.0,
      "step": 22270
    },
    {
      "entropy": 5.464755201339722,
      "epoch": 2.2019572953736652,
      "grad_norm": 1.40625,
      "learning_rate": 0.00045173071436441665,
      "loss": 5.1265,
      "mean_token_accuracy": 0.20448900908231735,
      "num_tokens": 51025368.0,
      "step": 22275
    },
    {
      "entropy": 5.522630929946899,
      "epoch": 2.202451561882167,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004517088712229507,
      "loss": 5.1027,
      "mean_token_accuracy": 0.20338659435510636,
      "num_tokens": 51035964.0,
      "step": 22280
    },
    {
      "entropy": 5.535574340820313,
      "epoch": 2.202945828390668,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045168702373441904,
      "loss": 5.1153,
      "mean_token_accuracy": 0.1960943043231964,
      "num_tokens": 51048313.0,
      "step": 22285
    },
    {
      "entropy": 5.482205104827881,
      "epoch": 2.2034400948991695,
      "grad_norm": 1.25,
      "learning_rate": 0.00045166517189935924,
      "loss": 5.121,
      "mean_token_accuracy": 0.20774466693401336,
      "num_tokens": 51060503.0,
      "step": 22290
    },
    {
      "entropy": 5.436234045028686,
      "epoch": 2.203934361407671,
      "grad_norm": 1.125,
      "learning_rate": 0.00045164331571830875,
      "loss": 5.0379,
      "mean_token_accuracy": 0.20530752390623092,
      "num_tokens": 51070865.0,
      "step": 22295
    },
    {
      "entropy": 5.568738698959351,
      "epoch": 2.2044286279161724,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045162145519180534,
      "loss": 5.148,
      "mean_token_accuracy": 0.19823688715696336,
      "num_tokens": 51081158.0,
      "step": 22300
    },
    {
      "entropy": 5.46003098487854,
      "epoch": 2.2049228944246737,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004515995903203867,
      "loss": 5.0211,
      "mean_token_accuracy": 0.2167837902903557,
      "num_tokens": 51090661.0,
      "step": 22305
    },
    {
      "entropy": 5.4276769161224365,
      "epoch": 2.205417160933175,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004515777211045908,
      "loss": 5.127,
      "mean_token_accuracy": 0.19683904349803924,
      "num_tokens": 51101825.0,
      "step": 22310
    },
    {
      "entropy": 5.470873737335205,
      "epoch": 2.2059114274416767,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045155584754495546,
      "loss": 5.1635,
      "mean_token_accuracy": 0.20107779055833816,
      "num_tokens": 51114546.0,
      "step": 22315
    },
    {
      "entropy": 5.537352800369263,
      "epoch": 2.206405693950178,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004515339696420189,
      "loss": 5.1222,
      "mean_token_accuracy": 0.201772840321064,
      "num_tokens": 51124888.0,
      "step": 22320
    },
    {
      "entropy": 5.430833196640014,
      "epoch": 2.206899960458679,
      "grad_norm": 1.125,
      "learning_rate": 0.0004515120873963194,
      "loss": 5.0289,
      "mean_token_accuracy": 0.21231637597084047,
      "num_tokens": 51135892.0,
      "step": 22325
    },
    {
      "entropy": 5.490753078460694,
      "epoch": 2.207394226967181,
      "grad_norm": 1.125,
      "learning_rate": 0.0004514902008083951,
      "loss": 5.1369,
      "mean_token_accuracy": 0.19497858732938766,
      "num_tokens": 51147513.0,
      "step": 22330
    },
    {
      "entropy": 5.543672513961792,
      "epoch": 2.207888493475682,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004514683098787845,
      "loss": 5.1545,
      "mean_token_accuracy": 0.20054354816675185,
      "num_tokens": 51159514.0,
      "step": 22335
    },
    {
      "entropy": 5.566498517990112,
      "epoch": 2.2083827599841834,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004514464146080261,
      "loss": 5.2124,
      "mean_token_accuracy": 0.19934438169002533,
      "num_tokens": 51171245.0,
      "step": 22340
    },
    {
      "entropy": 5.507154893875122,
      "epoch": 2.2088770264926847,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045142451499665846,
      "loss": 5.1167,
      "mean_token_accuracy": 0.20402316600084305,
      "num_tokens": 51182763.0,
      "step": 22345
    },
    {
      "entropy": 5.561678552627564,
      "epoch": 2.2093712930011864,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045140261104522044,
      "loss": 5.1567,
      "mean_token_accuracy": 0.19880735427141188,
      "num_tokens": 51193811.0,
      "step": 22350
    },
    {
      "entropy": 5.451747703552246,
      "epoch": 2.2098655595096877,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004513807027542508,
      "loss": 5.066,
      "mean_token_accuracy": 0.20757050663232804,
      "num_tokens": 51203963.0,
      "step": 22355
    },
    {
      "entropy": 5.455369281768799,
      "epoch": 2.210359826018189,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004513587901242886,
      "loss": 5.0383,
      "mean_token_accuracy": 0.20728101432323456,
      "num_tokens": 51215128.0,
      "step": 22360
    },
    {
      "entropy": 5.4400492191314695,
      "epoch": 2.21085409252669,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045133687315587263,
      "loss": 5.063,
      "mean_token_accuracy": 0.20726129710674285,
      "num_tokens": 51226773.0,
      "step": 22365
    },
    {
      "entropy": 5.443158340454102,
      "epoch": 2.211348359035192,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004513149518495424,
      "loss": 5.1112,
      "mean_token_accuracy": 0.2014615133404732,
      "num_tokens": 51238226.0,
      "step": 22370
    },
    {
      "entropy": 5.5417955875396725,
      "epoch": 2.211842625543693,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045129302620583683,
      "loss": 5.1354,
      "mean_token_accuracy": 0.20179601162672042,
      "num_tokens": 51249096.0,
      "step": 22375
    },
    {
      "entropy": 5.542620038986206,
      "epoch": 2.2123368920521944,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045127109622529563,
      "loss": 5.1383,
      "mean_token_accuracy": 0.20518684834241868,
      "num_tokens": 51259709.0,
      "step": 22380
    },
    {
      "entropy": 5.427329730987549,
      "epoch": 2.212831158560696,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045124916190845797,
      "loss": 5.1409,
      "mean_token_accuracy": 0.20966077893972396,
      "num_tokens": 51271787.0,
      "step": 22385
    },
    {
      "entropy": 5.616235589981079,
      "epoch": 2.2133254250691974,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045122722325586373,
      "loss": 5.1933,
      "mean_token_accuracy": 0.1969411239027977,
      "num_tokens": 51284093.0,
      "step": 22390
    },
    {
      "entropy": 5.584792518615723,
      "epoch": 2.2138196915776986,
      "grad_norm": 1.4375,
      "learning_rate": 0.00045120528026805235,
      "loss": 5.1653,
      "mean_token_accuracy": 0.20590700507164,
      "num_tokens": 51295659.0,
      "step": 22395
    },
    {
      "entropy": 5.476453065872192,
      "epoch": 2.2143139580862,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045118333294556377,
      "loss": 5.2022,
      "mean_token_accuracy": 0.19549500495195388,
      "num_tokens": 51307960.0,
      "step": 22400
    },
    {
      "entropy": 5.459325885772705,
      "epoch": 2.2148082245947016,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004511613812889379,
      "loss": 5.0289,
      "mean_token_accuracy": 0.21107092052698134,
      "num_tokens": 51319366.0,
      "step": 22405
    },
    {
      "entropy": 5.477718544006348,
      "epoch": 2.215302491103203,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045113942529871473,
      "loss": 5.0859,
      "mean_token_accuracy": 0.20231460630893708,
      "num_tokens": 51332445.0,
      "step": 22410
    },
    {
      "entropy": 5.42995319366455,
      "epoch": 2.215796757611704,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004511174649754343,
      "loss": 5.0418,
      "mean_token_accuracy": 0.20715775340795517,
      "num_tokens": 51342900.0,
      "step": 22415
    },
    {
      "entropy": 5.41667742729187,
      "epoch": 2.2162910241202054,
      "grad_norm": 1.125,
      "learning_rate": 0.000451095500319637,
      "loss": 5.1392,
      "mean_token_accuracy": 0.2025244042277336,
      "num_tokens": 51354437.0,
      "step": 22420
    },
    {
      "entropy": 5.424160051345825,
      "epoch": 2.216785290628707,
      "grad_norm": 1.078125,
      "learning_rate": 0.000451073531331863,
      "loss": 5.0326,
      "mean_token_accuracy": 0.20858192443847656,
      "num_tokens": 51364630.0,
      "step": 22425
    },
    {
      "entropy": 5.527617788314819,
      "epoch": 2.2172795571372084,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004510515580126528,
      "loss": 5.2078,
      "mean_token_accuracy": 0.19514022916555404,
      "num_tokens": 51376547.0,
      "step": 22430
    },
    {
      "entropy": 5.600302267074585,
      "epoch": 2.2177738236457096,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000451029580362547,
      "loss": 5.1648,
      "mean_token_accuracy": 0.20315398275852203,
      "num_tokens": 51388704.0,
      "step": 22435
    },
    {
      "entropy": 5.65431170463562,
      "epoch": 2.2182680901542113,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004510075983820862,
      "loss": 5.3464,
      "mean_token_accuracy": 0.1848840370774269,
      "num_tokens": 51400884.0,
      "step": 22440
    },
    {
      "entropy": 5.50028715133667,
      "epoch": 2.2187623566627126,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045098561207181114,
      "loss": 5.06,
      "mean_token_accuracy": 0.20976512879133224,
      "num_tokens": 51412946.0,
      "step": 22445
    },
    {
      "entropy": 5.4981241703033445,
      "epoch": 2.219256623171214,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004509636214322628,
      "loss": 5.2081,
      "mean_token_accuracy": 0.20081951171159745,
      "num_tokens": 51425860.0,
      "step": 22450
    },
    {
      "entropy": 5.51390233039856,
      "epoch": 2.219750889679715,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004509416264639821,
      "loss": 5.127,
      "mean_token_accuracy": 0.20241011530160904,
      "num_tokens": 51438812.0,
      "step": 22455
    },
    {
      "entropy": 5.5867030143737795,
      "epoch": 2.220245156188217,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004509196271675101,
      "loss": 5.2399,
      "mean_token_accuracy": 0.19625402688980104,
      "num_tokens": 51449881.0,
      "step": 22460
    },
    {
      "entropy": 5.484631109237671,
      "epoch": 2.220739422696718,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045089762354338784,
      "loss": 5.0769,
      "mean_token_accuracy": 0.2054135710000992,
      "num_tokens": 51460403.0,
      "step": 22465
    },
    {
      "entropy": 5.507558298110962,
      "epoch": 2.2212336892052194,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045087561559215683,
      "loss": 5.1563,
      "mean_token_accuracy": 0.19342798739671707,
      "num_tokens": 51471414.0,
      "step": 22470
    },
    {
      "entropy": 5.41144208908081,
      "epoch": 2.221727955713721,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045085360331435835,
      "loss": 5.0816,
      "mean_token_accuracy": 0.20704265236854552,
      "num_tokens": 51482855.0,
      "step": 22475
    },
    {
      "entropy": 5.598956155776977,
      "epoch": 2.2222222222222223,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000450831586710534,
      "loss": 5.213,
      "mean_token_accuracy": 0.1966195896267891,
      "num_tokens": 51495363.0,
      "step": 22480
    },
    {
      "entropy": 5.498496723175049,
      "epoch": 2.2227164887307236,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045080956578122535,
      "loss": 5.1111,
      "mean_token_accuracy": 0.2004269391298294,
      "num_tokens": 51506678.0,
      "step": 22485
    },
    {
      "entropy": 5.519071817398071,
      "epoch": 2.223210755239225,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045078754052697406,
      "loss": 5.176,
      "mean_token_accuracy": 0.19880853593349457,
      "num_tokens": 51518596.0,
      "step": 22490
    },
    {
      "entropy": 5.553566312789917,
      "epoch": 2.2237050217477266,
      "grad_norm": 1.625,
      "learning_rate": 0.00045076551094832206,
      "loss": 5.2656,
      "mean_token_accuracy": 0.18756288588047026,
      "num_tokens": 51531292.0,
      "step": 22495
    },
    {
      "entropy": 5.490634965896606,
      "epoch": 2.224199288256228,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045074347704581114,
      "loss": 5.0935,
      "mean_token_accuracy": 0.2068154737353325,
      "num_tokens": 51542507.0,
      "step": 22500
    },
    {
      "entropy": 5.494726228713989,
      "epoch": 2.224693554764729,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045072143881998345,
      "loss": 5.1149,
      "mean_token_accuracy": 0.20549058318138122,
      "num_tokens": 51553895.0,
      "step": 22505
    },
    {
      "entropy": 5.518519735336303,
      "epoch": 2.2251878212732303,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004506993962713812,
      "loss": 5.2088,
      "mean_token_accuracy": 0.19290893524885178,
      "num_tokens": 51566533.0,
      "step": 22510
    },
    {
      "entropy": 5.533729934692383,
      "epoch": 2.225682087781732,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045067734940054644,
      "loss": 5.1399,
      "mean_token_accuracy": 0.19883281290531157,
      "num_tokens": 51576308.0,
      "step": 22515
    },
    {
      "entropy": 5.480108737945557,
      "epoch": 2.2261763542902333,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045065529820802165,
      "loss": 5.0762,
      "mean_token_accuracy": 0.2030370369553566,
      "num_tokens": 51589844.0,
      "step": 22520
    },
    {
      "entropy": 5.4806859493255615,
      "epoch": 2.2266706207987346,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045063324269434924,
      "loss": 5.0694,
      "mean_token_accuracy": 0.20250414311885834,
      "num_tokens": 51601685.0,
      "step": 22525
    },
    {
      "entropy": 5.505438137054443,
      "epoch": 2.227164887307236,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004506111828600719,
      "loss": 5.1401,
      "mean_token_accuracy": 0.20779650211334227,
      "num_tokens": 51613258.0,
      "step": 22530
    },
    {
      "entropy": 5.585435056686402,
      "epoch": 2.2276591538157375,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004505891187057322,
      "loss": 5.1681,
      "mean_token_accuracy": 0.2065178394317627,
      "num_tokens": 51624088.0,
      "step": 22535
    },
    {
      "entropy": 5.498156404495239,
      "epoch": 2.228153420324239,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004505670502318729,
      "loss": 5.1006,
      "mean_token_accuracy": 0.20892043709754943,
      "num_tokens": 51636617.0,
      "step": 22540
    },
    {
      "entropy": 5.5169140815734865,
      "epoch": 2.22864768683274,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045054497743903687,
      "loss": 5.1876,
      "mean_token_accuracy": 0.1927280232310295,
      "num_tokens": 51647777.0,
      "step": 22545
    },
    {
      "entropy": 5.510875511169433,
      "epoch": 2.2291419533412418,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004505229003277673,
      "loss": 5.201,
      "mean_token_accuracy": 0.18906791061162947,
      "num_tokens": 51660505.0,
      "step": 22550
    },
    {
      "entropy": 5.574018764495849,
      "epoch": 2.229636219849743,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045050081889860713,
      "loss": 5.1678,
      "mean_token_accuracy": 0.2038664773106575,
      "num_tokens": 51672296.0,
      "step": 22555
    },
    {
      "entropy": 5.482110118865966,
      "epoch": 2.2301304863582443,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045047873315209945,
      "loss": 5.145,
      "mean_token_accuracy": 0.2070688009262085,
      "num_tokens": 51684437.0,
      "step": 22560
    },
    {
      "entropy": 5.428293085098266,
      "epoch": 2.2306247528667456,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004504566430887878,
      "loss": 4.9732,
      "mean_token_accuracy": 0.21568287312984466,
      "num_tokens": 51694919.0,
      "step": 22565
    },
    {
      "entropy": 5.4811194896697994,
      "epoch": 2.2311190193752473,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045043454870921553,
      "loss": 5.1101,
      "mean_token_accuracy": 0.20445645451545716,
      "num_tokens": 51705273.0,
      "step": 22570
    },
    {
      "entropy": 5.53201994895935,
      "epoch": 2.2316132858837485,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004504124500139261,
      "loss": 5.1519,
      "mean_token_accuracy": 0.19600190222263336,
      "num_tokens": 51717048.0,
      "step": 22575
    },
    {
      "entropy": 5.5662352561950685,
      "epoch": 2.23210755239225,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004503903470034632,
      "loss": 5.2155,
      "mean_token_accuracy": 0.1910916820168495,
      "num_tokens": 51729208.0,
      "step": 22580
    },
    {
      "entropy": 5.4474719047546385,
      "epoch": 2.2326018189007515,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045036823967837066,
      "loss": 5.0749,
      "mean_token_accuracy": 0.21341143548488617,
      "num_tokens": 51740393.0,
      "step": 22585
    },
    {
      "entropy": 5.535370540618897,
      "epoch": 2.2330960854092528,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004503461280391921,
      "loss": 5.1325,
      "mean_token_accuracy": 0.20634863674640655,
      "num_tokens": 51751044.0,
      "step": 22590
    },
    {
      "entropy": 5.5139857769012455,
      "epoch": 2.233590351917754,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004503240120864716,
      "loss": 5.1398,
      "mean_token_accuracy": 0.19938633292913438,
      "num_tokens": 51761969.0,
      "step": 22595
    },
    {
      "entropy": 5.510321378707886,
      "epoch": 2.2340846184262553,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004503018918207532,
      "loss": 5.1617,
      "mean_token_accuracy": 0.19652422815561293,
      "num_tokens": 51774185.0,
      "step": 22600
    },
    {
      "entropy": 5.55922384262085,
      "epoch": 2.234578884934757,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004502797672425811,
      "loss": 5.1112,
      "mean_token_accuracy": 0.20445195883512496,
      "num_tokens": 51786155.0,
      "step": 22605
    },
    {
      "entropy": 5.472774600982666,
      "epoch": 2.2350731514432582,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045025763835249957,
      "loss": 5.1444,
      "mean_token_accuracy": 0.20036866068840026,
      "num_tokens": 51797125.0,
      "step": 22610
    },
    {
      "entropy": 5.555926036834717,
      "epoch": 2.2355674179517595,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004502355051510528,
      "loss": 5.2559,
      "mean_token_accuracy": 0.19540160447359084,
      "num_tokens": 51809046.0,
      "step": 22615
    },
    {
      "entropy": 5.586334991455078,
      "epoch": 2.2360616844602608,
      "grad_norm": 1.125,
      "learning_rate": 0.00045021336763878555,
      "loss": 5.2851,
      "mean_token_accuracy": 0.1851371705532074,
      "num_tokens": 51821857.0,
      "step": 22620
    },
    {
      "entropy": 5.598964262008667,
      "epoch": 2.2365559509687625,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00045019122581624224,
      "loss": 5.1011,
      "mean_token_accuracy": 0.20264934599399567,
      "num_tokens": 51833504.0,
      "step": 22625
    },
    {
      "entropy": 5.407631254196167,
      "epoch": 2.2370502174772637,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045016907968396764,
      "loss": 5.1649,
      "mean_token_accuracy": 0.20398636013269425,
      "num_tokens": 51844166.0,
      "step": 22630
    },
    {
      "entropy": 5.5200096607208256,
      "epoch": 2.237544483985765,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004501469292425064,
      "loss": 5.1457,
      "mean_token_accuracy": 0.20060763955116273,
      "num_tokens": 51855674.0,
      "step": 22635
    },
    {
      "entropy": 5.474892950057983,
      "epoch": 2.2380387504942667,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045012477449240363,
      "loss": 5.1089,
      "mean_token_accuracy": 0.204365636408329,
      "num_tokens": 51866128.0,
      "step": 22640
    },
    {
      "entropy": 5.49126558303833,
      "epoch": 2.238533017002768,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045010261543420416,
      "loss": 5.2041,
      "mean_token_accuracy": 0.2029258891940117,
      "num_tokens": 51877036.0,
      "step": 22645
    },
    {
      "entropy": 5.556707668304443,
      "epoch": 2.2390272835112692,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004500804520684532,
      "loss": 5.1224,
      "mean_token_accuracy": 0.2015409290790558,
      "num_tokens": 51888331.0,
      "step": 22650
    },
    {
      "entropy": 5.545964145660401,
      "epoch": 2.2395215500197705,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045005828439569594,
      "loss": 5.1574,
      "mean_token_accuracy": 0.19484515339136124,
      "num_tokens": 51899843.0,
      "step": 22655
    },
    {
      "entropy": 5.576833772659302,
      "epoch": 2.240015816528272,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045003611241647783,
      "loss": 5.186,
      "mean_token_accuracy": 0.19256659895181655,
      "num_tokens": 51911429.0,
      "step": 22660
    },
    {
      "entropy": 5.4972223281860355,
      "epoch": 2.2405100830367735,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045001393613134413,
      "loss": 5.1267,
      "mean_token_accuracy": 0.19981478601694108,
      "num_tokens": 51923154.0,
      "step": 22665
    },
    {
      "entropy": 5.520114755630493,
      "epoch": 2.2410043495452747,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004499917555408405,
      "loss": 5.1595,
      "mean_token_accuracy": 0.1968149498105049,
      "num_tokens": 51935126.0,
      "step": 22670
    },
    {
      "entropy": 5.537693977355957,
      "epoch": 2.241498616053776,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00044996957064551244,
      "loss": 5.1756,
      "mean_token_accuracy": 0.19135907888412476,
      "num_tokens": 51947629.0,
      "step": 22675
    },
    {
      "entropy": 5.519244766235351,
      "epoch": 2.2419928825622777,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004499473814459059,
      "loss": 5.166,
      "mean_token_accuracy": 0.19891617000102996,
      "num_tokens": 51957631.0,
      "step": 22680
    },
    {
      "entropy": 5.5661708354949955,
      "epoch": 2.242487149070779,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044992518794256657,
      "loss": 5.184,
      "mean_token_accuracy": 0.19683127254247665,
      "num_tokens": 51969854.0,
      "step": 22685
    },
    {
      "entropy": 5.561238622665405,
      "epoch": 2.24298141557928,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044990299013604045,
      "loss": 5.1745,
      "mean_token_accuracy": 0.19718017429113388,
      "num_tokens": 51980501.0,
      "step": 22690
    },
    {
      "entropy": 5.512120389938355,
      "epoch": 2.243475682087782,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044988078802687373,
      "loss": 5.1536,
      "mean_token_accuracy": 0.19805895984172822,
      "num_tokens": 51991336.0,
      "step": 22695
    },
    {
      "entropy": 5.533081817626953,
      "epoch": 2.243969948596283,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004498585816156124,
      "loss": 5.2178,
      "mean_token_accuracy": 0.19303175657987595,
      "num_tokens": 52003215.0,
      "step": 22700
    },
    {
      "entropy": 5.587298107147217,
      "epoch": 2.2444642151047844,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004498363709028029,
      "loss": 5.1825,
      "mean_token_accuracy": 0.19353781193494796,
      "num_tokens": 52014772.0,
      "step": 22705
    },
    {
      "entropy": 5.525957250595093,
      "epoch": 2.2449584816132857,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044981415588899153,
      "loss": 5.1658,
      "mean_token_accuracy": 0.19591038823127746,
      "num_tokens": 52025808.0,
      "step": 22710
    },
    {
      "entropy": 5.512095022201538,
      "epoch": 2.2454527481217874,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004497919365747248,
      "loss": 5.1987,
      "mean_token_accuracy": 0.19656058847904206,
      "num_tokens": 52037397.0,
      "step": 22715
    },
    {
      "entropy": 5.4825758934021,
      "epoch": 2.2459470146302887,
      "grad_norm": 1.484375,
      "learning_rate": 0.00044976971296054936,
      "loss": 5.0948,
      "mean_token_accuracy": 0.2112068608403206,
      "num_tokens": 52048874.0,
      "step": 22720
    },
    {
      "entropy": 5.514915657043457,
      "epoch": 2.24644128113879,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004497474850470118,
      "loss": 5.228,
      "mean_token_accuracy": 0.19076688587665558,
      "num_tokens": 52061180.0,
      "step": 22725
    },
    {
      "entropy": 5.537062120437622,
      "epoch": 2.2469355476472916,
      "grad_norm": 1.3125,
      "learning_rate": 0.000449725252834659,
      "loss": 5.1209,
      "mean_token_accuracy": 0.19823426753282547,
      "num_tokens": 52073400.0,
      "step": 22730
    },
    {
      "entropy": 5.515083885192871,
      "epoch": 2.247429814155793,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044970301632403796,
      "loss": 5.1852,
      "mean_token_accuracy": 0.20362466424703599,
      "num_tokens": 52084737.0,
      "step": 22735
    },
    {
      "entropy": 5.4838320255279545,
      "epoch": 2.247924080664294,
      "grad_norm": 1.125,
      "learning_rate": 0.00044968077551569554,
      "loss": 5.1114,
      "mean_token_accuracy": 0.2066636100411415,
      "num_tokens": 52096044.0,
      "step": 22740
    },
    {
      "entropy": 5.577515268325806,
      "epoch": 2.2484183471727954,
      "grad_norm": 1.078125,
      "learning_rate": 0.000449658530410179,
      "loss": 5.1216,
      "mean_token_accuracy": 0.20124562680721284,
      "num_tokens": 52108701.0,
      "step": 22745
    },
    {
      "entropy": 5.523634719848633,
      "epoch": 2.248912613681297,
      "grad_norm": 1.125,
      "learning_rate": 0.0004496362810080354,
      "loss": 5.1604,
      "mean_token_accuracy": 0.20224442183971406,
      "num_tokens": 52119797.0,
      "step": 22750
    },
    {
      "entropy": 5.494365501403808,
      "epoch": 2.2494068801897984,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004496140273098123,
      "loss": 5.2043,
      "mean_token_accuracy": 0.19335196912288666,
      "num_tokens": 52132557.0,
      "step": 22755
    },
    {
      "entropy": 5.453974819183349,
      "epoch": 2.2499011466982997,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044959176931605704,
      "loss": 5.0079,
      "mean_token_accuracy": 0.20661089271306993,
      "num_tokens": 52144058.0,
      "step": 22760
    },
    {
      "entropy": 5.515575885772705,
      "epoch": 2.250395413206801,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044956950702731714,
      "loss": 5.1447,
      "mean_token_accuracy": 0.20024342834949493,
      "num_tokens": 52155432.0,
      "step": 22765
    },
    {
      "entropy": 5.466350889205932,
      "epoch": 2.2508896797153026,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004495472404441403,
      "loss": 5.1019,
      "mean_token_accuracy": 0.20359287410974503,
      "num_tokens": 52166925.0,
      "step": 22770
    },
    {
      "entropy": 5.418187522888184,
      "epoch": 2.251383946223804,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004495249695670742,
      "loss": 5.1046,
      "mean_token_accuracy": 0.19748883694410324,
      "num_tokens": 52177452.0,
      "step": 22775
    },
    {
      "entropy": 5.524319791793824,
      "epoch": 2.251878212732305,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004495026943966669,
      "loss": 5.1067,
      "mean_token_accuracy": 0.19903458505868912,
      "num_tokens": 52189288.0,
      "step": 22780
    },
    {
      "entropy": 5.449623775482178,
      "epoch": 2.2523724792408064,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044948041493346615,
      "loss": 5.0824,
      "mean_token_accuracy": 0.208235402405262,
      "num_tokens": 52199963.0,
      "step": 22785
    },
    {
      "entropy": 5.447674798965454,
      "epoch": 2.252866745749308,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004494581311780202,
      "loss": 5.1508,
      "mean_token_accuracy": 0.197057144343853,
      "num_tokens": 52210727.0,
      "step": 22790
    },
    {
      "entropy": 5.535758781433105,
      "epoch": 2.2533610122578094,
      "grad_norm": 1.28125,
      "learning_rate": 0.00044943584313087705,
      "loss": 5.1514,
      "mean_token_accuracy": 0.1999640628695488,
      "num_tokens": 52222211.0,
      "step": 22795
    },
    {
      "entropy": 5.530038452148437,
      "epoch": 2.2538552787663106,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044941355079258517,
      "loss": 5.1971,
      "mean_token_accuracy": 0.1978902041912079,
      "num_tokens": 52234453.0,
      "step": 22800
    },
    {
      "entropy": 5.458242797851563,
      "epoch": 2.2543495452748123,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004493912541636928,
      "loss": 5.1292,
      "mean_token_accuracy": 0.2034463718533516,
      "num_tokens": 52246324.0,
      "step": 22805
    },
    {
      "entropy": 5.5805047988891605,
      "epoch": 2.2548438117833136,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004493689532447486,
      "loss": 5.1982,
      "mean_token_accuracy": 0.19785178899765016,
      "num_tokens": 52258223.0,
      "step": 22810
    },
    {
      "entropy": 5.531656694412232,
      "epoch": 2.255338078291815,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004493466480363011,
      "loss": 5.1814,
      "mean_token_accuracy": 0.2084525614976883,
      "num_tokens": 52270066.0,
      "step": 22815
    },
    {
      "entropy": 5.599063205718994,
      "epoch": 2.255832344800316,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044932433853889903,
      "loss": 5.2021,
      "mean_token_accuracy": 0.19684071838855743,
      "num_tokens": 52281366.0,
      "step": 22820
    },
    {
      "entropy": 5.526471471786499,
      "epoch": 2.256326611308818,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004493020247530912,
      "loss": 5.1876,
      "mean_token_accuracy": 0.1980804532766342,
      "num_tokens": 52292109.0,
      "step": 22825
    },
    {
      "entropy": 5.4550696849823,
      "epoch": 2.256820877817319,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004492797066794264,
      "loss": 5.0005,
      "mean_token_accuracy": 0.21735651195049285,
      "num_tokens": 52302707.0,
      "step": 22830
    },
    {
      "entropy": 5.467882537841797,
      "epoch": 2.2573151443258204,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044925738431845385,
      "loss": 5.1137,
      "mean_token_accuracy": 0.20674028098583222,
      "num_tokens": 52313715.0,
      "step": 22835
    },
    {
      "entropy": 5.518535566329956,
      "epoch": 2.257809410834322,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044923505767072267,
      "loss": 5.1412,
      "mean_token_accuracy": 0.19324890226125718,
      "num_tokens": 52325612.0,
      "step": 22840
    },
    {
      "entropy": 5.39333930015564,
      "epoch": 2.2583036773428233,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044921272673678197,
      "loss": 5.0204,
      "mean_token_accuracy": 0.21226562708616256,
      "num_tokens": 52338233.0,
      "step": 22845
    },
    {
      "entropy": 5.438125228881836,
      "epoch": 2.2587979438513246,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004491903915171812,
      "loss": 5.0594,
      "mean_token_accuracy": 0.20817974358797073,
      "num_tokens": 52348792.0,
      "step": 22850
    },
    {
      "entropy": 5.519145822525024,
      "epoch": 2.259292210359826,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044916805201246975,
      "loss": 5.1888,
      "mean_token_accuracy": 0.19983006715774537,
      "num_tokens": 52359334.0,
      "step": 22855
    },
    {
      "entropy": 5.510724115371704,
      "epoch": 2.2597864768683276,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044914570822319723,
      "loss": 5.1675,
      "mean_token_accuracy": 0.19261475503444672,
      "num_tokens": 52370600.0,
      "step": 22860
    },
    {
      "entropy": 5.573188018798828,
      "epoch": 2.260280743376829,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004491233601499132,
      "loss": 5.1712,
      "mean_token_accuracy": 0.19635389149188995,
      "num_tokens": 52382783.0,
      "step": 22865
    },
    {
      "entropy": 5.471719980239868,
      "epoch": 2.26077500988533,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044910100779316757,
      "loss": 5.0286,
      "mean_token_accuracy": 0.21403787583112716,
      "num_tokens": 52393372.0,
      "step": 22870
    },
    {
      "entropy": 5.4192633628845215,
      "epoch": 2.261269276393832,
      "grad_norm": 1.125,
      "learning_rate": 0.0004490786511535102,
      "loss": 5.0947,
      "mean_token_accuracy": 0.20197679698467255,
      "num_tokens": 52405181.0,
      "step": 22875
    },
    {
      "entropy": 5.521694707870483,
      "epoch": 2.261763542902333,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044905629023149084,
      "loss": 5.1169,
      "mean_token_accuracy": 0.2001598671078682,
      "num_tokens": 52417371.0,
      "step": 22880
    },
    {
      "entropy": 5.537524890899658,
      "epoch": 2.2622578094108343,
      "grad_norm": 1.125,
      "learning_rate": 0.0004490339250276598,
      "loss": 5.1677,
      "mean_token_accuracy": 0.19754072427749633,
      "num_tokens": 52428712.0,
      "step": 22885
    },
    {
      "entropy": 5.4326623439788815,
      "epoch": 2.2627520759193356,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004490115555425672,
      "loss": 5.1899,
      "mean_token_accuracy": 0.1943548008799553,
      "num_tokens": 52439764.0,
      "step": 22890
    },
    {
      "entropy": 5.508354759216308,
      "epoch": 2.263246342427837,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004489891817767634,
      "loss": 5.1601,
      "mean_token_accuracy": 0.19721818566322327,
      "num_tokens": 52451409.0,
      "step": 22895
    },
    {
      "entropy": 5.537233161926269,
      "epoch": 2.2637406089363385,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044896680373079874,
      "loss": 5.1405,
      "mean_token_accuracy": 0.19666795134544374,
      "num_tokens": 52461471.0,
      "step": 22900
    },
    {
      "entropy": 5.492299890518188,
      "epoch": 2.26423487544484,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044894442140522365,
      "loss": 5.1408,
      "mean_token_accuracy": 0.19994232803583145,
      "num_tokens": 52472693.0,
      "step": 22905
    },
    {
      "entropy": 5.450560235977173,
      "epoch": 2.264729141953341,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004489220348005889,
      "loss": 5.0447,
      "mean_token_accuracy": 0.20792285203933716,
      "num_tokens": 52483449.0,
      "step": 22910
    },
    {
      "entropy": 5.461733865737915,
      "epoch": 2.265223408461843,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000448899643917445,
      "loss": 5.1567,
      "mean_token_accuracy": 0.2060289889574051,
      "num_tokens": 52494048.0,
      "step": 22915
    },
    {
      "entropy": 5.463822650909424,
      "epoch": 2.265717674970344,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004488772487563429,
      "loss": 5.1581,
      "mean_token_accuracy": 0.19794767946004868,
      "num_tokens": 52506224.0,
      "step": 22920
    },
    {
      "entropy": 5.541604137420654,
      "epoch": 2.2662119414788453,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004488548493178336,
      "loss": 5.1224,
      "mean_token_accuracy": 0.19185584932565689,
      "num_tokens": 52517157.0,
      "step": 22925
    },
    {
      "entropy": 5.5390482425689695,
      "epoch": 2.2667062079873466,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044883244560246793,
      "loss": 5.1923,
      "mean_token_accuracy": 0.1975897505879402,
      "num_tokens": 52528826.0,
      "step": 22930
    },
    {
      "entropy": 5.4939353466033936,
      "epoch": 2.2672004744958483,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004488100376107972,
      "loss": 5.0324,
      "mean_token_accuracy": 0.20322360545396806,
      "num_tokens": 52540068.0,
      "step": 22935
    },
    {
      "entropy": 5.419287586212159,
      "epoch": 2.2676947410043495,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004487876253433726,
      "loss": 5.0789,
      "mean_token_accuracy": 0.20609167367219924,
      "num_tokens": 52552611.0,
      "step": 22940
    },
    {
      "entropy": 5.447744369506836,
      "epoch": 2.268189007512851,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004487652088007454,
      "loss": 5.137,
      "mean_token_accuracy": 0.20004348158836366,
      "num_tokens": 52564532.0,
      "step": 22945
    },
    {
      "entropy": 5.553086376190185,
      "epoch": 2.2686832740213525,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004487427879834671,
      "loss": 5.1752,
      "mean_token_accuracy": 0.19216422885656356,
      "num_tokens": 52576110.0,
      "step": 22950
    },
    {
      "entropy": 5.503455114364624,
      "epoch": 2.2691775405298538,
      "grad_norm": 1.125,
      "learning_rate": 0.00044872036289208926,
      "loss": 5.1293,
      "mean_token_accuracy": 0.19535443782806397,
      "num_tokens": 52586890.0,
      "step": 22955
    },
    {
      "entropy": 5.500127077102661,
      "epoch": 2.269671807038355,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004486979335271635,
      "loss": 5.096,
      "mean_token_accuracy": 0.20473039001226426,
      "num_tokens": 52598597.0,
      "step": 22960
    },
    {
      "entropy": 5.487302589416504,
      "epoch": 2.2701660735468563,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044867549988924176,
      "loss": 5.1021,
      "mean_token_accuracy": 0.20413933098316192,
      "num_tokens": 52609361.0,
      "step": 22965
    },
    {
      "entropy": 5.564129877090454,
      "epoch": 2.270660340055358,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044865306197887573,
      "loss": 5.1871,
      "mean_token_accuracy": 0.19465567469596862,
      "num_tokens": 52620954.0,
      "step": 22970
    },
    {
      "entropy": 5.533117437362671,
      "epoch": 2.2711546065638593,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044863061979661745,
      "loss": 5.2166,
      "mean_token_accuracy": 0.19608749449253082,
      "num_tokens": 52631944.0,
      "step": 22975
    },
    {
      "entropy": 5.546460723876953,
      "epoch": 2.2716488730723605,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004486081733430189,
      "loss": 5.1554,
      "mean_token_accuracy": 0.19546399414539337,
      "num_tokens": 52642517.0,
      "step": 22980
    },
    {
      "entropy": 5.534183692932129,
      "epoch": 2.2721431395808622,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00044858572261863245,
      "loss": 5.2213,
      "mean_token_accuracy": 0.1979176253080368,
      "num_tokens": 52653654.0,
      "step": 22985
    },
    {
      "entropy": 5.535181665420533,
      "epoch": 2.2726374060893635,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004485632676240102,
      "loss": 5.1827,
      "mean_token_accuracy": 0.1969092771410942,
      "num_tokens": 52666068.0,
      "step": 22990
    },
    {
      "entropy": 5.525499296188355,
      "epoch": 2.2731316725978647,
      "grad_norm": 1.25,
      "learning_rate": 0.0004485408083597047,
      "loss": 5.1275,
      "mean_token_accuracy": 0.1998702585697174,
      "num_tokens": 52677797.0,
      "step": 22995
    },
    {
      "entropy": 5.495450735092163,
      "epoch": 2.273625939106366,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004485183448262684,
      "loss": 5.0917,
      "mean_token_accuracy": 0.20326270908117294,
      "num_tokens": 52688947.0,
      "step": 23000
    },
    {
      "entropy": 5.468483924865723,
      "epoch": 2.2741202056148677,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004484958770242538,
      "loss": 5.1516,
      "mean_token_accuracy": 0.19325748682022095,
      "num_tokens": 52700357.0,
      "step": 23005
    },
    {
      "entropy": 5.555229949951172,
      "epoch": 2.274614472123369,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044847340495421384,
      "loss": 5.2149,
      "mean_token_accuracy": 0.19237735420465468,
      "num_tokens": 52712602.0,
      "step": 23010
    },
    {
      "entropy": 5.545125246047974,
      "epoch": 2.2751087386318702,
      "grad_norm": 1.125,
      "learning_rate": 0.00044845092861670114,
      "loss": 5.117,
      "mean_token_accuracy": 0.20498985499143602,
      "num_tokens": 52723480.0,
      "step": 23015
    },
    {
      "entropy": 5.464772462844849,
      "epoch": 2.2756030051403715,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004484284480122688,
      "loss": 5.116,
      "mean_token_accuracy": 0.19864934980869292,
      "num_tokens": 52735418.0,
      "step": 23020
    },
    {
      "entropy": 5.438072538375854,
      "epoch": 2.276097271648873,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044840596314146955,
      "loss": 5.0741,
      "mean_token_accuracy": 0.20668940246105194,
      "num_tokens": 52746308.0,
      "step": 23025
    },
    {
      "entropy": 5.442927551269531,
      "epoch": 2.2765915381573745,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004483834740048568,
      "loss": 5.0099,
      "mean_token_accuracy": 0.21110846549272538,
      "num_tokens": 52756834.0,
      "step": 23030
    },
    {
      "entropy": 5.47628870010376,
      "epoch": 2.2770858046658757,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004483609806029837,
      "loss": 5.1298,
      "mean_token_accuracy": 0.19860559701919556,
      "num_tokens": 52768398.0,
      "step": 23035
    },
    {
      "entropy": 5.54872145652771,
      "epoch": 2.277580071174377,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044833848293640354,
      "loss": 5.2009,
      "mean_token_accuracy": 0.1981884017586708,
      "num_tokens": 52780062.0,
      "step": 23040
    },
    {
      "entropy": 5.513368606567383,
      "epoch": 2.2780743376828787,
      "grad_norm": 1.4375,
      "learning_rate": 0.00044831598100566984,
      "loss": 5.09,
      "mean_token_accuracy": 0.20428134500980377,
      "num_tokens": 52791760.0,
      "step": 23045
    },
    {
      "entropy": 5.465321445465088,
      "epoch": 2.27856860419138,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044829347481133613,
      "loss": 5.036,
      "mean_token_accuracy": 0.20732480138540268,
      "num_tokens": 52802638.0,
      "step": 23050
    },
    {
      "entropy": 5.427728366851807,
      "epoch": 2.2790628706998812,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000448270964353956,
      "loss": 5.1121,
      "mean_token_accuracy": 0.20298752635717393,
      "num_tokens": 52813590.0,
      "step": 23055
    },
    {
      "entropy": 5.48470025062561,
      "epoch": 2.279557137208383,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004482484496340833,
      "loss": 5.1594,
      "mean_token_accuracy": 0.2083876460790634,
      "num_tokens": 52824177.0,
      "step": 23060
    },
    {
      "entropy": 5.59813232421875,
      "epoch": 2.280051403716884,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004482259306522719,
      "loss": 5.2527,
      "mean_token_accuracy": 0.19162196964025496,
      "num_tokens": 52834637.0,
      "step": 23065
    },
    {
      "entropy": 5.567384958267212,
      "epoch": 2.2805456702253855,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044820340740907565,
      "loss": 5.1801,
      "mean_token_accuracy": 0.1954122707247734,
      "num_tokens": 52847893.0,
      "step": 23070
    },
    {
      "entropy": 5.489695930480957,
      "epoch": 2.2810399367338867,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004481808799050488,
      "loss": 5.0891,
      "mean_token_accuracy": 0.2005574718117714,
      "num_tokens": 52859339.0,
      "step": 23075
    },
    {
      "entropy": 5.5077661037445065,
      "epoch": 2.2815342032423884,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044815834814074544,
      "loss": 5.1531,
      "mean_token_accuracy": 0.19729001671075821,
      "num_tokens": 52872405.0,
      "step": 23080
    },
    {
      "entropy": 5.452637004852295,
      "epoch": 2.2820284697508897,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004481358121167198,
      "loss": 5.0118,
      "mean_token_accuracy": 0.2136598214507103,
      "num_tokens": 52882187.0,
      "step": 23085
    },
    {
      "entropy": 5.449151563644409,
      "epoch": 2.282522736259391,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044811327183352627,
      "loss": 5.0662,
      "mean_token_accuracy": 0.2134748011827469,
      "num_tokens": 52894004.0,
      "step": 23090
    },
    {
      "entropy": 5.593410968780518,
      "epoch": 2.2830170027678927,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004480907272917195,
      "loss": 5.181,
      "mean_token_accuracy": 0.19056494385004044,
      "num_tokens": 52904749.0,
      "step": 23095
    },
    {
      "entropy": 5.482621431350708,
      "epoch": 2.283511269276394,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004480681784918539,
      "loss": 5.1273,
      "mean_token_accuracy": 0.19679268300533295,
      "num_tokens": 52915794.0,
      "step": 23100
    },
    {
      "entropy": 5.47328667640686,
      "epoch": 2.284005535784895,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044804562543448436,
      "loss": 5.0791,
      "mean_token_accuracy": 0.20068948417901994,
      "num_tokens": 52926695.0,
      "step": 23105
    },
    {
      "entropy": 5.484558534622193,
      "epoch": 2.2844998022933964,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004480230681201655,
      "loss": 5.0685,
      "mean_token_accuracy": 0.20456236600875854,
      "num_tokens": 52937824.0,
      "step": 23110
    },
    {
      "entropy": 5.471761465072632,
      "epoch": 2.284994068801898,
      "grad_norm": 1.125,
      "learning_rate": 0.00044800050654945233,
      "loss": 5.1097,
      "mean_token_accuracy": 0.20041943937540055,
      "num_tokens": 52949174.0,
      "step": 23115
    },
    {
      "entropy": 5.470703506469727,
      "epoch": 2.2854883353103994,
      "grad_norm": 1.3125,
      "learning_rate": 0.00044797794072289987,
      "loss": 5.1861,
      "mean_token_accuracy": 0.20034543722867965,
      "num_tokens": 52960675.0,
      "step": 23120
    },
    {
      "entropy": 5.4820891380310055,
      "epoch": 2.2859826018189007,
      "grad_norm": 1.3125,
      "learning_rate": 0.00044795537064106335,
      "loss": 5.1212,
      "mean_token_accuracy": 0.20402951240539552,
      "num_tokens": 52971705.0,
      "step": 23125
    },
    {
      "entropy": 5.525818157196045,
      "epoch": 2.2864768683274024,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044793279630449777,
      "loss": 5.1499,
      "mean_token_accuracy": 0.19926250576972962,
      "num_tokens": 52983025.0,
      "step": 23130
    },
    {
      "entropy": 5.524250793457031,
      "epoch": 2.2869711348359036,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004479102177137586,
      "loss": 5.131,
      "mean_token_accuracy": 0.19605901688337327,
      "num_tokens": 52993810.0,
      "step": 23135
    },
    {
      "entropy": 5.489673519134522,
      "epoch": 2.287465401344405,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004478876348694012,
      "loss": 5.1244,
      "mean_token_accuracy": 0.19711627215147018,
      "num_tokens": 53005902.0,
      "step": 23140
    },
    {
      "entropy": 5.5297948837280275,
      "epoch": 2.287959667852906,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004478650477719812,
      "loss": 5.1486,
      "mean_token_accuracy": 0.19510836154222488,
      "num_tokens": 53017511.0,
      "step": 23145
    },
    {
      "entropy": 5.435140991210938,
      "epoch": 2.2884539343614074,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004478424564220542,
      "loss": 5.0261,
      "mean_token_accuracy": 0.2113171711564064,
      "num_tokens": 53028514.0,
      "step": 23150
    },
    {
      "entropy": 5.4668066024780275,
      "epoch": 2.288948200869909,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044781986082017595,
      "loss": 5.0937,
      "mean_token_accuracy": 0.20100184082984923,
      "num_tokens": 53039338.0,
      "step": 23155
    },
    {
      "entropy": 5.510946655273438,
      "epoch": 2.2894424673784104,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044779726096690227,
      "loss": 5.0705,
      "mean_token_accuracy": 0.2000217169523239,
      "num_tokens": 53050877.0,
      "step": 23160
    },
    {
      "entropy": 5.512349271774292,
      "epoch": 2.2899367338869117,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004477746568627892,
      "loss": 5.128,
      "mean_token_accuracy": 0.19688339978456498,
      "num_tokens": 53062819.0,
      "step": 23165
    },
    {
      "entropy": 5.520561408996582,
      "epoch": 2.2904310003954134,
      "grad_norm": 1.28125,
      "learning_rate": 0.00044775204850839277,
      "loss": 5.1851,
      "mean_token_accuracy": 0.20069462358951567,
      "num_tokens": 53073486.0,
      "step": 23170
    },
    {
      "entropy": 5.466833448410034,
      "epoch": 2.2909252669039146,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004477294359042692,
      "loss": 5.0868,
      "mean_token_accuracy": 0.20201147049665452,
      "num_tokens": 53084499.0,
      "step": 23175
    },
    {
      "entropy": 5.455143833160401,
      "epoch": 2.291419533412416,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004477068190509747,
      "loss": 5.1603,
      "mean_token_accuracy": 0.1973249763250351,
      "num_tokens": 53097181.0,
      "step": 23180
    },
    {
      "entropy": 5.588787698745728,
      "epoch": 2.291913799920917,
      "grad_norm": 1.3125,
      "learning_rate": 0.00044768419794906556,
      "loss": 5.2094,
      "mean_token_accuracy": 0.20154581665992738,
      "num_tokens": 53108586.0,
      "step": 23185
    },
    {
      "entropy": 5.55732741355896,
      "epoch": 2.292408066429419,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044766157259909853,
      "loss": 5.0929,
      "mean_token_accuracy": 0.19937171489000322,
      "num_tokens": 53119481.0,
      "step": 23190
    },
    {
      "entropy": 5.515818977355957,
      "epoch": 2.29290233293792,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004476389430016299,
      "loss": 5.0843,
      "mean_token_accuracy": 0.20422664731740953,
      "num_tokens": 53129482.0,
      "step": 23195
    },
    {
      "entropy": 5.4313812255859375,
      "epoch": 2.2933965994464214,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044761630915721645,
      "loss": 5.1087,
      "mean_token_accuracy": 0.20126821249723434,
      "num_tokens": 53142311.0,
      "step": 23200
    },
    {
      "entropy": 5.453479480743408,
      "epoch": 2.293890865954923,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004475936710664151,
      "loss": 5.0825,
      "mean_token_accuracy": 0.1988782837986946,
      "num_tokens": 53153485.0,
      "step": 23205
    },
    {
      "entropy": 5.480115556716919,
      "epoch": 2.2943851324634243,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004475710287297826,
      "loss": 5.1159,
      "mean_token_accuracy": 0.2007051572203636,
      "num_tokens": 53164121.0,
      "step": 23210
    },
    {
      "entropy": 5.506258010864258,
      "epoch": 2.2948793989719256,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004475483821478761,
      "loss": 5.1076,
      "mean_token_accuracy": 0.19337636530399321,
      "num_tokens": 53174949.0,
      "step": 23215
    },
    {
      "entropy": 5.462037229537964,
      "epoch": 2.295373665480427,
      "grad_norm": 1.125,
      "learning_rate": 0.00044752573132125263,
      "loss": 5.0712,
      "mean_token_accuracy": 0.2067439004778862,
      "num_tokens": 53186762.0,
      "step": 23220
    },
    {
      "entropy": 5.4800573825836185,
      "epoch": 2.2958679319889286,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004475030762504694,
      "loss": 5.1167,
      "mean_token_accuracy": 0.20075072348117828,
      "num_tokens": 53198261.0,
      "step": 23225
    },
    {
      "entropy": 5.4227396011352536,
      "epoch": 2.29636219849743,
      "grad_norm": 1.125,
      "learning_rate": 0.0004474804169360836,
      "loss": 5.0288,
      "mean_token_accuracy": 0.20761310756206514,
      "num_tokens": 53209143.0,
      "step": 23230
    },
    {
      "entropy": 5.5744444847106935,
      "epoch": 2.296856465005931,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044745775337865293,
      "loss": 5.2812,
      "mean_token_accuracy": 0.1950062021613121,
      "num_tokens": 53222619.0,
      "step": 23235
    },
    {
      "entropy": 5.494481086730957,
      "epoch": 2.297350731514433,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044743508557873473,
      "loss": 5.0608,
      "mean_token_accuracy": 0.20697817504405974,
      "num_tokens": 53233829.0,
      "step": 23240
    },
    {
      "entropy": 5.536629629135132,
      "epoch": 2.297844998022934,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044741241353688665,
      "loss": 5.1472,
      "mean_token_accuracy": 0.19830814599990845,
      "num_tokens": 53244248.0,
      "step": 23245
    },
    {
      "entropy": 5.5566905498504635,
      "epoch": 2.2983392645314353,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004473897372536665,
      "loss": 5.1643,
      "mean_token_accuracy": 0.1976460799574852,
      "num_tokens": 53255638.0,
      "step": 23250
    },
    {
      "entropy": 5.483745908737182,
      "epoch": 2.2988335310399366,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000447367056729632,
      "loss": 5.1147,
      "mean_token_accuracy": 0.19571523219347,
      "num_tokens": 53265798.0,
      "step": 23255
    },
    {
      "entropy": 5.465081357955933,
      "epoch": 2.2993277975484383,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004473443719653412,
      "loss": 5.141,
      "mean_token_accuracy": 0.19944766014814377,
      "num_tokens": 53278001.0,
      "step": 23260
    },
    {
      "entropy": 5.589875316619873,
      "epoch": 2.2998220640569396,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044732168296135217,
      "loss": 5.1673,
      "mean_token_accuracy": 0.2008986458182335,
      "num_tokens": 53289742.0,
      "step": 23265
    },
    {
      "entropy": 5.528914499282837,
      "epoch": 2.300316330565441,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044729898971822296,
      "loss": 5.0955,
      "mean_token_accuracy": 0.20144775360822678,
      "num_tokens": 53301285.0,
      "step": 23270
    },
    {
      "entropy": 5.405538702011109,
      "epoch": 2.300810597073942,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004472762922365118,
      "loss": 5.0221,
      "mean_token_accuracy": 0.20915012955665588,
      "num_tokens": 53312883.0,
      "step": 23275
    },
    {
      "entropy": 5.51207013130188,
      "epoch": 2.301304863582444,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004472535905167773,
      "loss": 5.1753,
      "mean_token_accuracy": 0.195101198554039,
      "num_tokens": 53325215.0,
      "step": 23280
    },
    {
      "entropy": 5.562971067428589,
      "epoch": 2.301799130090945,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004472308845595776,
      "loss": 5.186,
      "mean_token_accuracy": 0.19648344069719315,
      "num_tokens": 53337104.0,
      "step": 23285
    },
    {
      "entropy": 5.499239587783814,
      "epoch": 2.3022933965994463,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004472081743654715,
      "loss": 5.1639,
      "mean_token_accuracy": 0.1996285080909729,
      "num_tokens": 53348097.0,
      "step": 23290
    },
    {
      "entropy": 5.487429618835449,
      "epoch": 2.3027876631079476,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044718545993501754,
      "loss": 5.0453,
      "mean_token_accuracy": 0.21309808343648912,
      "num_tokens": 53358405.0,
      "step": 23295
    },
    {
      "entropy": 5.486831521987915,
      "epoch": 2.3032819296164493,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004471627412687745,
      "loss": 5.1517,
      "mean_token_accuracy": 0.19874357283115388,
      "num_tokens": 53370012.0,
      "step": 23300
    },
    {
      "entropy": 5.5608631610870365,
      "epoch": 2.3037761961249505,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044714001836730145,
      "loss": 5.2464,
      "mean_token_accuracy": 0.19255422353744506,
      "num_tokens": 53383214.0,
      "step": 23305
    },
    {
      "entropy": 5.5540711879730225,
      "epoch": 2.304270462633452,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044711729123115716,
      "loss": 5.1807,
      "mean_token_accuracy": 0.19741398990154266,
      "num_tokens": 53395672.0,
      "step": 23310
    },
    {
      "entropy": 5.493957901000977,
      "epoch": 2.3047647291419535,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044709455986090084,
      "loss": 5.1103,
      "mean_token_accuracy": 0.20291218608617784,
      "num_tokens": 53407577.0,
      "step": 23315
    },
    {
      "entropy": 5.465126323699951,
      "epoch": 2.305258995650455,
      "grad_norm": 1.125,
      "learning_rate": 0.0004470718242570916,
      "loss": 5.0536,
      "mean_token_accuracy": 0.20453666001558304,
      "num_tokens": 53417772.0,
      "step": 23320
    },
    {
      "entropy": 5.504280376434326,
      "epoch": 2.305753262158956,
      "grad_norm": 1.125,
      "learning_rate": 0.0004470490844202889,
      "loss": 5.1383,
      "mean_token_accuracy": 0.1986495777964592,
      "num_tokens": 53429675.0,
      "step": 23325
    },
    {
      "entropy": 5.475494337081909,
      "epoch": 2.3062475286674573,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004470263403510519,
      "loss": 5.1488,
      "mean_token_accuracy": 0.19634604156017305,
      "num_tokens": 53442352.0,
      "step": 23330
    },
    {
      "entropy": 5.575124979019165,
      "epoch": 2.306741795175959,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044700359204994034,
      "loss": 5.297,
      "mean_token_accuracy": 0.19076389372348784,
      "num_tokens": 53453710.0,
      "step": 23335
    },
    {
      "entropy": 5.560633802413941,
      "epoch": 2.3072360616844603,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044698083951751365,
      "loss": 5.0994,
      "mean_token_accuracy": 0.21239494532346725,
      "num_tokens": 53464764.0,
      "step": 23340
    },
    {
      "entropy": 5.466793012619019,
      "epoch": 2.3077303281929615,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004469580827543318,
      "loss": 5.0604,
      "mean_token_accuracy": 0.20844493955373763,
      "num_tokens": 53476495.0,
      "step": 23345
    },
    {
      "entropy": 5.478285503387451,
      "epoch": 2.3082245947014632,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004469353217609542,
      "loss": 5.1509,
      "mean_token_accuracy": 0.19525968730449678,
      "num_tokens": 53488613.0,
      "step": 23350
    },
    {
      "entropy": 5.556738948822021,
      "epoch": 2.3087188612099645,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004469125565379411,
      "loss": 5.2414,
      "mean_token_accuracy": 0.1921929031610489,
      "num_tokens": 53501851.0,
      "step": 23355
    },
    {
      "entropy": 5.520310640335083,
      "epoch": 2.3092131277184658,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004468897870858525,
      "loss": 5.1446,
      "mean_token_accuracy": 0.19475848972797394,
      "num_tokens": 53514325.0,
      "step": 23360
    },
    {
      "entropy": 5.602771520614624,
      "epoch": 2.309707394226967,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044686701340524836,
      "loss": 5.1037,
      "mean_token_accuracy": 0.2014873206615448,
      "num_tokens": 53525717.0,
      "step": 23365
    },
    {
      "entropy": 5.486196565628052,
      "epoch": 2.3102016607354687,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004468442354966891,
      "loss": 5.1321,
      "mean_token_accuracy": 0.19803317934274672,
      "num_tokens": 53536488.0,
      "step": 23370
    },
    {
      "entropy": 5.55515341758728,
      "epoch": 2.31069592724397,
      "grad_norm": 1.125,
      "learning_rate": 0.000446821453360735,
      "loss": 5.2405,
      "mean_token_accuracy": 0.19185070991516112,
      "num_tokens": 53547230.0,
      "step": 23375
    },
    {
      "entropy": 5.568988847732544,
      "epoch": 2.3111901937524713,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044679866699794644,
      "loss": 5.243,
      "mean_token_accuracy": 0.19400170147418977,
      "num_tokens": 53558673.0,
      "step": 23380
    },
    {
      "entropy": 5.493989706039429,
      "epoch": 2.311684460260973,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004467758764088841,
      "loss": 5.0809,
      "mean_token_accuracy": 0.20522097051143645,
      "num_tokens": 53570088.0,
      "step": 23385
    },
    {
      "entropy": 5.450780439376831,
      "epoch": 2.3121787267694742,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004467530815941085,
      "loss": 5.1452,
      "mean_token_accuracy": 0.20699031352996827,
      "num_tokens": 53582115.0,
      "step": 23390
    },
    {
      "entropy": 5.466097831726074,
      "epoch": 2.3126729932779755,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00044673028255418047,
      "loss": 5.1537,
      "mean_token_accuracy": 0.1980408251285553,
      "num_tokens": 53594719.0,
      "step": 23395
    },
    {
      "entropy": 5.394616746902466,
      "epoch": 2.3131672597864767,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004467074792896608,
      "loss": 4.9782,
      "mean_token_accuracy": 0.21327086687088012,
      "num_tokens": 53606453.0,
      "step": 23400
    },
    {
      "entropy": 5.504375743865967,
      "epoch": 2.313661526294978,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044668467180111054,
      "loss": 5.1087,
      "mean_token_accuracy": 0.20375683009624482,
      "num_tokens": 53617038.0,
      "step": 23405
    },
    {
      "entropy": 5.565079164505005,
      "epoch": 2.3141557928034797,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044666186008909077,
      "loss": 5.1332,
      "mean_token_accuracy": 0.20720221996307372,
      "num_tokens": 53627230.0,
      "step": 23410
    },
    {
      "entropy": 5.537656307220459,
      "epoch": 2.314650059311981,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044663904415416245,
      "loss": 5.2022,
      "mean_token_accuracy": 0.19191043674945832,
      "num_tokens": 53638247.0,
      "step": 23415
    },
    {
      "entropy": 5.55374002456665,
      "epoch": 2.3151443258204822,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004466162239968871,
      "loss": 5.1032,
      "mean_token_accuracy": 0.20352677553892135,
      "num_tokens": 53648643.0,
      "step": 23420
    },
    {
      "entropy": 5.445544576644897,
      "epoch": 2.315638592328984,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004465933996178261,
      "loss": 5.0522,
      "mean_token_accuracy": 0.21596429944038392,
      "num_tokens": 53659715.0,
      "step": 23425
    },
    {
      "entropy": 5.412439489364624,
      "epoch": 2.316132858837485,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044657057101754073,
      "loss": 5.0942,
      "mean_token_accuracy": 0.20142505466938018,
      "num_tokens": 53671119.0,
      "step": 23430
    },
    {
      "entropy": 5.525711870193481,
      "epoch": 2.3166271253459865,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044654773819659277,
      "loss": 5.098,
      "mean_token_accuracy": 0.20247145742177963,
      "num_tokens": 53682145.0,
      "step": 23435
    },
    {
      "entropy": 5.512628793716431,
      "epoch": 2.3171213918544877,
      "grad_norm": 1.125,
      "learning_rate": 0.00044652490115554386,
      "loss": 5.0789,
      "mean_token_accuracy": 0.20571796894073485,
      "num_tokens": 53693882.0,
      "step": 23440
    },
    {
      "entropy": 5.491055965423584,
      "epoch": 2.3176156583629894,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004465020598949557,
      "loss": 5.1466,
      "mean_token_accuracy": 0.1952258065342903,
      "num_tokens": 53706010.0,
      "step": 23445
    },
    {
      "entropy": 5.506443643569947,
      "epoch": 2.3181099248714907,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004464792144153903,
      "loss": 5.1684,
      "mean_token_accuracy": 0.19902730435132981,
      "num_tokens": 53717657.0,
      "step": 23450
    },
    {
      "entropy": 5.496538877487183,
      "epoch": 2.318604191379992,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004464563647174096,
      "loss": 5.1168,
      "mean_token_accuracy": 0.20432027280330659,
      "num_tokens": 53729181.0,
      "step": 23455
    },
    {
      "entropy": 5.463153886795044,
      "epoch": 2.3190984578884937,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004464335108015757,
      "loss": 5.1124,
      "mean_token_accuracy": 0.20904105603694917,
      "num_tokens": 53740864.0,
      "step": 23460
    },
    {
      "entropy": 5.427383089065552,
      "epoch": 2.319592724396995,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004464106526684509,
      "loss": 5.0849,
      "mean_token_accuracy": 0.19871028363704682,
      "num_tokens": 53751726.0,
      "step": 23465
    },
    {
      "entropy": 5.548200511932373,
      "epoch": 2.320086990905496,
      "grad_norm": 1.125,
      "learning_rate": 0.0004463877903185974,
      "loss": 5.1608,
      "mean_token_accuracy": 0.20029897391796112,
      "num_tokens": 53763629.0,
      "step": 23470
    },
    {
      "entropy": 5.4977686405181885,
      "epoch": 2.3205812574139975,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004463649237525777,
      "loss": 5.1087,
      "mean_token_accuracy": 0.2064816877245903,
      "num_tokens": 53774814.0,
      "step": 23475
    },
    {
      "entropy": 5.508954572677612,
      "epoch": 2.321075523922499,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004463420529709543,
      "loss": 5.1188,
      "mean_token_accuracy": 0.20268789529800416,
      "num_tokens": 53786298.0,
      "step": 23480
    },
    {
      "entropy": 5.4936525344848635,
      "epoch": 2.3215697904310004,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044631917797428985,
      "loss": 5.1336,
      "mean_token_accuracy": 0.20196058750152587,
      "num_tokens": 53798509.0,
      "step": 23485
    },
    {
      "entropy": 5.538892650604248,
      "epoch": 2.3220640569395017,
      "grad_norm": 1.09375,
      "learning_rate": 0.000446296298763147,
      "loss": 5.2041,
      "mean_token_accuracy": 0.19422084987163543,
      "num_tokens": 53810689.0,
      "step": 23490
    },
    {
      "entropy": 5.552839946746826,
      "epoch": 2.3225583234480034,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004462734153380886,
      "loss": 5.167,
      "mean_token_accuracy": 0.19996706545352935,
      "num_tokens": 53822432.0,
      "step": 23495
    },
    {
      "entropy": 5.472344160079956,
      "epoch": 2.3230525899565047,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044625052769967766,
      "loss": 5.1063,
      "mean_token_accuracy": 0.20293254554271697,
      "num_tokens": 53833701.0,
      "step": 23500
    },
    {
      "entropy": 5.4395078182220455,
      "epoch": 2.323546856465006,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044622763584847716,
      "loss": 5.0618,
      "mean_token_accuracy": 0.20664691776037217,
      "num_tokens": 53844413.0,
      "step": 23505
    },
    {
      "entropy": 5.444320011138916,
      "epoch": 2.324041122973507,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00044620473978505023,
      "loss": 5.0284,
      "mean_token_accuracy": 0.20945281833410262,
      "num_tokens": 53856392.0,
      "step": 23510
    },
    {
      "entropy": 5.525319910049438,
      "epoch": 2.324535389482009,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044618183950996017,
      "loss": 5.1319,
      "mean_token_accuracy": 0.20306480377912522,
      "num_tokens": 53867938.0,
      "step": 23515
    },
    {
      "entropy": 5.41705470085144,
      "epoch": 2.32502965599051,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004461589350237702,
      "loss": 4.9996,
      "mean_token_accuracy": 0.21202801167964935,
      "num_tokens": 53878153.0,
      "step": 23520
    },
    {
      "entropy": 5.391370391845703,
      "epoch": 2.3255239224990114,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000446136026327044,
      "loss": 5.0413,
      "mean_token_accuracy": 0.20732551962137222,
      "num_tokens": 53889090.0,
      "step": 23525
    },
    {
      "entropy": 5.479256916046142,
      "epoch": 2.3260181890075127,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044611311342034494,
      "loss": 5.0974,
      "mean_token_accuracy": 0.20184465795755385,
      "num_tokens": 53900459.0,
      "step": 23530
    },
    {
      "entropy": 5.505434036254883,
      "epoch": 2.3265124555160144,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004460901963042368,
      "loss": 5.0879,
      "mean_token_accuracy": 0.2103040859103203,
      "num_tokens": 53912815.0,
      "step": 23535
    },
    {
      "entropy": 5.393052387237549,
      "epoch": 2.3270067220245156,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044606727497928324,
      "loss": 5.0335,
      "mean_token_accuracy": 0.2069545015692711,
      "num_tokens": 53924387.0,
      "step": 23540
    },
    {
      "entropy": 5.533142518997193,
      "epoch": 2.327500988533017,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044604434944604826,
      "loss": 5.1462,
      "mean_token_accuracy": 0.19943773448467256,
      "num_tokens": 53936147.0,
      "step": 23545
    },
    {
      "entropy": 5.487486839294434,
      "epoch": 2.327995255041518,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004460214197050956,
      "loss": 5.1502,
      "mean_token_accuracy": 0.20245511084794998,
      "num_tokens": 53948102.0,
      "step": 23550
    },
    {
      "entropy": 5.495618772506714,
      "epoch": 2.32848952155002,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004459984857569896,
      "loss": 5.16,
      "mean_token_accuracy": 0.1969030112028122,
      "num_tokens": 53959721.0,
      "step": 23555
    },
    {
      "entropy": 5.493838167190551,
      "epoch": 2.328983788058521,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044597554760229433,
      "loss": 5.1014,
      "mean_token_accuracy": 0.20428686290979386,
      "num_tokens": 53970302.0,
      "step": 23560
    },
    {
      "entropy": 5.56297116279602,
      "epoch": 2.3294780545670224,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000445952605241574,
      "loss": 5.2023,
      "mean_token_accuracy": 0.19855889081954955,
      "num_tokens": 53982198.0,
      "step": 23565
    },
    {
      "entropy": 5.57087082862854,
      "epoch": 2.329972321075524,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044592965867539306,
      "loss": 5.2452,
      "mean_token_accuracy": 0.1899275541305542,
      "num_tokens": 53994443.0,
      "step": 23570
    },
    {
      "entropy": 5.49559383392334,
      "epoch": 2.3304665875840254,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044590670790431604,
      "loss": 5.0914,
      "mean_token_accuracy": 0.1998365491628647,
      "num_tokens": 54005666.0,
      "step": 23575
    },
    {
      "entropy": 5.490523338317871,
      "epoch": 2.3309608540925266,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004458837529289074,
      "loss": 5.06,
      "mean_token_accuracy": 0.20578380674123764,
      "num_tokens": 54016173.0,
      "step": 23580
    },
    {
      "entropy": 5.459340524673462,
      "epoch": 2.331455120601028,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004458607937497321,
      "loss": 5.1898,
      "mean_token_accuracy": 0.1971178501844406,
      "num_tokens": 54027743.0,
      "step": 23585
    },
    {
      "entropy": 5.569420003890992,
      "epoch": 2.3319493871095296,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004458378303673546,
      "loss": 5.2056,
      "mean_token_accuracy": 0.19677744656801224,
      "num_tokens": 54039804.0,
      "step": 23590
    },
    {
      "entropy": 5.541219663619995,
      "epoch": 2.332443653618031,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004458148627823401,
      "loss": 5.1573,
      "mean_token_accuracy": 0.2018059328198433,
      "num_tokens": 54050202.0,
      "step": 23595
    },
    {
      "entropy": 5.513109111785889,
      "epoch": 2.332937920126532,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044579189099525334,
      "loss": 5.0865,
      "mean_token_accuracy": 0.20463416576385499,
      "num_tokens": 54061088.0,
      "step": 23600
    },
    {
      "entropy": 5.384110307693481,
      "epoch": 2.333432186635034,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044576891500665967,
      "loss": 5.0507,
      "mean_token_accuracy": 0.2078115776181221,
      "num_tokens": 54072524.0,
      "step": 23605
    },
    {
      "entropy": 5.491816186904908,
      "epoch": 2.333926453143535,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004457459348171242,
      "loss": 5.1273,
      "mean_token_accuracy": 0.20243655890226364,
      "num_tokens": 54083265.0,
      "step": 23610
    },
    {
      "entropy": 5.497485733032226,
      "epoch": 2.3344207196520363,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044572295042721215,
      "loss": 5.0544,
      "mean_token_accuracy": 0.21348097324371337,
      "num_tokens": 54093848.0,
      "step": 23615
    },
    {
      "entropy": 5.3828592777252195,
      "epoch": 2.3349149861605376,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044569996183748913,
      "loss": 5.0863,
      "mean_token_accuracy": 0.21013254076242446,
      "num_tokens": 54104595.0,
      "step": 23620
    },
    {
      "entropy": 5.390149164199829,
      "epoch": 2.3354092526690393,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044567696904852057,
      "loss": 4.9934,
      "mean_token_accuracy": 0.2143635034561157,
      "num_tokens": 54115577.0,
      "step": 23625
    },
    {
      "entropy": 5.453815269470215,
      "epoch": 2.3359035191775406,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000445653972060872,
      "loss": 5.0591,
      "mean_token_accuracy": 0.21048001050949097,
      "num_tokens": 54125579.0,
      "step": 23630
    },
    {
      "entropy": 5.481170225143432,
      "epoch": 2.336397785686042,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044563097087510934,
      "loss": 5.1833,
      "mean_token_accuracy": 0.1960402324795723,
      "num_tokens": 54137895.0,
      "step": 23635
    },
    {
      "entropy": 5.5386457443237305,
      "epoch": 2.3368920521945435,
      "grad_norm": 1.125,
      "learning_rate": 0.00044560796549179826,
      "loss": 5.1736,
      "mean_token_accuracy": 0.19894063472747803,
      "num_tokens": 54149423.0,
      "step": 23640
    },
    {
      "entropy": 5.584706878662109,
      "epoch": 2.337386318703045,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004455849559115048,
      "loss": 5.1806,
      "mean_token_accuracy": 0.19322724342346193,
      "num_tokens": 54160743.0,
      "step": 23645
    },
    {
      "entropy": 5.536889839172363,
      "epoch": 2.337880585211546,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004455619421347949,
      "loss": 5.179,
      "mean_token_accuracy": 0.2005209058523178,
      "num_tokens": 54171517.0,
      "step": 23650
    },
    {
      "entropy": 5.476043319702148,
      "epoch": 2.3383748517200473,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044553892416223485,
      "loss": 5.0418,
      "mean_token_accuracy": 0.20947880148887635,
      "num_tokens": 54182959.0,
      "step": 23655
    },
    {
      "entropy": 5.507992935180664,
      "epoch": 2.3388691182285486,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044551590199439075,
      "loss": 5.1434,
      "mean_token_accuracy": 0.2021603763103485,
      "num_tokens": 54194577.0,
      "step": 23660
    },
    {
      "entropy": 5.538587045669556,
      "epoch": 2.3393633847370503,
      "grad_norm": 1.0625,
      "learning_rate": 0.000445492875631829,
      "loss": 5.2715,
      "mean_token_accuracy": 0.18710128962993622,
      "num_tokens": 54207423.0,
      "step": 23665
    },
    {
      "entropy": 5.534144544601441,
      "epoch": 2.3398576512455516,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004454698450751161,
      "loss": 5.1083,
      "mean_token_accuracy": 0.2027324616909027,
      "num_tokens": 54218776.0,
      "step": 23670
    },
    {
      "entropy": 5.4754339218139645,
      "epoch": 2.340351917754053,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004454468103248186,
      "loss": 5.0326,
      "mean_token_accuracy": 0.21028360724449158,
      "num_tokens": 54229508.0,
      "step": 23675
    },
    {
      "entropy": 5.491382312774658,
      "epoch": 2.3408461842625545,
      "grad_norm": 1.046875,
      "learning_rate": 0.000445423771381503,
      "loss": 5.2049,
      "mean_token_accuracy": 0.19385700821876525,
      "num_tokens": 54240546.0,
      "step": 23680
    },
    {
      "entropy": 5.504705762863159,
      "epoch": 2.341340450771056,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004454007282457364,
      "loss": 5.1434,
      "mean_token_accuracy": 0.20375026017427444,
      "num_tokens": 54252486.0,
      "step": 23685
    },
    {
      "entropy": 5.573255968093872,
      "epoch": 2.341834717279557,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044537768091808516,
      "loss": 5.1426,
      "mean_token_accuracy": 0.19717117697000502,
      "num_tokens": 54264240.0,
      "step": 23690
    },
    {
      "entropy": 5.467287826538086,
      "epoch": 2.3423289837880583,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004453546293991168,
      "loss": 5.0862,
      "mean_token_accuracy": 0.2083623930811882,
      "num_tokens": 54275894.0,
      "step": 23695
    },
    {
      "entropy": 5.546685075759887,
      "epoch": 2.34282325029656,
      "grad_norm": 1.109375,
      "learning_rate": 0.000445331573689398,
      "loss": 5.1677,
      "mean_token_accuracy": 0.19838058948516846,
      "num_tokens": 54287697.0,
      "step": 23700
    },
    {
      "entropy": 5.477003860473633,
      "epoch": 2.3433175168050613,
      "grad_norm": 1.21875,
      "learning_rate": 0.000445308513789496,
      "loss": 5.052,
      "mean_token_accuracy": 0.21468464881181717,
      "num_tokens": 54300327.0,
      "step": 23705
    },
    {
      "entropy": 5.496341800689697,
      "epoch": 2.3438117833135625,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044528544969997823,
      "loss": 5.1978,
      "mean_token_accuracy": 0.19531141966581345,
      "num_tokens": 54311949.0,
      "step": 23710
    },
    {
      "entropy": 5.486055850982666,
      "epoch": 2.3443060498220643,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044526238142141196,
      "loss": 5.0846,
      "mean_token_accuracy": 0.20470067113637924,
      "num_tokens": 54322134.0,
      "step": 23715
    },
    {
      "entropy": 5.525734996795654,
      "epoch": 2.3448003163305655,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044523930895436475,
      "loss": 5.2081,
      "mean_token_accuracy": 0.1906544268131256,
      "num_tokens": 54334579.0,
      "step": 23720
    },
    {
      "entropy": 5.53402533531189,
      "epoch": 2.3452945828390668,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044521623229940407,
      "loss": 5.126,
      "mean_token_accuracy": 0.19673534482717514,
      "num_tokens": 54346624.0,
      "step": 23725
    },
    {
      "entropy": 5.449519395828247,
      "epoch": 2.345788849347568,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00044519315145709765,
      "loss": 5.0788,
      "mean_token_accuracy": 0.2061922326683998,
      "num_tokens": 54356455.0,
      "step": 23730
    },
    {
      "entropy": 5.579105615615845,
      "epoch": 2.3462831158560697,
      "grad_norm": 1.125,
      "learning_rate": 0.0004451700664280133,
      "loss": 5.2369,
      "mean_token_accuracy": 0.19721528589725495,
      "num_tokens": 54368141.0,
      "step": 23735
    },
    {
      "entropy": 5.509838581085205,
      "epoch": 2.346777382364571,
      "grad_norm": 1.4375,
      "learning_rate": 0.000445146977212719,
      "loss": 5.1471,
      "mean_token_accuracy": 0.20064401030540466,
      "num_tokens": 54379304.0,
      "step": 23740
    },
    {
      "entropy": 5.4684936046600345,
      "epoch": 2.3472716488730723,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004451238838117826,
      "loss": 5.0844,
      "mean_token_accuracy": 0.21057066172361374,
      "num_tokens": 54391462.0,
      "step": 23745
    },
    {
      "entropy": 5.455559349060058,
      "epoch": 2.347765915381574,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004451007862257723,
      "loss": 5.0684,
      "mean_token_accuracy": 0.21585457473993303,
      "num_tokens": 54402325.0,
      "step": 23750
    },
    {
      "entropy": 5.400908279418945,
      "epoch": 2.3482601818900752,
      "grad_norm": 1.25,
      "learning_rate": 0.0004450776844552562,
      "loss": 5.0447,
      "mean_token_accuracy": 0.2044763073325157,
      "num_tokens": 54413874.0,
      "step": 23755
    },
    {
      "entropy": 5.490333938598633,
      "epoch": 2.3487544483985765,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004450545785008027,
      "loss": 5.147,
      "mean_token_accuracy": 0.20435102581977843,
      "num_tokens": 54426266.0,
      "step": 23760
    },
    {
      "entropy": 5.468487787246704,
      "epoch": 2.3492487149070778,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044503146836298026,
      "loss": 5.1317,
      "mean_token_accuracy": 0.20383675694465636,
      "num_tokens": 54436682.0,
      "step": 23765
    },
    {
      "entropy": 5.477557945251465,
      "epoch": 2.3497429814155795,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044500835404235727,
      "loss": 5.1036,
      "mean_token_accuracy": 0.20383972078561782,
      "num_tokens": 54447462.0,
      "step": 23770
    },
    {
      "entropy": 5.581313896179199,
      "epoch": 2.3502372479240807,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004449852355395024,
      "loss": 5.1637,
      "mean_token_accuracy": 0.20379960238933564,
      "num_tokens": 54460116.0,
      "step": 23775
    },
    {
      "entropy": 5.523346328735352,
      "epoch": 2.350731514432582,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044496211285498434,
      "loss": 5.1205,
      "mean_token_accuracy": 0.20514213144779206,
      "num_tokens": 54471016.0,
      "step": 23780
    },
    {
      "entropy": 5.434674167633057,
      "epoch": 2.3512257809410833,
      "grad_norm": 1.140625,
      "learning_rate": 0.000444938985989372,
      "loss": 5.0443,
      "mean_token_accuracy": 0.20904642194509507,
      "num_tokens": 54482410.0,
      "step": 23785
    },
    {
      "entropy": 5.456105661392212,
      "epoch": 2.351720047449585,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044491585494323415,
      "loss": 5.1075,
      "mean_token_accuracy": 0.20677542686462402,
      "num_tokens": 54495555.0,
      "step": 23790
    },
    {
      "entropy": 5.499029684066772,
      "epoch": 2.3522143139580862,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044489271971713995,
      "loss": 5.1345,
      "mean_token_accuracy": 0.20220936685800553,
      "num_tokens": 54505658.0,
      "step": 23795
    },
    {
      "entropy": 5.398524618148803,
      "epoch": 2.3527085804665875,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00044486958031165856,
      "loss": 5.0665,
      "mean_token_accuracy": 0.20745669007301332,
      "num_tokens": 54518633.0,
      "step": 23800
    },
    {
      "entropy": 5.556973934173584,
      "epoch": 2.3532028469750887,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044484643672735907,
      "loss": 5.2666,
      "mean_token_accuracy": 0.1949191078543663,
      "num_tokens": 54531451.0,
      "step": 23805
    },
    {
      "entropy": 5.485000848770142,
      "epoch": 2.3536971134835905,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004448232889648109,
      "loss": 5.051,
      "mean_token_accuracy": 0.209907403588295,
      "num_tokens": 54542798.0,
      "step": 23810
    },
    {
      "entropy": 5.486566829681396,
      "epoch": 2.3541913799920917,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004448001370245835,
      "loss": 5.163,
      "mean_token_accuracy": 0.19694053679704665,
      "num_tokens": 54554551.0,
      "step": 23815
    },
    {
      "entropy": 5.596773386001587,
      "epoch": 2.354685646500593,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044477698090724635,
      "loss": 5.1654,
      "mean_token_accuracy": 0.20056671947240828,
      "num_tokens": 54564711.0,
      "step": 23820
    },
    {
      "entropy": 5.516957187652588,
      "epoch": 2.3551799130090947,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044475382061336913,
      "loss": 5.1803,
      "mean_token_accuracy": 0.19257331490516663,
      "num_tokens": 54576096.0,
      "step": 23825
    },
    {
      "entropy": 5.505116510391235,
      "epoch": 2.355674179517596,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004447306561435215,
      "loss": 5.1496,
      "mean_token_accuracy": 0.19757141470909118,
      "num_tokens": 54587939.0,
      "step": 23830
    },
    {
      "entropy": 5.549318838119507,
      "epoch": 2.356168446026097,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044470748749827354,
      "loss": 5.0406,
      "mean_token_accuracy": 0.20754163712263107,
      "num_tokens": 54598470.0,
      "step": 23835
    },
    {
      "entropy": 5.523328971862793,
      "epoch": 2.3566627125345985,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044468431467819505,
      "loss": 5.1368,
      "mean_token_accuracy": 0.2010859191417694,
      "num_tokens": 54608840.0,
      "step": 23840
    },
    {
      "entropy": 5.464947605133057,
      "epoch": 2.3571569790431,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000444661137683856,
      "loss": 5.0881,
      "mean_token_accuracy": 0.20052892714738846,
      "num_tokens": 54619319.0,
      "step": 23845
    },
    {
      "entropy": 5.4659031391143795,
      "epoch": 2.3576512455516014,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004446379565158268,
      "loss": 5.1063,
      "mean_token_accuracy": 0.2052032843232155,
      "num_tokens": 54630410.0,
      "step": 23850
    },
    {
      "entropy": 5.542837047576905,
      "epoch": 2.3581455120601027,
      "grad_norm": 1.125,
      "learning_rate": 0.0004446147711746774,
      "loss": 5.2246,
      "mean_token_accuracy": 0.19140036553144454,
      "num_tokens": 54642376.0,
      "step": 23855
    },
    {
      "entropy": 5.474863529205322,
      "epoch": 2.3586397785686044,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004445915816609784,
      "loss": 5.1145,
      "mean_token_accuracy": 0.2047329604625702,
      "num_tokens": 54654378.0,
      "step": 23860
    },
    {
      "entropy": 5.492846441268921,
      "epoch": 2.3591340450771057,
      "grad_norm": 1.28125,
      "learning_rate": 0.00044456838797530006,
      "loss": 5.0813,
      "mean_token_accuracy": 0.20095590949058534,
      "num_tokens": 54664879.0,
      "step": 23865
    },
    {
      "entropy": 5.458337736129761,
      "epoch": 2.359628311585607,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004445451901182132,
      "loss": 5.1164,
      "mean_token_accuracy": 0.2028850108385086,
      "num_tokens": 54677251.0,
      "step": 23870
    },
    {
      "entropy": 5.487436437606812,
      "epoch": 2.360122578094108,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004445219880902884,
      "loss": 5.0734,
      "mean_token_accuracy": 0.20168025493621827,
      "num_tokens": 54688827.0,
      "step": 23875
    },
    {
      "entropy": 5.470152473449707,
      "epoch": 2.36061684460261,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004444987818920963,
      "loss": 5.1431,
      "mean_token_accuracy": 0.1998108595609665,
      "num_tokens": 54701046.0,
      "step": 23880
    },
    {
      "entropy": 5.56888370513916,
      "epoch": 2.361111111111111,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004444755715242078,
      "loss": 5.2098,
      "mean_token_accuracy": 0.19205886870622635,
      "num_tokens": 54713254.0,
      "step": 23885
    },
    {
      "entropy": 5.500604152679443,
      "epoch": 2.3616053776196124,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044445235698719406,
      "loss": 5.0713,
      "mean_token_accuracy": 0.20636613816022872,
      "num_tokens": 54723734.0,
      "step": 23890
    },
    {
      "entropy": 5.497933959960937,
      "epoch": 2.362099644128114,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000444429138281626,
      "loss": 5.1357,
      "mean_token_accuracy": 0.19890673756599425,
      "num_tokens": 54735080.0,
      "step": 23895
    },
    {
      "entropy": 5.430173635482788,
      "epoch": 2.3625939106366154,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044440591540807496,
      "loss": 5.0233,
      "mean_token_accuracy": 0.21318188756704332,
      "num_tokens": 54745164.0,
      "step": 23900
    },
    {
      "entropy": 5.465091753005981,
      "epoch": 2.3630881771451167,
      "grad_norm": 1.171875,
      "learning_rate": 0.000444382688367112,
      "loss": 5.1448,
      "mean_token_accuracy": 0.20126674324274063,
      "num_tokens": 54757350.0,
      "step": 23905
    },
    {
      "entropy": 5.4750429630279545,
      "epoch": 2.363582443653618,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00044435945715930864,
      "loss": 5.0279,
      "mean_token_accuracy": 0.20452284514904023,
      "num_tokens": 54768290.0,
      "step": 23910
    },
    {
      "entropy": 5.501429843902588,
      "epoch": 2.364076710162119,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004443362217852364,
      "loss": 5.1582,
      "mean_token_accuracy": 0.20043061673641205,
      "num_tokens": 54779053.0,
      "step": 23915
    },
    {
      "entropy": 5.588014411926269,
      "epoch": 2.364570976670621,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004443129822454668,
      "loss": 5.2659,
      "mean_token_accuracy": 0.18776739984750748,
      "num_tokens": 54789787.0,
      "step": 23920
    },
    {
      "entropy": 5.546839427947998,
      "epoch": 2.365065243179122,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044428973854057164,
      "loss": 5.1557,
      "mean_token_accuracy": 0.20045240223407745,
      "num_tokens": 54802036.0,
      "step": 23925
    },
    {
      "entropy": 5.505240774154663,
      "epoch": 2.3655595096876234,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044426649067112257,
      "loss": 5.1542,
      "mean_token_accuracy": 0.20359872430562972,
      "num_tokens": 54811765.0,
      "step": 23930
    },
    {
      "entropy": 5.453181314468384,
      "epoch": 2.366053776196125,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004442432386376916,
      "loss": 5.0991,
      "mean_token_accuracy": 0.20627350509166717,
      "num_tokens": 54823057.0,
      "step": 23935
    },
    {
      "entropy": 5.502063417434693,
      "epoch": 2.3665480427046264,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044421998244085066,
      "loss": 5.1144,
      "mean_token_accuracy": 0.20075902044773103,
      "num_tokens": 54834070.0,
      "step": 23940
    },
    {
      "entropy": 5.486255168914795,
      "epoch": 2.3670423092131276,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044419672208117204,
      "loss": 5.1167,
      "mean_token_accuracy": 0.1963752716779709,
      "num_tokens": 54845913.0,
      "step": 23945
    },
    {
      "entropy": 5.483513355255127,
      "epoch": 2.367536575721629,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004441734575592276,
      "loss": 5.1383,
      "mean_token_accuracy": 0.201385959982872,
      "num_tokens": 54856721.0,
      "step": 23950
    },
    {
      "entropy": 5.565858983993531,
      "epoch": 2.3680308422301306,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044415018887559,
      "loss": 5.1626,
      "mean_token_accuracy": 0.1989971309900284,
      "num_tokens": 54868652.0,
      "step": 23955
    },
    {
      "entropy": 5.462923812866211,
      "epoch": 2.368525108738632,
      "grad_norm": 1.125,
      "learning_rate": 0.0004441269160308315,
      "loss": 5.0395,
      "mean_token_accuracy": 0.21349249929189681,
      "num_tokens": 54880753.0,
      "step": 23960
    },
    {
      "entropy": 5.409200239181518,
      "epoch": 2.369019375247133,
      "grad_norm": 1.578125,
      "learning_rate": 0.00044410363902552466,
      "loss": 5.0181,
      "mean_token_accuracy": 0.21645736396312715,
      "num_tokens": 54891724.0,
      "step": 23965
    },
    {
      "entropy": 5.492698335647583,
      "epoch": 2.369513641755635,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000444080357860242,
      "loss": 5.1398,
      "mean_token_accuracy": 0.20223337709903716,
      "num_tokens": 54902516.0,
      "step": 23970
    },
    {
      "entropy": 5.485365915298462,
      "epoch": 2.370007908264136,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044405707253555626,
      "loss": 5.1465,
      "mean_token_accuracy": 0.20234702080488204,
      "num_tokens": 54914392.0,
      "step": 23975
    },
    {
      "entropy": 5.468743562698364,
      "epoch": 2.3705021747726374,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004440337830520404,
      "loss": 5.024,
      "mean_token_accuracy": 0.21231903433799743,
      "num_tokens": 54925209.0,
      "step": 23980
    },
    {
      "entropy": 5.500132465362549,
      "epoch": 2.3709964412811386,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044401048941026725,
      "loss": 5.0484,
      "mean_token_accuracy": 0.201961487531662,
      "num_tokens": 54935972.0,
      "step": 23985
    },
    {
      "entropy": 5.45325722694397,
      "epoch": 2.3714907077896403,
      "grad_norm": 1.125,
      "learning_rate": 0.0004439871916108098,
      "loss": 5.0542,
      "mean_token_accuracy": 0.2077438160777092,
      "num_tokens": 54945631.0,
      "step": 23990
    },
    {
      "entropy": 5.474172449111938,
      "epoch": 2.3719849742981416,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044396388965424124,
      "loss": 5.1131,
      "mean_token_accuracy": 0.19927688390016557,
      "num_tokens": 54955734.0,
      "step": 23995
    },
    {
      "entropy": 5.484754180908203,
      "epoch": 2.372479240806643,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004439405835411348,
      "loss": 5.1763,
      "mean_token_accuracy": 0.19966885596513748,
      "num_tokens": 54968506.0,
      "step": 24000
    },
    {
      "epoch": 2.372479240806643,
      "eval_entropy": 5.347727731174502,
      "eval_loss": 5.2096967697143555,
      "eval_mean_token_accuracy": 0.2052016352283852,
      "eval_num_tokens": 54968506.0,
      "eval_runtime": 26.489,
      "eval_samples_per_second": 1227.417,
      "eval_steps_per_second": 153.46,
      "step": 24000
    },
    {
      "entropy": 5.4953700542449955,
      "epoch": 2.3729735073151446,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004439172732720637,
      "loss": 5.0653,
      "mean_token_accuracy": 0.21215397566556932,
      "num_tokens": 54979809.0,
      "step": 24005
    },
    {
      "entropy": 5.538318395614624,
      "epoch": 2.373467773823646,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004438939588476016,
      "loss": 5.2399,
      "mean_token_accuracy": 0.19654985666275024,
      "num_tokens": 54991276.0,
      "step": 24010
    },
    {
      "entropy": 5.516181516647339,
      "epoch": 2.373962040332147,
      "grad_norm": 1.390625,
      "learning_rate": 0.00044387064026832183,
      "loss": 5.096,
      "mean_token_accuracy": 0.2019803062081337,
      "num_tokens": 55002320.0,
      "step": 24015
    },
    {
      "entropy": 5.362782430648804,
      "epoch": 2.3744563068406483,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044384731753479813,
      "loss": 4.9517,
      "mean_token_accuracy": 0.22503859400749207,
      "num_tokens": 55012995.0,
      "step": 24020
    },
    {
      "entropy": 5.486984014511108,
      "epoch": 2.37495057334915,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044382399064760424,
      "loss": 5.1643,
      "mean_token_accuracy": 0.19842659384012223,
      "num_tokens": 55025342.0,
      "step": 24025
    },
    {
      "entropy": 5.454856729507446,
      "epoch": 2.3754448398576513,
      "grad_norm": 1.09375,
      "learning_rate": 0.000443800659607314,
      "loss": 5.1311,
      "mean_token_accuracy": 0.20662947446107865,
      "num_tokens": 55036925.0,
      "step": 24030
    },
    {
      "entropy": 5.516967248916626,
      "epoch": 2.3759391063661526,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004437773244145013,
      "loss": 5.1137,
      "mean_token_accuracy": 0.20011163949966432,
      "num_tokens": 55048913.0,
      "step": 24035
    },
    {
      "entropy": 5.48542537689209,
      "epoch": 2.376433372874654,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004437539850697403,
      "loss": 5.0949,
      "mean_token_accuracy": 0.2043862521648407,
      "num_tokens": 55059654.0,
      "step": 24040
    },
    {
      "entropy": 5.557176780700684,
      "epoch": 2.3769276393831555,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044373064157360504,
      "loss": 5.178,
      "mean_token_accuracy": 0.1968182221055031,
      "num_tokens": 55071077.0,
      "step": 24045
    },
    {
      "entropy": 5.50376615524292,
      "epoch": 2.377421905891657,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004437072939266698,
      "loss": 5.1216,
      "mean_token_accuracy": 0.20217968374490738,
      "num_tokens": 55082659.0,
      "step": 24050
    },
    {
      "entropy": 5.400593900680542,
      "epoch": 2.377916172400158,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044368394212950896,
      "loss": 5.1698,
      "mean_token_accuracy": 0.20113613605499267,
      "num_tokens": 55093938.0,
      "step": 24055
    },
    {
      "entropy": 5.5777500629425045,
      "epoch": 2.3784104389086593,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044366058618269707,
      "loss": 5.1513,
      "mean_token_accuracy": 0.20259769409894943,
      "num_tokens": 55105491.0,
      "step": 24060
    },
    {
      "entropy": 5.467372941970825,
      "epoch": 2.378904705417161,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004436372260868085,
      "loss": 4.9973,
      "mean_token_accuracy": 0.21586406230926514,
      "num_tokens": 55116941.0,
      "step": 24065
    },
    {
      "entropy": 5.559657764434815,
      "epoch": 2.3793989719256623,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044361386184241805,
      "loss": 5.1512,
      "mean_token_accuracy": 0.19971656054258347,
      "num_tokens": 55128294.0,
      "step": 24070
    },
    {
      "entropy": 5.55187201499939,
      "epoch": 2.3798932384341636,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044359049345010044,
      "loss": 5.154,
      "mean_token_accuracy": 0.20051767081022262,
      "num_tokens": 55139781.0,
      "step": 24075
    },
    {
      "entropy": 5.472703742980957,
      "epoch": 2.3803875049426653,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004435671209104305,
      "loss": 5.0995,
      "mean_token_accuracy": 0.20488187223672866,
      "num_tokens": 55151517.0,
      "step": 24080
    },
    {
      "entropy": 5.390579032897949,
      "epoch": 2.3808817714511665,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004435437442239833,
      "loss": 5.0892,
      "mean_token_accuracy": 0.20299521833658218,
      "num_tokens": 55163029.0,
      "step": 24085
    },
    {
      "entropy": 5.469687604904175,
      "epoch": 2.381376037959668,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004435203633913337,
      "loss": 5.0533,
      "mean_token_accuracy": 0.21242140382528304,
      "num_tokens": 55174262.0,
      "step": 24090
    },
    {
      "entropy": 5.4816405296325685,
      "epoch": 2.381870304468169,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044349697841305715,
      "loss": 5.0926,
      "mean_token_accuracy": 0.2117959141731262,
      "num_tokens": 55185047.0,
      "step": 24095
    },
    {
      "entropy": 5.449206686019897,
      "epoch": 2.3823645709766708,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044347358928972883,
      "loss": 5.1236,
      "mean_token_accuracy": 0.20197117030620576,
      "num_tokens": 55196396.0,
      "step": 24100
    },
    {
      "entropy": 5.529635572433472,
      "epoch": 2.382858837485172,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004434501960219239,
      "loss": 5.1444,
      "mean_token_accuracy": 0.20481221824884416,
      "num_tokens": 55208774.0,
      "step": 24105
    },
    {
      "entropy": 5.52050371170044,
      "epoch": 2.3833531039936733,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004434267986102181,
      "loss": 5.1317,
      "mean_token_accuracy": 0.19644552022218703,
      "num_tokens": 55219946.0,
      "step": 24110
    },
    {
      "entropy": 5.472808074951172,
      "epoch": 2.383847370502175,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000443403397055187,
      "loss": 5.1969,
      "mean_token_accuracy": 0.20116210579872132,
      "num_tokens": 55231926.0,
      "step": 24115
    },
    {
      "entropy": 5.489148712158203,
      "epoch": 2.3843416370106763,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044337999135740614,
      "loss": 5.0589,
      "mean_token_accuracy": 0.2045910194516182,
      "num_tokens": 55245240.0,
      "step": 24120
    },
    {
      "entropy": 5.537826061248779,
      "epoch": 2.3848359035191775,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004433565815174514,
      "loss": 5.1473,
      "mean_token_accuracy": 0.20090622305870057,
      "num_tokens": 55256862.0,
      "step": 24125
    },
    {
      "entropy": 5.507908821105957,
      "epoch": 2.3853301700276788,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044333316753589864,
      "loss": 5.0999,
      "mean_token_accuracy": 0.19921667575836183,
      "num_tokens": 55268157.0,
      "step": 24130
    },
    {
      "entropy": 5.53131136894226,
      "epoch": 2.3858244365361805,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004433097494133238,
      "loss": 5.1637,
      "mean_token_accuracy": 0.19946774691343308,
      "num_tokens": 55279580.0,
      "step": 24135
    },
    {
      "entropy": 5.617420244216919,
      "epoch": 2.3863187030446817,
      "grad_norm": 1.25,
      "learning_rate": 0.00044328632715030303,
      "loss": 5.33,
      "mean_token_accuracy": 0.18546686172485352,
      "num_tokens": 55292067.0,
      "step": 24140
    },
    {
      "entropy": 5.460070943832397,
      "epoch": 2.386812969553183,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044326290074741257,
      "loss": 5.0961,
      "mean_token_accuracy": 0.20935119390487672,
      "num_tokens": 55303338.0,
      "step": 24145
    },
    {
      "entropy": 5.492969608306884,
      "epoch": 2.3873072360616843,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004432394702052286,
      "loss": 5.1718,
      "mean_token_accuracy": 0.19402682185173034,
      "num_tokens": 55315165.0,
      "step": 24150
    },
    {
      "entropy": 5.54754672050476,
      "epoch": 2.387801502570186,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044321603552432753,
      "loss": 5.145,
      "mean_token_accuracy": 0.19918230026960373,
      "num_tokens": 55327928.0,
      "step": 24155
    },
    {
      "entropy": 5.449443340301514,
      "epoch": 2.3882957690786872,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044319259670528595,
      "loss": 5.0567,
      "mean_token_accuracy": 0.2062981069087982,
      "num_tokens": 55338438.0,
      "step": 24160
    },
    {
      "entropy": 5.417128801345825,
      "epoch": 2.3887900355871885,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044316915374868033,
      "loss": 5.0296,
      "mean_token_accuracy": 0.20826223641633987,
      "num_tokens": 55349513.0,
      "step": 24165
    },
    {
      "entropy": 5.460167026519775,
      "epoch": 2.3892843020956898,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044314570665508746,
      "loss": 5.1546,
      "mean_token_accuracy": 0.20052433609962464,
      "num_tokens": 55362498.0,
      "step": 24170
    },
    {
      "entropy": 5.576184177398682,
      "epoch": 2.3897785686041915,
      "grad_norm": 1.125,
      "learning_rate": 0.0004431222554250841,
      "loss": 5.1317,
      "mean_token_accuracy": 0.19695977717638016,
      "num_tokens": 55374065.0,
      "step": 24175
    },
    {
      "entropy": 5.468780374526977,
      "epoch": 2.3902728351126927,
      "grad_norm": 1.125,
      "learning_rate": 0.00044309880005924726,
      "loss": 5.1166,
      "mean_token_accuracy": 0.20670635849237443,
      "num_tokens": 55385175.0,
      "step": 24180
    },
    {
      "entropy": 5.544585561752319,
      "epoch": 2.390767101621194,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044307534055815375,
      "loss": 5.1864,
      "mean_token_accuracy": 0.1976812779903412,
      "num_tokens": 55395457.0,
      "step": 24185
    },
    {
      "entropy": 5.5038244247436525,
      "epoch": 2.3912613681296957,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044305187692238087,
      "loss": 5.1242,
      "mean_token_accuracy": 0.1999911606311798,
      "num_tokens": 55406341.0,
      "step": 24190
    },
    {
      "entropy": 5.50658950805664,
      "epoch": 2.391755634638197,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044302840915250575,
      "loss": 5.092,
      "mean_token_accuracy": 0.19989708214998245,
      "num_tokens": 55417893.0,
      "step": 24195
    },
    {
      "entropy": 5.518565368652344,
      "epoch": 2.392249901146698,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004430049372491056,
      "loss": 5.1715,
      "mean_token_accuracy": 0.19694544076919557,
      "num_tokens": 55428286.0,
      "step": 24200
    },
    {
      "entropy": 5.399940872192383,
      "epoch": 2.3927441676551995,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044298146121275804,
      "loss": 5.0358,
      "mean_token_accuracy": 0.20724110305309296,
      "num_tokens": 55439861.0,
      "step": 24205
    },
    {
      "entropy": 5.5473103523254395,
      "epoch": 2.393238434163701,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004429579810440404,
      "loss": 5.115,
      "mean_token_accuracy": 0.19754144847393035,
      "num_tokens": 55450971.0,
      "step": 24210
    },
    {
      "entropy": 5.4810456275939945,
      "epoch": 2.3937327006722025,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004429344967435305,
      "loss": 5.0928,
      "mean_token_accuracy": 0.19796846359968184,
      "num_tokens": 55460584.0,
      "step": 24215
    },
    {
      "entropy": 5.570596027374267,
      "epoch": 2.3942269671807037,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044291100831180577,
      "loss": 5.2274,
      "mean_token_accuracy": 0.1915025904774666,
      "num_tokens": 55471576.0,
      "step": 24220
    },
    {
      "entropy": 5.52740511894226,
      "epoch": 2.3947212336892054,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004428875157494442,
      "loss": 5.1252,
      "mean_token_accuracy": 0.20337176620960234,
      "num_tokens": 55483807.0,
      "step": 24225
    },
    {
      "entropy": 5.466325902938843,
      "epoch": 2.3952155001977067,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044286401905702376,
      "loss": 5.1302,
      "mean_token_accuracy": 0.20709213018417358,
      "num_tokens": 55495130.0,
      "step": 24230
    },
    {
      "entropy": 5.47646746635437,
      "epoch": 2.395709766706208,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004428405182351224,
      "loss": 5.0855,
      "mean_token_accuracy": 0.20969182401895523,
      "num_tokens": 55506460.0,
      "step": 24235
    },
    {
      "entropy": 5.5734693050384525,
      "epoch": 2.396204033214709,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004428170132843182,
      "loss": 5.1847,
      "mean_token_accuracy": 0.201817487180233,
      "num_tokens": 55518401.0,
      "step": 24240
    },
    {
      "entropy": 5.5229939937591555,
      "epoch": 2.396698299723211,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044279350420518956,
      "loss": 5.097,
      "mean_token_accuracy": 0.20685291439294815,
      "num_tokens": 55530403.0,
      "step": 24245
    },
    {
      "entropy": 5.536352062225342,
      "epoch": 2.397192566231712,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044276999099831463,
      "loss": 5.1471,
      "mean_token_accuracy": 0.20163918435573577,
      "num_tokens": 55541422.0,
      "step": 24250
    },
    {
      "entropy": 5.446721172332763,
      "epoch": 2.3976868327402134,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044274647366427186,
      "loss": 5.0659,
      "mean_token_accuracy": 0.20509060174226762,
      "num_tokens": 55553798.0,
      "step": 24255
    },
    {
      "entropy": 5.455235195159912,
      "epoch": 2.398181099248715,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044272295220363977,
      "loss": 5.1239,
      "mean_token_accuracy": 0.20419947803020477,
      "num_tokens": 55565970.0,
      "step": 24260
    },
    {
      "entropy": 5.505720472335815,
      "epoch": 2.3986753657572164,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004426994266169971,
      "loss": 5.1781,
      "mean_token_accuracy": 0.19681297689676286,
      "num_tokens": 55576385.0,
      "step": 24265
    },
    {
      "entropy": 5.449960517883301,
      "epoch": 2.3991696322657177,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004426758969049225,
      "loss": 5.0451,
      "mean_token_accuracy": 0.21620315909385682,
      "num_tokens": 55588752.0,
      "step": 24270
    },
    {
      "entropy": 5.451366329193116,
      "epoch": 2.399663898774219,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004426523630679949,
      "loss": 5.0965,
      "mean_token_accuracy": 0.20305163264274598,
      "num_tokens": 55599348.0,
      "step": 24275
    },
    {
      "entropy": 5.473493909835815,
      "epoch": 2.4001581652827206,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004426288251067931,
      "loss": 5.0732,
      "mean_token_accuracy": 0.20013322234153746,
      "num_tokens": 55610168.0,
      "step": 24280
    },
    {
      "entropy": 5.496099424362183,
      "epoch": 2.400652431791222,
      "grad_norm": 1.375,
      "learning_rate": 0.00044260528302189617,
      "loss": 5.1399,
      "mean_token_accuracy": 0.20125294476747513,
      "num_tokens": 55622487.0,
      "step": 24285
    },
    {
      "entropy": 5.559060001373291,
      "epoch": 2.401146698299723,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004425817368138833,
      "loss": 5.1556,
      "mean_token_accuracy": 0.2029305398464203,
      "num_tokens": 55634486.0,
      "step": 24290
    },
    {
      "entropy": 5.510579347610474,
      "epoch": 2.4016409648082244,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004425581864833337,
      "loss": 5.1473,
      "mean_token_accuracy": 0.19999028295278548,
      "num_tokens": 55646642.0,
      "step": 24295
    },
    {
      "entropy": 5.397876358032226,
      "epoch": 2.402135231316726,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004425346320308267,
      "loss": 4.9986,
      "mean_token_accuracy": 0.21957552284002305,
      "num_tokens": 55657921.0,
      "step": 24300
    },
    {
      "entropy": 5.441567993164062,
      "epoch": 2.4026294978252274,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004425110734569418,
      "loss": 5.1156,
      "mean_token_accuracy": 0.20996900349855424,
      "num_tokens": 55669414.0,
      "step": 24305
    },
    {
      "entropy": 5.458219718933106,
      "epoch": 2.4031237643337287,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044248751076225853,
      "loss": 5.056,
      "mean_token_accuracy": 0.20717837959527968,
      "num_tokens": 55681854.0,
      "step": 24310
    },
    {
      "entropy": 5.49722900390625,
      "epoch": 2.40361803084223,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044246394394735646,
      "loss": 5.0375,
      "mean_token_accuracy": 0.20497379899024964,
      "num_tokens": 55692458.0,
      "step": 24315
    },
    {
      "entropy": 5.525239181518555,
      "epoch": 2.4041122973507316,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044244037301281535,
      "loss": 5.1282,
      "mean_token_accuracy": 0.2021837279200554,
      "num_tokens": 55703676.0,
      "step": 24320
    },
    {
      "entropy": 5.514711332321167,
      "epoch": 2.404606563859233,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004424167979592152,
      "loss": 5.1258,
      "mean_token_accuracy": 0.19684544950723648,
      "num_tokens": 55714372.0,
      "step": 24325
    },
    {
      "entropy": 5.469604349136352,
      "epoch": 2.405100830367734,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004423932187871358,
      "loss": 5.0726,
      "mean_token_accuracy": 0.20633845180273055,
      "num_tokens": 55724342.0,
      "step": 24330
    },
    {
      "entropy": 5.489522981643677,
      "epoch": 2.405595096876236,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004423696354971572,
      "loss": 5.1874,
      "mean_token_accuracy": 0.19932929128408433,
      "num_tokens": 55736272.0,
      "step": 24335
    },
    {
      "entropy": 5.507037591934204,
      "epoch": 2.406089363384737,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004423460480898596,
      "loss": 5.0819,
      "mean_token_accuracy": 0.21597340255975722,
      "num_tokens": 55747174.0,
      "step": 24340
    },
    {
      "entropy": 5.40907654762268,
      "epoch": 2.4065836298932384,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004423224565658233,
      "loss": 5.0199,
      "mean_token_accuracy": 0.20998756289482118,
      "num_tokens": 55759676.0,
      "step": 24345
    },
    {
      "entropy": 5.454511117935181,
      "epoch": 2.4070778964017396,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004422988609256286,
      "loss": 5.1376,
      "mean_token_accuracy": 0.19554717242717742,
      "num_tokens": 55772140.0,
      "step": 24350
    },
    {
      "entropy": 5.467319822311401,
      "epoch": 2.4075721629102413,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000442275261169856,
      "loss": 5.0505,
      "mean_token_accuracy": 0.20849403291940688,
      "num_tokens": 55783639.0,
      "step": 24355
    },
    {
      "entropy": 5.532367849349976,
      "epoch": 2.4080664294187426,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044225165729908604,
      "loss": 5.2054,
      "mean_token_accuracy": 0.19538605064153672,
      "num_tokens": 55795867.0,
