jpn-jpan-10mb-100mb_seed10/checkpoint-24000/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.7697634160415463,
  "eval_steps": 3000,
  "global_step": 24000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 10.74217529296875,
      "epoch": 0.0005770340450086555,
      "grad_norm": 4.34375,
      "learning_rate": 2e-06,
      "loss": 10.8326,
      "mean_token_accuracy": 0.0,
      "num_tokens": 14330.0,
      "step": 5
    },
    {
      "entropy": 10.742151832580566,
      "epoch": 0.001154068090017311,
      "grad_norm": 4.65625,
      "learning_rate": 4.5e-06,
      "loss": 10.8057,
      "mean_token_accuracy": 8.11688310932368e-05,
      "num_tokens": 26284.0,
      "step": 10
    },
    {
      "entropy": 10.742152214050293,
      "epoch": 0.0017311021350259665,
      "grad_norm": 5.125,
      "learning_rate": 7e-06,
      "loss": 10.7851,
      "mean_token_accuracy": 0.00030720371869392694,
      "num_tokens": 38394.0,
      "step": 15
    },
    {
      "entropy": 10.742230319976807,
      "epoch": 0.002308136180034622,
      "grad_norm": 4.8125,
      "learning_rate": 9.5e-06,
      "loss": 10.7433,
      "mean_token_accuracy": 0.0,
      "num_tokens": 51341.0,
      "step": 20
    },
    {
      "entropy": 10.742213916778564,
      "epoch": 0.0028851702250432777,
      "grad_norm": 4.4375,
      "learning_rate": 1.2e-05,
      "loss": 10.64,
      "mean_token_accuracy": 0.003926298324950039,
      "num_tokens": 64559.0,
      "step": 25
    },
    {
      "entropy": 10.741898727416991,
      "epoch": 0.003462204270051933,
      "grad_norm": 3.921875,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 10.5216,
      "mean_token_accuracy": 0.03795875590294599,
      "num_tokens": 76688.0,
      "step": 30
    },
    {
      "entropy": 10.740620136260986,
      "epoch": 0.004039238315060588,
      "grad_norm": 3.46875,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 10.3693,
      "mean_token_accuracy": 0.05394260324537754,
      "num_tokens": 88392.0,
      "step": 35
    },
    {
      "entropy": 10.736398124694825,
      "epoch": 0.004616272360069244,
      "grad_norm": 2.578125,
      "learning_rate": 1.95e-05,
      "loss": 10.278,
      "mean_token_accuracy": 0.052969465777277944,
      "num_tokens": 100542.0,
      "step": 40
    },
    {
      "entropy": 10.728193473815917,
      "epoch": 0.0051933064050779,
      "grad_norm": 2.34375,
      "learning_rate": 2.2e-05,
      "loss": 10.0828,
      "mean_token_accuracy": 0.05751506164669991,
      "num_tokens": 113190.0,
      "step": 45
    },
    {
      "entropy": 10.716622734069825,
      "epoch": 0.005770340450086555,
      "grad_norm": 2.109375,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 10.0239,
      "mean_token_accuracy": 0.05997682772576809,
      "num_tokens": 125554.0,
      "step": 50
    },
    {
      "entropy": 10.704752922058105,
      "epoch": 0.006347374495095211,
      "grad_norm": 1.9765625,
      "learning_rate": 2.7e-05,
      "loss": 9.9661,
      "mean_token_accuracy": 0.06483750529587269,
      "num_tokens": 138527.0,
      "step": 55
    },
    {
      "entropy": 10.696050262451172,
      "epoch": 0.006924408540103866,
      "grad_norm": 1.9375,
      "learning_rate": 2.95e-05,
      "loss": 10.0065,
      "mean_token_accuracy": 0.0495222195982933,
      "num_tokens": 150404.0,
      "step": 60
    },
    {
      "entropy": 10.69492769241333,
      "epoch": 0.0075014425851125215,
      "grad_norm": 1.859375,
      "learning_rate": 3.2e-05,
      "loss": 9.9359,
      "mean_token_accuracy": 0.055068480968475345,
      "num_tokens": 163743.0,
      "step": 65
    },
    {
      "entropy": 10.692989349365234,
      "epoch": 0.008078476630121177,
      "grad_norm": 1.8671875,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 9.9027,
      "mean_token_accuracy": 0.052650950476527215,
      "num_tokens": 175657.0,
      "step": 70
    },
    {
      "entropy": 10.689637851715087,
      "epoch": 0.008655510675129832,
      "grad_norm": 1.828125,
      "learning_rate": 3.7e-05,
      "loss": 9.8572,
      "mean_token_accuracy": 0.053944576904177664,
      "num_tokens": 188756.0,
      "step": 75
    },
    {
      "entropy": 10.678643798828125,
      "epoch": 0.009232544720138488,
      "grad_norm": 1.8984375,
      "learning_rate": 3.95e-05,
      "loss": 9.801,
      "mean_token_accuracy": 0.053839181736111644,
      "num_tokens": 200548.0,
      "step": 80
    },
    {
      "entropy": 10.661964607238769,
      "epoch": 0.009809578765147143,
      "grad_norm": 1.859375,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 9.7136,
      "mean_token_accuracy": 0.06665473207831382,
      "num_tokens": 213307.0,
      "step": 85
    },
    {
      "entropy": 10.63293104171753,
      "epoch": 0.0103866128101558,
      "grad_norm": 1.7890625,
      "learning_rate": 4.45e-05,
      "loss": 9.6765,
      "mean_token_accuracy": 0.06476713679730892,
      "num_tokens": 226863.0,
      "step": 90
    },
    {
      "entropy": 10.615675067901611,
      "epoch": 0.010963646855164455,
      "grad_norm": 1.6015625,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 9.632,
      "mean_token_accuracy": 0.06121704056859016,
      "num_tokens": 238655.0,
      "step": 95
    },
    {
      "entropy": 10.613676834106446,
      "epoch": 0.01154068090017311,
      "grad_norm": 1.734375,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 9.5655,
      "mean_token_accuracy": 0.0704349122941494,
      "num_tokens": 251118.0,
      "step": 100
    },
    {
      "entropy": 10.56223793029785,
      "epoch": 0.012117714945181766,
      "grad_norm": 1.953125,
      "learning_rate": 5.2e-05,
      "loss": 9.4708,
      "mean_token_accuracy": 0.07793616428971291,
      "num_tokens": 263665.0,
      "step": 105
    },
    {
      "entropy": 10.50258274078369,
      "epoch": 0.012694748990190421,
      "grad_norm": 1.6796875,
      "learning_rate": 5.45e-05,
      "loss": 9.4156,
      "mean_token_accuracy": 0.07520635947585105,
      "num_tokens": 276531.0,
      "step": 110
    },
    {
      "entropy": 10.506224632263184,
      "epoch": 0.013271783035199077,
      "grad_norm": 1.5625,
      "learning_rate": 5.7e-05,
      "loss": 9.367,
      "mean_token_accuracy": 0.0757908646017313,
      "num_tokens": 289878.0,
      "step": 115
    },
    {
      "entropy": 10.472508811950684,
      "epoch": 0.013848817080207732,
      "grad_norm": 1.5546875,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 9.2817,
      "mean_token_accuracy": 0.07052403129637241,
      "num_tokens": 302705.0,
      "step": 120
    },
    {
      "entropy": 10.418921184539794,
      "epoch": 0.014425851125216388,
      "grad_norm": 1.59375,
      "learning_rate": 6.2e-05,
      "loss": 9.1572,
      "mean_token_accuracy": 0.08047313243150711,
      "num_tokens": 314775.0,
      "step": 125
    },
    {
      "entropy": 10.330512046813965,
      "epoch": 0.015002885170225043,
      "grad_norm": 1.2890625,
      "learning_rate": 6.450000000000001e-05,
      "loss": 9.1596,
      "mean_token_accuracy": 0.07958029508590699,
      "num_tokens": 328563.0,
      "step": 130
    },
    {
      "entropy": 10.242788600921632,
      "epoch": 0.015579919215233698,
      "grad_norm": 1.3984375,
      "learning_rate": 6.7e-05,
      "loss": 9.0052,
      "mean_token_accuracy": 0.08549408987164497,
      "num_tokens": 341820.0,
      "step": 135
    },
    {
      "entropy": 10.257353591918946,
      "epoch": 0.016156953260242354,
      "grad_norm": 1.34375,
      "learning_rate": 6.950000000000001e-05,
      "loss": 8.9633,
      "mean_token_accuracy": 0.08436901345849038,
      "num_tokens": 354994.0,
      "step": 140
    },
    {
      "entropy": 10.1291428565979,
      "epoch": 0.01673398730525101,
      "grad_norm": 1.203125,
      "learning_rate": 7.2e-05,
      "loss": 8.8768,
      "mean_token_accuracy": 0.08217294588685035,
      "num_tokens": 365852.0,
      "step": 145
    },
    {
      "entropy": 10.04762372970581,
      "epoch": 0.017311021350259664,
      "grad_norm": 1.4609375,
      "learning_rate": 7.45e-05,
      "loss": 8.8657,
      "mean_token_accuracy": 0.07920254692435265,
      "num_tokens": 379147.0,
      "step": 150
    },
    {
      "entropy": 10.023890781402589,
      "epoch": 0.01788805539526832,
      "grad_norm": 1.3359375,
      "learning_rate": 7.7e-05,
      "loss": 8.8181,
      "mean_token_accuracy": 0.0735938459634781,
      "num_tokens": 392257.0,
      "step": 155
    },
    {
      "entropy": 9.896601295471191,
      "epoch": 0.018465089440276975,
      "grad_norm": 1.25,
      "learning_rate": 7.950000000000001e-05,
      "loss": 8.7072,
      "mean_token_accuracy": 0.07832617908716202,
      "num_tokens": 405249.0,
      "step": 160
    },
    {
      "entropy": 9.709205722808838,
      "epoch": 0.019042123485285632,
      "grad_norm": 0.9765625,
      "learning_rate": 8.2e-05,
      "loss": 8.6516,
      "mean_token_accuracy": 0.08040346056222916,
      "num_tokens": 417819.0,
      "step": 165
    },
    {
      "entropy": 9.60182809829712,
      "epoch": 0.019619157530294286,
      "grad_norm": 0.98828125,
      "learning_rate": 8.450000000000001e-05,
      "loss": 8.5351,
      "mean_token_accuracy": 0.08392796665430069,
      "num_tokens": 429364.0,
      "step": 170
    },
    {
      "entropy": 9.431674385070801,
      "epoch": 0.020196191575302943,
      "grad_norm": 1.140625,
      "learning_rate": 8.7e-05,
      "loss": 8.5151,
      "mean_token_accuracy": 0.08911501616239548,
      "num_tokens": 441804.0,
      "step": 175
    },
    {
      "entropy": 9.291469860076905,
      "epoch": 0.0207732256203116,
      "grad_norm": 1.140625,
      "learning_rate": 8.95e-05,
      "loss": 8.4571,
      "mean_token_accuracy": 0.08912580311298371,
      "num_tokens": 453886.0,
      "step": 180
    },
    {
      "entropy": 9.148678588867188,
      "epoch": 0.021350259665320254,
      "grad_norm": 0.89453125,
      "learning_rate": 9.2e-05,
      "loss": 8.4362,
      "mean_token_accuracy": 0.08946132510900498,
      "num_tokens": 465828.0,
      "step": 185
    },
    {
      "entropy": 9.080477714538574,
      "epoch": 0.02192729371032891,
      "grad_norm": 0.796875,
      "learning_rate": 9.45e-05,
      "loss": 8.3788,
      "mean_token_accuracy": 0.09273218810558319,
      "num_tokens": 478783.0,
      "step": 190
    },
    {
      "entropy": 8.97773780822754,
      "epoch": 0.022504327755337564,
      "grad_norm": 1.0625,
      "learning_rate": 9.7e-05,
      "loss": 8.4168,
      "mean_token_accuracy": 0.08846393302083015,
      "num_tokens": 491167.0,
      "step": 195
    },
    {
      "entropy": 8.89993076324463,
      "epoch": 0.02308136180034622,
      "grad_norm": 1.3359375,
      "learning_rate": 9.95e-05,
      "loss": 8.3844,
      "mean_token_accuracy": 0.09320372641086579,
      "num_tokens": 503511.0,
      "step": 200
    },
    {
      "entropy": 8.787849521636963,
      "epoch": 0.023658395845354875,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000102,
      "loss": 8.2705,
      "mean_token_accuracy": 0.0978940948843956,
      "num_tokens": 515816.0,
      "step": 205
    },
    {
      "entropy": 8.848995685577393,
      "epoch": 0.024235429890363532,
      "grad_norm": 0.921875,
      "learning_rate": 0.00010449999999999999,
      "loss": 8.3781,
      "mean_token_accuracy": 0.08895341083407401,
      "num_tokens": 528919.0,
      "step": 210
    },
    {
      "entropy": 8.744617557525634,
      "epoch": 0.024812463935372186,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000107,
      "loss": 8.2875,
      "mean_token_accuracy": 0.10044435858726501,
      "num_tokens": 540973.0,
      "step": 215
    },
    {
      "entropy": 8.707606983184814,
      "epoch": 0.025389497980380843,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0001095,
      "loss": 8.2458,
      "mean_token_accuracy": 0.10567026063799859,
      "num_tokens": 553265.0,
      "step": 220
    },
    {
      "entropy": 8.753429126739501,
      "epoch": 0.025966532025389497,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000112,
      "loss": 8.3474,
      "mean_token_accuracy": 0.0927657037973404,
      "num_tokens": 567534.0,
      "step": 225
    },
    {
      "entropy": 8.750615215301513,
      "epoch": 0.026543566070398154,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0001145,
      "loss": 8.2643,
      "mean_token_accuracy": 0.09682516902685165,
      "num_tokens": 579297.0,
      "step": 230
    },
    {
      "entropy": 8.695602512359619,
      "epoch": 0.027120600115406807,
      "grad_norm": 0.875,
      "learning_rate": 0.00011700000000000001,
      "loss": 8.2384,
      "mean_token_accuracy": 0.09817592948675155,
      "num_tokens": 592554.0,
      "step": 235
    },
    {
      "entropy": 8.732758808135987,
      "epoch": 0.027697634160415464,
      "grad_norm": 0.921875,
      "learning_rate": 0.00011949999999999999,
      "loss": 8.2291,
      "mean_token_accuracy": 0.09817634895443916,
      "num_tokens": 605038.0,
      "step": 240
    },
    {
      "entropy": 8.617175579071045,
      "epoch": 0.02827466820542412,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000122,
      "loss": 8.2451,
      "mean_token_accuracy": 0.09733725637197495,
      "num_tokens": 617172.0,
      "step": 245
    },
    {
      "entropy": 8.57886381149292,
      "epoch": 0.028851702250432775,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0001245,
      "loss": 8.1868,
      "mean_token_accuracy": 0.09926503598690033,
      "num_tokens": 629243.0,
      "step": 250
    },
    {
      "entropy": 8.597729873657226,
      "epoch": 0.029428736295441432,
      "grad_norm": 0.921875,
      "learning_rate": 0.000127,
      "loss": 8.2453,
      "mean_token_accuracy": 0.09898478463292122,
      "num_tokens": 641487.0,
      "step": 255
    },
    {
      "entropy": 8.657223415374755,
      "epoch": 0.030005770340450086,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0001295,
      "loss": 8.2382,
      "mean_token_accuracy": 0.08884634375572205,
      "num_tokens": 653214.0,
      "step": 260
    },
    {
      "entropy": 8.59217348098755,
      "epoch": 0.030582804385458743,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000132,
      "loss": 8.1923,
      "mean_token_accuracy": 0.11060753613710403,
      "num_tokens": 666549.0,
      "step": 265
    },
    {
      "entropy": 8.513946914672852,
      "epoch": 0.031159838430467397,
      "grad_norm": 0.828125,
      "learning_rate": 0.00013450000000000002,
      "loss": 8.2137,
      "mean_token_accuracy": 0.10257397592067719,
      "num_tokens": 679443.0,
      "step": 270
    },
    {
      "entropy": 8.499358367919921,
      "epoch": 0.031736872475476054,
      "grad_norm": 1.015625,
      "learning_rate": 0.00013700000000000002,
      "loss": 8.2136,
      "mean_token_accuracy": 0.09787259846925736,
      "num_tokens": 692929.0,
      "step": 275
    },
    {
      "entropy": 8.5993070602417,
      "epoch": 0.03231390652048471,
      "grad_norm": 1.078125,
      "learning_rate": 0.0001395,
      "loss": 8.133,
      "mean_token_accuracy": 0.1011851780116558,
      "num_tokens": 705605.0,
      "step": 280
    },
    {
      "entropy": 8.563784217834472,
      "epoch": 0.03289094056549336,
      "grad_norm": 0.7109375,
      "learning_rate": 0.00014199999999999998,
      "loss": 8.2181,
      "mean_token_accuracy": 0.09577601477503776,
      "num_tokens": 719064.0,
      "step": 285
    },
    {
      "entropy": 8.528156089782716,
      "epoch": 0.03346797461050202,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0001445,
      "loss": 8.148,
      "mean_token_accuracy": 0.10154027119278908,
      "num_tokens": 732572.0,
      "step": 290
    },
    {
      "entropy": 8.501526927947998,
      "epoch": 0.034045008655510675,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000147,
      "loss": 8.1452,
      "mean_token_accuracy": 0.09947796389460564,
      "num_tokens": 745161.0,
      "step": 295
    },
    {
      "entropy": 8.492893695831299,
      "epoch": 0.03462204270051933,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0001495,
      "loss": 8.1182,
      "mean_token_accuracy": 0.1088537909090519,
      "num_tokens": 756180.0,
      "step": 300
    },
    {
      "entropy": 8.508162879943848,
      "epoch": 0.03519907674552799,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000152,
      "loss": 8.142,
      "mean_token_accuracy": 0.10471795424818993,
      "num_tokens": 768568.0,
      "step": 305
    },
    {
      "entropy": 8.514046001434327,
      "epoch": 0.03577611079053664,
      "grad_norm": 1.078125,
      "learning_rate": 0.00015450000000000001,
      "loss": 8.1756,
      "mean_token_accuracy": 0.10089959502220154,
      "num_tokens": 781580.0,
      "step": 310
    },
    {
      "entropy": 8.390854835510254,
      "epoch": 0.0363531448355453,
      "grad_norm": 0.9375,
      "learning_rate": 0.000157,
      "loss": 8.0541,
      "mean_token_accuracy": 0.11660909354686737,
      "num_tokens": 793486.0,
      "step": 315
    },
    {
      "entropy": 8.494402980804443,
      "epoch": 0.03693017888055395,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0001595,
      "loss": 8.1249,
      "mean_token_accuracy": 0.10236181169748307,
      "num_tokens": 805997.0,
      "step": 320
    },
    {
      "entropy": 8.422712516784667,
      "epoch": 0.03750721292556261,
      "grad_norm": 1.125,
      "learning_rate": 0.000162,
      "loss": 8.0829,
      "mean_token_accuracy": 0.10648757740855216,
      "num_tokens": 817681.0,
      "step": 325
    },
    {
      "entropy": 8.429091262817384,
      "epoch": 0.038084246970571264,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00016450000000000001,
      "loss": 8.0916,
      "mean_token_accuracy": 0.11145856082439423,
      "num_tokens": 830027.0,
      "step": 330
    },
    {
      "entropy": 8.464876174926758,
      "epoch": 0.03866128101557992,
      "grad_norm": 0.78125,
      "learning_rate": 0.00016700000000000002,
      "loss": 8.1081,
      "mean_token_accuracy": 0.10330765470862388,
      "num_tokens": 842862.0,
      "step": 335
    },
    {
      "entropy": 8.401940250396729,
      "epoch": 0.03923831506058857,
      "grad_norm": 1.015625,
      "learning_rate": 0.00016950000000000003,
      "loss": 8.0324,
      "mean_token_accuracy": 0.10951320230960845,
      "num_tokens": 856548.0,
      "step": 340
    },
    {
      "entropy": 8.391893863677979,
      "epoch": 0.03981534910559723,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00017199999999999998,
      "loss": 8.0417,
      "mean_token_accuracy": 0.10960106775164605,
      "num_tokens": 869651.0,
      "step": 345
    },
    {
      "entropy": 8.451942253112794,
      "epoch": 0.040392383150605886,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00017449999999999999,
      "loss": 8.0689,
      "mean_token_accuracy": 0.1082707829773426,
      "num_tokens": 880717.0,
      "step": 350
    },
    {
      "entropy": 8.369545364379883,
      "epoch": 0.04096941719561454,
      "grad_norm": 0.921875,
      "learning_rate": 0.000177,
      "loss": 8.0521,
      "mean_token_accuracy": 0.11259332075715064,
      "num_tokens": 894066.0,
      "step": 355
    },
    {
      "entropy": 8.282257270812988,
      "epoch": 0.0415464512406232,
      "grad_norm": 1.046875,
      "learning_rate": 0.0001795,
      "loss": 7.9567,
      "mean_token_accuracy": 0.11248702630400657,
      "num_tokens": 905287.0,
      "step": 360
    },
    {
      "entropy": 8.459172248840332,
      "epoch": 0.042123485285631854,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000182,
      "loss": 8.0392,
      "mean_token_accuracy": 0.11402102708816528,
      "num_tokens": 917512.0,
      "step": 365
    },
    {
      "entropy": 8.318458461761475,
      "epoch": 0.04270051933064051,
      "grad_norm": 1.265625,
      "learning_rate": 0.0001845,
      "loss": 7.9899,
      "mean_token_accuracy": 0.10831499472260475,
      "num_tokens": 930263.0,
      "step": 370
    },
    {
      "entropy": 8.339896297454834,
      "epoch": 0.04327755337564916,
      "grad_norm": 0.90625,
      "learning_rate": 0.000187,
      "loss": 8.0425,
      "mean_token_accuracy": 0.11203662231564522,
      "num_tokens": 942330.0,
      "step": 375
    },
    {
      "entropy": 8.366154861450195,
      "epoch": 0.04385458742065782,
      "grad_norm": 1.03125,
      "learning_rate": 0.0001895,
      "loss": 8.0316,
      "mean_token_accuracy": 0.10789357051253319,
      "num_tokens": 954005.0,
      "step": 380
    },
    {
      "entropy": 8.355004215240479,
      "epoch": 0.044431621465666475,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000192,
      "loss": 8.026,
      "mean_token_accuracy": 0.1113132230937481,
      "num_tokens": 966725.0,
      "step": 385
    },
    {
      "entropy": 8.353032684326172,
      "epoch": 0.04500865551067513,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0001945,
      "loss": 8.0044,
      "mean_token_accuracy": 0.1139283761382103,
      "num_tokens": 978021.0,
      "step": 390
    },
    {
      "entropy": 8.344230270385742,
      "epoch": 0.04558568955568378,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00019700000000000002,
      "loss": 7.9905,
      "mean_token_accuracy": 0.11658623144030571,
      "num_tokens": 991321.0,
      "step": 395
    },
    {
      "entropy": 8.339315605163574,
      "epoch": 0.04616272360069244,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00019950000000000002,
      "loss": 7.9818,
      "mean_token_accuracy": 0.11692097708582878,
      "num_tokens": 1004172.0,
      "step": 400
    },
    {
      "entropy": 8.253484535217286,
      "epoch": 0.0467397576457011,
      "grad_norm": 1.015625,
      "learning_rate": 0.000202,
      "loss": 7.936,
      "mean_token_accuracy": 0.11915154382586479,
      "num_tokens": 1016513.0,
      "step": 405
    },
    {
      "entropy": 8.23137845993042,
      "epoch": 0.04731679169070975,
      "grad_norm": 1.265625,
      "learning_rate": 0.00020449999999999998,
      "loss": 7.9323,
      "mean_token_accuracy": 0.11277970448136329,
      "num_tokens": 1028536.0,
      "step": 410
    },
    {
      "entropy": 8.298001146316528,
      "epoch": 0.04789382573571841,
      "grad_norm": 1.421875,
      "learning_rate": 0.000207,
      "loss": 7.8757,
      "mean_token_accuracy": 0.12688373401761055,
      "num_tokens": 1040534.0,
      "step": 415
    },
    {
      "entropy": 8.242660903930664,
      "epoch": 0.048470859780727064,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002095,
      "loss": 7.9649,
      "mean_token_accuracy": 0.11573690325021743,
      "num_tokens": 1054995.0,
      "step": 420
    },
    {
      "entropy": 8.27421875,
      "epoch": 0.04904789382573572,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000212,
      "loss": 7.9873,
      "mean_token_accuracy": 0.11399574726819992,
      "num_tokens": 1067722.0,
      "step": 425
    },
    {
      "entropy": 8.373875331878661,
      "epoch": 0.04962492787074437,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002145,
      "loss": 7.9637,
      "mean_token_accuracy": 0.10732617229223251,
      "num_tokens": 1079788.0,
      "step": 430
    },
    {
      "entropy": 8.326363945007325,
      "epoch": 0.05020196191575303,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00021700000000000002,
      "loss": 7.9539,
      "mean_token_accuracy": 0.1163248635828495,
      "num_tokens": 1092016.0,
      "step": 435
    },
    {
      "entropy": 8.17879877090454,
      "epoch": 0.050778995960761686,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002195,
      "loss": 7.9654,
      "mean_token_accuracy": 0.11178191602230073,
      "num_tokens": 1104522.0,
      "step": 440
    },
    {
      "entropy": 8.363206481933593,
      "epoch": 0.05135603000577034,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000222,
      "loss": 7.979,
      "mean_token_accuracy": 0.12337485775351524,
      "num_tokens": 1117239.0,
      "step": 445
    },
    {
      "entropy": 8.268813133239746,
      "epoch": 0.05193306405077899,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002245,
      "loss": 7.9035,
      "mean_token_accuracy": 0.11866443902254105,
      "num_tokens": 1129061.0,
      "step": 450
    },
    {
      "entropy": 8.266489696502685,
      "epoch": 0.052510098095787654,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00022700000000000002,
      "loss": 7.8976,
      "mean_token_accuracy": 0.11053306236863136,
      "num_tokens": 1141409.0,
      "step": 455
    },
    {
      "entropy": 8.20300588607788,
      "epoch": 0.05308713214079631,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00022950000000000002,
      "loss": 7.9036,
      "mean_token_accuracy": 0.11711592748761177,
      "num_tokens": 1153623.0,
      "step": 460
    },
    {
      "entropy": 8.255742359161378,
      "epoch": 0.05366416618580496,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00023200000000000003,
      "loss": 7.9518,
      "mean_token_accuracy": 0.11555556952953339,
      "num_tokens": 1166877.0,
      "step": 465
    },
    {
      "entropy": 8.26829776763916,
      "epoch": 0.054241200230813615,
      "grad_norm": 0.7734375,
      "learning_rate": 0.00023449999999999998,
      "loss": 7.9228,
      "mean_token_accuracy": 0.11642130017280579,
      "num_tokens": 1179625.0,
      "step": 470
    },
    {
      "entropy": 8.19015064239502,
      "epoch": 0.054818234275822275,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000237,
      "loss": 7.9401,
      "mean_token_accuracy": 0.11644659489393235,
      "num_tokens": 1193586.0,
      "step": 475
    },
    {
      "entropy": 8.207570362091065,
      "epoch": 0.05539526832083093,
      "grad_norm": 0.90625,
      "learning_rate": 0.0002395,
      "loss": 7.8841,
      "mean_token_accuracy": 0.12152690812945366,
      "num_tokens": 1207086.0,
      "step": 480
    },
    {
      "entropy": 8.263338947296143,
      "epoch": 0.05597230236583958,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000242,
      "loss": 7.8895,
      "mean_token_accuracy": 0.11943561136722565,
      "num_tokens": 1219389.0,
      "step": 485
    },
    {
      "entropy": 8.21270751953125,
      "epoch": 0.05654933641084824,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002445,
      "loss": 7.825,
      "mean_token_accuracy": 0.12387257441878319,
      "num_tokens": 1232080.0,
      "step": 490
    },
    {
      "entropy": 8.094493389129639,
      "epoch": 0.0571263704558569,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000247,
      "loss": 7.874,
      "mean_token_accuracy": 0.12492263168096543,
      "num_tokens": 1246775.0,
      "step": 495
    },
    {
      "entropy": 8.204738426208497,
      "epoch": 0.05770340450086555,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0002495,
      "loss": 7.9068,
      "mean_token_accuracy": 0.114556435495615,
      "num_tokens": 1258839.0,
      "step": 500
    },
    {
      "entropy": 8.169677925109863,
      "epoch": 0.058280438545874204,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000252,
      "loss": 7.8279,
      "mean_token_accuracy": 0.12553127557039262,
      "num_tokens": 1270710.0,
      "step": 505
    },
    {
      "entropy": 8.218509674072266,
      "epoch": 0.058857472590882864,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002545,
      "loss": 7.786,
      "mean_token_accuracy": 0.12104848474264145,
      "num_tokens": 1281930.0,
      "step": 510
    },
    {
      "entropy": 8.104724454879761,
      "epoch": 0.05943450663589152,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000257,
      "loss": 7.8534,
      "mean_token_accuracy": 0.11715041622519493,
      "num_tokens": 1294548.0,
      "step": 515
    },
    {
      "entropy": 8.166370916366578,
      "epoch": 0.06001154068090017,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002595,
      "loss": 7.8064,
      "mean_token_accuracy": 0.11803867742419243,
      "num_tokens": 1307588.0,
      "step": 520
    },
    {
      "entropy": 8.132765865325927,
      "epoch": 0.060588574725908825,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000262,
      "loss": 7.7863,
      "mean_token_accuracy": 0.12690332680940627,
      "num_tokens": 1319478.0,
      "step": 525
    },
    {
      "entropy": 8.216028451919556,
      "epoch": 0.061165608770917486,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00026450000000000003,
      "loss": 7.8708,
      "mean_token_accuracy": 0.11611150279641151,
      "num_tokens": 1332950.0,
      "step": 530
    },
    {
      "entropy": 8.090120029449462,
      "epoch": 0.06174264281592614,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00026700000000000004,
      "loss": 7.7615,
      "mean_token_accuracy": 0.12739897295832633,
      "num_tokens": 1345510.0,
      "step": 535
    },
    {
      "entropy": 8.01339612007141,
      "epoch": 0.06231967686093479,
      "grad_norm": 1.1875,
      "learning_rate": 0.00026950000000000005,
      "loss": 7.8042,
      "mean_token_accuracy": 0.12997828647494317,
      "num_tokens": 1359388.0,
      "step": 540
    },
    {
      "entropy": 8.145500087738037,
      "epoch": 0.06289671090594345,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00027200000000000005,
      "loss": 7.8149,
      "mean_token_accuracy": 0.12175369709730148,
      "num_tokens": 1373067.0,
      "step": 545
    },
    {
      "entropy": 8.04567756652832,
      "epoch": 0.06347374495095211,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0002745,
      "loss": 7.7835,
      "mean_token_accuracy": 0.12414814531803131,
      "num_tokens": 1385718.0,
      "step": 550
    },
    {
      "entropy": 8.235342979431152,
      "epoch": 0.06405077899596076,
      "grad_norm": 1.171875,
      "learning_rate": 0.000277,
      "loss": 7.7898,
      "mean_token_accuracy": 0.12540573999285698,
      "num_tokens": 1398788.0,
      "step": 555
    },
    {
      "entropy": 8.071519756317139,
      "epoch": 0.06462781304096941,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002795,
      "loss": 7.7882,
      "mean_token_accuracy": 0.12845603749155998,
      "num_tokens": 1411752.0,
      "step": 560
    },
    {
      "entropy": 8.075310802459716,
      "epoch": 0.06520484708597807,
      "grad_norm": 1.015625,
      "learning_rate": 0.00028199999999999997,
      "loss": 7.8443,
      "mean_token_accuracy": 0.12352012917399406,
      "num_tokens": 1423979.0,
      "step": 565
    },
    {
      "entropy": 8.11637830734253,
      "epoch": 0.06578188113098672,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0002845,
      "loss": 7.8721,
      "mean_token_accuracy": 0.11852672100067138,
      "num_tokens": 1437009.0,
      "step": 570
    },
    {
      "entropy": 8.124029541015625,
      "epoch": 0.06635891517599539,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000287,
      "loss": 7.7174,
      "mean_token_accuracy": 0.12357776835560799,
      "num_tokens": 1448220.0,
      "step": 575
    },
    {
      "entropy": 8.00622100830078,
      "epoch": 0.06693594922100404,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002895,
      "loss": 7.7566,
      "mean_token_accuracy": 0.12290028035640717,
      "num_tokens": 1460024.0,
      "step": 580
    },
    {
      "entropy": 8.137111854553222,
      "epoch": 0.0675129832660127,
      "grad_norm": 0.984375,
      "learning_rate": 0.000292,
      "loss": 7.7451,
      "mean_token_accuracy": 0.1224621057510376,
      "num_tokens": 1472283.0,
      "step": 585
    },
    {
      "entropy": 8.003714799880981,
      "epoch": 0.06809001731102135,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0002945,
      "loss": 7.7908,
      "mean_token_accuracy": 0.12675876170396805,
      "num_tokens": 1486565.0,
      "step": 590
    },
    {
      "entropy": 7.9941528797149655,
      "epoch": 0.06866705135603,
      "grad_norm": 1.03125,
      "learning_rate": 0.000297,
      "loss": 7.7704,
      "mean_token_accuracy": 0.1255685530602932,
      "num_tokens": 1498708.0,
      "step": 595
    },
    {
      "entropy": 8.076406526565552,
      "epoch": 0.06924408540103866,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002995,
      "loss": 7.7121,
      "mean_token_accuracy": 0.1268436163663864,
      "num_tokens": 1511318.0,
      "step": 600
    },
    {
      "entropy": 7.977707815170288,
      "epoch": 0.06982111944604731,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000302,
      "loss": 7.695,
      "mean_token_accuracy": 0.13601174131035804,
      "num_tokens": 1525415.0,
      "step": 605
    },
    {
      "entropy": 8.098061847686768,
      "epoch": 0.07039815349105598,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003045,
      "loss": 7.7433,
      "mean_token_accuracy": 0.12430487498641014,
      "num_tokens": 1538757.0,
      "step": 610
    },
    {
      "entropy": 7.966400146484375,
      "epoch": 0.07097518753606463,
      "grad_norm": 1.140625,
      "learning_rate": 0.000307,
      "loss": 7.6977,
      "mean_token_accuracy": 0.12538538724184037,
      "num_tokens": 1550619.0,
      "step": 615
    },
    {
      "entropy": 8.005772590637207,
      "epoch": 0.07155222158107329,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003095,
      "loss": 7.7496,
      "mean_token_accuracy": 0.12077321037650109,
      "num_tokens": 1562588.0,
      "step": 620
    },
    {
      "entropy": 7.989212322235107,
      "epoch": 0.07212925562608194,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000312,
      "loss": 7.6787,
      "mean_token_accuracy": 0.12967532649636268,
      "num_tokens": 1574892.0,
      "step": 625
    },
    {
      "entropy": 7.974766254425049,
      "epoch": 0.0727062896710906,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003145,
      "loss": 7.7117,
      "mean_token_accuracy": 0.12752822190523147,
      "num_tokens": 1587525.0,
      "step": 630
    },
    {
      "entropy": 8.038787698745727,
      "epoch": 0.07328332371609925,
      "grad_norm": 1.265625,
      "learning_rate": 0.000317,
      "loss": 7.6921,
      "mean_token_accuracy": 0.12393600195646286,
      "num_tokens": 1600350.0,
      "step": 635
    },
    {
      "entropy": 7.9325361251831055,
      "epoch": 0.0738603577611079,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003195,
      "loss": 7.6823,
      "mean_token_accuracy": 0.12833620086312295,
      "num_tokens": 1613318.0,
      "step": 640
    },
    {
      "entropy": 8.086976289749146,
      "epoch": 0.07443739180611655,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000322,
      "loss": 7.8105,
      "mean_token_accuracy": 0.12034560292959214,
      "num_tokens": 1626757.0,
      "step": 645
    },
    {
      "entropy": 7.982501125335693,
      "epoch": 0.07501442585112522,
      "grad_norm": 0.984375,
      "learning_rate": 0.00032450000000000003,
      "loss": 7.6224,
      "mean_token_accuracy": 0.13317669555544853,
      "num_tokens": 1639185.0,
      "step": 650
    },
    {
      "entropy": 7.921093416213989,
      "epoch": 0.07559145989613388,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00032700000000000003,
      "loss": 7.727,
      "mean_token_accuracy": 0.12360085397958756,
      "num_tokens": 1652009.0,
      "step": 655
    },
    {
      "entropy": 7.985281324386596,
      "epoch": 0.07616849394114253,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00032950000000000004,
      "loss": 7.7099,
      "mean_token_accuracy": 0.12278636991977691,
      "num_tokens": 1665604.0,
      "step": 660
    },
    {
      "entropy": 7.997657918930054,
      "epoch": 0.07674552798615118,
      "grad_norm": 0.953125,
      "learning_rate": 0.00033200000000000005,
      "loss": 7.6893,
      "mean_token_accuracy": 0.12272793427109718,
      "num_tokens": 1679036.0,
      "step": 665
    },
    {
      "entropy": 7.8749652862548825,
      "epoch": 0.07732256203115984,
      "grad_norm": 1.0625,
      "learning_rate": 0.00033450000000000005,
      "loss": 7.664,
      "mean_token_accuracy": 0.13161094784736632,
      "num_tokens": 1691512.0,
      "step": 670
    },
    {
      "entropy": 8.132879447937011,
      "epoch": 0.07789959607616849,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000337,
      "loss": 7.7292,
      "mean_token_accuracy": 0.12437204793095588,
      "num_tokens": 1704589.0,
      "step": 675
    },
    {
      "entropy": 7.819326114654541,
      "epoch": 0.07847663012117714,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003395,
      "loss": 7.7067,
      "mean_token_accuracy": 0.12104201540350915,
      "num_tokens": 1718612.0,
      "step": 680
    },
    {
      "entropy": 7.861719846725464,
      "epoch": 0.07905366416618581,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000342,
      "loss": 7.6127,
      "mean_token_accuracy": 0.12460441142320633,
      "num_tokens": 1730472.0,
      "step": 685
    },
    {
      "entropy": 8.00782494544983,
      "epoch": 0.07963069821119446,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00034449999999999997,
      "loss": 7.7467,
      "mean_token_accuracy": 0.12077393978834153,
      "num_tokens": 1742535.0,
      "step": 690
    },
    {
      "entropy": 8.006076240539551,
      "epoch": 0.08020773225620312,
      "grad_norm": 1.03125,
      "learning_rate": 0.000347,
      "loss": 7.7141,
      "mean_token_accuracy": 0.12221984565258026,
      "num_tokens": 1753877.0,
      "step": 695
    },
    {
      "entropy": 7.926956653594971,
      "epoch": 0.08078476630121177,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0003495,
      "loss": 7.6293,
      "mean_token_accuracy": 0.12848915457725524,
      "num_tokens": 1766413.0,
      "step": 700
    },
    {
      "entropy": 7.813401508331299,
      "epoch": 0.08136180034622043,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000352,
      "loss": 7.4754,
      "mean_token_accuracy": 0.13737191781401634,
      "num_tokens": 1779719.0,
      "step": 705
    },
    {
      "entropy": 7.915401935577393,
      "epoch": 0.08193883439122908,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003545,
      "loss": 7.6389,
      "mean_token_accuracy": 0.13027774319052696,
      "num_tokens": 1791448.0,
      "step": 710
    },
    {
      "entropy": 7.866633939743042,
      "epoch": 0.08251586843623773,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000357,
      "loss": 7.6429,
      "mean_token_accuracy": 0.12651973143219947,
      "num_tokens": 1803802.0,
      "step": 715
    },
    {
      "entropy": 7.941481637954712,
      "epoch": 0.0830929024812464,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003595,
      "loss": 7.6641,
      "mean_token_accuracy": 0.12703499272465707,
      "num_tokens": 1817545.0,
      "step": 720
    },
    {
      "entropy": 7.91947979927063,
      "epoch": 0.08366993652625505,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000362,
      "loss": 7.6464,
      "mean_token_accuracy": 0.12531809359788895,
      "num_tokens": 1830115.0,
      "step": 725
    },
    {
      "entropy": 7.799381828308105,
      "epoch": 0.08424697057126371,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003645,
      "loss": 7.5711,
      "mean_token_accuracy": 0.12959943637251853,
      "num_tokens": 1844707.0,
      "step": 730
    },
    {
      "entropy": 7.925750207901001,
      "epoch": 0.08482400461627236,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000367,
      "loss": 7.6557,
      "mean_token_accuracy": 0.12370599582791328,
      "num_tokens": 1858719.0,
      "step": 735
    },
    {
      "entropy": 7.862723875045776,
      "epoch": 0.08540103866128101,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003695,
      "loss": 7.5893,
      "mean_token_accuracy": 0.12989407479763032,
      "num_tokens": 1870421.0,
      "step": 740
    },
    {
      "entropy": 7.863739395141602,
      "epoch": 0.08597807270628967,
      "grad_norm": 0.890625,
      "learning_rate": 0.000372,
      "loss": 7.6572,
      "mean_token_accuracy": 0.1180440478026867,
      "num_tokens": 1883792.0,
      "step": 745
    },
    {
      "entropy": 7.822341585159302,
      "epoch": 0.08655510675129832,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003745,
      "loss": 7.6349,
      "mean_token_accuracy": 0.12786092832684517,
      "num_tokens": 1896209.0,
      "step": 750
    },
    {
      "entropy": 7.870438051223755,
      "epoch": 0.08713214079630698,
      "grad_norm": 1.046875,
      "learning_rate": 0.000377,
      "loss": 7.5787,
      "mean_token_accuracy": 0.1304923750460148,
      "num_tokens": 1908520.0,
      "step": 755
    },
    {
      "entropy": 7.85968918800354,
      "epoch": 0.08770917484131564,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003795,
      "loss": 7.5898,
      "mean_token_accuracy": 0.12856273204088212,
      "num_tokens": 1922489.0,
      "step": 760
    },
    {
      "entropy": 7.790146064758301,
      "epoch": 0.0882862088863243,
      "grad_norm": 1.015625,
      "learning_rate": 0.000382,
      "loss": 7.5841,
      "mean_token_accuracy": 0.1291157804429531,
      "num_tokens": 1935206.0,
      "step": 765
    },
    {
      "entropy": 7.834140348434448,
      "epoch": 0.08886324293133295,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003845,
      "loss": 7.5396,
      "mean_token_accuracy": 0.13339237496256828,
      "num_tokens": 1948798.0,
      "step": 770
    },
    {
      "entropy": 7.79986310005188,
      "epoch": 0.0894402769763416,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038700000000000003,
      "loss": 7.537,
      "mean_token_accuracy": 0.13368105366826058,
      "num_tokens": 1961095.0,
      "step": 775
    },
    {
      "entropy": 7.886825084686279,
      "epoch": 0.09001731102135026,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038950000000000003,
      "loss": 7.5479,
      "mean_token_accuracy": 0.12701786160469056,
      "num_tokens": 1974229.0,
      "step": 780
    },
    {
      "entropy": 7.805331611633301,
      "epoch": 0.09059434506635891,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00039200000000000004,
      "loss": 7.6401,
      "mean_token_accuracy": 0.12609824389219285,
      "num_tokens": 1986334.0,
      "step": 785
    },
    {
      "entropy": 7.807611322402954,
      "epoch": 0.09117137911136756,
      "grad_norm": 0.828125,
      "learning_rate": 0.00039450000000000005,
      "loss": 7.5884,
      "mean_token_accuracy": 0.12830588668584825,
      "num_tokens": 2000337.0,
      "step": 790
    },
    {
      "entropy": 7.795743417739868,
      "epoch": 0.09174841315637623,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00039700000000000005,
      "loss": 7.51,
      "mean_token_accuracy": 0.13368136137723924,
      "num_tokens": 2012048.0,
      "step": 795
    },
    {
      "entropy": 7.784616470336914,
      "epoch": 0.09232544720138489,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003995,
      "loss": 7.529,
      "mean_token_accuracy": 0.13036411479115487,
      "num_tokens": 2025244.0,
      "step": 800
    },
    {
      "entropy": 7.810312700271607,
      "epoch": 0.09290248124639354,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000402,
      "loss": 7.5725,
      "mean_token_accuracy": 0.13225525096058846,
      "num_tokens": 2036542.0,
      "step": 805
    },
    {
      "entropy": 7.813270044326782,
      "epoch": 0.0934795152914022,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004045,
      "loss": 7.5115,
      "mean_token_accuracy": 0.13005766347050668,
      "num_tokens": 2049358.0,
      "step": 810
    },
    {
      "entropy": 7.700181245803833,
      "epoch": 0.09405654933641085,
      "grad_norm": 0.953125,
      "learning_rate": 0.00040699999999999997,
      "loss": 7.5409,
      "mean_token_accuracy": 0.12904200404882432,
      "num_tokens": 2062114.0,
      "step": 815
    },
    {
      "entropy": 7.861858272552491,
      "epoch": 0.0946335833814195,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004095,
      "loss": 7.5139,
      "mean_token_accuracy": 0.13364754766225814,
      "num_tokens": 2074236.0,
      "step": 820
    },
    {
      "entropy": 7.705525732040405,
      "epoch": 0.09521061742642815,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000412,
      "loss": 7.4266,
      "mean_token_accuracy": 0.13591486886143683,
      "num_tokens": 2086320.0,
      "step": 825
    },
    {
      "entropy": 7.766620826721192,
      "epoch": 0.09578765147143682,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004145,
      "loss": 7.4885,
      "mean_token_accuracy": 0.13510627821087837,
      "num_tokens": 2099119.0,
      "step": 830
    },
    {
      "entropy": 7.711123943328857,
      "epoch": 0.09636468551644548,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000417,
      "loss": 7.5148,
      "mean_token_accuracy": 0.1323984071612358,
      "num_tokens": 2111795.0,
      "step": 835
    },
    {
      "entropy": 7.660836982727051,
      "epoch": 0.09694171956145413,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004195,
      "loss": 7.4957,
      "mean_token_accuracy": 0.11923368275165558,
      "num_tokens": 2126621.0,
      "step": 840
    },
    {
      "entropy": 7.764021539688111,
      "epoch": 0.09751875360646278,
      "grad_norm": 0.9375,
      "learning_rate": 0.000422,
      "loss": 7.3964,
      "mean_token_accuracy": 0.1446671850979328,
      "num_tokens": 2139770.0,
      "step": 845
    },
    {
      "entropy": 7.73313717842102,
      "epoch": 0.09809578765147144,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004245,
      "loss": 7.4817,
      "mean_token_accuracy": 0.13560690730810165,
      "num_tokens": 2151683.0,
      "step": 850
    },
    {
      "entropy": 7.648449945449829,
      "epoch": 0.09867282169648009,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000427,
      "loss": 7.4604,
      "mean_token_accuracy": 0.1294977843761444,
      "num_tokens": 2163178.0,
      "step": 855
    },
    {
      "entropy": 7.755085849761963,
      "epoch": 0.09924985574148874,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004295,
      "loss": 7.5485,
      "mean_token_accuracy": 0.12189057022333145,
      "num_tokens": 2176312.0,
      "step": 860
    },
    {
      "entropy": 7.789842224121093,
      "epoch": 0.0998268897864974,
      "grad_norm": 1.390625,
      "learning_rate": 0.000432,
      "loss": 7.48,
      "mean_token_accuracy": 0.12981949299573897,
      "num_tokens": 2189045.0,
      "step": 865
    },
    {
      "entropy": 7.571734762191772,
      "epoch": 0.10040392383150606,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004345,
      "loss": 7.4754,
      "mean_token_accuracy": 0.12462260350584983,
      "num_tokens": 2202083.0,
      "step": 870
    },
    {
      "entropy": 7.715145778656006,
      "epoch": 0.10098095787651472,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000437,
      "loss": 7.5248,
      "mean_token_accuracy": 0.1264673389494419,
      "num_tokens": 2215775.0,
      "step": 875
    },
    {
      "entropy": 7.77250919342041,
      "epoch": 0.10155799192152337,
      "grad_norm": 1.0,
      "learning_rate": 0.0004395,
      "loss": 7.4774,
      "mean_token_accuracy": 0.12318376153707504,
      "num_tokens": 2228387.0,
      "step": 880
    },
    {
      "entropy": 7.6760975360870365,
      "epoch": 0.10213502596653203,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000442,
      "loss": 7.4827,
      "mean_token_accuracy": 0.13410628139972686,
      "num_tokens": 2241491.0,
      "step": 885
    },
    {
      "entropy": 7.5755048274993895,
      "epoch": 0.10271206001154068,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004445,
      "loss": 7.4057,
      "mean_token_accuracy": 0.13969002217054366,
      "num_tokens": 2255282.0,
      "step": 890
    },
    {
      "entropy": 7.67344913482666,
      "epoch": 0.10328909405654933,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000447,
      "loss": 7.4078,
      "mean_token_accuracy": 0.13210881277918815,
      "num_tokens": 2268081.0,
      "step": 895
    },
    {
      "entropy": 7.641108894348145,
      "epoch": 0.10386612810155799,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044950000000000003,
      "loss": 7.3767,
      "mean_token_accuracy": 0.1385306254029274,
      "num_tokens": 2281069.0,
      "step": 900
    },
    {
      "entropy": 7.662174987792969,
      "epoch": 0.10444316214656665,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00045200000000000004,
      "loss": 7.4474,
      "mean_token_accuracy": 0.12799655497074128,
      "num_tokens": 2293606.0,
      "step": 905
    },
    {
      "entropy": 7.689595937728882,
      "epoch": 0.10502019619157531,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045450000000000004,
      "loss": 7.4478,
      "mean_token_accuracy": 0.12830137014389037,
      "num_tokens": 2304973.0,
      "step": 910
    },
    {
      "entropy": 7.686489248275757,
      "epoch": 0.10559723023658396,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00045700000000000005,
      "loss": 7.4863,
      "mean_token_accuracy": 0.12266820147633553,
      "num_tokens": 2319016.0,
      "step": 915
    },
    {
      "entropy": 7.6014503002166744,
      "epoch": 0.10617426428159261,
      "grad_norm": 1.0,
      "learning_rate": 0.00045950000000000006,
      "loss": 7.4463,
      "mean_token_accuracy": 0.1375369183719158,
      "num_tokens": 2330599.0,
      "step": 920
    },
    {
      "entropy": 7.717125272750854,
      "epoch": 0.10675129832660127,
      "grad_norm": 1.09375,
      "learning_rate": 0.000462,
      "loss": 7.4444,
      "mean_token_accuracy": 0.12856984063982962,
      "num_tokens": 2343153.0,
      "step": 925
    },
    {
      "entropy": 7.697896146774292,
      "epoch": 0.10732833237160992,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004645,
      "loss": 7.4695,
      "mean_token_accuracy": 0.12443050369620323,
      "num_tokens": 2356762.0,
      "step": 930
    },
    {
      "entropy": 7.53573899269104,
      "epoch": 0.10790536641661858,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000467,
      "loss": 7.4916,
      "mean_token_accuracy": 0.13005805611610413,
      "num_tokens": 2369581.0,
      "step": 935
    },
    {
      "entropy": 7.7132821559906,
      "epoch": 0.10848240046162723,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004695,
      "loss": 7.4025,
      "mean_token_accuracy": 0.1298571690917015,
      "num_tokens": 2381419.0,
      "step": 940
    },
    {
      "entropy": 7.599806404113769,
      "epoch": 0.1090594345066359,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000472,
      "loss": 7.404,
      "mean_token_accuracy": 0.1387295603752136,
      "num_tokens": 2393586.0,
      "step": 945
    },
    {
      "entropy": 7.5154454708099365,
      "epoch": 0.10963646855164455,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004745,
      "loss": 7.3841,
      "mean_token_accuracy": 0.13711024671792985,
      "num_tokens": 2407720.0,
      "step": 950
    },
    {
      "entropy": 7.691217041015625,
      "epoch": 0.1102135025966532,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000477,
      "loss": 7.4341,
      "mean_token_accuracy": 0.12648762315511702,
      "num_tokens": 2421089.0,
      "step": 955
    },
    {
      "entropy": 7.593528985977173,
      "epoch": 0.11079053664166186,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004795,
      "loss": 7.3719,
      "mean_token_accuracy": 0.13786724731326103,
      "num_tokens": 2432859.0,
      "step": 960
    },
    {
      "entropy": 7.563003826141357,
      "epoch": 0.11136757068667051,
      "grad_norm": 0.96875,
      "learning_rate": 0.000482,
      "loss": 7.3823,
      "mean_token_accuracy": 0.1313467375934124,
      "num_tokens": 2445072.0,
      "step": 965
    },
    {
      "entropy": 7.626921939849853,
      "epoch": 0.11194460473167916,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004845,
      "loss": 7.3802,
      "mean_token_accuracy": 0.13656689301133157,
      "num_tokens": 2458724.0,
      "step": 970
    },
    {
      "entropy": 7.551381254196167,
      "epoch": 0.11252163877668782,
      "grad_norm": 0.953125,
      "learning_rate": 0.000487,
      "loss": 7.4385,
      "mean_token_accuracy": 0.13167190551757812,
      "num_tokens": 2472976.0,
      "step": 975
    },
    {
      "entropy": 7.675592613220215,
      "epoch": 0.11309867282169649,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004895,
      "loss": 7.4166,
      "mean_token_accuracy": 0.12921208068728446,
      "num_tokens": 2486164.0,
      "step": 980
    },
    {
      "entropy": 7.476428079605102,
      "epoch": 0.11367570686670514,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000492,
      "loss": 7.3646,
      "mean_token_accuracy": 0.13289769142866134,
      "num_tokens": 2499395.0,
      "step": 985
    },
    {
      "entropy": 7.640331506729126,
      "epoch": 0.1142527409117138,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004945,
      "loss": 7.3659,
      "mean_token_accuracy": 0.13304105922579765,
      "num_tokens": 2512163.0,
      "step": 990
    },
    {
      "entropy": 7.560707950592041,
      "epoch": 0.11482977495672245,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000497,
      "loss": 7.3714,
      "mean_token_accuracy": 0.12891542091965674,
      "num_tokens": 2523634.0,
      "step": 995
    },
    {
      "entropy": 7.603393411636352,
      "epoch": 0.1154068090017311,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004995,
      "loss": 7.4008,
      "mean_token_accuracy": 0.1331055149435997,
      "num_tokens": 2536237.0,
      "step": 1000
    },
    {
      "entropy": 7.625240325927734,
      "epoch": 0.11598384304673975,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999999975783157,
      "loss": 7.3994,
      "mean_token_accuracy": 0.11986244916915893,
      "num_tokens": 2548035.0,
      "step": 1005
    },
    {
      "entropy": 7.594330739974976,
      "epoch": 0.11656087709174841,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004999999877402236,
      "loss": 7.3344,
      "mean_token_accuracy": 0.13319490626454353,
      "num_tokens": 2561670.0,
      "step": 1010
    },
    {
      "entropy": 7.533303833007812,
      "epoch": 0.11713791113675708,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999999703343686,
      "loss": 7.3709,
      "mean_token_accuracy": 0.1304928630590439,
      "num_tokens": 2573471.0,
      "step": 1015
    },
    {
      "entropy": 7.531730508804321,
      "epoch": 0.11771494518176573,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999999453607515,
      "loss": 7.3175,
      "mean_token_accuracy": 0.13719066530466079,
      "num_tokens": 2586534.0,
      "step": 1020
    },
    {
      "entropy": 7.636745643615723,
      "epoch": 0.11829197922677438,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999999128193729,
      "loss": 7.3719,
      "mean_token_accuracy": 0.12794509902596474,
      "num_tokens": 2597970.0,
      "step": 1025
    },
    {
      "entropy": 7.373585796356201,
      "epoch": 0.11886901327178304,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000499999872710234,
      "loss": 7.308,
      "mean_token_accuracy": 0.13802397325634957,
      "num_tokens": 2610924.0,
      "step": 1030
    },
    {
      "entropy": 7.593305397033691,
      "epoch": 0.11944604731679169,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004999998250333361,
      "loss": 7.3737,
      "mean_token_accuracy": 0.13383316844701768,
      "num_tokens": 2623571.0,
      "step": 1035
    },
    {
      "entropy": 7.540350818634034,
      "epoch": 0.12002308136180034,
      "grad_norm": 1.0,
      "learning_rate": 0.000499999769788681,
      "loss": 7.3823,
      "mean_token_accuracy": 0.1317036457359791,
      "num_tokens": 2636063.0,
      "step": 1040
    },
    {
      "entropy": 7.56674108505249,
      "epoch": 0.120600115406809,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004999997069762703,
      "loss": 7.3334,
      "mean_token_accuracy": 0.1320666678249836,
      "num_tokens": 2648621.0,
      "step": 1045
    },
    {
      "entropy": 7.478792190551758,
      "epoch": 0.12117714945181765,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999996365961062,
      "loss": 7.2548,
      "mean_token_accuracy": 0.13802519217133521,
      "num_tokens": 2661794.0,
      "step": 1050
    },
    {
      "entropy": 7.471138048171997,
      "epoch": 0.12175418349682632,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999995586481912,
      "loss": 7.3311,
      "mean_token_accuracy": 0.1260454349219799,
      "num_tokens": 2674048.0,
      "step": 1055
    },
    {
      "entropy": 7.464170122146607,
      "epoch": 0.12233121754183497,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999994731325276,
      "loss": 7.2485,
      "mean_token_accuracy": 0.13324502855539322,
      "num_tokens": 2685510.0,
      "step": 1060
    },
    {
      "entropy": 7.5221740245819095,
      "epoch": 0.12290825158684363,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004999993800491187,
      "loss": 7.3328,
      "mean_token_accuracy": 0.13226443603634835,
      "num_tokens": 2698414.0,
      "step": 1065
    },
    {
      "entropy": 7.50278525352478,
      "epoch": 0.12348528563185228,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004999992793979672,
      "loss": 7.3332,
      "mean_token_accuracy": 0.1338995724916458,
      "num_tokens": 2711343.0,
      "step": 1070
    },
    {
      "entropy": 7.443923377990723,
      "epoch": 0.12406231967686093,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999991711790769,
      "loss": 7.2217,
      "mean_token_accuracy": 0.14058572947978973,
      "num_tokens": 2723123.0,
      "step": 1075
    },
    {
      "entropy": 7.387674856185913,
      "epoch": 0.12463935372186959,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999990553924511,
      "loss": 7.2621,
      "mean_token_accuracy": 0.13784897848963737,
      "num_tokens": 2736015.0,
      "step": 1080
    },
    {
      "entropy": 7.483056497573853,
      "epoch": 0.12521638776687824,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999989320380939,
      "loss": 7.2327,
      "mean_token_accuracy": 0.1457214944064617,
      "num_tokens": 2747561.0,
      "step": 1085
    },
    {
      "entropy": 7.438428640365601,
      "epoch": 0.1257934218118869,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999988011160094,
      "loss": 7.2835,
      "mean_token_accuracy": 0.13277626112103463,
      "num_tokens": 2758992.0,
      "step": 1090
    },
    {
      "entropy": 7.489367580413818,
      "epoch": 0.12637045585689555,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499998662626202,
      "loss": 7.2773,
      "mean_token_accuracy": 0.13354492634534837,
      "num_tokens": 2771164.0,
      "step": 1095
    },
    {
      "entropy": 7.357281446456909,
      "epoch": 0.12694748990190421,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999985165686764,
      "loss": 7.2575,
      "mean_token_accuracy": 0.1360307179391384,
      "num_tokens": 2784400.0,
      "step": 1100
    },
    {
      "entropy": 7.427149534225464,
      "epoch": 0.12752452394691285,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999983629434373,
      "loss": 7.223,
      "mean_token_accuracy": 0.13857290968298913,
      "num_tokens": 2795729.0,
      "step": 1105
    },
    {
      "entropy": 7.424337244033813,
      "epoch": 0.12810155799192152,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999982017504901,
      "loss": 7.307,
      "mean_token_accuracy": 0.13230945765972138,
      "num_tokens": 2808118.0,
      "step": 1110
    },
    {
      "entropy": 7.445949077606201,
      "epoch": 0.1286785920369302,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999980329898401,
      "loss": 7.236,
      "mean_token_accuracy": 0.1337020590901375,
      "num_tokens": 2821118.0,
      "step": 1115
    },
    {
      "entropy": 7.446945095062256,
      "epoch": 0.12925562608193883,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000499997856661493,
      "loss": 7.233,
      "mean_token_accuracy": 0.13448369577527047,
      "num_tokens": 2833636.0,
      "step": 1120
    },
    {
      "entropy": 7.35188136100769,
      "epoch": 0.1298326601269475,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499997672765455,
      "loss": 7.2509,
      "mean_token_accuracy": 0.1398313045501709,
      "num_tokens": 2846617.0,
      "step": 1125
    },
    {
      "entropy": 7.367458200454712,
      "epoch": 0.13040969417195614,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999974813017318,
      "loss": 7.2097,
      "mean_token_accuracy": 0.14283193722367288,
      "num_tokens": 2859191.0,
      "step": 1130
    },
    {
      "entropy": 7.45753436088562,
      "epoch": 0.1309867282169648,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999972822703301,
      "loss": 7.2303,
      "mean_token_accuracy": 0.13343884348869323,
      "num_tokens": 2872592.0,
      "step": 1135
    },
    {
      "entropy": 7.310500621795654,
      "epoch": 0.13156376226197344,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999970756712567,
      "loss": 7.2068,
      "mean_token_accuracy": 0.14036742150783538,
      "num_tokens": 2884094.0,
      "step": 1140
    },
    {
      "entropy": 7.466288280487061,
      "epoch": 0.1321407963069821,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999968615045183,
      "loss": 7.265,
      "mean_token_accuracy": 0.1353508099913597,
      "num_tokens": 2896947.0,
      "step": 1145
    },
    {
      "entropy": 7.352451229095459,
      "epoch": 0.13271783035199078,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999966397701222,
      "loss": 7.2677,
      "mean_token_accuracy": 0.13296897783875467,
      "num_tokens": 2909029.0,
      "step": 1150
    },
    {
      "entropy": 7.461985635757446,
      "epoch": 0.13329486439699942,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999964104680759,
      "loss": 7.2432,
      "mean_token_accuracy": 0.13714048266410828,
      "num_tokens": 2920778.0,
      "step": 1155
    },
    {
      "entropy": 7.354297304153443,
      "epoch": 0.13387189844200809,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999961735983871,
      "loss": 7.1539,
      "mean_token_accuracy": 0.14073749631643295,
      "num_tokens": 2933754.0,
      "step": 1160
    },
    {
      "entropy": 7.476262760162354,
      "epoch": 0.13444893248701673,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999959291610639,
      "loss": 7.2608,
      "mean_token_accuracy": 0.13915467262268066,
      "num_tokens": 2946019.0,
      "step": 1165
    },
    {
      "entropy": 7.351610231399536,
      "epoch": 0.1350259665320254,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999956771561143,
      "loss": 7.2192,
      "mean_token_accuracy": 0.13505645617842674,
      "num_tokens": 2959688.0,
      "step": 1170
    },
    {
      "entropy": 7.36425518989563,
      "epoch": 0.13560300057703403,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999954175835469,
      "loss": 7.1361,
      "mean_token_accuracy": 0.14523073583841323,
      "num_tokens": 2971211.0,
      "step": 1175
    },
    {
      "entropy": 7.323524904251099,
      "epoch": 0.1361800346220427,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999951504433703,
      "loss": 7.2293,
      "mean_token_accuracy": 0.13175321519374847,
      "num_tokens": 2983559.0,
      "step": 1180
    },
    {
      "entropy": 7.404617214202881,
      "epoch": 0.13675706866705137,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999948757355935,
      "loss": 7.2025,
      "mean_token_accuracy": 0.1324743650853634,
      "num_tokens": 2995053.0,
      "step": 1185
    },
    {
      "entropy": 7.352153348922729,
      "epoch": 0.13733410271206,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499994593460226,
      "loss": 7.1457,
      "mean_token_accuracy": 0.13915842920541763,
      "num_tokens": 3006143.0,
      "step": 1190
    },
    {
      "entropy": 7.277075576782226,
      "epoch": 0.13791113675706868,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999943036172771,
      "loss": 7.147,
      "mean_token_accuracy": 0.14147392511367798,
      "num_tokens": 3018793.0,
      "step": 1195
    },
    {
      "entropy": 7.392578411102295,
      "epoch": 0.13848817080207732,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999940062067565,
      "loss": 7.1464,
      "mean_token_accuracy": 0.13925720006227493,
      "num_tokens": 3031547.0,
      "step": 1200
    },
    {
      "entropy": 7.269504356384277,
      "epoch": 0.13906520484708598,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999937012286742,
      "loss": 7.1604,
      "mean_token_accuracy": 0.13719142302870752,
      "num_tokens": 3044035.0,
      "step": 1205
    },
    {
      "entropy": 7.297099208831787,
      "epoch": 0.13964223889209462,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999933886830405,
      "loss": 7.1862,
      "mean_token_accuracy": 0.14078905582427978,
      "num_tokens": 3056238.0,
      "step": 1210
    },
    {
      "entropy": 7.321411800384522,
      "epoch": 0.1402192729371033,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499993068569866,
      "loss": 7.1833,
      "mean_token_accuracy": 0.14090409278869628,
      "num_tokens": 3069250.0,
      "step": 1215
    },
    {
      "entropy": 7.357383728027344,
      "epoch": 0.14079630698211196,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999927408891614,
      "loss": 7.1557,
      "mean_token_accuracy": 0.13715188950300217,
      "num_tokens": 3082048.0,
      "step": 1220
    },
    {
      "entropy": 7.301809167861938,
      "epoch": 0.1413733410271206,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999924056409378,
      "loss": 7.2215,
      "mean_token_accuracy": 0.13659722208976746,
      "num_tokens": 3094588.0,
      "step": 1225
    },
    {
      "entropy": 7.427079439163208,
      "epoch": 0.14195037507212926,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999920628252063,
      "loss": 7.2441,
      "mean_token_accuracy": 0.128413824737072,
      "num_tokens": 3108067.0,
      "step": 1230
    },
    {
      "entropy": 7.301552772521973,
      "epoch": 0.1425274091171379,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004999917124419785,
      "loss": 7.1715,
      "mean_token_accuracy": 0.1430498704314232,
      "num_tokens": 3120524.0,
      "step": 1235
    },
    {
      "entropy": 7.375691175460815,
      "epoch": 0.14310444316214657,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999913544912664,
      "loss": 7.2377,
      "mean_token_accuracy": 0.13174993246793748,
      "num_tokens": 3133533.0,
      "step": 1240
    },
    {
      "entropy": 7.397565317153931,
      "epoch": 0.1436814772071552,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999909889730817,
      "loss": 7.2182,
      "mean_token_accuracy": 0.137331123650074,
      "num_tokens": 3146183.0,
      "step": 1245
    },
    {
      "entropy": 7.296873569488525,
      "epoch": 0.14425851125216388,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999906158874368,
      "loss": 7.1694,
      "mean_token_accuracy": 0.13684176057577133,
      "num_tokens": 3158436.0,
      "step": 1250
    },
    {
      "entropy": 7.299059915542602,
      "epoch": 0.14483554529717255,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999902352343444,
      "loss": 7.1194,
      "mean_token_accuracy": 0.1443310186266899,
      "num_tokens": 3170426.0,
      "step": 1255
    },
    {
      "entropy": 7.344976425170898,
      "epoch": 0.1454125793421812,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999898470138171,
      "loss": 7.1264,
      "mean_token_accuracy": 0.13932598382234573,
      "num_tokens": 3182865.0,
      "step": 1260
    },
    {
      "entropy": 7.26788363456726,
      "epoch": 0.14598961338718985,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999894512258681,
      "loss": 7.1381,
      "mean_token_accuracy": 0.13955464959144592,
      "num_tokens": 3195674.0,
      "step": 1265
    },
    {
      "entropy": 7.33721137046814,
      "epoch": 0.1465666474321985,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004999890478705107,
      "loss": 7.114,
      "mean_token_accuracy": 0.14049444273114203,
      "num_tokens": 3208848.0,
      "step": 1270
    },
    {
      "entropy": 7.3063184261322025,
      "epoch": 0.14714368147720716,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999886369477585,
      "loss": 7.2017,
      "mean_token_accuracy": 0.1356468491256237,
      "num_tokens": 3221179.0,
      "step": 1275
    },
    {
      "entropy": 7.336249351501465,
      "epoch": 0.1477207155222158,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999882184576251,
      "loss": 7.2815,
      "mean_token_accuracy": 0.13257396966218948,
      "num_tokens": 3233671.0,
      "step": 1280
    },
    {
      "entropy": 7.405120134353638,
      "epoch": 0.14829774956722447,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004999877924001248,
      "loss": 7.1979,
      "mean_token_accuracy": 0.13243449851870537,
      "num_tokens": 3247219.0,
      "step": 1285
    },
    {
      "entropy": 7.313788509368896,
      "epoch": 0.1488747836122331,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999873587752718,
      "loss": 7.1673,
      "mean_token_accuracy": 0.13630512207746506,
      "num_tokens": 3259444.0,
      "step": 1290
    },
    {
      "entropy": 7.180786943435669,
      "epoch": 0.14945181765724178,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999869175830808,
      "loss": 7.019,
      "mean_token_accuracy": 0.14473430812358856,
      "num_tokens": 3272092.0,
      "step": 1295
    },
    {
      "entropy": 7.377786588668823,
      "epoch": 0.15002885170225044,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999864688235666,
      "loss": 7.1995,
      "mean_token_accuracy": 0.12725819498300553,
      "num_tokens": 3284260.0,
      "step": 1300
    },
    {
      "entropy": 7.249001407623291,
      "epoch": 0.15060588574725908,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999860124967442,
      "loss": 7.1613,
      "mean_token_accuracy": 0.13158362433314325,
      "num_tokens": 3296662.0,
      "step": 1305
    },
    {
      "entropy": 7.297910118103028,
      "epoch": 0.15118291979226775,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999855486026291,
      "loss": 7.1213,
      "mean_token_accuracy": 0.14045775979757308,
      "num_tokens": 3307782.0,
      "step": 1310
    },
    {
      "entropy": 7.283909845352173,
      "epoch": 0.1517599538372764,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999850771412369,
      "loss": 7.1728,
      "mean_token_accuracy": 0.13270583972334862,
      "num_tokens": 3320519.0,
      "step": 1315
    },
    {
      "entropy": 7.31526288986206,
      "epoch": 0.15233698788228506,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999845981125832,
      "loss": 7.0413,
      "mean_token_accuracy": 0.14500000029802323,
      "num_tokens": 3331722.0,
      "step": 1320
    },
    {
      "entropy": 7.212509202957153,
      "epoch": 0.1529140219272937,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999841115166844,
      "loss": 7.1385,
      "mean_token_accuracy": 0.13171057179570198,
      "num_tokens": 3345143.0,
      "step": 1325
    },
    {
      "entropy": 7.255769968032837,
      "epoch": 0.15349105597230236,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999836173535568,
      "loss": 7.0707,
      "mean_token_accuracy": 0.14495590403676034,
      "num_tokens": 3358946.0,
      "step": 1330
    },
    {
      "entropy": 7.313462114334106,
      "epoch": 0.15406809001731103,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999831156232169,
      "loss": 7.1788,
      "mean_token_accuracy": 0.13734060302376747,
      "num_tokens": 3372465.0,
      "step": 1335
    },
    {
      "entropy": 7.221960115432739,
      "epoch": 0.15464512406231967,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999826063256818,
      "loss": 7.0636,
      "mean_token_accuracy": 0.1419673442840576,
      "num_tokens": 3384392.0,
      "step": 1340
    },
    {
      "entropy": 7.322636604309082,
      "epoch": 0.15522215810732834,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004999820894609683,
      "loss": 7.1346,
      "mean_token_accuracy": 0.12882967665791512,
      "num_tokens": 3398137.0,
      "step": 1345
    },
    {
      "entropy": 7.281394720077515,
      "epoch": 0.15579919215233698,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999815650290941,
      "loss": 7.1003,
      "mean_token_accuracy": 0.13911654129624368,
      "num_tokens": 3410353.0,
      "step": 1350
    },
    {
      "entropy": 7.242393922805786,
      "epoch": 0.15637622619734565,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999810330300766,
      "loss": 7.0515,
      "mean_token_accuracy": 0.13970422446727754,
      "num_tokens": 3423206.0,
      "step": 1355
    },
    {
      "entropy": 7.206804037094116,
      "epoch": 0.1569532602423543,
      "grad_norm": 0.828125,
      "learning_rate": 0.000499980493463934,
      "loss": 7.1134,
      "mean_token_accuracy": 0.1346282660961151,
      "num_tokens": 3436780.0,
      "step": 1360
    },
    {
      "entropy": 7.3222535133361815,
      "epoch": 0.15753029428736295,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999799463306841,
      "loss": 7.1791,
      "mean_token_accuracy": 0.13330562338232993,
      "num_tokens": 3450285.0,
      "step": 1365
    },
    {
      "entropy": 7.289401626586914,
      "epoch": 0.15810732833237162,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999793916303455,
      "loss": 7.0678,
      "mean_token_accuracy": 0.14209839105606079,
      "num_tokens": 3462157.0,
      "step": 1370
    },
    {
      "entropy": 7.195488595962525,
      "epoch": 0.15868436237738026,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999788293629368,
      "loss": 7.0299,
      "mean_token_accuracy": 0.14806710481643676,
      "num_tokens": 3475027.0,
      "step": 1375
    },
    {
      "entropy": 7.218692922592163,
      "epoch": 0.15926139642238893,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000499978259528477,
      "loss": 7.1068,
      "mean_token_accuracy": 0.14241924211382867,
      "num_tokens": 3489111.0,
      "step": 1380
    },
    {
      "entropy": 7.240436649322509,
      "epoch": 0.15983843046739757,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999776821269852,
      "loss": 7.0521,
      "mean_token_accuracy": 0.13724546879529953,
      "num_tokens": 3501673.0,
      "step": 1385
    },
    {
      "entropy": 7.110876035690308,
      "epoch": 0.16041546451240624,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999770971584807,
      "loss": 6.9797,
      "mean_token_accuracy": 0.1431804358959198,
      "num_tokens": 3513511.0,
      "step": 1390
    },
    {
      "entropy": 7.26352219581604,
      "epoch": 0.16099249855741488,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999765046229834,
      "loss": 7.1193,
      "mean_token_accuracy": 0.13453444615006446,
      "num_tokens": 3525218.0,
      "step": 1395
    },
    {
      "entropy": 7.1966602325439455,
      "epoch": 0.16156953260242354,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499975904520513,
      "loss": 6.9928,
      "mean_token_accuracy": 0.1506774343550205,
      "num_tokens": 3537006.0,
      "step": 1400
    },
    {
      "entropy": 7.221553230285645,
      "epoch": 0.1621465666474322,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00049997529685109,
      "loss": 7.1169,
      "mean_token_accuracy": 0.13538158535957337,
      "num_tokens": 3549873.0,
      "step": 1405
    },
    {
      "entropy": 7.317210006713867,
      "epoch": 0.16272360069244085,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999746816147344,
      "loss": 7.1801,
      "mean_token_accuracy": 0.1276666961610317,
      "num_tokens": 3563542.0,
      "step": 1410
    },
    {
      "entropy": 7.172958326339722,
      "epoch": 0.16330063473744952,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004999740588114673,
      "loss": 7.0726,
      "mean_token_accuracy": 0.13991388827562332,
      "num_tokens": 3575679.0,
      "step": 1415
    },
    {
      "entropy": 7.290738105773926,
      "epoch": 0.16387766878245816,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999734284413095,
      "loss": 7.0954,
      "mean_token_accuracy": 0.14116299375891686,
      "num_tokens": 3588605.0,
      "step": 1420
    },
    {
      "entropy": 7.13003716468811,
      "epoch": 0.16445470282746683,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499972790504282,
      "loss": 7.0764,
      "mean_token_accuracy": 0.1506669543683529,
      "num_tokens": 3601894.0,
      "step": 1425
    },
    {
      "entropy": 7.193125915527344,
      "epoch": 0.16503173687247547,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004999721450004067,
      "loss": 7.0103,
      "mean_token_accuracy": 0.14832852333784102,
      "num_tokens": 3614436.0,
      "step": 1430
    },
    {
      "entropy": 7.113531923294067,
      "epoch": 0.16560877091748413,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999714919297049,
      "loss": 6.9698,
      "mean_token_accuracy": 0.14278605505824088,
      "num_tokens": 3627590.0,
      "step": 1435
    },
    {
      "entropy": 7.303047561645508,
      "epoch": 0.1661858049624928,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004999708312921987,
      "loss": 7.1398,
      "mean_token_accuracy": 0.1305020570755005,
      "num_tokens": 3640834.0,
      "step": 1440
    },
    {
      "entropy": 7.112206935882568,
      "epoch": 0.16676283900750144,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999701630879103,
      "loss": 6.9822,
      "mean_token_accuracy": 0.14259293228387832,
      "num_tokens": 3653024.0,
      "step": 1445
    },
    {
      "entropy": 7.1947370052337645,
      "epoch": 0.1673398730525101,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999694873168623,
      "loss": 7.0366,
      "mean_token_accuracy": 0.1446026474237442,
      "num_tokens": 3665311.0,
      "step": 1450
    },
    {
      "entropy": 7.208151340484619,
      "epoch": 0.16791690709751875,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004999688039790773,
      "loss": 7.0803,
      "mean_token_accuracy": 0.1384113185107708,
      "num_tokens": 3679102.0,
      "step": 1455
    },
    {
      "entropy": 7.164588069915771,
      "epoch": 0.16849394114252741,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999681130745784,
      "loss": 6.9906,
      "mean_token_accuracy": 0.14679074734449388,
      "num_tokens": 3690046.0,
      "step": 1460
    },
    {
      "entropy": 7.1119578838348385,
      "epoch": 0.16907097518753605,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999674146033888,
      "loss": 7.0146,
      "mean_token_accuracy": 0.1386751838028431,
      "num_tokens": 3703668.0,
      "step": 1465
    },
    {
      "entropy": 7.215666484832764,
      "epoch": 0.16964800923254472,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004999667085655318,
      "loss": 7.0317,
      "mean_token_accuracy": 0.142906092107296,
      "num_tokens": 3715852.0,
      "step": 1470
    },
    {
      "entropy": 7.163801479339599,
      "epoch": 0.1702250432775534,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999659949610313,
      "loss": 6.9624,
      "mean_token_accuracy": 0.1409413382411003,
      "num_tokens": 3728310.0,
      "step": 1475
    },
    {
      "entropy": 7.225057554244995,
      "epoch": 0.17080207732256203,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999652737899114,