      "step": 24360
    },
    {
      "entropy": 5.373350048065186,
      "epoch": 2.408560695927244,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044222804931389934,
      "loss": 5.0225,
      "mean_token_accuracy": 0.2171737089753151,
      "num_tokens": 55807528.0,
      "step": 24365
    },
    {
      "entropy": 5.502852153778076,
      "epoch": 2.4090549624357456,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004422044372148766,
      "loss": 5.0497,
      "mean_token_accuracy": 0.20848052054643632,
      "num_tokens": 55817547.0,
      "step": 24370
    },
    {
      "entropy": 5.419193935394287,
      "epoch": 2.409549228944247,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044218082100259883,
      "loss": 5.0321,
      "mean_token_accuracy": 0.21471346616744996,
      "num_tokens": 55828474.0,
      "step": 24375
    },
    {
      "entropy": 5.444436120986938,
      "epoch": 2.410043495452748,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044215720067764696,
      "loss": 5.0505,
      "mean_token_accuracy": 0.20983597487211228,
      "num_tokens": 55838485.0,
      "step": 24380
    },
    {
      "entropy": 5.4874927520751955,
      "epoch": 2.4105377619612494,
      "grad_norm": 1.140625,
      "learning_rate": 0.000442133576240602,
      "loss": 5.1367,
      "mean_token_accuracy": 0.21096194088459014,
      "num_tokens": 55850219.0,
      "step": 24385
    },
    {
      "entropy": 5.449599742889404,
      "epoch": 2.411032028469751,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004421099476920451,
      "loss": 5.0031,
      "mean_token_accuracy": 0.2150494083762169,
      "num_tokens": 55860241.0,
      "step": 24390
    },
    {
      "entropy": 5.5127543926239015,
      "epoch": 2.4115262949782523,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004420863150325575,
      "loss": 5.2229,
      "mean_token_accuracy": 0.19013428092002868,
      "num_tokens": 55871201.0,
      "step": 24395
    },
    {
      "entropy": 5.457510232925415,
      "epoch": 2.4120205614867536,
      "grad_norm": 1.125,
      "learning_rate": 0.0004420626782627208,
      "loss": 5.0179,
      "mean_token_accuracy": 0.2111206203699112,
      "num_tokens": 55881758.0,
      "step": 24400
    },
    {
      "entropy": 5.4877049922943115,
      "epoch": 2.412514827995255,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044203903738311616,
      "loss": 5.1013,
      "mean_token_accuracy": 0.20375279933214188,
      "num_tokens": 55892931.0,
      "step": 24405
    },
    {
      "entropy": 5.53276309967041,
      "epoch": 2.4130090945037566,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044201539239432534,
      "loss": 5.0876,
      "mean_token_accuracy": 0.20771762132644653,
      "num_tokens": 55903376.0,
      "step": 24410
    },
    {
      "entropy": 5.5666814804077145,
      "epoch": 2.413503361012258,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044199174329693,
      "loss": 5.1631,
      "mean_token_accuracy": 0.19881929010152816,
      "num_tokens": 55913821.0,
      "step": 24415
    },
    {
      "entropy": 5.442328023910522,
      "epoch": 2.413997627520759,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044196809009151183,
      "loss": 5.1209,
      "mean_token_accuracy": 0.20072400122880935,
      "num_tokens": 55925085.0,
      "step": 24420
    },
    {
      "entropy": 5.467133474349976,
      "epoch": 2.4144918940292603,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004419444327786527,
      "loss": 5.129,
      "mean_token_accuracy": 0.19822475761175157,
      "num_tokens": 55937572.0,
      "step": 24425
    },
    {
      "entropy": 5.480417394638062,
      "epoch": 2.414986160537762,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004419207713589347,
      "loss": 5.1034,
      "mean_token_accuracy": 0.20375400185585021,
      "num_tokens": 55948890.0,
      "step": 24430
    },
    {
      "entropy": 5.469589138031006,
      "epoch": 2.4154804270462633,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004418971058329398,
      "loss": 5.1769,
      "mean_token_accuracy": 0.19648048132658005,
      "num_tokens": 55959319.0,
      "step": 24435
    },
    {
      "entropy": 5.534429931640625,
      "epoch": 2.4159746935547646,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044187343620125016,
      "loss": 5.1612,
      "mean_token_accuracy": 0.1967574030160904,
      "num_tokens": 55970922.0,
      "step": 24440
    },
    {
      "entropy": 5.45665340423584,
      "epoch": 2.4164689600632663,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044184976246444815,
      "loss": 5.1177,
      "mean_token_accuracy": 0.20160169154405594,
      "num_tokens": 55982610.0,
      "step": 24445
    },
    {
      "entropy": 5.47747220993042,
      "epoch": 2.4169632265717675,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044182608462311606,
      "loss": 5.0358,
      "mean_token_accuracy": 0.2117295503616333,
      "num_tokens": 55994393.0,
      "step": 24450
    },
    {
      "entropy": 5.427250719070434,
      "epoch": 2.417457493080269,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044180240267783637,
      "loss": 5.0084,
      "mean_token_accuracy": 0.21002840995788574,
      "num_tokens": 56005705.0,
      "step": 24455
    },
    {
      "entropy": 5.497283935546875,
      "epoch": 2.41795175958877,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044177871662919184,
      "loss": 5.1164,
      "mean_token_accuracy": 0.20299535542726516,
      "num_tokens": 56016942.0,
      "step": 24460
    },
    {
      "entropy": 5.517076730728149,
      "epoch": 2.4184460260972718,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044175502647776486,
      "loss": 5.1307,
      "mean_token_accuracy": 0.19543941617012023,
      "num_tokens": 56028495.0,
      "step": 24465
    },
    {
      "entropy": 5.521891403198242,
      "epoch": 2.418940292605773,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044173133222413845,
      "loss": 5.1734,
      "mean_token_accuracy": 0.20578213781118393,
      "num_tokens": 56039654.0,
      "step": 24470
    },
    {
      "entropy": 5.4873748302459715,
      "epoch": 2.4194345591142743,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004417076338688954,
      "loss": 5.0872,
      "mean_token_accuracy": 0.20901118814945222,
      "num_tokens": 56051392.0,
      "step": 24475
    },
    {
      "entropy": 5.406898450851441,
      "epoch": 2.419928825622776,
      "grad_norm": 1.125,
      "learning_rate": 0.00044168393141261864,
      "loss": 4.9999,
      "mean_token_accuracy": 0.21457445174455642,
      "num_tokens": 56062574.0,
      "step": 24480
    },
    {
      "entropy": 5.456973505020142,
      "epoch": 2.4204230921312773,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044166022485589136,
      "loss": 5.0162,
      "mean_token_accuracy": 0.20823191553354264,
      "num_tokens": 56072637.0,
      "step": 24485
    },
    {
      "entropy": 5.4129579067230225,
      "epoch": 2.4209173586397785,
      "grad_norm": 1.125,
      "learning_rate": 0.00044163651419929667,
      "loss": 5.0523,
      "mean_token_accuracy": 0.20744042992591857,
      "num_tokens": 56083573.0,
      "step": 24490
    },
    {
      "entropy": 5.446558046340942,
      "epoch": 2.42141162514828,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044161279944341785,
      "loss": 5.0881,
      "mean_token_accuracy": 0.20395224839448928,
      "num_tokens": 56094833.0,
      "step": 24495
    },
    {
      "entropy": 5.43768081665039,
      "epoch": 2.4219058916567815,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004415890805888383,
      "loss": 5.0732,
      "mean_token_accuracy": 0.20282193422317504,
      "num_tokens": 56107750.0,
      "step": 24500
    },
    {
      "entropy": 5.49199481010437,
      "epoch": 2.4224001581652828,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00044156535763614167,
      "loss": 5.0553,
      "mean_token_accuracy": 0.2118619292974472,
      "num_tokens": 56118815.0,
      "step": 24505
    },
    {
      "entropy": 5.460046339035034,
      "epoch": 2.422894424673784,
      "grad_norm": 1.125,
      "learning_rate": 0.0004415416305859113,
      "loss": 5.1299,
      "mean_token_accuracy": 0.20109503716230392,
      "num_tokens": 56129120.0,
      "step": 24510
    },
    {
      "entropy": 5.417664670944214,
      "epoch": 2.4233886911822857,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044151789943873095,
      "loss": 5.0363,
      "mean_token_accuracy": 0.20678567439317702,
      "num_tokens": 56140293.0,
      "step": 24515
    },
    {
      "entropy": 5.591537427902222,
      "epoch": 2.423882957690787,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004414941641951845,
      "loss": 5.166,
      "mean_token_accuracy": 0.19427590668201447,
      "num_tokens": 56151942.0,
      "step": 24520
    },
    {
      "entropy": 5.491158437728882,
      "epoch": 2.4243772241992882,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044147042485585575,
      "loss": 5.0884,
      "mean_token_accuracy": 0.20904184877872467,
      "num_tokens": 56163088.0,
      "step": 24525
    },
    {
      "entropy": 5.439865303039551,
      "epoch": 2.4248714907077895,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004414466814213287,
      "loss": 5.1084,
      "mean_token_accuracy": 0.20518146455287933,
      "num_tokens": 56174879.0,
      "step": 24530
    },
    {
      "entropy": 5.527640914916992,
      "epoch": 2.425365757216291,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004414229338921875,
      "loss": 5.1311,
      "mean_token_accuracy": 0.20211424827575683,
      "num_tokens": 56187407.0,
      "step": 24535
    },
    {
      "entropy": 5.472452688217163,
      "epoch": 2.4258600237247925,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044139918226901627,
      "loss": 5.0298,
      "mean_token_accuracy": 0.21303242594003677,
      "num_tokens": 56198830.0,
      "step": 24540
    },
    {
      "entropy": 5.552091598510742,
      "epoch": 2.4263542902332937,
      "grad_norm": 1.0,
      "learning_rate": 0.00044137542655239945,
      "loss": 5.194,
      "mean_token_accuracy": 0.19474708735942842,
      "num_tokens": 56210779.0,
      "step": 24545
    },
    {
      "entropy": 5.43374981880188,
      "epoch": 2.426848556741795,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004413516667429212,
      "loss": 5.0951,
      "mean_token_accuracy": 0.20230225324630738,
      "num_tokens": 56222258.0,
      "step": 24550
    },
    {
      "entropy": 5.400051641464233,
      "epoch": 2.4273428232502967,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004413279028411662,
      "loss": 5.0546,
      "mean_token_accuracy": 0.2163394182920456,
      "num_tokens": 56232834.0,
      "step": 24555
    },
    {
      "entropy": 5.436425590515137,
      "epoch": 2.427837089758798,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000441304134847719,
      "loss": 5.0192,
      "mean_token_accuracy": 0.20468443781137466,
      "num_tokens": 56244517.0,
      "step": 24560
    },
    {
      "entropy": 5.582173871994018,
      "epoch": 2.4283313562672992,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044128036276316424,
      "loss": 5.2103,
      "mean_token_accuracy": 0.1959197700023651,
      "num_tokens": 56256629.0,
      "step": 24565
    },
    {
      "entropy": 5.553802061080932,
      "epoch": 2.4288256227758005,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044125658658808676,
      "loss": 5.2411,
      "mean_token_accuracy": 0.19048468619585038,
      "num_tokens": 56268992.0,
      "step": 24570
    },
    {
      "entropy": 5.430667400360107,
      "epoch": 2.429319889284302,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004412328063230715,
      "loss": 5.0315,
      "mean_token_accuracy": 0.2153841346502304,
      "num_tokens": 56281588.0,
      "step": 24575
    },
    {
      "entropy": 5.5612630367279055,
      "epoch": 2.4298141557928035,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004412090219687034,
      "loss": 5.2208,
      "mean_token_accuracy": 0.19102327823638915,
      "num_tokens": 56292587.0,
      "step": 24580
    },
    {
      "entropy": 5.505627155303955,
      "epoch": 2.4303084223013047,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044118523352556756,
      "loss": 5.0932,
      "mean_token_accuracy": 0.20477480739355086,
      "num_tokens": 56304873.0,
      "step": 24585
    },
    {
      "entropy": 5.413132286071777,
      "epoch": 2.4308026888098064,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004411614409942492,
      "loss": 5.0737,
      "mean_token_accuracy": 0.20850208401679993,
      "num_tokens": 56316048.0,
      "step": 24590
    },
    {
      "entropy": 5.474745225906372,
      "epoch": 2.4312969553183077,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044113764437533356,
      "loss": 5.0977,
      "mean_token_accuracy": 0.20837050825357437,
      "num_tokens": 56328185.0,
      "step": 24595
    },
    {
      "entropy": 5.536024951934815,
      "epoch": 2.431791221826809,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004411138436694062,
      "loss": 5.1992,
      "mean_token_accuracy": 0.19488661885261535,
      "num_tokens": 56341533.0,
      "step": 24600
    },
    {
      "entropy": 5.623372173309326,
      "epoch": 2.43228548833531,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00044109003887705245,
      "loss": 5.2632,
      "mean_token_accuracy": 0.19504843652248383,
      "num_tokens": 56351610.0,
      "step": 24605
    },
    {
      "entropy": 5.517627906799317,
      "epoch": 2.432779754843812,
      "grad_norm": 1.109375,
      "learning_rate": 0.000441066229998858,
      "loss": 5.1318,
      "mean_token_accuracy": 0.21082397550344467,
      "num_tokens": 56361822.0,
      "step": 24610
    },
    {
      "entropy": 5.4779948711395265,
      "epoch": 2.433274021352313,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044104241703540854,
      "loss": 5.0989,
      "mean_token_accuracy": 0.2033616706728935,
      "num_tokens": 56373432.0,
      "step": 24615
    },
    {
      "entropy": 5.458730173110962,
      "epoch": 2.4337682878608144,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004410185999872898,
      "loss": 5.0957,
      "mean_token_accuracy": 0.20270515233278275,
      "num_tokens": 56385086.0,
      "step": 24620
    },
    {
      "entropy": 5.515968894958496,
      "epoch": 2.434262554369316,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044099477885508774,
      "loss": 5.0693,
      "mean_token_accuracy": 0.20813625454902648,
      "num_tokens": 56395963.0,
      "step": 24625
    },
    {
      "entropy": 5.521343183517456,
      "epoch": 2.4347568208778174,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044097095363938843,
      "loss": 5.1533,
      "mean_token_accuracy": 0.19692583233118058,
      "num_tokens": 56407685.0,
      "step": 24630
    },
    {
      "entropy": 5.524341297149658,
      "epoch": 2.4352510873863187,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004409471243407779,
      "loss": 5.1008,
      "mean_token_accuracy": 0.20853097438812257,
      "num_tokens": 56418570.0,
      "step": 24635
    },
    {
      "entropy": 5.505349016189575,
      "epoch": 2.43574535389482,
      "grad_norm": 1.125,
      "learning_rate": 0.0004409232909598424,
      "loss": 5.1409,
      "mean_token_accuracy": 0.20602805465459822,
      "num_tokens": 56430312.0,
      "step": 24640
    },
    {
      "entropy": 5.582175302505493,
      "epoch": 2.4362396204033216,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004408994534971682,
      "loss": 5.1993,
      "mean_token_accuracy": 0.1931255653500557,
      "num_tokens": 56441847.0,
      "step": 24645
    },
    {
      "entropy": 5.46953239440918,
      "epoch": 2.436733886911823,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044087561195334165,
      "loss": 5.1078,
      "mean_token_accuracy": 0.20379755645990372,
      "num_tokens": 56453567.0,
      "step": 24650
    },
    {
      "entropy": 5.47821159362793,
      "epoch": 2.437228153420324,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004408517663289493,
      "loss": 5.1033,
      "mean_token_accuracy": 0.19907173961400987,
      "num_tokens": 56464726.0,
      "step": 24655
    },
    {
      "entropy": 5.492896032333374,
      "epoch": 2.4377224199288254,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004408279166245778,
      "loss": 5.1088,
      "mean_token_accuracy": 0.20372008681297302,
      "num_tokens": 56475989.0,
      "step": 24660
    },
    {
      "entropy": 5.540158462524414,
      "epoch": 2.438216686437327,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044080406284081387,
      "loss": 5.1158,
      "mean_token_accuracy": 0.20692139863967896,
      "num_tokens": 56487782.0,
      "step": 24665
    },
    {
      "entropy": 5.479180479049683,
      "epoch": 2.4387109529458284,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004407802049782442,
      "loss": 5.1135,
      "mean_token_accuracy": 0.20168498754501343,
      "num_tokens": 56500529.0,
      "step": 24670
    },
    {
      "entropy": 5.390605449676514,
      "epoch": 2.4392052194543297,
      "grad_norm": 1.0,
      "learning_rate": 0.0004407563430374558,
      "loss": 5.044,
      "mean_token_accuracy": 0.20880429893732072,
      "num_tokens": 56512411.0,
      "step": 24675
    },
    {
      "entropy": 5.428992176055909,
      "epoch": 2.439699485962831,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004407324770190357,
      "loss": 4.973,
      "mean_token_accuracy": 0.21480999141931534,
      "num_tokens": 56522983.0,
      "step": 24680
    },
    {
      "entropy": 5.430551528930664,
      "epoch": 2.4401937524713326,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044070860692357086,
      "loss": 5.142,
      "mean_token_accuracy": 0.2067538946866989,
      "num_tokens": 56534380.0,
      "step": 24685
    },
    {
      "entropy": 5.515534734725952,
      "epoch": 2.440688018979834,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044068473275164863,
      "loss": 5.1171,
      "mean_token_accuracy": 0.20013733953237534,
      "num_tokens": 56546003.0,
      "step": 24690
    },
    {
      "entropy": 5.444821453094482,
      "epoch": 2.441182285488335,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004406608545038563,
      "loss": 5.007,
      "mean_token_accuracy": 0.2121601089835167,
      "num_tokens": 56556006.0,
      "step": 24695
    },
    {
      "entropy": 5.430108880996704,
      "epoch": 2.441676551996837,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004406369721807812,
      "loss": 5.0358,
      "mean_token_accuracy": 0.20903393775224685,
      "num_tokens": 56567569.0,
      "step": 24700
    },
    {
      "entropy": 5.478749513626099,
      "epoch": 2.442170818505338,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004406130857830109,
      "loss": 5.1485,
      "mean_token_accuracy": 0.20527566224336624,
      "num_tokens": 56580143.0,
      "step": 24705
    },
    {
      "entropy": 5.455153560638427,
      "epoch": 2.4426650850138394,
      "grad_norm": 1.03125,
      "learning_rate": 0.000440589195311133,
      "loss": 5.0382,
      "mean_token_accuracy": 0.21119300723075868,
      "num_tokens": 56591169.0,
      "step": 24710
    },
    {
      "entropy": 5.4638916015625,
      "epoch": 2.4431593515223406,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044056530076573515,
      "loss": 5.1079,
      "mean_token_accuracy": 0.20415205359458924,
      "num_tokens": 56602556.0,
      "step": 24715
    },
    {
      "entropy": 5.522335195541382,
      "epoch": 2.4436536180308424,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004405414021474053,
      "loss": 5.0889,
      "mean_token_accuracy": 0.20856422334909439,
      "num_tokens": 56614854.0,
      "step": 24720
    },
    {
      "entropy": 5.440070009231567,
      "epoch": 2.4441478845393436,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004405174994567313,
      "loss": 5.0132,
      "mean_token_accuracy": 0.2098423272371292,
      "num_tokens": 56626004.0,
      "step": 24725
    },
    {
      "entropy": 5.560739040374756,
      "epoch": 2.444642151047845,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044049359269430105,
      "loss": 5.1482,
      "mean_token_accuracy": 0.19802184402942657,
      "num_tokens": 56636373.0,
      "step": 24730
    },
    {
      "entropy": 5.441765069961548,
      "epoch": 2.4451364175563466,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004404696818607028,
      "loss": 4.9975,
      "mean_token_accuracy": 0.20928096473217012,
      "num_tokens": 56648414.0,
      "step": 24735
    },
    {
      "entropy": 5.454145193099976,
      "epoch": 2.445630684064848,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044044576695652463,
      "loss": 5.0901,
      "mean_token_accuracy": 0.2045057475566864,
      "num_tokens": 56660430.0,
      "step": 24740
    },
    {
      "entropy": 5.5101110458374025,
      "epoch": 2.446124950573349,
      "grad_norm": 1.109375,
      "learning_rate": 0.000440421847982355,
      "loss": 5.0962,
      "mean_token_accuracy": 0.20624202191829683,
      "num_tokens": 56671133.0,
      "step": 24745
    },
    {
      "entropy": 5.491914939880371,
      "epoch": 2.4466192170818504,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004403979249387822,
      "loss": 5.1786,
      "mean_token_accuracy": 0.19465420842170716,
      "num_tokens": 56682788.0,
      "step": 24750
    },
    {
      "entropy": 5.434906959533691,
      "epoch": 2.447113483590352,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004403739978263948,
      "loss": 5.1049,
      "mean_token_accuracy": 0.20493227243423462,
      "num_tokens": 56693832.0,
      "step": 24755
    },
    {
      "entropy": 5.5496758937835695,
      "epoch": 2.4476077500988533,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044035006664578145,
      "loss": 5.1738,
      "mean_token_accuracy": 0.19455254226922988,
      "num_tokens": 56705915.0,
      "step": 24760
    },
    {
      "entropy": 5.468791103363037,
      "epoch": 2.4481020166073546,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044032613139753077,
      "loss": 5.0909,
      "mean_token_accuracy": 0.20960814356803895,
      "num_tokens": 56717461.0,
      "step": 24765
    },
    {
      "entropy": 5.411565494537354,
      "epoch": 2.4485962831158563,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004403021920822316,
      "loss": 4.9504,
      "mean_token_accuracy": 0.2178207442164421,
      "num_tokens": 56728695.0,
      "step": 24770
    },
    {
      "entropy": 5.412420845031738,
      "epoch": 2.4490905496243576,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004402782487004728,
      "loss": 5.0535,
      "mean_token_accuracy": 0.203393816947937,
      "num_tokens": 56740520.0,
      "step": 24775
    },
    {
      "entropy": 5.396411514282226,
      "epoch": 2.449584816132859,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044025430125284353,
      "loss": 5.0021,
      "mean_token_accuracy": 0.21679751127958297,
      "num_tokens": 56751071.0,
      "step": 24780
    },
    {
      "entropy": 5.5069197654724125,
      "epoch": 2.45007908264136,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004402303497399329,
      "loss": 5.1577,
      "mean_token_accuracy": 0.19686210602521897,
      "num_tokens": 56764759.0,
      "step": 24785
    },
    {
      "entropy": 5.555088710784912,
      "epoch": 2.4505733491498614,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044020639416232984,
      "loss": 5.1564,
      "mean_token_accuracy": 0.2036934807896614,
      "num_tokens": 56775592.0,
      "step": 24790
    },
    {
      "entropy": 5.51851372718811,
      "epoch": 2.451067615658363,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000440182434520624,
      "loss": 5.1963,
      "mean_token_accuracy": 0.19388588964939119,
      "num_tokens": 56786860.0,
      "step": 24795
    },
    {
      "entropy": 5.528635692596436,
      "epoch": 2.4515618821668643,
      "grad_norm": 1.125,
      "learning_rate": 0.0004401584708154045,
      "loss": 5.1398,
      "mean_token_accuracy": 0.2015665054321289,
      "num_tokens": 56798741.0,
      "step": 24800
    },
    {
      "entropy": 5.510555458068848,
      "epoch": 2.4520561486753656,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044013450304726114,
      "loss": 5.0734,
      "mean_token_accuracy": 0.20401668846607207,
      "num_tokens": 56810612.0,
      "step": 24805
    },
    {
      "entropy": 5.480600357055664,
      "epoch": 2.4525504151838673,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004401105312167833,
      "loss": 5.1233,
      "mean_token_accuracy": 0.2056616112589836,
      "num_tokens": 56822169.0,
      "step": 24810
    },
    {
      "entropy": 5.446529817581177,
      "epoch": 2.4530446816923686,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004400865553245608,
      "loss": 5.0524,
      "mean_token_accuracy": 0.20780978202819825,
      "num_tokens": 56833873.0,
      "step": 24815
    },
    {
      "entropy": 5.495194149017334,
      "epoch": 2.45353894820087,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044006257537118344,
      "loss": 5.0034,
      "mean_token_accuracy": 0.21442195475101472,
      "num_tokens": 56844616.0,
      "step": 24820
    },
    {
      "entropy": 5.456418752670288,
      "epoch": 2.454033214709371,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004400385913572412,
      "loss": 5.0941,
      "mean_token_accuracy": 0.1964623361825943,
      "num_tokens": 56855877.0,
      "step": 24825
    },
    {
      "entropy": 5.507020425796509,
      "epoch": 2.454527481217873,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044001460328332387,
      "loss": 5.1006,
      "mean_token_accuracy": 0.20555953681468964,
      "num_tokens": 56868103.0,
      "step": 24830
    },
    {
      "entropy": 5.523243713378906,
      "epoch": 2.455021747726374,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004399906111500218,
      "loss": 5.21,
      "mean_token_accuracy": 0.19425486028194427,
      "num_tokens": 56878957.0,
      "step": 24835
    },
    {
      "entropy": 5.5970892906188965,
      "epoch": 2.4555160142348753,
      "grad_norm": 1.125,
      "learning_rate": 0.000439966614957925,
      "loss": 5.1359,
      "mean_token_accuracy": 0.19487901479005815,
      "num_tokens": 56891915.0,
      "step": 24840
    },
    {
      "entropy": 5.481054449081421,
      "epoch": 2.456010280743377,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004399426147076239,
      "loss": 5.0808,
      "mean_token_accuracy": 0.20580538511276245,
      "num_tokens": 56902322.0,
      "step": 24845
    },
    {
      "entropy": 5.522133684158325,
      "epoch": 2.4565045472518783,
      "grad_norm": 1.0625,
      "learning_rate": 0.000439918610399709,
      "loss": 5.1613,
      "mean_token_accuracy": 0.1908208027482033,
      "num_tokens": 56913560.0,
      "step": 24850
    },
    {
      "entropy": 5.490698385238647,
      "epoch": 2.4569988137603795,
      "grad_norm": 1.125,
      "learning_rate": 0.00043989460203477053,
      "loss": 5.0835,
      "mean_token_accuracy": 0.20346077531576157,
      "num_tokens": 56924235.0,
      "step": 24855
    },
    {
      "entropy": 5.387242698669434,
      "epoch": 2.457493080268881,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043987058961339937,
      "loss": 4.9477,
      "mean_token_accuracy": 0.21069643795490264,
      "num_tokens": 56935257.0,
      "step": 24860
    },
    {
      "entropy": 5.461850261688232,
      "epoch": 2.4579873467773825,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043984657313618616,
      "loss": 5.1089,
      "mean_token_accuracy": 0.20711364448070527,
      "num_tokens": 56945896.0,
      "step": 24865
    },
    {
      "entropy": 5.526152944564819,
      "epoch": 2.4584816132858838,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004398225526037216,
      "loss": 5.2315,
      "mean_token_accuracy": 0.1939054474234581,
      "num_tokens": 56957274.0,
      "step": 24870
    },
    {
      "entropy": 5.474370336532592,
      "epoch": 2.458975879794385,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043979852801659673,
      "loss": 5.0508,
      "mean_token_accuracy": 0.20531777739524842,
      "num_tokens": 56969647.0,
      "step": 24875
    },
    {
      "entropy": 5.5560894966125485,
      "epoch": 2.4594701463028867,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004397744993754025,
      "loss": 5.0783,
      "mean_token_accuracy": 0.20568324029445648,
      "num_tokens": 56981443.0,
      "step": 24880
    },
    {
      "entropy": 5.455016422271728,
      "epoch": 2.459964412811388,
      "grad_norm": 1.125,
      "learning_rate": 0.00043975046668072996,
      "loss": 5.0613,
      "mean_token_accuracy": 0.21121497750282286,
      "num_tokens": 56992482.0,
      "step": 24885
    },
    {
      "entropy": 5.478019952774048,
      "epoch": 2.4604586793198893,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043972642993317043,
      "loss": 5.1238,
      "mean_token_accuracy": 0.2063273996114731,
      "num_tokens": 57003172.0,
      "step": 24890
    },
    {
      "entropy": 5.553106594085693,
      "epoch": 2.4609529458283905,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043970238913331513,
      "loss": 5.2015,
      "mean_token_accuracy": 0.19956048130989074,
      "num_tokens": 57014763.0,
      "step": 24895
    },
    {
      "entropy": 5.472220802307129,
      "epoch": 2.4614472123368922,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043967834428175565,
      "loss": 5.1008,
      "mean_token_accuracy": 0.21199352741241456,
      "num_tokens": 57025564.0,
      "step": 24900
    },
    {
      "entropy": 5.475388669967652,
      "epoch": 2.4619414788453935,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043965429537908316,
      "loss": 5.0859,
      "mean_token_accuracy": 0.20371516942977905,
      "num_tokens": 57036121.0,
      "step": 24905
    },
    {
      "entropy": 5.504273414611816,
      "epoch": 2.4624357453538948,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004396302424258896,
      "loss": 5.23,
      "mean_token_accuracy": 0.19250302910804748,
      "num_tokens": 57048759.0,
      "step": 24910
    },
    {
      "entropy": 5.4025146007537845,
      "epoch": 2.462930011862396,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004396061854227665,
      "loss": 5.0325,
      "mean_token_accuracy": 0.20790684819221497,
      "num_tokens": 57060791.0,
      "step": 24915
    },
    {
      "entropy": 5.4440961360931395,
      "epoch": 2.4634242783708977,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043958212437030566,
      "loss": 5.0462,
      "mean_token_accuracy": 0.20593809932470322,
      "num_tokens": 57070788.0,
      "step": 24920
    },
    {
      "entropy": 5.49532527923584,
      "epoch": 2.463918544879399,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000439558059269099,
      "loss": 5.1473,
      "mean_token_accuracy": 0.20224593728780746,
      "num_tokens": 57082008.0,
      "step": 24925
    },
    {
      "entropy": 5.496189165115356,
      "epoch": 2.4644128113879002,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004395339901197386,
      "loss": 5.0753,
      "mean_token_accuracy": 0.2049644336104393,
      "num_tokens": 57093081.0,
      "step": 24930
    },
    {
      "entropy": 5.390702390670777,
      "epoch": 2.4649070778964015,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004395099169228166,
      "loss": 5.0253,
      "mean_token_accuracy": 0.20953360497951506,
      "num_tokens": 57103294.0,
      "step": 24935
    },
    {
      "entropy": 5.440811967849731,
      "epoch": 2.465401344404903,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043948583967892495,
      "loss": 5.006,
      "mean_token_accuracy": 0.21172278076410295,
      "num_tokens": 57114509.0,
      "step": 24940
    },
    {
      "entropy": 5.5770281791687015,
      "epoch": 2.4658956109134045,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004394617583886563,
      "loss": 5.1707,
      "mean_token_accuracy": 0.19137718975543977,
      "num_tokens": 57126481.0,
      "step": 24945
    },
    {
      "entropy": 5.461055946350098,
      "epoch": 2.4663898774219057,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004394376730526028,
      "loss": 5.0777,
      "mean_token_accuracy": 0.2014887735247612,
      "num_tokens": 57138235.0,
      "step": 24950
    },
    {
      "entropy": 5.475442743301391,
      "epoch": 2.4668841439304074,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00043941358367135706,
      "loss": 5.0945,
      "mean_token_accuracy": 0.2053341507911682,
      "num_tokens": 57150731.0,
      "step": 24955
    },
    {
      "entropy": 5.49628381729126,
      "epoch": 2.4673784104389087,
      "grad_norm": 1.40625,
      "learning_rate": 0.00043938949024551155,
      "loss": 5.1084,
      "mean_token_accuracy": 0.19958603531122207,
      "num_tokens": 57162729.0,
      "step": 24960
    },
    {
      "entropy": 5.497325372695923,
      "epoch": 2.46787267694741,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043936539277565916,
      "loss": 5.1003,
      "mean_token_accuracy": 0.2059482827782631,
      "num_tokens": 57173855.0,
      "step": 24965
    },
    {
      "entropy": 5.431111669540405,
      "epoch": 2.4683669434559112,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043934129126239256,
      "loss": 5.0562,
      "mean_token_accuracy": 0.20922463536262512,
      "num_tokens": 57184829.0,
      "step": 24970
    },
    {
      "entropy": 5.548756122589111,
      "epoch": 2.468861209964413,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004393171857063048,
      "loss": 5.1462,
      "mean_token_accuracy": 0.20031761080026628,
      "num_tokens": 57195679.0,
      "step": 24975
    },
    {
      "entropy": 5.521723079681396,
      "epoch": 2.469355476472914,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043929307610798867,
      "loss": 5.0809,
      "mean_token_accuracy": 0.20107638835906982,
      "num_tokens": 57208022.0,
      "step": 24980
    },
    {
      "entropy": 5.412291574478149,
      "epoch": 2.4698497429814155,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004392689624680375,
      "loss": 5.0094,
      "mean_token_accuracy": 0.21692616790533065,
      "num_tokens": 57218699.0,
      "step": 24985
    },
    {
      "entropy": 5.460743713378906,
      "epoch": 2.470344009489917,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004392448447870443,
      "loss": 5.1466,
      "mean_token_accuracy": 0.20303266644477844,
      "num_tokens": 57231333.0,
      "step": 24990
    },
    {
      "entropy": 5.563371562957764,
      "epoch": 2.4708382759984184,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004392207230656024,
      "loss": 5.1274,
      "mean_token_accuracy": 0.199388787150383,
      "num_tokens": 57242748.0,
      "step": 24995
    },
    {
      "entropy": 5.479325628280639,
      "epoch": 2.4713325425069197,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004391965973043053,
      "loss": 5.0923,
      "mean_token_accuracy": 0.20689608454704284,
      "num_tokens": 57255438.0,
      "step": 25000
    },
    {
      "entropy": 5.554194355010987,
      "epoch": 2.471826809015421,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004391724675037464,
      "loss": 5.1902,
      "mean_token_accuracy": 0.1978789046406746,
      "num_tokens": 57267472.0,
      "step": 25005
    },
    {
      "entropy": 5.4828859806060795,
      "epoch": 2.4723210755239227,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043914833366451934,
      "loss": 5.0851,
      "mean_token_accuracy": 0.20473093986511232,
      "num_tokens": 57277921.0,
      "step": 25010
    },
    {
      "entropy": 5.46669750213623,
      "epoch": 2.472815342032424,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004391241957872178,
      "loss": 5.0955,
      "mean_token_accuracy": 0.20800813734531404,
      "num_tokens": 57290207.0,
      "step": 25015
    },
    {
      "entropy": 5.494811391830444,
      "epoch": 2.473309608540925,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043910005387243564,
      "loss": 5.0396,
      "mean_token_accuracy": 0.20656308978796006,
      "num_tokens": 57300599.0,
      "step": 25020
    },
    {
      "entropy": 5.477337074279785,
      "epoch": 2.473803875049427,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043907590792076665,
      "loss": 5.163,
      "mean_token_accuracy": 0.19489261507987976,
      "num_tokens": 57313035.0,
      "step": 25025
    },
    {
      "entropy": 5.505033302307129,
      "epoch": 2.474298141557928,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004390517579328048,
      "loss": 5.0978,
      "mean_token_accuracy": 0.2065257340669632,
      "num_tokens": 57325141.0,
      "step": 25030
    },
    {
      "entropy": 5.513743925094604,
      "epoch": 2.4747924080664294,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004390276039091443,
      "loss": 5.131,
      "mean_token_accuracy": 0.20462456345558167,
      "num_tokens": 57336824.0,
      "step": 25035
    },
    {
      "entropy": 5.538777112960815,
      "epoch": 2.4752866745749307,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043900344585037933,
      "loss": 5.1935,
      "mean_token_accuracy": 0.19034961313009263,
      "num_tokens": 57347078.0,
      "step": 25040
    },
    {
      "entropy": 5.459711074829102,
      "epoch": 2.475780941083432,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043897928375710417,
      "loss": 5.0509,
      "mean_token_accuracy": 0.20723038166761398,
      "num_tokens": 57359320.0,
      "step": 25045
    },
    {
      "entropy": 5.523319673538208,
      "epoch": 2.4762752075919336,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004389551176299132,
      "loss": 5.1199,
      "mean_token_accuracy": 0.1993309736251831,
      "num_tokens": 57370642.0,
      "step": 25050
    },
    {
      "entropy": 5.465310955047608,
      "epoch": 2.476769474100435,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004389309474694008,
      "loss": 5.097,
      "mean_token_accuracy": 0.2100017175078392,
      "num_tokens": 57382082.0,
      "step": 25055
    },
    {
      "entropy": 5.47874345779419,
      "epoch": 2.477263740608936,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004389067732761617,
      "loss": 5.1853,
      "mean_token_accuracy": 0.19917632192373275,
      "num_tokens": 57393591.0,
      "step": 25060
    },
    {
      "entropy": 5.44642972946167,
      "epoch": 2.477758007117438,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043888259505079063,
      "loss": 5.0274,
      "mean_token_accuracy": 0.2053271070122719,
      "num_tokens": 57404999.0,
      "step": 25065
    },
    {
      "entropy": 5.47453818321228,
      "epoch": 2.478252273625939,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004388584127938822,
      "loss": 5.1419,
      "mean_token_accuracy": 0.2057407319545746,
      "num_tokens": 57416467.0,
      "step": 25070
    },
    {
      "entropy": 5.556550598144531,
      "epoch": 2.4787465401344404,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004388342265060314,
      "loss": 5.1772,
      "mean_token_accuracy": 0.20119862109422684,
      "num_tokens": 57427201.0,
      "step": 25075
    },
    {
      "entropy": 5.550575399398804,
      "epoch": 2.4792408066429417,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043881003618783326,
      "loss": 5.1619,
      "mean_token_accuracy": 0.19863881915807724,
      "num_tokens": 57438345.0,
      "step": 25080
    },
    {
      "entropy": 5.504270887374878,
      "epoch": 2.4797350731514434,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004387858418398828,
      "loss": 5.1548,
      "mean_token_accuracy": 0.2047478049993515,
      "num_tokens": 57449818.0,
      "step": 25085
    },
    {
      "entropy": 5.515871524810791,
      "epoch": 2.4802293396599446,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004387616434627753,
      "loss": 5.1212,
      "mean_token_accuracy": 0.2020190730690956,
      "num_tokens": 57461961.0,
      "step": 25090
    },
    {
      "entropy": 5.449320268630982,
      "epoch": 2.480723606168446,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043873744105710585,
      "loss": 5.0566,
      "mean_token_accuracy": 0.2140592947602272,
      "num_tokens": 57473422.0,
      "step": 25095
    },
    {
      "entropy": 5.507005786895752,
      "epoch": 2.4812178726769476,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043871323462347,
      "loss": 5.0359,
      "mean_token_accuracy": 0.22032095193862916,
      "num_tokens": 57484618.0,
      "step": 25100
    },
    {
      "entropy": 5.46153450012207,
      "epoch": 2.481712139185449,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004386890241624631,
      "loss": 4.9377,
      "mean_token_accuracy": 0.2145286336541176,
      "num_tokens": 57493521.0,
      "step": 25105
    },
    {
      "entropy": 5.436525344848633,
      "epoch": 2.48220640569395,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004386648096746809,
      "loss": 5.1013,
      "mean_token_accuracy": 0.20619928240776061,
      "num_tokens": 57505693.0,
      "step": 25110
    },
    {
      "entropy": 5.460772132873535,
      "epoch": 2.4827006722024514,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043864059116071897,
      "loss": 5.0549,
      "mean_token_accuracy": 0.21243076026439667,
      "num_tokens": 57517734.0,
      "step": 25115
    },
    {
      "entropy": 5.4907745838165285,
      "epoch": 2.483194938710953,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004386163686211731,
      "loss": 5.1261,
      "mean_token_accuracy": 0.20214656442403794,
      "num_tokens": 57529393.0,
      "step": 25120
    },
    {
      "entropy": 5.397996234893799,
      "epoch": 2.4836892052194544,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043859214205663915,
      "loss": 5.0012,
      "mean_token_accuracy": 0.21629233956336974,
      "num_tokens": 57541109.0,
      "step": 25125
    },
    {
      "entropy": 5.561869525909424,
      "epoch": 2.4841834717279556,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043856791146771327,
      "loss": 5.165,
      "mean_token_accuracy": 0.19821611493825914,
      "num_tokens": 57551869.0,
      "step": 25130
    },
    {
      "entropy": 5.480923461914062,
      "epoch": 2.4846777382364573,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004385436768549912,
      "loss": 5.0822,
      "mean_token_accuracy": 0.2116673082113266,
      "num_tokens": 57564174.0,
      "step": 25135
    },
    {
      "entropy": 5.543782186508179,
      "epoch": 2.4851720047449586,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043851943821906944,
      "loss": 5.1575,
      "mean_token_accuracy": 0.2018676593899727,
      "num_tokens": 57575110.0,
      "step": 25140
    },
    {
      "entropy": 5.452737665176391,
      "epoch": 2.48566627125346,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043849519556054406,
      "loss": 5.0748,
      "mean_token_accuracy": 0.20832909494638444,
      "num_tokens": 57587143.0,
      "step": 25145
    },
    {
      "entropy": 5.418619108200073,
      "epoch": 2.486160537761961,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043847094888001156,
      "loss": 5.0489,
      "mean_token_accuracy": 0.20842526108026505,
      "num_tokens": 57598828.0,
      "step": 25150
    },
    {
      "entropy": 5.545168352127075,
      "epoch": 2.486654804270463,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004384466981780683,
      "loss": 5.1117,
      "mean_token_accuracy": 0.19912556558847427,
      "num_tokens": 57609598.0,
      "step": 25155
    },
    {
      "entropy": 5.507103300094604,
      "epoch": 2.487149070778964,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043842244345531094,
      "loss": 5.0746,
      "mean_token_accuracy": 0.207445028424263,
      "num_tokens": 57620702.0,
      "step": 25160
    },
    {
      "entropy": 5.493671989440918,
      "epoch": 2.4876433372874653,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043839818471233614,
      "loss": 5.1259,
      "mean_token_accuracy": 0.19632306843996047,
      "num_tokens": 57632996.0,
      "step": 25165
    },
    {
      "entropy": 5.4382984161376955,
      "epoch": 2.4881376037959666,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004383739219497406,
      "loss": 5.109,
      "mean_token_accuracy": 0.203692626953125,
      "num_tokens": 57644222.0,
      "step": 25170
    },
    {
      "entropy": 5.450685834884643,
      "epoch": 2.4886318703044683,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043834965516812124,
      "loss": 5.0794,
      "mean_token_accuracy": 0.20437074452638626,
      "num_tokens": 57656399.0,
      "step": 25175
    },
    {
      "entropy": 5.586425685882569,
      "epoch": 2.4891261368129696,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043832538436807506,
      "loss": 5.1498,
      "mean_token_accuracy": 0.19669721722602845,
      "num_tokens": 57668229.0,
      "step": 25180
    },
    {
      "entropy": 5.4174130916595455,
      "epoch": 2.489620403321471,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043830110955019903,
      "loss": 5.0521,
      "mean_token_accuracy": 0.20735956728458405,
      "num_tokens": 57679828.0,
      "step": 25185
    },
    {
      "entropy": 5.515197992324829,
      "epoch": 2.490114669829972,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043827683071509044,
      "loss": 5.1247,
      "mean_token_accuracy": 0.19781017452478408,
      "num_tokens": 57692248.0,
      "step": 25190
    },
    {
      "entropy": 5.516388273239135,
      "epoch": 2.490608936338474,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043825254786334646,
      "loss": 5.1188,
      "mean_token_accuracy": 0.20316905677318572,
      "num_tokens": 57704123.0,
      "step": 25195
    },
    {
      "entropy": 5.403844404220581,
      "epoch": 2.491103202846975,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043822826099556444,
      "loss": 4.9788,
      "mean_token_accuracy": 0.20711625814437867,
      "num_tokens": 57715867.0,
      "step": 25200
    },
    {
      "entropy": 5.547531270980835,
      "epoch": 2.4915974693554763,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043820397011234193,
      "loss": 5.1802,
      "mean_token_accuracy": 0.193709833920002,
      "num_tokens": 57727551.0,
      "step": 25205
    },
    {
      "entropy": 5.504286289215088,
      "epoch": 2.492091735863978,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004381796752142764,
      "loss": 5.0491,
      "mean_token_accuracy": 0.2111051544547081,
      "num_tokens": 57738872.0,
      "step": 25210
    },
    {
      "entropy": 5.411252832412719,
      "epoch": 2.4925860023724793,
      "grad_norm": 1.125,
      "learning_rate": 0.0004381553763019655,
      "loss": 4.9921,
      "mean_token_accuracy": 0.21423764079809188,
      "num_tokens": 57748323.0,
      "step": 25215
    },
    {
      "entropy": 5.469523859024048,
      "epoch": 2.4930802688809806,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043813107337600717,
      "loss": 5.1332,
      "mean_token_accuracy": 0.2021828681230545,
      "num_tokens": 57759697.0,
      "step": 25220
    },
    {
      "entropy": 5.465771865844727,
      "epoch": 2.493574535389482,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000438106766436999,
      "loss": 5.0622,
      "mean_token_accuracy": 0.20388637334108353,
      "num_tokens": 57772532.0,
      "step": 25225
    },
    {
      "entropy": 5.378614711761474,
      "epoch": 2.4940688018979835,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043808245548553916,
      "loss": 5.0229,
      "mean_token_accuracy": 0.21199197471141815,
      "num_tokens": 57783895.0,
      "step": 25230
    },
    {
      "entropy": 5.402005910873413,
      "epoch": 2.494563068406485,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043805814052222554,
      "loss": 5.0323,
      "mean_token_accuracy": 0.20967167168855666,
      "num_tokens": 57795566.0,
      "step": 25235
    },
    {
      "entropy": 5.46422176361084,
      "epoch": 2.495057334914986,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004380338215476564,
      "loss": 5.1338,
      "mean_token_accuracy": 0.2008281096816063,
      "num_tokens": 57807792.0,
      "step": 25240
    },
    {
      "entropy": 5.505619764328003,
      "epoch": 2.4955516014234878,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043800949856242995,
      "loss": 5.0604,
      "mean_token_accuracy": 0.21342795938253403,
      "num_tokens": 57817771.0,
      "step": 25245
    },
    {
      "entropy": 5.525988483428955,
      "epoch": 2.496045867931989,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004379851715671445,
      "loss": 5.1338,
      "mean_token_accuracy": 0.2059691146016121,
      "num_tokens": 57828442.0,
      "step": 25250
    },
    {
      "entropy": 5.5657049179077145,
      "epoch": 2.4965401344404903,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043796084056239864,
      "loss": 5.2022,
      "mean_token_accuracy": 0.1981976345181465,
      "num_tokens": 57840537.0,
      "step": 25255
    },
    {
      "entropy": 5.548855590820312,
      "epoch": 2.4970344009489915,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043793650554879075,
      "loss": 5.2333,
      "mean_token_accuracy": 0.1963667690753937,
      "num_tokens": 57851645.0,
      "step": 25260
    },
    {
      "entropy": 5.454617071151733,
      "epoch": 2.4975286674574932,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004379121665269196,
      "loss": 5.0197,
      "mean_token_accuracy": 0.20639844387769699,
      "num_tokens": 57863812.0,
      "step": 25265
    },
    {
      "entropy": 5.498637437820435,
      "epoch": 2.4980229339659945,
      "grad_norm": 1.515625,
      "learning_rate": 0.00043788782349738387,
      "loss": 5.2147,
      "mean_token_accuracy": 0.19475086331367492,
      "num_tokens": 57876472.0,
      "step": 25270
    },
    {
      "entropy": 5.5039009094238285,
      "epoch": 2.4985172004744958,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004378634764607824,
      "loss": 5.0363,
      "mean_token_accuracy": 0.20650025755167006,
      "num_tokens": 57887905.0,
      "step": 25275
    },
    {
      "entropy": 5.440869140625,
      "epoch": 2.4990114669829975,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004378391254177141,
      "loss": 5.0229,
      "mean_token_accuracy": 0.2094441443681717,
      "num_tokens": 57899084.0,
      "step": 25280
    },
    {
      "entropy": 5.387057638168335,
      "epoch": 2.4995057334914987,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043781477036877807,
      "loss": 5.0569,
      "mean_token_accuracy": 0.2079085662961006,
      "num_tokens": 57910134.0,
      "step": 25285
    },
    {
      "entropy": 5.505024719238281,
      "epoch": 2.5,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004377904113145735,
      "loss": 5.0823,
      "mean_token_accuracy": 0.21174312382936478,
      "num_tokens": 57921957.0,
      "step": 25290
    },
    {
      "entropy": 5.404270601272583,
      "epoch": 2.5004942665085013,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004377660482556995,
      "loss": 5.0239,
      "mean_token_accuracy": 0.21161014437675477,
      "num_tokens": 57932924.0,
      "step": 25295
    },
    {
      "entropy": 5.4153509616851805,
      "epoch": 2.5009885330170025,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043774168119275553,
      "loss": 5.0758,
      "mean_token_accuracy": 0.2098454013466835,
      "num_tokens": 57945275.0,
      "step": 25300
    },
    {
      "entropy": 5.446534633636475,
      "epoch": 2.5014827995255042,
      "grad_norm": 1.296875,
      "learning_rate": 0.00043771731012634085,
      "loss": 5.0935,
      "mean_token_accuracy": 0.21138486713171006,
      "num_tokens": 57957933.0,
      "step": 25305
    },
    {
      "entropy": 5.468757963180542,
      "epoch": 2.5019770660340055,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004376929350570551,
      "loss": 5.0753,
      "mean_token_accuracy": 0.20457536429166795,
      "num_tokens": 57970401.0,
      "step": 25310
    },
    {
      "entropy": 5.535120010375977,
      "epoch": 2.5024713325425068,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000437668555985498,
      "loss": 5.0973,
      "mean_token_accuracy": 0.20259498655796052,
      "num_tokens": 57981433.0,
      "step": 25315
    },
    {
      "entropy": 5.488575649261475,
      "epoch": 2.5029655990510085,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004376441729122691,
      "loss": 5.074,
      "mean_token_accuracy": 0.20786820650100707,
      "num_tokens": 57992288.0,
      "step": 25320
    },
    {
      "entropy": 5.508282947540283,
      "epoch": 2.5034598655595097,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004376197858379684,
      "loss": 5.1707,
      "mean_token_accuracy": 0.2027887687087059,
      "num_tokens": 58004156.0,
      "step": 25325
    },
    {
      "entropy": 5.448185443878174,
      "epoch": 2.503954132068011,
      "grad_norm": 6.09375,
      "learning_rate": 0.0004375953947631957,
      "loss": 5.0545,
      "mean_token_accuracy": 0.21164609044790267,
      "num_tokens": 58014949.0,
      "step": 25330
    },
    {
      "entropy": 5.438953495025634,
      "epoch": 2.5044483985765122,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043757099968855104,
      "loss": 5.0338,
      "mean_token_accuracy": 0.2044109120965004,
      "num_tokens": 58025664.0,
      "step": 25335
    },
    {
      "entropy": 5.480745267868042,
      "epoch": 2.504942665085014,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004375466006146346,
      "loss": 5.1223,
      "mean_token_accuracy": 0.20828879177570342,
      "num_tokens": 58037550.0,
      "step": 25340
    },
    {
      "entropy": 5.4701801300048825,
      "epoch": 2.505436931593515,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004375221975420465,
      "loss": 5.0367,
      "mean_token_accuracy": 0.2033163473010063,
      "num_tokens": 58048292.0,
      "step": 25345
    },
    {
      "entropy": 5.442317199707031,
      "epoch": 2.5059311981020165,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004374977904713872,
      "loss": 5.0713,
      "mean_token_accuracy": 0.2078281044960022,
      "num_tokens": 58060471.0,
      "step": 25350
    },
    {
      "entropy": 5.486228609085083,
      "epoch": 2.506425464610518,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000437473379403257,
      "loss": 5.0606,
      "mean_token_accuracy": 0.2062642306089401,
      "num_tokens": 58072205.0,
      "step": 25355
    },
    {
      "entropy": 5.484630346298218,
      "epoch": 2.5069197311190194,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043744896433825655,
      "loss": 5.0677,
      "mean_token_accuracy": 0.20787267684936522,
      "num_tokens": 58083789.0,
      "step": 25360
    },
    {
      "entropy": 5.45786485671997,
      "epoch": 2.5074139976275207,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004374245452769863,
      "loss": 5.2117,
      "mean_token_accuracy": 0.20523021817207338,
      "num_tokens": 58096064.0,
      "step": 25365
    },
    {
      "entropy": 5.533016157150269,
      "epoch": 2.507908264136022,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043740012222004697,
      "loss": 5.1528,
      "mean_token_accuracy": 0.20319538712501525,
      "num_tokens": 58106935.0,
      "step": 25370
    },
    {
      "entropy": 5.444779109954834,
      "epoch": 2.5084025306445237,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043737569516803955,
      "loss": 5.0765,
      "mean_token_accuracy": 0.20627350956201554,
      "num_tokens": 58119204.0,
      "step": 25375
    },
    {
      "entropy": 5.446405506134033,
      "epoch": 2.508896797153025,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043735126412156484,
      "loss": 5.0601,
      "mean_token_accuracy": 0.20352013707160949,
      "num_tokens": 58132006.0,
      "step": 25380
    },
    {
      "entropy": 5.534933280944824,
      "epoch": 2.509391063661526,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004373268290812238,
      "loss": 5.169,
      "mean_token_accuracy": 0.19522662162780763,
      "num_tokens": 58143845.0,
      "step": 25385
    },
    {
      "entropy": 5.407278394699096,
      "epoch": 2.509885330170028,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043730239004761756,
      "loss": 4.9695,
      "mean_token_accuracy": 0.216986683011055,
      "num_tokens": 58154838.0,
      "step": 25390
    },
    {
      "entropy": 5.549399900436401,
      "epoch": 2.510379596678529,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043727794702134743,
      "loss": 5.1686,
      "mean_token_accuracy": 0.20060989409685134,
      "num_tokens": 58165446.0,
      "step": 25395
    },
    {
      "entropy": 5.3782330513000485,
      "epoch": 2.5108738631870304,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043725350000301455,
      "loss": 4.9852,
      "mean_token_accuracy": 0.2172129347920418,
      "num_tokens": 58177815.0,
      "step": 25400
    },
    {
      "entropy": 5.463104295730591,
      "epoch": 2.5113681296955317,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004372290489932204,
      "loss": 5.0953,
      "mean_token_accuracy": 0.20444085896015168,
      "num_tokens": 58190296.0,
      "step": 25405
    },
    {
      "entropy": 5.480475330352784,
      "epoch": 2.511862396204033,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004372045939925665,
      "loss": 5.0466,
      "mean_token_accuracy": 0.2100695088505745,
      "num_tokens": 58200293.0,
      "step": 25410
    },
    {
      "entropy": 5.416175651550293,
      "epoch": 2.5123566627125347,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004371801350016545,
      "loss": 5.1276,
      "mean_token_accuracy": 0.199514901638031,
      "num_tokens": 58212263.0,
      "step": 25415
    },
    {
      "entropy": 5.579802703857422,
      "epoch": 2.512850929221036,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043715567202108597,
      "loss": 5.146,
      "mean_token_accuracy": 0.20056370198726653,
      "num_tokens": 58222961.0,
      "step": 25420
    },
    {
      "entropy": 5.57877287864685,
      "epoch": 2.5133451957295376,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004371312050514626,
      "loss": 5.1003,
      "mean_token_accuracy": 0.20480694770812988,
      "num_tokens": 58233514.0,
      "step": 25425
    },
    {
      "entropy": 5.460144901275635,
      "epoch": 2.513839462238039,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043710673409338665,
      "loss": 5.0228,
      "mean_token_accuracy": 0.21181508302688598,
      "num_tokens": 58243637.0,
      "step": 25430
    },
    {
      "entropy": 5.412405490875244,
      "epoch": 2.51433372874654,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004370822591474598,
      "loss": 5.0505,
      "mean_token_accuracy": 0.2127783253788948,
      "num_tokens": 58256603.0,
      "step": 25435
    },
    {
      "entropy": 5.396742534637451,
      "epoch": 2.5148279952550414,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004370577802142842,
      "loss": 5.0146,
      "mean_token_accuracy": 0.20932749956846236,
      "num_tokens": 58268112.0,
      "step": 25440
    },
    {
      "entropy": 5.523680925369263,
      "epoch": 2.5153222617635427,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043703329729446213,
      "loss": 5.1644,
      "mean_token_accuracy": 0.20320967584848404,
      "num_tokens": 58281026.0,
      "step": 25445
    },
    {
      "entropy": 5.478973627090454,
      "epoch": 2.5158165282720444,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004370088103885958,
      "loss": 5.1327,
      "mean_token_accuracy": 0.20354149341583253,
      "num_tokens": 58292749.0,
      "step": 25450
    },
    {
      "entropy": 5.5251110076904295,
      "epoch": 2.5163107947805456,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004369843194972876,
      "loss": 5.1493,
      "mean_token_accuracy": 0.19624797701835633,
      "num_tokens": 58304019.0,
      "step": 25455
    },
    {
      "entropy": 5.519231748580933,
      "epoch": 2.516805061289047,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043695982462114,
      "loss": 5.1413,
      "mean_token_accuracy": 0.20545008331537246,
      "num_tokens": 58315998.0,
      "step": 25460
    },
    {
      "entropy": 5.507953643798828,
      "epoch": 2.5172993277975486,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043693532576075556,
      "loss": 5.1076,
      "mean_token_accuracy": 0.19968424141407012,
      "num_tokens": 58327839.0,
      "step": 25465
    },
    {
      "entropy": 5.460997438430786,
      "epoch": 2.51779359430605,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000436910822916737,
      "loss": 5.0473,
      "mean_token_accuracy": 0.20713820308446884,
      "num_tokens": 58339847.0,
      "step": 25470
    },
    {
      "entropy": 5.4520563125610355,
      "epoch": 2.518287860814551,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000436886316089687,
      "loss": 5.0223,
      "mean_token_accuracy": 0.21081761121749878,
      "num_tokens": 58351727.0,
      "step": 25475
    },
    {
      "entropy": 5.397040796279907,
      "epoch": 2.5187821273230524,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043686180528020865,
      "loss": 5.0752,
      "mean_token_accuracy": 0.2062039703130722,
      "num_tokens": 58362519.0,
      "step": 25480
    },
    {
      "entropy": 5.507416820526123,
      "epoch": 2.519276393831554,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004368372904889047,
      "loss": 5.1373,
      "mean_token_accuracy": 0.20928605943918227,
      "num_tokens": 58373747.0,
      "step": 25485
    },
    {
      "entropy": 5.454503345489502,
      "epoch": 2.5197706603400554,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004368127717163783,
      "loss": 5.0417,
      "mean_token_accuracy": 0.20597781985998154,
      "num_tokens": 58385196.0,
      "step": 25490
    },
    {
      "entropy": 5.4363383769989015,
      "epoch": 2.5202649268485566,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004367882489632327,
      "loss": 5.1329,
      "mean_token_accuracy": 0.20246631950139998,
      "num_tokens": 58397005.0,
      "step": 25495
    },
    {
      "entropy": 5.362036323547363,
      "epoch": 2.5207591933570583,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043676372223007095,
      "loss": 5.0151,
      "mean_token_accuracy": 0.21005190461874007,
      "num_tokens": 58407685.0,
      "step": 25500
    },
    {
      "entropy": 5.418642473220825,
      "epoch": 2.5212534598655596,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004367391915174966,
      "loss": 5.0683,
      "mean_token_accuracy": 0.20876320004463195,
      "num_tokens": 58419250.0,
      "step": 25505
    },
    {
      "entropy": 5.535060787200928,
      "epoch": 2.521747726374061,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004367146568261131,
      "loss": 5.0959,
      "mean_token_accuracy": 0.20361196994781494,
      "num_tokens": 58431135.0,
      "step": 25510
    },
    {
      "entropy": 5.5453839778900145,
      "epoch": 2.522241992882562,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004366901181565239,
      "loss": 5.1526,
      "mean_token_accuracy": 0.19703492969274522,
      "num_tokens": 58443865.0,
      "step": 25515
    },
    {
      "entropy": 5.531009864807129,
      "epoch": 2.5227362593910634,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043666557550933266,
      "loss": 5.1433,
      "mean_token_accuracy": 0.2097615659236908,
      "num_tokens": 58456484.0,
      "step": 25520
    },
    {
      "entropy": 5.4832288265228275,
      "epoch": 2.523230525899565,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004366410288851433,
      "loss": 5.1147,
      "mean_token_accuracy": 0.20213446915149688,
      "num_tokens": 58468384.0,
      "step": 25525
    },
    {
      "entropy": 5.442923259735108,
      "epoch": 2.5237247924080664,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043661647828455944,
      "loss": 5.0743,
      "mean_token_accuracy": 0.20265950113534928,
      "num_tokens": 58479703.0,
      "step": 25530
    },
    {
      "entropy": 5.5766970157623295,
      "epoch": 2.524219058916568,
      "grad_norm": 1.125,
      "learning_rate": 0.0004365919237081852,
      "loss": 5.2102,
      "mean_token_accuracy": 0.19806513786315919,
      "num_tokens": 58491313.0,
      "step": 25535
    },
    {
      "entropy": 5.525884294509888,
      "epoch": 2.5247133254250693,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043656736515662457,
      "loss": 5.1232,
      "mean_token_accuracy": 0.1975715383887291,
      "num_tokens": 58503323.0,
      "step": 25540
    },
    {
      "entropy": 5.483131456375122,
      "epoch": 2.5252075919335706,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004365428026304816,
      "loss": 5.0568,
      "mean_token_accuracy": 0.210589799284935,
      "num_tokens": 58514935.0,
      "step": 25545
    },
    {
      "entropy": 5.620908164978028,
      "epoch": 2.525701858442072,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004365182361303607,
      "loss": 5.2412,
      "mean_token_accuracy": 0.19156759679317475,
      "num_tokens": 58525724.0,
      "step": 25550
    },
    {
      "entropy": 5.5009866714477536,
      "epoch": 2.526196124950573,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043649366565686615,
      "loss": 5.1263,
      "mean_token_accuracy": 0.2070714831352234,
      "num_tokens": 58536868.0,
      "step": 25555
    },
    {
      "entropy": 5.443649673461914,
      "epoch": 2.526690391459075,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043646909121060237,
      "loss": 5.1073,
      "mean_token_accuracy": 0.20141696631908418,
      "num_tokens": 58547285.0,
      "step": 25560
    },
    {
      "entropy": 5.5117950439453125,
      "epoch": 2.527184657967576,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004364445127921738,
      "loss": 5.1133,
      "mean_token_accuracy": 0.20147760659456254,
      "num_tokens": 58559960.0,
      "step": 25565
    },
    {
      "entropy": 5.536963081359863,
      "epoch": 2.5276789244760773,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004364199304021853,
      "loss": 5.071,
      "mean_token_accuracy": 0.2032789409160614,
      "num_tokens": 58571229.0,
      "step": 25570
    },
    {
      "entropy": 5.477812957763672,
      "epoch": 2.528173190984579,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004363953440412414,
      "loss": 5.1563,
      "mean_token_accuracy": 0.20263191908597947,
      "num_tokens": 58580988.0,
      "step": 25575
    },
    {
      "entropy": 5.481512832641601,
      "epoch": 2.5286674574930803,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000436370753709947,
      "loss": 5.1044,
      "mean_token_accuracy": 0.207915036380291,
      "num_tokens": 58592547.0,
      "step": 25580
    },
    {
      "entropy": 5.525057554244995,
      "epoch": 2.5291617240015816,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004363461594089071,
      "loss": 5.1423,
      "mean_token_accuracy": 0.20642145872116088,
      "num_tokens": 58604820.0,
      "step": 25585
    },
    {
      "entropy": 5.452645254135132,
      "epoch": 2.529655990510083,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043632156113872656,
      "loss": 5.0395,
      "mean_token_accuracy": 0.21131696850061416,
      "num_tokens": 58615635.0,
      "step": 25590
    },
    {
      "entropy": 5.507434940338134,
      "epoch": 2.5301502570185845,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043629695890001064,
      "loss": 5.0877,
      "mean_token_accuracy": 0.20697855055332184,
      "num_tokens": 58628799.0,
      "step": 25595
    },
    {
      "entropy": 5.431208372116089,
      "epoch": 2.530644523527086,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004362723526933645,
      "loss": 5.0812,
      "mean_token_accuracy": 0.2099124938249588,
      "num_tokens": 58641514.0,
      "step": 25600
    },
    {
      "entropy": 5.393680763244629,
      "epoch": 2.531138790035587,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004362477425193934,
      "loss": 5.0277,
      "mean_token_accuracy": 0.21237764060497283,
      "num_tokens": 58651719.0,
      "step": 25605
    },
    {
      "entropy": 5.52293643951416,
      "epoch": 2.5316330565440888,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004362231283787029,
      "loss": 5.0727,
      "mean_token_accuracy": 0.2066674590110779,
      "num_tokens": 58663032.0,
      "step": 25610
    },
    {
      "entropy": 5.39792103767395,
      "epoch": 2.53212732305259,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004361985102718984,
      "loss": 5.007,
      "mean_token_accuracy": 0.21407516300678253,
      "num_tokens": 58674188.0,
      "step": 25615
    },
    {
      "entropy": 5.392745447158814,
      "epoch": 2.5326215895610913,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043617388819958554,
      "loss": 4.97,
      "mean_token_accuracy": 0.2179355338215828,
      "num_tokens": 58684892.0,
      "step": 25620
    },
    {
      "entropy": 5.408490800857544,
      "epoch": 2.5331158560695926,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043614926216237,
      "loss": 5.0126,
      "mean_token_accuracy": 0.21216772347688675,
      "num_tokens": 58695520.0,
      "step": 25625
    },
    {
      "entropy": 5.450341463088989,
      "epoch": 2.5336101225780943,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004361246321608577,
      "loss": 5.0068,
      "mean_token_accuracy": 0.21002770513296126,
      "num_tokens": 58706579.0,
      "step": 25630
    },
    {
      "entropy": 5.502708530426025,
      "epoch": 2.5341043890865955,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004360999981956544,
      "loss": 5.1039,
      "mean_token_accuracy": 0.20209251791238786,
      "num_tokens": 58717292.0,
      "step": 25635
    },
    {
      "entropy": 5.473971319198609,
      "epoch": 2.534598655595097,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043607536026736603,
      "loss": 5.131,
      "mean_token_accuracy": 0.19884008467197417,
      "num_tokens": 58728761.0,
      "step": 25640
    },
    {
      "entropy": 5.470366907119751,
      "epoch": 2.5350929221035985,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000436050718376599,
      "loss": 4.9813,
      "mean_token_accuracy": 0.2147522062063217,
      "num_tokens": 58739494.0,
      "step": 25645
    },
    {
      "entropy": 5.528844547271729,
      "epoch": 2.5355871886120998,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004360260725239592,
      "loss": 5.1839,
      "mean_token_accuracy": 0.20093917101621628,
      "num_tokens": 58750119.0,
      "step": 25650
    },
    {
      "entropy": 5.47432050704956,
      "epoch": 2.536081455120601,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043600142271005304,
      "loss": 5.0427,
      "mean_token_accuracy": 0.2014262318611145,
      "num_tokens": 58760438.0,
      "step": 25655
    },
    {
      "entropy": 5.518746757507325,
      "epoch": 2.5365757216291023,
      "grad_norm": 1.078125,
      "learning_rate": 0.000435976768935487,
      "loss": 5.1483,
      "mean_token_accuracy": 0.1982058048248291,
      "num_tokens": 58771484.0,
      "step": 25660
    },
    {
      "entropy": 5.504497051239014,
      "epoch": 2.5370699881376035,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004359521112008674,
      "loss": 5.0963,
      "mean_token_accuracy": 0.20272475928068162,
      "num_tokens": 58782716.0,
      "step": 25665
    },
    {
      "entropy": 5.517631769180298,
      "epoch": 2.5375642546461052,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004359274495068009,
      "loss": 5.1422,
      "mean_token_accuracy": 0.20292294919490814,
      "num_tokens": 58793557.0,
      "step": 25670
    },
    {
      "entropy": 5.445505905151367,
      "epoch": 2.5380585211546065,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004359027838538941,
      "loss": 5.0566,
      "mean_token_accuracy": 0.2045275539159775,
      "num_tokens": 58804225.0,
      "step": 25675
    },
    {
      "entropy": 5.440919351577759,
      "epoch": 2.538552787663108,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000435878114242754,
      "loss": 5.0209,
      "mean_token_accuracy": 0.21195252537727355,
      "num_tokens": 58814368.0,
      "step": 25680
    },
    {
      "entropy": 5.464352607727051,
      "epoch": 2.5390470541716095,
      "grad_norm": 1.125,
      "learning_rate": 0.00043585344067398725,
      "loss": 5.0326,
      "mean_token_accuracy": 0.21023503839969634,
      "num_tokens": 58825925.0,
      "step": 25685
    },
    {
      "entropy": 5.52657790184021,
      "epoch": 2.5395413206801107,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004358287631482009,
      "loss": 5.134,
      "mean_token_accuracy": 0.19989823997020723,
      "num_tokens": 58837334.0,
      "step": 25690
    },
    {
      "entropy": 5.507832813262939,
      "epoch": 2.540035587188612,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043580408166600206,
      "loss": 5.0708,
      "mean_token_accuracy": 0.20861420929431915,
      "num_tokens": 58848045.0,
      "step": 25695
    },
    {
      "entropy": 5.462322854995728,
      "epoch": 2.5405298536971133,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043577939622799786,
      "loss": 5.0475,
      "mean_token_accuracy": 0.21379576176404952,
      "num_tokens": 58858186.0,
      "step": 25700
    },
    {
      "entropy": 5.409792137145996,
      "epoch": 2.541024120205615,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004357547068347955,
      "loss": 5.011,
      "mean_token_accuracy": 0.21509841233491897,
      "num_tokens": 58870102.0,
      "step": 25705
    },
    {
      "entropy": 5.525062131881714,
      "epoch": 2.5415183867141162,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004357300134870025,
      "loss": 5.0979,
      "mean_token_accuracy": 0.1990061953663826,
      "num_tokens": 58881246.0,
      "step": 25710
    },
    {
      "entropy": 5.527791547775268,
      "epoch": 2.5420126532226175,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004357053161852262,
      "loss": 5.1938,
      "mean_token_accuracy": 0.1975237935781479,
      "num_tokens": 58893722.0,
      "step": 25715
    },
    {
      "entropy": 5.439000606536865,
      "epoch": 2.542506919731119,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004356806149300742,
      "loss": 4.9523,
      "mean_token_accuracy": 0.2172499492764473,
      "num_tokens": 58904703.0,
      "step": 25720
    },
    {
      "entropy": 5.37617244720459,
      "epoch": 2.5430011862396205,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004356559097221541,
      "loss": 4.9975,
      "mean_token_accuracy": 0.21296435445547104,
      "num_tokens": 58916626.0,
      "step": 25725
    },
    {
      "entropy": 5.451753091812134,
      "epoch": 2.5434954527481217,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043563120056207375,
      "loss": 5.125,
      "mean_token_accuracy": 0.2034957453608513,
      "num_tokens": 58927943.0,
      "step": 25730
    },
    {
      "entropy": 5.577938032150269,
      "epoch": 2.543989719256623,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004356064874504409,
      "loss": 5.1738,
      "mean_token_accuracy": 0.20399570614099502,
      "num_tokens": 58940086.0,
      "step": 25735
    },
    {
      "entropy": 5.550048303604126,
      "epoch": 2.5444839857651247,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004355817703878637,
      "loss": 5.0875,
      "mean_token_accuracy": 0.20873271822929382,
      "num_tokens": 58952280.0,
      "step": 25740
    },
    {
      "entropy": 5.520841217041015,
      "epoch": 2.544978252273626,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004355570493749498,
      "loss": 5.108,
      "mean_token_accuracy": 0.20368714034557342,
      "num_tokens": 58963726.0,
      "step": 25745
    },
    {
      "entropy": 5.477183151245117,
      "epoch": 2.545472518782127,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004355323244123077,
      "loss": 5.105,
      "mean_token_accuracy": 0.20496445000171662,
      "num_tokens": 58974941.0,
      "step": 25750
    },
    {
      "entropy": 5.446059703826904,
      "epoch": 2.545966785290629,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043550759550054554,
      "loss": 5.0849,
      "mean_token_accuracy": 0.20758278220891951,
      "num_tokens": 58985234.0,
      "step": 25755
    },
    {
      "entropy": 5.441429471969604,
      "epoch": 2.54646105179913,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043548286264027155,
      "loss": 5.0062,
      "mean_token_accuracy": 0.2124189779162407,
      "num_tokens": 58996781.0,
      "step": 25760
    },
    {
      "entropy": 5.4362006187438965,
      "epoch": 2.5469553183076314,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043545812583209426,
      "loss": 5.0581,
      "mean_token_accuracy": 0.20466673225164414,
      "num_tokens": 59008044.0,
      "step": 25765
    },
    {
      "entropy": 5.455527400970459,
      "epoch": 2.5474495848161327,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043543338507662227,
      "loss": 5.1323,
      "mean_token_accuracy": 0.20451315343379975,
      "num_tokens": 59021579.0,
      "step": 25770
    },
    {
      "entropy": 5.528790092468261,
      "epoch": 2.547943851324634,
      "grad_norm": 1.125,
      "learning_rate": 0.0004354086403744641,
      "loss": 5.1221,
      "mean_token_accuracy": 0.20080033540725709,
      "num_tokens": 59033541.0,
      "step": 25775
    },
    {
      "entropy": 5.470633602142334,
      "epoch": 2.5484381178331357,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004353838917262284,
      "loss": 5.0637,
      "mean_token_accuracy": 0.20321372300386428,
      "num_tokens": 59044067.0,
      "step": 25780
    },
    {
      "entropy": 5.359145307540894,
      "epoch": 2.548932384341637,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004353591391325241,
      "loss": 4.9939,
      "mean_token_accuracy": 0.21232887208461762,
      "num_tokens": 59056758.0,
      "step": 25785
    },
    {
      "entropy": 5.3897607803344725,
      "epoch": 2.5494266508501386,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043533438259396015,
      "loss": 4.9958,
      "mean_token_accuracy": 0.21601679027080536,
      "num_tokens": 59067681.0,
      "step": 25790
    },
    {
      "entropy": 5.483454275131225,
      "epoch": 2.54992091735864,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043530962211114547,
      "loss": 5.0963,
      "mean_token_accuracy": 0.21015080362558364,
      "num_tokens": 59079148.0,
      "step": 25795
    },
    {
      "entropy": 5.516743135452271,
      "epoch": 2.550415183867141,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004352848576846893,
      "loss": 5.1451,
      "mean_token_accuracy": 0.20016193985939026,
      "num_tokens": 59090597.0,
      "step": 25800
    },
    {
      "entropy": 5.43180570602417,
      "epoch": 2.5509094503756424,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00043526008931520067,
      "loss": 5.0112,
      "mean_token_accuracy": 0.21306267976760865,
      "num_tokens": 59102445.0,
      "step": 25805
    },
    {
      "entropy": 5.475431108474732,
      "epoch": 2.5514037168841437,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000435235317003289,
      "loss": 5.1398,
      "mean_token_accuracy": 0.1937178537249565,
      "num_tokens": 59113470.0,
      "step": 25810
    },
    {
      "entropy": 5.398521852493286,
      "epoch": 2.5518979833926454,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043521054074956367,
      "loss": 4.9623,
      "mean_token_accuracy": 0.2216824099421501,
      "num_tokens": 59124370.0,
      "step": 25815
    },
    {
      "entropy": 5.497758388519287,
      "epoch": 2.5523922499011467,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043518576055463423,
      "loss": 5.1437,
      "mean_token_accuracy": 0.2056589886546135,
      "num_tokens": 59135783.0,
      "step": 25820
    },
    {
      "entropy": 5.484558820724487,
      "epoch": 2.552886516409648,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004351609764191102,
      "loss": 5.0883,
      "mean_token_accuracy": 0.20761982798576356,
      "num_tokens": 59146729.0,
      "step": 25825
    },
    {
      "entropy": 5.561499691009521,
      "epoch": 2.5533807829181496,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004351361883436014,
      "loss": 5.1488,
      "mean_token_accuracy": 0.20425196141004562,
      "num_tokens": 59159413.0,
      "step": 25830
    },
    {
      "entropy": 5.431789302825928,
      "epoch": 2.553875049426651,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004351113963287173,
      "loss": 5.0332,
      "mean_token_accuracy": 0.20214271396398545,
      "num_tokens": 59171013.0,
      "step": 25835
    },
    {
      "entropy": 5.48479323387146,
      "epoch": 2.554369315935152,
      "grad_norm": 1.0,
      "learning_rate": 0.00043508660037506813,
      "loss": 5.105,
      "mean_token_accuracy": 0.20411466658115388,
      "num_tokens": 59182571.0,
      "step": 25840
    },
    {
      "entropy": 5.436587142944336,
      "epoch": 2.5548635824436534,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004350618004832638,
      "loss": 5.0124,
      "mean_token_accuracy": 0.2166304498910904,
      "num_tokens": 59194310.0,
      "step": 25845
    },
    {
      "entropy": 5.472873210906982,
      "epoch": 2.555357848952155,
      "grad_norm": 1.0,
      "learning_rate": 0.0004350369966539143,
      "loss": 5.0917,
      "mean_token_accuracy": 0.2050582692027092,
      "num_tokens": 59206656.0,
      "step": 25850
    },
    {
      "entropy": 5.512592363357544,
      "epoch": 2.5558521154606564,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043501218888762985,
      "loss": 5.1855,
      "mean_token_accuracy": 0.20180006474256515,
      "num_tokens": 59219180.0,
      "step": 25855
    },
    {
      "entropy": 5.413117742538452,
      "epoch": 2.5563463819691576,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004349873771850208,
      "loss": 5.0288,
      "mean_token_accuracy": 0.21067819446325303,
      "num_tokens": 59230879.0,
      "step": 25860
    },
    {
      "entropy": 5.576191663742065,
      "epoch": 2.5568406484776594,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004349625615466974,
      "loss": 5.1423,
      "mean_token_accuracy": 0.19706527143716812,
      "num_tokens": 59242283.0,
      "step": 25865
    },
    {
      "entropy": 5.427069187164307,
      "epoch": 2.5573349149861606,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043493774197327014,
      "loss": 4.999,
      "mean_token_accuracy": 0.21658221185207366,
      "num_tokens": 59253655.0,
      "step": 25870
    },
    {
      "entropy": 5.489561891555786,
      "epoch": 2.557829181494662,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043491291846534965,
      "loss": 5.0811,
      "mean_token_accuracy": 0.19935698956251144,
      "num_tokens": 59266600.0,
      "step": 25875
    },
    {
      "entropy": 5.545461559295655,
      "epoch": 2.558323448003163,
      "grad_norm": 1.125,
      "learning_rate": 0.00043488809102354653,
      "loss": 5.1603,
      "mean_token_accuracy": 0.19824885725975036,
      "num_tokens": 59277422.0,
      "step": 25880
    },
    {
      "entropy": 5.420630836486817,
      "epoch": 2.558817714511665,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043486325964847163,
      "loss": 5.0252,
      "mean_token_accuracy": 0.21108732372522354,
      "num_tokens": 59289179.0,
      "step": 25885
    },
    {
      "entropy": 5.38097128868103,
      "epoch": 2.559311981020166,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00043483842434073567,
      "loss": 4.9641,
      "mean_token_accuracy": 0.22336167544126512,
      "num_tokens": 59299692.0,
      "step": 25890
    },
    {
      "entropy": 5.481780576705932,
      "epoch": 2.5598062475286674,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004348135851009497,
      "loss": 5.0482,
      "mean_token_accuracy": 0.203604955971241,
      "num_tokens": 59310797.0,
      "step": 25895
    },
    {
      "entropy": 5.3913411617279055,
      "epoch": 2.560300514037169,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004347887419297247,
      "loss": 5.0623,
      "mean_token_accuracy": 0.20897647589445115,
      "num_tokens": 59323167.0,
      "step": 25900
    },
    {
      "entropy": 5.419101095199585,
      "epoch": 2.5607947805456703,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043476389482767193,
      "loss": 5.0867,
      "mean_token_accuracy": 0.20849373936653137,
      "num_tokens": 59335956.0,
      "step": 25905
    },
    {
      "entropy": 5.551991510391235,
      "epoch": 2.5612890470541716,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004347390437954026,
      "loss": 5.1173,
      "mean_token_accuracy": 0.20107119530439377,
      "num_tokens": 59348090.0,
      "step": 25910
    },
    {
      "entropy": 5.447291660308838,
      "epoch": 2.561783313562673,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004347141888335279,
      "loss": 5.0887,
      "mean_token_accuracy": 0.2058752343058586,
      "num_tokens": 59360276.0,
      "step": 25915
    },
    {
      "entropy": 5.499720335006714,
      "epoch": 2.562277580071174,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004346893299426595,
      "loss": 5.1277,
      "mean_token_accuracy": 0.19745977371931075,
      "num_tokens": 59372049.0,
      "step": 25920
    },
    {
      "entropy": 5.5342559814453125,
      "epoch": 2.562771846579676,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043466446712340866,
      "loss": 5.0763,
      "mean_token_accuracy": 0.2055215984582901,
      "num_tokens": 59384017.0,
      "step": 25925
    },
    {
      "entropy": 5.546181678771973,
      "epoch": 2.563266113088177,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004346396003763873,
      "loss": 5.1655,
      "mean_token_accuracy": 0.19704237431287766,
      "num_tokens": 59395966.0,
      "step": 25930
    },
    {
      "entropy": 5.436600971221924,
      "epoch": 2.563760379596679,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004346147297022069,
      "loss": 5.0676,
      "mean_token_accuracy": 0.20828012526035308,
      "num_tokens": 59405815.0,
      "step": 25935
    },
    {
      "entropy": 5.47698941230774,
      "epoch": 2.56425464610518,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043458985510147946,
      "loss": 5.074,
      "mean_token_accuracy": 0.20358948707580565,
      "num_tokens": 59416662.0,
      "step": 25940
    },
    {
      "entropy": 5.521800994873047,
      "epoch": 2.5647489126136813,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004345649765748168,
      "loss": 5.1075,
      "mean_token_accuracy": 0.20760411620140076,
      "num_tokens": 59427143.0,
      "step": 25945
    },
    {
      "entropy": 5.4276646137237545,
      "epoch": 2.5652431791221826,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043454009412283097,
      "loss": 5.0809,
      "mean_token_accuracy": 0.20239093601703645,
      "num_tokens": 59439657.0,
      "step": 25950
    },
    {
      "entropy": 5.483403158187866,
      "epoch": 2.565737445630684,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043451520774613405,
      "loss": 5.0886,
      "mean_token_accuracy": 0.20373429358005524,
      "num_tokens": 59452313.0,
      "step": 25955
    },
    {
      "entropy": 5.437200307846069,
      "epoch": 2.5662317121391856,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043449031744533824,
      "loss": 5.0041,
      "mean_token_accuracy": 0.2102823704481125,
      "num_tokens": 59462412.0,
      "step": 25960
    },
    {
      "entropy": 5.491439437866211,
      "epoch": 2.566725978647687,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00043446542322105595,
      "loss": 5.0782,
      "mean_token_accuracy": 0.20713732689619063,
      "num_tokens": 59474784.0,
      "step": 25965
    },
    {
      "entropy": 5.420276355743408,
      "epoch": 2.567220245156188,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004344405250738994,
      "loss": 4.9843,
      "mean_token_accuracy": 0.21068878173828126,
      "num_tokens": 59487006.0,
      "step": 25970
    },
    {
      "entropy": 5.44613208770752,
      "epoch": 2.56771451166469,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043441562300448134,
      "loss": 5.0726,
      "mean_token_accuracy": 0.20465949177742004,
      "num_tokens": 59498625.0,
      "step": 25975
    },
    {
      "entropy": 5.413465929031372,
      "epoch": 2.568208778173191,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004343907170134142,
      "loss": 4.954,
      "mean_token_accuracy": 0.21782627254724501,
      "num_tokens": 59509257.0,
      "step": 25980
    },
    {
      "entropy": 5.471138620376587,
      "epoch": 2.5687030446816923,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004343658071013106,
      "loss": 5.1869,
      "mean_token_accuracy": 0.19876190423965454,
      "num_tokens": 59520876.0,
      "step": 25985
    },
    {
      "entropy": 5.442031145095825,
      "epoch": 2.5691973111901936,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043434089326878356,
      "loss": 5.0171,
      "mean_token_accuracy": 0.20690550804138183,
      "num_tokens": 59532047.0,
      "step": 25990
    },
    {
      "entropy": 5.464065408706665,
      "epoch": 2.5696915776986953,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004343159755164457,
      "loss": 5.103,
      "mean_token_accuracy": 0.20479633957147597,
      "num_tokens": 59543831.0,
      "step": 25995
    },
    {
      "entropy": 5.433709287643433,
      "epoch": 2.5701858442071965,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004342910538449102,
      "loss": 4.9804,
      "mean_token_accuracy": 0.2126169756054878,
      "num_tokens": 59555298.0,
      "step": 26000
    },
    {
      "entropy": 5.512145757675171,
      "epoch": 2.570680110715698,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043426612825479003,
      "loss": 5.0807,
      "mean_token_accuracy": 0.20384642630815505,
      "num_tokens": 59566204.0,
      "step": 26005
    },
    {
      "entropy": 5.4690773487091064,
      "epoch": 2.5711743772241995,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004342411987466984,
      "loss": 5.1668,
      "mean_token_accuracy": 0.19438787847757338,
      "num_tokens": 59577505.0,
      "step": 26010
    },
    {
      "entropy": 5.505500650405883,
      "epoch": 2.5716686437327008,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004342162653212486,
      "loss": 5.1348,
      "mean_token_accuracy": 0.20243036299943923,
      "num_tokens": 59589430.0,
      "step": 26015
    },
    {
      "entropy": 5.4255115509033205,
      "epoch": 2.572162910241202,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004341913279790539,
      "loss": 5.0199,
      "mean_token_accuracy": 0.21482257395982743,
      "num_tokens": 59600328.0,
      "step": 26020
    },
    {
      "entropy": 5.4620726108551025,
      "epoch": 2.5726571767497033,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004341663867207279,
      "loss": 5.0698,
      "mean_token_accuracy": 0.20392953604459763,
      "num_tokens": 59611643.0,
      "step": 26025
    },
    {
      "entropy": 5.510752820968628,
      "epoch": 2.5731514432582046,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043414144154688405,
      "loss": 5.1701,
      "mean_token_accuracy": 0.208053095638752,
      "num_tokens": 59622586.0,
      "step": 26030
    },
    {
      "entropy": 5.577480125427246,
      "epoch": 2.5736457097667063,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043411649245813613,
      "loss": 5.1597,
      "mean_token_accuracy": 0.20617804527282715,
      "num_tokens": 59632719.0,
      "step": 26035
    },
    {
      "entropy": 5.47836856842041,
      "epoch": 2.5741399762752075,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004340915394550977,
      "loss": 5.024,
      "mean_token_accuracy": 0.21188221722841263,
      "num_tokens": 59643788.0,
      "step": 26040
    },
    {
      "entropy": 5.448640537261963,
      "epoch": 2.5746342427837092,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004340665825383828,
      "loss": 5.0571,
      "mean_token_accuracy": 0.21054611951112748,
      "num_tokens": 59655355.0,
      "step": 26045
    },
    {
      "entropy": 5.530148649215699,
      "epoch": 2.5751285092922105,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043404162170860525,
      "loss": 5.1857,
      "mean_token_accuracy": 0.19971430003643037,
      "num_tokens": 59667099.0,
      "step": 26050
    },
    {
      "entropy": 5.464069175720215,
      "epoch": 2.5756227758007118,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004340166569663792,
      "loss": 5.1498,
      "mean_token_accuracy": 0.19777952879667282,
      "num_tokens": 59679514.0,
      "step": 26055
    },
    {
      "entropy": 5.6031379222869875,
      "epoch": 2.576117042309213,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043399168831231865,
      "loss": 5.218,
      "mean_token_accuracy": 0.19847166091203688,
      "num_tokens": 59691569.0,
      "step": 26060
    },
    {
      "entropy": 5.5251699924469,
      "epoch": 2.5766113088177143,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004339667157470379,
      "loss": 5.145,
      "mean_token_accuracy": 0.2035716250538826,
      "num_tokens": 59702841.0,
      "step": 26065
    },
    {
      "entropy": 5.4570536613464355,
      "epoch": 2.577105575326216,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004339417392711513,
      "loss": 5.0584,
      "mean_token_accuracy": 0.20833969712257386,
      "num_tokens": 59714834.0,
      "step": 26070
    },
    {
      "entropy": 5.525702953338623,
      "epoch": 2.5775998418347172,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004339167588852733,
      "loss": 5.122,
      "mean_token_accuracy": 0.1981703370809555,
      "num_tokens": 59726863.0,
      "step": 26075
    },
    {
      "entropy": 5.468962955474853,
      "epoch": 2.5780941083432185,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004338917745900183,
      "loss": 5.1427,
      "mean_token_accuracy": 0.20276612937450408,
      "num_tokens": 59738952.0,
      "step": 26080
    },
    {
      "entropy": 5.452627944946289,
      "epoch": 2.57858837485172,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004338667863860011,
      "loss": 5.054,
      "mean_token_accuracy": 0.2119163081049919,
      "num_tokens": 59750445.0,
      "step": 26085
    },
    {
      "entropy": 5.4511988162994385,
      "epoch": 2.5790826413602215,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004338417942738362,
      "loss": 5.0739,
      "mean_token_accuracy": 0.20618994683027267,
      "num_tokens": 59762447.0,
      "step": 26090
    },
    {
      "entropy": 5.394779300689697,
      "epoch": 2.5795769078687227,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004338167982541386,
      "loss": 4.9822,
      "mean_token_accuracy": 0.21389454156160354,
      "num_tokens": 59773122.0,
      "step": 26095
    },
    {
      "entropy": 5.560962152481079,
      "epoch": 2.580071174377224,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043379179832752306,
      "loss": 5.1372,
      "mean_token_accuracy": 0.20742413103580476,
      "num_tokens": 59784355.0,
      "step": 26100
    },
    {
      "entropy": 5.464340782165527,
      "epoch": 2.5805654408857257,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043376679449460463,
      "loss": 4.9969,
      "mean_token_accuracy": 0.2158476710319519,
      "num_tokens": 59796529.0,
      "step": 26105
    },
    {
      "entropy": 5.466800260543823,
      "epoch": 2.581059707394227,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043374178675599853,
      "loss": 5.059,
      "mean_token_accuracy": 0.20189632177352906,
      "num_tokens": 59807510.0,
      "step": 26110
    },
    {
      "entropy": 5.449657154083252,
      "epoch": 2.5815539739027282,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043371677511231977,
      "loss": 5.0751,
      "mean_token_accuracy": 0.20399401634931563,
      "num_tokens": 59819355.0,
      "step": 26115
    },
    {
      "entropy": 5.434161186218262,
      "epoch": 2.58204824041123,
      "grad_norm": 1.125,
      "learning_rate": 0.0004336917595641838,
      "loss": 5.0438,
      "mean_token_accuracy": 0.2003992572426796,
      "num_tokens": 59830780.0,
      "step": 26120
    },
    {
      "entropy": 5.452296018600464,
      "epoch": 2.582542506919731,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004336667401122058,
      "loss": 5.1161,
      "mean_token_accuracy": 0.2084857478737831,
      "num_tokens": 59842031.0,
      "step": 26125
    },
    {
      "entropy": 5.516363096237183,
      "epoch": 2.5830367734282325,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004336417167570015,
      "loss": 5.0795,
      "mean_token_accuracy": 0.2099551409482956,
      "num_tokens": 59852788.0,
      "step": 26130
    },
    {
      "entropy": 5.4264037132263185,
      "epoch": 2.5835310399367337,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043361668949918627,
      "loss": 5.0877,
      "mean_token_accuracy": 0.2121927633881569,
      "num_tokens": 59865618.0,
      "step": 26135
    },
    {
      "entropy": 5.457371187210083,
      "epoch": 2.5840253064452354,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004335916583393759,
      "loss": 5.0609,
      "mean_token_accuracy": 0.20406586974859237,
      "num_tokens": 59877053.0,
      "step": 26140
    },
    {
      "entropy": 5.4266472339630125,
      "epoch": 2.5845195729537367,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004335666232781861,
      "loss": 4.9977,
      "mean_token_accuracy": 0.21431368440389634,
      "num_tokens": 59888730.0,
      "step": 26145
    },
    {
      "entropy": 5.458841991424561,
      "epoch": 2.585013839462238,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004335415843162328,
      "loss": 5.1012,
      "mean_token_accuracy": 0.20518300384283067,
      "num_tokens": 59900255.0,
      "step": 26150
    },
    {
      "entropy": 5.472054195404053,
      "epoch": 2.5855081059707397,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043351654145413197,
      "loss": 5.1191,
      "mean_token_accuracy": 0.20492274910211564,
      "num_tokens": 59910705.0,
      "step": 26155
    },
    {
      "entropy": 5.378151369094849,
      "epoch": 2.586002372479241,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004334914946924996,
      "loss": 4.966,
      "mean_token_accuracy": 0.2115619659423828,
      "num_tokens": 59923199.0,
      "step": 26160
    },
    {
      "entropy": 5.515431547164917,
      "epoch": 2.586496638987742,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043346644403195186,
      "loss": 5.0516,
      "mean_token_accuracy": 0.20314325243234635,
      "num_tokens": 59934098.0,
      "step": 26165
    },
    {
      "entropy": 5.508603906631469,
      "epoch": 2.5869909054962434,
      "grad_norm": 0.984375,
      "learning_rate": 0.000433441389473105,
      "loss": 5.0894,
      "mean_token_accuracy": 0.2059381127357483,
      "num_tokens": 59945566.0,
      "step": 26170
    },
    {
      "entropy": 5.507591390609742,
      "epoch": 2.5874851720047447,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004334163310165754,
      "loss": 5.1511,
      "mean_token_accuracy": 0.2025526985526085,
      "num_tokens": 59956354.0,
      "step": 26175
    },
    {
      "entropy": 5.444117021560669,
      "epoch": 2.5879794385132464,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043339126866297943,
      "loss": 5.0909,
      "mean_token_accuracy": 0.2051892399787903,
      "num_tokens": 59967413.0,
      "step": 26180
    },
    {
      "entropy": 5.513080072402954,
      "epoch": 2.5884737050217477,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004333662024129337,
      "loss": 5.0713,
      "mean_token_accuracy": 0.20803989619016647,
      "num_tokens": 59977936.0,
      "step": 26185
    },
    {
      "entropy": 5.50823860168457,
      "epoch": 2.5889679715302494,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004333411322670549,
      "loss": 5.0927,
      "mean_token_accuracy": 0.205030357837677,
      "num_tokens": 59989566.0,
      "step": 26190
    },
    {
      "entropy": 5.451539993286133,
      "epoch": 2.5894622380387506,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043331605822595964,
      "loss": 5.0615,
      "mean_token_accuracy": 0.2113240748643875,
      "num_tokens": 60000339.0,
      "step": 26195
    },
    {
      "entropy": 5.455885124206543,
      "epoch": 2.589956504547252,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043329098029026484,
      "loss": 5.0294,
      "mean_token_accuracy": 0.2087863564491272,
      "num_tokens": 60012903.0,
      "step": 26200
    },
    {
      "entropy": 5.471307945251465,
      "epoch": 2.590450771055753,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043326589846058724,
      "loss": 5.1217,
      "mean_token_accuracy": 0.2106748953461647,
      "num_tokens": 60024135.0,
      "step": 26205
    },
    {
      "entropy": 5.433697700500488,
      "epoch": 2.5909450375642544,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043324081273754403,
      "loss": 5.0857,
      "mean_token_accuracy": 0.20609477162361145,
      "num_tokens": 60036314.0,
      "step": 26210
    },
    {
      "entropy": 5.531763982772827,
      "epoch": 2.591439304072756,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043321572312175234,
      "loss": 5.0631,
      "mean_token_accuracy": 0.20795445293188095,
      "num_tokens": 60047337.0,
      "step": 26215
    },
    {
      "entropy": 5.528925609588623,
      "epoch": 2.5919335705812574,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043319062961382924,
      "loss": 5.1085,
      "mean_token_accuracy": 0.2030316486954689,
      "num_tokens": 60057875.0,
      "step": 26220
    },
    {
      "entropy": 5.4021995067596436,
      "epoch": 2.5924278370897587,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043316553221439214,
      "loss": 5.0849,
      "mean_token_accuracy": 0.20612587332725524,
      "num_tokens": 60069577.0,
      "step": 26225
    },
    {
      "entropy": 5.524349498748779,
      "epoch": 2.5929221035982604,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043314043092405836,
      "loss": 5.145,
      "mean_token_accuracy": 0.19848928600549698,
      "num_tokens": 60081744.0,
      "step": 26230
    },
    {
      "entropy": 5.538342905044556,
      "epoch": 2.5934163701067616,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004331153257434455,
      "loss": 5.1336,
      "mean_token_accuracy": 0.20093371868133544,
      "num_tokens": 60092837.0,
      "step": 26235
    },
    {
      "entropy": 5.4161949634552,
      "epoch": 2.593910636615263,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004330902166731711,
      "loss": 5.0822,
      "mean_token_accuracy": 0.20455933064222337,
      "num_tokens": 60104512.0,
      "step": 26240
    },
    {
      "entropy": 5.413430261611938,
      "epoch": 2.594404903123764,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004330651037138529,
      "loss": 5.0509,
      "mean_token_accuracy": 0.2118354469537735,
      "num_tokens": 60115652.0,
      "step": 26245
    },
    {
      "entropy": 5.494190359115601,
      "epoch": 2.594899169632266,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004330399868661085,
      "loss": 5.0933,
      "mean_token_accuracy": 0.20735854506492615,
      "num_tokens": 60127834.0,
      "step": 26250
    },
    {
      "entropy": 5.498034858703614,
      "epoch": 2.595393436140767,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000433014866130556,
      "loss": 5.1078,
      "mean_token_accuracy": 0.20279533565044403,
      "num_tokens": 60140200.0,
      "step": 26255
    },
    {
      "entropy": 5.507016563415528,
      "epoch": 2.5958877026492684,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043298974150781326,
      "loss": 5.1522,
      "mean_token_accuracy": 0.208327217400074,
      "num_tokens": 60150506.0,
      "step": 26260
    },
    {
      "entropy": 5.520722246170044,
      "epoch": 2.59638196915777,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043296461299849835,
      "loss": 5.2044,
      "mean_token_accuracy": 0.19973582327365874,
      "num_tokens": 60162107.0,
      "step": 26265
    },
    {
      "entropy": 5.550403022766114,
      "epoch": 2.5968762356662713,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043293948060322944,
      "loss": 5.1625,
      "mean_token_accuracy": 0.20080641359090806,
      "num_tokens": 60174435.0,
      "step": 26270
    },
    {
      "entropy": 5.498379039764404,
      "epoch": 2.5973705021747726,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004329143443226249,
      "loss": 5.2042,
      "mean_token_accuracy": 0.19971993565559387,
      "num_tokens": 60186319.0,
      "step": 26275
    },
    {
      "entropy": 5.512925815582276,
      "epoch": 2.597864768683274,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004328892041573029,
      "loss": 5.1285,
      "mean_token_accuracy": 0.20593624114990233,
      "num_tokens": 60197981.0,
      "step": 26280
    },
    {
      "entropy": 5.477794742584228,
      "epoch": 2.598359035191775,
      "grad_norm": 1.0,
      "learning_rate": 0.000432864060107882,
      "loss": 4.9758,
      "mean_token_accuracy": 0.21213001608848572,
      "num_tokens": 60209086.0,
      "step": 26285
    },
    {
      "entropy": 5.420455884933472,
      "epoch": 2.598853301700277,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043283891217498074,
      "loss": 5.053,
      "mean_token_accuracy": 0.21611824184656142,
      "num_tokens": 60221040.0,
      "step": 26290
    },
    {
      "entropy": 5.376713609695434,
      "epoch": 2.599347568208778,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004328137603592177,
      "loss": 4.9808,
      "mean_token_accuracy": 0.2165214866399765,
      "num_tokens": 60232709.0,
      "step": 26295
    },
    {
      "entropy": 5.487830781936646,
      "epoch": 2.59984183471728,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004327886046612117,
      "loss": 5.0771,
      "mean_token_accuracy": 0.20377163589000702,
      "num_tokens": 60244539.0,
      "step": 26300
    },
    {
      "entropy": 5.501164484024048,
      "epoch": 2.600336101225781,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004327634450815817,
      "loss": 5.0579,
      "mean_token_accuracy": 0.2056467801332474,
      "num_tokens": 60255124.0,
      "step": 26305
    },
    {
      "entropy": 5.545708274841308,
      "epoch": 2.6008303677342823,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043273828162094624,
      "loss": 5.2157,
      "mean_token_accuracy": 0.19296193718910218,
      "num_tokens": 60267819.0,
      "step": 26310
    },
    {
      "entropy": 5.514840126037598,
      "epoch": 2.6013246342427836,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004327131142799247,
      "loss": 5.1171,
      "mean_token_accuracy": 0.20716441720724105,
      "num_tokens": 60279694.0,
      "step": 26315
    },
    {
      "entropy": 5.508053398132324,
      "epoch": 2.601818900751285,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004326879430591361,
      "loss": 5.1317,
      "mean_token_accuracy": 0.19925848692655562,
      "num_tokens": 60291215.0,
      "step": 26320
    },
    {
      "entropy": 5.474817609786987,
      "epoch": 2.6023131672597866,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043266276795919954,
      "loss": 5.0348,
      "mean_token_accuracy": 0.20297530591487883,
      "num_tokens": 60303639.0,
      "step": 26325
    },
    {
      "entropy": 5.462808895111084,
      "epoch": 2.602807433768288,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004326375889807345,
      "loss": 5.0705,
      "mean_token_accuracy": 0.20270152091979982,
      "num_tokens": 60315290.0,
      "step": 26330
    },
    {
      "entropy": 5.4940886974334715,
      "epoch": 2.603301700276789,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004326124061243603,
      "loss": 5.0741,
      "mean_token_accuracy": 0.21007844507694245,
      "num_tokens": 60325391.0,
      "step": 26335
    },
    {
      "entropy": 5.43610258102417,
      "epoch": 2.603795966785291,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004325872193906965,
      "loss": 4.9974,
      "mean_token_accuracy": 0.2205174446105957,
      "num_tokens": 60335758.0,
      "step": 26340
    },
    {
      "entropy": 5.497511386871338,
      "epoch": 2.604290233293792,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043256202878036264,
      "loss": 5.0612,
      "mean_token_accuracy": 0.20659103840589524,
      "num_tokens": 60347365.0,
      "step": 26345
    },
    {
      "entropy": 5.38678731918335,
      "epoch": 2.6047844998022933,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043253683429397843,
      "loss": 4.9503,
      "mean_token_accuracy": 0.21219963431358338,
      "num_tokens": 60359222.0,
      "step": 26350
    },
    {
      "entropy": 5.499383592605591,
      "epoch": 2.6052787663107946,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043251163593216365,
      "loss": 5.1614,
      "mean_token_accuracy": 0.1955256789922714,
      "num_tokens": 60370730.0,
      "step": 26355
    },
    {
      "entropy": 5.40680947303772,
      "epoch": 2.6057730328192963,
      "grad_norm": 1.0,
      "learning_rate": 0.00043248643369553813,
      "loss": 5.0133,
      "mean_token_accuracy": 0.2191762626171112,
      "num_tokens": 60381955.0,
      "step": 26360
    },
    {
      "entropy": 5.513626623153686,
      "epoch": 2.6062672993277975,
      "grad_norm": 1.078125,
      "learning_rate": 0.000432461227584722,
      "loss": 5.1445,
      "mean_token_accuracy": 0.19840544164180757,
      "num_tokens": 60395331.0,
      "step": 26365
    },
    {
      "entropy": 5.442679119110108,
      "epoch": 2.606761565836299,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004324360176003352,
      "loss": 5.0153,
      "mean_token_accuracy": 0.2100110948085785,
      "num_tokens": 60407179.0,
      "step": 26370
    },
    {
      "entropy": 5.393788528442383,
      "epoch": 2.6072558323448005,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004324108037429979,
      "loss": 5.0632,
      "mean_token_accuracy": 0.2075355812907219,
      "num_tokens": 60418377.0,
      "step": 26375
    },
    {
      "entropy": 5.4761566638946535,
      "epoch": 2.607750098853302,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004323855860133305,
      "loss": 5.0548,
      "mean_token_accuracy": 0.20840090811252593,
      "num_tokens": 60428802.0,
      "step": 26380
    },
    {
      "entropy": 5.514241933822632,
      "epoch": 2.608244365361803,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004323603644119532,
      "loss": 5.1036,
      "mean_token_accuracy": 0.20701534152030945,
      "num_tokens": 60440445.0,
      "step": 26385
    },
    {
      "entropy": 5.424113368988037,
      "epoch": 2.6087386318703043,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043233513893948654,
      "loss": 5.0391,
      "mean_token_accuracy": 0.20562277138233184,
      "num_tokens": 60451040.0,
      "step": 26390
    },
    {
      "entropy": 5.508267259597778,
      "epoch": 2.609232898378806,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004323099095965511,
      "loss": 5.1796,
      "mean_token_accuracy": 0.19670177549123763,
      "num_tokens": 60463820.0,
      "step": 26395
    },
    {
      "entropy": 5.480975866317749,
      "epoch": 2.6097271648873073,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004322846763837674,
      "loss": 5.0508,
      "mean_token_accuracy": 0.21566741913557053,
      "num_tokens": 60474285.0,
      "step": 26400
    },
    {
      "entropy": 5.381169176101684,
      "epoch": 2.6102214313958085,
      "grad_norm": 1.125,
      "learning_rate": 0.00043225943930175626,
      "loss": 4.993,
      "mean_token_accuracy": 0.22108036875724793,
      "num_tokens": 60484475.0,
      "step": 26405
    },
    {
      "entropy": 5.465008306503296,
      "epoch": 2.6107156979043102,
      "grad_norm": 1.125,
      "learning_rate": 0.0004322341983511386,
      "loss": 5.0155,
      "mean_token_accuracy": 0.21417828947305678,
      "num_tokens": 60494182.0,
      "step": 26410
    },
    {
      "entropy": 5.417870378494262,
      "epoch": 2.6112099644128115,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043220895353253516,
      "loss": 5.0619,
      "mean_token_accuracy": 0.21032250672578812,
      "num_tokens": 60505455.0,
      "step": 26415
    },
    {
      "entropy": 5.393094491958618,
      "epoch": 2.6117042309213128,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043218370484656717,
      "loss": 4.9958,
      "mean_token_accuracy": 0.21256716251373292,
      "num_tokens": 60518040.0,
      "step": 26420
    },
    {
      "entropy": 5.475756120681763,
      "epoch": 2.612198497429814,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043215845229385567,
      "loss": 5.0994,
      "mean_token_accuracy": 0.2092757284641266,
      "num_tokens": 60528896.0,
      "step": 26425
    },
    {
      "entropy": 5.454882574081421,
      "epoch": 2.6126927639383153,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043213319587502173,
      "loss": 5.0122,
      "mean_token_accuracy": 0.2114014893770218,
      "num_tokens": 60539505.0,
      "step": 26430
    },
    {
      "entropy": 5.437011003494263,
      "epoch": 2.613187030446817,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004321079355906869,
      "loss": 5.0648,
      "mean_token_accuracy": 0.2086468294262886,
      "num_tokens": 60550544.0,
      "step": 26435
    },
    {
      "entropy": 5.544268035888672,
      "epoch": 2.6136812969553183,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043208267144147244,
      "loss": 5.163,
      "mean_token_accuracy": 0.1958056643605232,
      "num_tokens": 60561697.0,
      "step": 26440
    },
    {
      "entropy": 5.389890193939209,
      "epoch": 2.61417556346382,
      "grad_norm": 1.0,
      "learning_rate": 0.00043205740342799995,
      "loss": 4.9761,
      "mean_token_accuracy": 0.21673928052186966,
      "num_tokens": 60572817.0,
      "step": 26445
    },
    {
      "entropy": 5.477241182327271,
      "epoch": 2.6146698299723212,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043203213155089095,
      "loss": 5.0631,
      "mean_token_accuracy": 0.207886503636837,
      "num_tokens": 60584679.0,
      "step": 26450
    },
    {
      "entropy": 5.440163421630859,
      "epoch": 2.6151640964808225,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004320068558107671,
      "loss": 5.0891,
      "mean_token_accuracy": 0.20481555610895158,
      "num_tokens": 60594871.0,
      "step": 26455
    },
    {
      "entropy": 5.43923716545105,
      "epoch": 2.6156583629893237,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043198157620825023,
      "loss": 5.0671,
      "mean_token_accuracy": 0.21116907596588136,
      "num_tokens": 60605626.0,
      "step": 26460
    },
    {
      "entropy": 5.436730098724365,
      "epoch": 2.616152629497825,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043195629274396237,
      "loss": 5.0885,
      "mean_token_accuracy": 0.21786459684371948,
      "num_tokens": 60617802.0,
      "step": 26465
    },
    {
      "entropy": 5.502986764907837,
      "epoch": 2.6166468960063267,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043193100541852526,
      "loss": 5.1397,
      "mean_token_accuracy": 0.20379660576581954,
      "num_tokens": 60629044.0,
      "step": 26470
    },
    {
      "entropy": 5.429233694076538,
      "epoch": 2.617141162514828,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000431905714232561,
      "loss": 5.0217,
      "mean_token_accuracy": 0.21311260610818863,
      "num_tokens": 60639729.0,
      "step": 26475
    },
    {
      "entropy": 5.4653191566467285,
      "epoch": 2.6176354290233292,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043188041918669197,
      "loss": 5.1361,
      "mean_token_accuracy": 0.2030935451388359,
      "num_tokens": 60650499.0,
      "step": 26480
    },
    {
      "entropy": 5.5962292671203615,
      "epoch": 2.618129695531831,
      "grad_norm": 3.390625,
      "learning_rate": 0.0004318551202815402,
      "loss": 5.2017,
      "mean_token_accuracy": 0.19315394014120102,
      "num_tokens": 60661453.0,
      "step": 26485
    },
    {
      "entropy": 5.480583620071411,
      "epoch": 2.618623962040332,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043182981751772816,
      "loss": 5.0554,
      "mean_token_accuracy": 0.20832203328609467,
      "num_tokens": 60672748.0,
      "step": 26490
    },
    {
      "entropy": 5.437837219238281,
      "epoch": 2.6191182285488335,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004318045108958783,
      "loss": 5.0296,
      "mean_token_accuracy": 0.211034794151783,
      "num_tokens": 60683679.0,
      "step": 26495
    },
    {
      "entropy": 5.4222112655639645,
      "epoch": 2.6196124950573347,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043177920041661313,
      "loss": 5.0262,
      "mean_token_accuracy": 0.2086041808128357,
      "num_tokens": 60696988.0,
      "step": 26500
    },
    {
      "entropy": 5.466078042984009,
      "epoch": 2.6201067615658364,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004317538860805553,
      "loss": 5.1961,
      "mean_token_accuracy": 0.19980721473693847,
      "num_tokens": 60709817.0,
      "step": 26505
    },
    {
      "entropy": 5.531041431427002,
      "epoch": 2.6206010280743377,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043172856788832755,
      "loss": 5.1067,
      "mean_token_accuracy": 0.21104227155447006,
      "num_tokens": 60720941.0,
      "step": 26510
    },
    {
      "entropy": 5.5365966796875,
      "epoch": 2.621095294582839,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043170324584055275,
      "loss": 5.2056,
      "mean_token_accuracy": 0.19685383290052413,
      "num_tokens": 60731873.0,
      "step": 26515
    },
    {
      "entropy": 5.434489774703979,
      "epoch": 2.6215895610913407,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043167791993785375,
      "loss": 5.0962,
      "mean_token_accuracy": 0.2112856924533844,
      "num_tokens": 60744361.0,
      "step": 26520
    },
    {
      "entropy": 5.5674567222595215,
      "epoch": 2.622083827599842,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004316525901808536,
      "loss": 5.1244,
      "mean_token_accuracy": 0.20559146702289582,
      "num_tokens": 60756111.0,
      "step": 26525
    },
    {
      "entropy": 5.4759602546691895,
      "epoch": 2.622578094108343,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043162725657017544,
      "loss": 4.9882,
      "mean_token_accuracy": 0.2184994637966156,
      "num_tokens": 60766602.0,
      "step": 26530
    },
    {
      "entropy": 5.463848733901978,
      "epoch": 2.6230723606168445,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043160191910644253,
      "loss": 5.1117,
      "mean_token_accuracy": 0.20356844067573548,
      "num_tokens": 60778670.0,
      "step": 26535
    },
    {
      "entropy": 5.488467216491699,
      "epoch": 2.6235666271253457,
      "grad_norm": 1.125,
      "learning_rate": 0.0004315765777902781,
      "loss": 5.1716,
      "mean_token_accuracy": 0.19680868536233903,
      "num_tokens": 60790026.0,
      "step": 26540
    },
    {
      "entropy": 5.472838830947876,
      "epoch": 2.6240608936338474,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004315512326223055,
      "loss": 5.1312,
      "mean_token_accuracy": 0.20088608264923097,
      "num_tokens": 60801325.0,
      "step": 26545
    },
    {
      "entropy": 5.52078104019165,
      "epoch": 2.6245551601423487,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004315258836031483,
      "loss": 5.1208,
      "mean_token_accuracy": 0.20107238590717316,
      "num_tokens": 60813421.0,
      "step": 26550
    },
    {
      "entropy": 5.54648175239563,
      "epoch": 2.6250494266508504,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004315005307334301,
      "loss": 5.1191,
      "mean_token_accuracy": 0.19864647835493088,
      "num_tokens": 60826447.0,
      "step": 26555
    },
    {
      "entropy": 5.461081171035767,
      "epoch": 2.6255436931593517,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004314751740137746,
      "loss": 5.0733,
      "mean_token_accuracy": 0.20090051591396332,
      "num_tokens": 60837663.0,
      "step": 26560
    },
    {
      "entropy": 5.46001935005188,
      "epoch": 2.626037959667853,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043144981344480553,
      "loss": 5.0635,
      "mean_token_accuracy": 0.19907326698303224,
      "num_tokens": 60849040.0,
      "step": 26565
    },
    {
      "entropy": 5.442444801330566,
      "epoch": 2.626532226176354,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043142444902714676,
      "loss": 5.1298,
      "mean_token_accuracy": 0.20175640434026718,
      "num_tokens": 60862112.0,
      "step": 26570
    },
    {
      "entropy": 5.440461301803589,
      "epoch": 2.6270264926848554,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043139908076142224,
      "loss": 5.049,
      "mean_token_accuracy": 0.2075262725353241,
      "num_tokens": 60872822.0,
      "step": 26575
    },
    {
      "entropy": 5.49421877861023,
      "epoch": 2.627520759193357,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004313737086482562,
      "loss": 5.1459,
      "mean_token_accuracy": 0.2012254476547241,
      "num_tokens": 60884374.0,
      "step": 26580
    },
    {
      "entropy": 5.469945573806763,
      "epoch": 2.6280150257018584,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004313483326882725,
      "loss": 5.1148,
      "mean_token_accuracy": 0.2062963977456093,
      "num_tokens": 60895490.0,
      "step": 26585
    },
    {
      "entropy": 5.382683801651001,
      "epoch": 2.6285092922103597,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004313229528820957,
      "loss": 4.9652,
      "mean_token_accuracy": 0.2191467374563217,
      "num_tokens": 60906190.0,
      "step": 26590
    },
    {
      "entropy": 5.479799509048462,
      "epoch": 2.6290035587188614,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043129756923034995,
      "loss": 5.0981,
      "mean_token_accuracy": 0.21025551855564117,
      "num_tokens": 60918466.0,
      "step": 26595
    },
    {
      "entropy": 5.386405181884766,
      "epoch": 2.6294978252273626,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004312721817336597,
      "loss": 4.9745,
      "mean_token_accuracy": 0.21411477476358415,
      "num_tokens": 60930202.0,
      "step": 26600
    },
    {
      "entropy": 5.492415094375611,
      "epoch": 2.629992091735864,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043124679039264964,
      "loss": 5.1096,
      "mean_token_accuracy": 0.20260215997695924,
      "num_tokens": 60942089.0,
      "step": 26605
    },
    {
      "entropy": 5.488260698318482,
      "epoch": 2.630486358244365,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043122139520794425,
      "loss": 5.0397,
      "mean_token_accuracy": 0.20462982654571532,
      "num_tokens": 60954493.0,
      "step": 26610
    },
    {
      "entropy": 5.418860816955567,
      "epoch": 2.630980624752867,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00043119599618016824,
      "loss": 5.0818,
      "mean_token_accuracy": 0.20976886600255967,
      "num_tokens": 60965888.0,
      "step": 26615
    },
    {
      "entropy": 5.438356876373291,
      "epoch": 2.631474891261368,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004311705933099466,
      "loss": 5.0117,
      "mean_token_accuracy": 0.2073403149843216,
      "num_tokens": 60976414.0,
      "step": 26620
    },
    {
      "entropy": 5.448660182952881,
      "epoch": 2.6319691577698694,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004311451865979041,
      "loss": 5.012,
      "mean_token_accuracy": 0.2099453791975975,
      "num_tokens": 60986353.0,
      "step": 26625
    },
    {
      "entropy": 5.45160813331604,
      "epoch": 2.632463424278371,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043111977604466575,
      "loss": 5.0387,
      "mean_token_accuracy": 0.21607030928134918,
      "num_tokens": 60996916.0,
      "step": 26630
    },
    {
      "entropy": 5.342118644714356,
      "epoch": 2.6329576907868724,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004310943616508567,
      "loss": 4.9957,
      "mean_token_accuracy": 0.21445899456739426,
      "num_tokens": 61008741.0,
      "step": 26635
    },
    {
      "entropy": 5.511223268508911,
      "epoch": 2.6334519572953736,
      "grad_norm": 1.125,
      "learning_rate": 0.0004310689434171022,
      "loss": 5.1891,
      "mean_token_accuracy": 0.1967450961470604,
      "num_tokens": 61019988.0,
      "step": 26640
    },
    {
      "entropy": 5.539672708511352,
      "epoch": 2.633946223803875,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043104352134402734,
      "loss": 5.1157,
      "mean_token_accuracy": 0.20142392218112945,
      "num_tokens": 61031667.0,
      "step": 26645
    },
    {
      "entropy": 5.456628370285034,
      "epoch": 2.6344404903123766,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043101809543225777,
      "loss": 5.0283,
      "mean_token_accuracy": 0.21000753939151764,
      "num_tokens": 61042203.0,
      "step": 26650
    },
    {
      "entropy": 5.443127679824829,
      "epoch": 2.634934756820878,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043099266568241885,
      "loss": 5.0423,
      "mean_token_accuracy": 0.2038487523794174,
      "num_tokens": 61054243.0,
      "step": 26655
    },
    {
      "entropy": 5.342503356933594,
      "epoch": 2.635429023329379,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043096723209513606,
      "loss": 4.8997,
      "mean_token_accuracy": 0.22634548097848892,
      "num_tokens": 61064068.0,
      "step": 26660
    },
    {
      "entropy": 5.357486343383789,
      "epoch": 2.635923289837881,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004309417946710352,
      "loss": 4.9875,
      "mean_token_accuracy": 0.21721184253692627,
      "num_tokens": 61075273.0,
      "step": 26665
    },
    {
      "entropy": 5.499697875976563,
      "epoch": 2.636417556346382,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000430916353410742,
      "loss": 5.0841,
      "mean_token_accuracy": 0.20613777339458467,
      "num_tokens": 61086536.0,
      "step": 26670
    },
    {
      "entropy": 5.44303674697876,
      "epoch": 2.6369118228548833,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004308909083148824,
      "loss": 5.0504,
      "mean_token_accuracy": 0.2095375418663025,
      "num_tokens": 61097567.0,
      "step": 26675
    },
    {
      "entropy": 5.459296464920044,
      "epoch": 2.6374060893633846,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004308654593840822,
      "loss": 5.0471,
      "mean_token_accuracy": 0.20929599404335023,
      "num_tokens": 61108379.0,
      "step": 26680
    },
    {
      "entropy": 5.350851202011109,
      "epoch": 2.637900355871886,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043084000661896757,
      "loss": 5.0269,
      "mean_token_accuracy": 0.20912112146615983,
      "num_tokens": 61119715.0,
      "step": 26685
    },
    {
      "entropy": 5.519884395599365,
      "epoch": 2.6383946223803876,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004308145500201645,
      "loss": 5.1525,
      "mean_token_accuracy": 0.2016537070274353,
      "num_tokens": 61131810.0,
      "step": 26690
    },
    {
      "entropy": 5.433495283126831,
      "epoch": 2.638888888888889,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043078908958829945,
      "loss": 5.0097,
      "mean_token_accuracy": 0.21264057457447053,
      "num_tokens": 61142771.0,
      "step": 26695
    },
    {
      "entropy": 5.382149171829224,
      "epoch": 2.6393831553973905,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043076362532399855,
      "loss": 5.0759,
      "mean_token_accuracy": 0.204647296667099,
      "num_tokens": 61154105.0,
      "step": 26700
    },
    {
      "entropy": 5.490643835067749,
      "epoch": 2.639877421905892,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004307381572278882,
      "loss": 4.9739,
      "mean_token_accuracy": 0.21394731253385543,
      "num_tokens": 61164574.0,
      "step": 26705
    },
    {
      "entropy": 5.45215539932251,
      "epoch": 2.640371688414393,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004307126853005952,
      "loss": 5.1332,
      "mean_token_accuracy": 0.20109274089336396,
      "num_tokens": 61177085.0,
      "step": 26710
    },
    {
      "entropy": 5.440183925628662,
      "epoch": 2.6408659549228943,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004306872095427459,
      "loss": 5.0312,
      "mean_token_accuracy": 0.2124553382396698,
      "num_tokens": 61188313.0,
      "step": 26715
    },
    {
      "entropy": 5.392519187927246,
      "epoch": 2.6413602214313956,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004306617299549672,
      "loss": 4.9626,
      "mean_token_accuracy": 0.21787741929292678,
      "num_tokens": 61199726.0,
      "step": 26720
    },
    {
      "entropy": 5.410009431838989,
      "epoch": 2.6418544879398973,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004306362465378857,
      "loss": 5.0746,
      "mean_token_accuracy": 0.20545369237661362,
      "num_tokens": 61211861.0,
      "step": 26725
    },
    {
      "entropy": 5.420197296142578,
      "epoch": 2.6423487544483986,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043061075929212836,
      "loss": 5.0362,
      "mean_token_accuracy": 0.2091488778591156,
      "num_tokens": 61222464.0,
      "step": 26730
    },
    {
      "entropy": 5.490419244766235,
      "epoch": 2.6428430209569,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004305852682183222,
      "loss": 5.028,
      "mean_token_accuracy": 0.20403672009706497,
      "num_tokens": 61233678.0,
      "step": 26735
    },
    {
      "entropy": 5.432313632965088,
      "epoch": 2.6433372874654015,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004305597733170944,
      "loss": 5.0998,
      "mean_token_accuracy": 0.20257449746131898,
      "num_tokens": 61244636.0,
      "step": 26740
    },
    {
      "entropy": 5.529808282852173,
      "epoch": 2.643831553973903,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043053427458907196,
      "loss": 5.078,
      "mean_token_accuracy": 0.20320031940937042,
      "num_tokens": 61256334.0,
      "step": 26745
    },
    {
      "entropy": 5.47851333618164,
      "epoch": 2.644325820482404,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004305087720348823,
      "loss": 5.0992,
      "mean_token_accuracy": 0.21004230976104737,
      "num_tokens": 61268614.0,
      "step": 26750
    },
    {
      "entropy": 5.484437370300293,
      "epoch": 2.6448200869909053,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043048326565515267,
      "loss": 5.1569,
      "mean_token_accuracy": 0.19599247276782988,
      "num_tokens": 61280220.0,
      "step": 26755
    },
    {
      "entropy": 5.474890804290771,
      "epoch": 2.645314353499407,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043045775545051055,
      "loss": 4.9947,
      "mean_token_accuracy": 0.21295276135206223,
      "num_tokens": 61290796.0,
      "step": 26760
    },
    {
      "entropy": 5.437591934204102,
      "epoch": 2.6458086200079083,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004304322414215836,
      "loss": 5.0929,
      "mean_token_accuracy": 0.20901717990636826,
      "num_tokens": 61303645.0,
      "step": 26765
    },
    {
      "entropy": 5.446726894378662,
      "epoch": 2.6463028865164095,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043040672356899946,
      "loss": 5.0861,
      "mean_token_accuracy": 0.20499373972415924,
      "num_tokens": 61316641.0,
      "step": 26770
    },
    {
      "entropy": 5.511597728729248,
      "epoch": 2.6467971530249113,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043038120189338566,
      "loss": 5.1653,
      "mean_token_accuracy": 0.19674562364816667,
      "num_tokens": 61329314.0,
      "step": 26775
    },
    {
      "entropy": 5.503765201568603,
      "epoch": 2.6472914195334125,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043035567639537026,
      "loss": 5.16,
      "mean_token_accuracy": 0.20136458426713943,
      "num_tokens": 61340847.0,
      "step": 26780
    },
    {
      "entropy": 5.565564870834351,
      "epoch": 2.6477856860419138,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043033014707558113,
      "loss": 5.1369,
      "mean_token_accuracy": 0.20283611863851547,
      "num_tokens": 61351875.0,
      "step": 26785
    },
    {
      "entropy": 5.46535062789917,
      "epoch": 2.648279952550415,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004303046139346462,
      "loss": 5.1046,
      "mean_token_accuracy": 0.2105036199092865,
      "num_tokens": 61364441.0,
      "step": 26790
    },
    {
      "entropy": 5.412400817871093,
      "epoch": 2.6487742190589163,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004302790769731938,
      "loss": 5.0532,
      "mean_token_accuracy": 0.20345621854066848,
      "num_tokens": 61375084.0,
      "step": 26795
    },
    {
      "entropy": 5.517037677764892,
      "epoch": 2.649268485567418,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043025353619185195,
      "loss": 5.1402,
      "mean_token_accuracy": 0.20046569854021073,
      "num_tokens": 61386036.0,
      "step": 26800
    },
    {
      "entropy": 5.508811283111572,
      "epoch": 2.6497627520759193,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043022799159124904,
      "loss": 5.0772,
      "mean_token_accuracy": 0.2088192507624626,
      "num_tokens": 61396006.0,
      "step": 26805
    },
    {
      "entropy": 5.483269786834716,
      "epoch": 2.650257018584421,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043020244317201344,
      "loss": 5.1567,
      "mean_token_accuracy": 0.1989763006567955,
      "num_tokens": 61406453.0,
      "step": 26810
    },
    {
      "entropy": 5.51816234588623,
      "epoch": 2.6507512850929222,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004301768909347737,
      "loss": 5.1564,
      "mean_token_accuracy": 0.20571520328521728,
      "num_tokens": 61418271.0,
      "step": 26815
    },
    {
      "entropy": 5.55323224067688,
      "epoch": 2.6512455516014235,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004301513348801583,
      "loss": 5.0889,
      "mean_token_accuracy": 0.199857459962368,
      "num_tokens": 61430575.0,
      "step": 26820
    },
    {
      "entropy": 5.360191583633423,
      "epoch": 2.6517398181099248,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000430125775008796,
      "loss": 4.9965,
      "mean_token_accuracy": 0.21166138648986815,
      "num_tokens": 61442336.0,
      "step": 26825
    },
    {
      "entropy": 5.469023084640503,
      "epoch": 2.652234084618426,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043010021132131564,
      "loss": 5.0747,
      "mean_token_accuracy": 0.20623867511749266,
      "num_tokens": 61453945.0,
      "step": 26830
    },
    {
      "entropy": 5.489079141616822,
      "epoch": 2.6527283511269277,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000430074643818346,
      "loss": 5.0425,
      "mean_token_accuracy": 0.21198657900094986,
      "num_tokens": 61463877.0,
      "step": 26835
    },
    {
      "entropy": 5.446172046661377,
      "epoch": 2.653222617635429,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043004907250051606,
      "loss": 4.9775,
      "mean_token_accuracy": 0.21135786771774293,
      "num_tokens": 61473278.0,
      "step": 26840
    },
    {
      "entropy": 5.550709342956543,
      "epoch": 2.6537168841439303,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004300234973684548,
      "loss": 5.168,
      "mean_token_accuracy": 0.201534004509449,
      "num_tokens": 61486038.0,
      "step": 26845
    },
    {
      "entropy": 5.468103408813477,
      "epoch": 2.654211150652432,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004299979184227916,
      "loss": 5.0906,
      "mean_token_accuracy": 0.20338857173919678,
      "num_tokens": 61497722.0,
      "step": 26850
    },
    {
      "entropy": 5.415162372589111,
      "epoch": 2.6547054171609332,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042997233566415547,
      "loss": 5.0362,
      "mean_token_accuracy": 0.20714619755744934,
      "num_tokens": 61508791.0,
      "step": 26855
    },
    {
      "entropy": 5.467224073410034,
      "epoch": 2.6551996836694345,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042994674909317584,
      "loss": 5.0141,
      "mean_token_accuracy": 0.21556785702705383,
      "num_tokens": 61520047.0,
      "step": 26860
    },
    {
      "entropy": 5.444084167480469,
      "epoch": 2.6556939501779357,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042992115871048214,
      "loss": 5.0257,
      "mean_token_accuracy": 0.2081543743610382,
      "num_tokens": 61531005.0,
      "step": 26865
    },
    {
      "entropy": 5.409723520278931,
      "epoch": 2.6561882166864375,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042989556451670396,
      "loss": 5.0781,
      "mean_token_accuracy": 0.20670622438192368,
      "num_tokens": 61542087.0,
      "step": 26870
    },
    {
      "entropy": 5.452761220932007,
      "epoch": 2.6566824831949387,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004298699665124708,
      "loss": 5.155,
      "mean_token_accuracy": 0.20421219915151595,
      "num_tokens": 61553575.0,
      "step": 26875
    },
    {
      "entropy": 5.582581949234009,
      "epoch": 2.65717674970344,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042984436469841245,
      "loss": 5.1373,
      "mean_token_accuracy": 0.2018783912062645,
      "num_tokens": 61564833.0,
      "step": 26880
    },
    {
      "entropy": 5.415939044952393,
      "epoch": 2.6576710162119417,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042981875907515874,
      "loss": 4.9372,
      "mean_token_accuracy": 0.21965519785881044,
      "num_tokens": 61575721.0,
      "step": 26885
    },
    {
      "entropy": 5.471623086929322,
      "epoch": 2.658165282720443,
      "grad_norm": 1.125,
      "learning_rate": 0.00042979314964333953,
      "loss": 5.0992,
      "mean_token_accuracy": 0.20798233300447463,
      "num_tokens": 61587022.0,
      "step": 26890
    },
    {
      "entropy": 5.452496528625488,
      "epoch": 2.658659549228944,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004297675364035848,
      "loss": 5.0425,
      "mean_token_accuracy": 0.20675941407680512,
      "num_tokens": 61597610.0,
      "step": 26895
    },
    {
      "entropy": 5.549894571304321,
      "epoch": 2.6591538157374455,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004297419193565247,
      "loss": 5.1466,
      "mean_token_accuracy": 0.19942129850387574,
      "num_tokens": 61608254.0,
      "step": 26900
    },
    {
      "entropy": 5.477271461486817,
      "epoch": 2.659648082245947,
      "grad_norm": 1.0,
      "learning_rate": 0.00042971629850278924,
      "loss": 5.0503,
      "mean_token_accuracy": 0.20788105428218842,
      "num_tokens": 61619937.0,
      "step": 26905
    },
    {
      "entropy": 5.419719552993774,
      "epoch": 2.6601423487544484,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042969067384300905,
      "loss": 5.0893,
      "mean_token_accuracy": 0.20655132234096527,
      "num_tokens": 61630460.0,
      "step": 26910
    },
    {
      "entropy": 5.455670738220215,
      "epoch": 2.6606366152629497,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042966504537781416,
      "loss": 5.0139,
      "mean_token_accuracy": 0.21229053139686585,
      "num_tokens": 61640855.0,
      "step": 26915
    },
    {
      "entropy": 5.56550350189209,
      "epoch": 2.6611308817714514,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004296394131078351,
      "loss": 5.1356,
      "mean_token_accuracy": 0.20096256732940673,
      "num_tokens": 61652405.0,
      "step": 26920
    },
    {
      "entropy": 5.459198141098023,
      "epoch": 2.6616251482799527,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042961377703370265,
      "loss": 5.0121,
      "mean_token_accuracy": 0.20885648280382157,
      "num_tokens": 61664379.0,
      "step": 26925
    },
    {
      "entropy": 5.444211769104004,
      "epoch": 2.662119414788454,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004295881371560472,
      "loss": 5.0363,
      "mean_token_accuracy": 0.21038832217454911,
      "num_tokens": 61675259.0,
      "step": 26930
    },
    {
      "entropy": 5.418305683135986,
      "epoch": 2.662613681296955,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042956249347549964,
      "loss": 5.0662,
      "mean_token_accuracy": 0.20536253452301026,
      "num_tokens": 61687035.0,
      "step": 26935
    },
    {
      "entropy": 5.529556465148926,
      "epoch": 2.6631079478054565,
      "grad_norm": 1.125,
      "learning_rate": 0.0004295368459926907,
      "loss": 5.103,
      "mean_token_accuracy": 0.20379627794027327,
      "num_tokens": 61698082.0,
      "step": 26940
    },
    {
      "entropy": 5.485279846191406,
      "epoch": 2.663602214313958,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042951119470825146,
      "loss": 5.0762,
      "mean_token_accuracy": 0.19980324804782867,
      "num_tokens": 61708533.0,
      "step": 26945
    },
    {
      "entropy": 5.520138072967529,
      "epoch": 2.6640964808224594,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042948553962281286,
      "loss": 5.0981,
      "mean_token_accuracy": 0.2021261289715767,
      "num_tokens": 61719301.0,
      "step": 26950
    },
    {
      "entropy": 5.4330301761627195,
      "epoch": 2.664590747330961,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004294598807370059,
      "loss": 5.0357,
      "mean_token_accuracy": 0.21138217598199843,
      "num_tokens": 61732225.0,
      "step": 26955
    },
    {
      "entropy": 5.364568758010864,
      "epoch": 2.6650850138394624,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000429434218051462,
      "loss": 5.0185,
      "mean_token_accuracy": 0.21513096392154693,
      "num_tokens": 61743901.0,
      "step": 26960
    },
    {
      "entropy": 5.517559814453125,
      "epoch": 2.6655792803479637,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004294085515668123,
      "loss": 5.0511,
      "mean_token_accuracy": 0.21002853959798812,
      "num_tokens": 61756047.0,
      "step": 26965
    },
    {
      "entropy": 5.472736930847168,
      "epoch": 2.666073546856465,
      "grad_norm": 1.0,
      "learning_rate": 0.00042938288128368835,
      "loss": 5.0314,
      "mean_token_accuracy": 0.21113575696945192,
      "num_tokens": 61767784.0,
      "step": 26970
    },
    {
      "entropy": 5.440064191818237,
      "epoch": 2.666567813364966,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042935720720272156,
      "loss": 5.1486,
      "mean_token_accuracy": 0.20169439613819123,
      "num_tokens": 61778915.0,
      "step": 26975
    },
    {
      "entropy": 5.424635314941407,
      "epoch": 2.667062079873468,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004293315293245435,
      "loss": 5.0386,
      "mean_token_accuracy": 0.21064051985740662,
      "num_tokens": 61791046.0,
      "step": 26980
    },
    {
      "entropy": 5.468916416168213,
      "epoch": 2.667556346381969,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042930584764978597,
      "loss": 5.0401,
      "mean_token_accuracy": 0.2153873398900032,
      "num_tokens": 61802135.0,
      "step": 26985
    },
    {
      "entropy": 5.5550189971923825,
      "epoch": 2.6680506128904704,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042928016217908053,
      "loss": 5.1007,
      "mean_token_accuracy": 0.2072929099202156,
      "num_tokens": 61815077.0,
      "step": 26990
    },
    {
      "entropy": 5.405262088775634,
      "epoch": 2.668544879398972,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004292544729130592,
      "loss": 4.9867,
      "mean_token_accuracy": 0.2124576225876808,
      "num_tokens": 61826031.0,
      "step": 26995
    },
    {
      "entropy": 5.393457460403442,
      "epoch": 2.6690391459074734,
      "grad_norm": 1.0,
      "learning_rate": 0.00042922877985235395,
      "loss": 5.0121,
      "mean_token_accuracy": 0.21686503291130066,
      "num_tokens": 61836372.0,
      "step": 27000
    },
    {
      "epoch": 2.6690391459074734,
      "eval_entropy": 5.19592115905452,
      "eval_loss": 5.143263816833496,
      "eval_mean_token_accuracy": 0.2119038778257546,
      "eval_num_tokens": 61836372.0,
      "eval_runtime": 26.4994,
      "eval_samples_per_second": 1226.933,
      "eval_steps_per_second": 153.4,
      "step": 27000
    },
    {
      "entropy": 5.376849269866943,
      "epoch": 2.6695334124159746,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042920308299759677,
      "loss": 4.995,
      "mean_token_accuracy": 0.21264728009700776,
      "num_tokens": 61847829.0,
      "step": 27005
    },
    {
      "entropy": 5.4404786109924315,
      "epoch": 2.670027678924476,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004291773823494198,
      "loss": 5.072,
      "mean_token_accuracy": 0.20994795113801956,
      "num_tokens": 61859074.0,
      "step": 27010
    },
    {
      "entropy": 5.5471666812896725,
      "epoch": 2.6705219454329776,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042915167790845535,
      "loss": 5.1732,
      "mean_token_accuracy": 0.19429955631494522,
      "num_tokens": 61868596.0,
      "step": 27015
    },
    {
      "entropy": 5.411706495285034,
      "epoch": 2.671016211941479,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004291259696753356,
      "loss": 5.0581,
      "mean_token_accuracy": 0.21059284508228301,
      "num_tokens": 61879173.0,
      "step": 27020
    },
    {
      "entropy": 5.4422389507293705,
      "epoch": 2.67151047844998,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004291002576506932,
      "loss": 5.0285,
      "mean_token_accuracy": 0.21663536578416825,
      "num_tokens": 61889784.0,
      "step": 27025
    },
    {
      "entropy": 5.544349336624146,
      "epoch": 2.672004744958482,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042907454183516055,
      "loss": 5.0784,
      "mean_token_accuracy": 0.21403243094682695,
      "num_tokens": 61900295.0,
      "step": 27030
    },
    {
      "entropy": 5.407369327545166,
      "epoch": 2.672499011466983,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004290488222293702,
      "loss": 5.0137,
      "mean_token_accuracy": 0.21434864401817322,
      "num_tokens": 61911938.0,
      "step": 27035
    },
    {
      "entropy": 5.396992111206055,
      "epoch": 2.6729932779754844,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042902309883395497,
      "loss": 5.0672,
      "mean_token_accuracy": 0.20752326399087906,
      "num_tokens": 61922609.0,
      "step": 27040
    },
    {
      "entropy": 5.39989709854126,
      "epoch": 2.6734875444839856,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004289973716495476,
      "loss": 5.013,
      "mean_token_accuracy": 0.20684352219104768,
      "num_tokens": 61932744.0,
      "step": 27045
    },
    {
      "entropy": 5.403465700149536,
      "epoch": 2.673981810992487,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042897164067678094,
      "loss": 5.0443,
      "mean_token_accuracy": 0.21280886083841324,
      "num_tokens": 61944099.0,
      "step": 27050
    },
    {
      "entropy": 5.499967098236084,
      "epoch": 2.6744760775009886,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042894590591628813,
      "loss": 5.135,
      "mean_token_accuracy": 0.20270054638385773,
      "num_tokens": 61955241.0,
      "step": 27055
    },
    {
      "entropy": 5.3820994853973385,
      "epoch": 2.67497034400949,
      "grad_norm": 1.0,
      "learning_rate": 0.00042892016736870213,
      "loss": 5.0122,
      "mean_token_accuracy": 0.2121971696615219,
      "num_tokens": 61966578.0,
      "step": 27060
    },
    {
      "entropy": 5.396243524551392,
      "epoch": 2.6754646105179916,
      "grad_norm": 1.0,
      "learning_rate": 0.00042889442503465616,
      "loss": 4.9962,
      "mean_token_accuracy": 0.2186695635318756,
      "num_tokens": 61977628.0,
      "step": 27065
    },
    {
      "entropy": 5.600973892211914,
      "epoch": 2.675958877026493,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042886867891478344,
      "loss": 5.16,
      "mean_token_accuracy": 0.1974322810769081,
      "num_tokens": 61988904.0,
      "step": 27070
    },
    {
      "entropy": 5.475010585784912,
      "epoch": 2.676453143534994,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004288429290097173,
      "loss": 5.1404,
      "mean_token_accuracy": 0.20540257543325424,
      "num_tokens": 62000977.0,
      "step": 27075
    },
    {
      "entropy": 5.45012321472168,
      "epoch": 2.6769474100434953,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042881717532009125,
      "loss": 5.0962,
      "mean_token_accuracy": 0.20464272648096085,
      "num_tokens": 62011023.0,
      "step": 27080
    },
    {
      "entropy": 5.477319812774658,
      "epoch": 2.6774416765519966,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042879141784653887,
      "loss": 5.0876,
      "mean_token_accuracy": 0.2068314343690872,
      "num_tokens": 62022679.0,
      "step": 27085
    },
    {
      "entropy": 5.46135835647583,
      "epoch": 2.6779359430604983,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042876565658969375,
      "loss": 4.9993,
      "mean_token_accuracy": 0.2068876639008522,
      "num_tokens": 62033922.0,
      "step": 27090
    },
    {
      "entropy": 5.471785974502564,
      "epoch": 2.6784302095689996,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042873989155018955,
      "loss": 5.1392,
      "mean_token_accuracy": 0.20213527977466583,
      "num_tokens": 62045511.0,
      "step": 27095
    },
    {
      "entropy": 5.373157072067261,
      "epoch": 2.678924476077501,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004287141227286602,
      "loss": 4.9359,
      "mean_token_accuracy": 0.22093926519155502,
      "num_tokens": 62055113.0,
      "step": 27100
    },
    {
      "entropy": 5.559252309799194,
      "epoch": 2.6794187425860025,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004286883501257396,
      "loss": 5.1986,
      "mean_token_accuracy": 0.19876435548067092,
      "num_tokens": 62067808.0,
      "step": 27105
    },
    {
      "entropy": 5.4390770435333256,
      "epoch": 2.679913009094504,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042866257374206174,
      "loss": 5.0101,
      "mean_token_accuracy": 0.20870517641305925,
      "num_tokens": 62078266.0,
      "step": 27110
    },
    {
      "entropy": 5.460529327392578,
      "epoch": 2.680407275603005,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042863679357826076,
      "loss": 5.1102,
      "mean_token_accuracy": 0.1978542074561119,
      "num_tokens": 62091612.0,
      "step": 27115
    },
    {
      "entropy": 5.442612171173096,
      "epoch": 2.6809015421115063,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004286110096349708,
      "loss": 5.0819,
      "mean_token_accuracy": 0.20319596081972122,
      "num_tokens": 62104198.0,
      "step": 27120
    },
    {
      "entropy": 5.4053990840911865,
      "epoch": 2.681395808620008,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004285852219128261,
      "loss": 5.0108,
      "mean_token_accuracy": 0.20899137407541274,
      "num_tokens": 62115155.0,
      "step": 27125
    },
    {
      "entropy": 5.485495328903198,
      "epoch": 2.6818900751285093,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004285594304124612,
      "loss": 5.126,
      "mean_token_accuracy": 0.20180237144231797,
      "num_tokens": 62126555.0,
      "step": 27130
    },
    {
      "entropy": 5.505683279037475,
      "epoch": 2.6823843416370106,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042853363513451056,
      "loss": 5.0797,
      "mean_token_accuracy": 0.20305417627096176,
      "num_tokens": 62138976.0,
      "step": 27135
    },
    {
      "entropy": 5.480543851852417,
      "epoch": 2.6828786081455123,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042850783607960855,
      "loss": 5.1428,
      "mean_token_accuracy": 0.20103807747364044,
      "num_tokens": 62150453.0,
      "step": 27140
    },
    {
      "entropy": 5.386187314987183,
      "epoch": 2.6833728746540135,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042848203324839,
      "loss": 4.9057,
      "mean_token_accuracy": 0.22359127700328826,
      "num_tokens": 62162342.0,
      "step": 27145
    },
    {
      "entropy": 5.40433349609375,
      "epoch": 2.683867141162515,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004284562266414896,
      "loss": 5.0805,
      "mean_token_accuracy": 0.20785312056541444,
      "num_tokens": 62173850.0,
      "step": 27150
    },
    {
      "entropy": 5.3674815654754635,
      "epoch": 2.684361407671016,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042843041625954223,
      "loss": 4.989,
      "mean_token_accuracy": 0.20567018836736678,
      "num_tokens": 62185410.0,
      "step": 27155
    },
    {
      "entropy": 5.442239999771118,
      "epoch": 2.6848556741795178,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004284046021031828,
      "loss": 5.031,
      "mean_token_accuracy": 0.21022606492042542,
      "num_tokens": 62197620.0,
      "step": 27160
    },
    {
      "entropy": 5.4070922374725345,
      "epoch": 2.685349940688019,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004283787841730464,
      "loss": 5.0416,
      "mean_token_accuracy": 0.210693122446537,
      "num_tokens": 62209435.0,
      "step": 27165
    },
    {
      "entropy": 5.415926694869995,
      "epoch": 2.6858442071965203,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042835296246976806,
      "loss": 5.0616,
      "mean_token_accuracy": 0.20783940702676773,
      "num_tokens": 62219948.0,
      "step": 27170
    },
    {
      "entropy": 5.4536645889282225,
      "epoch": 2.686338473705022,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004283271369939831,
      "loss": 5.072,
      "mean_token_accuracy": 0.20454054921865464,
      "num_tokens": 62230299.0,
      "step": 27175
    },
    {
      "entropy": 5.492748689651489,
      "epoch": 2.6868327402135233,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042830130774632687,
      "loss": 5.0476,
      "mean_token_accuracy": 0.21218666732311248,
      "num_tokens": 62241952.0,
      "step": 27180
    },
    {
      "entropy": 5.465061044692993,
      "epoch": 2.6873270067220245,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004282754747274345,
      "loss": 5.0821,
      "mean_token_accuracy": 0.20951579362154008,
      "num_tokens": 62253200.0,
      "step": 27185
    },
    {
      "entropy": 5.4542553424835205,
      "epoch": 2.6878212732305258,
      "grad_norm": 1.125,
      "learning_rate": 0.00042824963793794174,
      "loss": 5.0542,
      "mean_token_accuracy": 0.21183974295854568,
      "num_tokens": 62264203.0,
      "step": 27190
    },
    {
      "entropy": 5.5155284881591795,
      "epoch": 2.688315539739027,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042822379737848424,
      "loss": 5.0442,
      "mean_token_accuracy": 0.2059824988245964,
      "num_tokens": 62276296.0,
      "step": 27195
    },
    {
      "entropy": 5.496863651275635,
      "epoch": 2.6888098062475287,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004281979530496974,
      "loss": 5.0938,
      "mean_token_accuracy": 0.20391029268503189,
      "num_tokens": 62287827.0,
      "step": 27200
    },
    {
      "entropy": 5.518351745605469,
      "epoch": 2.68930407275603,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004281721049522172,
      "loss": 5.1173,
      "mean_token_accuracy": 0.19997261017560958,
      "num_tokens": 62299451.0,
      "step": 27205
    },
    {
      "entropy": 5.4263482093811035,
      "epoch": 2.6897983392645317,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042814625308667945,
      "loss": 5.0588,
      "mean_token_accuracy": 0.20434576272964478,
      "num_tokens": 62311508.0,
      "step": 27210
    },
    {
      "entropy": 5.481621170043946,
      "epoch": 2.690292605773033,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042812039745372004,
      "loss": 5.0282,
      "mean_token_accuracy": 0.22057853490114213,
      "num_tokens": 62323279.0,
      "step": 27215
    },
    {
      "entropy": 5.553976106643677,
      "epoch": 2.6907868722815342,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042809453805397525,
      "loss": 5.1674,
      "mean_token_accuracy": 0.20084467232227327,
      "num_tokens": 62335897.0,
      "step": 27220
    },
    {
      "entropy": 5.468023204803467,
      "epoch": 2.6912811387900355,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000428068674888081,
      "loss": 5.0745,
      "mean_token_accuracy": 0.20778805017471313,
      "num_tokens": 62347090.0,
      "step": 27225
    },
    {
      "entropy": 5.480611276626587,
      "epoch": 2.6917754052985368,
      "grad_norm": 1.125,
      "learning_rate": 0.00042804280795667363,
      "loss": 5.0644,
      "mean_token_accuracy": 0.20592082291841507,
      "num_tokens": 62358211.0,
      "step": 27230
    },
    {
      "entropy": 5.440846920013428,
      "epoch": 2.6922696718070385,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004280169372603894,
      "loss": 5.0577,
      "mean_token_accuracy": 0.21047901064157487,
      "num_tokens": 62368628.0,
      "step": 27235
    },
    {
      "entropy": 5.444861125946045,
      "epoch": 2.6927639383155397,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004279910627998647,
      "loss": 5.066,
      "mean_token_accuracy": 0.20430624783039092,
      "num_tokens": 62380633.0,
      "step": 27240
    },
    {
      "entropy": 5.511349391937256,
      "epoch": 2.693258204824041,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042796518457573625,
      "loss": 5.1343,
      "mean_token_accuracy": 0.20013356506824492,
      "num_tokens": 62392615.0,
      "step": 27245
    },
    {
      "entropy": 5.460319995880127,
      "epoch": 2.6937524713325427,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004279393025886405,
      "loss": 5.0446,
      "mean_token_accuracy": 0.21847428679466246,
      "num_tokens": 62403087.0,
      "step": 27250
    },
    {
      "entropy": 5.512298202514648,
      "epoch": 2.694246737841044,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042791341683921417,
      "loss": 5.0894,
      "mean_token_accuracy": 0.2014314502477646,
      "num_tokens": 62415487.0,
      "step": 27255
    },
    {
      "entropy": 5.389772081375122,
      "epoch": 2.694741004349545,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004278875273280941,
      "loss": 5.0265,
      "mean_token_accuracy": 0.21235078871250151,
      "num_tokens": 62428123.0,
      "step": 27260
    },
    {
      "entropy": 5.458454656600952,
      "epoch": 2.6952352708580465,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004278616340559171,
      "loss": 5.0621,
      "mean_token_accuracy": 0.20406317114830017,
      "num_tokens": 62440250.0,
      "step": 27265
    },
    {
      "entropy": 5.457391452789307,
      "epoch": 2.695729537366548,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042783573702332014,
      "loss": 5.0262,
      "mean_token_accuracy": 0.2083328604698181,
      "num_tokens": 62451500.0,
      "step": 27270
    },
    {
      "entropy": 5.446199989318847,
      "epoch": 2.6962238038750495,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042780983623094037,
      "loss": 5.0518,
      "mean_token_accuracy": 0.21355418711900712,
      "num_tokens": 62461839.0,
      "step": 27275
    },
    {
      "entropy": 5.3950074195861815,
      "epoch": 2.6967180703835507,
      "grad_norm": 1.125,
      "learning_rate": 0.0004277839316794149,
      "loss": 5.0532,
      "mean_token_accuracy": 0.20838071256875992,
      "num_tokens": 62473650.0,
      "step": 27280
    },
    {
      "entropy": 5.475646591186523,
      "epoch": 2.6972123368920524,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042775802336938106,
      "loss": 5.0796,
      "mean_token_accuracy": 0.20373637229204178,
      "num_tokens": 62484024.0,
      "step": 27285
    },
    {
      "entropy": 5.504436016082764,
      "epoch": 2.6977066034005537,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004277321113014762,
      "loss": 5.0704,
      "mean_token_accuracy": 0.2018016442656517,
      "num_tokens": 62495958.0,
      "step": 27290
    },
    {
      "entropy": 5.459137296676635,
      "epoch": 2.698200869909055,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042770619547633767,
      "loss": 5.1683,
      "mean_token_accuracy": 0.20532295107841492,
      "num_tokens": 62508474.0,
      "step": 27295
    },
    {
      "entropy": 5.470155715942383,
      "epoch": 2.698695136417556,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042768027589460295,
      "loss": 5.0796,
      "mean_token_accuracy": 0.21067889779806137,
      "num_tokens": 62520281.0,
      "step": 27300
    },
    {
      "entropy": 5.514051628112793,
      "epoch": 2.6991894029260575,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004276543525569098,
      "loss": 5.083,
      "mean_token_accuracy": 0.2080689936876297,
      "num_tokens": 62532242.0,
      "step": 27305
    },
    {
      "entropy": 5.416023111343383,
      "epoch": 2.699683669434559,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042762842546389593,
      "loss": 5.0818,
      "mean_token_accuracy": 0.2080347642302513,
      "num_tokens": 62544416.0,
      "step": 27310
    },
    {
      "entropy": 5.469554901123047,
      "epoch": 2.7001779359430604,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042760249461619914,
      "loss": 5.0897,
      "mean_token_accuracy": 0.20498355925083162,
      "num_tokens": 62555932.0,
      "step": 27315
    },
    {
      "entropy": 5.36395788192749,
      "epoch": 2.700672202451562,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042757656001445727,
      "loss": 4.9818,
      "mean_token_accuracy": 0.21798075735569,
      "num_tokens": 62567322.0,
      "step": 27320
    },
    {
      "entropy": 5.404558372497559,
      "epoch": 2.7011664689600634,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004275506216593083,
      "loss": 5.019,
      "mean_token_accuracy": 0.2134648784995079,
      "num_tokens": 62577491.0,
      "step": 27325
    },
    {
      "entropy": 5.45589690208435,
      "epoch": 2.7016607354685647,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004275246795513904,
      "loss": 5.0758,
      "mean_token_accuracy": 0.21003155708312987,
      "num_tokens": 62589912.0,
      "step": 27330
    },
    {
      "entropy": 5.398430824279785,
      "epoch": 2.702155001977066,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004274987336913418,
      "loss": 5.0774,
      "mean_token_accuracy": 0.21025702804327012,
      "num_tokens": 62601571.0,
      "step": 27335
    },
    {
      "entropy": 5.444499397277832,
      "epoch": 2.702649268485567,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004274727840798007,
      "loss": 5.0811,
      "mean_token_accuracy": 0.20036977976560594,
      "num_tokens": 62615006.0,
      "step": 27340
    },
    {
      "entropy": 5.524222040176392,
      "epoch": 2.703143534994069,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042744683071740524,
      "loss": 5.1513,
      "mean_token_accuracy": 0.2061564400792122,
      "num_tokens": 62626678.0,
      "step": 27345
    },
    {
      "entropy": 5.3929688930511475,
      "epoch": 2.70363780150257,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004274208736047944,
      "loss": 5.011,
      "mean_token_accuracy": 0.21418582797050476,
      "num_tokens": 62637819.0,
      "step": 27350
    },
    {
      "entropy": 5.354187965393066,
      "epoch": 2.7041320680110714,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042739491274260615,
      "loss": 4.9462,
      "mean_token_accuracy": 0.2202274441719055,
      "num_tokens": 62649705.0,
      "step": 27355
    },
    {
      "entropy": 5.379169130325318,
      "epoch": 2.704626334519573,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042736894813147954,
      "loss": 4.9763,
      "mean_token_accuracy": 0.20851344615221024,
      "num_tokens": 62661695.0,
      "step": 27360
    },
    {
      "entropy": 5.575306749343872,
      "epoch": 2.7051206010280744,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042734297977205325,
      "loss": 5.1858,
      "mean_token_accuracy": 0.19385486841201782,
      "num_tokens": 62674102.0,
      "step": 27365
    },
    {
      "entropy": 5.534251880645752,
      "epoch": 2.7056148675365757,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042731700766496586,
      "loss": 5.121,
      "mean_token_accuracy": 0.2099744826555252,
      "num_tokens": 62685562.0,
      "step": 27370
    },
    {
      "entropy": 5.451116847991943,
      "epoch": 2.706109134045077,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004272910318108565,
      "loss": 5.06,
      "mean_token_accuracy": 0.20875783413648605,
      "num_tokens": 62696653.0,
      "step": 27375
    },
    {
      "entropy": 5.464148378372192,
      "epoch": 2.7066034005535786,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042726505221036415,
      "loss": 5.0649,
      "mean_token_accuracy": 0.2096223697066307,
      "num_tokens": 62708760.0,
      "step": 27380
    },
    {
      "entropy": 5.5038727760314945,
      "epoch": 2.70709766706208,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004272390688641279,
      "loss": 5.0587,
      "mean_token_accuracy": 0.2037187173962593,
      "num_tokens": 62718548.0,
      "step": 27385
    },
    {
      "entropy": 5.4387062072753904,
      "epoch": 2.707591933570581,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042721308177278696,
      "loss": 5.0368,
      "mean_token_accuracy": 0.20857273787260056,
      "num_tokens": 62730792.0,
      "step": 27390
    },
    {
      "entropy": 5.499172115325928,
      "epoch": 2.708086200079083,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042718709093698054,
      "loss": 5.0955,
      "mean_token_accuracy": 0.2078446701169014,
      "num_tokens": 62741528.0,
      "step": 27395
    },
    {
      "entropy": 5.4290628910064695,
      "epoch": 2.708580466587584,
      "grad_norm": 1.125,
      "learning_rate": 0.00042716109635734806,
      "loss": 5.0048,
      "mean_token_accuracy": 0.2135871633887291,
      "num_tokens": 62751650.0,
      "step": 27400
    },
    {
      "entropy": 5.441471338272095,
      "epoch": 2.7090747330960854,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042713509803452904,
      "loss": 5.048,
      "mean_token_accuracy": 0.21088856011629104,
      "num_tokens": 62763248.0,
      "step": 27405
    },
    {
      "entropy": 5.467253971099853,
      "epoch": 2.7095689996045866,
      "grad_norm": 1.0,
      "learning_rate": 0.00042710909596916307,
      "loss": 5.0747,
      "mean_token_accuracy": 0.20864329040050505,
      "num_tokens": 62775696.0,
      "step": 27410
    },
    {
      "entropy": 5.43760838508606,
      "epoch": 2.7100632661130883,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042708309016188965,
      "loss": 4.9803,
      "mean_token_accuracy": 0.2213533490896225,
      "num_tokens": 62786564.0,
      "step": 27415
    },
    {
      "entropy": 5.416453504562378,
      "epoch": 2.7105575326215896,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004270570806133486,
      "loss": 5.0772,
      "mean_token_accuracy": 0.2080509305000305,
      "num_tokens": 62799235.0,
      "step": 27420
    },
    {
      "entropy": 5.412689971923828,
      "epoch": 2.711051799130091,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004270310673241799,
      "loss": 5.0303,
      "mean_token_accuracy": 0.2080045834183693,
      "num_tokens": 62809605.0,
      "step": 27425
    },
    {
      "entropy": 5.442185163497925,
      "epoch": 2.7115460656385926,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004270050502950233,
      "loss": 5.0367,
      "mean_token_accuracy": 0.2093183919787407,
      "num_tokens": 62820900.0,
      "step": 27430
    },
    {
      "entropy": 5.376932048797608,
      "epoch": 2.712040332147094,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004269790295265189,
      "loss": 4.9765,
      "mean_token_accuracy": 0.21994495391845703,
      "num_tokens": 62831761.0,
      "step": 27435
    },
    {
      "entropy": 5.526027488708496,
      "epoch": 2.712534598655595,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042695300501930674,
      "loss": 5.1199,
      "mean_token_accuracy": 0.20050384551286698,
      "num_tokens": 62842512.0,
      "step": 27440
    },
    {
      "entropy": 5.463005018234253,
      "epoch": 2.7130288651640964,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004269269767740271,
      "loss": 5.0904,
      "mean_token_accuracy": 0.20866186767816544,
      "num_tokens": 62853495.0,
      "step": 27445
    },
    {
      "entropy": 5.393105792999267,
      "epoch": 2.7135231316725976,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042690094479132023,
      "loss": 4.9651,
      "mean_token_accuracy": 0.22447549998760224,
      "num_tokens": 62864524.0,
      "step": 27450
    },
    {
      "entropy": 5.474991416931152,
      "epoch": 2.7140173981810993,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042687490907182665,
      "loss": 5.1734,
      "mean_token_accuracy": 0.19940508008003235,
      "num_tokens": 62874522.0,
      "step": 27455
    },
    {
      "entropy": 5.479139280319214,
      "epoch": 2.7145116646896006,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042684886961618664,
      "loss": 5.0437,
      "mean_token_accuracy": 0.20800797492265702,
      "num_tokens": 62885477.0,
      "step": 27460
    },
    {
      "entropy": 5.4705112934112545,
      "epoch": 2.715005931198102,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042682282642504096,
      "loss": 5.079,
      "mean_token_accuracy": 0.21376798897981644,
      "num_tokens": 62896140.0,
      "step": 27465
    },
    {
      "entropy": 5.538098764419556,
      "epoch": 2.7155001977066036,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042679677949903017,
      "loss": 5.1384,
      "mean_token_accuracy": 0.2000359058380127,
      "num_tokens": 62907707.0,
      "step": 27470
    },
    {
      "entropy": 5.530870151519776,
      "epoch": 2.715994464215105,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042677072883879496,
      "loss": 5.1663,
      "mean_token_accuracy": 0.20053110122680665,
      "num_tokens": 62919209.0,
      "step": 27475
    },
    {
      "entropy": 5.3769608497619625,
      "epoch": 2.716488730723606,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042674467444497644,
      "loss": 4.9039,
      "mean_token_accuracy": 0.22954662889242172,
      "num_tokens": 62931971.0,
      "step": 27480
    },
    {
      "entropy": 5.45075421333313,
      "epoch": 2.7169829972321073,
      "grad_norm": 1.0,
      "learning_rate": 0.0004267186163182153,
      "loss": 5.1173,
      "mean_token_accuracy": 0.20684964209794998,
      "num_tokens": 62944246.0,
      "step": 27485
    },
    {
      "entropy": 5.489340114593506,
      "epoch": 2.717477263740609,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004266925544591527,
      "loss": 5.1544,
      "mean_token_accuracy": 0.20345284044742584,
      "num_tokens": 62956973.0,
      "step": 27490
    },
    {
      "entropy": 5.530171728134155,
      "epoch": 2.7179715302491103,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004266664888684297,
      "loss": 5.1356,
      "mean_token_accuracy": 0.1969824105501175,
      "num_tokens": 62968590.0,
      "step": 27495
    },
    {
      "entropy": 5.491403293609619,
      "epoch": 2.7184657967576116,
      "grad_norm": 1.0,
      "learning_rate": 0.00042664041954668753,
      "loss": 5.0693,
      "mean_token_accuracy": 0.21268833428621292,
      "num_tokens": 62980382.0,
      "step": 27500
    },
    {
      "entropy": 5.420768737792969,
      "epoch": 2.7189600632661133,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004266143464945675,
      "loss": 5.0492,
      "mean_token_accuracy": 0.20863761156797409,
      "num_tokens": 62990432.0,
      "step": 27505
    },
    {
      "entropy": 5.466956424713135,
      "epoch": 2.7194543297746145,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042658826971271103,
      "loss": 5.1129,
      "mean_token_accuracy": 0.20202534049749374,
      "num_tokens": 63003341.0,
      "step": 27510
    },
    {
      "entropy": 5.52876844406128,
      "epoch": 2.719948596283116,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042656218920175963,
      "loss": 5.1073,
      "mean_token_accuracy": 0.2001764729619026,
      "num_tokens": 63015527.0,
      "step": 27515
    },
    {
      "entropy": 5.497215366363525,
      "epoch": 2.720442862791617,
      "grad_norm": 1.0,
      "learning_rate": 0.000426536104962355,
      "loss": 5.0633,
      "mean_token_accuracy": 0.2107557475566864,
      "num_tokens": 63026995.0,
      "step": 27520
    },
    {
      "entropy": 5.41289439201355,
      "epoch": 2.7209371293001188,
      "grad_norm": 1.0,
      "learning_rate": 0.0004265100169951385,
      "loss": 4.9735,
      "mean_token_accuracy": 0.21499338746070862,
      "num_tokens": 63037156.0,
      "step": 27525
    },
    {
      "entropy": 5.4538195610046385,
      "epoch": 2.72143139580862,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004264839253007521,
      "loss": 5.0969,
      "mean_token_accuracy": 0.21453964114189147,
      "num_tokens": 63047513.0,
      "step": 27530
    },
    {
      "entropy": 5.44812273979187,
      "epoch": 2.7219256623171213,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042645782987983763,
      "loss": 5.0133,
      "mean_token_accuracy": 0.20725918412208558,
      "num_tokens": 63059968.0,
      "step": 27535
    },
    {
      "entropy": 5.415910005569458,
      "epoch": 2.722419928825623,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042643173073303713,
      "loss": 5.051,
      "mean_token_accuracy": 0.20527620911598204,
      "num_tokens": 63071514.0,
      "step": 27540
    },
    {
      "entropy": 5.516444492340088,
      "epoch": 2.7229141953341243,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042640562786099245,
      "loss": 5.079,
      "mean_token_accuracy": 0.2103402942419052,
      "num_tokens": 63082810.0,
      "step": 27545
    },
    {
      "entropy": 5.489263868331909,
      "epoch": 2.7234084618426255,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042637952126434593,
      "loss": 5.0581,
      "mean_token_accuracy": 0.21370716840028764,
      "num_tokens": 63093324.0,
      "step": 27550
    },
    {
      "entropy": 5.4249077320098875,
      "epoch": 2.723902728351127,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004263534109437397,
      "loss": 5.0982,
      "mean_token_accuracy": 0.2104247182607651,
      "num_tokens": 63104834.0,
      "step": 27555
    },
    {
      "entropy": 5.443571281433106,
      "epoch": 2.724396994859628,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042632729689981604,
      "loss": 5.0736,
      "mean_token_accuracy": 0.207258041203022,
      "num_tokens": 63116625.0,
      "step": 27560
    },
    {
      "entropy": 5.470551252365112,
      "epoch": 2.7248912613681298,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042630117913321734,
      "loss": 5.0632,
      "mean_token_accuracy": 0.20198529958724976,
      "num_tokens": 63127616.0,
      "step": 27565
    },
    {
      "entropy": 5.384239768981933,
      "epoch": 2.725385527876631,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042627505764458626,
      "loss": 5.0587,
      "mean_token_accuracy": 0.21460792124271394,
      "num_tokens": 63140546.0,
      "step": 27570
    },
    {
      "entropy": 5.411870956420898,
      "epoch": 2.7258797943851327,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004262489324345652,
      "loss": 4.9648,
      "mean_token_accuracy": 0.2166174441576004,
      "num_tokens": 63151693.0,
      "step": 27575
    },
    {
      "entropy": 5.448649835586548,
      "epoch": 2.726374060893634,
      "grad_norm": 1.125,
      "learning_rate": 0.0004262228035037971,
      "loss": 4.9926,
      "mean_token_accuracy": 0.20525157898664476,
      "num_tokens": 63162569.0,
      "step": 27580
    },
    {
      "entropy": 5.324292802810669,
      "epoch": 2.7268683274021353,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004261966708529244,
      "loss": 4.9166,
      "mean_token_accuracy": 0.22584639340639115,
      "num_tokens": 63174073.0,
      "step": 27585
    },
    {
      "entropy": 5.462039232254028,
      "epoch": 2.7273625939106365,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042617053448259026,
      "loss": 5.0254,
      "mean_token_accuracy": 0.21396956145763396,
      "num_tokens": 63184752.0,
      "step": 27590
    },
    {
      "entropy": 5.391912269592285,
      "epoch": 2.7278568604191378,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004261443943934375,
      "loss": 4.9913,
      "mean_token_accuracy": 0.21298307776451111,
      "num_tokens": 63197682.0,
      "step": 27595
    },
    {
      "entropy": 5.519440174102783,
      "epoch": 2.7283511269276395,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042611825058610917,
      "loss": 5.1228,
      "mean_token_accuracy": 0.19199850857257844,
      "num_tokens": 63208868.0,
      "step": 27600
    },
    {
      "entropy": 5.4924345970153805,
      "epoch": 2.7288453934361407,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004260921030612483,
      "loss": 5.153,
      "mean_token_accuracy": 0.20959761440753938,
      "num_tokens": 63220418.0,
      "step": 27605
    },
    {
      "entropy": 5.45736026763916,
      "epoch": 2.729339659944642,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004260659518194985,
      "loss": 5.0231,
      "mean_token_accuracy": 0.21358337253332138,
      "num_tokens": 63231431.0,
      "step": 27610
    },
    {
      "entropy": 5.424257850646972,
      "epoch": 2.7298339264531437,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004260397968615027,
      "loss": 5.0552,
      "mean_token_accuracy": 0.2042704552412033,
      "num_tokens": 63243994.0,
      "step": 27615
    },
    {
      "entropy": 5.3735960006713865,
      "epoch": 2.730328192961645,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004260136381879045,
      "loss": 4.9751,
      "mean_token_accuracy": 0.22144197076559066,
      "num_tokens": 63255414.0,
      "step": 27620
    },
    {
      "entropy": 5.493383836746216,
      "epoch": 2.7308224594701462,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042598747579934736,
      "loss": 5.156,
      "mean_token_accuracy": 0.20417538583278655,
      "num_tokens": 63266289.0,
      "step": 27625
    },
    {
      "entropy": 5.448935508728027,
      "epoch": 2.7313167259786475,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042596130969647494,
      "loss": 4.9666,
      "mean_token_accuracy": 0.21250412315130235,
      "num_tokens": 63276497.0,
      "step": 27630
    },
    {
      "entropy": 5.43988184928894,
      "epoch": 2.731810992487149,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004259351398799308,
      "loss": 5.0213,
      "mean_token_accuracy": 0.21391626596450805,
      "num_tokens": 63286892.0,
      "step": 27635
    },
    {
      "entropy": 5.471703815460205,
      "epoch": 2.7323052589956505,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00042590896635035885,
      "loss": 5.0975,
      "mean_token_accuracy": 0.2043243631720543,
      "num_tokens": 63299978.0,
      "step": 27640
    },
    {
      "entropy": 5.45467095375061,
      "epoch": 2.7327995255041517,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042588278910840296,
      "loss": 5.0739,
      "mean_token_accuracy": 0.2128247633576393,
      "num_tokens": 63311916.0,
      "step": 27645
    },
    {
      "entropy": 5.445849752426147,
      "epoch": 2.7332937920126534,
      "grad_norm": 1.078125,
      "learning_rate": 0.000425856608154707,
      "loss": 5.027,
      "mean_token_accuracy": 0.20853590071201325,
      "num_tokens": 63322994.0,
      "step": 27650
    },
    {
      "entropy": 5.524616289138794,
      "epoch": 2.7337880585211547,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004258304234899151,
      "loss": 5.108,
      "mean_token_accuracy": 0.20361389517784118,
      "num_tokens": 63334708.0,
      "step": 27655
    },
    {
      "entropy": 5.420170688629151,
      "epoch": 2.734282325029656,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004258042351146713,
      "loss": 5.0071,
      "mean_token_accuracy": 0.20895163118839263,
      "num_tokens": 63346823.0,
      "step": 27660
    },
    {
      "entropy": 5.48543791770935,
      "epoch": 2.734776591538157,
      "grad_norm": 1.125,
      "learning_rate": 0.00042577804302961994,
      "loss": 5.056,
      "mean_token_accuracy": 0.20693513453006745,
      "num_tokens": 63357677.0,
      "step": 27665
    },
    {
      "entropy": 5.35520920753479,
      "epoch": 2.7352708580466585,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042575184723540546,
      "loss": 4.9336,
      "mean_token_accuracy": 0.21335779279470443,
      "num_tokens": 63369685.0,
      "step": 27670
    },
    {
      "entropy": 5.487816953659058,
      "epoch": 2.73576512455516,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042572564773267203,
      "loss": 5.0938,
      "mean_token_accuracy": 0.20345963537693024,
      "num_tokens": 63380340.0,
      "step": 27675
    },
    {
      "entropy": 5.438711023330688,
      "epoch": 2.7362593910636615,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004256994445220643,
      "loss": 5.0813,
      "mean_token_accuracy": 0.21313631683588027,
      "num_tokens": 63390868.0,
      "step": 27680
    },
    {
      "entropy": 5.439948415756225,
      "epoch": 2.736753657572163,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004256732376042269,
      "loss": 5.0371,
      "mean_token_accuracy": 0.21079247295856476,
      "num_tokens": 63401669.0,
      "step": 27685
    },
    {
      "entropy": 5.436453533172608,
      "epoch": 2.7372479240806644,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004256470269798044,
      "loss": 5.0228,
      "mean_token_accuracy": 0.21370188444852828,
      "num_tokens": 63412593.0,
      "step": 27690
    },
    {
      "entropy": 5.449593734741211,
      "epoch": 2.7377421905891657,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042562081264944175,
      "loss": 5.0853,
      "mean_token_accuracy": 0.21506880670785905,
      "num_tokens": 63423952.0,
      "step": 27695
    },
    {
      "entropy": 5.484494066238403,
      "epoch": 2.738236457097667,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042559459461378367,
      "loss": 5.0926,
      "mean_token_accuracy": 0.20260501205921172,
      "num_tokens": 63435152.0,
      "step": 27700
    },
    {
      "entropy": 5.40783724784851,
      "epoch": 2.738730723606168,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042556837287347524,
      "loss": 4.9209,
      "mean_token_accuracy": 0.22029463052749634,
      "num_tokens": 63445977.0,
      "step": 27705
    },
    {
      "entropy": 5.444705581665039,
      "epoch": 2.73922499011467,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042554214742916154,
      "loss": 5.1358,
      "mean_token_accuracy": 0.20119498819112777,
      "num_tokens": 63456790.0,
      "step": 27710
    },
    {
      "entropy": 5.448411178588867,
      "epoch": 2.739719256623171,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004255159182814876,
      "loss": 5.0343,
      "mean_token_accuracy": 0.2181098908185959,
      "num_tokens": 63467775.0,
      "step": 27715
    },
    {
      "entropy": 5.518222951889038,
      "epoch": 2.7402135231316724,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004254896854310987,
      "loss": 5.1537,
      "mean_token_accuracy": 0.19921877533197402,
      "num_tokens": 63479445.0,
      "step": 27720
    },
    {
      "entropy": 5.510286331176758,
      "epoch": 2.740707789640174,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004254634488786402,
      "loss": 5.143,
      "mean_token_accuracy": 0.1973639115691185,
      "num_tokens": 63489938.0,
      "step": 27725
    },
    {
      "entropy": 5.4932373046875,
      "epoch": 2.7412020561486754,
      "grad_norm": 1.59375,
      "learning_rate": 0.00042543720862475747,
      "loss": 5.0473,
      "mean_token_accuracy": 0.21136311441659927,
      "num_tokens": 63501509.0,
      "step": 27730
    },
    {
      "entropy": 5.411088848114014,
      "epoch": 2.7416963226571767,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042541096467009617,
      "loss": 5.0005,
      "mean_token_accuracy": 0.21368778347969056,
      "num_tokens": 63512222.0,
      "step": 27735
    },
    {
      "entropy": 5.449278211593628,
      "epoch": 2.742190589165678,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042538471701530174,
      "loss": 5.155,
      "mean_token_accuracy": 0.20174836069345475,
      "num_tokens": 63524841.0,
      "step": 27740
    },
    {
      "entropy": 5.489743947982788,
      "epoch": 2.7426848556741796,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042535846566102,
      "loss": 5.026,
      "mean_token_accuracy": 0.20561160296201705,
      "num_tokens": 63536167.0,
      "step": 27745
    },
    {
      "entropy": 5.453299570083618,
      "epoch": 2.743179122182681,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004253322106078966,
      "loss": 5.0374,
      "mean_token_accuracy": 0.2143303319811821,
      "num_tokens": 63547774.0,
      "step": 27750
    },
    {
      "entropy": 5.490023469924926,
      "epoch": 2.743673388691182,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042530595185657756,
      "loss": 5.1652,
      "mean_token_accuracy": 0.20143587440252303,
      "num_tokens": 63560712.0,
      "step": 27755
    },
    {
      "entropy": 5.472793054580689,
      "epoch": 2.744167655199684,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042527968940770874,
      "loss": 5.1069,
      "mean_token_accuracy": 0.2109362468123436,
      "num_tokens": 63572587.0,
      "step": 27760
    },
    {
      "entropy": 5.519396781921387,
      "epoch": 2.744661921708185,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042525342326193624,
      "loss": 5.1323,
      "mean_token_accuracy": 0.20220906138420106,
      "num_tokens": 63584694.0,
      "step": 27765
    },
    {
      "entropy": 5.463712692260742,
      "epoch": 2.7451561882166864,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00042522715341990625,
      "loss": 5.0849,
      "mean_token_accuracy": 0.20479255467653273,
      "num_tokens": 63597894.0,
      "step": 27770
    },
    {
      "entropy": 5.4457048892974855,
      "epoch": 2.7456504547251877,
      "grad_norm": 1.0625,
      "learning_rate": 0.000425200879882265,
      "loss": 5.1013,
      "mean_token_accuracy": 0.2055336356163025,
      "num_tokens": 63608873.0,
      "step": 27775
    },
    {
      "entropy": 5.531204605102539,
      "epoch": 2.7461447212336894,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004251746026496588,
      "loss": 5.1105,
      "mean_token_accuracy": 0.2069559320807457,
      "num_tokens": 63621013.0,
      "step": 27780
    },
    {
      "entropy": 5.52852611541748,
      "epoch": 2.7466389877421906,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042514832172273404,
      "loss": 5.1722,
      "mean_token_accuracy": 0.19898469895124435,
      "num_tokens": 63633438.0,
      "step": 27785
    },
    {
      "entropy": 5.460723781585694,
      "epoch": 2.747133254250692,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042512203710213733,
      "loss": 4.9892,
      "mean_token_accuracy": 0.20972118079662322,
      "num_tokens": 63645032.0,
      "step": 27790
    },
    {
      "entropy": 5.425799751281739,
      "epoch": 2.7476275207591936,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004250957487885151,
      "loss": 5.0036,
      "mean_token_accuracy": 0.21543105244636535,
      "num_tokens": 63657747.0,
      "step": 27795
    },
    {
      "entropy": 5.412010526657104,
      "epoch": 2.748121787267695,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004250694567825143,
      "loss": 5.0146,
      "mean_token_accuracy": 0.20998825132846832,
      "num_tokens": 63669472.0,
      "step": 27800
    },
    {
      "entropy": 5.424338054656983,
      "epoch": 2.748616053776196,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042504316108478153,
      "loss": 5.0666,
      "mean_token_accuracy": 0.21615904718637466,
      "num_tokens": 63681498.0,
      "step": 27805
    },
    {
      "entropy": 5.349238491058349,
      "epoch": 2.7491103202846974,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042501686169596367,
      "loss": 4.9549,
      "mean_token_accuracy": 0.22451326996088028,
      "num_tokens": 63692628.0,
      "step": 27810
    },
    {
      "entropy": 5.458430385589599,
      "epoch": 2.7496045867931986,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042499055861670783,
      "loss": 5.0326,
      "mean_token_accuracy": 0.20796682238578795,
      "num_tokens": 63704732.0,
      "step": 27815
    },
    {
      "entropy": 5.5318440914154055,
      "epoch": 2.7500988533017003,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004249642518476609,
      "loss": 5.1085,
      "mean_token_accuracy": 0.20424683392047882,
      "num_tokens": 63716346.0,
      "step": 27820
    },
    {
      "entropy": 5.473616027832032,
      "epoch": 2.7505931198102016,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042493794138947013,
      "loss": 5.0789,
      "mean_token_accuracy": 0.2096447989344597,
      "num_tokens": 63728320.0,
      "step": 27825
    },
    {
      "entropy": 5.426529121398926,
      "epoch": 2.7510873863187033,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042491162724278276,
      "loss": 5.112,
      "mean_token_accuracy": 0.20659882873296737,
      "num_tokens": 63740211.0,
      "step": 27830
    },
    {
      "entropy": 5.430647134780884,
      "epoch": 2.7515816528272046,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042488530940824616,
      "loss": 4.9574,
      "mean_token_accuracy": 0.2154908522963524,
      "num_tokens": 63751706.0,
      "step": 27835
    },
    {
      "entropy": 5.458653593063355,
      "epoch": 2.752075919335706,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004248589878865076,
      "loss": 5.076,
      "mean_token_accuracy": 0.20776806771755219,
      "num_tokens": 63764045.0,
      "step": 27840
    },
    {
      "entropy": 5.407937383651733,
      "epoch": 2.752570185844207,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004248326626782147,
      "loss": 4.9849,
      "mean_token_accuracy": 0.21759278923273087,
      "num_tokens": 63774148.0,
      "step": 27845
    },
    {
      "entropy": 5.450107145309448,
      "epoch": 2.7530644523527084,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042480633378401503,
      "loss": 5.008,
      "mean_token_accuracy": 0.21652012914419175,
      "num_tokens": 63785202.0,
      "step": 27850
    },
    {
      "entropy": 5.486587619781494,
      "epoch": 2.75355871886121,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004247800012045564,
      "loss": 5.0352,
      "mean_token_accuracy": 0.20686526000499725,
      "num_tokens": 63796990.0,
      "step": 27855
    },
    {
      "entropy": 5.385894393920898,
      "epoch": 2.7540529853697113,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042475366494048645,
      "loss": 4.9545,
      "mean_token_accuracy": 0.22018436789512635,
      "num_tokens": 63807201.0,
      "step": 27860
    },
    {
      "entropy": 5.407347297668457,
      "epoch": 2.7545472518782126,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042472732499245316,
      "loss": 5.0352,
      "mean_token_accuracy": 0.2168760195374489,
      "num_tokens": 63818805.0,
      "step": 27865
    },
    {
      "entropy": 5.438308429718018,
      "epoch": 2.7550415183867143,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042470098136110444,
      "loss": 4.987,
      "mean_token_accuracy": 0.21575102359056472,
      "num_tokens": 63829444.0,
      "step": 27870
    },
    {
      "entropy": 5.389496183395385,
      "epoch": 2.7555357848952156,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042467463404708826,
      "loss": 5.0168,
      "mean_token_accuracy": 0.21383088678121567,
      "num_tokens": 63840868.0,
      "step": 27875
    },
    {
      "entropy": 5.41587986946106,
      "epoch": 2.756030051403717,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042464828305105283,
      "loss": 5.0487,
      "mean_token_accuracy": 0.2067001923918724,
      "num_tokens": 63851992.0,
      "step": 27880
    },
    {
      "entropy": 5.463610696792602,
      "epoch": 2.756524317912218,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042462192837364654,
      "loss": 5.0521,
      "mean_token_accuracy": 0.20601622462272645,
      "num_tokens": 63863362.0,
      "step": 27885
    },
    {
      "entropy": 5.4817568302154545,
      "epoch": 2.75701858442072,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00042459557001551764,
      "loss": 5.119,
      "mean_token_accuracy": 0.20808735340833664,
      "num_tokens": 63875712.0,
      "step": 27890
    },
    {
      "entropy": 5.502285003662109,
      "epoch": 2.757512850929221,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042456920797731443,
      "loss": 5.021,
      "mean_token_accuracy": 0.2129630535840988,
      "num_tokens": 63887253.0,
      "step": 27895
    },
    {
      "entropy": 5.50234088897705,
      "epoch": 2.7580071174377223,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042454284225968553,
      "loss": 5.138,
      "mean_token_accuracy": 0.20281088948249817,
      "num_tokens": 63898328.0,
      "step": 27900
    },
    {
      "entropy": 5.396198081970215,
      "epoch": 2.758501383946224,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004245164728632795,
      "loss": 4.9217,
      "mean_token_accuracy": 0.22561803013086318,
      "num_tokens": 63908080.0,
      "step": 27905
    },
    {
      "entropy": 5.4357017993927,
      "epoch": 2.7589956504547253,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000424490099788745,
      "loss": 5.1029,
      "mean_token_accuracy": 0.20584024488925934,
      "num_tokens": 63918860.0,
      "step": 27910
    },
    {
      "entropy": 5.459130859375,
      "epoch": 2.7594899169632265,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004244637230367309,
      "loss": 5.0101,
      "mean_token_accuracy": 0.21050338745117186,
      "num_tokens": 63929601.0,
      "step": 27915
    },
    {
      "entropy": 5.456412220001221,
      "epoch": 2.759984183471728,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000424437342607886,
      "loss": 5.0588,
      "mean_token_accuracy": 0.21508972048759462,
      "num_tokens": 63941170.0,
      "step": 27920
    },
    {
      "entropy": 5.392394924163819,
      "epoch": 2.760478449980229,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004244109585028593,
      "loss": 5.0093,
      "mean_token_accuracy": 0.2207762598991394,
      "num_tokens": 63951907.0,
      "step": 27925
    },
    {
      "entropy": 5.433092880249023,
      "epoch": 2.7609727164887308,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042438457072229986,
      "loss": 5.0568,
      "mean_token_accuracy": 0.2078884318470955,
      "num_tokens": 63963140.0,
      "step": 27930
    },
    {
      "entropy": 5.4470666408538815,
      "epoch": 2.761466982997232,
      "grad_norm": 1.0,
      "learning_rate": 0.00042435817926685684,
      "loss": 5.0383,
      "mean_token_accuracy": 0.20354804992675782,
      "num_tokens": 63974445.0,
      "step": 27935
    },
    {
      "entropy": 5.440000295639038,
      "epoch": 2.7619612495057337,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004243317841371793,
      "loss": 4.999,
      "mean_token_accuracy": 0.21949042826890947,
      "num_tokens": 63985463.0,
      "step": 27940
    },
    {
      "entropy": 5.453473520278931,
      "epoch": 2.762455516014235,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042430538533391683,
      "loss": 5.0757,
      "mean_token_accuracy": 0.20680534690618516,
      "num_tokens": 63997120.0,
      "step": 27945
    },
    {
      "entropy": 5.424569368362427,
      "epoch": 2.7629497825227363,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042427898285771865,
      "loss": 5.0475,
      "mean_token_accuracy": 0.2080203875899315,
      "num_tokens": 64009195.0,
      "step": 27950
    },
    {
      "entropy": 5.492969036102295,
      "epoch": 2.7634440490312375,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004242525767092344,
      "loss": 5.0701,
      "mean_token_accuracy": 0.21110433489084243,
      "num_tokens": 64020915.0,
      "step": 27955
    },
    {
      "entropy": 5.420156574249267,
      "epoch": 2.763938315539739,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004242261668891135,
      "loss": 5.0045,
      "mean_token_accuracy": 0.21134259253740312,
      "num_tokens": 64031187.0,
      "step": 27960
    },
    {
      "entropy": 5.415789651870727,
      "epoch": 2.7644325820482405,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004241997533980059,
      "loss": 5.0574,
      "mean_token_accuracy": 0.2055559501051903,
      "num_tokens": 64043045.0,
      "step": 27965
    },
    {
      "entropy": 5.578073787689209,
      "epoch": 2.7649268485567418,
      "grad_norm": 1.125,
      "learning_rate": 0.00042417333623656113,
      "loss": 5.1845,
      "mean_token_accuracy": 0.20065457373857498,
      "num_tokens": 64054030.0,
      "step": 27970
    },
    {
      "entropy": 5.520448637008667,
      "epoch": 2.765421115065243,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042414691540542917,
      "loss": 5.0666,
      "mean_token_accuracy": 0.2161511093378067,
      "num_tokens": 64064179.0,
      "step": 27975
    },
    {
      "entropy": 5.50725417137146,
      "epoch": 2.7659153815737447,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042412049090525996,
      "loss": 5.1302,
      "mean_token_accuracy": 0.19654364734888077,
      "num_tokens": 64075812.0,
      "step": 27980
    },
    {
      "entropy": 5.406628799438477,
      "epoch": 2.766409648082246,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004240940627367035,
      "loss": 5.0437,
      "mean_token_accuracy": 0.21193777769804,
      "num_tokens": 64086848.0,
      "step": 27985
    },
    {
      "entropy": 5.450288724899292,
      "epoch": 2.7669039145907472,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042406763090040996,
      "loss": 5.0645,
      "mean_token_accuracy": 0.20912527590990065,
      "num_tokens": 64098879.0,
      "step": 27990
    },
    {
      "entropy": 5.4657680034637455,
      "epoch": 2.7673981810992485,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042404119539702956,
      "loss": 5.0815,
      "mean_token_accuracy": 0.20548973828554154,
      "num_tokens": 64109946.0,
      "step": 27995
    },
    {
      "entropy": 5.519073534011841,
      "epoch": 2.76789244760775,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042401475622721264,
      "loss": 5.1419,
      "mean_token_accuracy": 0.20594579875469207,
      "num_tokens": 64122219.0,
      "step": 28000
    },
    {
      "entropy": 5.466803693771363,
      "epoch": 2.7683867141162515,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004239883133916097,
      "loss": 5.0273,
      "mean_token_accuracy": 0.21092600524425506,
      "num_tokens": 64132950.0,
      "step": 28005
    },
    {
      "entropy": 5.470190191268921,
      "epoch": 2.7688809806247527,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004239618668908711,
      "loss": 5.1212,
      "mean_token_accuracy": 0.20221826583147048,
      "num_tokens": 64145060.0,
      "step": 28010
    },
    {
      "entropy": 5.5050252914428714,
      "epoch": 2.7693752471332544,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004239354167256474,
      "loss": 5.1326,
      "mean_token_accuracy": 0.20182606130838393,
      "num_tokens": 64157294.0,
      "step": 28015
    },
    {
      "entropy": 5.492866945266724,
      "epoch": 2.7698695136417557,
      "grad_norm": 1.0,
      "learning_rate": 0.00042390896289658936,
      "loss": 5.0841,
      "mean_token_accuracy": 0.20653266608715057,
      "num_tokens": 64169001.0,
      "step": 28020
    },
    {
      "entropy": 5.3837135314941404,
      "epoch": 2.770363780150257,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004238825054043477,
      "loss": 5.0074,
      "mean_token_accuracy": 0.21395400762557984,
      "num_tokens": 64180525.0,
      "step": 28025
    },
    {
      "entropy": 5.419885635375977,
      "epoch": 2.7708580466587582,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042385604424957337,
      "loss": 4.9943,
      "mean_token_accuracy": 0.21156293600797654,
      "num_tokens": 64192398.0,
      "step": 28030
    },
    {
      "entropy": 5.436616373062134,
      "epoch": 2.77135231316726,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004238295794329172,
      "loss": 5.0749,
      "mean_token_accuracy": 0.20443775355815888,
      "num_tokens": 64203485.0,
      "step": 28035
    },
    {
      "entropy": 5.485711193084716,
      "epoch": 2.771846579675761,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042380311095503037,
      "loss": 5.0232,
      "mean_token_accuracy": 0.21363509446382523,
      "num_tokens": 64214226.0,
      "step": 28040
    },
    {
      "entropy": 5.429607725143432,
      "epoch": 2.7723408461842625,
      "grad_norm": 1.0,
      "learning_rate": 0.00042377663881656394,
      "loss": 4.9907,
      "mean_token_accuracy": 0.21144504398107528,
      "num_tokens": 64225873.0,
      "step": 28045
    },
    {
      "entropy": 5.569371891021729,
      "epoch": 2.772835112692764,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042375016301816903,
      "loss": 5.2284,
      "mean_token_accuracy": 0.19292194098234178,
      "num_tokens": 64238023.0,
      "step": 28050
    },
    {
      "entropy": 5.4937074184417725,
      "epoch": 2.7733293792012654,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004237236835604971,
      "loss": 5.0886,
      "mean_token_accuracy": 0.19927995949983596,
      "num_tokens": 64248533.0,
      "step": 28055
    },
    {
      "entropy": 5.499326610565186,
      "epoch": 2.7738236457097667,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004236972004441994,
      "loss": 5.1322,
      "mean_token_accuracy": 0.20386985093355178,
      "num_tokens": 64260539.0,
      "step": 28060
    },
    {
      "entropy": 5.440857410430908,
      "epoch": 2.774317912218268,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042367071366992753,
      "loss": 5.0397,
      "mean_token_accuracy": 0.20603834241628646,
      "num_tokens": 64271280.0,
      "step": 28065
    },
    {
      "entropy": 5.453833103179932,
      "epoch": 2.774812178726769,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042364422323833313,
      "loss": 5.0905,
      "mean_token_accuracy": 0.20637732446193696,
      "num_tokens": 64282728.0,
      "step": 28070
    },
    {
      "entropy": 5.470165538787842,
      "epoch": 2.775306445235271,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042361772915006763,
      "loss": 5.0852,
      "mean_token_accuracy": 0.20544882863759995,
      "num_tokens": 64294728.0,
      "step": 28075
    },
    {
      "entropy": 5.452522945404053,
      "epoch": 2.775800711743772,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004235912314057831,
      "loss": 5.1287,
      "mean_token_accuracy": 0.20635941177606582,
      "num_tokens": 64305873.0,
      "step": 28080
    },
    {
      "entropy": 5.401246690750122,
      "epoch": 2.776294978252274,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004235647300061311,
      "loss": 4.9733,
      "mean_token_accuracy": 0.21105220168828964,
      "num_tokens": 64318892.0,
      "step": 28085
    },
    {
      "entropy": 5.392233037948609,
      "epoch": 2.776789244760775,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004235382249517637,
      "loss": 4.9692,
      "mean_token_accuracy": 0.21376692056655883,
      "num_tokens": 64328918.0,
      "step": 28090
    },
    {
      "entropy": 5.383206176757812,
      "epoch": 2.7772835112692764,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042351171624333293,
      "loss": 5.0177,
      "mean_token_accuracy": 0.21142184734344482,
      "num_tokens": 64338670.0,
      "step": 28095
    },
    {
      "entropy": 5.45309591293335,
      "epoch": 2.7777777777777777,
      "grad_norm": 0.9375,
      "learning_rate": 0.00042348520388149093,
      "loss": 5.0173,
      "mean_token_accuracy": 0.20620488822460176,
      "num_tokens": 64350503.0,
      "step": 28100
    },
    {
      "entropy": 5.422789144515991,
      "epoch": 2.778272044286279,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004234586878668899,
      "loss": 4.9506,
      "mean_token_accuracy": 0.22056611776351928,
      "num_tokens": 64361176.0,
      "step": 28105
    },
    {
      "entropy": 5.434823989868164,
      "epoch": 2.7787663107947806,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004234321682001821,
      "loss": 5.0892,
      "mean_token_accuracy": 0.20408917665481568,
      "num_tokens": 64373397.0,
      "step": 28110
    },
    {
      "entropy": 5.379253149032593,
      "epoch": 2.779260577303282,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042340564488201987,
      "loss": 5.0072,
      "mean_token_accuracy": 0.20698851495981216,
      "num_tokens": 64384910.0,
      "step": 28115
    },
    {
      "entropy": 5.369167327880859,
      "epoch": 2.779754843811783,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004233791179130559,
      "loss": 5.0321,
      "mean_token_accuracy": 0.20789069086313247,
      "num_tokens": 64396787.0,
      "step": 28120
    },
    {
      "entropy": 5.438919878005981,
      "epoch": 2.780249110320285,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004233525872939424,
      "loss": 5.0751,
      "mean_token_accuracy": 0.20410680025815964,
      "num_tokens": 64407911.0,
      "step": 28125
    },
    {
      "entropy": 5.572028493881225,
      "epoch": 2.780743376828786,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042332605302533246,
      "loss": 5.1452,
      "mean_token_accuracy": 0.20519044548273085,
      "num_tokens": 64420697.0,
      "step": 28130
    },
    {
      "entropy": 5.446814012527466,
      "epoch": 2.7812376433372874,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00042329951510787847,
      "loss": 4.9825,
      "mean_token_accuracy": 0.21768329441547393,
      "num_tokens": 64431675.0,
      "step": 28135
    },
    {
      "entropy": 5.445724821090698,
      "epoch": 2.7817319098457887,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00042327297354223346,
      "loss": 5.0513,
      "mean_token_accuracy": 0.21032214760780335,
      "num_tokens": 64443398.0,
      "step": 28140
    },
    {
      "entropy": 5.4173455238342285,
      "epoch": 2.7822261763542904,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004232464283290502,
      "loss": 5.0681,
      "mean_token_accuracy": 0.2088153064250946,
      "num_tokens": 64455306.0,
      "step": 28145
    },
    {
      "entropy": 5.407170963287354,
      "epoch": 2.7827204428627916,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004232198794689819,
      "loss": 5.0175,
      "mean_token_accuracy": 0.22061744928359986,
      "num_tokens": 64466475.0,
      "step": 28150
    },
    {
      "entropy": 5.40111517906189,
      "epoch": 2.783214709371293,
      "grad_norm": 1.125,
      "learning_rate": 0.0004231933269626815,
      "loss": 5.0219,
      "mean_token_accuracy": 0.21246253848075866,
      "num_tokens": 64476729.0,
      "step": 28155
    },
    {
      "entropy": 5.528533267974853,
      "epoch": 2.7837089758797946,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004231667708108024,
      "loss": 5.1127,
      "mean_token_accuracy": 0.20727159976959228,
      "num_tokens": 64489403.0,
      "step": 28160
    },
    {
      "entropy": 5.445553827285766,
      "epoch": 2.784203242388296,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042314021101399763,
      "loss": 5.0273,
      "mean_token_accuracy": 0.22125994861125947,
      "num_tokens": 64501000.0,
      "step": 28165
    },
    {
      "entropy": 5.408338499069214,
      "epoch": 2.784697508896797,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004231136475729207,
      "loss": 5.0032,
      "mean_token_accuracy": 0.21040546000003815,
      "num_tokens": 64513504.0,
      "step": 28170
    },
    {
      "entropy": 5.378835201263428,
      "epoch": 2.7851917754052984,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042308708048822513,
      "loss": 4.9783,
      "mean_token_accuracy": 0.21552161872386932,
      "num_tokens": 64525522.0,
      "step": 28175
    },
    {
      "entropy": 5.48802227973938,
      "epoch": 2.7856860419137996,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004230605097605644,
      "loss": 5.096,
      "mean_token_accuracy": 0.20565718710422515,
      "num_tokens": 64537261.0,
      "step": 28180
    },
    {
      "entropy": 5.456069278717041,
      "epoch": 2.7861803084223014,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042303393539059204,
      "loss": 4.9789,
      "mean_token_accuracy": 0.21787464916706084,
      "num_tokens": 64548527.0,
      "step": 28185
    },
    {
      "entropy": 5.332040166854858,
      "epoch": 2.7866745749308026,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000423007357378962,
      "loss": 4.9562,
      "mean_token_accuracy": 0.22125230282545089,
      "num_tokens": 64560196.0,
      "step": 28190
    },
    {
      "entropy": 5.49363431930542,
      "epoch": 2.7871688414393043,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000422980775726328,
      "loss": 5.105,
      "mean_token_accuracy": 0.21046043336391448,
      "num_tokens": 64570828.0,
      "step": 28195
    },
    {
      "entropy": 5.516827487945557,
      "epoch": 2.7876631079478056,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004229541904333439,
      "loss": 5.1219,
      "mean_token_accuracy": 0.2023715227842331,
      "num_tokens": 64581845.0,
      "step": 28200
    },
    {
      "entropy": 5.458334732055664,
      "epoch": 2.788157374456307,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042292760150066386,
      "loss": 5.0773,
      "mean_token_accuracy": 0.20569535344839096,
      "num_tokens": 64593105.0,
      "step": 28205
    },
    {
      "entropy": 5.440145874023438,
      "epoch": 2.788651640964808,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042290100892894184,
      "loss": 5.0509,
      "mean_token_accuracy": 0.2111356034874916,
      "num_tokens": 64605075.0,
      "step": 28210
    },
    {
      "entropy": 5.420182895660401,
      "epoch": 2.7891459074733094,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042287441271883204,
      "loss": 5.0719,
      "mean_token_accuracy": 0.2085297644138336,
      "num_tokens": 64617091.0,
      "step": 28215
    },
    {
      "entropy": 5.461166000366211,
      "epoch": 2.789640173981811,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004228478128709888,
      "loss": 4.978,
      "mean_token_accuracy": 0.2131640285253525,
      "num_tokens": 64627870.0,
      "step": 28220
    },
    {
      "entropy": 5.379143905639649,
      "epoch": 2.7901344404903123,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004228212093860664,
      "loss": 5.0623,
      "mean_token_accuracy": 0.20517639964818954,
      "num_tokens": 64639583.0,
      "step": 28225
    },
    {
      "entropy": 5.465548849105835,
      "epoch": 2.7906287069988136,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004227946022647192,
      "loss": 5.0039,
      "mean_token_accuracy": 0.21015277355909348,
      "num_tokens": 64650477.0,
      "step": 28230
    },
    {
      "entropy": 5.475518894195557,
      "epoch": 2.7911229735073153,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000422767991507602,
      "loss": 5.0474,
      "mean_token_accuracy": 0.21396049857139587,
      "num_tokens": 64660994.0,
      "step": 28235
    },
    {
      "entropy": 5.417670440673828,
      "epoch": 2.7916172400158166,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004227413771153692,
      "loss": 5.0727,
      "mean_token_accuracy": 0.21750295907258987,
      "num_tokens": 64672348.0,
      "step": 28240
    },
    {
      "entropy": 5.459055376052857,
      "epoch": 2.792111506524318,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042271475908867563,
      "loss": 5.0456,
      "mean_token_accuracy": 0.20169767886400222,
      "num_tokens": 64683352.0,
      "step": 28245
    },
    {
      "entropy": 5.481772375106812,
      "epoch": 2.792605773032819,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004226881374281761,
      "loss": 5.0918,
      "mean_token_accuracy": 0.20454653054475785,
      "num_tokens": 64695429.0,
      "step": 28250
    },
    {
      "entropy": 5.3621031761169435,
      "epoch": 2.793100039541321,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004226615121345254,
      "loss": 4.9596,
      "mean_token_accuracy": 0.21337895095348358,
      "num_tokens": 64706438.0,
      "step": 28255
    },
    {
      "entropy": 5.451272964477539,
      "epoch": 2.793594306049822,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004226348832083787,
      "loss": 5.0313,
      "mean_token_accuracy": 0.20661709308624268,
      "num_tokens": 64717991.0,
      "step": 28260
    },
    {
      "entropy": 5.450286197662353,
      "epoch": 2.7940885725583233,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042260825065039095,
      "loss": 5.1039,
      "mean_token_accuracy": 0.20481336563825608,
      "num_tokens": 64730196.0,
      "step": 28265
    },
    {
      "entropy": 5.411657094955444,
      "epoch": 2.794582839066825,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004225816144612173,
      "loss": 5.0777,
      "mean_token_accuracy": 0.20436010360717774,
      "num_tokens": 64742977.0,
      "step": 28270
    },
    {
      "entropy": 5.484560871124268,
      "epoch": 2.7950771055753263,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042255497464151307,
      "loss": 5.0664,
      "mean_token_accuracy": 0.20762816220521926,
      "num_tokens": 64754101.0,
      "step": 28275
    },
    {
      "entropy": 5.415189981460571,
      "epoch": 2.7955713720838276,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042252833119193345,
      "loss": 5.0577,
      "mean_token_accuracy": 0.20904034227132798,
      "num_tokens": 64766611.0,
      "step": 28280
    },
    {
      "entropy": 5.416021013259888,
      "epoch": 2.796065638592329,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042250168411313413,
      "loss": 5.065,
      "mean_token_accuracy": 0.21142401546239853,
      "num_tokens": 64778764.0,
      "step": 28285
    },
    {
      "entropy": 5.434243440628052,
      "epoch": 2.7965599051008305,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042247503340577046,
      "loss": 5.0162,
      "mean_token_accuracy": 0.21078031957149507,
      "num_tokens": 64790208.0,
      "step": 28290
    },
    {
      "entropy": 5.428766965866089,
      "epoch": 2.797054171609332,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004224483790704981,
      "loss": 5.0932,
      "mean_token_accuracy": 0.20621077865362167,
      "num_tokens": 64801692.0,
      "step": 28295
    },
    {
      "entropy": 5.44076361656189,
      "epoch": 2.797548438117833,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042242172110797273,
      "loss": 4.9692,
      "mean_token_accuracy": 0.21063510030508042,
      "num_tokens": 64811919.0,
      "step": 28300
    },
    {
      "entropy": 5.423096132278443,
      "epoch": 2.7980427046263348,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042239505951885013,
      "loss": 4.9955,
      "mean_token_accuracy": 0.2161523461341858,
      "num_tokens": 64821679.0,
      "step": 28305
    },
    {
      "entropy": 5.407114171981812,
      "epoch": 2.798536971134836,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042236839430378624,
      "loss": 5.0423,
      "mean_token_accuracy": 0.21406506597995759,
      "num_tokens": 64833295.0,
      "step": 28310
    },
    {
      "entropy": 5.381708288192749,
      "epoch": 2.7990312376433373,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042234172546343695,
      "loss": 4.9453,
      "mean_token_accuracy": 0.2197033077478409,
      "num_tokens": 64844970.0,
      "step": 28315
    },
    {
      "entropy": 5.434743928909302,
      "epoch": 2.7995255041518385,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042231505299845836,
      "loss": 4.9401,
      "mean_token_accuracy": 0.21513905823230745,
      "num_tokens": 64855496.0,
      "step": 28320
    },
    {
      "entropy": 5.39394793510437,
      "epoch": 2.80001977066034,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042228837690950664,
      "loss": 5.0022,
      "mean_token_accuracy": 0.2134407252073288,
      "num_tokens": 64866836.0,
      "step": 28325
    },
    {
      "entropy": 5.359603548049927,
      "epoch": 2.8005140371688415,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042226169719723793,
      "loss": 5.0357,
      "mean_token_accuracy": 0.2154480680823326,
      "num_tokens": 64879105.0,
      "step": 28330
    },
    {
      "entropy": 5.5322788715362545,
      "epoch": 2.8010083036773428,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004222350138623087,
      "loss": 5.0502,
      "mean_token_accuracy": 0.21348078697919845,
      "num_tokens": 64891227.0,
      "step": 28335
    },
    {
      "entropy": 5.418788433074951,
      "epoch": 2.8015025701858445,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042220832690537525,
      "loss": 5.0299,
      "mean_token_accuracy": 0.20542630702257156,
      "num_tokens": 64904070.0,
      "step": 28340
    },
    {
      "entropy": 5.397738885879517,
      "epoch": 2.8019968366943457,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004221816363270941,
      "loss": 4.9814,
      "mean_token_accuracy": 0.2145321935415268,
      "num_tokens": 64915087.0,
      "step": 28345
    },
    {
      "entropy": 5.371378374099732,
      "epoch": 2.802491103202847,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042215494212812193,
      "loss": 4.9284,
      "mean_token_accuracy": 0.22041096836328505,
      "num_tokens": 64926163.0,
      "step": 28350
    },
    {
      "entropy": 5.366522884368896,
      "epoch": 2.8029853697113483,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042212824430911533,
      "loss": 4.9901,
      "mean_token_accuracy": 0.21784412115812302,
      "num_tokens": 64936513.0,
      "step": 28355
    },
    {
      "entropy": 5.318961334228516,
      "epoch": 2.8034796362198495,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004221015428707311,
      "loss": 4.9056,
      "mean_token_accuracy": 0.22387145906686784,
      "num_tokens": 64947232.0,
      "step": 28360
    },
    {
      "entropy": 5.504414463043213,
      "epoch": 2.8039739027283512,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000422074837813626,
      "loss": 5.0786,
      "mean_token_accuracy": 0.20795378237962722,
      "num_tokens": 64958622.0,
      "step": 28365
    },
    {
      "entropy": 5.477715539932251,
      "epoch": 2.8044681692368525,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042204812913845726,
      "loss": 5.0509,
      "mean_token_accuracy": 0.21152992099523543,
      "num_tokens": 64969975.0,
      "step": 28370
    },
    {
      "entropy": 5.407004928588867,
      "epoch": 2.8049624357453538,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042202141684588156,
      "loss": 4.9566,
      "mean_token_accuracy": 0.2120935320854187,
      "num_tokens": 64981195.0,
      "step": 28375
    },
    {
      "entropy": 5.492878818511963,
      "epoch": 2.8054567022538555,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004219947009365563,
      "loss": 5.1183,
      "mean_token_accuracy": 0.20944650173187257,
      "num_tokens": 64992265.0,
      "step": 28380
    },
    {
      "entropy": 5.47233476638794,
      "epoch": 2.8059509687623567,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004219679814111386,
      "loss": 5.0757,
      "mean_token_accuracy": 0.2086569920182228,
      "num_tokens": 65003613.0,
      "step": 28385
    },
    {
      "entropy": 5.376001310348511,
      "epoch": 2.806445235270858,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004219412582702857,
      "loss": 5.0015,
      "mean_token_accuracy": 0.21682388633489608,
      "num_tokens": 65014768.0,
      "step": 28390
    },
    {
      "entropy": 5.423770427703857,
      "epoch": 2.8069395017793592,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042191453151465506,
      "loss": 5.0232,
      "mean_token_accuracy": 0.20888259261846542,
      "num_tokens": 65026425.0,
      "step": 28395
    },
    {
      "entropy": 5.472204303741455,
      "epoch": 2.807433768287861,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042188780114490413,
      "loss": 5.0379,
      "mean_token_accuracy": 0.20887178778648377,
      "num_tokens": 65037624.0,
      "step": 28400
    },
    {
      "entropy": 5.470787906646729,
      "epoch": 2.807928034796362,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004218610671616905,
      "loss": 5.1279,
      "mean_token_accuracy": 0.21145582497119902,
      "num_tokens": 65048659.0,
      "step": 28405
    },
    {
      "entropy": 5.431805229187011,
      "epoch": 2.8084223013048635,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042183432956567184,
      "loss": 4.9314,
      "mean_token_accuracy": 0.22886100560426711,
      "num_tokens": 65059084.0,
      "step": 28410
    },
    {
      "entropy": 5.430448961257935,
      "epoch": 2.808916567813365,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042180758835750595,
      "loss": 5.0147,
      "mean_token_accuracy": 0.21786174178123474,
      "num_tokens": 65071525.0,
      "step": 28415
    },
    {
      "entropy": 5.410004329681397,
      "epoch": 2.8094108343218664,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004217808435378506,
      "loss": 5.0765,
      "mean_token_accuracy": 0.21156519651412964,
      "num_tokens": 65082738.0,
      "step": 28420
    },
    {
      "entropy": 5.483507108688355,
      "epoch": 2.8099051008303677,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004217540951073637,
      "loss": 5.081,
      "mean_token_accuracy": 0.20905945748090743,
      "num_tokens": 65094988.0,
      "step": 28425
    },
    {
      "entropy": 5.430683851242065,
      "epoch": 2.810399367338869,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004217273430667033,
      "loss": 5.0533,
      "mean_token_accuracy": 0.20987478494644166,
      "num_tokens": 65106412.0,
      "step": 28430
    },
    {
      "entropy": 5.41854248046875,
      "epoch": 2.8108936338473702,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004217005874165274,
      "loss": 5.0448,
      "mean_token_accuracy": 0.2156478375196457,
      "num_tokens": 65117756.0,
      "step": 28435
    },
    {
      "entropy": 5.457470035552978,
      "epoch": 2.811387900355872,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004216738281574944,
      "loss": 5.0944,
      "mean_token_accuracy": 0.20926827043294907,
      "num_tokens": 65129368.0,
      "step": 28440
    },
    {
      "entropy": 5.4730747699737545,
      "epoch": 2.811882166864373,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042164706529026255,
      "loss": 5.0858,
      "mean_token_accuracy": 0.20692606419324874,
      "num_tokens": 65140553.0,
      "step": 28445
    },
    {
      "entropy": 5.398244857788086,
      "epoch": 2.812376433372875,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004216202988154901,
      "loss": 4.9965,
      "mean_token_accuracy": 0.21188055723905563,
      "num_tokens": 65152047.0,
      "step": 28450
    },
    {
      "entropy": 5.427066135406494,
      "epoch": 2.812870699881376,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004215935287338354,
      "loss": 5.0105,
      "mean_token_accuracy": 0.21325698047876357,
      "num_tokens": 65164001.0,
      "step": 28455
    },
    {
      "entropy": 5.419220924377441,
      "epoch": 2.8133649663898774,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004215667550459573,
      "loss": 5.0261,
      "mean_token_accuracy": 0.2111202135682106,
      "num_tokens": 65175949.0,
      "step": 28460
    },
    {
      "entropy": 5.469588136672973,
      "epoch": 2.8138592328983787,
      "grad_norm": 1.125,
      "learning_rate": 0.0004215399777525142,
      "loss": 5.0348,
      "mean_token_accuracy": 0.2180962383747101,
      "num_tokens": 65186725.0,
      "step": 28465
    },
    {
      "entropy": 5.456300973892212,
      "epoch": 2.81435349940688,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000421513196854165,
      "loss": 5.0273,
      "mean_token_accuracy": 0.21134413927793502,
      "num_tokens": 65198489.0,
      "step": 28470
    },
    {
      "entropy": 5.356208610534668,
      "epoch": 2.8148477659153817,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004214864123515684,
      "loss": 4.9651,
      "mean_token_accuracy": 0.21576543748378754,
      "num_tokens": 65210572.0,
      "step": 28475
    },
    {
      "entropy": 5.423062467575074,
      "epoch": 2.815342032423883,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042145962424538336,
      "loss": 5.0535,
      "mean_token_accuracy": 0.20608833581209182,
      "num_tokens": 65221785.0,
      "step": 28480
    },
    {
      "entropy": 5.502148151397705,
      "epoch": 2.815836298932384,
      "grad_norm": 1.0,
      "learning_rate": 0.0004214328325362687,
      "loss": 5.085,
      "mean_token_accuracy": 0.2033339187502861,
      "num_tokens": 65233998.0,
      "step": 28485
    },
    {
      "entropy": 5.477193450927734,
      "epoch": 2.816330565440886,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004214060372248838,
      "loss": 5.0574,
      "mean_token_accuracy": 0.2020837187767029,
      "num_tokens": 65243907.0,
      "step": 28490
    },
    {
      "entropy": 5.422337436676026,
      "epoch": 2.816824831949387,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042137923831188763,
      "loss": 5.0218,
      "mean_token_accuracy": 0.21004620641469957,
      "num_tokens": 65254555.0,
      "step": 28495
    },
    {
      "entropy": 5.4710005760192875,
      "epoch": 2.8173190984578884,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042135243579793946,
      "loss": 5.0387,
      "mean_token_accuracy": 0.20894381850957872,
      "num_tokens": 65265751.0,
      "step": 28500
    },
    {
      "entropy": 5.389526271820069,
      "epoch": 2.8178133649663897,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042132562968369864,
      "loss": 4.9751,
      "mean_token_accuracy": 0.21284334361553192,
      "num_tokens": 65275839.0,
      "step": 28505
    },
    {
      "entropy": 5.426601552963257,
      "epoch": 2.8183076314748914,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004212988199698247,
      "loss": 5.0075,
      "mean_token_accuracy": 0.21365518122911453,
      "num_tokens": 65287796.0,
      "step": 28510
    },
    {
      "entropy": 5.48567566871643,
      "epoch": 2.8188018979833926,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00042127200665697715,
      "loss": 5.0818,
      "mean_token_accuracy": 0.21423664540052414,
      "num_tokens": 65299014.0,
      "step": 28515
    },
    {
      "entropy": 5.412119722366333,
      "epoch": 2.819296164491894,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042124518974581536,
      "loss": 5.0079,
      "mean_token_accuracy": 0.2147382378578186,
      "num_tokens": 65309830.0,
      "step": 28520
    },
    {
      "entropy": 5.482333850860596,
      "epoch": 2.8197904310003956,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004212183692369994,
      "loss": 5.1074,
      "mean_token_accuracy": 0.20084495842456818,
      "num_tokens": 65321921.0,
      "step": 28525
    },
    {
      "entropy": 5.436908149719239,
      "epoch": 2.820284697508897,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004211915451311887,
      "loss": 5.0313,
      "mean_token_accuracy": 0.2132962614297867,
      "num_tokens": 65334175.0,
      "step": 28530
    },
    {
      "entropy": 5.462993240356445,
      "epoch": 2.820778964017398,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004211647174290434,
      "loss": 5.0934,
      "mean_token_accuracy": 0.2148503839969635,
      "num_tokens": 65345460.0,
      "step": 28535
    },
    {
      "entropy": 5.464183950424195,
      "epoch": 2.8212732305258994,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042113788613122333,
      "loss": 5.0084,
      "mean_token_accuracy": 0.2155225932598114,
      "num_tokens": 65357252.0,
      "step": 28540
    },
    {
      "entropy": 5.47107892036438,
      "epoch": 2.821767497034401,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004211110512383887,
      "loss": 5.1547,
      "mean_token_accuracy": 0.20428680628538132,
      "num_tokens": 65370195.0,
      "step": 28545
    },
    {
      "entropy": 5.368879413604736,
      "epoch": 2.8222617635429024,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004210842127511995,
      "loss": 5.0087,