      "loss": 7.0428,
      "mean_token_accuracy": 0.13530816361308098,
      "num_tokens": 3741323.0,
      "step": 1480
    },
    {
      "entropy": 7.164874649047851,
      "epoch": 0.1713791113675707,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004999645450521963,
      "loss": 7.0429,
      "mean_token_accuracy": 0.13799375295639038,
      "num_tokens": 3752667.0,
      "step": 1485
    },
    {
      "entropy": 7.169693231582642,
      "epoch": 0.17195614541257934,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999638087479104,
      "loss": 7.0422,
      "mean_token_accuracy": 0.13898203894495964,
      "num_tokens": 3764282.0,
      "step": 1490
    },
    {
      "entropy": 7.190469121932983,
      "epoch": 0.172533179457588,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004999630648770786,
      "loss": 7.0687,
      "mean_token_accuracy": 0.13370918706059456,
      "num_tokens": 3778280.0,
      "step": 1495
    },
    {
      "entropy": 7.144445991516113,
      "epoch": 0.17311021350259664,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999623134397257,
      "loss": 7.0879,
      "mean_token_accuracy": 0.14445693269371987,
      "num_tokens": 3791010.0,
      "step": 1500
    },
    {
      "entropy": 7.24975266456604,
      "epoch": 0.1736872475476053,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999615544358774,
      "loss": 7.0116,
      "mean_token_accuracy": 0.13824615478515626,
      "num_tokens": 3802738.0,
      "step": 1505
    },
    {
      "entropy": 7.121045970916748,
      "epoch": 0.17426428159261395,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999607878655587,
      "loss": 7.0366,
      "mean_token_accuracy": 0.14151245802640916,
      "num_tokens": 3815448.0,
      "step": 1510
    },
    {
      "entropy": 7.1481671810150145,
      "epoch": 0.17484131563762262,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004999600137287958,
      "loss": 7.0332,
      "mean_token_accuracy": 0.1367475040256977,
      "num_tokens": 3829833.0,
      "step": 1515
    },
    {
      "entropy": 7.121943235397339,
      "epoch": 0.17541834968263129,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004999592320256146,
      "loss": 7.0023,
      "mean_token_accuracy": 0.1391633428633213,
      "num_tokens": 3841544.0,
      "step": 1520
    },
    {
      "entropy": 7.198780822753906,
      "epoch": 0.17599538372763993,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999584427560412,
      "loss": 6.9958,
      "mean_token_accuracy": 0.13729065656661987,
      "num_tokens": 3853886.0,
      "step": 1525
    },
    {
      "entropy": 7.095717239379883,
      "epoch": 0.1765724177726486,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004999576459201024,
      "loss": 7.0104,
      "mean_token_accuracy": 0.1445713609457016,
      "num_tokens": 3867942.0,
      "step": 1530
    },
    {
      "entropy": 7.133176279067993,
      "epoch": 0.17714945181765723,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499956841517825,
      "loss": 7.0257,
      "mean_token_accuracy": 0.14167340844869614,
      "num_tokens": 3881764.0,
      "step": 1535
    },
    {
      "entropy": 7.127892589569091,
      "epoch": 0.1777264858626659,
      "grad_norm": 0.875,
      "learning_rate": 0.000499956029549236,
      "loss": 6.9289,
      "mean_token_accuracy": 0.1465323746204376,
      "num_tokens": 3893979.0,
      "step": 1540
    },
    {
      "entropy": 7.17246675491333,
      "epoch": 0.17830351990767454,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004999552100143625,
      "loss": 7.0449,
      "mean_token_accuracy": 0.13901807218790055,
      "num_tokens": 3905934.0,
      "step": 1545
    },
    {
      "entropy": 7.145467090606689,
      "epoch": 0.1788805539526832,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999543829132324,
      "loss": 6.9948,
      "mean_token_accuracy": 0.14433147236704827,
      "num_tokens": 3918348.0,
      "step": 1550
    },
    {
      "entropy": 7.112019491195679,
      "epoch": 0.17945758799769188,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999535482458734,
      "loss": 6.9814,
      "mean_token_accuracy": 0.13943516984581947,
      "num_tokens": 3931863.0,
      "step": 1555
    },
    {
      "entropy": 7.105517101287842,
      "epoch": 0.18003462204270052,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004999527060123135,
      "loss": 6.9699,
      "mean_token_accuracy": 0.1462463617324829,
      "num_tokens": 3945108.0,
      "step": 1560
    },
    {
      "entropy": 7.1699309825897215,
      "epoch": 0.18061165608770918,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999518562125811,
      "loss": 6.9963,
      "mean_token_accuracy": 0.13957771062850952,
      "num_tokens": 3956695.0,
      "step": 1565
    },
    {
      "entropy": 7.086351537704468,
      "epoch": 0.18118869013271782,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004999509988467047,
      "loss": 6.9873,
      "mean_token_accuracy": 0.1385016493499279,
      "num_tokens": 3970458.0,
      "step": 1570
    },
    {
      "entropy": 7.148052930831909,
      "epoch": 0.1817657241777265,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999501339147132,
      "loss": 6.9739,
      "mean_token_accuracy": 0.13998864516615867,
      "num_tokens": 3981962.0,
      "step": 1575
    },
    {
      "entropy": 7.1447553634643555,
      "epoch": 0.18234275822273513,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999492614166357,
      "loss": 6.982,
      "mean_token_accuracy": 0.13720383271574974,
      "num_tokens": 3993512.0,
      "step": 1580
    },
    {
      "entropy": 7.046021413803101,
      "epoch": 0.1829197922677438,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999483813525014,
      "loss": 6.8679,
      "mean_token_accuracy": 0.15238967761397362,
      "num_tokens": 4005826.0,
      "step": 1585
    },
    {
      "entropy": 7.077240610122681,
      "epoch": 0.18349682631275246,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999474937223403,
      "loss": 6.9461,
      "mean_token_accuracy": 0.14126977398991586,
      "num_tokens": 4018927.0,
      "step": 1590
    },
    {
      "entropy": 7.139278173446655,
      "epoch": 0.1840738603577611,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004999465985261818,
      "loss": 6.9951,
      "mean_token_accuracy": 0.13713297992944717,
      "num_tokens": 4030970.0,
      "step": 1595
    },
    {
      "entropy": 7.041838884353638,
      "epoch": 0.18465089440276977,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004999456957640562,
      "loss": 6.9665,
      "mean_token_accuracy": 0.14336296245455743,
      "num_tokens": 4043396.0,
      "step": 1600
    },
    {
      "entropy": 7.167615985870361,
      "epoch": 0.1852279284477784,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004999447854359939,
      "loss": 6.9946,
      "mean_token_accuracy": 0.13789970949292182,
      "num_tokens": 4055588.0,
      "step": 1605
    },
    {
      "entropy": 7.082289409637451,
      "epoch": 0.18580496249278708,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004999438675420255,
      "loss": 6.9775,
      "mean_token_accuracy": 0.1350480891764164,
      "num_tokens": 4068036.0,
      "step": 1610
    },
    {
      "entropy": 7.123587226867675,
      "epoch": 0.18638199653779572,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999429420821818,
      "loss": 7.015,
      "mean_token_accuracy": 0.1355690099298954,
      "num_tokens": 4080531.0,
      "step": 1615
    },
    {
      "entropy": 7.1394895076751705,
      "epoch": 0.1869590305828044,
      "grad_norm": 0.80078125,
      "learning_rate": 0.000499942009056494,
      "loss": 6.9259,
      "mean_token_accuracy": 0.14576323702931404,
      "num_tokens": 4093614.0,
      "step": 1620
    },
    {
      "entropy": 7.157642984390259,
      "epoch": 0.18753606462781305,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999410684649935,
      "loss": 7.0282,
      "mean_token_accuracy": 0.13709268644452094,
      "num_tokens": 4106836.0,
      "step": 1625
    },
    {
      "entropy": 7.0767253875732425,
      "epoch": 0.1881130986728217,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999401203077119,
      "loss": 6.9509,
      "mean_token_accuracy": 0.14356981962919235,
      "num_tokens": 4118739.0,
      "step": 1630
    },
    {
      "entropy": 7.107544803619385,
      "epoch": 0.18869013271783036,
      "grad_norm": 0.859375,
      "learning_rate": 0.000499939164584681,
      "loss": 6.9214,
      "mean_token_accuracy": 0.14475391134619714,
      "num_tokens": 4131174.0,
      "step": 1635
    },
    {
      "entropy": 7.091096496582031,
      "epoch": 0.189267166762839,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999382012959331,
      "loss": 6.9713,
      "mean_token_accuracy": 0.13550762236118316,
      "num_tokens": 4143459.0,
      "step": 1640
    },
    {
      "entropy": 7.121346950531006,
      "epoch": 0.18984420080784767,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999372304415005,
      "loss": 7.0043,
      "mean_token_accuracy": 0.13536836579442024,
      "num_tokens": 4155864.0,
      "step": 1645
    },
    {
      "entropy": 6.936057138442993,
      "epoch": 0.1904212348528563,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999362520214159,
      "loss": 6.8156,
      "mean_token_accuracy": 0.14860668629407883,
      "num_tokens": 4168433.0,
      "step": 1650
    },
    {
      "entropy": 7.137879848480225,
      "epoch": 0.19099826889786498,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999352660357122,
      "loss": 6.9852,
      "mean_token_accuracy": 0.13815735951066016,
      "num_tokens": 4180852.0,
      "step": 1655
    },
    {
      "entropy": 7.092313289642334,
      "epoch": 0.19157530294287364,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999342724844226,
      "loss": 6.9307,
      "mean_token_accuracy": 0.14323657527565956,
      "num_tokens": 4194712.0,
      "step": 1660
    },
    {
      "entropy": 7.107403707504273,
      "epoch": 0.19215233698788228,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999332713675804,
      "loss": 6.9964,
      "mean_token_accuracy": 0.14108359590172767,
      "num_tokens": 4207970.0,
      "step": 1665
    },
    {
      "entropy": 7.045479440689087,
      "epoch": 0.19272937103289095,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999322626852193,
      "loss": 6.9161,
      "mean_token_accuracy": 0.1442576117813587,
      "num_tokens": 4220893.0,
      "step": 1670
    },
    {
      "entropy": 7.002764892578125,
      "epoch": 0.1933064050778996,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999312464373734,
      "loss": 6.8981,
      "mean_token_accuracy": 0.13980508148670195,
      "num_tokens": 4232463.0,
      "step": 1675
    },
    {
      "entropy": 7.067962741851806,
      "epoch": 0.19388343912290826,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999302226240767,
      "loss": 6.8634,
      "mean_token_accuracy": 0.1389753997325897,
      "num_tokens": 4244604.0,
      "step": 1680
    },
    {
      "entropy": 7.014446783065796,
      "epoch": 0.1944604731679169,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004999291912453637,
      "loss": 6.9458,
      "mean_token_accuracy": 0.14495371729135514,
      "num_tokens": 4257260.0,
      "step": 1685
    },
    {
      "entropy": 7.177657508850098,
      "epoch": 0.19503750721292556,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999281523012691,
      "loss": 6.9024,
      "mean_token_accuracy": 0.13922334164381028,
      "num_tokens": 4269885.0,
      "step": 1690
    },
    {
      "entropy": 7.044678115844727,
      "epoch": 0.1956145412579342,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999271057918278,
      "loss": 7.0126,
      "mean_token_accuracy": 0.13741706758737565,
      "num_tokens": 4281560.0,
      "step": 1695
    },
    {
      "entropy": 7.127787733078003,
      "epoch": 0.19619157530294287,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999260517170751,
      "loss": 6.8977,
      "mean_token_accuracy": 0.14665021896362304,
      "num_tokens": 4293746.0,
      "step": 1700
    },
    {
      "entropy": 7.0871453285217285,
      "epoch": 0.19676860934795154,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999249900770463,
      "loss": 6.9355,
      "mean_token_accuracy": 0.1397123083472252,
      "num_tokens": 4306595.0,
      "step": 1705
    },
    {
      "entropy": 7.123568153381347,
      "epoch": 0.19734564339296018,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999239208717772,
      "loss": 6.8848,
      "mean_token_accuracy": 0.14523950070142747,
      "num_tokens": 4318705.0,
      "step": 1710
    },
    {
      "entropy": 6.925599145889282,
      "epoch": 0.19792267743796885,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999228441013037,
      "loss": 6.9293,
      "mean_token_accuracy": 0.14079967737197877,
      "num_tokens": 4331321.0,
      "step": 1715
    },
    {
      "entropy": 7.063972091674804,
      "epoch": 0.1984997114829775,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499921759765662,
      "loss": 6.8839,
      "mean_token_accuracy": 0.14801763668656348,
      "num_tokens": 4343517.0,
      "step": 1720
    },
    {
      "entropy": 7.0327746868133545,
      "epoch": 0.19907674552798615,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004999206678648888,
      "loss": 6.9032,
      "mean_token_accuracy": 0.14797452092170715,
      "num_tokens": 4356342.0,
      "step": 1725
    },
    {
      "entropy": 6.983541250228882,
      "epoch": 0.1996537795729948,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999195683990206,
      "loss": 6.8887,
      "mean_token_accuracy": 0.1441090725362301,
      "num_tokens": 4369078.0,
      "step": 1730
    },
    {
      "entropy": 7.019094848632813,
      "epoch": 0.20023081361800346,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999184613680945,
      "loss": 6.8854,
      "mean_token_accuracy": 0.13662683367729186,
      "num_tokens": 4382091.0,
      "step": 1735
    },
    {
      "entropy": 7.010761690139771,
      "epoch": 0.20080784766301213,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999173467721476,
      "loss": 6.9208,
      "mean_token_accuracy": 0.1393657937645912,
      "num_tokens": 4393919.0,
      "step": 1740
    },
    {
      "entropy": 7.006711435317993,
      "epoch": 0.20138488170802077,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004999162246112175,
      "loss": 6.8789,
      "mean_token_accuracy": 0.1454889804124832,
      "num_tokens": 4407054.0,
      "step": 1745
    },
    {
      "entropy": 7.003733777999878,
      "epoch": 0.20196191575302944,
      "grad_norm": 1.375,
      "learning_rate": 0.0004999150948853419,
      "loss": 6.94,
      "mean_token_accuracy": 0.137217216193676,
      "num_tokens": 4420180.0,
      "step": 1750
    },
    {
      "entropy": 7.14780068397522,
      "epoch": 0.20253894979803808,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999139575945587,
      "loss": 6.9242,
      "mean_token_accuracy": 0.150433312356472,
      "num_tokens": 4433248.0,
      "step": 1755
    },
    {
      "entropy": 6.930108451843262,
      "epoch": 0.20311598384304674,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999128127389065,
      "loss": 6.9106,
      "mean_token_accuracy": 0.14306253790855408,
      "num_tokens": 4446340.0,
      "step": 1760
    },
    {
      "entropy": 7.059494829177856,
      "epoch": 0.20369301788805538,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999116603184233,
      "loss": 6.9249,
      "mean_token_accuracy": 0.140160171687603,
      "num_tokens": 4459918.0,
      "step": 1765
    },
    {
      "entropy": 6.981036758422851,
      "epoch": 0.20427005193306405,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999105003331482,
      "loss": 6.8797,
      "mean_token_accuracy": 0.1448831871151924,
      "num_tokens": 4472020.0,
      "step": 1770
    },
    {
      "entropy": 7.015055799484253,
      "epoch": 0.20484708597807272,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999093327831202,
      "loss": 6.8445,
      "mean_token_accuracy": 0.14826457872986792,
      "num_tokens": 4484269.0,
      "step": 1775
    },
    {
      "entropy": 6.998699283599853,
      "epoch": 0.20542412002308136,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999081576683784,
      "loss": 6.9411,
      "mean_token_accuracy": 0.13955088704824448,
      "num_tokens": 4498040.0,
      "step": 1780
    },
    {
      "entropy": 7.111977243423462,
      "epoch": 0.20600115406809003,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999069749889626,
      "loss": 6.8938,
      "mean_token_accuracy": 0.14736564159393312,
      "num_tokens": 4509701.0,
      "step": 1785
    },
    {
      "entropy": 6.967676115036011,
      "epoch": 0.20657818811309867,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999057847449123,
      "loss": 6.9249,
      "mean_token_accuracy": 0.13947466611862183,
      "num_tokens": 4521840.0,
      "step": 1790
    },
    {
      "entropy": 7.0544905185699465,
      "epoch": 0.20715522215810733,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999045869362678,
      "loss": 6.9169,
      "mean_token_accuracy": 0.1435271330177784,
      "num_tokens": 4534241.0,
      "step": 1795
    },
    {
      "entropy": 7.035863065719605,
      "epoch": 0.20773225620311597,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499903381563069,
      "loss": 6.8669,
      "mean_token_accuracy": 0.1447395406663418,
      "num_tokens": 4547255.0,
      "step": 1800
    },
    {
      "entropy": 7.061683273315429,
      "epoch": 0.20830929024812464,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999021686253568,
      "loss": 6.902,
      "mean_token_accuracy": 0.14312518313527106,
      "num_tokens": 4559538.0,
      "step": 1805
    },
    {
      "entropy": 6.928671979904175,
      "epoch": 0.2088863242931333,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999009481231719,
      "loss": 6.8802,
      "mean_token_accuracy": 0.14481844902038574,
      "num_tokens": 4571444.0,
      "step": 1810
    },
    {
      "entropy": 6.93266978263855,
      "epoch": 0.20946335833814195,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004998997200565553,
      "loss": 6.7182,
      "mean_token_accuracy": 0.15741085708141328,
      "num_tokens": 4586294.0,
      "step": 1815
    },
    {
      "entropy": 7.03538408279419,
      "epoch": 0.21004039238315061,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004998984844255483,
      "loss": 6.8886,
      "mean_token_accuracy": 0.14464563131332397,
      "num_tokens": 4599661.0,
      "step": 1820
    },
    {
      "entropy": 6.98320894241333,
      "epoch": 0.21061742642815925,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004998972412301925,
      "loss": 6.8964,
      "mean_token_accuracy": 0.14106034338474274,
      "num_tokens": 4611305.0,
      "step": 1825
    },
    {
      "entropy": 7.015894365310669,
      "epoch": 0.21119446047316792,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998959904705297,
      "loss": 6.9004,
      "mean_token_accuracy": 0.13522190302610398,
      "num_tokens": 4623992.0,
      "step": 1830
    },
    {
      "entropy": 7.014666414260864,
      "epoch": 0.21177149451817656,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998947321466021,
      "loss": 6.8942,
      "mean_token_accuracy": 0.1404628872871399,
      "num_tokens": 4636495.0,
      "step": 1835
    },
    {
      "entropy": 6.996430206298828,
      "epoch": 0.21234852856318523,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004998934662584518,
      "loss": 6.8858,
      "mean_token_accuracy": 0.14315241277217866,
      "num_tokens": 4650266.0,
      "step": 1840
    },
    {
      "entropy": 7.023065519332886,
      "epoch": 0.2129255626081939,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998921928061214,
      "loss": 6.8519,
      "mean_token_accuracy": 0.14534742385149002,
      "num_tokens": 4663238.0,
      "step": 1845
    },
    {
      "entropy": 6.998432397842407,
      "epoch": 0.21350259665320254,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998909117896539,
      "loss": 6.8968,
      "mean_token_accuracy": 0.14261196106672286,
      "num_tokens": 4675501.0,
      "step": 1850
    },
    {
      "entropy": 6.991396522521972,
      "epoch": 0.2140796306982112,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998896232090922,
      "loss": 6.8619,
      "mean_token_accuracy": 0.13769274204969406,
      "num_tokens": 4688996.0,
      "step": 1855
    },
    {
      "entropy": 7.000917768478393,
      "epoch": 0.21465666474321984,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004998883270644798,
      "loss": 6.9589,
      "mean_token_accuracy": 0.138343658298254,
      "num_tokens": 4703226.0,
      "step": 1860
    },
    {
      "entropy": 7.035671234130859,
      "epoch": 0.2152336987882285,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004998870233558602,
      "loss": 6.8455,
      "mean_token_accuracy": 0.14217291474342347,
      "num_tokens": 4715825.0,
      "step": 1865
    },
    {
      "entropy": 6.937395477294922,
      "epoch": 0.21581073283323715,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998857120832774,
      "loss": 6.842,
      "mean_token_accuracy": 0.14556335806846618,
      "num_tokens": 4727819.0,
      "step": 1870
    },
    {
      "entropy": 7.042612791061401,
      "epoch": 0.21638776687824582,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004998843932467751,
      "loss": 6.875,
      "mean_token_accuracy": 0.14139556288719177,
      "num_tokens": 4742890.0,
      "step": 1875
    },
    {
      "entropy": 6.9743971824646,
      "epoch": 0.21696480092325446,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998830668463982,
      "loss": 6.8641,
      "mean_token_accuracy": 0.1403570555150509,
      "num_tokens": 4755609.0,
      "step": 1880
    },
    {
      "entropy": 6.923630332946777,
      "epoch": 0.21754183496826313,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004998817328821909,
      "loss": 6.842,
      "mean_token_accuracy": 0.14689554423093795,
      "num_tokens": 4766674.0,
      "step": 1885
    },
    {
      "entropy": 6.97991976737976,
      "epoch": 0.2181188690132718,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998803913541983,
      "loss": 6.8786,
      "mean_token_accuracy": 0.14042821601033212,
      "num_tokens": 4779639.0,
      "step": 1890
    },
    {
      "entropy": 7.112171745300293,
      "epoch": 0.21869590305828043,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998790422624654,
      "loss": 6.9663,
      "mean_token_accuracy": 0.13750633150339125,
      "num_tokens": 4791702.0,
      "step": 1895
    },
    {
      "entropy": 6.947701549530029,
      "epoch": 0.2192729371032891,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998776856070376,
      "loss": 6.7836,
      "mean_token_accuracy": 0.15090153217315674,
      "num_tokens": 4805756.0,
      "step": 1900
    },
    {
      "entropy": 6.948662805557251,
      "epoch": 0.21984997114829774,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004998763213879606,
      "loss": 6.843,
      "mean_token_accuracy": 0.145341220498085,
      "num_tokens": 4817941.0,
      "step": 1905
    },
    {
      "entropy": 6.965862226486206,
      "epoch": 0.2204270051933064,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998749496052803,
      "loss": 6.7269,
      "mean_token_accuracy": 0.14902952387928964,
      "num_tokens": 4832298.0,
      "step": 1910
    },
    {
      "entropy": 6.848330736160278,
      "epoch": 0.22100403923831505,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998735702590426,
      "loss": 6.8195,
      "mean_token_accuracy": 0.1508714735507965,
      "num_tokens": 4844814.0,
      "step": 1915
    },
    {
      "entropy": 7.058192014694214,
      "epoch": 0.22158107328332372,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998721833492942,
      "loss": 6.8856,
      "mean_token_accuracy": 0.1431492656469345,
      "num_tokens": 4857888.0,
      "step": 1920
    },
    {
      "entropy": 6.893492078781128,
      "epoch": 0.22215810732833238,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998707888760816,
      "loss": 6.7919,
      "mean_token_accuracy": 0.14359032064676286,
      "num_tokens": 4870479.0,
      "step": 1925
    },
    {
      "entropy": 6.971369791030884,
      "epoch": 0.22273514137334102,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998693868394516,
      "loss": 6.7938,
      "mean_token_accuracy": 0.1480065330862999,
      "num_tokens": 4883294.0,
      "step": 1930
    },
    {
      "entropy": 7.05402398109436,
      "epoch": 0.2233121754183497,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004998679772394516,
      "loss": 6.9319,
      "mean_token_accuracy": 0.13698863312602044,
      "num_tokens": 4896331.0,
      "step": 1935
    },
    {
      "entropy": 6.9416258335113525,
      "epoch": 0.22388920946335833,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004998665600761289,
      "loss": 6.8675,
      "mean_token_accuracy": 0.14291590824723244,
      "num_tokens": 4908857.0,
      "step": 1940
    },
    {
      "entropy": 6.996976089477539,
      "epoch": 0.224466243508367,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499865135349531,
      "loss": 6.8545,
      "mean_token_accuracy": 0.14505003690719603,
      "num_tokens": 4921564.0,
      "step": 1945
    },
    {
      "entropy": 6.967352771759034,
      "epoch": 0.22504327755337564,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004998637030597061,
      "loss": 6.8762,
      "mean_token_accuracy": 0.14787980914115906,
      "num_tokens": 4934960.0,
      "step": 1950
    },
    {
      "entropy": 6.854097366333008,
      "epoch": 0.2256203115983843,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998622632067022,
      "loss": 6.7272,
      "mean_token_accuracy": 0.15127141177654266,
      "num_tokens": 4946942.0,
      "step": 1955
    },
    {
      "entropy": 6.964528608322143,
      "epoch": 0.22619734564339297,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998608157905678,
      "loss": 6.8454,
      "mean_token_accuracy": 0.13881651386618615,
      "num_tokens": 4960751.0,
      "step": 1960
    },
    {
      "entropy": 6.9480382919311525,
      "epoch": 0.2267743796884016,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998593608113515,
      "loss": 6.8417,
      "mean_token_accuracy": 0.14182508513331413,
      "num_tokens": 4973551.0,
      "step": 1965
    },
    {
      "entropy": 6.950195169448852,
      "epoch": 0.22735141373341028,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998578982691024,
      "loss": 6.87,
      "mean_token_accuracy": 0.14090688973665239,
      "num_tokens": 4986224.0,
      "step": 1970
    },
    {
      "entropy": 6.947819328308105,
      "epoch": 0.22792844777841892,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004998564281638694,
      "loss": 6.8368,
      "mean_token_accuracy": 0.14499269127845765,
      "num_tokens": 4999874.0,
      "step": 1975
    },
    {
      "entropy": 6.946592903137207,
      "epoch": 0.2285054818234276,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004998549504957023,
      "loss": 6.8382,
      "mean_token_accuracy": 0.1452639691531658,
      "num_tokens": 5013106.0,
      "step": 1980
    },
    {
      "entropy": 6.935692501068115,
      "epoch": 0.22908251586843623,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004998534652646506,
      "loss": 6.8023,
      "mean_token_accuracy": 0.14065282493829728,
      "num_tokens": 5026145.0,
      "step": 1985
    },
    {
      "entropy": 6.941158723831177,
      "epoch": 0.2296595499134449,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004998519724707642,
      "loss": 6.8182,
      "mean_token_accuracy": 0.14938395619392394,
      "num_tokens": 5039216.0,
      "step": 1990
    },
    {
      "entropy": 6.966537952423096,
      "epoch": 0.23023658395845356,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004998504721140934,
      "loss": 6.795,
      "mean_token_accuracy": 0.14359883442521096,
      "num_tokens": 5051956.0,
      "step": 1995
    },
    {
      "entropy": 6.932427787780762,
      "epoch": 0.2308136180034622,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004998489641946887,
      "loss": 6.8764,
      "mean_token_accuracy": 0.14441612362861633,
      "num_tokens": 5065325.0,
      "step": 2000
    },
    {
      "entropy": 6.94489541053772,
      "epoch": 0.23139065204847087,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998474487126009,
      "loss": 6.875,
      "mean_token_accuracy": 0.14884231835603715,
      "num_tokens": 5077776.0,
      "step": 2005
    },
    {
      "entropy": 6.97218074798584,
      "epoch": 0.2319676860934795,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998459256678806,
      "loss": 6.8698,
      "mean_token_accuracy": 0.14031682312488555,
      "num_tokens": 5091167.0,
      "step": 2010
    },
    {
      "entropy": 7.01816759109497,
      "epoch": 0.23254472013848818,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004998443950605796,
      "loss": 6.8068,
      "mean_token_accuracy": 0.14413101896643638,
      "num_tokens": 5103321.0,
      "step": 2015
    },
    {
      "entropy": 6.925113582611084,
      "epoch": 0.23312175418349682,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004998428568907488,
      "loss": 6.8077,
      "mean_token_accuracy": 0.14342234432697296,
      "num_tokens": 5116999.0,
      "step": 2020
    },
    {
      "entropy": 6.899875020980835,
      "epoch": 0.23369878822850548,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998413111584403,
      "loss": 6.803,
      "mean_token_accuracy": 0.14782628118991853,
      "num_tokens": 5130429.0,
      "step": 2025
    },
    {
      "entropy": 6.895891237258911,
      "epoch": 0.23427582227351415,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998397578637059,
      "loss": 6.7544,
      "mean_token_accuracy": 0.14964505434036254,
      "num_tokens": 5142728.0,
      "step": 2030
    },
    {
      "entropy": 6.994881963729858,
      "epoch": 0.2348528563185228,
      "grad_norm": 0.859375,
      "learning_rate": 0.000499838197006598,
      "loss": 6.854,
      "mean_token_accuracy": 0.14518777132034302,
      "num_tokens": 5155105.0,
      "step": 2035
    },
    {
      "entropy": 6.9891077995300295,
      "epoch": 0.23542989036353146,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000499836628587169,
      "loss": 6.8814,
      "mean_token_accuracy": 0.13513262197375298,
      "num_tokens": 5169206.0,
      "step": 2040
    },
    {
      "entropy": 6.860302734375,
      "epoch": 0.2360069244085401,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004998350526054716,
      "loss": 6.7535,
      "mean_token_accuracy": 0.1476128190755844,
      "num_tokens": 5182007.0,
      "step": 2045
    },
    {
      "entropy": 6.977161026000976,
      "epoch": 0.23658395845354876,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004998334690615591,
      "loss": 6.7814,
      "mean_token_accuracy": 0.14299210011959076,
      "num_tokens": 5195137.0,
      "step": 2050
    },
    {
      "entropy": 6.875006675720215,
      "epoch": 0.2371609924985574,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004998318779554844,
      "loss": 6.7999,
      "mean_token_accuracy": 0.14403216391801835,
      "num_tokens": 5207713.0,
      "step": 2055
    },
    {
      "entropy": 6.91954345703125,
      "epoch": 0.23773802654356607,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998302792873012,
      "loss": 6.7539,
      "mean_token_accuracy": 0.1439061000943184,
      "num_tokens": 5220125.0,
      "step": 2060
    },
    {
      "entropy": 6.926095628738404,
      "epoch": 0.23831506058857474,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004998286730570631,
      "loss": 6.8071,
      "mean_token_accuracy": 0.1458573669195175,
      "num_tokens": 5232142.0,
      "step": 2065
    },
    {
      "entropy": 6.897360849380493,
      "epoch": 0.23889209463358338,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998270592648245,
      "loss": 6.7912,
      "mean_token_accuracy": 0.15026629641652106,
      "num_tokens": 5245517.0,
      "step": 2070
    },
    {
      "entropy": 6.937630748748779,
      "epoch": 0.23946912867859205,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004998254379106394,
      "loss": 6.7863,
      "mean_token_accuracy": 0.14699850231409073,
      "num_tokens": 5257862.0,
      "step": 2075
    },
    {
      "entropy": 6.865565061569214,
      "epoch": 0.2400461627236007,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998238089945622,
      "loss": 6.7348,
      "mean_token_accuracy": 0.15117276906967164,
      "num_tokens": 5270141.0,
      "step": 2080
    },
    {
      "entropy": 6.878887939453125,
      "epoch": 0.24062319676860935,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004998221725166479,
      "loss": 6.7593,
      "mean_token_accuracy": 0.15452364832162857,
      "num_tokens": 5283143.0,
      "step": 2085
    },
    {
      "entropy": 6.871646356582642,
      "epoch": 0.241200230813618,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004998205284769516,
      "loss": 6.8235,
      "mean_token_accuracy": 0.13959436044096946,
      "num_tokens": 5295340.0,
      "step": 2090
    },
    {
      "entropy": 6.964116764068604,
      "epoch": 0.24177726485862666,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998188768755285,
      "loss": 6.7683,
      "mean_token_accuracy": 0.14472750276327134,
      "num_tokens": 5306821.0,
      "step": 2095
    },
    {
      "entropy": 6.834314155578613,
      "epoch": 0.2423542989036353,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004998172177124341,
      "loss": 6.7865,
      "mean_token_accuracy": 0.14368025809526444,
      "num_tokens": 5319428.0,
      "step": 2100
    },
    {
      "entropy": 7.012720727920533,
      "epoch": 0.24293133294864397,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004998155509877242,
      "loss": 6.8145,
      "mean_token_accuracy": 0.13865589946508408,
      "num_tokens": 5331777.0,
      "step": 2105
    },
    {
      "entropy": 6.713190269470215,
      "epoch": 0.24350836699365264,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499813876701455,
      "loss": 6.7742,
      "mean_token_accuracy": 0.14583554863929749,
      "num_tokens": 5344533.0,
      "step": 2110
    },
    {
      "entropy": 6.8935747146606445,
      "epoch": 0.24408540103866128,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998121948536828,
      "loss": 6.7137,
      "mean_token_accuracy": 0.15403580144047738,
      "num_tokens": 5357123.0,
      "step": 2115
    },
    {
      "entropy": 6.924336910247803,
      "epoch": 0.24466243508366994,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998105054444639,
      "loss": 6.8329,
      "mean_token_accuracy": 0.13986173048615455,
      "num_tokens": 5370123.0,
      "step": 2120
    },
    {
      "entropy": 6.897236251831055,
      "epoch": 0.24523946912867858,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004998088084738555,
      "loss": 6.7763,
      "mean_token_accuracy": 0.15366744846105576,
      "num_tokens": 5382870.0,
      "step": 2125
    },
    {
      "entropy": 6.823340272903442,
      "epoch": 0.24581650317368725,
      "grad_norm": 0.875,
      "learning_rate": 0.0004998071039419144,
      "loss": 6.7582,
      "mean_token_accuracy": 0.14249610602855683,
      "num_tokens": 5395052.0,
      "step": 2130
    },
    {
      "entropy": 6.9604456424713135,
      "epoch": 0.2463935372186959,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000499805391848698,
      "loss": 6.7837,
      "mean_token_accuracy": 0.1490890696644783,
      "num_tokens": 5406720.0,
      "step": 2135
    },
    {
      "entropy": 6.8235283374786375,
      "epoch": 0.24697057126370456,
      "grad_norm": 0.828125,
      "learning_rate": 0.000499803672194264,
      "loss": 6.7838,
      "mean_token_accuracy": 0.14721129089593887,
      "num_tokens": 5419465.0,
      "step": 2140
    },
    {
      "entropy": 6.8882369041442875,
      "epoch": 0.24754760530871323,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004998019449786701,
      "loss": 6.7894,
      "mean_token_accuracy": 0.14069402888417243,
      "num_tokens": 5432434.0,
      "step": 2145
    },
    {
      "entropy": 6.860111331939697,
      "epoch": 0.24812463935372187,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998002102019744,
      "loss": 6.7522,
      "mean_token_accuracy": 0.15137931406497956,
      "num_tokens": 5445249.0,
      "step": 2150
    },
    {
      "entropy": 6.983528804779053,
      "epoch": 0.24870167339873053,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004997984678642354,
      "loss": 6.7916,
      "mean_token_accuracy": 0.1442440040409565,
      "num_tokens": 5458996.0,
      "step": 2155
    },
    {
      "entropy": 6.853938817977905,
      "epoch": 0.24927870744373917,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004997967179655115,
      "loss": 6.7744,
      "mean_token_accuracy": 0.14349499493837356,
      "num_tokens": 5472174.0,
      "step": 2160
    },
    {
      "entropy": 6.808877658843994,
      "epoch": 0.24985574148874784,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004997949605058617,
      "loss": 6.7069,
      "mean_token_accuracy": 0.15367118269205093,
      "num_tokens": 5484464.0,
      "step": 2165
    },
    {
      "entropy": 6.934421110153198,
      "epoch": 0.2504327755337565,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997931954853451,
      "loss": 6.8061,
      "mean_token_accuracy": 0.14632220715284347,
      "num_tokens": 5497278.0,
      "step": 2170
    },
    {
      "entropy": 6.86837854385376,
      "epoch": 0.2510098095787651,
      "grad_norm": 0.890625,
      "learning_rate": 0.000499791422904021,
      "loss": 6.8287,
      "mean_token_accuracy": 0.1460747793316841,
      "num_tokens": 5510847.0,
      "step": 2175
    },
    {
      "entropy": 6.842457818984985,
      "epoch": 0.2515868436237738,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997896427619491,
      "loss": 6.7482,
      "mean_token_accuracy": 0.14972207546234131,
      "num_tokens": 5522543.0,
      "step": 2180
    },
    {
      "entropy": 6.892645740509034,
      "epoch": 0.25216387766878245,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004997878550591892,
      "loss": 6.7552,
      "mean_token_accuracy": 0.1436435416340828,
      "num_tokens": 5534856.0,
      "step": 2185
    },
    {
      "entropy": 6.831497383117676,
      "epoch": 0.2527409117137911,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004997860597958013,
      "loss": 6.6717,
      "mean_token_accuracy": 0.1602167546749115,
      "num_tokens": 5546881.0,
      "step": 2190
    },
    {
      "entropy": 6.895463514328003,
      "epoch": 0.2533179457587998,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004997842569718461,
      "loss": 6.7887,
      "mean_token_accuracy": 0.14391066506505013,
      "num_tokens": 5559711.0,
      "step": 2195
    },
    {
      "entropy": 6.865831708908081,
      "epoch": 0.25389497980380843,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000499782446587384,
      "loss": 6.7843,
      "mean_token_accuracy": 0.1506132774055004,
      "num_tokens": 5573553.0,
      "step": 2200
    },
    {
      "entropy": 6.900459480285645,
      "epoch": 0.25447201384881707,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000499780628642476,
      "loss": 6.7603,
      "mean_token_accuracy": 0.13870777040719987,
      "num_tokens": 5586046.0,
      "step": 2205
    },
    {
      "entropy": 6.854037666320801,
      "epoch": 0.2550490478938257,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997788031371834,
      "loss": 6.7603,
      "mean_token_accuracy": 0.14891895428299903,
      "num_tokens": 5597876.0,
      "step": 2210
    },
    {
      "entropy": 6.933327293395996,
      "epoch": 0.2556260819388344,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997769700715672,
      "loss": 6.6716,
      "mean_token_accuracy": 0.1543987885117531,
      "num_tokens": 5611264.0,
      "step": 2215
    },
    {
      "entropy": 6.804627752304077,
      "epoch": 0.25620311598384304,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004997751294456892,
      "loss": 6.7679,
      "mean_token_accuracy": 0.1464955747127533,
      "num_tokens": 5625093.0,
      "step": 2220
    },
    {
      "entropy": 6.852876806259156,
      "epoch": 0.2567801500288517,
      "grad_norm": 0.875,
      "learning_rate": 0.0004997732812596114,
      "loss": 6.7744,
      "mean_token_accuracy": 0.14379966408014297,
      "num_tokens": 5637206.0,
      "step": 2225
    },
    {
      "entropy": 6.857150936126709,
      "epoch": 0.2573571840738604,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004997714255133961,
      "loss": 6.6563,
      "mean_token_accuracy": 0.15202945172786714,
      "num_tokens": 5649136.0,
      "step": 2230
    },
    {
      "entropy": 6.837795639038086,
      "epoch": 0.257934218118869,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997695622071054,
      "loss": 6.6956,
      "mean_token_accuracy": 0.15131831020116807,
      "num_tokens": 5660714.0,
      "step": 2235
    },
    {
      "entropy": 6.87720136642456,
      "epoch": 0.25851125216387766,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997676913408021,
      "loss": 6.7654,
      "mean_token_accuracy": 0.14454615712165833,
      "num_tokens": 5673295.0,
      "step": 2240
    },
    {
      "entropy": 6.8868872165679935,
      "epoch": 0.2590882862088863,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004997658129145493,
      "loss": 6.777,
      "mean_token_accuracy": 0.14726119190454484,
      "num_tokens": 5686822.0,
      "step": 2245
    },
    {
      "entropy": 6.793786144256591,
      "epoch": 0.259665320253895,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997639269284099,
      "loss": 6.8295,
      "mean_token_accuracy": 0.1436148166656494,
      "num_tokens": 5700621.0,
      "step": 2250
    },
    {
      "entropy": 6.9439198017120365,
      "epoch": 0.26024235429890363,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997620333824476,
      "loss": 6.7379,
      "mean_token_accuracy": 0.15637929141521453,
      "num_tokens": 5713565.0,
      "step": 2255
    },
    {
      "entropy": 6.800307655334473,
      "epoch": 0.2608193883439123,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004997601322767258,
      "loss": 6.783,
      "mean_token_accuracy": 0.1416866213083267,
      "num_tokens": 5726487.0,
      "step": 2260
    },
    {
      "entropy": 6.966365957260132,
      "epoch": 0.26139642238892097,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004997582236113087,
      "loss": 6.8581,
      "mean_token_accuracy": 0.1411590591073036,
      "num_tokens": 5739934.0,
      "step": 2265
    },
    {
      "entropy": 6.865359163284301,
      "epoch": 0.2619734564339296,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004997563073862603,
      "loss": 6.8048,
      "mean_token_accuracy": 0.13894407004117965,
      "num_tokens": 5752187.0,
      "step": 2270
    },
    {
      "entropy": 6.836117839813232,
      "epoch": 0.26255049047893825,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004997543836016453,
      "loss": 6.7345,
      "mean_token_accuracy": 0.14709190279245377,
      "num_tokens": 5764850.0,
      "step": 2275
    },
    {
      "entropy": 6.9021507740020756,
      "epoch": 0.2631275245239469,
      "grad_norm": 0.71484375,
      "learning_rate": 0.0004997524522575281,
      "loss": 6.8019,
      "mean_token_accuracy": 0.14260414987802505,
      "num_tokens": 5778577.0,
      "step": 2280
    },
    {
      "entropy": 6.944571781158447,
      "epoch": 0.2637045585689556,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004997505133539738,
      "loss": 6.7487,
      "mean_token_accuracy": 0.1511955589056015,
      "num_tokens": 5791699.0,
      "step": 2285
    },
    {
      "entropy": 6.796835660934448,
      "epoch": 0.2642815926139642,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004997485668910476,
      "loss": 6.7555,
      "mean_token_accuracy": 0.14254892766475677,
      "num_tokens": 5804553.0,
      "step": 2290
    },
    {
      "entropy": 6.993517017364502,
      "epoch": 0.26485862665897286,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997466128688151,
      "loss": 6.7579,
      "mean_token_accuracy": 0.14862642660737038,
      "num_tokens": 5816758.0,
      "step": 2295
    },
    {
      "entropy": 6.853436231613159,
      "epoch": 0.26543566070398156,
      "grad_norm": 0.828125,
      "learning_rate": 0.000499744651287342,
      "loss": 6.7812,
      "mean_token_accuracy": 0.13678565695881845,
      "num_tokens": 5831048.0,
      "step": 2300
    },
    {
      "entropy": 6.919308710098266,
      "epoch": 0.2660126947489902,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000499742682146694,
      "loss": 6.7909,
      "mean_token_accuracy": 0.13784368932247162,
      "num_tokens": 5843159.0,
      "step": 2305
    },
    {
      "entropy": 6.807005167007446,
      "epoch": 0.26658972879399884,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004997407054469377,
      "loss": 6.6791,
      "mean_token_accuracy": 0.15264758318662644,
      "num_tokens": 5856078.0,
      "step": 2310
    },
    {
      "entropy": 6.854742860794067,
      "epoch": 0.2671667628390075,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004997387211881392,
      "loss": 6.706,
      "mean_token_accuracy": 0.14707566574215888,
      "num_tokens": 5867835.0,
      "step": 2315
    },
    {
      "entropy": 6.8351672172546385,
      "epoch": 0.26774379688401617,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004997367293703656,
      "loss": 6.7444,
      "mean_token_accuracy": 0.13954102098941804,
      "num_tokens": 5881179.0,
      "step": 2320
    },
    {
      "entropy": 6.878024482727051,
      "epoch": 0.2683208309290248,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004997347299936837,
      "loss": 6.7195,
      "mean_token_accuracy": 0.14241515398025512,
      "num_tokens": 5892774.0,
      "step": 2325
    },
    {
      "entropy": 6.870539665222168,
      "epoch": 0.26889786497403345,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004997327230581608,
      "loss": 6.756,
      "mean_token_accuracy": 0.1463689960539341,
      "num_tokens": 5905857.0,
      "step": 2330
    },
    {
      "entropy": 6.805995798110962,
      "epoch": 0.26947489901904215,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004997307085638644,
      "loss": 6.7166,
      "mean_token_accuracy": 0.14379769265651704,
      "num_tokens": 5918128.0,
      "step": 2335
    },
    {
      "entropy": 6.918767642974854,
      "epoch": 0.2700519330640508,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004997286865108621,
      "loss": 6.8232,
      "mean_token_accuracy": 0.14087387174367905,
      "num_tokens": 5930606.0,
      "step": 2340
    },
    {
      "entropy": 6.841729688644409,
      "epoch": 0.2706289671090594,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004997266568992222,
      "loss": 6.7696,
      "mean_token_accuracy": 0.13774231299757958,
      "num_tokens": 5944139.0,
      "step": 2345
    },
    {
      "entropy": 6.887184953689575,
      "epoch": 0.27120600115406807,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004997246197290128,
      "loss": 6.7646,
      "mean_token_accuracy": 0.14507462382316588,
      "num_tokens": 5956779.0,
      "step": 2350
    },
    {
      "entropy": 6.837090969085693,
      "epoch": 0.27178303519907676,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997225750003024,
      "loss": 6.7083,
      "mean_token_accuracy": 0.15230057314038276,
      "num_tokens": 5971510.0,
      "step": 2355
    },
    {
      "entropy": 6.695552968978882,
      "epoch": 0.2723600692440854,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004997205227131599,
      "loss": 6.7371,
      "mean_token_accuracy": 0.1434318020939827,
      "num_tokens": 5983877.0,
      "step": 2360
    },
    {
      "entropy": 6.974739694595337,
      "epoch": 0.27293710328909404,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004997184628676542,
      "loss": 6.7617,
      "mean_token_accuracy": 0.14533916413784026,
      "num_tokens": 5996739.0,
      "step": 2365
    },
    {
      "entropy": 6.768356561660767,
      "epoch": 0.27351413733410274,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004997163954638546,
      "loss": 6.7255,
      "mean_token_accuracy": 0.15101923793554306,
      "num_tokens": 6009289.0,
      "step": 2370
    },
    {
      "entropy": 6.87177038192749,
      "epoch": 0.2740911713791114,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997143205018306,
      "loss": 6.794,
      "mean_token_accuracy": 0.14188004285097122,
      "num_tokens": 6021375.0,
      "step": 2375
    },
    {
      "entropy": 6.898721408843994,
      "epoch": 0.27466820542412,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004997122379816523,
      "loss": 6.7271,
      "mean_token_accuracy": 0.14577853456139564,
      "num_tokens": 6034182.0,
      "step": 2380
    },
    {
      "entropy": 6.774568223953247,
      "epoch": 0.27524523946912866,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004997101479033893,
      "loss": 6.7104,
      "mean_token_accuracy": 0.1492339551448822,
      "num_tokens": 6046486.0,
      "step": 2385
    },
    {
      "entropy": 6.845656013488769,
      "epoch": 0.27582227351413735,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004997080502671122,
      "loss": 6.7644,
      "mean_token_accuracy": 0.14610921069979668,
      "num_tokens": 6060867.0,
      "step": 2390
    },
    {
      "entropy": 6.7973473072052,
      "epoch": 0.276399307559146,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004997059450728913,
      "loss": 6.6264,
      "mean_token_accuracy": 0.15345287024974824,
      "num_tokens": 6073000.0,
      "step": 2395
    },
    {
      "entropy": 6.76253228187561,
      "epoch": 0.27697634160415463,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004997038323207977,
      "loss": 6.6906,
      "mean_token_accuracy": 0.14774474650621414,
      "num_tokens": 6085884.0,
      "step": 2400
    },
    {
      "entropy": 6.787846565246582,
      "epoch": 0.2775533756491633,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004997017120109022,
      "loss": 6.6146,
      "mean_token_accuracy": 0.15057797580957413,
      "num_tokens": 6098230.0,
      "step": 2405
    },
    {
      "entropy": 6.782491827011109,
      "epoch": 0.27813040969417197,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004996995841432762,
      "loss": 6.6701,
      "mean_token_accuracy": 0.15104399472475052,
      "num_tokens": 6111568.0,
      "step": 2410
    },
    {
      "entropy": 6.779537677764893,
      "epoch": 0.2787074437391806,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004996974487179915,
      "loss": 6.646,
      "mean_token_accuracy": 0.14741345196962358,
      "num_tokens": 6123992.0,
      "step": 2415
    },
    {
      "entropy": 6.796364402770996,
      "epoch": 0.27928447778418924,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004996953057351195,
      "loss": 6.686,
      "mean_token_accuracy": 0.1465798109769821,
      "num_tokens": 6135452.0,
      "step": 2420
    },
    {
      "entropy": 6.774402904510498,
      "epoch": 0.27986151182919794,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004996931551947327,
      "loss": 6.5866,
      "mean_token_accuracy": 0.15778433084487914,
      "num_tokens": 6150144.0,
      "step": 2425
    },
    {
      "entropy": 6.837385368347168,
      "epoch": 0.2804385458742066,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004996909970969031,
      "loss": 6.7329,
      "mean_token_accuracy": 0.14868736639618874,
      "num_tokens": 6163397.0,
      "step": 2430
    },
    {
      "entropy": 6.832604503631591,
      "epoch": 0.2810155799192152,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004996888314417034,
      "loss": 6.7233,
      "mean_token_accuracy": 0.14309920221567154,
      "num_tokens": 6175579.0,
      "step": 2435
    },
    {
      "entropy": 6.790614795684815,
      "epoch": 0.2815926139642239,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004996866582292067,
      "loss": 6.7063,
      "mean_token_accuracy": 0.15070078819990157,
      "num_tokens": 6189229.0,
      "step": 2440
    },
    {
      "entropy": 6.777169132232666,
      "epoch": 0.28216964800923255,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996844774594856,
      "loss": 6.6934,
      "mean_token_accuracy": 0.1502875879406929,
      "num_tokens": 6203817.0,
      "step": 2445
    },
    {
      "entropy": 6.82915768623352,
      "epoch": 0.2827466820542412,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004996822891326138,
      "loss": 6.7206,
      "mean_token_accuracy": 0.14171916842460633,
      "num_tokens": 6216098.0,
      "step": 2450
    },
    {
      "entropy": 6.812518548965454,
      "epoch": 0.28332371609924983,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004996800932486648,
      "loss": 6.7166,
      "mean_token_accuracy": 0.14603719413280486,
      "num_tokens": 6228987.0,
      "step": 2455
    },
    {
      "entropy": 6.742992639541626,
      "epoch": 0.28390075014425853,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004996778898077126,
      "loss": 6.6685,
      "mean_token_accuracy": 0.15222399532794953,
      "num_tokens": 6241676.0,
      "step": 2460
    },
    {
      "entropy": 6.889522314071655,
      "epoch": 0.28447778418926717,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004996756788098309,
      "loss": 6.702,
      "mean_token_accuracy": 0.14694569408893585,
      "num_tokens": 6254744.0,
      "step": 2465
    },
    {
      "entropy": 6.758044672012329,
      "epoch": 0.2850548182342758,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004996734602550945,
      "loss": 6.6596,
      "mean_token_accuracy": 0.1465069644153118,
      "num_tokens": 6268242.0,
      "step": 2470
    },
    {
      "entropy": 6.747747182846069,
      "epoch": 0.2856318522792845,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004996712341435779,
      "loss": 6.5867,
      "mean_token_accuracy": 0.1545226328074932,
      "num_tokens": 6279904.0,
      "step": 2475
    },
    {
      "entropy": 6.745885562896729,
      "epoch": 0.28620888632429314,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004996690004753559,
      "loss": 6.7018,
      "mean_token_accuracy": 0.14820884317159652,
      "num_tokens": 6292885.0,
      "step": 2480
    },
    {
      "entropy": 6.753550338745117,
      "epoch": 0.2867859203693018,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004996667592505037,
      "loss": 6.6887,
      "mean_token_accuracy": 0.15229137539863585,
      "num_tokens": 6306278.0,
      "step": 2485
    },
    {
      "entropy": 6.884258556365967,
      "epoch": 0.2873629544143104,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004996645104690967,
      "loss": 6.7179,
      "mean_token_accuracy": 0.13899706602096557,
      "num_tokens": 6320360.0,
      "step": 2490
    },
    {
      "entropy": 6.700476408004761,
      "epoch": 0.2879399884593191,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004996622541312103,
      "loss": 6.7035,
      "mean_token_accuracy": 0.144860952347517,
      "num_tokens": 6332778.0,
      "step": 2495
    },
    {
      "entropy": 6.817806053161621,
      "epoch": 0.28851702250432776,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004996599902369207,
      "loss": 6.6808,
      "mean_token_accuracy": 0.14956472143530847,
      "num_tokens": 6345028.0,
      "step": 2500
    },
    {
      "entropy": 6.786530876159668,
      "epoch": 0.2890940565493364,
      "grad_norm": 0.875,
      "learning_rate": 0.0004996577187863039,
      "loss": 6.6425,
      "mean_token_accuracy": 0.15257902592420577,
      "num_tokens": 6356931.0,
      "step": 2505
    },
    {
      "entropy": 6.698649883270264,
      "epoch": 0.2896710905943451,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004996554397794364,
      "loss": 6.5883,
      "mean_token_accuracy": 0.15760106444358826,
      "num_tokens": 6370247.0,
      "step": 2510
    },
    {
      "entropy": 6.7704665184021,
      "epoch": 0.29024812463935373,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004996531532163947,
      "loss": 6.6603,
      "mean_token_accuracy": 0.14782440811395645,
      "num_tokens": 6382352.0,
      "step": 2515
    },
    {
      "entropy": 6.754695749282837,
      "epoch": 0.2908251586843624,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996508590972558,
      "loss": 6.6397,
      "mean_token_accuracy": 0.14761101007461547,
      "num_tokens": 6395949.0,
      "step": 2520
    },
    {
      "entropy": 6.754686546325684,
      "epoch": 0.291402192729371,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004996485574220969,
      "loss": 6.6647,
      "mean_token_accuracy": 0.15044545978307725,
      "num_tokens": 6408469.0,
      "step": 2525
    },
    {
      "entropy": 6.768384695053101,
      "epoch": 0.2919792267743797,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004996462481909953,
      "loss": 6.6579,
      "mean_token_accuracy": 0.1553287237882614,
      "num_tokens": 6419953.0,
      "step": 2530
    },
    {
      "entropy": 6.790286922454834,
      "epoch": 0.29255626081938835,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996439314040287,
      "loss": 6.7094,
      "mean_token_accuracy": 0.143075567483902,
      "num_tokens": 6432567.0,
      "step": 2535
    },
    {
      "entropy": 6.826392459869385,
      "epoch": 0.293133294864397,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499641607061275,
      "loss": 6.6549,
      "mean_token_accuracy": 0.1453825443983078,
      "num_tokens": 6445951.0,
      "step": 2540
    },
    {
      "entropy": 6.742476511001587,
      "epoch": 0.2937103289094057,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996392751628126,
      "loss": 6.7032,
      "mean_token_accuracy": 0.15503485202789308,
      "num_tokens": 6458799.0,
      "step": 2545
    },
    {
      "entropy": 6.815782499313355,
      "epoch": 0.2942873629544143,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996369357087196,
      "loss": 6.7065,
      "mean_token_accuracy": 0.1430730052292347,
      "num_tokens": 6471423.0,
      "step": 2550
    },
    {
      "entropy": 6.799359512329102,
      "epoch": 0.29486439699942296,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499634588699075,
      "loss": 6.7392,
      "mean_token_accuracy": 0.14322351813316345,
      "num_tokens": 6484248.0,
      "step": 2555
    },
    {
      "entropy": 6.833653163909912,
      "epoch": 0.2954414310444316,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004996322341339575,
      "loss": 6.6637,
      "mean_token_accuracy": 0.14406396448612213,
      "num_tokens": 6497048.0,
      "step": 2560
    },
    {
      "entropy": 6.775566720962525,
      "epoch": 0.2960184650894403,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004996298720134465,
      "loss": 6.6848,
      "mean_token_accuracy": 0.14442287534475326,
      "num_tokens": 6508840.0,
      "step": 2565
    },
    {
      "entropy": 6.7831347465515135,
      "epoch": 0.29659549913444894,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004996275023376213,
      "loss": 6.6583,
      "mean_token_accuracy": 0.1537433847784996,
      "num_tokens": 6520976.0,
      "step": 2570
    },
    {
      "entropy": 6.741522121429443,
      "epoch": 0.2971725331794576,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004996251251065615,
      "loss": 6.673,
      "mean_token_accuracy": 0.15090716034173965,
      "num_tokens": 6533009.0,
      "step": 2575
    },
    {
      "entropy": 6.822439098358155,
      "epoch": 0.2977495672244662,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996227403203473,
      "loss": 6.6662,
      "mean_token_accuracy": 0.14721700698137283,
      "num_tokens": 6545196.0,
      "step": 2580
    },
    {
      "entropy": 6.812913274765014,
      "epoch": 0.2983266012694749,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004996203479790589,
      "loss": 6.7017,
      "mean_token_accuracy": 0.1475054517388344,
      "num_tokens": 6557479.0,
      "step": 2585
    },
    {
      "entropy": 6.819405889511108,
      "epoch": 0.29890363531448355,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996179480827764,
      "loss": 6.7114,
      "mean_token_accuracy": 0.14923257529735565,
      "num_tokens": 6569832.0,
      "step": 2590
    },
    {
      "entropy": 6.738608884811401,
      "epoch": 0.2994806693594922,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000499615540631581,
      "loss": 6.6737,
      "mean_token_accuracy": 0.14806562215089797,
      "num_tokens": 6581651.0,
      "step": 2595
    },
    {
      "entropy": 6.774818849563599,
      "epoch": 0.3000577034045009,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004996131256255534,
      "loss": 6.7423,
      "mean_token_accuracy": 0.14710890799760817,
      "num_tokens": 6596331.0,
      "step": 2600
    },
    {
      "entropy": 6.817371892929077,
      "epoch": 0.3006347374495095,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004996107030647749,
      "loss": 6.6685,
      "mean_token_accuracy": 0.14770287573337554,
      "num_tokens": 6609712.0,
      "step": 2605
    },
    {
      "entropy": 6.759057664871216,
      "epoch": 0.30121177149451817,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004996082729493269,
      "loss": 6.6962,
      "mean_token_accuracy": 0.1502513587474823,
      "num_tokens": 6622780.0,
      "step": 2610
    },
    {
      "entropy": 6.784829711914062,
      "epoch": 0.3017888055395268,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004996058352792913,
      "loss": 6.7561,
      "mean_token_accuracy": 0.1468655101954937,
      "num_tokens": 6636288.0,
      "step": 2615
    },
    {
      "entropy": 6.858647060394287,
      "epoch": 0.3023658395845355,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00049960339005475,
      "loss": 6.7313,
      "mean_token_accuracy": 0.1382056713104248,
      "num_tokens": 6649883.0,
      "step": 2620
    },
    {
      "entropy": 6.787996578216553,
      "epoch": 0.30294287362954414,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004996009372757852,
      "loss": 6.6803,
      "mean_token_accuracy": 0.14866930097341538,
      "num_tokens": 6662241.0,
      "step": 2625
    },
    {
      "entropy": 6.705225419998169,
      "epoch": 0.3035199076745528,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004995984769424795,
      "loss": 6.6389,
      "mean_token_accuracy": 0.14856984168291093,
      "num_tokens": 6674790.0,
      "step": 2630
    },
    {
      "entropy": 6.879620838165283,
      "epoch": 0.3040969417195615,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004995960090549155,
      "loss": 6.6451,
      "mean_token_accuracy": 0.14790016412734985,
      "num_tokens": 6687713.0,
      "step": 2635
    },
    {
      "entropy": 6.629987955093384,
      "epoch": 0.3046739757645701,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004995935336131764,
      "loss": 6.6574,
      "mean_token_accuracy": 0.1474847599864006,
      "num_tokens": 6701123.0,
      "step": 2640
    },
    {
      "entropy": 6.796099090576172,
      "epoch": 0.30525100980957875,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004995910506173452,
      "loss": 6.6029,
      "mean_token_accuracy": 0.15525631308555604,
      "num_tokens": 6712674.0,
      "step": 2645
    },
    {
      "entropy": 6.721958589553833,
      "epoch": 0.3058280438545874,
      "grad_norm": 0.875,
      "learning_rate": 0.0004995885600675059,
      "loss": 6.601,
      "mean_token_accuracy": 0.15212517231702805,
      "num_tokens": 6724840.0,
      "step": 2650
    },
    {
      "entropy": 6.760172510147095,
      "epoch": 0.3064050778995961,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004995860619637415,
      "loss": 6.7425,
      "mean_token_accuracy": 0.13689624965190889,
      "num_tokens": 6738746.0,
      "step": 2655
    },
    {
      "entropy": 6.875043773651123,
      "epoch": 0.30698211194460473,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004995835563061367,
      "loss": 6.6436,
      "mean_token_accuracy": 0.1482662469148636,
      "num_tokens": 6751620.0,
      "step": 2660
    },
    {
      "entropy": 6.727645397186279,
      "epoch": 0.30755914598961337,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004995810430947756,
      "loss": 6.6337,
      "mean_token_accuracy": 0.1513223595917225,
      "num_tokens": 6763957.0,
      "step": 2665
    },
    {
      "entropy": 6.559813404083252,
      "epoch": 0.30813618003462206,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004995785223297426,
      "loss": 6.5468,
      "mean_token_accuracy": 0.1547765851020813,
      "num_tokens": 6776211.0,
      "step": 2670
    },
    {
      "entropy": 6.862458086013794,
      "epoch": 0.3087132140796307,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004995759940111225,
      "loss": 6.786,
      "mean_token_accuracy": 0.14168819040060043,
      "num_tokens": 6789745.0,
      "step": 2675
    },
    {
      "entropy": 6.805125570297241,
      "epoch": 0.30929024812463934,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004995734581390005,
      "loss": 6.7054,
      "mean_token_accuracy": 0.14849800616502762,
      "num_tokens": 6801771.0,
      "step": 2680
    },
    {
      "entropy": 6.708049964904785,
      "epoch": 0.309867282169648,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004995709147134617,
      "loss": 6.6365,
      "mean_token_accuracy": 0.14683517664670945,
      "num_tokens": 6813993.0,
      "step": 2685
    },
    {
      "entropy": 6.818436479568481,
      "epoch": 0.3104443162146567,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004995683637345919,
      "loss": 6.7144,
      "mean_token_accuracy": 0.14922749251127243,
      "num_tokens": 6826833.0,
      "step": 2690
    },
    {
      "entropy": 6.754158878326416,
      "epoch": 0.3110213502596653,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004995658052024765,
      "loss": 6.6478,
      "mean_token_accuracy": 0.1457843691110611,
      "num_tokens": 6838947.0,
      "step": 2695
    },
    {
      "entropy": 6.80776834487915,
      "epoch": 0.31159838430467396,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004995632391172019,
      "loss": 6.7147,
      "mean_token_accuracy": 0.14534788727760314,
      "num_tokens": 6851781.0,
      "step": 2700
    },
    {
      "entropy": 6.8032268524169925,
      "epoch": 0.31217541834968265,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995606654788543,
      "loss": 6.601,
      "mean_token_accuracy": 0.1524400979280472,
      "num_tokens": 6863479.0,
      "step": 2705
    },
    {
      "entropy": 6.714714527130127,
      "epoch": 0.3127524523946913,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004995580842875202,
      "loss": 6.593,
      "mean_token_accuracy": 0.15386273488402366,
      "num_tokens": 6877000.0,
      "step": 2710
    },
    {
      "entropy": 6.8048097610473635,
      "epoch": 0.31332948643969993,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995554955432866,
      "loss": 6.6981,
      "mean_token_accuracy": 0.14284373223781585,
      "num_tokens": 6888878.0,
      "step": 2715
    },
    {
      "entropy": 6.769490671157837,
      "epoch": 0.3139065204847086,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004995528992462404,
      "loss": 6.6196,
      "mean_token_accuracy": 0.14831875115633011,
      "num_tokens": 6902147.0,
      "step": 2720
    },
    {
      "entropy": 6.750906562805175,
      "epoch": 0.31448355452971727,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000499550295396469,
      "loss": 6.6661,
      "mean_token_accuracy": 0.15283936485648156,
      "num_tokens": 6914927.0,
      "step": 2725
    },
    {
      "entropy": 6.761933422088623,
      "epoch": 0.3150605885747259,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004995476839940598,
      "loss": 6.6683,
      "mean_token_accuracy": 0.1505434513092041,
      "num_tokens": 6927757.0,
      "step": 2730
    },
    {
      "entropy": 6.762182760238647,
      "epoch": 0.31563762261973455,
      "grad_norm": 0.79296875,
      "learning_rate": 0.000499545065039101,
      "loss": 6.6725,
      "mean_token_accuracy": 0.1466390758752823,
      "num_tokens": 6941039.0,
      "step": 2735
    },
    {
      "entropy": 6.781767749786377,
      "epoch": 0.31621465666474324,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004995424385316803,
      "loss": 6.7293,
      "mean_token_accuracy": 0.14698702692985535,
      "num_tokens": 6954876.0,
      "step": 2740
    },
    {
      "entropy": 6.725474309921265,
      "epoch": 0.3167916907097519,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004995398044718863,
      "loss": 6.5811,
      "mean_token_accuracy": 0.1516340285539627,
      "num_tokens": 6966889.0,
      "step": 2745
    },
    {
      "entropy": 6.753961610794067,
      "epoch": 0.3173687247547605,
      "grad_norm": 0.875,
      "learning_rate": 0.0004995371628598074,
      "loss": 6.6151,
      "mean_token_accuracy": 0.1514896869659424,
      "num_tokens": 6979894.0,
      "step": 2750
    },
    {
      "entropy": 6.722421312332154,
      "epoch": 0.31794575879976916,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004995345136955328,
      "loss": 6.6356,
      "mean_token_accuracy": 0.1484585776925087,
      "num_tokens": 6992734.0,
      "step": 2755
    },
    {
      "entropy": 6.755990600585937,
      "epoch": 0.31852279284477786,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004995318569791512,
      "loss": 6.663,
      "mean_token_accuracy": 0.14797065705060958,
      "num_tokens": 7006504.0,
      "step": 2760
    },
    {
      "entropy": 6.7291755199432375,
      "epoch": 0.3190998268897865,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004995291927107522,
      "loss": 6.5971,
      "mean_token_accuracy": 0.15441082864999772,
      "num_tokens": 7018978.0,
      "step": 2765
    },
    {
      "entropy": 6.710022735595703,
      "epoch": 0.31967686093479514,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0004995265208904252,
      "loss": 6.6868,
      "mean_token_accuracy": 0.1457875519990921,
      "num_tokens": 7031294.0,
      "step": 2770
    },
    {
      "entropy": 6.80343861579895,
      "epoch": 0.32025389497980383,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004995238415182603,
      "loss": 6.6386,
      "mean_token_accuracy": 0.1497604802250862,
      "num_tokens": 7044139.0,
      "step": 2775
    },
    {
      "entropy": 6.643225526809692,
      "epoch": 0.3208309290248125,
      "grad_norm": 0.875,
      "learning_rate": 0.0004995211545943476,
      "loss": 6.5431,
      "mean_token_accuracy": 0.15468174815177918,
      "num_tokens": 7056232.0,
      "step": 2780
    },
    {
      "entropy": 6.770214605331421,
      "epoch": 0.3214079630698211,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004995184601187772,
      "loss": 6.6978,
      "mean_token_accuracy": 0.1458260402083397,
      "num_tokens": 7070125.0,
      "step": 2785
    },
    {
      "entropy": 6.735332918167114,
      "epoch": 0.32198499711482975,
      "grad_norm": 0.8125,
      "learning_rate": 0.00049951575809164,
      "loss": 6.5678,
      "mean_token_accuracy": 0.15374045222997665,
      "num_tokens": 7083535.0,
      "step": 2790
    },
    {
      "entropy": 6.679257535934449,
      "epoch": 0.32256203115983845,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004995130485130269,
      "loss": 6.537,
      "mean_token_accuracy": 0.15215112119913102,
      "num_tokens": 7095081.0,
      "step": 2795
    },
    {
      "entropy": 6.756899642944336,
      "epoch": 0.3231390652048471,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004995103313830289,
      "loss": 6.6676,
      "mean_token_accuracy": 0.14167121201753616,
      "num_tokens": 7107764.0,
      "step": 2800
    },
    {
      "entropy": 6.7260784149169925,
      "epoch": 0.3237160992498557,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004995076067017373,
      "loss": 6.6009,
      "mean_token_accuracy": 0.1534109964966774,
      "num_tokens": 7119754.0,
      "step": 2805
    },
    {
      "entropy": 6.6373598098754885,
      "epoch": 0.3242931332948644,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004995048744692439,
      "loss": 6.6299,
      "mean_token_accuracy": 0.15635569840669633,
      "num_tokens": 7132087.0,
      "step": 2810
    },
    {
      "entropy": 6.691185474395752,
      "epoch": 0.32487016733987306,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004995021346856405,
      "loss": 6.6053,
      "mean_token_accuracy": 0.14959825202822685,
      "num_tokens": 7143325.0,
      "step": 2815
    },
    {
      "entropy": 6.7681121826171875,
      "epoch": 0.3254472013848817,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004994993873510196,
      "loss": 6.5619,
      "mean_token_accuracy": 0.15113845020532607,
      "num_tokens": 7154771.0,
      "step": 2820
    },
    {
      "entropy": 6.670676565170288,
      "epoch": 0.32602423542989034,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004994966324654731,
      "loss": 6.6585,
      "mean_token_accuracy": 0.14671303182840348,
      "num_tokens": 7168352.0,
      "step": 2825
    },
    {
      "entropy": 6.619832849502563,
      "epoch": 0.32660126947489904,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000499493870029094,
      "loss": 6.4456,
      "mean_token_accuracy": 0.16549365520477294,
      "num_tokens": 7180618.0,
      "step": 2830
    },
    {
      "entropy": 6.737293481826782,
      "epoch": 0.3271783035199077,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004994911000419749,
      "loss": 6.6031,
      "mean_token_accuracy": 0.14624983444809914,
      "num_tokens": 7192634.0,
      "step": 2835
    },
    {
      "entropy": 6.660342025756836,
      "epoch": 0.3277553375649163,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004994883225042093,
      "loss": 6.6186,
      "mean_token_accuracy": 0.15046972185373306,
      "num_tokens": 7205839.0,
      "step": 2840
    },
    {
      "entropy": 6.736796808242798,
      "epoch": 0.328332371609925,
      "grad_norm": 0.73828125,
      "learning_rate": 0.0004994855374158905,
      "loss": 6.627,
      "mean_token_accuracy": 0.15292316824197769,
      "num_tokens": 7218371.0,
      "step": 2845
    },
    {
      "entropy": 6.6943048477172855,
      "epoch": 0.32890940565493365,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000499482744777112,
      "loss": 6.4322,
      "mean_token_accuracy": 0.1693370446562767,
      "num_tokens": 7231506.0,
      "step": 2850
    },
    {
      "entropy": 6.649769687652588,
      "epoch": 0.3294864396999423,
      "grad_norm": 0.8125,
      "learning_rate": 0.000499479944587968,
      "loss": 6.5888,
      "mean_token_accuracy": 0.1511973798274994,
      "num_tokens": 7243088.0,
      "step": 2855
    },
    {
      "entropy": 6.694813299179077,
      "epoch": 0.33006347374495093,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004994771368485526,
      "loss": 6.5163,
      "mean_token_accuracy": 0.1597664922475815,
      "num_tokens": 7255383.0,
      "step": 2860
    },
    {
      "entropy": 6.647224283218383,
      "epoch": 0.3306405077899596,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004994743215589602,
      "loss": 6.608,
      "mean_token_accuracy": 0.15329548567533494,
      "num_tokens": 7268270.0,
      "step": 2865
    },
    {
      "entropy": 6.673927307128906,
      "epoch": 0.33121754183496827,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004994714987192854,
      "loss": 6.5362,
      "mean_token_accuracy": 0.1510409414768219,
      "num_tokens": 7281108.0,
      "step": 2870
    },
    {
      "entropy": 6.702598619461059,
      "epoch": 0.3317945758799769,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004994686683296234,
      "loss": 6.5981,
      "mean_token_accuracy": 0.14916437938809396,
      "num_tokens": 7293961.0,
      "step": 2875
    },
    {
      "entropy": 6.701837539672852,
      "epoch": 0.3323716099249856,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004994658303900691,
      "loss": 6.6155,
      "mean_token_accuracy": 0.15184523910284042,
      "num_tokens": 7306574.0,
      "step": 2880
    },
    {
      "entropy": 6.728485918045044,
      "epoch": 0.33294864396999424,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004994629849007182,
      "loss": 6.6466,
      "mean_token_accuracy": 0.14797799810767173,
      "num_tokens": 7319046.0,
      "step": 2885
    },
    {
      "entropy": 6.79762864112854,
      "epoch": 0.3335256780150029,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004994601318616663,
      "loss": 6.64,
      "mean_token_accuracy": 0.14297626465559005,
      "num_tokens": 7332033.0,
      "step": 2890
    },
    {
      "entropy": 6.6803630828857425,
      "epoch": 0.3341027120600115,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004994572712730092,
      "loss": 6.6512,
      "mean_token_accuracy": 0.14246861189603804,
      "num_tokens": 7344658.0,
      "step": 2895
    },
    {
      "entropy": 6.816131925582885,
      "epoch": 0.3346797461050202,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004994544031348434,
      "loss": 6.5279,
      "mean_token_accuracy": 0.15453597754240037,
      "num_tokens": 7357252.0,
      "step": 2900
    },
    {
      "entropy": 6.603551912307739,
      "epoch": 0.33525678015002885,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004994515274472653,
      "loss": 6.4869,
      "mean_token_accuracy": 0.1610173389315605,
      "num_tokens": 7369280.0,
      "step": 2905
    },
    {
      "entropy": 6.595618057250976,
      "epoch": 0.3358338141950375,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004994486442103715,
      "loss": 6.4916,
      "mean_token_accuracy": 0.1597081333398819,
      "num_tokens": 7383017.0,
      "step": 2910
    },
    {
      "entropy": 6.698597812652588,
      "epoch": 0.3364108482400462,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499445753424259,
      "loss": 6.6324,
      "mean_token_accuracy": 0.1377485029399395,
      "num_tokens": 7395634.0,
      "step": 2915
    },
    {
      "entropy": 6.708563899993896,
      "epoch": 0.33698788228505483,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004994428550890251,
      "loss": 6.601,
      "mean_token_accuracy": 0.15198962539434432,
      "num_tokens": 7408590.0,
      "step": 2920
    },
    {
      "entropy": 6.720327949523925,
      "epoch": 0.33756491633006347,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004994399492047672,
      "loss": 6.5912,
      "mean_token_accuracy": 0.1499352663755417,
      "num_tokens": 7420202.0,
      "step": 2925
    },
    {
      "entropy": 6.670965766906738,
      "epoch": 0.3381419503750721,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004994370357715832,
      "loss": 6.5504,
      "mean_token_accuracy": 0.16089099049568176,