      "mean_token_accuracy": 0.21406644433736802,
      "num_tokens": 65382531.0,
      "step": 28550
    },
    {
      "entropy": 5.516090154647827,
      "epoch": 2.8227560300514036,
      "grad_norm": 1.125,
      "learning_rate": 0.0004210573706703159,
      "loss": 5.1205,
      "mean_token_accuracy": 0.20873817205429077,
      "num_tokens": 65393829.0,
      "step": 28555
    },
    {
      "entropy": 5.442989110946655,
      "epoch": 2.8232502965599053,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004210305249963984,
      "loss": 5.0316,
      "mean_token_accuracy": 0.21122439652681352,
      "num_tokens": 65404479.0,
      "step": 28560
    },
    {
      "entropy": 5.487104177474976,
      "epoch": 2.8237445630684066,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004210036757301073,
      "loss": 5.1095,
      "mean_token_accuracy": 0.20648325830698014,
      "num_tokens": 65415468.0,
      "step": 28565
    },
    {
      "entropy": 5.3704681396484375,
      "epoch": 2.824238829576908,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004209768228721032,
      "loss": 4.9328,
      "mean_token_accuracy": 0.22169162034988404,
      "num_tokens": 65425994.0,
      "step": 28570
    },
    {
      "entropy": 5.407183504104614,
      "epoch": 2.824733096085409,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042094996642304657,
      "loss": 5.0608,
      "mean_token_accuracy": 0.20824696570634843,
      "num_tokens": 65438394.0,
      "step": 28575
    },
    {
      "entropy": 5.458730554580688,
      "epoch": 2.8252273625939104,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004209231063835981,
      "loss": 5.1024,
      "mean_token_accuracy": 0.20825848430395127,
      "num_tokens": 65450248.0,
      "step": 28580
    },
    {
      "entropy": 5.454391670227051,
      "epoch": 2.825721629102412,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042089624275441865,
      "loss": 5.0393,
      "mean_token_accuracy": 0.20841936022043228,
      "num_tokens": 65460803.0,
      "step": 28585
    },
    {
      "entropy": 5.4210961818695065,
      "epoch": 2.8262158956109134,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004208693755361689,
      "loss": 5.0462,
      "mean_token_accuracy": 0.2115780904889107,
      "num_tokens": 65471899.0,
      "step": 28590
    },
    {
      "entropy": 5.444215106964111,
      "epoch": 2.826710162119415,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004208425047295099,
      "loss": 5.0551,
      "mean_token_accuracy": 0.213706636428833,
      "num_tokens": 65483767.0,
      "step": 28595
    },
    {
      "entropy": 5.392724227905274,
      "epoch": 2.8272044286279163,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004208156303351027,
      "loss": 4.9165,
      "mean_token_accuracy": 0.21858488768339157,
      "num_tokens": 65494859.0,
      "step": 28600
    },
    {
      "entropy": 5.510323762893677,
      "epoch": 2.8276986951364176,
      "grad_norm": 1.0,
      "learning_rate": 0.00042078875235360835,
      "loss": 5.1335,
      "mean_token_accuracy": 0.20347557961940765,
      "num_tokens": 65506411.0,
      "step": 28605
    },
    {
      "entropy": 5.491088438034057,
      "epoch": 2.828192961644919,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042076187078568804,
      "loss": 4.9975,
      "mean_token_accuracy": 0.21039482355117797,
      "num_tokens": 65516655.0,
      "step": 28610
    },
    {
      "entropy": 5.286016368865967,
      "epoch": 2.82868722815342,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004207349856320031,
      "loss": 4.975,
      "mean_token_accuracy": 0.21509698927402496,
      "num_tokens": 65529305.0,
      "step": 28615
    },
    {
      "entropy": 5.462048482894898,
      "epoch": 2.829181494661922,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042070809689321497,
      "loss": 5.0466,
      "mean_token_accuracy": 0.2105020761489868,
      "num_tokens": 65542129.0,
      "step": 28620
    },
    {
      "entropy": 5.408416318893432,
      "epoch": 2.829675761170423,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000420681204569985,
      "loss": 4.9923,
      "mean_token_accuracy": 0.2233053922653198,
      "num_tokens": 65554240.0,
      "step": 28625
    },
    {
      "entropy": 5.3401312828063965,
      "epoch": 2.8301700276789243,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042065430866297486,
      "loss": 4.9445,
      "mean_token_accuracy": 0.2213255926966667,
      "num_tokens": 65564994.0,
      "step": 28630
    },
    {
      "entropy": 5.445550966262817,
      "epoch": 2.830664294187426,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004206274091728461,
      "loss": 5.0144,
      "mean_token_accuracy": 0.2138695552945137,
      "num_tokens": 65575986.0,
      "step": 28635
    },
    {
      "entropy": 5.520591068267822,
      "epoch": 2.8311585606959273,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004206005061002604,
      "loss": 5.127,
      "mean_token_accuracy": 0.21299002021551133,
      "num_tokens": 65587792.0,
      "step": 28640
    },
    {
      "entropy": 5.527397346496582,
      "epoch": 2.8316528272044286,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004205735994458798,
      "loss": 5.2198,
      "mean_token_accuracy": 0.19350966364145278,
      "num_tokens": 65599845.0,
      "step": 28645
    },
    {
      "entropy": 5.493664312362671,
      "epoch": 2.83214709371293,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004205466892103659,
      "loss": 5.0909,
      "mean_token_accuracy": 0.20469158589839936,
      "num_tokens": 65611796.0,
      "step": 28650
    },
    {
      "entropy": 5.589280033111573,
      "epoch": 2.8326413602214315,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042051977539438103,
      "loss": 5.138,
      "mean_token_accuracy": 0.19820379316806794,
      "num_tokens": 65622319.0,
      "step": 28655
    },
    {
      "entropy": 5.403728342056274,
      "epoch": 2.833135626729933,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042049285799858707,
      "loss": 5.0045,
      "mean_token_accuracy": 0.2140220209956169,
      "num_tokens": 65634403.0,
      "step": 28660
    },
    {
      "entropy": 5.399525594711304,
      "epoch": 2.833629893238434,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042046593702364614,
      "loss": 5.035,
      "mean_token_accuracy": 0.21251134276390077,
      "num_tokens": 65644719.0,
      "step": 28665
    },
    {
      "entropy": 5.597980213165283,
      "epoch": 2.8341241597469358,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004204390124702207,
      "loss": 5.1401,
      "mean_token_accuracy": 0.20101604014635086,
      "num_tokens": 65657083.0,
      "step": 28670
    },
    {
      "entropy": 5.413203954696655,
      "epoch": 2.834618426255437,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004204120843389729,
      "loss": 5.0593,
      "mean_token_accuracy": 0.21174242943525315,
      "num_tokens": 65667579.0,
      "step": 28675
    },
    {
      "entropy": 5.425848817825317,
      "epoch": 2.8351126927639383,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042038515263056537,
      "loss": 5.0218,
      "mean_token_accuracy": 0.21818395406007768,
      "num_tokens": 65679066.0,
      "step": 28680
    },
    {
      "entropy": 5.491365098953247,
      "epoch": 2.8356069592724396,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004203582173456604,
      "loss": 5.1196,
      "mean_token_accuracy": 0.2053641274571419,
      "num_tokens": 65690785.0,
      "step": 28685
    },
    {
      "entropy": 5.440450763702392,
      "epoch": 2.836101225780941,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042033127848492086,
      "loss": 5.0539,
      "mean_token_accuracy": 0.21911954432725905,
      "num_tokens": 65701613.0,
      "step": 28690
    },
    {
      "entropy": 5.458510875701904,
      "epoch": 2.8365954922894425,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004203043360490093,
      "loss": 5.043,
      "mean_token_accuracy": 0.21063696444034577,
      "num_tokens": 65713146.0,
      "step": 28695
    },
    {
      "entropy": 5.46273193359375,
      "epoch": 2.837089758797944,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042027739003858855,
      "loss": 5.0669,
      "mean_token_accuracy": 0.20314407497644424,
      "num_tokens": 65724216.0,
      "step": 28700
    },
    {
      "entropy": 5.37620325088501,
      "epoch": 2.8375840253064455,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004202504404543214,
      "loss": 4.9685,
      "mean_token_accuracy": 0.21216566860675812,
      "num_tokens": 65735357.0,
      "step": 28705
    },
    {
      "entropy": 5.441872882843017,
      "epoch": 2.8380782918149468,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004202234872968709,
      "loss": 5.0238,
      "mean_token_accuracy": 0.20495211333036423,
      "num_tokens": 65747679.0,
      "step": 28710
    },
    {
      "entropy": 5.467099714279175,
      "epoch": 2.838572558323448,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004201965305669001,
      "loss": 5.0027,
      "mean_token_accuracy": 0.21413648277521133,
      "num_tokens": 65759126.0,
      "step": 28715
    },
    {
      "entropy": 5.457605504989624,
      "epoch": 2.8390668248319493,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042016957026507206,
      "loss": 5.0821,
      "mean_token_accuracy": 0.2066372349858284,
      "num_tokens": 65771240.0,
      "step": 28720
    },
    {
      "entropy": 5.4145081520080565,
      "epoch": 2.8395610913404505,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004201426063920501,
      "loss": 4.9547,
      "mean_token_accuracy": 0.2189878061413765,
      "num_tokens": 65783323.0,
      "step": 28725
    },
    {
      "entropy": 5.371854448318482,
      "epoch": 2.8400553578489522,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042011563894849756,
      "loss": 4.9632,
      "mean_token_accuracy": 0.21307497322559357,
      "num_tokens": 65794165.0,
      "step": 28730
    },
    {
      "entropy": 5.373121643066407,
      "epoch": 2.8405496243574535,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004200886679350778,
      "loss": 4.9595,
      "mean_token_accuracy": 0.21789026111364365,
      "num_tokens": 65804541.0,
      "step": 28735
    },
    {
      "entropy": 5.413978815078735,
      "epoch": 2.8410438908659548,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004200616933524542,
      "loss": 5.0579,
      "mean_token_accuracy": 0.20826001912355424,
      "num_tokens": 65816986.0,
      "step": 28740
    },
    {
      "entropy": 5.406021976470948,
      "epoch": 2.8415381573744565,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004200347152012905,
      "loss": 5.0276,
      "mean_token_accuracy": 0.20835131108760835,
      "num_tokens": 65828711.0,
      "step": 28745
    },
    {
      "entropy": 5.496225643157959,
      "epoch": 2.8420324238829577,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004200077334822503,
      "loss": 5.0378,
      "mean_token_accuracy": 0.21068355441093445,
      "num_tokens": 65840525.0,
      "step": 28750
    },
    {
      "entropy": 5.393509340286255,
      "epoch": 2.842526690391459,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004199807481959973,
      "loss": 4.9461,
      "mean_token_accuracy": 0.22170021682977675,
      "num_tokens": 65851663.0,
      "step": 28755
    },
    {
      "entropy": 5.406895637512207,
      "epoch": 2.8430209568999603,
      "grad_norm": 1.125,
      "learning_rate": 0.0004199537593431955,
      "loss": 5.0142,
      "mean_token_accuracy": 0.21615724563598632,
      "num_tokens": 65863337.0,
      "step": 28760
    },
    {
      "entropy": 5.545959329605102,
      "epoch": 2.843515223408462,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004199267669245087,
      "loss": 5.1735,
      "mean_token_accuracy": 0.19357152432203292,
      "num_tokens": 65875581.0,
      "step": 28765
    },
    {
      "entropy": 5.4703444480896,
      "epoch": 2.8440094899169632,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041989977094060096,
      "loss": 5.0366,
      "mean_token_accuracy": 0.20807341933250428,
      "num_tokens": 65886577.0,
      "step": 28770
    },
    {
      "entropy": 5.360303735733032,
      "epoch": 2.8445037564254645,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004198727713921364,
      "loss": 5.0367,
      "mean_token_accuracy": 0.20737203508615493,
      "num_tokens": 65897914.0,
      "step": 28775
    },
    {
      "entropy": 5.410736417770385,
      "epoch": 2.844998022933966,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004198457682797791,
      "loss": 5.0354,
      "mean_token_accuracy": 0.216299407184124,
      "num_tokens": 65909781.0,
      "step": 28780
    },
    {
      "entropy": 5.449419927597046,
      "epoch": 2.8454922894424675,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041981876160419353,
      "loss": 5.0091,
      "mean_token_accuracy": 0.2140233263373375,
      "num_tokens": 65921107.0,
      "step": 28785
    },
    {
      "entropy": 5.475045871734619,
      "epoch": 2.8459865559509687,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041979175136604385,
      "loss": 5.0377,
      "mean_token_accuracy": 0.2162925958633423,
      "num_tokens": 65931166.0,
      "step": 28790
    },
    {
      "entropy": 5.509069871902466,
      "epoch": 2.84648082245947,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004197647375659947,
      "loss": 5.1517,
      "mean_token_accuracy": 0.20443524569272994,
      "num_tokens": 65943835.0,
      "step": 28795
    },
    {
      "entropy": 5.408158588409424,
      "epoch": 2.8469750889679717,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00041973772020471054,
      "loss": 5.0078,
      "mean_token_accuracy": 0.21722477674484253,
      "num_tokens": 65956961.0,
      "step": 28800
    },
    {
      "entropy": 5.488613128662109,
      "epoch": 2.847469355476473,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00041971069928285596,
      "loss": 5.0728,
      "mean_token_accuracy": 0.20766416788101197,
      "num_tokens": 65967774.0,
      "step": 28805
    },
    {
      "entropy": 5.395430374145508,
      "epoch": 2.847963621984974,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041968367480109577,
      "loss": 5.0164,
      "mean_token_accuracy": 0.21161174774169922,
      "num_tokens": 65978318.0,
      "step": 28810
    },
    {
      "entropy": 5.454450607299805,
      "epoch": 2.848457888493476,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004196566467600947,
      "loss": 4.9899,
      "mean_token_accuracy": 0.21338155418634414,
      "num_tokens": 65989763.0,
      "step": 28815
    },
    {
      "entropy": 5.374890422821045,
      "epoch": 2.848952155001977,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041962961516051777,
      "loss": 4.9676,
      "mean_token_accuracy": 0.21946410983800888,
      "num_tokens": 66001117.0,
      "step": 28820
    },
    {
      "entropy": 5.460238122940064,
      "epoch": 2.8494464215104784,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004196025800030298,
      "loss": 5.0642,
      "mean_token_accuracy": 0.20710013061761856,
      "num_tokens": 66012340.0,
      "step": 28825
    },
    {
      "entropy": 5.380323600769043,
      "epoch": 2.8499406880189797,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004195755412882959,
      "loss": 5.0285,
      "mean_token_accuracy": 0.2171403020620346,
      "num_tokens": 66023151.0,
      "step": 28830
    },
    {
      "entropy": 5.348619365692139,
      "epoch": 2.850434954527481,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041954849901698133,
      "loss": 4.8907,
      "mean_token_accuracy": 0.22517501413822175,
      "num_tokens": 66033399.0,
      "step": 28835
    },
    {
      "entropy": 5.373459339141846,
      "epoch": 2.8509292210359827,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041952145318975114,
      "loss": 5.0377,
      "mean_token_accuracy": 0.2155236452817917,
      "num_tokens": 66045817.0,
      "step": 28840
    },
    {
      "entropy": 5.535502767562866,
      "epoch": 2.851423487544484,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004194944038072709,
      "loss": 5.1554,
      "mean_token_accuracy": 0.20260321497917175,
      "num_tokens": 66058696.0,
      "step": 28845
    },
    {
      "entropy": 5.4878850936889645,
      "epoch": 2.8519177540529856,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041946735087020586,
      "loss": 4.9903,
      "mean_token_accuracy": 0.2148006796836853,
      "num_tokens": 66070197.0,
      "step": 28850
    },
    {
      "entropy": 5.358147144317627,
      "epoch": 2.852412020561487,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004194402943792216,
      "loss": 4.8468,
      "mean_token_accuracy": 0.2272876963019371,
      "num_tokens": 66080250.0,
      "step": 28855
    },
    {
      "entropy": 5.425247240066528,
      "epoch": 2.852906287069988,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004194132343349837,
      "loss": 5.0649,
      "mean_token_accuracy": 0.21172251999378205,
      "num_tokens": 66090957.0,
      "step": 28860
    },
    {
      "entropy": 5.378398466110229,
      "epoch": 2.8534005535784894,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004193861707381578,
      "loss": 4.9615,
      "mean_token_accuracy": 0.21455155164003373,
      "num_tokens": 66103042.0,
      "step": 28865
    },
    {
      "entropy": 5.437839984893799,
      "epoch": 2.8538948200869907,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041935910358940974,
      "loss": 5.0211,
      "mean_token_accuracy": 0.21466942131519318,
      "num_tokens": 66113832.0,
      "step": 28870
    },
    {
      "entropy": 5.407901763916016,
      "epoch": 2.8543890865954924,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004193320328894053,
      "loss": 4.946,
      "mean_token_accuracy": 0.22254039943218232,
      "num_tokens": 66126268.0,
      "step": 28875
    },
    {
      "entropy": 5.369946622848511,
      "epoch": 2.8548833531039937,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004193049586388105,
      "loss": 4.9414,
      "mean_token_accuracy": 0.20930675268173218,
      "num_tokens": 66137590.0,
      "step": 28880
    },
    {
      "entropy": 5.299560451507569,
      "epoch": 2.855377619612495,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041927788083829127,
      "loss": 4.9763,
      "mean_token_accuracy": 0.21806904971599578,
      "num_tokens": 66149110.0,
      "step": 28885
    },
    {
      "entropy": 5.416442966461181,
      "epoch": 2.8558718861209966,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004192507994885137,
      "loss": 4.9582,
      "mean_token_accuracy": 0.21779021322727204,
      "num_tokens": 66160116.0,
      "step": 28890
    },
    {
      "entropy": 5.518072319030762,
      "epoch": 2.856366152629498,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004192237145901443,
      "loss": 5.1029,
      "mean_token_accuracy": 0.20488617122173308,
      "num_tokens": 66170743.0,
      "step": 28895
    },
    {
      "entropy": 5.509412479400635,
      "epoch": 2.856860419137999,
      "grad_norm": 1.015625,
      "learning_rate": 0.000419196626143849,
      "loss": 5.0308,
      "mean_token_accuracy": 0.2063233435153961,
      "num_tokens": 66182280.0,
      "step": 28900
    },
    {
      "entropy": 5.44746675491333,
      "epoch": 2.8573546856465004,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041916953415029447,
      "loss": 5.0845,
      "mean_token_accuracy": 0.20638009905815125,
      "num_tokens": 66193387.0,
      "step": 28905
    },
    {
      "entropy": 5.420180797576904,
      "epoch": 2.857848952155002,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004191424386101469,
      "loss": 5.0304,
      "mean_token_accuracy": 0.20419544875621795,
      "num_tokens": 66206180.0,
      "step": 28910
    },
    {
      "entropy": 5.362782382965088,
      "epoch": 2.8583432186635034,
      "grad_norm": 1.09375,
      "learning_rate": 0.000419115339524073,
      "loss": 4.96,
      "mean_token_accuracy": 0.2170044645667076,
      "num_tokens": 66217495.0,
      "step": 28915
    },
    {
      "entropy": 5.455839490890503,
      "epoch": 2.8588374851720046,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004190882368927394,
      "loss": 5.051,
      "mean_token_accuracy": 0.2046976640820503,
      "num_tokens": 66229107.0,
      "step": 28920
    },
    {
      "entropy": 5.423680019378662,
      "epoch": 2.8593317516805064,
      "grad_norm": 1.0,
      "learning_rate": 0.0004190611307168128,
      "loss": 4.9937,
      "mean_token_accuracy": 0.2228888288140297,
      "num_tokens": 66240480.0,
      "step": 28925
    },
    {
      "entropy": 5.322403860092163,
      "epoch": 2.8598260181890076,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041903402099696007,
      "loss": 4.9747,
      "mean_token_accuracy": 0.22297030687332153,
      "num_tokens": 66251785.0,
      "step": 28930
    },
    {
      "entropy": 5.411808586120605,
      "epoch": 2.860320284697509,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004190069077338481,
      "loss": 4.9429,
      "mean_token_accuracy": 0.2188699409365654,
      "num_tokens": 66262557.0,
      "step": 28935
    },
    {
      "entropy": 5.47333402633667,
      "epoch": 2.86081455120601,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004189797909281438,
      "loss": 5.0855,
      "mean_token_accuracy": 0.20751052498817443,
      "num_tokens": 66272931.0,
      "step": 28940
    },
    {
      "entropy": 5.441748142242432,
      "epoch": 2.8613088177145114,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041895267058051434,
      "loss": 5.0151,
      "mean_token_accuracy": 0.21667002588510514,
      "num_tokens": 66283550.0,
      "step": 28945
    },
    {
      "entropy": 5.3790491104125975,
      "epoch": 2.861803084223013,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041892554669162687,
      "loss": 5.0467,
      "mean_token_accuracy": 0.21014563292264937,
      "num_tokens": 66294740.0,
      "step": 28950
    },
    {
      "entropy": 5.3559685230255125,
      "epoch": 2.8622973507315144,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041889841926214856,
      "loss": 4.9323,
      "mean_token_accuracy": 0.21505227088928222,
      "num_tokens": 66306241.0,
      "step": 28955
    },
    {
      "entropy": 5.488435745239258,
      "epoch": 2.862791617240016,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00041887128829274685,
      "loss": 5.0506,
      "mean_token_accuracy": 0.20568871647119522,
      "num_tokens": 66320006.0,
      "step": 28960
    },
    {
      "entropy": 5.457212352752686,
      "epoch": 2.8632858837485173,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041884415378408917,
      "loss": 5.0068,
      "mean_token_accuracy": 0.2104433223605156,
      "num_tokens": 66330810.0,
      "step": 28965
    },
    {
      "entropy": 5.451360559463501,
      "epoch": 2.8637801502570186,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004188170157368429,
      "loss": 5.0485,
      "mean_token_accuracy": 0.21119442880153655,
      "num_tokens": 66341858.0,
      "step": 28970
    },
    {
      "entropy": 5.399813270568847,
      "epoch": 2.86427441676552,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004187898741516758,
      "loss": 5.0278,
      "mean_token_accuracy": 0.21459172666072845,
      "num_tokens": 66353175.0,
      "step": 28975
    },
    {
      "entropy": 5.452212810516357,
      "epoch": 2.864768683274021,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004187627290292555,
      "loss": 5.0227,
      "mean_token_accuracy": 0.21673197001218797,
      "num_tokens": 66364245.0,
      "step": 28980
    },
    {
      "entropy": 5.3820422172546385,
      "epoch": 2.865262949782523,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041873558037024974,
      "loss": 5.05,
      "mean_token_accuracy": 0.20762303918600084,
      "num_tokens": 66376793.0,
      "step": 28985
    },
    {
      "entropy": 5.472010374069214,
      "epoch": 2.865757216291024,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004187084281753264,
      "loss": 5.0659,
      "mean_token_accuracy": 0.20549626499414445,
      "num_tokens": 66388130.0,
      "step": 28990
    },
    {
      "entropy": 5.5225176334381105,
      "epoch": 2.8662514827995254,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041868127244515336,
      "loss": 5.0763,
      "mean_token_accuracy": 0.20701929926872253,
      "num_tokens": 66400784.0,
      "step": 28995
    },
    {
      "entropy": 5.435283279418945,
      "epoch": 2.866745749308027,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041865411318039883,
      "loss": 5.0256,
      "mean_token_accuracy": 0.21093764007091523,
      "num_tokens": 66412159.0,
      "step": 29000
    },
    {
      "entropy": 5.442742919921875,
      "epoch": 2.8672400158165283,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041862695038173085,
      "loss": 4.9939,
      "mean_token_accuracy": 0.21506983786821365,
      "num_tokens": 66423471.0,
      "step": 29005
    },
    {
      "entropy": 5.534247159957886,
      "epoch": 2.8677342823250296,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004185997840498175,
      "loss": 5.1456,
      "mean_token_accuracy": 0.19756857454776763,
      "num_tokens": 66436889.0,
      "step": 29010
    },
    {
      "entropy": 5.494853401184082,
      "epoch": 2.868228548833531,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004185726141853272,
      "loss": 5.0857,
      "mean_token_accuracy": 0.2067474365234375,
      "num_tokens": 66448157.0,
      "step": 29015
    },
    {
      "entropy": 5.319740676879883,
      "epoch": 2.8687228153420326,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004185454407889284,
      "loss": 4.9137,
      "mean_token_accuracy": 0.22362101525068284,
      "num_tokens": 66459877.0,
      "step": 29020
    },
    {
      "entropy": 5.463305711746216,
      "epoch": 2.869217081850534,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004185182638612895,
      "loss": 4.974,
      "mean_token_accuracy": 0.21965937465429305,
      "num_tokens": 66469648.0,
      "step": 29025
    },
    {
      "entropy": 5.448493480682373,
      "epoch": 2.869711348359035,
      "grad_norm": 0.9375,
      "learning_rate": 0.00041849108340307893,
      "loss": 5.1009,
      "mean_token_accuracy": 0.20633881241083146,
      "num_tokens": 66480878.0,
      "step": 29030
    },
    {
      "entropy": 5.52951488494873,
      "epoch": 2.870205614867537,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041846389941496557,
      "loss": 5.1282,
      "mean_token_accuracy": 0.20393907725811006,
      "num_tokens": 66492099.0,
      "step": 29035
    },
    {
      "entropy": 5.524483633041382,
      "epoch": 2.870699881376038,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041843671189761795,
      "loss": 5.0704,
      "mean_token_accuracy": 0.20663715451955794,
      "num_tokens": 66503194.0,
      "step": 29040
    },
    {
      "entropy": 5.466691875457764,
      "epoch": 2.8711941478845393,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000418409520851705,
      "loss": 5.0651,
      "mean_token_accuracy": 0.20576851814985275,
      "num_tokens": 66513393.0,
      "step": 29045
    },
    {
      "entropy": 5.415179014205933,
      "epoch": 2.8716884143930406,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004183823262778955,
      "loss": 5.0136,
      "mean_token_accuracy": 0.2205430820584297,
      "num_tokens": 66524042.0,
      "step": 29050
    },
    {
      "entropy": 5.491376972198486,
      "epoch": 2.8721826809015423,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004183551281768586,
      "loss": 5.0668,
      "mean_token_accuracy": 0.20981571972370147,
      "num_tokens": 66534111.0,
      "step": 29055
    },
    {
      "entropy": 5.444293785095215,
      "epoch": 2.8726769474100435,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041832792654926333,
      "loss": 5.0503,
      "mean_token_accuracy": 0.2121535837650299,
      "num_tokens": 66545605.0,
      "step": 29060
    },
    {
      "entropy": 5.397779512405395,
      "epoch": 2.873171213918545,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041830072139577886,
      "loss": 5.0129,
      "mean_token_accuracy": 0.21744753122329713,
      "num_tokens": 66556579.0,
      "step": 29065
    },
    {
      "entropy": 5.420639133453369,
      "epoch": 2.8736654804270465,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041827351271707435,
      "loss": 5.0416,
      "mean_token_accuracy": 0.20862719118595124,
      "num_tokens": 66570262.0,
      "step": 29070
    },
    {
      "entropy": 5.52435712814331,
      "epoch": 2.8741597469355478,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041824630051381917,
      "loss": 5.0917,
      "mean_token_accuracy": 0.19899357557296754,
      "num_tokens": 66581159.0,
      "step": 29075
    },
    {
      "entropy": 5.406038093566894,
      "epoch": 2.874654013444049,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041821908478668283,
      "loss": 5.0044,
      "mean_token_accuracy": 0.22269929945468903,
      "num_tokens": 66591641.0,
      "step": 29080
    },
    {
      "entropy": 5.349451398849487,
      "epoch": 2.8751482799525503,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041819186553633476,
      "loss": 5.0629,
      "mean_token_accuracy": 0.2039193645119667,
      "num_tokens": 66603737.0,
      "step": 29085
    },
    {
      "entropy": 5.468878746032715,
      "epoch": 2.8756425464610516,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041816464276344463,
      "loss": 5.0534,
      "mean_token_accuracy": 0.20736332088708878,
      "num_tokens": 66615874.0,
      "step": 29090
    },
    {
      "entropy": 5.5321571826934814,
      "epoch": 2.8761368129695533,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041813741646868204,
      "loss": 5.0871,
      "mean_token_accuracy": 0.20669715851545334,
      "num_tokens": 66627297.0,
      "step": 29095
    },
    {
      "entropy": 5.521060848236084,
      "epoch": 2.8766310794780545,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004181101866527168,
      "loss": 5.1055,
      "mean_token_accuracy": 0.2055671274662018,
      "num_tokens": 66638342.0,
      "step": 29100
    },
    {
      "entropy": 5.453798151016235,
      "epoch": 2.8771253459865562,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004180829533162188,
      "loss": 4.9793,
      "mean_token_accuracy": 0.21441750675439836,
      "num_tokens": 66648599.0,
      "step": 29105
    },
    {
      "entropy": 5.308062553405762,
      "epoch": 2.8776196124950575,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004180557164598579,
      "loss": 4.9655,
      "mean_token_accuracy": 0.21880108416080474,
      "num_tokens": 66660850.0,
      "step": 29110
    },
    {
      "entropy": 5.361252927780152,
      "epoch": 2.8781138790035588,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004180284760843042,
      "loss": 5.0115,
      "mean_token_accuracy": 0.21193771213293075,
      "num_tokens": 66672577.0,
      "step": 29115
    },
    {
      "entropy": 5.532903861999512,
      "epoch": 2.87860814551206,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041800123219022786,
      "loss": 5.0706,
      "mean_token_accuracy": 0.20973120182752608,
      "num_tokens": 66683720.0,
      "step": 29120
    },
    {
      "entropy": 5.514287614822388,
      "epoch": 2.8791024120205613,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004179739847782989,
      "loss": 5.1069,
      "mean_token_accuracy": 0.20648871958255768,
      "num_tokens": 66696492.0,
      "step": 29125
    },
    {
      "entropy": 5.402522277832031,
      "epoch": 2.879596678529063,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004179467338491878,
      "loss": 5.0591,
      "mean_token_accuracy": 0.2091243639588356,
      "num_tokens": 66707203.0,
      "step": 29130
    },
    {
      "entropy": 5.4554685115814205,
      "epoch": 2.8800909450375642,
      "grad_norm": 1.125,
      "learning_rate": 0.0004179194794035649,
      "loss": 5.0239,
      "mean_token_accuracy": 0.20720912516117096,
      "num_tokens": 66717433.0,
      "step": 29135
    },
    {
      "entropy": 5.389675760269165,
      "epoch": 2.8805852115460655,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041789222144210066,
      "loss": 4.9107,
      "mean_token_accuracy": 0.22658638805150985,
      "num_tokens": 66729452.0,
      "step": 29140
    },
    {
      "entropy": 5.383780765533447,
      "epoch": 2.881079478054567,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041786495996546546,
      "loss": 5.0445,
      "mean_token_accuracy": 0.21312773525714873,
      "num_tokens": 66741360.0,
      "step": 29145
    },
    {
      "entropy": 5.422603464126587,
      "epoch": 2.8815737445630685,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004178376949743302,
      "loss": 5.0384,
      "mean_token_accuracy": 0.2102280631661415,
      "num_tokens": 66752585.0,
      "step": 29150
    },
    {
      "entropy": 5.407947969436646,
      "epoch": 2.8820680110715697,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004178104264693655,
      "loss": 5.016,
      "mean_token_accuracy": 0.21438546627759933,
      "num_tokens": 66763898.0,
      "step": 29155
    },
    {
      "entropy": 5.357155990600586,
      "epoch": 2.882562277580071,
      "grad_norm": 1.03125,
      "learning_rate": 0.000417783154451242,
      "loss": 4.8964,
      "mean_token_accuracy": 0.22477778494358064,
      "num_tokens": 66774272.0,
      "step": 29160
    },
    {
      "entropy": 5.414708948135376,
      "epoch": 2.8830565440885727,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041775587892063095,
      "loss": 4.9726,
      "mean_token_accuracy": 0.2168130725622177,
      "num_tokens": 66784943.0,
      "step": 29165
    },
    {
      "entropy": 5.572418737411499,
      "epoch": 2.883550810597074,
      "grad_norm": 1.015625,
      "learning_rate": 0.000417728599878203,
      "loss": 5.1095,
      "mean_token_accuracy": 0.20130929946899415,
      "num_tokens": 66796449.0,
      "step": 29170
    },
    {
      "entropy": 5.412405633926392,
      "epoch": 2.8840450771055752,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004177013173246294,
      "loss": 5.0455,
      "mean_token_accuracy": 0.21262743473052978,
      "num_tokens": 66809936.0,
      "step": 29175
    },
    {
      "entropy": 5.483620548248291,
      "epoch": 2.884539343614077,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004176740312605813,
      "loss": 5.0774,
      "mean_token_accuracy": 0.20734924226999282,
      "num_tokens": 66820846.0,
      "step": 29180
    },
    {
      "entropy": 5.499366760253906,
      "epoch": 2.885033610122578,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004176467416867298,
      "loss": 5.038,
      "mean_token_accuracy": 0.20632170885801315,
      "num_tokens": 66832919.0,
      "step": 29185
    },
    {
      "entropy": 5.382720184326172,
      "epoch": 2.8855278766310795,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041761944860374637,
      "loss": 4.9553,
      "mean_token_accuracy": 0.22057608515024185,
      "num_tokens": 66844944.0,
      "step": 29190
    },
    {
      "entropy": 5.383631372451783,
      "epoch": 2.8860221431395807,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004175921520123023,
      "loss": 5.0302,
      "mean_token_accuracy": 0.20763741433620453,
      "num_tokens": 66857987.0,
      "step": 29195
    },
    {
      "entropy": 5.4278984546661375,
      "epoch": 2.886516409648082,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004175648519130693,
      "loss": 4.9715,
      "mean_token_accuracy": 0.21713627129793167,
      "num_tokens": 66869100.0,
      "step": 29200
    },
    {
      "entropy": 5.386370754241943,
      "epoch": 2.8870106761565837,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041753754830671876,
      "loss": 4.9808,
      "mean_token_accuracy": 0.21103835552930833,
      "num_tokens": 66881168.0,
      "step": 29205
    },
    {
      "entropy": 5.4128196239471436,
      "epoch": 2.887504942665085,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004175102411939224,
      "loss": 5.0532,
      "mean_token_accuracy": 0.21214286386966705,
      "num_tokens": 66894080.0,
      "step": 29210
    },
    {
      "entropy": 5.425430679321289,
      "epoch": 2.8879992091735867,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041748293057535207,
      "loss": 5.0386,
      "mean_token_accuracy": 0.21126446574926377,
      "num_tokens": 66906155.0,
      "step": 29215
    },
    {
      "entropy": 5.434283590316772,
      "epoch": 2.888493475682088,
      "grad_norm": 1.125,
      "learning_rate": 0.00041745561645167953,
      "loss": 4.956,
      "mean_token_accuracy": 0.2202584519982338,
      "num_tokens": 66915745.0,
      "step": 29220
    },
    {
      "entropy": 5.472662353515625,
      "epoch": 2.888987742190589,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004174282988235767,
      "loss": 5.0641,
      "mean_token_accuracy": 0.21087280958890914,
      "num_tokens": 66927722.0,
      "step": 29225
    },
    {
      "entropy": 5.447684574127197,
      "epoch": 2.8894820086990904,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004174009776917156,
      "loss": 5.0496,
      "mean_token_accuracy": 0.20940049290657042,
      "num_tokens": 66939063.0,
      "step": 29230
    },
    {
      "entropy": 5.396454763412476,
      "epoch": 2.8899762752075917,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004173736530567684,
      "loss": 4.9423,
      "mean_token_accuracy": 0.2206297680735588,
      "num_tokens": 66950596.0,
      "step": 29235
    },
    {
      "entropy": 5.369441366195678,
      "epoch": 2.8904705417160934,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004173463249194072,
      "loss": 4.9796,
      "mean_token_accuracy": 0.2126283824443817,
      "num_tokens": 66961836.0,
      "step": 29240
    },
    {
      "entropy": 5.481157302856445,
      "epoch": 2.8909648082245947,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041731899328030443,
      "loss": 5.1406,
      "mean_token_accuracy": 0.20322689414024353,
      "num_tokens": 66976064.0,
      "step": 29245
    },
    {
      "entropy": 5.398590755462647,
      "epoch": 2.891459074733096,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004172916581401322,
      "loss": 4.9137,
      "mean_token_accuracy": 0.22454738318920137,
      "num_tokens": 66987572.0,
      "step": 29250
    },
    {
      "entropy": 5.363443326950073,
      "epoch": 2.8919533412415976,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004172643194995632,
      "loss": 4.9412,
      "mean_token_accuracy": 0.2234678491950035,
      "num_tokens": 66998668.0,
      "step": 29255
    },
    {
      "entropy": 5.414989471435547,
      "epoch": 2.892447607750099,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004172369773592699,
      "loss": 5.0579,
      "mean_token_accuracy": 0.21215982586145402,
      "num_tokens": 67010140.0,
      "step": 29260
    },
    {
      "entropy": 5.326801490783692,
      "epoch": 2.8929418742586,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004172096317199248,
      "loss": 4.878,
      "mean_token_accuracy": 0.2305400550365448,
      "num_tokens": 67021727.0,
      "step": 29265
    },
    {
      "entropy": 5.379196310043335,
      "epoch": 2.8934361407671014,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004171822825822007,
      "loss": 5.0152,
      "mean_token_accuracy": 0.21398012936115265,
      "num_tokens": 67033327.0,
      "step": 29270
    },
    {
      "entropy": 5.503091335296631,
      "epoch": 2.893930407275603,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004171549299467704,
      "loss": 5.0833,
      "mean_token_accuracy": 0.21331964135169984,
      "num_tokens": 67044827.0,
      "step": 29275
    },
    {
      "entropy": 5.465588188171386,
      "epoch": 2.8944246737841044,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004171275738143068,
      "loss": 4.9373,
      "mean_token_accuracy": 0.21313460171222687,
      "num_tokens": 67056042.0,
      "step": 29280
    },
    {
      "entropy": 5.467501831054688,
      "epoch": 2.8949189402926057,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041710021418548283,
      "loss": 5.0435,
      "mean_token_accuracy": 0.21317075788974763,
      "num_tokens": 67068346.0,
      "step": 29285
    },
    {
      "entropy": 5.366368341445923,
      "epoch": 2.8954132068011074,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004170728510609716,
      "loss": 4.9473,
      "mean_token_accuracy": 0.2216492161154747,
      "num_tokens": 67078282.0,
      "step": 29290
    },
    {
      "entropy": 5.403209543228149,
      "epoch": 2.8959074733096086,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041704548444144605,
      "loss": 5.0205,
      "mean_token_accuracy": 0.2127322033047676,
      "num_tokens": 67089958.0,
      "step": 29295
    },
    {
      "entropy": 5.456916475296021,
      "epoch": 2.89640173981811,
      "grad_norm": 1.0,
      "learning_rate": 0.00041701811432757964,
      "loss": 5.0311,
      "mean_token_accuracy": 0.212491375207901,
      "num_tokens": 67101672.0,
      "step": 29300
    },
    {
      "entropy": 5.404790544509888,
      "epoch": 2.896896006326611,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004169907407200456,
      "loss": 4.9678,
      "mean_token_accuracy": 0.20835311859846115,
      "num_tokens": 67113334.0,
      "step": 29305
    },
    {
      "entropy": 5.480239343643189,
      "epoch": 2.897390272835113,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041696336361951727,
      "loss": 5.0483,
      "mean_token_accuracy": 0.20734936892986297,
      "num_tokens": 67123880.0,
      "step": 29310
    },
    {
      "entropy": 5.421977853775024,
      "epoch": 2.897884539343614,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004169359830266682,
      "loss": 5.05,
      "mean_token_accuracy": 0.20636602491140366,
      "num_tokens": 67134966.0,
      "step": 29315
    },
    {
      "entropy": 5.403618240356446,
      "epoch": 2.8983788058521154,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004169085989421719,
      "loss": 4.9974,
      "mean_token_accuracy": 0.21643251031637192,
      "num_tokens": 67145407.0,
      "step": 29320
    },
    {
      "entropy": 5.408173561096191,
      "epoch": 2.898873072360617,
      "grad_norm": 1.0,
      "learning_rate": 0.0004168812113667021,
      "loss": 5.0213,
      "mean_token_accuracy": 0.21050399988889695,
      "num_tokens": 67156201.0,
      "step": 29325
    },
    {
      "entropy": 5.390816497802734,
      "epoch": 2.8993673388691183,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041685382030093245,
      "loss": 5.0013,
      "mean_token_accuracy": 0.21631201952695847,
      "num_tokens": 67168851.0,
      "step": 29330
    },
    {
      "entropy": 5.335356569290161,
      "epoch": 2.8998616053776196,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041682642574553683,
      "loss": 4.8736,
      "mean_token_accuracy": 0.23464400619268416,
      "num_tokens": 67180559.0,
      "step": 29335
    },
    {
      "entropy": 5.434173631668091,
      "epoch": 2.900355871886121,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041679902770118914,
      "loss": 5.0402,
      "mean_token_accuracy": 0.20814952850341797,
      "num_tokens": 67192494.0,
      "step": 29340
    },
    {
      "entropy": 5.410094404220581,
      "epoch": 2.900850138394622,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041677162616856337,
      "loss": 5.0181,
      "mean_token_accuracy": 0.21588039994239808,
      "num_tokens": 67203846.0,
      "step": 29345
    },
    {
      "entropy": 5.481628179550171,
      "epoch": 2.901344404903124,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00041674422114833357,
      "loss": 5.0053,
      "mean_token_accuracy": 0.20902076810598375,
      "num_tokens": 67214528.0,
      "step": 29350
    },
    {
      "entropy": 5.40119423866272,
      "epoch": 2.901838671411625,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004167168126411739,
      "loss": 5.0556,
      "mean_token_accuracy": 0.21299450546503068,
      "num_tokens": 67225798.0,
      "step": 29355
    },
    {
      "entropy": 5.525965738296509,
      "epoch": 2.902332937920127,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004166894006477587,
      "loss": 5.1571,
      "mean_token_accuracy": 0.20394525974988936,
      "num_tokens": 67237245.0,
      "step": 29360
    },
    {
      "entropy": 5.491499567031861,
      "epoch": 2.902827204428628,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004166619851687623,
      "loss": 5.0392,
      "mean_token_accuracy": 0.20619229674339296,
      "num_tokens": 67249741.0,
      "step": 29365
    },
    {
      "entropy": 5.458441877365113,
      "epoch": 2.9033214709371293,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041663456620485913,
      "loss": 5.0922,
      "mean_token_accuracy": 0.2089219644665718,
      "num_tokens": 67261452.0,
      "step": 29370
    },
    {
      "entropy": 5.479209613800049,
      "epoch": 2.9038157374456306,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004166071437567236,
      "loss": 5.0353,
      "mean_token_accuracy": 0.21224239766597747,
      "num_tokens": 67273420.0,
      "step": 29375
    },
    {
      "entropy": 5.478652715682983,
      "epoch": 2.904310003954132,
      "grad_norm": 1.0,
      "learning_rate": 0.0004165797178250303,
      "loss": 5.0892,
      "mean_token_accuracy": 0.20319828689098357,
      "num_tokens": 67284853.0,
      "step": 29380
    },
    {
      "entropy": 5.32659683227539,
      "epoch": 2.9048042704626336,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041655228841045404,
      "loss": 4.8902,
      "mean_token_accuracy": 0.2282964378595352,
      "num_tokens": 67295956.0,
      "step": 29385
    },
    {
      "entropy": 5.436525583267212,
      "epoch": 2.905298536971135,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041652485551366955,
      "loss": 5.029,
      "mean_token_accuracy": 0.21798741519451142,
      "num_tokens": 67307150.0,
      "step": 29390
    },
    {
      "entropy": 5.407930469512939,
      "epoch": 2.905792803479636,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004164974191353516,
      "loss": 4.9151,
      "mean_token_accuracy": 0.22059123516082763,
      "num_tokens": 67318012.0,
      "step": 29395
    },
    {
      "entropy": 5.35202693939209,
      "epoch": 2.906287069988138,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004164699792761753,
      "loss": 5.0086,
      "mean_token_accuracy": 0.21235252767801285,
      "num_tokens": 67331335.0,
      "step": 29400
    },
    {
      "entropy": 5.489719438552856,
      "epoch": 2.906781336496639,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041644253593681543,
      "loss": 5.051,
      "mean_token_accuracy": 0.20618800818920135,
      "num_tokens": 67342595.0,
      "step": 29405
    },
    {
      "entropy": 5.460517358779907,
      "epoch": 2.9072756030051403,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00041641508911794724,
      "loss": 5.0206,
      "mean_token_accuracy": 0.21091959327459336,
      "num_tokens": 67355620.0,
      "step": 29410
    },
    {
      "entropy": 5.444975519180298,
      "epoch": 2.9077698695136416,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000416387638820246,
      "loss": 5.0063,
      "mean_token_accuracy": 0.21085948050022124,
      "num_tokens": 67366600.0,
      "step": 29415
    },
    {
      "entropy": 5.4302386283874515,
      "epoch": 2.9082641360221433,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041636018504438683,
      "loss": 5.016,
      "mean_token_accuracy": 0.21166083663702012,
      "num_tokens": 67377122.0,
      "step": 29420
    },
    {
      "entropy": 5.431444549560547,
      "epoch": 2.9087584025306445,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004163327277910452,
      "loss": 5.0779,
      "mean_token_accuracy": 0.20935009568929672,
      "num_tokens": 67387572.0,
      "step": 29425
    },
    {
      "entropy": 5.399361181259155,
      "epoch": 2.909252669039146,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041630526706089657,
      "loss": 4.9452,
      "mean_token_accuracy": 0.21774937510490416,
      "num_tokens": 67399608.0,
      "step": 29430
    },
    {
      "entropy": 5.472567081451416,
      "epoch": 2.9097469355476475,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041627780285461645,
      "loss": 5.106,
      "mean_token_accuracy": 0.1999306008219719,
      "num_tokens": 67410144.0,
      "step": 29435
    },
    {
      "entropy": 5.454007482528686,
      "epoch": 2.910241202056149,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041625033517288045,
      "loss": 5.0067,
      "mean_token_accuracy": 0.2140176549553871,
      "num_tokens": 67422777.0,
      "step": 29440
    },
    {
      "entropy": 5.43764009475708,
      "epoch": 2.91073546856465,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004162228640163643,
      "loss": 5.0945,
      "mean_token_accuracy": 0.20747531354427337,
      "num_tokens": 67434226.0,
      "step": 29445
    },
    {
      "entropy": 5.513816976547242,
      "epoch": 2.9112297350731513,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004161953893857438,
      "loss": 5.1023,
      "mean_token_accuracy": 0.20979815423488618,
      "num_tokens": 67445316.0,
      "step": 29450
    },
    {
      "entropy": 5.450288200378418,
      "epoch": 2.9117240015816526,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004161679112816948,
      "loss": 5.0611,
      "mean_token_accuracy": 0.20746947526931764,
      "num_tokens": 67456973.0,
      "step": 29455
    },
    {
      "entropy": 5.339000797271728,
      "epoch": 2.9122182680901543,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041614042970489326,
      "loss": 4.9281,
      "mean_token_accuracy": 0.22249667197465897,
      "num_tokens": 67468134.0,
      "step": 29460
    },
    {
      "entropy": 5.464120578765869,
      "epoch": 2.9127125345986555,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041611294465601525,
      "loss": 5.0719,
      "mean_token_accuracy": 0.2054212436079979,
      "num_tokens": 67479973.0,
      "step": 29465
    },
    {
      "entropy": 5.461737489700317,
      "epoch": 2.9132068011071572,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041608545613573694,
      "loss": 5.0215,
      "mean_token_accuracy": 0.2158416450023651,
      "num_tokens": 67491469.0,
      "step": 29470
    },
    {
      "entropy": 5.4495023727417,
      "epoch": 2.9137010676156585,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004160579641447345,
      "loss": 5.103,
      "mean_token_accuracy": 0.20878621190786362,
      "num_tokens": 67502911.0,
      "step": 29475
    },
    {
      "entropy": 5.496672058105469,
      "epoch": 2.9141953341241598,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041603046868368427,
      "loss": 5.1399,
      "mean_token_accuracy": 0.20336408913135529,
      "num_tokens": 67513958.0,
      "step": 29480
    },
    {
      "entropy": 5.4117344379425045,
      "epoch": 2.914689600632661,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041600296975326267,
      "loss": 5.0302,
      "mean_token_accuracy": 0.21220000088214874,
      "num_tokens": 67524894.0,
      "step": 29485
    },
    {
      "entropy": 5.406762170791626,
      "epoch": 2.9151838671411623,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004159754673541461,
      "loss": 4.9453,
      "mean_token_accuracy": 0.22098466604948044,
      "num_tokens": 67536256.0,
      "step": 29490
    },
    {
      "entropy": 5.430414056777954,
      "epoch": 2.915678133649664,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041594796148701116,
      "loss": 4.9947,
      "mean_token_accuracy": 0.21433212459087372,
      "num_tokens": 67548064.0,
      "step": 29495
    },
    {
      "entropy": 5.470881938934326,
      "epoch": 2.9161724001581653,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004159204521525345,
      "loss": 5.0271,
      "mean_token_accuracy": 0.2102196842432022,
      "num_tokens": 67559530.0,
      "step": 29500
    },
    {
      "entropy": 5.45843095779419,
      "epoch": 2.9166666666666665,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004158929393513929,
      "loss": 5.1169,
      "mean_token_accuracy": 0.21055484414100648,
      "num_tokens": 67571633.0,
      "step": 29505
    },
    {
      "entropy": 5.467301845550537,
      "epoch": 2.9171609331751682,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041586542308426306,
      "loss": 4.9826,
      "mean_token_accuracy": 0.21196904629468918,
      "num_tokens": 67583778.0,
      "step": 29510
    },
    {
      "entropy": 5.3847614288330075,
      "epoch": 2.9176551996836695,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041583790335182196,
      "loss": 4.9428,
      "mean_token_accuracy": 0.22037430107593536,
      "num_tokens": 67595803.0,
      "step": 29515
    },
    {
      "entropy": 5.35503625869751,
      "epoch": 2.9181494661921707,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041581038015474673,
      "loss": 5.0153,
      "mean_token_accuracy": 0.2126976504921913,
      "num_tokens": 67606030.0,
      "step": 29520
    },
    {
      "entropy": 5.471170043945312,
      "epoch": 2.918643732700672,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041578285349371414,
      "loss": 5.1335,
      "mean_token_accuracy": 0.20745330601930617,
      "num_tokens": 67617632.0,
      "step": 29525
    },
    {
      "entropy": 5.472629690170288,
      "epoch": 2.9191379992091737,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004157553233694016,
      "loss": 4.9894,
      "mean_token_accuracy": 0.21809024661779403,
      "num_tokens": 67627954.0,
      "step": 29530
    },
    {
      "entropy": 5.399608850479126,
      "epoch": 2.919632265717675,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004157277897824862,
      "loss": 5.0145,
      "mean_token_accuracy": 0.21048288494348527,
      "num_tokens": 67639180.0,
      "step": 29535
    },
    {
      "entropy": 5.388995552062989,
      "epoch": 2.9201265322261762,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004157002527336454,
      "loss": 5.106,
      "mean_token_accuracy": 0.20704077929258347,
      "num_tokens": 67651834.0,
      "step": 29540
    },
    {
      "entropy": 5.442848491668701,
      "epoch": 2.920620798734678,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041567271222355657,
      "loss": 4.9851,
      "mean_token_accuracy": 0.2132923573255539,
      "num_tokens": 67663263.0,
      "step": 29545
    },
    {
      "entropy": 5.473592853546142,
      "epoch": 2.921115065243179,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004156451682528971,
      "loss": 5.0195,
      "mean_token_accuracy": 0.20816118121147156,
      "num_tokens": 67673970.0,
      "step": 29550
    },
    {
      "entropy": 5.541548538208008,
      "epoch": 2.9216093317516805,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041561762082234467,
      "loss": 5.1055,
      "mean_token_accuracy": 0.20621173977851867,
      "num_tokens": 67685668.0,
      "step": 29555
    },
    {
      "entropy": 5.3782542705535885,
      "epoch": 2.9221035982601817,
      "grad_norm": 1.046875,
      "learning_rate": 0.000415590069932577,
      "loss": 4.9419,
      "mean_token_accuracy": 0.22660989463329315,
      "num_tokens": 67696169.0,
      "step": 29560
    },
    {
      "entropy": 5.498178100585937,
      "epoch": 2.9225978647686834,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004155625155842719,
      "loss": 5.0873,
      "mean_token_accuracy": 0.20074700117111205,
      "num_tokens": 67709211.0,
      "step": 29565
    },
    {
      "entropy": 5.469609546661377,
      "epoch": 2.9230921312771847,
      "grad_norm": 1.046875,
      "learning_rate": 0.000415534957778107,
      "loss": 5.0974,
      "mean_token_accuracy": 0.2067451223731041,
      "num_tokens": 67720967.0,
      "step": 29570
    },
    {
      "entropy": 5.430124378204345,
      "epoch": 2.923586397785686,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004155073965147604,
      "loss": 4.945,
      "mean_token_accuracy": 0.21931995302438737,
      "num_tokens": 67733099.0,
      "step": 29575
    },
    {
      "entropy": 5.390460348129272,
      "epoch": 2.9240806642941877,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041547983179491,
      "loss": 5.0073,
      "mean_token_accuracy": 0.21467345654964448,
      "num_tokens": 67744739.0,
      "step": 29580
    },
    {
      "entropy": 5.403063631057739,
      "epoch": 2.924574930802689,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041545226361923394,
      "loss": 5.0877,
      "mean_token_accuracy": 0.21051979959011077,
      "num_tokens": 67756944.0,
      "step": 29585
    },
    {
      "entropy": 5.508056592941284,
      "epoch": 2.92506919731119,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041542469198841043,
      "loss": 5.11,
      "mean_token_accuracy": 0.2051416426897049,
      "num_tokens": 67769018.0,
      "step": 29590
    },
    {
      "entropy": 5.469253206253052,
      "epoch": 2.9255634638196915,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004153971169031177,
      "loss": 5.0022,
      "mean_token_accuracy": 0.21390393525362014,
      "num_tokens": 67780164.0,
      "step": 29595
    },
    {
      "entropy": 5.462141847610473,
      "epoch": 2.9260577303281927,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004153695383640341,
      "loss": 5.0358,
      "mean_token_accuracy": 0.21115020513534546,
      "num_tokens": 67790841.0,
      "step": 29600
    },
    {
      "entropy": 5.4433776378631595,
      "epoch": 2.9265519968366944,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041534195637183807,
      "loss": 5.0333,
      "mean_token_accuracy": 0.2089438945055008,
      "num_tokens": 67802397.0,
      "step": 29605
    },
    {
      "entropy": 5.437507629394531,
      "epoch": 2.9270462633451957,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004153143709272082,
      "loss": 5.0451,
      "mean_token_accuracy": 0.21321855783462523,
      "num_tokens": 67814079.0,
      "step": 29610
    },
    {
      "entropy": 5.473219871520996,
      "epoch": 2.927540529853697,
      "grad_norm": 1.015625,
      "learning_rate": 0.000415286782030823,
      "loss": 5.0869,
      "mean_token_accuracy": 0.20599968880414962,
      "num_tokens": 67825377.0,
      "step": 29615
    },
    {
      "entropy": 5.370988655090332,
      "epoch": 2.9280347963621987,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041525918968336116,
      "loss": 4.9968,
      "mean_token_accuracy": 0.22287837862968446,
      "num_tokens": 67838277.0,
      "step": 29620
    },
    {
      "entropy": 5.4456212520599365,
      "epoch": 2.9285290628707,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041523159388550155,
      "loss": 5.0376,
      "mean_token_accuracy": 0.21056950688362122,
      "num_tokens": 67849720.0,
      "step": 29625
    },
    {
      "entropy": 5.4076502323150635,
      "epoch": 2.929023329379201,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041520399463792297,
      "loss": 4.9846,
      "mean_token_accuracy": 0.2173058345913887,
      "num_tokens": 67860994.0,
      "step": 29630
    },
    {
      "entropy": 5.394793605804443,
      "epoch": 2.9295175958877024,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00041517639194130435,
      "loss": 5.1036,
      "mean_token_accuracy": 0.20377420485019684,
      "num_tokens": 67874445.0,
      "step": 29635
    },
    {
      "entropy": 5.426474905014038,
      "epoch": 2.930011862396204,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041514878579632477,
      "loss": 4.9666,
      "mean_token_accuracy": 0.21433853656053542,
      "num_tokens": 67886789.0,
      "step": 29640
    },
    {
      "entropy": 5.513182497024536,
      "epoch": 2.9305061289047054,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004151211762036632,
      "loss": 5.0627,
      "mean_token_accuracy": 0.2147467166185379,
      "num_tokens": 67898839.0,
      "step": 29645
    },
    {
      "entropy": 5.43680510520935,
      "epoch": 2.9310003954132067,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041509356316399916,
      "loss": 5.0764,
      "mean_token_accuracy": 0.20469371676445008,
      "num_tokens": 67909419.0,
      "step": 29650
    },
    {
      "entropy": 5.370168781280517,
      "epoch": 2.9314946619217084,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041506594667801164,
      "loss": 5.0222,
      "mean_token_accuracy": 0.2114386036992073,
      "num_tokens": 67922011.0,
      "step": 29655
    },
    {
      "entropy": 5.510838317871094,
      "epoch": 2.9319889284302096,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041503832674638,
      "loss": 5.1727,
      "mean_token_accuracy": 0.19935372620820999,
      "num_tokens": 67934445.0,
      "step": 29660
    },
    {
      "entropy": 5.5285179138183596,
      "epoch": 2.932483194938711,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004150107033697839,
      "loss": 5.0974,
      "mean_token_accuracy": 0.2022954374551773,
      "num_tokens": 67944914.0,
      "step": 29665
    },
    {
      "entropy": 5.417542791366577,
      "epoch": 2.932977461447212,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004149830765489028,
      "loss": 5.0579,
      "mean_token_accuracy": 0.21152153611183167,
      "num_tokens": 67955375.0,
      "step": 29670
    },
    {
      "entropy": 5.486369180679321,
      "epoch": 2.933471727955714,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041495544628441627,
      "loss": 5.0195,
      "mean_token_accuracy": 0.21479301750659943,
      "num_tokens": 67968453.0,
      "step": 29675
    },
    {
      "entropy": 5.47714319229126,
      "epoch": 2.933965994464215,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041492781257700405,
      "loss": 5.0378,
      "mean_token_accuracy": 0.20948146730661393,
      "num_tokens": 67979099.0,
      "step": 29680
    },
    {
      "entropy": 5.426479864120483,
      "epoch": 2.9344602609727164,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00041490017542734594,
      "loss": 5.0022,
      "mean_token_accuracy": 0.21084415763616562,
      "num_tokens": 67989399.0,
      "step": 29685
    },
    {
      "entropy": 5.432713890075684,
      "epoch": 2.934954527481218,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041487253483612173,
      "loss": 4.9375,
      "mean_token_accuracy": 0.22332856953144073,
      "num_tokens": 68001255.0,
      "step": 29690
    },
    {
      "entropy": 5.423839569091797,
      "epoch": 2.9354487939897194,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004148448908040116,
      "loss": 5.0283,
      "mean_token_accuracy": 0.21043030321598052,
      "num_tokens": 68012844.0,
      "step": 29695
    },
    {
      "entropy": 5.377053594589233,
      "epoch": 2.9359430604982206,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004148172433316954,
      "loss": 5.025,
      "mean_token_accuracy": 0.21104152947664262,
      "num_tokens": 68024986.0,
      "step": 29700
    },
    {
      "entropy": 5.496196317672729,
      "epoch": 2.936437327006722,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004147895924198533,
      "loss": 5.0002,
      "mean_token_accuracy": 0.2192436307668686,
      "num_tokens": 68036081.0,
      "step": 29705
    },
    {
      "entropy": 5.456634140014648,
      "epoch": 2.936931593515223,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00041476193806916557,
      "loss": 5.0424,
      "mean_token_accuracy": 0.2120230183005333,
      "num_tokens": 68049054.0,
      "step": 29710
    },
    {
      "entropy": 5.479152870178223,
      "epoch": 2.937425860023725,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004147342802803124,
      "loss": 5.0652,
      "mean_token_accuracy": 0.21160530000925065,
      "num_tokens": 68060440.0,
      "step": 29715
    },
    {
      "entropy": 5.34980263710022,
      "epoch": 2.937920126532226,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004147066190539744,
      "loss": 5.053,
      "mean_token_accuracy": 0.21492915451526642,
      "num_tokens": 68072260.0,
      "step": 29720
    },
    {
      "entropy": 5.431468152999878,
      "epoch": 2.938414393040728,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004146789543908318,
      "loss": 5.0006,
      "mean_token_accuracy": 0.20896161049604417,
      "num_tokens": 68084737.0,
      "step": 29725
    },
    {
      "entropy": 5.444734525680542,
      "epoch": 2.938908659549229,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004146512862915653,
      "loss": 4.9945,
      "mean_token_accuracy": 0.21797845065593718,
      "num_tokens": 68097146.0,
      "step": 29730
    },
    {
      "entropy": 5.523804616928101,
      "epoch": 2.9394029260577303,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004146236147568554,
      "loss": 5.1032,
      "mean_token_accuracy": 0.20520014315843582,
      "num_tokens": 68108158.0,
      "step": 29735
    },
    {
      "entropy": 5.357573461532593,
      "epoch": 2.9398971925662316,
      "grad_norm": 1.0,
      "learning_rate": 0.0004145959397873831,
      "loss": 4.9716,
      "mean_token_accuracy": 0.2156193658709526,
      "num_tokens": 68120242.0,
      "step": 29740
    },
    {
      "entropy": 5.45639181137085,
      "epoch": 2.940391459074733,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041456826138382884,
      "loss": 5.0986,
      "mean_token_accuracy": 0.20665988177061081,
      "num_tokens": 68131372.0,
      "step": 29745
    },
    {
      "entropy": 5.4489704132080075,
      "epoch": 2.9408857255832346,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004145405795468738,
      "loss": 5.0631,
      "mean_token_accuracy": 0.21649622917175293,
      "num_tokens": 68142828.0,
      "step": 29750
    },
    {
      "entropy": 5.445520401000977,
      "epoch": 2.941379992091736,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004145128942771988,
      "loss": 5.1082,
      "mean_token_accuracy": 0.2005729541182518,
      "num_tokens": 68154398.0,
      "step": 29755
    },
    {
      "entropy": 5.4622945308685305,
      "epoch": 2.941874258600237,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000414485205575485,
      "loss": 5.0797,
      "mean_token_accuracy": 0.20611651688814164,
      "num_tokens": 68166645.0,
      "step": 29760
    },
    {
      "entropy": 5.480318880081176,
      "epoch": 2.942368525108739,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041445751344241344,
      "loss": 4.9827,
      "mean_token_accuracy": 0.2160264030098915,
      "num_tokens": 68176800.0,
      "step": 29765
    },
    {
      "entropy": 5.446160173416137,
      "epoch": 2.94286279161724,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041442981787866547,
      "loss": 5.0479,
      "mean_token_accuracy": 0.21194081157445907,
      "num_tokens": 68187933.0,
      "step": 29770
    },
    {
      "entropy": 5.458788824081421,
      "epoch": 2.9433570581257413,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004144021188849223,
      "loss": 5.1331,
      "mean_token_accuracy": 0.20279710590839387,
      "num_tokens": 68199925.0,
      "step": 29775
    },
    {
      "entropy": 5.37613525390625,
      "epoch": 2.9438513246342426,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004143744164618655,
      "loss": 4.9579,
      "mean_token_accuracy": 0.2154856115579605,
      "num_tokens": 68211313.0,
      "step": 29780
    },
    {
      "entropy": 5.450648450851441,
      "epoch": 2.9443455911427443,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041434671061017634,
      "loss": 5.0538,
      "mean_token_accuracy": 0.20749825239181519,
      "num_tokens": 68223148.0,
      "step": 29785
    },
    {
      "entropy": 5.4024779319763185,
      "epoch": 2.9448398576512456,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004143190013305364,
      "loss": 5.0097,
      "mean_token_accuracy": 0.21971790939569474,
      "num_tokens": 68234369.0,
      "step": 29790
    },
    {
      "entropy": 5.428864908218384,
      "epoch": 2.945334124159747,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041429128862362756,
      "loss": 5.009,
      "mean_token_accuracy": 0.21110998839139938,
      "num_tokens": 68243903.0,
      "step": 29795
    },
    {
      "entropy": 5.454585361480713,
      "epoch": 2.9458283906682485,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004142635724901313,
      "loss": 5.0245,
      "mean_token_accuracy": 0.20784134566783904,
      "num_tokens": 68254997.0,
      "step": 29800
    },
    {
      "entropy": 5.381480026245117,
      "epoch": 2.94632265717675,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004142358529307296,
      "loss": 4.9511,
      "mean_token_accuracy": 0.21778082102537155,
      "num_tokens": 68266203.0,
      "step": 29805
    },
    {
      "entropy": 5.444616413116455,
      "epoch": 2.946816923685251,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004142081299461044,
      "loss": 5.0297,
      "mean_token_accuracy": 0.21376969218254088,
      "num_tokens": 68275930.0,
      "step": 29810
    },
    {
      "entropy": 5.432279491424561,
      "epoch": 2.9473111901937523,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004141804035369375,
      "loss": 5.0059,
      "mean_token_accuracy": 0.21652431339025496,
      "num_tokens": 68287468.0,
      "step": 29815
    },
    {
      "entropy": 5.416721439361572,
      "epoch": 2.9478054567022536,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004141526737039111,
      "loss": 4.9901,
      "mean_token_accuracy": 0.20955535471439363,
      "num_tokens": 68299151.0,
      "step": 29820
    },
    {
      "entropy": 5.475926780700684,
      "epoch": 2.9482997232107553,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004141249404477074,
      "loss": 5.0812,
      "mean_token_accuracy": 0.19960540682077407,
      "num_tokens": 68310874.0,
      "step": 29825
    },
    {
      "entropy": 5.474725008010864,
      "epoch": 2.9487939897192565,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004140972037690085,
      "loss": 5.0172,
      "mean_token_accuracy": 0.21428060084581374,
      "num_tokens": 68323449.0,
      "step": 29830
    },
    {
      "entropy": 5.402032852172852,
      "epoch": 2.9492882562277583,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041406946366849686,
      "loss": 5.0158,
      "mean_token_accuracy": 0.217159466445446,
      "num_tokens": 68334799.0,
      "step": 29835
    },
    {
      "entropy": 5.371946382522583,
      "epoch": 2.9497825227362595,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041404172014685483,
      "loss": 4.9902,
      "mean_token_accuracy": 0.2194390505552292,
      "num_tokens": 68348019.0,
      "step": 29840
    },
    {
      "entropy": 5.523121643066406,
      "epoch": 2.9502767892447608,
      "grad_norm": 1.0,
      "learning_rate": 0.0004140139732047649,
      "loss": 5.0408,
      "mean_token_accuracy": 0.20583546608686448,
      "num_tokens": 68360969.0,
      "step": 29845
    },
    {
      "entropy": 5.450361061096191,
      "epoch": 2.950771055753262,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004139862228429097,
      "loss": 5.0556,
      "mean_token_accuracy": 0.21377750337123871,
      "num_tokens": 68373245.0,
      "step": 29850
    },
    {
      "entropy": 5.3475220680236815,
      "epoch": 2.9512653222617633,
      "grad_norm": 1.0,
      "learning_rate": 0.00041395846906197174,
      "loss": 4.972,
      "mean_token_accuracy": 0.22119609117507935,
      "num_tokens": 68385278.0,
      "step": 29855
    },
    {
      "entropy": 5.428192281723023,
      "epoch": 2.951759588770265,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041393071186263396,
      "loss": 4.9914,
      "mean_token_accuracy": 0.20961822867393493,
      "num_tokens": 68396624.0,
      "step": 29860
    },
    {
      "entropy": 5.452015829086304,
      "epoch": 2.9522538552787663,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004139029512455791,
      "loss": 4.9811,
      "mean_token_accuracy": 0.21468058675527574,
      "num_tokens": 68408656.0,
      "step": 29865
    },
    {
      "entropy": 5.423299551010132,
      "epoch": 2.9527481217872675,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004138751872114902,
      "loss": 4.9609,
      "mean_token_accuracy": 0.21442710161209105,
      "num_tokens": 68421150.0,
      "step": 29870
    },
    {
      "entropy": 5.496262407302856,
      "epoch": 2.9532423882957692,
      "grad_norm": 1.0,
      "learning_rate": 0.00041384741976105004,
      "loss": 5.0633,
      "mean_token_accuracy": 0.20297735780477524,
      "num_tokens": 68432446.0,
      "step": 29875
    },
    {
      "entropy": 5.375298595428466,
      "epoch": 2.9537366548042705,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041381964889494185,
      "loss": 5.0744,
      "mean_token_accuracy": 0.2089361071586609,
      "num_tokens": 68443508.0,
      "step": 29880
    },
    {
      "entropy": 5.421127462387085,
      "epoch": 2.9542309213127718,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004137918746138487,
      "loss": 5.0491,
      "mean_token_accuracy": 0.20866889953613282,
      "num_tokens": 68454425.0,
      "step": 29885
    },
    {
      "entropy": 5.423910856246948,
      "epoch": 2.954725187821273,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041376409691845395,
      "loss": 4.9974,
      "mean_token_accuracy": 0.21270785480737686,
      "num_tokens": 68467759.0,
      "step": 29890
    },
    {
      "entropy": 5.433946228027343,
      "epoch": 2.9552194543297747,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041373631580944093,
      "loss": 5.0112,
      "mean_token_accuracy": 0.2177659884095192,
      "num_tokens": 68478447.0,
      "step": 29895
    },
    {
      "entropy": 5.4856034278869625,
      "epoch": 2.955713720838276,
      "grad_norm": 1.078125,
      "learning_rate": 0.000413708531287493,
      "loss": 5.1154,
      "mean_token_accuracy": 0.20548634678125383,
      "num_tokens": 68489056.0,
      "step": 29900
    },
    {
      "entropy": 5.495143890380859,
      "epoch": 2.9562079873467773,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004136807433532936,
      "loss": 5.1022,
      "mean_token_accuracy": 0.2039782300591469,
      "num_tokens": 68500697.0,
      "step": 29905
    },
    {
      "entropy": 5.455454730987549,
      "epoch": 2.956702253855279,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004136529520075265,
      "loss": 4.9922,
      "mean_token_accuracy": 0.2124455526471138,
      "num_tokens": 68510594.0,
      "step": 29910
    },
    {
      "entropy": 5.410472202301025,
      "epoch": 2.9571965203637802,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041362515725087523,
      "loss": 5.0492,
      "mean_token_accuracy": 0.20675116777420044,
      "num_tokens": 68521293.0,
      "step": 29915
    },
    {
      "entropy": 5.399221277236938,
      "epoch": 2.9576907868722815,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041359735908402356,
      "loss": 4.9828,
      "mean_token_accuracy": 0.22450968325138093,
      "num_tokens": 68532380.0,
      "step": 29920
    },
    {
      "entropy": 5.445450782775879,
      "epoch": 2.9581850533807827,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041356955750765535,
      "loss": 5.0248,
      "mean_token_accuracy": 0.21461350172758104,
      "num_tokens": 68543510.0,
      "step": 29925
    },
    {
      "entropy": 5.4660967826843265,
      "epoch": 2.9586793198892845,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004135417525224546,
      "loss": 5.0983,
      "mean_token_accuracy": 0.20471111238002776,
      "num_tokens": 68556191.0,
      "step": 29930
    },
    {
      "entropy": 5.502834177017212,
      "epoch": 2.9591735863977857,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004135139441291052,
      "loss": 5.1435,
      "mean_token_accuracy": 0.2034518301486969,
      "num_tokens": 68568225.0,
      "step": 29935
    },
    {
      "entropy": 5.549093866348267,
      "epoch": 2.959667852906287,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004134861323282913,
      "loss": 5.0724,
      "mean_token_accuracy": 0.20381063967943192,
      "num_tokens": 68580040.0,
      "step": 29940
    },
    {
      "entropy": 5.409531450271606,
      "epoch": 2.9601621194147887,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041345831712069696,
      "loss": 4.9844,
      "mean_token_accuracy": 0.20694458335638047,
      "num_tokens": 68590338.0,
      "step": 29945
    },
    {
      "entropy": 5.413264179229737,
      "epoch": 2.96065638592329,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041343049850700673,
      "loss": 5.0419,
      "mean_token_accuracy": 0.2098769351840019,
      "num_tokens": 68601153.0,
      "step": 29950
    },
    {
      "entropy": 5.451947116851807,
      "epoch": 2.961150652431791,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041340267648790464,
      "loss": 5.0294,
      "mean_token_accuracy": 0.21531984806060792,
      "num_tokens": 68612665.0,
      "step": 29955
    },
    {
      "entropy": 5.376000118255615,
      "epoch": 2.9616449189402925,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004133748510640752,
      "loss": 4.9735,
      "mean_token_accuracy": 0.2169503316283226,
      "num_tokens": 68624229.0,
      "step": 29960
    },
    {
      "entropy": 5.4725548267364506,
      "epoch": 2.9621391854487937,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041334702223620306,
      "loss": 5.054,
      "mean_token_accuracy": 0.20738171339035033,
      "num_tokens": 68635811.0,
      "step": 29965
    },
    {
      "entropy": 5.405178546905518,
      "epoch": 2.9626334519572954,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041331919000497263,
      "loss": 4.9824,
      "mean_token_accuracy": 0.2144791975617409,
      "num_tokens": 68647518.0,
      "step": 29970
    },
    {
      "entropy": 5.4682763576507565,
      "epoch": 2.9631277184657967,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041329135437106866,
      "loss": 5.1861,
      "mean_token_accuracy": 0.20350570380687713,
      "num_tokens": 68658739.0,
      "step": 29975
    },
    {
      "entropy": 5.414107608795166,
      "epoch": 2.9636219849742984,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041326351533517604,
      "loss": 4.978,
      "mean_token_accuracy": 0.22001911848783492,
      "num_tokens": 68670317.0,
      "step": 29980
    },
    {
      "entropy": 5.420254755020141,
      "epoch": 2.9641162514827997,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004132356728979794,
      "loss": 4.9938,
      "mean_token_accuracy": 0.22395044714212417,
      "num_tokens": 68681584.0,
      "step": 29985
    },
    {
      "entropy": 5.354911613464355,
      "epoch": 2.964610517991301,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00041320782706016375,
      "loss": 4.9501,
      "mean_token_accuracy": 0.2232272669672966,
      "num_tokens": 68694296.0,
      "step": 29990
    },
    {
      "entropy": 5.3138494968414305,
      "epoch": 2.965104784499802,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004131799778224142,
      "loss": 4.8981,
      "mean_token_accuracy": 0.2237188771367073,
      "num_tokens": 68705281.0,
      "step": 29995
    },
    {
      "entropy": 5.365420198440551,
      "epoch": 2.9655990510083035,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041315212518541565,
      "loss": 4.8691,
      "mean_token_accuracy": 0.22169061601161957,
      "num_tokens": 68716146.0,
      "step": 30000
    },
    {
      "epoch": 2.9655990510083035,
      "eval_entropy": 5.2244182545613835,
      "eval_loss": 5.084600925445557,
      "eval_mean_token_accuracy": 0.21783759948622256,
      "eval_num_tokens": 68716146.0,
      "eval_runtime": 26.5243,
      "eval_samples_per_second": 1225.783,
      "eval_steps_per_second": 153.256,
      "step": 30000
    },
    {
      "entropy": 5.374813413619995,
      "epoch": 2.966093317516805,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004131242691498534,
      "loss": 5.0081,
      "mean_token_accuracy": 0.21937230080366135,
      "num_tokens": 68728812.0,
      "step": 30005
    },
    {
      "entropy": 5.398124265670776,
      "epoch": 2.9665875840253064,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041309640971641276,
      "loss": 5.0504,
      "mean_token_accuracy": 0.2035959243774414,
      "num_tokens": 68741283.0,
      "step": 30010
    },
    {
      "entropy": 5.433569812774659,
      "epoch": 2.9670818505338077,
      "grad_norm": 1.09375,
      "learning_rate": 0.000413068546885779,
      "loss": 4.9419,
      "mean_token_accuracy": 0.2216383546590805,
      "num_tokens": 68752701.0,
      "step": 30015
    },
    {
      "entropy": 5.339234113693237,
      "epoch": 2.9675761170423094,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004130406806586375,
      "loss": 4.9358,
      "mean_token_accuracy": 0.22480463236570358,
      "num_tokens": 68763133.0,
      "step": 30020
    },
    {
      "entropy": 5.34029369354248,
      "epoch": 2.9680703835508107,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000413012811035674,
      "loss": 5.0004,
      "mean_token_accuracy": 0.21680634766817092,
      "num_tokens": 68774692.0,
      "step": 30025
    },
    {
      "entropy": 5.447988653182984,
      "epoch": 2.968564650059312,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004129849380175738,
      "loss": 5.0195,
      "mean_token_accuracy": 0.20923581421375276,
      "num_tokens": 68786233.0,
      "step": 30030
    },
    {
      "entropy": 5.386737871170044,
      "epoch": 2.969058916567813,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004129570616050227,
      "loss": 5.0184,
      "mean_token_accuracy": 0.20901082456111908,
      "num_tokens": 68799674.0,
      "step": 30035
    },
    {
      "entropy": 5.428809309005738,
      "epoch": 2.969553183076315,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041292918179870665,
      "loss": 5.0359,
      "mean_token_accuracy": 0.21005220115184783,
      "num_tokens": 68811163.0,
      "step": 30040
    },
    {
      "entropy": 5.478037738800049,
      "epoch": 2.970047449584816,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004129012985993112,
      "loss": 5.0011,
      "mean_token_accuracy": 0.211514151096344,
      "num_tokens": 68822781.0,
      "step": 30045
    },
    {
      "entropy": 5.418920373916626,
      "epoch": 2.9705417160933174,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041287341200752235,
      "loss": 5.0359,
      "mean_token_accuracy": 0.21865557432174682,
      "num_tokens": 68834908.0,
      "step": 30050
    },
    {
      "entropy": 5.336770915985108,
      "epoch": 2.971035982601819,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041284552202402634,
      "loss": 4.8938,
      "mean_token_accuracy": 0.2242880642414093,
      "num_tokens": 68844253.0,
      "step": 30055
    },
    {
      "entropy": 5.394455051422119,
      "epoch": 2.9715302491103204,
      "grad_norm": 1.09375,
      "learning_rate": 0.000412817628649509,
      "loss": 4.9613,
      "mean_token_accuracy": 0.21909925937652588,
      "num_tokens": 68856132.0,
      "step": 30060
    },
    {
      "entropy": 5.431913423538208,
      "epoch": 2.9720245156188216,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041278973188465655,
      "loss": 5.0175,
      "mean_token_accuracy": 0.21180320531129837,
      "num_tokens": 68867536.0,
      "step": 30065
    },
    {
      "entropy": 5.338440608978272,
      "epoch": 2.972518782127323,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004127618317301555,
      "loss": 5.0174,
      "mean_token_accuracy": 0.21350359320640563,
      "num_tokens": 68880544.0,
      "step": 30070
    },
    {
      "entropy": 5.437804174423218,
      "epoch": 2.973013048635824,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041273392818669193,
      "loss": 5.1065,
      "mean_token_accuracy": 0.21149717569351195,
      "num_tokens": 68890755.0,
      "step": 30075
    },
    {
      "entropy": 5.527105283737183,
      "epoch": 2.973507315144326,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041270602125495227,
      "loss": 5.0946,
      "mean_token_accuracy": 0.19919875562191008,
      "num_tokens": 68902280.0,
      "step": 30080
    },
    {
      "entropy": 5.500098180770874,
      "epoch": 2.974001581652827,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041267811093562325,
      "loss": 5.0941,
      "mean_token_accuracy": 0.21007563024759293,
      "num_tokens": 68913674.0,
      "step": 30085
    },
    {
      "entropy": 5.484152030944824,
      "epoch": 2.974495848161329,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041265019722939127,
      "loss": 5.0329,
      "mean_token_accuracy": 0.2125246748328209,
      "num_tokens": 68925129.0,
      "step": 30090
    },
    {
      "entropy": 5.382931661605835,
      "epoch": 2.97499011466983,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041262228013694307,
      "loss": 4.9408,
      "mean_token_accuracy": 0.21873839199543,
      "num_tokens": 68935157.0,
      "step": 30095
    },
    {
      "entropy": 5.386719608306885,
      "epoch": 2.9754843811783314,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041259435965896555,
      "loss": 4.9996,
      "mean_token_accuracy": 0.2097771406173706,
      "num_tokens": 68946752.0,
      "step": 30100
    },
    {
      "entropy": 5.447910165786743,
      "epoch": 2.9759786476868326,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004125664357961453,
      "loss": 5.0005,
      "mean_token_accuracy": 0.2104752317070961,
      "num_tokens": 68958078.0,
      "step": 30105
    },
    {
      "entropy": 5.484065532684326,
      "epoch": 2.976472914195334,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041253850854916946,
      "loss": 5.0522,
      "mean_token_accuracy": 0.20692237764596938,
      "num_tokens": 68970384.0,
      "step": 30110
    },
    {
      "entropy": 5.409961891174317,
      "epoch": 2.9769671807038356,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00041251057791872497,
      "loss": 4.9962,
      "mean_token_accuracy": 0.2119045689702034,
      "num_tokens": 68983209.0,
      "step": 30115
    },
    {
      "entropy": 5.517865610122681,
      "epoch": 2.977461447212337,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041248264390549884,
      "loss": 5.135,
      "mean_token_accuracy": 0.20450677573680878,
      "num_tokens": 68994710.0,
      "step": 30120
    },
    {
      "entropy": 5.441586971282959,
      "epoch": 2.977955713720838,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041245470651017846,
      "loss": 5.0686,
      "mean_token_accuracy": 0.2117442160844803,
      "num_tokens": 69005457.0,
      "step": 30125
    },
    {
      "entropy": 5.43292818069458,
      "epoch": 2.97844998022934,
      "grad_norm": 0.953125,
      "learning_rate": 0.00041242676573345096,
      "loss": 4.9552,
      "mean_token_accuracy": 0.2197311520576477,
      "num_tokens": 69018379.0,
      "step": 30130
    },
    {
      "entropy": 5.499438238143921,
      "epoch": 2.978944246737841,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004123988215760036,
      "loss": 5.0582,
      "mean_token_accuracy": 0.20787317603826522,
      "num_tokens": 69029608.0,
      "step": 30135
    },
    {
      "entropy": 5.388748836517334,
      "epoch": 2.9794385132463423,
      "grad_norm": 1.078125,
      "learning_rate": 0.000412370874038524,
      "loss": 5.0204,
      "mean_token_accuracy": 0.21486295908689498,
      "num_tokens": 69040341.0,
      "step": 30140
    },
    {
      "entropy": 5.397605943679809,
      "epoch": 2.9799327797548436,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041234292312169964,
      "loss": 5.0725,
      "mean_token_accuracy": 0.20575768798589705,
      "num_tokens": 69050814.0,
      "step": 30145
    },
    {
      "entropy": 5.532184886932373,
      "epoch": 2.9804270462633453,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004123149688262179,
      "loss": 5.1236,
      "mean_token_accuracy": 0.20521850287914276,
      "num_tokens": 69063087.0,
      "step": 30150
    },
    {
      "entropy": 5.524083662033081,
      "epoch": 2.9809213127718466,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004122870111527668,
      "loss": 5.0957,
      "mean_token_accuracy": 0.20572248995304107,
      "num_tokens": 69075342.0,
      "step": 30155
    },
    {
      "entropy": 5.508407354354858,
      "epoch": 2.981415579280348,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041225905010203385,
      "loss": 5.0773,
      "mean_token_accuracy": 0.205494287610054,
      "num_tokens": 69087616.0,
      "step": 30160
    },
    {
      "entropy": 5.508612108230591,
      "epoch": 2.9819098457888495,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041223108567470696,
      "loss": 5.0288,
      "mean_token_accuracy": 0.20066885501146317,
      "num_tokens": 69099816.0,
      "step": 30165
    },
    {
      "entropy": 5.345908784866333,
      "epoch": 2.982404112297351,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041220311787147407,
      "loss": 4.8819,
      "mean_token_accuracy": 0.2250371754169464,
      "num_tokens": 69111806.0,
      "step": 30170
    },
    {
      "entropy": 5.364367914199829,
      "epoch": 2.982898378805852,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004121751466930232,
      "loss": 5.0271,
      "mean_token_accuracy": 0.21128880828619004,
      "num_tokens": 69123479.0,
      "step": 30175
    },
    {
      "entropy": 5.5502983093261715,
      "epoch": 2.9833926453143533,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004121471721400424,
      "loss": 5.1208,
      "mean_token_accuracy": 0.20105748772621154,
      "num_tokens": 69134541.0,
      "step": 30180
    },
    {
      "entropy": 5.449119997024536,
      "epoch": 2.983886911822855,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041211919421322,
      "loss": 4.9867,
      "mean_token_accuracy": 0.21370634883642198,
      "num_tokens": 69144765.0,
      "step": 30185
    },
    {
      "entropy": 5.478382730484009,
      "epoch": 2.9843811783313563,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004120912129132441,
      "loss": 5.0276,
      "mean_token_accuracy": 0.20873027443885803,
      "num_tokens": 69156761.0,
      "step": 30190
    },
    {
      "entropy": 5.399588775634766,
      "epoch": 2.9848754448398576,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004120632282408031,
      "loss": 4.9931,
      "mean_token_accuracy": 0.21574910283088683,
      "num_tokens": 69167968.0,
      "step": 30195
    },
    {
      "entropy": 5.461322355270386,
      "epoch": 2.9853697113483593,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004120352401965854,
      "loss": 5.0288,
      "mean_token_accuracy": 0.22156336456537246,
      "num_tokens": 69178356.0,
      "step": 30200
    },
    {
      "entropy": 5.44198784828186,
      "epoch": 2.9858639778568605,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041200724878127953,
      "loss": 4.9395,
      "mean_token_accuracy": 0.22379438430070878,
      "num_tokens": 69188777.0,
      "step": 30205
    },
    {
      "entropy": 5.42715744972229,
      "epoch": 2.986358244365362,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041197925399557407,
      "loss": 4.9515,
      "mean_token_accuracy": 0.21258749365806578,
      "num_tokens": 69199166.0,
      "step": 30210
    },
    {
      "entropy": 5.368791484832764,
      "epoch": 2.986852510873863,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004119512558401578,
      "loss": 4.969,
      "mean_token_accuracy": 0.21709192246198655,
      "num_tokens": 69211062.0,
      "step": 30215
    },
    {
      "entropy": 5.487198972702027,
      "epoch": 2.9873467773823643,
      "grad_norm": 1.125,
      "learning_rate": 0.00041192325431571933,
      "loss": 5.0695,
      "mean_token_accuracy": 0.21078742146492005,
      "num_tokens": 69222263.0,
      "step": 30220
    },
    {
      "entropy": 5.458101272583008,
      "epoch": 2.987841043890866,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041189524942294753,
      "loss": 5.0663,
      "mean_token_accuracy": 0.2123659998178482,
      "num_tokens": 69234888.0,
      "step": 30225
    },
    {
      "entropy": 5.47477388381958,
      "epoch": 2.9883353103993673,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004118672411625314,
      "loss": 5.0614,
      "mean_token_accuracy": 0.2158815085887909,
      "num_tokens": 69244618.0,
      "step": 30230
    },
    {
      "entropy": 5.4674101829528805,
      "epoch": 2.988829576907869,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041183922953515987,
      "loss": 4.9613,
      "mean_token_accuracy": 0.21999464482069014,
      "num_tokens": 69254687.0,
      "step": 30235
    },
    {
      "entropy": 5.407430362701416,
      "epoch": 2.9893238434163703,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000411811214541522,
      "loss": 5.076,
      "mean_token_accuracy": 0.20625722855329515,
      "num_tokens": 69267029.0,
      "step": 30240
    },
    {
      "entropy": 5.461917400360107,
      "epoch": 2.9898181099248715,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004117831961823072,
      "loss": 5.0344,
      "mean_token_accuracy": 0.21315219551324843,
      "num_tokens": 69278001.0,
      "step": 30245
    },
    {
      "entropy": 5.425647115707397,
      "epoch": 2.9903123764333728,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004117551744582045,
      "loss": 4.9945,
      "mean_token_accuracy": 0.2177058830857277,
      "num_tokens": 69290591.0,
      "step": 30250
    },
    {
      "entropy": 5.461457061767578,
      "epoch": 2.990806642941874,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004117271493699032,
      "loss": 5.1132,
      "mean_token_accuracy": 0.19877080321311952,
      "num_tokens": 69301977.0,
      "step": 30255
    },
    {
      "entropy": 5.398529958724976,
      "epoch": 2.9913009094503757,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041169912091809294,
      "loss": 5.0178,
      "mean_token_accuracy": 0.21599252820014953,
      "num_tokens": 69313457.0,
      "step": 30260
    },
    {
      "entropy": 5.390436410903931,
      "epoch": 2.991795175958877,
      "grad_norm": 1.203125,
      "learning_rate": 0.000411671089103463,
      "loss": 5.0043,
      "mean_token_accuracy": 0.22034430950880052,
      "num_tokens": 69324860.0,
      "step": 30265
    },
    {
      "entropy": 5.482173252105713,
      "epoch": 2.9922894424673783,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004116430539267031,
      "loss": 5.0296,
      "mean_token_accuracy": 0.20768877267837524,
      "num_tokens": 69335609.0,
      "step": 30270
    },
    {
      "entropy": 5.471250343322754,
      "epoch": 2.99278370897588,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041161501538850286,
      "loss": 5.0855,
      "mean_token_accuracy": 0.2077333837747574,
      "num_tokens": 69346952.0,
      "step": 30275
    },
    {
      "entropy": 5.365226888656617,
      "epoch": 2.9932779754843812,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000411586973489552,
      "loss": 4.9269,
      "mean_token_accuracy": 0.22048699110746384,
      "num_tokens": 69357875.0,
      "step": 30280
    },
    {
      "entropy": 5.390284776687622,
      "epoch": 2.9937722419928825,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041155892823054046,
      "loss": 5.0124,
      "mean_token_accuracy": 0.21204376220703125,
      "num_tokens": 69370371.0,
      "step": 30285
    },
    {
      "entropy": 5.474743509292603,
      "epoch": 2.9942665085013838,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004115308796121582,
      "loss": 5.0906,
      "mean_token_accuracy": 0.20433961153030394,
      "num_tokens": 69381567.0,
      "step": 30290
    },
    {
      "entropy": 5.4392476081848145,
      "epoch": 2.9947607750098855,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000411502827635095,
      "loss": 5.0011,
      "mean_token_accuracy": 0.20772162079811096,
      "num_tokens": 69393559.0,
      "step": 30295
    },
    {
      "entropy": 5.451076221466065,
      "epoch": 2.9952550415183867,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041147477230004103,
      "loss": 5.0336,
      "mean_token_accuracy": 0.21130805909633638,
      "num_tokens": 69404727.0,
      "step": 30300
    },
    {
      "entropy": 5.416012811660766,
      "epoch": 2.995749308026888,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041144671360768655,
      "loss": 4.9729,
      "mean_token_accuracy": 0.2153286412358284,
      "num_tokens": 69414919.0,
      "step": 30305
    },
    {
      "entropy": 5.424667596817017,
      "epoch": 2.9962435745353897,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004114186515587218,
      "loss": 4.9847,
      "mean_token_accuracy": 0.22588408440351487,
      "num_tokens": 69426362.0,
      "step": 30310
    },
    {
      "entropy": 5.40551643371582,
      "epoch": 2.996737841043891,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041139058615383703,
      "loss": 5.0441,
      "mean_token_accuracy": 0.20775208324193956,
      "num_tokens": 69438641.0,
      "step": 30315
    },
    {
      "entropy": 5.412407732009887,
      "epoch": 2.997232107552392,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004113625173937226,
      "loss": 5.0183,
      "mean_token_accuracy": 0.21528330445289612,
      "num_tokens": 69449237.0,
      "step": 30320
    },
    {
      "entropy": 5.455360984802246,
      "epoch": 2.9977263740608935,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004113344452790691,
      "loss": 5.023,
      "mean_token_accuracy": 0.2116187885403633,
      "num_tokens": 69460423.0,
      "step": 30325
    },
    {
      "entropy": 5.382966041564941,
      "epoch": 2.9982206405693947,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041130636981056716,
      "loss": 4.9035,
      "mean_token_accuracy": 0.22321608960628508,
      "num_tokens": 69470999.0,
      "step": 30330
    },
    {
      "entropy": 5.506158018112183,
      "epoch": 2.9987149070778965,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004112782909889074,
      "loss": 5.1062,
      "mean_token_accuracy": 0.20230512470006942,
      "num_tokens": 69481517.0,
      "step": 30335
    },
    {
      "entropy": 5.409990024566651,
      "epoch": 2.9992091735863977,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004112502088147804,
      "loss": 5.0697,
      "mean_token_accuracy": 0.21306455731391907,
      "num_tokens": 69493387.0,
      "step": 30340
    },
    {
      "entropy": 5.407167720794678,
      "epoch": 2.9997034400948994,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041122212328887725,
      "loss": 4.9653,
      "mean_token_accuracy": 0.21770255416631698,
      "num_tokens": 69504715.0,
      "step": 30345
    },
    {
      "entropy": 5.470080137252808,
      "epoch": 3.0001977066034007,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041119403441188874,
      "loss": 5.1344,
      "mean_token_accuracy": 0.20098523646593094,
      "num_tokens": 69515558.0,
      "step": 30350
    },
    {
      "entropy": 5.479523229598999,
      "epoch": 3.000691973111902,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004111659421845058,
      "loss": 5.0024,
      "mean_token_accuracy": 0.21217188239097595,
      "num_tokens": 69528195.0,
      "step": 30355
    },
    {
      "entropy": 5.5197916507720945,
      "epoch": 3.001186239620403,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041113784660741954,
      "loss": 4.977,
      "mean_token_accuracy": 0.2103230059146881,
      "num_tokens": 69539565.0,
      "step": 30360
    },
    {
      "entropy": 5.337203741073608,
      "epoch": 3.001680506128905,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004111097476813212,
      "loss": 4.8943,
      "mean_token_accuracy": 0.21753544807434083,
      "num_tokens": 69550827.0,
      "step": 30365
    },
    {
      "entropy": 5.476755571365357,
      "epoch": 3.002174772637406,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041108164540690196,
      "loss": 5.0289,
      "mean_token_accuracy": 0.216603983938694,
      "num_tokens": 69564205.0,
      "step": 30370
    },
    {
      "entropy": 5.486319541931152,
      "epoch": 3.0026690391459074,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041105353978485303,
      "loss": 4.9702,
      "mean_token_accuracy": 0.22316693365573884,
      "num_tokens": 69577353.0,
      "step": 30375
    },
    {
      "entropy": 5.368816471099853,
      "epoch": 3.0031633056544087,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041102543081586603,
      "loss": 4.9209,
      "mean_token_accuracy": 0.22770458608865737,
      "num_tokens": 69588401.0,
      "step": 30380
    },
    {
      "entropy": 5.444299364089966,
      "epoch": 3.0036575721629104,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004109973185006322,
      "loss": 4.9754,
      "mean_token_accuracy": 0.21429192125797272,
      "num_tokens": 69599966.0,
      "step": 30385
    },
    {
      "entropy": 5.438836336135864,
      "epoch": 3.0041518386714117,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041096920283984335,
      "loss": 4.9641,
      "mean_token_accuracy": 0.2159617781639099,
      "num_tokens": 69611264.0,
      "step": 30390
    },
    {
      "entropy": 5.408601760864258,
      "epoch": 3.004646105179913,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000410941083834191,
      "loss": 4.9557,
      "mean_token_accuracy": 0.21699421852827072,
      "num_tokens": 69621638.0,
      "step": 30395
    },
    {
      "entropy": 5.39887752532959,
      "epoch": 3.005140371688414,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004109129614843669,
      "loss": 4.9003,
      "mean_token_accuracy": 0.2281687006354332,
      "num_tokens": 69633236.0,
      "step": 30400
    },
    {
      "entropy": 5.392551565170288,
      "epoch": 3.005634638196916,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004108848357910628,
      "loss": 4.9095,
      "mean_token_accuracy": 0.21926873326301574,
      "num_tokens": 69644672.0,
      "step": 30405
    },
    {
      "entropy": 5.411789083480835,
      "epoch": 3.006128904705417,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00041085670675497065,
      "loss": 4.9462,
      "mean_token_accuracy": 0.21901183873414992,
      "num_tokens": 69655580.0,
      "step": 30410
    },
    {
      "entropy": 5.422664213180542,
      "epoch": 3.0066231712139184,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004108285743767825,
      "loss": 4.9873,
      "mean_token_accuracy": 0.21415219604969024,
      "num_tokens": 69668692.0,
      "step": 30415
    },
    {
      "entropy": 5.482473659515381,
      "epoch": 3.00711743772242,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041080043865719033,
      "loss": 5.0156,
      "mean_token_accuracy": 0.21315995305776597,
      "num_tokens": 69679885.0,
      "step": 30420
    },
    {
      "entropy": 5.429695320129395,
      "epoch": 3.0076117042309214,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004107722995968863,
      "loss": 5.0141,
      "mean_token_accuracy": 0.21658846437931062,
      "num_tokens": 69691395.0,
      "step": 30425
    },
    {
      "entropy": 5.370427513122559,
      "epoch": 3.0081059707394227,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004107441571965627,
      "loss": 4.8535,
      "mean_token_accuracy": 0.21821608245372773,
      "num_tokens": 69701777.0,
      "step": 30430
    },
    {
      "entropy": 5.410035943984985,
      "epoch": 3.008600237247924,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004107160114569117,
      "loss": 4.9403,
      "mean_token_accuracy": 0.22258511334657669,
      "num_tokens": 69712922.0,
      "step": 30435
    },
    {
      "entropy": 5.40771894454956,
      "epoch": 3.0090945037564256,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004106878623786258,
      "loss": 4.9111,
      "mean_token_accuracy": 0.21988439559936523,
      "num_tokens": 69724342.0,
      "step": 30440
    },
    {
      "entropy": 5.351280736923218,
      "epoch": 3.009588770264927,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004106597099623974,
      "loss": 4.904,
      "mean_token_accuracy": 0.22288720458745956,
      "num_tokens": 69735650.0,
      "step": 30445
    },
    {
      "entropy": 5.385583019256591,
      "epoch": 3.010083036773428,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004106315542089191,
      "loss": 4.887,
      "mean_token_accuracy": 0.22392853796482087,
      "num_tokens": 69746745.0,
      "step": 30450
    },
    {
      "entropy": 5.437001132965088,
      "epoch": 3.0105773032819294,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004106033951188835,
      "loss": 4.8734,
      "mean_token_accuracy": 0.21626591235399245,
      "num_tokens": 69759315.0,
      "step": 30455
    },
    {
      "entropy": 5.404405546188355,
      "epoch": 3.011071569790431,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041057523269298336,
      "loss": 4.9352,
      "mean_token_accuracy": 0.21534991562366484,
      "num_tokens": 69769883.0,
      "step": 30460
    },
    {
      "entropy": 5.3852126121521,
      "epoch": 3.0115658362989324,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004105470669319115,
      "loss": 4.9482,
      "mean_token_accuracy": 0.21387778371572494,
      "num_tokens": 69781449.0,
      "step": 30465
    },
    {
      "entropy": 5.314585876464844,
      "epoch": 3.0120601028074336,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004105188978363608,
      "loss": 4.8801,
      "mean_token_accuracy": 0.2294148787856102,
      "num_tokens": 69793952.0,
      "step": 30470
    },
    {
      "entropy": 5.422433185577392,
      "epoch": 3.0125543693159353,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004104907254070241,
      "loss": 4.9342,
      "mean_token_accuracy": 0.2228323370218277,
      "num_tokens": 69805705.0,
      "step": 30475
    },
    {
      "entropy": 5.4462769508361815,
      "epoch": 3.0130486358244366,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004104625496445946,
      "loss": 5.027,
      "mean_token_accuracy": 0.2082013815641403,
      "num_tokens": 69817298.0,
      "step": 30480
    },
    {
      "entropy": 5.472079658508301,
      "epoch": 3.013542902332938,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041043437054976524,
      "loss": 5.0091,
      "mean_token_accuracy": 0.21154625713825226,
      "num_tokens": 69829280.0,
      "step": 30485
    },
    {
      "entropy": 5.35842752456665,
      "epoch": 3.014037168841439,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004104061881232295,
      "loss": 4.8667,
      "mean_token_accuracy": 0.2250947117805481,
      "num_tokens": 69841500.0,
      "step": 30490
    },
    {
      "entropy": 5.404494905471802,
      "epoch": 3.014531435349941,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041037800236568044,
      "loss": 4.9184,
      "mean_token_accuracy": 0.22083764672279357,
      "num_tokens": 69852651.0,
      "step": 30495
    },
    {
      "entropy": 5.398337507247925,
      "epoch": 3.015025701858442,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004103498132778116,
      "loss": 4.942,
      "mean_token_accuracy": 0.2207941621541977,
      "num_tokens": 69864746.0,
      "step": 30500
    },
    {
      "entropy": 5.333256673812866,
      "epoch": 3.0155199683669434,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004103216208603163,
      "loss": 4.8734,
      "mean_token_accuracy": 0.22742423713207244,
      "num_tokens": 69875696.0,
      "step": 30505
    },
    {
      "entropy": 5.409759950637818,
      "epoch": 3.0160142348754446,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004102934251138881,
      "loss": 4.8463,
      "mean_token_accuracy": 0.2300860121846199,
      "num_tokens": 69887363.0,
      "step": 30510
    },
    {
      "entropy": 5.4117772579193115,
      "epoch": 3.0165085013839463,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041026522603922075,
      "loss": 4.9517,
      "mean_token_accuracy": 0.21733166873455048,
      "num_tokens": 69898561.0,
      "step": 30515
    },
    {
      "entropy": 5.404744338989258,
      "epoch": 3.0170027678924476,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004102370236370077,
      "loss": 5.008,
      "mean_token_accuracy": 0.21028438508510588,
      "num_tokens": 69910297.0,
      "step": 30520
    },
    {
      "entropy": 5.416772794723511,
      "epoch": 3.017497034400949,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041020881790794305,
      "loss": 4.9829,
      "mean_token_accuracy": 0.21758916079998017,
      "num_tokens": 69922435.0,
      "step": 30525
    },
    {
      "entropy": 5.383396434783935,
      "epoch": 3.0179913009094506,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00041018060885272047,
      "loss": 4.9189,
      "mean_token_accuracy": 0.22540755420923234,
      "num_tokens": 69935151.0,
      "step": 30530
    },
    {
      "entropy": 5.434050130844116,
      "epoch": 3.018485567417952,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041015239647203386,
      "loss": 4.96,
      "mean_token_accuracy": 0.21290067285299302,
      "num_tokens": 69945915.0,
      "step": 30535
    },
    {
      "entropy": 5.391369152069092,
      "epoch": 3.018979833926453,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041012418076657737,
      "loss": 4.9103,
      "mean_token_accuracy": 0.22301845997571945,
      "num_tokens": 69957016.0,
      "step": 30540
    },
    {
      "entropy": 5.32572112083435,
      "epoch": 3.0194741004349543,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041009596173704506,
      "loss": 4.8365,
      "mean_token_accuracy": 0.22583547979593277,
      "num_tokens": 69967863.0,
      "step": 30545
    },
    {
      "entropy": 5.44273009300232,
      "epoch": 3.019968366943456,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041006773938413114,
      "loss": 5.0109,
      "mean_token_accuracy": 0.20767912417650222,
      "num_tokens": 69978901.0,
      "step": 30550
    },
    {
      "entropy": 5.34404821395874,
      "epoch": 3.0204626334519573,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004100395137085299,
      "loss": 4.9098,
      "mean_token_accuracy": 0.2203651860356331,
      "num_tokens": 69990829.0,
      "step": 30555
    },
    {
      "entropy": 5.426419591903686,
      "epoch": 3.0209568999604586,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004100112847109357,
      "loss": 4.896,
      "mean_token_accuracy": 0.22002136409282685,
      "num_tokens": 70002130.0,
      "step": 30560
    },
    {
      "entropy": 5.429311513900757,
      "epoch": 3.0214511664689603,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004099830523920429,
      "loss": 4.9508,
      "mean_token_accuracy": 0.2205308571457863,
      "num_tokens": 70012913.0,
      "step": 30565
    },
    {
      "entropy": 5.405665874481201,
      "epoch": 3.0219454329774615,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040995481675254616,
      "loss": 4.9496,
      "mean_token_accuracy": 0.21888624578714372,
      "num_tokens": 70024689.0,
      "step": 30570
    },
    {
      "entropy": 5.393888711929321,
      "epoch": 3.022439699485963,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004099265777931398,
      "loss": 4.9389,
      "mean_token_accuracy": 0.21826994568109512,
      "num_tokens": 70036153.0,
      "step": 30575
    },
    {
      "entropy": 5.365143585205078,
      "epoch": 3.022933965994464,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040989833551451885,
      "loss": 4.9663,
      "mean_token_accuracy": 0.2144753485918045,
      "num_tokens": 70049190.0,
      "step": 30580
    },
    {
      "entropy": 5.399091529846191,
      "epoch": 3.0234282325029658,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040987008991737786,
      "loss": 4.8693,
      "mean_token_accuracy": 0.21980566680431365,
      "num_tokens": 70061391.0,
      "step": 30585
    },
    {
      "entropy": 5.329920721054077,
      "epoch": 3.023922499011467,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004098418410024118,
      "loss": 4.8351,
      "mean_token_accuracy": 0.22599011808633804,
      "num_tokens": 70072811.0,
      "step": 30590
    },
    {
      "entropy": 5.372668600082397,
      "epoch": 3.0244167655199683,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040981358877031546,
      "loss": 4.8567,
      "mean_token_accuracy": 0.22677309811115265,
      "num_tokens": 70083776.0,
      "step": 30595
    },
    {
      "entropy": 5.3922264099121096,
      "epoch": 3.0249110320284696,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004097853332217839,
      "loss": 4.9897,
      "mean_token_accuracy": 0.21422107815742492,
      "num_tokens": 70095720.0,
      "step": 30600
    },
    {
      "entropy": 5.41444239616394,
      "epoch": 3.0254052985369713,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004097570743575123,
      "loss": 4.9744,
      "mean_token_accuracy": 0.2142170250415802,
      "num_tokens": 70106315.0,
      "step": 30605
    },
    {
      "entropy": 5.4558929920196535,
      "epoch": 3.0258995650454725,
      "grad_norm": 1.0,
      "learning_rate": 0.00040972881217819573,
      "loss": 4.9907,
      "mean_token_accuracy": 0.21169532984495162,
      "num_tokens": 70118018.0,
      "step": 30610
    },
    {
      "entropy": 5.415301704406739,
      "epoch": 3.026393831553974,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004097005466845295,
      "loss": 4.9226,
      "mean_token_accuracy": 0.22231138944625856,
      "num_tokens": 70130543.0,
      "step": 30615
    },
    {
      "entropy": 5.3317783832550045,
      "epoch": 3.0268880980624755,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004096722778772089,
      "loss": 4.8639,
      "mean_token_accuracy": 0.22819124311208724,
      "num_tokens": 70142582.0,
      "step": 30620
    },
    {
      "entropy": 5.436104011535645,
      "epoch": 3.0273823645709768,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00040964400575692946,
      "loss": 4.9807,
      "mean_token_accuracy": 0.21840066760778426,
      "num_tokens": 70155002.0,
      "step": 30625
    },
    {
      "entropy": 5.478664445877075,
      "epoch": 3.027876631079478,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040961573032438647,
      "loss": 4.9619,
      "mean_token_accuracy": 0.21520168632268905,
      "num_tokens": 70167354.0,
      "step": 30630
    },
    {
      "entropy": 5.397228240966797,
      "epoch": 3.0283708975879793,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004095874515802757,
      "loss": 5.0005,
      "mean_token_accuracy": 0.2179683968424797,
      "num_tokens": 70178561.0,
      "step": 30635
    },
    {
      "entropy": 5.412921810150147,
      "epoch": 3.028865164096481,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00040955916952529277,
      "loss": 4.9377,
      "mean_token_accuracy": 0.22295175939798356,
      "num_tokens": 70189954.0,
      "step": 30640
    },
    {
      "entropy": 5.437738561630249,
      "epoch": 3.0293594306049823,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040953088416013325,
      "loss": 5.0096,
      "mean_token_accuracy": 0.20835305899381637,
      "num_tokens": 70200595.0,
      "step": 30645
    },
    {
      "entropy": 5.407878112792969,
      "epoch": 3.0298536971134835,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004095025954854933,
      "loss": 4.9579,
      "mean_token_accuracy": 0.21726207733154296,
      "num_tokens": 70211757.0,
      "step": 30650
    },
    {
      "entropy": 5.257646322250366,
      "epoch": 3.0303479636219848,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004094743035020685,
      "loss": 4.8358,
      "mean_token_accuracy": 0.22766544818878173,
      "num_tokens": 70223404.0,
      "step": 30655
    },
    {
      "entropy": 5.532404994964599,
      "epoch": 3.0308422301304865,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040944600821055506,
      "loss": 5.0552,
      "mean_token_accuracy": 0.2062969148159027,
      "num_tokens": 70234022.0,
      "step": 30660
    },
    {
      "entropy": 5.438463401794434,
      "epoch": 3.0313364966389877,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004094177096116489,
      "loss": 4.958,
      "mean_token_accuracy": 0.21756165474653244,
      "num_tokens": 70245279.0,
      "step": 30665
    },
    {
      "entropy": 5.3319604873657225,
      "epoch": 3.031830763147489,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040938940770604624,
      "loss": 4.9262,
      "mean_token_accuracy": 0.22059249728918076,
      "num_tokens": 70256561.0,
      "step": 30670
    },
    {
      "entropy": 5.355877113342285,
      "epoch": 3.0323250296559907,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00040936110249444323,
      "loss": 4.8562,
      "mean_token_accuracy": 0.22982880622148513,
      "num_tokens": 70268535.0,
      "step": 30675
    },
    {
      "entropy": 5.366777515411377,
      "epoch": 3.032819296164492,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004093327939775364,
      "loss": 4.9119,
      "mean_token_accuracy": 0.23053721934556962,
      "num_tokens": 70279681.0,
      "step": 30680
    },
    {
      "entropy": 5.359801483154297,
      "epoch": 3.0333135626729932,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004093044821560219,
      "loss": 4.8859,
      "mean_token_accuracy": 0.22680379450321198,
      "num_tokens": 70290796.0,
      "step": 30685
    },
    {
      "entropy": 5.366371917724609,
      "epoch": 3.0338078291814945,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040927616703059636,
      "loss": 4.8738,
      "mean_token_accuracy": 0.22066937386989594,
      "num_tokens": 70302254.0,
      "step": 30690
    },
    {
      "entropy": 5.30592098236084,
      "epoch": 3.034302095689996,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004092478486019562,
      "loss": 4.842,
      "mean_token_accuracy": 0.23007274121046067,
      "num_tokens": 70313022.0,
      "step": 30695
    },
    {
      "entropy": 5.413453483581543,
      "epoch": 3.0347963621984975,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040921952687079824,
      "loss": 4.9298,
      "mean_token_accuracy": 0.21921290159225465,
      "num_tokens": 70325356.0,
      "step": 30700
    },
    {
      "entropy": 5.4563432216644285,
      "epoch": 3.0352906287069987,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000409191201837819,
      "loss": 5.0157,
      "mean_token_accuracy": 0.21103538423776627,
      "num_tokens": 70336622.0,
      "step": 30705
    },
    {
      "entropy": 5.434011030197143,
      "epoch": 3.0357848952155,
      "grad_norm": 1.125,
      "learning_rate": 0.00040916287350371546,
      "loss": 4.9493,
      "mean_token_accuracy": 0.21375397741794586,
      "num_tokens": 70347799.0,
      "step": 30710
    },
    {
      "entropy": 5.371381378173828,
      "epoch": 3.0362791617240017,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040913454186918436,
      "loss": 4.9203,
      "mean_token_accuracy": 0.22277666628360748,
      "num_tokens": 70358918.0,
      "step": 30715
    },
    {
      "entropy": 5.416863536834716,
      "epoch": 3.036773428232503,
      "grad_norm": 1.0,
      "learning_rate": 0.00040910620693492277,
      "loss": 5.04,
      "mean_token_accuracy": 0.20856893807649612,
      "num_tokens": 70371853.0,
      "step": 30720
    },
    {
      "entropy": 5.435876703262329,
      "epoch": 3.037267694741004,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004090778687016276,
      "loss": 4.9792,
      "mean_token_accuracy": 0.21008629053831102,
      "num_tokens": 70383696.0,
      "step": 30725
    },
    {
      "entropy": 5.360678434371948,
      "epoch": 3.037761961249506,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00040904952716999617,
      "loss": 4.8971,
      "mean_token_accuracy": 0.21712495386600494,
      "num_tokens": 70395651.0,
      "step": 30730
    },
    {
      "entropy": 5.424138832092285,
      "epoch": 3.038256227758007,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004090211823407255,
      "loss": 4.9624,
      "mean_token_accuracy": 0.21351974308490754,
      "num_tokens": 70408621.0,
      "step": 30735
    },
    {
      "entropy": 5.306193542480469,
      "epoch": 3.0387504942665085,
      "grad_norm": 0.984375,
      "learning_rate": 0.000408992834214513,
      "loss": 4.8243,
      "mean_token_accuracy": 0.22526150792837143,
      "num_tokens": 70419627.0,
      "step": 30740
    },
    {
      "entropy": 5.381680536270141,
      "epoch": 3.0392447607750097,
      "grad_norm": 1.0,
      "learning_rate": 0.0004089644827920559,
      "loss": 4.9145,
      "mean_token_accuracy": 0.22642378360033036,
      "num_tokens": 70431795.0,
      "step": 30745
    },
    {
      "entropy": 5.372473192214966,
      "epoch": 3.0397390272835114,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040893612807405184,
      "loss": 4.8781,
      "mean_token_accuracy": 0.23239432871341706,
      "num_tokens": 70444086.0,
      "step": 30750
    },
    {
      "entropy": 5.422641086578369,
      "epoch": 3.0402332937920127,
      "grad_norm": 1.0,
      "learning_rate": 0.0004089077700611982,
      "loss": 4.9372,
      "mean_token_accuracy": 0.22059000432491302,
      "num_tokens": 70454844.0,
      "step": 30755
    },
    {
      "entropy": 5.439218473434448,
      "epoch": 3.040727560300514,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040887940875419267,
      "loss": 5.0055,
      "mean_token_accuracy": 0.21494870334863664,
      "num_tokens": 70466003.0,
      "step": 30760
    },
    {
      "entropy": 5.461134624481201,
      "epoch": 3.041221826809015,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004088510441537329,
      "loss": 4.9111,
      "mean_token_accuracy": 0.2241834282875061,
      "num_tokens": 70476586.0,
      "step": 30765
    },
    {
      "entropy": 5.41762228012085,
      "epoch": 3.041716093317517,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040882267626051664,
      "loss": 5.0205,
      "mean_token_accuracy": 0.2086621031165123,
      "num_tokens": 70488561.0,
      "step": 30770
    },
    {
      "entropy": 5.399963903427124,
      "epoch": 3.042210359826018,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004087943050752419,
      "loss": 4.9468,
      "mean_token_accuracy": 0.20916391015052796,
      "num_tokens": 70499454.0,
      "step": 30775
    },
    {
      "entropy": 5.466132020950317,
      "epoch": 3.0427046263345194,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004087659305986064,
      "loss": 4.9669,
      "mean_token_accuracy": 0.21167234033346177,
      "num_tokens": 70510880.0,
      "step": 30780
    },
    {
      "entropy": 5.411253929138184,
      "epoch": 3.043198892843021,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00040873755283130824,
      "loss": 4.9514,
      "mean_token_accuracy": 0.21867942959070205,
      "num_tokens": 70523199.0,
      "step": 30785
    },
    {
      "entropy": 5.3632300853729244,
      "epoch": 3.0436931593515224,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004087091717740456,
      "loss": 4.895,
      "mean_token_accuracy": 0.21445599794387818,
      "num_tokens": 70535124.0,
      "step": 30790
    },
    {
      "entropy": 5.383294916152954,
      "epoch": 3.0441874258600237,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040868078742751655,
      "loss": 4.9058,
      "mean_token_accuracy": 0.22451711744070052,
      "num_tokens": 70547543.0,
      "step": 30795
    },
    {
      "entropy": 5.396477365493775,
      "epoch": 3.044681692368525,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00040865239979241946,
      "loss": 4.906,
      "mean_token_accuracy": 0.22317237854003907,
      "num_tokens": 70559788.0,
      "step": 30800
    },
    {
      "entropy": 5.3785358428955075,
      "epoch": 3.0451759588770266,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004086240088694525,
      "loss": 4.9204,
      "mean_token_accuracy": 0.22607837617397308,
      "num_tokens": 70570303.0,
      "step": 30805
    },
    {
      "entropy": 5.312200355529785,
      "epoch": 3.045670225385528,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040859561465931426,
      "loss": 4.8663,
      "mean_token_accuracy": 0.22607066184282304,
      "num_tokens": 70580924.0,
      "step": 30810
    },
    {
      "entropy": 5.430265235900879,
      "epoch": 3.046164491894029,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040856721716270316,
      "loss": 4.9288,
      "mean_token_accuracy": 0.22272013127803802,
      "num_tokens": 70592557.0,
      "step": 30815
    },
    {
      "entropy": 5.279453992843628,
      "epoch": 3.046658758402531,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004085388163803178,
      "loss": 4.8054,
      "mean_token_accuracy": 0.2347233086824417,
      "num_tokens": 70604374.0,
      "step": 30820
    },
    {
      "entropy": 5.39760103225708,
      "epoch": 3.047153024911032,
      "grad_norm": 1.125,
      "learning_rate": 0.0004085104123128568,
      "loss": 4.9426,
      "mean_token_accuracy": 0.2144983798265457,
      "num_tokens": 70616380.0,
      "step": 30825
    },
    {
      "entropy": 5.437837171554565,
      "epoch": 3.0476472914195334,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040848200496101895,
      "loss": 4.9902,
      "mean_token_accuracy": 0.21941965967416763,
      "num_tokens": 70628297.0,
      "step": 30830
    },
    {
      "entropy": 5.498921918869018,
      "epoch": 3.0481415579280347,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004084535943255031,
      "loss": 4.963,
      "mean_token_accuracy": 0.21315976828336716,
      "num_tokens": 70639815.0,
      "step": 30835
    },
    {
      "entropy": 5.413691711425781,
      "epoch": 3.0486358244365364,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040842518040700815,
      "loss": 4.9753,
      "mean_token_accuracy": 0.21351510137319565,
      "num_tokens": 70651283.0,
      "step": 30840
    },
    {
      "entropy": 5.392809772491455,
      "epoch": 3.0491300909450376,
      "grad_norm": 1.125,
      "learning_rate": 0.000408396763206233,
      "loss": 4.8942,
      "mean_token_accuracy": 0.2247121050953865,
      "num_tokens": 70662465.0,
      "step": 30845
    },
    {
      "entropy": 5.4868518829345705,
      "epoch": 3.049624357453539,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004083683427238768,
      "loss": 5.0363,
      "mean_token_accuracy": 0.21093931198120117,
      "num_tokens": 70674301.0,
      "step": 30850
    },
    {
      "entropy": 5.390825653076172,
      "epoch": 3.05011862396204,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004083399189606387,
      "loss": 4.9514,
      "mean_token_accuracy": 0.22158412039279937,
      "num_tokens": 70685423.0,
      "step": 30855
    },
    {
      "entropy": 5.416312408447266,
      "epoch": 3.050612890470542,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004083114919172179,
      "loss": 4.8487,
      "mean_token_accuracy": 0.22390309125185012,
      "num_tokens": 70697043.0,
      "step": 30860
    },
    {
      "entropy": 5.4309204578399655,
      "epoch": 3.051107156979043,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004082830615943137,
      "loss": 4.9734,
      "mean_token_accuracy": 0.21805118024349213,
      "num_tokens": 70708585.0,
      "step": 30865
    },
    {
      "entropy": 5.330524110794068,
      "epoch": 3.0516014234875444,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004082546279926256,
      "loss": 4.8688,
      "mean_token_accuracy": 0.2197417512536049,
      "num_tokens": 70720703.0,
      "step": 30870
    },
    {
      "entropy": 5.413475751876831,
      "epoch": 3.052095689996046,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040822619111285284,
      "loss": 4.9585,
      "mean_token_accuracy": 0.21656543463468553,
      "num_tokens": 70731146.0,
      "step": 30875
    },
    {
      "entropy": 5.355567407608032,
      "epoch": 3.0525899565045473,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00040819775095569515,
      "loss": 4.9074,
      "mean_token_accuracy": 0.22070125192403794,
      "num_tokens": 70742393.0,
      "step": 30880
    },
    {
      "entropy": 5.379467248916626,
      "epoch": 3.0530842230130486,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004081693075218522,
      "loss": 4.8261,
      "mean_token_accuracy": 0.22518666088581085,
      "num_tokens": 70753419.0,
      "step": 30885
    },
    {
      "entropy": 5.338520956039429,
      "epoch": 3.05357848952155,
      "grad_norm": 0.9375,
      "learning_rate": 0.00040814086081202364,
      "loss": 4.8771,
      "mean_token_accuracy": 0.23138557374477386,
      "num_tokens": 70765445.0,
      "step": 30890
    },
    {
      "entropy": 5.396488857269287,
      "epoch": 3.0540727560300516,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040811241082690914,
      "loss": 4.9309,
      "mean_token_accuracy": 0.2177235007286072,
      "num_tokens": 70776598.0,
      "step": 30895
    },
    {
      "entropy": 5.426545333862305,
      "epoch": 3.054567022538553,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004080839575672087,
      "loss": 4.9552,
      "mean_token_accuracy": 0.21597821563482283,
      "num_tokens": 70787982.0,
      "step": 30900
    },
    {
      "entropy": 5.38010311126709,
      "epoch": 3.055061289047054,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040805550103362225,
      "loss": 4.8911,
      "mean_token_accuracy": 0.22319617867469788,
      "num_tokens": 70799080.0,
      "step": 30905
    },
    {
      "entropy": 5.3277545928955075,
      "epoch": 3.0555555555555554,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00040802704122684994,
      "loss": 4.898,
      "mean_token_accuracy": 0.2252265527844429,
      "num_tokens": 70809831.0,
      "step": 30910
    },
    {
      "entropy": 5.3703197002410885,
      "epoch": 3.056049822064057,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040799857814759167,
      "loss": 4.9964,
      "mean_token_accuracy": 0.21615443229675294,
      "num_tokens": 70821623.0,
      "step": 30915
    },
    {
      "entropy": 5.515080261230469,
      "epoch": 3.0565440885725583,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004079701117965477,
      "loss": 5.0308,
      "mean_token_accuracy": 0.20916198641061784,
      "num_tokens": 70833875.0,
      "step": 30920
    },
    {
      "entropy": 5.366548156738281,
      "epoch": 3.0570383550810596,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004079416421744185,
      "loss": 4.9406,
      "mean_token_accuracy": 0.21883416622877122,
      "num_tokens": 70845287.0,
      "step": 30925
    },
    {
      "entropy": 5.475229024887085,
      "epoch": 3.0575326215895613,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040791316928190423,
      "loss": 4.9492,
      "mean_token_accuracy": 0.21836733520030976,
      "num_tokens": 70856294.0,
      "step": 30930
    },
    {
      "entropy": 5.381893444061279,
      "epoch": 3.0580268880980626,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040788469311970525,
      "loss": 4.9624,
      "mean_token_accuracy": 0.21809504479169844,
      "num_tokens": 70867490.0,
      "step": 30935
    },
    {
      "entropy": 5.3736796379089355,
      "epoch": 3.058521154606564,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040785621368852234,
      "loss": 4.9473,
      "mean_token_accuracy": 0.22061517387628554,
      "num_tokens": 70877991.0,
      "step": 30940
    },
    {
      "entropy": 5.431356954574585,
      "epoch": 3.059015421115065,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040782773098905594,
      "loss": 4.9651,
      "mean_token_accuracy": 0.21861248314380646,
      "num_tokens": 70888439.0,
      "step": 30945
    },
    {
      "entropy": 5.369113779067993,
      "epoch": 3.059509687623567,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004077992450220068,
      "loss": 4.8822,
      "mean_token_accuracy": 0.22857068926095964,
      "num_tokens": 70898331.0,
      "step": 30950
    },
    {
      "entropy": 5.380788660049438,
      "epoch": 3.060003954132068,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004077707557880755,
      "loss": 4.8872,
      "mean_token_accuracy": 0.22493885457515717,
      "num_tokens": 70908221.0,
      "step": 30955
    },
    {
      "entropy": 5.394108057022095,
      "epoch": 3.0604982206405693,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040774226328796315,
      "loss": 4.8639,
      "mean_token_accuracy": 0.22668589800596237,
      "num_tokens": 70919720.0,
      "step": 30960
    },
    {
      "entropy": 5.463345384597778,
      "epoch": 3.0609924871490706,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004077137675223704,
      "loss": 4.9375,
      "mean_token_accuracy": 0.21904116421937941,
      "num_tokens": 70931322.0,
      "step": 30965
    },
    {
      "entropy": 5.366185092926026,
      "epoch": 3.0614867536575723,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004076852684919985,
      "loss": 4.9716,
      "mean_token_accuracy": 0.21896986067295074,
      "num_tokens": 70943124.0,
      "step": 30970
    },
    {
      "entropy": 5.411995077133179,
      "epoch": 3.0619810201660735,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004076567661975484,
      "loss": 4.9034,
      "mean_token_accuracy": 0.22310545593500136,
      "num_tokens": 70954986.0,
      "step": 30975
    },
    {
      "entropy": 5.3533241748809814,
      "epoch": 3.062475286674575,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004076282606397211,
      "loss": 4.8592,
      "mean_token_accuracy": 0.22208243906497954,
      "num_tokens": 70965500.0,
      "step": 30980
    },
    {
      "entropy": 5.361663341522217,
      "epoch": 3.0629695531830765,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004075997518192182,
      "loss": 4.8759,
      "mean_token_accuracy": 0.21764241755008698,
      "num_tokens": 70976717.0,
      "step": 30985
    },
    {
      "entropy": 5.364130306243896,
      "epoch": 3.0634638196915778,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004075712397367408,
      "loss": 4.9403,
      "mean_token_accuracy": 0.2222544103860855,
      "num_tokens": 70987286.0,
      "step": 30990
    },
    {
      "entropy": 5.517619943618774,
      "epoch": 3.063958086200079,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040754272439299025,
      "loss": 5.0224,
      "mean_token_accuracy": 0.20722428411245347,
      "num_tokens": 70998710.0,
      "step": 30995
    },
    {
      "entropy": 5.37889256477356,
      "epoch": 3.0644523527085803,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040751420578866816,
      "loss": 4.9573,
      "mean_token_accuracy": 0.2150046482682228,
      "num_tokens": 71010725.0,
      "step": 31000
    },
    {
      "entropy": 5.406433534622193,
      "epoch": 3.064946619217082,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004074856839244761,
      "loss": 4.9983,
      "mean_token_accuracy": 0.2200597956776619,
      "num_tokens": 71021723.0,
      "step": 31005
    },
    {
      "entropy": 5.397740221023559,
      "epoch": 3.0654408857255833,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004074571588011156,
      "loss": 4.8892,
      "mean_token_accuracy": 0.2221280738711357,
      "num_tokens": 71032744.0,
      "step": 31010
    },
    {
      "entropy": 5.339925718307495,
      "epoch": 3.0659351522340845,
      "grad_norm": 1.125,
      "learning_rate": 0.00040742863041928854,
      "loss": 4.8455,
      "mean_token_accuracy": 0.22274887859821318,
      "num_tokens": 71044296.0,
      "step": 31015
    },
    {
      "entropy": 5.385367155075073,
      "epoch": 3.066429418742586,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040740009877969654,
      "loss": 4.9754,
      "mean_token_accuracy": 0.22018209397792815,
      "num_tokens": 71054636.0,
      "step": 31020
    },
    {
      "entropy": 5.443326568603515,
      "epoch": 3.0669236852510875,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004073715638830415,
      "loss": 4.9921,
      "mean_token_accuracy": 0.22188208252191544,
      "num_tokens": 71067350.0,
      "step": 31025
    },
    {
      "entropy": 5.461763191223144,
      "epoch": 3.0674179517595888,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040734302573002565,
      "loss": 4.9734,
      "mean_token_accuracy": 0.2189631298184395,
      "num_tokens": 71078951.0,
      "step": 31030
    },
    {
      "entropy": 5.351203966140747,
      "epoch": 3.06791221826809,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004073144843213507,
      "loss": 4.8951,
      "mean_token_accuracy": 0.22432831674814224,
      "num_tokens": 71090039.0,
      "step": 31035
    },
    {
      "entropy": 5.2922282218933105,
      "epoch": 3.0684064847765917,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040728593965771895,
      "loss": 4.8832,
      "mean_token_accuracy": 0.22834322601556778,
      "num_tokens": 71101203.0,
      "step": 31040
    },
    {
      "entropy": 5.361834383010864,
      "epoch": 3.068900751285093,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040725739173983253,
      "loss": 4.8568,
      "mean_token_accuracy": 0.22096928656101228,
      "num_tokens": 71112720.0,
      "step": 31045
    },
    {
      "entropy": 5.462894058227539,
      "epoch": 3.0693950177935942,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004072288405683938,
      "loss": 5.0034,
      "mean_token_accuracy": 0.21408717036247255,
      "num_tokens": 71123126.0,
      "step": 31050
    },
    {
      "entropy": 5.387087774276734,
      "epoch": 3.0698892843020955,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00040720028614410497,
      "loss": 4.8691,
      "mean_token_accuracy": 0.21910379976034164,
      "num_tokens": 71135555.0,
      "step": 31055
    },
    {
      "entropy": 5.378819370269776,
      "epoch": 3.070383550810597,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004071717284676687,
      "loss": 4.8809,
      "mean_token_accuracy": 0.2264954060316086,
      "num_tokens": 71147304.0,
      "step": 31060
    },
    {
      "entropy": 5.393623876571655,
      "epoch": 3.0708778173190985,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004071431675397874,
      "loss": 4.9361,
      "mean_token_accuracy": 0.21454462707042693,
      "num_tokens": 71157564.0,
      "step": 31065
    },
    {
      "entropy": 5.2822497367858885,
      "epoch": 3.0713720838275997,
      "grad_norm": 1.0,
      "learning_rate": 0.00040711460336116364,
      "loss": 4.7838,
      "mean_token_accuracy": 0.23773422986268997,
      "num_tokens": 71168994.0,
      "step": 31070
    },
    {
      "entropy": 5.422020387649536,
      "epoch": 3.0718663503361014,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040708603593250016,
      "loss": 4.9406,
      "mean_token_accuracy": 0.22453244030475616,
      "num_tokens": 71180736.0,
      "step": 31075
    },
    {
      "entropy": 5.3104777336120605,
      "epoch": 3.0723606168446027,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040705746525449964,
      "loss": 4.8672,
      "mean_token_accuracy": 0.22726356089115143,
      "num_tokens": 71193187.0,
      "step": 31080
    },
    {
      "entropy": 5.355903625488281,
      "epoch": 3.072854883353104,
      "grad_norm": 1.0,
      "learning_rate": 0.00040702889132786507,
      "loss": 4.8529,
      "mean_token_accuracy": 0.23288288861513137,
      "num_tokens": 71204294.0,
      "step": 31085
    },
    {
      "entropy": 5.395114326477051,
      "epoch": 3.0733491498616052,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004070003141532994,
      "loss": 4.912,
      "mean_token_accuracy": 0.2149900808930397,
      "num_tokens": 71215259.0,
      "step": 31090
    },
    {
      "entropy": 5.360031986236573,
      "epoch": 3.073843416370107,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004069717337315053,
      "loss": 4.9146,
      "mean_token_accuracy": 0.22238492965698242,
      "num_tokens": 71227008.0,
      "step": 31095
    },
    {
      "entropy": 5.404832744598389,
      "epoch": 3.074337682878608,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00040694315006318616,
      "loss": 4.9444,
      "mean_token_accuracy": 0.2207683727145195,
      "num_tokens": 71238190.0,
      "step": 31100
    },
    {
      "entropy": 5.465323400497437,
      "epoch": 3.0748319493871095,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040691456314904506,
      "loss": 4.9754,
      "mean_token_accuracy": 0.2101695105433464,
      "num_tokens": 71250534.0,
      "step": 31105
    },
    {
      "entropy": 5.354959344863891,
      "epoch": 3.0753262158956107,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040688597298978526,
      "loss": 4.8938,
      "mean_token_accuracy": 0.22911267429590226,
      "num_tokens": 71261432.0,
      "step": 31110
    },
    {
      "entropy": 5.371334505081177,
      "epoch": 3.0758204824041124,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004068573795861101,
      "loss": 4.9785,
      "mean_token_accuracy": 0.2108052685856819,
      "num_tokens": 71272090.0,
      "step": 31115
    },
    {
      "entropy": 5.381827688217163,
      "epoch": 3.0763147489126137,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004068287829387229,
      "loss": 4.9233,
      "mean_token_accuracy": 0.2179894044995308,
      "num_tokens": 71284696.0,
      "step": 31120
    },
    {
      "entropy": 5.381552028656006,
      "epoch": 3.076809015421115,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040680018304832717,
      "loss": 4.8956,
      "mean_token_accuracy": 0.2265668570995331,
      "num_tokens": 71295701.0,
      "step": 31125
    },
    {
      "entropy": 5.379792308807373,
      "epoch": 3.0773032819296167,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040677157991562653,
      "loss": 4.918,
      "mean_token_accuracy": 0.23010421097278594,
      "num_tokens": 71306798.0,
      "step": 31130
    },
    {
      "entropy": 5.4131556987762455,
      "epoch": 3.077797548438118,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004067429735413245,
      "loss": 4.9092,
      "mean_token_accuracy": 0.21743375808000565,
      "num_tokens": 71319171.0,
      "step": 31135
    },
    {
      "entropy": 5.399747848510742,
      "epoch": 3.078291814946619,
      "grad_norm": 1.0,
      "learning_rate": 0.000406714363926125,
      "loss": 4.9369,
      "mean_token_accuracy": 0.21440124809741973,
      "num_tokens": 71331479.0,
      "step": 31140
    },
    {
      "entropy": 5.346318340301513,
      "epoch": 3.0787860814551204,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00040668575107073165,
      "loss": 4.9041,
      "mean_token_accuracy": 0.22555935978889466,
      "num_tokens": 71343044.0,
      "step": 31145
    },
    {
      "entropy": 5.524312686920166,
      "epoch": 3.079280347963622,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040665713497584836,
      "loss": 4.9776,
      "mean_token_accuracy": 0.21446940153837205,
      "num_tokens": 71354552.0,
      "step": 31150
    },
    {
      "entropy": 5.341790008544922,
      "epoch": 3.0797746144721234,
      "grad_norm": 1.0,
      "learning_rate": 0.00040662851564217915,
      "loss": 4.8821,
      "mean_token_accuracy": 0.2279602110385895,
      "num_tokens": 71366201.0,
      "step": 31155
    },
    {
      "entropy": 5.384427165985107,
      "epoch": 3.0802688809806247,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00040659989307042816,
      "loss": 4.8928,
      "mean_token_accuracy": 0.2159039244055748,
      "num_tokens": 71377966.0,
      "step": 31160
    },
    {
      "entropy": 5.424661779403687,
      "epoch": 3.080763147489126,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040657126726129924,
      "loss": 4.9465,
      "mean_token_accuracy": 0.22257007658481598,
      "num_tokens": 71388927.0,
      "step": 31165
    },
    {
      "entropy": 5.378629970550537,
      "epoch": 3.0812574139976276,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004065426382154968,
      "loss": 4.9496,
      "mean_token_accuracy": 0.22233583331108092,
      "num_tokens": 71401728.0,
      "step": 31170
    },
    {
      "entropy": 5.457248687744141,
      "epoch": 3.081751680506129,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000406514005933725,
      "loss": 5.0882,
      "mean_token_accuracy": 0.2037815496325493,
      "num_tokens": 71413694.0,
      "step": 31175
    },
    {
      "entropy": 5.37805700302124,
      "epoch": 3.08224594701463,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004064853704166883,
      "loss": 4.8822,
      "mean_token_accuracy": 0.22784789204597472,
      "num_tokens": 71425485.0,
      "step": 31180
    },
    {
      "entropy": 5.404892349243164,
      "epoch": 3.082740213523132,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000406456731665091,
      "loss": 4.9193,
      "mean_token_accuracy": 0.2195939764380455,
      "num_tokens": 71437320.0,
      "step": 31185
    },
    {
      "entropy": 5.423341894149781,
      "epoch": 3.083234480031633,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040642808967963787,
      "loss": 4.9461,
      "mean_token_accuracy": 0.21681167781352997,
      "num_tokens": 71448948.0,
      "step": 31190
    },
    {
      "entropy": 5.385015249252319,
      "epoch": 3.0837287465401344,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040639944446103324,
      "loss": 4.8721,
      "mean_token_accuracy": 0.2233794316649437,
      "num_tokens": 71460997.0,
      "step": 31195
    },
    {
      "entropy": 5.344717121124267,
      "epoch": 3.0842230130486357,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040637079600998186,
      "loss": 4.8856,
      "mean_token_accuracy": 0.22196568101644515,
      "num_tokens": 71472174.0,
      "step": 31200
    },
    {
      "entropy": 5.376114988327027,
      "epoch": 3.0847172795571374,
      "grad_norm": 1.0,
      "learning_rate": 0.0004063421443271885,
      "loss": 4.9573,
      "mean_token_accuracy": 0.21956509947776795,
      "num_tokens": 71484306.0,
      "step": 31205
    },
    {
      "entropy": 5.395713472366333,
      "epoch": 3.0852115460656386,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040631348941335814,
      "loss": 4.889,
      "mean_token_accuracy": 0.22074636369943618,
      "num_tokens": 71494866.0,
      "step": 31210
    },
    {
      "entropy": 5.409148502349853,
      "epoch": 3.08570581257414,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004062848312691955,
      "loss": 4.8938,
      "mean_token_accuracy": 0.22326148003339769,
      "num_tokens": 71505591.0,
      "step": 31215
    },
    {
      "entropy": 5.407227277755737,
      "epoch": 3.086200079082641,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040625616989540565,
      "loss": 4.9401,
      "mean_token_accuracy": 0.21782163232564927,
      "num_tokens": 71517651.0,
      "step": 31220
    },
    {
      "entropy": 5.447109365463257,
      "epoch": 3.086694345591143,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004062275052926936,
      "loss": 4.9522,
      "mean_token_accuracy": 0.21395166367292404,
      "num_tokens": 71529301.0,
      "step": 31225
    },
    {
      "entropy": 5.2454078674316404,
      "epoch": 3.087188612099644,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004061988374617646,
      "loss": 4.7839,
      "mean_token_accuracy": 0.23325910419225693,
      "num_tokens": 71539971.0,
      "step": 31230
    },
    {
      "entropy": 5.385706424713135,
      "epoch": 3.0876828786081454,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004061701664033239,
      "loss": 4.9797,
      "mean_token_accuracy": 0.21628005653619767,
      "num_tokens": 71551905.0,
      "step": 31235
    },
    {
      "entropy": 5.398733282089234,
      "epoch": 3.088177145116647,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004061414921180767,
      "loss": 4.8953,
      "mean_token_accuracy": 0.22348379641771315,
      "num_tokens": 71562945.0,
      "step": 31240
    },
    {
      "entropy": 5.416925144195557,
      "epoch": 3.0886714116251484,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040611281460672846,
      "loss": 5.014,
      "mean_token_accuracy": 0.21070448756217958,
      "num_tokens": 71574429.0,
      "step": 31245
    },
    {
      "entropy": 5.4225428104400635,
      "epoch": 3.0891656781336496,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004060841338699847,
      "loss": 4.9695,
      "mean_token_accuracy": 0.21581717282533647,
      "num_tokens": 71584827.0,
      "step": 31250
    },
    {
      "entropy": 5.402391767501831,
      "epoch": 3.089659944642151,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040605544990855074,
      "loss": 4.8709,
      "mean_token_accuracy": 0.21529546827077867,
      "num_tokens": 71595729.0,
      "step": 31255
    },
    {
      "entropy": 5.386199188232422,
      "epoch": 3.0901542111506526,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040602676272313257,
      "loss": 4.925,
      "mean_token_accuracy": 0.2200094774365425,
      "num_tokens": 71606912.0,
      "step": 31260
    },
    {
      "entropy": 5.3592897891998295,
      "epoch": 3.090648477659154,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040599807231443566,
      "loss": 4.9733,
      "mean_token_accuracy": 0.22121639251708985,
      "num_tokens": 71618706.0,
      "step": 31265
    },
    {
      "entropy": 5.368315076828003,
      "epoch": 3.091142744167655,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004059693786831658,
      "loss": 4.8915,
      "mean_token_accuracy": 0.22769914269447328,
      "num_tokens": 71630826.0,
      "step": 31270
    },
    {
      "entropy": 5.456730222702026,
      "epoch": 3.0916370106761564,
      "grad_norm": 1.03125,
      "learning_rate": 0.000405940681830029,
      "loss": 4.99,
      "mean_token_accuracy": 0.2148710310459137,
      "num_tokens": 71643914.0,
      "step": 31275
    },
    {
      "entropy": 5.393510341644287,
      "epoch": 3.092131277184658,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000405911981755731,
      "loss": 4.8724,
      "mean_token_accuracy": 0.2283584237098694,
      "num_tokens": 71655523.0,
      "step": 31280
    },
    {
      "entropy": 5.350870990753174,
      "epoch": 3.0926255436931593,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004058832784609779,
      "loss": 4.8912,
      "mean_token_accuracy": 0.2224075213074684,
      "num_tokens": 71667423.0,
      "step": 31285
    },
    {
      "entropy": 5.3328886985778805,
      "epoch": 3.0931198102016606,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040585457194647595,
      "loss": 4.8832,
      "mean_token_accuracy": 0.21815989166498184,
      "num_tokens": 71678116.0,
      "step": 31290
    },
    {
      "entropy": 5.420734596252442,
      "epoch": 3.0936140767101623,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040582586221293127,
      "loss": 4.9783,
      "mean_token_accuracy": 0.2150306671857834,
      "num_tokens": 71690109.0,
      "step": 31295
    },
    {
      "entropy": 5.439567804336548,
      "epoch": 3.0941083432186636,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040579714926105,
      "loss": 5.0233,
      "mean_token_accuracy": 0.21238171607255935,
      "num_tokens": 71702640.0,
      "step": 31300
    },
    {
      "entropy": 5.398792934417725,
      "epoch": 3.094602609727165,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004057684330915386,
      "loss": 4.8879,
      "mean_token_accuracy": 0.2197289064526558,
      "num_tokens": 71714537.0,
      "step": 31305
    },
    {
      "entropy": 5.367176675796509,
      "epoch": 3.095096876235666,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040573971370510344,
      "loss": 4.9955,
      "mean_token_accuracy": 0.2143272802233696,
      "num_tokens": 71724901.0,
      "step": 31310
    },
    {
      "entropy": 5.427648258209229,
      "epoch": 3.095591142744168,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000405710991102451,
      "loss": 4.9692,
      "mean_token_accuracy": 0.21403862088918685,
      "num_tokens": 71736640.0,
      "step": 31315
    },
    {
      "entropy": 5.432495880126953,
      "epoch": 3.096085409252669,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040568226528428795,
      "loss": 4.9688,
      "mean_token_accuracy": 0.21255689710378647,
      "num_tokens": 71747193.0,
      "step": 31320
    },
    {
      "entropy": 5.420435237884521,
      "epoch": 3.0965796757611703,
      "grad_norm": 1.03125,
      "learning_rate": 0.000405653536251321,
      "loss": 5.024,
      "mean_token_accuracy": 0.20597499012947082,
      "num_tokens": 71759073.0,
      "step": 31325
    },
    {
      "entropy": 5.454598999023437,
      "epoch": 3.097073942269672,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004056248040042566,
      "loss": 4.9412,
      "mean_token_accuracy": 0.21655234545469285,
      "num_tokens": 71770486.0,
      "step": 31330
    },
    {
      "entropy": 5.466549205780029,
      "epoch": 3.0975682087781733,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040559606854380186,
      "loss": 5.0116,
      "mean_token_accuracy": 0.2191604793071747,
      "num_tokens": 71781536.0,
      "step": 31335
    },
    {
      "entropy": 5.421922922134399,
      "epoch": 3.0980624752866746,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040556732987066357,
      "loss": 4.992,
      "mean_token_accuracy": 0.21619280576705932,
      "num_tokens": 71792521.0,
      "step": 31340
    },
    {
      "entropy": 5.3545595645904545,
      "epoch": 3.098556741795176,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004055385879855487,
      "loss": 4.9587,
      "mean_token_accuracy": 0.21848182529211044,
      "num_tokens": 71804150.0,
      "step": 31345
    },
    {
      "entropy": 5.363057565689087,
      "epoch": 3.0990510083036775,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040550984288916436,
      "loss": 4.91,
      "mean_token_accuracy": 0.230670328438282,
      "num_tokens": 71815842.0,
      "step": 31350
    },
    {
      "entropy": 5.4817955017089846,
      "epoch": 3.099545274812179,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040548109458221756,
      "loss": 4.945,
      "mean_token_accuracy": 0.2173105150461197,
      "num_tokens": 71827243.0,
      "step": 31355
    },
    {
      "entropy": 5.40636100769043,
      "epoch": 3.10003954132068,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004054523430654156,
      "loss": 4.986,
      "mean_token_accuracy": 0.21955959349870682,
      "num_tokens": 71839424.0,
      "step": 31360
    },
    {
      "entropy": 5.391243124008179,
      "epoch": 3.1005338078291813,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040542358833946577,
      "loss": 4.9982,
      "mean_token_accuracy": 0.2187706470489502,
      "num_tokens": 71851709.0,
      "step": 31365
    },
    {
      "entropy": 5.391156339645386,
      "epoch": 3.101028074337683,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004053948304050755,
      "loss": 4.9196,
      "mean_token_accuracy": 0.2206217274069786,
      "num_tokens": 71862841.0,
      "step": 31370
    },
    {
      "entropy": 5.403653764724732,
      "epoch": 3.1015223408461843,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040536606926295203,
      "loss": 4.911,
      "mean_token_accuracy": 0.2273534953594208,
      "num_tokens": 71873104.0,
      "step": 31375
    },
    {
      "entropy": 5.449539995193481,
      "epoch": 3.1020166073546855,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040533730491380316,
      "loss": 5.0252,
      "mean_token_accuracy": 0.21418839693069458,
      "num_tokens": 71884809.0,
      "step": 31380
    },
    {
      "entropy": 5.403022241592407,
      "epoch": 3.102510873863187,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004053085373583363,
      "loss": 4.925,
      "mean_token_accuracy": 0.22159834653139115,
      "num_tokens": 71896625.0,
      "step": 31385
    },
    {
      "entropy": 5.371267318725586,
      "epoch": 3.1030051403716885,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004052797665972593,
      "loss": 4.9183,
      "mean_token_accuracy": 0.2240040495991707,
      "num_tokens": 71908628.0,
      "step": 31390
    },
    {
      "entropy": 5.404375076293945,
      "epoch": 3.1034994068801898,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040525099263127976,
      "loss": 4.944,
      "mean_token_accuracy": 0.2187753990292549,
      "num_tokens": 71919971.0,
      "step": 31395
    },
    {
      "entropy": 5.420210647583008,
      "epoch": 3.103993673388691,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040522221546110556,
      "loss": 5.0062,
      "mean_token_accuracy": 0.21275546848773957,
      "num_tokens": 71931240.0,
      "step": 31400
    },
    {
      "entropy": 5.343015098571778,
      "epoch": 3.1044879398971927,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00040519343508744473,
      "loss": 4.9394,
      "mean_token_accuracy": 0.22236475944519044,
      "num_tokens": 71942735.0,
      "step": 31405
    },
    {
      "entropy": 5.32021427154541,
      "epoch": 3.104982206405694,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040516465151100515,
      "loss": 4.875,
      "mean_token_accuracy": 0.2349018082022667,
      "num_tokens": 71953685.0,
      "step": 31410
    },
    {
      "entropy": 5.411223554611206,
      "epoch": 3.1054764729141953,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040513586473249496,
      "loss": 4.9482,
      "mean_token_accuracy": 0.21441251188516616,
      "num_tokens": 71966061.0,
      "step": 31415
    },
    {
      "entropy": 5.453322887420654,
      "epoch": 3.1059707394226965,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004051070747526223,
      "loss": 4.9788,
      "mean_token_accuracy": 0.21059480160474778,
      "num_tokens": 71978002.0,
      "step": 31420
    },
    {
      "entropy": 5.345702886581421,
      "epoch": 3.1064650059311982,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040507828157209546,
      "loss": 4.8359,
      "mean_token_accuracy": 0.23058629482984544,
      "num_tokens": 71988504.0,
      "step": 31425
    },
    {
      "entropy": 5.35727858543396,
      "epoch": 3.1069592724396995,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040504948519162276,
      "loss": 4.9915,
      "mean_token_accuracy": 0.2080179750919342,
      "num_tokens": 72000900.0,
      "step": 31430
    },
    {
      "entropy": 5.463846635818482,
      "epoch": 3.1074535389482008,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004050206856119124,
      "loss": 4.992,
      "mean_token_accuracy": 0.21868247985839845,
      "num_tokens": 72011134.0,
      "step": 31435
    },
    {
      "entropy": 5.375261640548706,
      "epoch": 3.1079478054567025,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004049918828336732,
      "loss": 4.892,
      "mean_token_accuracy": 0.2202080473303795,
      "num_tokens": 72022784.0,
      "step": 31440
    },
    {
      "entropy": 5.332051944732666,
      "epoch": 3.1084420719652037,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004049630768576133,
      "loss": 4.8455,
      "mean_token_accuracy": 0.2305077686905861,
      "num_tokens": 72033353.0,
      "step": 31445
    },
    {
      "entropy": 5.4143595695495605,
      "epoch": 3.108936338473705,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040493426768444164,
      "loss": 4.935,
      "mean_token_accuracy": 0.21477026492357254,
      "num_tokens": 72044182.0,
      "step": 31450
    },
    {
      "entropy": 5.455833053588867,
      "epoch": 3.1094306049822062,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004049054553148668,
      "loss": 4.992,
      "mean_token_accuracy": 0.21190426349639893,
      "num_tokens": 72056365.0,
      "step": 31455
    },
    {
      "entropy": 5.328071784973145,
      "epoch": 3.109924871490708,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004048766397495977,
      "loss": 4.9127,
      "mean_token_accuracy": 0.21887015104293822,
      "num_tokens": 72066580.0,
      "step": 31460
    },
    {
      "entropy": 5.364016199111939,
      "epoch": 3.110419137999209,
      "grad_norm": 1.0,
      "learning_rate": 0.00040484782098934303,
      "loss": 4.9479,
      "mean_token_accuracy": 0.22352479100227357,
      "num_tokens": 72077731.0,
      "step": 31465
    },
    {
      "entropy": 5.446867942810059,
      "epoch": 3.1109134045077105,
      "grad_norm": 1.25,
      "learning_rate": 0.0004048189990348119,
      "loss": 4.9528,
      "mean_token_accuracy": 0.21751641035079955,
      "num_tokens": 72088785.0,
      "step": 31470
    },
    {
      "entropy": 5.362438631057739,
      "epoch": 3.1114076710162117,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040479017388671317,
      "loss": 4.9167,
      "mean_token_accuracy": 0.22526904940605164,
      "num_tokens": 72100235.0,
      "step": 31475
    },
    {
      "entropy": 5.321478366851807,
      "epoch": 3.1119019375247134,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040476134554575613,
      "loss": 4.8661,
      "mean_token_accuracy": 0.2253323972225189,
      "num_tokens": 72112124.0,
      "step": 31480
    },
    {
      "entropy": 5.447280359268189,
      "epoch": 3.1123962040332147,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004047325140126497,
      "loss": 4.928,
      "mean_token_accuracy": 0.21651099473237992,
      "num_tokens": 72124138.0,
      "step": 31485
    },
    {
      "entropy": 5.428809785842896,
      "epoch": 3.112890470541716,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040470367928810343,
      "loss": 4.9175,
      "mean_token_accuracy": 0.2161221444606781,
      "num_tokens": 72134330.0,
      "step": 31490
    },
    {
      "entropy": 5.477111148834228,
      "epoch": 3.1133847370502177,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00040467484137282644,
      "loss": 5.0524,
      "mean_token_accuracy": 0.20785557329654694,
      "num_tokens": 72146999.0,
      "step": 31495
    },
    {
      "entropy": 5.3855945587158205,
      "epoch": 3.113879003558719,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040464600026752834,
      "loss": 4.9342,
      "mean_token_accuracy": 0.2198129877448082,
      "num_tokens": 72158648.0,
      "step": 31500
    },
    {
      "entropy": 5.412158632278443,
      "epoch": 3.11437327006722,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004046171559729184,
      "loss": 4.9338,
      "mean_token_accuracy": 0.22022552341222762,
      "num_tokens": 72169621.0,
      "step": 31505
    },
    {
      "entropy": 5.352029800415039,
      "epoch": 3.1148675365757215,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004045883084897064,
      "loss": 4.8555,
      "mean_token_accuracy": 0.22678750306367873,
      "num_tokens": 72179928.0,
      "step": 31510
    },
    {
      "entropy": 5.354543972015381,
      "epoch": 3.115361803084223,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040455945781860185,
      "loss": 4.8919,
      "mean_token_accuracy": 0.22088756412267685,
      "num_tokens": 72190699.0,
      "step": 31515
    },
    {
      "entropy": 5.374576663970947,
      "epoch": 3.1158560695927244,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004045306039603146,
      "loss": 4.9549,
      "mean_token_accuracy": 0.2160742238163948,
      "num_tokens": 72202893.0,
      "step": 31520
    },
    {
      "entropy": 5.416742992401123,
      "epoch": 3.1163503361012257,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040450174691555434,
      "loss": 5.0209,
      "mean_token_accuracy": 0.215982648730278,
      "num_tokens": 72214614.0,
      "step": 31525
    },
    {
      "entropy": 5.371134567260742,
      "epoch": 3.116844602609727,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004044728866850311,
      "loss": 4.9136,
      "mean_token_accuracy": 0.22836748510599136,
      "num_tokens": 72225882.0,
      "step": 31530
    },
    {
      "entropy": 5.445662593841552,
      "epoch": 3.1173388691182287,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004044440232694547,
      "loss": 4.9401,
      "mean_token_accuracy": 0.21767502427101135,
      "num_tokens": 72236264.0,
      "step": 31535
    },
    {
      "entropy": 5.399515914916992,
      "epoch": 3.11783313562673,
      "grad_norm": 1.0,
      "learning_rate": 0.0004044151566695353,
      "loss": 5.0062,
      "mean_token_accuracy": 0.2138653352856636,
      "num_tokens": 72247380.0,
      "step": 31540
    },
    {
      "entropy": 5.389342212677002,
      "epoch": 3.118327402135231,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040438628688598287,
      "loss": 4.8524,
      "mean_token_accuracy": 0.22732233852148057,
      "num_tokens": 72259141.0,
      "step": 31545
    },
    {
      "entropy": 5.3546709537506105,
      "epoch": 3.118821668643733,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004043574139195078,
      "loss": 4.8124,
      "mean_token_accuracy": 0.2260691612958908,
      "num_tokens": 72269549.0,
      "step": 31550
    },
    {
      "entropy": 5.313196992874145,
      "epoch": 3.119315935152234,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040432853777082025,
      "loss": 4.915,
      "mean_token_accuracy": 0.22924906015396118,
      "num_tokens": 72281589.0,
      "step": 31555
    },
    {
      "entropy": 5.3704828262329105,
      "epoch": 3.1198102016607354,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004042996584406305,
      "loss": 5.0143,
      "mean_token_accuracy": 0.20906869620084761,
      "num_tokens": 72294239.0,
      "step": 31560
    },
    {
      "entropy": 5.368531560897827,
      "epoch": 3.1203044681692367,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040427077592964923,
      "loss": 4.9811,
      "mean_token_accuracy": 0.21467864960432054,
      "num_tokens": 72306153.0,
      "step": 31565
    },
    {
      "entropy": 5.444409036636353,
      "epoch": 3.1207987346777384,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004042418902385868,
      "loss": 4.924,
      "mean_token_accuracy": 0.21923563927412032,
      "num_tokens": 72317270.0,
      "step": 31570
    },
    {
      "entropy": 5.399342727661133,
      "epoch": 3.1212930011862396,
      "grad_norm": 1.0,
      "learning_rate": 0.00040421300136815375,
      "loss": 4.961,
      "mean_token_accuracy": 0.21719055920839309,
      "num_tokens": 72329073.0,
      "step": 31575
    },
    {
      "entropy": 5.32893853187561,
      "epoch": 3.121787267694741,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004041841093190609,
      "loss": 4.8975,
      "mean_token_accuracy": 0.22465674430131913,
      "num_tokens": 72339677.0,
      "step": 31580
    },
    {
      "entropy": 5.417914962768554,
      "epoch": 3.1222815342032426,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004041552140920189,
      "loss": 4.9468,
      "mean_token_accuracy": 0.22710075378417968,
      "num_tokens": 72350527.0,
      "step": 31585
    },
    {
      "entropy": 5.439721775054932,
      "epoch": 3.122775800711744,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004041263156877386,
      "loss": 4.948,
      "mean_token_accuracy": 0.22070085257291794,
      "num_tokens": 72362740.0,
      "step": 31590
    },
    {
      "entropy": 5.354420804977417,
      "epoch": 3.123270067220245,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040409741410693084,
      "loss": 4.919,
      "mean_token_accuracy": 0.2167448952794075,
      "num_tokens": 72372705.0,
      "step": 31595
    },
    {
      "entropy": 5.251502418518067,
      "epoch": 3.1237643337287464,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040406850935030677,
      "loss": 4.8502,
      "mean_token_accuracy": 0.23087685853242873,
      "num_tokens": 72385085.0,
      "step": 31600
    },
    {
      "entropy": 5.402199172973633,
      "epoch": 3.124258600237248,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040403960141857724,
      "loss": 4.9619,
      "mean_token_accuracy": 0.2238198161125183,
      "num_tokens": 72397017.0,
      "step": 31605
    },
    {
      "entropy": 5.39735746383667,
      "epoch": 3.1247528667457494,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004040106903124536,
      "loss": 4.9275,
      "mean_token_accuracy": 0.22010526955127716,
      "num_tokens": 72409378.0,
      "step": 31610
    },
    {
      "entropy": 5.406848955154419,
      "epoch": 3.1252471332542506,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004039817760326469,
      "loss": 4.9687,
      "mean_token_accuracy": 0.21038103550672532,
      "num_tokens": 72420585.0,
      "step": 31615
    },
    {
      "entropy": 5.493139743804932,
      "epoch": 3.125741399762752,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040395285857986845,
      "loss": 5.0677,
      "mean_token_accuracy": 0.2071381613612175,
      "num_tokens": 72431779.0,
      "step": 31620
    },
    {
      "entropy": 5.404197645187378,
      "epoch": 3.1262356662712536,
      "grad_norm": 1.0,
      "learning_rate": 0.0004039239379548298,
      "loss": 4.8775,
      "mean_token_accuracy": 0.22189362496137618,
      "num_tokens": 72441421.0,
      "step": 31625
    },
    {
      "entropy": 5.4007243633270265,
      "epoch": 3.126729932779755,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004038950141582422,
      "loss": 4.9403,
      "mean_token_accuracy": 0.2159806951880455,
      "num_tokens": 72452664.0,
      "step": 31630
    },
    {
      "entropy": 5.350979661941528,
      "epoch": 3.127224199288256,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004038660871908173,
      "loss": 4.8457,
      "mean_token_accuracy": 0.22348916977643968,
      "num_tokens": 72464008.0,
      "step": 31635
    },
    {
      "entropy": 5.460098218917847,
      "epoch": 3.1277184657967574,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040383715705326665,
      "loss": 5.0126,
      "mean_token_accuracy": 0.2143450438976288,
      "num_tokens": 72475782.0,
      "step": 31640
    },
    {
      "entropy": 5.360867547988891,
      "epoch": 3.128212732305259,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004038082237463019,
      "loss": 4.9662,
      "mean_token_accuracy": 0.21342115700244904,
      "num_tokens": 72487770.0,
      "step": 31645
    },
    {
      "entropy": 5.3050302028656,
      "epoch": 3.1287069988137604,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00040377928727063485,
      "loss": 4.8601,
      "mean_token_accuracy": 0.2245774507522583,
      "num_tokens": 72499336.0,
      "step": 31650
    },
    {
      "entropy": 5.353214168548584,
      "epoch": 3.1292012653222616,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00040375034762697736,
      "loss": 4.9044,
      "mean_token_accuracy": 0.21769764870405198,
      "num_tokens": 72510658.0,
      "step": 31655
    },
    {
      "entropy": 5.400566720962525,
      "epoch": 3.1296955318307633,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004037214048160415,
      "loss": 4.9384,
      "mean_token_accuracy": 0.21332341134548188,
      "num_tokens": 72522617.0,
      "step": 31660
    },
    {
      "entropy": 5.379146718978882,
      "epoch": 3.1301897983392646,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004036924588385389,
      "loss": 4.8237,
      "mean_token_accuracy": 0.23479813039302827,
      "num_tokens": 72532829.0,
      "step": 31665
    },
    {
      "entropy": 5.350657653808594,
      "epoch": 3.130684064847766,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040366350969518196,
      "loss": 4.9004,
      "mean_token_accuracy": 0.22724444568157195,
      "num_tokens": 72543556.0,
      "step": 31670
    },
    {
      "entropy": 5.446150922775269,
      "epoch": 3.131178331356267,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004036345573866827,
      "loss": 4.9468,
      "mean_token_accuracy": 0.21405071914196014,
      "num_tokens": 72554545.0,
      "step": 31675
    },
    {
      "entropy": 5.4809637546539305,
      "epoch": 3.131672597864769,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040360560191375335,
      "loss": 5.0305,
      "mean_token_accuracy": 0.21018345803022384,
      "num_tokens": 72566158.0,
      "step": 31680
    },
    {
      "entropy": 5.426585483551025,
      "epoch": 3.13216686437327,
      "grad_norm": 1.0,
      "learning_rate": 0.0004035766432771063,
      "loss": 5.057,
      "mean_token_accuracy": 0.21679708808660508,
      "num_tokens": 72578229.0,
      "step": 31685
    },
    {
      "entropy": 5.3285027027130125,
      "epoch": 3.1326611308817713,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040354768147745387,
      "loss": 4.8339,
      "mean_token_accuracy": 0.22811528593301772,
      "num_tokens": 72589237.0,
      "step": 31690
    },
    {
      "entropy": 5.420715808868408,
      "epoch": 3.133155397390273,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004035187165155085,
      "loss": 4.9489,
      "mean_token_accuracy": 0.2195846125483513,
      "num_tokens": 72600531.0,
      "step": 31695
    },
    {
      "entropy": 5.37408480644226,
      "epoch": 3.1336496638987743,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004034897483919827,
      "loss": 4.8816,
      "mean_token_accuracy": 0.22504348456859588,
      "num_tokens": 72611861.0,
      "step": 31700
    },
    {
      "entropy": 5.399639129638672,
      "epoch": 3.1341439304072756,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004034607771075893,
      "loss": 4.9678,
      "mean_token_accuracy": 0.22025686651468276,
      "num_tokens": 72624142.0,
      "step": 31705
    },
    {
      "entropy": 5.369750118255615,
      "epoch": 3.134638196915777,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040343180266304074,
      "loss": 4.873,
      "mean_token_accuracy": 0.2192170351743698,
      "num_tokens": 72634995.0,
      "step": 31710
    },
    {
      "entropy": 5.423182582855224,
      "epoch": 3.1351324634242785,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00040340282505904997,
      "loss": 5.0036,
      "mean_token_accuracy": 0.21890877038240433,
      "num_tokens": 72646798.0,
      "step": 31715
    },
    {
      "entropy": 5.428723955154419,
      "epoch": 3.13562672993278,
      "grad_norm": 1.0,
      "learning_rate": 0.00040337384429632974,
      "loss": 4.9292,
      "mean_token_accuracy": 0.22421688884496688,
      "num_tokens": 72657938.0,
      "step": 31720
    },
    {
      "entropy": 5.454217100143433,
      "epoch": 3.136120996441281,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000403344860375593,
      "loss": 5.003,
      "mean_token_accuracy": 0.2139722988009453,
      "num_tokens": 72671069.0,
      "step": 31725
    },
    {
      "entropy": 5.37121171951294,
      "epoch": 3.1366152629497823,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004033158732975529,
      "loss": 4.9461,
      "mean_token_accuracy": 0.2167557120323181,
      "num_tokens": 72682695.0,
      "step": 31730
    },
    {
      "entropy": 5.40613489151001,
      "epoch": 3.137109529458284,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004032868830629223,
      "loss": 4.976,
      "mean_token_accuracy": 0.22074110209941863,
      "num_tokens": 72694514.0,
      "step": 31735
    },
    {
      "entropy": 5.360488414764404,
      "epoch": 3.1376037959667853,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040325788967241453,
      "loss": 4.872,
      "mean_token_accuracy": 0.21971508413553237,
      "num_tokens": 72705427.0,
      "step": 31740
    },
    {
      "entropy": 5.319887733459472,
      "epoch": 3.1380980624752866,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040322889312674275,
      "loss": 4.8495,
      "mean_token_accuracy": 0.22729218751192093,
      "num_tokens": 72716877.0,
      "step": 31745
    },
    {
      "entropy": 5.331980514526367,
      "epoch": 3.1385923289837883,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040319989342662036,
      "loss": 4.872,
      "mean_token_accuracy": 0.22580246329307557,
      "num_tokens": 72727927.0,
      "step": 31750
    },
    {
      "entropy": 5.428868579864502,
      "epoch": 3.1390865954922895,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040317089057276067,
      "loss": 5.0252,
      "mean_token_accuracy": 0.21703898459672927,
      "num_tokens": 72739645.0,
      "step": 31755
    },
    {
      "entropy": 5.461847829818725,
      "epoch": 3.139580862000791,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004031418845658772,
      "loss": 5.0022,
      "mean_token_accuracy": 0.20884791016578674,
      "num_tokens": 72750961.0,
      "step": 31760
    },
    {
      "entropy": 5.440127515792847,
      "epoch": 3.140075128509292,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040311287540668346,
      "loss": 4.9883,
      "mean_token_accuracy": 0.21199798583984375,
      "num_tokens": 72763599.0,
      "step": 31765
    },
    {
      "entropy": 5.377119350433349,
      "epoch": 3.1405693950177938,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00040308386309589314,
      "loss": 4.8961,
      "mean_token_accuracy": 0.22561048567295075,
      "num_tokens": 72775807.0,
      "step": 31770
    },
    {
      "entropy": 5.389883232116699,
      "epoch": 3.141063661526295,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004030548476342199,
      "loss": 4.9096,
      "mean_token_accuracy": 0.22096057534217833,
      "num_tokens": 72786161.0,
      "step": 31775
    },
    {
      "entropy": 5.403966283798217,
      "epoch": 3.1415579280347963,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040302582902237755,
      "loss": 4.9828,
      "mean_token_accuracy": 0.21934870034456252,
      "num_tokens": 72798211.0,
      "step": 31780
    },
    {
      "entropy": 5.434576845169067,
      "epoch": 3.1420521945432975,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00040299680726107995,
      "loss": 4.9783,
      "mean_token_accuracy": 0.21553680002689363,
      "num_tokens": 72809187.0,
      "step": 31785
    },
    {
      "entropy": 5.373623180389404,
      "epoch": 3.1425464610517992,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00040296778235104105,
      "loss": 4.9146,
      "mean_token_accuracy": 0.22289526462554932,
      "num_tokens": 72820486.0,
      "step": 31790
    },
    {
      "entropy": 5.322180080413818,
      "epoch": 3.1430407275603005,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004029387542929748,
      "loss": 4.8525,
      "mean_token_accuracy": 0.23025185018777847,
      "num_tokens": 72831442.0,
      "step": 31795
    },
    {
      "entropy": 5.368121814727783,
      "epoch": 3.1435349940688018,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040290972308759535,
      "loss": 4.8835,
      "mean_token_accuracy": 0.2250526249408722,
      "num_tokens": 72841902.0,
      "step": 31800
    },
    {
      "entropy": 5.288333320617676,
      "epoch": 3.1440292605773035,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040288068873561687,
      "loss": 4.7973,
      "mean_token_accuracy": 0.23452527672052384,
      "num_tokens": 72852327.0,
      "step": 31805
    },
    {
      "entropy": 5.406289434432983,
      "epoch": 3.1445235270858047,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004028516512377537,
      "loss": 4.9435,
      "mean_token_accuracy": 0.20876182466745377,
      "num_tokens": 72862643.0,
      "step": 31810
    },
    {
      "entropy": 5.346348810195923,
      "epoch": 3.145017793594306,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00040282261059471997,
      "loss": 4.9435,
      "mean_token_accuracy": 0.21532326936721802,
      "num_tokens": 72874081.0,
      "step": 31815
    },
    {
      "entropy": 5.411605739593506,
      "epoch": 3.1455120601028073,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00040279356680723025,
      "loss": 4.9311,
      "mean_token_accuracy": 0.2198442131280899,
      "num_tokens": 72885472.0,
      "step": 31820
    },
    {
      "entropy": 5.401024675369262,
      "epoch": 3.146006326611309,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004027645198759988,
      "loss": 4.9145,
      "mean_token_accuracy": 0.2271820932626724,
      "num_tokens": 72896626.0,
      "step": 31825
    },
    {
      "entropy": 5.380189085006714,
      "epoch": 3.1465005931198102,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004027354698017405,
      "loss": 4.9274,
      "mean_token_accuracy": 0.22124247550964354,
      "num_tokens": 72907782.0,
      "step": 31830
    },
    {
      "entropy": 5.353755712509155,
      "epoch": 3.1469948596283115,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040270641658516986,
      "loss": 4.8882,
      "mean_token_accuracy": 0.21811637729406358,
      "num_tokens": 72919024.0,
      "step": 31835
    },
    {
      "entropy": 5.393744707107544,
      "epoch": 3.147489126136813,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004026773602270015,
      "loss": 4.9435,
      "mean_token_accuracy": 0.2233407124876976,
      "num_tokens": 72930576.0,
      "step": 31840
    },
    {
      "entropy": 5.416038560867309,
      "epoch": 3.1479833926453145,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040264830072795027,
      "loss": 4.9904,
      "mean_token_accuracy": 0.21649879217147827,
      "num_tokens": 72942489.0,
      "step": 31845
    },
    {
      "entropy": 5.44717001914978,
      "epoch": 3.1484776591538157,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004026192380887311,
      "loss": 5.0102,
      "mean_token_accuracy": 0.20751241594552994,
      "num_tokens": 72954209.0,
      "step": 31850
    },
    {
      "entropy": 5.45564866065979,
      "epoch": 3.148971925662317,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004025901723100588,
      "loss": 5.0308,
      "mean_token_accuracy": 0.21578427255153657,
      "num_tokens": 72965939.0,
      "step": 31855
    },
    {
      "entropy": 5.375657749176026,
      "epoch": 3.1494661921708187,
      "grad_norm": 1.0,
      "learning_rate": 0.0004025611033926484,
      "loss": 4.8697,
      "mean_token_accuracy": 0.22631241828203202,
      "num_tokens": 72977114.0,
      "step": 31860
    },
    {
      "entropy": 5.429088401794433,
      "epoch": 3.14996045867932,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040253203133721515,
      "loss": 4.9948,
      "mean_token_accuracy": 0.21727503091096878,
      "num_tokens": 72989726.0,
      "step": 31865
    },
    {
      "entropy": 5.303113174438477,
      "epoch": 3.150454725187821,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004025029561444742,
      "loss": 4.9174,
      "mean_token_accuracy": 0.2209486961364746,
      "num_tokens": 73000055.0,
      "step": 31870
    },
    {
      "entropy": 5.313371324539185,
      "epoch": 3.1509489916963225,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040247387781514065,
      "loss": 4.9158,
      "mean_token_accuracy": 0.221088208258152,
      "num_tokens": 73010769.0,
      "step": 31875
    },
    {
      "entropy": 5.46993932723999,
      "epoch": 3.151443258204824,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00040244479634992997,
      "loss": 4.9782,
      "mean_token_accuracy": 0.21680374145507814,
      "num_tokens": 73022032.0,
      "step": 31880
    },
    {
      "entropy": 5.358133220672608,
      "epoch": 3.1519375247133254,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040241571174955755,
      "loss": 4.9203,
      "mean_token_accuracy": 0.21705729514360428,
      "num_tokens": 73033877.0,
      "step": 31885
    },
    {
      "entropy": 5.351234340667725,
      "epoch": 3.1524317912218267,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004023866240147389,
      "loss": 4.8757,
      "mean_token_accuracy": 0.21882398128509523,
      "num_tokens": 73045670.0,
      "step": 31890
    },
    {
      "entropy": 5.301354646682739,
      "epoch": 3.152926057730328,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004023575331461894,
      "loss": 4.92,
      "mean_token_accuracy": 0.2239979699254036,
      "num_tokens": 73056879.0,
      "step": 31895
    },
    {
      "entropy": 5.388696146011353,
      "epoch": 3.1534203242388297,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040232843914462493,
      "loss": 4.9598,
      "mean_token_accuracy": 0.212164968252182,
      "num_tokens": 73068422.0,
      "step": 31900
    },
    {
      "entropy": 5.407004117965698,
      "epoch": 3.153914590747331,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004022993420107611,
      "loss": 4.8649,
      "mean_token_accuracy": 0.2228280708193779,
      "num_tokens": 73079883.0,
      "step": 31905
    },
    {
      "entropy": 5.354489421844482,
      "epoch": 3.154408857255832,
      "grad_norm": 1.125,
      "learning_rate": 0.00040227024174531364,
      "loss": 4.934,
      "mean_token_accuracy": 0.2211434632539749,
      "num_tokens": 73090832.0,
      "step": 31910
    },
    {
      "entropy": 5.334781455993652,
      "epoch": 3.154903123764334,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004022411383489985,
      "loss": 4.8762,
      "mean_token_accuracy": 0.22714847773313523,
      "num_tokens": 73102154.0,
      "step": 31915
    },
    {
      "entropy": 5.40356183052063,
      "epoch": 3.155397390272835,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004022120318225317,
      "loss": 4.9525,
      "mean_token_accuracy": 0.22355523258447646,
      "num_tokens": 73113530.0,
      "step": 31920
    },
    {
      "entropy": 5.464871644973755,
      "epoch": 3.1558916567813364,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040218292216662903,
      "loss": 5.0099,
      "mean_token_accuracy": 0.21506910920143127,
      "num_tokens": 73124379.0,
      "step": 31925
    },
    {
      "entropy": 5.337626218795776,
      "epoch": 3.1563859232898377,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040215380938200693,
      "loss": 4.9465,
      "mean_token_accuracy": 0.21807491332292556,
      "num_tokens": 73136406.0,
      "step": 31930
    },
    {
      "entropy": 5.4622725486755375,
      "epoch": 3.1568801897983394,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00040212469346938127,
      "loss": 4.9911,
      "mean_token_accuracy": 0.21763025969266891,
      "num_tokens": 73149302.0,
      "step": 31935
    },
    {
      "entropy": 5.361596584320068,
      "epoch": 3.1573744563068407,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004020955744294685,
      "loss": 4.9368,
      "mean_token_accuracy": 0.2116974651813507,
      "num_tokens": 73160462.0,
      "step": 31940
    },
    {
      "entropy": 5.325572109222412,
      "epoch": 3.157868722815342,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040206645226298484,
      "loss": 4.832,
      "mean_token_accuracy": 0.22986166179180145,
      "num_tokens": 73171849.0,
      "step": 31945
    },
    {
      "entropy": 5.326461219787598,
      "epoch": 3.1583629893238436,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004020373269706468,
      "loss": 4.8834,
      "mean_token_accuracy": 0.22431587129831315,
      "num_tokens": 73182132.0,
      "step": 31950
    },
    {
      "entropy": 5.4314525604248045,
      "epoch": 3.158857255832345,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004020081985531706,
      "loss": 4.9867,
      "mean_token_accuracy": 0.21719207167625426,
      "num_tokens": 73194527.0,
      "step": 31955
    },
    {
      "entropy": 5.338670158386231,
      "epoch": 3.159351522340846,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004019790670112733,
      "loss": 4.9093,
      "mean_token_accuracy": 0.22252341955900193,
      "num_tokens": 73206104.0,
      "step": 31960
    },
    {
      "entropy": 5.308073091506958,
      "epoch": 3.1598457888493474,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040194993234567107,
      "loss": 4.9158,
      "mean_token_accuracy": 0.22295881509780885,
      "num_tokens": 73217665.0,
      "step": 31965
    },
    {
      "entropy": 5.413907527923584,
      "epoch": 3.160340055357849,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040192079455708083,
      "loss": 4.9342,
      "mean_token_accuracy": 0.21799944043159486,
      "num_tokens": 73229677.0,
      "step": 31970
    },
    {
      "entropy": 5.442161321640015,
      "epoch": 3.1608343218663504,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040189165364621943,
      "loss": 4.9647,
      "mean_token_accuracy": 0.21902752071619033,
      "num_tokens": 73240806.0,
      "step": 31975
    },
    {
      "entropy": 5.363771724700928,
      "epoch": 3.1613285883748516,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004018625096138036,
      "loss": 4.8617,
      "mean_token_accuracy": 0.2169555678963661,
      "num_tokens": 73252016.0,
      "step": 31980
    },
    {
      "entropy": 5.392846345901489,
      "epoch": 3.161822854883353,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040183336246055033,
      "loss": 4.8977,
      "mean_token_accuracy": 0.21520236283540725,
      "num_tokens": 73262790.0,
      "step": 31985
    },
    {
      "entropy": 5.413765335083008,
      "epoch": 3.1623171213918546,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004018042121871767,
      "loss": 5.0099,
      "mean_token_accuracy": 0.21065073162317277,
      "num_tokens": 73275424.0,
      "step": 31990
    },
    {
      "entropy": 5.3123939037323,
      "epoch": 3.162811387900356,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040177505879439984,
      "loss": 4.901,
      "mean_token_accuracy": 0.22359256446361542,
      "num_tokens": 73286790.0,
      "step": 31995
    },
    {
      "entropy": 5.453336715698242,
      "epoch": 3.163305654408857,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004017459022829368,
      "loss": 5.0565,
      "mean_token_accuracy": 0.20937722325325012,
      "num_tokens": 73298876.0,
      "step": 32000
    },
    {
      "entropy": 5.389720010757446,
      "epoch": 3.163799920917359,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00040171674265350487,
      "loss": 4.9266,
      "mean_token_accuracy": 0.21936736404895782,
      "num_tokens": 73310635.0,
      "step": 32005
    },
    {
      "entropy": 5.450805377960205,
      "epoch": 3.16429418742586,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004016875799068215,
      "loss": 4.9988,
      "mean_token_accuracy": 0.21439515948295593,
      "num_tokens": 73321998.0,
      "step": 32010
    },
    {
      "entropy": 5.359823179244995,
      "epoch": 3.1647884539343614,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004016584140436039,
      "loss": 4.9091,
      "mean_token_accuracy": 0.22614518254995347,
      "num_tokens": 73332979.0,
      "step": 32015
    },
    {
      "entropy": 5.371836948394775,
      "epoch": 3.1652827204428626,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004016292450645698,
      "loss": 4.9673,
      "mean_token_accuracy": 0.21279709041118622,
      "num_tokens": 73344405.0,
      "step": 32020
    },
    {
      "entropy": 5.407127141952515,
      "epoch": 3.1657769869513643,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040160007297043655,
      "loss": 4.9479,
      "mean_token_accuracy": 0.22183094620704652,
      "num_tokens": 73356291.0,
      "step": 32025
    },
    {
      "entropy": 5.418062734603882,
      "epoch": 3.1662712534598656,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040157089776192184,
      "loss": 4.9688,
      "mean_token_accuracy": 0.2266653746366501,
      "num_tokens": 73368299.0,
      "step": 32030
    },
    {
      "entropy": 5.404416704177857,
      "epoch": 3.166765519968367,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040154171943974345,
      "loss": 4.9896,
      "mean_token_accuracy": 0.2193855032324791,
      "num_tokens": 73380527.0,
      "step": 32035
    },
    {
      "entropy": 5.370492839813233,
      "epoch": 3.167259786476868,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00040151253800461913,
      "loss": 4.9496,
      "mean_token_accuracy": 0.2171730473637581,
      "num_tokens": 73392547.0,
      "step": 32040
    },
    {
      "entropy": 5.343085718154907,
      "epoch": 3.16775405298537,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004014833534572667,
      "loss": 4.8933,
      "mean_token_accuracy": 0.22689200192689896,
      "num_tokens": 73404293.0,
      "step": 32045
    },
    {
      "entropy": 5.398412609100342,
      "epoch": 3.168248319493871,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004014541657984043,
      "loss": 4.848,
      "mean_token_accuracy": 0.22503772079944612,
      "num_tokens": 73414581.0,
      "step": 32050
    },
    {
      "entropy": 5.384239530563354,
      "epoch": 3.1687425860023724,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040142497502874963,
      "loss": 5.0017,
      "mean_token_accuracy": 0.21043816059827805,
      "num_tokens": 73426485.0,
      "step": 32055
    },
    {
      "entropy": 5.422813844680786,
      "epoch": 3.169236852510874,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040139578114902104,
      "loss": 5.0201,
      "mean_token_accuracy": 0.2133195236325264,
      "num_tokens": 73438481.0,
      "step": 32060
    },
    {
      "entropy": 5.442287635803223,
      "epoch": 3.1697311190193753,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040136658415993663,
      "loss": 4.8985,
      "mean_token_accuracy": 0.21722345054149628,
      "num_tokens": 73448552.0,
      "step": 32065
    },
    {
      "entropy": 5.359316396713257,
      "epoch": 3.1702253855278766,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004013373840622146,
      "loss": 4.905,
      "mean_token_accuracy": 0.21567890793085098,
      "num_tokens": 73459354.0,
      "step": 32070
    },
    {
      "entropy": 5.3637388229370115,
      "epoch": 3.170719652036378,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040130818085657333,
      "loss": 4.9151,
      "mean_token_accuracy": 0.2169932022690773,
      "num_tokens": 73470718.0,
      "step": 32075
    },
    {
      "entropy": 5.4414391040802,
      "epoch": 3.1712139185448796,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004012789745437313,
      "loss": 5.0483,
      "mean_token_accuracy": 0.20973199605941772,
      "num_tokens": 73483133.0,
      "step": 32080
    },
    {
      "entropy": 5.414180326461792,
      "epoch": 3.171708185053381,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004012497651244068,
      "loss": 4.9298,
      "mean_token_accuracy": 0.21286708414554595,
      "num_tokens": 73494675.0,
      "step": 32085
    },
    {
      "entropy": 5.340225172042847,
      "epoch": 3.172202451561882,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040122055259931856,
      "loss": 4.904,
      "mean_token_accuracy": 0.23085100650787355,
      "num_tokens": 73505412.0,
      "step": 32090
    },
    {
      "entropy": 5.293953609466553,
      "epoch": 3.172696718070384,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040119133696918515,
      "loss": 4.811,
      "mean_token_accuracy": 0.23139822334051133,
      "num_tokens": 73515835.0,
      "step": 32095
    },
    {
      "entropy": 5.400915431976318,
      "epoch": 3.173190984578885,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004011621182347252,
      "loss": 4.9264,
      "mean_token_accuracy": 0.22256221771240234,
      "num_tokens": 73526943.0,
      "step": 32100
    },
    {
      "entropy": 5.394599533081054,
      "epoch": 3.1736852510873863,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00040113289639665765,
      "loss": 4.9203,
      "mean_token_accuracy": 0.22208186388015747,
      "num_tokens": 73540076.0,
      "step": 32105
    },
    {
      "entropy": 5.450634717941284,
      "epoch": 3.1741795175958876,
      "grad_norm": 0.9375,
      "learning_rate": 0.00040110367145570125,
      "loss": 4.988,
      "mean_token_accuracy": 0.21733183711767196,
      "num_tokens": 73552265.0,
      "step": 32110
    },
    {
      "entropy": 5.349441719055176,
      "epoch": 3.1746737841043893,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000401074443412575,
      "loss": 4.9203,
      "mean_token_accuracy": 0.22119634747505187,
      "num_tokens": 73562581.0,
      "step": 32115
    },
    {
      "entropy": 5.4000505924224855,
      "epoch": 3.1751680506128905,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040104521226799777,
      "loss": 4.915,
      "mean_token_accuracy": 0.2272097185254097,
      "num_tokens": 73573478.0,
      "step": 32120
    },
    {
      "entropy": 5.436168813705445,
      "epoch": 3.175662317121392,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040101597802268883,
      "loss": 4.9395,
      "mean_token_accuracy": 0.219424007833004,
      "num_tokens": 73584639.0,
      "step": 32125
    },
    {
      "entropy": 5.367035675048828,
      "epoch": 3.176156583629893,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004009867406773673,
      "loss": 4.9673,
      "mean_token_accuracy": 0.2182232677936554,
      "num_tokens": 73596084.0,
      "step": 32130
    },
    {
      "entropy": 5.40050368309021,
      "epoch": 3.1766508501383948,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040095750023275243,
      "loss": 4.9402,
      "mean_token_accuracy": 0.22098592668771744,
      "num_tokens": 73606551.0,
      "step": 32135
    },
    {
      "entropy": 5.395291757583618,
      "epoch": 3.177145116646896,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004009282566895635,
      "loss": 4.8628,
      "mean_token_accuracy": 0.22460100799798965,
      "num_tokens": 73616134.0,
      "step": 32140
    },
    {
      "entropy": 5.320925998687744,
      "epoch": 3.1776393831553973,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004008990100485199,
      "loss": 4.8145,
      "mean_token_accuracy": 0.23015804439783097,
      "num_tokens": 73626813.0,
      "step": 32145
    },
    {
      "entropy": 5.330163621902466,
      "epoch": 3.1781336496638986,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040086976031034104,
      "loss": 4.9287,
      "mean_token_accuracy": 0.223099085688591,
      "num_tokens": 73638205.0,
      "step": 32150
    },
    {
      "entropy": 5.5047728538513185,
      "epoch": 3.1786279161724003,
      "grad_norm": 0.9375,
      "learning_rate": 0.00040084050747574666,
      "loss": 5.0448,
      "mean_token_accuracy": 0.20947058796882628,
      "num_tokens": 73652276.0,
      "step": 32155
    },
    {
      "entropy": 5.452353239059448,
      "epoch": 3.1791221826809015,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004008112515454562,
      "loss": 5.0694,
      "mean_token_accuracy": 0.21353176534175872,
      "num_tokens": 73665516.0,
      "step": 32160
    },
    {
      "entropy": 5.3728560447692875,
      "epoch": 3.179616449189403,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004007819925201895,
      "loss": 4.8885,
      "mean_token_accuracy": 0.23020555675029755,
      "num_tokens": 73675783.0,
      "step": 32165
    },
    {
      "entropy": 5.436239004135132,
      "epoch": 3.1801107156979045,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004007527304006663,
      "loss": 4.9953,
      "mean_token_accuracy": 0.2108194947242737,
      "num_tokens": 73687853.0,
      "step": 32170
    },
    {
      "entropy": 5.372262048721313,
      "epoch": 3.1806049822064058,
      "grad_norm": 1.0,
      "learning_rate": 0.00040072346518760634,
      "loss": 4.8765,
      "mean_token_accuracy": 0.22489214688539505,
      "num_tokens": 73699189.0,
      "step": 32175
    },
    {
      "entropy": 5.449042654037475,
      "epoch": 3.181099248714907,
      "grad_norm": 0.890625,
      "learning_rate": 0.00040069419688172967,
      "loss": 5.0609,
      "mean_token_accuracy": 0.21041840463876724,
      "num_tokens": 73711003.0,
      "step": 32180
    },
    {
      "entropy": 5.454993152618409,
      "epoch": 3.1815935152234083,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040066492548375627,
      "loss": 5.0211,
      "mean_token_accuracy": 0.21032641977071762,
      "num_tokens": 73723311.0,
      "step": 32185
    },
    {
      "entropy": 5.454977083206177,
      "epoch": 3.18208778173191,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040063565099440616,
      "loss": 4.9962,
      "mean_token_accuracy": 0.2210969015955925,
      "num_tokens": 73734155.0,
      "step": 32190
    },
    {
      "entropy": 5.323065567016601,
      "epoch": 3.1825820482404112,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004006063734143996,
      "loss": 4.8956,
      "mean_token_accuracy": 0.22565834522247313,
      "num_tokens": 73745289.0,
      "step": 32195
    },
    {
      "entropy": 5.347533941268921,
      "epoch": 3.1830763147489125,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004005770927444567,
      "loss": 4.9211,
      "mean_token_accuracy": 0.21814856231212615,
      "num_tokens": 73756955.0,
      "step": 32200
    },
    {
      "entropy": 5.418862056732178,
      "epoch": 3.183570581257414,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004005478089852979,
      "loss": 4.9057,
      "mean_token_accuracy": 0.21892037540674208,
      "num_tokens": 73768676.0,
      "step": 32205
    },
    {
      "entropy": 5.334467363357544,
      "epoch": 3.1840648477659155,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004005185221376435,
      "loss": 4.9082,
      "mean_token_accuracy": 0.22764068096876144,
      "num_tokens": 73779505.0,
      "step": 32210
    },
    {
      "entropy": 5.388941049575806,
      "epoch": 3.1845591142744167,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040048923220221397,
      "loss": 4.9573,
      "mean_token_accuracy": 0.21513389497995378,
      "num_tokens": 73790785.0,
      "step": 32215
    },
    {
      "entropy": 5.396496391296386,
      "epoch": 3.185053380782918,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040045993917972985,
      "loss": 4.9242,
      "mean_token_accuracy": 0.21771674305200578,
      "num_tokens": 73802689.0,
      "step": 32220
    },
    {
      "entropy": 5.331830883026123,
      "epoch": 3.1855476472914197,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004004306430709118,
      "loss": 4.9076,
      "mean_token_accuracy": 0.22164315730333328,
      "num_tokens": 73815029.0,
      "step": 32225
    },
    {
      "entropy": 5.4280784130096436,
      "epoch": 3.186041913799921,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040040134387648045,
      "loss": 4.9463,
      "mean_token_accuracy": 0.22016408592462539,
      "num_tokens": 73825509.0,
      "step": 32230
    },
    {
      "entropy": 5.410914611816406,
      "epoch": 3.1865361803084222,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004003720415971566,
      "loss": 4.967,
      "mean_token_accuracy": 0.21901264935731887,
      "num_tokens": 73836014.0,
      "step": 32235
    },
    {
      "entropy": 5.306758117675781,
      "epoch": 3.1870304468169235,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004003427362336611,
      "loss": 4.8711,
      "mean_token_accuracy": 0.2245720788836479,
      "num_tokens": 73847044.0,
      "step": 32240
    },
    {
      "entropy": 5.38459849357605,
      "epoch": 3.187524713325425,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040031342778671476,
      "loss": 4.9229,
      "mean_token_accuracy": 0.2230529084801674,
      "num_tokens": 73859748.0,
      "step": 32245
    },
    {
      "entropy": 5.366387891769409,
      "epoch": 3.1880189798339265,
      "grad_norm": 1.0,
      "learning_rate": 0.0004002841162570387,
      "loss": 4.8874,
      "mean_token_accuracy": 0.22334673702716829,
      "num_tokens": 73870450.0,
      "step": 32250
    },
    {
      "entropy": 5.41474723815918,
      "epoch": 3.1885132463424277,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040025480164535405,
      "loss": 5.0072,
      "mean_token_accuracy": 0.21244468092918395,
      "num_tokens": 73882533.0,
      "step": 32255
    },
    {
      "entropy": 5.441440057754517,
      "epoch": 3.1890075128509294,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040022548395238173,
      "loss": 4.954,
      "mean_token_accuracy": 0.22020991295576095,
      "num_tokens": 73894247.0,
      "step": 32260
    },
    {
      "entropy": 5.396122217178345,
      "epoch": 3.1895017793594307,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004001961631788431,
      "loss": 4.9165,
      "mean_token_accuracy": 0.22600238621234894,
      "num_tokens": 73905246.0,
      "step": 32265
    },
    {
      "entropy": 5.332942390441895,
      "epoch": 3.189996045867932,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040016683932545937,
      "loss": 4.8861,
      "mean_token_accuracy": 0.22444458305835724,
      "num_tokens": 73915906.0,
      "step": 32270
    },
    {
      "entropy": 5.390687894821167,
      "epoch": 3.190490312376433,
      "grad_norm": 1.09375,
      "learning_rate": 0.000400137512392952,
      "loss": 4.9243,
      "mean_token_accuracy": 0.21948201805353165,
      "num_tokens": 73928629.0,
      "step": 32275
    },
    {
      "entropy": 5.47108039855957,
      "epoch": 3.190984578884935,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040010818238204256,
      "loss": 4.9977,
      "mean_token_accuracy": 0.21575254052877427,
      "num_tokens": 73940474.0,
      "step": 32280
    },
    {
      "entropy": 5.367278003692627,
      "epoch": 3.191478845393436,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00040007884929345224,
      "loss": 4.8889,
      "mean_token_accuracy": 0.2212388977408409,
      "num_tokens": 73951636.0,
      "step": 32285
    },
    {
      "entropy": 5.319686937332153,
      "epoch": 3.1919731119019374,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004000495131279029,
      "loss": 4.868,
      "mean_token_accuracy": 0.227410988509655,
      "num_tokens": 73963253.0,
      "step": 32290
    },
    {
      "entropy": 5.337161350250244,
      "epoch": 3.1924673784104387,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004000201738861162,
      "loss": 4.8528,
      "mean_token_accuracy": 0.22322608530521393,
      "num_tokens": 73974215.0,
      "step": 32295
    },
    {
      "entropy": 5.420774650573731,
      "epoch": 3.1929616449189404,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003999908315688137,
      "loss": 4.9595,
      "mean_token_accuracy": 0.22171608954668046,
      "num_tokens": 73986127.0,
      "step": 32300
    },
    {
      "entropy": 5.454508829116821,
      "epoch": 3.1934559114274417,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039996148617671743,
      "loss": 4.941,
      "mean_token_accuracy": 0.22343740016222,
      "num_tokens": 73996998.0,
      "step": 32305
    },
    {
      "entropy": 5.366781949996948,
      "epoch": 3.193950177935943,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003999321377105492,
      "loss": 4.9136,
      "mean_token_accuracy": 0.2248337149620056,
      "num_tokens": 74009432.0,
      "step": 32310
    },
    {
      "entropy": 5.3012172222137455,
      "epoch": 3.1944444444444446,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000399902786171031,
      "loss": 4.921,
      "mean_token_accuracy": 0.21944795846939086,
      "num_tokens": 74021242.0,
      "step": 32315
    },
    {
      "entropy": 5.4300682067871096,
      "epoch": 3.194938710952946,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00039987343155888484,
      "loss": 4.942,
      "mean_token_accuracy": 0.20683617442846297,
      "num_tokens": 74032865.0,
      "step": 32320
    },
    {
      "entropy": 5.389030694961548,
      "epoch": 3.195432977461447,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000399844073874833,
      "loss": 4.8883,
      "mean_token_accuracy": 0.2226030319929123,
      "num_tokens": 74044505.0,
      "step": 32325
    },
    {
      "entropy": 5.3718267440795895,
      "epoch": 3.1959272439699484,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039981471311959746,
      "loss": 4.9364,
      "mean_token_accuracy": 0.21708377301692963,
      "num_tokens": 74056551.0,
      "step": 32330
    },
    {
      "entropy": 5.390479278564453,
      "epoch": 3.19642151047845,
      "grad_norm": 1.0,
      "learning_rate": 0.00039978534929390066,
      "loss": 4.8975,
      "mean_token_accuracy": 0.22459932863712312,
      "num_tokens": 74066746.0,
      "step": 32335
    },
    {
      "entropy": 5.392752027511596,
      "epoch": 3.1969157769869514,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039975598239846494,
      "loss": 4.9516,
      "mean_token_accuracy": 0.21600686758756638,
      "num_tokens": 74078493.0,
      "step": 32340
    },
    {
      "entropy": 5.3416375637054445,
      "epoch": 3.1974100434954527,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039972661243401267,
      "loss": 4.853,
      "mean_token_accuracy": 0.22139085233211517,
      "num_tokens": 74089484.0,
      "step": 32345
    },
    {
      "entropy": 5.442918491363526,
      "epoch": 3.1979043100039544,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003996972394012664,
      "loss": 4.9454,
      "mean_token_accuracy": 0.219377800822258,
      "num_tokens": 74101038.0,
      "step": 32350
    },
    {
      "entropy": 5.473010778427124,
      "epoch": 3.1983985765124556,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003996678633009486,
      "loss": 5.023,
      "mean_token_accuracy": 0.21315684765577317,
      "num_tokens": 74112708.0,
      "step": 32355
    },
    {
      "entropy": 5.396411323547364,
      "epoch": 3.198892843020957,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003996384841337822,
      "loss": 4.9571,
      "mean_token_accuracy": 0.21799690872430802,
      "num_tokens": 74123899.0,
      "step": 32360
    },
    {
      "entropy": 5.357379388809204,
      "epoch": 3.199387109529458,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003996091019004896,
      "loss": 4.9318,
      "mean_token_accuracy": 0.21749658286571502,
      "num_tokens": 74136025.0,
      "step": 32365
    },
    {
      "entropy": 5.470479440689087,
      "epoch": 3.19988137603796,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003995797166017938,
      "loss": 5.0779,
      "mean_token_accuracy": 0.21105027943849564,
      "num_tokens": 74149140.0,
      "step": 32370
    },
    {
      "entropy": 5.478125476837159,
      "epoch": 3.200375642546461,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039955032823841766,
      "loss": 4.9423,
      "mean_token_accuracy": 0.2196016415953636,
      "num_tokens": 74160849.0,
      "step": 32375
    },
    {
      "entropy": 5.428229331970215,
      "epoch": 3.2008699090549624,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039952093681108416,
      "loss": 4.9609,
      "mean_token_accuracy": 0.2232427030801773,
      "num_tokens": 74171678.0,
      "step": 32380
    },
    {
      "entropy": 5.328751564025879,
      "epoch": 3.2013641755634636,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003994915423205162,
      "loss": 4.8935,
      "mean_token_accuracy": 0.2230215400457382,
      "num_tokens": 74183000.0,
      "step": 32385
    },
    {
      "entropy": 5.468348550796509,
      "epoch": 3.2018584420719653,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003994621447674371,
      "loss": 5.0012,
      "mean_token_accuracy": 0.21578442454338073,
      "num_tokens": 74195495.0,
      "step": 32390
    },
    {
      "entropy": 5.429653835296631,
      "epoch": 3.2023527085804666,
      "grad_norm": 1.125,
      "learning_rate": 0.0003994327441525699,
      "loss": 4.9702,
      "mean_token_accuracy": 0.21309554427862168,
      "num_tokens": 74206023.0,
      "step": 32395
    },
    {
      "entropy": 5.399688339233398,
      "epoch": 3.202846975088968,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003994033404766378,
      "loss": 4.9696,
      "mean_token_accuracy": 0.21778280884027482,
      "num_tokens": 74217943.0,
      "step": 32400
    },
    {
      "entropy": 5.422229480743408,
      "epoch": 3.203341241597469,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003993739337403643,
      "loss": 4.9939,
      "mean_token_accuracy": 0.2150809422135353,
      "num_tokens": 74228142.0,
      "step": 32405
    },
    {
      "entropy": 5.455906915664673,
      "epoch": 3.203835508105971,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003993445239444728,
      "loss": 4.9907,
      "mean_token_accuracy": 0.22167946100234986,
      "num_tokens": 74239692.0,
      "step": 32410
    },
    {
      "entropy": 5.426267433166504,
      "epoch": 3.204329774614472,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003993151110896866,
      "loss": 4.9575,
      "mean_token_accuracy": 0.22039847671985627,
      "num_tokens": 74251010.0,
      "step": 32415
    },
    {
      "entropy": 5.314350891113281,
      "epoch": 3.2048240411229734,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003992856951767294,
      "loss": 4.8286,
      "mean_token_accuracy": 0.22832494527101516,
      "num_tokens": 74262579.0,
      "step": 32420
    },
    {
      "entropy": 5.395722723007202,
      "epoch": 3.205318307631475,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039925627620632486,
      "loss": 4.8619,
      "mean_token_accuracy": 0.22499454468488694,
      "num_tokens": 74272833.0,
      "step": 32425
    },
    {
      "entropy": 5.410754442214966,
      "epoch": 3.2058125741399763,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003992268541791967,
      "loss": 4.9854,
      "mean_token_accuracy": 0.21471409350633622,
      "num_tokens": 74284565.0,
      "step": 32430
    },
    {
      "entropy": 5.285041999816895,
      "epoch": 3.2063068406484776,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003991974290960685,
      "loss": 4.8126,
      "mean_token_accuracy": 0.24158210903406144,
      "num_tokens": 74295473.0,
      "step": 32435
    },
    {
      "entropy": 5.42221155166626,
      "epoch": 3.206801107156979,
      "grad_norm": 1.125,
      "learning_rate": 0.00039916800095766443,
      "loss": 5.0117,
      "mean_token_accuracy": 0.20981270521879197,
      "num_tokens": 74305688.0,
      "step": 32440
    },
    {
      "entropy": 5.348579406738281,
      "epoch": 3.2072953736654806,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00039913856976470823,
      "loss": 4.8916,
      "mean_token_accuracy": 0.22253837883472444,
      "num_tokens": 74318268.0,
      "step": 32445
    },
    {
      "entropy": 5.452938032150269,
      "epoch": 3.207789640173982,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003991091355179239,
      "loss": 5.01,
      "mean_token_accuracy": 0.20986150950193405,
      "num_tokens": 74330310.0,
      "step": 32450
    },
    {
      "entropy": 5.404800891876221,
      "epoch": 3.208283906682483,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003990796982180356,
      "loss": 4.9708,
      "mean_token_accuracy": 0.21881929636001587,
      "num_tokens": 74342040.0,
      "step": 32455
    },
    {
      "entropy": 5.476469278335571,
      "epoch": 3.208778173190985,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039905025786576755,
      "loss": 5.0246,
      "mean_token_accuracy": 0.2059953436255455,
      "num_tokens": 74354502.0,
      "step": 32460
    },
    {
      "entropy": 5.45098910331726,
      "epoch": 3.209272439699486,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003990208144618438,
      "loss": 4.9305,
      "mean_token_accuracy": 0.21893686503171922,
      "num_tokens": 74366100.0,
      "step": 32465
    },
    {
      "entropy": 5.459633684158325,
      "epoch": 3.2097667062079873,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039899136800698883,
      "loss": 5.031,
      "mean_token_accuracy": 0.20270186960697173,
      "num_tokens": 74377627.0,
      "step": 32470
    },
    {
      "entropy": 5.34204134941101,
      "epoch": 3.2102609727164886,
      "grad_norm": 1.0,
      "learning_rate": 0.000398961918501927,
      "loss": 4.9088,
      "mean_token_accuracy": 0.22563174664974212,
      "num_tokens": 74389689.0,
      "step": 32475
    },
    {
      "entropy": 5.463394498825073,
      "epoch": 3.2107552392249903,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003989324659473827,
      "loss": 5.0459,
      "mean_token_accuracy": 0.20966026335954666,
      "num_tokens": 74399902.0,
      "step": 32480
    },
    {
      "entropy": 5.4572875022888185,
      "epoch": 3.2112495057334915,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039890301034408056,
      "loss": 4.9974,
      "mean_token_accuracy": 0.21277348548173905,
      "num_tokens": 74410935.0,
      "step": 32485
    },
    {
      "entropy": 5.4490913391113285,
      "epoch": 3.211743772241993,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003988735516927451,
      "loss": 4.9105,
      "mean_token_accuracy": 0.22563932538032533,
      "num_tokens": 74421850.0,
      "step": 32490
    },
    {
      "entropy": 5.371338653564453,
      "epoch": 3.212238038750494,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00039884408999410107,
      "loss": 4.9024,
      "mean_token_accuracy": 0.2234208583831787,
      "num_tokens": 74434104.0,
      "step": 32495
    },
    {
      "entropy": 5.412356615066528,
      "epoch": 3.212732305258996,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003988146252488732,
      "loss": 5.0078,
      "mean_token_accuracy": 0.2166041389107704,
      "num_tokens": 74446738.0,
      "step": 32500
    },
    {
      "entropy": 5.535458421707153,
      "epoch": 3.213226571767497,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003987851574577863,
      "loss": 5.0898,
      "mean_token_accuracy": 0.2095863252878189,
      "num_tokens": 74458202.0,
      "step": 32505
    },
    {
      "entropy": 5.472123146057129,
      "epoch": 3.2137208382759983,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003987556866215654,
      "loss": 4.9999,
      "mean_token_accuracy": 0.21904619336128234,
      "num_tokens": 74470349.0,
      "step": 32510
    },
    {
      "entropy": 5.346045398712159,
      "epoch": 3.2142151047845,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003987262127409354,
      "loss": 4.8714,
      "mean_token_accuracy": 0.2259410172700882,
      "num_tokens": 74482569.0,
      "step": 32515
    },
    {
      "entropy": 5.37182207107544,
      "epoch": 3.2147093712930013,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003986967358166213,
      "loss": 4.9002,
      "mean_token_accuracy": 0.21836068332195283,
      "num_tokens": 74495425.0,
      "step": 32520
    },
    {
      "entropy": 5.460014390945434,
      "epoch": 3.2152036378015025,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039866725584934834,
      "loss": 4.9953,
      "mean_token_accuracy": 0.21120498180389405,
      "num_tokens": 74506429.0,
      "step": 32525
    },
    {
      "entropy": 5.443718290328979,
      "epoch": 3.215697904310004,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039863777283984174,
      "loss": 4.9768,
      "mean_token_accuracy": 0.21571776419878005,
      "num_tokens": 74517426.0,
      "step": 32530
    },
    {
      "entropy": 5.446964931488037,
      "epoch": 3.2161921708185055,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039860828678882674,
      "loss": 4.9676,
      "mean_token_accuracy": 0.21266179233789445,
      "num_tokens": 74528777.0,
      "step": 32535
    },
    {
      "entropy": 5.404529190063476,
      "epoch": 3.2166864373270068,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003985787976970287,
      "loss": 4.917,
      "mean_token_accuracy": 0.21307890564203263,
      "num_tokens": 74540384.0,
      "step": 32540
    },
    {
      "entropy": 5.373161125183105,
      "epoch": 3.217180703835508,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000398549305565173,
      "loss": 4.87,
      "mean_token_accuracy": 0.22038756757974626,
      "num_tokens": 74551698.0,
      "step": 32545
    },
    {
      "entropy": 5.376799631118774,
      "epoch": 3.2176749703440093,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039851981039398527,
      "loss": 4.9433,
      "mean_token_accuracy": 0.2257494166493416,
      "num_tokens": 74563467.0,
      "step": 32550
    },
    {
      "entropy": 5.310831785202026,
      "epoch": 3.218169236852511,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00039849031218419114,
      "loss": 4.8745,
      "mean_token_accuracy": 0.22689485996961595,
      "num_tokens": 74575341.0,
      "step": 32555
    },
    {
      "entropy": 5.330077981948852,
      "epoch": 3.2186635033610123,
      "grad_norm": 1.0,
      "learning_rate": 0.000398460810936516,
      "loss": 4.813,
      "mean_token_accuracy": 0.22998543530702592,
      "num_tokens": 74587706.0,
      "step": 32560
    },
    {
      "entropy": 5.400575828552246,
      "epoch": 3.2191577698695135,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039843130665168585,
      "loss": 4.9246,
      "mean_token_accuracy": 0.22073059231042863,
      "num_tokens": 74599310.0,
      "step": 32565
    },
    {
      "entropy": 5.449219751358032,
      "epoch": 3.2196520363780152,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00039840179933042644,
      "loss": 5.0508,
      "mean_token_accuracy": 0.2097255527973175,
      "num_tokens": 74610436.0,
      "step": 32570
    },
    {
      "entropy": 5.47999792098999,
      "epoch": 3.2201463028865165,
      "grad_norm": 0.984375,
      "learning_rate": 0.00039837228897346357,
      "loss": 4.9256,
      "mean_token_accuracy": 0.21416904032230377,
      "num_tokens": 74621788.0,
      "step": 32575
    },
    {
      "entropy": 5.426194286346435,
      "epoch": 3.2206405693950177,
      "grad_norm": 1.0,
      "learning_rate": 0.00039834277558152326,
      "loss": 4.9523,
      "mean_token_accuracy": 0.22077723741531372,
      "num_tokens": 74633345.0,
      "step": 32580
    },
    {
      "entropy": 5.388727617263794,
      "epoch": 3.221134835903519,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003983132591553315,
      "loss": 4.9749,
      "mean_token_accuracy": 0.21887448132038118,
      "num_tokens": 74646217.0,
      "step": 32585
    },
    {
      "entropy": 5.378500890731812,
      "epoch": 3.2216291024120207,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003982837396956145,
      "loss": 4.943,
      "mean_token_accuracy": 0.22191304862499237,
      "num_tokens": 74657935.0,
      "step": 32590
    },
    {
      "entropy": 5.377217435836792,
      "epoch": 3.222123368920522,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039825421720309835,
      "loss": 4.9778,
      "mean_token_accuracy": 0.21892588287591935,
      "num_tokens": 74668547.0,
      "step": 32595
    },
    {
      "entropy": 5.491641187667847,
      "epoch": 3.2226176354290232,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003982246916785093,
      "loss": 5.0227,
      "mean_token_accuracy": 0.20635757893323897,
      "num_tokens": 74680013.0,
      "step": 32600
    },
    {
      "entropy": 5.432769536972046,
      "epoch": 3.223111901937525,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003981951631225738,
      "loss": 4.9012,
      "mean_token_accuracy": 0.22637124061584474,
      "num_tokens": 74689927.0,
      "step": 32605
    },
    {
      "entropy": 5.396711826324463,
      "epoch": 3.223606168446026,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003981656315360181,
      "loss": 4.9684,
      "mean_token_accuracy": 0.21589536666870118,
      "num_tokens": 74701854.0,
      "step": 32610
    },
    {
      "entropy": 5.356378984451294,
      "epoch": 3.2241004349545275,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039813609691956883,
      "loss": 4.9306,
      "mean_token_accuracy": 0.220461343228817,
      "num_tokens": 74713974.0,
      "step": 32615
    },
    {
      "entropy": 5.456284427642823,
      "epoch": 3.2245947014630287,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003981065592739524,
      "loss": 5.0203,
      "mean_token_accuracy": 0.21066245436668396,
      "num_tokens": 74725367.0,
      "step": 32620
    },
    {
      "entropy": 5.452782535552979,
      "epoch": 3.2250889679715304,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039807701859989554,
      "loss": 4.9356,
      "mean_token_accuracy": 0.2184978798031807,
      "num_tokens": 74736801.0,
      "step": 32625
    },
    {
      "entropy": 5.364739561080933,
      "epoch": 3.2255832344800317,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039804747489812495,
      "loss": 4.929,
      "mean_token_accuracy": 0.2210173040628433,
      "num_tokens": 74747949.0,
      "step": 32630
    },
    {
      "entropy": 5.343549156188965,
      "epoch": 3.226077500988533,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003980179281693674,
      "loss": 4.9228,
      "mean_token_accuracy": 0.2271590694785118,
      "num_tokens": 74759985.0,
      "step": 32635
    },
    {
      "entropy": 5.386345243453979,
      "epoch": 3.2265717674970342,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039798837841434966,
      "loss": 4.9059,
      "mean_token_accuracy": 0.21878810673952104,
      "num_tokens": 74770981.0,
      "step": 32640
    },
    {
      "entropy": 5.3279430866241455,
      "epoch": 3.227066034005536,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039795882563379876,
      "loss": 4.8664,
      "mean_token_accuracy": 0.23160792589187623,
      "num_tokens": 74783120.0,
      "step": 32645
    },
    {
      "entropy": 5.31374454498291,
      "epoch": 3.227560300514037,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003979292698284417,
      "loss": 4.8478,
      "mean_token_accuracy": 0.22931768149137496,
      "num_tokens": 74793492.0,
      "step": 32650
    },
    {
      "entropy": 5.319516801834107,
      "epoch": 3.2280545670225385,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039789971099900545,
      "loss": 4.9108,
      "mean_token_accuracy": 0.22170383483171463,
      "num_tokens": 74805333.0,
      "step": 32655
    },
    {
      "entropy": 5.377570867538452,
      "epoch": 3.2285488335310397,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003978701491462174,
      "loss": 4.8781,
      "mean_token_accuracy": 0.22457063049077988,
      "num_tokens": 74816269.0,
      "step": 32660
    },
    {
      "entropy": 5.40594072341919,
      "epoch": 3.2290431000395414,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003978405842708045,
      "loss": 4.8964,
      "mean_token_accuracy": 0.226781465113163,
      "num_tokens": 74826901.0,
      "step": 32665
    },
    {
      "entropy": 5.355334234237671,
      "epoch": 3.2295373665480427,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039781101637349417,
      "loss": 4.9403,
      "mean_token_accuracy": 0.22345403879880904,
      "num_tokens": 74837373.0,
      "step": 32670
    },
    {
      "entropy": 5.40598201751709,
      "epoch": 3.230031633056544,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039778144545501374,
      "loss": 4.9128,
      "mean_token_accuracy": 0.21659032851457596,
      "num_tokens": 74848609.0,
      "step": 32675
    },
    {
      "entropy": 5.3725837707519535,
      "epoch": 3.2305258995650457,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003977518715160908,
      "loss": 4.9401,
      "mean_token_accuracy": 0.21664578169584275,
      "num_tokens": 74859953.0,
      "step": 32680
    },
    {
      "entropy": 5.406901454925537,
      "epoch": 3.231020166073547,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039772229455745285,
      "loss": 4.9444,
      "mean_token_accuracy": 0.21986195743083953,
      "num_tokens": 74872173.0,
      "step": 32685
    },
    {
      "entropy": 5.391617774963379,
      "epoch": 3.231514432582048,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039769271457982733,
      "loss": 4.8884,
      "mean_token_accuracy": 0.22647606134414672,
      "num_tokens": 74884060.0,
      "step": 32690
    },
    {
      "entropy": 5.391271114349365,
      "epoch": 3.2320086990905494,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003976631315839419,
      "loss": 4.9213,
      "mean_token_accuracy": 0.22046476751565933,
      "num_tokens": 74895660.0,
      "step": 32695
    },
    {
      "entropy": 5.394249391555786,
      "epoch": 3.232502965599051,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003976335455705245,
      "loss": 4.9339,
      "mean_token_accuracy": 0.2172132357954979,
      "num_tokens": 74907228.0,
      "step": 32700
    },
    {
      "entropy": 5.368260860443115,
      "epoch": 3.2329972321075524,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003976039565403029,
      "loss": 4.9667,
      "mean_token_accuracy": 0.21886109411716462,
      "num_tokens": 74919062.0,
      "step": 32705
    },
    {
      "entropy": 5.354167509078979,
      "epoch": 3.2334914986160537,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039757436449400484,
      "loss": 4.887,
      "mean_token_accuracy": 0.22820591777563096,
      "num_tokens": 74930551.0,
      "step": 32710
    },
    {
      "entropy": 5.388807725906372,
      "epoch": 3.2339857651245554,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003975447694323585,
      "loss": 4.8764,
      "mean_token_accuracy": 0.22623901218175888,
      "num_tokens": 74941219.0,
      "step": 32715
    },
    {
      "entropy": 5.4520079612731935,
      "epoch": 3.2344800316330566,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003975151713560918,
      "loss": 4.9636,
      "mean_token_accuracy": 0.21810445487499236,
      "num_tokens": 74951857.0,
      "step": 32720
    },
    {
      "entropy": 5.327975082397461,
      "epoch": 3.234974298141558,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003974855702659328,
      "loss": 4.7741,
      "mean_token_accuracy": 0.22978077381849288,
      "num_tokens": 74963807.0,
      "step": 32725
    },
    {
      "entropy": 5.303542423248291,
      "epoch": 3.235468564650059,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039745596616260983,
      "loss": 4.8858,
      "mean_token_accuracy": 0.22933170795440674,
      "num_tokens": 74975646.0,
      "step": 32730
    },
    {
      "entropy": 5.387570381164551,
      "epoch": 3.235962831158561,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003974263590468511,
      "loss": 4.9939,
      "mean_token_accuracy": 0.2190198227763176,
      "num_tokens": 74987197.0,
      "step": 32735
    },
    {
      "entropy": 5.375151300430298,