      "num_tokens": 7432664.0,
      "step": 2930
    },
    {
      "entropy": 6.661357879638672,
      "epoch": 0.3387189844200808,
      "grad_norm": 0.75390625,
      "learning_rate": 0.000499434114789571,
      "loss": 6.7051,
      "mean_token_accuracy": 0.1464277185499668,
      "num_tokens": 7446084.0,
      "step": 2935
    },
    {
      "entropy": 6.714499425888062,
      "epoch": 0.33929601846508944,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004994311862588286,
      "loss": 6.5508,
      "mean_token_accuracy": 0.15420828610658646,
      "num_tokens": 7458314.0,
      "step": 2940
    },
    {
      "entropy": 6.676367473602295,
      "epoch": 0.3398730525100981,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004994282501794551,
      "loss": 6.5995,
      "mean_token_accuracy": 0.15544719249010086,
      "num_tokens": 7471995.0,
      "step": 2945
    },
    {
      "entropy": 6.703731489181519,
      "epoch": 0.3404500865551068,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004994253065515486,
      "loss": 6.622,
      "mean_token_accuracy": 0.1454193964600563,
      "num_tokens": 7485676.0,
      "step": 2950
    },
    {
      "entropy": 6.740993881225586,
      "epoch": 0.3410271206001154,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004994223553752084,
      "loss": 6.5061,
      "mean_token_accuracy": 0.15575395077466964,
      "num_tokens": 7497057.0,
      "step": 2955
    },
    {
      "entropy": 6.6203021049499515,
      "epoch": 0.34160415464512406,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004994193966505339,
      "loss": 6.5331,
      "mean_token_accuracy": 0.15374020338058472,
      "num_tokens": 7510625.0,
      "step": 2960
    },
    {
      "entropy": 6.623828363418579,
      "epoch": 0.3421811886901327,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004994164303776244,
      "loss": 6.5414,
      "mean_token_accuracy": 0.15313206017017364,
      "num_tokens": 7522677.0,
      "step": 2965
    },
    {
      "entropy": 6.613079309463501,
      "epoch": 0.3427582227351414,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004994134565565797,
      "loss": 6.4729,
      "mean_token_accuracy": 0.15777821689844132,
      "num_tokens": 7535406.0,
      "step": 2970
    },
    {
      "entropy": 6.725028848648071,
      "epoch": 0.34333525678015003,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004994104751875,
      "loss": 6.6259,
      "mean_token_accuracy": 0.1496363550424576,
      "num_tokens": 7547577.0,
      "step": 2975
    },
    {
      "entropy": 6.7461694240570065,
      "epoch": 0.3439122908251587,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004994074862704854,
      "loss": 6.6151,
      "mean_token_accuracy": 0.15172602161765097,
      "num_tokens": 7561734.0,
      "step": 2980
    },
    {
      "entropy": 6.712504768371582,
      "epoch": 0.3444893248701673,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004994044898056363,
      "loss": 6.6621,
      "mean_token_accuracy": 0.14886820018291474,
      "num_tokens": 7574256.0,
      "step": 2985
    },
    {
      "entropy": 6.752849674224853,
      "epoch": 0.345066358915176,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004994014857930538,
      "loss": 6.5801,
      "mean_token_accuracy": 0.15615265667438508,
      "num_tokens": 7587149.0,
      "step": 2990
    },
    {
      "entropy": 6.642135143280029,
      "epoch": 0.34564339296018465,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004993984742328387,
      "loss": 6.5502,
      "mean_token_accuracy": 0.15114353597164154,
      "num_tokens": 7599124.0,
      "step": 2995
    },
    {
      "entropy": 6.641741514205933,
      "epoch": 0.3462204270051933,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004993954551250925,
      "loss": 6.5556,
      "mean_token_accuracy": 0.1523570418357849,
      "num_tokens": 7612070.0,
      "step": 3000
    },
    {
      "epoch": 0.3462204270051933,
      "eval_entropy": 6.5966810582204065,
      "eval_loss": 6.6427154541015625,
      "eval_mean_token_accuracy": 0.1525940384301639,
      "eval_num_tokens": 7612070.0,
      "eval_runtime": 17.8197,
      "eval_samples_per_second": 1578.922,
      "eval_steps_per_second": 197.365,
      "step": 3000
    },
    {
      "entropy": 6.603645992279053,
      "epoch": 0.346797461050202,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004993924284699165,
      "loss": 6.4531,
      "mean_token_accuracy": 0.15813793241977692,
      "num_tokens": 7626096.0,
      "step": 3005
    },
    {
      "entropy": 6.70902099609375,
      "epoch": 0.3473744950952106,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004993893942674127,
      "loss": 6.5844,
      "mean_token_accuracy": 0.15069538950920106,
      "num_tokens": 7639666.0,
      "step": 3010
    },
    {
      "entropy": 6.644587516784668,
      "epoch": 0.34795152914021926,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499386352517683,
      "loss": 6.5407,
      "mean_token_accuracy": 0.15207042545080185,
      "num_tokens": 7652904.0,
      "step": 3015
    },
    {
      "entropy": 6.669216442108154,
      "epoch": 0.3485285631852279,
      "grad_norm": 0.82421875,
      "learning_rate": 0.00049938330322083,
      "loss": 6.5692,
      "mean_token_accuracy": 0.15089073702692984,
      "num_tokens": 7665523.0,
      "step": 3020
    },
    {
      "entropy": 6.743306303024292,
      "epoch": 0.3491055972302366,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004993802463769558,
      "loss": 6.6148,
      "mean_token_accuracy": 0.14982035011053085,
      "num_tokens": 7677063.0,
      "step": 3025
    },
    {
      "entropy": 6.61863112449646,
      "epoch": 0.34968263127524524,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004993771819861636,
      "loss": 6.5418,
      "mean_token_accuracy": 0.15196796357631684,
      "num_tokens": 7688980.0,
      "step": 3030
    },
    {
      "entropy": 6.773717403411865,
      "epoch": 0.3502596653202539,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004993741100485564,
      "loss": 6.664,
      "mean_token_accuracy": 0.15363381579518318,
      "num_tokens": 7702604.0,
      "step": 3035
    },
    {
      "entropy": 6.7194211959838865,
      "epoch": 0.35083669936526257,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004993710305642374,
      "loss": 6.6045,
      "mean_token_accuracy": 0.148685784637928,
      "num_tokens": 7715324.0,
      "step": 3040
    },
    {
      "entropy": 6.778428649902343,
      "epoch": 0.3514137334102712,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004993679435333102,
      "loss": 6.6311,
      "mean_token_accuracy": 0.15059239715337752,
      "num_tokens": 7729098.0,
      "step": 3045
    },
    {
      "entropy": 6.679904794692993,
      "epoch": 0.35199076745527985,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004993648489558788,
      "loss": 6.5728,
      "mean_token_accuracy": 0.14844775199890137,
      "num_tokens": 7740717.0,
      "step": 3050
    },
    {
      "entropy": 6.585766887664795,
      "epoch": 0.3525678015002885,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004993617468320471,
      "loss": 6.5,
      "mean_token_accuracy": 0.1497762158513069,
      "num_tokens": 7753928.0,
      "step": 3055
    },
    {
      "entropy": 6.741098833084107,
      "epoch": 0.3531448355452972,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004993586371619193,
      "loss": 6.5951,
      "mean_token_accuracy": 0.14959297627210616,
      "num_tokens": 7765728.0,
      "step": 3060
    },
    {
      "entropy": 6.595219230651855,
      "epoch": 0.3537218695903058,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004993555199456004,
      "loss": 6.54,
      "mean_token_accuracy": 0.15218116194009781,
      "num_tokens": 7778199.0,
      "step": 3065
    },
    {
      "entropy": 6.61001672744751,
      "epoch": 0.35429890363531447,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004993523951831949,
      "loss": 6.5172,
      "mean_token_accuracy": 0.16153395771980286,
      "num_tokens": 7791026.0,
      "step": 3070
    },
    {
      "entropy": 6.656833124160767,
      "epoch": 0.35487593768032316,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004993492628748081,
      "loss": 6.5961,
      "mean_token_accuracy": 0.1556033119559288,
      "num_tokens": 7804333.0,
      "step": 3075
    },
    {
      "entropy": 6.677346849441529,
      "epoch": 0.3554529717253318,
      "grad_norm": 0.73046875,
      "learning_rate": 0.0004993461230205453,
      "loss": 6.597,
      "mean_token_accuracy": 0.1483662024140358,
      "num_tokens": 7818339.0,
      "step": 3080
    },
    {
      "entropy": 6.623977136611939,
      "epoch": 0.35603000577034044,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004993429756205121,
      "loss": 6.5718,
      "mean_token_accuracy": 0.15360086113214494,
      "num_tokens": 7831537.0,
      "step": 3085
    },
    {
      "entropy": 6.702095079421997,
      "epoch": 0.3566070398153491,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004993398206748142,
      "loss": 6.576,
      "mean_token_accuracy": 0.15210341215133666,
      "num_tokens": 7843705.0,
      "step": 3090
    },
    {
      "entropy": 6.664427471160889,
      "epoch": 0.3571840738603578,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004993366581835581,
      "loss": 6.5747,
      "mean_token_accuracy": 0.15079948231577872,
      "num_tokens": 7856481.0,
      "step": 3095
    },
    {
      "entropy": 6.682329511642456,
      "epoch": 0.3577611079053664,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004993334881468498,
      "loss": 6.5722,
      "mean_token_accuracy": 0.1513454347848892,
      "num_tokens": 7868732.0,
      "step": 3100
    },
    {
      "entropy": 6.700321531295776,
      "epoch": 0.35833814195037506,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004993303105647961,
      "loss": 6.5369,
      "mean_token_accuracy": 0.1532818630337715,
      "num_tokens": 7881470.0,
      "step": 3105
    },
    {
      "entropy": 6.685085391998291,
      "epoch": 0.35891517599538375,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004993271254375038,
      "loss": 6.6055,
      "mean_token_accuracy": 0.14828379899263383,
      "num_tokens": 7895074.0,
      "step": 3110
    },
    {
      "entropy": 6.695715379714966,
      "epoch": 0.3594922100403924,
      "grad_norm": 0.80078125,
      "learning_rate": 0.00049932393276508,
      "loss": 6.5743,
      "mean_token_accuracy": 0.15514225661754608,
      "num_tokens": 7909140.0,
      "step": 3115
    },
    {
      "entropy": 6.622471666336059,
      "epoch": 0.36006924408540103,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004993207325476323,
      "loss": 6.5438,
      "mean_token_accuracy": 0.15456567257642745,
      "num_tokens": 7922689.0,
      "step": 3120
    },
    {
      "entropy": 6.6618585109710695,
      "epoch": 0.36064627813040967,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004993175247852681,
      "loss": 6.5498,
      "mean_token_accuracy": 0.14956910461187362,
      "num_tokens": 7935207.0,
      "step": 3125
    },
    {
      "entropy": 6.619128084182739,
      "epoch": 0.36122331217541837,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004993143094780954,
      "loss": 6.4961,
      "mean_token_accuracy": 0.15582271665334702,
      "num_tokens": 7947491.0,
      "step": 3130
    },
    {
      "entropy": 6.628898763656617,
      "epoch": 0.361800346220427,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004993110866262224,
      "loss": 6.502,
      "mean_token_accuracy": 0.15252822190523146,
      "num_tokens": 7959329.0,
      "step": 3135
    },
    {
      "entropy": 6.6644287109375,
      "epoch": 0.36237738026543564,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004993078562297573,
      "loss": 6.5191,
      "mean_token_accuracy": 0.15571288764476776,
      "num_tokens": 7971794.0,
      "step": 3140
    },
    {
      "entropy": 6.633200788497925,
      "epoch": 0.36295441431044434,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004993046182888089,
      "loss": 6.658,
      "mean_token_accuracy": 0.15406894385814668,
      "num_tokens": 7983942.0,
      "step": 3145
    },
    {
      "entropy": 6.721062707901001,
      "epoch": 0.363531448355453,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004993013728034861,
      "loss": 6.527,
      "mean_token_accuracy": 0.15496130287647247,
      "num_tokens": 7996266.0,
      "step": 3150
    },
    {
      "entropy": 6.672237873077393,
      "epoch": 0.3641084824004616,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000499298119773898,
      "loss": 6.5901,
      "mean_token_accuracy": 0.14736275672912597,
      "num_tokens": 8009318.0,
      "step": 3155
    },
    {
      "entropy": 6.651838397979736,
      "epoch": 0.36468551644547026,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004992948592001541,
      "loss": 6.5862,
      "mean_token_accuracy": 0.15489773005247115,
      "num_tokens": 8021213.0,
      "step": 3160
    },
    {
      "entropy": 6.635162830352783,
      "epoch": 0.36526255049047895,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000499291591082364,
      "loss": 6.5594,
      "mean_token_accuracy": 0.14926373437047005,
      "num_tokens": 8033927.0,
      "step": 3165
    },
    {
      "entropy": 6.569822597503662,
      "epoch": 0.3658395845354876,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004992883154206377,
      "loss": 6.4402,
      "mean_token_accuracy": 0.15944022238254546,
      "num_tokens": 8046709.0,
      "step": 3170
    },
    {
      "entropy": 6.684438991546631,
      "epoch": 0.36641661858049623,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004992850322150853,
      "loss": 6.5909,
      "mean_token_accuracy": 0.14909223467111588,
      "num_tokens": 8060127.0,
      "step": 3175
    },
    {
      "entropy": 6.648997974395752,
      "epoch": 0.36699365262550493,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004992817414658172,
      "loss": 6.5246,
      "mean_token_accuracy": 0.15754903107881546,
      "num_tokens": 8072927.0,
      "step": 3180
    },
    {
      "entropy": 6.563595008850098,
      "epoch": 0.36757068667051357,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004992784431729442,
      "loss": 6.4497,
      "mean_token_accuracy": 0.15927306860685347,
      "num_tokens": 8083973.0,
      "step": 3185
    },
    {
      "entropy": 6.6668370246887205,
      "epoch": 0.3681477207155222,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004992751373365771,
      "loss": 6.5226,
      "mean_token_accuracy": 0.15365545451641083,
      "num_tokens": 8096801.0,
      "step": 3190
    },
    {
      "entropy": 6.642384719848633,
      "epoch": 0.36872475476053085,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004992718239568273,
      "loss": 6.5395,
      "mean_token_accuracy": 0.15474597811698915,
      "num_tokens": 8109099.0,
      "step": 3195
    },
    {
      "entropy": 6.654610347747803,
      "epoch": 0.36930178880553954,
      "grad_norm": 0.8125,
      "learning_rate": 0.000499268503033806,
      "loss": 6.4807,
      "mean_token_accuracy": 0.15879638940095903,
      "num_tokens": 8120712.0,
      "step": 3200
    },
    {
      "entropy": 6.599427938461304,
      "epoch": 0.3698788228505482,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004992651745676249,
      "loss": 6.5206,
      "mean_token_accuracy": 0.15537554174661636,
      "num_tokens": 8132228.0,
      "step": 3205
    },
    {
      "entropy": 6.6891693592071535,
      "epoch": 0.3704558568955568,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004992618385583962,
      "loss": 6.531,
      "mean_token_accuracy": 0.1528848886489868,
      "num_tokens": 8145015.0,
      "step": 3210
    },
    {
      "entropy": 6.579451227188111,
      "epoch": 0.3710328909405655,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004992584950062319,
      "loss": 6.4827,
      "mean_token_accuracy": 0.1555587813258171,
      "num_tokens": 8156985.0,
      "step": 3215
    },
    {
      "entropy": 6.705874490737915,
      "epoch": 0.37160992498557416,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004992551439112446,
      "loss": 6.6359,
      "mean_token_accuracy": 0.14975506514310838,
      "num_tokens": 8170633.0,
      "step": 3220
    },
    {
      "entropy": 6.702548408508301,
      "epoch": 0.3721869590305828,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004992517852735469,
      "loss": 6.5145,
      "mean_token_accuracy": 0.15716839730739593,
      "num_tokens": 8182096.0,
      "step": 3225
    },
    {
      "entropy": 6.643147325515747,
      "epoch": 0.37276399307559144,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004992484190932517,
      "loss": 6.5591,
      "mean_token_accuracy": 0.1519121289253235,
      "num_tokens": 8193768.0,
      "step": 3230
    },
    {
      "entropy": 6.594755172729492,
      "epoch": 0.37334102712060013,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004992450453704723,
      "loss": 6.4174,
      "mean_token_accuracy": 0.1655030742287636,
      "num_tokens": 8206237.0,
      "step": 3235
    },
    {
      "entropy": 6.629844379425049,
      "epoch": 0.3739180611656088,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004992416641053223,
      "loss": 6.5086,
      "mean_token_accuracy": 0.15595395714044571,
      "num_tokens": 8218066.0,
      "step": 3240
    },
    {
      "entropy": 6.635203266143799,
      "epoch": 0.3744950952106174,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004992382752979153,
      "loss": 6.4969,
      "mean_token_accuracy": 0.1581435188651085,
      "num_tokens": 8230976.0,
      "step": 3245
    },
    {
      "entropy": 6.595430040359497,
      "epoch": 0.3750721292556261,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004992348789483651,
      "loss": 6.4836,
      "mean_token_accuracy": 0.16031598895788193,
      "num_tokens": 8242713.0,
      "step": 3250
    },
    {
      "entropy": 6.556427621841431,
      "epoch": 0.37564916330063475,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004992314750567864,
      "loss": 6.4254,
      "mean_token_accuracy": 0.1674373909831047,
      "num_tokens": 8255698.0,
      "step": 3255
    },
    {
      "entropy": 6.6029973983764645,
      "epoch": 0.3762261973456434,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004992280636232933,
      "loss": 6.5007,
      "mean_token_accuracy": 0.158811055123806,
      "num_tokens": 8267138.0,
      "step": 3260
    },
    {
      "entropy": 6.55572772026062,
      "epoch": 0.376803231390652,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004992246446480007,
      "loss": 6.5147,
      "mean_token_accuracy": 0.1620226725935936,
      "num_tokens": 8280902.0,
      "step": 3265
    },
    {
      "entropy": 6.678423690795898,
      "epoch": 0.3773802654356607,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004992212181310235,
      "loss": 6.5346,
      "mean_token_accuracy": 0.1485668234527111,
      "num_tokens": 8293027.0,
      "step": 3270
    },
    {
      "entropy": 6.711181497573852,
      "epoch": 0.37795729948066936,
      "grad_norm": 0.828125,
      "learning_rate": 0.000499217784072477,
      "loss": 6.5666,
      "mean_token_accuracy": 0.1553458794951439,
      "num_tokens": 8306676.0,
      "step": 3275
    },
    {
      "entropy": 6.473035192489624,
      "epoch": 0.378534333525678,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004992143424724768,
      "loss": 6.5013,
      "mean_token_accuracy": 0.16312037110328675,
      "num_tokens": 8319134.0,
      "step": 3280
    },
    {
      "entropy": 6.6510356903076175,
      "epoch": 0.3791113675706867,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004992108933311385,
      "loss": 6.471,
      "mean_token_accuracy": 0.15111429542303084,
      "num_tokens": 8333010.0,
      "step": 3285
    },
    {
      "entropy": 6.661249351501465,
      "epoch": 0.37968840161569534,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004992074366485782,
      "loss": 6.5438,
      "mean_token_accuracy": 0.15512855350971222,
      "num_tokens": 8345576.0,
      "step": 3290
    },
    {
      "entropy": 6.524981737136841,
      "epoch": 0.380265435660704,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000499203972424912,
      "loss": 6.5412,
      "mean_token_accuracy": 0.1497179873287678,
      "num_tokens": 8359278.0,
      "step": 3295
    },
    {
      "entropy": 6.744177532196045,
      "epoch": 0.3808424697057126,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004992005006602567,
      "loss": 6.5289,
      "mean_token_accuracy": 0.1496485650539398,
      "num_tokens": 8372159.0,
      "step": 3300
    },
    {
      "entropy": 6.55145263671875,
      "epoch": 0.3814195037507213,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004991970213547289,
      "loss": 6.5775,
      "mean_token_accuracy": 0.14963392093777655,
      "num_tokens": 8385735.0,
      "step": 3305
    },
    {
      "entropy": 6.705130910873413,
      "epoch": 0.38199653779572995,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004991935345084457,
      "loss": 6.5483,
      "mean_token_accuracy": 0.14926708191633226,
      "num_tokens": 8397714.0,
      "step": 3310
    },
    {
      "entropy": 6.710420894622803,
      "epoch": 0.3825735718407386,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004991900401215243,
      "loss": 6.5794,
      "mean_token_accuracy": 0.14717367589473723,
      "num_tokens": 8410812.0,
      "step": 3315
    },
    {
      "entropy": 6.6759387969970705,
      "epoch": 0.3831506058857473,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004991865381940822,
      "loss": 6.5506,
      "mean_token_accuracy": 0.15251373499631882,
      "num_tokens": 8422400.0,
      "step": 3320
    },
    {
      "entropy": 6.678918695449829,
      "epoch": 0.3837276399307559,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004991830287262374,
      "loss": 6.5629,
      "mean_token_accuracy": 0.1532184287905693,
      "num_tokens": 8435080.0,
      "step": 3325
    },
    {
      "entropy": 6.6434855461120605,
      "epoch": 0.38430467397576457,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004991795117181077,
      "loss": 6.4836,
      "mean_token_accuracy": 0.16186368614435195,
      "num_tokens": 8447866.0,
      "step": 3330
    },
    {
      "entropy": 6.498814868927002,
      "epoch": 0.3848817080207732,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004991759871698113,
      "loss": 6.4477,
      "mean_token_accuracy": 0.16394371688365936,
      "num_tokens": 8461447.0,
      "step": 3335
    },
    {
      "entropy": 6.702191495895386,
      "epoch": 0.3854587420657819,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004991724550814671,
      "loss": 6.6218,
      "mean_token_accuracy": 0.15160019621253013,
      "num_tokens": 8475389.0,
      "step": 3340
    },
    {
      "entropy": 6.667648410797119,
      "epoch": 0.38603577611079054,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004991689154531937,
      "loss": 6.5244,
      "mean_token_accuracy": 0.15132362991571427,
      "num_tokens": 8487271.0,
      "step": 3345
    },
    {
      "entropy": 6.618583345413208,
      "epoch": 0.3866128101557992,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004991653682851102,
      "loss": 6.5175,
      "mean_token_accuracy": 0.15368050336837769,
      "num_tokens": 8498914.0,
      "step": 3350
    },
    {
      "entropy": 6.5969428539276125,
      "epoch": 0.3871898442008078,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499161813577336,
      "loss": 6.5066,
      "mean_token_accuracy": 0.15542698353528978,
      "num_tokens": 8510254.0,
      "step": 3355
    },
    {
      "entropy": 6.588715410232544,
      "epoch": 0.3877668782458165,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004991582513299903,
      "loss": 6.5463,
      "mean_token_accuracy": 0.15643467903137206,
      "num_tokens": 8522612.0,
      "step": 3360
    },
    {
      "entropy": 6.603983736038208,
      "epoch": 0.38834391229082516,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004991546815431932,
      "loss": 6.4701,
      "mean_token_accuracy": 0.15623842924833298,
      "num_tokens": 8534952.0,
      "step": 3365
    },
    {
      "entropy": 6.585997819900513,
      "epoch": 0.3889209463358338,
      "grad_norm": 0.8125,
      "learning_rate": 0.000499151104217065,
      "loss": 6.5975,
      "mean_token_accuracy": 0.15297566056251527,
      "num_tokens": 8548388.0,
      "step": 3370
    },
    {
      "entropy": 6.616267681121826,
      "epoch": 0.3894979803808425,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991475193517254,
      "loss": 6.4104,
      "mean_token_accuracy": 0.16712062805891037,
      "num_tokens": 8561576.0,
      "step": 3375
    },
    {
      "entropy": 6.5804126262664795,
      "epoch": 0.39007501442585113,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004991439269472956,
      "loss": 6.5575,
      "mean_token_accuracy": 0.15445181280374526,
      "num_tokens": 8574253.0,
      "step": 3380
    },
    {
      "entropy": 6.6331945896148685,
      "epoch": 0.39065204847085977,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004991403270038961,
      "loss": 6.458,
      "mean_token_accuracy": 0.16088307946920394,
      "num_tokens": 8585708.0,
      "step": 3385
    },
    {
      "entropy": 6.60706205368042,
      "epoch": 0.3912290825158684,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000499136719521648,
      "loss": 6.5642,
      "mean_token_accuracy": 0.14254746586084366,
      "num_tokens": 8598327.0,
      "step": 3390
    },
    {
      "entropy": 6.649341821670532,
      "epoch": 0.3918061165608771,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004991331045006726,
      "loss": 6.569,
      "mean_token_accuracy": 0.1483719229698181,
      "num_tokens": 8610606.0,
      "step": 3395
    },
    {
      "entropy": 6.717316722869873,
      "epoch": 0.39238315060588574,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004991294819410916,
      "loss": 6.5855,
      "mean_token_accuracy": 0.14726121872663497,
      "num_tokens": 8625201.0,
      "step": 3400
    },
    {
      "entropy": 6.582614088058472,
      "epoch": 0.3929601846508944,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004991258518430268,
      "loss": 6.5456,
      "mean_token_accuracy": 0.16018152683973313,
      "num_tokens": 8638288.0,
      "step": 3405
    },
    {
      "entropy": 6.658870792388916,
      "epoch": 0.3935372186959031,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004991222142066003,
      "loss": 6.4757,
      "mean_token_accuracy": 0.15923092365264893,
      "num_tokens": 8649550.0,
      "step": 3410
    },
    {
      "entropy": 6.565553760528564,
      "epoch": 0.3941142527409117,
      "grad_norm": 0.875,
      "learning_rate": 0.0004991185690319345,
      "loss": 6.4775,
      "mean_token_accuracy": 0.1563425600528717,
      "num_tokens": 8660956.0,
      "step": 3415
    },
    {
      "entropy": 6.676780319213867,
      "epoch": 0.39469128678592036,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499114916319152,
      "loss": 6.5616,
      "mean_token_accuracy": 0.14661440700292588,
      "num_tokens": 8675130.0,
      "step": 3420
    },
    {
      "entropy": 6.595038652420044,
      "epoch": 0.395268320830929,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004991112560683755,
      "loss": 6.4548,
      "mean_token_accuracy": 0.1577983319759369,
      "num_tokens": 8687482.0,
      "step": 3425
    },
    {
      "entropy": 6.5211607933044435,
      "epoch": 0.3958453548759377,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004991075882797283,
      "loss": 6.4495,
      "mean_token_accuracy": 0.16318398714065552,
      "num_tokens": 8700202.0,
      "step": 3430
    },
    {
      "entropy": 6.686310291290283,
      "epoch": 0.39642238892094633,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004991039129533337,
      "loss": 6.6103,
      "mean_token_accuracy": 0.1472760871052742,
      "num_tokens": 8713602.0,
      "step": 3435
    },
    {
      "entropy": 6.605076885223388,
      "epoch": 0.396999422965955,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004991002300893152,
      "loss": 6.5444,
      "mean_token_accuracy": 0.15414105653762816,
      "num_tokens": 8727318.0,
      "step": 3440
    },
    {
      "entropy": 6.642222690582275,
      "epoch": 0.39757645701096367,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004990965396877969,
      "loss": 6.4035,
      "mean_token_accuracy": 0.16198018193244934,
      "num_tokens": 8739941.0,
      "step": 3445
    },
    {
      "entropy": 6.616573715209961,
      "epoch": 0.3981534910559723,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004990928417489027,
      "loss": 6.5685,
      "mean_token_accuracy": 0.1492309018969536,
      "num_tokens": 8753012.0,
      "step": 3450
    },
    {
      "entropy": 6.5814603805542,
      "epoch": 0.39873052510098095,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004990891362727572,
      "loss": 6.4986,
      "mean_token_accuracy": 0.15747803896665574,
      "num_tokens": 8766998.0,
      "step": 3455
    },
    {
      "entropy": 6.569999074935913,
      "epoch": 0.3993075591459896,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004990854232594848,
      "loss": 6.4675,
      "mean_token_accuracy": 0.15911515578627586,
      "num_tokens": 8778993.0,
      "step": 3460
    },
    {
      "entropy": 6.586950969696045,
      "epoch": 0.3998845931909983,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004990817027092106,
      "loss": 6.46,
      "mean_token_accuracy": 0.15577831119298935,
      "num_tokens": 8790612.0,
      "step": 3465
    },
    {
      "entropy": 6.649136638641357,
      "epoch": 0.4004616272360069,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004990779746220595,
      "loss": 6.5891,
      "mean_token_accuracy": 0.15244996547698975,
      "num_tokens": 8803040.0,
      "step": 3470
    },
    {
      "entropy": 6.64963927268982,
      "epoch": 0.40103866128101556,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004990742389981572,
      "loss": 6.5433,
      "mean_token_accuracy": 0.1518448457121849,
      "num_tokens": 8816179.0,
      "step": 3475
    },
    {
      "entropy": 6.681894111633301,
      "epoch": 0.40161569532602426,
      "grad_norm": 0.8046875,
      "learning_rate": 0.000499070495837629,
      "loss": 6.5148,
      "mean_token_accuracy": 0.1452247053384781,
      "num_tokens": 8828779.0,
      "step": 3480
    },
    {
      "entropy": 6.678278493881225,
      "epoch": 0.4021927293710329,
      "grad_norm": 0.7265625,
      "learning_rate": 0.0004990667451406012,
      "loss": 6.5099,
      "mean_token_accuracy": 0.1595836415886879,
      "num_tokens": 8842354.0,
      "step": 3485
    },
    {
      "entropy": 6.577237606048584,
      "epoch": 0.40276976341604154,
      "grad_norm": 0.74609375,
      "learning_rate": 0.0004990629869071995,
      "loss": 6.5268,
      "mean_token_accuracy": 0.14757629632949829,
      "num_tokens": 8855462.0,
      "step": 3490
    },
    {
      "entropy": 6.64400954246521,
      "epoch": 0.4033467974610502,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004990592211375506,
      "loss": 6.5004,
      "mean_token_accuracy": 0.15390561521053314,
      "num_tokens": 8867882.0,
      "step": 3495
    },
    {
      "entropy": 6.5557056903839115,
      "epoch": 0.4039238315060589,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004990554478317811,
      "loss": 6.4443,
      "mean_token_accuracy": 0.1584362655878067,
      "num_tokens": 8880906.0,
      "step": 3500
    },
    {
      "entropy": 6.489110326766967,
      "epoch": 0.4045008655510675,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004990516669900179,
      "loss": 6.3377,
      "mean_token_accuracy": 0.16366536319255828,
      "num_tokens": 8893005.0,
      "step": 3505
    },
    {
      "entropy": 6.614466857910156,
      "epoch": 0.40507789959607615,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004990478786123882,
      "loss": 6.5371,
      "mean_token_accuracy": 0.14857055991888046,
      "num_tokens": 8905611.0,
      "step": 3510
    },
    {
      "entropy": 6.641386365890503,
      "epoch": 0.40565493364108485,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004990440826990193,
      "loss": 6.468,
      "mean_token_accuracy": 0.15724545121192932,
      "num_tokens": 8918454.0,
      "step": 3515
    },
    {
      "entropy": 6.577886247634888,
      "epoch": 0.4062319676860935,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004990402792500392,
      "loss": 6.4571,
      "mean_token_accuracy": 0.1559569999575615,
      "num_tokens": 8931871.0,
      "step": 3520
    },
    {
      "entropy": 6.567701816558838,
      "epoch": 0.4068090017311021,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004990364682655754,
      "loss": 6.5097,
      "mean_token_accuracy": 0.15251438468694686,
      "num_tokens": 8944813.0,
      "step": 3525
    },
    {
      "entropy": 6.611566114425659,
      "epoch": 0.40738603577611077,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004990326497457564,
      "loss": 6.5267,
      "mean_token_accuracy": 0.15673490911722182,
      "num_tokens": 8958902.0,
      "step": 3530
    },
    {
      "entropy": 6.5758881092071535,
      "epoch": 0.40796306982111946,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004990288236907104,
      "loss": 6.5406,
      "mean_token_accuracy": 0.15299722254276277,
      "num_tokens": 8971390.0,
      "step": 3535
    },
    {
      "entropy": 6.597666025161743,
      "epoch": 0.4085401038661281,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004990249901005661,
      "loss": 6.4701,
      "mean_token_accuracy": 0.15784490257501602,
      "num_tokens": 8983510.0,
      "step": 3540
    },
    {
      "entropy": 6.558063411712647,
      "epoch": 0.40911713791113674,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004990211489754527,
      "loss": 6.4978,
      "mean_token_accuracy": 0.1489548921585083,
      "num_tokens": 8997039.0,
      "step": 3545
    },
    {
      "entropy": 6.638218259811401,
      "epoch": 0.40969417195614544,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004990173003154993,
      "loss": 6.5936,
      "mean_token_accuracy": 0.14932667464017868,
      "num_tokens": 9009672.0,
      "step": 3550
    },
    {
      "entropy": 6.660990905761719,
      "epoch": 0.4102712060011541,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004990134441208351,
      "loss": 6.5222,
      "mean_token_accuracy": 0.15500398725271225,
      "num_tokens": 9023477.0,
      "step": 3555
    },
    {
      "entropy": 6.542504739761353,
      "epoch": 0.4108482400461627,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004990095803915901,
      "loss": 6.3898,
      "mean_token_accuracy": 0.16572054326534272,
      "num_tokens": 9035914.0,
      "step": 3560
    },
    {
      "entropy": 6.596242094039917,
      "epoch": 0.41142527409117136,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004990057091278942,
      "loss": 6.4774,
      "mean_token_accuracy": 0.15172698199748993,
      "num_tokens": 9048205.0,
      "step": 3565
    },
    {
      "entropy": 6.598004865646362,
      "epoch": 0.41200230813618005,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004990018303298773,
      "loss": 6.5425,
      "mean_token_accuracy": 0.15234472006559371,
      "num_tokens": 9061467.0,
      "step": 3570
    },
    {
      "entropy": 6.677182149887085,
      "epoch": 0.4125793421811887,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004989979439976703,
      "loss": 6.504,
      "mean_token_accuracy": 0.15467257797718048,
      "num_tokens": 9074434.0,
      "step": 3575
    },
    {
      "entropy": 6.4798554420471195,
      "epoch": 0.41315637622619733,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004989940501314037,
      "loss": 6.3941,
      "mean_token_accuracy": 0.16334292888641358,
      "num_tokens": 9087745.0,
      "step": 3580
    },
    {
      "entropy": 6.609991073608398,
      "epoch": 0.413733410271206,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004989901487312084,
      "loss": 6.4618,
      "mean_token_accuracy": 0.15822864919900895,
      "num_tokens": 9099424.0,
      "step": 3585
    },
    {
      "entropy": 6.562083673477173,
      "epoch": 0.41431044431621467,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004989862397972157,
      "loss": 6.4246,
      "mean_token_accuracy": 0.1650937482714653,
      "num_tokens": 9111056.0,
      "step": 3590
    },
    {
      "entropy": 6.623899793624878,
      "epoch": 0.4148874783612233,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004989823233295572,
      "loss": 6.5662,
      "mean_token_accuracy": 0.15530386716127395,
      "num_tokens": 9124539.0,
      "step": 3595
    },
    {
      "entropy": 6.6407468795776365,
      "epoch": 0.41546451240623195,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004989783993283644,
      "loss": 6.4541,
      "mean_token_accuracy": 0.15695587545633316,
      "num_tokens": 9136580.0,
      "step": 3600
    },
    {
      "entropy": 6.544411754608154,
      "epoch": 0.41604154645124064,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004989744677937694,
      "loss": 6.5157,
      "mean_token_accuracy": 0.15283992290496826,
      "num_tokens": 9148411.0,
      "step": 3605
    },
    {
      "entropy": 6.5882384300231935,
      "epoch": 0.4166185804962493,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004989705287259046,
      "loss": 6.4407,
      "mean_token_accuracy": 0.1591333732008934,
      "num_tokens": 9160778.0,
      "step": 3610
    },
    {
      "entropy": 6.622359991073608,
      "epoch": 0.4171956145412579,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004989665821249021,
      "loss": 6.4965,
      "mean_token_accuracy": 0.15284003764390947,
      "num_tokens": 9174174.0,
      "step": 3615
    },
    {
      "entropy": 6.63104214668274,
      "epoch": 0.4177726485862666,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000498962627990895,
      "loss": 6.473,
      "mean_token_accuracy": 0.1580877423286438,
      "num_tokens": 9186465.0,
      "step": 3620
    },
    {
      "entropy": 6.522603988647461,
      "epoch": 0.41834968263127525,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004989586663240161,
      "loss": 6.4506,
      "mean_token_accuracy": 0.15930631905794143,
      "num_tokens": 9198900.0,
      "step": 3625
    },
    {
      "entropy": 6.6117743968963625,
      "epoch": 0.4189267166762839,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004989546971243988,
      "loss": 6.4617,
      "mean_token_accuracy": 0.15446389317512513,
      "num_tokens": 9211148.0,
      "step": 3630
    },
    {
      "entropy": 6.6376965045928955,
      "epoch": 0.41950375072129253,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004989507203921763,
      "loss": 6.5696,
      "mean_token_accuracy": 0.1492958277463913,
      "num_tokens": 9226216.0,
      "step": 3635
    },
    {
      "entropy": 6.662663078308105,
      "epoch": 0.42008078476630123,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004989467361274828,
      "loss": 6.5074,
      "mean_token_accuracy": 0.15405069887638093,
      "num_tokens": 9238361.0,
      "step": 3640
    },
    {
      "entropy": 6.5690850734710695,
      "epoch": 0.42065781881130987,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004989427443304519,
      "loss": 6.428,
      "mean_token_accuracy": 0.1591155469417572,
      "num_tokens": 9251828.0,
      "step": 3645
    },
    {
      "entropy": 6.554191303253174,
      "epoch": 0.4212348528563185,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000498938745001218,
      "loss": 6.475,
      "mean_token_accuracy": 0.1533695265650749,
      "num_tokens": 9264433.0,
      "step": 3650
    },
    {
      "entropy": 6.662342739105225,
      "epoch": 0.4218118869013272,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004989347381399158,
      "loss": 6.506,
      "mean_token_accuracy": 0.15566451102495193,
      "num_tokens": 9277859.0,
      "step": 3655
    },
    {
      "entropy": 6.549519300460815,
      "epoch": 0.42238892094633584,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0004989307237466799,
      "loss": 6.524,
      "mean_token_accuracy": 0.1491675093770027,
      "num_tokens": 9290585.0,
      "step": 3660
    },
    {
      "entropy": 6.6417014598846436,
      "epoch": 0.4229659549913445,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004989267018216453,
      "loss": 6.5246,
      "mean_token_accuracy": 0.1502677731215954,
      "num_tokens": 9302633.0,
      "step": 3665
    },
    {
      "entropy": 6.578337478637695,
      "epoch": 0.4235429890363531,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004989226723649473,
      "loss": 6.5606,
      "mean_token_accuracy": 0.14945575445890427,
      "num_tokens": 9314577.0,
      "step": 3670
    },
    {
      "entropy": 6.560885238647461,
      "epoch": 0.4241200230813618,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004989186353767214,
      "loss": 6.421,
      "mean_token_accuracy": 0.16296332329511642,
      "num_tokens": 9327260.0,
      "step": 3675
    },
    {
      "entropy": 6.561640357971191,
      "epoch": 0.42469705712637046,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004989145908571035,
      "loss": 6.485,
      "mean_token_accuracy": 0.16140415668487548,
      "num_tokens": 9339684.0,
      "step": 3680
    },
    {
      "entropy": 6.605708312988281,
      "epoch": 0.4252740911713791,
      "grad_norm": 0.875,
      "learning_rate": 0.0004989105388062295,
      "loss": 6.4979,
      "mean_token_accuracy": 0.1537920966744423,
      "num_tokens": 9352532.0,
      "step": 3685
    },
    {
      "entropy": 6.655845975875854,
      "epoch": 0.4258511252163878,
      "grad_norm": 0.875,
      "learning_rate": 0.0004989064792242358,
      "loss": 6.5319,
      "mean_token_accuracy": 0.1520333930850029,
      "num_tokens": 9364680.0,
      "step": 3690
    },
    {
      "entropy": 6.6132406234741214,
      "epoch": 0.42642815926139643,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004989024121112589,
      "loss": 6.4949,
      "mean_token_accuracy": 0.1559430792927742,
      "num_tokens": 9377472.0,
      "step": 3695
    },
    {
      "entropy": 6.559667921066284,
      "epoch": 0.4270051933064051,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004988983374674356,
      "loss": 6.5138,
      "mean_token_accuracy": 0.15148606151342392,
      "num_tokens": 9391646.0,
      "step": 3700
    },
    {
      "entropy": 6.607914209365845,
      "epoch": 0.4275822273514137,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004988942552929029,
      "loss": 6.461,
      "mean_token_accuracy": 0.1514444500207901,
      "num_tokens": 9402864.0,
      "step": 3705
    },
    {
      "entropy": 6.598295068740844,
      "epoch": 0.4281592613964224,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004988901655877981,
      "loss": 6.4723,
      "mean_token_accuracy": 0.1514604777097702,
      "num_tokens": 9414498.0,
      "step": 3710
    },
    {
      "entropy": 6.556909513473511,
      "epoch": 0.42873629544143105,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004988860683522589,
      "loss": 6.5078,
      "mean_token_accuracy": 0.15649012476205826,
      "num_tokens": 9427903.0,
      "step": 3715
    },
    {
      "entropy": 6.611214685440063,
      "epoch": 0.4293133294864397,
      "grad_norm": 0.7734375,
      "learning_rate": 0.000498881963586423,
      "loss": 6.4587,
      "mean_token_accuracy": 0.16037748903036117,
      "num_tokens": 9441566.0,
      "step": 3720
    },
    {
      "entropy": 6.466671895980835,
      "epoch": 0.4298903635314484,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004988778512904282,
      "loss": 6.3517,
      "mean_token_accuracy": 0.1674039363861084,
      "num_tokens": 9453614.0,
      "step": 3725
    },
    {
      "entropy": 6.64181957244873,
      "epoch": 0.430467397576457,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004988737314644134,
      "loss": 6.5038,
      "mean_token_accuracy": 0.15030203759670258,
      "num_tokens": 9468046.0,
      "step": 3730
    },
    {
      "entropy": 6.4948383331298825,
      "epoch": 0.43104443162146566,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004988696041085168,
      "loss": 6.3915,
      "mean_token_accuracy": 0.16536259204149245,
      "num_tokens": 9481469.0,
      "step": 3735
    },
    {
      "entropy": 6.563466310501099,
      "epoch": 0.4316214656664743,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004988654692228772,
      "loss": 6.4936,
      "mean_token_accuracy": 0.15875237435102463,
      "num_tokens": 9492833.0,
      "step": 3740
    },
    {
      "entropy": 6.588851833343506,
      "epoch": 0.432198499711483,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004988613268076335,
      "loss": 6.4178,
      "mean_token_accuracy": 0.15342947393655776,
      "num_tokens": 9504955.0,
      "step": 3745
    },
    {
      "entropy": 6.5497125625610355,
      "epoch": 0.43277553375649164,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004988571768629257,
      "loss": 6.425,
      "mean_token_accuracy": 0.16287132054567338,
      "num_tokens": 9516830.0,
      "step": 3750
    },
    {
      "entropy": 6.517823219299316,
      "epoch": 0.4333525678015003,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004988530193888927,
      "loss": 6.4592,
      "mean_token_accuracy": 0.15513070076704025,
      "num_tokens": 9528965.0,
      "step": 3755
    },
    {
      "entropy": 6.682994794845581,
      "epoch": 0.4339296018465089,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004988488543856747,
      "loss": 6.5369,
      "mean_token_accuracy": 0.1505722388625145,
      "num_tokens": 9541838.0,
      "step": 3760
    },
    {
      "entropy": 6.553580951690674,
      "epoch": 0.4345066358915176,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0004988446818534115,
      "loss": 6.4714,
      "mean_token_accuracy": 0.15720604211091996,
      "num_tokens": 9554502.0,
      "step": 3765
    },
    {
      "entropy": 6.539259767532348,
      "epoch": 0.43508366993652625,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004988405017922438,
      "loss": 6.3194,
      "mean_token_accuracy": 0.16168797761201859,
      "num_tokens": 9567028.0,
      "step": 3770
    },
    {
      "entropy": 6.590677881240845,
      "epoch": 0.4356607039815349,
      "grad_norm": 0.79296875,
      "learning_rate": 0.000498836314202312,
      "loss": 6.4877,
      "mean_token_accuracy": 0.14764939695596696,
      "num_tokens": 9580336.0,
      "step": 3775
    },
    {
      "entropy": 6.573682975769043,
      "epoch": 0.4362377380265436,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004988321190837568,
      "loss": 6.4161,
      "mean_token_accuracy": 0.1552751526236534,
      "num_tokens": 9592410.0,
      "step": 3780
    },
    {
      "entropy": 6.57819676399231,
      "epoch": 0.4368147720715522,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004988279164367196,
      "loss": 6.4278,
      "mean_token_accuracy": 0.15180183351039886,
      "num_tokens": 9605693.0,
      "step": 3785
    },
    {
      "entropy": 6.623029613494873,
      "epoch": 0.43739180611656087,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004988237062613415,
      "loss": 6.4494,
      "mean_token_accuracy": 0.1581479698419571,
      "num_tokens": 9618555.0,
      "step": 3790
    },
    {
      "entropy": 6.549624633789063,
      "epoch": 0.4379688401615695,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004988194885577644,
      "loss": 6.4702,
      "mean_token_accuracy": 0.15138823837041854,
      "num_tokens": 9630330.0,
      "step": 3795
    },
    {
      "entropy": 6.601396465301514,
      "epoch": 0.4385458742065782,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004988152633261299,
      "loss": 6.4799,
      "mean_token_accuracy": 0.15994462221860886,
      "num_tokens": 9642516.0,
      "step": 3800
    },
    {
      "entropy": 6.569391345977783,
      "epoch": 0.43912290825158684,
      "grad_norm": 0.875,
      "learning_rate": 0.00049881103056658,
      "loss": 6.4483,
      "mean_token_accuracy": 0.15572853684425353,
      "num_tokens": 9654886.0,
      "step": 3805
    },
    {
      "entropy": 6.535688161849976,
      "epoch": 0.4396999422965955,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004988067902792575,
      "loss": 6.4739,
      "mean_token_accuracy": 0.1606386885046959,
      "num_tokens": 9666346.0,
      "step": 3810
    },
    {
      "entropy": 6.550254011154175,
      "epoch": 0.4402769763416042,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004988025424643047,
      "loss": 6.4505,
      "mean_token_accuracy": 0.15357128232717515,
      "num_tokens": 9679971.0,
      "step": 3815
    },
    {
      "entropy": 6.616861248016358,
      "epoch": 0.4408540103866128,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004987982871218645,
      "loss": 6.5232,
      "mean_token_accuracy": 0.15241505280137063,
      "num_tokens": 9693032.0,
      "step": 3820
    },
    {
      "entropy": 6.587024784088134,
      "epoch": 0.44143104443162146,
      "grad_norm": 0.7421875,
      "learning_rate": 0.0004987940242520802,
      "loss": 6.4342,
      "mean_token_accuracy": 0.16022422015666962,
      "num_tokens": 9705706.0,
      "step": 3825
    },
    {
      "entropy": 6.461678075790405,
      "epoch": 0.4420080784766301,
      "grad_norm": 0.796875,
      "learning_rate": 0.000498789753855095,
      "loss": 6.3687,
      "mean_token_accuracy": 0.16021044701337814,
      "num_tokens": 9718235.0,
      "step": 3830
    },
    {
      "entropy": 6.554477596282959,
      "epoch": 0.4425851125216388,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004987854759310526,
      "loss": 6.5244,
      "mean_token_accuracy": 0.1447369821369648,
      "num_tokens": 9730758.0,
      "step": 3835
    },
    {
      "entropy": 6.565464019775391,
      "epoch": 0.44316214656664743,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004987811904800968,
      "loss": 6.3592,
      "mean_token_accuracy": 0.16637056618928908,
      "num_tokens": 9743131.0,
      "step": 3840
    },
    {
      "entropy": 6.500174474716187,
      "epoch": 0.44373918061165607,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004987768975023719,
      "loss": 6.4459,
      "mean_token_accuracy": 0.15130365639925003,
      "num_tokens": 9755965.0,
      "step": 3845
    },
    {
      "entropy": 6.591799068450928,
      "epoch": 0.44431621465666477,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004987725969980222,
      "loss": 6.3492,
      "mean_token_accuracy": 0.16706654876470567,
      "num_tokens": 9768460.0,
      "step": 3850
    },
    {
      "entropy": 6.561370229721069,
      "epoch": 0.4448932487016734,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004987682889671923,
      "loss": 6.5234,
      "mean_token_accuracy": 0.15095077604055404,
      "num_tokens": 9781653.0,
      "step": 3855
    },
    {
      "entropy": 6.50119457244873,
      "epoch": 0.44547028274668204,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004987639734100272,
      "loss": 6.3379,
      "mean_token_accuracy": 0.1660802647471428,
      "num_tokens": 9793814.0,
      "step": 3860
    },
    {
      "entropy": 6.53929648399353,
      "epoch": 0.4460473167916907,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004987596503266721,
      "loss": 6.4071,
      "mean_token_accuracy": 0.159231798350811,
      "num_tokens": 9806635.0,
      "step": 3865
    },
    {
      "entropy": 6.531664180755615,
      "epoch": 0.4466243508366994,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004987553197172722,
      "loss": 6.3851,
      "mean_token_accuracy": 0.15614811927080155,
      "num_tokens": 9819213.0,
      "step": 3870
    },
    {
      "entropy": 6.460581731796265,
      "epoch": 0.447201384881708,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004987509815819732,
      "loss": 6.3818,
      "mean_token_accuracy": 0.15675631165504456,
      "num_tokens": 9831823.0,
      "step": 3875
    },
    {
      "entropy": 6.535300016403198,
      "epoch": 0.44777841892671666,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004987466359209213,
      "loss": 6.3739,
      "mean_token_accuracy": 0.15925684422254563,
      "num_tokens": 9844993.0,
      "step": 3880
    },
    {
      "entropy": 6.474202013015747,
      "epoch": 0.44835545297172535,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004987422827342622,
      "loss": 6.2131,
      "mean_token_accuracy": 0.17272613048553467,
      "num_tokens": 9857853.0,
      "step": 3885
    },
    {
      "entropy": 6.5748802661895756,
      "epoch": 0.448932487016734,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004987379220221426,
      "loss": 6.4607,
      "mean_token_accuracy": 0.15414051860570907,
      "num_tokens": 9871465.0,
      "step": 3890
    },
    {
      "entropy": 6.5424511432647705,
      "epoch": 0.44950952106174263,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004987335537847092,
      "loss": 6.443,
      "mean_token_accuracy": 0.16424099653959273,
      "num_tokens": 9884973.0,
      "step": 3895
    },
    {
      "entropy": 6.639307117462158,
      "epoch": 0.4500865551067513,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004987291780221088,
      "loss": 6.5656,
      "mean_token_accuracy": 0.14660622775554658,
      "num_tokens": 9898143.0,
      "step": 3900
    },
    {
      "entropy": 6.629957818984986,
      "epoch": 0.45066358915175997,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004987247947344887,
      "loss": 6.4866,
      "mean_token_accuracy": 0.154781673848629,
      "num_tokens": 9910584.0,
      "step": 3905
    },
    {
      "entropy": 6.60298523902893,
      "epoch": 0.4512406231967686,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004987204039219962,
      "loss": 6.5319,
      "mean_token_accuracy": 0.14886820018291474,
      "num_tokens": 9922901.0,
      "step": 3910
    },
    {
      "entropy": 6.651839780807495,
      "epoch": 0.45181765724177725,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004987160055847791,
      "loss": 6.523,
      "mean_token_accuracy": 0.14895009398460388,
      "num_tokens": 9936131.0,
      "step": 3915
    },
    {
      "entropy": 6.493554782867432,
      "epoch": 0.45239469128678594,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004987115997229852,
      "loss": 6.2838,
      "mean_token_accuracy": 0.1695529192686081,
      "num_tokens": 9950292.0,
      "step": 3920
    },
    {
      "entropy": 6.512192010879517,
      "epoch": 0.4529717253317946,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004987071863367629,
      "loss": 6.4452,
      "mean_token_accuracy": 0.15502430945634843,
      "num_tokens": 9963765.0,
      "step": 3925
    },
    {
      "entropy": 6.560297203063965,
      "epoch": 0.4535487593768032,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004987027654262604,
      "loss": 6.4376,
      "mean_token_accuracy": 0.1608374685049057,
      "num_tokens": 9975981.0,
      "step": 3930
    },
    {
      "entropy": 6.54636697769165,
      "epoch": 0.45412579342181186,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004986983369916264,
      "loss": 6.3945,
      "mean_token_accuracy": 0.15865204036235808,
      "num_tokens": 9988763.0,
      "step": 3935
    },
    {
      "entropy": 6.480308389663696,
      "epoch": 0.45470282746682056,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004986939010330102,
      "loss": 6.4012,
      "mean_token_accuracy": 0.1632219061255455,
      "num_tokens": 10000663.0,
      "step": 3940
    },
    {
      "entropy": 6.530785274505615,
      "epoch": 0.4552798615118292,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004986894575505606,
      "loss": 6.4096,
      "mean_token_accuracy": 0.15825921148061753,
      "num_tokens": 10012673.0,
      "step": 3945
    },
    {
      "entropy": 6.528098106384277,
      "epoch": 0.45585689555683784,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004986850065444272,
      "loss": 6.4407,
      "mean_token_accuracy": 0.15066046863794327,
      "num_tokens": 10024410.0,
      "step": 3950
    },
    {
      "entropy": 6.55545563697815,
      "epoch": 0.45643392960184653,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004986805480147598,
      "loss": 6.3386,
      "mean_token_accuracy": 0.16893032789230347,
      "num_tokens": 10036782.0,
      "step": 3955
    },
    {
      "entropy": 6.51139645576477,
      "epoch": 0.4570109636468552,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004986760819617082,
      "loss": 6.4205,
      "mean_token_accuracy": 0.15875447541475296,
      "num_tokens": 10049234.0,
      "step": 3960
    },
    {
      "entropy": 6.578258800506592,
      "epoch": 0.4575879976918638,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004986716083854228,
      "loss": 6.4511,
      "mean_token_accuracy": 0.15605315566062927,
      "num_tokens": 10061853.0,
      "step": 3965
    },
    {
      "entropy": 6.626540565490723,
      "epoch": 0.45816503173687245,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004986671272860538,
      "loss": 6.47,
      "mean_token_accuracy": 0.15049075484275817,
      "num_tokens": 10074890.0,
      "step": 3970
    },
    {
      "entropy": 6.581723546981811,
      "epoch": 0.45874206578188115,
      "grad_norm": 0.7421875,
      "learning_rate": 0.0004986626386637521,
      "loss": 6.5189,
      "mean_token_accuracy": 0.1504073806107044,
      "num_tokens": 10090226.0,
      "step": 3975
    },
    {
      "entropy": 6.5390383243560795,
      "epoch": 0.4593190998268898,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004986581425186688,
      "loss": 6.47,
      "mean_token_accuracy": 0.15238555520772934,
      "num_tokens": 10103716.0,
      "step": 3980
    },
    {
      "entropy": 6.602221393585205,
      "epoch": 0.4598961338718984,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004986536388509548,
      "loss": 6.5884,
      "mean_token_accuracy": 0.14533913433551787,
      "num_tokens": 10117553.0,
      "step": 3985
    },
    {
      "entropy": 6.543276596069336,
      "epoch": 0.4604731679169071,
      "grad_norm": 0.875,
      "learning_rate": 0.0004986491276607618,
      "loss": 6.4153,
      "mean_token_accuracy": 0.15889854580163956,
      "num_tokens": 10129897.0,
      "step": 3990
    },
    {
      "entropy": 6.616291904449463,
      "epoch": 0.46105020196191576,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004986446089482416,
      "loss": 6.5394,
      "mean_token_accuracy": 0.1505708009004593,
      "num_tokens": 10141974.0,
      "step": 3995
    },
    {
      "entropy": 6.616165256500244,
      "epoch": 0.4616272360069244,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004986400827135459,
      "loss": 6.4595,
      "mean_token_accuracy": 0.14896104633808135,
      "num_tokens": 10155471.0,
      "step": 4000
    },
    {
      "entropy": 6.565793371200561,
      "epoch": 0.46220427005193304,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004986355489568272,
      "loss": 6.4254,
      "mean_token_accuracy": 0.15408960729837418,
      "num_tokens": 10167362.0,
      "step": 4005
    },
    {
      "entropy": 6.527460670471191,
      "epoch": 0.46278130409694174,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004986310076782379,
      "loss": 6.4344,
      "mean_token_accuracy": 0.15517805516719818,
      "num_tokens": 10180706.0,
      "step": 4010
    },
    {
      "entropy": 6.633301687240601,
      "epoch": 0.4633583381419504,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004986264588779307,
      "loss": 6.4721,
      "mean_token_accuracy": 0.15134465247392653,
      "num_tokens": 10193642.0,
      "step": 4015
    },
    {
      "entropy": 6.590562105178833,
      "epoch": 0.463935372186959,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004986219025560586,
      "loss": 6.4913,
      "mean_token_accuracy": 0.15406545549631118,
      "num_tokens": 10206552.0,
      "step": 4020
    },
    {
      "entropy": 6.54869499206543,
      "epoch": 0.4645124062319677,
      "grad_norm": 0.81640625,
      "learning_rate": 0.000498617338712775,
      "loss": 6.4934,
      "mean_token_accuracy": 0.157432359457016,
      "num_tokens": 10219576.0,
      "step": 4025
    },
    {
      "entropy": 6.630867338180542,
      "epoch": 0.46508944027697635,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004986127673482332,
      "loss": 6.3994,
      "mean_token_accuracy": 0.15101128816604614,
      "num_tokens": 10231816.0,
      "step": 4030
    },
    {
      "entropy": 6.429687786102295,
      "epoch": 0.465666474321985,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004986081884625871,
      "loss": 6.2776,
      "mean_token_accuracy": 0.16077178120613098,
      "num_tokens": 10244329.0,
      "step": 4035
    },
    {
      "entropy": 6.522308301925659,
      "epoch": 0.46624350836699363,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004986036020559906,
      "loss": 6.4183,
      "mean_token_accuracy": 0.1520432010293007,
      "num_tokens": 10256521.0,
      "step": 4040
    },
    {
      "entropy": 6.558711004257202,
      "epoch": 0.4668205424120023,
      "grad_norm": 0.765625,
      "learning_rate": 0.000498599008128598,
      "loss": 6.4135,
      "mean_token_accuracy": 0.15548562109470368,
      "num_tokens": 10269023.0,
      "step": 4045
    },
    {
      "entropy": 6.542877626419068,
      "epoch": 0.46739757645701097,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004985944066805639,
      "loss": 6.4606,
      "mean_token_accuracy": 0.1481800675392151,
      "num_tokens": 10281476.0,
      "step": 4050
    },
    {
      "entropy": 6.557100534439087,
      "epoch": 0.4679746105020196,
      "grad_norm": 0.921875,
      "learning_rate": 0.000498589797712043,
      "loss": 6.3751,
      "mean_token_accuracy": 0.15909120291471482,
      "num_tokens": 10293749.0,
      "step": 4055
    },
    {
      "entropy": 6.567505502700806,
      "epoch": 0.4685516445470283,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004985851812231903,
      "loss": 6.3867,
      "mean_token_accuracy": 0.16700287014245987,
      "num_tokens": 10306554.0,
      "step": 4060
    },
    {
      "entropy": 6.560662698745728,
      "epoch": 0.46912867859203694,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004985805572141611,
      "loss": 6.4409,
      "mean_token_accuracy": 0.15463736951351165,
      "num_tokens": 10319374.0,
      "step": 4065
    },
    {
      "entropy": 6.51935601234436,
      "epoch": 0.4697057126370456,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004985759256851108,
      "loss": 6.4818,
      "mean_token_accuracy": 0.15578267723321915,
      "num_tokens": 10333358.0,
      "step": 4070
    },
    {
      "entropy": 6.555917501449585,
      "epoch": 0.4702827466820542,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004985712866361954,
      "loss": 6.4144,
      "mean_token_accuracy": 0.16267106384038926,
      "num_tokens": 10344830.0,
      "step": 4075
    },
    {
      "entropy": 6.5785972595214846,
      "epoch": 0.4708597807270629,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004985666400675709,
      "loss": 6.4821,
      "mean_token_accuracy": 0.15228304713964463,
      "num_tokens": 10357813.0,
      "step": 4080
    },
    {
      "entropy": 6.615689039230347,
      "epoch": 0.47143681477207156,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004985619859793934,
      "loss": 6.5128,
      "mean_token_accuracy": 0.15323988795280458,
      "num_tokens": 10370925.0,
      "step": 4085
    },
    {
      "entropy": 6.606573104858398,
      "epoch": 0.4720138488170802,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0004985573243718195,
      "loss": 6.4639,
      "mean_token_accuracy": 0.1552817478775978,
      "num_tokens": 10384172.0,
      "step": 4090
    },
    {
      "entropy": 6.58382625579834,
      "epoch": 0.4725908828620889,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004985526552450061,
      "loss": 6.406,
      "mean_token_accuracy": 0.15829960033297538,
      "num_tokens": 10396797.0,
      "step": 4095
    },
    {
      "entropy": 6.483374834060669,
      "epoch": 0.47316791690709753,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004985479785991103,
      "loss": 6.4509,
      "mean_token_accuracy": 0.1525266408920288,
      "num_tokens": 10409139.0,
      "step": 4100
    },
    {
      "entropy": 6.621217346191406,
      "epoch": 0.47374495095210617,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004985432944342892,
      "loss": 6.3974,
      "mean_token_accuracy": 0.15837369561195375,
      "num_tokens": 10421442.0,
      "step": 4105
    },
    {
      "entropy": 6.532503461837768,
      "epoch": 0.4743219849971148,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004985386027507003,
      "loss": 6.3919,
      "mean_token_accuracy": 0.16080056130886078,
      "num_tokens": 10433761.0,
      "step": 4110
    },
    {
      "entropy": 6.5195691108703615,
      "epoch": 0.4748990190421235,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004985339035485018,
      "loss": 6.4894,
      "mean_token_accuracy": 0.15351008325815202,
      "num_tokens": 10446508.0,
      "step": 4115
    },
    {
      "entropy": 6.506284713745117,
      "epoch": 0.47547605308713214,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004985291968278513,
      "loss": 6.3903,
      "mean_token_accuracy": 0.1637458860874176,
      "num_tokens": 10459352.0,
      "step": 4120
    },
    {
      "entropy": 6.540371036529541,
      "epoch": 0.4760530871321408,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004985244825889074,
      "loss": 6.4646,
      "mean_token_accuracy": 0.150443135201931,
      "num_tokens": 10470324.0,
      "step": 4125
    },
    {
      "entropy": 6.587960052490234,
      "epoch": 0.4766301211771495,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004985197608318284,
      "loss": 6.4331,
      "mean_token_accuracy": 0.15636301636695862,
      "num_tokens": 10482715.0,
      "step": 4130
    },
    {
      "entropy": 6.566105747222901,
      "epoch": 0.4772071552221581,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004985150315567735,
      "loss": 6.4303,
      "mean_token_accuracy": 0.1571477860212326,
      "num_tokens": 10493207.0,
      "step": 4135
    },
    {
      "entropy": 6.46118950843811,
      "epoch": 0.47778418926716676,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004985102947639013,
      "loss": 6.3845,
      "mean_token_accuracy": 0.1525936096906662,
      "num_tokens": 10505187.0,
      "step": 4140
    },
    {
      "entropy": 6.585748958587646,
      "epoch": 0.4783612233121754,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004985055504533715,
      "loss": 6.422,
      "mean_token_accuracy": 0.16249419003725052,
      "num_tokens": 10518301.0,
      "step": 4145
    },
    {
      "entropy": 6.492847347259522,
      "epoch": 0.4789382573571841,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004985007986253435,
      "loss": 6.4236,
      "mean_token_accuracy": 0.15842338651418686,
      "num_tokens": 10531288.0,
      "step": 4150
    },
    {
      "entropy": 6.4829020500183105,
      "epoch": 0.47951529140219273,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004984960392799772,
      "loss": 6.3133,
      "mean_token_accuracy": 0.16635616421699523,
      "num_tokens": 10544120.0,
      "step": 4155
    },
    {
      "entropy": 6.4867268085479735,
      "epoch": 0.4800923254472014,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004984912724174326,
      "loss": 6.406,
      "mean_token_accuracy": 0.16807787865400314,
      "num_tokens": 10556453.0,
      "step": 4160
    },
    {
      "entropy": 6.576405239105225,
      "epoch": 0.48066935949221,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004984864980378701,
      "loss": 6.5096,
      "mean_token_accuracy": 0.14650319889187813,
      "num_tokens": 10569404.0,
      "step": 4165
    },
    {
      "entropy": 6.551389598846436,
      "epoch": 0.4812463935372187,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004984817161414502,
      "loss": 6.4266,
      "mean_token_accuracy": 0.15406363904476167,
      "num_tokens": 10582312.0,
      "step": 4170
    },
    {
      "entropy": 6.541248607635498,
      "epoch": 0.48182342758222735,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004984769267283338,
      "loss": 6.4758,
      "mean_token_accuracy": 0.15606760680675508,
      "num_tokens": 10594391.0,
      "step": 4175
    },
    {
      "entropy": 6.58020920753479,
      "epoch": 0.482400461627236,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004984721297986819,
      "loss": 6.4124,
      "mean_token_accuracy": 0.15352201610803604,
      "num_tokens": 10606354.0,
      "step": 4180
    },
    {
      "entropy": 6.512746667861938,
      "epoch": 0.4829774956722447,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004984673253526561,
      "loss": 6.4036,
      "mean_token_accuracy": 0.1524976074695587,
      "num_tokens": 10618051.0,
      "step": 4185
    },
    {
      "entropy": 6.496927738189697,
      "epoch": 0.4835545297172533,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004984625133904176,
      "loss": 6.3149,
      "mean_token_accuracy": 0.16221490651369094,
      "num_tokens": 10629688.0,
      "step": 4190
    },
    {
      "entropy": 6.55139570236206,
      "epoch": 0.48413156376226196,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004984576939121286,
      "loss": 6.4507,
      "mean_token_accuracy": 0.1502139836549759,
      "num_tokens": 10642249.0,
      "step": 4195
    },
    {
      "entropy": 6.47601842880249,
      "epoch": 0.4847085978072706,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004984528669179511,
      "loss": 6.3853,
      "mean_token_accuracy": 0.1614321783185005,
      "num_tokens": 10655231.0,
      "step": 4200
    },
    {
      "entropy": 6.433983659744262,
      "epoch": 0.4852856318522793,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004984480324080472,
      "loss": 6.3893,
      "mean_token_accuracy": 0.16549270153045653,
      "num_tokens": 10668443.0,
      "step": 4205
    },
    {
      "entropy": 6.628914546966553,
      "epoch": 0.48586266589728794,
      "grad_norm": 0.84375,
      "learning_rate": 0.00049844319038258,
      "loss": 6.3931,
      "mean_token_accuracy": 0.15670084208250046,
      "num_tokens": 10680523.0,
      "step": 4210
    },
    {
      "entropy": 6.616771078109741,
      "epoch": 0.4864396999422966,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0004984383408417119,
      "loss": 6.4454,
      "mean_token_accuracy": 0.1573428303003311,
      "num_tokens": 10693223.0,
      "step": 4215
    },
    {
      "entropy": 6.509914207458496,
      "epoch": 0.4870167339873053,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004984334837856064,
      "loss": 6.3977,
      "mean_token_accuracy": 0.16693476736545562,
      "num_tokens": 10705276.0,
      "step": 4220
    },
    {
      "entropy": 6.503002166748047,
      "epoch": 0.4875937680323139,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004984286192144263,
      "loss": 6.434,
      "mean_token_accuracy": 0.16314385533332826,
      "num_tokens": 10718495.0,
      "step": 4225
    },
    {
      "entropy": 6.521592330932617,
      "epoch": 0.48817080207732255,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004984237471283359,
      "loss": 6.3447,
      "mean_token_accuracy": 0.16712241023778915,
      "num_tokens": 10730885.0,
      "step": 4230
    },
    {
      "entropy": 6.5000770568847654,
      "epoch": 0.4887478361223312,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004984188675274986,
      "loss": 6.3908,
      "mean_token_accuracy": 0.15794526487588884,
      "num_tokens": 10742906.0,
      "step": 4235
    },
    {
      "entropy": 6.528513193130493,
      "epoch": 0.4893248701673399,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004984139804120787,
      "loss": 6.3446,
      "mean_token_accuracy": 0.1597832128405571,
      "num_tokens": 10757050.0,
      "step": 4240
    },
    {
      "entropy": 6.557890224456787,
      "epoch": 0.4899019042123485,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004984090857822406,
      "loss": 6.3975,
      "mean_token_accuracy": 0.1596458524465561,
      "num_tokens": 10768667.0,
      "step": 4245
    },
    {
      "entropy": 6.591907787322998,
      "epoch": 0.49047893825735717,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004984041836381488,
      "loss": 6.4248,
      "mean_token_accuracy": 0.15413296818733216,
      "num_tokens": 10781687.0,
      "step": 4250
    },
    {
      "entropy": 6.402188920974732,
      "epoch": 0.49105597230236586,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004983992739799682,
      "loss": 6.3199,
      "mean_token_accuracy": 0.16379206255078316,
      "num_tokens": 10794966.0,
      "step": 4255
    },
    {
      "entropy": 6.588558530807495,
      "epoch": 0.4916330063473745,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004983943568078642,
      "loss": 6.4368,
      "mean_token_accuracy": 0.1489301383495331,
      "num_tokens": 10807946.0,
      "step": 4260
    },
    {
      "entropy": 6.531991815567016,
      "epoch": 0.49221004039238314,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004983894321220017,
      "loss": 6.3998,
      "mean_token_accuracy": 0.15459950715303422,
      "num_tokens": 10819258.0,
      "step": 4265
    },
    {
      "entropy": 6.468906021118164,
      "epoch": 0.4927870744373918,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004983844999225468,
      "loss": 6.3978,
      "mean_token_accuracy": 0.15649582892656327,
      "num_tokens": 10831439.0,
      "step": 4270
    },
    {
      "entropy": 6.5455316543579105,
      "epoch": 0.4933641084824005,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0004983795602096651,
      "loss": 6.3436,
      "mean_token_accuracy": 0.16237957179546356,
      "num_tokens": 10844214.0,
      "step": 4275
    },
    {
      "entropy": 6.526888799667359,
      "epoch": 0.4939411425274091,
      "grad_norm": 0.828125,
      "learning_rate": 0.000498374612983523,
      "loss": 6.4648,
      "mean_token_accuracy": 0.15004564374685286,
      "num_tokens": 10856675.0,
      "step": 4280
    },
    {
      "entropy": 6.541804265975952,
      "epoch": 0.49451817657241776,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004983696582442866,
      "loss": 6.389,
      "mean_token_accuracy": 0.15704446136951447,
      "num_tokens": 10869296.0,
      "step": 4285
    },
    {
      "entropy": 6.478774929046631,
      "epoch": 0.49509521061742645,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004983646959921227,
      "loss": 6.4454,
      "mean_token_accuracy": 0.16479874402284622,
      "num_tokens": 10881484.0,
      "step": 4290
    },
    {
      "entropy": 6.578076696395874,
      "epoch": 0.4956722446624351,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004983597262271984,
      "loss": 6.4188,
      "mean_token_accuracy": 0.16067702323198318,
      "num_tokens": 10894322.0,
      "step": 4295
    },
    {
      "entropy": 6.481278276443481,
      "epoch": 0.49624927870744373,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004983547489496804,
      "loss": 6.3652,
      "mean_token_accuracy": 0.17005881816148757,
      "num_tokens": 10907568.0,
      "step": 4300
    },
    {
      "entropy": 6.5737041473388675,
      "epoch": 0.49682631275245237,
      "grad_norm": 0.71875,
      "learning_rate": 0.0004983497641597365,
      "loss": 6.5261,
      "mean_token_accuracy": 0.1484769120812416,
      "num_tokens": 10921371.0,
      "step": 4305
    },
    {
      "entropy": 6.64163818359375,
      "epoch": 0.49740334679746107,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004983447718575342,
      "loss": 6.4558,
      "mean_token_accuracy": 0.153278249502182,
      "num_tokens": 10934725.0,
      "step": 4310
    },
    {
      "entropy": 6.529429626464844,
      "epoch": 0.4979803808424697,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004983397720432414,
      "loss": 6.4106,
      "mean_token_accuracy": 0.15111638754606246,
      "num_tokens": 10947985.0,
      "step": 4315
    },
    {
      "entropy": 6.3701372146606445,
      "epoch": 0.49855741488747835,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004983347647170264,
      "loss": 6.2466,
      "mean_token_accuracy": 0.17306942492723465,
      "num_tokens": 10961112.0,
      "step": 4320
    },
    {
      "entropy": 6.4976856231689455,
      "epoch": 0.49913444893248704,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004983297498790574,
      "loss": 6.3554,
      "mean_token_accuracy": 0.16256517618894578,
      "num_tokens": 10973414.0,
      "step": 4325
    },
    {
      "entropy": 6.610036849975586,
      "epoch": 0.4997114829774957,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004983247275295034,
      "loss": 6.3368,
      "mean_token_accuracy": 0.16400094330310822,
      "num_tokens": 10985609.0,
      "step": 4330
    },
    {
      "entropy": 6.477565145492553,
      "epoch": 0.5002885170225043,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004983196976685329,
      "loss": 6.4223,
      "mean_token_accuracy": 0.16130915582180022,
      "num_tokens": 10998928.0,
      "step": 4335
    },
    {
      "entropy": 6.518366050720215,
      "epoch": 0.500865551067513,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004983146602963155,
      "loss": 6.3347,
      "mean_token_accuracy": 0.15564725548028946,
      "num_tokens": 11012254.0,
      "step": 4340
    },
    {
      "entropy": 6.5654683113098145,
      "epoch": 0.5014425851125216,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004983096154130203,
      "loss": 6.3981,
      "mean_token_accuracy": 0.15106508135795593,
      "num_tokens": 11024040.0,
      "step": 4345
    },
    {
      "entropy": 6.603226995468139,
      "epoch": 0.5020196191575302,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004983045630188171,
      "loss": 6.5357,
      "mean_token_accuracy": 0.14584791511297227,
      "num_tokens": 11037207.0,
      "step": 4350
    },
    {
      "entropy": 6.5520185947418215,
      "epoch": 0.502596653202539,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004982995031138759,
      "loss": 6.4134,
      "mean_token_accuracy": 0.1584774896502495,
      "num_tokens": 11049990.0,
      "step": 4355
    },
    {
      "entropy": 6.566530513763428,
      "epoch": 0.5031736872475476,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004982944356983666,
      "loss": 6.4835,
      "mean_token_accuracy": 0.15407072454690934,
      "num_tokens": 11062365.0,
      "step": 4360
    },
    {
      "entropy": 6.509780597686768,
      "epoch": 0.5037507212925563,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004982893607724602,
      "loss": 6.3554,
      "mean_token_accuracy": 0.1576795235276222,
      "num_tokens": 11073721.0,
      "step": 4365
    },
    {
      "entropy": 6.46516637802124,
      "epoch": 0.5043277553375649,
      "grad_norm": 0.84375,