      "epoch": 3.236457097667062,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003973967489193849,
      "loss": 4.934,
      "mean_token_accuracy": 0.22435315549373627,
      "num_tokens": 74999020.0,
      "step": 32740
    },
    {
      "entropy": 5.415841960906983,
      "epoch": 3.2369513641755634,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039736713578093975,
      "loss": 4.9683,
      "mean_token_accuracy": 0.22276031076908112,
      "num_tokens": 75010889.0,
      "step": 32745
    },
    {
      "entropy": 5.421621036529541,
      "epoch": 3.2374456306840647,
      "grad_norm": 1.0,
      "learning_rate": 0.000397337519632244,
      "loss": 4.9601,
      "mean_token_accuracy": 0.21610680222511292,
      "num_tokens": 75021934.0,
      "step": 32750
    },
    {
      "entropy": 5.303825426101684,
      "epoch": 3.2379398971925664,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039730790047402635,
      "loss": 4.8692,
      "mean_token_accuracy": 0.2264394298195839,
      "num_tokens": 75032337.0,
      "step": 32755
    },
    {
      "entropy": 5.374797630310058,
      "epoch": 3.2384341637010676,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039727827830701533,
      "loss": 4.8576,
      "mean_token_accuracy": 0.22494217455387117,
      "num_tokens": 75043094.0,
      "step": 32760
    },
    {
      "entropy": 5.332032632827759,
      "epoch": 3.238928430209569,
      "grad_norm": 1.125,
      "learning_rate": 0.00039724865313193976,
      "loss": 4.8684,
      "mean_token_accuracy": 0.22277786135673522,
      "num_tokens": 75055108.0,
      "step": 32765
    },
    {
      "entropy": 5.3155396461486815,
      "epoch": 3.23942269671807,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003972190249495283,
      "loss": 4.8991,
      "mean_token_accuracy": 0.2223450005054474,
      "num_tokens": 75065388.0,
      "step": 32770
    },
    {
      "entropy": 5.408094120025635,
      "epoch": 3.239916963226572,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003971893937605099,
      "loss": 4.9246,
      "mean_token_accuracy": 0.21939627826213837,
      "num_tokens": 75077390.0,
      "step": 32775
    },
    {
      "entropy": 5.420478439331054,
      "epoch": 3.240411229735073,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039715975956561343,
      "loss": 4.9099,
      "mean_token_accuracy": 0.21839433461427687,
      "num_tokens": 75088795.0,
      "step": 32780
    },
    {
      "entropy": 5.361070823669434,
      "epoch": 3.2409054962435744,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039713012236556797,
      "loss": 4.9693,
      "mean_token_accuracy": 0.2168997898697853,
      "num_tokens": 75100336.0,
      "step": 32785
    },
    {
      "entropy": 5.428211641311646,
      "epoch": 3.241399762752076,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003971004821611025,
      "loss": 4.9762,
      "mean_token_accuracy": 0.21381666362285615,
      "num_tokens": 75111639.0,
      "step": 32790
    },
    {
      "entropy": 5.368367099761963,
      "epoch": 3.2418940292605773,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039707083895294625,
      "loss": 4.8383,
      "mean_token_accuracy": 0.2314461812376976,
      "num_tokens": 75122006.0,
      "step": 32795
    },
    {
      "entropy": 5.407904529571534,
      "epoch": 3.2423882957690786,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039704119274182835,
      "loss": 4.9763,
      "mean_token_accuracy": 0.21481959968805314,
      "num_tokens": 75132406.0,
      "step": 32800
    },
    {
      "entropy": 5.361754369735718,
      "epoch": 3.24288256227758,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003970115435284783,
      "loss": 4.9176,
      "mean_token_accuracy": 0.22165888249874116,
      "num_tokens": 75145444.0,
      "step": 32805
    },
    {
      "entropy": 5.394265174865723,
      "epoch": 3.2433768287860816,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039698189131362526,
      "loss": 4.8663,
      "mean_token_accuracy": 0.22916646748781205,
      "num_tokens": 75156069.0,
      "step": 32810
    },
    {
      "entropy": 5.3999810218811035,
      "epoch": 3.243871095294583,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039695223609799885,
      "loss": 4.9563,
      "mean_token_accuracy": 0.21861810535192489,
      "num_tokens": 75167075.0,
      "step": 32815
    },
    {
      "entropy": 5.324251461029053,
      "epoch": 3.244365361803084,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003969225778823285,
      "loss": 4.8574,
      "mean_token_accuracy": 0.22367113679647446,
      "num_tokens": 75178464.0,
      "step": 32820
    },
    {
      "entropy": 5.293191242218017,
      "epoch": 3.244859628311586,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003968929166673438,
      "loss": 4.7987,
      "mean_token_accuracy": 0.22804713249206543,
      "num_tokens": 75189803.0,
      "step": 32825
    },
    {
      "entropy": 5.400164175033569,
      "epoch": 3.245353894820087,
      "grad_norm": 1.0,
      "learning_rate": 0.00039686325245377437,
      "loss": 4.9826,
      "mean_token_accuracy": 0.2165562093257904,
      "num_tokens": 75201207.0,
      "step": 32830
    },
    {
      "entropy": 5.417500782012939,
      "epoch": 3.2458481613285883,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003968335852423501,
      "loss": 5.0082,
      "mean_token_accuracy": 0.21177451759576799,
      "num_tokens": 75214092.0,
      "step": 32835
    },
    {
      "entropy": 5.390856313705444,
      "epoch": 3.2463424278370896,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003968039150338007,
      "loss": 4.8983,
      "mean_token_accuracy": 0.21975431740283966,
      "num_tokens": 75226139.0,
      "step": 32840
    },
    {
      "entropy": 5.395218467712402,
      "epoch": 3.2468366943455913,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039677424182885606,
      "loss": 4.8732,
      "mean_token_accuracy": 0.22600615471601487,
      "num_tokens": 75238273.0,
      "step": 32845
    },
    {
      "entropy": 5.403037071228027,
      "epoch": 3.2473309608540926,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003967445656282462,
      "loss": 4.9977,
      "mean_token_accuracy": 0.21605613827705383,
      "num_tokens": 75250356.0,
      "step": 32850
    },
    {
      "entropy": 5.399277591705323,
      "epoch": 3.247825227362594,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003967148864327012,
      "loss": 4.9335,
      "mean_token_accuracy": 0.21497711688280105,
      "num_tokens": 75260329.0,
      "step": 32855
    },
    {
      "entropy": 5.408873128890991,
      "epoch": 3.248319493871095,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003966852042429511,
      "loss": 4.9941,
      "mean_token_accuracy": 0.213995960354805,
      "num_tokens": 75271405.0,
      "step": 32860
    },
    {
      "entropy": 5.459861183166504,
      "epoch": 3.248813760379597,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039665551905972607,
      "loss": 5.077,
      "mean_token_accuracy": 0.20924915373325348,
      "num_tokens": 75283033.0,
      "step": 32865
    },
    {
      "entropy": 5.442206907272339,
      "epoch": 3.249308026888098,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003966258308837564,
      "loss": 4.935,
      "mean_token_accuracy": 0.21764641851186753,
      "num_tokens": 75293582.0,
      "step": 32870
    },
    {
      "entropy": 5.385214567184448,
      "epoch": 3.2498022933965993,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003965961397157724,
      "loss": 4.9321,
      "mean_token_accuracy": 0.22123805284500123,
      "num_tokens": 75306500.0,
      "step": 32875
    },
    {
      "entropy": 5.36683874130249,
      "epoch": 3.250296559905101,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039656644555650456,
      "loss": 4.8915,
      "mean_token_accuracy": 0.224776192009449,
      "num_tokens": 75318562.0,
      "step": 32880
    },
    {
      "entropy": 5.389555597305298,
      "epoch": 3.2507908264136023,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039653674840668326,
      "loss": 4.9689,
      "mean_token_accuracy": 0.21936971694231033,
      "num_tokens": 75330211.0,
      "step": 32885
    },
    {
      "entropy": 5.388100242614746,
      "epoch": 3.2512850929221035,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003965070482670391,
      "loss": 4.8617,
      "mean_token_accuracy": 0.22920366525650024,
      "num_tokens": 75341244.0,
      "step": 32890
    },
    {
      "entropy": 5.3360130310058596,
      "epoch": 3.251779359430605,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003964773451383027,
      "loss": 4.8599,
      "mean_token_accuracy": 0.22387697398662568,
      "num_tokens": 75353381.0,
      "step": 32895
    },
    {
      "entropy": 5.351694583892822,
      "epoch": 3.2522736259391065,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039644763902120473,
      "loss": 4.8789,
      "mean_token_accuracy": 0.22591210454702376,
      "num_tokens": 75364777.0,
      "step": 32900
    },
    {
      "entropy": 5.389423179626465,
      "epoch": 3.252767892447608,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039641792991647605,
      "loss": 4.9414,
      "mean_token_accuracy": 0.22352657467126846,
      "num_tokens": 75375677.0,
      "step": 32905
    },
    {
      "entropy": 5.427908802032471,
      "epoch": 3.253262158956109,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003963882178248475,
      "loss": 4.9536,
      "mean_token_accuracy": 0.2233499363064766,
      "num_tokens": 75387925.0,
      "step": 32910
    },
    {
      "entropy": 5.283987331390381,
      "epoch": 3.2537564254646103,
      "grad_norm": 1.0,
      "learning_rate": 0.0003963585027470498,
      "loss": 4.7727,
      "mean_token_accuracy": 0.23631056696176528,
      "num_tokens": 75399000.0,
      "step": 32915
    },
    {
      "entropy": 5.327047300338745,
      "epoch": 3.254250691973112,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003963287846838142,
      "loss": 4.8504,
      "mean_token_accuracy": 0.22150594145059585,
      "num_tokens": 75409930.0,
      "step": 32920
    },
    {
      "entropy": 5.407949209213257,
      "epoch": 3.2547449584816133,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039629906363587177,
      "loss": 4.9509,
      "mean_token_accuracy": 0.21806704998016357,
      "num_tokens": 75421621.0,
      "step": 32925
    },
    {
      "entropy": 5.483580493927002,
      "epoch": 3.2552392249901145,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039626933960395345,
      "loss": 5.0444,
      "mean_token_accuracy": 0.21340982466936112,
      "num_tokens": 75434339.0,
      "step": 32930
    },
    {
      "entropy": 5.384823799133301,
      "epoch": 3.2557334914986162,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039623961258879057,
      "loss": 4.9982,
      "mean_token_accuracy": 0.22143218964338302,
      "num_tokens": 75446720.0,
      "step": 32935
    },
    {
      "entropy": 5.337284708023072,
      "epoch": 3.2562277580071175,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003962098825911144,
      "loss": 4.8874,
      "mean_token_accuracy": 0.22609578669071198,
      "num_tokens": 75459354.0,
      "step": 32940
    },
    {
      "entropy": 5.454944658279419,
      "epoch": 3.2567220245156188,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039618014961165633,
      "loss": 5.0068,
      "mean_token_accuracy": 0.21245534121990203,
      "num_tokens": 75471704.0,
      "step": 32945
    },
    {
      "entropy": 5.479271316528321,
      "epoch": 3.25721629102412,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003961504136511478,
      "loss": 5.0383,
      "mean_token_accuracy": 0.21522038131952287,
      "num_tokens": 75483145.0,
      "step": 32950
    },
    {
      "entropy": 5.387699460983276,
      "epoch": 3.2577105575326217,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039612067471032024,
      "loss": 4.893,
      "mean_token_accuracy": 0.22120036631822587,
      "num_tokens": 75493678.0,
      "step": 32955
    },
    {
      "entropy": 5.362654542922973,
      "epoch": 3.258204824041123,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003960909327899053,
      "loss": 4.8968,
      "mean_token_accuracy": 0.22074384093284607,
      "num_tokens": 75504581.0,
      "step": 32960
    },
    {
      "entropy": 5.335285758972168,
      "epoch": 3.2586990905496243,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003960611878906347,
      "loss": 4.8394,
      "mean_token_accuracy": 0.2247749850153923,
      "num_tokens": 75515116.0,
      "step": 32965
    },
    {
      "entropy": 5.341491842269898,
      "epoch": 3.259193357058126,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003960314400132401,
      "loss": 4.8567,
      "mean_token_accuracy": 0.22782252430915834,
      "num_tokens": 75527001.0,
      "step": 32970
    },
    {
      "entropy": 5.400572299957275,
      "epoch": 3.2596876235666272,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003960016891584532,
      "loss": 4.9137,
      "mean_token_accuracy": 0.22772835046052933,
      "num_tokens": 75537641.0,
      "step": 32975
    },
    {
      "entropy": 5.349691247940063,
      "epoch": 3.2601818900751285,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003959719353270061,
      "loss": 4.9123,
      "mean_token_accuracy": 0.2220082625746727,
      "num_tokens": 75551225.0,
      "step": 32980
    },
    {
      "entropy": 5.296876382827759,
      "epoch": 3.2606761565836297,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039594217851963054,
      "loss": 4.877,
      "mean_token_accuracy": 0.22837100327014923,
      "num_tokens": 75563112.0,
      "step": 32985
    },
    {
      "entropy": 5.326445865631103,
      "epoch": 3.2611704230921315,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039591241873705864,
      "loss": 4.8494,
      "mean_token_accuracy": 0.22929680496454238,
      "num_tokens": 75573709.0,
      "step": 32990
    },
    {
      "entropy": 5.394075059890747,
      "epoch": 3.2616646896006327,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003958826559800225,
      "loss": 4.9538,
      "mean_token_accuracy": 0.22167688757181167,
      "num_tokens": 75585269.0,
      "step": 32995
    },
    {
      "entropy": 5.396078062057495,
      "epoch": 3.262158956109134,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003958528902492543,
      "loss": 4.9532,
      "mean_token_accuracy": 0.2148666575551033,
      "num_tokens": 75597084.0,
      "step": 33000
    },
    {
      "epoch": 3.262158956109134,
      "eval_entropy": 5.189067420924282,
      "eval_loss": 5.038053035736084,
      "eval_mean_token_accuracy": 0.22311703603609315,
      "eval_num_tokens": 75597084.0,
      "eval_runtime": 26.9434,
      "eval_samples_per_second": 1206.715,
      "eval_steps_per_second": 150.872,
      "step": 33000
    },
    {
      "entropy": 5.333249092102051,
      "epoch": 3.2626532226176352,
      "grad_norm": 0.984375,
      "learning_rate": 0.00039582312154548614,
      "loss": 4.8716,
      "mean_token_accuracy": 0.23040277510881424,
      "num_tokens": 75608948.0,
      "step": 33005
    },
    {
      "entropy": 5.396775484085083,
      "epoch": 3.263147489126137,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039579334986945055,
      "loss": 4.9344,
      "mean_token_accuracy": 0.2203484371304512,
      "num_tokens": 75619845.0,
      "step": 33010
    },
    {
      "entropy": 5.401003646850586,
      "epoch": 3.263641755634638,
      "grad_norm": 1.0,
      "learning_rate": 0.0003957635752218797,
      "loss": 5.0082,
      "mean_token_accuracy": 0.21123508512973785,
      "num_tokens": 75632116.0,
      "step": 33015
    },
    {
      "entropy": 5.377741050720215,
      "epoch": 3.2641360221431395,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003957337976035063,
      "loss": 4.9135,
      "mean_token_accuracy": 0.2185782015323639,
      "num_tokens": 75643675.0,
      "step": 33020
    },
    {
      "entropy": 5.434565830230713,
      "epoch": 3.2646302886516407,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003957040170150626,
      "loss": 4.9373,
      "mean_token_accuracy": 0.22162001430988312,
      "num_tokens": 75654451.0,
      "step": 33025
    },
    {
      "entropy": 5.306909465789795,
      "epoch": 3.2651245551601424,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003956742334572815,
      "loss": 4.833,
      "mean_token_accuracy": 0.23065955936908722,
      "num_tokens": 75666260.0,
      "step": 33030
    },
    {
      "entropy": 5.3937944889068605,
      "epoch": 3.2656188216686437,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003956444469308955,
      "loss": 4.9523,
      "mean_token_accuracy": 0.21924472153186797,
      "num_tokens": 75678785.0,
      "step": 33035
    },
    {
      "entropy": 5.376312351226806,
      "epoch": 3.266113088177145,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003956146574366374,
      "loss": 4.8952,
      "mean_token_accuracy": 0.22618096619844436,
      "num_tokens": 75690500.0,
      "step": 33040
    },
    {
      "entropy": 5.2899340152740475,
      "epoch": 3.2666073546856467,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003955848649752399,
      "loss": 4.8799,
      "mean_token_accuracy": 0.2316228747367859,
      "num_tokens": 75703616.0,
      "step": 33045
    },
    {
      "entropy": 5.330752801895142,
      "epoch": 3.267101621194148,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039555506954743613,
      "loss": 4.883,
      "mean_token_accuracy": 0.22277265340089797,
      "num_tokens": 75713996.0,
      "step": 33050
    },
    {
      "entropy": 5.447803068161011,
      "epoch": 3.267595887702649,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039552527115395896,
      "loss": 5.0082,
      "mean_token_accuracy": 0.2129393383860588,
      "num_tokens": 75726746.0,
      "step": 33055
    },
    {
      "entropy": 5.447826862335205,
      "epoch": 3.2680901542111505,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039549546979554133,
      "loss": 5.0132,
      "mean_token_accuracy": 0.21259796172380446,
      "num_tokens": 75738441.0,
      "step": 33060
    },
    {
      "entropy": 5.328391361236572,
      "epoch": 3.268584420719652,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003954656654729166,
      "loss": 4.9824,
      "mean_token_accuracy": 0.21579799056053162,
      "num_tokens": 75749592.0,
      "step": 33065
    },
    {
      "entropy": 5.501325416564941,
      "epoch": 3.2690786872281534,
      "grad_norm": 1.0,
      "learning_rate": 0.00039543585818681764,
      "loss": 5.082,
      "mean_token_accuracy": 0.20892251580953597,
      "num_tokens": 75761668.0,
      "step": 33070
    },
    {
      "entropy": 5.450222539901733,
      "epoch": 3.2695729537366547,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003954060479379779,
      "loss": 4.9634,
      "mean_token_accuracy": 0.21851823031902312,
      "num_tokens": 75773277.0,
      "step": 33075
    },
    {
      "entropy": 5.334537601470947,
      "epoch": 3.2700672202451564,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003953762347271308,
      "loss": 4.9197,
      "mean_token_accuracy": 0.22264908403158187,
      "num_tokens": 75783888.0,
      "step": 33080
    },
    {
      "entropy": 5.31234040260315,
      "epoch": 3.2705614867536577,
      "grad_norm": 0.984375,
      "learning_rate": 0.00039534641855500964,
      "loss": 4.918,
      "mean_token_accuracy": 0.22322579473257065,
      "num_tokens": 75796157.0,
      "step": 33085
    },
    {
      "entropy": 5.315906238555908,
      "epoch": 3.271055753262159,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039531659942234785,
      "loss": 4.7878,
      "mean_token_accuracy": 0.2390444815158844,
      "num_tokens": 75807523.0,
      "step": 33090
    },
    {
      "entropy": 5.434315299987793,
      "epoch": 3.27155001977066,
      "grad_norm": 0.984375,
      "learning_rate": 0.00039528677732987913,
      "loss": 4.9892,
      "mean_token_accuracy": 0.21224308758974075,
      "num_tokens": 75819383.0,
      "step": 33095
    },
    {
      "entropy": 5.355528879165649,
      "epoch": 3.272044286279162,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000395256952278337,
      "loss": 4.9885,
      "mean_token_accuracy": 0.22014871686697007,
      "num_tokens": 75830957.0,
      "step": 33100
    },
    {
      "entropy": 5.398595809936523,
      "epoch": 3.272538552787663,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039522712426845507,
      "loss": 4.9023,
      "mean_token_accuracy": 0.22013138830661774,
      "num_tokens": 75841260.0,
      "step": 33105
    },
    {
      "entropy": 5.41156964302063,
      "epoch": 3.2730328192961644,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039519729330096735,
      "loss": 4.9388,
      "mean_token_accuracy": 0.22359984666109084,
      "num_tokens": 75852803.0,
      "step": 33110
    },
    {
      "entropy": 5.425392150878906,
      "epoch": 3.273527085804666,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003951674593766076,
      "loss": 4.9389,
      "mean_token_accuracy": 0.21920476257801055,
      "num_tokens": 75863989.0,
      "step": 33115
    },
    {
      "entropy": 5.333084201812744,
      "epoch": 3.2740213523131674,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003951376224961096,
      "loss": 4.8652,
      "mean_token_accuracy": 0.22816912680864335,
      "num_tokens": 75875381.0,
      "step": 33120
    },
    {
      "entropy": 5.321984958648682,
      "epoch": 3.2745156188216686,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003951077826602075,
      "loss": 4.8376,
      "mean_token_accuracy": 0.2276059091091156,
      "num_tokens": 75887204.0,
      "step": 33125
    },
    {
      "entropy": 5.356298351287842,
      "epoch": 3.27500988533017,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003950779398696353,
      "loss": 4.9493,
      "mean_token_accuracy": 0.21966141909360887,
      "num_tokens": 75898666.0,
      "step": 33130
    },
    {
      "entropy": 5.359775733947754,
      "epoch": 3.275504151838671,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039504809412512724,
      "loss": 4.8834,
      "mean_token_accuracy": 0.22104412615299224,
      "num_tokens": 75910151.0,
      "step": 33135
    },
    {
      "entropy": 5.394236993789673,
      "epoch": 3.275998418347173,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003950182454274173,
      "loss": 4.9513,
      "mean_token_accuracy": 0.22307956963777542,
      "num_tokens": 75920533.0,
      "step": 33140
    },
    {
      "entropy": 5.400747680664063,
      "epoch": 3.276492684855674,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00039498839377723996,
      "loss": 4.8987,
      "mean_token_accuracy": 0.2286425918340683,
      "num_tokens": 75932335.0,
      "step": 33145
    },
    {
      "entropy": 5.472214078903198,
      "epoch": 3.2769869513641754,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003949585391753295,
      "loss": 4.9762,
      "mean_token_accuracy": 0.222672675549984,
      "num_tokens": 75943741.0,
      "step": 33150
    },
    {
      "entropy": 5.323811674118042,
      "epoch": 3.277481217872677,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003949286816224204,
      "loss": 4.9179,
      "mean_token_accuracy": 0.22773498743772508,
      "num_tokens": 75955755.0,
      "step": 33155
    },
    {
      "entropy": 5.372470188140869,
      "epoch": 3.2779754843811784,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003948988211192471,
      "loss": 4.8702,
      "mean_token_accuracy": 0.22397494316101074,
      "num_tokens": 75967057.0,
      "step": 33160
    },
    {
      "entropy": 5.342099380493164,
      "epoch": 3.2784697508896796,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003948689576665443,
      "loss": 4.9184,
      "mean_token_accuracy": 0.22534862905740738,
      "num_tokens": 75978367.0,
      "step": 33165
    },
    {
      "entropy": 5.464762020111084,
      "epoch": 3.278964017398181,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039483909126504643,
      "loss": 5.0162,
      "mean_token_accuracy": 0.21369384378194808,
      "num_tokens": 75990738.0,
      "step": 33170
    },
    {
      "entropy": 5.405552721023559,
      "epoch": 3.2794582839066826,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003948092219154883,
      "loss": 4.9756,
      "mean_token_accuracy": 0.2139458552002907,
      "num_tokens": 76003058.0,
      "step": 33175
    },
    {
      "entropy": 5.37271203994751,
      "epoch": 3.279952550415184,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003947793496186048,
      "loss": 4.8752,
      "mean_token_accuracy": 0.22626155912876128,
      "num_tokens": 76013743.0,
      "step": 33180
    },
    {
      "entropy": 5.414726066589355,
      "epoch": 3.280446816923685,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003947494743751307,
      "loss": 4.9461,
      "mean_token_accuracy": 0.21824132651090622,
      "num_tokens": 76024346.0,
      "step": 33185
    },
    {
      "entropy": 5.421266269683838,
      "epoch": 3.280941083432187,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039471959618580095,
      "loss": 4.9412,
      "mean_token_accuracy": 0.21763127148151398,
      "num_tokens": 76035444.0,
      "step": 33190
    },
    {
      "entropy": 5.280162906646728,
      "epoch": 3.281435349940688,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003946897150513506,
      "loss": 4.8708,
      "mean_token_accuracy": 0.23251291066408158,
      "num_tokens": 76047363.0,
      "step": 33195
    },
    {
      "entropy": 5.3771833896636965,
      "epoch": 3.2819296164491893,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003946598309725146,
      "loss": 4.9583,
      "mean_token_accuracy": 0.21591981649398803,
      "num_tokens": 76059581.0,
      "step": 33200
    },
    {
      "entropy": 5.366854286193847,
      "epoch": 3.2824238829576906,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003946299439500283,
      "loss": 4.8647,
      "mean_token_accuracy": 0.23014160692691804,
      "num_tokens": 76071455.0,
      "step": 33205
    },
    {
      "entropy": 5.307225704193115,
      "epoch": 3.2829181494661923,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003946000539846267,
      "loss": 4.8654,
      "mean_token_accuracy": 0.2253640830516815,
      "num_tokens": 76083794.0,
      "step": 33210
    },
    {
      "entropy": 5.4104081153869625,
      "epoch": 3.2834124159746936,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003945701610770453,
      "loss": 4.9444,
      "mean_token_accuracy": 0.21654590368270873,
      "num_tokens": 76095690.0,
      "step": 33215
    },
    {
      "entropy": 5.342202615737915,
      "epoch": 3.283906682483195,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003945402652280193,
      "loss": 4.8679,
      "mean_token_accuracy": 0.22139070332050323,
      "num_tokens": 76106730.0,
      "step": 33220
    },
    {
      "entropy": 5.3227495670318605,
      "epoch": 3.2844009489916965,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003945103664382844,
      "loss": 4.9515,
      "mean_token_accuracy": 0.21724514663219452,
      "num_tokens": 76117333.0,
      "step": 33225
    },
    {
      "entropy": 5.347023677825928,
      "epoch": 3.284895215500198,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003944804647085759,
      "loss": 4.8623,
      "mean_token_accuracy": 0.22841713279485704,
      "num_tokens": 76129060.0,
      "step": 33230
    },
    {
      "entropy": 5.354047203063965,
      "epoch": 3.285389482008699,
      "grad_norm": 1.125,
      "learning_rate": 0.0003944505600396294,
      "loss": 4.8513,
      "mean_token_accuracy": 0.2321026399731636,
      "num_tokens": 76140724.0,
      "step": 33235
    },
    {
      "entropy": 5.276392889022827,
      "epoch": 3.2858837485172003,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003944206524321805,
      "loss": 4.8738,
      "mean_token_accuracy": 0.230803744494915,
      "num_tokens": 76153021.0,
      "step": 33240
    },
    {
      "entropy": 5.359737825393677,
      "epoch": 3.286378015025702,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039439074188696516,
      "loss": 4.9115,
      "mean_token_accuracy": 0.2254365250468254,
      "num_tokens": 76165759.0,
      "step": 33245
    },
    {
      "entropy": 5.418773746490478,
      "epoch": 3.2868722815342033,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003943608284047189,
      "loss": 4.9172,
      "mean_token_accuracy": 0.21975175738334657,
      "num_tokens": 76178645.0,
      "step": 33250
    },
    {
      "entropy": 5.376545047760009,
      "epoch": 3.2873665480427046,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003943309119861779,
      "loss": 4.875,
      "mean_token_accuracy": 0.22064872980117797,
      "num_tokens": 76189320.0,
      "step": 33255
    },
    {
      "entropy": 5.433485460281372,
      "epoch": 3.287860814551206,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003943009926320778,
      "loss": 5.0367,
      "mean_token_accuracy": 0.2148431658744812,
      "num_tokens": 76200672.0,
      "step": 33260
    },
    {
      "entropy": 5.343992805480957,
      "epoch": 3.2883550810597075,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003942710703431548,
      "loss": 4.8888,
      "mean_token_accuracy": 0.22837442457675933,
      "num_tokens": 76211035.0,
      "step": 33265
    },
    {
      "entropy": 5.398343658447265,
      "epoch": 3.288849347568209,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003942411451201451,
      "loss": 4.9467,
      "mean_token_accuracy": 0.21951565593481065,
      "num_tokens": 76222262.0,
      "step": 33270
    },
    {
      "entropy": 5.4242250442504885,
      "epoch": 3.28934361407671,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039421121696378466,
      "loss": 4.9654,
      "mean_token_accuracy": 0.2216469407081604,
      "num_tokens": 76234346.0,
      "step": 33275
    },
    {
      "entropy": 5.3626867771148685,
      "epoch": 3.2898378805852113,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003941812858748098,
      "loss": 4.9136,
      "mean_token_accuracy": 0.22503640204668046,
      "num_tokens": 76244536.0,
      "step": 33280
    },
    {
      "entropy": 5.360766744613647,
      "epoch": 3.290332147093713,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003941513518539568,
      "loss": 4.9299,
      "mean_token_accuracy": 0.22251774370670319,
      "num_tokens": 76256080.0,
      "step": 33285
    },
    {
      "entropy": 5.35719633102417,
      "epoch": 3.2908264136022143,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039412141490196206,
      "loss": 4.8865,
      "mean_token_accuracy": 0.22919877618551254,
      "num_tokens": 76266741.0,
      "step": 33290
    },
    {
      "entropy": 5.293213367462158,
      "epoch": 3.2913206801107155,
      "grad_norm": 1.046875,
      "learning_rate": 0.000394091475019562,
      "loss": 4.9008,
      "mean_token_accuracy": 0.2169655054807663,
      "num_tokens": 76279099.0,
      "step": 33295
    },
    {
      "entropy": 5.428159141540528,
      "epoch": 3.2918149466192173,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003940615322074932,
      "loss": 5.05,
      "mean_token_accuracy": 0.21708894670009612,
      "num_tokens": 76290692.0,
      "step": 33300
    },
    {
      "entropy": 5.41946234703064,
      "epoch": 3.2923092131277185,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00039403158646649227,
      "loss": 4.9044,
      "mean_token_accuracy": 0.22026312947273255,
      "num_tokens": 76302003.0,
      "step": 33305
    },
    {
      "entropy": 5.303877639770508,
      "epoch": 3.2928034796362198,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003940016377972958,
      "loss": 4.9124,
      "mean_token_accuracy": 0.22323793768882752,
      "num_tokens": 76313255.0,
      "step": 33310
    },
    {
      "entropy": 5.3094110012054445,
      "epoch": 3.293297746144721,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039397168620064065,
      "loss": 4.8586,
      "mean_token_accuracy": 0.23006775081157685,
      "num_tokens": 76323154.0,
      "step": 33315
    },
    {
      "entropy": 5.449469757080078,
      "epoch": 3.2937920126532227,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039394173167726346,
      "loss": 4.9465,
      "mean_token_accuracy": 0.2136124223470688,
      "num_tokens": 76333168.0,
      "step": 33320
    },
    {
      "entropy": 5.434146451950073,
      "epoch": 3.294286279161724,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003939117742279013,
      "loss": 4.9932,
      "mean_token_accuracy": 0.21460400968790055,
      "num_tokens": 76344916.0,
      "step": 33325
    },
    {
      "entropy": 5.453251647949219,
      "epoch": 3.2947805456702253,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000393881813853291,
      "loss": 4.9332,
      "mean_token_accuracy": 0.21773354709148407,
      "num_tokens": 76357124.0,
      "step": 33330
    },
    {
      "entropy": 5.447827339172363,
      "epoch": 3.295274812178727,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00039385185055416965,
      "loss": 5.0019,
      "mean_token_accuracy": 0.2152120754122734,
      "num_tokens": 76368784.0,
      "step": 33335
    },
    {
      "entropy": 5.4210724353790285,
      "epoch": 3.2957690786872282,
      "grad_norm": 1.0,
      "learning_rate": 0.00039382188433127437,
      "loss": 4.9705,
      "mean_token_accuracy": 0.21075177788734437,
      "num_tokens": 76379864.0,
      "step": 33340
    },
    {
      "entropy": 5.431576681137085,
      "epoch": 3.2962633451957295,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003937919151853422,
      "loss": 4.934,
      "mean_token_accuracy": 0.21875980198383332,
      "num_tokens": 76392184.0,
      "step": 33345
    },
    {
      "entropy": 5.43218765258789,
      "epoch": 3.2967576117042308,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003937619431171107,
      "loss": 4.8875,
      "mean_token_accuracy": 0.2202820137143135,
      "num_tokens": 76404312.0,
      "step": 33350
    },
    {
      "entropy": 5.351992845535278,
      "epoch": 3.2972518782127325,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003937319681273168,
      "loss": 4.994,
      "mean_token_accuracy": 0.21723760515451432,
      "num_tokens": 76414351.0,
      "step": 33355
    },
    {
      "entropy": 5.430846405029297,
      "epoch": 3.2977461447212337,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003937019902166981,
      "loss": 4.9861,
      "mean_token_accuracy": 0.21242332309484482,
      "num_tokens": 76425851.0,
      "step": 33360
    },
    {
      "entropy": 5.3695393085479735,
      "epoch": 3.298240411229735,
      "grad_norm": 1.046875,
      "learning_rate": 0.000393672009385992,
      "loss": 4.8973,
      "mean_token_accuracy": 0.226909901201725,
      "num_tokens": 76437669.0,
      "step": 33365
    },
    {
      "entropy": 5.3629718780517575,
      "epoch": 3.2987346777382367,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003936420256359361,
      "loss": 4.8502,
      "mean_token_accuracy": 0.2292870655655861,
      "num_tokens": 76449589.0,
      "step": 33370
    },
    {
      "entropy": 5.40869255065918,
      "epoch": 3.299228944246738,
      "grad_norm": 1.015625,
      "learning_rate": 0.000393612038967268,
      "loss": 5.0212,
      "mean_token_accuracy": 0.21748004257678985,
      "num_tokens": 76462143.0,
      "step": 33375
    },
    {
      "entropy": 5.39310998916626,
      "epoch": 3.2997232107552392,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039358204938072527,
      "loss": 4.9117,
      "mean_token_accuracy": 0.22305780351161958,
      "num_tokens": 76473731.0,
      "step": 33380
    },
    {
      "entropy": 5.3966583728790285,
      "epoch": 3.3002174772637405,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00039355205687704575,
      "loss": 4.9191,
      "mean_token_accuracy": 0.2204673781991005,
      "num_tokens": 76484287.0,
      "step": 33385
    },
    {
      "entropy": 5.339991331100464,
      "epoch": 3.3007117437722417,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003935220614569672,
      "loss": 4.9028,
      "mean_token_accuracy": 0.2242744043469429,
      "num_tokens": 76495940.0,
      "step": 33390
    },
    {
      "entropy": 5.37202091217041,
      "epoch": 3.3012060102807435,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039349206312122765,
      "loss": 4.8527,
      "mean_token_accuracy": 0.23563077449798583,
      "num_tokens": 76507890.0,
      "step": 33395
    },
    {
      "entropy": 5.362610101699829,
      "epoch": 3.3017002767892447,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003934620618705649,
      "loss": 4.9089,
      "mean_token_accuracy": 0.22457995265722275,
      "num_tokens": 76519654.0,
      "step": 33400
    },
    {
      "entropy": 5.362840890884399,
      "epoch": 3.302194543297746,
      "grad_norm": 1.046875,
      "learning_rate": 0.000393432057705717,
      "loss": 4.8911,
      "mean_token_accuracy": 0.21823182255029677,
      "num_tokens": 76532026.0,
      "step": 33405
    },
    {
      "entropy": 5.347477006912231,
      "epoch": 3.3026888098062477,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003934020506274222,
      "loss": 4.8595,
      "mean_token_accuracy": 0.23167002499103545,
      "num_tokens": 76543629.0,
      "step": 33410
    },
    {
      "entropy": 5.450390863418579,
      "epoch": 3.303183076314749,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003933720406364185,
      "loss": 4.9403,
      "mean_token_accuracy": 0.22513831555843353,
      "num_tokens": 76554069.0,
      "step": 33415
    },
    {
      "entropy": 5.345998096466064,
      "epoch": 3.30367734282325,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039334202773344437,
      "loss": 4.8741,
      "mean_token_accuracy": 0.2238425150513649,
      "num_tokens": 76566148.0,
      "step": 33420
    },
    {
      "entropy": 5.342526197433472,
      "epoch": 3.3041716093317515,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039331201191923794,
      "loss": 4.8986,
      "mean_token_accuracy": 0.2182041421532631,
      "num_tokens": 76577208.0,
      "step": 33425
    },
    {
      "entropy": 5.312380123138428,
      "epoch": 3.304665875840253,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003932819931945375,
      "loss": 4.8683,
      "mean_token_accuracy": 0.23641083687543868,
      "num_tokens": 76588532.0,
      "step": 33430
    },
    {
      "entropy": 5.416561651229858,
      "epoch": 3.3051601423487544,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003932519715600819,
      "loss": 4.9788,
      "mean_token_accuracy": 0.21638747900724412,
      "num_tokens": 76600054.0,
      "step": 33435
    },
    {
      "entropy": 5.44054765701294,
      "epoch": 3.3056544088572557,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039322194701660936,
      "loss": 4.9512,
      "mean_token_accuracy": 0.22497459650039672,
      "num_tokens": 76612571.0,
      "step": 33440
    },
    {
      "entropy": 5.426646518707275,
      "epoch": 3.3061486753657574,
      "grad_norm": 1.0,
      "learning_rate": 0.00039319191956485854,
      "loss": 4.9827,
      "mean_token_accuracy": 0.21273379623889924,
      "num_tokens": 76625435.0,
      "step": 33445
    },
    {
      "entropy": 5.349206018447876,
      "epoch": 3.3066429418742587,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003931618892055682,
      "loss": 4.8585,
      "mean_token_accuracy": 0.23163534253835677,
      "num_tokens": 76635529.0,
      "step": 33450
    },
    {
      "entropy": 5.251764392852783,
      "epoch": 3.30713720838276,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00039313185593947713,
      "loss": 4.7924,
      "mean_token_accuracy": 0.23351703584194183,
      "num_tokens": 76645819.0,
      "step": 33455
    },
    {
      "entropy": 5.483918046951294,
      "epoch": 3.307631474891261,
      "grad_norm": 0.953125,
      "learning_rate": 0.00039310181976732404,
      "loss": 5.0777,
      "mean_token_accuracy": 0.20413907021284103,
      "num_tokens": 76657803.0,
      "step": 33460
    },
    {
      "entropy": 5.48245062828064,
      "epoch": 3.308125741399763,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039307178068984783,
      "loss": 5.0505,
      "mean_token_accuracy": 0.20711697190999984,
      "num_tokens": 76670260.0,
      "step": 33465
    },
    {
      "entropy": 5.346644258499145,
      "epoch": 3.308620007908264,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003930417387077875,
      "loss": 4.8822,
      "mean_token_accuracy": 0.22450132369995118,
      "num_tokens": 76682591.0,
      "step": 33470
    },
    {
      "entropy": 5.3553873062133786,
      "epoch": 3.3091142744167654,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003930116938218821,
      "loss": 4.8914,
      "mean_token_accuracy": 0.22261531203985213,
      "num_tokens": 76694147.0,
      "step": 33475
    },
    {
      "entropy": 5.389763259887696,
      "epoch": 3.309608540925267,
      "grad_norm": 0.9375,
      "learning_rate": 0.00039298164603287063,
      "loss": 5.0014,
      "mean_token_accuracy": 0.22221826612949372,
      "num_tokens": 76706325.0,
      "step": 33480
    },
    {
      "entropy": 5.421973657608032,
      "epoch": 3.3101028074337684,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003929515953414926,
      "loss": 4.9382,
      "mean_token_accuracy": 0.22265520095825195,
      "num_tokens": 76719574.0,
      "step": 33485
    },
    {
      "entropy": 5.332542037963867,
      "epoch": 3.3105970739422697,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003929215417484868,
      "loss": 4.8567,
      "mean_token_accuracy": 0.2235609397292137,
      "num_tokens": 76731820.0,
      "step": 33490
    },
    {
      "entropy": 5.27767424583435,
      "epoch": 3.311091340450771,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039289148525459283,
      "loss": 4.8013,
      "mean_token_accuracy": 0.23237260431051254,
      "num_tokens": 76742849.0,
      "step": 33495
    },
    {
      "entropy": 5.442330646514892,
      "epoch": 3.3115856069592726,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039286142586055005,
      "loss": 4.9921,
      "mean_token_accuracy": 0.21917519569396973,
      "num_tokens": 76754747.0,
      "step": 33500
    },
    {
      "entropy": 5.32992582321167,
      "epoch": 3.312079873467774,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000392831363567098,
      "loss": 4.8508,
      "mean_token_accuracy": 0.22816484421491623,
      "num_tokens": 76766738.0,
      "step": 33505
    },
    {
      "entropy": 5.316068887710571,
      "epoch": 3.312574139976275,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039280129837497595,
      "loss": 4.9319,
      "mean_token_accuracy": 0.21944709569215776,
      "num_tokens": 76777457.0,
      "step": 33510
    },
    {
      "entropy": 5.433663654327392,
      "epoch": 3.3130684064847764,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039277123028492376,
      "loss": 4.9983,
      "mean_token_accuracy": 0.20827572345733641,
      "num_tokens": 76788335.0,
      "step": 33515
    },
    {
      "entropy": 5.471884393692017,
      "epoch": 3.313562672993278,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003927411592976811,
      "loss": 4.8966,
      "mean_token_accuracy": 0.22105845510959626,
      "num_tokens": 76800471.0,
      "step": 33520
    },
    {
      "entropy": 5.306353425979614,
      "epoch": 3.3140569395017794,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039271108541398753,
      "loss": 4.9065,
      "mean_token_accuracy": 0.23003062903881072,
      "num_tokens": 76811792.0,
      "step": 33525
    },
    {
      "entropy": 5.318279695510864,
      "epoch": 3.3145512060102806,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003926810086345831,
      "loss": 4.839,
      "mean_token_accuracy": 0.22707404643297197,
      "num_tokens": 76824332.0,
      "step": 33530
    },
    {
      "entropy": 5.451320314407349,
      "epoch": 3.315045472518782,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039265092896020767,
      "loss": 4.9444,
      "mean_token_accuracy": 0.2167900413274765,
      "num_tokens": 76835712.0,
      "step": 33535
    },
    {
      "entropy": 5.344161748886108,
      "epoch": 3.3155397390272836,
      "grad_norm": 1.015625,
      "learning_rate": 0.000392620846391601,
      "loss": 4.8768,
      "mean_token_accuracy": 0.23346701860427857,
      "num_tokens": 76847915.0,
      "step": 33540
    },
    {
      "entropy": 5.338288497924805,
      "epoch": 3.316034005535785,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039259076092950333,
      "loss": 4.9512,
      "mean_token_accuracy": 0.21920739710330964,
      "num_tokens": 76859982.0,
      "step": 33545
    },
    {
      "entropy": 5.300202178955078,
      "epoch": 3.316528272044286,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003925606725746547,
      "loss": 4.869,
      "mean_token_accuracy": 0.23055555373430253,
      "num_tokens": 76870743.0,
      "step": 33550
    },
    {
      "entropy": 5.242681550979614,
      "epoch": 3.317022538552788,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003925305813277952,
      "loss": 4.8327,
      "mean_token_accuracy": 0.22515660226345063,
      "num_tokens": 76882118.0,
      "step": 33555
    },
    {
      "entropy": 5.366752481460571,
      "epoch": 3.317516805061289,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039250048718966524,
      "loss": 4.927,
      "mean_token_accuracy": 0.2165424183011055,
      "num_tokens": 76893658.0,
      "step": 33560
    },
    {
      "entropy": 5.425343322753906,
      "epoch": 3.3180110715697904,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039247039016100514,
      "loss": 4.9071,
      "mean_token_accuracy": 0.2201566904783249,
      "num_tokens": 76904974.0,
      "step": 33565
    },
    {
      "entropy": 5.368563795089722,
      "epoch": 3.3185053380782916,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039244029024255506,
      "loss": 4.9054,
      "mean_token_accuracy": 0.22674650847911834,
      "num_tokens": 76915972.0,
      "step": 33570
    },
    {
      "entropy": 5.3593607425689695,
      "epoch": 3.3189996045867933,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003924101874350557,
      "loss": 4.9148,
      "mean_token_accuracy": 0.21904488205909728,
      "num_tokens": 76927397.0,
      "step": 33575
    },
    {
      "entropy": 5.390121936798096,
      "epoch": 3.3194938710952946,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003923800817392475,
      "loss": 4.9053,
      "mean_token_accuracy": 0.2155536949634552,
      "num_tokens": 76939744.0,
      "step": 33580
    },
    {
      "entropy": 5.442476797103882,
      "epoch": 3.319988137603796,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039234997315587105,
      "loss": 5.041,
      "mean_token_accuracy": 0.21334196478128434,
      "num_tokens": 76951747.0,
      "step": 33585
    },
    {
      "entropy": 5.393838119506836,
      "epoch": 3.3204824041122976,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003923198616856671,
      "loss": 4.92,
      "mean_token_accuracy": 0.22689845860004426,
      "num_tokens": 76961734.0,
      "step": 33590
    },
    {
      "entropy": 5.396941137313843,
      "epoch": 3.320976670620799,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039228974732937637,
      "loss": 4.9382,
      "mean_token_accuracy": 0.21828213185071946,
      "num_tokens": 76972566.0,
      "step": 33595
    },
    {
      "entropy": 5.407994174957276,
      "epoch": 3.3214709371293,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003922596300877396,
      "loss": 4.9418,
      "mean_token_accuracy": 0.2156861662864685,
      "num_tokens": 76983446.0,
      "step": 33600
    },
    {
      "entropy": 5.3977141857147215,
      "epoch": 3.3219652036378013,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003922295099614978,
      "loss": 4.8974,
      "mean_token_accuracy": 0.21930639892816545,
      "num_tokens": 76994292.0,
      "step": 33605
    },
    {
      "entropy": 5.356265068054199,
      "epoch": 3.322459470146303,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039219938695139186,
      "loss": 4.9351,
      "mean_token_accuracy": 0.2184906616806984,
      "num_tokens": 77007095.0,
      "step": 33610
    },
    {
      "entropy": 5.386044025421143,
      "epoch": 3.3229537366548043,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003921692610581628,
      "loss": 4.8916,
      "mean_token_accuracy": 0.21931139677762984,
      "num_tokens": 77017358.0,
      "step": 33615
    },
    {
      "entropy": 5.336177015304566,
      "epoch": 3.3234480031633056,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003921391322825517,
      "loss": 4.8381,
      "mean_token_accuracy": 0.22498152405023575,
      "num_tokens": 77028794.0,
      "step": 33620
    },
    {
      "entropy": 5.271046543121338,
      "epoch": 3.3239422696718073,
      "grad_norm": 1.0,
      "learning_rate": 0.00039210900062529986,
      "loss": 4.8912,
      "mean_token_accuracy": 0.2239728018641472,
      "num_tokens": 77040795.0,
      "step": 33625
    },
    {
      "entropy": 5.390525197982788,
      "epoch": 3.3244365361803085,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003920788660871483,
      "loss": 4.9159,
      "mean_token_accuracy": 0.22753553390502929,
      "num_tokens": 77051810.0,
      "step": 33630
    },
    {
      "entropy": 5.416570615768433,
      "epoch": 3.32493080268881,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039204872866883863,
      "loss": 5.0195,
      "mean_token_accuracy": 0.21929581165313722,
      "num_tokens": 77064504.0,
      "step": 33635
    },
    {
      "entropy": 5.413747644424438,
      "epoch": 3.325425069197311,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000392018588371112,
      "loss": 4.9209,
      "mean_token_accuracy": 0.21761177033185958,
      "num_tokens": 77075163.0,
      "step": 33640
    },
    {
      "entropy": 5.382626724243164,
      "epoch": 3.3259193357058123,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039198844519470994,
      "loss": 4.9325,
      "mean_token_accuracy": 0.2229388937354088,
      "num_tokens": 77086871.0,
      "step": 33645
    },
    {
      "entropy": 5.304977893829346,
      "epoch": 3.326413602214314,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000391958299140374,
      "loss": 4.8426,
      "mean_token_accuracy": 0.23287004828453065,
      "num_tokens": 77097285.0,
      "step": 33650
    },
    {
      "entropy": 5.252664184570312,
      "epoch": 3.3269078687228153,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003919281502088457,
      "loss": 4.794,
      "mean_token_accuracy": 0.2384445458650589,
      "num_tokens": 77107035.0,
      "step": 33655
    },
    {
      "entropy": 5.41862850189209,
      "epoch": 3.3274021352313166,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003918979984008668,
      "loss": 4.9472,
      "mean_token_accuracy": 0.21224371045827867,
      "num_tokens": 77119449.0,
      "step": 33660
    },
    {
      "entropy": 5.415667486190796,
      "epoch": 3.3278964017398183,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039186784371717903,
      "loss": 4.9272,
      "mean_token_accuracy": 0.22217332273721696,
      "num_tokens": 77130508.0,
      "step": 33665
    },
    {
      "entropy": 5.392439460754394,
      "epoch": 3.3283906682483195,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003918376861585242,
      "loss": 4.9313,
      "mean_token_accuracy": 0.21787269711494445,
      "num_tokens": 77142080.0,
      "step": 33670
    },
    {
      "entropy": 5.4059410572052,
      "epoch": 3.328884934756821,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039180752572564405,
      "loss": 4.9724,
      "mean_token_accuracy": 0.21855347156524657,
      "num_tokens": 77152366.0,
      "step": 33675
    },
    {
      "entropy": 5.384652709960937,
      "epoch": 3.329379201265322,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003917773624192808,
      "loss": 5.0076,
      "mean_token_accuracy": 0.21514663994312286,
      "num_tokens": 77164984.0,
      "step": 33680
    },
    {
      "entropy": 5.44873194694519,
      "epoch": 3.3298734677738238,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003917471962401762,
      "loss": 4.9911,
      "mean_token_accuracy": 0.21500820070505142,
      "num_tokens": 77176870.0,
      "step": 33685
    },
    {
      "entropy": 5.315370893478393,
      "epoch": 3.330367734282325,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003917170271890724,
      "loss": 4.826,
      "mean_token_accuracy": 0.23330404609441757,
      "num_tokens": 77188376.0,
      "step": 33690
    },
    {
      "entropy": 5.336959934234619,
      "epoch": 3.3308620007908263,
      "grad_norm": 1.125,
      "learning_rate": 0.0003916868552667118,
      "loss": 4.8716,
      "mean_token_accuracy": 0.23045846074819565,
      "num_tokens": 77198556.0,
      "step": 33695
    },
    {
      "entropy": 5.3706443309783936,
      "epoch": 3.331356267299328,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039165668047383634,
      "loss": 4.9689,
      "mean_token_accuracy": 0.2217753201723099,
      "num_tokens": 77209409.0,
      "step": 33700
    },
    {
      "entropy": 5.418371868133545,
      "epoch": 3.3318505338078293,
      "grad_norm": 1.0,
      "learning_rate": 0.0003916265028111886,
      "loss": 4.9434,
      "mean_token_accuracy": 0.21837315261363982,
      "num_tokens": 77219986.0,
      "step": 33705
    },
    {
      "entropy": 5.399960041046143,
      "epoch": 3.3323448003163305,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039159632227951066,
      "loss": 4.9446,
      "mean_token_accuracy": 0.21917613744735717,
      "num_tokens": 77231380.0,
      "step": 33710
    },
    {
      "entropy": 5.325772428512574,
      "epoch": 3.3328390668248318,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000391566138879545,
      "loss": 4.9422,
      "mean_token_accuracy": 0.2255724176764488,
      "num_tokens": 77241765.0,
      "step": 33715
    },
    {
      "entropy": 5.368354558944702,
      "epoch": 3.3333333333333335,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003915359526120344,
      "loss": 4.964,
      "mean_token_accuracy": 0.21786848306655884,
      "num_tokens": 77256091.0,
      "step": 33720
    },
    {
      "entropy": 5.485012817382812,
      "epoch": 3.3338275998418347,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003915057634777213,
      "loss": 4.993,
      "mean_token_accuracy": 0.21501253247261048,
      "num_tokens": 77267084.0,
      "step": 33725
    },
    {
      "entropy": 5.515783357620239,
      "epoch": 3.334321866350336,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003914755714773483,
      "loss": 4.9605,
      "mean_token_accuracy": 0.21667275726795196,
      "num_tokens": 77278435.0,
      "step": 33730
    },
    {
      "entropy": 5.350251579284668,
      "epoch": 3.3348161328588377,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003914453766116581,
      "loss": 4.9563,
      "mean_token_accuracy": 0.21631327122449875,
      "num_tokens": 77290971.0,
      "step": 33735
    },
    {
      "entropy": 5.351337862014771,
      "epoch": 3.335310399367339,
      "grad_norm": 1.0,
      "learning_rate": 0.0003914151788813937,
      "loss": 4.9033,
      "mean_token_accuracy": 0.22512815445661544,
      "num_tokens": 77302704.0,
      "step": 33740
    },
    {
      "entropy": 5.3679790019989015,
      "epoch": 3.3358046658758402,
      "grad_norm": 0.953125,
      "learning_rate": 0.00039138497828729773,
      "loss": 4.8536,
      "mean_token_accuracy": 0.23604413568973542,
      "num_tokens": 77314107.0,
      "step": 33745
    },
    {
      "entropy": 5.44464693069458,
      "epoch": 3.3362989323843415,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003913547748301133,
      "loss": 4.9148,
      "mean_token_accuracy": 0.22416321635246278,
      "num_tokens": 77324040.0,
      "step": 33750
    },
    {
      "entropy": 5.279737567901611,
      "epoch": 3.336793198892843,
      "grad_norm": 1.0,
      "learning_rate": 0.0003913245685105834,
      "loss": 4.8683,
      "mean_token_accuracy": 0.23330911844968796,
      "num_tokens": 77336471.0,
      "step": 33755
    },
    {
      "entropy": 5.397892093658447,
      "epoch": 3.3372874654013445,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039129435932945094,
      "loss": 4.952,
      "mean_token_accuracy": 0.22048970311880112,
      "num_tokens": 77346947.0,
      "step": 33760
    },
    {
      "entropy": 5.382558059692383,
      "epoch": 3.3377817319098457,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003912641472874593,
      "loss": 4.8554,
      "mean_token_accuracy": 0.23307389318943023,
      "num_tokens": 77357995.0,
      "step": 33765
    },
    {
      "entropy": 5.32887225151062,
      "epoch": 3.338275998418347,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003912339323853516,
      "loss": 4.8779,
      "mean_token_accuracy": 0.22487757205963135,
      "num_tokens": 77368910.0,
      "step": 33770
    },
    {
      "entropy": 5.428550100326538,
      "epoch": 3.3387702649268487,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039120371462387105,
      "loss": 4.9732,
      "mean_token_accuracy": 0.21811838150024415,
      "num_tokens": 77381010.0,
      "step": 33775
    },
    {
      "entropy": 5.406692600250244,
      "epoch": 3.33926453143535,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039117349400376116,
      "loss": 5.0094,
      "mean_token_accuracy": 0.21354680955410005,
      "num_tokens": 77392182.0,
      "step": 33780
    },
    {
      "entropy": 5.436107158660889,
      "epoch": 3.339758797943851,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039114327052576525,
      "loss": 5.0195,
      "mean_token_accuracy": 0.21025025546550752,
      "num_tokens": 77403574.0,
      "step": 33785
    },
    {
      "entropy": 5.277062273025512,
      "epoch": 3.3402530644523525,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039111304419062686,
      "loss": 4.7949,
      "mean_token_accuracy": 0.22950299829244614,
      "num_tokens": 77414941.0,
      "step": 33790
    },
    {
      "entropy": 5.388889360427856,
      "epoch": 3.340747330960854,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003910828149990897,
      "loss": 4.9596,
      "mean_token_accuracy": 0.21883899569511414,
      "num_tokens": 77426361.0,
      "step": 33795
    },
    {
      "entropy": 5.398156595230103,
      "epoch": 3.3412415974693555,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039105258295189717,
      "loss": 4.9743,
      "mean_token_accuracy": 0.22055567502975465,
      "num_tokens": 77438366.0,
      "step": 33800
    },
    {
      "entropy": 5.355645084381104,
      "epoch": 3.3417358639778567,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039102234804979307,
      "loss": 4.8865,
      "mean_token_accuracy": 0.23479216694831848,
      "num_tokens": 77448662.0,
      "step": 33805
    },
    {
      "entropy": 5.353066778182983,
      "epoch": 3.3422301304863584,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003909921102935213,
      "loss": 4.8408,
      "mean_token_accuracy": 0.23334318846464158,
      "num_tokens": 77459806.0,
      "step": 33810
    },
    {
      "entropy": 5.3951879978179935,
      "epoch": 3.3427243969948597,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039096186968382555,
      "loss": 4.9798,
      "mean_token_accuracy": 0.22152237296104432,
      "num_tokens": 77470571.0,
      "step": 33815
    },
    {
      "entropy": 5.373051738739013,
      "epoch": 3.343218663503361,
      "grad_norm": 1.0,
      "learning_rate": 0.0003909316262214498,
      "loss": 4.9064,
      "mean_token_accuracy": 0.22443777173757554,
      "num_tokens": 77483365.0,
      "step": 33820
    },
    {
      "entropy": 5.3990129947662355,
      "epoch": 3.343712930011862,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039090137990713817,
      "loss": 4.8675,
      "mean_token_accuracy": 0.22349267452955246,
      "num_tokens": 77493710.0,
      "step": 33825
    },
    {
      "entropy": 5.3637778759002686,
      "epoch": 3.344207196520364,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003908711307416345,
      "loss": 4.8706,
      "mean_token_accuracy": 0.22645164132118226,
      "num_tokens": 77505153.0,
      "step": 33830
    },
    {
      "entropy": 5.434598207473755,
      "epoch": 3.344701463028865,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003908408787256831,
      "loss": 5.0168,
      "mean_token_accuracy": 0.21272971630096435,
      "num_tokens": 77516370.0,
      "step": 33835
    },
    {
      "entropy": 5.433268594741821,
      "epoch": 3.3451957295373664,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003908106238600281,
      "loss": 4.9967,
      "mean_token_accuracy": 0.22129547744989395,
      "num_tokens": 77526660.0,
      "step": 33840
    },
    {
      "entropy": 5.29403166770935,
      "epoch": 3.345689996045868,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039078036614541376,
      "loss": 4.8235,
      "mean_token_accuracy": 0.23171166330575943,
      "num_tokens": 77538112.0,
      "step": 33845
    },
    {
      "entropy": 5.2873132705688475,
      "epoch": 3.3461842625543694,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003907501055825845,
      "loss": 4.8856,
      "mean_token_accuracy": 0.22381557673215866,
      "num_tokens": 77548807.0,
      "step": 33850
    },
    {
      "entropy": 5.308782863616943,
      "epoch": 3.3466785290628707,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039071984217228464,
      "loss": 4.906,
      "mean_token_accuracy": 0.22866106629371644,
      "num_tokens": 77559803.0,
      "step": 33855
    },
    {
      "entropy": 5.415783977508545,
      "epoch": 3.347172795571372,
      "grad_norm": 1.0,
      "learning_rate": 0.00039068957591525873,
      "loss": 4.9141,
      "mean_token_accuracy": 0.22099430561065675,
      "num_tokens": 77570973.0,
      "step": 33860
    },
    {
      "entropy": 5.391210556030273,
      "epoch": 3.3476670620798736,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003906593068122513,
      "loss": 4.949,
      "mean_token_accuracy": 0.21853359639644623,
      "num_tokens": 77583614.0,
      "step": 33865
    },
    {
      "entropy": 5.367411375045776,
      "epoch": 3.348161328588375,
      "grad_norm": 1.078125,
      "learning_rate": 0.000390629034864007,
      "loss": 4.9546,
      "mean_token_accuracy": 0.22670968621969223,
      "num_tokens": 77595086.0,
      "step": 33870
    },
    {
      "entropy": 5.404358053207398,
      "epoch": 3.348655595096876,
      "grad_norm": 1.125,
      "learning_rate": 0.00039059876007127044,
      "loss": 4.9249,
      "mean_token_accuracy": 0.2182835668325424,
      "num_tokens": 77605492.0,
      "step": 33875
    },
    {
      "entropy": 5.354220771789551,
      "epoch": 3.349149861605378,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003905684824347865,
      "loss": 4.8997,
      "mean_token_accuracy": 0.22284364104270935,
      "num_tokens": 77618123.0,
      "step": 33880
    },
    {
      "entropy": 5.420915365219116,
      "epoch": 3.349644128113879,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003905382019553,
      "loss": 4.9492,
      "mean_token_accuracy": 0.22250049859285354,
      "num_tokens": 77630733.0,
      "step": 33885
    },
    {
      "entropy": 5.391465473175049,
      "epoch": 3.3501383946223804,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039050791863355573,
      "loss": 4.9675,
      "mean_token_accuracy": 0.22079341262578964,
      "num_tokens": 77641337.0,
      "step": 33890
    },
    {
      "entropy": 5.396787834167481,
      "epoch": 3.3506326611308817,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039047763247029876,
      "loss": 4.9635,
      "mean_token_accuracy": 0.2206554263830185,
      "num_tokens": 77652567.0,
      "step": 33895
    },
    {
      "entropy": 5.422830247879029,
      "epoch": 3.351126927639383,
      "grad_norm": 1.0,
      "learning_rate": 0.0003904473434662741,
      "loss": 4.9973,
      "mean_token_accuracy": 0.21841063648462294,
      "num_tokens": 77664146.0,
      "step": 33900
    },
    {
      "entropy": 5.452880239486694,
      "epoch": 3.3516211941478846,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003904170516222269,
      "loss": 4.9461,
      "mean_token_accuracy": 0.22160077840089798,
      "num_tokens": 77675406.0,
      "step": 33905
    },
    {
      "entropy": 5.348075723648071,
      "epoch": 3.352115460656386,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003903867569389024,
      "loss": 4.9357,
      "mean_token_accuracy": 0.2238214209675789,
      "num_tokens": 77686846.0,
      "step": 33910
    },
    {
      "entropy": 5.387111520767212,
      "epoch": 3.352609727164887,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003903564594170457,
      "loss": 4.9086,
      "mean_token_accuracy": 0.21831238269805908,
      "num_tokens": 77697136.0,
      "step": 33915
    },
    {
      "entropy": 5.49753589630127,
      "epoch": 3.353103993673389,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039032615905740226,
      "loss": 4.994,
      "mean_token_accuracy": 0.21247973293066025,
      "num_tokens": 77708056.0,
      "step": 33920
    },
    {
      "entropy": 5.361322355270386,
      "epoch": 3.35359826018189,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003902958558607173,
      "loss": 4.9409,
      "mean_token_accuracy": 0.21844601929187774,
      "num_tokens": 77718582.0,
      "step": 33925
    },
    {
      "entropy": 5.379973840713501,
      "epoch": 3.3540925266903914,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003902655498277366,
      "loss": 4.9328,
      "mean_token_accuracy": 0.2242838129401207,
      "num_tokens": 77729388.0,
      "step": 33930
    },
    {
      "entropy": 5.417075109481812,
      "epoch": 3.3545867931988926,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039023524095920534,
      "loss": 4.8292,
      "mean_token_accuracy": 0.22499683797359465,
      "num_tokens": 77739952.0,
      "step": 33935
    },
    {
      "entropy": 5.302096176147461,
      "epoch": 3.3550810597073943,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003902049292558693,
      "loss": 4.9108,
      "mean_token_accuracy": 0.22599294781684875,
      "num_tokens": 77752415.0,
      "step": 33940
    },
    {
      "entropy": 5.293969774246216,
      "epoch": 3.3555753262158956,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039017461471847417,
      "loss": 4.8217,
      "mean_token_accuracy": 0.22671417891979218,
      "num_tokens": 77764102.0,
      "step": 33945
    },
    {
      "entropy": 5.359054756164551,
      "epoch": 3.356069592724397,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003901442973477656,
      "loss": 4.9089,
      "mean_token_accuracy": 0.22155204117298127,
      "num_tokens": 77774908.0,
      "step": 33950
    },
    {
      "entropy": 5.415456485748291,
      "epoch": 3.3565638592328986,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003901139771444894,
      "loss": 4.9841,
      "mean_token_accuracy": 0.2191945418715477,
      "num_tokens": 77785913.0,
      "step": 33955
    },
    {
      "entropy": 5.369298124313355,
      "epoch": 3.3570581257414,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003900836541093916,
      "loss": 4.9224,
      "mean_token_accuracy": 0.2211829200387001,
      "num_tokens": 77797539.0,
      "step": 33960
    },
    {
      "entropy": 5.3753527164459225,
      "epoch": 3.357552392249901,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000390053328243218,
      "loss": 4.9391,
      "mean_token_accuracy": 0.2193540796637535,
      "num_tokens": 77809777.0,
      "step": 33965
    },
    {
      "entropy": 5.378765869140625,
      "epoch": 3.3580466587584024,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003900229995467147,
      "loss": 4.864,
      "mean_token_accuracy": 0.22592062056064605,
      "num_tokens": 77820401.0,
      "step": 33970
    },
    {
      "entropy": 5.371637344360352,
      "epoch": 3.358540925266904,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003899926680206278,
      "loss": 4.8979,
      "mean_token_accuracy": 0.2189282774925232,
      "num_tokens": 77832235.0,
      "step": 33975
    },
    {
      "entropy": 5.375926876068116,
      "epoch": 3.3590351917754053,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00038996233366570335,
      "loss": 4.9442,
      "mean_token_accuracy": 0.22109284549951552,
      "num_tokens": 77843938.0,
      "step": 33980
    },
    {
      "entropy": 5.402053451538086,
      "epoch": 3.3595294582839066,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038993199648268775,
      "loss": 4.937,
      "mean_token_accuracy": 0.21671762466430664,
      "num_tokens": 77853921.0,
      "step": 33985
    },
    {
      "entropy": 5.311585330963135,
      "epoch": 3.3600237247924083,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038990165647232713,
      "loss": 4.7784,
      "mean_token_accuracy": 0.23718335181474687,
      "num_tokens": 77864661.0,
      "step": 33990
    },
    {
      "entropy": 5.517502069473267,
      "epoch": 3.3605179913009096,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003898713136353679,
      "loss": 5.0692,
      "mean_token_accuracy": 0.20248904824256897,
      "num_tokens": 77875743.0,
      "step": 33995
    },
    {
      "entropy": 5.370690107345581,
      "epoch": 3.361012257809411,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003898409679725566,
      "loss": 4.9169,
      "mean_token_accuracy": 0.2217498689889908,
      "num_tokens": 77888575.0,
      "step": 34000
    },
    {
      "entropy": 5.487114524841308,
      "epoch": 3.361506524317912,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003898106194846396,
      "loss": 5.0735,
      "mean_token_accuracy": 0.21143750697374344,
      "num_tokens": 77901212.0,
      "step": 34005
    },
    {
      "entropy": 5.352934122085571,
      "epoch": 3.362000790826414,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038978026817236365,
      "loss": 4.8812,
      "mean_token_accuracy": 0.2320416271686554,
      "num_tokens": 77911905.0,
      "step": 34010
    },
    {
      "entropy": 5.425566053390503,
      "epoch": 3.362495057334915,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038974991403647536,
      "loss": 4.9569,
      "mean_token_accuracy": 0.21426244229078292,
      "num_tokens": 77923710.0,
      "step": 34015
    },
    {
      "entropy": 5.412097549438476,
      "epoch": 3.3629893238434163,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038971955707772126,
      "loss": 4.9209,
      "mean_token_accuracy": 0.22057076096534728,
      "num_tokens": 77934501.0,
      "step": 34020
    },
    {
      "entropy": 5.329768657684326,
      "epoch": 3.3634835903519176,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003896891972968482,
      "loss": 4.886,
      "mean_token_accuracy": 0.22223179936408996,
      "num_tokens": 77944512.0,
      "step": 34025
    },
    {
      "entropy": 5.361877822875977,
      "epoch": 3.3639778568604193,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003896588346946033,
      "loss": 4.8644,
      "mean_token_accuracy": 0.23420026749372483,
      "num_tokens": 77955253.0,
      "step": 34030
    },
    {
      "entropy": 5.318459177017212,
      "epoch": 3.3644721233689205,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038962846927173306,
      "loss": 4.8199,
      "mean_token_accuracy": 0.2307724416255951,
      "num_tokens": 77967193.0,
      "step": 34035
    },
    {
      "entropy": 5.354242420196533,
      "epoch": 3.364966389877422,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003895981010289848,
      "loss": 4.9987,
      "mean_token_accuracy": 0.21727477759122849,
      "num_tokens": 77977105.0,
      "step": 34040
    },
    {
      "entropy": 5.384107065200806,
      "epoch": 3.365460656385923,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038956772996710554,
      "loss": 4.9492,
      "mean_token_accuracy": 0.2220601335167885,
      "num_tokens": 77989517.0,
      "step": 34045
    },
    {
      "entropy": 5.45135498046875,
      "epoch": 3.3659549228944248,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00038953735608684216,
      "loss": 4.9278,
      "mean_token_accuracy": 0.2151881605386734,
      "num_tokens": 78000797.0,
      "step": 34050
    },
    {
      "entropy": 5.275989770889282,
      "epoch": 3.366449189402926,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003895069793889422,
      "loss": 4.7745,
      "mean_token_accuracy": 0.2426856517791748,
      "num_tokens": 78012470.0,
      "step": 34055
    },
    {
      "entropy": 5.353229951858521,
      "epoch": 3.3669434559114273,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003894765998741528,
      "loss": 4.8679,
      "mean_token_accuracy": 0.22156235277652742,
      "num_tokens": 78023774.0,
      "step": 34060
    },
    {
      "entropy": 5.3370583057403564,
      "epoch": 3.367437722419929,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003894462175432212,
      "loss": 4.968,
      "mean_token_accuracy": 0.21899116039276123,
      "num_tokens": 78035181.0,
      "step": 34065
    },
    {
      "entropy": 5.415081739425659,
      "epoch": 3.3679319889284303,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038941583239689495,
      "loss": 4.925,
      "mean_token_accuracy": 0.21794160455465317,
      "num_tokens": 78045728.0,
      "step": 34070
    },
    {
      "entropy": 5.439558029174805,
      "epoch": 3.3684262554369315,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038938544443592146,
      "loss": 4.9613,
      "mean_token_accuracy": 0.21791402697563172,
      "num_tokens": 78056675.0,
      "step": 34075
    },
    {
      "entropy": 5.341196537017822,
      "epoch": 3.368920521945433,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003893550536610483,
      "loss": 4.8371,
      "mean_token_accuracy": 0.23003219366073607,
      "num_tokens": 78067686.0,
      "step": 34080
    },
    {
      "entropy": 5.298730659484863,
      "epoch": 3.3694147884539345,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000389324660073023,
      "loss": 4.884,
      "mean_token_accuracy": 0.23259907215833664,
      "num_tokens": 78078063.0,
      "step": 34085
    },
    {
      "entropy": 5.44101128578186,
      "epoch": 3.3699090549624358,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003892942636725934,
      "loss": 5.0217,
      "mean_token_accuracy": 0.2214057818055153,
      "num_tokens": 78088729.0,
      "step": 34090
    },
    {
      "entropy": 5.28401370048523,
      "epoch": 3.370403321470937,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003892638644605072,
      "loss": 4.8409,
      "mean_token_accuracy": 0.23488483130931853,
      "num_tokens": 78100468.0,
      "step": 34095
    },
    {
      "entropy": 5.355393648147583,
      "epoch": 3.3708975879794387,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038923346243751217,
      "loss": 4.9442,
      "mean_token_accuracy": 0.22354098856449128,
      "num_tokens": 78111911.0,
      "step": 34100
    },
    {
      "entropy": 5.3077263832092285,
      "epoch": 3.37139185448794,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038920305760435626,
      "loss": 4.829,
      "mean_token_accuracy": 0.22860943973064424,
      "num_tokens": 78122411.0,
      "step": 34105
    },
    {
      "entropy": 5.45547251701355,
      "epoch": 3.3718861209964412,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003891726499617874,
      "loss": 4.9935,
      "mean_token_accuracy": 0.21136883348226548,
      "num_tokens": 78134313.0,
      "step": 34110
    },
    {
      "entropy": 5.262450790405273,
      "epoch": 3.3723803875049425,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003891422395105536,
      "loss": 4.7271,
      "mean_token_accuracy": 0.23711652755737306,
      "num_tokens": 78144753.0,
      "step": 34115
    },
    {
      "entropy": 5.271003150939942,
      "epoch": 3.372874654013444,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038911182625140305,
      "loss": 4.8246,
      "mean_token_accuracy": 0.23160648196935654,
      "num_tokens": 78154772.0,
      "step": 34120
    },
    {
      "entropy": 5.332530736923218,
      "epoch": 3.3733689205219455,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038908141018508387,
      "loss": 4.8975,
      "mean_token_accuracy": 0.23049889355897904,
      "num_tokens": 78166876.0,
      "step": 34125
    },
    {
      "entropy": 5.381187486648559,
      "epoch": 3.3738631870304467,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038905099131234433,
      "loss": 4.8825,
      "mean_token_accuracy": 0.23446121364831923,
      "num_tokens": 78177530.0,
      "step": 34130
    },
    {
      "entropy": 5.332113265991211,
      "epoch": 3.3743574535389484,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00038902056963393265,
      "loss": 4.8659,
      "mean_token_accuracy": 0.22584918290376663,
      "num_tokens": 78187986.0,
      "step": 34135
    },
    {
      "entropy": 5.305330753326416,
      "epoch": 3.3748517200474497,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00038899014515059735,
      "loss": 4.8961,
      "mean_token_accuracy": 0.22376441061496735,
      "num_tokens": 78199874.0,
      "step": 34140
    },
    {
      "entropy": 5.3618138313293455,
      "epoch": 3.375345986555951,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003889597178630867,
      "loss": 4.8785,
      "mean_token_accuracy": 0.22781001776456833,
      "num_tokens": 78210995.0,
      "step": 34145
    },
    {
      "entropy": 5.36954779624939,
      "epoch": 3.3758402530644522,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003889292877721493,
      "loss": 4.8542,
      "mean_token_accuracy": 0.22493356168270112,
      "num_tokens": 78221792.0,
      "step": 34150
    },
    {
      "entropy": 5.285779666900635,
      "epoch": 3.3763345195729535,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038889885487853384,
      "loss": 4.8771,
      "mean_token_accuracy": 0.2388073295354843,
      "num_tokens": 78232799.0,
      "step": 34155
    },
    {
      "entropy": 5.412586688995361,
      "epoch": 3.376828786081455,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003888684191829889,
      "loss": 4.9448,
      "mean_token_accuracy": 0.2174442321062088,
      "num_tokens": 78243784.0,
      "step": 34160
    },
    {
      "entropy": 5.3730456829071045,
      "epoch": 3.3773230525899565,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038883798068626305,
      "loss": 4.8222,
      "mean_token_accuracy": 0.23194272667169571,
      "num_tokens": 78255161.0,
      "step": 34165
    },
    {
      "entropy": 5.280057096481324,
      "epoch": 3.3778173190984577,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038880753938910523,
      "loss": 4.86,
      "mean_token_accuracy": 0.22354026287794113,
      "num_tokens": 78266462.0,
      "step": 34170
    },
    {
      "entropy": 5.3405842781066895,
      "epoch": 3.3783115856069594,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038877709529226446,
      "loss": 4.9237,
      "mean_token_accuracy": 0.21859470009803772,
      "num_tokens": 78278136.0,
      "step": 34175
    },
    {
      "entropy": 5.420118093490601,
      "epoch": 3.3788058521154607,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003887466483964893,
      "loss": 4.9368,
      "mean_token_accuracy": 0.215442955493927,
      "num_tokens": 78290046.0,
      "step": 34180
    },
    {
      "entropy": 5.375010347366333,
      "epoch": 3.379300118623962,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00038871619870252904,
      "loss": 4.8909,
      "mean_token_accuracy": 0.22480558604002,
      "num_tokens": 78301636.0,
      "step": 34185
    },
    {
      "entropy": 5.382464838027954,
      "epoch": 3.379794385132463,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038868574621113265,
      "loss": 4.9007,
      "mean_token_accuracy": 0.22122861444950104,
      "num_tokens": 78313326.0,
      "step": 34190
    },
    {
      "entropy": 5.392512369155884,
      "epoch": 3.380288651640965,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003886552909230492,
      "loss": 4.9613,
      "mean_token_accuracy": 0.2138638973236084,
      "num_tokens": 78325214.0,
      "step": 34195
    },
    {
      "entropy": 5.325533723831176,
      "epoch": 3.380782918149466,
      "grad_norm": 1.03125,
      "learning_rate": 0.000388624832839028,
      "loss": 4.8566,
      "mean_token_accuracy": 0.22787874937057495,
      "num_tokens": 78336461.0,
      "step": 34200
    },
    {
      "entropy": 5.374352645874024,
      "epoch": 3.3812771846579674,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038859437195981833,
      "loss": 4.9311,
      "mean_token_accuracy": 0.22009425163269042,
      "num_tokens": 78347852.0,
      "step": 34205
    },
    {
      "entropy": 5.428445720672608,
      "epoch": 3.381771451166469,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038856390828616945,
      "loss": 4.9114,
      "mean_token_accuracy": 0.21112205386161803,
      "num_tokens": 78358736.0,
      "step": 34210
    },
    {
      "entropy": 5.468721151351929,
      "epoch": 3.3822657176749704,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003885334418188308,
      "loss": 5.0686,
      "mean_token_accuracy": 0.21147384643554687,
      "num_tokens": 78370842.0,
      "step": 34215
    },
    {
      "entropy": 5.297071504592895,
      "epoch": 3.3827599841834717,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038850297255855186,
      "loss": 4.8045,
      "mean_token_accuracy": 0.22888566106557845,
      "num_tokens": 78382159.0,
      "step": 34220
    },
    {
      "entropy": 5.385731363296509,
      "epoch": 3.383254250691973,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038847250050608217,
      "loss": 5.0142,
      "mean_token_accuracy": 0.21420636773109436,
      "num_tokens": 78392130.0,
      "step": 34225
    },
    {
      "entropy": 5.373178958892822,
      "epoch": 3.3837485172004746,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038844202566217146,
      "loss": 4.9099,
      "mean_token_accuracy": 0.22126138508319854,
      "num_tokens": 78403271.0,
      "step": 34230
    },
    {
      "entropy": 5.392863702774048,
      "epoch": 3.384242783708976,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038841154802756925,
      "loss": 4.9223,
      "mean_token_accuracy": 0.21964584290981293,
      "num_tokens": 78414621.0,
      "step": 34235
    },
    {
      "entropy": 5.493626737594605,
      "epoch": 3.384737050217477,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00038838106760302537,
      "loss": 5.0524,
      "mean_token_accuracy": 0.21165242195129394,
      "num_tokens": 78427708.0,
      "step": 34240
    },
    {
      "entropy": 5.383589410781861,
      "epoch": 3.385231316725979,
      "grad_norm": 1.0,
      "learning_rate": 0.00038835058438928965,
      "loss": 4.8728,
      "mean_token_accuracy": 0.2254788801074028,
      "num_tokens": 78438736.0,
      "step": 34245
    },
    {
      "entropy": 5.31712965965271,
      "epoch": 3.38572558323448,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000388320098387112,
      "loss": 4.9109,
      "mean_token_accuracy": 0.22821611613035203,
      "num_tokens": 78449118.0,
      "step": 34250
    },
    {
      "entropy": 5.398679590225219,
      "epoch": 3.3862198497429814,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038828960959724225,
      "loss": 4.9621,
      "mean_token_accuracy": 0.2192313238978386,
      "num_tokens": 78460027.0,
      "step": 34255
    },
    {
      "entropy": 5.327192926406861,
      "epoch": 3.3867141162514827,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038825911802043057,
      "loss": 4.8343,
      "mean_token_accuracy": 0.22823193818330764,
      "num_tokens": 78471251.0,
      "step": 34260
    },
    {
      "entropy": 5.385414695739746,
      "epoch": 3.3872083827599844,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038822862365742704,
      "loss": 4.9488,
      "mean_token_accuracy": 0.2186719924211502,
      "num_tokens": 78482042.0,
      "step": 34265
    },
    {
      "entropy": 5.330732965469361,
      "epoch": 3.3877026492684856,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003881981265089818,
      "loss": 4.8427,
      "mean_token_accuracy": 0.2274144247174263,
      "num_tokens": 78492941.0,
      "step": 34270
    },
    {
      "entropy": 5.3614898204803465,
      "epoch": 3.388196915776987,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003881676265758451,
      "loss": 4.9176,
      "mean_token_accuracy": 0.21566414088010788,
      "num_tokens": 78506108.0,
      "step": 34275
    },
    {
      "entropy": 5.390552663803101,
      "epoch": 3.388691182285488,
      "grad_norm": 0.96875,
      "learning_rate": 0.00038813712385876714,
      "loss": 4.8909,
      "mean_token_accuracy": 0.22801575660705567,
      "num_tokens": 78517784.0,
      "step": 34280
    },
    {
      "entropy": 5.361949157714844,
      "epoch": 3.38918544879399,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038810661835849845,
      "loss": 4.951,
      "mean_token_accuracy": 0.21690807938575746,
      "num_tokens": 78528795.0,
      "step": 34285
    },
    {
      "entropy": 5.368957281112671,
      "epoch": 3.389679715302491,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00038807611007578947,
      "loss": 4.8305,
      "mean_token_accuracy": 0.23518289476633072,
      "num_tokens": 78540724.0,
      "step": 34290
    },
    {
      "entropy": 5.4847158908844,
      "epoch": 3.3901739818109924,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003880455990113905,
      "loss": 5.0081,
      "mean_token_accuracy": 0.21639263778924941,
      "num_tokens": 78552569.0,
      "step": 34295
    },
    {
      "entropy": 5.361520385742187,
      "epoch": 3.3906682483194936,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003880150851660524,
      "loss": 4.9444,
      "mean_token_accuracy": 0.22136489003896714,
      "num_tokens": 78564785.0,
      "step": 34300
    },
    {
      "entropy": 5.3645061492919925,
      "epoch": 3.3911625148279954,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038798456854052555,
      "loss": 4.9401,
      "mean_token_accuracy": 0.2191367119550705,
      "num_tokens": 78576323.0,
      "step": 34305
    },
    {
      "entropy": 5.4595530986785885,
      "epoch": 3.3916567813364966,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038795404913556083,
      "loss": 4.996,
      "mean_token_accuracy": 0.2143404632806778,
      "num_tokens": 78586978.0,
      "step": 34310
    },
    {
      "entropy": 5.390563440322876,
      "epoch": 3.392151047844998,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038792352695190895,
      "loss": 4.9493,
      "mean_token_accuracy": 0.22530075609683992,
      "num_tokens": 78597591.0,
      "step": 34315
    },
    {
      "entropy": 5.368131828308106,
      "epoch": 3.3926453143534996,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003878930019903209,
      "loss": 4.8681,
      "mean_token_accuracy": 0.22552868723869324,
      "num_tokens": 78608189.0,
      "step": 34320
    },
    {
      "entropy": 5.395864534378052,
      "epoch": 3.393139580862001,
      "grad_norm": 1.125,
      "learning_rate": 0.0003878624742515474,
      "loss": 4.9676,
      "mean_token_accuracy": 0.22675816267728804,
      "num_tokens": 78620170.0,
      "step": 34325
    },
    {
      "entropy": 5.408623838424683,
      "epoch": 3.393633847370502,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038783194373633956,
      "loss": 4.9327,
      "mean_token_accuracy": 0.22023035883903502,
      "num_tokens": 78632423.0,
      "step": 34330
    },
    {
      "entropy": 5.380444431304932,
      "epoch": 3.3941281138790034,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038780141044544836,
      "loss": 4.9527,
      "mean_token_accuracy": 0.22786918729543687,
      "num_tokens": 78644138.0,
      "step": 34335
    },
    {
      "entropy": 5.3805702209472654,
      "epoch": 3.394622380387505,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000387770874379625,
      "loss": 4.9427,
      "mean_token_accuracy": 0.21810621470212938,
      "num_tokens": 78655659.0,
      "step": 34340
    },
    {
      "entropy": 5.302137327194214,
      "epoch": 3.3951166468960063,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003877403355396206,
      "loss": 4.8884,
      "mean_token_accuracy": 0.22721552848815918,
      "num_tokens": 78665625.0,
      "step": 34345
    },
    {
      "entropy": 5.377222537994385,
      "epoch": 3.3956109134045076,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003877097939261865,
      "loss": 4.919,
      "mean_token_accuracy": 0.22234344482421875,
      "num_tokens": 78677253.0,
      "step": 34350
    },
    {
      "entropy": 5.43311448097229,
      "epoch": 3.3961051799130093,
      "grad_norm": 1.0,
      "learning_rate": 0.00038767924954007387,
      "loss": 4.9034,
      "mean_token_accuracy": 0.22785402536392213,
      "num_tokens": 78688574.0,
      "step": 34355
    },
    {
      "entropy": 5.361625242233276,
      "epoch": 3.3965994464215106,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003876487023820343,
      "loss": 4.8596,
      "mean_token_accuracy": 0.2304435685276985,
      "num_tokens": 78700537.0,
      "step": 34360
    },
    {
      "entropy": 5.290830659866333,
      "epoch": 3.397093712930012,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038761815245281925,
      "loss": 4.8416,
      "mean_token_accuracy": 0.23075627982616426,
      "num_tokens": 78710257.0,
      "step": 34365
    },
    {
      "entropy": 5.364655303955078,
      "epoch": 3.397587979438513,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038758759975318006,
      "loss": 4.8775,
      "mean_token_accuracy": 0.2282530203461647,
      "num_tokens": 78722503.0,
      "step": 34370
    },
    {
      "entropy": 5.355743408203125,
      "epoch": 3.398082245947015,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003875570442838686,
      "loss": 4.9273,
      "mean_token_accuracy": 0.22537087351083757,
      "num_tokens": 78734476.0,
      "step": 34375
    },
    {
      "entropy": 5.4768712520599365,
      "epoch": 3.398576512455516,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003875264860456362,
      "loss": 4.9847,
      "mean_token_accuracy": 0.2231973811984062,
      "num_tokens": 78745685.0,
      "step": 34380
    },
    {
      "entropy": 5.444960737228394,
      "epoch": 3.3990707789640173,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00038749592503923484,
      "loss": 4.9647,
      "mean_token_accuracy": 0.21849963963031768,
      "num_tokens": 78758977.0,
      "step": 34385
    },
    {
      "entropy": 5.344271469116211,
      "epoch": 3.399565045472519,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003874653612654164,
      "loss": 4.893,
      "mean_token_accuracy": 0.2171176090836525,
      "num_tokens": 78770702.0,
      "step": 34390
    },
    {
      "entropy": 5.367283582687378,
      "epoch": 3.4000593119810203,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003874347947249325,
      "loss": 4.9295,
      "mean_token_accuracy": 0.21808202415704728,
      "num_tokens": 78781763.0,
      "step": 34395
    },
    {
      "entropy": 5.442790508270264,
      "epoch": 3.4005535784895216,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038740422541853523,
      "loss": 4.9304,
      "mean_token_accuracy": 0.22123486399650574,
      "num_tokens": 78792341.0,
      "step": 34400
    },
    {
      "entropy": 5.375040102005005,
      "epoch": 3.401047844998023,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00038737365334697656,
      "loss": 4.8727,
      "mean_token_accuracy": 0.2255553662776947,
      "num_tokens": 78803947.0,
      "step": 34405
    },
    {
      "entropy": 5.2559435844421385,
      "epoch": 3.401542111506524,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038734307851100863,
      "loss": 4.7495,
      "mean_token_accuracy": 0.23775913566350937,
      "num_tokens": 78815503.0,
      "step": 34410
    },
    {
      "entropy": 5.3592897891998295,
      "epoch": 3.402036378015026,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003873125009113835,
      "loss": 4.9463,
      "mean_token_accuracy": 0.2192987620830536,
      "num_tokens": 78826206.0,
      "step": 34415
    },
    {
      "entropy": 5.3875932693481445,
      "epoch": 3.402530644523527,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038728192054885346,
      "loss": 4.8755,
      "mean_token_accuracy": 0.22509803175926207,
      "num_tokens": 78837472.0,
      "step": 34420
    },
    {
      "entropy": 5.453868627548218,
      "epoch": 3.4030249110320283,
      "grad_norm": 1.0,
      "learning_rate": 0.0003872513374241707,
      "loss": 5.0213,
      "mean_token_accuracy": 0.21520039886236192,
      "num_tokens": 78848005.0,
      "step": 34425
    },
    {
      "entropy": 5.37598614692688,
      "epoch": 3.40351917754053,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038722075153808764,
      "loss": 4.8834,
      "mean_token_accuracy": 0.22352307438850402,
      "num_tokens": 78859577.0,
      "step": 34430
    },
    {
      "entropy": 5.373977565765381,
      "epoch": 3.4040134440490313,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003871901628913568,
      "loss": 4.9316,
      "mean_token_accuracy": 0.21942528188228608,
      "num_tokens": 78871527.0,
      "step": 34435
    },
    {
      "entropy": 5.359360313415527,
      "epoch": 3.4045077105575325,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003871595714847304,
      "loss": 4.918,
      "mean_token_accuracy": 0.22698676586151123,
      "num_tokens": 78882557.0,
      "step": 34440
    },
    {
      "entropy": 5.340625,
      "epoch": 3.405001977066034,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038712897731896126,
      "loss": 4.8864,
      "mean_token_accuracy": 0.22179235517978668,
      "num_tokens": 78893191.0,
      "step": 34445
    },
    {
      "entropy": 5.380088853836059,
      "epoch": 3.4054962435745355,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003870983803948017,
      "loss": 4.9413,
      "mean_token_accuracy": 0.22095131874084473,
      "num_tokens": 78904397.0,
      "step": 34450
    },
    {
      "entropy": 5.421801137924194,
      "epoch": 3.4059905100830368,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003870677807130047,
      "loss": 4.9357,
      "mean_token_accuracy": 0.2246629536151886,
      "num_tokens": 78916765.0,
      "step": 34455
    },
    {
      "entropy": 5.434466028213501,
      "epoch": 3.406484776591538,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000387037178274323,
      "loss": 4.8977,
      "mean_token_accuracy": 0.21882708221673966,
      "num_tokens": 78929745.0,
      "step": 34460
    },
    {
      "entropy": 5.350763845443725,
      "epoch": 3.4069790431000397,
      "grad_norm": 0.96875,
      "learning_rate": 0.00038700657307950926,
      "loss": 4.8943,
      "mean_token_accuracy": 0.2250893533229828,
      "num_tokens": 78941158.0,
      "step": 34465
    },
    {
      "entropy": 5.315621042251587,
      "epoch": 3.407473309608541,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038697596512931643,
      "loss": 4.8717,
      "mean_token_accuracy": 0.221297724545002,
      "num_tokens": 78953465.0,
      "step": 34470
    },
    {
      "entropy": 5.384997844696045,
      "epoch": 3.4079675761170423,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003869453544244975,
      "loss": 4.9159,
      "mean_token_accuracy": 0.22321389764547347,
      "num_tokens": 78964841.0,
      "step": 34475
    },
    {
      "entropy": 5.407142591476441,
      "epoch": 3.4084618426255435,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038691474096580545,
      "loss": 4.9232,
      "mean_token_accuracy": 0.226350137591362,
      "num_tokens": 78975173.0,
      "step": 34480
    },
    {
      "entropy": 5.343042802810669,
      "epoch": 3.4089561091340452,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003868841247539934,
      "loss": 4.9084,
      "mean_token_accuracy": 0.22551754862070084,
      "num_tokens": 78986020.0,
      "step": 34485
    },
    {
      "entropy": 5.375292253494263,
      "epoch": 3.4094503756425465,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003868535057898146,
      "loss": 4.8983,
      "mean_token_accuracy": 0.2185765489935875,
      "num_tokens": 78997906.0,
      "step": 34490
    },
    {
      "entropy": 5.45292067527771,
      "epoch": 3.4099446421510478,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038682288407402215,
      "loss": 4.9785,
      "mean_token_accuracy": 0.21217100322246552,
      "num_tokens": 79009322.0,
      "step": 34495
    },
    {
      "entropy": 5.313228178024292,
      "epoch": 3.4104389086595495,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003867922596073694,
      "loss": 4.8306,
      "mean_token_accuracy": 0.22980786710977555,
      "num_tokens": 79020940.0,
      "step": 34500
    },
    {
      "entropy": 5.408861112594605,
      "epoch": 3.4109331751680507,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003867616323906098,
      "loss": 4.9656,
      "mean_token_accuracy": 0.21851980835199356,
      "num_tokens": 79033002.0,
      "step": 34505
    },
    {
      "entropy": 5.290494441986084,
      "epoch": 3.411427441676552,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00038673100242449666,
      "loss": 4.8414,
      "mean_token_accuracy": 0.22964343577623367,
      "num_tokens": 79044129.0,
      "step": 34510
    },
    {
      "entropy": 5.330547046661377,
      "epoch": 3.4119217081850532,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003867003697097835,
      "loss": 4.8822,
      "mean_token_accuracy": 0.22968688160181044,
      "num_tokens": 79054961.0,
      "step": 34515
    },
    {
      "entropy": 5.431367444992065,
      "epoch": 3.412415974693555,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003866697342472239,
      "loss": 4.9456,
      "mean_token_accuracy": 0.21347442418336868,
      "num_tokens": 79066136.0,
      "step": 34520
    },
    {
      "entropy": 5.32498140335083,
      "epoch": 3.412910241202056,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003866390960375716,
      "loss": 4.931,
      "mean_token_accuracy": 0.2204502761363983,
      "num_tokens": 79077821.0,
      "step": 34525
    },
    {
      "entropy": 5.356904792785644,
      "epoch": 3.4134045077105575,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003866084550815801,
      "loss": 4.8688,
      "mean_token_accuracy": 0.22440022230148315,
      "num_tokens": 79087554.0,
      "step": 34530
    },
    {
      "entropy": 5.471297311782837,
      "epoch": 3.4138987742190587,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038657781138000337,
      "loss": 4.9972,
      "mean_token_accuracy": 0.2158384919166565,
      "num_tokens": 79098434.0,
      "step": 34535
    },
    {
      "entropy": 5.392957067489624,
      "epoch": 3.4143930407275604,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038654716493359515,
      "loss": 4.8861,
      "mean_token_accuracy": 0.2250680595636368,
      "num_tokens": 79108962.0,
      "step": 34540
    },
    {
      "entropy": 5.291246938705444,
      "epoch": 3.4148873072360617,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003865165157431093,
      "loss": 4.8818,
      "mean_token_accuracy": 0.22493674904108046,
      "num_tokens": 79119592.0,
      "step": 34545
    },
    {
      "entropy": 5.365667390823364,
      "epoch": 3.415381573744563,
      "grad_norm": 1.0,
      "learning_rate": 0.0003864858638093,
      "loss": 4.849,
      "mean_token_accuracy": 0.2237905368208885,
      "num_tokens": 79130170.0,
      "step": 34550
    },
    {
      "entropy": 5.394690465927124,
      "epoch": 3.4158758402530642,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038645520913292117,
      "loss": 4.904,
      "mean_token_accuracy": 0.2250271201133728,
      "num_tokens": 79140949.0,
      "step": 34555
    },
    {
      "entropy": 5.3765043258667,
      "epoch": 3.416370106761566,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038642455171472676,
      "loss": 4.915,
      "mean_token_accuracy": 0.21994384974241257,
      "num_tokens": 79151508.0,
      "step": 34560
    },
    {
      "entropy": 5.407494068145752,
      "epoch": 3.416864373270067,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003863938915554711,
      "loss": 4.9224,
      "mean_token_accuracy": 0.22397559881210327,
      "num_tokens": 79163062.0,
      "step": 34565
    },
    {
      "entropy": 5.387100553512573,
      "epoch": 3.4173586397785685,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003863632286559085,
      "loss": 4.9053,
      "mean_token_accuracy": 0.22685438096523286,
      "num_tokens": 79174025.0,
      "step": 34570
    },
    {
      "entropy": 5.343644475936889,
      "epoch": 3.41785290628707,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003863325630167932,
      "loss": 4.801,
      "mean_token_accuracy": 0.23368772864341736,
      "num_tokens": 79185169.0,
      "step": 34575
    },
    {
      "entropy": 5.3502466678619385,
      "epoch": 3.4183471727955714,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038630189463887955,
      "loss": 4.9055,
      "mean_token_accuracy": 0.22347963154315947,
      "num_tokens": 79196503.0,
      "step": 34580
    },
    {
      "entropy": 5.3405845165252686,
      "epoch": 3.4188414393040727,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038627122352292206,
      "loss": 4.8976,
      "mean_token_accuracy": 0.2198534280061722,
      "num_tokens": 79207887.0,
      "step": 34585
    },
    {
      "entropy": 5.340464353561401,
      "epoch": 3.419335705812574,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038624054966967516,
      "loss": 4.8637,
      "mean_token_accuracy": 0.219783715903759,
      "num_tokens": 79219339.0,
      "step": 34590
    },
    {
      "entropy": 5.368650817871094,
      "epoch": 3.4198299723210757,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00038620987307989347,
      "loss": 4.9314,
      "mean_token_accuracy": 0.21905310004949569,
      "num_tokens": 79231454.0,
      "step": 34595
    },
    {
      "entropy": 5.426078033447266,
      "epoch": 3.420324238829577,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038617919375433176,
      "loss": 4.9058,
      "mean_token_accuracy": 0.22011983394622803,
      "num_tokens": 79241761.0,
      "step": 34600
    },
    {
      "entropy": 5.402372264862061,
      "epoch": 3.420818505338078,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00038614851169374456,
      "loss": 4.8793,
      "mean_token_accuracy": 0.22196827083826065,
      "num_tokens": 79252272.0,
      "step": 34605
    },
    {
      "entropy": 5.34308762550354,
      "epoch": 3.42131277184658,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003861178268988867,
      "loss": 4.8583,
      "mean_token_accuracy": 0.2242792949080467,
      "num_tokens": 79263612.0,
      "step": 34610
    },
    {
      "entropy": 5.294901704788208,
      "epoch": 3.421807038355081,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003860871393705131,
      "loss": 4.8686,
      "mean_token_accuracy": 0.22705257683992386,
      "num_tokens": 79275361.0,
      "step": 34615
    },
    {
      "entropy": 5.377409887313843,
      "epoch": 3.4223013048635824,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038605644910937856,
      "loss": 4.8927,
      "mean_token_accuracy": 0.2278641551733017,
      "num_tokens": 79285277.0,
      "step": 34620
    },
    {
      "entropy": 5.250478315353393,
      "epoch": 3.4227955713720837,
      "grad_norm": 1.0,
      "learning_rate": 0.00038602575611623827,
      "loss": 4.75,
      "mean_token_accuracy": 0.23823626190423966,
      "num_tokens": 79296713.0,
      "step": 34625
    },
    {
      "entropy": 5.37995138168335,
      "epoch": 3.4232898378805854,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000385995060391847,
      "loss": 4.9246,
      "mean_token_accuracy": 0.21905192136764526,
      "num_tokens": 79308363.0,
      "step": 34630
    },
    {
      "entropy": 5.390201377868652,
      "epoch": 3.4237841043890866,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038596436193696015,
      "loss": 4.9834,
      "mean_token_accuracy": 0.2163274735212326,
      "num_tokens": 79319626.0,
      "step": 34635
    },
    {
      "entropy": 5.362008285522461,
      "epoch": 3.424278370897588,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003859336607523327,
      "loss": 4.8872,
      "mean_token_accuracy": 0.22105125188827515,
      "num_tokens": 79330628.0,
      "step": 34640
    },
    {
      "entropy": 5.316167545318604,
      "epoch": 3.424772637406089,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038590295683872005,
      "loss": 4.8742,
      "mean_token_accuracy": 0.22724528908729552,
      "num_tokens": 79342323.0,
      "step": 34645
    },
    {
      "entropy": 5.290068578720093,
      "epoch": 3.425266903914591,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003858722501968774,
      "loss": 4.7944,
      "mean_token_accuracy": 0.23574014753103256,
      "num_tokens": 79354108.0,
      "step": 34650
    },
    {
      "entropy": 5.361473560333252,
      "epoch": 3.425761170423092,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00038584154082756036,
      "loss": 4.9303,
      "mean_token_accuracy": 0.22985867857933046,
      "num_tokens": 79365494.0,
      "step": 34655
    },
    {
      "entropy": 5.265077400207519,
      "epoch": 3.4262554369315934,
      "grad_norm": 1.078125,
      "learning_rate": 0.000385810828731524,
      "loss": 4.7353,
      "mean_token_accuracy": 0.23472481071949006,
      "num_tokens": 79376881.0,
      "step": 34660
    },
    {
      "entropy": 5.3192946434021,
      "epoch": 3.4267497034400947,
      "grad_norm": 1.0,
      "learning_rate": 0.0003857801139095242,
      "loss": 4.844,
      "mean_token_accuracy": 0.23531267940998077,
      "num_tokens": 79388154.0,
      "step": 34665
    },
    {
      "entropy": 5.429132413864136,
      "epoch": 3.4272439699485964,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038574939636231635,
      "loss": 4.9345,
      "mean_token_accuracy": 0.22201031744480132,
      "num_tokens": 79398995.0,
      "step": 34670
    },
    {
      "entropy": 5.389501714706421,
      "epoch": 3.4277382364570976,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003857186760906562,
      "loss": 4.9626,
      "mean_token_accuracy": 0.22202810943126677,
      "num_tokens": 79410697.0,
      "step": 34675
    },
    {
      "entropy": 5.379842329025268,
      "epoch": 3.428232502965599,
      "grad_norm": 0.96875,
      "learning_rate": 0.00038568795309529947,
      "loss": 4.9337,
      "mean_token_accuracy": 0.22680349946022033,
      "num_tokens": 79421862.0,
      "step": 34680
    },
    {
      "entropy": 5.418475246429443,
      "epoch": 3.4287267694741006,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038565722737700185,
      "loss": 4.9212,
      "mean_token_accuracy": 0.2204930379986763,
      "num_tokens": 79432610.0,
      "step": 34685
    },
    {
      "entropy": 5.418197345733643,
      "epoch": 3.429221035982602,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038562649893651924,
      "loss": 4.9427,
      "mean_token_accuracy": 0.22385623604059218,
      "num_tokens": 79443386.0,
      "step": 34690
    },
    {
      "entropy": 5.38770432472229,
      "epoch": 3.429715302491103,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003855957677746077,
      "loss": 4.9573,
      "mean_token_accuracy": 0.22566352039575577,
      "num_tokens": 79454334.0,
      "step": 34695
    },
    {
      "entropy": 5.453788328170776,
      "epoch": 3.4302095689996044,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003855650338920231,
      "loss": 4.9028,
      "mean_token_accuracy": 0.2223539501428604,
      "num_tokens": 79465489.0,
      "step": 34700
    },
    {
      "entropy": 5.404337596893311,
      "epoch": 3.430703835508106,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038553429728952144,
      "loss": 4.9346,
      "mean_token_accuracy": 0.21787726432085036,
      "num_tokens": 79477293.0,
      "step": 34705
    },
    {
      "entropy": 5.407230234146118,
      "epoch": 3.4311981020166074,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038550355796785897,
      "loss": 4.9782,
      "mean_token_accuracy": 0.21467052698135375,
      "num_tokens": 79488995.0,
      "step": 34710
    },
    {
      "entropy": 5.432978725433349,
      "epoch": 3.4316923685251086,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038547281592779184,
      "loss": 5.0104,
      "mean_token_accuracy": 0.21857332140207292,
      "num_tokens": 79502757.0,
      "step": 34715
    },
    {
      "entropy": 5.412739610671997,
      "epoch": 3.4321866350336103,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003854420711700763,
      "loss": 4.9399,
      "mean_token_accuracy": 0.2203126460313797,
      "num_tokens": 79515220.0,
      "step": 34720
    },
    {
      "entropy": 5.452446508407593,
      "epoch": 3.4326809015421116,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038541132369546865,
      "loss": 4.9898,
      "mean_token_accuracy": 0.2169725701212883,
      "num_tokens": 79527577.0,
      "step": 34725
    },
    {
      "entropy": 5.38204870223999,
      "epoch": 3.433175168050613,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003853805735047253,
      "loss": 4.8912,
      "mean_token_accuracy": 0.2210562616586685,
      "num_tokens": 79539136.0,
      "step": 34730
    },
    {
      "entropy": 5.315843343734741,
      "epoch": 3.433669434559114,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038534982059860274,
      "loss": 4.8004,
      "mean_token_accuracy": 0.237005053460598,
      "num_tokens": 79550123.0,
      "step": 34735
    },
    {
      "entropy": 5.422584962844849,
      "epoch": 3.434163701067616,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038531906497785753,
      "loss": 5.0047,
      "mean_token_accuracy": 0.21353789418935776,
      "num_tokens": 79563072.0,
      "step": 34740
    },
    {
      "entropy": 5.391173458099365,
      "epoch": 3.434657967576117,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038528830664324613,
      "loss": 4.8913,
      "mean_token_accuracy": 0.2232232227921486,
      "num_tokens": 79573901.0,
      "step": 34745
    },
    {
      "entropy": 5.329513835906982,
      "epoch": 3.4351522340846183,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038525754559552526,
      "loss": 4.8869,
      "mean_token_accuracy": 0.22773543000221252,
      "num_tokens": 79585384.0,
      "step": 34750
    },
    {
      "entropy": 5.391636943817138,
      "epoch": 3.43564650059312,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003852267818354517,
      "loss": 4.9792,
      "mean_token_accuracy": 0.2183568462729454,
      "num_tokens": 79597277.0,
      "step": 34755
    },
    {
      "entropy": 5.433878660202026,
      "epoch": 3.4361407671016213,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003851960153637822,
      "loss": 5.0053,
      "mean_token_accuracy": 0.21485636085271836,
      "num_tokens": 79609797.0,
      "step": 34760
    },
    {
      "entropy": 5.384763097763061,
      "epoch": 3.4366350336101226,
      "grad_norm": 0.96875,
      "learning_rate": 0.00038516524618127364,
      "loss": 4.897,
      "mean_token_accuracy": 0.2271730661392212,
      "num_tokens": 79621247.0,
      "step": 34765
    },
    {
      "entropy": 5.423787307739258,
      "epoch": 3.437129300118624,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003851344742886829,
      "loss": 4.9656,
      "mean_token_accuracy": 0.21451084911823273,
      "num_tokens": 79632892.0,
      "step": 34770
    },
    {
      "entropy": 5.395720672607422,
      "epoch": 3.4376235666271255,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00038510369968676705,
      "loss": 4.9566,
      "mean_token_accuracy": 0.2155747652053833,
      "num_tokens": 79644426.0,
      "step": 34775
    },
    {
      "entropy": 5.377327156066895,
      "epoch": 3.438117833135627,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003850729223762831,
      "loss": 4.8378,
      "mean_token_accuracy": 0.23031335026025773,
      "num_tokens": 79656414.0,
      "step": 34780
    },
    {
      "entropy": 5.346804475784301,
      "epoch": 3.438612099644128,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003850421423579882,
      "loss": 4.9233,
      "mean_token_accuracy": 0.22731201350688934,
      "num_tokens": 79667244.0,
      "step": 34785
    },
    {
      "entropy": 5.373436689376831,
      "epoch": 3.4391063661526293,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038501135963263947,
      "loss": 4.9186,
      "mean_token_accuracy": 0.221695414185524,
      "num_tokens": 79677837.0,
      "step": 34790
    },
    {
      "entropy": 5.429844522476197,
      "epoch": 3.439600632661131,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038498057420099425,
      "loss": 4.9699,
      "mean_token_accuracy": 0.21675306260585786,
      "num_tokens": 79689500.0,
      "step": 34795
    },
    {
      "entropy": 5.3419231414794925,
      "epoch": 3.4400948991696323,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038494978606380984,
      "loss": 4.8506,
      "mean_token_accuracy": 0.22827258706092834,
      "num_tokens": 79701621.0,
      "step": 34800
    },
    {
      "entropy": 5.306184768676758,
      "epoch": 3.4405891656781336,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003849189952218436,
      "loss": 4.913,
      "mean_token_accuracy": 0.2285364493727684,
      "num_tokens": 79712635.0,
      "step": 34805
    },
    {
      "entropy": 5.434918737411499,
      "epoch": 3.441083432186635,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003848882016758531,
      "loss": 4.9763,
      "mean_token_accuracy": 0.21368080526590347,
      "num_tokens": 79723626.0,
      "step": 34810
    },
    {
      "entropy": 5.40279598236084,
      "epoch": 3.4415776986951365,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038485740542659573,
      "loss": 4.9526,
      "mean_token_accuracy": 0.2160220503807068,
      "num_tokens": 79734983.0,
      "step": 34815
    },
    {
      "entropy": 5.429786157608032,
      "epoch": 3.442071965203638,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003848266064748292,
      "loss": 4.9316,
      "mean_token_accuracy": 0.218422794342041,
      "num_tokens": 79746767.0,
      "step": 34820
    },
    {
      "entropy": 5.415806818008423,
      "epoch": 3.442566231712139,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003847958048213111,
      "loss": 4.9324,
      "mean_token_accuracy": 0.2198669731616974,
      "num_tokens": 79759062.0,
      "step": 34825
    },
    {
      "entropy": 5.334010314941406,