      "learning_rate": 0.000498284278336327,
      "loss": 6.3865,
      "mean_token_accuracy": 0.15484800785779954,
      "num_tokens": 11085834.0,
      "step": 4370
    },
    {
      "entropy": 6.561235237121582,
      "epoch": 0.5049047893825735,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004982791883901379,
      "loss": 6.4379,
      "mean_token_accuracy": 0.1539237193763256,
      "num_tokens": 11099432.0,
      "step": 4375
    },
    {
      "entropy": 6.417895078659058,
      "epoch": 0.5054818234275822,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004982740909340643,
      "loss": 6.2179,
      "mean_token_accuracy": 0.178082175552845,
      "num_tokens": 11111169.0,
      "step": 4380
    },
    {
      "entropy": 6.4774627685546875,
      "epoch": 0.5060588574725908,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004982689859682775,
      "loss": 6.3629,
      "mean_token_accuracy": 0.15544058978557587,
      "num_tokens": 11124131.0,
      "step": 4385
    },
    {
      "entropy": 6.549461221694946,
      "epoch": 0.5066358915175996,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004982638734929492,
      "loss": 6.3555,
      "mean_token_accuracy": 0.15999710261821748,
      "num_tokens": 11135810.0,
      "step": 4390
    },
    {
      "entropy": 6.5114202976226805,
      "epoch": 0.5072129255626082,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004982587535082515,
      "loss": 6.4399,
      "mean_token_accuracy": 0.16011044085025788,
      "num_tokens": 11148081.0,
      "step": 4395
    },
    {
      "entropy": 6.538697814941406,
      "epoch": 0.5077899596076169,
      "grad_norm": 0.7421875,
      "learning_rate": 0.0004982536260143565,
      "loss": 6.3381,
      "mean_token_accuracy": 0.1549960732460022,
      "num_tokens": 11161427.0,
      "step": 4400
    },
    {
      "entropy": 6.452679586410523,
      "epoch": 0.5083669936526255,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004982484910114367,
      "loss": 6.2565,
      "mean_token_accuracy": 0.16689955294132233,
      "num_tokens": 11174183.0,
      "step": 4405
    },
    {
      "entropy": 6.442296552658081,
      "epoch": 0.5089440276976341,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004982433484996648,
      "loss": 6.418,
      "mean_token_accuracy": 0.15510635450482368,
      "num_tokens": 11187151.0,
      "step": 4410
    },
    {
      "entropy": 6.545526695251465,
      "epoch": 0.5095210617426428,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004982381984792138,
      "loss": 6.298,
      "mean_token_accuracy": 0.16621606498956681,
      "num_tokens": 11199603.0,
      "step": 4415
    },
    {
      "entropy": 6.4789220809936525,
      "epoch": 0.5100980957876514,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004982330409502568,
      "loss": 6.3847,
      "mean_token_accuracy": 0.16437848657369614,
      "num_tokens": 11211709.0,
      "step": 4420
    },
    {
      "entropy": 6.491673469543457,
      "epoch": 0.5106751298326602,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004982278759129674,
      "loss": 6.373,
      "mean_token_accuracy": 0.16032237857580184,
      "num_tokens": 11224192.0,
      "step": 4425
    },
    {
      "entropy": 6.517372989654541,
      "epoch": 0.5112521638776688,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004982227033675194,
      "loss": 6.3591,
      "mean_token_accuracy": 0.16160479038953782,
      "num_tokens": 11237026.0,
      "step": 4430
    },
    {
      "entropy": 6.414926385879516,
      "epoch": 0.5118291979226774,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004982175233140865,
      "loss": 6.3635,
      "mean_token_accuracy": 0.15989502817392348,
      "num_tokens": 11250646.0,
      "step": 4435
    },
    {
      "entropy": 6.640326833724975,
      "epoch": 0.5124062319676861,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004982123357528431,
      "loss": 6.4629,
      "mean_token_accuracy": 0.15322466492652892,
      "num_tokens": 11262991.0,
      "step": 4440
    },
    {
      "entropy": 6.547874546051025,
      "epoch": 0.5129832660126947,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004982071406839636,
      "loss": 6.3598,
      "mean_token_accuracy": 0.15413384437561034,
      "num_tokens": 11275535.0,
      "step": 4445
    },
    {
      "entropy": 6.51468071937561,
      "epoch": 0.5135603000577034,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004982019381076229,
      "loss": 6.5046,
      "mean_token_accuracy": 0.15111950039863586,
      "num_tokens": 11288473.0,
      "step": 4450
    },
    {
      "entropy": 6.622634696960449,
      "epoch": 0.514137334102712,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004981967280239958,
      "loss": 6.5094,
      "mean_token_accuracy": 0.15655739828944207,
      "num_tokens": 11300755.0,
      "step": 4455
    },
    {
      "entropy": 6.58624119758606,
      "epoch": 0.5147143681477208,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004981915104332578,
      "loss": 6.3661,
      "mean_token_accuracy": 0.1594516083598137,
      "num_tokens": 11312912.0,
      "step": 4460
    },
    {
      "entropy": 6.457141399383545,
      "epoch": 0.5152914021927294,
      "grad_norm": 0.81640625,
      "learning_rate": 0.000498186285335584,
      "loss": 6.4066,
      "mean_token_accuracy": 0.1596623405814171,
      "num_tokens": 11325714.0,
      "step": 4465
    },
    {
      "entropy": 6.578498840332031,
      "epoch": 0.515868436237738,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004981810527311505,
      "loss": 6.3365,
      "mean_token_accuracy": 0.16815614998340606,
      "num_tokens": 11337820.0,
      "step": 4470
    },
    {
      "entropy": 6.45233736038208,
      "epoch": 0.5164454702827467,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000498175812620133,
      "loss": 6.3944,
      "mean_token_accuracy": 0.15770285874605178,
      "num_tokens": 11350005.0,
      "step": 4475
    },
    {
      "entropy": 6.5402703285217285,
      "epoch": 0.5170225043277553,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0004981705650027081,
      "loss": 6.4988,
      "mean_token_accuracy": 0.15301365703344344,
      "num_tokens": 11362918.0,
      "step": 4480
    },
    {
      "entropy": 6.550092267990112,
      "epoch": 0.517599538372764,
      "grad_norm": 0.8046875,
      "learning_rate": 0.000498165309879052,
      "loss": 6.3836,
      "mean_token_accuracy": 0.15631941705942154,
      "num_tokens": 11375186.0,
      "step": 4485
    },
    {
      "entropy": 6.527862882614135,
      "epoch": 0.5181765724177726,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004981600472493415,
      "loss": 6.4194,
      "mean_token_accuracy": 0.15206652581691743,
      "num_tokens": 11387607.0,
      "step": 4490
    },
    {
      "entropy": 6.546878957748413,
      "epoch": 0.5187536064627813,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004981547771137537,
      "loss": 6.437,
      "mean_token_accuracy": 0.1564124882221222,
      "num_tokens": 11400090.0,
      "step": 4495
    },
    {
      "entropy": 6.549637079238892,
      "epoch": 0.51933064050779,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004981494994724658,
      "loss": 6.4771,
      "mean_token_accuracy": 0.15591733604669572,
      "num_tokens": 11413220.0,
      "step": 4500
    },
    {
      "entropy": 6.531276607513428,
      "epoch": 0.5199076745527986,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004981442143256554,
      "loss": 6.3598,
      "mean_token_accuracy": 0.1604811280965805,
      "num_tokens": 11425043.0,
      "step": 4505
    },
    {
      "entropy": 6.486877870559693,
      "epoch": 0.5204847085978073,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004981389216735001,
      "loss": 6.3656,
      "mean_token_accuracy": 0.16314008086919785,
      "num_tokens": 11438097.0,
      "step": 4510
    },
    {
      "entropy": 6.48299241065979,
      "epoch": 0.5210617426428159,
      "grad_norm": 0.81640625,
      "learning_rate": 0.000498133621516178,
      "loss": 6.3548,
      "mean_token_accuracy": 0.16053190380334853,
      "num_tokens": 11450903.0,
      "step": 4515
    },
    {
      "entropy": 6.5034607410430905,
      "epoch": 0.5216387766878245,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004981283138538675,
      "loss": 6.4428,
      "mean_token_accuracy": 0.15942756831645966,
      "num_tokens": 11464566.0,
      "step": 4520
    },
    {
      "entropy": 6.546222782135009,
      "epoch": 0.5222158107328332,
      "grad_norm": 0.7734375,
      "learning_rate": 0.000498122998686747,
      "loss": 6.3884,
      "mean_token_accuracy": 0.1575324922800064,
      "num_tokens": 11476664.0,
      "step": 4525
    },
    {
      "entropy": 6.478293800354004,
      "epoch": 0.5227928447778419,
      "grad_norm": 0.875,
      "learning_rate": 0.0004981176760149951,
      "loss": 6.4246,
      "mean_token_accuracy": 0.15483495444059373,
      "num_tokens": 11489052.0,
      "step": 4530
    },
    {
      "entropy": 6.555764818191529,
      "epoch": 0.5233698788228506,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004981123458387911,
      "loss": 6.3773,
      "mean_token_accuracy": 0.1606239289045334,
      "num_tokens": 11501774.0,
      "step": 4535
    },
    {
      "entropy": 6.521160745620728,
      "epoch": 0.5239469128678592,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004981070081583142,
      "loss": 6.4976,
      "mean_token_accuracy": 0.15571219325065613,
      "num_tokens": 11514397.0,
      "step": 4540
    },
    {
      "entropy": 6.519374561309815,
      "epoch": 0.5245239469128679,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000498101662973744,
      "loss": 6.3928,
      "mean_token_accuracy": 0.15284379273653032,
      "num_tokens": 11527413.0,
      "step": 4545
    },
    {
      "entropy": 6.563087224960327,
      "epoch": 0.5251009809578765,
      "grad_norm": 0.7890625,
      "learning_rate": 0.00049809631028526,
      "loss": 6.4284,
      "mean_token_accuracy": 0.15203100591897964,
      "num_tokens": 11540423.0,
      "step": 4550
    },
    {
      "entropy": 6.506260728836059,
      "epoch": 0.5256780150028851,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004980909500930424,
      "loss": 6.3276,
      "mean_token_accuracy": 0.16709806472063066,
      "num_tokens": 11552825.0,
      "step": 4555
    },
    {
      "entropy": 6.4506385803222654,
      "epoch": 0.5262550490478938,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004980855823972717,
      "loss": 6.2924,
      "mean_token_accuracy": 0.16380308270454408,
      "num_tokens": 11565556.0,
      "step": 4560
    },
    {
      "entropy": 6.501029014587402,
      "epoch": 0.5268320830929025,
      "grad_norm": 0.76171875,
      "learning_rate": 0.000498080207198128,
      "loss": 6.382,
      "mean_token_accuracy": 0.16203862726688384,
      "num_tokens": 11580056.0,
      "step": 4565
    },
    {
      "entropy": 6.510374975204468,
      "epoch": 0.5274091171379112,
      "grad_norm": 0.7421875,
      "learning_rate": 0.0004980748244957925,
      "loss": 6.4179,
      "mean_token_accuracy": 0.1628774508833885,
      "num_tokens": 11592104.0,
      "step": 4570
    },
    {
      "entropy": 6.455078506469727,
      "epoch": 0.5279861511829198,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004980694342904461,
      "loss": 6.3118,
      "mean_token_accuracy": 0.15631030946969987,
      "num_tokens": 11605349.0,
      "step": 4575
    },
    {
      "entropy": 6.481748342514038,
      "epoch": 0.5285631852279284,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004980640365822701,
      "loss": 6.2951,
      "mean_token_accuracy": 0.1727045714855194,
      "num_tokens": 11617787.0,
      "step": 4580
    },
    {
      "entropy": 6.375321769714356,
      "epoch": 0.5291402192729371,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000498058631371446,
      "loss": 6.3024,
      "mean_token_accuracy": 0.16541579365730286,
      "num_tokens": 11630279.0,
      "step": 4585
    },
    {
      "entropy": 6.560987758636474,
      "epoch": 0.5297172533179457,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004980532186581555,
      "loss": 6.4075,
      "mean_token_accuracy": 0.16303146183490752,
      "num_tokens": 11643378.0,
      "step": 4590
    },
    {
      "entropy": 6.474534416198731,
      "epoch": 0.5302942873629544,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000498047798442581,
      "loss": 6.3745,
      "mean_token_accuracy": 0.16842037290334702,
      "num_tokens": 11656472.0,
      "step": 4595
    },
    {
      "entropy": 6.472138118743897,
      "epoch": 0.5308713214079631,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004980423707249044,
      "loss": 6.3307,
      "mean_token_accuracy": 0.16127256602048873,
      "num_tokens": 11669265.0,
      "step": 4600
    },
    {
      "entropy": 6.472711849212646,
      "epoch": 0.5314483554529718,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004980369355053086,
      "loss": 6.3725,
      "mean_token_accuracy": 0.15782807618379593,
      "num_tokens": 11681620.0,
      "step": 4605
    },
    {
      "entropy": 6.379305410385132,
      "epoch": 0.5320253894979804,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004980314927839763,
      "loss": 6.2888,
      "mean_token_accuracy": 0.16392303705215455,
      "num_tokens": 11695720.0,
      "step": 4610
    },
    {
      "entropy": 6.473198652267456,
      "epoch": 0.532602423542989,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004980260425610903,
      "loss": 6.3356,
      "mean_token_accuracy": 0.15988775640726088,
      "num_tokens": 11709563.0,
      "step": 4615
    },
    {
      "entropy": 6.553856039047242,
      "epoch": 0.5331794575879977,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004980205848368343,
      "loss": 6.3965,
      "mean_token_accuracy": 0.16205898225307463,
      "num_tokens": 11722401.0,
      "step": 4620
    },
    {
      "entropy": 6.4375251770019535,
      "epoch": 0.5337564916330063,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004980151196113917,
      "loss": 6.4087,
      "mean_token_accuracy": 0.16084825843572617,
      "num_tokens": 11735183.0,
      "step": 4625
    },
    {
      "entropy": 6.566055488586426,
      "epoch": 0.534333525678015,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004980096468849464,
      "loss": 6.3688,
      "mean_token_accuracy": 0.15413255393505096,
      "num_tokens": 11748282.0,
      "step": 4630
    },
    {
      "entropy": 6.47422194480896,
      "epoch": 0.5349105597230237,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004980041666576823,
      "loss": 6.3513,
      "mean_token_accuracy": 0.15646012276411056,
      "num_tokens": 11760208.0,
      "step": 4635
    },
    {
      "entropy": 6.466552066802978,
      "epoch": 0.5354875937680323,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004979986789297837,
      "loss": 6.4018,
      "mean_token_accuracy": 0.1543569967150688,
      "num_tokens": 11772509.0,
      "step": 4640
    },
    {
      "entropy": 6.59739408493042,
      "epoch": 0.536064627813041,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004979931837014355,
      "loss": 6.3366,
      "mean_token_accuracy": 0.1621736466884613,
      "num_tokens": 11786101.0,
      "step": 4645
    },
    {
      "entropy": 6.488185453414917,
      "epoch": 0.5366416618580496,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004979876809728223,
      "loss": 6.3939,
      "mean_token_accuracy": 0.15879629403352738,
      "num_tokens": 11798325.0,
      "step": 4650
    },
    {
      "entropy": 6.516392993927002,
      "epoch": 0.5372186959030583,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004979821707441292,
      "loss": 6.3919,
      "mean_token_accuracy": 0.15318617820739747,
      "num_tokens": 11812905.0,
      "step": 4655
    },
    {
      "entropy": 6.4433094501495365,
      "epoch": 0.5377957299480669,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004979766530155416,
      "loss": 6.2937,
      "mean_token_accuracy": 0.15997868925333023,
      "num_tokens": 11826495.0,
      "step": 4660
    },
    {
      "entropy": 6.516284656524658,
      "epoch": 0.5383727639930755,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004979711277872449,
      "loss": 6.4204,
      "mean_token_accuracy": 0.14770983308553695,
      "num_tokens": 11840472.0,
      "step": 4665
    },
    {
      "entropy": 6.557969570159912,
      "epoch": 0.5389497980380843,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004979655950594252,
      "loss": 6.3612,
      "mean_token_accuracy": 0.1606568932533264,
      "num_tokens": 11852800.0,
      "step": 4670
    },
    {
      "entropy": 6.572749853134155,
      "epoch": 0.5395268320830929,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004979600548322684,
      "loss": 6.3818,
      "mean_token_accuracy": 0.15735995769500732,
      "num_tokens": 11866111.0,
      "step": 4675
    },
    {
      "entropy": 6.396126937866211,
      "epoch": 0.5401038661281016,
      "grad_norm": 0.81640625,
      "learning_rate": 0.000497954507105961,
      "loss": 6.3845,
      "mean_token_accuracy": 0.15580861121416092,
      "num_tokens": 11878103.0,
      "step": 4680
    },
    {
      "entropy": 6.506599283218383,
      "epoch": 0.5406809001731102,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004979489518806893,
      "loss": 6.3564,
      "mean_token_accuracy": 0.16247568577528,
      "num_tokens": 11890851.0,
      "step": 4685
    },
    {
      "entropy": 6.555721282958984,
      "epoch": 0.5412579342181189,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004979433891566405,
      "loss": 6.3961,
      "mean_token_accuracy": 0.1531717985868454,
      "num_tokens": 11903351.0,
      "step": 4690
    },
    {
      "entropy": 6.504356765747071,
      "epoch": 0.5418349682631275,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004979378189340015,
      "loss": 6.434,
      "mean_token_accuracy": 0.1565772294998169,
      "num_tokens": 11915239.0,
      "step": 4695
    },
    {
      "entropy": 6.47059440612793,
      "epoch": 0.5424120023081361,
      "grad_norm": 1.125,
      "learning_rate": 0.0004979322412129597,
      "loss": 6.307,
      "mean_token_accuracy": 0.16283725053071976,
      "num_tokens": 11928678.0,
      "step": 4700
    },
    {
      "entropy": 6.532925081253052,
      "epoch": 0.5429890363531449,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004979266559937028,
      "loss": 6.3443,
      "mean_token_accuracy": 0.1592644363641739,
      "num_tokens": 11940547.0,
      "step": 4705
    },
    {
      "entropy": 6.495752763748169,
      "epoch": 0.5435660703981535,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004979210632764185,
      "loss": 6.3841,
      "mean_token_accuracy": 0.15600561648607253,
      "num_tokens": 11953069.0,
      "step": 4710
    },
    {
      "entropy": 6.4508391380310055,
      "epoch": 0.5441431044431622,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004979154630612949,
      "loss": 6.3361,
      "mean_token_accuracy": 0.16159862279891968,
      "num_tokens": 11964301.0,
      "step": 4715
    },
    {
      "entropy": 6.479263782501221,
      "epoch": 0.5447201384881708,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004979098553485205,
      "loss": 6.3493,
      "mean_token_accuracy": 0.15931008011102676,
      "num_tokens": 11976311.0,
      "step": 4720
    },
    {
      "entropy": 6.455762338638306,
      "epoch": 0.5452971725331794,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004979042401382838,
      "loss": 6.3316,
      "mean_token_accuracy": 0.1660098060965538,
      "num_tokens": 11989749.0,
      "step": 4725
    },
    {
      "entropy": 6.427380084991455,
      "epoch": 0.5458742065781881,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004978986174307737,
      "loss": 6.2826,
      "mean_token_accuracy": 0.16364272236824035,
      "num_tokens": 12002384.0,
      "step": 4730
    },
    {
      "entropy": 6.565292167663574,
      "epoch": 0.5464512406231967,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004978929872261794,
      "loss": 6.3748,
      "mean_token_accuracy": 0.1662522479891777,
      "num_tokens": 12014573.0,
      "step": 4735
    },
    {
      "entropy": 6.344450426101685,
      "epoch": 0.5470282746682055,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004978873495246901,
      "loss": 6.2243,
      "mean_token_accuracy": 0.16644109189510345,
      "num_tokens": 12027340.0,
      "step": 4740
    },
    {
      "entropy": 6.545307683944702,
      "epoch": 0.5476053087132141,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004978817043264955,
      "loss": 6.3823,
      "mean_token_accuracy": 0.159617817401886,
      "num_tokens": 12039498.0,
      "step": 4745
    },
    {
      "entropy": 6.530856990814209,
      "epoch": 0.5481823427582228,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004978760516317856,
      "loss": 6.3984,
      "mean_token_accuracy": 0.15844097435474397,
      "num_tokens": 12051426.0,
      "step": 4750
    },
    {
      "entropy": 6.496254634857178,
      "epoch": 0.5487593768032314,
      "grad_norm": 0.875,
      "learning_rate": 0.0004978703914407503,
      "loss": 6.3698,
      "mean_token_accuracy": 0.15838246494531633,
      "num_tokens": 12063975.0,
      "step": 4755
    },
    {
      "entropy": 6.514954042434693,
      "epoch": 0.54933641084824,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004978647237535802,
      "loss": 6.3347,
      "mean_token_accuracy": 0.16198110431432725,
      "num_tokens": 12075159.0,
      "step": 4760
    },
    {
      "entropy": 6.464260864257812,
      "epoch": 0.5499134448932487,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004978590485704657,
      "loss": 6.331,
      "mean_token_accuracy": 0.16103683412075043,
      "num_tokens": 12087052.0,
      "step": 4765
    },
    {
      "entropy": 6.4438018798828125,
      "epoch": 0.5504904789382573,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004978533658915979,
      "loss": 6.4177,
      "mean_token_accuracy": 0.15159978345036507,
      "num_tokens": 12100021.0,
      "step": 4770
    },
    {
      "entropy": 6.500536775588989,
      "epoch": 0.5510675129832661,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004978476757171678,
      "loss": 6.3866,
      "mean_token_accuracy": 0.1626763090491295,
      "num_tokens": 12113067.0,
      "step": 4775
    },
    {
      "entropy": 6.49054799079895,
      "epoch": 0.5516445470282747,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004978419780473668,
      "loss": 6.3858,
      "mean_token_accuracy": 0.16142747104167937,
      "num_tokens": 12126108.0,
      "step": 4780
    },
    {
      "entropy": 6.436980056762695,
      "epoch": 0.5522215810732833,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004978362728823865,
      "loss": 6.2368,
      "mean_token_accuracy": 0.16867295056581497,
      "num_tokens": 12138821.0,
      "step": 4785
    },
    {
      "entropy": 6.561331415176392,
      "epoch": 0.552798615118292,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004978305602224189,
      "loss": 6.4602,
      "mean_token_accuracy": 0.15451819002628325,
      "num_tokens": 12151621.0,
      "step": 4790
    },
    {
      "entropy": 6.565451955795288,
      "epoch": 0.5533756491633006,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004978248400676562,
      "loss": 6.4049,
      "mean_token_accuracy": 0.15947678834199905,
      "num_tokens": 12163232.0,
      "step": 4795
    },
    {
      "entropy": 6.437511682510376,
      "epoch": 0.5539526832083093,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004978191124182905,
      "loss": 6.3774,
      "mean_token_accuracy": 0.15531412959098817,
      "num_tokens": 12176857.0,
      "step": 4800
    },
    {
      "entropy": 6.3844664096832275,
      "epoch": 0.5545297172533179,
      "grad_norm": 0.74609375,
      "learning_rate": 0.0004978133772745149,
      "loss": 6.2797,
      "mean_token_accuracy": 0.16717115342617034,
      "num_tokens": 12190311.0,
      "step": 4805
    },
    {
      "entropy": 6.522913789749145,
      "epoch": 0.5551067512983267,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004978076346365218,
      "loss": 6.3607,
      "mean_token_accuracy": 0.15984816551208497,
      "num_tokens": 12202490.0,
      "step": 4810
    },
    {
      "entropy": 6.519549751281739,
      "epoch": 0.5556837853433353,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004978018845045047,
      "loss": 6.3681,
      "mean_token_accuracy": 0.1594349056482315,
      "num_tokens": 12214759.0,
      "step": 4815
    },
    {
      "entropy": 6.502993392944336,
      "epoch": 0.5562608193883439,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004977961268786568,
      "loss": 6.3037,
      "mean_token_accuracy": 0.1682477280497551,
      "num_tokens": 12227639.0,
      "step": 4820
    },
    {
      "entropy": 6.485041999816895,
      "epoch": 0.5568378534333526,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004977903617591719,
      "loss": 6.3981,
      "mean_token_accuracy": 0.15737620294094085,
      "num_tokens": 12241049.0,
      "step": 4825
    },
    {
      "entropy": 6.483181476593018,
      "epoch": 0.5574148874783612,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004977845891462439,
      "loss": 6.2891,
      "mean_token_accuracy": 0.16137616634368895,
      "num_tokens": 12253980.0,
      "step": 4830
    },
    {
      "entropy": 6.440955638885498,
      "epoch": 0.5579919215233698,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004977788090400668,
      "loss": 6.3626,
      "mean_token_accuracy": 0.16536644846200943,
      "num_tokens": 12265103.0,
      "step": 4835
    },
    {
      "entropy": 6.425109720230102,
      "epoch": 0.5585689555683785,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004977730214408352,
      "loss": 6.3493,
      "mean_token_accuracy": 0.1596547245979309,
      "num_tokens": 12278574.0,
      "step": 4840
    },
    {
      "entropy": 6.427152872085571,
      "epoch": 0.5591459896133872,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004977672263487435,
      "loss": 6.3435,
      "mean_token_accuracy": 0.16322456002235414,
      "num_tokens": 12291038.0,
      "step": 4845
    },
    {
      "entropy": 6.527709913253784,
      "epoch": 0.5597230236583959,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000497761423763987,
      "loss": 6.4031,
      "mean_token_accuracy": 0.1564134478569031,
      "num_tokens": 12304339.0,
      "step": 4850
    },
    {
      "entropy": 6.523615217208862,
      "epoch": 0.5603000577034045,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004977556136867606,
      "loss": 6.4104,
      "mean_token_accuracy": 0.15531465411186218,
      "num_tokens": 12317927.0,
      "step": 4855
    },
    {
      "entropy": 6.5276378154754635,
      "epoch": 0.5608770917484132,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0004977497961172598,
      "loss": 6.3349,
      "mean_token_accuracy": 0.15690107643604279,
      "num_tokens": 12332904.0,
      "step": 4860
    },
    {
      "entropy": 6.502278852462768,
      "epoch": 0.5614541257934218,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004977439710556802,
      "loss": 6.2659,
      "mean_token_accuracy": 0.16384697556495667,
      "num_tokens": 12345449.0,
      "step": 4865
    },
    {
      "entropy": 6.3931052684783936,
      "epoch": 0.5620311598384304,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004977381385022179,
      "loss": 6.319,
      "mean_token_accuracy": 0.1633118733763695,
      "num_tokens": 12357334.0,
      "step": 4870
    },
    {
      "entropy": 6.561559820175171,
      "epoch": 0.5626081938834391,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004977322984570688,
      "loss": 6.3612,
      "mean_token_accuracy": 0.15845298767089844,
      "num_tokens": 12370094.0,
      "step": 4875
    },
    {
      "entropy": 6.466009855270386,
      "epoch": 0.5631852279284478,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004977264509204296,
      "loss": 6.3963,
      "mean_token_accuracy": 0.15249388813972473,
      "num_tokens": 12382875.0,
      "step": 4880
    },
    {
      "entropy": 6.431724596023559,
      "epoch": 0.5637622619734565,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004977205958924967,
      "loss": 6.3243,
      "mean_token_accuracy": 0.15520370602607728,
      "num_tokens": 12395534.0,
      "step": 4885
    },
    {
      "entropy": 6.492003107070923,
      "epoch": 0.5643392960184651,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004977147333734673,
      "loss": 6.3458,
      "mean_token_accuracy": 0.15953391343355178,
      "num_tokens": 12407300.0,
      "step": 4890
    },
    {
      "entropy": 6.409242057800293,
      "epoch": 0.5649163300634737,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004977088633635385,
      "loss": 6.2632,
      "mean_token_accuracy": 0.1659919276833534,
      "num_tokens": 12419499.0,
      "step": 4895
    },
    {
      "entropy": 6.421889972686768,
      "epoch": 0.5654933641084824,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004977029858629076,
      "loss": 6.2121,
      "mean_token_accuracy": 0.1685824289917946,
      "num_tokens": 12431279.0,
      "step": 4900
    },
    {
      "entropy": 6.295782613754272,
      "epoch": 0.566070398153491,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004976971008717726,
      "loss": 6.1719,
      "mean_token_accuracy": 0.17155274003744125,
      "num_tokens": 12443976.0,
      "step": 4905
    },
    {
      "entropy": 6.393910932540893,
      "epoch": 0.5666474321984997,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004976912083903309,
      "loss": 6.3299,
      "mean_token_accuracy": 0.16340630799531936,
      "num_tokens": 12457752.0,
      "step": 4910
    },
    {
      "entropy": 6.492015361785889,
      "epoch": 0.5672244662435084,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004976853084187814,
      "loss": 6.34,
      "mean_token_accuracy": 0.15709138363599778,
      "num_tokens": 12469776.0,
      "step": 4915
    },
    {
      "entropy": 6.52558970451355,
      "epoch": 0.5678015002885171,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004976794009573219,
      "loss": 6.3669,
      "mean_token_accuracy": 0.14939526617527008,
      "num_tokens": 12483300.0,
      "step": 4920
    },
    {
      "entropy": 6.38107852935791,
      "epoch": 0.5683785343335257,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004976734860061515,
      "loss": 6.3129,
      "mean_token_accuracy": 0.16156516522169112,
      "num_tokens": 12494564.0,
      "step": 4925
    },
    {
      "entropy": 6.423398447036743,
      "epoch": 0.5689555683785343,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004976675635654688,
      "loss": 6.3394,
      "mean_token_accuracy": 0.16027214378118515,
      "num_tokens": 12507761.0,
      "step": 4930
    },
    {
      "entropy": 6.51806640625,
      "epoch": 0.569532602423543,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004976616336354733,
      "loss": 6.331,
      "mean_token_accuracy": 0.16270025670528412,
      "num_tokens": 12520045.0,
      "step": 4935
    },
    {
      "entropy": 6.4369653224945065,
      "epoch": 0.5701096364685516,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004976556962163645,
      "loss": 6.2544,
      "mean_token_accuracy": 0.16926467120647432,
      "num_tokens": 12532471.0,
      "step": 4940
    },
    {
      "entropy": 6.468795585632324,
      "epoch": 0.5706866705135603,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004976497513083419,
      "loss": 6.3643,
      "mean_token_accuracy": 0.15875103324651718,
      "num_tokens": 12544610.0,
      "step": 4945
    },
    {
      "entropy": 6.472221326828003,
      "epoch": 0.571263704558569,
      "grad_norm": 0.875,
      "learning_rate": 0.0004976437989116055,
      "loss": 6.3786,
      "mean_token_accuracy": 0.16317480951547622,
      "num_tokens": 12558682.0,
      "step": 4950
    },
    {
      "entropy": 6.466315221786499,
      "epoch": 0.5718407386035776,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004976378390263554,
      "loss": 6.2759,
      "mean_token_accuracy": 0.15960074067115784,
      "num_tokens": 12570711.0,
      "step": 4955
    },
    {
      "entropy": 6.410310411453247,
      "epoch": 0.5724177726485863,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004976318716527924,
      "loss": 6.3023,
      "mean_token_accuracy": 0.16249395459890364,
      "num_tokens": 12582468.0,
      "step": 4960
    },
    {
      "entropy": 6.483807325363159,
      "epoch": 0.5729948066935949,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004976258967911168,
      "loss": 6.3381,
      "mean_token_accuracy": 0.16272847652435302,
      "num_tokens": 12594991.0,
      "step": 4965
    },
    {
      "entropy": 6.498748826980591,
      "epoch": 0.5735718407386036,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004976199144415298,
      "loss": 6.3175,
      "mean_token_accuracy": 0.16240322440862656,
      "num_tokens": 12606928.0,
      "step": 4970
    },
    {
      "entropy": 6.455127906799317,
      "epoch": 0.5741488747836122,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004976139246042325,
      "loss": 6.4012,
      "mean_token_accuracy": 0.15524941086769103,
      "num_tokens": 12620539.0,
      "step": 4975
    },
    {
      "entropy": 6.4773979663848875,
      "epoch": 0.5747259088286208,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004976079272794265,
      "loss": 6.3446,
      "mean_token_accuracy": 0.15420007407665254,
      "num_tokens": 12632629.0,
      "step": 4980
    },
    {
      "entropy": 6.43410210609436,
      "epoch": 0.5753029428736296,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004976019224673134,
      "loss": 6.3069,
      "mean_token_accuracy": 0.16904855519533157,
      "num_tokens": 12646631.0,
      "step": 4985
    },
    {
      "entropy": 6.449543762207031,
      "epoch": 0.5758799769186382,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004975959101680953,
      "loss": 6.3019,
      "mean_token_accuracy": 0.16361594051122666,
      "num_tokens": 12659154.0,
      "step": 4990
    },
    {
      "entropy": 6.447082471847534,
      "epoch": 0.5764570109636469,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004975898903819742,
      "loss": 6.3907,
      "mean_token_accuracy": 0.1598036080598831,
      "num_tokens": 12671796.0,
      "step": 4995
    },
    {
      "entropy": 6.5214070796966555,
      "epoch": 0.5770340450086555,
      "grad_norm": 0.875,
      "learning_rate": 0.0004975838631091527,
      "loss": 6.3528,
      "mean_token_accuracy": 0.15806207805871964,
      "num_tokens": 12684432.0,
      "step": 5000
    },
    {
      "entropy": 6.488364028930664,
      "epoch": 0.5776110790536642,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004975778283498336,
      "loss": 6.3571,
      "mean_token_accuracy": 0.16473528295755385,
      "num_tokens": 12697585.0,
      "step": 5005
    },
    {
      "entropy": 6.400859832763672,
      "epoch": 0.5781881130986728,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004975717861042198,
      "loss": 6.2783,
      "mean_token_accuracy": 0.1685378298163414,
      "num_tokens": 12710654.0,
      "step": 5010
    },
    {
      "entropy": 6.401386976242065,
      "epoch": 0.5787651471436814,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004975657363725146,
      "loss": 6.3226,
      "mean_token_accuracy": 0.15781749784946442,
      "num_tokens": 12723141.0,
      "step": 5015
    },
    {
      "entropy": 6.534639596939087,
      "epoch": 0.5793421811886902,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004975596791549213,
      "loss": 6.4103,
      "mean_token_accuracy": 0.15141381472349166,
      "num_tokens": 12735921.0,
      "step": 5020
    },
    {
      "entropy": 6.395477533340454,
      "epoch": 0.5799192152336988,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004975536144516437,
      "loss": 6.3753,
      "mean_token_accuracy": 0.15583303123712539,
      "num_tokens": 12747704.0,
      "step": 5025
    },
    {
      "entropy": 6.563265943527222,
      "epoch": 0.5804962492787075,
      "grad_norm": 1.109375,
      "learning_rate": 0.000497547542262886,
      "loss": 6.3668,
      "mean_token_accuracy": 0.15204933732748033,
      "num_tokens": 12759368.0,
      "step": 5030
    },
    {
      "entropy": 6.520508766174316,
      "epoch": 0.5810732833237161,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004975414625888521,
      "loss": 6.3235,
      "mean_token_accuracy": 0.16076601594686507,
      "num_tokens": 12772226.0,
      "step": 5035
    },
    {
      "entropy": 6.515491724014282,
      "epoch": 0.5816503173687247,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004975353754297468,
      "loss": 6.3922,
      "mean_token_accuracy": 0.15815138816833496,
      "num_tokens": 12784817.0,
      "step": 5040
    },
    {
      "entropy": 6.4733335971832275,
      "epoch": 0.5822273514137334,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004975292807857745,
      "loss": 6.3285,
      "mean_token_accuracy": 0.16127829551696776,
      "num_tokens": 12796996.0,
      "step": 5045
    },
    {
      "entropy": 6.462998390197754,
      "epoch": 0.582804385458742,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004975231786571406,
      "loss": 6.3803,
      "mean_token_accuracy": 0.15985668301582337,
      "num_tokens": 12809581.0,
      "step": 5050
    },
    {
      "entropy": 6.427072811126709,
      "epoch": 0.5833814195037508,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004975170690440499,
      "loss": 6.3413,
      "mean_token_accuracy": 0.1596603661775589,
      "num_tokens": 12822060.0,
      "step": 5055
    },
    {
      "entropy": 6.490553283691407,
      "epoch": 0.5839584535487594,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004975109519467083,
      "loss": 6.3249,
      "mean_token_accuracy": 0.1605900213122368,
      "num_tokens": 12835554.0,
      "step": 5060
    },
    {
      "entropy": 6.465603160858154,
      "epoch": 0.584535487593768,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004975048273653212,
      "loss": 6.2566,
      "mean_token_accuracy": 0.17011079341173171,
      "num_tokens": 12848187.0,
      "step": 5065
    },
    {
      "entropy": 6.394630336761475,
      "epoch": 0.5851125216387767,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004974986953000948,
      "loss": 6.2778,
      "mean_token_accuracy": 0.16856665015220643,
      "num_tokens": 12861498.0,
      "step": 5070
    },
    {
      "entropy": 6.557682037353516,
      "epoch": 0.5856895556837853,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004974925557512354,
      "loss": 6.459,
      "mean_token_accuracy": 0.1463148832321167,
      "num_tokens": 12874906.0,
      "step": 5075
    },
    {
      "entropy": 6.386854028701782,
      "epoch": 0.586266589728794,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004974864087189493,
      "loss": 6.2865,
      "mean_token_accuracy": 0.16323204934597016,
      "num_tokens": 12887169.0,
      "step": 5080
    },
    {
      "entropy": 6.466772794723511,
      "epoch": 0.5868436237738026,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004974802542034434,
      "loss": 6.3339,
      "mean_token_accuracy": 0.16101205796003343,
      "num_tokens": 12899639.0,
      "step": 5085
    },
    {
      "entropy": 6.465685510635376,
      "epoch": 0.5874206578188114,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004974740922049246,
      "loss": 6.2465,
      "mean_token_accuracy": 0.16204580515623093,
      "num_tokens": 12911913.0,
      "step": 5090
    },
    {
      "entropy": 6.532586860656738,
      "epoch": 0.58799769186382,
      "grad_norm": 0.74609375,
      "learning_rate": 0.0004974679227236004,
      "loss": 6.4113,
      "mean_token_accuracy": 0.15404328852891921,
      "num_tokens": 12924200.0,
      "step": 5095
    },
    {
      "entropy": 6.4691802024841305,
      "epoch": 0.5885747259088286,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004974617457596779,
      "loss": 6.3971,
      "mean_token_accuracy": 0.15657893866300582,
      "num_tokens": 12936726.0,
      "step": 5100
    },
    {
      "entropy": 6.508554601669312,
      "epoch": 0.5891517599538373,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004974555613133652,
      "loss": 6.2811,
      "mean_token_accuracy": 0.159566231071949,
      "num_tokens": 12949751.0,
      "step": 5105
    },
    {
      "entropy": 6.475770330429077,
      "epoch": 0.5897287939988459,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004974493693848702,
      "loss": 6.3737,
      "mean_token_accuracy": 0.1601461336016655,
      "num_tokens": 12962573.0,
      "step": 5110
    },
    {
      "entropy": 6.506417083740234,
      "epoch": 0.5903058280438546,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004974431699744011,
      "loss": 6.4022,
      "mean_token_accuracy": 0.15874024480581284,
      "num_tokens": 12975635.0,
      "step": 5115
    },
    {
      "entropy": 6.464160919189453,
      "epoch": 0.5908828620888632,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004974369630821665,
      "loss": 6.3972,
      "mean_token_accuracy": 0.15452799648046495,
      "num_tokens": 12987822.0,
      "step": 5120
    },
    {
      "entropy": 6.50497260093689,
      "epoch": 0.5914598961338718,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004974307487083752,
      "loss": 6.347,
      "mean_token_accuracy": 0.16242809891700744,
      "num_tokens": 13000082.0,
      "step": 5125
    },
    {
      "entropy": 6.421209764480591,
      "epoch": 0.5920369301788806,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004974245268532361,
      "loss": 6.2443,
      "mean_token_accuracy": 0.16429829150438308,
      "num_tokens": 13011902.0,
      "step": 5130
    },
    {
      "entropy": 6.432503414154053,
      "epoch": 0.5926139642238892,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0004974182975169585,
      "loss": 6.2979,
      "mean_token_accuracy": 0.16235667765140532,
      "num_tokens": 13024507.0,
      "step": 5135
    },
    {
      "entropy": 6.4553704261779785,
      "epoch": 0.5931909982688979,
      "grad_norm": 0.8125,
      "learning_rate": 0.000497412060699752,
      "loss": 6.4016,
      "mean_token_accuracy": 0.16003545448184014,
      "num_tokens": 13037955.0,
      "step": 5140
    },
    {
      "entropy": 6.525091123580933,
      "epoch": 0.5937680323139065,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004974058164018263,
      "loss": 6.351,
      "mean_token_accuracy": 0.1640540525317192,
      "num_tokens": 13050631.0,
      "step": 5145
    },
    {
      "entropy": 6.477291822433472,
      "epoch": 0.5943450663589152,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004973995646233914,
      "loss": 6.2996,
      "mean_token_accuracy": 0.16100106835365297,
      "num_tokens": 13062566.0,
      "step": 5150
    },
    {
      "entropy": 6.407656812667847,
      "epoch": 0.5949221004039238,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004973933053646576,
      "loss": 6.2747,
      "mean_token_accuracy": 0.16398902982473373,
      "num_tokens": 13075454.0,
      "step": 5155
    },
    {
      "entropy": 6.498917055130005,
      "epoch": 0.5954991344489324,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004973870386258355,
      "loss": 6.3468,
      "mean_token_accuracy": 0.15738814622163771,
      "num_tokens": 13087780.0,
      "step": 5160
    },
    {
      "entropy": 6.477192354202271,
      "epoch": 0.5960761684939412,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004973807644071357,
      "loss": 6.3504,
      "mean_token_accuracy": 0.15569237619638443,
      "num_tokens": 13100957.0,
      "step": 5165
    },
    {
      "entropy": 6.460265016555786,
      "epoch": 0.5966532025389498,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0004973744827087695,
      "loss": 6.2701,
      "mean_token_accuracy": 0.16348017752170563,
      "num_tokens": 13113232.0,
      "step": 5170
    },
    {
      "entropy": 6.407651424407959,
      "epoch": 0.5972302365839585,
      "grad_norm": 0.84375,
      "learning_rate": 0.000497368193530948,
      "loss": 6.2438,
      "mean_token_accuracy": 0.16760450899600982,
      "num_tokens": 13125950.0,
      "step": 5175
    },
    {
      "entropy": 6.380605173110962,
      "epoch": 0.5978072706289671,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004973618968738828,
      "loss": 6.2429,
      "mean_token_accuracy": 0.16398492008447646,
      "num_tokens": 13138936.0,
      "step": 5180
    },
    {
      "entropy": 6.498394584655761,
      "epoch": 0.5983843046739757,
      "grad_norm": 0.71875,
      "learning_rate": 0.0004973555927377856,
      "loss": 6.3801,
      "mean_token_accuracy": 0.1577525481581688,
      "num_tokens": 13152787.0,
      "step": 5185
    },
    {
      "entropy": 6.453382253646851,
      "epoch": 0.5989613387189844,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004973492811228685,
      "loss": 6.3077,
      "mean_token_accuracy": 0.162015600502491,
      "num_tokens": 13165459.0,
      "step": 5190
    },
    {
      "entropy": 6.52237606048584,
      "epoch": 0.599538372763993,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004973429620293438,
      "loss": 6.3203,
      "mean_token_accuracy": 0.1562004953622818,
      "num_tokens": 13178766.0,
      "step": 5195
    },
    {
      "entropy": 6.502490377426147,
      "epoch": 0.6001154068090018,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004973366354574239,
      "loss": 6.4284,
      "mean_token_accuracy": 0.1566441461443901,
      "num_tokens": 13193712.0,
      "step": 5200
    },
    {
      "entropy": 6.505635213851929,
      "epoch": 0.6006924408540104,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004973303014073219,
      "loss": 6.4283,
      "mean_token_accuracy": 0.1619483232498169,
      "num_tokens": 13207363.0,
      "step": 5205
    },
    {
      "entropy": 6.441206932067871,
      "epoch": 0.601269474899019,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004973239598792504,
      "loss": 6.2197,
      "mean_token_accuracy": 0.16894505620002748,
      "num_tokens": 13219131.0,
      "step": 5210
    },
    {
      "entropy": 6.462107229232788,
      "epoch": 0.6018465089440277,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004973176108734232,
      "loss": 6.3294,
      "mean_token_accuracy": 0.1623179242014885,
      "num_tokens": 13231595.0,
      "step": 5215
    },
    {
      "entropy": 6.4451995372772215,
      "epoch": 0.6024235429890363,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004973112543900535,
      "loss": 6.3424,
      "mean_token_accuracy": 0.16196154356002807,
      "num_tokens": 13243727.0,
      "step": 5220
    },
    {
      "entropy": 6.453670310974121,
      "epoch": 0.603000577034045,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004973048904293551,
      "loss": 6.3048,
      "mean_token_accuracy": 0.16469427496194838,
      "num_tokens": 13255002.0,
      "step": 5225
    },
    {
      "entropy": 6.488593482971192,
      "epoch": 0.6035776110790536,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004972985189915422,
      "loss": 6.3417,
      "mean_token_accuracy": 0.16113921850919724,
      "num_tokens": 13268523.0,
      "step": 5230
    },
    {
      "entropy": 6.368151140213013,
      "epoch": 0.6041546451240624,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000497292140076829,
      "loss": 6.2508,
      "mean_token_accuracy": 0.1658511444926262,
      "num_tokens": 13280625.0,
      "step": 5235
    },
    {
      "entropy": 6.551406383514404,
      "epoch": 0.604731679169071,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004972857536854301,
      "loss": 6.4091,
      "mean_token_accuracy": 0.15450685024261473,
      "num_tokens": 13294880.0,
      "step": 5240
    },
    {
      "entropy": 6.51069598197937,
      "epoch": 0.6053087132140796,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004972793598175603,
      "loss": 6.3424,
      "mean_token_accuracy": 0.15950118750333786,
      "num_tokens": 13307875.0,
      "step": 5245
    },
    {
      "entropy": 6.417842197418213,
      "epoch": 0.6058857472590883,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004972729584734347,
      "loss": 6.2173,
      "mean_token_accuracy": 0.16876524686813354,
      "num_tokens": 13320067.0,
      "step": 5250
    },
    {
      "entropy": 6.375647735595703,
      "epoch": 0.6064627813040969,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004972665496532685,
      "loss": 6.2833,
      "mean_token_accuracy": 0.16125652939081192,
      "num_tokens": 13332036.0,
      "step": 5255
    },
    {
      "entropy": 6.473684883117675,
      "epoch": 0.6070398153491056,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004972601333572774,
      "loss": 6.3059,
      "mean_token_accuracy": 0.1631132811307907,
      "num_tokens": 13343831.0,
      "step": 5260
    },
    {
      "entropy": 6.508532238006592,
      "epoch": 0.6076168493941142,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004972537095856769,
      "loss": 6.3611,
      "mean_token_accuracy": 0.15429210513830185,
      "num_tokens": 13356452.0,
      "step": 5265
    },
    {
      "entropy": 6.461349248886108,
      "epoch": 0.608193883439123,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004972472783386834,
      "loss": 6.2744,
      "mean_token_accuracy": 0.1614803835749626,
      "num_tokens": 13369052.0,
      "step": 5270
    },
    {
      "entropy": 6.479144048690796,
      "epoch": 0.6087709174841316,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004972408396165132,
      "loss": 6.3656,
      "mean_token_accuracy": 0.15780351161956788,
      "num_tokens": 13381943.0,
      "step": 5275
    },
    {
      "entropy": 6.455678510665893,
      "epoch": 0.6093479515291402,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004972343934193826,
      "loss": 6.3247,
      "mean_token_accuracy": 0.15625445544719696,
      "num_tokens": 13394069.0,
      "step": 5280
    },
    {
      "entropy": 6.525052213668824,
      "epoch": 0.6099249855741489,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004972279397475086,
      "loss": 6.3845,
      "mean_token_accuracy": 0.1577905982732773,
      "num_tokens": 13407397.0,
      "step": 5285
    },
    {
      "entropy": 6.425345516204834,
      "epoch": 0.6105020196191575,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004972214786011083,
      "loss": 6.3092,
      "mean_token_accuracy": 0.1553454354405403,
      "num_tokens": 13419716.0,
      "step": 5290
    },
    {
      "entropy": 6.471757507324218,
      "epoch": 0.6110790536641661,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000497215009980399,
      "loss": 6.2705,
      "mean_token_accuracy": 0.15723506659269332,
      "num_tokens": 13432839.0,
      "step": 5295
    },
    {
      "entropy": 6.493448495864868,
      "epoch": 0.6116560877091748,
      "grad_norm": 0.7734375,
      "learning_rate": 0.000497208533885598,
      "loss": 6.314,
      "mean_token_accuracy": 0.15734073221683503,
      "num_tokens": 13445318.0,
      "step": 5300
    },
    {
      "entropy": 6.421839809417724,
      "epoch": 0.6122331217541835,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004972020503169235,
      "loss": 6.2926,
      "mean_token_accuracy": 0.1610089898109436,
      "num_tokens": 13456652.0,
      "step": 5305
    },
    {
      "entropy": 6.419455242156983,
      "epoch": 0.6128101557991922,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004971955592745934,
      "loss": 6.2935,
      "mean_token_accuracy": 0.15970533713698387,
      "num_tokens": 13469133.0,
      "step": 5310
    },
    {
      "entropy": 6.470340442657471,
      "epoch": 0.6133871898442008,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000497189060758826,
      "loss": 6.3026,
      "mean_token_accuracy": 0.16227268874645234,
      "num_tokens": 13481741.0,
      "step": 5315
    },
    {
      "entropy": 6.43679633140564,
      "epoch": 0.6139642238892095,
      "grad_norm": 0.984375,
      "learning_rate": 0.00049718255476984,
      "loss": 6.3954,
      "mean_token_accuracy": 0.15243706703186036,
      "num_tokens": 13494257.0,
      "step": 5320
    },
    {
      "entropy": 6.415068626403809,
      "epoch": 0.6145412579342181,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004971760413078539,
      "loss": 6.2579,
      "mean_token_accuracy": 0.1738950192928314,
      "num_tokens": 13507714.0,
      "step": 5325
    },
    {
      "entropy": 6.428360319137573,
      "epoch": 0.6151182919792267,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004971695203730872,
      "loss": 6.3064,
      "mean_token_accuracy": 0.1591736286878586,
      "num_tokens": 13519499.0,
      "step": 5330
    },
    {
      "entropy": 6.4912420272827145,
      "epoch": 0.6156953260242354,
      "grad_norm": 0.80078125,
      "learning_rate": 0.000497162991965759,
      "loss": 6.2825,
      "mean_token_accuracy": 0.16679684966802596,
      "num_tokens": 13531538.0,
      "step": 5335
    },
    {
      "entropy": 6.553953790664673,
      "epoch": 0.6162723600692441,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004971564560860889,
      "loss": 6.352,
      "mean_token_accuracy": 0.15729986131191254,
      "num_tokens": 13544576.0,
      "step": 5340
    },
    {
      "entropy": 6.415745735168457,
      "epoch": 0.6168493941142528,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004971499127342968,
      "loss": 6.297,
      "mean_token_accuracy": 0.15716785192489624,
      "num_tokens": 13557318.0,
      "step": 5345
    },
    {
      "entropy": 6.352963447570801,
      "epoch": 0.6174264281592614,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004971433619106027,
      "loss": 6.1886,
      "mean_token_accuracy": 0.16966686099767686,
      "num_tokens": 13569334.0,
      "step": 5350
    },
    {
      "entropy": 6.437016868591309,
      "epoch": 0.61800346220427,
      "grad_norm": 0.875,
      "learning_rate": 0.000497136803615227,
      "loss": 6.2662,
      "mean_token_accuracy": 0.1659908726811409,
      "num_tokens": 13581748.0,
      "step": 5355
    },
    {
      "entropy": 6.47385802268982,
      "epoch": 0.6185804962492787,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004971302378483902,
      "loss": 6.3469,
      "mean_token_accuracy": 0.16016255766153337,
      "num_tokens": 13593892.0,
      "step": 5360
    },
    {
      "entropy": 6.445749044418335,
      "epoch": 0.6191575302942873,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004971236646103132,
      "loss": 6.2198,
      "mean_token_accuracy": 0.16264959871768953,
      "num_tokens": 13605593.0,
      "step": 5365
    },
    {
      "entropy": 6.426254367828369,
      "epoch": 0.619734564339296,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004971170839012171,
      "loss": 6.402,
      "mean_token_accuracy": 0.1573261246085167,
      "num_tokens": 13618356.0,
      "step": 5370
    },
    {
      "entropy": 6.485814094543457,
      "epoch": 0.6203115983843047,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004971104957213233,
      "loss": 6.3457,
      "mean_token_accuracy": 0.1604474663734436,
      "num_tokens": 13629939.0,
      "step": 5375
    },
    {
      "entropy": 6.3256737232208256,
      "epoch": 0.6208886324293134,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004971039000708531,
      "loss": 6.1998,
      "mean_token_accuracy": 0.17142249196767806,
      "num_tokens": 13642123.0,
      "step": 5380
    },
    {
      "entropy": 6.41780834197998,
      "epoch": 0.621465666474322,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004970972969500286,
      "loss": 6.2908,
      "mean_token_accuracy": 0.16614297032356262,
      "num_tokens": 13654956.0,
      "step": 5385
    },
    {
      "entropy": 6.4260025978088375,
      "epoch": 0.6220427005193306,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000497090686359072,
      "loss": 6.2563,
      "mean_token_accuracy": 0.16691885590553285,
      "num_tokens": 13667549.0,
      "step": 5390
    },
    {
      "entropy": 6.461333417892456,
      "epoch": 0.6226197345643393,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004970840682982054,
      "loss": 6.3172,
      "mean_token_accuracy": 0.16463381201028823,
      "num_tokens": 13680747.0,
      "step": 5395
    },
    {
      "entropy": 6.487573194503784,
      "epoch": 0.6231967686093479,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004970774427676514,
      "loss": 6.3576,
      "mean_token_accuracy": 0.15117861852049827,
      "num_tokens": 13695500.0,
      "step": 5400
    },
    {
      "entropy": 6.485534191131592,
      "epoch": 0.6237738026543566,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000497070809767633,
      "loss": 6.327,
      "mean_token_accuracy": 0.15981363877654076,
      "num_tokens": 13707807.0,
      "step": 5405
    },
    {
      "entropy": 6.407062005996704,
      "epoch": 0.6243508366993653,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004970641692983731,
      "loss": 6.2726,
      "mean_token_accuracy": 0.16532813906669616,
      "num_tokens": 13720401.0,
      "step": 5410
    },
    {
      "entropy": 6.414212942123413,
      "epoch": 0.624927870744374,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004970575213600954,
      "loss": 6.2596,
      "mean_token_accuracy": 0.16263928413391113,
      "num_tokens": 13732611.0,
      "step": 5415
    },
    {
      "entropy": 6.413683795928955,
      "epoch": 0.6255049047893826,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004970508659530231,
      "loss": 6.3135,
      "mean_token_accuracy": 0.1561568185687065,
      "num_tokens": 13744876.0,
      "step": 5420
    },
    {
      "entropy": 6.420420742034912,
      "epoch": 0.6260819388343912,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004970442030773802,
      "loss": 6.3382,
      "mean_token_accuracy": 0.16061879694461823,
      "num_tokens": 13757265.0,
      "step": 5425
    },
    {
      "entropy": 6.492178726196289,
      "epoch": 0.6266589728793999,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004970375327333909,
      "loss": 6.2707,
      "mean_token_accuracy": 0.16342684477567673,
      "num_tokens": 13770101.0,
      "step": 5430
    },
    {
      "entropy": 6.485668134689331,
      "epoch": 0.6272360069244085,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004970308549212796,
      "loss": 6.3115,
      "mean_token_accuracy": 0.16133440434932708,
      "num_tokens": 13784104.0,
      "step": 5435
    },
    {
      "entropy": 6.386432552337647,
      "epoch": 0.6278130409694171,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004970241696412705,
      "loss": 6.2359,
      "mean_token_accuracy": 0.16974283754825592,
      "num_tokens": 13796901.0,
      "step": 5440
    },
    {
      "entropy": 6.43957667350769,
      "epoch": 0.6283900750144259,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000497017476893589,
      "loss": 6.3112,
      "mean_token_accuracy": 0.1625409170985222,
      "num_tokens": 13809640.0,
      "step": 5445
    },
    {
      "entropy": 6.53232970237732,
      "epoch": 0.6289671090594345,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004970107766784598,
      "loss": 6.3116,
      "mean_token_accuracy": 0.16599140763282777,
      "num_tokens": 13821684.0,
      "step": 5450
    },
    {
      "entropy": 6.448949337005615,
      "epoch": 0.6295441431044432,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004970040689961084,
      "loss": 6.3363,
      "mean_token_accuracy": 0.1553775981068611,
      "num_tokens": 13833973.0,
      "step": 5455
    },
    {
      "entropy": 6.375447082519531,
      "epoch": 0.6301211771494518,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004969973538467605,
      "loss": 6.2622,
      "mean_token_accuracy": 0.1685728073120117,
      "num_tokens": 13845954.0,
      "step": 5460
    },
    {
      "entropy": 6.517121410369873,
      "epoch": 0.6306982111944605,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004969906312306419,
      "loss": 6.3,
      "mean_token_accuracy": 0.15561674758791924,
      "num_tokens": 13858261.0,
      "step": 5465
    },
    {
      "entropy": 6.478764772415161,
      "epoch": 0.6312752452394691,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004969839011479786,
      "loss": 6.2791,
      "mean_token_accuracy": 0.16333419382572173,
      "num_tokens": 13870409.0,
      "step": 5470
    },
    {
      "entropy": 6.430654668807984,
      "epoch": 0.6318522792844777,
      "grad_norm": 0.78125,
      "learning_rate": 0.000496977163598997,
      "loss": 6.3614,
      "mean_token_accuracy": 0.15769426077604293,
      "num_tokens": 13883980.0,
      "step": 5475
    },
    {
      "entropy": 6.476246643066406,
      "epoch": 0.6324293133294865,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004969704185839239,
      "loss": 6.3736,
      "mean_token_accuracy": 0.16204104721546173,
      "num_tokens": 13897901.0,
      "step": 5480
    },
    {
      "entropy": 6.482294368743896,
      "epoch": 0.6330063473744951,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004969636661029859,
      "loss": 6.3509,
      "mean_token_accuracy": 0.1583631828427315,
      "num_tokens": 13909655.0,
      "step": 5485
    },
    {
      "entropy": 6.494528293609619,
      "epoch": 0.6335833814195038,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004969569061564103,
      "loss": 6.3003,
      "mean_token_accuracy": 0.16088411957025528,
      "num_tokens": 13923121.0,
      "step": 5490
    },
    {
      "entropy": 6.520718622207641,
      "epoch": 0.6341604154645124,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004969501387444245,
      "loss": 6.3925,
      "mean_token_accuracy": 0.1511763960123062,
      "num_tokens": 13936100.0,
      "step": 5495
    },
    {
      "entropy": 6.4032745361328125,
      "epoch": 0.634737449509521,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004969433638672559,
      "loss": 6.2821,
      "mean_token_accuracy": 0.15816063433885574,
      "num_tokens": 13949799.0,
      "step": 5500
    },
    {
      "entropy": 6.450932168960572,
      "epoch": 0.6353144835545297,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004969365815251325,
      "loss": 6.267,
      "mean_token_accuracy": 0.16805639117956161,
      "num_tokens": 13961805.0,
      "step": 5505
    },
    {
      "entropy": 6.401394653320312,
      "epoch": 0.6358915175995383,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004969297917182825,
      "loss": 6.3204,
      "mean_token_accuracy": 0.1655355915427208,
      "num_tokens": 13974264.0,
      "step": 5510
    },
    {
      "entropy": 6.427269029617309,
      "epoch": 0.6364685516445471,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004969229944469342,
      "loss": 6.2694,
      "mean_token_accuracy": 0.16473935395479203,
      "num_tokens": 13986582.0,
      "step": 5515
    },
    {
      "entropy": 6.460540533065796,
      "epoch": 0.6370455856895557,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004969161897113162,
      "loss": 6.2857,
      "mean_token_accuracy": 0.157159423828125,
      "num_tokens": 13998675.0,
      "step": 5520
    },
    {
      "entropy": 6.466981744766235,
      "epoch": 0.6376226197345644,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004969093775116574,
      "loss": 6.3225,
      "mean_token_accuracy": 0.1617126539349556,
      "num_tokens": 14010744.0,
      "step": 5525
    },
    {
      "entropy": 6.524676513671875,
      "epoch": 0.638199653779573,
      "grad_norm": 0.75,
      "learning_rate": 0.0004969025578481869,
      "loss": 6.3096,
      "mean_token_accuracy": 0.1576560452580452,
      "num_tokens": 14023721.0,
      "step": 5530
    },
    {
      "entropy": 6.431024599075317,
      "epoch": 0.6387766878245816,
      "grad_norm": 0.79296875,
      "learning_rate": 0.000496895730721134,
      "loss": 6.2184,
      "mean_token_accuracy": 0.16521818935871124,
      "num_tokens": 14035998.0,
      "step": 5535
    },
    {
      "entropy": 6.338489437103272,
      "epoch": 0.6393537218695903,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004968888961307286,
      "loss": 6.2027,
      "mean_token_accuracy": 0.16932371407747268,
      "num_tokens": 14048407.0,
      "step": 5540
    },
    {
      "entropy": 6.3543487071990965,
      "epoch": 0.6399307559145989,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004968820540772003,
      "loss": 6.1883,
      "mean_token_accuracy": 0.1674530401825905,
      "num_tokens": 14060898.0,
      "step": 5545
    },
    {
      "entropy": 6.428665018081665,
      "epoch": 0.6405077899596077,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004968752045607794,
      "loss": 6.2746,
      "mean_token_accuracy": 0.16548453867435456,
      "num_tokens": 14073474.0,
      "step": 5550
    },
    {
      "entropy": 6.425108814239502,
      "epoch": 0.6410848240046163,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004968683475816961,
      "loss": 6.353,
      "mean_token_accuracy": 0.15872218310832978,
      "num_tokens": 14085229.0,
      "step": 5555
    },
    {
      "entropy": 6.432684373855591,
      "epoch": 0.641661858049625,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004968614831401811,
      "loss": 6.2776,
      "mean_token_accuracy": 0.16353354454040528,
      "num_tokens": 14097943.0,
      "step": 5560
    },
    {
      "entropy": 6.543388414382934,
      "epoch": 0.6422388920946336,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004968546112364654,
      "loss": 6.4176,
      "mean_token_accuracy": 0.1529288560152054,
      "num_tokens": 14112397.0,
      "step": 5565
    },
    {
      "entropy": 6.441380167007447,
      "epoch": 0.6428159261396422,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00049684773187078,
      "loss": 6.2476,
      "mean_token_accuracy": 0.16730485260486602,
      "num_tokens": 14127096.0,
      "step": 5570
    },
    {
      "entropy": 6.497290849685669,
      "epoch": 0.6433929601846509,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004968408450433565,
      "loss": 6.3284,
      "mean_token_accuracy": 0.15766475945711136,
      "num_tokens": 14139325.0,
      "step": 5575
    },
    {
      "entropy": 6.50930609703064,
      "epoch": 0.6439699942296595,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004968339507544263,
      "loss": 6.3056,
      "mean_token_accuracy": 0.15905881971120833,
      "num_tokens": 14151334.0,
      "step": 5580
    },
    {
      "entropy": 6.486502265930175,
      "epoch": 0.6445470282746683,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004968270490042212,
      "loss": 6.2151,
      "mean_token_accuracy": 0.17320423871278762,
      "num_tokens": 14164195.0,
      "step": 5585
    },
    {
      "entropy": 6.368127059936524,
      "epoch": 0.6451240623196769,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004968201397929737,
      "loss": 6.2717,
      "mean_token_accuracy": 0.16237038969993592,
      "num_tokens": 14175937.0,
      "step": 5590
    },
    {
      "entropy": 6.4224189758300785,
      "epoch": 0.6457010963646855,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004968132231209158,
      "loss": 6.2844,
      "mean_token_accuracy": 0.16292650997638702,
      "num_tokens": 14189288.0,
      "step": 5595
    },
    {
      "entropy": 6.4685217380523685,
      "epoch": 0.6462781304096942,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004968062989882803,
      "loss": 6.2794,
      "mean_token_accuracy": 0.16588214188814163,
      "num_tokens": 14202428.0,
      "step": 5600
    },
    {
      "entropy": 6.327192735671997,
      "epoch": 0.6468551644547028,
      "grad_norm": 0.73828125,
      "learning_rate": 0.0004967993673953003,
      "loss": 6.273,
      "mean_token_accuracy": 0.16544562429189683,
      "num_tokens": 14216871.0,
      "step": 5605
    },
    {
      "entropy": 6.401887845993042,
      "epoch": 0.6474321984997115,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004967924283422087,
      "loss": 6.2498,
      "mean_token_accuracy": 0.16043491661548615,
      "num_tokens": 14228503.0,
      "step": 5610
    },
    {
      "entropy": 6.42364354133606,
      "epoch": 0.6480092325447201,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000496785481829239,
      "loss": 6.1793,
      "mean_token_accuracy": 0.16777475029230118,
      "num_tokens": 14240239.0,
      "step": 5615
    },
    {
      "entropy": 6.395428085327149,
      "epoch": 0.6485862665897288,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004967785278566245,
      "loss": 6.2541,
      "mean_token_accuracy": 0.16892644315958022,
      "num_tokens": 14253180.0,
      "step": 5620
    },
    {
      "entropy": 6.3999152183532715,
      "epoch": 0.6491633006347375,
      "grad_norm": 0.7109375,
      "learning_rate": 0.0004967715664245997,
      "loss": 6.3116,
      "mean_token_accuracy": 0.15934029072523118,
      "num_tokens": 14267275.0,
      "step": 5625
    },
    {
      "entropy": 6.455025959014892,
      "epoch": 0.6497403346797461,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004967645975333983,
      "loss": 6.3232,
      "mean_token_accuracy": 0.1586598202586174,
      "num_tokens": 14279650.0,
      "step": 5630
    },
    {
      "entropy": 6.418586301803589,
      "epoch": 0.6503173687247548,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004967576211832548,
      "loss": 6.2187,
      "mean_token_accuracy": 0.16348557323217391,
      "num_tokens": 14291647.0,
      "step": 5635
    },
    {
      "entropy": 6.325366878509522,
      "epoch": 0.6508944027697634,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004967506373744039,
      "loss": 6.2198,
      "mean_token_accuracy": 0.16452054381370546,
      "num_tokens": 14304335.0,
      "step": 5640
    },
    {
      "entropy": 6.430365800857544,
      "epoch": 0.651471436814772,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004967436461070805,
      "loss": 6.2481,
      "mean_token_accuracy": 0.16254803836345671,
      "num_tokens": 14317226.0,
      "step": 5645
    },
    {
      "entropy": 6.434140729904175,
      "epoch": 0.6520484708597807,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004967366473815196,
      "loss": 6.2797,
      "mean_token_accuracy": 0.1666984811425209,
      "num_tokens": 14329000.0,
      "step": 5650
    },
    {
      "entropy": 6.386842870712281,
      "epoch": 0.6526255049047894,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004967296411979568,
      "loss": 6.2577,
      "mean_token_accuracy": 0.16278322041034698,
      "num_tokens": 14341714.0,
      "step": 5655
    },
    {
      "entropy": 6.412582921981811,
      "epoch": 0.6532025389497981,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004967226275566275,
      "loss": 6.2462,
      "mean_token_accuracy": 0.15906601399183273,
      "num_tokens": 14353889.0,
      "step": 5660
    },
    {
      "entropy": 6.535210132598877,
      "epoch": 0.6537795729948067,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000496715606457768,
      "loss": 6.3295,
      "mean_token_accuracy": 0.1594246506690979,
      "num_tokens": 14367093.0,
      "step": 5665
    },
    {
      "entropy": 6.274228048324585,
      "epoch": 0.6543566070398154,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004967085779016142,
      "loss": 6.0987,
      "mean_token_accuracy": 0.17560223937034608,
      "num_tokens": 14380064.0,
      "step": 5670
    },
    {
      "entropy": 6.444216394424439,
      "epoch": 0.654933641084824,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004967015418884022,
      "loss": 6.2893,
      "mean_token_accuracy": 0.16281547397375107,
      "num_tokens": 14392026.0,
      "step": 5675
    },
    {
      "entropy": 6.419415473937988,
      "epoch": 0.6555106751298326,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004966944984183692,
      "loss": 6.2463,
      "mean_token_accuracy": 0.15920784771442414,
      "num_tokens": 14404011.0,
      "step": 5680
    },
    {
      "entropy": 6.416391611099243,
      "epoch": 0.6560877091748413,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004966874474917518,
      "loss": 6.2416,
      "mean_token_accuracy": 0.16538350880146027,
      "num_tokens": 14416356.0,
      "step": 5685
    },
    {
      "entropy": 6.384862327575684,
      "epoch": 0.65666474321985,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004966803891087873,
      "loss": 6.2679,
      "mean_token_accuracy": 0.15895518362522126,
      "num_tokens": 14428873.0,
      "step": 5690
    },
    {
      "entropy": 6.373016309738159,
      "epoch": 0.6572417772648587,
      "grad_norm": 0.76953125,
      "learning_rate": 0.000496673323269713,
      "loss": 6.2474,
      "mean_token_accuracy": 0.1652483820915222,
      "num_tokens": 14441703.0,
      "step": 5695
    },
    {
      "entropy": 6.473770570755005,
      "epoch": 0.6578188113098673,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004966662499747666,
      "loss": 6.2524,
      "mean_token_accuracy": 0.16153307110071183,
      "num_tokens": 14453701.0,
      "step": 5700
    },
    {
      "entropy": 6.456718921661377,
      "epoch": 0.6583958453548759,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004966591692241859,
      "loss": 6.2848,
      "mean_token_accuracy": 0.1631843477487564,
      "num_tokens": 14466108.0,
      "step": 5705
    },
    {
      "entropy": 6.405096006393433,
      "epoch": 0.6589728793998846,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004966520810182092,
      "loss": 6.3424,
      "mean_token_accuracy": 0.16256508827209473,
      "num_tokens": 14479130.0,
      "step": 5710
    },
    {
      "entropy": 6.451769876480102,
      "epoch": 0.6595499134448932,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004966449853570749,
      "loss": 6.2904,
      "mean_token_accuracy": 0.16311692148447038,
      "num_tokens": 14492797.0,
      "step": 5715
    },
    {
      "entropy": 6.406583213806153,
      "epoch": 0.6601269474899019,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004966378822410215,
      "loss": 6.2139,
      "mean_token_accuracy": 0.1620306298136711,
      "num_tokens": 14505279.0,
      "step": 5720
    },
    {
      "entropy": 6.40832781791687,
      "epoch": 0.6607039815349106,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004966307716702881,
      "loss": 6.3031,
      "mean_token_accuracy": 0.16416211426258087,
      "num_tokens": 14517867.0,
      "step": 5725
    },
    {
      "entropy": 6.400218391418457,
      "epoch": 0.6612810155799193,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004966236536451138,
      "loss": 6.2295,
      "mean_token_accuracy": 0.16840277910232543,
      "num_tokens": 14529804.0,
      "step": 5730
    },
    {
      "entropy": 6.332806348800659,
      "epoch": 0.6618580496249279,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000496616528165738,
      "loss": 6.1157,
      "mean_token_accuracy": 0.17764069586992265,
      "num_tokens": 14541484.0,
      "step": 5735
    },
    {
      "entropy": 6.35449628829956,
      "epoch": 0.6624350836699365,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004966093952324003,
      "loss": 6.1861,
      "mean_token_accuracy": 0.16744254529476166,
      "num_tokens": 14554796.0,
      "step": 5740
    },
    {
      "entropy": 6.292174339294434,
      "epoch": 0.6630121177149452,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004966022548453406,
      "loss": 6.1281,
      "mean_token_accuracy": 0.18883802443742753,
      "num_tokens": 14567090.0,
      "step": 5745
    },
    {
      "entropy": 6.370087051391602,
      "epoch": 0.6635891517599538,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004965951070047993,
      "loss": 6.2575,
      "mean_token_accuracy": 0.16039278656244277,
      "num_tokens": 14580334.0,
      "step": 5750
    },
    {
      "entropy": 6.384637451171875,
      "epoch": 0.6641661858049625,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004965879517110166,
      "loss": 6.1818,
      "mean_token_accuracy": 0.17177506238222123,
      "num_tokens": 14593670.0,
      "step": 5755
    },
    {
      "entropy": 6.3392963886260985,
      "epoch": 0.6647432198499712,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004965807889642333,
      "loss": 6.2103,
      "mean_token_accuracy": 0.16691275388002397,
      "num_tokens": 14606612.0,
      "step": 5760
    },
    {
      "entropy": 6.490033864974976,
      "epoch": 0.6653202538949798,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00049657361876469,
      "loss": 6.2239,
      "mean_token_accuracy": 0.16626789271831513,
      "num_tokens": 14618323.0,
      "step": 5765
    },
    {
      "entropy": 6.27760009765625,
      "epoch": 0.6658972879399885,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004965664411126282,
      "loss": 6.2311,
      "mean_token_accuracy": 0.16586460769176484,
      "num_tokens": 14631580.0,
      "step": 5770
    },
    {
      "entropy": 6.381122303009033,
      "epoch": 0.6664743219849971,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004965592560082894,
      "loss": 6.1559,
      "mean_token_accuracy": 0.16871390789747237,
      "num_tokens": 14643429.0,
      "step": 5775
    },
    {
      "entropy": 6.4877759456634525,
      "epoch": 0.6670513560300058,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004965520634519149,
      "loss": 6.3704,
      "mean_token_accuracy": 0.15916894674301146,
      "num_tokens": 14656890.0,
      "step": 5780
    },
    {
      "entropy": 6.332120084762574,
      "epoch": 0.6676283900750144,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004965448634437468,
      "loss": 6.2516,
      "mean_token_accuracy": 0.1616871252655983,
      "num_tokens": 14670721.0,
      "step": 5785
    },
    {
      "entropy": 6.45994815826416,
      "epoch": 0.668205424120023,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004965376559840272,
      "loss": 6.2788,
      "mean_token_accuracy": 0.15587088614702224,
      "num_tokens": 14682808.0,
      "step": 5790
    },
    {
      "entropy": 6.4104235649108885,
      "epoch": 0.6687824581650318,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004965304410729986,
      "loss": 6.2592,
      "mean_token_accuracy": 0.17194210290908812,
      "num_tokens": 14694580.0,
      "step": 5795
    },
    {
      "entropy": 6.341176223754883,
      "epoch": 0.6693594922100404,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004965232187109037,
      "loss": 6.2736,
      "mean_token_accuracy": 0.1583656281232834,
      "num_tokens": 14707832.0,
      "step": 5800
    },
    {
      "entropy": 6.460653781890869,
      "epoch": 0.6699365262550491,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004965159888979854,
      "loss": 6.2805,
      "mean_token_accuracy": 0.16277113258838655,
      "num_tokens": 14719293.0,
      "step": 5805
    },
    {
      "entropy": 6.33139591217041,
      "epoch": 0.6705135603000577,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004965087516344869,
      "loss": 6.1755,
      "mean_token_accuracy": 0.16717009246349335,
      "num_tokens": 14731421.0,
      "step": 5810
    },
    {
      "entropy": 6.35786280632019,
      "epoch": 0.6710905943450663,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004965015069206515,
      "loss": 6.2799,
      "mean_token_accuracy": 0.16111143082380294,
      "num_tokens": 14744160.0,
      "step": 5815
    },
    {
      "entropy": 6.518506193161011,
      "epoch": 0.671667628390075,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000496494254756723,
      "loss": 6.2824,
      "mean_token_accuracy": 0.15849184840917588,
      "num_tokens": 14756820.0,
      "step": 5820
    },
    {
      "entropy": 6.372296142578125,
      "epoch": 0.6722446624350836,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004964869951429451,
      "loss": 6.2227,
      "mean_token_accuracy": 0.1700335294008255,
      "num_tokens": 14770000.0,
      "step": 5825
    },
    {
      "entropy": 6.415471744537354,
      "epoch": 0.6728216964800924,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004964797280795622,
      "loss": 6.3075,
      "mean_token_accuracy": 0.1579108163714409,
      "num_tokens": 14782777.0,
      "step": 5830
    },
    {
      "entropy": 6.452236700057983,
      "epoch": 0.673398730525101,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004964724535668188,
      "loss": 6.3368,
      "mean_token_accuracy": 0.16094402670860292,
      "num_tokens": 14796561.0,
      "step": 5835
    },
    {
      "entropy": 6.412081909179688,
      "epoch": 0.6739757645701097,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004964651716049593,
      "loss": 6.2962,
      "mean_token_accuracy": 0.16613052189350128,
      "num_tokens": 14809975.0,
      "step": 5840
    },
    {
      "entropy": 6.4701982021331785,
      "epoch": 0.6745527986151183,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004964578821942288,
      "loss": 6.318,
      "mean_token_accuracy": 0.15860529243946075,
      "num_tokens": 14823578.0,
      "step": 5845
    },
    {
      "entropy": 6.476156282424927,
      "epoch": 0.6751298326601269,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004964505853348724,
      "loss": 6.2851,
      "mean_token_accuracy": 0.16648429483175278,
      "num_tokens": 14836520.0,
      "step": 5850
    },
    {
      "entropy": 6.431745862960815,
      "epoch": 0.6757068667051356,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004964432810271356,
      "loss": 6.321,
      "mean_token_accuracy": 0.15974137783050538,
      "num_tokens": 14847806.0,
      "step": 5855
    },
    {
      "entropy": 6.3844091415405275,
      "epoch": 0.6762839007501442,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004964359692712641,
      "loss": 6.2074,
      "mean_token_accuracy": 0.16706834882497787,
      "num_tokens": 14859751.0,
      "step": 5860
    },
    {
      "entropy": 6.481093883514404,
      "epoch": 0.676860934795153,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004964286500675037,
      "loss": 6.2351,
      "mean_token_accuracy": 0.165215802192688,
      "num_tokens": 14872215.0,
      "step": 5865
    },
    {
      "entropy": 6.417674732208252,
      "epoch": 0.6774379688401616,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004964213234161007,
      "loss": 6.2849,
      "mean_token_accuracy": 0.16408911049365998,
      "num_tokens": 14884895.0,
      "step": 5870
    },
    {
      "entropy": 6.39830961227417,
      "epoch": 0.6780150028851702,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004964139893173014,
      "loss": 6.2486,
      "mean_token_accuracy": 0.16585659235715866,
      "num_tokens": 14897131.0,
      "step": 5875
    },
    {
      "entropy": 6.350745296478271,
      "epoch": 0.6785920369301789,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004964066477713525,
      "loss": 6.1867,
      "mean_token_accuracy": 0.1712081626057625,
      "num_tokens": 14908835.0,
      "step": 5880
    },
    {
      "entropy": 6.394719171524048,
      "epoch": 0.6791690709751875,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004963992987785011,
      "loss": 6.2861,
      "mean_token_accuracy": 0.1636969566345215,
      "num_tokens": 14920808.0,
      "step": 5885
    },
    {
      "entropy": 6.453951072692871,
      "epoch": 0.6797461050201962,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004963919423389942,
      "loss": 6.1362,
      "mean_token_accuracy": 0.17062204480171203,
      "num_tokens": 14932558.0,
      "step": 5890
    },
    {
      "entropy": 6.310215997695923,
      "epoch": 0.6803231390652048,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004963845784530794,
      "loss": 6.2496,
      "mean_token_accuracy": 0.16623902767896653,
      "num_tokens": 14946623.0,
      "step": 5895
    },
    {
      "entropy": 6.384698295593262,
      "epoch": 0.6809001731102136,
      "grad_norm": 0.8046875,
      "learning_rate": 0.000496377207121004,
      "loss": 6.2711,
      "mean_token_accuracy": 0.1658199280500412,
      "num_tokens": 14959168.0,
      "step": 5900
    },
    {
      "entropy": 6.477806615829468,
      "epoch": 0.6814772071552222,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004963698283430164,
      "loss": 6.3015,
      "mean_token_accuracy": 0.15701866149902344,
      "num_tokens": 14972710.0,
      "step": 5905
    },
    {
      "entropy": 6.399756193161011,
      "epoch": 0.6820542412002308,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004963624421193646,
      "loss": 6.1501,
      "mean_token_accuracy": 0.16826880425214769,
      "num_tokens": 14984707.0,
      "step": 5910
    },
    {
      "entropy": 6.298456907272339,
      "epoch": 0.6826312752452395,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000496355048450297,
      "loss": 6.1769,
      "mean_token_accuracy": 0.17006256580352783,
      "num_tokens": 14996623.0,
      "step": 5915
    },
    {
      "entropy": 6.430842161178589,
      "epoch": 0.6832083092902481,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004963476473360623,
      "loss": 6.2945,
      "mean_token_accuracy": 0.1616099148988724,
      "num_tokens": 15009579.0,
      "step": 5920
    },
    {
      "entropy": 6.403625202178955,
      "epoch": 0.6837853433352568,
      "grad_norm": 0.875,
      "learning_rate": 0.0004963402387769094,
      "loss": 6.2285,
      "mean_token_accuracy": 0.16827364563941954,
      "num_tokens": 15022239.0,
      "step": 5925
    },
    {
      "entropy": 6.363240528106689,
      "epoch": 0.6843623773802654,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004963328227730876,
      "loss": 6.2556,
      "mean_token_accuracy": 0.16110678166151046,
      "num_tokens": 15034724.0,
      "step": 5930
    },
    {
      "entropy": 6.31289849281311,
      "epoch": 0.684939411425274,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004963253993248461,
      "loss": 6.1712,
      "mean_token_accuracy": 0.16620973199605943,
      "num_tokens": 15047534.0,
      "step": 5935
    },
    {
      "entropy": 6.4068629264831545,
      "epoch": 0.6855164454702828,
      "grad_norm": 0.875,
      "learning_rate": 0.0004963179684324349,
      "loss": 6.2699,
      "mean_token_accuracy": 0.16617678850889206,
      "num_tokens": 15060468.0,
      "step": 5940
    },
    {
      "entropy": 6.435975074768066,
      "epoch": 0.6860934795152914,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004963105300961036,
      "loss": 6.2452,
      "mean_token_accuracy": 0.1608870580792427,
      "num_tokens": 15073424.0,
      "step": 5945
    },
    {
      "entropy": 6.380932950973511,
      "epoch": 0.6866705135603001,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004963030843161026,
      "loss": 6.287,
      "mean_token_accuracy": 0.15832511186599732,
      "num_tokens": 15086582.0,
      "step": 5950
    },
    {
      "entropy": 6.479296588897705,
      "epoch": 0.6872475476053087,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004962956310926823,
      "loss": 6.2874,
      "mean_token_accuracy": 0.1592211276292801,
      "num_tokens": 15099288.0,
      "step": 5955
    },
    {
      "entropy": 6.414581060409546,
      "epoch": 0.6878245816503173,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004962881704260934,
      "loss": 6.2051,
      "mean_token_accuracy": 0.16755102276802064,
      "num_tokens": 15111800.0,
      "step": 5960
    },
    {
      "entropy": 6.396193742752075,
      "epoch": 0.688401615695326,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004962807023165868,
      "loss": 6.2563,
      "mean_token_accuracy": 0.16351549625396727,
      "num_tokens": 15123384.0,
      "step": 5965
    },
    {
      "entropy": 6.36266450881958,
      "epoch": 0.6889786497403346,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004962732267644138,
      "loss": 6.1621,
      "mean_token_accuracy": 0.1664055049419403,
      "num_tokens": 15134857.0,
      "step": 5970
    },
    {
      "entropy": 6.395646381378174,
      "epoch": 0.6895556837853434,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004962657437698254,
      "loss": 6.3036,
      "mean_token_accuracy": 0.16814867109060289,
      "num_tokens": 15149124.0,
      "step": 5975
    },
    {
      "entropy": 6.244737005233764,
      "epoch": 0.690132717830352,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004962582533330739,
      "loss": 6.0647,
      "mean_token_accuracy": 0.18276433199644088,
      "num_tokens": 15163315.0,
      "step": 5980
    },
    {
      "entropy": 6.433995866775513,
      "epoch": 0.6907097518753607,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004962507554544109,
      "loss": 6.2554,
      "mean_token_accuracy": 0.16413442492485047,
      "num_tokens": 15175371.0,
      "step": 5985
    },
    {
      "entropy": 6.423586463928222,
      "epoch": 0.6912867859203693,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004962432501340886,
      "loss": 6.2297,
      "mean_token_accuracy": 0.16294970810413362,
      "num_tokens": 15187633.0,
      "step": 5990
    },
    {
      "entropy": 6.39416732788086,
      "epoch": 0.6918638199653779,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004962357373723596,
      "loss": 6.2701,
      "mean_token_accuracy": 0.1605140045285225,
      "num_tokens": 15199652.0,
      "step": 5995
    },
    {
      "entropy": 6.471811056137085,
      "epoch": 0.6924408540103866,
      "grad_norm": 0.75,
      "learning_rate": 0.0004962282171694763,
      "loss": 6.2885,
      "mean_token_accuracy": 0.16474244892597198,
      "num_tokens": 15213309.0,
      "step": 6000
    },
    {
      "epoch": 0.6924408540103866,
      "eval_entropy": 6.305234524279722,
      "eval_loss": 6.288375377655029,
      "eval_mean_token_accuracy": 0.166228380397424,
      "eval_num_tokens": 15213309.0,
      "eval_runtime": 17.8169,
      "eval_samples_per_second": 1579.178,
      "eval_steps_per_second": 197.397,
      "step": 6000
    },
    {
      "entropy": 6.407313680648803,
      "epoch": 0.6930178880553952,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004962206895256919,
      "loss": 6.2349,
      "mean_token_accuracy": 0.16205479800701142,
      "num_tokens": 15226466.0,
      "step": 6005
    },
    {
      "entropy": 6.275579786300659,
      "epoch": 0.693594922100404,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004962131544412595,
      "loss": 6.1267,
      "mean_token_accuracy": 0.17538843154907227,
      "num_tokens": 15239306.0,
      "step": 6010
    },
    {
      "entropy": 6.371297407150268,
      "epoch": 0.6941719561454126,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004962056119164325,
      "loss": 6.2535,
      "mean_token_accuracy": 0.16581086218357086,
      "num_tokens": 15252365.0,
      "step": 6015
    },
    {
      "entropy": 6.46108570098877,
      "epoch": 0.6947489901904212,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004961980619514646,
      "loss": 6.2761,
      "mean_token_accuracy": 0.16486027836799622,
      "num_tokens": 15265770.0,
      "step": 6020
    },
    {
      "entropy": 6.398010969161987,
      "epoch": 0.6953260242354299,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004961905045466098,
      "loss": 6.2754,
      "mean_token_accuracy": 0.16328403055667878,
      "num_tokens": 15278766.0,
      "step": 6025
    },
    {
      "entropy": 6.503208446502685,
      "epoch": 0.6959030582804385,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004961829397021222,
      "loss": 6.2858,
      "mean_token_accuracy": 0.15995303988456727,
      "num_tokens": 15292180.0,
      "step": 6030
    },
    {
      "entropy": 6.451317501068115,
      "epoch": 0.6964800923254472,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004961753674182564,
      "loss": 6.3007,
      "mean_token_accuracy": 0.15992565751075744,
      "num_tokens": 15305313.0,
      "step": 6035
    },
    {
      "entropy": 6.455318593978882,
      "epoch": 0.6970571263704558,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004961677876952669,
      "loss": 6.3497,
      "mean_token_accuracy": 0.15649957954883575,
      "num_tokens": 15318734.0,
      "step": 6040
    },
    {
      "entropy": 6.367619514465332,
      "epoch": 0.6976341604154646,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004961602005334087,
      "loss": 6.2162,
      "mean_token_accuracy": 0.1630083441734314,
      "num_tokens": 15330915.0,
      "step": 6045
    },
    {
      "entropy": 6.420135116577148,
      "epoch": 0.6982111944604732,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000496152605932937,
      "loss": 6.3083,
      "mean_token_accuracy": 0.15993753224611282,
      "num_tokens": 15343634.0,
      "step": 6050
    },
    {
      "entropy": 6.443638706207276,
      "epoch": 0.6987882285054818,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004961450038941074,
      "loss": 6.2466,
      "mean_token_accuracy": 0.16260004937648773,
      "num_tokens": 15356049.0,
      "step": 6055
    },
    {
      "entropy": 6.369221115112305,
      "epoch": 0.6993652625504905,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004961373944171754,
      "loss": 6.1004,
      "mean_token_accuracy": 0.1690909057855606,
      "num_tokens": 15367804.0,
      "step": 6060
    },
    {
      "entropy": 6.348468399047851,
      "epoch": 0.6999422965954991,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004961297775023968,
      "loss": 6.2341,
      "mean_token_accuracy": 0.1663442924618721,
      "num_tokens": 15379921.0,
      "step": 6065
    },
    {
      "entropy": 6.332316446304321,
      "epoch": 0.7005193306405078,
      "grad_norm": 0.7578125,
      "learning_rate": 0.000496122153150028,
      "loss": 6.2063,
      "mean_token_accuracy": 0.1757096141576767,
      "num_tokens": 15393043.0,
      "step": 6070
    },
    {
      "entropy": 6.452652359008789,
      "epoch": 0.7010963646855164,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004961145213603255,
      "loss": 6.2926,
      "mean_token_accuracy": 0.1563486099243164,
      "num_tokens": 15407577.0,
      "step": 6075
    },
    {
      "entropy": 6.411776113510132,
      "epoch": 0.7016733987305251,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000496106882133546,
      "loss": 6.2942,
      "mean_token_accuracy": 0.16173667162656785,
      "num_tokens": 15420157.0,
      "step": 6080
    },
    {
      "entropy": 6.3121175289154055,
      "epoch": 0.7022504327755338,
      "grad_norm": 0.875,
      "learning_rate": 0.0004960992354699463,
      "loss": 6.2521,
      "mean_token_accuracy": 0.1679789364337921,
      "num_tokens": 15432882.0,
      "step": 6085
    },
    {
      "entropy": 6.40244779586792,
      "epoch": 0.7028274668205424,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004960915813697836,
      "loss": 6.1713,
      "mean_token_accuracy": 0.16617501974105836,
      "num_tokens": 15444541.0,
      "step": 6090
    },
    {
      "entropy": 6.37771053314209,
      "epoch": 0.7034045008655511,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004960839198333154,
      "loss": 6.3018,
      "mean_token_accuracy": 0.16228880733251572,
      "num_tokens": 15457104.0,
      "step": 6095
    },
    {
      "entropy": 6.42156720161438,
      "epoch": 0.7039815349105597,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004960762508607994,
      "loss": 6.2456,
      "mean_token_accuracy": 0.16384343653917313,
      "num_tokens": 15469724.0,
      "step": 6100
    },
    {
      "entropy": 6.416327810287475,
      "epoch": 0.7045585689555683,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004960685744524934,
      "loss": 6.2881,
      "mean_token_accuracy": 0.16480329632759094,
      "num_tokens": 15481261.0,
      "step": 6105
    },
    {
      "entropy": 6.350079822540283,
      "epoch": 0.705135603000577,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004960608906086558,
      "loss": 6.2208,
      "mean_token_accuracy": 0.16420309692621232,
      "num_tokens": 15494996.0,
      "step": 6110
    },
    {
      "entropy": 6.404261541366577,
      "epoch": 0.7057126370455857,
      "grad_norm": 0.7890625,
      "learning_rate": 0.000496053199329545,
      "loss": 6.1642,
      "mean_token_accuracy": 0.16822902262210845,
      "num_tokens": 15506772.0,
      "step": 6115
    },
    {
      "entropy": 6.3732819080352785,
      "epoch": 0.7062896710905944,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004960455006154196,
      "loss": 6.2876,
      "mean_token_accuracy": 0.16168780773878097,
      "num_tokens": 15520336.0,
      "step": 6120
    },
    {
      "entropy": 6.474810409545898,
      "epoch": 0.706866705135603,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004960377944665386,
      "loss": 6.3679,
      "mean_token_accuracy": 0.15211256816983224,
      "num_tokens": 15535139.0,
      "step": 6125
    },
    {
      "entropy": 6.396150255203247,
      "epoch": 0.7074437391806117,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004960300808831611,
      "loss": 6.2355,
      "mean_token_accuracy": 0.1614559680223465,
      "num_tokens": 15547523.0,
      "step": 6130
    },
    {
      "entropy": 6.351924133300781,
      "epoch": 0.7080207732256203,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004960223598655467,
      "loss": 6.2385,
      "mean_token_accuracy": 0.16700737476348876,
      "num_tokens": 15560816.0,
      "step": 6135
    },
    {
      "entropy": 6.430610704421997,
      "epoch": 0.7085978072706289,
      "grad_norm": 0.7890625,
      "learning_rate": 0.000496014631413955,
      "loss": 6.2409,
      "mean_token_accuracy": 0.1643073469400406,
      "num_tokens": 15574364.0,
      "step": 6140
    },
    {
      "entropy": 6.307897710800171,
      "epoch": 0.7091748413156376,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004960068955286459,
      "loss": 6.0504,
      "mean_token_accuracy": 0.18151674717664718,
      "num_tokens": 15586969.0,
      "step": 6145
    },
    {
      "entropy": 6.4239410877227785,
      "epoch": 0.7097518753606463,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004959991522098797,
      "loss": 6.2119,
      "mean_token_accuracy": 0.16022869125008582,
      "num_tokens": 15600459.0,
      "step": 6150
    },
    {
      "entropy": 6.444924354553223,
      "epoch": 0.710328909405655,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004959914014579168,
      "loss": 6.2652,
      "mean_token_accuracy": 0.15775836259126663,
      "num_tokens": 15613484.0,
      "step": 6155
    },
    {
      "entropy": 6.353692579269409,
      "epoch": 0.7109059434506636,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004959836432730178,
      "loss": 6.282,
      "mean_token_accuracy": 0.16222158819437027,
      "num_tokens": 15626036.0,
      "step": 6160
    },
    {
      "entropy": 6.286346817016602,
      "epoch": 0.7114829774956722,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004959758776554438,
      "loss": 6.2124,
      "mean_token_accuracy": 0.17564715892076493,
      "num_tokens": 15640229.0,
      "step": 6165
    },
    {
      "entropy": 6.401192998886108,
      "epoch": 0.7120600115406809,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000495968104605456,
      "loss": 6.1071,
      "mean_token_accuracy": 0.17238144278526307,
      "num_tokens": 15652655.0,
      "step": 6170
    },
    {
      "entropy": 6.407490825653076,
      "epoch": 0.7126370455856895,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004959603241233157,
      "loss": 6.2637,
      "mean_token_accuracy": 0.16557672023773193,
      "num_tokens": 15665535.0,
      "step": 6175
    },
    {
      "entropy": 6.40711145401001,
      "epoch": 0.7132140796306982,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004959525362092846,
      "loss": 6.2342,
      "mean_token_accuracy": 0.16254877895116807,
      "num_tokens": 15678631.0,
      "step": 6180
    },
    {
      "entropy": 6.4322343349456785,
      "epoch": 0.7137911136757069,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004959447408636247,
      "loss": 6.3179,
      "mean_token_accuracy": 0.16229903995990752,
      "num_tokens": 15691662.0,
      "step": 6185
    },
    {
      "entropy": 6.395531558990479,
      "epoch": 0.7143681477207156,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004959369380865983,
      "loss": 6.2408,
      "mean_token_accuracy": 0.1694466605782509,
      "num_tokens": 15705017.0,
      "step": 6190
    },
    {
      "entropy": 6.42110686302185,
      "epoch": 0.7149451817657242,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004959291278784676,
      "loss": 6.2544,
      "mean_token_accuracy": 0.16188978999853135,
      "num_tokens": 15717537.0,
      "step": 6195
    },
    {
      "entropy": 6.332228946685791,
      "epoch": 0.7155222158107328,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004959213102394954,
      "loss": 6.0963,
      "mean_token_accuracy": 0.16999780088663102,
      "num_tokens": 15730075.0,
      "step": 6200
    },
    {
      "entropy": 6.376552534103394,
      "epoch": 0.7160992498557415,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004959134851699449,
      "loss": 6.1954,
      "mean_token_accuracy": 0.16710759848356246,
      "num_tokens": 15743187.0,
      "step": 6205
    },
    {
      "entropy": 6.411378002166748,
      "epoch": 0.7166762839007501,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004959056526700788,
      "loss": 6.2817,
      "mean_token_accuracy": 0.1558884635567665,
      "num_tokens": 15755459.0,
      "step": 6210
    },
    {
      "entropy": 6.355316114425659,
      "epoch": 0.7172533179457588,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004958978127401609,
      "loss": 6.1278,
      "mean_token_accuracy": 0.16565006971359253,
      "num_tokens": 15767929.0,
      "step": 6215
    },
    {
      "entropy": 6.33356614112854,
      "epoch": 0.7178303519907675,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004958899653804548,
      "loss": 6.1179,
      "mean_token_accuracy": 0.1736751303076744,
      "num_tokens": 15780079.0,
      "step": 6220
    },
    {
      "entropy": 6.420310592651367,
      "epoch": 0.7184073860357761,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004958821105912246,
      "loss": 6.217,
      "mean_token_accuracy": 0.16546618044376374,
      "num_tokens": 15792417.0,
      "step": 6225
    },
    {
      "entropy": 6.43408350944519,
      "epoch": 0.7189844200807848,
      "grad_norm": 0.7734375,
      "learning_rate": 0.000495874248372734,
      "loss": 6.2172,
      "mean_token_accuracy": 0.1711224213242531,
      "num_tokens": 15804993.0,
      "step": 6230
    },
    {
      "entropy": 6.350424289703369,
      "epoch": 0.7195614541257934,
      "grad_norm": 0.77734375,
      "learning_rate": 0.000495866378725248,
      "loss": 6.2582,
      "mean_token_accuracy": 0.16318586766719817,
      "num_tokens": 15819765.0,
      "step": 6235
    },
    {
      "entropy": 6.361863899230957,
      "epoch": 0.7201384881708021,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000495858501649031,
      "loss": 6.2619,
      "mean_token_accuracy": 0.16482506394386293,
      "num_tokens": 15831992.0,
      "step": 6240
    },
    {
      "entropy": 6.335883855819702,
      "epoch": 0.7207155222158107,
      "grad_norm": 0.81640625,
      "learning_rate": 0.000495850617144348,
      "loss": 6.2084,
      "mean_token_accuracy": 0.17357497811317443,
      "num_tokens": 15845348.0,
      "step": 6245
    },
    {
      "entropy": 6.4281090259552,
      "epoch": 0.7212925562608193,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004958427252114643,
      "loss": 6.2316,
      "mean_token_accuracy": 0.16620133072137833,
      "num_tokens": 15857677.0,
      "step": 6250
    },
    {
      "entropy": 6.310732650756836,
      "epoch": 0.7218695903058281,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000495834825850645,
      "loss": 6.1912,
      "mean_token_accuracy": 0.17365667521953582,
      "num_tokens": 15871626.0,
      "step": 6255
    },
    {
      "entropy": 6.438030433654785,
      "epoch": 0.7224466243508367,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004958269190621562,
      "loss": 6.2999,
      "mean_token_accuracy": 0.16112429201602935,
      "num_tokens": 15884372.0,
      "step": 6260
    },
    {
      "entropy": 6.265491580963134,
      "epoch": 0.7230236583958454,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004958190048462637,
      "loss": 6.1757,
      "mean_token_accuracy": 0.16929175704717636,
      "num_tokens": 15896079.0,
      "step": 6265
    },
    {
      "entropy": 6.373610734939575,
      "epoch": 0.723600692440854,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004958110832032336,
      "loss": 6.2789,
      "mean_token_accuracy": 0.16542789340019226,
      "num_tokens": 15909528.0,
      "step": 6270
    },
    {
      "entropy": 6.391331529617309,
      "epoch": 0.7241777264858626,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004958031541333322,
      "loss": 6.203,
      "mean_token_accuracy": 0.16602921783924102,
      "num_tokens": 15921247.0,
      "step": 6275
    },
    {
      "entropy": 6.411171197891235,
      "epoch": 0.7247547605308713,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004957952176368267,
      "loss": 6.2785,
      "mean_token_accuracy": 0.16443133652210234,
      "num_tokens": 15933093.0,
      "step": 6280
    },
    {
      "entropy": 6.3585645198822025,
      "epoch": 0.7253317945758799,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004957872737139836,
      "loss": 6.2239,
      "mean_token_accuracy": 0.16355139017105103,
      "num_tokens": 15944888.0,
      "step": 6285
    },
    {
      "entropy": 6.40592737197876,
      "epoch": 0.7259088286208887,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004957793223650702,
      "loss": 6.2547,
      "mean_token_accuracy": 0.16579299718141555,
      "num_tokens": 15957793.0,
      "step": 6290
    },
    {
      "entropy": 6.399983835220337,
      "epoch": 0.7264858626658973,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000495771363590354,
      "loss": 6.2577,
      "mean_token_accuracy": 0.15640069246292115,
      "num_tokens": 15970152.0,
      "step": 6295
    },
    {
      "entropy": 6.358456373214722,
      "epoch": 0.727062896710906,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004957633973901027,
      "loss": 6.2443,
      "mean_token_accuracy": 0.16428245455026627,
      "num_tokens": 15983347.0,
      "step": 6300
    },
    {
      "entropy": 6.383837699890137,
      "epoch": 0.7276399307559146,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004957554237645841,
      "loss": 6.212,
      "mean_token_accuracy": 0.1730270951986313,
      "num_tokens": 15996148.0,
      "step": 6305
    },
    {
      "entropy": 6.391459274291992,
      "epoch": 0.7282169648009232,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004957474427140665,
      "loss": 6.2195,
      "mean_token_accuracy": 0.16915312260389329,
      "num_tokens": 16008833.0,
      "step": 6310
    },
    {
      "entropy": 6.3193872451782225,
      "epoch": 0.7287939988459319,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004957394542388182,
      "loss": 6.1426,
      "mean_token_accuracy": 0.172357177734375,
      "num_tokens": 16020674.0,
      "step": 6315
    },
    {
      "entropy": 6.316094303131104,
      "epoch": 0.7293710328909405,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004957314583391082,
      "loss": 6.1191,
      "mean_token_accuracy": 0.1790053740143776,
      "num_tokens": 16033229.0,
      "step": 6320
    },
    {
      "entropy": 6.403400850296021,
      "epoch": 0.7299480669359493,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004957234550152052,
      "loss": 6.1338,
      "mean_token_accuracy": 0.16456628888845443,
      "num_tokens": 16045646.0,
      "step": 6325
    },
    {
      "entropy": 6.404736471176148,
      "epoch": 0.7305251009809579,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004957154442673784,
      "loss": 6.2346,
      "mean_token_accuracy": 0.16127407997846605,
      "num_tokens": 16058400.0,
      "step": 6330
    },
    {
      "entropy": 6.295368671417236,
      "epoch": 0.7311021350259665,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004957074260958972,
      "loss": 6.2097,
      "mean_token_accuracy": 0.16328535377979278,
      "num_tokens": 16071067.0,
      "step": 6335
    },
    {
      "entropy": 6.4290313720703125,
      "epoch": 0.7316791690709752,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004956994005010315,
      "loss": 6.2093,
      "mean_token_accuracy": 0.16073839813470842,
      "num_tokens": 16083760.0,
      "step": 6340
    },
    {
      "entropy": 6.383505821228027,
      "epoch": 0.7322562031159838,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000495691367483051,
      "loss": 6.1714,
      "mean_token_accuracy": 0.17427341043949127,
      "num_tokens": 16095469.0,
      "step": 6345
    },
    {
      "entropy": 6.330617809295655,
      "epoch": 0.7328332371609925,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004956833270422259,
      "loss": 6.1945,
      "mean_token_accuracy": 0.167451611161232,
      "num_tokens": 16108215.0,
      "step": 6350
    },
    {
      "entropy": 6.456769227981567,
      "epoch": 0.7334102712060011,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004956752791788269,
      "loss": 6.2532,
      "mean_token_accuracy": 0.16016148626804352,
      "num_tokens": 16121050.0,
      "step": 6355
    },
    {
      "entropy": 6.417612981796265,
      "epoch": 0.7339873052510099,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004956672238931244,
      "loss": 6.2578,
      "mean_token_accuracy": 0.1622995361685753,
      "num_tokens": 16133603.0,
      "step": 6360
    },
    {
      "entropy": 6.339549922943116,
      "epoch": 0.7345643392960185,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004956591611853894,
      "loss": 6.1409,
      "mean_token_accuracy": 0.171478071808815,
      "num_tokens": 16146797.0,
      "step": 6365
    },
    {
      "entropy": 6.4115142822265625,
      "epoch": 0.7351413733410271,
      "grad_norm": 0.73828125,
      "learning_rate": 0.0004956510910558931,
      "loss": 6.3133,
      "mean_token_accuracy": 0.1641661122441292,
      "num_tokens": 16159223.0,
      "step": 6370
    },
    {
      "entropy": 6.351845216751099,
      "epoch": 0.7357184073860358,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004956430135049071,
      "loss": 6.2143,
      "mean_token_accuracy": 0.1620581865310669,
      "num_tokens": 16172305.0,
      "step": 6375
    },
    {
      "entropy": 6.428864145278931,
      "epoch": 0.7362954414310444,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004956349285327028,
      "loss": 6.2265,
      "mean_token_accuracy": 0.16274599581956864,
      "num_tokens": 16185575.0,
      "step": 6380
    },
    {
      "entropy": 6.340291213989258,
      "epoch": 0.7368724754760531,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004956268361395523,
      "loss": 6.1484,
      "mean_token_accuracy": 0.17272731512784958,
      "num_tokens": 16198014.0,
      "step": 6385
    },
    {
      "entropy": 6.40310697555542,
      "epoch": 0.7374495095210617,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004956187363257278,
      "loss": 6.2038,
      "mean_token_accuracy": 0.16336548924446107,
      "num_tokens": 16211717.0,
      "step": 6390
    },
    {
      "entropy": 6.410084342956543,
      "epoch": 0.7380265435660704,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004956106290915017,
      "loss": 6.2492,
      "mean_token_accuracy": 0.16823570430278778,
      "num_tokens": 16225039.0,
      "step": 6395
    },
    {
      "entropy": 6.312919092178345,
      "epoch": 0.7386035776110791,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004956025144371467,
      "loss": 6.1904,
      "mean_token_accuracy": 0.16095079034566878,
      "num_tokens": 16236832.0,
      "step": 6400
    },
    {
      "entropy": 6.389898061752319,
      "epoch": 0.7391806116560877,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004955943923629356,
      "loss": 6.1755,
      "mean_token_accuracy": 0.1681758537888527,
      "num_tokens": 16250093.0,
      "step": 6405
    },
    {
      "entropy": 6.388084173202515,
      "epoch": 0.7397576457010964,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004955862628691417,
      "loss": 6.1642,
      "mean_token_accuracy": 0.1677396222949028,
      "num_tokens": 16263051.0,
      "step": 6410
    },
    {
      "entropy": 6.302446174621582,
      "epoch": 0.740334679746105,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004955781259560386,
      "loss": 6.2202,
      "mean_token_accuracy": 0.16396949291229249,
      "num_tokens": 16276545.0,
      "step": 6415
    },
    {
      "entropy": 6.393322086334228,
      "epoch": 0.7409117137911136,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004955699816238995,
      "loss": 6.2185,
      "mean_token_accuracy": 0.16904087066650392,
      "num_tokens": 16288900.0,
      "step": 6420
    },
    {
      "entropy": 6.370112466812134,
      "epoch": 0.7414887478361223,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004955618298729988,
      "loss": 6.2267,
      "mean_token_accuracy": 0.1618413209915161,
      "num_tokens": 16302066.0,
      "step": 6425
    },
    {
      "entropy": 6.420699691772461,
      "epoch": 0.742065781881131,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004955536707036105,
      "loss": 6.3154,
      "mean_token_accuracy": 0.1590120330452919,
      "num_tokens": 16316255.0,
      "step": 6430
    },
    {
      "entropy": 6.3836956977844235,
      "epoch": 0.7426428159261397,
      "grad_norm": 0.796875,
      "learning_rate": 0.000495545504116009,
      "loss": 6.2086,
      "mean_token_accuracy": 0.16407999843358995,
      "num_tokens": 16328419.0,
      "step": 6435
    },
    {
      "entropy": 6.333625841140747,
      "epoch": 0.7432198499711483,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004955373301104691,
      "loss": 6.1446,
      "mean_token_accuracy": 0.17361177057027816,
      "num_tokens": 16339864.0,
      "step": 6440
    },
    {
      "entropy": 6.400269651412964,
      "epoch": 0.743796884016157,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004955291486872657,
      "loss": 6.2265,
      "mean_token_accuracy": 0.16552126109600068,
      "num_tokens": 16354364.0,
      "step": 6445
    },
    {
      "entropy": 6.3924407958984375,
      "epoch": 0.7443739180611656,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004955209598466738,
      "loss": 6.2129,
      "mean_token_accuracy": 0.15995777845382692,
      "num_tokens": 16366118.0,
      "step": 6450
    },
    {
      "entropy": 6.413617372512817,
      "epoch": 0.7449509521061742,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000495512763588969,
      "loss": 6.2586,
      "mean_token_accuracy": 0.16330986469984055,
      "num_tokens": 16378805.0,
      "step": 6455
    },
    {
      "entropy": 6.362340259552002,
      "epoch": 0.7455279861511829,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004955045599144269,
      "loss": 6.2099,
      "mean_token_accuracy": 0.17093328684568404,
      "num_tokens": 16391202.0,
      "step": 6460
    },
    {
      "entropy": 6.318493318557739,
      "epoch": 0.7461050201961916,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004954963488233235,
      "loss": 6.1521,
      "mean_token_accuracy": 0.16465499103069306,
      "num_tokens": 16402409.0,
      "step": 6465
    },
    {
      "entropy": 6.378865432739258,
      "epoch": 0.7466820542412003,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000495488130315935,
      "loss": 6.1678,
      "mean_token_accuracy": 0.166618475317955,
      "num_tokens": 16415710.0,
      "step": 6470
    },
    {
      "entropy": 6.359663486480713,
      "epoch": 0.7472590882862089,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004954799043925377,
      "loss": 6.1871,
      "mean_token_accuracy": 0.1670057252049446,
      "num_tokens": 16428135.0,
      "step": 6475
    },
    {
      "entropy": 6.366726970672607,
      "epoch": 0.7478361223312175,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004954716710534082,
      "loss": 6.2828,
      "mean_token_accuracy": 0.16121697127819062,
      "num_tokens": 16440052.0,
      "step": 6480
    },
    {
      "entropy": 6.375087308883667,
      "epoch": 0.7484131563762262,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004954634302988237,
      "loss": 6.194,
      "mean_token_accuracy": 0.1659095734357834,
      "num_tokens": 16453188.0,
      "step": 6485
    },
    {
      "entropy": 6.43290228843689,
      "epoch": 0.7489901904212348,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004954551821290612,
      "loss": 6.2665,
      "mean_token_accuracy": 0.1626218318939209,
      "num_tokens": 16464946.0,
      "step": 6490
    },
    {
      "entropy": 6.450923252105713,
      "epoch": 0.7495672244662435,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004954469265443981,
      "loss": 6.2631,
      "mean_token_accuracy": 0.1620316132903099,
      "num_tokens": 16476957.0,
      "step": 6495
    },
    {
      "entropy": 6.407968378067016,
      "epoch": 0.7501442585112522,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004954386635451123,
      "loss": 6.1228,
      "mean_token_accuracy": 0.17593680024147035,
      "num_tokens": 16488847.0,
      "step": 6500
    },
    {
      "entropy": 6.364045333862305,
      "epoch": 0.7507212925562609,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004954303931314814,
      "loss": 6.1777,
      "mean_token_accuracy": 0.17224719375371933,
      "num_tokens": 16502497.0,
      "step": 6505
    },
    {
      "entropy": 6.375355052947998,
      "epoch": 0.7512983266012695,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004954221153037837,
      "loss": 6.2474,
      "mean_token_accuracy": 0.16628716737031937,
      "num_tokens": 16514799.0,
      "step": 6510
    },
    {
      "entropy": 6.4122003555297855,
      "epoch": 0.7518753606462781,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004954138300622978,
      "loss": 6.2757,
      "mean_token_accuracy": 0.16112708896398545,
      "num_tokens": 16528051.0,
      "step": 6515
    },
    {
      "entropy": 6.426703834533692,
      "epoch": 0.7524523946912868,
      "grad_norm": 0.78515625,
      "learning_rate": 0.000495405537407302,
      "loss": 6.2066,
      "mean_token_accuracy": 0.16361595392227174,
      "num_tokens": 16540127.0,
      "step": 6520
    },
    {
      "entropy": 6.361682319641114,
      "epoch": 0.7530294287362954,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004953972373390755,
      "loss": 6.1774,
      "mean_token_accuracy": 0.1759776309132576,
      "num_tokens": 16553419.0,
      "step": 6525
    },
    {
      "entropy": 6.464717721939087,
      "epoch": 0.753606462781304,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004953889298578975,
      "loss": 6.2795,
      "mean_token_accuracy": 0.16779824048280717,
      "num_tokens": 16567068.0,
      "step": 6530
    },
    {
      "entropy": 6.353485202789306,
      "epoch": 0.7541834968263128,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004953806149640473,
      "loss": 6.1126,
      "mean_token_accuracy": 0.1709718957543373,
      "num_tokens": 16579516.0,
      "step": 6535
    },
    {
      "entropy": 6.367344331741333,
      "epoch": 0.7547605308713214,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004953722926578045,
      "loss": 6.2389,
      "mean_token_accuracy": 0.1674762025475502,
      "num_tokens": 16592489.0,
      "step": 6540
    },
    {
      "entropy": 6.301311111450195,
      "epoch": 0.7553375649163301,
      "grad_norm": 0.875,
      "learning_rate": 0.0004953639629394492,
      "loss": 6.1814,
      "mean_token_accuracy": 0.17027597576379777,
      "num_tokens": 16604848.0,
      "step": 6545
    },
    {
      "entropy": 6.436667346954346,
      "epoch": 0.7559145989613387,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004953556258092613,
      "loss": 6.2132,
      "mean_token_accuracy": 0.16695142388343812,
      "num_tokens": 16617426.0,
      "step": 6550
    },
    {
      "entropy": 6.425419187545776,
      "epoch": 0.7564916330063474,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004953472812675216,
      "loss": 6.217,
      "mean_token_accuracy": 0.1640801399946213,
      "num_tokens": 16629676.0,
      "step": 6555
    },
    {
      "entropy": 6.4231878280639645,
      "epoch": 0.757068667051356,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004953389293145104,
      "loss": 6.2477,
      "mean_token_accuracy": 0.15641462951898574,
      "num_tokens": 16642113.0,
      "step": 6560
    },
    {
      "entropy": 6.338461065292359,
      "epoch": 0.7576457010963646,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004953305699505088,
      "loss": 6.179,
      "mean_token_accuracy": 0.16644170880317688,
      "num_tokens": 16653866.0,
      "step": 6565
    },
    {
      "entropy": 6.3428082942962645,
      "epoch": 0.7582227351413734,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004953222031757979,
      "loss": 6.2116,
      "mean_token_accuracy": 0.16792171001434325,
      "num_tokens": 16666917.0,
      "step": 6570
    },
    {
      "entropy": 6.365523910522461,
      "epoch": 0.758799769186382,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004953138289906592,
      "loss": 6.155,
      "mean_token_accuracy": 0.16653495728969575,
      "num_tokens": 16678752.0,
      "step": 6575
    },
    {
      "entropy": 6.335975551605225,
      "epoch": 0.7593768032313907,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004953054473953742,
      "loss": 6.1744,
      "mean_token_accuracy": 0.17191429883241655,
      "num_tokens": 16690942.0,
      "step": 6580
    },
    {
      "entropy": 6.373892307281494,
      "epoch": 0.7599538372763993,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004952970583902251,
      "loss": 6.1776,
      "mean_token_accuracy": 0.1657567486166954,
      "num_tokens": 16703009.0,
      "step": 6585
    },
    {
      "entropy": 6.329607057571411,
      "epoch": 0.760530871321408,
      "grad_norm": 0.7265625,
      "learning_rate": 0.0004952886619754937,
      "loss": 6.1528,
      "mean_token_accuracy": 0.16590441018342972,
      "num_tokens": 16714881.0,
      "step": 6590
    },
    {
      "entropy": 6.353391504287719,
      "epoch": 0.7611079053664166,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004952802581514625,
      "loss": 6.2282,
      "mean_token_accuracy": 0.16431566178798676,
      "num_tokens": 16729892.0,
      "step": 6595
    },
    {
      "entropy": 6.42394437789917,
      "epoch": 0.7616849394114252,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004952718469184144,
      "loss": 6.1992,
      "mean_token_accuracy": 0.16453457325696946,
      "num_tokens": 16742974.0,
      "step": 6600
    },
    {
      "entropy": 6.389783954620361,
      "epoch": 0.762261973456434,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004952634282766322,
      "loss": 6.2503,
      "mean_token_accuracy": 0.158389113843441,
      "num_tokens": 16756758.0,
      "step": 6605
    },
    {
      "entropy": 6.463902425765991,
      "epoch": 0.7628390075014426,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004952550022263988,
      "loss": 6.2623,
      "mean_token_accuracy": 0.16274497509002686,
      "num_tokens": 16768875.0,
      "step": 6610
    },
    {
      "entropy": 6.353977489471435,
      "epoch": 0.7634160415464513,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004952465687679979,
      "loss": 6.1515,
      "mean_token_accuracy": 0.16281863003969194,
      "num_tokens": 16780526.0,
      "step": 6615
    },
    {
      "entropy": 6.3285585880279545,
      "epoch": 0.7639930755914599,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000495238127901713,
      "loss": 6.1239,
      "mean_token_accuracy": 0.1764816462993622,
      "num_tokens": 16793778.0,
      "step": 6620
    },
    {
      "entropy": 6.302238130569458,
      "epoch": 0.7645701096364685,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004952296796278282,
      "loss": 6.1349,
      "mean_token_accuracy": 0.16890315860509872,
      "num_tokens": 16805367.0,
      "step": 6625
    },
    {
      "entropy": 6.379828214645386,
      "epoch": 0.7651471436814772,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004952212239466274,
      "loss": 6.1833,
      "mean_token_accuracy": 0.16319627091288566,
      "num_tokens": 16818298.0,
      "step": 6630
    },
    {
      "entropy": 6.403966951370239,
      "epoch": 0.7657241777264858,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004952127608583953,
      "loss": 6.2453,
      "mean_token_accuracy": 0.16557506173849107,
      "num_tokens": 16831031.0,
      "step": 6635
    },
    {
      "entropy": 6.322995996475219,
      "epoch": 0.7663012117714946,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004952042903634162,
      "loss": 6.2186,
      "mean_token_accuracy": 0.17293581366539001,
      "num_tokens": 16844143.0,
      "step": 6640
    },
    {
      "entropy": 6.311652565002442,
      "epoch": 0.7668782458165032,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004951958124619752,
      "loss": 6.1241,
      "mean_token_accuracy": 0.18038636595010757,
      "num_tokens": 16857639.0,
      "step": 6645
    },
    {
      "entropy": 6.274792861938477,
      "epoch": 0.7674552798615119,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004951873271543573,
      "loss": 6.1833,
      "mean_token_accuracy": 0.1697569027543068,
      "num_tokens": 16870479.0,
      "step": 6650
    },
    {
      "entropy": 6.442476463317871,
      "epoch": 0.7680323139065205,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004951788344408483,
      "loss": 6.1996,
      "mean_token_accuracy": 0.17086183428764343,
      "num_tokens": 16883876.0,
      "step": 6655
    },
    {
      "entropy": 6.388926219940186,
      "epoch": 0.7686093479515291,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004951703343217335,
      "loss": 6.1687,
      "mean_token_accuracy": 0.1668828547000885,
      "num_tokens": 16895516.0,
      "step": 6660
    },
    {
      "entropy": 6.394172859191895,
      "epoch": 0.7691863819965378,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004951618267972987,
      "loss": 6.231,
      "mean_token_accuracy": 0.1687481701374054,
      "num_tokens": 16906558.0,
      "step": 6665
    },
    {
      "entropy": 6.362077713012695,
      "epoch": 0.7697634160415464,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004951533118678304,
      "loss": 6.1972,
      "mean_token_accuracy": 0.17087253034114838,
      "num_tokens": 16919951.0,
      "step": 6670
    },
    {
      "entropy": 6.459950017929077,
      "epoch": 0.7703404500865552,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004951447895336147,
      "loss": 6.2211,
      "mean_token_accuracy": 0.16551271378993987,
      "num_tokens": 16932885.0,
      "step": 6675
    },
    {
      "entropy": 6.363070869445801,
      "epoch": 0.7709174841315638,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004951362597949384,
      "loss": 6.2169,
      "mean_token_accuracy": 0.1604614496231079,
      "num_tokens": 16944769.0,
      "step": 6680
    },
    {
      "entropy": 6.330994749069214,
      "epoch": 0.7714945181765724,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004951277226520883,
      "loss": 6.2045,
      "mean_token_accuracy": 0.162946917116642,
      "num_tokens": 16958652.0,
      "step": 6685
    },
    {
      "entropy": 6.24558916091919,
      "epoch": 0.7720715522215811,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004951191781053517,
      "loss": 6.0816,
      "mean_token_accuracy": 0.18285587728023528,
      "num_tokens": 16971850.0,
      "step": 6690
    },
    {
      "entropy": 6.400682544708252,
      "epoch": 0.7726485862665897,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004951106261550157,
      "loss": 6.1938,
      "mean_token_accuracy": 0.16207986921072007,
      "num_tokens": 16984164.0,
      "step": 6695
    },
    {
      "entropy": 6.372224807739258,
      "epoch": 0.7732256203115984,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004951020668013683,
      "loss": 6.2855,
      "mean_token_accuracy": 0.15490710884332656,
      "num_tokens": 16999140.0,
      "step": 6700
    },
    {
      "entropy": 6.408423328399659,
      "epoch": 0.773802654356607,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004950935000446972,
      "loss": 6.1631,
      "mean_token_accuracy": 0.17136792242527008,
      "num_tokens": 17012137.0,
      "step": 6705
    },
    {
      "entropy": 6.317019891738892,
      "epoch": 0.7743796884016156,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004950849258852905,
      "loss": 6.1201,
      "mean_token_accuracy": 0.1736213445663452,
      "num_tokens": 17024594.0,
      "step": 6710
    },
    {
      "entropy": 6.341596508026123,
      "epoch": 0.7749567224466244,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004950763443234366,
      "loss": 6.2444,
      "mean_token_accuracy": 0.1684948444366455,
      "num_tokens": 17036927.0,
      "step": 6715
    },
    {
      "entropy": 6.291156196594239,
      "epoch": 0.775533756491633,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004950677553594243,
      "loss": 6.1711,
      "mean_token_accuracy": 0.17052215486764907,
      "num_tokens": 17049393.0,
      "step": 6720
    },
    {
      "entropy": 6.433130264282227,
      "epoch": 0.7761107905366417,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004950591589935422,
      "loss": 6.2294,
      "mean_token_accuracy": 0.16084382981061934,
      "num_tokens": 17063053.0,
      "step": 6725
    },
    {
      "entropy": 6.465151453018189,
      "epoch": 0.7766878245816503,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004950505552260798,
      "loss": 6.2362,
      "mean_token_accuracy": 0.16392787098884581,
      "num_tokens": 17075700.0,
      "step": 6730
    },
    {
      "entropy": 6.315752601623535,
      "epoch": 0.777264858626659,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004950419440573261,
      "loss": 6.1095,
      "mean_token_accuracy": 0.16548786759376527,
      "num_tokens": 17088089.0,
      "step": 6735
    },
    {
      "entropy": 6.378903341293335,
      "epoch": 0.7778418926716676,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000495033325487571,
      "loss": 6.1886,
      "mean_token_accuracy": 0.17116651982069014,
      "num_tokens": 17100769.0,
      "step": 6740
    },
    {
      "entropy": 6.349705028533935,
      "epoch": 0.7784189267166762,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004950246995171042,
      "loss": 6.2394,
      "mean_token_accuracy": 0.1664668083190918,
      "num_tokens": 17113107.0,
      "step": 6745
    },
    {
      "entropy": 6.290965127944946,
      "epoch": 0.778995960761685,
      "grad_norm": 0.81640625,
      "learning_rate": 0.000495016066146216,
      "loss": 6.1825,
      "mean_token_accuracy": 0.17528556734323503,
      "num_tokens": 17125856.0,
      "step": 6750
    },
    {
      "entropy": 6.3704639911651615,
      "epoch": 0.7795729948066936,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004950074253751968,
      "loss": 6.1766,
      "mean_token_accuracy": 0.16756103485822677,
      "num_tokens": 17139931.0,
      "step": 6755
    },
    {
      "entropy": 6.3566728115081785,
      "epoch": 0.7801500288517023,
      "grad_norm": 0.7890625,
      "learning_rate": 0.000494998777204337,
      "loss": 6.1251,
      "mean_token_accuracy": 0.16913940459489823,
      "num_tokens": 17151669.0,
      "step": 6760
    },
    {
      "entropy": 6.3416680812835695,
      "epoch": 0.7807270628967109,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004949901216339276,
      "loss": 6.142,
      "mean_token_accuracy": 0.17505425065755845,
      "num_tokens": 17164434.0,
      "step": 6765
    },
    {
      "entropy": 6.324961948394775,
      "epoch": 0.7813040969417195,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004949814586642598,
      "loss": 6.2053,
      "mean_token_accuracy": 0.16748605370521547,
      "num_tokens": 17176732.0,
      "step": 6770
    },
    {
      "entropy": 6.379989337921143,
      "epoch": 0.7818811309867282,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000494972788295625,
      "loss": 6.1319,
      "mean_token_accuracy": 0.16905532032251358,
      "num_tokens": 17188698.0,
      "step": 6775
    },
    {
      "entropy": 6.338295555114746,
      "epoch": 0.7824581650317368,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004949641105283144,
      "loss": 6.1267,
      "mean_token_accuracy": 0.17444328367710113,
      "num_tokens": 17201219.0,
      "step": 6780
    },
    {
      "entropy": 6.366789197921753,
      "epoch": 0.7830351990767456,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004949554253626205,
      "loss": 6.235,
      "mean_token_accuracy": 0.16587198823690413,
      "num_tokens": 17213318.0,
      "step": 6785
    },
    {
      "entropy": 6.349771308898926,
      "epoch": 0.7836122331217542,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004949467327988351,
      "loss": 6.1603,
      "mean_token_accuracy": 0.16665525138378143,
      "num_tokens": 17226080.0,
      "step": 6790
    },
    {
      "entropy": 6.377096557617188,
      "epoch": 0.7841892671667628,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004949380328372505,
      "loss": 6.1844,
      "mean_token_accuracy": 0.16439106017351152,
      "num_tokens": 17237840.0,
      "step": 6795
    },
    {
      "entropy": 6.384302854537964,
      "epoch": 0.7847663012117715,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004949293254781595,
      "loss": 6.1254,
      "mean_token_accuracy": 0.17288285493850708,
      "num_tokens": 17249972.0,
      "step": 6800
    },
    {
      "entropy": 6.378346014022827,
      "epoch": 0.7853433352567801,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004949206107218547,
      "loss": 6.1792,
      "mean_token_accuracy": 0.16759558916091918,
      "num_tokens": 17261605.0,
      "step": 6805
    },
    {
      "entropy": 6.337935590744019,
      "epoch": 0.7859203693017888,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004949118885686296,
      "loss": 6.1345,
      "mean_token_accuracy": 0.16815117448568345,
      "num_tokens": 17273992.0,
      "step": 6810
    },
    {
      "entropy": 6.391584205627441,
      "epoch": 0.7864974033467974,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004949031590187774,
      "loss": 6.174,
      "mean_token_accuracy": 0.16896240413188934,
      "num_tokens": 17286978.0,
      "step": 6815
    },
    {
      "entropy": 6.266331195831299,
      "epoch": 0.7870744373918062,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004948944220725915,
      "loss": 6.1817,
      "mean_token_accuracy": 0.17290505915880203,
      "num_tokens": 17298998.0,
      "step": 6820
    },
    {
      "entropy": 6.404459142684937,
      "epoch": 0.7876514714368148,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000494885677730366,
      "loss": 6.2042,
      "mean_token_accuracy": 0.16686858087778092,
      "num_tokens": 17311614.0,
      "step": 6825
    },
    {
      "entropy": 6.361331844329834,
      "epoch": 0.7882285054818234,
      "grad_norm": 0.81640625,
      "learning_rate": 0.000494876925992395,
      "loss": 6.1912,
      "mean_token_accuracy": 0.16285034716129304,
      "num_tokens": 17323603.0,
      "step": 6830
    },
    {
      "entropy": 6.365416574478149,
      "epoch": 0.7888055395268321,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004948681668589728,
      "loss": 6.1573,
      "mean_token_accuracy": 0.17288437932729722,
      "num_tokens": 17334916.0,
      "step": 6835
    },
    {
      "entropy": 6.240475463867187,
      "epoch": 0.7893825735718407,
      "grad_norm": 0.81640625,
      "learning_rate": 0.000494859400330394,
      "loss": 6.1895,
      "mean_token_accuracy": 0.17153511196374893,
      "num_tokens": 17346777.0,
      "step": 6840
    },
    {
      "entropy": 6.372337913513183,
      "epoch": 0.7899596076168494,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004948506264069535,
      "loss": 6.1922,
      "mean_token_accuracy": 0.17216622829437256,
      "num_tokens": 17359144.0,
      "step": 6845
    },
    {
      "entropy": 6.4244777202606205,
      "epoch": 0.790536641661858,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004948418450889464,
      "loss": 6.2456,
      "mean_token_accuracy": 0.16362111568450927,
      "num_tokens": 17372992.0,
      "step": 6850
    },
    {
      "entropy": 6.280022144317627,
      "epoch": 0.7911136757068667,
      "grad_norm": 0.78125,
      "learning_rate": 0.000494833056376668,
      "loss": 6.0957,
      "mean_token_accuracy": 0.17566343545913696,
      "num_tokens": 17385413.0,
      "step": 6855
    },
    {
      "entropy": 6.367961072921753,
      "epoch": 0.7916907097518754,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004948242602704139,
      "loss": 6.2238,
      "mean_token_accuracy": 0.16973874419927598,
      "num_tokens": 17397882.0,
      "step": 6860
    },
    {
      "entropy": 6.288860702514649,
      "epoch": 0.792267743796884,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004948154567704801,
      "loss": 6.1115,
      "mean_token_accuracy": 0.16626325845718384,
      "num_tokens": 17410662.0,
      "step": 6865
    },
    {
      "entropy": 6.3847719669342045,
      "epoch": 0.7928447778418927,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004948066458771625,
      "loss": 6.2049,
      "mean_token_accuracy": 0.16260354667901994,
      "num_tokens": 17423301.0,
      "step": 6870
    },
    {
      "entropy": 6.400874137878418,
      "epoch": 0.7934218118869013,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004947978275907577,
      "loss": 6.1772,
      "mean_token_accuracy": 0.1704396441578865,
      "num_tokens": 17435733.0,
      "step": 6875
    },
    {
      "entropy": 6.368968200683594,
      "epoch": 0.79399884593191,
      "grad_norm": 0.828125,
      "learning_rate": 0.000494789001911562,
      "loss": 6.2607,
      "mean_token_accuracy": 0.15757424235343934,
      "num_tokens": 17448516.0,
      "step": 6880
    },
    {
      "entropy": 6.3357648849487305,
      "epoch": 0.7945758799769186,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004947801688398725,
      "loss": 6.1557,
      "mean_token_accuracy": 0.16735616326332092,
      "num_tokens": 17462199.0,
      "step": 6885
    },
    {
      "entropy": 6.445990657806396,
      "epoch": 0.7951529140219273,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004947713283759862,
      "loss": 6.2338,
      "mean_token_accuracy": 0.16647788584232331,
      "num_tokens": 17475651.0,
      "step": 6890
    },
    {
      "entropy": 6.448093700408935,
      "epoch": 0.795729948066936,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004947624805202003,
      "loss": 6.2342,
      "mean_token_accuracy": 0.16142688542604447,
      "num_tokens": 17488264.0,
      "step": 6895
    },
    {
      "entropy": 6.368506669998169,
      "epoch": 0.7963069821119446,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004947536252728126,
      "loss": 6.1983,
      "mean_token_accuracy": 0.16493333429098128,
      "num_tokens": 17500691.0,
      "step": 6900
    },
    {
      "entropy": 6.353544139862061,
      "epoch": 0.7968840161569533,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0004947447626341209,
      "loss": 6.1957,
      "mean_token_accuracy": 0.1679762914776802,
      "num_tokens": 17513957.0,
      "step": 6905
    },
    {
      "entropy": 6.4467274188995365,
      "epoch": 0.7974610502019619,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004947358926044232,
      "loss": 6.2185,
      "mean_token_accuracy": 0.16547130346298217,
      "num_tokens": 17527057.0,
      "step": 6910
    },
    {
      "entropy": 6.373523235321045,
      "epoch": 0.7980380842469705,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004947270151840179,
      "loss": 6.1973,
      "mean_token_accuracy": 0.15839087814092637,
      "num_tokens": 17541215.0,
      "step": 6915
    },
    {
      "entropy": 6.35720386505127,
      "epoch": 0.7986151182919792,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004947181303732038,
      "loss": 6.21,
      "mean_token_accuracy": 0.16345522552728653,
      "num_tokens": 17553116.0,
      "step": 6920
    },
    {
      "entropy": 6.306585454940796,
      "epoch": 0.7991921523369879,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004947092381722793,
      "loss": 6.1259,
      "mean_token_accuracy": 0.16889618337154388,
      "num_tokens": 17565480.0,
      "step": 6925
    },
    {
      "entropy": 6.295399761199951,
      "epoch": 0.7997691863819966,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004947003385815438,
      "loss": 6.2127,
      "mean_token_accuracy": 0.16810959428548813,
      "num_tokens": 17577706.0,
      "step": 6930
    },
    {
      "entropy": 6.434894371032715,
      "epoch": 0.8003462204270052,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004946914316012964,
      "loss": 6.2355,
      "mean_token_accuracy": 0.16640186607837676,
      "num_tokens": 17591469.0,
      "step": 6935
    },
    {
      "entropy": 6.428550434112549,
      "epoch": 0.8009232544720138,
      "grad_norm": 0.7890625,
      "learning_rate": 0.000494682517231837,
      "loss": 6.164,
      "mean_token_accuracy": 0.17068625688552858,
      "num_tokens": 17605225.0,
      "step": 6940
    },
    {
      "entropy": 6.245584726333618,
      "epoch": 0.8015002885170225,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004946735954734652,
      "loss": 6.0642,
      "mean_token_accuracy": 0.17588913589715957,
      "num_tokens": 17618509.0,
      "step": 6945
    },
    {
      "entropy": 6.3537209033966064,
      "epoch": 0.8020773225620311,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004946646663264811,
      "loss": 6.2486,
      "mean_token_accuracy": 0.1653627946972847,
      "num_tokens": 17630652.0,
      "step": 6950
    },
    {
      "entropy": 6.327255773544311,
      "epoch": 0.8026543566070398,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004946557297911852,
      "loss": 6.079,
      "mean_token_accuracy": 0.17674667686223983,
      "num_tokens": 17641935.0,
      "step": 6955
    },
    {
      "entropy": 6.3670917510986325,
      "epoch": 0.8032313906520485,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004946467858678777,
      "loss": 6.2835,
      "mean_token_accuracy": 0.15880746990442277,
      "num_tokens": 17654043.0,
      "step": 6960
    },
    {
      "entropy": 6.446195363998413,
      "epoch": 0.8038084246970572,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004946378345568597,
      "loss": 6.2089,
      "mean_token_accuracy": 0.16398643404245378,
      "num_tokens": 17666566.0,
      "step": 6965
    },
    {
      "entropy": 6.234373331069946,
      "epoch": 0.8043854587420658,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004946288758584324,
      "loss": 6.0275,
      "mean_token_accuracy": 0.17151330709457396,
      "num_tokens": 17678463.0,
      "step": 6970
    },
    {
      "entropy": 6.405322551727295,
      "epoch": 0.8049624927870744,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004946199097728968,
      "loss": 6.2667,
      "mean_token_accuracy": 0.16178204566240312,
      "num_tokens": 17691663.0,
      "step": 6975
    },
    {
      "entropy": 6.395699882507325,
      "epoch": 0.8055395268320831,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004946109363005548,
      "loss": 6.2324,
      "mean_token_accuracy": 0.16503103375434874,
      "num_tokens": 17705048.0,
      "step": 6980
    },
    {
      "entropy": 6.363728809356689,
      "epoch": 0.8061165608770917,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000494601955441708,
      "loss": 6.211,
      "mean_token_accuracy": 0.16915686428546906,
      "num_tokens": 17717865.0,
      "step": 6985
    },
    {
      "entropy": 6.375814628601074,
      "epoch": 0.8066935949221004,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004945929671966585,
      "loss": 6.2272,
      "mean_token_accuracy": 0.16043869405984879,
      "num_tokens": 17730568.0,
      "step": 6990
    },
    {
      "entropy": 6.314331912994385,
      "epoch": 0.8072706289671091,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004945839715657085,
      "loss": 6.1815,
      "mean_token_accuracy": 0.16977579444646834,
      "num_tokens": 17743303.0,
      "step": 6995
    },
    {
      "entropy": 6.427794599533081,
      "epoch": 0.8078476630121177,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004945749685491607,
      "loss": 6.2577,
      "mean_token_accuracy": 0.1618587702512741,
      "num_tokens": 17756552.0,
      "step": 7000
    },
    {
      "entropy": 6.3551558494567875,
      "epoch": 0.8084246970571264,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004945659581473181,
      "loss": 6.2785,
      "mean_token_accuracy": 0.16330808997154236,
      "num_tokens": 17770783.0,
      "step": 7005
    },
    {
      "entropy": 6.383954095840454,
      "epoch": 0.809001731102135,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004945569403604833,
      "loss": 6.1562,
      "mean_token_accuracy": 0.16762078553438187,
      "num_tokens": 17784573.0,
      "step": 7010
    },
    {
      "entropy": 6.424306869506836,
      "epoch": 0.8095787651471437,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004945479151889601,
      "loss": 6.1923,
      "mean_token_accuracy": 0.16073544770479203,
      "num_tokens": 17797304.0,
      "step": 7015
    },
    {
      "entropy": 6.291573762893677,
      "epoch": 0.8101557991921523,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004945388826330517,
      "loss": 6.0622,
      "mean_token_accuracy": 0.17693958580493926,
      "num_tokens": 17810401.0,
      "step": 7020
    },
    {
      "entropy": 6.305244064331054,
      "epoch": 0.8107328332371609,
      "grad_norm": 0.7421875,
      "learning_rate": 0.0004945298426930621,
      "loss": 6.1434,
      "mean_token_accuracy": 0.17025542855262757,
      "num_tokens": 17824120.0,
      "step": 7025
    },
    {
      "entropy": 6.380581474304199,
      "epoch": 0.8113098672821697,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004945207953692952,
      "loss": 6.1799,
      "mean_token_accuracy": 0.16469264179468154,
      "num_tokens": 17836865.0,
      "step": 7030
    },
    {
      "entropy": 6.311848449707031,
      "epoch": 0.8118869013271783,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004945117406620556,
      "loss": 6.2074,
      "mean_token_accuracy": 0.1644309490919113,
      "num_tokens": 17849128.0,
      "step": 7035
    },
    {
      "entropy": 6.367767953872681,
      "epoch": 0.812463935372187,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004945026785716474,
      "loss": 6.1677,
      "mean_token_accuracy": 0.1677388444542885,
      "num_tokens": 17861151.0,
      "step": 7040
    },
    {
      "entropy": 6.327512121200561,
      "epoch": 0.8130409694171956,
      "grad_norm": 0.875,
      "learning_rate": 0.0004944936090983757,
      "loss": 6.1501,
      "mean_token_accuracy": 0.1745579496026039,
      "num_tokens": 17874043.0,
      "step": 7045
    },
    {
      "entropy": 6.328273725509644,
      "epoch": 0.8136180034622043,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004944845322425455,
      "loss": 6.1685,
      "mean_token_accuracy": 0.1651006817817688,
      "num_tokens": 17887176.0,
      "step": 7050
    },
    {
      "entropy": 6.4299534320831295,
      "epoch": 0.8141950375072129,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004944754480044621,
      "loss": 6.2311,
      "mean_token_accuracy": 0.1592079609632492,
      "num_tokens": 17899708.0,
      "step": 7055
    },
    {
      "entropy": 6.320512151718139,
      "epoch": 0.8147720715522215,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004944663563844311,
      "loss": 6.1268,
      "mean_token_accuracy": 0.17118301540613173,
      "num_tokens": 17912303.0,
      "step": 7060
    },
    {
      "entropy": 6.379102373123169,
      "epoch": 0.8153491055972303,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004944572573827581,
      "loss": 6.2577,
      "mean_token_accuracy": 0.15897056609392166,
      "num_tokens": 17925963.0,
      "step": 7065
    },
    {
      "entropy": 6.393832969665527,
      "epoch": 0.8159261396422389,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004944481509997494,
      "loss": 6.2386,
      "mean_token_accuracy": 0.16631796956062317,
      "num_tokens": 17938786.0,
      "step": 7070
    },
    {
      "entropy": 6.40287504196167,
      "epoch": 0.8165031736872476,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000494439037235711,
      "loss": 6.1682,
      "mean_token_accuracy": 0.1683258295059204,
      "num_tokens": 17950646.0,
      "step": 7075
    },
    {
      "entropy": 6.379344654083252,
      "epoch": 0.8170802077322562,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004944299160909495,
      "loss": 6.1901,
      "mean_token_accuracy": 0.16389957815408707,
      "num_tokens": 17963115.0,
      "step": 7080
    },
    {
      "entropy": 6.328940725326538,
      "epoch": 0.8176572417772648,
      "grad_norm": 0.8125,
      "learning_rate": 0.000494420787565772,
      "loss": 6.1938,
      "mean_token_accuracy": 0.1683003857731819,
      "num_tokens": 17976068.0,
      "step": 7085
    },
    {
      "entropy": 6.358946800231934,
      "epoch": 0.8182342758222735,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004944116516604852,
      "loss": 6.1481,
      "mean_token_accuracy": 0.17259256839752196,
      "num_tokens": 17988487.0,
      "step": 7090
    },
    {
      "entropy": 6.357552099227905,
      "epoch": 0.8188113098672821,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004944025083753965,
      "loss": 6.1936,
      "mean_token_accuracy": 0.1638806402683258,
      "num_tokens": 18001664.0,
      "step": 7095
    },
    {
      "entropy": 6.372899293899536,
      "epoch": 0.8193883439122909,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004943933577108133,
      "loss": 6.2075,
      "mean_token_accuracy": 0.16567729860544206,
      "num_tokens": 18014384.0,
      "step": 7100
    },
    {
      "entropy": 6.366517114639282,
      "epoch": 0.8199653779572995,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004943841996670436,
      "loss": 6.1411,
      "mean_token_accuracy": 0.16836489140987396,
      "num_tokens": 18027516.0,
      "step": 7105
    },
    {
      "entropy": 6.323846769332886,
      "epoch": 0.8205424120023082,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004943750342443953,
      "loss": 6.0784,
      "mean_token_accuracy": 0.1761428400874138,
      "num_tokens": 18039771.0,
      "step": 7110
    },
    {
      "entropy": 6.335892629623413,
      "epoch": 0.8211194460473168,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004943658614431767,
      "loss": 6.2552,
      "mean_token_accuracy": 0.1610368736088276,
      "num_tokens": 18052620.0,
      "step": 7115
    },
    {
      "entropy": 6.236982154846191,
      "epoch": 0.8216964800923254,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004943566812636963,
      "loss": 6.0156,
      "mean_token_accuracy": 0.17757391780614853,
      "num_tokens": 18065723.0,
      "step": 7120
    },
    {
      "entropy": 6.309774351119995,
      "epoch": 0.8222735141373341,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000494347493706263,
      "loss": 6.1139,
      "mean_token_accuracy": 0.16816986054182054,
      "num_tokens": 18078335.0,
      "step": 7125
    },
    {
      "entropy": 6.3064737796783445,
      "epoch": 0.8228505481823427,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004943382987711856,
      "loss": 6.1359,
      "mean_token_accuracy": 0.17049853056669234,
      "num_tokens": 18089943.0,
      "step": 7130
    },
    {
      "entropy": 6.325173425674438,
      "epoch": 0.8234275822273515,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004943290964587734,
      "loss": 6.0903,
      "mean_token_accuracy": 0.16795280575752258,
      "num_tokens": 18101843.0,
      "step": 7135
    },
    {
      "entropy": 6.241851854324341,
      "epoch": 0.8240046162723601,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004943198867693361,
      "loss": 6.0009,
      "mean_token_accuracy": 0.18136388063430786,
      "num_tokens": 18113170.0,
      "step": 7140
    },
    {
      "entropy": 6.325794649124146,
      "epoch": 0.8245816503173687,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004943106697031833,
      "loss": 6.1785,
      "mean_token_accuracy": 0.16728792637586593,
      "num_tokens": 18124936.0,
      "step": 7145
    },
    {
      "entropy": 6.3345221996307375,
      "epoch": 0.8251586843623774,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004943014452606251,
      "loss": 6.1313,
      "mean_token_accuracy": 0.16424822360277175,
      "num_tokens": 18137894.0,
      "step": 7150
    },
    {
      "entropy": 6.357614278793335,
      "epoch": 0.825735718407386,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004942922134419717,
      "loss": 6.1923,
      "mean_token_accuracy": 0.16184918731451034,
      "num_tokens": 18149170.0,
      "step": 7155
    },
    {
      "entropy": 6.346024084091186,
      "epoch": 0.8263127524523947,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004942829742475336,
      "loss": 6.2033,
      "mean_token_accuracy": 0.1654985547065735,
      "num_tokens": 18160694.0,
      "step": 7160
    },
    {
      "entropy": 6.423761940002441,
      "epoch": 0.8268897864974033,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004942737276776217,
      "loss": 6.3124,
      "mean_token_accuracy": 0.16496190130710603,
      "num_tokens": 18173619.0,
      "step": 7165
    },
    {
      "entropy": 6.449676036834717,
      "epoch": 0.827466820542412,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004942644737325468,
      "loss": 6.2614,
      "mean_token_accuracy": 0.16386640071868896,
      "num_tokens": 18186751.0,
      "step": 7170
    },
    {
      "entropy": 6.325207805633545,
      "epoch": 0.8280438545874207,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004942552124126202,
      "loss": 6.1525,
      "mean_token_accuracy": 0.16702402085065843,
      "num_tokens": 18198701.0,
      "step": 7175
    },
    {
      "entropy": 6.350707149505615,
      "epoch": 0.8286208886324293,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004942459437181535,
      "loss": 6.1251,
      "mean_token_accuracy": 0.17409539818763733,
      "num_tokens": 18211422.0,
      "step": 7180
    },
    {
      "entropy": 6.320394611358642,
      "epoch": 0.829197922677438,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004942366676494584,
      "loss": 6.2097,
      "mean_token_accuracy": 0.1659229278564453,
      "num_tokens": 18223427.0,
      "step": 7185
    },
    {
      "entropy": 6.277689838409424,
      "epoch": 0.8297749567224466,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004942273842068469,
      "loss": 6.1679,
      "mean_token_accuracy": 0.17429596334695815,
      "num_tokens": 18235234.0,
      "step": 7190
    },
    {
      "entropy": 6.374852085113526,
      "epoch": 0.8303519907674553,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004942180933906311,
      "loss": 6.0967,
      "mean_token_accuracy": 0.17455530613660813,
      "num_tokens": 18247669.0,
      "step": 7195
    },
    {
      "entropy": 6.261248683929443,
      "epoch": 0.8309290248124639,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004942087952011237,
      "loss": 6.0792,
      "mean_token_accuracy": 0.17736732363700866,
      "num_tokens": 18260987.0,
      "step": 7200
    },
    {
      "entropy": 6.378260564804077,
      "epoch": 0.8315060588574726,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004941994896386374,
      "loss": 6.1867,
      "mean_token_accuracy": 0.1577957659959793,
      "num_tokens": 18274805.0,
      "step": 7205
    },
    {
      "entropy": 6.360244607925415,
      "epoch": 0.8320830929024813,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004941901767034851,
      "loss": 6.1673,
      "mean_token_accuracy": 0.16796040832996367,
      "num_tokens": 18286438.0,
      "step": 7210
    },
    {
      "entropy": 6.357167530059814,
      "epoch": 0.8326601269474899,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004941808563959802,
      "loss": 6.1856,
      "mean_token_accuracy": 0.16499463766813277,
      "num_tokens": 18299310.0,
      "step": 7215
    },
    {
      "entropy": 6.29741940498352,
      "epoch": 0.8332371609924986,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004941715287164359,
      "loss": 6.1245,
      "mean_token_accuracy": 0.16438130140304566,
      "num_tokens": 18311286.0,
      "step": 7220
    },
    {
      "entropy": 6.32746958732605,
      "epoch": 0.8338141950375072,
      "grad_norm": 0.75,
      "learning_rate": 0.0004941621936651661,
      "loss": 6.1803,
      "mean_token_accuracy": 0.16897405236959456,
      "num_tokens": 18325759.0,
      "step": 7225
    },
    {
      "entropy": 6.34986686706543,
      "epoch": 0.8343912290825158,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004941528512424849,
      "loss": 6.1252,
      "mean_token_accuracy": 0.1749942496418953,
      "num_tokens": 18338907.0,
      "step": 7230
    },
    {
      "entropy": 6.402791833877563,
      "epoch": 0.8349682631275245,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004941435014487064,
      "loss": 6.1905,
      "mean_token_accuracy": 0.16975402384996413,
      "num_tokens": 18351760.0,
      "step": 7235
    },
    {
      "entropy": 6.343279838562012,
      "epoch": 0.8355452971725332,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000494134144284145,
      "loss": 6.1873,
      "mean_token_accuracy": 0.1620405212044716,
      "num_tokens": 18363348.0,
      "step": 7240
    },
    {
      "entropy": 6.356585359573364,
      "epoch": 0.8361223312175419,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004941247797491156,
      "loss": 6.2152,
      "mean_token_accuracy": 0.16385446041822432,
      "num_tokens": 18375233.0,
      "step": 7245
    },
    {
      "entropy": 6.361816120147705,