      "epoch": 3.4430604982206408,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00038476500046679915,
      "loss": 4.8459,
      "mean_token_accuracy": 0.2314682587981224,
      "num_tokens": 79769962.0,
      "step": 34830
    },
    {
      "entropy": 5.293807888031006,
      "epoch": 3.443554764729142,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038473419341205115,
      "loss": 4.8751,
      "mean_token_accuracy": 0.22174903452396394,
      "num_tokens": 79781279.0,
      "step": 34835
    },
    {
      "entropy": 5.38449444770813,
      "epoch": 3.4440490312376433,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000384703383657825,
      "loss": 4.9469,
      "mean_token_accuracy": 0.22278484851121902,
      "num_tokens": 79792406.0,
      "step": 34840
    },
    {
      "entropy": 5.433879566192627,
      "epoch": 3.4445432977461445,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003846725712048786,
      "loss": 4.9669,
      "mean_token_accuracy": 0.2145937889814377,
      "num_tokens": 79804150.0,
      "step": 34845
    },
    {
      "entropy": 5.377838277816773,
      "epoch": 3.4450375642546462,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003846417560539699,
      "loss": 4.8412,
      "mean_token_accuracy": 0.23103207647800444,
      "num_tokens": 79814267.0,
      "step": 34850
    },
    {
      "entropy": 5.32079849243164,
      "epoch": 3.4455318307631475,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000384610938205857,
      "loss": 4.8888,
      "mean_token_accuracy": 0.22784971445798874,
      "num_tokens": 79824887.0,
      "step": 34855
    },
    {
      "entropy": 5.399475336074829,
      "epoch": 3.4460260972716488,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038458011766129795,
      "loss": 4.9568,
      "mean_token_accuracy": 0.21606847196817397,
      "num_tokens": 79836840.0,
      "step": 34860
    },
    {
      "entropy": 5.358064651489258,
      "epoch": 3.4465203637801505,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038454929442105105,
      "loss": 4.8479,
      "mean_token_accuracy": 0.2297658547759056,
      "num_tokens": 79846876.0,
      "step": 34865
    },
    {
      "entropy": 5.387265205383301,
      "epoch": 3.4470146302886517,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003845184684858746,
      "loss": 4.9453,
      "mean_token_accuracy": 0.21692655980587006,
      "num_tokens": 79859271.0,
      "step": 34870
    },
    {
      "entropy": 5.389416742324829,
      "epoch": 3.447508896797153,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00038448763985652674,
      "loss": 4.9953,
      "mean_token_accuracy": 0.21735395044088363,
      "num_tokens": 79871938.0,
      "step": 34875
    },
    {
      "entropy": 5.443049192428589,
      "epoch": 3.4480031633056543,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003844568085337659,
      "loss": 4.9559,
      "mean_token_accuracy": 0.21973032951354982,
      "num_tokens": 79883202.0,
      "step": 34880
    },
    {
      "entropy": 5.351394319534302,
      "epoch": 3.448497429814156,
      "grad_norm": 0.96875,
      "learning_rate": 0.00038442597451835076,
      "loss": 4.8857,
      "mean_token_accuracy": 0.227353535592556,
      "num_tokens": 79895794.0,
      "step": 34885
    },
    {
      "entropy": 5.375788307189941,
      "epoch": 3.4489916963226572,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003843951378110396,
      "loss": 4.8724,
      "mean_token_accuracy": 0.225347638130188,
      "num_tokens": 79906548.0,
      "step": 34890
    },
    {
      "entropy": 5.363926362991333,
      "epoch": 3.4494859628311585,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000384364298412591,
      "loss": 4.9583,
      "mean_token_accuracy": 0.21985134184360505,
      "num_tokens": 79918116.0,
      "step": 34895
    },
    {
      "entropy": 5.361332845687866,
      "epoch": 3.4499802293396598,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038433345632376376,
      "loss": 4.9503,
      "mean_token_accuracy": 0.22752606272697448,
      "num_tokens": 79930791.0,
      "step": 34900
    },
    {
      "entropy": 5.305279636383057,
      "epoch": 3.4504744958481615,
      "grad_norm": 3.546875,
      "learning_rate": 0.00038430261154531655,
      "loss": 4.8206,
      "mean_token_accuracy": 0.23466309905052185,
      "num_tokens": 79942359.0,
      "step": 34905
    },
    {
      "entropy": 5.327594995498657,
      "epoch": 3.4509687623566627,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003842717640780081,
      "loss": 4.8593,
      "mean_token_accuracy": 0.22191991209983825,
      "num_tokens": 79954424.0,
      "step": 34910
    },
    {
      "entropy": 5.340651988983154,
      "epoch": 3.451463028865164,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038424091392259737,
      "loss": 4.8565,
      "mean_token_accuracy": 0.22657064348459244,
      "num_tokens": 79965704.0,
      "step": 34915
    },
    {
      "entropy": 5.370573329925537,
      "epoch": 3.4519572953736652,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003842100610798431,
      "loss": 4.8827,
      "mean_token_accuracy": 0.22257652282714843,
      "num_tokens": 79977891.0,
      "step": 34920
    },
    {
      "entropy": 5.334797763824463,
      "epoch": 3.452451561882167,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038417920555050447,
      "loss": 4.9096,
      "mean_token_accuracy": 0.22360444962978362,
      "num_tokens": 79988709.0,
      "step": 34925
    },
    {
      "entropy": 5.380033826828003,
      "epoch": 3.452945828390668,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038414834733534046,
      "loss": 4.8938,
      "mean_token_accuracy": 0.22453713715076445,
      "num_tokens": 80000349.0,
      "step": 34930
    },
    {
      "entropy": 5.330774116516113,
      "epoch": 3.4534400948991695,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003841174864351101,
      "loss": 4.7868,
      "mean_token_accuracy": 0.23438074290752411,
      "num_tokens": 80011317.0,
      "step": 34935
    },
    {
      "entropy": 5.375869703292847,
      "epoch": 3.453934361407671,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038408662285057267,
      "loss": 4.9693,
      "mean_token_accuracy": 0.22311795502901077,
      "num_tokens": 80023471.0,
      "step": 34940
    },
    {
      "entropy": 5.4005248069763185,
      "epoch": 3.4544286279161724,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038405575658248737,
      "loss": 4.9064,
      "mean_token_accuracy": 0.2192446067929268,
      "num_tokens": 80035937.0,
      "step": 34945
    },
    {
      "entropy": 5.357674694061279,
      "epoch": 3.4549228944246737,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003840248876316135,
      "loss": 4.8915,
      "mean_token_accuracy": 0.22633458375930787,
      "num_tokens": 80048435.0,
      "step": 34950
    },
    {
      "entropy": 5.340887928009034,
      "epoch": 3.455417160933175,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003839940159987106,
      "loss": 4.8972,
      "mean_token_accuracy": 0.22777303010225297,
      "num_tokens": 80059675.0,
      "step": 34955
    },
    {
      "entropy": 5.449123239517212,
      "epoch": 3.4559114274416767,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038396314168453793,
      "loss": 4.9536,
      "mean_token_accuracy": 0.21783632785081863,
      "num_tokens": 80071232.0,
      "step": 34960
    },
    {
      "entropy": 5.347108364105225,
      "epoch": 3.456405693950178,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003839322646898551,
      "loss": 4.8735,
      "mean_token_accuracy": 0.2254775047302246,
      "num_tokens": 80083331.0,
      "step": 34965
    },
    {
      "entropy": 5.431919908523559,
      "epoch": 3.456899960458679,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038390138501542164,
      "loss": 4.9686,
      "mean_token_accuracy": 0.22217265218496324,
      "num_tokens": 80095883.0,
      "step": 34970
    },
    {
      "entropy": 5.340282011032104,
      "epoch": 3.457394226967181,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003838705026619972,
      "loss": 4.9147,
      "mean_token_accuracy": 0.22169393599033355,
      "num_tokens": 80108098.0,
      "step": 34975
    },
    {
      "entropy": 5.325607490539551,
      "epoch": 3.457888493475682,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003838396176303416,
      "loss": 4.9074,
      "mean_token_accuracy": 0.22087820023298263,
      "num_tokens": 80121184.0,
      "step": 34980
    },
    {
      "entropy": 5.407978153228759,
      "epoch": 3.4583827599841834,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003838087299212144,
      "loss": 4.862,
      "mean_token_accuracy": 0.22226207554340363,
      "num_tokens": 80132862.0,
      "step": 34985
    },
    {
      "entropy": 5.329229879379272,
      "epoch": 3.4588770264926847,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038377783953537565,
      "loss": 4.8866,
      "mean_token_accuracy": 0.22380609363317489,
      "num_tokens": 80143756.0,
      "step": 34990
    },
    {
      "entropy": 5.352037668228149,
      "epoch": 3.4593712930011864,
      "grad_norm": 1.0,
      "learning_rate": 0.0003837469464735851,
      "loss": 4.8866,
      "mean_token_accuracy": 0.22633173763751985,
      "num_tokens": 80154570.0,
      "step": 34995
    },
    {
      "entropy": 5.357908868789673,
      "epoch": 3.4598655595096877,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00038371605073660284,
      "loss": 4.8879,
      "mean_token_accuracy": 0.22478317767381667,
      "num_tokens": 80165968.0,
      "step": 35000
    },
    {
      "entropy": 5.343455982208252,
      "epoch": 3.460359826018189,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038368515232518885,
      "loss": 4.8937,
      "mean_token_accuracy": 0.21958194822072982,
      "num_tokens": 80177584.0,
      "step": 35005
    },
    {
      "entropy": 5.3588173389434814,
      "epoch": 3.4608540925266906,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003836542512401033,
      "loss": 4.9053,
      "mean_token_accuracy": 0.22101389318704606,
      "num_tokens": 80190113.0,
      "step": 35010
    },
    {
      "entropy": 5.433740901947021,
      "epoch": 3.461348359035192,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003836233474821063,
      "loss": 4.9684,
      "mean_token_accuracy": 0.21682923138141633,
      "num_tokens": 80201074.0,
      "step": 35015
    },
    {
      "entropy": 5.335023641586304,
      "epoch": 3.461842625543693,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038359244105195807,
      "loss": 4.9322,
      "mean_token_accuracy": 0.22437410950660705,
      "num_tokens": 80212248.0,
      "step": 35020
    },
    {
      "entropy": 5.335826206207275,
      "epoch": 3.4623368920521944,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038356153195041886,
      "loss": 4.8362,
      "mean_token_accuracy": 0.2207690566778183,
      "num_tokens": 80223151.0,
      "step": 35025
    },
    {
      "entropy": 5.333377981185913,
      "epoch": 3.462831158560696,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038353062017824923,
      "loss": 4.8079,
      "mean_token_accuracy": 0.22640860229730606,
      "num_tokens": 80234978.0,
      "step": 35030
    },
    {
      "entropy": 5.343447303771972,
      "epoch": 3.4633254250691974,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003834997057362096,
      "loss": 4.8664,
      "mean_token_accuracy": 0.22885859459638597,
      "num_tokens": 80246297.0,
      "step": 35035
    },
    {
      "entropy": 5.433044004440307,
      "epoch": 3.4638196915776986,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003834687886250602,
      "loss": 4.9611,
      "mean_token_accuracy": 0.21328379213809967,
      "num_tokens": 80258628.0,
      "step": 35040
    },
    {
      "entropy": 5.386111927032471,
      "epoch": 3.4643139580862,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003834378688455618,
      "loss": 4.8582,
      "mean_token_accuracy": 0.22875118553638457,
      "num_tokens": 80270694.0,
      "step": 35045
    },
    {
      "entropy": 5.32470121383667,
      "epoch": 3.4648082245947016,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003834069463984751,
      "loss": 4.912,
      "mean_token_accuracy": 0.22508965730667113,
      "num_tokens": 80282032.0,
      "step": 35050
    },
    {
      "entropy": 5.334949254989624,
      "epoch": 3.465302491103203,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003833760212845605,
      "loss": 4.8701,
      "mean_token_accuracy": 0.2228470116853714,
      "num_tokens": 80294033.0,
      "step": 35055
    },
    {
      "entropy": 5.346255540847778,
      "epoch": 3.465796757611704,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038334509350457913,
      "loss": 4.7949,
      "mean_token_accuracy": 0.2312503233551979,
      "num_tokens": 80304273.0,
      "step": 35060
    },
    {
      "entropy": 5.250077915191651,
      "epoch": 3.4662910241202054,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003833141630592916,
      "loss": 4.8149,
      "mean_token_accuracy": 0.23373170495033263,
      "num_tokens": 80314049.0,
      "step": 35065
    },
    {
      "entropy": 5.416063451766968,
      "epoch": 3.466785290628707,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038328322994945883,
      "loss": 4.9886,
      "mean_token_accuracy": 0.21375520676374435,
      "num_tokens": 80324632.0,
      "step": 35070
    },
    {
      "entropy": 5.337831830978393,
      "epoch": 3.4672795571372084,
      "grad_norm": 0.9375,
      "learning_rate": 0.00038325229417584194,
      "loss": 4.834,
      "mean_token_accuracy": 0.23133582025766372,
      "num_tokens": 80335824.0,
      "step": 35075
    },
    {
      "entropy": 5.333151960372925,
      "epoch": 3.4677738236457096,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038322135573920176,
      "loss": 4.8556,
      "mean_token_accuracy": 0.21455931216478347,
      "num_tokens": 80346537.0,
      "step": 35080
    },
    {
      "entropy": 5.363255405426026,
      "epoch": 3.4682680901542113,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003831904146402994,
      "loss": 4.8552,
      "mean_token_accuracy": 0.23108550608158113,
      "num_tokens": 80358446.0,
      "step": 35085
    },
    {
      "entropy": 5.311655950546265,
      "epoch": 3.4687623566627126,
      "grad_norm": 0.96875,
      "learning_rate": 0.000383159470879896,
      "loss": 4.87,
      "mean_token_accuracy": 0.23010483235120774,
      "num_tokens": 80370219.0,
      "step": 35090
    },
    {
      "entropy": 5.354640007019043,
      "epoch": 3.469256623171214,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003831285244587529,
      "loss": 4.9134,
      "mean_token_accuracy": 0.22385768592357635,
      "num_tokens": 80380886.0,
      "step": 35095
    },
    {
      "entropy": 5.337466430664063,
      "epoch": 3.469750889679715,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003830975753776314,
      "loss": 4.8335,
      "mean_token_accuracy": 0.22811726480722427,
      "num_tokens": 80391171.0,
      "step": 35100
    },
    {
      "entropy": 5.340829277038575,
      "epoch": 3.470245156188217,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003830666236372927,
      "loss": 4.8661,
      "mean_token_accuracy": 0.2247500643134117,
      "num_tokens": 80403527.0,
      "step": 35105
    },
    {
      "entropy": 5.411605882644653,
      "epoch": 3.470739422696718,
      "grad_norm": 1.0,
      "learning_rate": 0.00038303566923849827,
      "loss": 4.989,
      "mean_token_accuracy": 0.21747735589742662,
      "num_tokens": 80414992.0,
      "step": 35110
    },
    {
      "entropy": 5.359706544876099,
      "epoch": 3.4712336892052194,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038300471218200965,
      "loss": 4.8998,
      "mean_token_accuracy": 0.21905082911252977,
      "num_tokens": 80426167.0,
      "step": 35115
    },
    {
      "entropy": 5.3413280010223385,
      "epoch": 3.471727955713721,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038297375246858835,
      "loss": 4.877,
      "mean_token_accuracy": 0.22599556148052216,
      "num_tokens": 80438063.0,
      "step": 35120
    },
    {
      "entropy": 5.312160158157349,
      "epoch": 3.4722222222222223,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000382942790098996,
      "loss": 4.8972,
      "mean_token_accuracy": 0.21451212614774703,
      "num_tokens": 80449898.0,
      "step": 35125
    },
    {
      "entropy": 5.379366779327393,
      "epoch": 3.4727164887307236,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038291182507399425,
      "loss": 4.8729,
      "mean_token_accuracy": 0.23261666893959046,
      "num_tokens": 80461623.0,
      "step": 35130
    },
    {
      "entropy": 5.3715729236602785,
      "epoch": 3.473210755239225,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00038288085739434486,
      "loss": 4.8516,
      "mean_token_accuracy": 0.2293989896774292,
      "num_tokens": 80472627.0,
      "step": 35135
    },
    {
      "entropy": 5.318172216415405,
      "epoch": 3.4737050217477266,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003828498870608096,
      "loss": 4.8696,
      "mean_token_accuracy": 0.22925539016723634,
      "num_tokens": 80483494.0,
      "step": 35140
    },
    {
      "entropy": 5.32388277053833,
      "epoch": 3.474199288256228,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038281891407415046,
      "loss": 4.843,
      "mean_token_accuracy": 0.22897504419088363,
      "num_tokens": 80495122.0,
      "step": 35145
    },
    {
      "entropy": 5.294313383102417,
      "epoch": 3.474693554764729,
      "grad_norm": 0.96875,
      "learning_rate": 0.00038278793843512935,
      "loss": 4.8565,
      "mean_token_accuracy": 0.23140700459480285,
      "num_tokens": 80506639.0,
      "step": 35150
    },
    {
      "entropy": 5.367994737625122,
      "epoch": 3.4751878212732303,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003827569601445081,
      "loss": 4.895,
      "mean_token_accuracy": 0.22339883744716643,
      "num_tokens": 80517897.0,
      "step": 35155
    },
    {
      "entropy": 5.427991247177124,
      "epoch": 3.475682087781732,
      "grad_norm": 0.953125,
      "learning_rate": 0.000382725979203049,
      "loss": 4.8953,
      "mean_token_accuracy": 0.217093725502491,
      "num_tokens": 80530112.0,
      "step": 35160
    },
    {
      "entropy": 5.369055938720703,
      "epoch": 3.4761763542902333,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00038269499561151407,
      "loss": 4.8436,
      "mean_token_accuracy": 0.23295638114213943,
      "num_tokens": 80541811.0,
      "step": 35165
    },
    {
      "entropy": 5.3344573974609375,
      "epoch": 3.4766706207987346,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038266400937066555,
      "loss": 4.9158,
      "mean_token_accuracy": 0.23322578221559526,
      "num_tokens": 80553107.0,
      "step": 35170
    },
    {
      "entropy": 5.431958436965942,
      "epoch": 3.477164887307236,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038263302048126567,
      "loss": 4.9484,
      "mean_token_accuracy": 0.2215990886092186,
      "num_tokens": 80562916.0,
      "step": 35175
    },
    {
      "entropy": 5.347736883163452,
      "epoch": 3.4776591538157375,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003826020289440768,
      "loss": 4.8682,
      "mean_token_accuracy": 0.22408616542816162,
      "num_tokens": 80573179.0,
      "step": 35180
    },
    {
      "entropy": 5.2815220832824705,
      "epoch": 3.478153420324239,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00038257103475986134,
      "loss": 4.8841,
      "mean_token_accuracy": 0.22795872688293456,
      "num_tokens": 80583494.0,
      "step": 35185
    },
    {
      "entropy": 5.307571887969971,
      "epoch": 3.47864768683274,
      "grad_norm": 0.953125,
      "learning_rate": 0.00038254003792938176,
      "loss": 4.9274,
      "mean_token_accuracy": 0.22886021435260773,
      "num_tokens": 80595926.0,
      "step": 35190
    },
    {
      "entropy": 5.333722496032715,
      "epoch": 3.4791419533412418,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00038250903845340057,
      "loss": 4.8784,
      "mean_token_accuracy": 0.23345627039670944,
      "num_tokens": 80607449.0,
      "step": 35195
    },
    {
      "entropy": 5.375921678543091,
      "epoch": 3.479636219849743,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003824780363326803,
      "loss": 4.8961,
      "mean_token_accuracy": 0.22520331740379335,
      "num_tokens": 80619201.0,
      "step": 35200
    },
    {
      "entropy": 5.394715070724487,
      "epoch": 3.4801304863582443,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038244703156798364,
      "loss": 4.9819,
      "mean_token_accuracy": 0.21950269341468812,
      "num_tokens": 80632068.0,
      "step": 35205
    },
    {
      "entropy": 5.384284973144531,
      "epoch": 3.4806247528667456,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003824160241600734,
      "loss": 4.8777,
      "mean_token_accuracy": 0.216710364818573,
      "num_tokens": 80643569.0,
      "step": 35210
    },
    {
      "entropy": 5.380399131774903,
      "epoch": 3.4811190193752473,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038238501410971226,
      "loss": 4.8776,
      "mean_token_accuracy": 0.22461363971233367,
      "num_tokens": 80656000.0,
      "step": 35215
    },
    {
      "entropy": 5.350134086608887,
      "epoch": 3.4816132858837485,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038235400141766314,
      "loss": 4.8787,
      "mean_token_accuracy": 0.22634776830673217,
      "num_tokens": 80667473.0,
      "step": 35220
    },
    {
      "entropy": 5.488603258132935,
      "epoch": 3.48210755239225,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00038232298608468894,
      "loss": 5.0155,
      "mean_token_accuracy": 0.22015490978956223,
      "num_tokens": 80679322.0,
      "step": 35225
    },
    {
      "entropy": 5.506798124313354,
      "epoch": 3.4826018189007515,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003822919681115526,
      "loss": 5.0154,
      "mean_token_accuracy": 0.21678259074687958,
      "num_tokens": 80690907.0,
      "step": 35230
    },
    {
      "entropy": 5.286512279510498,
      "epoch": 3.4830960854092528,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003822609474990172,
      "loss": 4.8006,
      "mean_token_accuracy": 0.2317992016673088,
      "num_tokens": 80702012.0,
      "step": 35235
    },
    {
      "entropy": 5.443157052993774,
      "epoch": 3.483590351917754,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000382229924247846,
      "loss": 5.0184,
      "mean_token_accuracy": 0.21097750216722488,
      "num_tokens": 80714283.0,
      "step": 35240
    },
    {
      "entropy": 5.322410345077515,
      "epoch": 3.4840846184262553,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003821988983588018,
      "loss": 4.8471,
      "mean_token_accuracy": 0.23458994477987288,
      "num_tokens": 80724964.0,
      "step": 35245
    },
    {
      "entropy": 5.3958815574646,
      "epoch": 3.484578884934757,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038216786983264817,
      "loss": 4.8924,
      "mean_token_accuracy": 0.22375074326992034,
      "num_tokens": 80735934.0,
      "step": 35250
    },
    {
      "entropy": 5.442958974838257,
      "epoch": 3.4850731514432582,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00038213683867014824,
      "loss": 4.9911,
      "mean_token_accuracy": 0.2130389168858528,
      "num_tokens": 80746988.0,
      "step": 35255
    },
    {
      "entropy": 5.424798440933228,
      "epoch": 3.4855674179517595,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00038210580487206563,
      "loss": 4.9724,
      "mean_token_accuracy": 0.2100682571530342,
      "num_tokens": 80759300.0,
      "step": 35260
    },
    {
      "entropy": 5.3273814678192135,
      "epoch": 3.486061684460261,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003820747684391636,
      "loss": 4.9307,
      "mean_token_accuracy": 0.22200098633766174,
      "num_tokens": 80771055.0,
      "step": 35265
    },
    {
      "entropy": 5.30106749534607,
      "epoch": 3.4865559509687625,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003820437293722054,
      "loss": 4.8218,
      "mean_token_accuracy": 0.2328087329864502,
      "num_tokens": 80782803.0,
      "step": 35270
    },
    {
      "entropy": 5.381383800506592,
      "epoch": 3.4870502174772637,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003820126876719551,
      "loss": 4.9052,
      "mean_token_accuracy": 0.22208563685417176,
      "num_tokens": 80793833.0,
      "step": 35275
    },
    {
      "entropy": 5.367293834686279,
      "epoch": 3.487544483985765,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000381981643339176,
      "loss": 4.8708,
      "mean_token_accuracy": 0.23207556009292601,
      "num_tokens": 80804800.0,
      "step": 35280
    },
    {
      "entropy": 5.401548528671265,
      "epoch": 3.4880387504942663,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003819505963746318,
      "loss": 4.8888,
      "mean_token_accuracy": 0.22003807127475739,
      "num_tokens": 80815934.0,
      "step": 35285
    },
    {
      "entropy": 5.386237049102784,
      "epoch": 3.488533017002768,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003819195467790864,
      "loss": 4.9165,
      "mean_token_accuracy": 0.21976712346076965,
      "num_tokens": 80826971.0,
      "step": 35290
    },
    {
      "entropy": 5.361319065093994,
      "epoch": 3.4890272835112692,
      "grad_norm": 1.0,
      "learning_rate": 0.0003818884945533036,
      "loss": 4.9559,
      "mean_token_accuracy": 0.21476338803768158,
      "num_tokens": 80839847.0,
      "step": 35295
    },
    {
      "entropy": 5.41700439453125,
      "epoch": 3.4895215500197705,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003818574396980472,
      "loss": 4.9214,
      "mean_token_accuracy": 0.22057391405105592,
      "num_tokens": 80852937.0,
      "step": 35300
    },
    {
      "entropy": 5.380382776260376,
      "epoch": 3.490015816528272,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003818263822140812,
      "loss": 4.8456,
      "mean_token_accuracy": 0.22506628483533858,
      "num_tokens": 80862978.0,
      "step": 35305
    },
    {
      "entropy": 5.336280632019043,
      "epoch": 3.4905100830367735,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038179532210216974,
      "loss": 4.8835,
      "mean_token_accuracy": 0.2255462199449539,
      "num_tokens": 80874446.0,
      "step": 35310
    },
    {
      "entropy": 5.299367380142212,
      "epoch": 3.4910043495452747,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003817642593630767,
      "loss": 4.9428,
      "mean_token_accuracy": 0.22643188685178756,
      "num_tokens": 80886289.0,
      "step": 35315
    },
    {
      "entropy": 5.38615984916687,
      "epoch": 3.491498616053776,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00038173319399756626,
      "loss": 4.9259,
      "mean_token_accuracy": 0.2203057363629341,
      "num_tokens": 80897363.0,
      "step": 35320
    },
    {
      "entropy": 5.435445070266724,
      "epoch": 3.4919928825622777,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003817021260064028,
      "loss": 4.9101,
      "mean_token_accuracy": 0.22135420739650727,
      "num_tokens": 80907850.0,
      "step": 35325
    },
    {
      "entropy": 5.39670958518982,
      "epoch": 3.492487149070779,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003816710553903504,
      "loss": 4.9232,
      "mean_token_accuracy": 0.21810300797224044,
      "num_tokens": 80918512.0,
      "step": 35330
    },
    {
      "entropy": 5.359760379791259,
      "epoch": 3.49298141557928,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003816399821501736,
      "loss": 4.8988,
      "mean_token_accuracy": 0.2290360614657402,
      "num_tokens": 80929550.0,
      "step": 35335
    },
    {
      "entropy": 5.338519430160522,
      "epoch": 3.493475682087782,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00038160890628663663,
      "loss": 4.8717,
      "mean_token_accuracy": 0.22589582651853563,
      "num_tokens": 80941129.0,
      "step": 35340
    },
    {
      "entropy": 5.360996532440185,
      "epoch": 3.493969948596283,
      "grad_norm": 0.953125,
      "learning_rate": 0.000381577827800504,
      "loss": 4.8599,
      "mean_token_accuracy": 0.22563497126102447,
      "num_tokens": 80951804.0,
      "step": 35345
    },
    {
      "entropy": 5.291131925582886,
      "epoch": 3.4944642151047844,
      "grad_norm": 1.0,
      "learning_rate": 0.0003815467466925402,
      "loss": 4.8482,
      "mean_token_accuracy": 0.23483843952417374,
      "num_tokens": 80962556.0,
      "step": 35350
    },
    {
      "entropy": 5.405055856704712,
      "epoch": 3.4949584816132857,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038151566296351,
      "loss": 5.0183,
      "mean_token_accuracy": 0.21757636368274688,
      "num_tokens": 80974550.0,
      "step": 35355
    },
    {
      "entropy": 5.408387756347656,
      "epoch": 3.4954527481217874,
      "grad_norm": 1.03125,
      "learning_rate": 0.000381484576614178,
      "loss": 4.8875,
      "mean_token_accuracy": 0.22002579569816588,
      "num_tokens": 80987033.0,
      "step": 35360
    },
    {
      "entropy": 5.343898868560791,
      "epoch": 3.4959470146302887,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038145348764530887,
      "loss": 4.829,
      "mean_token_accuracy": 0.22586225271224974,
      "num_tokens": 80998216.0,
      "step": 35365
    },
    {
      "entropy": 5.349352693557739,
      "epoch": 3.49644128113879,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038142239605766737,
      "loss": 4.8914,
      "mean_token_accuracy": 0.22209430634975433,
      "num_tokens": 81010276.0,
      "step": 35370
    },
    {
      "entropy": 5.341150522232056,
      "epoch": 3.4969355476472916,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038139130185201845,
      "loss": 4.8634,
      "mean_token_accuracy": 0.22152438014745712,
      "num_tokens": 81020892.0,
      "step": 35375
    },
    {
      "entropy": 5.2937572479248045,
      "epoch": 3.497429814155793,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038136020502912705,
      "loss": 4.8109,
      "mean_token_accuracy": 0.2360413044691086,
      "num_tokens": 81032390.0,
      "step": 35380
    },
    {
      "entropy": 5.330139684677124,
      "epoch": 3.497924080664294,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003813291055897581,
      "loss": 4.8452,
      "mean_token_accuracy": 0.23186170011758805,
      "num_tokens": 81043433.0,
      "step": 35385
    },
    {
      "entropy": 5.357116556167602,
      "epoch": 3.4984183471727954,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003812980035346766,
      "loss": 4.868,
      "mean_token_accuracy": 0.22528961598873137,
      "num_tokens": 81054798.0,
      "step": 35390
    },
    {
      "entropy": 5.316542148590088,
      "epoch": 3.498912613681297,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003812668988646477,
      "loss": 4.943,
      "mean_token_accuracy": 0.2176274314522743,
      "num_tokens": 81067746.0,
      "step": 35395
    },
    {
      "entropy": 5.397860193252564,
      "epoch": 3.4994068801897984,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003812357915804366,
      "loss": 4.9329,
      "mean_token_accuracy": 0.21818071305751802,
      "num_tokens": 81079328.0,
      "step": 35400
    },
    {
      "entropy": 5.407144165039062,
      "epoch": 3.4999011466982997,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003812046816828086,
      "loss": 4.9109,
      "mean_token_accuracy": 0.22632373571395875,
      "num_tokens": 81091447.0,
      "step": 35405
    },
    {
      "entropy": 5.413412857055664,
      "epoch": 3.5003954132068014,
      "grad_norm": 1.03125,
      "learning_rate": 0.000381173569172529,
      "loss": 4.97,
      "mean_token_accuracy": 0.2185871735215187,
      "num_tokens": 81103614.0,
      "step": 35410
    },
    {
      "entropy": 5.46408429145813,
      "epoch": 3.5008896797153026,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000381142454050363,
      "loss": 5.0109,
      "mean_token_accuracy": 0.21528606712818146,
      "num_tokens": 81115387.0,
      "step": 35415
    },
    {
      "entropy": 5.40613317489624,
      "epoch": 3.501383946223804,
      "grad_norm": 1.0,
      "learning_rate": 0.00038111133631707624,
      "loss": 4.9272,
      "mean_token_accuracy": 0.22027081102132798,
      "num_tokens": 81128063.0,
      "step": 35420
    },
    {
      "entropy": 5.41538405418396,
      "epoch": 3.501878212732305,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003810802159734342,
      "loss": 4.9126,
      "mean_token_accuracy": 0.22631302773952483,
      "num_tokens": 81139137.0,
      "step": 35425
    },
    {
      "entropy": 5.389663887023926,
      "epoch": 3.5023724792408064,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003810490930202023,
      "loss": 5.0344,
      "mean_token_accuracy": 0.21445220559835435,
      "num_tokens": 81151896.0,
      "step": 35430
    },
    {
      "entropy": 5.463431072235108,
      "epoch": 3.502866745749308,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003810179674581463,
      "loss": 4.8794,
      "mean_token_accuracy": 0.22319562882184982,
      "num_tokens": 81162843.0,
      "step": 35435
    },
    {
      "entropy": 5.463798379898071,
      "epoch": 3.5033610122578094,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038098683928803186,
      "loss": 5.0475,
      "mean_token_accuracy": 0.2118205189704895,
      "num_tokens": 81176316.0,
      "step": 35440
    },
    {
      "entropy": 5.453320503234863,
      "epoch": 3.5038552787663106,
      "grad_norm": 0.953125,
      "learning_rate": 0.00038095570851062467,
      "loss": 4.9006,
      "mean_token_accuracy": 0.22157826870679856,
      "num_tokens": 81187956.0,
      "step": 35445
    },
    {
      "entropy": 5.373100614547729,
      "epoch": 3.5043495452748123,
      "grad_norm": 0.9375,
      "learning_rate": 0.00038092457512669067,
      "loss": 4.8912,
      "mean_token_accuracy": 0.22848647385835646,
      "num_tokens": 81199575.0,
      "step": 35450
    },
    {
      "entropy": 5.2962781429290775,
      "epoch": 3.5048438117833136,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003808934391369957,
      "loss": 4.8405,
      "mean_token_accuracy": 0.23707397729158403,
      "num_tokens": 81211959.0,
      "step": 35455
    },
    {
      "entropy": 5.307772207260132,
      "epoch": 3.505338078291815,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003808623005423057,
      "loss": 4.8313,
      "mean_token_accuracy": 0.23189762532711028,
      "num_tokens": 81222358.0,
      "step": 35460
    },
    {
      "entropy": 5.40881118774414,
      "epoch": 3.505832344800316,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038083115934338667,
      "loss": 4.8963,
      "mean_token_accuracy": 0.22693750411272048,
      "num_tokens": 81233674.0,
      "step": 35465
    },
    {
      "entropy": 5.343675661087036,
      "epoch": 3.506326611308818,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038080001554100467,
      "loss": 4.8585,
      "mean_token_accuracy": 0.23475908041000365,
      "num_tokens": 81243885.0,
      "step": 35470
    },
    {
      "entropy": 5.439158010482788,
      "epoch": 3.506820877817319,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003807688691359259,
      "loss": 5.0236,
      "mean_token_accuracy": 0.21338081508874893,
      "num_tokens": 81255869.0,
      "step": 35475
    },
    {
      "entropy": 5.372809553146363,
      "epoch": 3.5073151443258204,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003807377201289165,
      "loss": 4.8966,
      "mean_token_accuracy": 0.21953511536121367,
      "num_tokens": 81266746.0,
      "step": 35480
    },
    {
      "entropy": 5.360100221633911,
      "epoch": 3.507809410834322,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003807065685207429,
      "loss": 4.8979,
      "mean_token_accuracy": 0.2245924264192581,
      "num_tokens": 81277356.0,
      "step": 35485
    },
    {
      "entropy": 5.397751140594482,
      "epoch": 3.5083036773428233,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003806754143121712,
      "loss": 4.9546,
      "mean_token_accuracy": 0.22313296794891357,
      "num_tokens": 81288926.0,
      "step": 35490
    },
    {
      "entropy": 5.4592162609100345,
      "epoch": 3.5087979438513246,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038064425750396793,
      "loss": 4.9237,
      "mean_token_accuracy": 0.22758299857378006,
      "num_tokens": 81300990.0,
      "step": 35495
    },
    {
      "entropy": 5.2832671165466305,
      "epoch": 3.509292210359826,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038061309809689955,
      "loss": 4.8364,
      "mean_token_accuracy": 0.23108351677656175,
      "num_tokens": 81312964.0,
      "step": 35500
    },
    {
      "entropy": 5.361547565460205,
      "epoch": 3.5097864768683276,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038058193609173256,
      "loss": 4.9203,
      "mean_token_accuracy": 0.22482313215732574,
      "num_tokens": 81324196.0,
      "step": 35505
    },
    {
      "entropy": 5.388671970367431,
      "epoch": 3.510280743376829,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003805507714892336,
      "loss": 4.8993,
      "mean_token_accuracy": 0.22257969975471498,
      "num_tokens": 81334723.0,
      "step": 35510
    },
    {
      "entropy": 5.372091436386109,
      "epoch": 3.51077500988533,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038051960429016923,
      "loss": 4.9409,
      "mean_token_accuracy": 0.21985040307044984,
      "num_tokens": 81346126.0,
      "step": 35515
    },
    {
      "entropy": 5.3360363960266115,
      "epoch": 3.511269276393832,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003804884344953063,
      "loss": 4.882,
      "mean_token_accuracy": 0.2258705511689186,
      "num_tokens": 81358093.0,
      "step": 35520
    },
    {
      "entropy": 5.40121898651123,
      "epoch": 3.511763542902333,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038045726210541144,
      "loss": 4.9516,
      "mean_token_accuracy": 0.22141797840595245,
      "num_tokens": 81369734.0,
      "step": 35525
    },
    {
      "entropy": 5.349990653991699,
      "epoch": 3.5122578094108343,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003804260871212516,
      "loss": 4.9087,
      "mean_token_accuracy": 0.22574088275432586,
      "num_tokens": 81381474.0,
      "step": 35530
    },
    {
      "entropy": 5.300096559524536,
      "epoch": 3.5127520759193356,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003803949095435937,
      "loss": 4.8499,
      "mean_token_accuracy": 0.22507714480161667,
      "num_tokens": 81394661.0,
      "step": 35535
    },
    {
      "entropy": 5.445417785644532,
      "epoch": 3.513246342427837,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003803637293732046,
      "loss": 4.9966,
      "mean_token_accuracy": 0.21807651668787004,
      "num_tokens": 81405591.0,
      "step": 35540
    },
    {
      "entropy": 5.40941162109375,
      "epoch": 3.5137406089363385,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00038033254661085146,
      "loss": 5.0131,
      "mean_token_accuracy": 0.21807652860879898,
      "num_tokens": 81417561.0,
      "step": 35545
    },
    {
      "entropy": 5.443903207778931,
      "epoch": 3.51423487544484,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038030136125730135,
      "loss": 4.9726,
      "mean_token_accuracy": 0.21926791667938234,
      "num_tokens": 81430502.0,
      "step": 35550
    },
    {
      "entropy": 5.3760340213775635,
      "epoch": 3.5147291419533415,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003802701733133214,
      "loss": 4.9141,
      "mean_token_accuracy": 0.22555883079767228,
      "num_tokens": 81441717.0,
      "step": 35555
    },
    {
      "entropy": 5.320801019668579,
      "epoch": 3.515223408461843,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038023898277967885,
      "loss": 4.8553,
      "mean_token_accuracy": 0.22992610931396484,
      "num_tokens": 81453282.0,
      "step": 35560
    },
    {
      "entropy": 5.3407166481018065,
      "epoch": 3.515717674970344,
      "grad_norm": 1.046875,
      "learning_rate": 0.000380207789657141,
      "loss": 4.8864,
      "mean_token_accuracy": 0.22186102569103242,
      "num_tokens": 81463582.0,
      "step": 35565
    },
    {
      "entropy": 5.326526165008545,
      "epoch": 3.5162119414788453,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003801765939464752,
      "loss": 4.8804,
      "mean_token_accuracy": 0.22806383967399596,
      "num_tokens": 81475490.0,
      "step": 35570
    },
    {
      "entropy": 5.3576818943023685,
      "epoch": 3.5167062079873466,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003801453956484489,
      "loss": 4.8435,
      "mean_token_accuracy": 0.23497522324323655,
      "num_tokens": 81486939.0,
      "step": 35575
    },
    {
      "entropy": 5.3534018993377686,
      "epoch": 3.5172004744958483,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003801141947638295,
      "loss": 4.9016,
      "mean_token_accuracy": 0.22537907510995864,
      "num_tokens": 81498260.0,
      "step": 35580
    },
    {
      "entropy": 5.379961109161377,
      "epoch": 3.5176947410043495,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038008299129338463,
      "loss": 4.8389,
      "mean_token_accuracy": 0.233942674100399,
      "num_tokens": 81508788.0,
      "step": 35585
    },
    {
      "entropy": 5.398923110961914,
      "epoch": 3.518189007512851,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038005178523788184,
      "loss": 4.9484,
      "mean_token_accuracy": 0.21832203567028047,
      "num_tokens": 81520500.0,
      "step": 35590
    },
    {
      "entropy": 5.293336868286133,
      "epoch": 3.5186832740213525,
      "grad_norm": 1.0,
      "learning_rate": 0.00038002057659808884,
      "loss": 4.8977,
      "mean_token_accuracy": 0.22243375033140184,
      "num_tokens": 81531617.0,
      "step": 35595
    },
    {
      "entropy": 5.40934009552002,
      "epoch": 3.5191775405298538,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003799893653747735,
      "loss": 4.8827,
      "mean_token_accuracy": 0.22816220670938492,
      "num_tokens": 81542814.0,
      "step": 35600
    },
    {
      "entropy": 5.32665867805481,
      "epoch": 3.519671807038355,
      "grad_norm": 1.125,
      "learning_rate": 0.00037995815156870334,
      "loss": 4.8074,
      "mean_token_accuracy": 0.235041606426239,
      "num_tokens": 81552551.0,
      "step": 35605
    },
    {
      "entropy": 5.388699626922607,
      "epoch": 3.5201660735468563,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037992693518064635,
      "loss": 5.0332,
      "mean_token_accuracy": 0.21169514656066896,
      "num_tokens": 81564576.0,
      "step": 35610
    },
    {
      "entropy": 5.36256594657898,
      "epoch": 3.520660340055358,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00037989571621137053,
      "loss": 4.8818,
      "mean_token_accuracy": 0.2299272358417511,
      "num_tokens": 81575639.0,
      "step": 35615
    },
    {
      "entropy": 5.418667793273926,
      "epoch": 3.5211546065638593,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037986449466164383,
      "loss": 4.9189,
      "mean_token_accuracy": 0.21851727068424226,
      "num_tokens": 81587012.0,
      "step": 35620
    },
    {
      "entropy": 5.369828701019287,
      "epoch": 3.5216488730723605,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003798332705322343,
      "loss": 4.8991,
      "mean_token_accuracy": 0.22809659391641618,
      "num_tokens": 81598194.0,
      "step": 35625
    },
    {
      "entropy": 5.359159469604492,
      "epoch": 3.5221431395808622,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003798020438239101,
      "loss": 4.9222,
      "mean_token_accuracy": 0.2231315165758133,
      "num_tokens": 81608702.0,
      "step": 35630
    },
    {
      "entropy": 5.3343226432800295,
      "epoch": 3.5226374060893635,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003797708145374392,
      "loss": 4.8925,
      "mean_token_accuracy": 0.2302893027663231,
      "num_tokens": 81619598.0,
      "step": 35635
    },
    {
      "entropy": 5.380286693572998,
      "epoch": 3.5231316725978647,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003797395826735901,
      "loss": 4.9128,
      "mean_token_accuracy": 0.21611912548542023,
      "num_tokens": 81631928.0,
      "step": 35640
    },
    {
      "entropy": 5.494489526748657,
      "epoch": 3.523625939106366,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00037970834823313106,
      "loss": 5.0482,
      "mean_token_accuracy": 0.2140463784337044,
      "num_tokens": 81644553.0,
      "step": 35645
    },
    {
      "entropy": 5.460727262496948,
      "epoch": 3.5241202056148673,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037967711121683037,
      "loss": 4.9965,
      "mean_token_accuracy": 0.2166570708155632,
      "num_tokens": 81656633.0,
      "step": 35650
    },
    {
      "entropy": 5.378783178329468,
      "epoch": 3.524614472123369,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003796458716254565,
      "loss": 4.8998,
      "mean_token_accuracy": 0.22698196917772293,
      "num_tokens": 81669287.0,
      "step": 35655
    },
    {
      "entropy": 5.303140354156494,
      "epoch": 3.5251087386318702,
      "grad_norm": 1.109375,
      "learning_rate": 0.00037961462945977794,
      "loss": 4.794,
      "mean_token_accuracy": 0.23135911375284196,
      "num_tokens": 81680298.0,
      "step": 35660
    },
    {
      "entropy": 5.375480222702026,
      "epoch": 3.525603005140372,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003795833847205633,
      "loss": 4.9641,
      "mean_token_accuracy": 0.2174853965640068,
      "num_tokens": 81691088.0,
      "step": 35665
    },
    {
      "entropy": 5.398610162734985,
      "epoch": 3.526097271648873,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003795521374085811,
      "loss": 4.914,
      "mean_token_accuracy": 0.22223890721797943,
      "num_tokens": 81701791.0,
      "step": 35670
    },
    {
      "entropy": 5.476521110534668,
      "epoch": 3.5265915381573745,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037952088752460025,
      "loss": 5.0159,
      "mean_token_accuracy": 0.216519333422184,
      "num_tokens": 81713682.0,
      "step": 35675
    },
    {
      "entropy": 5.3094910144805905,
      "epoch": 3.5270858046658757,
      "grad_norm": 1.125,
      "learning_rate": 0.0003794896350693893,
      "loss": 4.7579,
      "mean_token_accuracy": 0.2445613369345665,
      "num_tokens": 81724857.0,
      "step": 35680
    },
    {
      "entropy": 5.386565065383911,
      "epoch": 3.527580071174377,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000379458380043717,
      "loss": 4.9098,
      "mean_token_accuracy": 0.2245212808251381,
      "num_tokens": 81737074.0,
      "step": 35685
    },
    {
      "entropy": 5.452234077453613,
      "epoch": 3.5280743376828787,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003794271224483524,
      "loss": 4.9988,
      "mean_token_accuracy": 0.21030569076538086,
      "num_tokens": 81749679.0,
      "step": 35690
    },
    {
      "entropy": 5.368377447128296,
      "epoch": 3.52856860419138,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037939586228406445,
      "loss": 4.906,
      "mean_token_accuracy": 0.22342533320188523,
      "num_tokens": 81761165.0,
      "step": 35695
    },
    {
      "entropy": 5.329142141342163,
      "epoch": 3.5290628706998812,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00037936459955162204,
      "loss": 4.7725,
      "mean_token_accuracy": 0.2392613783478737,
      "num_tokens": 81773289.0,
      "step": 35700
    },
    {
      "entropy": 5.367126893997193,
      "epoch": 3.529557137208383,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003793333342517943,
      "loss": 4.856,
      "mean_token_accuracy": 0.22544315010309218,
      "num_tokens": 81785021.0,
      "step": 35705
    },
    {
      "entropy": 5.316310596466065,
      "epoch": 3.530051403716884,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003793020663853503,
      "loss": 4.8602,
      "mean_token_accuracy": 0.22383590638637543,
      "num_tokens": 81796111.0,
      "step": 35710
    },
    {
      "entropy": 5.356758117675781,
      "epoch": 3.5305456702253855,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003792707959530594,
      "loss": 4.8452,
      "mean_token_accuracy": 0.22855856418609619,
      "num_tokens": 81806927.0,
      "step": 35715
    },
    {
      "entropy": 5.410068702697754,
      "epoch": 3.5310399367338867,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003792395229556907,
      "loss": 5.0262,
      "mean_token_accuracy": 0.21530334055423736,
      "num_tokens": 81819942.0,
      "step": 35720
    },
    {
      "entropy": 5.3535853862762455,
      "epoch": 3.5315342032423884,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003792082473940135,
      "loss": 4.9107,
      "mean_token_accuracy": 0.22415890097618102,
      "num_tokens": 81831192.0,
      "step": 35725
    },
    {
      "entropy": 5.4085243225097654,
      "epoch": 3.5320284697508897,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003791769692687973,
      "loss": 4.9287,
      "mean_token_accuracy": 0.2243881031870842,
      "num_tokens": 81844684.0,
      "step": 35730
    },
    {
      "entropy": 5.400975704193115,
      "epoch": 3.532522736259391,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00037914568858081146,
      "loss": 4.9489,
      "mean_token_accuracy": 0.22354546338319778,
      "num_tokens": 81855379.0,
      "step": 35735
    },
    {
      "entropy": 5.321661520004272,
      "epoch": 3.5330170027678927,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003791144053308256,
      "loss": 4.7546,
      "mean_token_accuracy": 0.23168833255767823,
      "num_tokens": 81867560.0,
      "step": 35740
    },
    {
      "entropy": 5.321735525131226,
      "epoch": 3.533511269276394,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00037908311951960916,
      "loss": 4.9066,
      "mean_token_accuracy": 0.22142521440982818,
      "num_tokens": 81879298.0,
      "step": 35745
    },
    {
      "entropy": 5.35098786354065,
      "epoch": 3.534005535784895,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003790518311479318,
      "loss": 4.8961,
      "mean_token_accuracy": 0.22518471479415894,
      "num_tokens": 81891323.0,
      "step": 35750
    },
    {
      "entropy": 5.337911939620971,
      "epoch": 3.5344998022933964,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003790205402165633,
      "loss": 4.8164,
      "mean_token_accuracy": 0.23768679797649384,
      "num_tokens": 81901324.0,
      "step": 35755
    },
    {
      "entropy": 5.389211368560791,
      "epoch": 3.534994068801898,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037898924672627346,
      "loss": 4.9456,
      "mean_token_accuracy": 0.22096111625432968,
      "num_tokens": 81911740.0,
      "step": 35760
    },
    {
      "entropy": 5.3115562915802,
      "epoch": 3.5354883353103994,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037895795067783184,
      "loss": 4.8163,
      "mean_token_accuracy": 0.22595912963151932,
      "num_tokens": 81922685.0,
      "step": 35765
    },
    {
      "entropy": 5.326155948638916,
      "epoch": 3.5359826018189007,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003789266520720086,
      "loss": 4.9066,
      "mean_token_accuracy": 0.22974238842725753,
      "num_tokens": 81935822.0,
      "step": 35770
    },
    {
      "entropy": 5.327796888351441,
      "epoch": 3.5364768683274024,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003788953509095736,
      "loss": 4.8639,
      "mean_token_accuracy": 0.2318188801407814,
      "num_tokens": 81948420.0,
      "step": 35775
    },
    {
      "entropy": 5.400712013244629,
      "epoch": 3.5369711348359036,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003788640471912968,
      "loss": 4.9138,
      "mean_token_accuracy": 0.22621310353279114,
      "num_tokens": 81959740.0,
      "step": 35780
    },
    {
      "entropy": 5.333722400665283,
      "epoch": 3.537465401344405,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037883274091794837,
      "loss": 4.858,
      "mean_token_accuracy": 0.22978856712579726,
      "num_tokens": 81971334.0,
      "step": 35785
    },
    {
      "entropy": 5.357579040527344,
      "epoch": 3.537959667852906,
      "grad_norm": 1.109375,
      "learning_rate": 0.00037880143209029844,
      "loss": 4.8838,
      "mean_token_accuracy": 0.22318846881389617,
      "num_tokens": 81981740.0,
      "step": 35790
    },
    {
      "entropy": 5.363036060333252,
      "epoch": 3.5384539343614074,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003787701207091171,
      "loss": 4.9381,
      "mean_token_accuracy": 0.22171318680047988,
      "num_tokens": 81992987.0,
      "step": 35795
    },
    {
      "entropy": 5.379034042358398,
      "epoch": 3.538948200869909,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00037873880677517476,
      "loss": 4.9048,
      "mean_token_accuracy": 0.223313008248806,
      "num_tokens": 82003590.0,
      "step": 35800
    },
    {
      "entropy": 5.457319927215576,
      "epoch": 3.5394424673784104,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003787074902892416,
      "loss": 4.9589,
      "mean_token_accuracy": 0.21598458886146546,
      "num_tokens": 82014894.0,
      "step": 35805
    },
    {
      "entropy": 5.328961420059204,
      "epoch": 3.5399367338869117,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037867617125208813,
      "loss": 4.8392,
      "mean_token_accuracy": 0.22856746315956117,
      "num_tokens": 82025214.0,
      "step": 35810
    },
    {
      "entropy": 5.312014532089234,
      "epoch": 3.5404310003954134,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003786448496644848,
      "loss": 4.8642,
      "mean_token_accuracy": 0.22681592255830765,
      "num_tokens": 82038866.0,
      "step": 35815
    },
    {
      "entropy": 5.390064668655396,
      "epoch": 3.5409252669039146,
      "grad_norm": 1.0625,
      "learning_rate": 0.000378613525527202,
      "loss": 4.9874,
      "mean_token_accuracy": 0.21908834725618362,
      "num_tokens": 82051114.0,
      "step": 35820
    },
    {
      "entropy": 5.320617294311523,
      "epoch": 3.541419533412416,
      "grad_norm": 0.953125,
      "learning_rate": 0.00037858219884101047,
      "loss": 4.9006,
      "mean_token_accuracy": 0.22503003031015395,
      "num_tokens": 82063091.0,
      "step": 35825
    },
    {
      "entropy": 5.371501827239991,
      "epoch": 3.541913799920917,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003785508696066808,
      "loss": 4.8504,
      "mean_token_accuracy": 0.22753026932477952,
      "num_tokens": 82073363.0,
      "step": 35830
    },
    {
      "entropy": 5.409093809127808,
      "epoch": 3.542408066429419,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037851953782498367,
      "loss": 4.9432,
      "mean_token_accuracy": 0.21445583254098893,
      "num_tokens": 82083477.0,
      "step": 35835
    },
    {
      "entropy": 5.363201379776001,
      "epoch": 3.54290233293792,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003784882034966898,
      "loss": 4.8267,
      "mean_token_accuracy": 0.22567698806524278,
      "num_tokens": 82094738.0,
      "step": 35840
    },
    {
      "entropy": 5.362297487258911,
      "epoch": 3.5433965994464214,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037845686662257,
      "loss": 4.8927,
      "mean_token_accuracy": 0.2260923519730568,
      "num_tokens": 82105769.0,
      "step": 35845
    },
    {
      "entropy": 5.377329921722412,
      "epoch": 3.543890865954923,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003784255272033954,
      "loss": 4.937,
      "mean_token_accuracy": 0.22094734907150268,
      "num_tokens": 82117644.0,
      "step": 35850
    },
    {
      "entropy": 5.348103094100952,
      "epoch": 3.5443851324634243,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00037839418523993667,
      "loss": 4.8496,
      "mean_token_accuracy": 0.2295177161693573,
      "num_tokens": 82129716.0,
      "step": 35855
    },
    {
      "entropy": 5.265044832229615,
      "epoch": 3.5448793989719256,
      "grad_norm": 1.109375,
      "learning_rate": 0.000378362840732965,
      "loss": 4.7803,
      "mean_token_accuracy": 0.23901551216840744,
      "num_tokens": 82142378.0,
      "step": 35860
    },
    {
      "entropy": 5.284296607971191,
      "epoch": 3.545373665480427,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037833149368325143,
      "loss": 4.8001,
      "mean_token_accuracy": 0.23130851984024048,
      "num_tokens": 82152893.0,
      "step": 35865
    },
    {
      "entropy": 5.359723138809204,
      "epoch": 3.5458679319889286,
      "grad_norm": 0.96875,
      "learning_rate": 0.000378300144091567,
      "loss": 4.952,
      "mean_token_accuracy": 0.22596070170402527,
      "num_tokens": 82165813.0,
      "step": 35870
    },
    {
      "entropy": 5.33999719619751,
      "epoch": 3.54636219849743,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000378268791958683,
      "loss": 4.8975,
      "mean_token_accuracy": 0.21875406950712203,
      "num_tokens": 82177641.0,
      "step": 35875
    },
    {
      "entropy": 5.422315835952759,
      "epoch": 3.546856465005931,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003782374372853708,
      "loss": 4.8711,
      "mean_token_accuracy": 0.22439011484384536,
      "num_tokens": 82187594.0,
      "step": 35880
    },
    {
      "entropy": 5.341062450408936,
      "epoch": 3.547350731514433,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003782060800724016,
      "loss": 4.8912,
      "mean_token_accuracy": 0.22224678993225097,
      "num_tokens": 82199177.0,
      "step": 35885
    },
    {
      "entropy": 5.371145296096802,
      "epoch": 3.547844998022934,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003781747203205468,
      "loss": 4.9005,
      "mean_token_accuracy": 0.2173135071992874,
      "num_tokens": 82211176.0,
      "step": 35890
    },
    {
      "entropy": 5.325938320159912,
      "epoch": 3.5483392645314353,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00037814335803057773,
      "loss": 4.8029,
      "mean_token_accuracy": 0.22973656207323073,
      "num_tokens": 82222767.0,
      "step": 35895
    },
    {
      "entropy": 5.379412508010864,
      "epoch": 3.5488335310399366,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003781119932032662,
      "loss": 4.9242,
      "mean_token_accuracy": 0.22057341784238815,
      "num_tokens": 82234060.0,
      "step": 35900
    },
    {
      "entropy": 5.390061950683593,
      "epoch": 3.549327797548438,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037808062583938367,
      "loss": 4.9069,
      "mean_token_accuracy": 0.22166765630245208,
      "num_tokens": 82244910.0,
      "step": 35905
    },
    {
      "entropy": 5.405637407302857,
      "epoch": 3.5498220640569396,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003780492559397017,
      "loss": 4.9862,
      "mean_token_accuracy": 0.21612415313720704,
      "num_tokens": 82257730.0,
      "step": 35910
    },
    {
      "entropy": 5.339308452606201,
      "epoch": 3.550316330565441,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003780178835049921,
      "loss": 4.8383,
      "mean_token_accuracy": 0.2383595108985901,
      "num_tokens": 82268688.0,
      "step": 35915
    },
    {
      "entropy": 5.343853616714478,
      "epoch": 3.5508105970739425,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003779865085360265,
      "loss": 4.912,
      "mean_token_accuracy": 0.22326348721981049,
      "num_tokens": 82279562.0,
      "step": 35920
    },
    {
      "entropy": 5.384601879119873,
      "epoch": 3.551304863582444,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003779551310335768,
      "loss": 4.8489,
      "mean_token_accuracy": 0.22973953038454056,
      "num_tokens": 82289945.0,
      "step": 35925
    },
    {
      "entropy": 5.414995908737183,
      "epoch": 3.551799130090945,
      "grad_norm": 1.0,
      "learning_rate": 0.00037792375099841505,
      "loss": 4.9355,
      "mean_token_accuracy": 0.2279105380177498,
      "num_tokens": 82301878.0,
      "step": 35930
    },
    {
      "entropy": 5.372233486175537,
      "epoch": 3.5522933965994463,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037789236843131297,
      "loss": 4.9363,
      "mean_token_accuracy": 0.2245788171887398,
      "num_tokens": 82313565.0,
      "step": 35935
    },
    {
      "entropy": 5.382424640655517,
      "epoch": 3.5527876631079476,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003778609833330426,
      "loss": 4.9338,
      "mean_token_accuracy": 0.2176203414797783,
      "num_tokens": 82324375.0,
      "step": 35940
    },
    {
      "entropy": 5.455204963684082,
      "epoch": 3.5532819296164493,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003778295957043761,
      "loss": 5.0362,
      "mean_token_accuracy": 0.21008071750402452,
      "num_tokens": 82335642.0,
      "step": 35945
    },
    {
      "entropy": 5.345562076568603,
      "epoch": 3.5537761961249505,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003777982055460857,
      "loss": 4.7872,
      "mean_token_accuracy": 0.24031329303979873,
      "num_tokens": 82345847.0,
      "step": 35950
    },
    {
      "entropy": 5.465438938140869,
      "epoch": 3.554270462633452,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003777668128589434,
      "loss": 5.0075,
      "mean_token_accuracy": 0.21809862107038497,
      "num_tokens": 82358249.0,
      "step": 35955
    },
    {
      "entropy": 5.354618692398072,
      "epoch": 3.5547647291419535,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003777354176437216,
      "loss": 4.8856,
      "mean_token_accuracy": 0.22534047067165375,
      "num_tokens": 82369996.0,
      "step": 35960
    },
    {
      "entropy": 5.402975130081177,
      "epoch": 3.555258995650455,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003777040199011926,
      "loss": 4.8414,
      "mean_token_accuracy": 0.23164844810962676,
      "num_tokens": 82381569.0,
      "step": 35965
    },
    {
      "entropy": 5.312043905258179,
      "epoch": 3.555753262158956,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037767261963212866,
      "loss": 4.9236,
      "mean_token_accuracy": 0.21461534053087233,
      "num_tokens": 82392343.0,
      "step": 35970
    },
    {
      "entropy": 5.306873369216919,
      "epoch": 3.5562475286674573,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037764121683730247,
      "loss": 4.9222,
      "mean_token_accuracy": 0.2222127765417099,
      "num_tokens": 82403681.0,
      "step": 35975
    },
    {
      "entropy": 5.396149063110352,
      "epoch": 3.556741795175959,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00037760981151748645,
      "loss": 4.9545,
      "mean_token_accuracy": 0.21818052232265472,
      "num_tokens": 82415773.0,
      "step": 35980
    },
    {
      "entropy": 5.353687286376953,
      "epoch": 3.5572360616844603,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003775784036734531,
      "loss": 4.7992,
      "mean_token_accuracy": 0.2325186848640442,
      "num_tokens": 82426636.0,
      "step": 35985
    },
    {
      "entropy": 5.484178924560547,
      "epoch": 3.5577303281929615,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003775469933059751,
      "loss": 4.9894,
      "mean_token_accuracy": 0.21849699169397355,
      "num_tokens": 82438546.0,
      "step": 35990
    },
    {
      "entropy": 5.3433209419250485,
      "epoch": 3.5582245947014632,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037751558041582514,
      "loss": 4.9139,
      "mean_token_accuracy": 0.21960765570402146,
      "num_tokens": 82451911.0,
      "step": 35995
    },
    {
      "entropy": 5.370611572265625,
      "epoch": 3.5587188612099645,
      "grad_norm": 1.078125,
      "learning_rate": 0.000377484165003776,
      "loss": 4.8445,
      "mean_token_accuracy": 0.22619989663362502,
      "num_tokens": 82464691.0,
      "step": 36000
    },
    {
      "epoch": 3.5587188612099645,
      "eval_entropy": 5.1553237261074205,
      "eval_loss": 4.991928577423096,
      "eval_mean_token_accuracy": 0.22754630471727505,
      "eval_num_tokens": 82464691.0,
      "eval_runtime": 26.4644,
      "eval_samples_per_second": 1228.557,
      "eval_steps_per_second": 153.603,
      "step": 36000
    },
    {
      "entropy": 5.3582721710205075,
      "epoch": 3.5592131277184658,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037745274707060055,
      "loss": 4.9246,
      "mean_token_accuracy": 0.22260909229516984,
      "num_tokens": 82476498.0,
      "step": 36005
    },
    {
      "entropy": 5.296879386901855,
      "epoch": 3.559707394226967,
      "grad_norm": 0.9375,
      "learning_rate": 0.00037742132661707157,
      "loss": 4.8461,
      "mean_token_accuracy": 0.23117797076702118,
      "num_tokens": 82488434.0,
      "step": 36010
    },
    {
      "entropy": 5.335042428970337,
      "epoch": 3.5602016607354683,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00037738990364396207,
      "loss": 4.8197,
      "mean_token_accuracy": 0.2264535829424858,
      "num_tokens": 82500277.0,
      "step": 36015
    },
    {
      "entropy": 5.418668365478515,
      "epoch": 3.56069592724397,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037735847815204503,
      "loss": 4.955,
      "mean_token_accuracy": 0.22316774725914001,
      "num_tokens": 82513107.0,
      "step": 36020
    },
    {
      "entropy": 5.401876926422119,
      "epoch": 3.5611901937524713,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037732705014209366,
      "loss": 4.9222,
      "mean_token_accuracy": 0.22162585854530334,
      "num_tokens": 82524967.0,
      "step": 36025
    },
    {
      "entropy": 5.340512657165528,
      "epoch": 3.561684460260973,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003772956196148809,
      "loss": 4.8799,
      "mean_token_accuracy": 0.22124444097280502,
      "num_tokens": 82536721.0,
      "step": 36030
    },
    {
      "entropy": 5.429541969299317,
      "epoch": 3.5621787267694742,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00037726418657118004,
      "loss": 4.9385,
      "mean_token_accuracy": 0.2233225643634796,
      "num_tokens": 82549386.0,
      "step": 36035
    },
    {
      "entropy": 5.3624509334564205,
      "epoch": 3.5626729932779755,
      "grad_norm": 0.96875,
      "learning_rate": 0.00037723275101176436,
      "loss": 4.8076,
      "mean_token_accuracy": 0.23480848520994185,
      "num_tokens": 82560207.0,
      "step": 36040
    },
    {
      "entropy": 5.3558636665344235,
      "epoch": 3.5631672597864767,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003772013129374071,
      "loss": 4.9349,
      "mean_token_accuracy": 0.22056962847709655,
      "num_tokens": 82570484.0,
      "step": 36045
    },
    {
      "entropy": 5.347987461090088,
      "epoch": 3.563661526294978,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003771698723488817,
      "loss": 4.8538,
      "mean_token_accuracy": 0.221049901843071,
      "num_tokens": 82581833.0,
      "step": 36050
    },
    {
      "entropy": 5.28921594619751,
      "epoch": 3.5641557928034797,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037713842924696155,
      "loss": 4.8016,
      "mean_token_accuracy": 0.23402377665042878,
      "num_tokens": 82593351.0,
      "step": 36055
    },
    {
      "entropy": 5.395655393600464,
      "epoch": 3.564650059311981,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003771069836324203,
      "loss": 4.937,
      "mean_token_accuracy": 0.21493322253227234,
      "num_tokens": 82606644.0,
      "step": 36060
    },
    {
      "entropy": 5.326463794708252,
      "epoch": 3.5651443258204822,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003770755355060313,
      "loss": 4.933,
      "mean_token_accuracy": 0.22492736279964448,
      "num_tokens": 82618779.0,
      "step": 36065
    },
    {
      "entropy": 5.345552015304565,
      "epoch": 3.565638592328984,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003770440848685683,
      "loss": 4.8574,
      "mean_token_accuracy": 0.225765398144722,
      "num_tokens": 82629968.0,
      "step": 36070
    },
    {
      "entropy": 5.34907169342041,
      "epoch": 3.566132858837485,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00037701263172080514,
      "loss": 4.8672,
      "mean_token_accuracy": 0.23132342100143433,
      "num_tokens": 82641586.0,
      "step": 36075
    },
    {
      "entropy": 5.412793874740601,
      "epoch": 3.5666271253459865,
      "grad_norm": 1.0,
      "learning_rate": 0.00037698117606351525,
      "loss": 4.9796,
      "mean_token_accuracy": 0.21982377171516418,
      "num_tokens": 82651901.0,
      "step": 36080
    },
    {
      "entropy": 5.282834672927857,
      "epoch": 3.5671213918544877,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003769497178974726,
      "loss": 4.8374,
      "mean_token_accuracy": 0.2314256712794304,
      "num_tokens": 82663055.0,
      "step": 36085
    },
    {
      "entropy": 5.280278921127319,
      "epoch": 3.5676156583629894,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003769182572234512,
      "loss": 4.7923,
      "mean_token_accuracy": 0.23667036592960358,
      "num_tokens": 82674709.0,
      "step": 36090
    },
    {
      "entropy": 5.365299940109253,
      "epoch": 3.5681099248714907,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003768867940422248,
      "loss": 4.7973,
      "mean_token_accuracy": 0.2387310281395912,
      "num_tokens": 82685215.0,
      "step": 36095
    },
    {
      "entropy": 5.409591388702393,
      "epoch": 3.568604191379992,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003768553283545674,
      "loss": 4.9585,
      "mean_token_accuracy": 0.22036982625722884,
      "num_tokens": 82697473.0,
      "step": 36100
    },
    {
      "entropy": 5.278342580795288,
      "epoch": 3.5690984578884937,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003768238601612532,
      "loss": 4.8395,
      "mean_token_accuracy": 0.22933488339185715,
      "num_tokens": 82708033.0,
      "step": 36105
    },
    {
      "entropy": 5.3069933414459225,
      "epoch": 3.569592724396995,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003767923894630562,
      "loss": 4.8386,
      "mean_token_accuracy": 0.22580899745225907,
      "num_tokens": 82719410.0,
      "step": 36110
    },
    {
      "entropy": 5.388894987106323,
      "epoch": 3.570086990905496,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037676091626075056,
      "loss": 4.794,
      "mean_token_accuracy": 0.22915762960910796,
      "num_tokens": 82730648.0,
      "step": 36115
    },
    {
      "entropy": 5.3518335819244385,
      "epoch": 3.5705812574139975,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003767294405551107,
      "loss": 4.9016,
      "mean_token_accuracy": 0.22254659831523896,
      "num_tokens": 82742041.0,
      "step": 36120
    },
    {
      "entropy": 5.387453174591064,
      "epoch": 3.571075523922499,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003766979623469108,
      "loss": 4.929,
      "mean_token_accuracy": 0.2192847266793251,
      "num_tokens": 82754083.0,
      "step": 36125
    },
    {
      "entropy": 5.430908346176148,
      "epoch": 3.5715697904310004,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003766664816369252,
      "loss": 4.9768,
      "mean_token_accuracy": 0.21282288879156114,
      "num_tokens": 82766735.0,
      "step": 36130
    },
    {
      "entropy": 5.434899997711182,
      "epoch": 3.5720640569395017,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037663499842592845,
      "loss": 5.0163,
      "mean_token_accuracy": 0.21411536782979965,
      "num_tokens": 82778498.0,
      "step": 36135
    },
    {
      "entropy": 5.475568151473999,
      "epoch": 3.5725583234480034,
      "grad_norm": 1.109375,
      "learning_rate": 0.00037660351271469496,
      "loss": 5.001,
      "mean_token_accuracy": 0.2114973321557045,
      "num_tokens": 82789927.0,
      "step": 36140
    },
    {
      "entropy": 5.40434308052063,
      "epoch": 3.5730525899565047,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003765720245039993,
      "loss": 4.8875,
      "mean_token_accuracy": 0.22115093916654588,
      "num_tokens": 82800846.0,
      "step": 36145
    },
    {
      "entropy": 5.323550844192505,
      "epoch": 3.573546856465006,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037654053379461606,
      "loss": 4.9014,
      "mean_token_accuracy": 0.22038141191005706,
      "num_tokens": 82812912.0,
      "step": 36150
    },
    {
      "entropy": 5.386518096923828,
      "epoch": 3.574041122973507,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003765090405873199,
      "loss": 4.9479,
      "mean_token_accuracy": 0.2225004956126213,
      "num_tokens": 82824186.0,
      "step": 36155
    },
    {
      "entropy": 5.349151372909546,
      "epoch": 3.5745353894820084,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003764775448828856,
      "loss": 4.9261,
      "mean_token_accuracy": 0.22722647190093995,
      "num_tokens": 82835599.0,
      "step": 36160
    },
    {
      "entropy": 5.323567628860474,
      "epoch": 3.57502965599051,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037644604668208805,
      "loss": 4.7879,
      "mean_token_accuracy": 0.2287730097770691,
      "num_tokens": 82846272.0,
      "step": 36165
    },
    {
      "entropy": 5.3890900135040285,
      "epoch": 3.5755239224990114,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003764145459857019,
      "loss": 4.9139,
      "mean_token_accuracy": 0.21889124661684037,
      "num_tokens": 82858245.0,
      "step": 36170
    },
    {
      "entropy": 5.403174495697021,
      "epoch": 3.576018189007513,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037638304279450225,
      "loss": 4.925,
      "mean_token_accuracy": 0.2238876312971115,
      "num_tokens": 82869951.0,
      "step": 36175
    },
    {
      "entropy": 5.3100879192352295,
      "epoch": 3.5765124555160144,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037635153710926394,
      "loss": 4.8249,
      "mean_token_accuracy": 0.23399822413921356,
      "num_tokens": 82880489.0,
      "step": 36180
    },
    {
      "entropy": 5.3540952682495115,
      "epoch": 3.5770067220245156,
      "grad_norm": 0.984375,
      "learning_rate": 0.00037632002893076217,
      "loss": 4.9593,
      "mean_token_accuracy": 0.2208590552210808,
      "num_tokens": 82892188.0,
      "step": 36185
    },
    {
      "entropy": 5.350983333587647,
      "epoch": 3.577500988533017,
      "grad_norm": 1.109375,
      "learning_rate": 0.000376288518259772,
      "loss": 4.8617,
      "mean_token_accuracy": 0.22965113520622255,
      "num_tokens": 82903504.0,
      "step": 36190
    },
    {
      "entropy": 5.319399118423462,
      "epoch": 3.577995255041518,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003762570050970685,
      "loss": 4.8039,
      "mean_token_accuracy": 0.22818130552768706,
      "num_tokens": 82914396.0,
      "step": 36195
    },
    {
      "entropy": 5.2737305641174315,
      "epoch": 3.57848952155002,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000376225489443427,
      "loss": 4.7998,
      "mean_token_accuracy": 0.23316608518362045,
      "num_tokens": 82925759.0,
      "step": 36200
    },
    {
      "entropy": 5.330904245376587,
      "epoch": 3.578983788058521,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003761939712996227,
      "loss": 4.8618,
      "mean_token_accuracy": 0.22488100677728654,
      "num_tokens": 82937474.0,
      "step": 36205
    },
    {
      "entropy": 5.423814201354981,
      "epoch": 3.5794780545670224,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003761624506664311,
      "loss": 4.9433,
      "mean_token_accuracy": 0.21881335377693176,
      "num_tokens": 82948686.0,
      "step": 36210
    },
    {
      "entropy": 5.390163564682007,
      "epoch": 3.579972321075524,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003761309275446274,
      "loss": 4.9292,
      "mean_token_accuracy": 0.22747351825237275,
      "num_tokens": 82960099.0,
      "step": 36215
    },
    {
      "entropy": 5.353944635391235,
      "epoch": 3.5804665875840254,
      "grad_norm": 1.0,
      "learning_rate": 0.0003760994019349873,
      "loss": 4.8491,
      "mean_token_accuracy": 0.2233394980430603,
      "num_tokens": 82971415.0,
      "step": 36220
    },
    {
      "entropy": 5.375739526748657,
      "epoch": 3.5809608540925266,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003760678738382862,
      "loss": 4.9412,
      "mean_token_accuracy": 0.21731544584035872,
      "num_tokens": 82983581.0,
      "step": 36225
    },
    {
      "entropy": 5.393738698959351,
      "epoch": 3.581455120601028,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003760363432552997,
      "loss": 4.9001,
      "mean_token_accuracy": 0.2250516891479492,
      "num_tokens": 82994817.0,
      "step": 36230
    },
    {
      "entropy": 5.303756761550903,
      "epoch": 3.5819493871095296,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003760048101868035,
      "loss": 4.7979,
      "mean_token_accuracy": 0.23693961799144744,
      "num_tokens": 83004620.0,
      "step": 36235
    },
    {
      "entropy": 5.337575769424438,
      "epoch": 3.582443653618031,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00037597327463357336,
      "loss": 4.9117,
      "mean_token_accuracy": 0.2175038769841194,
      "num_tokens": 83015257.0,
      "step": 36240
    },
    {
      "entropy": 5.258402585983276,
      "epoch": 3.582937920126532,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003759417365963849,
      "loss": 4.7494,
      "mean_token_accuracy": 0.23803631961345673,
      "num_tokens": 83026765.0,
      "step": 36245
    },
    {
      "entropy": 5.373844003677368,
      "epoch": 3.583432186635034,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00037591019607601407,
      "loss": 4.9184,
      "mean_token_accuracy": 0.22234239876270295,
      "num_tokens": 83038893.0,
      "step": 36250
    },
    {
      "entropy": 5.2905220031738285,
      "epoch": 3.583926453143535,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00037587865307323683,
      "loss": 4.8669,
      "mean_token_accuracy": 0.23194096982479095,
      "num_tokens": 83050615.0,
      "step": 36255
    },
    {
      "entropy": 5.418975830078125,
      "epoch": 3.5844207196520363,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00037584710758882897,
      "loss": 4.9402,
      "mean_token_accuracy": 0.2200712338089943,
      "num_tokens": 83062459.0,
      "step": 36260
    },
    {
      "entropy": 5.445836067199707,
      "epoch": 3.5849149861605376,
      "grad_norm": 0.90625,
      "learning_rate": 0.00037581555962356666,
      "loss": 4.9978,
      "mean_token_accuracy": 0.21745892465114594,
      "num_tokens": 83073563.0,
      "step": 36265
    },
    {
      "entropy": 5.415084457397461,
      "epoch": 3.585409252669039,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037578400917822604,
      "loss": 4.9377,
      "mean_token_accuracy": 0.21675791144371032,
      "num_tokens": 83084831.0,
      "step": 36270
    },
    {
      "entropy": 5.36310715675354,
      "epoch": 3.5859035191775406,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000375752456253583,
      "loss": 4.8521,
      "mean_token_accuracy": 0.2282091796398163,
      "num_tokens": 83096504.0,
      "step": 36275
    },
    {
      "entropy": 5.327261257171631,
      "epoch": 3.586397785686042,
      "grad_norm": 1.0625,
      "learning_rate": 0.000375720900850414,
      "loss": 4.9297,
      "mean_token_accuracy": 0.22356237322092057,
      "num_tokens": 83108312.0,
      "step": 36280
    },
    {
      "entropy": 5.310328531265259,
      "epoch": 3.5868920521945435,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003756893429694952,
      "loss": 4.8304,
      "mean_token_accuracy": 0.2325567916035652,
      "num_tokens": 83119040.0,
      "step": 36285
    },
    {
      "entropy": 5.377159643173218,
      "epoch": 3.587386318703045,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037565778261160296,
      "loss": 4.7752,
      "mean_token_accuracy": 0.23957739770412445,
      "num_tokens": 83129803.0,
      "step": 36290
    },
    {
      "entropy": 5.303463315963745,
      "epoch": 3.587880585211546,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003756262197775137,
      "loss": 4.8519,
      "mean_token_accuracy": 0.2304352268576622,
      "num_tokens": 83142364.0,
      "step": 36295
    },
    {
      "entropy": 5.27993803024292,
      "epoch": 3.5883748517200473,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00037559465446800364,
      "loss": 4.8809,
      "mean_token_accuracy": 0.2229409098625183,
      "num_tokens": 83152813.0,
      "step": 36300
    },
    {
      "entropy": 5.465881490707398,
      "epoch": 3.5888691182285486,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037556308668384964,
      "loss": 5.0218,
      "mean_token_accuracy": 0.21707602590322495,
      "num_tokens": 83164052.0,
      "step": 36305
    },
    {
      "entropy": 5.398655748367309,
      "epoch": 3.5893633847370503,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00037553151642582807,
      "loss": 4.8686,
      "mean_token_accuracy": 0.22257529497146605,
      "num_tokens": 83176236.0,
      "step": 36310
    },
    {
      "entropy": 5.330732011795044,
      "epoch": 3.5898576512455516,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003754999436947156,
      "loss": 4.9124,
      "mean_token_accuracy": 0.22084331810474395,
      "num_tokens": 83187173.0,
      "step": 36315
    },
    {
      "entropy": 5.37593674659729,
      "epoch": 3.590351917754053,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003754683684912889,
      "loss": 4.8957,
      "mean_token_accuracy": 0.22767562866210939,
      "num_tokens": 83199628.0,
      "step": 36320
    },
    {
      "entropy": 5.449776220321655,
      "epoch": 3.5908461842625545,
      "grad_norm": 1.109375,
      "learning_rate": 0.00037543679081632474,
      "loss": 4.9336,
      "mean_token_accuracy": 0.22522548586130142,
      "num_tokens": 83211748.0,
      "step": 36325
    },
    {
      "entropy": 5.283000326156616,
      "epoch": 3.591340450771056,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003754052106706,
      "loss": 4.812,
      "mean_token_accuracy": 0.23188331723213196,
      "num_tokens": 83223078.0,
      "step": 36330
    },
    {
      "entropy": 5.363790273666382,
      "epoch": 3.591834717279557,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037537362805489143,
      "loss": 4.8745,
      "mean_token_accuracy": 0.2306692823767662,
      "num_tokens": 83234046.0,
      "step": 36335
    },
    {
      "entropy": 5.378996276855469,
      "epoch": 3.5923289837880583,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003753420429699761,
      "loss": 4.9269,
      "mean_token_accuracy": 0.21831649392843247,
      "num_tokens": 83244121.0,
      "step": 36340
    },
    {
      "entropy": 5.413927316665649,
      "epoch": 3.59282325029656,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003753104554166309,
      "loss": 4.8769,
      "mean_token_accuracy": 0.22316502779722214,
      "num_tokens": 83254182.0,
      "step": 36345
    },
    {
      "entropy": 5.356520700454712,
      "epoch": 3.5933175168050613,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003752788653956329,
      "loss": 4.8531,
      "mean_token_accuracy": 0.22837089151144027,
      "num_tokens": 83264527.0,
      "step": 36350
    },
    {
      "entropy": 5.395730304718017,
      "epoch": 3.5938117833135625,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003752472729077593,
      "loss": 4.8969,
      "mean_token_accuracy": 0.2169907882809639,
      "num_tokens": 83275186.0,
      "step": 36355
    },
    {
      "entropy": 5.451711654663086,
      "epoch": 3.5943060498220643,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003752156779537873,
      "loss": 5.0201,
      "mean_token_accuracy": 0.21423997282981871,
      "num_tokens": 83286919.0,
      "step": 36360
    },
    {
      "entropy": 5.390353536605835,
      "epoch": 3.5948003163305655,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037518408053449404,
      "loss": 4.8688,
      "mean_token_accuracy": 0.22217227518558502,
      "num_tokens": 83298112.0,
      "step": 36365
    },
    {
      "entropy": 5.332046318054199,
      "epoch": 3.5952945828390668,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003751524806506568,
      "loss": 4.9454,
      "mean_token_accuracy": 0.22163941115140914,
      "num_tokens": 83308978.0,
      "step": 36370
    },
    {
      "entropy": 5.397374010086059,
      "epoch": 3.595788849347568,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037512087830305306,
      "loss": 4.9842,
      "mean_token_accuracy": 0.2149580627679825,
      "num_tokens": 83320780.0,
      "step": 36375
    },
    {
      "entropy": 5.4503217220306395,
      "epoch": 3.5962831158560697,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003750892734924601,
      "loss": 4.9769,
      "mean_token_accuracy": 0.21490403711795808,
      "num_tokens": 83332070.0,
      "step": 36380
    },
    {
      "entropy": 5.403571033477784,
      "epoch": 3.596777382364571,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003750576662196556,
      "loss": 4.9235,
      "mean_token_accuracy": 0.22100280821323395,
      "num_tokens": 83345506.0,
      "step": 36385
    },
    {
      "entropy": 5.452732753753662,
      "epoch": 3.5972716488730723,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003750260564854169,
      "loss": 5.0079,
      "mean_token_accuracy": 0.21736557334661483,
      "num_tokens": 83357097.0,
      "step": 36390
    },
    {
      "entropy": 5.322962236404419,
      "epoch": 3.597765915381574,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003749944442905217,
      "loss": 4.7998,
      "mean_token_accuracy": 0.23713620454072953,
      "num_tokens": 83369088.0,
      "step": 36395
    },
    {
      "entropy": 5.382613277435302,
      "epoch": 3.5982601818900752,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003749628296357477,
      "loss": 4.9192,
      "mean_token_accuracy": 0.22792432010173796,
      "num_tokens": 83381249.0,
      "step": 36400
    },
    {
      "entropy": 5.333266019821167,
      "epoch": 3.5987544483985765,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037493121252187256,
      "loss": 4.8635,
      "mean_token_accuracy": 0.22568015605211258,
      "num_tokens": 83392000.0,
      "step": 36405
    },
    {
      "entropy": 5.324226665496826,
      "epoch": 3.5992487149070778,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003748995929496741,
      "loss": 4.7883,
      "mean_token_accuracy": 0.23469123542308806,
      "num_tokens": 83403841.0,
      "step": 36410
    },
    {
      "entropy": 5.385702753067017,
      "epoch": 3.599742981415579,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037486797091993015,
      "loss": 4.9064,
      "mean_token_accuracy": 0.22474417537450792,
      "num_tokens": 83415114.0,
      "step": 36415
    },
    {
      "entropy": 5.375254535675049,
      "epoch": 3.6002372479240807,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037483634643341866,
      "loss": 4.9531,
      "mean_token_accuracy": 0.21599696278572084,
      "num_tokens": 83428158.0,
      "step": 36420
    },
    {
      "entropy": 5.44150562286377,
      "epoch": 3.600731514432582,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003748047194909175,
      "loss": 4.9325,
      "mean_token_accuracy": 0.22725801169872284,
      "num_tokens": 83439535.0,
      "step": 36425
    },
    {
      "entropy": 5.354701089859009,
      "epoch": 3.6012257809410837,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037477309009320477,
      "loss": 4.9223,
      "mean_token_accuracy": 0.22424223124980927,
      "num_tokens": 83451215.0,
      "step": 36430
    },
    {
      "entropy": 5.402597904205322,
      "epoch": 3.601720047449585,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037474145824105854,
      "loss": 4.8879,
      "mean_token_accuracy": 0.2199832335114479,
      "num_tokens": 83461750.0,
      "step": 36435
    },
    {
      "entropy": 5.35029706954956,
      "epoch": 3.6022143139580862,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00037470982393525697,
      "loss": 4.8818,
      "mean_token_accuracy": 0.22946826964616776,
      "num_tokens": 83473790.0,
      "step": 36440
    },
    {
      "entropy": 5.342877960205078,
      "epoch": 3.6027085804665875,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037467818717657825,
      "loss": 4.8597,
      "mean_token_accuracy": 0.2287721484899521,
      "num_tokens": 83485470.0,
      "step": 36445
    },
    {
      "entropy": 5.3918922424316404,
      "epoch": 3.6032028469750887,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003746465479658007,
      "loss": 4.8911,
      "mean_token_accuracy": 0.22231501638889312,
      "num_tokens": 83496007.0,
      "step": 36450
    },
    {
      "entropy": 5.414427089691162,
      "epoch": 3.6036971134835905,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003746149063037026,
      "loss": 4.9833,
      "mean_token_accuracy": 0.21907393485307694,
      "num_tokens": 83508620.0,
      "step": 36455
    },
    {
      "entropy": 5.333643579483033,
      "epoch": 3.6041913799920917,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003745832621910623,
      "loss": 4.8658,
      "mean_token_accuracy": 0.22993077486753463,
      "num_tokens": 83520323.0,
      "step": 36460
    },
    {
      "entropy": 5.262320470809937,
      "epoch": 3.604685646500593,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037455161562865833,
      "loss": 4.8119,
      "mean_token_accuracy": 0.23296705186367034,
      "num_tokens": 83532471.0,
      "step": 36465
    },
    {
      "entropy": 5.337418937683106,
      "epoch": 3.6051799130090947,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003745199666172692,
      "loss": 4.8417,
      "mean_token_accuracy": 0.23423704653978347,
      "num_tokens": 83544156.0,
      "step": 36470
    },
    {
      "entropy": 5.418204212188721,
      "epoch": 3.605674179517596,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00037448831515767353,
      "loss": 4.8841,
      "mean_token_accuracy": 0.22467339783906937,
      "num_tokens": 83556321.0,
      "step": 36475
    },
    {
      "entropy": 5.338177728652954,
      "epoch": 3.606168446026097,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003744566612506497,
      "loss": 4.8737,
      "mean_token_accuracy": 0.22403129786252976,
      "num_tokens": 83566604.0,
      "step": 36480
    },
    {
      "entropy": 5.361013317108155,
      "epoch": 3.6066627125345985,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037442500489697666,
      "loss": 4.869,
      "mean_token_accuracy": 0.22904723435640334,
      "num_tokens": 83578349.0,
      "step": 36485
    },
    {
      "entropy": 5.37205286026001,
      "epoch": 3.6071569790431,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003743933460974331,
      "loss": 4.8877,
      "mean_token_accuracy": 0.22562311142683028,
      "num_tokens": 83589198.0,
      "step": 36490
    },
    {
      "entropy": 5.448660945892334,
      "epoch": 3.6076512455516014,
      "grad_norm": 0.96875,
      "learning_rate": 0.00037436168485279777,
      "loss": 4.9502,
      "mean_token_accuracy": 0.22046404480934143,
      "num_tokens": 83602076.0,
      "step": 36495
    },
    {
      "entropy": 5.377399253845215,
      "epoch": 3.6081455120601027,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037433002116384957,
      "loss": 4.861,
      "mean_token_accuracy": 0.22547683566808702,
      "num_tokens": 83613340.0,
      "step": 36500
    },
    {
      "entropy": 5.307501268386841,
      "epoch": 3.6086397785686044,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037429835503136744,
      "loss": 4.9111,
      "mean_token_accuracy": 0.2243644818663597,
      "num_tokens": 83624715.0,
      "step": 36505
    },
    {
      "entropy": 5.309962129592895,
      "epoch": 3.6091340450771057,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003742666864561304,
      "loss": 4.9161,
      "mean_token_accuracy": 0.2332532748579979,
      "num_tokens": 83635833.0,
      "step": 36510
    },
    {
      "entropy": 5.381388807296753,
      "epoch": 3.609628311585607,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003742350154389174,
      "loss": 4.8492,
      "mean_token_accuracy": 0.22707187831401826,
      "num_tokens": 83647281.0,
      "step": 36515
    },
    {
      "entropy": 5.393423843383789,
      "epoch": 3.610122578094108,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037420334198050775,
      "loss": 4.9302,
      "mean_token_accuracy": 0.223509019613266,
      "num_tokens": 83658745.0,
      "step": 36520
    },
    {
      "entropy": 5.389162158966064,
      "epoch": 3.6106168446026095,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003741716660816804,
      "loss": 4.9029,
      "mean_token_accuracy": 0.2229568988084793,
      "num_tokens": 83671580.0,
      "step": 36525
    },
    {
      "entropy": 5.389196157455444,
      "epoch": 3.611111111111111,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037413998774321475,
      "loss": 4.9107,
      "mean_token_accuracy": 0.2231159493327141,
      "num_tokens": 83682667.0,
      "step": 36530
    },
    {
      "entropy": 5.339846420288086,
      "epoch": 3.6116053776196124,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037410830696588994,
      "loss": 4.8356,
      "mean_token_accuracy": 0.2271808370947838,
      "num_tokens": 83693276.0,
      "step": 36535
    },
    {
      "entropy": 5.342306280136109,
      "epoch": 3.612099644128114,
      "grad_norm": 0.96875,
      "learning_rate": 0.00037407662375048544,
      "loss": 4.9308,
      "mean_token_accuracy": 0.22439689189195633,
      "num_tokens": 83705304.0,
      "step": 36540
    },
    {
      "entropy": 5.331595849990845,
      "epoch": 3.6125939106366154,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037404493809778063,
      "loss": 4.8989,
      "mean_token_accuracy": 0.2246134966611862,
      "num_tokens": 83717166.0,
      "step": 36545
    },
    {
      "entropy": 5.366345977783203,
      "epoch": 3.6130881771451167,
      "grad_norm": 1.109375,
      "learning_rate": 0.00037401325000855495,
      "loss": 4.8401,
      "mean_token_accuracy": 0.2327850878238678,
      "num_tokens": 83727322.0,
      "step": 36550
    },
    {
      "entropy": 5.36159315109253,
      "epoch": 3.613582443653618,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037398155948358805,
      "loss": 4.8438,
      "mean_token_accuracy": 0.22833312749862672,
      "num_tokens": 83738721.0,
      "step": 36555
    },
    {
      "entropy": 5.291883754730224,
      "epoch": 3.614076710162119,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003739498665236593,
      "loss": 4.8753,
      "mean_token_accuracy": 0.23352148979902268,
      "num_tokens": 83750622.0,
      "step": 36560
    },
    {
      "entropy": 5.382266044616699,
      "epoch": 3.614570976670621,
      "grad_norm": 1.0,
      "learning_rate": 0.0003739181711295485,
      "loss": 4.9561,
      "mean_token_accuracy": 0.2155703902244568,
      "num_tokens": 83761755.0,
      "step": 36565
    },
    {
      "entropy": 5.321734189987183,
      "epoch": 3.615065243179122,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037388647330203545,
      "loss": 4.8055,
      "mean_token_accuracy": 0.23482843935489656,
      "num_tokens": 83772335.0,
      "step": 36570
    },
    {
      "entropy": 5.325783920288086,
      "epoch": 3.6155595096876234,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003738547730418999,
      "loss": 4.926,
      "mean_token_accuracy": 0.22703629583120347,
      "num_tokens": 83784949.0,
      "step": 36575
    },
    {
      "entropy": 5.347792482376098,
      "epoch": 3.616053776196125,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003738230703499213,
      "loss": 4.8806,
      "mean_token_accuracy": 0.22740814387798308,
      "num_tokens": 83795555.0,
      "step": 36580
    },
    {
      "entropy": 5.327791404724121,
      "epoch": 3.6165480427046264,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037379136522688,
      "loss": 4.8393,
      "mean_token_accuracy": 0.22769590616226196,
      "num_tokens": 83806682.0,
      "step": 36585
    },
    {
      "entropy": 5.348587989807129,
      "epoch": 3.6170423092131276,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037375965767355584,
      "loss": 4.8652,
      "mean_token_accuracy": 0.2256679654121399,
      "num_tokens": 83818485.0,
      "step": 36590
    },
    {
      "entropy": 5.261403560638428,
      "epoch": 3.617536575721629,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003737279476907287,
      "loss": 4.802,
      "mean_token_accuracy": 0.23808386772871018,
      "num_tokens": 83829757.0,
      "step": 36595
    },
    {
      "entropy": 5.36194953918457,
      "epoch": 3.6180308422301306,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00037369623527917876,
      "loss": 4.915,
      "mean_token_accuracy": 0.22493782490491868,
      "num_tokens": 83841827.0,
      "step": 36600
    },
    {
      "entropy": 5.276743221282959,
      "epoch": 3.618525108738632,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003736645204396861,
      "loss": 4.731,
      "mean_token_accuracy": 0.2410330817103386,
      "num_tokens": 83852471.0,
      "step": 36605
    },
    {
      "entropy": 5.320824193954468,
      "epoch": 3.619019375247133,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003736328031730309,
      "loss": 4.8325,
      "mean_token_accuracy": 0.2330484449863434,
      "num_tokens": 83864322.0,
      "step": 36610
    },
    {
      "entropy": 5.374589300155639,
      "epoch": 3.619513641755635,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003736010834799936,
      "loss": 4.9269,
      "mean_token_accuracy": 0.2262087881565094,
      "num_tokens": 83875286.0,
      "step": 36615
    },
    {
      "entropy": 5.320151424407959,
      "epoch": 3.620007908264136,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003735693613613542,
      "loss": 4.7911,
      "mean_token_accuracy": 0.2374019667506218,
      "num_tokens": 83887794.0,
      "step": 36620
    },
    {
      "entropy": 5.371251058578491,
      "epoch": 3.6205021747726374,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003735376368178932,
      "loss": 4.896,
      "mean_token_accuracy": 0.22413357645273208,
      "num_tokens": 83898644.0,
      "step": 36625
    },
    {
      "entropy": 5.380478096008301,
      "epoch": 3.6209964412811386,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003735059098503911,
      "loss": 4.9225,
      "mean_token_accuracy": 0.21952511221170426,
      "num_tokens": 83909944.0,
      "step": 36630
    },
    {
      "entropy": 5.360162448883057,
      "epoch": 3.6214907077896403,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003734741804596284,
      "loss": 4.8936,
      "mean_token_accuracy": 0.2248098596930504,
      "num_tokens": 83920928.0,
      "step": 36635
    },
    {
      "entropy": 5.341647434234619,
      "epoch": 3.6219849742981416,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003734424486463855,
      "loss": 4.8068,
      "mean_token_accuracy": 0.23733924627304076,
      "num_tokens": 83933519.0,
      "step": 36640
    },
    {
      "entropy": 5.420714855194092,
      "epoch": 3.622479240806643,
      "grad_norm": 0.953125,
      "learning_rate": 0.00037341071441144316,
      "loss": 4.9682,
      "mean_token_accuracy": 0.2225632354617119,
      "num_tokens": 83946933.0,
      "step": 36645
    },
    {
      "entropy": 5.3279835224151615,
      "epoch": 3.6229735073151446,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003733789777555819,
      "loss": 4.8632,
      "mean_token_accuracy": 0.2297317162156105,
      "num_tokens": 83957858.0,
      "step": 36650
    },
    {
      "entropy": 5.326677227020264,
      "epoch": 3.623467773823646,
      "grad_norm": 1.0,
      "learning_rate": 0.00037334723867958256,
      "loss": 4.8429,
      "mean_token_accuracy": 0.23162795901298522,
      "num_tokens": 83968363.0,
      "step": 36655
    },
    {
      "entropy": 5.349777364730835,
      "epoch": 3.623962040332147,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003733154971842259,
      "loss": 4.8453,
      "mean_token_accuracy": 0.23122699558734894,
      "num_tokens": 83978719.0,
      "step": 36660
    },
    {
      "entropy": 5.301551580429077,
      "epoch": 3.6244563068406483,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037328375327029274,
      "loss": 4.8348,
      "mean_token_accuracy": 0.22420499622821807,
      "num_tokens": 83990195.0,
      "step": 36665
    },
    {
      "entropy": 5.3063499450683596,
      "epoch": 3.6249505733491496,
      "grad_norm": 0.984375,
      "learning_rate": 0.00037325200693856406,
      "loss": 4.9345,
      "mean_token_accuracy": 0.2208761379122734,
      "num_tokens": 84002068.0,
      "step": 36670
    },
    {
      "entropy": 5.357392263412476,
      "epoch": 3.6254448398576513,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003732202581898206,
      "loss": 4.8667,
      "mean_token_accuracy": 0.22409599423408508,
      "num_tokens": 84013405.0,
      "step": 36675
    },
    {
      "entropy": 5.4420802116394045,
      "epoch": 3.6259391063661526,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003731885070248437,
      "loss": 4.97,
      "mean_token_accuracy": 0.21321644484996796,
      "num_tokens": 84025650.0,
      "step": 36680
    },
    {
      "entropy": 5.265451574325562,
      "epoch": 3.6264333728746543,
      "grad_norm": 0.9375,
      "learning_rate": 0.00037315675344441426,
      "loss": 4.7806,
      "mean_token_accuracy": 0.24111033380031585,
      "num_tokens": 84036753.0,
      "step": 36685
    },
    {
      "entropy": 5.458128547668457,
      "epoch": 3.6269276393831555,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003731249974493135,
      "loss": 4.9771,
      "mean_token_accuracy": 0.22219886332750322,
      "num_tokens": 84049082.0,
      "step": 36690
    },
    {
      "entropy": 5.388784551620484,
      "epoch": 3.627421905891657,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037309323904032254,
      "loss": 4.8837,
      "mean_token_accuracy": 0.2212057039141655,
      "num_tokens": 84059471.0,
      "step": 36695
    },
    {
      "entropy": 5.322232389450074,
      "epoch": 3.627916172400158,
      "grad_norm": 1.0,
      "learning_rate": 0.0003730614782182226,
      "loss": 4.8996,
      "mean_token_accuracy": 0.2337486118078232,
      "num_tokens": 84071475.0,
      "step": 36700
    },
    {
      "entropy": 5.359746217727661,
      "epoch": 3.6284104389086593,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003730297149837952,
      "loss": 4.9073,
      "mean_token_accuracy": 0.22454703897237777,
      "num_tokens": 84082465.0,
      "step": 36705
    },
    {
      "entropy": 5.337502193450928,
      "epoch": 3.628904705417161,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003729979493378215,
      "loss": 4.8432,
      "mean_token_accuracy": 0.2250938132405281,
      "num_tokens": 84093845.0,
      "step": 36710
    },
    {
      "entropy": 5.386278963088989,
      "epoch": 3.6293989719256623,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037296618128108316,
      "loss": 4.8957,
      "mean_token_accuracy": 0.22552331984043122,
      "num_tokens": 84105865.0,
      "step": 36715
    },
    {
      "entropy": 5.362513780593872,
      "epoch": 3.6298932384341636,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00037293441081436154,
      "loss": 4.8928,
      "mean_token_accuracy": 0.2250848591327667,
      "num_tokens": 84116376.0,
      "step": 36720
    },
    {
      "entropy": 5.331195068359375,
      "epoch": 3.6303875049426653,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003729026379384381,
      "loss": 4.7998,
      "mean_token_accuracy": 0.2327029064297676,
      "num_tokens": 84126806.0,
      "step": 36725
    },
    {
      "entropy": 5.382214879989624,
      "epoch": 3.6308817714511665,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003728708626540948,
      "loss": 4.8927,
      "mean_token_accuracy": 0.23157044649124145,
      "num_tokens": 84136750.0,
      "step": 36730
    },
    {
      "entropy": 5.398047351837159,
      "epoch": 3.631376037959668,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037283908496211295,
      "loss": 4.9792,
      "mean_token_accuracy": 0.21441410183906556,
      "num_tokens": 84148614.0,
      "step": 36735
    },
    {
      "entropy": 5.4482917308807375,
      "epoch": 3.631870304468169,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003728073048632744,
      "loss": 4.9704,
      "mean_token_accuracy": 0.21500234156847,
      "num_tokens": 84160347.0,
      "step": 36740
    },
    {
      "entropy": 5.388708925247192,
      "epoch": 3.6323645709766708,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037277552235836104,
      "loss": 4.9327,
      "mean_token_accuracy": 0.22186036556959152,
      "num_tokens": 84172240.0,
      "step": 36745
    },
    {
      "entropy": 5.401154708862305,
      "epoch": 3.632858837485172,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037274373744815466,
      "loss": 4.9244,
      "mean_token_accuracy": 0.22853055745363235,
      "num_tokens": 84184050.0,
      "step": 36750
    },
    {
      "entropy": 5.370966386795044,
      "epoch": 3.6333531039936733,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003727119501334371,
      "loss": 4.8796,
      "mean_token_accuracy": 0.22564692795276642,
      "num_tokens": 84194661.0,
      "step": 36755
    },
    {
      "entropy": 5.353967189788818,
      "epoch": 3.633847370502175,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037268016041499055,
      "loss": 4.8641,
      "mean_token_accuracy": 0.22966067492961884,
      "num_tokens": 84205677.0,
      "step": 36760
    },
    {
      "entropy": 5.370859670639038,
      "epoch": 3.6343416370106763,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003726483682935968,
      "loss": 4.8734,
      "mean_token_accuracy": 0.22522821724414827,
      "num_tokens": 84217342.0,
      "step": 36765
    },
    {
      "entropy": 5.373274326324463,
      "epoch": 3.6348359035191775,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037261657377003803,
      "loss": 4.902,
      "mean_token_accuracy": 0.22258375883102416,
      "num_tokens": 84229637.0,
      "step": 36770
    },
    {
      "entropy": 5.422550916671753,
      "epoch": 3.6353301700276788,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003725847768450964,
      "loss": 4.9378,
      "mean_token_accuracy": 0.21780333817005157,
      "num_tokens": 84241156.0,
      "step": 36775
    },
    {
      "entropy": 5.438250494003296,
      "epoch": 3.63582443653618,
      "grad_norm": 1.0,
      "learning_rate": 0.0003725529775195542,
      "loss": 4.9215,
      "mean_token_accuracy": 0.2223355457186699,
      "num_tokens": 84253172.0,
      "step": 36780
    },
    {
      "entropy": 5.323154497146606,
      "epoch": 3.6363187030446817,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037252117579419355,
      "loss": 4.8387,
      "mean_token_accuracy": 0.22752950191497803,
      "num_tokens": 84265349.0,
      "step": 36785
    },
    {
      "entropy": 5.2843529224395756,
      "epoch": 3.636812969553183,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037248937166979687,
      "loss": 4.8131,
      "mean_token_accuracy": 0.23492369502782823,
      "num_tokens": 84275931.0,
      "step": 36790
    },
    {
      "entropy": 5.355429697036743,
      "epoch": 3.6373072360616847,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003724575651471465,
      "loss": 4.8951,
      "mean_token_accuracy": 0.22533716410398483,
      "num_tokens": 84286502.0,
      "step": 36795
    },
    {
      "entropy": 5.393918180465699,
      "epoch": 3.637801502570186,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003724257562270249,
      "loss": 4.9556,
      "mean_token_accuracy": 0.21121064126491546,
      "num_tokens": 84297748.0,
      "step": 36800
    },
    {
      "entropy": 5.392804431915283,
      "epoch": 3.6382957690786872,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037239394491021463,
      "loss": 4.9367,
      "mean_token_accuracy": 0.22384822517633438,
      "num_tokens": 84309733.0,
      "step": 36805
    },
    {
      "entropy": 5.434682703018188,
      "epoch": 3.6387900355871885,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037236213119749823,
      "loss": 4.9565,
      "mean_token_accuracy": 0.21578225642442703,
      "num_tokens": 84320606.0,
      "step": 36810
    },
    {
      "entropy": 5.332182598114014,
      "epoch": 3.6392843020956898,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037233031508965823,
      "loss": 4.8586,
      "mean_token_accuracy": 0.22566750049591064,
      "num_tokens": 84331149.0,
      "step": 36815
    },
    {
      "entropy": 5.397786808013916,
      "epoch": 3.6397785686041915,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003722984965874774,
      "loss": 4.9138,
      "mean_token_accuracy": 0.2253658413887024,
      "num_tokens": 84343016.0,
      "step": 36820
    },
    {
      "entropy": 5.397837495803833,
      "epoch": 3.6402728351126927,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003722666756917385,
      "loss": 4.9305,
      "mean_token_accuracy": 0.22284920364618302,
      "num_tokens": 84354846.0,
      "step": 36825
    },
    {
      "entropy": 5.338949537277221,
      "epoch": 3.640767101621194,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003722348524032242,
      "loss": 4.8771,
      "mean_token_accuracy": 0.22435445189476014,
      "num_tokens": 84366698.0,
      "step": 36830
    },
    {
      "entropy": 5.3655980110168455,
      "epoch": 3.6412613681296957,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00037220302672271756,
      "loss": 4.9012,
      "mean_token_accuracy": 0.2269018679857254,
      "num_tokens": 84378186.0,
      "step": 36835
    },
    {
      "entropy": 5.440581846237182,
      "epoch": 3.641755634638197,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037217119865100127,
      "loss": 4.9859,
      "mean_token_accuracy": 0.21867185682058335,
      "num_tokens": 84390301.0,
      "step": 36840
    },
    {
      "entropy": 5.331853151321411,
      "epoch": 3.642249901146698,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037213936818885853,
      "loss": 4.8997,
      "mean_token_accuracy": 0.2267444059252739,
      "num_tokens": 84401134.0,
      "step": 36845
    },
    {
      "entropy": 5.350849914550781,
      "epoch": 3.6427441676551995,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037210753533707214,
      "loss": 4.8607,
      "mean_token_accuracy": 0.2309836983680725,
      "num_tokens": 84411734.0,
      "step": 36850
    },
    {
      "entropy": 5.403667211532593,
      "epoch": 3.643238434163701,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037207570009642536,
      "loss": 4.9637,
      "mean_token_accuracy": 0.21293123960494995,
      "num_tokens": 84422952.0,
      "step": 36855
    },
    {
      "entropy": 5.384668350219727,
      "epoch": 3.6437327006722025,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003720438624677014,
      "loss": 4.8181,
      "mean_token_accuracy": 0.23322306871414183,
      "num_tokens": 84433048.0,
      "step": 36860
    },
    {
      "entropy": 5.369571256637573,
      "epoch": 3.6442269671807037,
      "grad_norm": 1.0,
      "learning_rate": 0.0003720120224516832,
      "loss": 4.8903,
      "mean_token_accuracy": 0.21995196491479874,
      "num_tokens": 84443318.0,
      "step": 36865
    },
    {
      "entropy": 5.372448539733886,
      "epoch": 3.6447212336892054,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003719801800491543,
      "loss": 4.8605,
      "mean_token_accuracy": 0.22919123619794846,
      "num_tokens": 84453441.0,
      "step": 36870
    },
    {
      "entropy": 5.312135410308838,
      "epoch": 3.6452155001977067,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037194833526089797,
      "loss": 4.8895,
      "mean_token_accuracy": 0.22450451403856278,
      "num_tokens": 84465290.0,
      "step": 36875
    },
    {
      "entropy": 5.399795293807983,
      "epoch": 3.645709766706208,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003719164880876975,
      "loss": 4.9191,
      "mean_token_accuracy": 0.22377658635377884,
      "num_tokens": 84476482.0,
      "step": 36880
    },
    {
      "entropy": 5.335363960266113,
      "epoch": 3.646204033214709,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003718846385303363,
      "loss": 4.8183,
      "mean_token_accuracy": 0.23134245425462724,
      "num_tokens": 84488641.0,
      "step": 36885
    },
    {
      "entropy": 5.345751237869263,
      "epoch": 3.646698299723211,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003718527865895981,
      "loss": 4.9313,
      "mean_token_accuracy": 0.22859867960214614,
      "num_tokens": 84501967.0,
      "step": 36890
    },
    {
      "entropy": 5.404613494873047,
      "epoch": 3.647192566231712,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003718209322662662,
      "loss": 4.8878,
      "mean_token_accuracy": 0.21708727926015853,
      "num_tokens": 84513982.0,
      "step": 36895
    },
    {
      "entropy": 5.3517796993255615,
      "epoch": 3.6476868327402134,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003717890755611244,
      "loss": 4.8622,
      "mean_token_accuracy": 0.2245517775416374,
      "num_tokens": 84523395.0,
      "step": 36900
    },
    {
      "entropy": 5.295890140533447,
      "epoch": 3.648181099248715,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037175721647495635,
      "loss": 4.8601,
      "mean_token_accuracy": 0.23574294298887252,
      "num_tokens": 84536019.0,
      "step": 36905
    },
    {
      "entropy": 5.337387752532959,
      "epoch": 3.6486753657572164,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003717253550085458,
      "loss": 4.8831,
      "mean_token_accuracy": 0.22999414801597595,
      "num_tokens": 84546882.0,
      "step": 36910
    },
    {
      "entropy": 5.391915798187256,
      "epoch": 3.6491696322657177,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037169349116267643,
      "loss": 4.9074,
      "mean_token_accuracy": 0.22693344801664353,
      "num_tokens": 84560195.0,
      "step": 36915