      "epoch": 0.8366993652625505,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004941154078439329,
      "loss": 6.129,
      "mean_token_accuracy": 0.16602645069360733,
      "num_tokens": 18388001.0,
      "step": 7250
    },
    {
      "entropy": 6.210663080215454,
      "epoch": 0.8372763993075591,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004941060285689126,
      "loss": 6.0004,
      "mean_token_accuracy": 0.1854765772819519,
      "num_tokens": 18399647.0,
      "step": 7255
    },
    {
      "entropy": 6.286527299880982,
      "epoch": 0.8378534333525678,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004940966419243695,
      "loss": 6.1585,
      "mean_token_accuracy": 0.17007975280284882,
      "num_tokens": 18412441.0,
      "step": 7260
    },
    {
      "entropy": 6.3775622844696045,
      "epoch": 0.8384304673975764,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00049408724791062,
      "loss": 6.1714,
      "mean_token_accuracy": 0.17022394239902497,
      "num_tokens": 18425232.0,
      "step": 7265
    },
    {
      "entropy": 6.389389944076538,
      "epoch": 0.8390075014425851,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004940778465279797,
      "loss": 6.1677,
      "mean_token_accuracy": 0.1612836867570877,
      "num_tokens": 18438148.0,
      "step": 7270
    },
    {
      "entropy": 6.276881313323974,
      "epoch": 0.8395845354875938,
      "grad_norm": 0.875,
      "learning_rate": 0.0004940684377767647,
      "loss": 6.006,
      "mean_token_accuracy": 0.17631301879882813,
      "num_tokens": 18449955.0,
      "step": 7275
    },
    {
      "entropy": 6.251585054397583,
      "epoch": 0.8401615695326025,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004940590216572916,
      "loss": 6.105,
      "mean_token_accuracy": 0.17243984788656236,
      "num_tokens": 18463125.0,
      "step": 7280
    },
    {
      "entropy": 6.367015790939331,
      "epoch": 0.8407386035776111,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004940495981698772,
      "loss": 6.2058,
      "mean_token_accuracy": 0.16317611634731294,
      "num_tokens": 18475280.0,
      "step": 7285
    },
    {
      "entropy": 6.35017147064209,
      "epoch": 0.8413156376226197,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0004940401673148384,
      "loss": 6.162,
      "mean_token_accuracy": 0.16571533679962158,
      "num_tokens": 18487406.0,
      "step": 7290
    },
    {
      "entropy": 6.275267219543457,
      "epoch": 0.8418926716676284,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004940307290924923,
      "loss": 6.1234,
      "mean_token_accuracy": 0.17094551920890808,
      "num_tokens": 18500288.0,
      "step": 7295
    },
    {
      "entropy": 6.404674673080445,
      "epoch": 0.842469705712637,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004940212835031565,
      "loss": 6.1988,
      "mean_token_accuracy": 0.1616864025592804,
      "num_tokens": 18514254.0,
      "step": 7300
    },
    {
      "entropy": 6.36431851387024,
      "epoch": 0.8430467397576457,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004940118305471486,
      "loss": 6.2378,
      "mean_token_accuracy": 0.15897418856620787,
      "num_tokens": 18526536.0,
      "step": 7305
    },
    {
      "entropy": 6.427651834487915,
      "epoch": 0.8436237738026544,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004940023702247866,
      "loss": 6.1441,
      "mean_token_accuracy": 0.16780431121587752,
      "num_tokens": 18540137.0,
      "step": 7310
    },
    {
      "entropy": 6.333643674850464,
      "epoch": 0.844200807847663,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004939929025363886,
      "loss": 6.1381,
      "mean_token_accuracy": 0.17276106625795365,
      "num_tokens": 18552975.0,
      "step": 7315
    },
    {
      "entropy": 6.247806549072266,
      "epoch": 0.8447778418926717,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004939834274822731,
      "loss": 6.0625,
      "mean_token_accuracy": 0.17385534346103668,
      "num_tokens": 18564653.0,
      "step": 7320
    },
    {
      "entropy": 6.364656734466553,
      "epoch": 0.8453548759376803,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004939739450627588,
      "loss": 6.2033,
      "mean_token_accuracy": 0.16594440788030623,
      "num_tokens": 18577226.0,
      "step": 7325
    },
    {
      "entropy": 6.347729778289795,
      "epoch": 0.845931909982689,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004939644552781647,
      "loss": 6.1359,
      "mean_token_accuracy": 0.16554813832044601,
      "num_tokens": 18591206.0,
      "step": 7330
    },
    {
      "entropy": 6.305720424652099,
      "epoch": 0.8465089440276976,
      "grad_norm": 0.875,
      "learning_rate": 0.0004939549581288099,
      "loss": 6.138,
      "mean_token_accuracy": 0.17282395362854003,
      "num_tokens": 18604182.0,
      "step": 7335
    },
    {
      "entropy": 6.379542970657349,
      "epoch": 0.8470859780727062,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004939454536150138,
      "loss": 6.2233,
      "mean_token_accuracy": 0.16193835139274598,
      "num_tokens": 18615830.0,
      "step": 7340
    },
    {
      "entropy": 6.351224136352539,
      "epoch": 0.847663012117715,
      "grad_norm": 0.828125,
      "learning_rate": 0.000493935941737096,
      "loss": 6.1833,
      "mean_token_accuracy": 0.16340697109699248,
      "num_tokens": 18628372.0,
      "step": 7345
    },
    {
      "entropy": 6.373519563674927,
      "epoch": 0.8482400461627236,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004939264224953768,
      "loss": 6.2525,
      "mean_token_accuracy": 0.1579662710428238,
      "num_tokens": 18641470.0,
      "step": 7350
    },
    {
      "entropy": 6.371709108352661,
      "epoch": 0.8488170802077323,
      "grad_norm": 0.8046875,
      "learning_rate": 0.000493916895890176,
      "loss": 6.0974,
      "mean_token_accuracy": 0.1789488598704338,
      "num_tokens": 18654660.0,
      "step": 7355
    },
    {
      "entropy": 6.333353853225708,
      "epoch": 0.8493941142527409,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000493907361921814,
      "loss": 6.1924,
      "mean_token_accuracy": 0.16414279490709305,
      "num_tokens": 18665626.0,
      "step": 7360
    },
    {
      "entropy": 6.318211460113526,
      "epoch": 0.8499711482977496,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004938978205906118,
      "loss": 6.1137,
      "mean_token_accuracy": 0.17439971268177032,
      "num_tokens": 18677175.0,
      "step": 7365
    },
    {
      "entropy": 6.343401908874512,
      "epoch": 0.8505481823427582,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004938882718968901,
      "loss": 6.1332,
      "mean_token_accuracy": 0.17032305002212525,
      "num_tokens": 18689005.0,
      "step": 7370
    },
    {
      "entropy": 6.2724262237548825,
      "epoch": 0.8511252163877668,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004938787158409702,
      "loss": 6.181,
      "mean_token_accuracy": 0.1706418961286545,
      "num_tokens": 18702085.0,
      "step": 7375
    },
    {
      "entropy": 6.351395463943481,
      "epoch": 0.8517022504327756,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004938691524231733,
      "loss": 6.1819,
      "mean_token_accuracy": 0.16577364206314088,
      "num_tokens": 18715064.0,
      "step": 7380
    },
    {
      "entropy": 6.367648410797119,
      "epoch": 0.8522792844777842,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004938595816438212,
      "loss": 6.1981,
      "mean_token_accuracy": 0.16841965466737746,
      "num_tokens": 18728105.0,
      "step": 7385
    },
    {
      "entropy": 6.286236524581909,
      "epoch": 0.8528563185227929,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004938500035032358,
      "loss": 6.1076,
      "mean_token_accuracy": 0.1745881989598274,
      "num_tokens": 18741021.0,
      "step": 7390
    },
    {
      "entropy": 6.283361291885376,
      "epoch": 0.8534333525678015,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004938404180017392,
      "loss": 6.1466,
      "mean_token_accuracy": 0.17935479879379274,
      "num_tokens": 18753271.0,
      "step": 7395
    },
    {
      "entropy": 6.326036405563355,
      "epoch": 0.8540103866128101,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004938308251396539,
      "loss": 6.1516,
      "mean_token_accuracy": 0.167026287317276,
      "num_tokens": 18765631.0,
      "step": 7400
    },
    {
      "entropy": 6.335333061218262,
      "epoch": 0.8545874206578188,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004938212249173024,
      "loss": 6.1485,
      "mean_token_accuracy": 0.1670267701148987,
      "num_tokens": 18778472.0,
      "step": 7405
    },
    {
      "entropy": 6.418694257736206,
      "epoch": 0.8551644547028274,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004938116173350078,
      "loss": 6.1868,
      "mean_token_accuracy": 0.16128545999526978,
      "num_tokens": 18791564.0,
      "step": 7410
    },
    {
      "entropy": 6.310502290725708,
      "epoch": 0.8557414887478362,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0004938020023930932,
      "loss": 6.1428,
      "mean_token_accuracy": 0.16949567347764968,
      "num_tokens": 18804749.0,
      "step": 7415
    },
    {
      "entropy": 6.373205137252808,
      "epoch": 0.8563185227928448,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004937923800918817,
      "loss": 6.1656,
      "mean_token_accuracy": 0.16341724693775178,
      "num_tokens": 18816863.0,
      "step": 7420
    },
    {
      "entropy": 6.354863500595092,
      "epoch": 0.8568955568378535,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004937827504316974,
      "loss": 6.1286,
      "mean_token_accuracy": 0.17136932164430618,
      "num_tokens": 18831543.0,
      "step": 7425
    },
    {
      "entropy": 6.3266009330749515,
      "epoch": 0.8574725908828621,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004937731134128639,
      "loss": 6.1649,
      "mean_token_accuracy": 0.16626855731010437,
      "num_tokens": 18844242.0,
      "step": 7430
    },
    {
      "entropy": 6.309177446365356,
      "epoch": 0.8580496249278707,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004937634690357054,
      "loss": 6.1785,
      "mean_token_accuracy": 0.16604990065097808,
      "num_tokens": 18856213.0,
      "step": 7435
    },
    {
      "entropy": 6.389643764495849,
      "epoch": 0.8586266589728794,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004937538173005462,
      "loss": 6.1669,
      "mean_token_accuracy": 0.16924736946821212,
      "num_tokens": 18869680.0,
      "step": 7440
    },
    {
      "entropy": 6.371203804016114,
      "epoch": 0.859203693017888,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004937441582077111,
      "loss": 6.1415,
      "mean_token_accuracy": 0.17059220224618912,
      "num_tokens": 18882360.0,
      "step": 7445
    },
    {
      "entropy": 6.28338508605957,
      "epoch": 0.8597807270628968,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004937344917575249,
      "loss": 6.1129,
      "mean_token_accuracy": 0.16984111964702606,
      "num_tokens": 18895138.0,
      "step": 7450
    },
    {
      "entropy": 6.285325384140014,
      "epoch": 0.8603577611079054,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004937248179503127,
      "loss": 6.1584,
      "mean_token_accuracy": 0.16774938404560089,
      "num_tokens": 18908849.0,
      "step": 7455
    },
    {
      "entropy": 6.353943300247193,
      "epoch": 0.860934795152914,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004937151367863998,
      "loss": 6.1267,
      "mean_token_accuracy": 0.17231214940547943,
      "num_tokens": 18920757.0,
      "step": 7460
    },
    {
      "entropy": 6.2311420917510985,
      "epoch": 0.8615118291979227,
      "grad_norm": 0.859375,
      "learning_rate": 0.000493705448266112,
      "loss": 6.0605,
      "mean_token_accuracy": 0.17295785695314408,
      "num_tokens": 18933239.0,
      "step": 7465
    },
    {
      "entropy": 6.357706356048584,
      "epoch": 0.8620888632429313,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004936957523897752,
      "loss": 6.177,
      "mean_token_accuracy": 0.16270439624786376,
      "num_tokens": 18945174.0,
      "step": 7470
    },
    {
      "entropy": 6.391672039031983,
      "epoch": 0.86266589728794,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004936860491577153,
      "loss": 6.2446,
      "mean_token_accuracy": 0.15656707137823106,
      "num_tokens": 18957478.0,
      "step": 7475
    },
    {
      "entropy": 6.338565874099731,
      "epoch": 0.8632429313329486,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004936763385702588,
      "loss": 6.1722,
      "mean_token_accuracy": 0.16844886690378189,
      "num_tokens": 18969461.0,
      "step": 7480
    },
    {
      "entropy": 6.447391748428345,
      "epoch": 0.8638199653779572,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004936666206277322,
      "loss": 6.1648,
      "mean_token_accuracy": 0.16605869233608245,
      "num_tokens": 18981891.0,
      "step": 7485
    },
    {
      "entropy": 6.316598176956177,
      "epoch": 0.864396999422966,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004936568953304624,
      "loss": 6.039,
      "mean_token_accuracy": 0.17837050408124924,
      "num_tokens": 18993376.0,
      "step": 7490
    },
    {
      "entropy": 6.228720474243164,
      "epoch": 0.8649740334679746,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004936471626787766,
      "loss": 6.1226,
      "mean_token_accuracy": 0.17042396664619447,
      "num_tokens": 19005485.0,
      "step": 7495
    },
    {
      "entropy": 6.285957956314087,
      "epoch": 0.8655510675129833,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004936374226730022,
      "loss": 6.1756,
      "mean_token_accuracy": 0.16250619143247605,
      "num_tokens": 19018509.0,
      "step": 7500
    },
    {
      "entropy": 6.430623006820679,
      "epoch": 0.8661281015579919,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004936276753134666,
      "loss": 6.1405,
      "mean_token_accuracy": 0.16987382471561432,
      "num_tokens": 19030595.0,
      "step": 7505
    },
    {
      "entropy": 6.373871994018555,
      "epoch": 0.8667051356030006,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004936179206004976,
      "loss": 6.2206,
      "mean_token_accuracy": 0.16224805116653443,
      "num_tokens": 19043441.0,
      "step": 7510
    },
    {
      "entropy": 6.303015947341919,
      "epoch": 0.8672821696480092,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004936081585344236,
      "loss": 6.1184,
      "mean_token_accuracy": 0.16949329972267152,
      "num_tokens": 19055761.0,
      "step": 7515
    },
    {
      "entropy": 6.329971075057983,
      "epoch": 0.8678592036930178,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004935983891155727,
      "loss": 6.1628,
      "mean_token_accuracy": 0.16706838011741637,
      "num_tokens": 19068211.0,
      "step": 7520
    },
    {
      "entropy": 6.333680200576782,
      "epoch": 0.8684362377380266,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004935886123442737,
      "loss": 6.1094,
      "mean_token_accuracy": 0.16620445251464844,
      "num_tokens": 19081487.0,
      "step": 7525
    },
    {
      "entropy": 6.3789692401885985,
      "epoch": 0.8690132717830352,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004935788282208551,
      "loss": 6.1258,
      "mean_token_accuracy": 0.16772944927215577,
      "num_tokens": 19094779.0,
      "step": 7530
    },
    {
      "entropy": 6.302552080154419,
      "epoch": 0.8695903058280439,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004935690367456464,
      "loss": 6.1898,
      "mean_token_accuracy": 0.1666984111070633,
      "num_tokens": 19106999.0,
      "step": 7535
    },
    {
      "entropy": 6.372557210922241,
      "epoch": 0.8701673398730525,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004935592379189766,
      "loss": 6.1381,
      "mean_token_accuracy": 0.1672473892569542,
      "num_tokens": 19119570.0,
      "step": 7540
    },
    {
      "entropy": 6.3990599632263185,
      "epoch": 0.8707443739180611,
      "grad_norm": 0.875,
      "learning_rate": 0.0004935494317411754,
      "loss": 6.1411,
      "mean_token_accuracy": 0.16526359766721727,
      "num_tokens": 19131101.0,
      "step": 7545
    },
    {
      "entropy": 6.244806814193725,
      "epoch": 0.8713214079630698,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004935396182125726,
      "loss": 6.1125,
      "mean_token_accuracy": 0.16791403144598008,
      "num_tokens": 19144159.0,
      "step": 7550
    },
    {
      "entropy": 6.387464237213135,
      "epoch": 0.8718984420080784,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004935297973334983,
      "loss": 6.1054,
      "mean_token_accuracy": 0.1751489594578743,
      "num_tokens": 19157034.0,
      "step": 7555
    },
    {
      "entropy": 6.3096170902252195,
      "epoch": 0.8724754760530872,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004935199691042828,
      "loss": 6.1511,
      "mean_token_accuracy": 0.17347051203250885,
      "num_tokens": 19170813.0,
      "step": 7560
    },
    {
      "entropy": 6.277336311340332,
      "epoch": 0.8730525100980958,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004935101335252568,
      "loss": 6.093,
      "mean_token_accuracy": 0.17822133302688598,
      "num_tokens": 19183060.0,
      "step": 7565
    },
    {
      "entropy": 6.308446788787842,
      "epoch": 0.8736295441431045,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004935002905967509,
      "loss": 6.0312,
      "mean_token_accuracy": 0.17758481651544572,
      "num_tokens": 19195615.0,
      "step": 7570
    },
    {
      "entropy": 6.345848464965821,
      "epoch": 0.8742065781881131,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004934904403190963,
      "loss": 6.1979,
      "mean_token_accuracy": 0.16513141840696335,
      "num_tokens": 19209531.0,
      "step": 7575
    },
    {
      "entropy": 6.288311004638672,
      "epoch": 0.8747836122331217,
      "grad_norm": 0.875,
      "learning_rate": 0.0004934805826926242,
      "loss": 6.1377,
      "mean_token_accuracy": 0.17015064060688018,
      "num_tokens": 19221658.0,
      "step": 7580
    },
    {
      "entropy": 6.373033761978149,
      "epoch": 0.8753606462781304,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004934707177176663,
      "loss": 6.1451,
      "mean_token_accuracy": 0.16882509142160415,
      "num_tokens": 19233491.0,
      "step": 7585
    },
    {
      "entropy": 6.36579999923706,
      "epoch": 0.875937680323139,
      "grad_norm": 0.875,
      "learning_rate": 0.0004934608453945543,
      "loss": 6.2309,
      "mean_token_accuracy": 0.1647283762693405,
      "num_tokens": 19246318.0,
      "step": 7590
    },
    {
      "entropy": 6.334402990341187,
      "epoch": 0.8765147143681478,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004934509657236203,
      "loss": 6.165,
      "mean_token_accuracy": 0.16658859997987746,
      "num_tokens": 19258735.0,
      "step": 7595
    },
    {
      "entropy": 6.417309999465942,
      "epoch": 0.8770917484131564,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004934410787051965,
      "loss": 6.2063,
      "mean_token_accuracy": 0.16228333860635757,
      "num_tokens": 19271640.0,
      "step": 7600
    },
    {
      "entropy": 6.300780439376831,
      "epoch": 0.877668782458165,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004934311843396157,
      "loss": 6.1118,
      "mean_token_accuracy": 0.17644276469945908,
      "num_tokens": 19283824.0,
      "step": 7605
    },
    {
      "entropy": 6.311488914489746,
      "epoch": 0.8782458165031737,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004934212826272104,
      "loss": 6.0685,
      "mean_token_accuracy": 0.17539282739162446,
      "num_tokens": 19298072.0,
      "step": 7610
    },
    {
      "entropy": 6.370089292526245,
      "epoch": 0.8788228505481823,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004934113735683137,
      "loss": 6.1521,
      "mean_token_accuracy": 0.16698871701955795,
      "num_tokens": 19311132.0,
      "step": 7615
    },
    {
      "entropy": 6.336485719680786,
      "epoch": 0.879399884593191,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000493401457163259,
      "loss": 6.1213,
      "mean_token_accuracy": 0.16742272078990936,
      "num_tokens": 19324342.0,
      "step": 7620
    },
    {
      "entropy": 6.300466108322143,
      "epoch": 0.8799769186381996,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004933915334123798,
      "loss": 6.1194,
      "mean_token_accuracy": 0.16596459597349167,
      "num_tokens": 19336629.0,
      "step": 7625
    },
    {
      "entropy": 6.315993547439575,
      "epoch": 0.8805539526832084,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004933816023160099,
      "loss": 6.1693,
      "mean_token_accuracy": 0.15985054075717925,
      "num_tokens": 19348247.0,
      "step": 7630
    },
    {
      "entropy": 6.3706707000732425,
      "epoch": 0.881130986728217,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004933716638744832,
      "loss": 6.0678,
      "mean_token_accuracy": 0.17529698759317397,
      "num_tokens": 19361230.0,
      "step": 7635
    },
    {
      "entropy": 6.294079065322876,
      "epoch": 0.8817080207732256,
      "grad_norm": 0.7890625,
      "learning_rate": 0.000493361718088134,
      "loss": 6.0966,
      "mean_token_accuracy": 0.177022348344326,
      "num_tokens": 19373803.0,
      "step": 7640
    },
    {
      "entropy": 6.273023843765259,
      "epoch": 0.8822850548182343,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000493351764957297,
      "loss": 6.051,
      "mean_token_accuracy": 0.17366888225078583,
      "num_tokens": 19385960.0,
      "step": 7645
    },
    {
      "entropy": 6.286736965179443,
      "epoch": 0.8828620888632429,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004933418044823068,
      "loss": 6.1875,
      "mean_token_accuracy": 0.1685008630156517,
      "num_tokens": 19399649.0,
      "step": 7650
    },
    {
      "entropy": 6.313032007217407,
      "epoch": 0.8834391229082516,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004933318366634984,
      "loss": 6.1604,
      "mean_token_accuracy": 0.1714610606431961,
      "num_tokens": 19413964.0,
      "step": 7655
    },
    {
      "entropy": 6.344218349456787,
      "epoch": 0.8840161569532602,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004933218615012072,
      "loss": 6.1798,
      "mean_token_accuracy": 0.1704179748892784,
      "num_tokens": 19425715.0,
      "step": 7660
    },
    {
      "entropy": 6.287836360931396,
      "epoch": 0.8845931909982689,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004933118789957687,
      "loss": 6.169,
      "mean_token_accuracy": 0.1677010625600815,
      "num_tokens": 19438733.0,
      "step": 7665
    },
    {
      "entropy": 6.394447135925293,
      "epoch": 0.8851702250432776,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004933018891475186,
      "loss": 6.2412,
      "mean_token_accuracy": 0.1652499720454216,
      "num_tokens": 19451924.0,
      "step": 7670
    },
    {
      "entropy": 6.310375356674195,
      "epoch": 0.8857472590882862,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004932918919567927,
      "loss": 6.0992,
      "mean_token_accuracy": 0.1740289106965065,
      "num_tokens": 19464331.0,
      "step": 7675
    },
    {
      "entropy": 6.290035152435303,
      "epoch": 0.8863242931332949,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004932818874239275,
      "loss": 6.1197,
      "mean_token_accuracy": 0.16606962829828262,
      "num_tokens": 19477127.0,
      "step": 7680
    },
    {
      "entropy": 6.310290384292602,
      "epoch": 0.8869013271783035,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004932718755492595,
      "loss": 6.089,
      "mean_token_accuracy": 0.17557645589113235,
      "num_tokens": 19490281.0,
      "step": 7685
    },
    {
      "entropy": 6.233003997802735,
      "epoch": 0.8874783612233121,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004932618563331254,
      "loss": 6.1019,
      "mean_token_accuracy": 0.17658591121435166,
      "num_tokens": 19503154.0,
      "step": 7690
    },
    {
      "entropy": 6.3082092761993405,
      "epoch": 0.8880553952683208,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004932518297758622,
      "loss": 6.2139,
      "mean_token_accuracy": 0.17067087143659593,
      "num_tokens": 19517363.0,
      "step": 7695
    },
    {
      "entropy": 6.33934588432312,
      "epoch": 0.8886324293133295,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004932417958778071,
      "loss": 6.1096,
      "mean_token_accuracy": 0.17136658877134323,
      "num_tokens": 19529785.0,
      "step": 7700
    },
    {
      "entropy": 6.30635895729065,
      "epoch": 0.8892094633583382,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004932317546392976,
      "loss": 6.173,
      "mean_token_accuracy": 0.17024747133255005,
      "num_tokens": 19542025.0,
      "step": 7705
    },
    {
      "entropy": 6.36300106048584,
      "epoch": 0.8897864974033468,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004932217060606714,
      "loss": 6.1706,
      "mean_token_accuracy": 0.16287696957588196,
      "num_tokens": 19553969.0,
      "step": 7710
    },
    {
      "entropy": 6.350517797470093,
      "epoch": 0.8903635314483554,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004932116501422665,
      "loss": 6.0277,
      "mean_token_accuracy": 0.17318399399518966,
      "num_tokens": 19566568.0,
      "step": 7715
    },
    {
      "entropy": 6.2909540176391605,
      "epoch": 0.8909405654933641,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004932015868844211,
      "loss": 6.1513,
      "mean_token_accuracy": 0.16795647740364075,
      "num_tokens": 19579947.0,
      "step": 7720
    },
    {
      "entropy": 6.366373825073242,
      "epoch": 0.8915175995383727,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004931915162874738,
      "loss": 6.1511,
      "mean_token_accuracy": 0.16321159303188323,
      "num_tokens": 19592457.0,
      "step": 7725
    },
    {
      "entropy": 6.3742717742919925,
      "epoch": 0.8920946335833814,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004931814383517632,
      "loss": 6.1153,
      "mean_token_accuracy": 0.1651516377925873,
      "num_tokens": 19604165.0,
      "step": 7730
    },
    {
      "entropy": 6.28874454498291,
      "epoch": 0.8926716676283901,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004931713530776284,
      "loss": 6.1508,
      "mean_token_accuracy": 0.16553493738174438,
      "num_tokens": 19617060.0,
      "step": 7735
    },
    {
      "entropy": 6.343902587890625,
      "epoch": 0.8932487016733988,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004931612604654083,
      "loss": 6.171,
      "mean_token_accuracy": 0.16089639961719512,
      "num_tokens": 19628724.0,
      "step": 7740
    },
    {
      "entropy": 6.39453706741333,
      "epoch": 0.8938257357184074,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004931511605154428,
      "loss": 6.1865,
      "mean_token_accuracy": 0.1709512323141098,
      "num_tokens": 19642281.0,
      "step": 7745
    },
    {
      "entropy": 6.229511404037476,
      "epoch": 0.894402769763416,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004931410532280711,
      "loss": 6.0995,
      "mean_token_accuracy": 0.17549168616533278,
      "num_tokens": 19655503.0,
      "step": 7750
    },
    {
      "entropy": 6.351932621002197,
      "epoch": 0.8949798038084247,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004931309386036337,
      "loss": 6.1062,
      "mean_token_accuracy": 0.16583985090255737,
      "num_tokens": 19667758.0,
      "step": 7755
    },
    {
      "entropy": 6.333207893371582,
      "epoch": 0.8955568378534333,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004931208166424704,
      "loss": 6.1706,
      "mean_token_accuracy": 0.17188671976327896,
      "num_tokens": 19679878.0,
      "step": 7760
    },
    {
      "entropy": 6.334847640991211,
      "epoch": 0.896133871898442,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004931106873449219,
      "loss": 6.2049,
      "mean_token_accuracy": 0.16228338330984116,
      "num_tokens": 19692278.0,
      "step": 7765
    },
    {
      "entropy": 6.2812264919281,
      "epoch": 0.8967109059434507,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004931005507113285,
      "loss": 6.0222,
      "mean_token_accuracy": 0.18351362347602845,
      "num_tokens": 19705319.0,
      "step": 7770
    },
    {
      "entropy": 6.338000631332397,
      "epoch": 0.8972879399884593,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004930904067420317,
      "loss": 6.0739,
      "mean_token_accuracy": 0.16478749513626098,
      "num_tokens": 19717908.0,
      "step": 7775
    },
    {
      "entropy": 6.409122657775879,
      "epoch": 0.897864974033468,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004930802554373723,
      "loss": 6.2259,
      "mean_token_accuracy": 0.1664321541786194,
      "num_tokens": 19730810.0,
      "step": 7780
    },
    {
      "entropy": 6.3239624977111815,
      "epoch": 0.8984420080784766,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004930700967976918,
      "loss": 6.1376,
      "mean_token_accuracy": 0.17125671803951265,
      "num_tokens": 19744176.0,
      "step": 7785
    },
    {
      "entropy": 6.298233366012573,
      "epoch": 0.8990190421234853,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000493059930823332,
      "loss": 6.0621,
      "mean_token_accuracy": 0.1712489977478981,
      "num_tokens": 19756072.0,
      "step": 7790
    },
    {
      "entropy": 6.259371900558472,
      "epoch": 0.8995960761684939,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004930497575146346,
      "loss": 6.1716,
      "mean_token_accuracy": 0.16485102325677872,
      "num_tokens": 19768170.0,
      "step": 7795
    },
    {
      "entropy": 6.293583345413208,
      "epoch": 0.9001731102135025,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004930395768719421,
      "loss": 6.1325,
      "mean_token_accuracy": 0.16966002881526948,
      "num_tokens": 19782029.0,
      "step": 7800
    },
    {
      "entropy": 6.358855533599853,
      "epoch": 0.9007501442585113,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004930293888955966,
      "loss": 6.0704,
      "mean_token_accuracy": 0.16987146884202958,
      "num_tokens": 19793973.0,
      "step": 7805
    },
    {
      "entropy": 6.232981252670288,
      "epoch": 0.9013271783035199,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000493019193585941,
      "loss": 6.062,
      "mean_token_accuracy": 0.17212969660758973,
      "num_tokens": 19806806.0,
      "step": 7810
    },
    {
      "entropy": 6.338475561141967,
      "epoch": 0.9019042123485286,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000493008990943318,
      "loss": 6.1776,
      "mean_token_accuracy": 0.16324291676282882,
      "num_tokens": 19821025.0,
      "step": 7815
    },
    {
      "entropy": 6.312364482879639,
      "epoch": 0.9024812463935372,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004929987809680709,
      "loss": 6.0055,
      "mean_token_accuracy": 0.17505942583084105,
      "num_tokens": 19832909.0,
      "step": 7820
    },
    {
      "entropy": 6.22635293006897,
      "epoch": 0.9030582804385459,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004929885636605432,
      "loss": 6.0019,
      "mean_token_accuracy": 0.1759738802909851,
      "num_tokens": 19846813.0,
      "step": 7825
    },
    {
      "entropy": 6.345417594909668,
      "epoch": 0.9036353144835545,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004929783390210784,
      "loss": 6.1491,
      "mean_token_accuracy": 0.16306580901145934,
      "num_tokens": 19860005.0,
      "step": 7830
    },
    {
      "entropy": 6.362852287292481,
      "epoch": 0.9042123485285631,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004929681070500204,
      "loss": 6.0974,
      "mean_token_accuracy": 0.16966380774974824,
      "num_tokens": 19871931.0,
      "step": 7835
    },
    {
      "entropy": 6.256046772003174,
      "epoch": 0.9047893825735719,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004929578677477135,
      "loss": 6.0799,
      "mean_token_accuracy": 0.17365067452192307,
      "num_tokens": 19884066.0,
      "step": 7840
    },
    {
      "entropy": 6.341046953201294,
      "epoch": 0.9053664166185805,
      "grad_norm": 0.875,
      "learning_rate": 0.0004929476211145019,
      "loss": 6.153,
      "mean_token_accuracy": 0.16422106325626373,
      "num_tokens": 19896416.0,
      "step": 7845
    },
    {
      "entropy": 6.42669677734375,
      "epoch": 0.9059434506635892,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004929373671507303,
      "loss": 6.1642,
      "mean_token_accuracy": 0.16458971053361893,
      "num_tokens": 19911036.0,
      "step": 7850
    },
    {
      "entropy": 6.351240587234497,
      "epoch": 0.9065204847085978,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004929271058567436,
      "loss": 6.1577,
      "mean_token_accuracy": 0.1675029769539833,
      "num_tokens": 19924516.0,
      "step": 7855
    },
    {
      "entropy": 6.252199649810791,
      "epoch": 0.9070975187536064,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000492916837232887,
      "loss": 6.0347,
      "mean_token_accuracy": 0.17800809144973756,
      "num_tokens": 19936994.0,
      "step": 7860
    },
    {
      "entropy": 6.266292953491211,
      "epoch": 0.9076745527986151,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004929065612795058,
      "loss": 6.0865,
      "mean_token_accuracy": 0.17676833420991897,
      "num_tokens": 19950304.0,
      "step": 7865
    },
    {
      "entropy": 6.254460954666138,
      "epoch": 0.9082515868436237,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004928962779969456,
      "loss": 6.1386,
      "mean_token_accuracy": 0.16655431985855101,
      "num_tokens": 19963842.0,
      "step": 7870
    },
    {
      "entropy": 6.3657478332519535,
      "epoch": 0.9088286208886325,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004928859873855524,
      "loss": 6.1584,
      "mean_token_accuracy": 0.16760774701833725,
      "num_tokens": 19975806.0,
      "step": 7875
    },
    {
      "entropy": 6.231832122802734,
      "epoch": 0.9094056549336411,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004928756894456723,
      "loss": 6.092,
      "mean_token_accuracy": 0.17502667307853698,
      "num_tokens": 19987647.0,
      "step": 7880
    },
    {
      "entropy": 6.357949447631836,
      "epoch": 0.9099826889786498,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004928653841776515,
      "loss": 6.1628,
      "mean_token_accuracy": 0.16325012296438218,
      "num_tokens": 20000591.0,
      "step": 7885
    },
    {
      "entropy": 6.406267166137695,
      "epoch": 0.9105597230236584,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004928550715818368,
      "loss": 6.0812,
      "mean_token_accuracy": 0.1728819116950035,
      "num_tokens": 20012657.0,
      "step": 7890
    },
    {
      "entropy": 6.382717943191528,
      "epoch": 0.911136757068667,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004928447516585749,
      "loss": 6.1694,
      "mean_token_accuracy": 0.16989215463399887,
      "num_tokens": 20024752.0,
      "step": 7895
    },
    {
      "entropy": 6.273079204559326,
      "epoch": 0.9117137911136757,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000492834424408213,
      "loss": 6.0692,
      "mean_token_accuracy": 0.1694467097520828,
      "num_tokens": 20036896.0,
      "step": 7900
    },
    {
      "entropy": 6.329956150054931,
      "epoch": 0.9122908251586843,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004928240898310984,
      "loss": 6.1359,
      "mean_token_accuracy": 0.16939880549907685,
      "num_tokens": 20049421.0,
      "step": 7905
    },
    {
      "entropy": 6.349690055847168,
      "epoch": 0.9128678592036931,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004928137479275789,
      "loss": 6.1758,
      "mean_token_accuracy": 0.1654338553547859,
      "num_tokens": 20062298.0,
      "step": 7910
    },
    {
      "entropy": 6.3017576217651365,
      "epoch": 0.9134448932487017,
      "grad_norm": 0.7734375,
      "learning_rate": 0.000492803398698002,
      "loss": 6.0854,
      "mean_token_accuracy": 0.1678279921412468,
      "num_tokens": 20075954.0,
      "step": 7915
    },
    {
      "entropy": 6.2847202777862545,
      "epoch": 0.9140219272937103,
      "grad_norm": 0.875,
      "learning_rate": 0.0004927930421427161,
      "loss": 6.0208,
      "mean_token_accuracy": 0.17434524595737458,
      "num_tokens": 20088035.0,
      "step": 7920
    },
    {
      "entropy": 6.294939088821411,
      "epoch": 0.914598961338719,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004927826782620694,
      "loss": 6.0776,
      "mean_token_accuracy": 0.17393822073936463,
      "num_tokens": 20100081.0,
      "step": 7925
    },
    {
      "entropy": 6.360169124603272,
      "epoch": 0.9151759953837276,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004927723070564104,
      "loss": 6.1337,
      "mean_token_accuracy": 0.17027704119682313,
      "num_tokens": 20113159.0,
      "step": 7930
    },
    {
      "entropy": 6.286882972717285,
      "epoch": 0.9157530294287363,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004927619285260881,
      "loss": 6.0953,
      "mean_token_accuracy": 0.1708056688308716,
      "num_tokens": 20125741.0,
      "step": 7935
    },
    {
      "entropy": 6.271196699142456,
      "epoch": 0.9163300634737449,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004927515426714515,
      "loss": 6.0599,
      "mean_token_accuracy": 0.17156231701374053,
      "num_tokens": 20137481.0,
      "step": 7940
    },
    {
      "entropy": 6.336545038223266,
      "epoch": 0.9169070975187537,
      "grad_norm": 0.8046875,
      "learning_rate": 0.00049274114949285,
      "loss": 6.1312,
      "mean_token_accuracy": 0.1683913141489029,
      "num_tokens": 20150188.0,
      "step": 7945
    },
    {
      "entropy": 6.371749639511108,
      "epoch": 0.9174841315637623,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000492730748990633,
      "loss": 6.1699,
      "mean_token_accuracy": 0.16632280945777894,
      "num_tokens": 20163012.0,
      "step": 7950
    },
    {
      "entropy": 6.326217842102051,
      "epoch": 0.9180611656087709,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004927203411651504,
      "loss": 6.1006,
      "mean_token_accuracy": 0.1692713424563408,
      "num_tokens": 20175468.0,
      "step": 7955
    },
    {
      "entropy": 6.294983530044556,
      "epoch": 0.9186381996537796,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004927099260167523,
      "loss": 6.1273,
      "mean_token_accuracy": 0.16911787688732147,
      "num_tokens": 20188382.0,
      "step": 7960
    },
    {
      "entropy": 6.343044328689575,
      "epoch": 0.9192152336987882,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004926995035457889,
      "loss": 6.0923,
      "mean_token_accuracy": 0.17835812717676164,
      "num_tokens": 20200091.0,
      "step": 7965
    },
    {
      "entropy": 6.284211587905884,
      "epoch": 0.9197922677437969,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000492689073752611,
      "loss": 6.1333,
      "mean_token_accuracy": 0.16809267848730086,
      "num_tokens": 20212004.0,
      "step": 7970
    },
    {
      "entropy": 6.268933963775635,
      "epoch": 0.9203693017888055,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004926786366375691,
      "loss": 6.1317,
      "mean_token_accuracy": 0.1755678579211235,
      "num_tokens": 20225214.0,
      "step": 7975
    },
    {
      "entropy": 6.2912781715393065,
      "epoch": 0.9209463358338142,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004926681922010145,
      "loss": 6.0733,
      "mean_token_accuracy": 0.17149922698736192,
      "num_tokens": 20236595.0,
      "step": 7980
    },
    {
      "entropy": 6.288493919372558,
      "epoch": 0.9215233698788229,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004926577404432982,
      "loss": 6.0899,
      "mean_token_accuracy": 0.16908173710107804,
      "num_tokens": 20248601.0,
      "step": 7985
    },
    {
      "entropy": 6.310372066497803,
      "epoch": 0.9221004039238315,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004926472813647721,
      "loss": 6.106,
      "mean_token_accuracy": 0.16341881453990936,
      "num_tokens": 20261314.0,
      "step": 7990
    },
    {
      "entropy": 6.356702184677124,
      "epoch": 0.9226774379688402,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004926368149657876,
      "loss": 6.1312,
      "mean_token_accuracy": 0.17094669938087464,
      "num_tokens": 20274797.0,
      "step": 7995
    },
    {
      "entropy": 6.321406745910645,
      "epoch": 0.9232544720138488,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004926263412466972,
      "loss": 6.1863,
      "mean_token_accuracy": 0.16469254940748215,
      "num_tokens": 20287031.0,
      "step": 8000
    },
    {
      "entropy": 6.2833436012268065,
      "epoch": 0.9238315060588574,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004926158602078527,
      "loss": 6.0814,
      "mean_token_accuracy": 0.16865895092487335,
      "num_tokens": 20298716.0,
      "step": 8005
    },
    {
      "entropy": 6.311143016815185,
      "epoch": 0.9244085401038661,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004926053718496069,
      "loss": 6.0557,
      "mean_token_accuracy": 0.18285336643457412,
      "num_tokens": 20311129.0,
      "step": 8010
    },
    {
      "entropy": 6.161644124984742,
      "epoch": 0.9249855741488748,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004925948761723126,
      "loss": 5.9655,
      "mean_token_accuracy": 0.17335004657506942,
      "num_tokens": 20323067.0,
      "step": 8015
    },
    {
      "entropy": 6.329570674896241,
      "epoch": 0.9255626081938835,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004925843731763226,
      "loss": 6.1137,
      "mean_token_accuracy": 0.1678135350346565,
      "num_tokens": 20335710.0,
      "step": 8020
    },
    {
      "entropy": 6.3780488014221195,
      "epoch": 0.9261396422388921,
      "grad_norm": 1.0,
      "learning_rate": 0.0004925738628619904,
      "loss": 6.1529,
      "mean_token_accuracy": 0.1679707333445549,
      "num_tokens": 20348436.0,
      "step": 8025
    },
    {
      "entropy": 6.2938000679016115,
      "epoch": 0.9267166762839008,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004925633452296693,
      "loss": 6.0995,
      "mean_token_accuracy": 0.16210879683494567,
      "num_tokens": 20362182.0,
      "step": 8030
    },
    {
      "entropy": 6.302653074264526,
      "epoch": 0.9272937103289094,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004925528202797131,
      "loss": 6.1194,
      "mean_token_accuracy": 0.16243739277124405,
      "num_tokens": 20374674.0,
      "step": 8035
    },
    {
      "entropy": 6.288832092285157,
      "epoch": 0.927870744373918,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004925422880124761,
      "loss": 6.0563,
      "mean_token_accuracy": 0.1735551044344902,
      "num_tokens": 20388918.0,
      "step": 8040
    },
    {
      "entropy": 6.276844120025634,
      "epoch": 0.9284477784189267,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004925317484283121,
      "loss": 6.0697,
      "mean_token_accuracy": 0.17131037414073944,
      "num_tokens": 20401428.0,
      "step": 8045
    },
    {
      "entropy": 6.325076341629028,
      "epoch": 0.9290248124639354,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004925212015275758,
      "loss": 6.1002,
      "mean_token_accuracy": 0.17513058632612227,
      "num_tokens": 20414276.0,
      "step": 8050
    },
    {
      "entropy": 6.340042304992676,
      "epoch": 0.9296018465089441,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000492510647310622,
      "loss": 6.0794,
      "mean_token_accuracy": 0.16827186644077302,
      "num_tokens": 20425515.0,
      "step": 8055
    },
    {
      "entropy": 6.31017689704895,
      "epoch": 0.9301788805539527,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004925000857778055,
      "loss": 6.1302,
      "mean_token_accuracy": 0.16977558434009551,
      "num_tokens": 20439137.0,
      "step": 8060
    },
    {
      "entropy": 6.269415760040284,
      "epoch": 0.9307559145989613,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004924895169294818,
      "loss": 6.0194,
      "mean_token_accuracy": 0.1778138369321823,
      "num_tokens": 20452059.0,
      "step": 8065
    },
    {
      "entropy": 6.357288503646851,
      "epoch": 0.93133294864397,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0004924789407660062,
      "loss": 6.192,
      "mean_token_accuracy": 0.15842437297105788,
      "num_tokens": 20465110.0,
      "step": 8070
    },
    {
      "entropy": 6.180405855178833,
      "epoch": 0.9319099826889786,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004924683572877345,
      "loss": 5.9699,
      "mean_token_accuracy": 0.17301704585552216,
      "num_tokens": 20477615.0,
      "step": 8075
    },
    {
      "entropy": 6.211012697219848,
      "epoch": 0.9324870167339873,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004924577664950225,
      "loss": 5.9691,
      "mean_token_accuracy": 0.18344345539808274,
      "num_tokens": 20490054.0,
      "step": 8080
    },
    {
      "entropy": 6.34092378616333,
      "epoch": 0.933064050778996,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004924471683882266,
      "loss": 6.1438,
      "mean_token_accuracy": 0.16662690043449402,
      "num_tokens": 20502439.0,
      "step": 8085
    },
    {
      "entropy": 6.311378908157349,
      "epoch": 0.9336410848240047,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004924365629677031,
      "loss": 6.1245,
      "mean_token_accuracy": 0.1725110724568367,
      "num_tokens": 20515094.0,
      "step": 8090
    },
    {
      "entropy": 6.306700801849365,
      "epoch": 0.9342181188690133,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000492425950233809,
      "loss": 6.04,
      "mean_token_accuracy": 0.1781902179121971,
      "num_tokens": 20527025.0,
      "step": 8095
    },
    {
      "entropy": 6.32002215385437,
      "epoch": 0.9347951529140219,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000492415330186901,
      "loss": 6.1692,
      "mean_token_accuracy": 0.16498290449380876,
      "num_tokens": 20540793.0,
      "step": 8100
    },
    {
      "entropy": 6.3359020233154295,
      "epoch": 0.9353721869590306,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004924047028273364,
      "loss": 6.1655,
      "mean_token_accuracy": 0.16523907482624053,
      "num_tokens": 20553425.0,
      "step": 8105
    },
    {
      "entropy": 6.360561370849609,
      "epoch": 0.9359492210040392,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004923940681554725,
      "loss": 6.1151,
      "mean_token_accuracy": 0.17208221107721328,
      "num_tokens": 20566794.0,
      "step": 8110
    },
    {
      "entropy": 6.330394124984741,
      "epoch": 0.9365262550490479,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004923834261716672,
      "loss": 6.1838,
      "mean_token_accuracy": 0.163845494389534,
      "num_tokens": 20579756.0,
      "step": 8115
    },
    {
      "entropy": 6.349153900146485,
      "epoch": 0.9371032890940566,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004923727768762782,
      "loss": 6.1509,
      "mean_token_accuracy": 0.16518385112285613,
      "num_tokens": 20592626.0,
      "step": 8120
    },
    {
      "entropy": 6.3262717723846436,
      "epoch": 0.9376803231390652,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000492362120269664,
      "loss": 6.0708,
      "mean_token_accuracy": 0.17791670858860015,
      "num_tokens": 20605023.0,
      "step": 8125
    },
    {
      "entropy": 6.345363521575928,
      "epoch": 0.9382573571840739,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004923514563521827,
      "loss": 6.1924,
      "mean_token_accuracy": 0.16382016837596894,
      "num_tokens": 20617668.0,
      "step": 8130
    },
    {
      "entropy": 6.29886908531189,
      "epoch": 0.9388343912290825,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004923407851241933,
      "loss": 6.0769,
      "mean_token_accuracy": 0.1696800783276558,
      "num_tokens": 20629848.0,
      "step": 8135
    },
    {
      "entropy": 6.284991836547851,
      "epoch": 0.9394114252740912,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004923301065860545,
      "loss": 6.0843,
      "mean_token_accuracy": 0.17413091361522676,
      "num_tokens": 20642695.0,
      "step": 8140
    },
    {
      "entropy": 6.334108829498291,
      "epoch": 0.9399884593190998,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004923194207381254,
      "loss": 6.1192,
      "mean_token_accuracy": 0.17119077295064927,
      "num_tokens": 20655844.0,
      "step": 8145
    },
    {
      "entropy": 6.342758464813232,
      "epoch": 0.9405654933641084,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004923087275807656,
      "loss": 6.2023,
      "mean_token_accuracy": 0.16785278022289277,
      "num_tokens": 20668380.0,
      "step": 8150
    },
    {
      "entropy": 6.322146558761597,
      "epoch": 0.9411425274091172,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004922980271143347,
      "loss": 6.0803,
      "mean_token_accuracy": 0.17897505462169647,
      "num_tokens": 20680766.0,
      "step": 8155
    },
    {
      "entropy": 6.282084703445435,
      "epoch": 0.9417195614541258,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004922873193391927,
      "loss": 6.0809,
      "mean_token_accuracy": 0.16945219039916992,
      "num_tokens": 20692538.0,
      "step": 8160
    },
    {
      "entropy": 6.333216905593872,
      "epoch": 0.9422965954991345,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004922766042556994,
      "loss": 6.0826,
      "mean_token_accuracy": 0.17635788023471832,
      "num_tokens": 20705832.0,
      "step": 8165
    },
    {
      "entropy": 6.336708974838257,
      "epoch": 0.9428736295441431,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004922658818642156,
      "loss": 6.1351,
      "mean_token_accuracy": 0.16822911351919173,
      "num_tokens": 20718741.0,
      "step": 8170
    },
    {
      "entropy": 6.332752132415772,
      "epoch": 0.9434506635891518,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004922551521651018,
      "loss": 6.1075,
      "mean_token_accuracy": 0.1693040058016777,
      "num_tokens": 20731516.0,
      "step": 8175
    },
    {
      "entropy": 6.334755182266235,
      "epoch": 0.9440276976341604,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004922444151587189,
      "loss": 6.1353,
      "mean_token_accuracy": 0.17082017064094543,
      "num_tokens": 20744835.0,
      "step": 8180
    },
    {
      "entropy": 6.292046594619751,
      "epoch": 0.944604731679169,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004922336708454279,
      "loss": 6.1189,
      "mean_token_accuracy": 0.16721362620592117,
      "num_tokens": 20758967.0,
      "step": 8185
    },
    {
      "entropy": 6.280726623535156,
      "epoch": 0.9451817657241778,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004922229192255903,
      "loss": 6.075,
      "mean_token_accuracy": 0.17387010008096696,
      "num_tokens": 20771660.0,
      "step": 8190
    },
    {
      "entropy": 6.358853816986084,
      "epoch": 0.9457587997691864,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004922121602995678,
      "loss": 6.1631,
      "mean_token_accuracy": 0.17015791386365892,
      "num_tokens": 20785366.0,
      "step": 8195
    },
    {
      "entropy": 6.381282901763916,
      "epoch": 0.9463358338141951,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004922013940677221,
      "loss": 6.1817,
      "mean_token_accuracy": 0.16708896309137344,
      "num_tokens": 20797996.0,
      "step": 8200
    },
    {
      "entropy": 6.308471012115478,
      "epoch": 0.9469128678592037,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004921906205304155,
      "loss": 6.1403,
      "mean_token_accuracy": 0.16666304767131807,
      "num_tokens": 20810921.0,
      "step": 8205
    },
    {
      "entropy": 6.313723754882813,
      "epoch": 0.9474899019042123,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004921798396880101,
      "loss": 5.995,
      "mean_token_accuracy": 0.17673814296722412,
      "num_tokens": 20823143.0,
      "step": 8210
    },
    {
      "entropy": 6.302704572677612,
      "epoch": 0.948066935949221,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004921690515408688,
      "loss": 6.1101,
      "mean_token_accuracy": 0.16627644300460814,
      "num_tokens": 20834604.0,
      "step": 8215
    },
    {
      "entropy": 6.333921718597412,
      "epoch": 0.9486439699942296,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004921582560893543,
      "loss": 6.1562,
      "mean_token_accuracy": 0.1714571610093117,
      "num_tokens": 20847224.0,
      "step": 8220
    },
    {
      "entropy": 6.32955641746521,
      "epoch": 0.9492210040392384,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004921474533338297,
      "loss": 6.14,
      "mean_token_accuracy": 0.1752069413661957,
      "num_tokens": 20859477.0,
      "step": 8225
    },
    {
      "entropy": 6.285461854934693,
      "epoch": 0.949798038084247,
      "grad_norm": 0.875,
      "learning_rate": 0.0004921366432746585,
      "loss": 6.0942,
      "mean_token_accuracy": 0.17483866810798646,
      "num_tokens": 20872144.0,
      "step": 8230
    },
    {
      "entropy": 6.339507579803467,
      "epoch": 0.9503750721292556,
      "grad_norm": 0.78515625,
      "learning_rate": 0.000492125825912204,
      "loss": 6.0751,
      "mean_token_accuracy": 0.17300888895988464,
      "num_tokens": 20885168.0,
      "step": 8235
    },
    {
      "entropy": 6.290178442001343,
      "epoch": 0.9509521061742643,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004921150012468302,
      "loss": 6.0957,
      "mean_token_accuracy": 0.1704855278134346,
      "num_tokens": 20896650.0,
      "step": 8240
    },
    {
      "entropy": 6.356605482101441,
      "epoch": 0.9515291402192729,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004921041692789013,
      "loss": 6.1853,
      "mean_token_accuracy": 0.16309679448604583,
      "num_tokens": 20908764.0,
      "step": 8245
    },
    {
      "entropy": 6.388557767868042,
      "epoch": 0.9521061742642816,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004920933300087813,
      "loss": 6.1032,
      "mean_token_accuracy": 0.16476467102766038,
      "num_tokens": 20920852.0,
      "step": 8250
    },
    {
      "entropy": 6.310107326507568,
      "epoch": 0.9526832083092902,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004920824834368353,
      "loss": 6.0673,
      "mean_token_accuracy": 0.17406057864427565,
      "num_tokens": 20933747.0,
      "step": 8255
    },
    {
      "entropy": 6.298854780197144,
      "epoch": 0.953260242354299,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004920716295634275,
      "loss": 6.1238,
      "mean_token_accuracy": 0.1746158018708229,
      "num_tokens": 20945563.0,
      "step": 8260
    },
    {
      "entropy": 6.278277587890625,
      "epoch": 0.9538372763993076,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004920607683889235,
      "loss": 6.09,
      "mean_token_accuracy": 0.16672398447990416,
      "num_tokens": 20957449.0,
      "step": 8265
    },
    {
      "entropy": 6.282410764694214,
      "epoch": 0.9544143104443162,
      "grad_norm": 0.875,
      "learning_rate": 0.0004920498999136882,
      "loss": 6.044,
      "mean_token_accuracy": 0.17534780949354173,
      "num_tokens": 20969548.0,
      "step": 8270
    },
    {
      "entropy": 6.251287317276001,
      "epoch": 0.9549913444893249,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004920390241380874,
      "loss": 6.0257,
      "mean_token_accuracy": 0.17555872350931168,
      "num_tokens": 20981712.0,
      "step": 8275
    },
    {
      "entropy": 6.305266666412353,
      "epoch": 0.9555683785343335,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004920281410624868,
      "loss": 6.1334,
      "mean_token_accuracy": 0.1694267451763153,
      "num_tokens": 20993997.0,
      "step": 8280
    },
    {
      "entropy": 6.3954089164733885,
      "epoch": 0.9561454125793422,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004920172506872525,
      "loss": 6.172,
      "mean_token_accuracy": 0.1699483409523964,
      "num_tokens": 21007158.0,
      "step": 8285
    },
    {
      "entropy": 6.303517913818359,
      "epoch": 0.9567224466243508,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004920063530127508,
      "loss": 6.0658,
      "mean_token_accuracy": 0.16866770535707473,
      "num_tokens": 21019592.0,
      "step": 8290
    },
    {
      "entropy": 6.278334474563598,
      "epoch": 0.9572994806693594,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004919954480393482,
      "loss": 6.0971,
      "mean_token_accuracy": 0.17727649658918382,
      "num_tokens": 21031850.0,
      "step": 8295
    },
    {
      "entropy": 6.324184608459473,
      "epoch": 0.9578765147143682,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004919845357674114,
      "loss": 6.0575,
      "mean_token_accuracy": 0.17361602932214737,
      "num_tokens": 21044348.0,
      "step": 8300
    },
    {
      "entropy": 6.316019916534424,
      "epoch": 0.9584535487593768,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004919736161973076,
      "loss": 6.1495,
      "mean_token_accuracy": 0.17223343402147293,
      "num_tokens": 21058017.0,
      "step": 8305
    },
    {
      "entropy": 6.355439186096191,
      "epoch": 0.9590305828043855,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000491962689329404,
      "loss": 6.0759,
      "mean_token_accuracy": 0.1674254924058914,
      "num_tokens": 21070443.0,
      "step": 8310
    },
    {
      "entropy": 6.29806694984436,
      "epoch": 0.9596076168493941,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004919517551640681,
      "loss": 6.1501,
      "mean_token_accuracy": 0.1736349418759346,
      "num_tokens": 21084033.0,
      "step": 8315
    },
    {
      "entropy": 6.336664342880249,
      "epoch": 0.9601846508944027,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004919408137016677,
      "loss": 6.0935,
      "mean_token_accuracy": 0.1669437140226364,
      "num_tokens": 21095737.0,
      "step": 8320
    },
    {
      "entropy": 6.397348928451538,
      "epoch": 0.9607616849394114,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004919298649425708,
      "loss": 6.0639,
      "mean_token_accuracy": 0.1697271764278412,
      "num_tokens": 21107769.0,
      "step": 8325
    },
    {
      "entropy": 6.216275596618653,
      "epoch": 0.96133871898442,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004919189088871456,
      "loss": 6.0161,
      "mean_token_accuracy": 0.17900092750787736,
      "num_tokens": 21119008.0,
      "step": 8330
    },
    {
      "entropy": 6.246855115890503,
      "epoch": 0.9619157530294288,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004919079455357608,
      "loss": 6.1334,
      "mean_token_accuracy": 0.16993842422962188,
      "num_tokens": 21131433.0,
      "step": 8335
    },
    {
      "entropy": 6.372685050964355,
      "epoch": 0.9624927870744374,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004918969748887847,
      "loss": 6.1928,
      "mean_token_accuracy": 0.16099000126123428,
      "num_tokens": 21144237.0,
      "step": 8340
    },
    {
      "entropy": 6.383633422851562,
      "epoch": 0.9630698211194461,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004918859969465868,
      "loss": 6.1689,
      "mean_token_accuracy": 0.16766799092292786,
      "num_tokens": 21157795.0,
      "step": 8345
    },
    {
      "entropy": 6.290271472930908,
      "epoch": 0.9636468551644547,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004918750117095361,
      "loss": 6.0985,
      "mean_token_accuracy": 0.1680177181959152,
      "num_tokens": 21169943.0,
      "step": 8350
    },
    {
      "entropy": 6.281228685379029,
      "epoch": 0.9642238892094633,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004918640191780021,
      "loss": 6.0017,
      "mean_token_accuracy": 0.17994622588157655,
      "num_tokens": 21184432.0,
      "step": 8355
    },
    {
      "entropy": 6.307895565032959,
      "epoch": 0.964800923254472,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004918530193523546,
      "loss": 6.0218,
      "mean_token_accuracy": 0.17425339967012404,
      "num_tokens": 21196539.0,
      "step": 8360
    },
    {
      "entropy": 6.293627548217773,
      "epoch": 0.9653779572994806,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004918420122329634,
      "loss": 6.1277,
      "mean_token_accuracy": 0.1632673978805542,
      "num_tokens": 21209550.0,
      "step": 8365
    },
    {
      "entropy": 6.234926080703735,
      "epoch": 0.9659549913444894,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004918309978201989,
      "loss": 5.986,
      "mean_token_accuracy": 0.18019188344478607,
      "num_tokens": 21220747.0,
      "step": 8370
    },
    {
      "entropy": 6.2483526229858395,
      "epoch": 0.966532025389498,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004918199761144315,
      "loss": 5.9756,
      "mean_token_accuracy": 0.18189812898635865,
      "num_tokens": 21233229.0,
      "step": 8375
    },
    {
      "entropy": 6.232093477249146,
      "epoch": 0.9671090594345066,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004918089471160318,
      "loss": 6.0097,
      "mean_token_accuracy": 0.17824846804141997,
      "num_tokens": 21244656.0,
      "step": 8380
    },
    {
      "entropy": 6.1882080078125,
      "epoch": 0.9676860934795153,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004917979108253709,
      "loss": 6.0632,
      "mean_token_accuracy": 0.16929151117801666,
      "num_tokens": 21257602.0,
      "step": 8385
    },
    {
      "entropy": 6.268950080871582,
      "epoch": 0.9682631275245239,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00049178686724282,
      "loss": 6.0456,
      "mean_token_accuracy": 0.1763706237077713,
      "num_tokens": 21270098.0,
      "step": 8390
    },
    {
      "entropy": 6.3219743251800535,
      "epoch": 0.9688401615695326,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004917758163687506,
      "loss": 6.1007,
      "mean_token_accuracy": 0.1630837947130203,
      "num_tokens": 21283516.0,
      "step": 8395
    },
    {
      "entropy": 6.242742538452148,
      "epoch": 0.9694171956145412,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004917647582035341,
      "loss": 6.026,
      "mean_token_accuracy": 0.17450403422117233,
      "num_tokens": 21296146.0,
      "step": 8400
    },
    {
      "entropy": 6.303759431838989,
      "epoch": 0.96999422965955,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004917536927475428,
      "loss": 6.1105,
      "mean_token_accuracy": 0.17302547097206117,
      "num_tokens": 21308537.0,
      "step": 8405
    },
    {
      "entropy": 6.3395380020141605,
      "epoch": 0.9705712637045586,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004917426200011486,
      "loss": 6.141,
      "mean_token_accuracy": 0.16933179795742034,
      "num_tokens": 21322375.0,
      "step": 8410
    },
    {
      "entropy": 6.35033950805664,
      "epoch": 0.9711482977495672,
      "grad_norm": 0.81640625,
      "learning_rate": 0.000491731539964724,
      "loss": 6.1101,
      "mean_token_accuracy": 0.17082908302545546,
      "num_tokens": 21335884.0,
      "step": 8415
    },
    {
      "entropy": 6.2293980598449705,
      "epoch": 0.9717253317945759,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004917204526386418,
      "loss": 5.9787,
      "mean_token_accuracy": 0.1721679225564003,
      "num_tokens": 21349106.0,
      "step": 8420
    },
    {
      "entropy": 6.288054132461548,
      "epoch": 0.9723023658395845,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004917093580232748,
      "loss": 6.095,
      "mean_token_accuracy": 0.1732641115784645,
      "num_tokens": 21362306.0,
      "step": 8425
    },
    {
      "entropy": 6.362192916870117,
      "epoch": 0.9728793998845932,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004916982561189962,
      "loss": 6.1051,
      "mean_token_accuracy": 0.16863009333610535,
      "num_tokens": 21375672.0,
      "step": 8430
    },
    {
      "entropy": 6.2904290676116945,
      "epoch": 0.9734564339296018,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004916871469261794,
      "loss": 6.0344,
      "mean_token_accuracy": 0.1826200544834137,
      "num_tokens": 21389174.0,
      "step": 8435
    },
    {
      "entropy": 6.269189786911011,
      "epoch": 0.9740334679746105,
      "grad_norm": 0.890625,
      "learning_rate": 0.000491676030445198,
      "loss": 6.1619,
      "mean_token_accuracy": 0.1706463173031807,
      "num_tokens": 21400027.0,
      "step": 8440
    },
    {
      "entropy": 6.293677568435669,
      "epoch": 0.9746105020196192,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004916649066764259,
      "loss": 6.1077,
      "mean_token_accuracy": 0.16729268431663513,
      "num_tokens": 21412843.0,
      "step": 8445
    },
    {
      "entropy": 6.347714185714722,
      "epoch": 0.9751875360646278,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004916537756202375,
      "loss": 6.1912,
      "mean_token_accuracy": 0.16289388090372087,
      "num_tokens": 21425253.0,
      "step": 8450
    },
    {
      "entropy": 6.295068073272705,
      "epoch": 0.9757645701096365,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004916426372770069,
      "loss": 5.997,
      "mean_token_accuracy": 0.1836947813630104,
      "num_tokens": 21436428.0,
      "step": 8455
    },
    {
      "entropy": 6.229881477355957,
      "epoch": 0.9763416041546451,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004916314916471087,
      "loss": 6.0812,
      "mean_token_accuracy": 0.17769705057144164,
      "num_tokens": 21448680.0,
      "step": 8460
    },
    {
      "entropy": 6.283309888839722,
      "epoch": 0.9769186381996537,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004916203387309179,
      "loss": 6.0943,
      "mean_token_accuracy": 0.17553262412548065,
      "num_tokens": 21460935.0,
      "step": 8465
    },
    {
      "entropy": 6.324648523330689,
      "epoch": 0.9774956722446624,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004916091785288096,
      "loss": 6.0911,
      "mean_token_accuracy": 0.17453473955392837,
      "num_tokens": 21473624.0,
      "step": 8470
    },
    {
      "entropy": 6.35996379852295,
      "epoch": 0.9780727062896711,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004915980110411593,
      "loss": 6.1947,
      "mean_token_accuracy": 0.16287709027528763,
      "num_tokens": 21486684.0,
      "step": 8475
    },
    {
      "entropy": 6.29085898399353,
      "epoch": 0.9786497403346798,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004915868362683425,
      "loss": 6.0445,
      "mean_token_accuracy": 0.18065717071294785,
      "num_tokens": 21500955.0,
      "step": 8480
    },
    {
      "entropy": 6.333687591552734,
      "epoch": 0.9792267743796884,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004915756542107349,
      "loss": 6.1112,
      "mean_token_accuracy": 0.16784068048000336,
      "num_tokens": 21512632.0,
      "step": 8485
    },
    {
      "entropy": 6.274033880233764,
      "epoch": 0.979803808424697,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004915644648687127,
      "loss": 6.0322,
      "mean_token_accuracy": 0.17534481436014177,
      "num_tokens": 21526455.0,
      "step": 8490
    },
    {
      "entropy": 6.342768144607544,
      "epoch": 0.9803808424697057,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004915532682426524,
      "loss": 6.114,
      "mean_token_accuracy": 0.16444663405418397,
      "num_tokens": 21538594.0,
      "step": 8495
    },
    {
      "entropy": 6.200225162506103,
      "epoch": 0.9809578765147143,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004915420643329306,
      "loss": 5.9968,
      "mean_token_accuracy": 0.1741709068417549,
      "num_tokens": 21551071.0,
      "step": 8500
    },
    {
      "entropy": 6.265279054641724,
      "epoch": 0.981534910559723,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000491530853139924,
      "loss": 6.0004,
      "mean_token_accuracy": 0.17739108949899673,
      "num_tokens": 21563241.0,
      "step": 8505
    },
    {
      "entropy": 6.25150089263916,
      "epoch": 0.9821119446047317,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004915196346640095,
      "loss": 6.0184,
      "mean_token_accuracy": 0.17257041484117508,
      "num_tokens": 21575634.0,
      "step": 8510
    },
    {
      "entropy": 6.281124877929687,
      "epoch": 0.9826889786497404,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004915084089055648,
      "loss": 6.093,
      "mean_token_accuracy": 0.17048918604850768,
      "num_tokens": 21587703.0,
      "step": 8515
    },
    {
      "entropy": 6.330684232711792,
      "epoch": 0.983266012694749,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004914971758649673,
      "loss": 6.1036,
      "mean_token_accuracy": 0.16494657844305038,
      "num_tokens": 21599048.0,
      "step": 8520
    },
    {
      "entropy": 6.28722071647644,
      "epoch": 0.9838430467397576,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004914859355425948,
      "loss": 5.978,
      "mean_token_accuracy": 0.17772875279188155,
      "num_tokens": 21611547.0,
      "step": 8525
    },
    {
      "entropy": 6.248496627807617,
      "epoch": 0.9844200807847663,
      "grad_norm": 0.875,
      "learning_rate": 0.0004914746879388255,
      "loss": 6.0668,
      "mean_token_accuracy": 0.18192058950662612,
      "num_tokens": 21623343.0,
      "step": 8530
    },
    {
      "entropy": 6.28045015335083,
      "epoch": 0.9849971148297749,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004914634330540373,
      "loss": 6.0795,
      "mean_token_accuracy": 0.16923675686120987,
      "num_tokens": 21635318.0,
      "step": 8535
    },
    {
      "entropy": 6.336523151397705,
      "epoch": 0.9855741488747836,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004914521708886093,
      "loss": 6.0891,
      "mean_token_accuracy": 0.16801240742206575,
      "num_tokens": 21647539.0,
      "step": 8540
    },
    {
      "entropy": 6.288561105728149,
      "epoch": 0.9861511829197923,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004914409014429201,
      "loss": 6.0868,
      "mean_token_accuracy": 0.1714344561100006,
      "num_tokens": 21660587.0,
      "step": 8545
    },
    {
      "entropy": 6.315939426422119,
      "epoch": 0.986728216964801,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004914296247173486,
      "loss": 6.0893,
      "mean_token_accuracy": 0.17230225205421448,
      "num_tokens": 21672123.0,
      "step": 8550
    },
    {
      "entropy": 6.329123783111572,
      "epoch": 0.9873052510098096,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004914183407122741,
      "loss": 6.0382,
      "mean_token_accuracy": 0.170340496301651,
      "num_tokens": 21685430.0,
      "step": 8555
    },
    {
      "entropy": 6.360370874404907,
      "epoch": 0.9878822850548182,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004914070494280763,
      "loss": 6.1429,
      "mean_token_accuracy": 0.16646113991737366,
      "num_tokens": 21698462.0,
      "step": 8560
    },
    {
      "entropy": 6.341304540634155,
      "epoch": 0.9884593190998269,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004913957508651349,
      "loss": 6.1809,
      "mean_token_accuracy": 0.1604655459523201,
      "num_tokens": 21711741.0,
      "step": 8565
    },
    {
      "entropy": 6.34149489402771,
      "epoch": 0.9890363531448355,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004913844450238299,
      "loss": 6.0853,
      "mean_token_accuracy": 0.16677425801753998,
      "num_tokens": 21723983.0,
      "step": 8570
    },
    {
      "entropy": 6.281299877166748,
      "epoch": 0.9896133871898442,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004913731319045416,
      "loss": 6.1262,
      "mean_token_accuracy": 0.17223995923995972,
      "num_tokens": 21736399.0,
      "step": 8575
    },
    {
      "entropy": 6.331571197509765,
      "epoch": 0.9901904212348529,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004913618115076505,
      "loss": 6.146,
      "mean_token_accuracy": 0.165971776843071,
      "num_tokens": 21748705.0,
      "step": 8580
    },
    {
      "entropy": 6.25674090385437,
      "epoch": 0.9907674552798615,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004913504838335372,
      "loss": 6.017,
      "mean_token_accuracy": 0.18154898434877395,
      "num_tokens": 21761659.0,
      "step": 8585
    },
    {
      "entropy": 6.277780055999756,
      "epoch": 0.9913444893248702,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004913391488825829,
      "loss": 6.0384,
      "mean_token_accuracy": 0.1767553433775902,
      "num_tokens": 21774217.0,
      "step": 8590
    },
    {
      "entropy": 6.284101390838623,
      "epoch": 0.9919215233698788,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004913278066551689,
      "loss": 6.0883,
      "mean_token_accuracy": 0.17325786352157593,
      "num_tokens": 21787417.0,
      "step": 8595
    },
    {
      "entropy": 6.308177947998047,
      "epoch": 0.9924985574148875,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004913164571516765,
      "loss": 6.1252,
      "mean_token_accuracy": 0.17321580201387404,
      "num_tokens": 21798869.0,
      "step": 8600
    },
    {
      "entropy": 6.301731824874878,
      "epoch": 0.9930755914598961,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004913051003724876,
      "loss": 6.0519,
      "mean_token_accuracy": 0.1723179042339325,
      "num_tokens": 21811622.0,
      "step": 8605
    },
    {
      "entropy": 6.314857387542725,
      "epoch": 0.9936526255049047,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004912937363179839,
      "loss": 6.0687,
      "mean_token_accuracy": 0.17025351524353027,
      "num_tokens": 21823401.0,
      "step": 8610
    },
    {
      "entropy": 6.277297592163086,
      "epoch": 0.9942296595499135,
      "grad_norm": 0.8125,
      "learning_rate": 0.000491282364988548,
      "loss": 6.064,
      "mean_token_accuracy": 0.17257231920957566,
      "num_tokens": 21835460.0,
      "step": 8615
    },
    {
      "entropy": 6.230294036865234,
      "epoch": 0.9948066935949221,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004912709863845621,
      "loss": 6.0106,
      "mean_token_accuracy": 0.17783847004175185,
      "num_tokens": 21847521.0,
      "step": 8620
    },
    {
      "entropy": 6.343156576156616,
      "epoch": 0.9953837276399308,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000491259600506409,
      "loss": 6.151,
      "mean_token_accuracy": 0.16660192161798476,
      "num_tokens": 21860861.0,
      "step": 8625
    },
    {
      "entropy": 6.298342561721801,
      "epoch": 0.9959607616849394,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004912482073544716,
      "loss": 5.988,
      "mean_token_accuracy": 0.1818713441491127,
      "num_tokens": 21872380.0,
      "step": 8630
    },
    {
      "entropy": 6.234196710586548,
      "epoch": 0.996537795729948,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004912368069291333,
      "loss": 6.085,
      "mean_token_accuracy": 0.17358871251344682,
      "num_tokens": 21885322.0,
      "step": 8635
    },
    {
      "entropy": 6.356041002273559,
      "epoch": 0.9971148297749567,
      "grad_norm": 0.875,
      "learning_rate": 0.0004912253992307773,
      "loss": 6.1464,
      "mean_token_accuracy": 0.1712245300412178,
      "num_tokens": 21898243.0,
      "step": 8640
    },
    {
      "entropy": 6.352480220794678,
      "epoch": 0.9976918638199653,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004912139842597875,
      "loss": 6.0877,
      "mean_token_accuracy": 0.16762156933546066,
      "num_tokens": 21911467.0,
      "step": 8645
    },
    {
      "entropy": 6.307034111022949,
      "epoch": 0.9982688978649741,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004912025620165477,
      "loss": 6.0945,
      "mean_token_accuracy": 0.16763445287942885,
      "num_tokens": 21925583.0,
      "step": 8650
    },
    {
      "entropy": 6.285261440277099,
      "epoch": 0.9988459319099827,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004911911325014421,
      "loss": 6.1352,
      "mean_token_accuracy": 0.16875722408294677,
      "num_tokens": 21937627.0,
      "step": 8655
    },
    {
      "entropy": 6.432320833206177,
      "epoch": 0.9994229659549914,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004911796957148552,
      "loss": 6.1302,
      "mean_token_accuracy": 0.16755087822675704,
      "num_tokens": 21950376.0,
      "step": 8660
    },
    {
      "entropy": 6.299717378616333,
      "epoch": 1.0,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004911682516571715,
      "loss": 6.0704,
      "mean_token_accuracy": 0.17948030233383178,
      "num_tokens": 21963117.0,
      "step": 8665
    },
    {
      "entropy": 6.186800336837768,
      "epoch": 1.0005770340450086,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004911568003287761,
      "loss": 5.9359,
      "mean_token_accuracy": 0.1857394814491272,
      "num_tokens": 21976256.0,
      "step": 8670
    },
    {
      "entropy": 6.2768176078796385,
      "epoch": 1.0011540680900173,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004911453417300541,
      "loss": 5.9421,
      "mean_token_accuracy": 0.1747553050518036,
      "num_tokens": 21988768.0,
      "step": 8675
    },
    {
      "entropy": 6.32557692527771,
      "epoch": 1.001731102135026,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004911338758613909,
      "loss": 5.9736,
      "mean_token_accuracy": 0.17850890904664993,
      "num_tokens": 22000570.0,
      "step": 8680
    },
    {
      "entropy": 6.319621515274048,
      "epoch": 1.0023081361800346,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004911224027231722,
      "loss": 6.0171,
      "mean_token_accuracy": 0.17065517157316207,
      "num_tokens": 22012914.0,
      "step": 8685
    },
    {
      "entropy": 6.350254249572754,
      "epoch": 1.0028851702250432,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004911109223157838,
      "loss": 6.0062,
      "mean_token_accuracy": 0.16737566739320756,
      "num_tokens": 22025474.0,