    },
    {
      "entropy": 5.449013423919678,
      "epoch": 3.649663898774219,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00037166162493813224,
      "loss": 4.9914,
      "mean_token_accuracy": 0.22236735373735428,
      "num_tokens": 84573165.0,
      "step": 36920
    },
    {
      "entropy": 5.307663440704346,
      "epoch": 3.65015816528272,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037162975633569704,
      "loss": 4.7478,
      "mean_token_accuracy": 0.2359714090824127,
      "num_tokens": 84584871.0,
      "step": 36925
    },
    {
      "entropy": 5.373053646087646,
      "epoch": 3.650652431791222,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037159788535615486,
      "loss": 4.9446,
      "mean_token_accuracy": 0.2162610948085785,
      "num_tokens": 84596122.0,
      "step": 36930
    },
    {
      "entropy": 5.372900295257568,
      "epoch": 3.651146698299723,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037156601200028975,
      "loss": 4.9032,
      "mean_token_accuracy": 0.22475857883691788,
      "num_tokens": 84607432.0,
      "step": 36935
    },
    {
      "entropy": 5.358575534820557,
      "epoch": 3.651640964808225,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003715341362688857,
      "loss": 4.8054,
      "mean_token_accuracy": 0.23150832056999207,
      "num_tokens": 84618440.0,
      "step": 36940
    },
    {
      "entropy": 5.2888916492462155,
      "epoch": 3.652135231316726,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003715022581627269,
      "loss": 4.8048,
      "mean_token_accuracy": 0.23704465627670288,
      "num_tokens": 84629173.0,
      "step": 36945
    },
    {
      "entropy": 5.314864540100098,
      "epoch": 3.6526294978252274,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003714703776825976,
      "loss": 4.7762,
      "mean_token_accuracy": 0.23005301356315613,
      "num_tokens": 84639770.0,
      "step": 36950
    },
    {
      "entropy": 5.352747106552124,
      "epoch": 3.6531237643337287,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003714384948292821,
      "loss": 4.9152,
      "mean_token_accuracy": 0.2208740457892418,
      "num_tokens": 84651405.0,
      "step": 36955
    },
    {
      "entropy": 5.3103289127349855,
      "epoch": 3.65361803084223,
      "grad_norm": 1.0,
      "learning_rate": 0.0003714066096035646,
      "loss": 4.7603,
      "mean_token_accuracy": 0.24185348749160768,
      "num_tokens": 84663136.0,
      "step": 36960
    },
    {
      "entropy": 5.317334365844727,
      "epoch": 3.6541122973507316,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003713747220062296,
      "loss": 4.8103,
      "mean_token_accuracy": 0.23086691796779632,
      "num_tokens": 84675432.0,
      "step": 36965
    },
    {
      "entropy": 5.30422625541687,
      "epoch": 3.654606563859233,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003713428320380614,
      "loss": 4.8161,
      "mean_token_accuracy": 0.2357401505112648,
      "num_tokens": 84686961.0,
      "step": 36970
    },
    {
      "entropy": 5.296778011322021,
      "epoch": 3.655100830367734,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003713109396998447,
      "loss": 4.7811,
      "mean_token_accuracy": 0.2357327163219452,
      "num_tokens": 84697316.0,
      "step": 36975
    },
    {
      "entropy": 5.371257638931274,
      "epoch": 3.655595096876236,
      "grad_norm": 1.0,
      "learning_rate": 0.00037127904499236386,
      "loss": 4.9525,
      "mean_token_accuracy": 0.22163959890604018,
      "num_tokens": 84709228.0,
      "step": 36980
    },
    {
      "entropy": 5.412351036071778,
      "epoch": 3.656089363384737,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037124714791640363,
      "loss": 4.8648,
      "mean_token_accuracy": 0.21919483691453934,
      "num_tokens": 84721300.0,
      "step": 36985
    },
    {
      "entropy": 5.3403558254241945,
      "epoch": 3.6565836298932384,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037121524847274855,
      "loss": 4.8797,
      "mean_token_accuracy": 0.22751695960760115,
      "num_tokens": 84732264.0,
      "step": 36990
    },
    {
      "entropy": 5.273571586608886,
      "epoch": 3.6570778964017396,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003711833466621835,
      "loss": 4.8327,
      "mean_token_accuracy": 0.222713865339756,
      "num_tokens": 84743636.0,
      "step": 36995
    },
    {
      "entropy": 5.329427337646484,
      "epoch": 3.6575721629102413,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037115144248549324,
      "loss": 4.8472,
      "mean_token_accuracy": 0.22944098860025405,
      "num_tokens": 84754696.0,
      "step": 37000
    },
    {
      "entropy": 5.450957202911377,
      "epoch": 3.6580664294187426,
      "grad_norm": 1.125,
      "learning_rate": 0.00037111953594346256,
      "loss": 4.8986,
      "mean_token_accuracy": 0.2210584431886673,
      "num_tokens": 84766026.0,
      "step": 37005
    },
    {
      "entropy": 5.384374380111694,
      "epoch": 3.658560695927244,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037108762703687634,
      "loss": 4.8743,
      "mean_token_accuracy": 0.23002879321575165,
      "num_tokens": 84777265.0,
      "step": 37010
    },
    {
      "entropy": 5.28236665725708,
      "epoch": 3.6590549624357456,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003710557157665196,
      "loss": 4.8443,
      "mean_token_accuracy": 0.2285000666975975,
      "num_tokens": 84787715.0,
      "step": 37015
    },
    {
      "entropy": 5.399721765518189,
      "epoch": 3.659549228944247,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003710238021331773,
      "loss": 4.9737,
      "mean_token_accuracy": 0.22135744243860245,
      "num_tokens": 84798874.0,
      "step": 37020
    },
    {
      "entropy": 5.412326955795288,
      "epoch": 3.660043495452748,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003709918861376346,
      "loss": 4.9712,
      "mean_token_accuracy": 0.2162778005003929,
      "num_tokens": 84810072.0,
      "step": 37025
    },
    {
      "entropy": 5.417651748657226,
      "epoch": 3.6605377619612494,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037095996778067664,
      "loss": 4.8633,
      "mean_token_accuracy": 0.232083822786808,
      "num_tokens": 84820831.0,
      "step": 37030
    },
    {
      "entropy": 5.285230207443237,
      "epoch": 3.6610320284697506,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003709280470630886,
      "loss": 4.7646,
      "mean_token_accuracy": 0.23941393047571183,
      "num_tokens": 84832329.0,
      "step": 37035
    },
    {
      "entropy": 5.34065523147583,
      "epoch": 3.6615262949782523,
      "grad_norm": 1.0,
      "learning_rate": 0.0003708961239856557,
      "loss": 4.9043,
      "mean_token_accuracy": 0.22205893993377684,
      "num_tokens": 84844239.0,
      "step": 37040
    },
    {
      "entropy": 5.317165660858154,
      "epoch": 3.6620205614867536,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00037086419854916327,
      "loss": 4.8689,
      "mean_token_accuracy": 0.2310461461544037,
      "num_tokens": 84855939.0,
      "step": 37045
    },
    {
      "entropy": 5.347628831863403,
      "epoch": 3.6625148279952553,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003708322707543967,
      "loss": 4.8523,
      "mean_token_accuracy": 0.2235839322209358,
      "num_tokens": 84867170.0,
      "step": 37050
    },
    {
      "entropy": 5.370975923538208,
      "epoch": 3.6630090945037566,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037080034060214134,
      "loss": 4.8773,
      "mean_token_accuracy": 0.23343184292316438,
      "num_tokens": 84877986.0,
      "step": 37055
    },
    {
      "entropy": 5.2930474281311035,
      "epoch": 3.663503361012258,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003707684080931827,
      "loss": 4.7803,
      "mean_token_accuracy": 0.23917830139398574,
      "num_tokens": 84888795.0,
      "step": 37060
    },
    {
      "entropy": 5.2725005626678465,
      "epoch": 3.663997627520759,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037073647322830657,
      "loss": 4.8444,
      "mean_token_accuracy": 0.23543160855770112,
      "num_tokens": 84900569.0,
      "step": 37065
    },
    {
      "entropy": 5.365699577331543,
      "epoch": 3.6644918940292603,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003707045360082982,
      "loss": 4.9457,
      "mean_token_accuracy": 0.22274716943502426,
      "num_tokens": 84912018.0,
      "step": 37070
    },
    {
      "entropy": 5.42576265335083,
      "epoch": 3.664986160537762,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003706725964339434,
      "loss": 4.9115,
      "mean_token_accuracy": 0.21912020891904832,
      "num_tokens": 84923088.0,
      "step": 37075
    },
    {
      "entropy": 5.330938196182251,
      "epoch": 3.6654804270462633,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000370640654506028,
      "loss": 4.8609,
      "mean_token_accuracy": 0.2320971205830574,
      "num_tokens": 84933838.0,
      "step": 37080
    },
    {
      "entropy": 5.319458961486816,
      "epoch": 3.6659746935547646,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003706087102253375,
      "loss": 4.8317,
      "mean_token_accuracy": 0.22839322537183762,
      "num_tokens": 84945208.0,
      "step": 37085
    },
    {
      "entropy": 5.271191263198853,
      "epoch": 3.6664689600632663,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000370576763592658,
      "loss": 4.7915,
      "mean_token_accuracy": 0.23921951055526733,
      "num_tokens": 84956224.0,
      "step": 37090
    },
    {
      "entropy": 5.394811820983887,
      "epoch": 3.6669632265717675,
      "grad_norm": 1.0,
      "learning_rate": 0.0003705448146087752,
      "loss": 4.9325,
      "mean_token_accuracy": 0.23088499903678894,
      "num_tokens": 84967507.0,
      "step": 37095
    },
    {
      "entropy": 5.351563787460327,
      "epoch": 3.667457493080269,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00037051286327447527,
      "loss": 4.8363,
      "mean_token_accuracy": 0.2297585517168045,
      "num_tokens": 84979578.0,
      "step": 37100
    },
    {
      "entropy": 5.459745979309082,
      "epoch": 3.66795175958877,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003704809095905439,
      "loss": 5.0377,
      "mean_token_accuracy": 0.21018216460943223,
      "num_tokens": 84990875.0,
      "step": 37105
    },
    {
      "entropy": 5.296231126785278,
      "epoch": 3.6684460260972718,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037044895355776747,
      "loss": 4.8057,
      "mean_token_accuracy": 0.23377428501844405,
      "num_tokens": 85001787.0,
      "step": 37110
    },
    {
      "entropy": 5.39904294013977,
      "epoch": 3.668940292605773,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000370416995176932,
      "loss": 4.9544,
      "mean_token_accuracy": 0.2177300363779068,
      "num_tokens": 85015189.0,
      "step": 37115
    },
    {
      "entropy": 5.377111196517944,
      "epoch": 3.6694345591142743,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003703850344488236,
      "loss": 4.8965,
      "mean_token_accuracy": 0.23029059916734695,
      "num_tokens": 85026267.0,
      "step": 37120
    },
    {
      "entropy": 5.335699605941772,
      "epoch": 3.669928825622776,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00037035307137422855,
      "loss": 4.8478,
      "mean_token_accuracy": 0.22860579788684846,
      "num_tokens": 85038296.0,
      "step": 37125
    },
    {
      "entropy": 5.318237066268921,
      "epoch": 3.6704230921312773,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037032110595393313,
      "loss": 4.8748,
      "mean_token_accuracy": 0.23026354014873504,
      "num_tokens": 85050518.0,
      "step": 37130
    },
    {
      "entropy": 5.3926276683807375,
      "epoch": 3.6709173586397785,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037028913818872375,
      "loss": 4.8754,
      "mean_token_accuracy": 0.22115449160337447,
      "num_tokens": 85062002.0,
      "step": 37135
    },
    {
      "entropy": 5.388245248794556,
      "epoch": 3.67141162514828,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037025716807938676,
      "loss": 4.9072,
      "mean_token_accuracy": 0.23021192103624344,
      "num_tokens": 85072818.0,
      "step": 37140
    },
    {
      "entropy": 5.315536832809448,
      "epoch": 3.6719058916567815,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003702251956267087,
      "loss": 4.8165,
      "mean_token_accuracy": 0.2368546187877655,
      "num_tokens": 85083747.0,
      "step": 37145
    },
    {
      "entropy": 5.338272571563721,
      "epoch": 3.6724001581652828,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000370193220831476,
      "loss": 4.8378,
      "mean_token_accuracy": 0.23062866926193237,
      "num_tokens": 85095546.0,
      "step": 37150
    },
    {
      "entropy": 5.345551490783691,
      "epoch": 3.672894424673784,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037016124369447533,
      "loss": 4.8893,
      "mean_token_accuracy": 0.22542488873004912,
      "num_tokens": 85105891.0,
      "step": 37155
    },
    {
      "entropy": 5.341178321838379,
      "epoch": 3.6733886911822857,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003701292642164933,
      "loss": 4.8985,
      "mean_token_accuracy": 0.2209042266011238,
      "num_tokens": 85118378.0,
      "step": 37160
    },
    {
      "entropy": 5.3939813613891605,
      "epoch": 3.673882957690787,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037009728239831665,
      "loss": 4.9374,
      "mean_token_accuracy": 0.22148457020521164,
      "num_tokens": 85130111.0,
      "step": 37165
    },
    {
      "entropy": 5.389163255691528,
      "epoch": 3.6743772241992882,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003700652982407321,
      "loss": 4.9425,
      "mean_token_accuracy": 0.22257527858018875,
      "num_tokens": 85140037.0,
      "step": 37170
    },
    {
      "entropy": 5.3506749153137205,
      "epoch": 3.6748714907077895,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003700333117445264,
      "loss": 4.8325,
      "mean_token_accuracy": 0.233109450340271,
      "num_tokens": 85152398.0,
      "step": 37175
    },
    {
      "entropy": 5.434806632995605,
      "epoch": 3.6753657572162908,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037000132291048655,
      "loss": 4.9393,
      "mean_token_accuracy": 0.22201527804136276,
      "num_tokens": 85164793.0,
      "step": 37180
    },
    {
      "entropy": 5.368464231491089,
      "epoch": 3.6758600237247925,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036996933173939935,
      "loss": 4.9324,
      "mean_token_accuracy": 0.21793834567070008,
      "num_tokens": 85176291.0,
      "step": 37185
    },
    {
      "entropy": 5.329867506027222,
      "epoch": 3.6763542902332937,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003699373382320519,
      "loss": 4.89,
      "mean_token_accuracy": 0.22061939686536788,
      "num_tokens": 85188706.0,
      "step": 37190
    },
    {
      "entropy": 5.401124382019043,
      "epoch": 3.6768485567417954,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036990534238923124,
      "loss": 4.9526,
      "mean_token_accuracy": 0.22581952214241027,
      "num_tokens": 85200771.0,
      "step": 37195
    },
    {
      "entropy": 5.367476320266723,
      "epoch": 3.6773428232502967,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036987334421172434,
      "loss": 4.9231,
      "mean_token_accuracy": 0.22476439774036408,
      "num_tokens": 85211209.0,
      "step": 37200
    },
    {
      "entropy": 5.331131410598755,
      "epoch": 3.677837089758798,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036984134370031846,
      "loss": 4.8203,
      "mean_token_accuracy": 0.23611721247434617,
      "num_tokens": 85222370.0,
      "step": 37205
    },
    {
      "entropy": 5.334611463546753,
      "epoch": 3.6783313562672992,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003698093408558009,
      "loss": 4.8096,
      "mean_token_accuracy": 0.22926597893238068,
      "num_tokens": 85234210.0,
      "step": 37210
    },
    {
      "entropy": 5.351852607727051,
      "epoch": 3.6788256227758005,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003697773356789588,
      "loss": 4.8795,
      "mean_token_accuracy": 0.23031645268201828,
      "num_tokens": 85245506.0,
      "step": 37215
    },
    {
      "entropy": 5.3380146503448485,
      "epoch": 3.679319889284302,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036974532817057946,
      "loss": 4.8541,
      "mean_token_accuracy": 0.22823161780834197,
      "num_tokens": 85256904.0,
      "step": 37220
    },
    {
      "entropy": 5.32533860206604,
      "epoch": 3.6798141557928035,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003697133183314505,
      "loss": 4.8754,
      "mean_token_accuracy": 0.2229703426361084,
      "num_tokens": 85268660.0,
      "step": 37225
    },
    {
      "entropy": 5.350935792922973,
      "epoch": 3.6803084223013047,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003696813061623591,
      "loss": 4.9164,
      "mean_token_accuracy": 0.22102297395467757,
      "num_tokens": 85280292.0,
      "step": 37230
    },
    {
      "entropy": 5.364089584350586,
      "epoch": 3.6808026888098064,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036964929166409294,
      "loss": 4.8966,
      "mean_token_accuracy": 0.22385205328464508,
      "num_tokens": 85292029.0,
      "step": 37235
    },
    {
      "entropy": 5.267552280426026,
      "epoch": 3.6812969553183077,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00036961727483743957,
      "loss": 4.7705,
      "mean_token_accuracy": 0.22518443316221237,
      "num_tokens": 85304350.0,
      "step": 37240
    },
    {
      "entropy": 5.341428947448731,
      "epoch": 3.681791221826809,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00036958525568318645,
      "loss": 4.8969,
      "mean_token_accuracy": 0.2259019359946251,
      "num_tokens": 85315646.0,
      "step": 37245
    },
    {
      "entropy": 5.469727468490601,
      "epoch": 3.68228548833531,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003695532342021214,
      "loss": 4.8912,
      "mean_token_accuracy": 0.2215048760175705,
      "num_tokens": 85328779.0,
      "step": 37250
    },
    {
      "entropy": 5.286429691314697,
      "epoch": 3.682779754843812,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003695212103950321,
      "loss": 4.8019,
      "mean_token_accuracy": 0.22999473065137863,
      "num_tokens": 85339916.0,
      "step": 37255
    },
    {
      "entropy": 5.34763879776001,
      "epoch": 3.683274021352313,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036948918426270646,
      "loss": 4.9578,
      "mean_token_accuracy": 0.2180489867925644,
      "num_tokens": 85351381.0,
      "step": 37260
    },
    {
      "entropy": 5.3334955215454105,
      "epoch": 3.6837682878608144,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00036945715580593213,
      "loss": 4.8975,
      "mean_token_accuracy": 0.2207340866327286,
      "num_tokens": 85362166.0,
      "step": 37265
    },
    {
      "entropy": 5.342495393753052,
      "epoch": 3.684262554369316,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003694251250254971,
      "loss": 4.8197,
      "mean_token_accuracy": 0.22968224287033082,
      "num_tokens": 85373626.0,
      "step": 37270
    },
    {
      "entropy": 5.390260791778564,
      "epoch": 3.6847568208778174,
      "grad_norm": 0.953125,
      "learning_rate": 0.00036939309192218946,
      "loss": 4.8884,
      "mean_token_accuracy": 0.21760563403367997,
      "num_tokens": 85385217.0,
      "step": 37275
    },
    {
      "entropy": 5.3575116157531735,
      "epoch": 3.6852510873863187,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000369361056496797,
      "loss": 4.8818,
      "mean_token_accuracy": 0.22889053225517272,
      "num_tokens": 85395944.0,
      "step": 37280
    },
    {
      "entropy": 5.36527099609375,
      "epoch": 3.68574535389482,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036932901875010797,
      "loss": 4.8948,
      "mean_token_accuracy": 0.22553078085184097,
      "num_tokens": 85407614.0,
      "step": 37285
    },
    {
      "entropy": 5.278856897354126,
      "epoch": 3.686239620403321,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036929697868291045,
      "loss": 4.8748,
      "mean_token_accuracy": 0.2310721442103386,
      "num_tokens": 85419006.0,
      "step": 37290
    },
    {
      "entropy": 5.344970750808716,
      "epoch": 3.686733886911823,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036926493629599257,
      "loss": 4.8397,
      "mean_token_accuracy": 0.24039324820041658,
      "num_tokens": 85430267.0,
      "step": 37295
    },
    {
      "entropy": 5.368427705764771,
      "epoch": 3.687228153420324,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003692328915901427,
      "loss": 4.8716,
      "mean_token_accuracy": 0.2288208469748497,
      "num_tokens": 85441424.0,
      "step": 37300
    },
    {
      "entropy": 5.294268417358398,
      "epoch": 3.687722419928826,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003692008445661491,
      "loss": 4.809,
      "mean_token_accuracy": 0.2325630307197571,
      "num_tokens": 85452606.0,
      "step": 37305
    },
    {
      "entropy": 5.324365472793579,
      "epoch": 3.688216686437327,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00036916879522480003,
      "loss": 4.8631,
      "mean_token_accuracy": 0.2328535497188568,
      "num_tokens": 85463532.0,
      "step": 37310
    },
    {
      "entropy": 5.397299575805664,
      "epoch": 3.6887109529458284,
      "grad_norm": 1.03125,
      "learning_rate": 0.000369136743566884,
      "loss": 4.9099,
      "mean_token_accuracy": 0.22854031026363372,
      "num_tokens": 85476014.0,
      "step": 37315
    },
    {
      "entropy": 5.339652681350708,
      "epoch": 3.6892052194543297,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003691046895931895,
      "loss": 4.8356,
      "mean_token_accuracy": 0.23448090106248856,
      "num_tokens": 85488296.0,
      "step": 37320
    },
    {
      "entropy": 5.392187643051147,
      "epoch": 3.689699485962831,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036907263330450505,
      "loss": 4.9393,
      "mean_token_accuracy": 0.22416726797819136,
      "num_tokens": 85498579.0,
      "step": 37325
    },
    {
      "entropy": 5.359186792373658,
      "epoch": 3.6901937524713326,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036904057470161914,
      "loss": 4.869,
      "mean_token_accuracy": 0.22490999847650528,
      "num_tokens": 85509978.0,
      "step": 37330
    },
    {
      "entropy": 5.382182979583741,
      "epoch": 3.690688018979834,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036900851378532065,
      "loss": 4.9204,
      "mean_token_accuracy": 0.22402578592300415,
      "num_tokens": 85522277.0,
      "step": 37335
    },
    {
      "entropy": 5.309179973602295,
      "epoch": 3.691182285488335,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036897645055639794,
      "loss": 4.8767,
      "mean_token_accuracy": 0.22820470929145814,
      "num_tokens": 85532880.0,
      "step": 37340
    },
    {
      "entropy": 5.311588954925537,
      "epoch": 3.691676551996837,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003689443850156401,
      "loss": 4.7849,
      "mean_token_accuracy": 0.23696347326040268,
      "num_tokens": 85543714.0,
      "step": 37345
    },
    {
      "entropy": 5.305025053024292,
      "epoch": 3.692170818505338,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003689123171638358,
      "loss": 4.8212,
      "mean_token_accuracy": 0.23167711347341538,
      "num_tokens": 85555870.0,
      "step": 37350
    },
    {
      "entropy": 5.248937320709229,
      "epoch": 3.6926650850138394,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000368880247001774,
      "loss": 4.7671,
      "mean_token_accuracy": 0.2432101011276245,
      "num_tokens": 85567607.0,
      "step": 37355
    },
    {
      "entropy": 5.296472215652466,
      "epoch": 3.6931593515223406,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003688481745302435,
      "loss": 4.8573,
      "mean_token_accuracy": 0.23187728822231293,
      "num_tokens": 85578448.0,
      "step": 37360
    },
    {
      "entropy": 5.397369432449341,
      "epoch": 3.6936536180308424,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003688160997500332,
      "loss": 4.9485,
      "mean_token_accuracy": 0.21830458790063859,
      "num_tokens": 85589220.0,
      "step": 37365
    },
    {
      "entropy": 5.451288843154908,
      "epoch": 3.6941478845393436,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036878402266193243,
      "loss": 5.0329,
      "mean_token_accuracy": 0.21796121299266816,
      "num_tokens": 85600095.0,
      "step": 37370
    },
    {
      "entropy": 5.379939174652099,
      "epoch": 3.694642151047845,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036875194326673025,
      "loss": 4.8851,
      "mean_token_accuracy": 0.2243768259882927,
      "num_tokens": 85612188.0,
      "step": 37375
    },
    {
      "entropy": 5.316216897964478,
      "epoch": 3.6951364175563466,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036871986156521554,
      "loss": 4.7904,
      "mean_token_accuracy": 0.23131428211927413,
      "num_tokens": 85623262.0,
      "step": 37380
    },
    {
      "entropy": 5.326744174957275,
      "epoch": 3.695630684064848,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036868777755817776,
      "loss": 4.8814,
      "mean_token_accuracy": 0.22689824253320695,
      "num_tokens": 85634811.0,
      "step": 37385
    },
    {
      "entropy": 5.3951582431793215,
      "epoch": 3.696124950573349,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003686556912464062,
      "loss": 4.9782,
      "mean_token_accuracy": 0.22041257917881013,
      "num_tokens": 85645739.0,
      "step": 37390
    },
    {
      "entropy": 5.370418691635132,
      "epoch": 3.6966192170818504,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036862360263069,
      "loss": 4.8712,
      "mean_token_accuracy": 0.2315293326973915,
      "num_tokens": 85657124.0,
      "step": 37395
    },
    {
      "entropy": 5.417504405975341,
      "epoch": 3.697113483590352,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003685915117118187,
      "loss": 4.8805,
      "mean_token_accuracy": 0.22560217529535292,
      "num_tokens": 85667750.0,
      "step": 37400
    },
    {
      "entropy": 5.295740938186645,
      "epoch": 3.6976077500988533,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003685594184905817,
      "loss": 4.874,
      "mean_token_accuracy": 0.2303482860326767,
      "num_tokens": 85679009.0,
      "step": 37405
    },
    {
      "entropy": 5.367545652389526,
      "epoch": 3.6981020166073546,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003685273229677684,
      "loss": 4.9233,
      "mean_token_accuracy": 0.22301653176546096,
      "num_tokens": 85690925.0,
      "step": 37410
    },
    {
      "entropy": 5.480771064758301,
      "epoch": 3.6985962831158563,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036849522514416854,
      "loss": 5.0001,
      "mean_token_accuracy": 0.21524980664253235,
      "num_tokens": 85702045.0,
      "step": 37415
    },
    {
      "entropy": 5.337428188323974,
      "epoch": 3.6990905496243576,
      "grad_norm": 1.0,
      "learning_rate": 0.0003684631250205716,
      "loss": 4.8878,
      "mean_token_accuracy": 0.22536301761865615,
      "num_tokens": 85713589.0,
      "step": 37420
    },
    {
      "entropy": 5.321285724639893,
      "epoch": 3.699584816132859,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00036843102259776736,
      "loss": 4.8243,
      "mean_token_accuracy": 0.23445389568805694,
      "num_tokens": 85726060.0,
      "step": 37425
    },
    {
      "entropy": 5.501253271102906,
      "epoch": 3.70007908264136,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003683989178765454,
      "loss": 4.9881,
      "mean_token_accuracy": 0.2160787671804428,
      "num_tokens": 85737473.0,
      "step": 37430
    },
    {
      "entropy": 5.394416570663452,
      "epoch": 3.7005733491498614,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003683668108576955,
      "loss": 4.9019,
      "mean_token_accuracy": 0.22621079236268998,
      "num_tokens": 85749392.0,
      "step": 37435
    },
    {
      "entropy": 5.373353672027588,
      "epoch": 3.701067615658363,
      "grad_norm": 0.953125,
      "learning_rate": 0.00036833470154200765,
      "loss": 4.9057,
      "mean_token_accuracy": 0.2211281478404999,
      "num_tokens": 85761180.0,
      "step": 37440
    },
    {
      "entropy": 5.271851396560669,
      "epoch": 3.7015618821668643,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003683025899302717,
      "loss": 4.8243,
      "mean_token_accuracy": 0.2373792812228203,
      "num_tokens": 85772230.0,
      "step": 37445
    },
    {
      "entropy": 5.377283525466919,
      "epoch": 3.702056148675366,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036827047602327757,
      "loss": 4.9038,
      "mean_token_accuracy": 0.22940150648355484,
      "num_tokens": 85783413.0,
      "step": 37450
    },
    {
      "entropy": 5.397137260437011,
      "epoch": 3.7025504151838673,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00036823835982181515,
      "loss": 4.8916,
      "mean_token_accuracy": 0.22874766141176223,
      "num_tokens": 85794057.0,
      "step": 37455
    },
    {
      "entropy": 5.369738054275513,
      "epoch": 3.7030446816923686,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036820624132667467,
      "loss": 4.8143,
      "mean_token_accuracy": 0.22291359901428223,
      "num_tokens": 85805379.0,
      "step": 37460
    },
    {
      "entropy": 5.404136800765992,
      "epoch": 3.70353894820087,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036817412053864626,
      "loss": 4.8703,
      "mean_token_accuracy": 0.22723647952079773,
      "num_tokens": 85817177.0,
      "step": 37465
    },
    {
      "entropy": 5.293812656402588,
      "epoch": 3.704033214709371,
      "grad_norm": 0.96875,
      "learning_rate": 0.00036814199745851997,
      "loss": 4.8663,
      "mean_token_accuracy": 0.2314177095890045,
      "num_tokens": 85830017.0,
      "step": 37470
    },
    {
      "entropy": 5.329754209518432,
      "epoch": 3.704527481217873,
      "grad_norm": 1.125,
      "learning_rate": 0.00036810987208708607,
      "loss": 4.976,
      "mean_token_accuracy": 0.21888710260391236,
      "num_tokens": 85840835.0,
      "step": 37475
    },
    {
      "entropy": 5.353043079376221,
      "epoch": 3.705021747726374,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036807774442513495,
      "loss": 4.859,
      "mean_token_accuracy": 0.23001932203769684,
      "num_tokens": 85852492.0,
      "step": 37480
    },
    {
      "entropy": 5.365503740310669,
      "epoch": 3.7055160142348753,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003680456144734567,
      "loss": 4.8556,
      "mean_token_accuracy": 0.22784620225429536,
      "num_tokens": 85864570.0,
      "step": 37485
    },
    {
      "entropy": 5.3175300598144535,
      "epoch": 3.706010280743377,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003680134822328421,
      "loss": 4.8595,
      "mean_token_accuracy": 0.2264627695083618,
      "num_tokens": 85875516.0,
      "step": 37490
    },
    {
      "entropy": 5.355171489715576,
      "epoch": 3.7065045472518783,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003679813477040814,
      "loss": 4.868,
      "mean_token_accuracy": 0.22997350841760636,
      "num_tokens": 85886051.0,
      "step": 37495
    },
    {
      "entropy": 5.373681116104126,
      "epoch": 3.7069988137603795,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000367949210887965,
      "loss": 4.9491,
      "mean_token_accuracy": 0.2235301598906517,
      "num_tokens": 85897996.0,
      "step": 37500
    },
    {
      "entropy": 5.315188884735107,
      "epoch": 3.707493080268881,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036791707178528355,
      "loss": 4.7756,
      "mean_token_accuracy": 0.2352730616927147,
      "num_tokens": 85908562.0,
      "step": 37505
    },
    {
      "entropy": 5.3776054859161375,
      "epoch": 3.7079873467773825,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003678849303968278,
      "loss": 4.9066,
      "mean_token_accuracy": 0.22013149708509444,
      "num_tokens": 85920859.0,
      "step": 37510
    },
    {
      "entropy": 5.259235286712647,
      "epoch": 3.7084816132858838,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003678527867233884,
      "loss": 4.8466,
      "mean_token_accuracy": 0.22778060883283616,
      "num_tokens": 85932919.0,
      "step": 37515
    },
    {
      "entropy": 5.295482206344604,
      "epoch": 3.708975879794385,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00036782064076575603,
      "loss": 4.7815,
      "mean_token_accuracy": 0.23084331601858138,
      "num_tokens": 85943399.0,
      "step": 37520
    },
    {
      "entropy": 5.413701438903809,
      "epoch": 3.7094701463028867,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036778849252472144,
      "loss": 4.8931,
      "mean_token_accuracy": 0.2234555169939995,
      "num_tokens": 85954110.0,
      "step": 37525
    },
    {
      "entropy": 5.390600919723511,
      "epoch": 3.709964412811388,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003677563420010755,
      "loss": 4.8787,
      "mean_token_accuracy": 0.23031336665153504,
      "num_tokens": 85965955.0,
      "step": 37530
    },
    {
      "entropy": 5.340165853500366,
      "epoch": 3.7104586793198893,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036772418919560937,
      "loss": 4.824,
      "mean_token_accuracy": 0.22824975997209548,
      "num_tokens": 85977475.0,
      "step": 37535
    },
    {
      "entropy": 5.318088102340698,
      "epoch": 3.7109529458283905,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003676920341091136,
      "loss": 4.8726,
      "mean_token_accuracy": 0.22235322147607803,
      "num_tokens": 85987995.0,
      "step": 37540
    },
    {
      "entropy": 5.355809116363526,
      "epoch": 3.711447212336892,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003676598767423795,
      "loss": 4.9973,
      "mean_token_accuracy": 0.21126776784658433,
      "num_tokens": 86000379.0,
      "step": 37545
    },
    {
      "entropy": 5.397480535507202,
      "epoch": 3.7119414788453935,
      "grad_norm": 1.125,
      "learning_rate": 0.000367627717096198,
      "loss": 4.8551,
      "mean_token_accuracy": 0.22699600458145142,
      "num_tokens": 86011959.0,
      "step": 37550
    },
    {
      "entropy": 5.40672197341919,
      "epoch": 3.7124357453538948,
      "grad_norm": 1.0,
      "learning_rate": 0.0003675955551713603,
      "loss": 4.8895,
      "mean_token_accuracy": 0.2211748942732811,
      "num_tokens": 86023015.0,
      "step": 37555
    },
    {
      "entropy": 5.2837486743927,
      "epoch": 3.7129300118623965,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003675633909686576,
      "loss": 4.8265,
      "mean_token_accuracy": 0.2310906708240509,
      "num_tokens": 86032802.0,
      "step": 37560
    },
    {
      "entropy": 5.373885011672973,
      "epoch": 3.7134242783708977,
      "grad_norm": 0.984375,
      "learning_rate": 0.00036753122448888114,
      "loss": 4.8699,
      "mean_token_accuracy": 0.2324053794145584,
      "num_tokens": 86043499.0,
      "step": 37565
    },
    {
      "entropy": 5.3693140029907225,
      "epoch": 3.713918544879399,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036749905573282214,
      "loss": 4.8885,
      "mean_token_accuracy": 0.2202165961265564,
      "num_tokens": 86054327.0,
      "step": 37570
    },
    {
      "entropy": 5.435192108154297,
      "epoch": 3.7144128113879002,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00036746688470127214,
      "loss": 5.0107,
      "mean_token_accuracy": 0.2128878727555275,
      "num_tokens": 86067023.0,
      "step": 37575
    },
    {
      "entropy": 5.377259922027588,
      "epoch": 3.7149070778964015,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036743471139502235,
      "loss": 4.8547,
      "mean_token_accuracy": 0.22782364785671233,
      "num_tokens": 86079906.0,
      "step": 37580
    },
    {
      "entropy": 5.370411968231201,
      "epoch": 3.715401344404903,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003674025358148643,
      "loss": 4.8779,
      "mean_token_accuracy": 0.2297048956155777,
      "num_tokens": 86090467.0,
      "step": 37585
    },
    {
      "entropy": 5.290926885604859,
      "epoch": 3.7158956109134045,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003673703579615895,
      "loss": 4.8481,
      "mean_token_accuracy": 0.2268904849886894,
      "num_tokens": 86102908.0,
      "step": 37590
    },
    {
      "entropy": 5.254464721679687,
      "epoch": 3.7163898774219057,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036733817783598973,
      "loss": 4.753,
      "mean_token_accuracy": 0.24840002804994582,
      "num_tokens": 86114588.0,
      "step": 37595
    },
    {
      "entropy": 5.370078897476196,
      "epoch": 3.7168841439304074,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003673059954388563,
      "loss": 4.8947,
      "mean_token_accuracy": 0.2252398133277893,
      "num_tokens": 86126800.0,
      "step": 37600
    },
    {
      "entropy": 5.373450660705567,
      "epoch": 3.7173784104389087,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003672738107709811,
      "loss": 4.8965,
      "mean_token_accuracy": 0.22675887048244475,
      "num_tokens": 86138703.0,
      "step": 37605
    },
    {
      "entropy": 5.3448103904724125,
      "epoch": 3.71787267694741,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036724162383315584,
      "loss": 4.8047,
      "mean_token_accuracy": 0.2295199990272522,
      "num_tokens": 86148856.0,
      "step": 37610
    },
    {
      "entropy": 5.452143716812134,
      "epoch": 3.7183669434559112,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036720943462617236,
      "loss": 4.9172,
      "mean_token_accuracy": 0.22458101361989974,
      "num_tokens": 86160108.0,
      "step": 37615
    },
    {
      "entropy": 5.362561511993408,
      "epoch": 3.718861209964413,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036717724315082245,
      "loss": 4.8892,
      "mean_token_accuracy": 0.2268182247877121,
      "num_tokens": 86170065.0,
      "step": 37620
    },
    {
      "entropy": 5.336143732070923,
      "epoch": 3.719355476472914,
      "grad_norm": 1.03125,
      "learning_rate": 0.000367145049407898,
      "loss": 4.8658,
      "mean_token_accuracy": 0.22561706304550172,
      "num_tokens": 86180469.0,
      "step": 37625
    },
    {
      "entropy": 5.422434234619141,
      "epoch": 3.7198497429814155,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036711285339819115,
      "loss": 4.9821,
      "mean_token_accuracy": 0.22278884798288345,
      "num_tokens": 86191801.0,
      "step": 37630
    },
    {
      "entropy": 5.3249329090118405,
      "epoch": 3.720344009489917,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003670806551224938,
      "loss": 4.8314,
      "mean_token_accuracy": 0.22526707798242568,
      "num_tokens": 86202967.0,
      "step": 37635
    },
    {
      "entropy": 5.398179149627685,
      "epoch": 3.7208382759984184,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036704845458159803,
      "loss": 4.9841,
      "mean_token_accuracy": 0.21622924655675888,
      "num_tokens": 86215225.0,
      "step": 37640
    },
    {
      "entropy": 5.411929798126221,
      "epoch": 3.7213325425069197,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036701625177629595,
      "loss": 4.9662,
      "mean_token_accuracy": 0.21607402414083482,
      "num_tokens": 86227417.0,
      "step": 37645
    },
    {
      "entropy": 5.348657751083374,
      "epoch": 3.721826809015421,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003669840467073798,
      "loss": 4.839,
      "mean_token_accuracy": 0.2351645991206169,
      "num_tokens": 86238749.0,
      "step": 37650
    },
    {
      "entropy": 5.334732341766357,
      "epoch": 3.7223210755239227,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003669518393756419,
      "loss": 4.943,
      "mean_token_accuracy": 0.22031576782464982,
      "num_tokens": 86250146.0,
      "step": 37655
    },
    {
      "entropy": 5.415461921691895,
      "epoch": 3.722815342032424,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003669196297818745,
      "loss": 4.9312,
      "mean_token_accuracy": 0.22262810319662094,
      "num_tokens": 86262086.0,
      "step": 37660
    },
    {
      "entropy": 5.42926025390625,
      "epoch": 3.723309608540925,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036688741792686985,
      "loss": 4.9532,
      "mean_token_accuracy": 0.22329955846071242,
      "num_tokens": 86273937.0,
      "step": 37665
    },
    {
      "entropy": 5.396841621398925,
      "epoch": 3.723803875049427,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003668552038114205,
      "loss": 4.9186,
      "mean_token_accuracy": 0.226019886136055,
      "num_tokens": 86284451.0,
      "step": 37670
    },
    {
      "entropy": 5.366309022903442,
      "epoch": 3.724298141557928,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003668229874363189,
      "loss": 4.8536,
      "mean_token_accuracy": 0.23373554944992064,
      "num_tokens": 86296406.0,
      "step": 37675
    },
    {
      "entropy": 5.307525491714477,
      "epoch": 3.7247924080664294,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036679076880235764,
      "loss": 4.8175,
      "mean_token_accuracy": 0.2322388544678688,
      "num_tokens": 86306922.0,
      "step": 37680
    },
    {
      "entropy": 5.317340230941772,
      "epoch": 3.7252866745749307,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036675854791032916,
      "loss": 4.8281,
      "mean_token_accuracy": 0.23478362262248992,
      "num_tokens": 86317785.0,
      "step": 37685
    },
    {
      "entropy": 5.469619226455689,
      "epoch": 3.725780941083432,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003667263247610261,
      "loss": 4.9695,
      "mean_token_accuracy": 0.21636180430650712,
      "num_tokens": 86329271.0,
      "step": 37690
    },
    {
      "entropy": 5.388055753707886,
      "epoch": 3.7262752075919336,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036669409935524136,
      "loss": 4.9219,
      "mean_token_accuracy": 0.22021329253911973,
      "num_tokens": 86339739.0,
      "step": 37695
    },
    {
      "entropy": 5.263347864151001,
      "epoch": 3.726769474100435,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003666618716937674,
      "loss": 4.8011,
      "mean_token_accuracy": 0.23168611973524095,
      "num_tokens": 86351111.0,
      "step": 37700
    },
    {
      "entropy": 5.3821934223175045,
      "epoch": 3.7272637406089366,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003666296417773973,
      "loss": 4.9257,
      "mean_token_accuracy": 0.2253758803009987,
      "num_tokens": 86363320.0,
      "step": 37705
    },
    {
      "entropy": 5.444052886962891,
      "epoch": 3.727758007117438,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003665974096069237,
      "loss": 4.932,
      "mean_token_accuracy": 0.22425027191638947,
      "num_tokens": 86374372.0,
      "step": 37710
    },
    {
      "entropy": 5.452753639221191,
      "epoch": 3.728252273625939,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003665651751831397,
      "loss": 4.9271,
      "mean_token_accuracy": 0.21717116683721543,
      "num_tokens": 86385545.0,
      "step": 37715
    },
    {
      "entropy": 5.317073106765747,
      "epoch": 3.7287465401344404,
      "grad_norm": 0.96875,
      "learning_rate": 0.00036653293850683803,
      "loss": 4.8728,
      "mean_token_accuracy": 0.23135519921779632,
      "num_tokens": 86398240.0,
      "step": 37720
    },
    {
      "entropy": 5.340314674377441,
      "epoch": 3.7292408066429417,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003665006995788121,
      "loss": 4.8857,
      "mean_token_accuracy": 0.22791634649038314,
      "num_tokens": 86409024.0,
      "step": 37725
    },
    {
      "entropy": 5.45699667930603,
      "epoch": 3.7297350731514434,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036646845839985464,
      "loss": 4.9097,
      "mean_token_accuracy": 0.22176747024059296,
      "num_tokens": 86422114.0,
      "step": 37730
    },
    {
      "entropy": 5.39141263961792,
      "epoch": 3.7302293396599446,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00036643621497075895,
      "loss": 4.936,
      "mean_token_accuracy": 0.21898404508829117,
      "num_tokens": 86435230.0,
      "step": 37735
    },
    {
      "entropy": 5.3408417224884035,
      "epoch": 3.730723606168446,
      "grad_norm": 1.0,
      "learning_rate": 0.00036640396929231813,
      "loss": 4.9824,
      "mean_token_accuracy": 0.21672510653734206,
      "num_tokens": 86446449.0,
      "step": 37740
    },
    {
      "entropy": 5.3964801788330075,
      "epoch": 3.7312178726769476,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003663717213653255,
      "loss": 4.8835,
      "mean_token_accuracy": 0.22500354498624803,
      "num_tokens": 86456814.0,
      "step": 37745
    },
    {
      "entropy": 5.390552043914795,
      "epoch": 3.731712139185449,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003663394711905744,
      "loss": 4.8402,
      "mean_token_accuracy": 0.22391951978206634,
      "num_tokens": 86468135.0,
      "step": 37750
    },
    {
      "entropy": 5.42072548866272,
      "epoch": 3.73220640569395,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036630721876885813,
      "loss": 4.9552,
      "mean_token_accuracy": 0.22633278071880342,
      "num_tokens": 86480268.0,
      "step": 37755
    },
    {
      "entropy": 5.323319721221924,
      "epoch": 3.7327006722024514,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036627496410097005,
      "loss": 4.811,
      "mean_token_accuracy": 0.23501592129468918,
      "num_tokens": 86492660.0,
      "step": 37760
    },
    {
      "entropy": 5.346138334274292,
      "epoch": 3.733194938710953,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036624270718770376,
      "loss": 4.9111,
      "mean_token_accuracy": 0.22300967425107956,
      "num_tokens": 86504638.0,
      "step": 37765
    },
    {
      "entropy": 5.314207887649536,
      "epoch": 3.7336892052194544,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003662104480298527,
      "loss": 4.8008,
      "mean_token_accuracy": 0.23697238564491271,
      "num_tokens": 86514225.0,
      "step": 37770
    },
    {
      "entropy": 5.3790826320648195,
      "epoch": 3.7341834717279556,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003661781866282104,
      "loss": 4.9003,
      "mean_token_accuracy": 0.2297806113958359,
      "num_tokens": 86525370.0,
      "step": 37775
    },
    {
      "entropy": 5.308582878112793,
      "epoch": 3.7346777382364573,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003661459229835707,
      "loss": 4.8067,
      "mean_token_accuracy": 0.23344469517469407,
      "num_tokens": 86535137.0,
      "step": 37780
    },
    {
      "entropy": 5.263848495483399,
      "epoch": 3.7351720047449586,
      "grad_norm": 1.125,
      "learning_rate": 0.00036611365709672707,
      "loss": 4.7681,
      "mean_token_accuracy": 0.23744950741529464,
      "num_tokens": 86546922.0,
      "step": 37785
    },
    {
      "entropy": 5.403316783905029,
      "epoch": 3.73566627125346,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003660813889684733,
      "loss": 4.9454,
      "mean_token_accuracy": 0.2184808999300003,
      "num_tokens": 86559356.0,
      "step": 37790
    },
    {
      "entropy": 5.30443172454834,
      "epoch": 3.736160537761961,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003660491185996033,
      "loss": 4.7535,
      "mean_token_accuracy": 0.24112751185894013,
      "num_tokens": 86571515.0,
      "step": 37795
    },
    {
      "entropy": 5.3521645069122314,
      "epoch": 3.7366548042704624,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003660168459909109,
      "loss": 4.8497,
      "mean_token_accuracy": 0.2254160851240158,
      "num_tokens": 86581641.0,
      "step": 37800
    },
    {
      "entropy": 5.290193748474121,
      "epoch": 3.737149070778964,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036598457114318984,
      "loss": 4.8661,
      "mean_token_accuracy": 0.22370484322309495,
      "num_tokens": 86594478.0,
      "step": 37805
    },
    {
      "entropy": 5.344572591781616,
      "epoch": 3.7376433372874653,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003659522940572343,
      "loss": 4.8347,
      "mean_token_accuracy": 0.22966650128364563,
      "num_tokens": 86605496.0,
      "step": 37810
    },
    {
      "entropy": 5.360835790634155,
      "epoch": 3.738137603795967,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036592001473383817,
      "loss": 4.8361,
      "mean_token_accuracy": 0.22982072681188584,
      "num_tokens": 86616745.0,
      "step": 37815
    },
    {
      "entropy": 5.374407958984375,
      "epoch": 3.7386318703044683,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003658877331737956,
      "loss": 4.9072,
      "mean_token_accuracy": 0.21839386820793152,
      "num_tokens": 86628094.0,
      "step": 37820
    },
    {
      "entropy": 5.365602111816406,
      "epoch": 3.7391261368129696,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036585544937790067,
      "loss": 4.8707,
      "mean_token_accuracy": 0.23322008699178695,
      "num_tokens": 86639421.0,
      "step": 37825
    },
    {
      "entropy": 5.270666027069092,
      "epoch": 3.739620403321471,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036582316334694766,
      "loss": 4.844,
      "mean_token_accuracy": 0.2352962464094162,
      "num_tokens": 86650653.0,
      "step": 37830
    },
    {
      "entropy": 5.40148229598999,
      "epoch": 3.740114669829972,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003657908750817307,
      "loss": 4.8453,
      "mean_token_accuracy": 0.23131305128335952,
      "num_tokens": 86660954.0,
      "step": 37835
    },
    {
      "entropy": 5.398519706726074,
      "epoch": 3.740608936338474,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00036575858458304407,
      "loss": 4.9471,
      "mean_token_accuracy": 0.2154250130057335,
      "num_tokens": 86671983.0,
      "step": 37840
    },
    {
      "entropy": 5.290555620193482,
      "epoch": 3.741103202846975,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036572629185168223,
      "loss": 4.7695,
      "mean_token_accuracy": 0.23971644788980484,
      "num_tokens": 86682634.0,
      "step": 37845
    },
    {
      "entropy": 5.359875345230103,
      "epoch": 3.7415974693554763,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003656939968884396,
      "loss": 4.903,
      "mean_token_accuracy": 0.2207307606935501,
      "num_tokens": 86693516.0,
      "step": 37850
    },
    {
      "entropy": 5.429385566711426,
      "epoch": 3.742091735863978,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036566169969411047,
      "loss": 4.9368,
      "mean_token_accuracy": 0.21972199231386186,
      "num_tokens": 86705350.0,
      "step": 37855
    },
    {
      "entropy": 5.3876160144805905,
      "epoch": 3.7425860023724793,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003656294002694895,
      "loss": 4.8447,
      "mean_token_accuracy": 0.22443049103021623,
      "num_tokens": 86715647.0,
      "step": 37860
    },
    {
      "entropy": 5.3528906345367435,
      "epoch": 3.7430802688809806,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003655970986153712,
      "loss": 4.9797,
      "mean_token_accuracy": 0.21694946736097337,
      "num_tokens": 86726453.0,
      "step": 37865
    },
    {
      "entropy": 5.333933925628662,
      "epoch": 3.743574535389482,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003655647947325503,
      "loss": 4.8831,
      "mean_token_accuracy": 0.2264312118291855,
      "num_tokens": 86737400.0,
      "step": 37870
    },
    {
      "entropy": 5.451200723648071,
      "epoch": 3.7440688018979835,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003655324886218213,
      "loss": 4.9975,
      "mean_token_accuracy": 0.22252598851919175,
      "num_tokens": 86748667.0,
      "step": 37875
    },
    {
      "entropy": 5.355406856536865,
      "epoch": 3.744563068406485,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036550018028397915,
      "loss": 4.8622,
      "mean_token_accuracy": 0.23090711534023284,
      "num_tokens": 86761081.0,
      "step": 37880
    },
    {
      "entropy": 5.421183252334595,
      "epoch": 3.745057334914986,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003654678697198184,
      "loss": 4.9303,
      "mean_token_accuracy": 0.2251483365893364,
      "num_tokens": 86770823.0,
      "step": 37885
    },
    {
      "entropy": 5.391869354248047,
      "epoch": 3.7455516014234878,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036543555693013406,
      "loss": 4.9073,
      "mean_token_accuracy": 0.22635042071342468,
      "num_tokens": 86784362.0,
      "step": 37890
    },
    {
      "entropy": 5.401109266281128,
      "epoch": 3.746045867931989,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003654032419157211,
      "loss": 4.9357,
      "mean_token_accuracy": 0.22845033556222916,
      "num_tokens": 86796679.0,
      "step": 37895
    },
    {
      "entropy": 5.3401796340942385,
      "epoch": 3.7465401344404903,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003653709246773743,
      "loss": 4.9115,
      "mean_token_accuracy": 0.23093120455741883,
      "num_tokens": 86808272.0,
      "step": 37900
    },
    {
      "entropy": 5.406108427047729,
      "epoch": 3.7470344009489915,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003653386052158886,
      "loss": 4.905,
      "mean_token_accuracy": 0.22383288890123368,
      "num_tokens": 86819428.0,
      "step": 37905
    },
    {
      "entropy": 5.327954530715942,
      "epoch": 3.7475286674574932,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003653062835320593,
      "loss": 4.8884,
      "mean_token_accuracy": 0.22125830501317978,
      "num_tokens": 86831373.0,
      "step": 37910
    },
    {
      "entropy": 5.288205671310425,
      "epoch": 3.7480229339659945,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036527395962668145,
      "loss": 4.7903,
      "mean_token_accuracy": 0.23108956664800645,
      "num_tokens": 86843196.0,
      "step": 37915
    },
    {
      "entropy": 5.4184177875518795,
      "epoch": 3.7485172004744958,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003652416335005502,
      "loss": 4.8846,
      "mean_token_accuracy": 0.22523870766162873,
      "num_tokens": 86855233.0,
      "step": 37920
    },
    {
      "entropy": 5.394596195220947,
      "epoch": 3.7490114669829975,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003652093051544607,
      "loss": 4.8376,
      "mean_token_accuracy": 0.22645198404788972,
      "num_tokens": 86865994.0,
      "step": 37925
    },
    {
      "entropy": 5.3088610649108885,
      "epoch": 3.7495057334914987,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036517697458920837,
      "loss": 4.8209,
      "mean_token_accuracy": 0.22754251658916474,
      "num_tokens": 86877451.0,
      "step": 37930
    },
    {
      "entropy": 5.311286926269531,
      "epoch": 3.75,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036514464180558836,
      "loss": 4.8438,
      "mean_token_accuracy": 0.22704618722200393,
      "num_tokens": 86888390.0,
      "step": 37935
    },
    {
      "entropy": 5.359607553482055,
      "epoch": 3.7504942665085013,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003651123068043963,
      "loss": 4.9451,
      "mean_token_accuracy": 0.2212749943137169,
      "num_tokens": 86899703.0,
      "step": 37940
    },
    {
      "entropy": 5.383024978637695,
      "epoch": 3.7509885330170025,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00036507996958642746,
      "loss": 4.8817,
      "mean_token_accuracy": 0.23018389642238618,
      "num_tokens": 86911753.0,
      "step": 37945
    },
    {
      "entropy": 5.5081065654754635,
      "epoch": 3.7514827995255042,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003650476301524773,
      "loss": 5.0456,
      "mean_token_accuracy": 0.22067539244890214,
      "num_tokens": 86924165.0,
      "step": 37950
    },
    {
      "entropy": 5.392571687698364,
      "epoch": 3.7519770660340055,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003650152885033416,
      "loss": 4.9233,
      "mean_token_accuracy": 0.2235248178243637,
      "num_tokens": 86935848.0,
      "step": 37955
    },
    {
      "entropy": 5.313085079193115,
      "epoch": 3.7524713325425068,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00036498294463981583,
      "loss": 4.8171,
      "mean_token_accuracy": 0.22586697190999985,
      "num_tokens": 86946816.0,
      "step": 37960
    },
    {
      "entropy": 5.418202495574951,
      "epoch": 3.7529655990510085,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036495059856269556,
      "loss": 4.9111,
      "mean_token_accuracy": 0.2251382663846016,
      "num_tokens": 86958436.0,
      "step": 37965
    },
    {
      "entropy": 5.4179126739501955,
      "epoch": 3.7534598655595097,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003649182502727768,
      "loss": 4.8706,
      "mean_token_accuracy": 0.22353545874357222,
      "num_tokens": 86970338.0,
      "step": 37970
    },
    {
      "entropy": 5.335784530639648,
      "epoch": 3.753954132068011,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036488589977085495,
      "loss": 4.8616,
      "mean_token_accuracy": 0.22124839425086976,
      "num_tokens": 86981796.0,
      "step": 37975
    },
    {
      "entropy": 5.257256984710693,
      "epoch": 3.7544483985765122,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036485354705772605,
      "loss": 4.8133,
      "mean_token_accuracy": 0.2379045844078064,
      "num_tokens": 86993681.0,
      "step": 37980
    },
    {
      "entropy": 5.4019153118133545,
      "epoch": 3.754942665085014,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036482119213418604,
      "loss": 5.0293,
      "mean_token_accuracy": 0.21565725207328795,
      "num_tokens": 87005621.0,
      "step": 37985
    },
    {
      "entropy": 5.346227216720581,
      "epoch": 3.755436931593515,
      "grad_norm": 1.0,
      "learning_rate": 0.00036478883500103067,
      "loss": 4.8778,
      "mean_token_accuracy": 0.23000551909208297,
      "num_tokens": 87016054.0,
      "step": 37990
    },
    {
      "entropy": 5.377190256118775,
      "epoch": 3.7559311981020165,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003647564756590561,
      "loss": 4.8593,
      "mean_token_accuracy": 0.22651178687810897,
      "num_tokens": 87026603.0,
      "step": 37995
    },
    {
      "entropy": 5.363053703308106,
      "epoch": 3.756425464610518,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003647241141090583,
      "loss": 4.8837,
      "mean_token_accuracy": 0.22475794702768326,
      "num_tokens": 87037036.0,
      "step": 38000
    },
    {
      "entropy": 5.3225304126739506,
      "epoch": 3.7569197311190194,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003646917503518333,
      "loss": 4.8107,
      "mean_token_accuracy": 0.23683008402585984,
      "num_tokens": 87047360.0,
      "step": 38005
    },
    {
      "entropy": 5.2651878833770756,
      "epoch": 3.7574139976275207,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036465938438817735,
      "loss": 4.8111,
      "mean_token_accuracy": 0.23411089330911636,
      "num_tokens": 87058580.0,
      "step": 38010
    },
    {
      "entropy": 5.318077468872071,
      "epoch": 3.757908264136022,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003646270162188867,
      "loss": 4.9272,
      "mean_token_accuracy": 0.22590129673480988,
      "num_tokens": 87070176.0,
      "step": 38015
    },
    {
      "entropy": 5.415104150772095,
      "epoch": 3.7584025306445237,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0003645946458447575,
      "loss": 4.8855,
      "mean_token_accuracy": 0.23313356339931487,
      "num_tokens": 87082188.0,
      "step": 38020
    },
    {
      "entropy": 5.393975114822387,
      "epoch": 3.758896797153025,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003645622732665862,
      "loss": 4.9015,
      "mean_token_accuracy": 0.22035285979509353,
      "num_tokens": 87093784.0,
      "step": 38025
    },
    {
      "entropy": 5.302675533294678,
      "epoch": 3.759391063661526,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036452989848516896,
      "loss": 4.8382,
      "mean_token_accuracy": 0.2278170794248581,
      "num_tokens": 87105102.0,
      "step": 38030
    },
    {
      "entropy": 5.285956859588623,
      "epoch": 3.759885330170028,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036449752150130235,
      "loss": 4.8005,
      "mean_token_accuracy": 0.23450107574462892,
      "num_tokens": 87115956.0,
      "step": 38035
    },
    {
      "entropy": 5.273462438583374,
      "epoch": 3.760379596678529,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036446514231578287,
      "loss": 4.8225,
      "mean_token_accuracy": 0.23392489701509475,
      "num_tokens": 87126514.0,
      "step": 38040
    },
    {
      "entropy": 5.336463928222656,
      "epoch": 3.7608738631870304,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036443276092940703,
      "loss": 4.855,
      "mean_token_accuracy": 0.22503628581762314,
      "num_tokens": 87137249.0,
      "step": 38045
    },
    {
      "entropy": 5.400457954406738,
      "epoch": 3.7613681296955317,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00036440037734297134,
      "loss": 4.9187,
      "mean_token_accuracy": 0.230281962454319,
      "num_tokens": 87148593.0,
      "step": 38050
    },
    {
      "entropy": 5.422585678100586,
      "epoch": 3.761862396204033,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036436799155727253,
      "loss": 4.9102,
      "mean_token_accuracy": 0.22507024705410003,
      "num_tokens": 87160085.0,
      "step": 38055
    },
    {
      "entropy": 5.362833261489868,
      "epoch": 3.7623566627125347,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00036433560357310727,
      "loss": 4.8397,
      "mean_token_accuracy": 0.22409479767084123,
      "num_tokens": 87171427.0,
      "step": 38060
    },
    {
      "entropy": 5.319900369644165,
      "epoch": 3.762850929221036,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003643032133912723,
      "loss": 4.8509,
      "mean_token_accuracy": 0.23050184100866317,
      "num_tokens": 87183073.0,
      "step": 38065
    },
    {
      "entropy": 5.354214096069336,
      "epoch": 3.7633451957295376,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003642708210125644,
      "loss": 4.8618,
      "mean_token_accuracy": 0.2307318389415741,
      "num_tokens": 87193199.0,
      "step": 38070
    },
    {
      "entropy": 5.415476369857788,
      "epoch": 3.763839462238039,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036423842643778045,
      "loss": 4.9146,
      "mean_token_accuracy": 0.22255075871944427,
      "num_tokens": 87205052.0,
      "step": 38075
    },
    {
      "entropy": 5.409682130813598,
      "epoch": 3.76433372874654,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036420602966771734,
      "loss": 4.9537,
      "mean_token_accuracy": 0.22160982936620713,
      "num_tokens": 87217485.0,
      "step": 38080
    },
    {
      "entropy": 5.340493011474609,
      "epoch": 3.7648279952550414,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003641736307031721,
      "loss": 4.8455,
      "mean_token_accuracy": 0.23244841545820236,
      "num_tokens": 87228184.0,
      "step": 38085
    },
    {
      "entropy": 5.336638975143432,
      "epoch": 3.7653222617635427,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00036414122954494177,
      "loss": 4.9452,
      "mean_token_accuracy": 0.22317275255918503,
      "num_tokens": 87240713.0,
      "step": 38090
    },
    {
      "entropy": 5.369870615005493,
      "epoch": 3.7658165282720444,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003641088261938233,
      "loss": 4.9503,
      "mean_token_accuracy": 0.21695732772350312,
      "num_tokens": 87252739.0,
      "step": 38095
    },
    {
      "entropy": 5.283787584304809,
      "epoch": 3.7663107947805456,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003640764206506138,
      "loss": 4.7652,
      "mean_token_accuracy": 0.24448940008878708,
      "num_tokens": 87263034.0,
      "step": 38100
    },
    {
      "entropy": 5.363323450088501,
      "epoch": 3.766805061289047,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003640440129161107,
      "loss": 4.8927,
      "mean_token_accuracy": 0.229938106238842,
      "num_tokens": 87275384.0,
      "step": 38105
    },
    {
      "entropy": 5.393299865722656,
      "epoch": 3.7672993277975486,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036401160299111095,
      "loss": 4.9507,
      "mean_token_accuracy": 0.21932601183652878,
      "num_tokens": 87286568.0,
      "step": 38110
    },
    {
      "entropy": 5.3443763732910154,
      "epoch": 3.76779359430605,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000363979190876412,
      "loss": 4.8495,
      "mean_token_accuracy": 0.23037409037351608,
      "num_tokens": 87297972.0,
      "step": 38115
    },
    {
      "entropy": 5.423915147781372,
      "epoch": 3.768287860814551,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036394677657281113,
      "loss": 4.8392,
      "mean_token_accuracy": 0.23136909753084184,
      "num_tokens": 87309137.0,
      "step": 38120
    },
    {
      "entropy": 5.324648714065551,
      "epoch": 3.7687821273230524,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036391436008110573,
      "loss": 4.9098,
      "mean_token_accuracy": 0.22938600927591324,
      "num_tokens": 87320097.0,
      "step": 38125
    },
    {
      "entropy": 5.297325658798218,
      "epoch": 3.769276393831554,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036388194140209336,
      "loss": 4.8531,
      "mean_token_accuracy": 0.2341489836573601,
      "num_tokens": 87332662.0,
      "step": 38130
    },
    {
      "entropy": 5.40741605758667,
      "epoch": 3.7697706603400554,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036384952053657144,
      "loss": 4.875,
      "mean_token_accuracy": 0.2403239369392395,
      "num_tokens": 87344118.0,
      "step": 38135
    },
    {
      "entropy": 5.268925285339355,
      "epoch": 3.7702649268485566,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036381709748533746,
      "loss": 4.8179,
      "mean_token_accuracy": 0.23086515218019485,
      "num_tokens": 87355787.0,
      "step": 38140
    },
    {
      "entropy": 5.330609321594238,
      "epoch": 3.7707591933570583,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003637846722491892,
      "loss": 4.8305,
      "mean_token_accuracy": 0.2314717710018158,
      "num_tokens": 87367332.0,
      "step": 38145
    },
    {
      "entropy": 5.345599222183227,
      "epoch": 3.7712534598655596,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003637522448289241,
      "loss": 4.834,
      "mean_token_accuracy": 0.23608337193727494,
      "num_tokens": 87378968.0,
      "step": 38150
    },
    {
      "entropy": 5.3243310928344725,
      "epoch": 3.771747726374061,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003637198152253401,
      "loss": 4.8685,
      "mean_token_accuracy": 0.23133375495672226,
      "num_tokens": 87391573.0,
      "step": 38155
    },
    {
      "entropy": 5.345226383209228,
      "epoch": 3.772241992882562,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003636873834392349,
      "loss": 4.8835,
      "mean_token_accuracy": 0.22699294686317445,
      "num_tokens": 87402431.0,
      "step": 38160
    },
    {
      "entropy": 5.325432682037354,
      "epoch": 3.7727362593910634,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003636549494714062,
      "loss": 4.918,
      "mean_token_accuracy": 0.22692957371473313,
      "num_tokens": 87414156.0,
      "step": 38165
    },
    {
      "entropy": 5.326112747192383,
      "epoch": 3.773230525899565,
      "grad_norm": 0.953125,
      "learning_rate": 0.00036362251332265204,
      "loss": 4.8509,
      "mean_token_accuracy": 0.22909048199653625,
      "num_tokens": 87425878.0,
      "step": 38170
    },
    {
      "entropy": 5.365592384338379,
      "epoch": 3.7737247924080664,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003635900749937703,
      "loss": 4.8864,
      "mean_token_accuracy": 0.23072891235351561,
      "num_tokens": 87437171.0,
      "step": 38175
    },
    {
      "entropy": 5.332992458343506,
      "epoch": 3.774219058916568,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000363557634485559,
      "loss": 4.8301,
      "mean_token_accuracy": 0.2271296426653862,
      "num_tokens": 87448465.0,
      "step": 38180
    },
    {
      "entropy": 5.293912267684936,
      "epoch": 3.7747133254250693,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003635251917988162,
      "loss": 4.8148,
      "mean_token_accuracy": 0.2318248674273491,
      "num_tokens": 87458800.0,
      "step": 38185
    },
    {
      "entropy": 5.3281519412994385,
      "epoch": 3.7752075919335706,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003634927469343399,
      "loss": 4.8141,
      "mean_token_accuracy": 0.23595484644174575,
      "num_tokens": 87470427.0,
      "step": 38190
    },
    {
      "entropy": 5.336515426635742,
      "epoch": 3.775701858442072,
      "grad_norm": 1.0,
      "learning_rate": 0.0003634602998929282,
      "loss": 4.837,
      "mean_token_accuracy": 0.23225877583026885,
      "num_tokens": 87481498.0,
      "step": 38195
    },
    {
      "entropy": 5.271830081939697,
      "epoch": 3.776196124950573,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003634278506753795,
      "loss": 4.8081,
      "mean_token_accuracy": 0.2363478273153305,
      "num_tokens": 87493578.0,
      "step": 38200
    },
    {
      "entropy": 5.402250671386719,
      "epoch": 3.776690391459075,
      "grad_norm": 1.125,
      "learning_rate": 0.000363395399282492,
      "loss": 4.9207,
      "mean_token_accuracy": 0.21721294522285461,
      "num_tokens": 87504387.0,
      "step": 38205
    },
    {
      "entropy": 5.4143274307250975,
      "epoch": 3.777184657967576,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036336294571506387,
      "loss": 4.8452,
      "mean_token_accuracy": 0.2269751861691475,
      "num_tokens": 87514231.0,
      "step": 38210
    },
    {
      "entropy": 5.3479022026062015,
      "epoch": 3.7776789244760773,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036333048997389363,
      "loss": 4.8981,
      "mean_token_accuracy": 0.2273597612977028,
      "num_tokens": 87526483.0,
      "step": 38215
    },
    {
      "entropy": 5.34328179359436,
      "epoch": 3.778173190984579,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003632980320597797,
      "loss": 4.8851,
      "mean_token_accuracy": 0.22264324128627777,
      "num_tokens": 87537931.0,
      "step": 38220
    },
    {
      "entropy": 5.343886089324951,
      "epoch": 3.7786674574930803,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003632655719735203,
      "loss": 4.8533,
      "mean_token_accuracy": 0.2284371718764305,
      "num_tokens": 87549486.0,
      "step": 38225
    },
    {
      "entropy": 5.421779584884644,
      "epoch": 3.7791617240015816,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003632331097159143,
      "loss": 4.9387,
      "mean_token_accuracy": 0.22603436708450317,
      "num_tokens": 87562232.0,
      "step": 38230
    },
    {
      "entropy": 5.394529485702515,
      "epoch": 3.779655990510083,
      "grad_norm": 0.96875,
      "learning_rate": 0.00036320064528776,
      "loss": 4.8211,
      "mean_token_accuracy": 0.23128775358200074,
      "num_tokens": 87574246.0,
      "step": 38235
    },
    {
      "entropy": 5.361046934127808,
      "epoch": 3.7801502570185845,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003631681786898562,
      "loss": 4.8475,
      "mean_token_accuracy": 0.22932169437408448,
      "num_tokens": 87586033.0,
      "step": 38240
    },
    {
      "entropy": 5.3254115104675295,
      "epoch": 3.780644523527086,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003631357099230015,
      "loss": 4.8678,
      "mean_token_accuracy": 0.22787068635225297,
      "num_tokens": 87596468.0,
      "step": 38245
    },
    {
      "entropy": 5.393712711334229,
      "epoch": 3.781138790035587,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003631032389879947,
      "loss": 4.9209,
      "mean_token_accuracy": 0.22521136254072188,
      "num_tokens": 87608577.0,
      "step": 38250
    },
    {
      "entropy": 5.336244201660156,
      "epoch": 3.7816330565440888,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003630707658856345,
      "loss": 4.9027,
      "mean_token_accuracy": 0.22325571924448012,
      "num_tokens": 87620773.0,
      "step": 38255
    },
    {
      "entropy": 5.415024900436402,
      "epoch": 3.78212732305259,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036303829061671986,
      "loss": 4.9224,
      "mean_token_accuracy": 0.22455300837755204,
      "num_tokens": 87633248.0,
      "step": 38260
    },
    {
      "entropy": 5.398361778259277,
      "epoch": 3.7826215895610913,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003630058131820495,
      "loss": 4.8928,
      "mean_token_accuracy": 0.22102607488632203,
      "num_tokens": 87644710.0,
      "step": 38265
    },
    {
      "entropy": 5.3185759544372555,
      "epoch": 3.7831158560695926,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003629733335824226,
      "loss": 4.8241,
      "mean_token_accuracy": 0.23241257071495056,
      "num_tokens": 87655297.0,
      "step": 38270
    },
    {
      "entropy": 5.324174880981445,
      "epoch": 3.7836101225780943,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000362940851818638,
      "loss": 4.8376,
      "mean_token_accuracy": 0.23247567117214202,
      "num_tokens": 87667595.0,
      "step": 38275
    },
    {
      "entropy": 5.327319860458374,
      "epoch": 3.7841043890865955,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00036290836789149483,
      "loss": 4.8494,
      "mean_token_accuracy": 0.22743949592113494,
      "num_tokens": 87680041.0,
      "step": 38280
    },
    {
      "entropy": 5.3841451644897464,
      "epoch": 3.784598655595097,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036287588180179215,
      "loss": 4.927,
      "mean_token_accuracy": 0.22256843894720077,
      "num_tokens": 87691287.0,
      "step": 38285
    },
    {
      "entropy": 5.367755794525147,
      "epoch": 3.7850929221035985,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003628433935503291,
      "loss": 4.9023,
      "mean_token_accuracy": 0.22961652427911758,
      "num_tokens": 87704224.0,
      "step": 38290
    },
    {
      "entropy": 5.353496122360229,
      "epoch": 3.7855871886120998,
      "grad_norm": 1.109375,
      "learning_rate": 0.000362810903137905,
      "loss": 4.8646,
      "mean_token_accuracy": 0.2328616887331009,
      "num_tokens": 87714449.0,
      "step": 38295
    },
    {
      "entropy": 5.364243698120117,
      "epoch": 3.786081455120601,
      "grad_norm": 1.0,
      "learning_rate": 0.000362778410565319,
      "loss": 4.8797,
      "mean_token_accuracy": 0.22384672611951828,
      "num_tokens": 87725348.0,
      "step": 38300
    },
    {
      "entropy": 5.373865127563477,
      "epoch": 3.7865757216291023,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036274591583337055,
      "loss": 4.871,
      "mean_token_accuracy": 0.2292153373360634,
      "num_tokens": 87737244.0,
      "step": 38305
    },
    {
      "entropy": 5.296612405776978,
      "epoch": 3.7870699881376035,
      "grad_norm": 0.953125,
      "learning_rate": 0.00036271341894285894,
      "loss": 4.8715,
      "mean_token_accuracy": 0.2200282484292984,
      "num_tokens": 87748710.0,
      "step": 38310
    },
    {
      "entropy": 5.336021804809571,
      "epoch": 3.7875642546461052,
      "grad_norm": 1.0,
      "learning_rate": 0.00036268091989458357,
      "loss": 4.9501,
      "mean_token_accuracy": 0.22037243247032165,
      "num_tokens": 87759944.0,
      "step": 38315
    },
    {
      "entropy": 5.418487501144409,
      "epoch": 3.7880585211546065,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003626484186893441,
      "loss": 4.8825,
      "mean_token_accuracy": 0.22596318274736404,
      "num_tokens": 87771201.0,
      "step": 38320
    },
    {
      "entropy": 5.368275690078735,
      "epoch": 3.788552787663108,
      "grad_norm": 1.0,
      "learning_rate": 0.0003626159153279398,
      "loss": 4.8358,
      "mean_token_accuracy": 0.2353709116578102,
      "num_tokens": 87782129.0,
      "step": 38325
    },
    {
      "entropy": 5.293320655822754,
      "epoch": 3.7890470541716095,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036258340981117044,
      "loss": 4.834,
      "mean_token_accuracy": 0.23037270158529283,
      "num_tokens": 87792915.0,
      "step": 38330
    },
    {
      "entropy": 5.326624584197998,
      "epoch": 3.7895413206801107,
      "grad_norm": 1.0,
      "learning_rate": 0.00036255090213983555,
      "loss": 4.8162,
      "mean_token_accuracy": 0.22929777055978776,
      "num_tokens": 87803301.0,
      "step": 38335
    },
    {
      "entropy": 5.4223309516906735,
      "epoch": 3.790035587188612,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036251839231473494,
      "loss": 4.947,
      "mean_token_accuracy": 0.2223178267478943,
      "num_tokens": 87815167.0,
      "step": 38340
    },
    {
      "entropy": 5.356593275070191,
      "epoch": 3.7905298536971133,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003624858803366684,
      "loss": 4.889,
      "mean_token_accuracy": 0.22587422281503677,
      "num_tokens": 87827348.0,
      "step": 38345
    },
    {
      "entropy": 5.296116781234741,
      "epoch": 3.791024120205615,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003624533662064355,
      "loss": 4.8227,
      "mean_token_accuracy": 0.24096131324768066,
      "num_tokens": 87837986.0,
      "step": 38350
    },
    {
      "entropy": 5.310147285461426,
      "epoch": 3.7915183867141162,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003624208499248362,
      "loss": 4.7822,
      "mean_token_accuracy": 0.23434026539325714,
      "num_tokens": 87847992.0,
      "step": 38355
    },
    {
      "entropy": 5.394367933273315,
      "epoch": 3.7920126532226175,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003623883314926705,
      "loss": 5.0051,
      "mean_token_accuracy": 0.21825725585222244,
      "num_tokens": 87859768.0,
      "step": 38360
    },
    {
      "entropy": 5.359997415542603,
      "epoch": 3.792506919731119,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003623558109107383,
      "loss": 4.8759,
      "mean_token_accuracy": 0.22261158674955367,
      "num_tokens": 87871366.0,
      "step": 38365
    },
    {
      "entropy": 5.325900602340698,
      "epoch": 3.7930011862396205,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003623232881798395,
      "loss": 4.8614,
      "mean_token_accuracy": 0.23714397549629213,
      "num_tokens": 87883074.0,
      "step": 38370
    },
    {
      "entropy": 5.324184799194336,
      "epoch": 3.7934954527481217,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036229076330077445,
      "loss": 4.8427,
      "mean_token_accuracy": 0.22796983271837234,
      "num_tokens": 87894906.0,
      "step": 38375
    },
    {
      "entropy": 5.344142818450928,
      "epoch": 3.793989719256623,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003622582362743429,
      "loss": 4.7806,
      "mean_token_accuracy": 0.23789840191602707,
      "num_tokens": 87906027.0,
      "step": 38380
    },
    {
      "entropy": 5.311930227279663,
      "epoch": 3.7944839857651247,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003622257071013453,
      "loss": 4.8668,
      "mean_token_accuracy": 0.22572862058877946,
      "num_tokens": 87916348.0,
      "step": 38385
    },
    {
      "entropy": 5.329049348831177,
      "epoch": 3.794978252273626,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003621931757825817,
      "loss": 4.8607,
      "mean_token_accuracy": 0.2262539356946945,
      "num_tokens": 87928397.0,
      "step": 38390
    },
    {
      "entropy": 5.409074878692627,
      "epoch": 3.795472518782127,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00036216064231885255,
      "loss": 4.8935,
      "mean_token_accuracy": 0.22128913402557374,
      "num_tokens": 87939925.0,
      "step": 38395
    },
    {
      "entropy": 5.338060426712036,
      "epoch": 3.795966785290629,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036212810671095797,
      "loss": 4.8852,
      "mean_token_accuracy": 0.23139142990112305,
      "num_tokens": 87951166.0,
      "step": 38400
    },
    {
      "entropy": 5.335096979141236,
      "epoch": 3.79646105179913,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036209556895969847,
      "loss": 4.8103,
      "mean_token_accuracy": 0.2354331061244011,
      "num_tokens": 87962861.0,
      "step": 38405
    },
    {
      "entropy": 5.282854509353638,
      "epoch": 3.7969553183076314,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003620630290658745,
      "loss": 4.717,
      "mean_token_accuracy": 0.2399861440062523,
      "num_tokens": 87974362.0,
      "step": 38410
    },
    {
      "entropy": 5.364345741271973,
      "epoch": 3.7974495848161327,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003620304870302865,
      "loss": 4.8671,
      "mean_token_accuracy": 0.2263699248433113,
      "num_tokens": 87984982.0,
      "step": 38415
    },
    {
      "entropy": 5.3159984111785885,
      "epoch": 3.797943851324634,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00036199794285373505,
      "loss": 4.8386,
      "mean_token_accuracy": 0.2292311742901802,
      "num_tokens": 87996980.0,
      "step": 38420
    },
    {
      "entropy": 5.360739040374756,
      "epoch": 3.7984381178331357,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036196539653702073,
      "loss": 4.8417,
      "mean_token_accuracy": 0.23049808889627457,
      "num_tokens": 88008585.0,
      "step": 38425
    },
    {
      "entropy": 5.2831724166870115,
      "epoch": 3.798932384341637,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000361932848080944,
      "loss": 4.8257,
      "mean_token_accuracy": 0.23755874037742614,
      "num_tokens": 88020779.0,
      "step": 38430
    },
    {
      "entropy": 5.380709791183472,
      "epoch": 3.7994266508501386,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003619002974863059,
      "loss": 4.9326,
      "mean_token_accuracy": 0.22018684148788453,
      "num_tokens": 88033003.0,
      "step": 38435
    },
    {
      "entropy": 5.381381177902222,
      "epoch": 3.79992091735864,
      "grad_norm": 0.921875,
      "learning_rate": 0.00036186774475390686,
      "loss": 4.8244,
      "mean_token_accuracy": 0.2334182381629944,
      "num_tokens": 88044095.0,
      "step": 38440
    },
    {
      "entropy": 5.393739652633667,
      "epoch": 3.800415183867141,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036183518988454794,
      "loss": 4.9341,
      "mean_token_accuracy": 0.22692109942436217,
      "num_tokens": 88056487.0,
      "step": 38445
    },
    {
      "entropy": 5.308628892898559,
      "epoch": 3.8009094503756424,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003618026328790299,
      "loss": 4.8881,
      "mean_token_accuracy": 0.23089361786842347,
      "num_tokens": 88067803.0,
      "step": 38450
    },
    {
      "entropy": 5.415417432785034,
      "epoch": 3.8014037168841437,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003617700737381535,
      "loss": 4.8922,
      "mean_token_accuracy": 0.23220450580120086,
      "num_tokens": 88078914.0,
      "step": 38455
    },
    {
      "entropy": 5.377992916107178,
      "epoch": 3.8018979833926454,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003617375124627199,
      "loss": 4.8629,
      "mean_token_accuracy": 0.22476836144924164,
      "num_tokens": 88090567.0,
      "step": 38460
    },
    {
      "entropy": 5.358003091812134,
      "epoch": 3.8023922499011467,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036170494905352996,
      "loss": 4.8804,
      "mean_token_accuracy": 0.22413631677627563,
      "num_tokens": 88101171.0,
      "step": 38465
    },
    {
      "entropy": 5.373101997375488,
      "epoch": 3.802886516409648,
      "grad_norm": 0.984375,
      "learning_rate": 0.00036167238351138487,
      "loss": 4.8891,
      "mean_token_accuracy": 0.22952985614538193,
      "num_tokens": 88111629.0,
      "step": 38470
    },
    {
      "entropy": 5.412649393081665,
      "epoch": 3.8033807829181496,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036163981583708574,
      "loss": 4.8945,
      "mean_token_accuracy": 0.22377551198005677,
      "num_tokens": 88122354.0,
      "step": 38475
    },
    {
      "entropy": 5.297402620315552,
      "epoch": 3.803875049426651,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003616072460314336,
      "loss": 4.8497,
      "mean_token_accuracy": 0.23153023272752762,
      "num_tokens": 88134830.0,
      "step": 38480
    },
    {
      "entropy": 5.300182867050171,
      "epoch": 3.804369315935152,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036157467409522986,
      "loss": 4.7884,
      "mean_token_accuracy": 0.23461138755083083,
      "num_tokens": 88145995.0,
      "step": 38485
    },
    {
      "entropy": 5.390954303741455,
      "epoch": 3.8048635824436534,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003615421000292757,
      "loss": 4.9015,
      "mean_token_accuracy": 0.22602294534444808,
      "num_tokens": 88157777.0,
      "step": 38490
    },
    {
      "entropy": 5.413525581359863,
      "epoch": 3.805357848952155,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036150952383437246,
      "loss": 4.9329,
      "mean_token_accuracy": 0.21816909015178682,
      "num_tokens": 88169971.0,
      "step": 38495
    },
    {
      "entropy": 5.315480756759643,
      "epoch": 3.8058521154606564,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003614769455113215,
      "loss": 4.7986,
      "mean_token_accuracy": 0.22978946417570115,
      "num_tokens": 88181317.0,
      "step": 38500
    },
    {
      "entropy": 5.4178986072540285,
      "epoch": 3.8063463819691576,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003614443650609242,
      "loss": 5.0009,
      "mean_token_accuracy": 0.21558564454317092,
      "num_tokens": 88194603.0,
      "step": 38505
    },
    {
      "entropy": 5.4000084400177,
      "epoch": 3.8068406484776594,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003614117824839822,
      "loss": 4.9455,
      "mean_token_accuracy": 0.2203699216246605,
      "num_tokens": 88205775.0,
      "step": 38510
    },
    {
      "entropy": 5.356663846969605,
      "epoch": 3.8073349149861606,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003613791977812969,
      "loss": 4.9034,
      "mean_token_accuracy": 0.2205299913883209,
      "num_tokens": 88216784.0,
      "step": 38515
    },
    {
      "entropy": 5.427613925933838,
      "epoch": 3.807829181494662,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036134661095367,
      "loss": 4.8446,
      "mean_token_accuracy": 0.23139554858207703,
      "num_tokens": 88228861.0,
      "step": 38520
    },
    {
      "entropy": 5.410721921920777,
      "epoch": 3.808323448003163,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00036131402200190296,
      "loss": 4.866,
      "mean_token_accuracy": 0.23002018481492997,
      "num_tokens": 88241134.0,
      "step": 38525
    },
    {
      "entropy": 5.316413879394531,
      "epoch": 3.808817714511665,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036128143092679764,
      "loss": 4.8536,
      "mean_token_accuracy": 0.22681862264871597,
      "num_tokens": 88253295.0,
      "step": 38530
    },
    {
      "entropy": 5.36937518119812,
      "epoch": 3.809311981020166,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00036124883772915573,
      "loss": 4.8939,
      "mean_token_accuracy": 0.2213737979531288,
      "num_tokens": 88265906.0,
      "step": 38535
    },
    {
      "entropy": 5.307774639129638,
      "epoch": 3.8098062475286674,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003612162424097791,
      "loss": 4.778,
      "mean_token_accuracy": 0.2374565064907074,
      "num_tokens": 88276928.0,
      "step": 38540
    },
    {
      "entropy": 5.297874450683594,
      "epoch": 3.810300514037169,
      "grad_norm": 1.0,
      "learning_rate": 0.00036118364496946943,
      "loss": 4.8348,
      "mean_token_accuracy": 0.23539720326662064,
      "num_tokens": 88288125.0,
      "step": 38545
    },
    {
      "entropy": 5.348841333389283,
      "epoch": 3.8107947805456703,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003611510454090288,
      "loss": 4.8706,
      "mean_token_accuracy": 0.22641536295413972,
      "num_tokens": 88298778.0,
      "step": 38550
    },
    {
      "entropy": 5.383422470092773,
      "epoch": 3.8112890470541716,
      "grad_norm": 1.09375,
      "learning_rate": 0.000361118443729259,
      "loss": 4.8676,
      "mean_token_accuracy": 0.23242599219083787,
      "num_tokens": 88309086.0,
      "step": 38555
    },
    {
      "entropy": 5.349318170547486,
      "epoch": 3.811783313562673,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003610858399309622,
      "loss": 4.9103,
      "mean_token_accuracy": 0.2235589638352394,
      "num_tokens": 88320358.0,
      "step": 38560
    },
    {
      "entropy": 5.34122257232666,
      "epoch": 3.812277580071174,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003610532340149404,
      "loss": 4.7822,
      "mean_token_accuracy": 0.23881173580884935,
      "num_tokens": 88331767.0,
      "step": 38565
    },
    {
      "entropy": 5.314900255203247,
      "epoch": 3.812771846579676,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036102062598199566,
      "loss": 4.774,
      "mean_token_accuracy": 0.23315003365278245,
      "num_tokens": 88342197.0,
      "step": 38570
    },
    {
      "entropy": 5.3448474407196045,
      "epoch": 3.813266113088177,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003609880158329301,
      "loss": 4.8438,
      "mean_token_accuracy": 0.22857076078653335,
      "num_tokens": 88353378.0,
      "step": 38575
    },
    {
      "entropy": 5.2246129512786865,
      "epoch": 3.813760379596679,
      "grad_norm": 1.0,
      "learning_rate": 0.0003609554035685462,
      "loss": 4.778,
      "mean_token_accuracy": 0.23535057306289672,
      "num_tokens": 88366006.0,
      "step": 38580
    },
    {
      "entropy": 5.401881885528565,
      "epoch": 3.81425464610518,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003609227891896459,
      "loss": 4.9599,
      "mean_token_accuracy": 0.22324532121419907,
      "num_tokens": 88378132.0,
      "step": 38585
    },
    {
      "entropy": 5.327497434616089,
      "epoch": 3.8147489126136813,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036089017269703164,
      "loss": 4.8015,
      "mean_token_accuracy": 0.23151977956295014,
      "num_tokens": 88389760.0,
      "step": 38590
    },
    {
      "entropy": 5.321097755432129,
      "epoch": 3.8152431791221826,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036085755409150577,
      "loss": 4.856,
      "mean_token_accuracy": 0.23330127894878389,
      "num_tokens": 88402518.0,
      "step": 38595
    },
    {
      "entropy": 5.379788875579834,
      "epoch": 3.815737445630684,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003608249333738709,
      "loss": 4.9005,
      "mean_token_accuracy": 0.21971990168094635,
      "num_tokens": 88413359.0,
      "step": 38600
    },
    {
      "entropy": 5.369346284866333,
      "epoch": 3.8162317121391856,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003607923105449292,
      "loss": 4.879,
      "mean_token_accuracy": 0.22521990537643433,
      "num_tokens": 88424966.0,
      "step": 38605
    },
    {
      "entropy": 5.275629043579102,
      "epoch": 3.816725978647687,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036075968560548346,
      "loss": 4.7323,
      "mean_token_accuracy": 0.23908917754888534,
      "num_tokens": 88435826.0,
      "step": 38610
    },
    {
      "entropy": 5.270106506347656,
      "epoch": 3.817220245156188,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036072705855633613,
      "loss": 4.7968,
      "mean_token_accuracy": 0.23496506214141846,
      "num_tokens": 88447136.0,
      "step": 38615
    },
    {
      "entropy": 5.216276359558106,
      "epoch": 3.81771451166469,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003606944293982898,
      "loss": 4.739,
      "mean_token_accuracy": 0.24501573890447617,
      "num_tokens": 88458394.0,
      "step": 38620
    },
    {
      "entropy": 5.370578861236572,
      "epoch": 3.818208778173191,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003606617981321473,
      "loss": 4.8558,
      "mean_token_accuracy": 0.22987488508224488,
      "num_tokens": 88469573.0,
      "step": 38625
    },
    {
      "entropy": 5.351563358306885,
      "epoch": 3.8187030446816923,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036062916475871124,
      "loss": 4.8048,
      "mean_token_accuracy": 0.23460220247507096,
      "num_tokens": 88479561.0,
      "step": 38630
    },
    {
      "entropy": 5.304766273498535,
      "epoch": 3.8191973111901936,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036059652927878446,
      "loss": 4.885,
      "mean_token_accuracy": 0.23363392949104309,
      "num_tokens": 88492296.0,
      "step": 38635
    },
    {
      "entropy": 5.446779251098633,
      "epoch": 3.8196915776986953,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036056389169316977,
      "loss": 4.987,
      "mean_token_accuracy": 0.22411785423755645,
      "num_tokens": 88504993.0,
      "step": 38640
    },
    {
      "entropy": 5.354978752136231,
      "epoch": 3.8201858442071965,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036053125200267,
      "loss": 4.8376,
      "mean_token_accuracy": 0.23644086122512817,
      "num_tokens": 88516126.0,
      "step": 38645
    },
    {
      "entropy": 5.3660839080810545,
      "epoch": 3.820680110715698,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003604986102080882,
      "loss": 4.8928,
      "mean_token_accuracy": 0.21806873828172685,
      "num_tokens": 88527327.0,
      "step": 38650
    },
    {
      "entropy": 5.269163179397583,
      "epoch": 3.8211743772241995,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003604659663102274,
      "loss": 4.7529,
      "mean_token_accuracy": 0.233850659430027,
      "num_tokens": 88539088.0,
      "step": 38655
    },
    {
      "entropy": 5.252378940582275,
      "epoch": 3.8216686437327008,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003604333203098905,
      "loss": 4.8215,
      "mean_token_accuracy": 0.2327428475022316,
      "num_tokens": 88551539.0,
      "step": 38660
    },
    {
      "entropy": 5.342427778244018,
      "epoch": 3.822162910241202,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00036040067220788075,
      "loss": 4.8429,
      "mean_token_accuracy": 0.22641442120075225,
      "num_tokens": 88563111.0,
      "step": 38665
    },
    {
      "entropy": 5.332728290557862,
      "epoch": 3.8226571767497033,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003603680220050011,
      "loss": 4.8759,
      "mean_token_accuracy": 0.2311147227883339,
      "num_tokens": 88574181.0,
      "step": 38670
    },
    {
      "entropy": 5.358493995666504,
      "epoch": 3.8231514432582046,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036033536970205487,
      "loss": 4.9027,
      "mean_token_accuracy": 0.22481439709663392,
      "num_tokens": 88584965.0,
      "step": 38675
    },
    {
      "entropy": 5.465213108062744,
      "epoch": 3.8236457097667063,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00036030271529984535,
      "loss": 4.9611,
      "mean_token_accuracy": 0.21721139997243882,
      "num_tokens": 88597390.0,
      "step": 38680
    },
    {
      "entropy": 5.412862110137939,
      "epoch": 3.8241399762752075,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003602700587991759,
      "loss": 4.8649,
      "mean_token_accuracy": 0.23177101612091064,
      "num_tokens": 88609470.0,
      "step": 38685
    },
    {
      "entropy": 5.3563840866088865,
      "epoch": 3.8246342427837092,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036023740020084963,
      "loss": 4.8482,
      "mean_token_accuracy": 0.23222074210643767,
      "num_tokens": 88620858.0,
      "step": 38690
    },
    {
      "entropy": 5.337638807296753,
      "epoch": 3.8251285092922105,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003602047395056701,
      "loss": 4.8995,
      "mean_token_accuracy": 0.23071579933166503,
      "num_tokens": 88632191.0,
      "step": 38695
    },
    {
      "entropy": 5.391883134841919,
      "epoch": 3.8256227758007118,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036017207671444086,
      "loss": 4.8842,
      "mean_token_accuracy": 0.22145015895366668,
      "num_tokens": 88643405.0,
      "step": 38700
    },
    {
      "entropy": 5.2909448623657225,
      "epoch": 3.826117042309213,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036013941182796516,
      "loss": 4.7111,
      "mean_token_accuracy": 0.2445885121822357,
      "num_tokens": 88653765.0,
      "step": 38705
    },
    {
      "entropy": 5.340324878692627,
      "epoch": 3.8266113088177143,
      "grad_norm": 1.0,
      "learning_rate": 0.00036010674484704684,
      "loss": 4.9087,
      "mean_token_accuracy": 0.22366599291563033,
      "num_tokens": 88665038.0,
      "step": 38710
    },
    {
      "entropy": 5.3206642150878904,
      "epoch": 3.827105575326216,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003600740757724894,
      "loss": 4.8182,
      "mean_token_accuracy": 0.2365850940346718,
      "num_tokens": 88675172.0,
      "step": 38715
    },
    {
      "entropy": 5.333687829971313,
      "epoch": 3.8275998418347172,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036004140460509646,
      "loss": 4.8654,
      "mean_token_accuracy": 0.22959543466567994,
      "num_tokens": 88687393.0,
      "step": 38720
    },
    {
      "entropy": 5.429917860031128,
      "epoch": 3.8280941083432185,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003600087313456718,
      "loss": 4.9491,
      "mean_token_accuracy": 0.2174512967467308,
      "num_tokens": 88700117.0,
      "step": 38725
    },
    {
      "entropy": 5.358299732208252,
      "epoch": 3.82858837485172,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00035997605599501914,
      "loss": 4.8521,
      "mean_token_accuracy": 0.23116564899682998,
      "num_tokens": 88711329.0,
      "step": 38730
    },
    {
      "entropy": 5.446119928359986,
      "epoch": 3.8290826413602215,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003599433785539423,
      "loss": 4.9222,
      "mean_token_accuracy": 0.2237999677658081,
      "num_tokens": 88722017.0,
      "step": 38735
    },
    {
      "entropy": 5.296647644042968,
      "epoch": 3.8295769078687227,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00035991069902324517,
      "loss": 4.8223,
      "mean_token_accuracy": 0.23357392251491546,
      "num_tokens": 88733749.0,
      "step": 38740
    },
    {
      "entropy": 5.367882013320923,
      "epoch": 3.830071174377224,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003598780174037318,
      "loss": 4.8423,
      "mean_token_accuracy": 0.2276182472705841,
      "num_tokens": 88745885.0,
      "step": 38745
    },
    {
      "entropy": 5.452992296218872,
      "epoch": 3.8305654408857257,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003598453336962059,
      "loss": 5.0202,
      "mean_token_accuracy": 0.21784991174936294,
      "num_tokens": 88759216.0,
      "step": 38750
    },
    {
      "entropy": 5.350735282897949,
      "epoch": 3.831059707394227,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035981264790147167,
      "loss": 4.9069,
      "mean_token_accuracy": 0.2278805047273636,
      "num_tokens": 88770272.0,
      "step": 38755
    },
    {
      "entropy": 5.278324174880981,
      "epoch": 3.8315539739027282,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003597799600203331,
      "loss": 4.7816,
      "mean_token_accuracy": 0.24329739511013032,
      "num_tokens": 88781509.0,
      "step": 38760
    },
    {
      "entropy": 5.328139734268189,
      "epoch": 3.83204824041123,
      "grad_norm": 1.171875,
      "learning_rate": 0.00035974727005359434,
      "loss": 4.8873,
      "mean_token_accuracy": 0.22978536039590836,
      "num_tokens": 88791770.0,
      "step": 38765
    },
    {
      "entropy": 5.360169363021851,
      "epoch": 3.832542506919731,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003597145780020597,
      "loss": 4.8782,
      "mean_token_accuracy": 0.2295731097459793,
      "num_tokens": 88804344.0,
      "step": 38770
    },
    {
      "entropy": 5.299268341064453,
      "epoch": 3.8330367734282325,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003596818838665333,
      "loss": 4.7899,
      "mean_token_accuracy": 0.2365250125527382,
      "num_tokens": 88816251.0,
      "step": 38775
    },
    {
      "entropy": 5.291674280166626,
      "epoch": 3.8335310399367337,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003596491876478193,
      "loss": 4.8398,
      "mean_token_accuracy": 0.22594382762908935,
      "num_tokens": 88827805.0,
      "step": 38780
    },
    {
      "entropy": 5.355212926864624,
      "epoch": 3.8340253064452354,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003596164893467222,
      "loss": 4.8451,
      "mean_token_accuracy": 0.23164054304361342,
      "num_tokens": 88839237.0,
      "step": 38785
    },
    {
      "entropy": 5.463304615020752,
      "epoch": 3.8345195729537367,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00035958378896404635,
      "loss": 4.9499,
      "mean_token_accuracy": 0.2250041037797928,
      "num_tokens": 88850950.0,
      "step": 38790
    },
    {
      "entropy": 5.330179214477539,
      "epoch": 3.835013839462238,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00035955108650059613,
      "loss": 4.8136,
      "mean_token_accuracy": 0.23652007132768632,
      "num_tokens": 88863743.0,
      "step": 38795
    },
    {
      "entropy": 5.284814929962158,
      "epoch": 3.8355081059707397,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00035951838195717614,
      "loss": 4.7795,
      "mean_token_accuracy": 0.23637140691280364,
      "num_tokens": 88875375.0,
      "step": 38800
    },
    {
      "entropy": 5.379338788986206,
      "epoch": 3.836002372479241,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035948567533459076,
      "loss": 4.9069,
      "mean_token_accuracy": 0.22986321300268173,
      "num_tokens": 88885890.0,
      "step": 38805
    },
    {
      "entropy": 5.385086297988892,
      "epoch": 3.836496638987742,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00035945296663364467,
      "loss": 4.8483,
      "mean_token_accuracy": 0.22942228615283966,
      "num_tokens": 88897783.0,
      "step": 38810
    },
    {
      "entropy": 5.368971633911133,
      "epoch": 3.8369909054962434,
      "grad_norm": 1.0,
      "learning_rate": 0.00035942025585514257,
      "loss": 4.8962,
      "mean_token_accuracy": 0.22794905453920364,
      "num_tokens": 88908982.0,
      "step": 38815
    },
    {
      "entropy": 5.263632106781006,
      "epoch": 3.8374851720047447,
      "grad_norm": 1.046875,
      "learning_rate": 0.000359387542999889,
      "loss": 4.7684,
      "mean_token_accuracy": 0.24553241729736328,
      "num_tokens": 88921059.0,
      "step": 38820
    },
    {
      "entropy": 5.38988151550293,
      "epoch": 3.8379794385132464,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003593548280686889,
      "loss": 4.9714,
      "mean_token_accuracy": 0.22016989141702653,
      "num_tokens": 88932988.0,
      "step": 38825
    },
    {
      "entropy": 5.359765148162841,
      "epoch": 3.8384737050217477,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003593221110623468,
      "loss": 4.8587,
      "mean_token_accuracy": 0.23190755099058152,
      "num_tokens": 88944890.0,
      "step": 38830
    },
    {
      "entropy": 5.341831254959106,
      "epoch": 3.8389679715302494,
      "grad_norm": 1.0,
      "learning_rate": 0.00035928939198166777,
      "loss": 4.8515,
      "mean_token_accuracy": 0.22803628891706468,
      "num_tokens": 88956936.0,
      "step": 38835
    },
    {
      "entropy": 5.345082187652588,
      "epoch": 3.8394622380387506,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003592566708274566,
      "loss": 4.9039,
      "mean_token_accuracy": 0.22817659825086595,
      "num_tokens": 88968705.0,
      "step": 38840
    },
    {
      "entropy": 5.334768199920655,
      "epoch": 3.839956504547252,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003592239476005183,
      "loss": 4.7929,
      "mean_token_accuracy": 0.23676215261220931,
      "num_tokens": 88979977.0,
      "step": 38845
    },
    {
      "entropy": 5.372999238967895,
      "epoch": 3.840450771055753,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003591912223016578,
      "loss": 4.9113,
      "mean_token_accuracy": 0.22296611070632935,
      "num_tokens": 88991974.0,
      "step": 38850
    },
    {
      "entropy": 5.281067419052124,
      "epoch": 3.8409450375642544,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003591584949316801,
      "loss": 4.7982,
      "mean_token_accuracy": 0.2399026036262512,
      "num_tokens": 89003733.0,
      "step": 38855
    },
    {
      "entropy": 5.4002148628234865,
      "epoch": 3.841439304072756,
      "grad_norm": 1.0,
      "learning_rate": 0.0003591257654913904,
      "loss": 4.9445,
      "mean_token_accuracy": 0.2151666358113289,
      "num_tokens": 89015701.0,
      "step": 38860
    },
    {
      "entropy": 5.409009647369385,
      "epoch": 3.8419335705812574,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003590930339815938,
      "loss": 4.9253,
      "mean_token_accuracy": 0.22643686830997467,
      "num_tokens": 89027216.0,
      "step": 38865
    },
    {
      "entropy": 5.366315174102783,
      "epoch": 3.8424278370897587,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003590603004030956,
      "loss": 4.8483,
      "mean_token_accuracy": 0.22920263260602952,
      "num_tokens": 89039305.0,
      "step": 38870
    },
    {
      "entropy": 5.300953006744384,
      "epoch": 3.8429221035982604,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000359027564756701,
      "loss": 4.7723,
      "mean_token_accuracy": 0.24024063646793364,
      "num_tokens": 89049982.0,
      "step": 38875
    },
    {
      "entropy": 5.3493781089782715,
      "epoch": 3.8434163701067616,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003589948270432151,
      "loss": 4.8543,
      "mean_token_accuracy": 0.23386740535497666,
      "num_tokens": 89062039.0,
      "step": 38880
    },
    {
      "entropy": 5.3854574203491214,
      "epoch": 3.843910636615263,
      "grad_norm": 1.0,
      "learning_rate": 0.0003589620872634436,
      "loss": 4.8478,
      "mean_token_accuracy": 0.22997320294380189,
      "num_tokens": 89072621.0,
      "step": 38885
    },
    {
      "entropy": 5.368063354492188,
      "epoch": 3.844404903123764,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003589293454181916,
      "loss": 4.8192,
      "mean_token_accuracy": 0.230879545211792,
      "num_tokens": 89084019.0,
      "step": 38890
    },
    {
      "entropy": 5.337488603591919,
      "epoch": 3.844899169632266,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035889660150826473,
      "loss": 4.8772,
      "mean_token_accuracy": 0.2260674476623535,
      "num_tokens": 89094305.0,
      "step": 38895
    },
    {
      "entropy": 5.3130467414855955,
      "epoch": 3.845393436140767,
      "grad_norm": 1.03125,
      "learning_rate": 0.00035886385553446853,
      "loss": 4.8662,
      "mean_token_accuracy": 0.2209334909915924,
      "num_tokens": 89106261.0,
      "step": 38900
    },
    {
      "entropy": 5.2739804744720455,
      "epoch": 3.8458877026492684,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00035883110749760836,
      "loss": 4.8112,
      "mean_token_accuracy": 0.2366209387779236,
      "num_tokens": 89118171.0,
      "step": 38905
    },
    {
      "entropy": 5.4105119705200195,
      "epoch": 3.84638196915777,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00035879835739848994,
      "loss": 4.9508,
      "mean_token_accuracy": 0.2265930637717247,
      "num_tokens": 89129953.0,
      "step": 38910
    },
    {
      "entropy": 5.293031835556031,
      "epoch": 3.8468762356662713,
      "grad_norm": 1.03125,
      "learning_rate": 0.000358765605237919,
      "loss": 4.7458,
      "mean_token_accuracy": 0.24137990772724152,
      "num_tokens": 89143284.0,
      "step": 38915
    },
    {
      "entropy": 5.339274978637695,
      "epoch": 3.8473705021747726,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003587328510167012,
      "loss": 4.893,
      "mean_token_accuracy": 0.22630639970302582,
      "num_tokens": 89154675.0,
      "step": 38920
    },
    {
      "entropy": 5.39530382156372,
      "epoch": 3.847864768683274,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003587000947356421,
      "loss": 4.9267,
      "mean_token_accuracy": 0.21936150044202804,
      "num_tokens": 89164576.0,
      "step": 38925
    },
    {
      "entropy": 5.344863986968994,
      "epoch": 3.848359035191775,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00035866733639554777,
      "loss": 4.9307,
      "mean_token_accuracy": 0.21964293122291564,
      "num_tokens": 89176106.0,
      "step": 38930
    },
    {
      "entropy": 5.359806537628174,
      "epoch": 3.848853301700277,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000358634575997224,
      "loss": 4.8484,
      "mean_token_accuracy": 0.22418383061885833,
      "num_tokens": 89187214.0,
      "step": 38935
    },
    {
      "entropy": 5.265488386154175,
      "epoch": 3.849347568208778,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003586018135414767,
      "loss": 4.7239,
      "mean_token_accuracy": 0.2363439306616783,
      "num_tokens": 89198366.0,
      "step": 38940
    },
    {
      "entropy": 5.314633560180664,
      "epoch": 3.84984183471728,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003585690490291119,
      "loss": 4.9177,
      "mean_token_accuracy": 0.22793875485658646,
      "num_tokens": 89209122.0,
      "step": 38945
    },
    {
      "entropy": 5.413237905502319,
      "epoch": 3.850336101225781,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003585362824609354,
      "loss": 4.9217,
      "mean_token_accuracy": 0.22476533204317092,
      "num_tokens": 89221104.0,
      "step": 38950
    },
    {
      "entropy": 5.410201597213745,
      "epoch": 3.8508303677342823,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003585035138377534,
      "loss": 4.8859,
      "mean_token_accuracy": 0.23163299560546874,
      "num_tokens": 89231087.0,
      "step": 38955
    },
    {
      "entropy": 5.380070447921753,
      "epoch": 3.8513246342427836,
      "grad_norm": 1.078125,
      "learning_rate": 0.00035847074316037206,
      "loss": 4.9493,
      "mean_token_accuracy": 0.2215514987707138,
      "num_tokens": 89243198.0,
      "step": 38960
    },
    {
      "entropy": 5.312656688690185,
      "epoch": 3.851818900751285,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003584379704295975,
      "loss": 4.8308,
      "mean_token_accuracy": 0.2345993012189865,
      "num_tokens": 89254408.0,
      "step": 38965
    },
    {
      "entropy": 5.409298133850098,
      "epoch": 3.8523131672597866,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003584051956462359,
      "loss": 4.9138,
      "mean_token_accuracy": 0.22906606048345565,
      "num_tokens": 89266087.0,
      "step": 38970
    },
    {
      "entropy": 5.31666841506958,
      "epoch": 3.852807433768288,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003583724188110936,
      "loss": 4.8014,
      "mean_token_accuracy": 0.22748318910598755,
      "num_tokens": 89277039.0,
      "step": 38975
    },
    {
      "entropy": 5.400131320953369,
      "epoch": 3.853301700276789,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00035833963992497667,
      "loss": 4.9247,
      "mean_token_accuracy": 0.22615182548761367,
      "num_tokens": 89288940.0,
      "step": 38980
    },
    {
      "entropy": 5.359665632247925,
      "epoch": 3.853795966785291,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003583068589886919,
      "loss": 4.8278,
      "mean_token_accuracy": 0.22810934782028197,
      "num_tokens": 89301460.0,
      "step": 38985
    },
    {
      "entropy": 5.353357267379761,
      "epoch": 3.854290233293792,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00035827407600304533,
      "loss": 4.8963,
      "mean_token_accuracy": 0.2213057443499565,
      "num_tokens": 89312421.0,
      "step": 38990
    },
    {
      "entropy": 5.3401388168334964,
      "epoch": 3.8547844998022933,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003582412909688436,
      "loss": 4.8549,
      "mean_token_accuracy": 0.2279989540576935,
      "num_tokens": 89323552.0,
      "step": 38995
    },
    {
      "entropy": 5.400350618362427,
      "epoch": 3.8552787663107946,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003582085038868932,
      "loss": 4.8827,
      "mean_token_accuracy": 0.22004474997520446,
      "num_tokens": 89333911.0,
      "step": 39000
    },
    {
      "epoch": 3.8552787663107946,
      "eval_entropy": 5.180456919394443,
      "eval_loss": 4.952737808227539,
      "eval_mean_token_accuracy": 0.2317748072292532,
      "eval_num_tokens": 89333911.0,
      "eval_runtime": 26.5918,
      "eval_samples_per_second": 1222.668,
      "eval_steps_per_second": 152.866,
      "step": 39000
    }
  ],
  "logging_steps": 5,
  "max_steps": 101150,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 3000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.39875019491456e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}