      "step": 8690
    },
    {
      "entropy": 6.246565246582032,
      "epoch": 1.0034622042700518,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004910994346396118,
      "loss": 5.873,
      "mean_token_accuracy": 0.1791422635316849,
      "num_tokens": 22039228.0,
      "step": 8695
    },
    {
      "entropy": 6.249841737747192,
      "epoch": 1.0040392383150605,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004910879396950427,
      "loss": 6.0373,
      "mean_token_accuracy": 0.17558213770389558,
      "num_tokens": 22050153.0,
      "step": 8700
    },
    {
      "entropy": 6.333817386627198,
      "epoch": 1.0046162723600693,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000491076437482463,
      "loss": 6.0313,
      "mean_token_accuracy": 0.1722121447324753,
      "num_tokens": 22063214.0,
      "step": 8705
    },
    {
      "entropy": 6.260972261428833,
      "epoch": 1.005193306405078,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004910649280022599,
      "loss": 5.9334,
      "mean_token_accuracy": 0.17726417928934096,
      "num_tokens": 22076032.0,
      "step": 8710
    },
    {
      "entropy": 6.274173450469971,
      "epoch": 1.0057703404500866,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004910534112548201,
      "loss": 5.9598,
      "mean_token_accuracy": 0.1748859703540802,
      "num_tokens": 22089134.0,
      "step": 8715
    },
    {
      "entropy": 6.288155460357666,
      "epoch": 1.0063473744950953,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004910418872405312,
      "loss": 5.9799,
      "mean_token_accuracy": 0.17057327181100845,
      "num_tokens": 22102772.0,
      "step": 8720
    },
    {
      "entropy": 6.227526712417602,
      "epoch": 1.006924408540104,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004910303559597806,
      "loss": 5.9699,
      "mean_token_accuracy": 0.1796274870634079,
      "num_tokens": 22117586.0,
      "step": 8725
    },
    {
      "entropy": 6.263715505599976,
      "epoch": 1.0075014425851125,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004910188174129564,
      "loss": 6.032,
      "mean_token_accuracy": 0.17084551006555557,
      "num_tokens": 22129229.0,
      "step": 8730
    },
    {
      "entropy": 6.3296393871307375,
      "epoch": 1.0080784766301212,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004910072716004466,
      "loss": 5.9848,
      "mean_token_accuracy": 0.17309292703866958,
      "num_tokens": 22142269.0,
      "step": 8735
    },
    {
      "entropy": 6.32510347366333,
      "epoch": 1.0086555106751298,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004909957185226394,
      "loss": 5.9946,
      "mean_token_accuracy": 0.17437432706356049,
      "num_tokens": 22153744.0,
      "step": 8740
    },
    {
      "entropy": 6.208776569366455,
      "epoch": 1.0092325447201385,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004909841581799236,
      "loss": 5.9027,
      "mean_token_accuracy": 0.17276026010513307,
      "num_tokens": 22166927.0,
      "step": 8745
    },
    {
      "entropy": 6.29905138015747,
      "epoch": 1.009809578765147,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004909725905726879,
      "loss": 5.9909,
      "mean_token_accuracy": 0.1712697610259056,
      "num_tokens": 22179003.0,
      "step": 8750
    },
    {
      "entropy": 6.315261030197144,
      "epoch": 1.0103866128101557,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004909610157013214,
      "loss": 6.0611,
      "mean_token_accuracy": 0.17488325536251068,
      "num_tokens": 22191221.0,
      "step": 8755
    },
    {
      "entropy": 6.272414875030518,
      "epoch": 1.0109636468551644,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004909494335662134,
      "loss": 6.0023,
      "mean_token_accuracy": 0.1766509383916855,
      "num_tokens": 22203387.0,
      "step": 8760
    },
    {
      "entropy": 6.285401821136475,
      "epoch": 1.011540680900173,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004909378441677535,
      "loss": 5.9543,
      "mean_token_accuracy": 0.1828874871134758,
      "num_tokens": 22215225.0,
      "step": 8765
    },
    {
      "entropy": 6.282279682159424,
      "epoch": 1.0121177149451817,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004909262475063314,
      "loss": 5.9417,
      "mean_token_accuracy": 0.18124219477176667,
      "num_tokens": 22226390.0,
      "step": 8770
    },
    {
      "entropy": 6.24841890335083,
      "epoch": 1.0126947489901905,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004909146435823373,
      "loss": 6.0019,
      "mean_token_accuracy": 0.176693457365036,
      "num_tokens": 22238884.0,
      "step": 8775
    },
    {
      "entropy": 6.337589979171753,
      "epoch": 1.0132717830351992,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004909030323961613,
      "loss": 6.1228,
      "mean_token_accuracy": 0.16325929015874863,
      "num_tokens": 22252214.0,
      "step": 8780
    },
    {
      "entropy": 6.35567684173584,
      "epoch": 1.0138488170802078,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004908914139481942,
      "loss": 5.9934,
      "mean_token_accuracy": 0.1741134375333786,
      "num_tokens": 22264814.0,
      "step": 8785
    },
    {
      "entropy": 6.22755389213562,
      "epoch": 1.0144258511252164,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004908797882388265,
      "loss": 5.9401,
      "mean_token_accuracy": 0.18325002491474152,
      "num_tokens": 22278737.0,
      "step": 8790
    },
    {
      "entropy": 6.229189825057984,
      "epoch": 1.015002885170225,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004908681552684495,
      "loss": 5.989,
      "mean_token_accuracy": 0.17604121416807175,
      "num_tokens": 22290974.0,
      "step": 8795
    },
    {
      "entropy": 6.2954566955566404,
      "epoch": 1.0155799192152337,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004908565150374542,
      "loss": 5.9761,
      "mean_token_accuracy": 0.1750330537557602,
      "num_tokens": 22303080.0,
      "step": 8800
    },
    {
      "entropy": 6.286450815200806,
      "epoch": 1.0161569532602424,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004908448675462323,
      "loss": 6.0355,
      "mean_token_accuracy": 0.1755423828959465,
      "num_tokens": 22315536.0,
      "step": 8805
    },
    {
      "entropy": 6.214225816726684,
      "epoch": 1.016733987305251,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004908332127951756,
      "loss": 5.9927,
      "mean_token_accuracy": 0.1759694129228592,
      "num_tokens": 22328114.0,
      "step": 8810
    },
    {
      "entropy": 6.273038387298584,
      "epoch": 1.0173110213502596,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004908215507846757,
      "loss": 5.9105,
      "mean_token_accuracy": 0.18296556174755096,
      "num_tokens": 22339608.0,
      "step": 8815
    },
    {
      "entropy": 6.29830846786499,
      "epoch": 1.0178880553952683,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004908098815151254,
      "loss": 5.9654,
      "mean_token_accuracy": 0.1758978173136711,
      "num_tokens": 22351594.0,
      "step": 8820
    },
    {
      "entropy": 6.278672885894776,
      "epoch": 1.018465089440277,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004907982049869168,
      "loss": 6.0408,
      "mean_token_accuracy": 0.17886385768651963,
      "num_tokens": 22363978.0,
      "step": 8825
    },
    {
      "entropy": 6.371420907974243,
      "epoch": 1.0190421234852856,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004907865212004428,
      "loss": 6.027,
      "mean_token_accuracy": 0.17178633213043212,
      "num_tokens": 22377845.0,
      "step": 8830
    },
    {
      "entropy": 6.272815847396851,
      "epoch": 1.0196191575302942,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004907748301560963,
      "loss": 5.9555,
      "mean_token_accuracy": 0.1750282183289528,
      "num_tokens": 22392654.0,
      "step": 8835
    },
    {
      "entropy": 6.312655305862426,
      "epoch": 1.0201961915753028,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004907631318542707,
      "loss": 6.0007,
      "mean_token_accuracy": 0.17550213485956193,
      "num_tokens": 22404713.0,
      "step": 8840
    },
    {
      "entropy": 6.37878384590149,
      "epoch": 1.0207732256203117,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004907514262953594,
      "loss": 6.0247,
      "mean_token_accuracy": 0.1659175708889961,
      "num_tokens": 22417879.0,
      "step": 8845
    },
    {
      "entropy": 6.241927146911621,
      "epoch": 1.0213502596653203,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004907397134797559,
      "loss": 5.9787,
      "mean_token_accuracy": 0.17025523334741594,
      "num_tokens": 22430674.0,
      "step": 8850
    },
    {
      "entropy": 6.322333812713623,
      "epoch": 1.021927293710329,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004907279934078543,
      "loss": 5.952,
      "mean_token_accuracy": 0.18116206377744676,
      "num_tokens": 22443135.0,
      "step": 8855
    },
    {
      "entropy": 6.23037805557251,
      "epoch": 1.0225043277553376,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004907162660800488,
      "loss": 5.9345,
      "mean_token_accuracy": 0.17716160118579866,
      "num_tokens": 22455832.0,
      "step": 8860
    },
    {
      "entropy": 6.340549373626709,
      "epoch": 1.0230813618003463,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004907045314967338,
      "loss": 5.992,
      "mean_token_accuracy": 0.17303016632795334,
      "num_tokens": 22467466.0,
      "step": 8865
    },
    {
      "entropy": 6.338985824584961,
      "epoch": 1.023658395845355,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004906927896583041,
      "loss": 6.0278,
      "mean_token_accuracy": 0.1670863524079323,
      "num_tokens": 22481127.0,
      "step": 8870
    },
    {
      "entropy": 6.263080072402954,
      "epoch": 1.0242354298903635,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004906810405651546,
      "loss": 6.0232,
      "mean_token_accuracy": 0.17175393402576447,
      "num_tokens": 22492894.0,
      "step": 8875
    },
    {
      "entropy": 6.383795213699341,
      "epoch": 1.0248124639353722,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004906692842176803,
      "loss": 6.0205,
      "mean_token_accuracy": 0.1698385149240494,
      "num_tokens": 22505369.0,
      "step": 8880
    },
    {
      "entropy": 6.239025115966797,
      "epoch": 1.0253894979803808,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004906575206162769,
      "loss": 5.9382,
      "mean_token_accuracy": 0.1769947186112404,
      "num_tokens": 22517163.0,
      "step": 8885
    },
    {
      "entropy": 6.221582269668579,
      "epoch": 1.0259665320253895,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004906457497613399,
      "loss": 6.017,
      "mean_token_accuracy": 0.17070773839950562,
      "num_tokens": 22531132.0,
      "step": 8890
    },
    {
      "entropy": 6.361068964004517,
      "epoch": 1.026543566070398,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004906339716532651,
      "loss": 6.0213,
      "mean_token_accuracy": 0.1671837493777275,
      "num_tokens": 22542663.0,
      "step": 8895
    },
    {
      "entropy": 6.330655860900879,
      "epoch": 1.0271206001154067,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004906221862924488,
      "loss": 6.0019,
      "mean_token_accuracy": 0.17416499108076094,
      "num_tokens": 22555397.0,
      "step": 8900
    },
    {
      "entropy": 6.191199731826782,
      "epoch": 1.0276976341604154,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004906103936792875,
      "loss": 5.9479,
      "mean_token_accuracy": 0.17581079453229903,
      "num_tokens": 22567373.0,
      "step": 8905
    },
    {
      "entropy": 6.29790563583374,
      "epoch": 1.028274668205424,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004905985938141777,
      "loss": 5.9528,
      "mean_token_accuracy": 0.1770506665110588,
      "num_tokens": 22579200.0,
      "step": 8910
    },
    {
      "entropy": 6.292901706695557,
      "epoch": 1.0288517022504329,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004905867866975163,
      "loss": 5.9836,
      "mean_token_accuracy": 0.17233404964208604,
      "num_tokens": 22592940.0,
      "step": 8915
    },
    {
      "entropy": 6.292080307006836,
      "epoch": 1.0294287362954415,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004905749723297003,
      "loss": 5.9885,
      "mean_token_accuracy": 0.17801400125026703,
      "num_tokens": 22605440.0,
      "step": 8920
    },
    {
      "entropy": 6.242578601837158,
      "epoch": 1.0300057703404502,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004905631507111271,
      "loss": 5.9847,
      "mean_token_accuracy": 0.17813263237476348,
      "num_tokens": 22618976.0,
      "step": 8925
    },
    {
      "entropy": 6.223312425613403,
      "epoch": 1.0305828043854588,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004905513218421946,
      "loss": 5.8729,
      "mean_token_accuracy": 0.18013641089200974,
      "num_tokens": 22632356.0,
      "step": 8930
    },
    {
      "entropy": 6.253937244415283,
      "epoch": 1.0311598384304674,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004905394857233004,
      "loss": 5.9736,
      "mean_token_accuracy": 0.1722057729959488,
      "num_tokens": 22645272.0,
      "step": 8935
    },
    {
      "entropy": 6.306738805770874,
      "epoch": 1.031736872475476,
      "grad_norm": 0.875,
      "learning_rate": 0.0004905276423548426,
      "loss": 5.9495,
      "mean_token_accuracy": 0.1715457633137703,
      "num_tokens": 22657101.0,
      "step": 8940
    },
    {
      "entropy": 6.301843118667603,
      "epoch": 1.0323139065204847,
      "grad_norm": 0.875,
      "learning_rate": 0.0004905157917372197,
      "loss": 5.9317,
      "mean_token_accuracy": 0.17873359471559525,
      "num_tokens": 22669107.0,
      "step": 8945
    },
    {
      "entropy": 6.304869365692139,
      "epoch": 1.0328909405654934,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004905039338708302,
      "loss": 5.9678,
      "mean_token_accuracy": 0.17514553815126419,
      "num_tokens": 22681633.0,
      "step": 8950
    },
    {
      "entropy": 6.355607652664185,
      "epoch": 1.033467974610502,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004904920687560728,
      "loss": 6.0256,
      "mean_token_accuracy": 0.17806240171194077,
      "num_tokens": 22694985.0,
      "step": 8955
    },
    {
      "entropy": 6.259906005859375,
      "epoch": 1.0340450086555106,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004904801963933469,
      "loss": 5.9828,
      "mean_token_accuracy": 0.174012191593647,
      "num_tokens": 22707333.0,
      "step": 8960
    },
    {
      "entropy": 6.276315689086914,
      "epoch": 1.0346220427005193,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004904683167830515,
      "loss": 5.9586,
      "mean_token_accuracy": 0.17578708678483962,
      "num_tokens": 22721030.0,
      "step": 8965
    },
    {
      "entropy": 6.2770734310150145,
      "epoch": 1.035199076745528,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004904564299255862,
      "loss": 5.9283,
      "mean_token_accuracy": 0.17963338494300843,
      "num_tokens": 22732711.0,
      "step": 8970
    },
    {
      "entropy": 6.2471513748168945,
      "epoch": 1.0357761107905366,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004904445358213511,
      "loss": 5.9418,
      "mean_token_accuracy": 0.18144085705280305,
      "num_tokens": 22745815.0,
      "step": 8975
    },
    {
      "entropy": 6.251951503753662,
      "epoch": 1.0363531448355452,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004904326344707461,
      "loss": 5.9865,
      "mean_token_accuracy": 0.1752885952591896,
      "num_tokens": 22757590.0,
      "step": 8980
    },
    {
      "entropy": 6.382661581039429,
      "epoch": 1.036930178880554,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004904207258741712,
      "loss": 6.044,
      "mean_token_accuracy": 0.16637400537729263,
      "num_tokens": 22769841.0,
      "step": 8985
    },
    {
      "entropy": 6.220981550216675,
      "epoch": 1.0375072129255627,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004904088100320274,
      "loss": 5.9489,
      "mean_token_accuracy": 0.18653181344270706,
      "num_tokens": 22781642.0,
      "step": 8990
    },
    {
      "entropy": 6.30796365737915,
      "epoch": 1.0380842469705713,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004903968869447151,
      "loss": 5.9683,
      "mean_token_accuracy": 0.18010073900222778,
      "num_tokens": 22794903.0,
      "step": 8995
    },
    {
      "entropy": 6.284050559997558,
      "epoch": 1.03866128101558,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004903849566126356,
      "loss": 6.0023,
      "mean_token_accuracy": 0.18212944269180298,
      "num_tokens": 22807839.0,
      "step": 9000
    },
    {
      "epoch": 1.03866128101558,
      "eval_entropy": 6.075341097021225,
      "eval_loss": 6.107968330383301,
      "eval_mean_token_accuracy": 0.17483100661346443,
      "eval_num_tokens": 22807839.0,
      "eval_runtime": 17.5957,
      "eval_samples_per_second": 1599.026,
      "eval_steps_per_second": 199.878,
      "step": 9000
    },
    {
      "entropy": 6.28920087814331,
      "epoch": 1.0392383150605886,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004903730190361902,
      "loss": 6.049,
      "mean_token_accuracy": 0.16872194558382034,
      "num_tokens": 22821037.0,
      "step": 9005
    },
    {
      "entropy": 6.310537385940552,
      "epoch": 1.0398153491055973,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00049036107421578,
      "loss": 6.0242,
      "mean_token_accuracy": 0.17456025928258895,
      "num_tokens": 22834225.0,
      "step": 9010
    },
    {
      "entropy": 6.295633745193482,
      "epoch": 1.040392383150606,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004903491221518073,
      "loss": 5.9698,
      "mean_token_accuracy": 0.18304451256990434,
      "num_tokens": 22847355.0,
      "step": 9015
    },
    {
      "entropy": 6.235943555831909,
      "epoch": 1.0409694171956145,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004903371628446737,
      "loss": 5.9699,
      "mean_token_accuracy": 0.17832315266132354,
      "num_tokens": 22859550.0,
      "step": 9020
    },
    {
      "entropy": 6.30399260520935,
      "epoch": 1.0415464512406232,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004903251962947817,
      "loss": 6.0482,
      "mean_token_accuracy": 0.16920068114995956,
      "num_tokens": 22871690.0,
      "step": 9025
    },
    {
      "entropy": 6.312657642364502,
      "epoch": 1.0421234852856318,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004903132225025335,
      "loss": 6.0506,
      "mean_token_accuracy": 0.17511861771345139,
      "num_tokens": 22883838.0,
      "step": 9030
    },
    {
      "entropy": 6.222030925750732,
      "epoch": 1.0427005193306405,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004903012414683322,
      "loss": 5.9589,
      "mean_token_accuracy": 0.17838482707738876,
      "num_tokens": 22896566.0,
      "step": 9035
    },
    {
      "entropy": 6.314267110824585,
      "epoch": 1.043277553375649,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004902892531925805,
      "loss": 5.9426,
      "mean_token_accuracy": 0.17883611619472503,
      "num_tokens": 22908480.0,
      "step": 9040
    },
    {
      "entropy": 6.292688226699829,
      "epoch": 1.0438545874206577,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004902772576756818,
      "loss": 6.0192,
      "mean_token_accuracy": 0.17431949079036713,
      "num_tokens": 22920923.0,
      "step": 9045
    },
    {
      "entropy": 6.328610754013061,
      "epoch": 1.0444316214656664,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004902652549180394,
      "loss": 6.0834,
      "mean_token_accuracy": 0.16915244311094285,
      "num_tokens": 22934991.0,
      "step": 9050
    },
    {
      "entropy": 6.247945976257324,
      "epoch": 1.0450086555106752,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004902532449200571,
      "loss": 5.9579,
      "mean_token_accuracy": 0.17632273137569426,
      "num_tokens": 22947223.0,
      "step": 9055
    },
    {
      "entropy": 6.269422769546509,
      "epoch": 1.0455856895556839,
      "grad_norm": 0.875,
      "learning_rate": 0.0004902412276821386,
      "loss": 6.0394,
      "mean_token_accuracy": 0.17386496663093567,
      "num_tokens": 22960320.0,
      "step": 9060
    },
    {
      "entropy": 6.267181730270385,
      "epoch": 1.0461627236006925,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004902292032046887,
      "loss": 5.9633,
      "mean_token_accuracy": 0.17790684700012208,
      "num_tokens": 22972373.0,
      "step": 9065
    },
    {
      "entropy": 6.352125930786133,
      "epoch": 1.0467397576457012,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004902171714881112,
      "loss": 6.0655,
      "mean_token_accuracy": 0.17117210924625398,
      "num_tokens": 22984936.0,
      "step": 9070
    },
    {
      "entropy": 6.308430576324463,
      "epoch": 1.0473167916907098,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004902051325328112,
      "loss": 5.9754,
      "mean_token_accuracy": 0.1752873942255974,
      "num_tokens": 22997027.0,
      "step": 9075
    },
    {
      "entropy": 6.249284076690674,
      "epoch": 1.0478938257357184,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004901930863391934,
      "loss": 5.9502,
      "mean_token_accuracy": 0.1768185690045357,
      "num_tokens": 23010861.0,
      "step": 9080
    },
    {
      "entropy": 6.295260381698609,
      "epoch": 1.048470859780727,
      "grad_norm": 0.84375,
      "learning_rate": 0.000490181032907663,
      "loss": 6.0785,
      "mean_token_accuracy": 0.1666529580950737,
      "num_tokens": 23023654.0,
      "step": 9085
    },
    {
      "entropy": 6.370468091964722,
      "epoch": 1.0490478938257357,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004901689722386255,
      "loss": 6.0577,
      "mean_token_accuracy": 0.1690685346722603,
      "num_tokens": 23036401.0,
      "step": 9090
    },
    {
      "entropy": 6.314890003204345,
      "epoch": 1.0496249278707444,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004901569043324864,
      "loss": 5.9686,
      "mean_token_accuracy": 0.1705662950873375,
      "num_tokens": 23048891.0,
      "step": 9095
    },
    {
      "entropy": 6.2467084407806395,
      "epoch": 1.050201961915753,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004901448291896518,
      "loss": 6.0284,
      "mean_token_accuracy": 0.17279767990112305,
      "num_tokens": 23062170.0,
      "step": 9100
    },
    {
      "entropy": 6.289740467071534,
      "epoch": 1.0507789959607616,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004901327468105277,
      "loss": 5.9663,
      "mean_token_accuracy": 0.1743663102388382,
      "num_tokens": 23075512.0,
      "step": 9105
    },
    {
      "entropy": 6.288102149963379,
      "epoch": 1.0513560300057703,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004901206571955205,
      "loss": 6.0543,
      "mean_token_accuracy": 0.16738157719373703,
      "num_tokens": 23086644.0,
      "step": 9110
    },
    {
      "entropy": 6.282284212112427,
      "epoch": 1.051933064050779,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004901085603450369,
      "loss": 5.8993,
      "mean_token_accuracy": 0.18418894112110137,
      "num_tokens": 23098596.0,
      "step": 9115
    },
    {
      "entropy": 6.259978818893432,
      "epoch": 1.0525100980957875,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004900964562594838,
      "loss": 5.9681,
      "mean_token_accuracy": 0.17490415424108505,
      "num_tokens": 23112746.0,
      "step": 9120
    },
    {
      "entropy": 6.273745822906494,
      "epoch": 1.0530871321407964,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000490084344939268,
      "loss": 5.9155,
      "mean_token_accuracy": 0.17623485028743743,
      "num_tokens": 23124297.0,
      "step": 9125
    },
    {
      "entropy": 6.273250675201416,
      "epoch": 1.053664166185805,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004900722263847973,
      "loss": 5.9252,
      "mean_token_accuracy": 0.17505841702222824,
      "num_tokens": 23135702.0,
      "step": 9130
    },
    {
      "entropy": 6.201640510559082,
      "epoch": 1.0542412002308137,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004900601005964791,
      "loss": 5.9241,
      "mean_token_accuracy": 0.17806514501571655,
      "num_tokens": 23149055.0,
      "step": 9135
    },
    {
      "entropy": 6.224706172943115,
      "epoch": 1.0548182342758223,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004900479675747212,
      "loss": 5.9314,
      "mean_token_accuracy": 0.1843092367053032,
      "num_tokens": 23162432.0,
      "step": 9140
    },
    {
      "entropy": 6.285671424865723,
      "epoch": 1.055395268320831,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004900358273199316,
      "loss": 6.0063,
      "mean_token_accuracy": 0.16955972015857695,
      "num_tokens": 23174187.0,
      "step": 9145
    },
    {
      "entropy": 6.261640024185181,
      "epoch": 1.0559723023658396,
      "grad_norm": 0.9375,
      "learning_rate": 0.000490023679832519,
      "loss": 5.9501,
      "mean_token_accuracy": 0.1814824089407921,
      "num_tokens": 23186584.0,
      "step": 9150
    },
    {
      "entropy": 6.1945452213287355,
      "epoch": 1.0565493364108482,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004900115251128916,
      "loss": 5.9121,
      "mean_token_accuracy": 0.18116955906152726,
      "num_tokens": 23197804.0,
      "step": 9155
    },
    {
      "entropy": 6.276153612136841,
      "epoch": 1.057126370455857,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004899993631614583,
      "loss": 5.9533,
      "mean_token_accuracy": 0.18248214274644853,
      "num_tokens": 23210262.0,
      "step": 9160
    },
    {
      "entropy": 6.257278537750244,
      "epoch": 1.0577034045008655,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004899871939786283,
      "loss": 5.9737,
      "mean_token_accuracy": 0.17818736732006074,
      "num_tokens": 23221922.0,
      "step": 9165
    },
    {
      "entropy": 6.192862701416016,
      "epoch": 1.0582804385458742,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004899750175648107,
      "loss": 5.953,
      "mean_token_accuracy": 0.18110017329454423,
      "num_tokens": 23232983.0,
      "step": 9170
    },
    {
      "entropy": 6.2926247119903564,
      "epoch": 1.0588574725908828,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004899628339204154,
      "loss": 6.0544,
      "mean_token_accuracy": 0.16700772047042847,
      "num_tokens": 23245692.0,
      "step": 9175
    },
    {
      "entropy": 6.373706245422364,
      "epoch": 1.0594345066358914,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004899506430458518,
      "loss": 5.914,
      "mean_token_accuracy": 0.18452391624450684,
      "num_tokens": 23256963.0,
      "step": 9180
    },
    {
      "entropy": 6.186127376556397,
      "epoch": 1.0600115406809,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004899384449415302,
      "loss": 5.9433,
      "mean_token_accuracy": 0.1747090682387352,
      "num_tokens": 23270226.0,
      "step": 9185
    },
    {
      "entropy": 6.316973161697388,
      "epoch": 1.0605885747259087,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004899262396078606,
      "loss": 6.0648,
      "mean_token_accuracy": 0.16537218689918518,
      "num_tokens": 23283666.0,
      "step": 9190
    },
    {
      "entropy": 6.315218734741211,
      "epoch": 1.0611656087709176,
      "grad_norm": 0.75,
      "learning_rate": 0.0004899140270452539,
      "loss": 5.994,
      "mean_token_accuracy": 0.17470985352993013,
      "num_tokens": 23295499.0,
      "step": 9195
    },
    {
      "entropy": 6.247121143341064,
      "epoch": 1.0617426428159262,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004899018072541204,
      "loss": 5.9745,
      "mean_token_accuracy": 0.17117512077093125,
      "num_tokens": 23308876.0,
      "step": 9200
    },
    {
      "entropy": 6.287448930740356,
      "epoch": 1.0623196768609349,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004898895802348715,
      "loss": 5.9946,
      "mean_token_accuracy": 0.17492063492536544,
      "num_tokens": 23321678.0,
      "step": 9205
    },
    {
      "entropy": 6.252198600769043,
      "epoch": 1.0628967109059435,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004898773459879183,
      "loss": 5.9443,
      "mean_token_accuracy": 0.18303306698799132,
      "num_tokens": 23334775.0,
      "step": 9210
    },
    {
      "entropy": 6.242930364608765,
      "epoch": 1.0634737449509521,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004898651045136724,
      "loss": 5.8649,
      "mean_token_accuracy": 0.18231088668107986,
      "num_tokens": 23347874.0,
      "step": 9215
    },
    {
      "entropy": 6.271909999847412,
      "epoch": 1.0640507789959608,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004898528558125453,
      "loss": 5.9815,
      "mean_token_accuracy": 0.1719153955578804,
      "num_tokens": 23360150.0,
      "step": 9220
    },
    {
      "entropy": 6.289255952835083,
      "epoch": 1.0646278130409694,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004898405998849492,
      "loss": 6.0067,
      "mean_token_accuracy": 0.1767814964056015,
      "num_tokens": 23373324.0,
      "step": 9225
    },
    {
      "entropy": 6.303161525726319,
      "epoch": 1.065204847085978,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004898283367312962,
      "loss": 6.0703,
      "mean_token_accuracy": 0.17053993195295333,
      "num_tokens": 23385576.0,
      "step": 9230
    },
    {
      "entropy": 6.293999862670899,
      "epoch": 1.0657818811309867,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004898160663519988,
      "loss": 5.9615,
      "mean_token_accuracy": 0.17673758119344712,
      "num_tokens": 23396793.0,
      "step": 9235
    },
    {
      "entropy": 6.301612710952758,
      "epoch": 1.0663589151759953,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004898037887474697,
      "loss": 6.0072,
      "mean_token_accuracy": 0.17729080319404603,
      "num_tokens": 23410034.0,
      "step": 9240
    },
    {
      "entropy": 6.265505599975586,
      "epoch": 1.066935949221004,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0004897915039181219,
      "loss": 5.9466,
      "mean_token_accuracy": 0.1732722267508507,
      "num_tokens": 23423857.0,
      "step": 9245
    },
    {
      "entropy": 6.289142179489136,
      "epoch": 1.0675129832660126,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004897792118643685,
      "loss": 6.0145,
      "mean_token_accuracy": 0.17617835998535156,
      "num_tokens": 23437075.0,
      "step": 9250
    },
    {
      "entropy": 6.2874561786651615,
      "epoch": 1.0680900173110213,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004897669125866231,
      "loss": 6.0174,
      "mean_token_accuracy": 0.16811560839414597,
      "num_tokens": 23450635.0,
      "step": 9255
    },
    {
      "entropy": 6.287706756591797,
      "epoch": 1.06866705135603,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004897546060852993,
      "loss": 5.9393,
      "mean_token_accuracy": 0.1778903305530548,
      "num_tokens": 23464341.0,
      "step": 9260
    },
    {
      "entropy": 6.318810749053955,
      "epoch": 1.0692440854010385,
      "grad_norm": 0.875,
      "learning_rate": 0.0004897422923608108,
      "loss": 6.0401,
      "mean_token_accuracy": 0.16921036690473557,
      "num_tokens": 23477462.0,
      "step": 9265
    },
    {
      "entropy": 6.309871530532837,
      "epoch": 1.0698211194460474,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004897299714135721,
      "loss": 6.0843,
      "mean_token_accuracy": 0.16747737675905228,
      "num_tokens": 23489990.0,
      "step": 9270
    },
    {
      "entropy": 6.295157384872437,
      "epoch": 1.070398153491056,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004897176432439974,
      "loss": 6.0181,
      "mean_token_accuracy": 0.17354026287794114,
      "num_tokens": 23502503.0,
      "step": 9275
    },
    {
      "entropy": 6.2775756359100345,
      "epoch": 1.0709751875360647,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004897053078525016,
      "loss": 5.9364,
      "mean_token_accuracy": 0.17987664341926574,
      "num_tokens": 23514903.0,
      "step": 9280
    },
    {
      "entropy": 6.298737668991089,
      "epoch": 1.0715522215810733,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004896929652394993,
      "loss": 5.9544,
      "mean_token_accuracy": 0.17657140791416168,
      "num_tokens": 23528563.0,
      "step": 9285
    },
    {
      "entropy": 6.278679037094117,
      "epoch": 1.072129255626082,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004896806154054057,
      "loss": 6.0616,
      "mean_token_accuracy": 0.1733594834804535,
      "num_tokens": 23541631.0,
      "step": 9290
    },
    {
      "entropy": 6.262742233276367,
      "epoch": 1.0727062896710906,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004896682583506363,
      "loss": 5.9021,
      "mean_token_accuracy": 0.18060626536607743,
      "num_tokens": 23553665.0,
      "step": 9295
    },
    {
      "entropy": 6.207019948959351,
      "epoch": 1.0732833237160992,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004896558940756067,
      "loss": 5.9404,
      "mean_token_accuracy": 0.1758260816335678,
      "num_tokens": 23567050.0,
      "step": 9300
    },
    {
      "entropy": 6.202441310882568,
      "epoch": 1.0738603577611079,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004896435225807327,
      "loss": 5.9671,
      "mean_token_accuracy": 0.1756475493311882,
      "num_tokens": 23580676.0,
      "step": 9305
    },
    {
      "entropy": 6.32417106628418,
      "epoch": 1.0744373918061165,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004896311438664304,
      "loss": 5.9776,
      "mean_token_accuracy": 0.17709142565727234,
      "num_tokens": 23592705.0,
      "step": 9310
    },
    {
      "entropy": 6.2693194389343265,
      "epoch": 1.0750144258511252,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004896187579331163,
      "loss": 6.0054,
      "mean_token_accuracy": 0.17715303897857665,
      "num_tokens": 23604365.0,
      "step": 9315
    },
    {
      "entropy": 6.246262836456299,
      "epoch": 1.0755914598961338,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004896063647812068,
      "loss": 6.0091,
      "mean_token_accuracy": 0.17169079631567002,
      "num_tokens": 23617657.0,
      "step": 9320
    },
    {
      "entropy": 6.27946252822876,
      "epoch": 1.0761684939411424,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004895939644111189,
      "loss": 5.9922,
      "mean_token_accuracy": 0.17139546871185302,
      "num_tokens": 23630215.0,
      "step": 9325
    },
    {
      "entropy": 6.311364984512329,
      "epoch": 1.076745527986151,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004895815568232694,
      "loss": 5.9789,
      "mean_token_accuracy": 0.17737127393484114,
      "num_tokens": 23643126.0,
      "step": 9330
    },
    {
      "entropy": 6.167384099960327,
      "epoch": 1.07732256203116,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004895691420180759,
      "loss": 5.9441,
      "mean_token_accuracy": 0.17400893270969392,
      "num_tokens": 23656156.0,
      "step": 9335
    },
    {
      "entropy": 6.2585046768188475,
      "epoch": 1.0778995960761686,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000489556719995956,
      "loss": 5.9616,
      "mean_token_accuracy": 0.1759215086698532,
      "num_tokens": 23668930.0,
      "step": 9340
    },
    {
      "entropy": 6.300254535675049,
      "epoch": 1.0784766301211772,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004895442907573274,
      "loss": 6.0009,
      "mean_token_accuracy": 0.17157413065433502,
      "num_tokens": 23681386.0,
      "step": 9345
    },
    {
      "entropy": 6.260494422912598,
      "epoch": 1.0790536641661859,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000489531854302608,
      "loss": 6.0394,
      "mean_token_accuracy": 0.16908756345510484,
      "num_tokens": 23695381.0,
      "step": 9350
    },
    {
      "entropy": 6.26428599357605,
      "epoch": 1.0796306982111945,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004895194106322164,
      "loss": 5.995,
      "mean_token_accuracy": 0.17683071345090867,
      "num_tokens": 23707259.0,
      "step": 9355
    },
    {
      "entropy": 6.371436262130738,
      "epoch": 1.0802077322562031,
      "grad_norm": 0.7421875,
      "learning_rate": 0.0004895069597465709,
      "loss": 6.026,
      "mean_token_accuracy": 0.17027197778224945,
      "num_tokens": 23720933.0,
      "step": 9360
    },
    {
      "entropy": 6.227053165435791,
      "epoch": 1.0807847663012118,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004894945016460904,
      "loss": 5.9677,
      "mean_token_accuracy": 0.17564187794923783,
      "num_tokens": 23733735.0,
      "step": 9365
    },
    {
      "entropy": 6.261332654953003,
      "epoch": 1.0813618003462204,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004894820363311938,
      "loss": 6.0141,
      "mean_token_accuracy": 0.173288731276989,
      "num_tokens": 23746231.0,
      "step": 9370
    },
    {
      "entropy": 6.316046094894409,
      "epoch": 1.081938834391229,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004894695638023005,
      "loss": 5.9359,
      "mean_token_accuracy": 0.1791118174791336,
      "num_tokens": 23758817.0,
      "step": 9375
    },
    {
      "entropy": 6.258134508132935,
      "epoch": 1.0825158684362377,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048945708405983,
      "loss": 5.9631,
      "mean_token_accuracy": 0.18306457698345185,
      "num_tokens": 23771983.0,
      "step": 9380
    },
    {
      "entropy": 6.270721912384033,
      "epoch": 1.0830929024812463,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004894445971042019,
      "loss": 6.0241,
      "mean_token_accuracy": 0.17005717158317565,
      "num_tokens": 23784112.0,
      "step": 9385
    },
    {
      "entropy": 6.339037942886352,
      "epoch": 1.083669936526255,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004894321029358364,
      "loss": 6.0191,
      "mean_token_accuracy": 0.16635214984416963,
      "num_tokens": 23796854.0,
      "step": 9390
    },
    {
      "entropy": 6.3159153938293455,
      "epoch": 1.0842469705712636,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004894196015551536,
      "loss": 6.0055,
      "mean_token_accuracy": 0.17898474484682084,
      "num_tokens": 23809450.0,
      "step": 9395
    },
    {
      "entropy": 6.296340751647949,
      "epoch": 1.0848240046162723,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000489407092962574,
      "loss": 6.0673,
      "mean_token_accuracy": 0.1673806294798851,
      "num_tokens": 23821742.0,
      "step": 9400
    },
    {
      "entropy": 6.257652044296265,
      "epoch": 1.085401038661281,
      "grad_norm": 0.875,
      "learning_rate": 0.0004893945771585183,
      "loss": 5.9193,
      "mean_token_accuracy": 0.17831383794546127,
      "num_tokens": 23834174.0,
      "step": 9405
    },
    {
      "entropy": 6.245878791809082,
      "epoch": 1.0859780727062898,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004893820541434075,
      "loss": 5.9412,
      "mean_token_accuracy": 0.18000270128250123,
      "num_tokens": 23847076.0,
      "step": 9410
    },
    {
      "entropy": 6.303604698181152,
      "epoch": 1.0865551067512984,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004893695239176629,
      "loss": 5.9605,
      "mean_token_accuracy": 0.17541009038686753,
      "num_tokens": 23860682.0,
      "step": 9415
    },
    {
      "entropy": 6.1643232822418215,
      "epoch": 1.087132140796307,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004893569864817057,
      "loss": 5.8431,
      "mean_token_accuracy": 0.18151384443044663,
      "num_tokens": 23873631.0,
      "step": 9420
    },
    {
      "entropy": 6.233168268203736,
      "epoch": 1.0877091748413157,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004893444418359579,
      "loss": 5.9243,
      "mean_token_accuracy": 0.17663467526435853,
      "num_tokens": 23885144.0,
      "step": 9425
    },
    {
      "entropy": 6.256800270080566,
      "epoch": 1.0882862088863243,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000489331889980841,
      "loss": 5.9955,
      "mean_token_accuracy": 0.17681868523359298,
      "num_tokens": 23898313.0,
      "step": 9430
    },
    {
      "entropy": 6.252180290222168,
      "epoch": 1.088863242931333,
      "grad_norm": 0.875,
      "learning_rate": 0.0004893193309167778,
      "loss": 5.9588,
      "mean_token_accuracy": 0.17652200907468796,
      "num_tokens": 23910002.0,
      "step": 9435
    },
    {
      "entropy": 6.266542387008667,
      "epoch": 1.0894402769763416,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004893067646441902,
      "loss": 6.0439,
      "mean_token_accuracy": 0.1703325092792511,
      "num_tokens": 23922770.0,
      "step": 9440
    },
    {
      "entropy": 6.327376222610473,
      "epoch": 1.0900173110213502,
      "grad_norm": 0.78515625,
      "learning_rate": 0.000489294191163501,
      "loss": 6.0293,
      "mean_token_accuracy": 0.1738341674208641,
      "num_tokens": 23936218.0,
      "step": 9445
    },
    {
      "entropy": 6.3197380065917965,
      "epoch": 1.0905943450663589,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004892816104751331,
      "loss": 6.0268,
      "mean_token_accuracy": 0.17544559985399247,
      "num_tokens": 23949250.0,
      "step": 9450
    },
    {
      "entropy": 6.346788835525513,
      "epoch": 1.0911713791113675,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004892690225795096,
      "loss": 5.9571,
      "mean_token_accuracy": 0.1708831012248993,
      "num_tokens": 23961112.0,
      "step": 9455
    },
    {
      "entropy": 6.313372278213501,
      "epoch": 1.0917484131563762,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004892564274770542,
      "loss": 6.0184,
      "mean_token_accuracy": 0.16648160070180892,
      "num_tokens": 23974450.0,
      "step": 9460
    },
    {
      "entropy": 6.311605501174927,
      "epoch": 1.0923254472013848,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004892438251681901,
      "loss": 5.9486,
      "mean_token_accuracy": 0.17587699443101884,
      "num_tokens": 23986368.0,
      "step": 9465
    },
    {
      "entropy": 6.303554916381836,
      "epoch": 1.0929024812463934,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004892312156533413,
      "loss": 6.0224,
      "mean_token_accuracy": 0.17484019100666046,
      "num_tokens": 23998186.0,
      "step": 9470
    },
    {
      "entropy": 6.253671360015869,
      "epoch": 1.0934795152914023,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004892185989329321,
      "loss": 6.0343,
      "mean_token_accuracy": 0.1720895916223526,
      "num_tokens": 24011909.0,
      "step": 9475
    },
    {
      "entropy": 6.279291820526123,
      "epoch": 1.094056549336411,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004892059750073868,
      "loss": 6.0182,
      "mean_token_accuracy": 0.1696061298251152,
      "num_tokens": 24024142.0,
      "step": 9480
    },
    {
      "entropy": 6.192721796035767,
      "epoch": 1.0946335833814196,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004891933438771299,
      "loss": 5.9322,
      "mean_token_accuracy": 0.17272215485572814,
      "num_tokens": 24037579.0,
      "step": 9485
    },
    {
      "entropy": 6.2890088081359865,
      "epoch": 1.0952106174264282,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004891807055425862,
      "loss": 5.9249,
      "mean_token_accuracy": 0.1830565094947815,
      "num_tokens": 24050707.0,
      "step": 9490
    },
    {
      "entropy": 6.295971488952636,
      "epoch": 1.0957876514714369,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004891680600041809,
      "loss": 6.0002,
      "mean_token_accuracy": 0.17599813193082808,
      "num_tokens": 24064148.0,
      "step": 9495
    },
    {
      "entropy": 6.1965096950531,
      "epoch": 1.0963646855164455,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004891554072623392,
      "loss": 5.9358,
      "mean_token_accuracy": 0.18336652219295502,
      "num_tokens": 24077769.0,
      "step": 9500
    },
    {
      "entropy": 6.3309009075164795,
      "epoch": 1.0969417195614541,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004891427473174869,
      "loss": 6.0218,
      "mean_token_accuracy": 0.17425058037042618,
      "num_tokens": 24089364.0,
      "step": 9505
    },
    {
      "entropy": 6.258981990814209,
      "epoch": 1.0975187536064628,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004891300801700496,
      "loss": 5.9553,
      "mean_token_accuracy": 0.1831838980317116,
      "num_tokens": 24102890.0,
      "step": 9510
    },
    {
      "entropy": 6.143836307525635,
      "epoch": 1.0980957876514714,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004891174058204534,
      "loss": 5.9647,
      "mean_token_accuracy": 0.17503372728824615,
      "num_tokens": 24116290.0,
      "step": 9515
    },
    {
      "entropy": 6.288706731796265,
      "epoch": 1.09867282169648,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004891047242691246,
      "loss": 5.9269,
      "mean_token_accuracy": 0.17750721275806428,
      "num_tokens": 24128318.0,
      "step": 9520
    },
    {
      "entropy": 6.221569299697876,
      "epoch": 1.0992498557414887,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004890920355164897,
      "loss": 5.9114,
      "mean_token_accuracy": 0.18555276691913605,
      "num_tokens": 24140947.0,
      "step": 9525
    },
    {
      "entropy": 6.231468296051025,
      "epoch": 1.0998268897864973,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004890793395629756,
      "loss": 5.9553,
      "mean_token_accuracy": 0.1733102098107338,
      "num_tokens": 24153254.0,
      "step": 9530
    },
    {
      "entropy": 6.391895532608032,
      "epoch": 1.100403923831506,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004890666364090093,
      "loss": 5.9852,
      "mean_token_accuracy": 0.17134765535593033,
      "num_tokens": 24166115.0,
      "step": 9535
    },
    {
      "entropy": 6.219246292114258,
      "epoch": 1.1009809578765146,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004890539260550181,
      "loss": 5.9368,
      "mean_token_accuracy": 0.18281740695238113,
      "num_tokens": 24178725.0,
      "step": 9540
    },
    {
      "entropy": 6.230550146102905,
      "epoch": 1.1015579919215233,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004890412085014292,
      "loss": 5.9611,
      "mean_token_accuracy": 0.18544753938913344,
      "num_tokens": 24191258.0,
      "step": 9545
    },
    {
      "entropy": 6.254304027557373,
      "epoch": 1.1021350259665321,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004890284837486706,
      "loss": 5.9631,
      "mean_token_accuracy": 0.17143959254026414,
      "num_tokens": 24203096.0,
      "step": 9550
    },
    {
      "entropy": 6.317858791351318,
      "epoch": 1.1027120600115408,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004890157517971704,
      "loss": 6.0098,
      "mean_token_accuracy": 0.17542506158351898,
      "num_tokens": 24214729.0,
      "step": 9555
    },
    {
      "entropy": 6.218365335464478,
      "epoch": 1.1032890940565494,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004890030126473566,
      "loss": 5.9449,
      "mean_token_accuracy": 0.1838381677865982,
      "num_tokens": 24227164.0,
      "step": 9560
    },
    {
      "entropy": 6.301871728897095,
      "epoch": 1.103866128101558,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004889902662996578,
      "loss": 6.0084,
      "mean_token_accuracy": 0.1720730483531952,
      "num_tokens": 24239240.0,
      "step": 9565
    },
    {
      "entropy": 6.2749342918396,
      "epoch": 1.1044431621465667,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004889775127545027,
      "loss": 5.9489,
      "mean_token_accuracy": 0.17174011915922166,
      "num_tokens": 24251562.0,
      "step": 9570
    },
    {
      "entropy": 6.220516014099121,
      "epoch": 1.1050201961915753,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004889647520123202,
      "loss": 5.996,
      "mean_token_accuracy": 0.1756235420703888,
      "num_tokens": 24265439.0,
      "step": 9575
    },
    {
      "entropy": 6.2107807159423825,
      "epoch": 1.105597230236584,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004889519840735396,
      "loss": 5.8746,
      "mean_token_accuracy": 0.17698525339365007,
      "num_tokens": 24278982.0,
      "step": 9580
    },
    {
      "entropy": 6.297426080703735,
      "epoch": 1.1061742642815926,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004889392089385903,
      "loss": 5.998,
      "mean_token_accuracy": 0.16790786385536194,
      "num_tokens": 24291119.0,
      "step": 9585
    },
    {
      "entropy": 6.284995889663696,
      "epoch": 1.1067512983266012,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004889264266079019,
      "loss": 5.9463,
      "mean_token_accuracy": 0.17347494512796402,
      "num_tokens": 24302646.0,
      "step": 9590
    },
    {
      "entropy": 6.230035257339478,
      "epoch": 1.1073283323716099,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004889136370819045,
      "loss": 5.9349,
      "mean_token_accuracy": 0.17844622880220412,
      "num_tokens": 24314754.0,
      "step": 9595
    },
    {
      "entropy": 6.291960000991821,
      "epoch": 1.1079053664166185,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004889008403610281,
      "loss": 5.9221,
      "mean_token_accuracy": 0.172476626932621,
      "num_tokens": 24328241.0,
      "step": 9600
    },
    {
      "entropy": 6.179753732681275,
      "epoch": 1.1084824004616272,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004888880364457033,
      "loss": 5.9554,
      "mean_token_accuracy": 0.17785455733537675,
      "num_tokens": 24340319.0,
      "step": 9605
    },
    {
      "entropy": 6.307469415664673,
      "epoch": 1.1090594345066358,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004888752253363604,
      "loss": 6.0234,
      "mean_token_accuracy": 0.1673170804977417,
      "num_tokens": 24352457.0,
      "step": 9610
    },
    {
      "entropy": 6.275782346725464,
      "epoch": 1.1096364685516447,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004888624070334308,
      "loss": 5.9054,
      "mean_token_accuracy": 0.18013895601034163,
      "num_tokens": 24363587.0,
      "step": 9615
    },
    {
      "entropy": 6.254109668731689,
      "epoch": 1.1102135025966533,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004888495815373452,
      "loss": 6.0126,
      "mean_token_accuracy": 0.17577215880155564,
      "num_tokens": 24376555.0,
      "step": 9620
    },
    {
      "entropy": 6.3164294242858885,
      "epoch": 1.110790536641662,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004888367488485351,
      "loss": 5.9698,
      "mean_token_accuracy": 0.17510447055101394,
      "num_tokens": 24388720.0,
      "step": 9625
    },
    {
      "entropy": 6.320489072799683,
      "epoch": 1.1113675706866706,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004888239089674323,
      "loss": 5.9155,
      "mean_token_accuracy": 0.18267911523580552,
      "num_tokens": 24400320.0,
      "step": 9630
    },
    {
      "entropy": 6.28203125,
      "epoch": 1.1119446047316792,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004888110618944686,
      "loss": 5.9813,
      "mean_token_accuracy": 0.18109282106161118,
      "num_tokens": 24414521.0,
      "step": 9635
    },
    {
      "entropy": 6.197869777679443,
      "epoch": 1.1125216387766879,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004887982076300759,
      "loss": 5.9748,
      "mean_token_accuracy": 0.1734583079814911,
      "num_tokens": 24426254.0,
      "step": 9640
    },
    {
      "entropy": 6.325536775588989,
      "epoch": 1.1130986728216965,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004887853461746867,
      "loss": 6.0058,
      "mean_token_accuracy": 0.1807387501001358,
      "num_tokens": 24438089.0,
      "step": 9645
    },
    {
      "entropy": 6.317915058135986,
      "epoch": 1.1136757068667051,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004887724775287336,
      "loss": 6.0462,
      "mean_token_accuracy": 0.1726977899670601,
      "num_tokens": 24450214.0,
      "step": 9650
    },
    {
      "entropy": 6.156339931488037,
      "epoch": 1.1142527409117138,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004887596016926494,
      "loss": 5.903,
      "mean_token_accuracy": 0.1844400867819786,
      "num_tokens": 24463577.0,
      "step": 9655
    },
    {
      "entropy": 6.1944958686828615,
      "epoch": 1.1148297749567224,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004887467186668673,
      "loss": 5.89,
      "mean_token_accuracy": 0.1837321028113365,
      "num_tokens": 24475533.0,
      "step": 9660
    },
    {
      "entropy": 6.331804418563843,
      "epoch": 1.115406809001731,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004887338284518204,
      "loss": 6.0242,
      "mean_token_accuracy": 0.1740209773182869,
      "num_tokens": 24488779.0,
      "step": 9665
    },
    {
      "entropy": 6.1834485054016115,
      "epoch": 1.1159838430467397,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004887209310479423,
      "loss": 5.8232,
      "mean_token_accuracy": 0.18647678643465043,
      "num_tokens": 24501967.0,
      "step": 9670
    },
    {
      "entropy": 6.206624746322632,
      "epoch": 1.1165608770917483,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004887080264556668,
      "loss": 6.0416,
      "mean_token_accuracy": 0.17203063666820526,
      "num_tokens": 24513743.0,
      "step": 9675
    },
    {
      "entropy": 6.337820100784302,
      "epoch": 1.117137911136757,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004886951146754282,
      "loss": 6.0012,
      "mean_token_accuracy": 0.17416843473911287,
      "num_tokens": 24524782.0,
      "step": 9680
    },
    {
      "entropy": 6.270286798477173,
      "epoch": 1.1177149451817656,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004886821957076603,
      "loss": 6.0378,
      "mean_token_accuracy": 0.17253595739603042,
      "num_tokens": 24537598.0,
      "step": 9685
    },
    {
      "entropy": 6.285682916641235,
      "epoch": 1.1182919792267745,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000488669269552798,
      "loss": 6.0296,
      "mean_token_accuracy": 0.1664452761411667,
      "num_tokens": 24551100.0,
      "step": 9690
    },
    {
      "entropy": 6.312683248519898,
      "epoch": 1.1188690132717831,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000488656336211276,
      "loss": 6.0474,
      "mean_token_accuracy": 0.17086057364940643,
      "num_tokens": 24562604.0,
      "step": 9695
    },
    {
      "entropy": 6.284778070449829,
      "epoch": 1.1194460473167918,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004886433956835292,
      "loss": 6.0066,
      "mean_token_accuracy": 0.17307329326868057,
      "num_tokens": 24573990.0,
      "step": 9700
    },
    {
      "entropy": 6.28976092338562,
      "epoch": 1.1200230813618004,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004886304479699929,
      "loss": 5.9857,
      "mean_token_accuracy": 0.17756191939115523,
      "num_tokens": 24586320.0,
      "step": 9705
    },
    {
      "entropy": 6.265582656860351,
      "epoch": 1.120600115406809,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004886174930711027,
      "loss": 5.953,
      "mean_token_accuracy": 0.18335249722003938,
      "num_tokens": 24597995.0,
      "step": 9710
    },
    {
      "entropy": 6.211058616638184,
      "epoch": 1.1211771494518177,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004886045309872941,
      "loss": 6.0002,
      "mean_token_accuracy": 0.17694109082221984,
      "num_tokens": 24609752.0,
      "step": 9715
    },
    {
      "entropy": 6.227928304672242,
      "epoch": 1.1217541834968263,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004885915617190034,
      "loss": 5.9134,
      "mean_token_accuracy": 0.17937326729297637,
      "num_tokens": 24622374.0,
      "step": 9720
    },
    {
      "entropy": 6.163439416885376,
      "epoch": 1.122331217541835,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004885785852666664,
      "loss": 5.7504,
      "mean_token_accuracy": 0.1951364681124687,
      "num_tokens": 24634877.0,
      "step": 9725
    },
    {
      "entropy": 6.205390739440918,
      "epoch": 1.1229082515868436,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004885656016307199,
      "loss": 5.9271,
      "mean_token_accuracy": 0.18244069665670395,
      "num_tokens": 24647498.0,
      "step": 9730
    },
    {
      "entropy": 6.170197486877441,
      "epoch": 1.1234852856318522,
      "grad_norm": 0.74609375,
      "learning_rate": 0.0004885526108116004,
      "loss": 5.9342,
      "mean_token_accuracy": 0.1805289387702942,
      "num_tokens": 24660331.0,
      "step": 9735
    },
    {
      "entropy": 6.309574604034424,
      "epoch": 1.1240623196768609,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000488539612809745,
      "loss": 5.9435,
      "mean_token_accuracy": 0.17946696281433105,
      "num_tokens": 24673169.0,
      "step": 9740
    },
    {
      "entropy": 6.188028383255005,
      "epoch": 1.1246393537218695,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004885266076255907,
      "loss": 5.8686,
      "mean_token_accuracy": 0.17608639299869538,
      "num_tokens": 24684900.0,
      "step": 9745
    },
    {
      "entropy": 6.24795560836792,
      "epoch": 1.1252163877668782,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000488513595259575,
      "loss": 5.8416,
      "mean_token_accuracy": 0.18763349801301957,
      "num_tokens": 24699135.0,
      "step": 9750
    },
    {
      "entropy": 6.301227998733521,
      "epoch": 1.125793421811887,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004885005757121357,
      "loss": 5.9887,
      "mean_token_accuracy": 0.17296784073114396,
      "num_tokens": 24711402.0,
      "step": 9755
    },
    {
      "entropy": 6.220034885406494,
      "epoch": 1.1263704558568954,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004884875489837105,
      "loss": 5.9736,
      "mean_token_accuracy": 0.17327672839164734,
      "num_tokens": 24723438.0,
      "step": 9760
    },
    {
      "entropy": 6.2852949619293215,
      "epoch": 1.1269474899019043,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004884745150747378,
      "loss": 5.9678,
      "mean_token_accuracy": 0.18274324685335158,
      "num_tokens": 24737939.0,
      "step": 9765
    },
    {
      "entropy": 6.229353189468384,
      "epoch": 1.127524523946913,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004884614739856556,
      "loss": 5.8845,
      "mean_token_accuracy": 0.17819585502147675,
      "num_tokens": 24751120.0,
      "step": 9770
    },
    {
      "entropy": 6.2253937244415285,
      "epoch": 1.1281015579919216,
      "grad_norm": 0.875,
      "learning_rate": 0.0004884484257169028,
      "loss": 5.8733,
      "mean_token_accuracy": 0.18652616888284684,
      "num_tokens": 24763156.0,
      "step": 9775
    },
    {
      "entropy": 6.247245836257934,
      "epoch": 1.1286785920369302,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004884353702689183,
      "loss": 5.995,
      "mean_token_accuracy": 0.17599666714668274,
      "num_tokens": 24775728.0,
      "step": 9780
    },
    {
      "entropy": 6.152251386642456,
      "epoch": 1.1292556260819389,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004884223076421411,
      "loss": 5.8939,
      "mean_token_accuracy": 0.18010073751211167,
      "num_tokens": 24789232.0,
      "step": 9785
    },
    {
      "entropy": 6.2644367694854735,
      "epoch": 1.1298326601269475,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004884092378370106,
      "loss": 5.9569,
      "mean_token_accuracy": 0.17408268004655839,
      "num_tokens": 24801188.0,
      "step": 9790
    },
    {
      "entropy": 6.232016277313233,
      "epoch": 1.1304096941719561,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004883961608539664,
      "loss": 5.9057,
      "mean_token_accuracy": 0.18096981942653656,
      "num_tokens": 24814482.0,
      "step": 9795
    },
    {
      "entropy": 6.2632325172424315,
      "epoch": 1.1309867282169648,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004883830766934484,
      "loss": 5.9835,
      "mean_token_accuracy": 0.17321178317070007,
      "num_tokens": 24829042.0,
      "step": 9800
    },
    {
      "entropy": 6.312577867507935,
      "epoch": 1.1315637622619734,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004883699853558965,
      "loss": 5.9507,
      "mean_token_accuracy": 0.17316993027925492,
      "num_tokens": 24842227.0,
      "step": 9805
    },
    {
      "entropy": 6.177276229858398,
      "epoch": 1.132140796306982,
      "grad_norm": 1.0,
      "learning_rate": 0.0004883568868417511,
      "loss": 5.8772,
      "mean_token_accuracy": 0.18718491941690446,
      "num_tokens": 24856225.0,
      "step": 9810
    },
    {
      "entropy": 6.271045732498169,
      "epoch": 1.1327178303519907,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004883437811514528,
      "loss": 5.9454,
      "mean_token_accuracy": 0.17582037746906282,
      "num_tokens": 24868889.0,
      "step": 9815
    },
    {
      "entropy": 6.276317453384399,
      "epoch": 1.1332948643969993,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004883306682854424,
      "loss": 5.8944,
      "mean_token_accuracy": 0.17765705287456512,
      "num_tokens": 24881738.0,
      "step": 9820
    },
    {
      "entropy": 6.29815149307251,
      "epoch": 1.133871898442008,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004883175482441611,
      "loss": 5.9537,
      "mean_token_accuracy": 0.17408571243286133,
      "num_tokens": 24893886.0,
      "step": 9825
    },
    {
      "entropy": 6.201640462875366,
      "epoch": 1.1344489324870168,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004883044210280499,
      "loss": 5.9868,
      "mean_token_accuracy": 0.17262559235095978,
      "num_tokens": 24908213.0,
      "step": 9830
    },
    {
      "entropy": 6.311663818359375,
      "epoch": 1.1350259665320255,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004882912866375505,
      "loss": 6.0331,
      "mean_token_accuracy": 0.17144252359867096,
      "num_tokens": 24921157.0,
      "step": 9835
    },
    {
      "entropy": 6.321885681152343,
      "epoch": 1.1356030005770341,
      "grad_norm": 0.828125,
      "learning_rate": 0.00048827814507310455,
      "loss": 5.9731,
      "mean_token_accuracy": 0.1762096717953682,
      "num_tokens": 24933922.0,
      "step": 9840
    },
    {
      "entropy": 6.22668604850769,
      "epoch": 1.1361800346220428,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048826499633515416,
      "loss": 5.9217,
      "mean_token_accuracy": 0.1811533421278,
      "num_tokens": 24945351.0,
      "step": 9845
    },
    {
      "entropy": 6.248162698745728,
      "epoch": 1.1367570686670514,
      "grad_norm": 0.84375,
      "learning_rate": 0.00048825184042414156,
      "loss": 5.9485,
      "mean_token_accuracy": 0.18205004185438156,
      "num_tokens": 24958099.0,
      "step": 9850
    },
    {
      "entropy": 6.259473514556885,
      "epoch": 1.13733410271206,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004882386773405092,
      "loss": 5.951,
      "mean_token_accuracy": 0.1808436021208763,
      "num_tokens": 24970622.0,
      "step": 9855
    },
    {
      "entropy": 6.347340154647827,
      "epoch": 1.1379111367570687,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004882255070846999,
      "loss": 6.0507,
      "mean_token_accuracy": 0.16723102182149888,
      "num_tokens": 24983456.0,
      "step": 9860
    },
    {
      "entropy": 6.251977062225341,
      "epoch": 1.1384881708020773,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00048821232965715663,
      "loss": 5.9674,
      "mean_token_accuracy": 0.18168559670448303,
      "num_tokens": 24995976.0,
      "step": 9865
    },
    {
      "entropy": 6.163515710830689,
      "epoch": 1.139065204847086,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004881991450583225,
      "loss": 5.943,
      "mean_token_accuracy": 0.1851297974586487,
      "num_tokens": 25008456.0,
      "step": 9870
    },
    {
      "entropy": 6.317627048492431,
      "epoch": 1.1396422388920946,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000488185953288641,
      "loss": 5.937,
      "mean_token_accuracy": 0.1743927538394928,
      "num_tokens": 25021186.0,
      "step": 9875
    },
    {
      "entropy": 6.253435707092285,
      "epoch": 1.1402192729371032,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004881727543485559,
      "loss": 6.0089,
      "mean_token_accuracy": 0.1727897360920906,
      "num_tokens": 25034561.0,
      "step": 9880
    },
    {
      "entropy": 6.223213577270508,
      "epoch": 1.1407963069821119,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048815954823851107,
      "loss": 5.929,
      "mean_token_accuracy": 0.17690471261739732,
      "num_tokens": 25046107.0,
      "step": 9885
    },
    {
      "entropy": 6.287363529205322,
      "epoch": 1.1413733410271205,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048814633495895067,
      "loss": 5.9566,
      "mean_token_accuracy": 0.17542949169874192,
      "num_tokens": 25058605.0,
      "step": 9890
    },
    {
      "entropy": 6.245450973510742,
      "epoch": 1.1419503750721294,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004881331145103192,
      "loss": 5.904,
      "mean_token_accuracy": 0.18198370784521103,
      "num_tokens": 25070730.0,
      "step": 9895
    },
    {
      "entropy": 6.239177989959717,
      "epoch": 1.1425274091171378,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004881198868930614,
      "loss": 5.9344,
      "mean_token_accuracy": 0.18006508350372313,
      "num_tokens": 25084121.0,
      "step": 9900
    },
    {
      "entropy": 6.264116048812866,
      "epoch": 1.1431044431621467,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048810665210762195,
      "loss": 5.941,
      "mean_token_accuracy": 0.18270495682954788,
      "num_tokens": 25096132.0,
      "step": 9905
    },
    {
      "entropy": 6.295691347122192,
      "epoch": 1.1436814772071553,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00048809341015444615,
      "loss": 6.063,
      "mean_token_accuracy": 0.17179990708827972,
      "num_tokens": 25108220.0,
      "step": 9910
    },
    {
      "entropy": 6.287291240692139,
      "epoch": 1.144258511252164,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048808016103397934,
      "loss": 6.0021,
      "mean_token_accuracy": 0.17640800923109054,
      "num_tokens": 25120955.0,
      "step": 9915
    },
    {
      "entropy": 6.336846494674683,
      "epoch": 1.1448355452971726,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004880669047466671,
      "loss": 6.0079,
      "mean_token_accuracy": 0.17164885699748994,
      "num_tokens": 25133644.0,
      "step": 9920
    },
    {
      "entropy": 6.164778423309326,
      "epoch": 1.1454125793421812,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00048805364129295554,
      "loss": 5.8454,
      "mean_token_accuracy": 0.18924638926982879,
      "num_tokens": 25147172.0,
      "step": 9925
    },
    {
      "entropy": 6.277313137054444,
      "epoch": 1.1459896133871899,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048804037067329037,
      "loss": 5.9499,
      "mean_token_accuracy": 0.18157438933849335,
      "num_tokens": 25159293.0,
      "step": 9930
    },
    {
      "entropy": 6.288929414749146,
      "epoch": 1.1465666474321985,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004880270928881183,
      "loss": 5.9989,
      "mean_token_accuracy": 0.17166264206171036,
      "num_tokens": 25172149.0,
      "step": 9935
    },
    {
      "entropy": 6.259950590133667,
      "epoch": 1.1471436814772071,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004880138079378857,
      "loss": 5.8348,
      "mean_token_accuracy": 0.17915803641080857,
      "num_tokens": 25184255.0,
      "step": 9940
    },
    {
      "entropy": 6.167291831970215,
      "epoch": 1.1477207155222158,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004880005158230395,
      "loss": 5.9105,
      "mean_token_accuracy": 0.17886314392089844,
      "num_tokens": 25197532.0,
      "step": 9945
    },
    {
      "entropy": 6.236536455154419,
      "epoch": 1.1482977495672244,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004879872165440268,
      "loss": 5.9421,
      "mean_token_accuracy": 0.1795220270752907,
      "num_tokens": 25210046.0,
      "step": 9950
    },
    {
      "entropy": 6.379000902175903,
      "epoch": 1.148874783612233,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004879739101012948,
      "loss": 6.0319,
      "mean_token_accuracy": 0.17408398389816285,
      "num_tokens": 25221123.0,
      "step": 9955
    },
    {
      "entropy": 6.206017351150512,
      "epoch": 1.1494518176572417,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004879605964952911,
      "loss": 5.9252,
      "mean_token_accuracy": 0.18063071370124817,
      "num_tokens": 25233196.0,
      "step": 9960
    },
    {
      "entropy": 6.240593004226684,
      "epoch": 1.1500288517022503,
      "grad_norm": 0.7734375,
      "learning_rate": 0.00048794727572646366,
      "loss": 5.9353,
      "mean_token_accuracy": 0.18255639672279358,
      "num_tokens": 25247921.0,
      "step": 9965
    },
    {
      "entropy": 6.255258989334107,
      "epoch": 1.1506058857472592,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004879339477952603,
      "loss": 5.9303,
      "mean_token_accuracy": 0.17574996054172515,
      "num_tokens": 25260999.0,
      "step": 9970
    },
    {
      "entropy": 6.160217332839966,
      "epoch": 1.1511829197922678,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048792061270212935,
      "loss": 5.9314,
      "mean_token_accuracy": 0.17898819595575333,
      "num_tokens": 25273435.0,
      "step": 9975
    },
    {
      "entropy": 6.338103199005127,
      "epoch": 1.1517599538372765,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004879072704475193,
      "loss": 6.0532,
      "mean_token_accuracy": 0.17280764877796173,
      "num_tokens": 25286071.0,
      "step": 9980
    },
    {
      "entropy": 6.353445053100586,
      "epoch": 1.1523369878822851,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048789392103187906,
      "loss": 5.9727,
      "mean_token_accuracy": 0.16990299075841903,
      "num_tokens": 25298083.0,
      "step": 9985
    },
    {
      "entropy": 6.3070615291595455,
      "epoch": 1.1529140219272938,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004878805644556575,
      "loss": 6.0159,
      "mean_token_accuracy": 0.1696576789021492,
      "num_tokens": 25310488.0,
      "step": 9990
    },
    {
      "entropy": 6.260022926330566,
      "epoch": 1.1534910559723024,
      "grad_norm": 0.8046875,
      "learning_rate": 0.000487867200719304,
      "loss": 5.944,
      "mean_token_accuracy": 0.17872022241353988,
      "num_tokens": 25323135.0,
      "step": 9995
    },
    {
      "entropy": 6.202866268157959,
      "epoch": 1.154068090017311,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004878538298232678,
      "loss": 5.9931,
      "mean_token_accuracy": 0.17249661833047866,
      "num_tokens": 25335349.0,
      "step": 10000
    },
    {
      "entropy": 6.328696584701538,
      "epoch": 1.1546451240623197,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004878404517679988,
      "loss": 5.9883,
      "mean_token_accuracy": 0.17477356344461442,
      "num_tokens": 25347717.0,
      "step": 10005
    },
    {
      "entropy": 6.2257850646972654,
      "epoch": 1.1552221581073283,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048782706655394695,
      "loss": 5.8876,
      "mean_token_accuracy": 0.18441763073205947,
      "num_tokens": 25360343.0,
      "step": 10010
    },
    {
      "entropy": 6.241095161437988,
      "epoch": 1.155799192152337,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004878136741815624,
      "loss": 5.9635,
      "mean_token_accuracy": 0.18109873235225676,
      "num_tokens": 25373516.0,
      "step": 10015
    },
    {
      "entropy": 6.361446714401245,
      "epoch": 1.1563762261973456,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004878002746512956,
      "loss": 6.0595,
      "mean_token_accuracy": 0.17292851507663726,
      "num_tokens": 25386728.0,
      "step": 10020
    },
    {
      "entropy": 6.218955707550049,
      "epoch": 1.1569532602423542,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004877868679635974,
      "loss": 5.9024,
      "mean_token_accuracy": 0.1897459074854851,
      "num_tokens": 25399419.0,
      "step": 10025
    },
    {
      "entropy": 6.189811897277832,
      "epoch": 1.1575302942873629,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004877734541189185,
      "loss": 5.9606,
      "mean_token_accuracy": 0.1754186123609543,
      "num_tokens": 25411922.0,
      "step": 10030
    },
    {
      "entropy": 6.3283487319946286,
      "epoch": 1.1581073283323717,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048776003311771013,
      "loss": 5.956,
      "mean_token_accuracy": 0.17566704303026198,
      "num_tokens": 25425430.0,
      "step": 10035
    },
    {
      "entropy": 6.30607500076294,
      "epoch": 1.1586843623773802,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004877466049604238,
      "loss": 6.0852,
      "mean_token_accuracy": 0.16973078697919847,
      "num_tokens": 25437222.0,
      "step": 10040
    },
    {
      "entropy": 6.215491485595703,
      "epoch": 1.159261396422389,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048773316964751106,
      "loss": 5.8869,
      "mean_token_accuracy": 0.1789697676897049,
      "num_tokens": 25448173.0,
      "step": 10045
    },
    {
      "entropy": 6.228227281570435,
      "epoch": 1.1598384304673977,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004877197271794239,
      "loss": 5.9049,
      "mean_token_accuracy": 0.1868424654006958,
      "num_tokens": 25458868.0,
      "step": 10050
    },
    {
      "entropy": 6.257163572311401,
      "epoch": 1.1604154645124063,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004877062775566142,
      "loss": 5.9607,
      "mean_token_accuracy": 0.17475768327713012,
      "num_tokens": 25470986.0,
      "step": 10055
    },
    {
      "entropy": 6.3420631885528564,
      "epoch": 1.160992498557415,
      "grad_norm": 0.80859375,
      "learning_rate": 0.00048769282077953464,
      "loss": 6.0805,
      "mean_token_accuracy": 0.17108901888132094,
      "num_tokens": 25483484.0,
      "step": 10060
    },
    {
      "entropy": 6.311474704742432,
      "epoch": 1.1615695326024236,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048767935684863775,
      "loss": 6.0074,
      "mean_token_accuracy": 0.1761361077427864,
      "num_tokens": 25496051.0,
      "step": 10065
    },
    {
      "entropy": 6.207817316055298,
      "epoch": 1.1621465666474322,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004876658857643762,
      "loss": 5.901,
      "mean_token_accuracy": 0.18453028351068496,
      "num_tokens": 25508543.0,
      "step": 10070
    },
    {
      "entropy": 6.313574409484863,
      "epoch": 1.1627236006924409,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004876524075272034,
      "loss": 6.0458,
      "mean_token_accuracy": 0.1723420113325119,
      "num_tokens": 25521306.0,
      "step": 10075
    },
    {
      "entropy": 6.213683414459228,
      "epoch": 1.1633006347374495,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048763892213757234,
      "loss": 5.8739,
      "mean_token_accuracy": 0.18324902951717376,
      "num_tokens": 25533428.0,
      "step": 10080
    },
    {
      "entropy": 6.2165604591369625,
      "epoch": 1.1638776687824581,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048762542959593683,
      "loss": 5.9177,
      "mean_token_accuracy": 0.17710841596126556,
      "num_tokens": 25545747.0,
      "step": 10085
    },
    {
      "entropy": 6.262358379364014,
      "epoch": 1.1644547028274668,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004876119299027506,
      "loss": 5.9793,
      "mean_token_accuracy": 0.1738194927573204,
      "num_tokens": 25558460.0,
      "step": 10090
    },
    {
      "entropy": 6.306369018554688,
      "epoch": 1.1650317368724754,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048759842305846766,
      "loss": 5.9543,
      "mean_token_accuracy": 0.17817896008491516,
      "num_tokens": 25571583.0,
      "step": 10095
    },
    {
      "entropy": 6.275261402130127,
      "epoch": 1.165608770917484,
      "grad_norm": 0.875,
      "learning_rate": 0.0004875849090635425,
      "loss": 5.9441,
      "mean_token_accuracy": 0.17820047736167907,
      "num_tokens": 25586133.0,
      "step": 10100
    },
    {
      "entropy": 6.1719725131988525,
      "epoch": 1.1661858049624927,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00048757138791842943,
      "loss": 5.8795,
      "mean_token_accuracy": 0.18058676272630692,
      "num_tokens": 25598935.0,
      "step": 10105
    },
    {
      "entropy": 6.281197595596313,
      "epoch": 1.1667628390075016,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004875578596235832,
      "loss": 5.9577,
      "mean_token_accuracy": 0.17972777932882308,
      "num_tokens": 25611400.0,
      "step": 10110
    },
    {
      "entropy": 6.264150381088257,
      "epoch": 1.1673398730525102,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004875443241794591,
      "loss": 5.9619,
      "mean_token_accuracy": 0.17912423759698867,
      "num_tokens": 25623389.0,
      "step": 10115
    },
    {
      "entropy": 6.290235376358032,
      "epoch": 1.1679169070975188,
      "grad_norm": 0.79296875,
      "learning_rate": 0.00048753078158651227,
      "loss": 5.9661,
      "mean_token_accuracy": 0.17819956839084625,
      "num_tokens": 25636702.0,
      "step": 10120
    },
    {
      "entropy": 6.168326663970947,
      "epoch": 1.1684939411425275,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000487517231845198,
      "loss": 5.9049,
      "mean_token_accuracy": 0.18582970052957534,
      "num_tokens": 25650216.0,
      "step": 10125
    },
    {
      "entropy": 6.188032722473144,
      "epoch": 1.169070975187536,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004875036749559724,
      "loss": 5.8525,
      "mean_token_accuracy": 0.18714485466480255,
      "num_tokens": 25662948.0,
      "step": 10130
    },
    {
      "entropy": 6.2656269550323485,
      "epoch": 1.1696480092325447,
      "grad_norm": 0.875,
      "learning_rate": 0.00048749011091929115,
      "loss": 5.9321,
      "mean_token_accuracy": 0.17210010439157486,
      "num_tokens": 25676984.0,
      "step": 10135
    },
    {
      "entropy": 6.225550413131714,
      "epoch": 1.1702250432775534,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004874765397356105,
      "loss": 5.9228,
      "mean_token_accuracy": 0.1823028579354286,
      "num_tokens": 25689422.0,
      "step": 10140
    },
    {
      "entropy": 6.189141035079956,
      "epoch": 1.170802077322562,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004874629614053871,
      "loss": 5.9438,
      "mean_token_accuracy": 0.17562722712755202,
      "num_tokens": 25701621.0,
      "step": 10145
    },
    {
      "entropy": 6.222708034515381,
      "epoch": 1.1713791113675707,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004874493759290775,
      "loss": 5.9719,
      "mean_token_accuracy": 0.1808430328965187,
      "num_tokens": 25716102.0,
      "step": 10150
    },
    {
      "entropy": 6.301426887512207,
      "epoch": 1.1719561454125793,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048743578330713854,
      "loss": 6.0343,
      "mean_token_accuracy": 0.1728496566414833,
      "num_tokens": 25728224.0,
      "step": 10155
    },
    {
      "entropy": 6.255610847473145,
      "epoch": 1.172533179457588,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004874221835400277,
      "loss": 5.9604,
      "mean_token_accuracy": 0.1778796449303627,
      "num_tokens": 25741709.0,
      "step": 10160
    },
    {
      "entropy": 6.272855520248413,
      "epoch": 1.1731102135025966,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004874085766282022,
      "loss": 5.9023,
      "mean_token_accuracy": 0.17917946428060533,
      "num_tokens": 25754565.0,
      "step": 10165
    },
    {
      "entropy": 6.303839254379272,
      "epoch": 1.1736872475476052,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048739496257211966,
      "loss": 6.0011,
      "mean_token_accuracy": 0.1711631953716278,
      "num_tokens": 25767329.0,
      "step": 10170
    },
    {
      "entropy": 6.270671081542969,
      "epoch": 1.1742642815926139,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048738134137223815,
      "loss": 5.9433,
      "mean_token_accuracy": 0.17837364673614503,
      "num_tokens": 25780253.0,
      "step": 10175
    },
    {
      "entropy": 6.211219215393067,
      "epoch": 1.1748413156376225,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048736771302901566,
      "loss": 5.8644,
      "mean_token_accuracy": 0.1847034215927124,
      "num_tokens": 25792712.0,
      "step": 10180
    },
    {
      "entropy": 6.237783908843994,
      "epoch": 1.1754183496826314,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00048735407754291063,
      "loss": 5.9562,
      "mean_token_accuracy": 0.1779343381524086,
      "num_tokens": 25804735.0,
      "step": 10185
    },
    {
      "entropy": 6.295846176147461,
      "epoch": 1.17599538372764,
      "grad_norm": 0.84375,
      "learning_rate": 0.00048734043491438175,
      "loss": 5.9706,
      "mean_token_accuracy": 0.178368416428566,
      "num_tokens": 25816880.0,
      "step": 10190
    },
    {
      "entropy": 6.2071069240570065,
      "epoch": 1.1765724177726486,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048732678514388773,
      "loss": 5.897,
      "mean_token_accuracy": 0.17773630768060683,
      "num_tokens": 25829756.0,
      "step": 10195
    },
    {
      "entropy": 6.221304512023925,
      "epoch": 1.1771494518176573,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004873131282318878,
      "loss": 5.8733,
      "mean_token_accuracy": 0.1843046262860298,
      "num_tokens": 25842318.0,
      "step": 10200
    },
    {
      "entropy": 6.237124586105347,
      "epoch": 1.177726485862666,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048729946417884126,
      "loss": 5.8587,
      "mean_token_accuracy": 0.17836786061525345,
      "num_tokens": 25854716.0,
      "step": 10205
    },
    {
      "entropy": 6.254219341278076,
      "epoch": 1.1783035199076746,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004872857929852076,
      "loss": 5.8758,
      "mean_token_accuracy": 0.18610414117574692,
      "num_tokens": 25867678.0,
      "step": 10210
    },
    {
      "entropy": 6.219987678527832,
      "epoch": 1.1788805539526832,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004872721146514469,
      "loss": 5.924,
      "mean_token_accuracy": 0.17481060624122619,
      "num_tokens": 25880845.0,
      "step": 10215
    },
    {
      "entropy": 6.169091033935547,
      "epoch": 1.1794575879976918,
      "grad_norm": 0.77734375,
      "learning_rate": 0.000487258429178019,
      "loss": 5.8927,
      "mean_token_accuracy": 0.18335053622722625,
      "num_tokens": 25893693.0,
      "step": 10220
    },
    {
      "entropy": 6.308605909347534,
      "epoch": 1.1800346220427005,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048724473656538427,
      "loss": 5.9838,
      "mean_token_accuracy": 0.1702417626976967,
      "num_tokens": 25906752.0,
      "step": 10225
    },
    {
      "entropy": 6.252112340927124,
      "epoch": 1.1806116560877091,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004872310368140032,
      "loss": 5.968,
      "mean_token_accuracy": 0.17995206713676454,
      "num_tokens": 25918343.0,
      "step": 10230
    },
    {
      "entropy": 6.320370054244995,
      "epoch": 1.1811886901327178,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004872173299243368,
      "loss": 5.9648,
      "mean_token_accuracy": 0.1704358845949173,
      "num_tokens": 25930521.0,
      "step": 10235
    },
    {
      "entropy": 6.319835519790649,
      "epoch": 1.1817657241777264,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048720361589684575,
      "loss": 5.9538,
      "mean_token_accuracy": 0.16953953057527543,
      "num_tokens": 25942618.0,
      "step": 10240
    },
    {
      "entropy": 6.239289712905884,
      "epoch": 1.182342758222735,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048718989473199147,
      "loss": 6.0173,
      "mean_token_accuracy": 0.17623186558485032,
      "num_tokens": 25956068.0,
      "step": 10245
    },
    {
      "entropy": 6.261501979827881,
      "epoch": 1.182919792267744,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004871761664302356,
      "loss": 5.9362,
      "mean_token_accuracy": 0.18427009284496307,
      "num_tokens": 25967639.0,
      "step": 10250
    },
    {
      "entropy": 6.202490139007568,
      "epoch": 1.1834968263127525,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004871624309920397,
      "loss": 5.8576,
      "mean_token_accuracy": 0.18903457522392272,
      "num_tokens": 25980493.0,
      "step": 10255
    },
    {
      "entropy": 6.278797340393067,
      "epoch": 1.1840738603577612,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00048714868841786586,
      "loss": 5.9615,
      "mean_token_accuracy": 0.17092124819755555,
      "num_tokens": 25994414.0,
      "step": 10260
    },
    {
      "entropy": 6.322451639175415,
      "epoch": 1.1846508944027698,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048713493870817626,
      "loss": 5.9682,
      "mean_token_accuracy": 0.17578869313001633,
      "num_tokens": 26008510.0,
      "step": 10265
    },
    {
      "entropy": 6.273719406127929,
      "epoch": 1.1852279284477785,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048712118186343336,
      "loss": 5.9798,
      "mean_token_accuracy": 0.17121853679418564,
      "num_tokens": 26022682.0,
      "step": 10270
    },
    {
      "entropy": 6.282404088973999,
      "epoch": 1.185804962492787,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048710741788409985,
      "loss": 5.9404,
      "mean_token_accuracy": 0.18331726789474487,
      "num_tokens": 26035476.0,
      "step": 10275
    },
    {
      "entropy": 6.266376304626465,
      "epoch": 1.1863819965377957,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004870936467706387,
      "loss": 5.9596,
      "mean_token_accuracy": 0.17464197874069215,
      "num_tokens": 26049253.0,
      "step": 10280
    },
    {
      "entropy": 6.241795921325684,
      "epoch": 1.1869590305828044,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004870798685235131,
      "loss": 5.9758,
      "mean_token_accuracy": 0.17531266808509827,
      "num_tokens": 26061343.0,
      "step": 10285
    },
    {
      "entropy": 6.286275720596313,
      "epoch": 1.187536064627813,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048706608314318654,
      "loss": 5.9276,
      "mean_token_accuracy": 0.17270954847335815,
      "num_tokens": 26074133.0,
      "step": 10290
    },
    {
      "entropy": 6.287650918960571,
      "epoch": 1.1881130986728217,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004870522906301225,
      "loss": 5.9051,
      "mean_token_accuracy": 0.1775716796517372,
      "num_tokens": 26086706.0,
      "step": 10295
    },
    {
      "entropy": 6.222880935668945,
      "epoch": 1.1886901327178303,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000487038490984785,
      "loss": 5.8803,
      "mean_token_accuracy": 0.1774602487683296,
      "num_tokens": 26100074.0,
      "step": 10300
    },
    {
      "entropy": 6.262276887893677,
      "epoch": 1.189267166762839,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048702468420763826,
      "loss": 5.9714,
      "mean_token_accuracy": 0.17627137899398804,
      "num_tokens": 26112758.0,
      "step": 10305
    },
    {
      "entropy": 6.281043195724488,
      "epoch": 1.1898442008078476,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048701087029914657,
      "loss": 5.9997,
      "mean_token_accuracy": 0.17558603584766388,
      "num_tokens": 26126088.0,
      "step": 10310
    },
    {
      "entropy": 6.287307453155518,
      "epoch": 1.1904212348528562,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004869970492597745,
      "loss": 5.9638,
      "mean_token_accuracy": 0.1700105145573616,
      "num_tokens": 26137793.0,
      "step": 10315
    },
    {
      "entropy": 6.2597325325012205,
      "epoch": 1.1909982688978649,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004869832210899871,
      "loss": 5.9329,
      "mean_token_accuracy": 0.1832864001393318,
      "num_tokens": 26150975.0,
      "step": 10320
    },
    {
      "entropy": 6.2944934368133545,
      "epoch": 1.1915753029428737,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048696938579024927,
      "loss": 5.9451,
      "mean_token_accuracy": 0.16789559572935103,
      "num_tokens": 26162658.0,
      "step": 10325
    },
    {
      "entropy": 6.289283800125122,
      "epoch": 1.1921523369878824,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048695554336102644,
      "loss": 5.9899,
      "mean_token_accuracy": 0.17547340244054793,
      "num_tokens": 26176089.0,
      "step": 10330
    },
    {
      "entropy": 6.292475605010987,
      "epoch": 1.192729371032891,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00048694169380278417,
      "loss": 6.0056,
      "mean_token_accuracy": 0.17644051909446717,
      "num_tokens": 26188974.0,
      "step": 10335
    },
    {
      "entropy": 6.325517845153809,
      "epoch": 1.1933064050778996,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004869278371159884,
      "loss": 5.9276,
      "mean_token_accuracy": 0.17661849409341812,
      "num_tokens": 26201628.0,
      "step": 10340
    },
    {
      "entropy": 6.220251035690308,
      "epoch": 1.1938834391229083,
      "grad_norm": 0.78515625,
      "learning_rate": 0.00048691397330110503,
      "loss": 5.8414,
      "mean_token_accuracy": 0.18477447777986528,
      "num_tokens": 26213540.0,
      "step": 10345
    },
    {
      "entropy": 6.30058274269104,
      "epoch": 1.194460473167917,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004869001023586005,
      "loss": 5.9964,
      "mean_token_accuracy": 0.17597270160913467,
      "num_tokens": 26226476.0,
      "step": 10350
    },
    {
      "entropy": 6.184869003295899,
      "epoch": 1.1950375072129256,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004868862242889413,
      "loss": 5.8766,
      "mean_token_accuracy": 0.18001159876585007,
      "num_tokens": 26238854.0,
      "step": 10355
    },
    {
      "entropy": 6.242160415649414,
      "epoch": 1.1956145412579342,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000486872339092594,
      "loss": 5.876,
      "mean_token_accuracy": 0.18178507834672927,
      "num_tokens": 26251279.0,
      "step": 10360
    },
    {
      "entropy": 6.350482940673828,
      "epoch": 1.1961915753029428,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000486858446770026,
      "loss": 5.969,
      "mean_token_accuracy": 0.16874755918979645,
      "num_tokens": 26263412.0,
      "step": 10365
    },
    {
      "entropy": 6.212138938903808,
      "epoch": 1.1967686093479515,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004868445473217043,
      "loss": 5.9521,
      "mean_token_accuracy": 0.17576365023851395,
      "num_tokens": 26276085.0,
      "step": 10370
    },
    {
      "entropy": 6.272251415252685,
      "epoch": 1.1973456433929601,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004868306407480965,
      "loss": 5.9841,
      "mean_token_accuracy": 0.17558115869760513,
      "num_tokens": 26288856.0,
      "step": 10375
    },
    {
      "entropy": 6.260346508026123,
      "epoch": 1.1979226774379688,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048681672704967036,
      "loss": 5.8918,
      "mean_token_accuracy": 0.18485673367977143,
      "num_tokens": 26300346.0,
      "step": 10380
    },
    {
      "entropy": 6.227512454986572,
      "epoch": 1.1984997114829774,
      "grad_norm": 0.7421875,
      "learning_rate": 0.0004868028062268938,
      "loss": 5.9256,
      "mean_token_accuracy": 0.19136003255844117,
      "num_tokens": 26313414.0,
      "step": 10385
    },
    {
      "entropy": 6.275210571289063,
      "epoch": 1.1990767455279863,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048678887828023504,
      "loss": 6.0018,
      "mean_token_accuracy": 0.1747821018099785,
      "num_tokens": 26325779.0,
      "step": 10390
    },
    {
      "entropy": 6.33063383102417,
      "epoch": 1.1996537795729947,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004867749432101626,
      "loss": 5.963,
      "mean_token_accuracy": 0.17039824426174163,
      "num_tokens": 26337513.0,
      "step": 10395
    },
    {
      "entropy": 6.262580442428589,
      "epoch": 1.2002308136180035,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004867610010171451,
      "loss": 5.9924,
      "mean_token_accuracy": 0.171356101334095,
      "num_tokens": 26350569.0,
      "step": 10400
    },
    {
      "entropy": 6.259071159362793,
      "epoch": 1.2008078476630122,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048674705170165147,
      "loss": 5.9607,
      "mean_token_accuracy": 0.1750052362680435,
      "num_tokens": 26362989.0,
      "step": 10405
    },
    {
      "entropy": 6.24370231628418,
      "epoch": 1.2013848817080208,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000486733095264151,
      "loss": 5.8552,
      "mean_token_accuracy": 0.18759616315364838,
      "num_tokens": 26376671.0,
      "step": 10410
    },
    {
      "entropy": 6.220479583740234,
      "epoch": 1.2019619157530295,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048671913170511306,
      "loss": 5.947,
      "mean_token_accuracy": 0.1848097637295723,
      "num_tokens": 26389437.0,
      "step": 10415
    },
    {
      "entropy": 6.22413477897644,
      "epoch": 1.202538949798038,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004867051610250073,
      "loss": 5.9054,
      "mean_token_accuracy": 0.17820365875959396,
      "num_tokens": 26401441.0,
      "step": 10420
    },
    {
      "entropy": 6.220470714569092,
      "epoch": 1.2031159838430467,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004866911832243035,
      "loss": 5.9023,
      "mean_token_accuracy": 0.1826564386487007,
      "num_tokens": 26413062.0,
      "step": 10425
    },
    {
      "entropy": 6.26506404876709,
      "epoch": 1.2036930178880554,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048667719830347203,
      "loss": 5.9178,
      "mean_token_accuracy": 0.186004638671875,
      "num_tokens": 26425822.0,
      "step": 10430
    },
    {
      "entropy": 6.231595230102539,
      "epoch": 1.204270051933064,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048666320626298307,
      "loss": 5.9005,
      "mean_token_accuracy": 0.18244679272174835,
      "num_tokens": 26438226.0,
      "step": 10435
    },
    {
      "entropy": 6.244556283950805,
      "epoch": 1.2048470859780727,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048664920710330735,
      "loss": 5.944,
      "mean_token_accuracy": 0.18163639754056932,
      "num_tokens": 26450851.0,
      "step": 10440
    },
    {
      "entropy": 6.342823171615601,
      "epoch": 1.2054241200230813,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048663520082491564,
      "loss": 6.0261,
      "mean_token_accuracy": 0.1734617084264755,
      "num_tokens": 26463292.0,
      "step": 10445
    },
    {
      "entropy": 6.288365459442138,
      "epoch": 1.20600115406809,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004866211874282791,
      "loss": 5.9855,
      "mean_token_accuracy": 0.17760066092014312,
      "num_tokens": 26476731.0,
      "step": 10450
    },
    {
      "entropy": 6.204863262176514,
      "epoch": 1.2065781881130986,
      "grad_norm": 0.875,
      "learning_rate": 0.000486607166913869,
      "loss": 5.9154,
      "mean_token_accuracy": 0.1823340192437172,
      "num_tokens": 26489295.0,
      "step": 10455
    },
    {
      "entropy": 6.265032958984375,
      "epoch": 1.2071552221581072,
      "grad_norm": 0.8203125,
      "learning_rate": 0.00048659313928215705,
      "loss": 5.9328,
      "mean_token_accuracy": 0.1746147707104683,
      "num_tokens": 26502729.0,
      "step": 10460
    },
    {
      "entropy": 6.255537033081055,
      "epoch": 1.207732256203116,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004865791045336149,
      "loss": 5.905,
      "mean_token_accuracy": 0.1834136500954628,
      "num_tokens": 26517442.0,
      "step": 10465
    },
    {
      "entropy": 6.270496034622193,
      "epoch": 1.2083092902481247,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004865650626687146,
      "loss": 5.9024,
      "mean_token_accuracy": 0.17524499744176864,
      "num_tokens": 26530980.0,
      "step": 10470
    },
    {
      "entropy": 6.211662769317627,
      "epoch": 1.2088863242931334,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048655101368792866,
      "loss": 5.9218,
      "mean_token_accuracy": 0.17961404323577881,
      "num_tokens": 26542143.0,
      "step": 10475
    },
    {
      "entropy": 6.274573040008545,
      "epoch": 1.209463358338142,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004865369575917293,
      "loss": 6.0033,
      "mean_token_accuracy": 0.17178912758827208,
      "num_tokens": 26555193.0,
      "step": 10480
    },
    {
      "entropy": 6.2629749298095705,
      "epoch": 1.2100403923831506,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048652289438058953,
      "loss": 5.9353,
      "mean_token_accuracy": 0.18222780376672745,
      "num_tokens": 26567015.0,
      "step": 10485
    },
    {
      "entropy": 6.2470403671264645,
      "epoch": 1.2106174264281593,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048650882405498226,
      "loss": 5.9635,
      "mean_token_accuracy": 0.1809882938861847,
      "num_tokens": 26579155.0,
      "step": 10490
    },
    {
      "entropy": 6.2330749988555905,
      "epoch": 1.211194460473168,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004864947466153808,
      "loss": 5.9678,
      "mean_token_accuracy": 0.17552987933158876,
      "num_tokens": 26592552.0,
      "step": 10495
    },
    {
      "entropy": 6.286801958084107,
      "epoch": 1.2117714945181766,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004864806620622585,
      "loss": 5.9867,
      "mean_token_accuracy": 0.17610828280448915,
      "num_tokens": 26604820.0,
      "step": 10500
    },
    {
      "entropy": 6.262265825271607,
      "epoch": 1.2123485285631852,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004864665703960892,
      "loss": 5.9559,
      "mean_token_accuracy": 0.17873747497797013,
      "num_tokens": 26617218.0,
      "step": 10505
    },
    {
      "entropy": 6.214716386795044,
      "epoch": 1.2129255626081938,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004864524716173469,
      "loss": 5.8617,
      "mean_token_accuracy": 0.17844094038009645,
      "num_tokens": 26628929.0,
      "step": 10510
    },
    {
      "entropy": 6.254156684875488,
      "epoch": 1.2135025966532025,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004864383657265058,
      "loss": 5.9793,
      "mean_token_accuracy": 0.17307288348674774,
      "num_tokens": 26640983.0,
      "step": 10515
    },
    {
      "entropy": 6.268857479095459,
      "epoch": 1.2140796306982111,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00048642425272404016,
      "loss": 5.9718,
      "mean_token_accuracy": 0.17783564031124116,
      "num_tokens": 26653931.0,
      "step": 10520
    },
    {
      "entropy": 6.255046319961548,
      "epoch": 1.2146566647432198,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004864101326104248,
      "loss": 5.9556,
      "mean_token_accuracy": 0.17700446248054505,
      "num_tokens": 26666316.0,
      "step": 10525
    },
    {
      "entropy": 6.311652421951294,
      "epoch": 1.2152336987882286,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004863960053861348,
      "loss": 5.9893,
      "mean_token_accuracy": 0.1785179167985916,
      "num_tokens": 26679818.0,
      "step": 10530
    },
    {
      "entropy": 6.252474546432495,
      "epoch": 1.215810732833237,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048638187105164507,
      "loss": 5.9331,
      "mean_token_accuracy": 0.18640975505113602,
      "num_tokens": 26693553.0,
      "step": 10535
    },
    {
      "entropy": 6.253302955627442,
      "epoch": 1.216387766878246,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004863677296074311,
      "loss": 5.9478,
      "mean_token_accuracy": 0.17934509366750717,
      "num_tokens": 26706037.0,
      "step": 10540
    },
    {
      "entropy": 6.329772233963013,
      "epoch": 1.2169648009232545,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004863535810539686,
      "loss": 5.9784,
      "mean_token_accuracy": 0.17564000189304352,
      "num_tokens": 26719071.0,
      "step": 10545
    },
    {
      "entropy": 6.203491926193237,
      "epoch": 1.2175418349682632,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048633942539173325,
      "loss": 5.8563,
      "mean_token_accuracy": 0.18047413378953933,
      "num_tokens": 26731367.0,
      "step": 10550
    },
    {
      "entropy": 6.22906174659729,
      "epoch": 1.2181188690132718,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048632526262120144,
      "loss": 5.9291,
      "mean_token_accuracy": 0.1793610692024231,
      "num_tokens": 26744945.0,
      "step": 10555
    },
    {
      "entropy": 6.238316583633423,
      "epoch": 1.2186959030582805,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004863110927428493,
      "loss": 5.9891,
      "mean_token_accuracy": 0.17158189862966539,
      "num_tokens": 26758237.0,
      "step": 10560
    },
    {
      "entropy": 6.379085302352905,
      "epoch": 1.219272937103289,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004862969157571536,
      "loss": 5.9938,
      "mean_token_accuracy": 0.17539911419153215,
      "num_tokens": 26770172.0,
      "step": 10565
    },
    {
      "entropy": 6.312456655502319,
      "epoch": 1.2198499711482977,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048628273166459105,
      "loss": 5.9186,
      "mean_token_accuracy": 0.18207830786705018,
      "num_tokens": 26783152.0,
      "step": 10570
    },
    {
      "entropy": 6.186334562301636,
      "epoch": 1.2204270051933064,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048626854046563876,
      "loss": 5.945,
      "mean_token_accuracy": 0.18191892355680467,
      "num_tokens": 26795023.0,
      "step": 10575
    },
    {
      "entropy": 6.285169792175293,
      "epoch": 1.221004039238315,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048625434216077404,
      "loss": 5.9909,
      "mean_token_accuracy": 0.1728364571928978,
      "num_tokens": 26807715.0,
      "step": 10580
    },
    {
      "entropy": 6.313040828704834,
      "epoch": 1.2215810732833237,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048624013675047453,
      "loss": 5.9668,
      "mean_token_accuracy": 0.17192235738039016,
      "num_tokens": 26819181.0,
      "step": 10585
    },
    {
      "entropy": 6.247681713104248,
      "epoch": 1.2221581073283323,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048622592423521783,
      "loss": 5.9288,
      "mean_token_accuracy": 0.17668328583240508,
      "num_tokens": 26831039.0,
      "step": 10590
    },
    {
      "entropy": 6.19183759689331,
      "epoch": 1.222735141373341,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004862117046154822,
      "loss": 5.8446,
      "mean_token_accuracy": 0.17904071509838104,
      "num_tokens": 26843715.0,
      "step": 10595
    },
    {
      "entropy": 6.204564189910888,
      "epoch": 1.2233121754183496,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048619747789174577,
      "loss": 5.8368,
      "mean_token_accuracy": 0.19233134090900422,
      "num_tokens": 26855856.0,
      "step": 10600
    },
    {
      "entropy": 6.258881902694702,
      "epoch": 1.2238892094633584,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004861832440644871,
      "loss": 5.9894,
      "mean_token_accuracy": 0.17488258183002472,
      "num_tokens": 26869368.0,
      "step": 10605
    },
    {
      "entropy": 6.175568771362305,
      "epoch": 1.224466243508367,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048616900313418486,
      "loss": 5.7949,
      "mean_token_accuracy": 0.18326447755098343,
      "num_tokens": 26882230.0,
      "step": 10610
    },
    {
      "entropy": 6.198123741149902,
      "epoch": 1.2250432775533757,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00048615475510131815,
      "loss": 5.9352,
      "mean_token_accuracy": 0.17948734313249587,
      "num_tokens": 26896307.0,
      "step": 10615
    },
    {
      "entropy": 6.278599119186401,
      "epoch": 1.2256203115983844,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048614049996636614,
      "loss": 6.0122,
      "mean_token_accuracy": 0.1749684914946556,
      "num_tokens": 26908759.0,
      "step": 10620
    },
    {
      "entropy": 6.305039787292481,
      "epoch": 1.226197345643393,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048612623772980825,
      "loss": 5.9409,
      "mean_token_accuracy": 0.1781001314520836,
      "num_tokens": 26922179.0,
      "step": 10625
    },
    {
      "entropy": 6.2818022727966305,
      "epoch": 1.2267743796884016,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048611196839212426,
      "loss": 5.9254,
      "mean_token_accuracy": 0.17407275289297103,
      "num_tokens": 26934225.0,
      "step": 10630
    },
    {
      "entropy": 6.200032091140747,
      "epoch": 1.2273514137334103,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000486097691953794,
      "loss": 5.9343,
      "mean_token_accuracy": 0.17591104358434678,
      "num_tokens": 26946644.0,
      "step": 10635
    },
    {
      "entropy": 6.283519649505616,
      "epoch": 1.227928447778419,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00048608340841529784,
      "loss": 5.9938,
      "mean_token_accuracy": 0.17831356674432755,
      "num_tokens": 26959504.0,
      "step": 10640
    },
    {
      "entropy": 6.276490831375122,
      "epoch": 1.2285054818234276,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000486069117777116,
      "loss": 5.9452,
      "mean_token_accuracy": 0.18053647875785828,
      "num_tokens": 26972274.0,
      "step": 10645
    },
    {
      "entropy": 6.258952665328979,
      "epoch": 1.2290825158684362,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004860548200397293,
      "loss": 5.9859,
      "mean_token_accuracy": 0.1751509740948677,
      "num_tokens": 26985800.0,
      "step": 10650
    },
    {
      "entropy": 6.334016227722168,
      "epoch": 1.2296595499134448,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048604051520361846,
      "loss": 5.9667,
      "mean_token_accuracy": 0.1742839276790619,
      "num_tokens": 26999058.0,
      "step": 10655
    },
    {
      "entropy": 6.243470478057861,
      "epoch": 1.2302365839584535,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048602620326926484,
      "loss": 5.9979,
      "mean_token_accuracy": 0.1792060136795044,
      "num_tokens": 27010142.0,
      "step": 10660
    },
    {
      "entropy": 6.227412986755371,
      "epoch": 1.2308136180034621,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048601188423714965,
      "loss": 5.8657,
      "mean_token_accuracy": 0.1805470257997513,
      "num_tokens": 27021547.0,
      "step": 10665
    },
    {
      "entropy": 6.315042877197266,
      "epoch": 1.231390652048471,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004859975581077545,
      "loss": 5.9662,
      "mean_token_accuracy": 0.18034473657608033,
      "num_tokens": 27034817.0,
      "step": 10670
    },
    {
      "entropy": 6.235047245025635,
      "epoch": 1.2319676860934794,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004859832248815614,
      "loss": 5.8851,
      "mean_token_accuracy": 0.18121038526296615,
      "num_tokens": 27047567.0,
      "step": 10675
    },
    {
      "entropy": 6.290113353729248,
      "epoch": 1.2325447201384883,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004859688845590522,
      "loss": 6.0442,
      "mean_token_accuracy": 0.17428648173809053,
      "num_tokens": 27060862.0,
      "step": 10680
    },
    {
      "entropy": 6.2551192283630375,
      "epoch": 1.233121754183497,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048595453714070944,
      "loss": 5.9642,
      "mean_token_accuracy": 0.17977361679077147,
      "num_tokens": 27073216.0,
      "step": 10685
    },
    {
      "entropy": 6.265455961227417,
      "epoch": 1.2336987882285055,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004859401826270156,
      "loss": 5.9053,
      "mean_token_accuracy": 0.17931997925043106,
      "num_tokens": 27085640.0,
      "step": 10690
    },
    {
      "entropy": 6.252834272384644,
      "epoch": 1.2342758222735142,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004859258210184536,
      "loss": 5.9758,
      "mean_token_accuracy": 0.1726067617535591,
      "num_tokens": 27097814.0,
      "step": 10695
    },
    {
      "entropy": 6.272702217102051,
      "epoch": 1.2348528563185228,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048591145231550623,
      "loss": 5.9918,
      "mean_token_accuracy": 0.1744575873017311,
      "num_tokens": 27110449.0,
      "step": 10700
    },
    {
      "entropy": 6.342760944366455,
      "epoch": 1.2354298903635315,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048589707651865697,
      "loss": 6.0019,
      "mean_token_accuracy": 0.17554773837327958,
      "num_tokens": 27123035.0,
      "step": 10705
    },
    {
      "entropy": 6.211860466003418,
      "epoch": 1.23600692440854,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004858826936283893,
      "loss": 5.8947,
      "mean_token_accuracy": 0.1764305353164673,
      "num_tokens": 27134794.0,
      "step": 10710
    },
    {
      "entropy": 6.274847316741943,
      "epoch": 1.2365839584535487,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000485868303645187,
      "loss": 5.9224,
      "mean_token_accuracy": 0.1810566082596779,
      "num_tokens": 27147492.0,
      "step": 10715
    },
    {
      "entropy": 6.254164600372315,
      "epoch": 1.2371609924985574,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048585390656953397,
      "loss": 5.978,
      "mean_token_accuracy": 0.17591814398765565,
      "num_tokens": 27159453.0,
      "step": 10720
    },
    {
      "entropy": 6.2425555229187015,
      "epoch": 1.237738026543566,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004858395024019147,
      "loss": 5.8714,
      "mean_token_accuracy": 0.18373606503009796,
      "num_tokens": 27175021.0,
      "step": 10725
    },
    {
      "entropy": 6.244323635101319,
      "epoch": 1.2383150605885747,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004858250911428133,
      "loss": 5.9083,
      "mean_token_accuracy": 0.18136540204286575,
      "num_tokens": 27187612.0,
      "step": 10730
    },
    {
      "entropy": 6.2558492660522464,
      "epoch": 1.2388920946335833,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004858106727927148,
      "loss": 5.8608,
      "mean_token_accuracy": 0.1807017669081688,
      "num_tokens": 27201045.0,
      "step": 10735
    },
    {
      "entropy": 6.284665250778199,
      "epoch": 1.239469128678592,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000485796247352104,
      "loss": 5.9948,
      "mean_token_accuracy": 0.1722082167863846,
      "num_tokens": 27213934.0,
      "step": 10740
    },
    {
      "entropy": 6.2472731590271,
      "epoch": 1.2400461627236008,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004857818148214662,
      "loss": 5.877,
      "mean_token_accuracy": 0.18405883610248566,
      "num_tokens": 27225695.0,
      "step": 10745
    },
    {
      "entropy": 6.2338207244873045,
      "epoch": 1.2406231967686094,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004857673752012866,
      "loss": 5.9229,
      "mean_token_accuracy": 0.17842336595058442,
      "num_tokens": 27238065.0,
      "step": 10750
    },
    {
      "entropy": 6.258692789077759,
      "epoch": 1.241200230813618,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00048575292849205114,
      "loss": 5.9237,
      "mean_token_accuracy": 0.17604467868804932,
      "num_tokens": 27250334.0,
      "step": 10755
    },
    {
      "entropy": 6.309586238861084,
      "epoch": 1.2417772648586267,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004857384746942456,
      "loss": 5.8988,
      "mean_token_accuracy": 0.18633213639259338,
      "num_tokens": 27262397.0,
      "step": 10760
    },
    {
      "entropy": 6.289945459365844,
      "epoch": 1.2423542989036354,
      "grad_norm": 0.875,
      "learning_rate": 0.0004857240138083562,
      "loss": 5.9365,
      "mean_token_accuracy": 0.17345996499061583,
      "num_tokens": 27274691.0,
      "step": 10765
    },
    {
      "entropy": 6.308668804168701,
      "epoch": 1.242931332948644,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004857095458348692,
      "loss": 5.9461,
      "mean_token_accuracy": 0.17695192098617554,
      "num_tokens": 27286970.0,
      "step": 10770
    },
    {
      "entropy": 6.153140926361084,
      "epoch": 1.2435083669936526,
      "grad_norm": 0.8203125,
      "learning_rate": 0.00048569507077427134,
      "loss": 5.7925,
      "mean_token_accuracy": 0.18906732350587846,
      "num_tokens": 27299788.0,
      "step": 10775
    },
    {
      "entropy": 6.22576937675476,
      "epoch": 1.2440854010386613,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004856805886270494,
      "loss": 5.8144,
      "mean_token_accuracy": 0.18642763644456864,
      "num_tokens": 27312040.0,
      "step": 10780
    },
    {
      "entropy": 6.322715330123901,
      "epoch": 1.24466243508367,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004856660993936905,
      "loss": 6.0647,
      "mean_token_accuracy": 0.1726178914308548,
      "num_tokens": 27325583.0,
      "step": 10785
    },
    {
      "entropy": 6.243177461624145,
      "epoch": 1.2452394691286786,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000485651603074682,
      "loss": 5.9373,
      "mean_token_accuracy": 0.17466154545545579,
      "num_tokens": 27337525.0,
      "step": 10790
    },
    {
      "entropy": 6.289048385620117,
      "epoch": 1.2458165031736872,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004856370996705115,
      "loss": 5.955,
      "mean_token_accuracy": 0.17251873314380645,
      "num_tokens": 27351368.0,
      "step": 10795
    },
    {
      "entropy": 6.30572543144226,
      "epoch": 1.2463935372186958,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004856225891816667,
      "loss": 5.9329,
      "mean_token_accuracy": 0.1764160618185997,
      "num_tokens": 27364497.0,
      "step": 10800
    },
    {
      "entropy": 6.2650816440582275,
      "epoch": 1.2469705712637045,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004856080716086358,
      "loss": 5.9098,
      "mean_token_accuracy": 0.17892101109027864,
      "num_tokens": 27377401.0,
      "step": 10805
    },
    {
      "entropy": 6.234683847427368,
      "epoch": 1.2475476053087133,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000485593546951907,
      "loss": 5.8836,
      "mean_token_accuracy": 0.1817560687661171,
      "num_tokens": 27389712.0,
      "step": 10810
    },
    {
      "entropy": 6.220177221298218,
      "epoch": 1.2481246393537218,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004855790152119688,
      "loss": 5.9099,
      "mean_token_accuracy": 0.1806002989411354,
      "num_tokens": 27403193.0,
      "step": 10815
    },
    {
      "entropy": 6.273285913467407,
      "epoch": 1.2487016733987306,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004855644763893102,
      "loss": 5.9624,
      "mean_token_accuracy": 0.1728382170200348,
      "num_tokens": 27415648.0,
      "step": 10820
    },
    {
      "entropy": 6.288532447814942,
      "epoch": 1.2492787074437393,
      "grad_norm": 0.875,
      "learning_rate": 0.00048554993048441987,
      "loss": 5.923,
      "mean_token_accuracy": 0.17991327494382858,
      "num_tokens": 27428825.0,
      "step": 10825
    },
    {
      "entropy": 6.268614768981934,
      "epoch": 1.249855741488748,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004855353774977873,
      "loss": 5.9259,
      "mean_token_accuracy": 0.1890061914920807,
      "num_tokens": 27441867.0,
      "step": 10830
    },
    {
      "entropy": 6.238099431991577,
      "epoch": 1.2504327755337565,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048552081742990184,
      "loss": 5.9305,
      "mean_token_accuracy": 0.17760562896728516,
      "num_tokens": 27454688.0,
      "step": 10835
    },
    {
      "entropy": 6.2731317520141605,
      "epoch": 1.2510098095787652,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048550625028125327,
      "loss": 6.0405,
      "mean_token_accuracy": 0.17051285952329637,
      "num_tokens": 27466585.0,
      "step": 10840
    },
    {
      "entropy": 6.293200397491455,
      "epoch": 1.2515868436237738,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004854916760523315,
      "loss": 5.9596,
      "mean_token_accuracy": 0.1799686074256897,
      "num_tokens": 27479111.0,
      "step": 10845
    },
    {
      "entropy": 6.257088565826416,
      "epoch": 1.2521638776687825,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048547709474362684,
      "loss": 5.8726,
      "mean_token_accuracy": 0.18254064321517943,
      "num_tokens": 27491359.0,
      "step": 10850
    },
    {
      "entropy": 6.252914571762085,
      "epoch": 1.252740911713791,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004854625063556296,
      "loss": 5.9206,
      "mean_token_accuracy": 0.17930027842521667,
      "num_tokens": 27504609.0,
      "step": 10855
    },
    {
      "entropy": 6.275938367843628,
      "epoch": 1.2533179457587997,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004854479108888305,
      "loss": 5.9204,
      "mean_token_accuracy": 0.1805565819144249,
      "num_tokens": 27517270.0,
      "step": 10860
    },
    {
      "entropy": 6.278730297088623,
      "epoch": 1.2538949798038084,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048543330834372047,
      "loss": 5.8923,
      "mean_token_accuracy": 0.17586951404809953,
      "num_tokens": 27530019.0,
      "step": 10865
    },
    {
      "entropy": 6.2833507537841795,
      "epoch": 1.254472013848817,
      "grad_norm": 0.81640625,
      "learning_rate": 0.00048541869872079064,
      "loss": 5.9811,
      "mean_token_accuracy": 0.17643605023622513,
      "num_tokens": 27542091.0,
      "step": 10870
    },
    {
      "entropy": 6.290549850463867,
      "epoch": 1.2550490478938257,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004854040820205324,
      "loss": 5.9995,
      "mean_token_accuracy": 0.17601939141750336,
      "num_tokens": 27554326.0,
      "step": 10875
    },
    {
      "entropy": 6.243378305435181,
      "epoch": 1.2556260819388343,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004853894582434373,
      "loss": 5.9981,
      "mean_token_accuracy": 0.1798421934247017,
      "num_tokens": 27567242.0,
      "step": 10880
    },
    {
      "entropy": 6.264650630950928,
      "epoch": 1.2562031159838432,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004853748273899974,
      "loss": 5.8837,
      "mean_token_accuracy": 0.18467852622270584,
      "num_tokens": 27579595.0,
      "step": 10885
    },
    {
      "entropy": 6.2654726028442385,
      "epoch": 1.2567801500288516,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004853601894607046,
      "loss": 5.9152,
      "mean_token_accuracy": 0.1795614629983902,
      "num_tokens": 27592694.0,
      "step": 10890
    },
    {
      "entropy": 6.226495456695557,
      "epoch": 1.2573571840738604,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004853455444560514,
      "loss": 5.8468,
      "mean_token_accuracy": 0.18729524165391923,
      "num_tokens": 27605532.0,
      "step": 10895
    },
    {
      "entropy": 6.22318925857544,
      "epoch": 1.257934218118869,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004853308923765302,
      "loss": 5.936,
      "mean_token_accuracy": 0.1763066902756691,
      "num_tokens": 27618394.0,
      "step": 10900
    },
    {
      "entropy": 6.259911680221558,
      "epoch": 1.2585112521638777,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000485316233222634,
      "loss": 5.9305,
      "mean_token_accuracy": 0.1735590636730194,
      "num_tokens": 27631293.0,
      "step": 10905
    },
    {
      "entropy": 6.275124645233154,
      "epoch": 1.2590882862088864,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004853015669948557,
      "loss": 5.9393,
      "mean_token_accuracy": 0.18006334453821182,
      "num_tokens": 27643972.0,
      "step": 10910
    },
    {
      "entropy": 6.174040031433106,
      "epoch": 1.259665320253895,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048528689369368863,
      "loss": 5.8363,
      "mean_token_accuracy": 0.18569005280733109,
      "num_tokens": 27656521.0,
      "step": 10915
    },
    {
      "entropy": 6.219848489761352,
      "epoch": 1.2602423542989036,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004852722133196264,
      "loss": 5.9161,
      "mean_token_accuracy": 0.17564667016267776,
      "num_tokens": 27670108.0,
      "step": 10920
    },
    {
      "entropy": 6.182351541519165,
      "epoch": 1.2608193883439123,
      "grad_norm": 1.125,
      "learning_rate": 0.0004852575258731627,
      "loss": 5.8366,
      "mean_token_accuracy": 0.19582096636295318,
      "num_tokens": 27683528.0,
      "step": 10925
    },
    {
      "entropy": 6.152352237701416,
      "epoch": 1.261396422388921,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004852428313547916,
      "loss": 5.7961,
      "mean_token_accuracy": 0.18625763207674026,
      "num_tokens": 27696172.0,
      "step": 10930
    },
    {
      "entropy": 6.15151572227478,
      "epoch": 1.2619734564339296,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048522812976500726,
      "loss": 5.8821,
      "mean_token_accuracy": 0.18320987075567247,
      "num_tokens": 27709514.0,
      "step": 10935
    },
    {
      "entropy": 6.264788818359375,
      "epoch": 1.2625504904789382,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048521342110430417,
      "loss": 5.9165,
      "mean_token_accuracy": 0.1847102239727974,
      "num_tokens": 27721743.0,
      "step": 10940
    },
    {
      "entropy": 6.220901203155518,
      "epoch": 1.2631275245239468,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048519870537317713,
      "loss": 5.9156,
      "mean_token_accuracy": 0.1821763515472412,
      "num_tokens": 27733802.0,
      "step": 10945
    },
    {
      "entropy": 6.246953773498535,
      "epoch": 1.2637045585689557,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00048518398257212103,
      "loss": 5.9518,
      "mean_token_accuracy": 0.18182697743177414,
      "num_tokens": 27746099.0,
      "step": 10950
    },
    {
      "entropy": 6.2778009414672855,
      "epoch": 1.2642815926139641,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004851692527016311,
      "loss": 5.9478,
      "mean_token_accuracy": 0.1744722381234169,
      "num_tokens": 27758334.0,
      "step": 10955
    },
    {
      "entropy": 6.296640300750733,
      "epoch": 1.264858626658973,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004851545157622027,
      "loss": 5.9157,
      "mean_token_accuracy": 0.17709164470434188,
      "num_tokens": 27770544.0,
      "step": 10960
    },
    {
      "entropy": 6.233324575424194,
      "epoch": 1.2654356607039816,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004851397717543316,
      "loss": 5.9101,
      "mean_token_accuracy": 0.1879000186920166,
      "num_tokens": 27783328.0,
      "step": 10965
    },
    {
      "entropy": 6.1756078720092775,
      "epoch": 1.2660126947489903,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004851250206785137,
      "loss": 5.9297,
      "mean_token_accuracy": 0.18229101747274398,
      "num_tokens": 27796771.0,
      "step": 10970
    },
    {
      "entropy": 6.259104633331299,
      "epoch": 1.266589728793999,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048511026253524503,
      "loss": 5.8745,
      "mean_token_accuracy": 0.18808848261833191,
      "num_tokens": 27809718.0,
      "step": 10975
    },
    {
      "entropy": 6.2750242233276365,
      "epoch": 1.2671667628390075,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004850954973250221,
      "loss": 5.9304,
      "mean_token_accuracy": 0.18131026923656463,
      "num_tokens": 27822750.0,
      "step": 10980
    },
    {
      "entropy": 6.143756961822509,
      "epoch": 1.2677437968840162,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048508072504834144,
      "loss": 5.9048,
      "mean_token_accuracy": 0.18114064484834672,
      "num_tokens": 27834783.0,
      "step": 10985
    },
    {
      "entropy": 6.252334308624268,
      "epoch": 1.2683208309290248,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048506594570569997,
      "loss": 5.9196,
      "mean_token_accuracy": 0.18419783264398576,
      "num_tokens": 27848355.0,
      "step": 10990
    },
    {
      "entropy": 6.2092420101165775,
      "epoch": 1.2688978649740335,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004850511592975947,
      "loss": 5.9262,
      "mean_token_accuracy": 0.18010048419237137,
      "num_tokens": 27859924.0,
      "step": 10995
    },
    {
      "entropy": 6.2110199451446535,
      "epoch": 1.269474899019042,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004850363658245231,
      "loss": 5.931,
      "mean_token_accuracy": 0.17980812340974808,
      "num_tokens": 27872186.0,
      "step": 11000
    },
    {
      "entropy": 6.280411291122436,
      "epoch": 1.2700519330640507,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004850215652869826,
      "loss": 5.9058,
      "mean_token_accuracy": 0.1814272880554199,
      "num_tokens": 27884512.0,
      "step": 11005
    },
    {
      "entropy": 6.299659395217896,
      "epoch": 1.2706289671090594,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004850067576854711,
      "loss": 5.942,
      "mean_token_accuracy": 0.1811836063861847,
      "num_tokens": 27898665.0,
      "step": 11010
    },
    {
      "entropy": 6.211774253845215,
      "epoch": 1.271206001154068,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004849919430204867,
      "loss": 5.8837,
      "mean_token_accuracy": 0.17753110080957413,
      "num_tokens": 27911198.0,
      "step": 11015
    },
    {
      "entropy": 6.193430089950562,
      "epoch": 1.2717830351990767,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004849771212925275,
      "loss": 5.9297,
      "mean_token_accuracy": 0.18194724917411803,
      "num_tokens": 27923677.0,
      "step": 11020
    },
    {
      "entropy": 6.183107089996338,
      "epoch": 1.2723600692440855,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048496229250209223,
      "loss": 5.8541,
      "mean_token_accuracy": 0.18819677382707595,
      "num_tokens": 27936269.0,
      "step": 11025
    },
    {
      "entropy": 6.271490049362183,
      "epoch": 1.272937103289094,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004849474566496795,
      "loss": 5.9791,
      "mean_token_accuracy": 0.182184799015522,
      "num_tokens": 27950463.0,
      "step": 11030
    },
    {
      "entropy": 6.240012836456299,
      "epoch": 1.2735141373341028,
      "grad_norm": 0.875,
      "learning_rate": 0.0004849326137357883,
      "loss": 5.9191,
      "mean_token_accuracy": 0.1788981929421425,
      "num_tokens": 27961948.0,
      "step": 11035
    },
    {
      "entropy": 6.258887386322021,
      "epoch": 1.2740911713791114,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000484917763760918,
      "loss": 5.9358,
      "mean_token_accuracy": 0.17902835607528686,
      "num_tokens": 27974075.0,
      "step": 11040
    },
    {
      "entropy": 6.351895093917847,
      "epoch": 1.27466820542412,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048490290672556784,
      "loss": 5.9836,
      "mean_token_accuracy": 0.17185096591711044,
      "num_tokens": 27987587.0,
      "step": 11045
    },
    {
      "entropy": 6.32787299156189,
      "epoch": 1.2752452394691287,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004848880426302378,
      "loss": 5.9565,
      "mean_token_accuracy": 0.1805362105369568,
      "num_tokens": 27999438.0,
      "step": 11050
    },
    {
      "entropy": 6.195203351974487,
      "epoch": 1.2758222735141374,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004848731714754277,
      "loss": 5.8534,
      "mean_token_accuracy": 0.1904847040772438,
      "num_tokens": 28012646.0,
      "step": 11055
    },
    {
      "entropy": 6.29269962310791,
      "epoch": 1.276399307559146,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004848582932616377,
      "loss": 6.0157,
      "mean_token_accuracy": 0.1755513846874237,
      "num_tokens": 28025133.0,
      "step": 11060
    },
    {
      "entropy": 6.287973022460937,
      "epoch": 1.2769763416041546,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004848434079893682,
      "loss": 5.9566,
      "mean_token_accuracy": 0.1762356460094452,
      "num_tokens": 28038062.0,
      "step": 11065
    },
    {
      "entropy": 6.3084269046783445,
      "epoch": 1.2775533756491633,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004848285156591201,
      "loss": 5.9849,
      "mean_token_accuracy": 0.17715671360492707,
      "num_tokens": 28050617.0,
      "step": 11070
    },
    {
      "entropy": 6.323256492614746,
      "epoch": 1.278130409694172,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048481361627139384,
      "loss": 6.0414,
      "mean_token_accuracy": 0.17031400352716447,
      "num_tokens": 28063692.0,
      "step": 11075
    },
    {
      "entropy": 6.296440744400025,
      "epoch": 1.2787074437391805,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048479870982669096,
      "loss": 5.9074,
      "mean_token_accuracy": 0.18544832915067672,
      "num_tokens": 28075628.0,
      "step": 11080
    },
    {
      "entropy": 6.234014177322388,
      "epoch": 1.2792844777841892,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048478379632551266,
      "loss": 5.9174,
      "mean_token_accuracy": 0.180943500995636,
      "num_tokens": 28087998.0,
      "step": 11085
    },
    {
      "entropy": 6.307333469390869,
      "epoch": 1.279861511829198,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004847688757683606,
      "loss": 6.0224,
      "mean_token_accuracy": 0.17671932131052018,
      "num_tokens": 28101260.0,
      "step": 11090
    },
    {
      "entropy": 6.303109073638916,
      "epoch": 1.2804385458742065,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004847539481557366,
      "loss": 5.9926,
      "mean_token_accuracy": 0.1822233498096466,
      "num_tokens": 28115077.0,
      "step": 11095
    },
    {
      "entropy": 6.266627597808838,
      "epoch": 1.2810155799192153,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048473901348814277,
      "loss": 5.9217,
      "mean_token_accuracy": 0.18121051490306855,
      "num_tokens": 28126661.0,
      "step": 11100
    },
    {
      "entropy": 6.239809274673462,
      "epoch": 1.281592613964224,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004847240717660814,
      "loss": 5.9857,
      "mean_token_accuracy": 0.17517693489789962,
      "num_tokens": 28140330.0,
      "step": 11105
    },
    {
      "entropy": 6.327423000335694,
      "epoch": 1.2821696480092326,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048470912299005493,
      "loss": 5.9408,
      "mean_token_accuracy": 0.17901820093393325,
      "num_tokens": 28153411.0,
      "step": 11110
    },
    {
      "entropy": 6.191156911849975,
      "epoch": 1.2827466820542412,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048469416716056635,
      "loss": 5.8422,
      "mean_token_accuracy": 0.18756602108478546,
      "num_tokens": 28165812.0,
      "step": 11115
    },
    {
      "entropy": 6.226385879516601,
      "epoch": 1.28332371609925,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004846792042781187,
      "loss": 5.9471,
      "mean_token_accuracy": 0.1808479607105255,
      "num_tokens": 28177702.0,
      "step": 11120
    },
    {
      "entropy": 6.226623916625977,
      "epoch": 1.2839007501442585,
      "grad_norm": 0.8203125,
      "learning_rate": 0.00048466423434321513,
      "loss": 5.8516,
      "mean_token_accuracy": 0.18334298580884933,
      "num_tokens": 28190712.0,
      "step": 11125
    },
    {
      "entropy": 6.325730323791504,
      "epoch": 1.2844777841892672,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048464925735635907,
      "loss": 5.9283,
      "mean_token_accuracy": 0.17690158188343047,
      "num_tokens": 28204410.0,
      "step": 11130
    },
    {
      "entropy": 6.317542886734008,
      "epoch": 1.2850548182342758,
      "grad_norm": 0.828125,
      "learning_rate": 0.00048463427331805445,
      "loss": 5.9749,
      "mean_token_accuracy": 0.1725844293832779,
      "num_tokens": 28217739.0,
      "step": 11135
    },
    {
      "entropy": 6.217558097839356,
      "epoch": 1.2856318522792844,
      "grad_norm": 0.875,
      "learning_rate": 0.0004846192822288052,
      "loss": 5.9108,
      "mean_token_accuracy": 0.18078666925430298,
      "num_tokens": 28230335.0,
      "step": 11140
    },
    {
      "entropy": 6.271143627166748,
      "epoch": 1.286208886324293,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004846042840891155,
      "loss": 5.935,
      "mean_token_accuracy": 0.17822139263153075,
      "num_tokens": 28243384.0,
      "step": 11145
    },
    {
      "entropy": 6.271648836135864,
      "epoch": 1.2867859203693017,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004845892788994899,
      "loss": 5.9132,
      "mean_token_accuracy": 0.18005290031433105,
      "num_tokens": 28254963.0,
      "step": 11150
    },
    {
      "entropy": 6.2827846050262455,
      "epoch": 1.2873629544143104,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004845742666604329,
      "loss": 5.9329,
      "mean_token_accuracy": 0.17585408240556716,
      "num_tokens": 28267320.0,
      "step": 11155
    },
    {
      "entropy": 6.285551738739014,
      "epoch": 1.287939988459319,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048455924737244953,
      "loss": 5.9349,
      "mean_token_accuracy": 0.1803737461566925,
      "num_tokens": 28279631.0,
      "step": 11160
    },
    {
      "entropy": 6.22711672782898,
      "epoch": 1.2885170225043279,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048454422103604505,
      "loss": 5.937,
      "mean_token_accuracy": 0.17457396388053895,
      "num_tokens": 28292356.0,
      "step": 11165
    },
    {
      "entropy": 6.2069940090179445,
      "epoch": 1.2890940565493363,
      "grad_norm": 1.25,
      "learning_rate": 0.0004845291876517247,
      "loss": 5.9427,
      "mean_token_accuracy": 0.1823141649365425,
      "num_tokens": 28306714.0,
      "step": 11170
    },
    {
      "entropy": 6.282351922988892,
      "epoch": 1.2896710905943451,
      "grad_norm": 0.875,
      "learning_rate": 0.00048451414721999424,
      "loss": 5.9322,
      "mean_token_accuracy": 0.1765112727880478,
      "num_tokens": 28319851.0,
      "step": 11175
    },
    {
      "entropy": 6.270438718795776,
      "epoch": 1.2902481246393538,
      "grad_norm": 0.828125,
      "learning_rate": 0.00048449909974135954,
      "loss": 5.9479,
      "mean_token_accuracy": 0.1769252136349678,
      "num_tokens": 28333048.0,
      "step": 11180
    },
    {
      "entropy": 6.221937274932861,
      "epoch": 1.2908251586843624,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004844840452163267,
      "loss": 5.8819,
      "mean_token_accuracy": 0.18255432099103927,
      "num_tokens": 28345215.0,
      "step": 11185
    },
    {
      "entropy": 6.173029613494873,
      "epoch": 1.291402192729371,
      "grad_norm": 0.90625,
      "learning_rate": 0.000484468983645402,
      "loss": 5.8402,
      "mean_token_accuracy": 0.1830596834421158,
      "num_tokens": 28356594.0,
      "step": 11190
    },
    {
      "entropy": 6.353919935226441,
      "epoch": 1.2919792267743797,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00048445391502909213,
      "loss": 5.9628,
      "mean_token_accuracy": 0.17117386758327485,
      "num_tokens": 28368728.0,
      "step": 11195
    },
    {
      "entropy": 6.314934778213501,
      "epoch": 1.2925562608193883,
      "grad_norm": 0.828125,
      "learning_rate": 0.00048443883936790386,
      "loss": 5.9534,
      "mean_token_accuracy": 0.17527435570955277,
      "num_tokens": 28382290.0,
      "step": 11200
    },
    {
      "entropy": 6.231500196456909,
      "epoch": 1.293133294864397,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048442375666234427,
      "loss": 5.908,
      "mean_token_accuracy": 0.18398450762033464,
      "num_tokens": 28394687.0,
      "step": 11205
    },
    {
      "entropy": 6.272225284576416,
      "epoch": 1.2937103289094056,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004844086669129206,
      "loss": 5.9707,
      "mean_token_accuracy": 0.17583649158477782,
      "num_tokens": 28407187.0,
      "step": 11210
    },
    {
      "entropy": 6.307099151611328,
      "epoch": 1.2942873629544143,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048439357012014045,
      "loss": 5.9863,
      "mean_token_accuracy": 0.17240729182958603,
      "num_tokens": 28419476.0,
      "step": 11215
    },
    {
      "entropy": 6.2180705070495605,
      "epoch": 1.294864396999423,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004843784662845116,
      "loss": 5.8619,
      "mean_token_accuracy": 0.18614352643489837,
      "num_tokens": 28433952.0,
      "step": 11220
    },
    {
      "entropy": 6.266728353500366,
      "epoch": 1.2954414310444315,
      "grad_norm": 0.953125,
      "learning_rate": 0.000484363355406542,
      "loss": 5.9587,
      "mean_token_accuracy": 0.1829248696565628,
      "num_tokens": 28445687.0,
      "step": 11225
    },
    {
      "entropy": 6.239110612869263,
      "epoch": 1.2960184650894404,
      "grad_norm": 0.8203125,
      "learning_rate": 0.00048434823748674,
      "loss": 5.895,
      "mean_token_accuracy": 0.17440749406814576,
      "num_tokens": 28458883.0,
      "step": 11230
    },
    {
      "entropy": 6.225468158721924,
      "epoch": 1.2965954991344488,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048433311252561403,
      "loss": 5.9128,
      "mean_token_accuracy": 0.17584753930568695,
      "num_tokens": 28471850.0,
      "step": 11235
    },
    {
      "entropy": 6.257522249221802,
      "epoch": 1.2971725331794577,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004843179805236728,
      "loss": 5.9568,
      "mean_token_accuracy": 0.1819494977593422,
      "num_tokens": 28483640.0,
      "step": 11240
    },
    {
      "entropy": 6.301819753646851,
      "epoch": 1.297749567224466,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00048430284148142516,
      "loss": 5.9648,
      "mean_token_accuracy": 0.17864782810211183,
      "num_tokens": 28497108.0,
      "step": 11245
    },
    {
      "entropy": 6.305349063873291,
      "epoch": 1.298326601269475,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004842876953993805,
      "loss": 5.9014,
      "mean_token_accuracy": 0.18031202405691146,
      "num_tokens": 28509071.0,
      "step": 11250
    },
    {
      "entropy": 6.268391466140747,
      "epoch": 1.2989036353144836,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004842725422780482,
      "loss": 5.8818,
      "mean_token_accuracy": 0.17940201461315156,
      "num_tokens": 28521822.0,
      "step": 11255
    },
    {
      "entropy": 6.333021831512451,
      "epoch": 1.2994806693594922,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004842573821179378,
      "loss": 6.0722,
      "mean_token_accuracy": 0.16750324070453643,
      "num_tokens": 28535809.0,
      "step": 11260
    },
    {
      "entropy": 6.191955423355102,
      "epoch": 1.3000577034045009,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004842422149195593,
      "loss": 5.8582,
      "mean_token_accuracy": 0.18359729945659636,
      "num_tokens": 28549397.0,
      "step": 11265
    },
    {
      "entropy": 6.292188835144043,
      "epoch": 1.3006347374495095,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048422704068342294,
      "loss": 5.8838,
      "mean_token_accuracy": 0.1853783443570137,
      "num_tokens": 28562490.0,
      "step": 11270
    },
    {
      "entropy": 6.246140241622925,
      "epoch": 1.3012117714945182,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004842118594100389,
      "loss": 5.9263,
      "mean_token_accuracy": 0.17879924476146697,
      "num_tokens": 28575039.0,
      "step": 11275
    },
    {
      "entropy": 6.230689954757691,
      "epoch": 1.3017888055395268,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048419667109991793,
      "loss": 5.869,
      "mean_token_accuracy": 0.18305207788944244,
      "num_tokens": 28587889.0,
      "step": 11280
    },
    {
      "entropy": 6.209166383743286,
      "epoch": 1.3023658395845354,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00048418147575357085,
      "loss": 5.8786,
      "mean_token_accuracy": 0.18345620334148408,
      "num_tokens": 28600440.0,
      "step": 11285
    },
    {
      "entropy": 6.279112243652344,
      "epoch": 1.302942873629544,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004841662733715087,
      "loss": 5.9199,
      "mean_token_accuracy": 0.1792471379041672,
      "num_tokens": 28613763.0,
      "step": 11290
    },
    {
      "entropy": 6.281858634948731,
      "epoch": 1.3035199076745527,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048415106395424294,
      "loss": 5.9479,
      "mean_token_accuracy": 0.18175759464502333,
      "num_tokens": 28625768.0,
      "step": 11295
    },
    {
      "entropy": 6.26454725265503,
      "epoch": 1.3040969417195614,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048413584750228494,
      "loss": 5.9353,
      "mean_token_accuracy": 0.17517566531896592,
      "num_tokens": 28638207.0,
      "step": 11300
    },
    {
      "entropy": 6.324590539932251,
      "epoch": 1.3046739757645702,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048412062401614653,
      "loss": 5.995,
      "mean_token_accuracy": 0.17503037452697753,
      "num_tokens": 28650348.0,
      "step": 11305
    },
    {
      "entropy": 6.286007690429687,
      "epoch": 1.3052510098095786,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048410539349634,
      "loss": 5.9438,
      "mean_token_accuracy": 0.18037812411785126,
      "num_tokens": 28663723.0,
      "step": 11310
    },
    {
      "entropy": 6.301861572265625,
      "epoch": 1.3058280438545875,
      "grad_norm": 0.84375,
      "learning_rate": 0.00048409015594337725,
      "loss": 6.0203,
      "mean_token_accuracy": 0.1764570191502571,
      "num_tokens": 28676877.0,
      "step": 11315
    },
    {
      "entropy": 6.3109055519104,
      "epoch": 1.3064050778995961,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000484074911357771,
      "loss": 5.96,
      "mean_token_accuracy": 0.17397731840610503,
      "num_tokens": 28689124.0,
      "step": 11320
    },
    {
      "entropy": 6.287434864044189,
      "epoch": 1.3069821119446048,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048405965974003404,
      "loss": 5.8659,
      "mean_token_accuracy": 0.19104320108890532,
      "num_tokens": 28701576.0,
      "step": 11325
    },
    {
      "entropy": 6.252522039413452,
      "epoch": 1.3075591459896134,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048404440109067927,
      "loss": 5.9361,
      "mean_token_accuracy": 0.17371978610754013,
      "num_tokens": 28715364.0,
      "step": 11330
    },
    {
      "entropy": 6.286492729187012,
      "epoch": 1.308136180034622,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004840291354102198,
      "loss": 5.9043,
      "mean_token_accuracy": 0.1816939204931259,
      "num_tokens": 28727390.0,
      "step": 11335
    },
    {
      "entropy": 6.252400159835815,
      "epoch": 1.3087132140796307,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004840138626991693,
      "loss": 5.9616,
      "mean_token_accuracy": 0.1764710545539856,
      "num_tokens": 28739310.0,
      "step": 11340
    },
    {
      "entropy": 6.308204650878906,
      "epoch": 1.3092902481246393,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004839985829580413,
      "loss": 5.9108,
      "mean_token_accuracy": 0.17781326472759246,
      "num_tokens": 28750517.0,
      "step": 11345
    },
    {
      "entropy": 6.238418865203857,
      "epoch": 1.309867282169648,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048398329618734977,
      "loss": 5.8885,
      "mean_token_accuracy": 0.1751066878437996,
      "num_tokens": 28763195.0,
      "step": 11350
    },
    {
      "entropy": 6.260350465774536,
      "epoch": 1.3104443162146566,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004839680023876089,
      "loss": 5.9859,
      "mean_token_accuracy": 0.18126043677330017,
      "num_tokens": 28776697.0,
      "step": 11355
    },
    {
      "entropy": 6.267036819458008,
      "epoch": 1.3110213502596653,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004839527015593331,
      "loss": 5.8906,
      "mean_token_accuracy": 0.18175694048404695,
      "num_tokens": 28788885.0,
      "step": 11360
    },
    {
      "entropy": 6.252719593048096,
      "epoch": 1.311598384304674,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048393739370303684,
      "loss": 5.9306,
      "mean_token_accuracy": 0.18300748169422149,
      "num_tokens": 28800502.0,
      "step": 11365
    },
    {
      "entropy": 6.253888273239136,
      "epoch": 1.3121754183496828,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004839220788192353,
      "loss": 5.8976,
      "mean_token_accuracy": 0.18338464945554733,
      "num_tokens": 28811985.0,
      "step": 11370
    },
    {
      "entropy": 6.239736890792846,
      "epoch": 1.3127524523946912,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048390675690844335,
      "loss": 5.94,
      "mean_token_accuracy": 0.18607171028852462,
      "num_tokens": 28825492.0,
      "step": 11375
    },
    {
      "entropy": 6.264604473114014,
      "epoch": 1.3133294864397,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004838914279711764,
      "loss": 5.9058,
      "mean_token_accuracy": 0.1811276286840439,
      "num_tokens": 28838889.0,
      "step": 11380
    },
    {
      "entropy": 6.31759090423584,
      "epoch": 1.3139065204847085,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048387609200795003,
      "loss": 5.9669,
      "mean_token_accuracy": 0.1744312524795532,
      "num_tokens": 28851341.0,
      "step": 11385
    },
    {
      "entropy": 6.271655178070068,
      "epoch": 1.3144835545297173,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048386074901928,
      "loss": 5.9294,
      "mean_token_accuracy": 0.1823418691754341,
      "num_tokens": 28862861.0,
      "step": 11390
    },
    {
      "entropy": 6.259878301620484,
      "epoch": 1.315060588574726,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004838453990056825,
      "loss": 5.9781,
      "mean_token_accuracy": 0.1788210690021515,
      "num_tokens": 28876261.0,
      "step": 11395
    },
    {
      "entropy": 6.253640127182007,
      "epoch": 1.3156376226197346,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048383004196767373,
      "loss": 5.9044,
      "mean_token_accuracy": 0.18201138377189635,
      "num_tokens": 28888710.0,
      "step": 11400
    },
    {
      "entropy": 6.226709461212158,
      "epoch": 1.3162146566647432,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004838146779057702,
      "loss": 5.907,
      "mean_token_accuracy": 0.18273279815912247,
      "num_tokens": 28900606.0,
      "step": 11405
    },
    {
      "entropy": 6.284420776367187,
      "epoch": 1.3167916907097519,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004837993068204887,
      "loss": 5.8961,
      "mean_token_accuracy": 0.18355491161346435,
      "num_tokens": 28912979.0,
      "step": 11410
    },
    {
      "entropy": 6.166273880004883,
      "epoch": 1.3173687247547605,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048378392871234634,
      "loss": 5.8684,
      "mean_token_accuracy": 0.18426571488380433,
      "num_tokens": 28927597.0,
      "step": 11415
    },
    {
      "entropy": 6.260102319717407,
      "epoch": 1.3179457587997692,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004837685435818601,
      "loss": 5.8415,
      "mean_token_accuracy": 0.18576819449663162,
      "num_tokens": 28938914.0,
      "step": 11420
    },
    {
      "entropy": 6.237336349487305,
      "epoch": 1.3185227928447778,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004837531514295477,
      "loss": 5.9616,
      "mean_token_accuracy": 0.17073754519224166,
      "num_tokens": 28951347.0,
      "step": 11425
    },
    {
      "entropy": 6.243986034393311,
      "epoch": 1.3190998268897864,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004837377522559267,
      "loss": 5.8933,
      "mean_token_accuracy": 0.18190166652202605,
      "num_tokens": 28964904.0,
      "step": 11430
    },
    {
      "entropy": 6.202196407318115,
      "epoch": 1.319676860934795,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00048372234606151507,
      "loss": 5.7398,
      "mean_token_accuracy": 0.19972495138645172,
      "num_tokens": 28977818.0,
      "step": 11435
    },
    {
      "entropy": 6.209478807449341,
      "epoch": 1.3202538949798037,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048370693284683106,
      "loss": 5.9015,
      "mean_token_accuracy": 0.18241266161203384,
      "num_tokens": 28991521.0,
      "step": 11440
    },
    {
      "entropy": 6.265409088134765,
      "epoch": 1.3208309290248126,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048369151261239303,
      "loss": 5.9561,
      "mean_token_accuracy": 0.173648901283741,
      "num_tokens": 29003339.0,
      "step": 11445
    },
    {
      "entropy": 6.305297374725342,
      "epoch": 1.321407963069821,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004836760853587196,
      "loss": 5.9985,
      "mean_token_accuracy": 0.17931702584028245,
      "num_tokens": 29017037.0,
      "step": 11450
    },
    {
      "entropy": 6.314496660232544,
      "epoch": 1.3219849971148299,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00048366065108632967,
      "loss": 5.8916,
      "mean_token_accuracy": 0.18250546604394913,
      "num_tokens": 29029496.0,
      "step": 11455
    },
    {
      "entropy": 6.156741714477539,
      "epoch": 1.3225620311598385,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00048364520979574246,
      "loss": 5.7652,
      "mean_token_accuracy": 0.1836748242378235,
      "num_tokens": 29042344.0,
      "step": 11460
    },
    {
      "entropy": 6.277041864395142,
      "epoch": 1.3231390652048471,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048362976148747715,
      "loss": 5.9035,
      "mean_token_accuracy": 0.1765417367219925,
      "num_tokens": 29055518.0,
      "step": 11465
    },
    {
      "entropy": 6.249780702590942,
      "epoch": 1.3237160992498558,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004836143061620536,
      "loss": 5.8702,
      "mean_token_accuracy": 0.19003255665302277,
      "num_tokens": 29067109.0,
      "step": 11470
    },
    {
      "entropy": 6.153318309783936,
      "epoch": 1.3242931332948644,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004835988438199914,
      "loss": 5.8012,
      "mean_token_accuracy": 0.19401812702417373,
      "num_tokens": 29079766.0,
      "step": 11475
    },
    {
      "entropy": 6.284279203414917,
      "epoch": 1.324870167339873,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004835833744618107,
      "loss": 5.9475,
      "mean_token_accuracy": 0.1757102370262146,
      "num_tokens": 29092371.0,
      "step": 11480
    },
    {
      "entropy": 6.324045658111572,
      "epoch": 1.3254472013848817,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004835678980880318,
      "loss": 5.9383,
      "mean_token_accuracy": 0.18364807367324829,
      "num_tokens": 29105551.0,
      "step": 11485
    },
    {
      "entropy": 6.315791082382202,
      "epoch": 1.3260242354298903,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004835524146991753,
      "loss": 5.968,
      "mean_token_accuracy": 0.1817828133702278,
      "num_tokens": 29118584.0,
      "step": 11490
    },
    {
      "entropy": 6.252958297729492,
      "epoch": 1.326601269474899,
      "grad_norm": 0.84375,
      "learning_rate": 0.00048353692429576185,
      "loss": 5.9323,
      "mean_token_accuracy": 0.18134985864162445,
      "num_tokens": 29131972.0,
      "step": 11495
    },
    {
      "entropy": 6.252517557144165,
      "epoch": 1.3271783035199076,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004835214268783126,
      "loss": 5.9058,
      "mean_token_accuracy": 0.1779496893286705,
      "num_tokens": 29143502.0,
      "step": 11500
    },
    {
      "entropy": 6.224537801742554,
      "epoch": 1.3277553375649163,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00048350592244734866,
      "loss": 5.9006,
      "mean_token_accuracy": 0.176849302649498,
      "num_tokens": 29156765.0,
      "step": 11505
    },
    {
      "entropy": 6.249957513809204,
      "epoch": 1.3283323716099251,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004834904110033917,
      "loss": 5.9128,
      "mean_token_accuracy": 0.1835545629262924,
      "num_tokens": 29168858.0,
      "step": 11510
    },
    {
      "entropy": 6.234279489517212,
      "epoch": 1.3289094056549335,
      "grad_norm": 2.390625,
      "learning_rate": 0.00048347489254696327,
      "loss": 5.9043,
      "mean_token_accuracy": 0.18543250858783722,
      "num_tokens": 29182497.0,
      "step": 11515
    },
    {
      "entropy": 6.23204607963562,
      "epoch": 1.3294864396999424,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004834593670785854,
      "loss": 5.8739,
      "mean_token_accuracy": 0.19426221251487732,
      "num_tokens": 29194659.0,
      "step": 11520
    },
    {
      "entropy": 6.29420223236084,
      "epoch": 1.3300634737449508,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048344383459878024,
      "loss": 5.959,
      "mean_token_accuracy": 0.17725456953048707,
      "num_tokens": 29208192.0,
      "step": 11525
    },
    {
      "entropy": 6.23691725730896,
      "epoch": 1.3306405077899597,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048342829510807024,
      "loss": 5.8889,
      "mean_token_accuracy": 0.1792220428586006,
      "num_tokens": 29220532.0,
      "step": 11530
    },
    {
      "entropy": 6.2727607727050785,
      "epoch": 1.3312175418349683,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000483412748606978,
      "loss": 5.9195,
      "mean_token_accuracy": 0.17411866188049316,
      "num_tokens": 29233802.0,
      "step": 11535
    },
    {
      "entropy": 6.256830835342408,
      "epoch": 1.331794575879977,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004833971950960266,
      "loss": 5.8412,
      "mean_token_accuracy": 0.18012107461690902,
      "num_tokens": 29246099.0,
      "step": 11540
    },
    {
      "entropy": 6.273479080200195,
      "epoch": 1.3323716099249856,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004833816345757391,
      "loss": 5.9419,
      "mean_token_accuracy": 0.17442608922719954,
      "num_tokens": 29259546.0,
      "step": 11545
    },
    {
      "entropy": 6.269906806945801,
      "epoch": 1.3329486439699942,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004833660670466387,
      "loss": 5.9005,
      "mean_token_accuracy": 0.18373001962900162,
      "num_tokens": 29271980.0,
      "step": 11550
    },
    {
      "entropy": 6.275066280364991,
      "epoch": 1.3335256780150029,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004833504925092492,
      "loss": 5.8445,
      "mean_token_accuracy": 0.17742277383804322,
      "num_tokens": 29284679.0,
      "step": 11555
    },
    {
      "entropy": 6.25799970626831,
      "epoch": 1.3341027120600115,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004833349109640944,
      "loss": 5.9152,
      "mean_token_accuracy": 0.18850205689668656,
      "num_tokens": 29296446.0,
      "step": 11560
    },
    {
      "entropy": 6.266789960861206,
      "epoch": 1.3346797461050202,
      "grad_norm": 0.875,
      "learning_rate": 0.0004833193224116983,
      "loss": 5.9458,
      "mean_token_accuracy": 0.17592986226081847,
      "num_tokens": 29311036.0,
      "step": 11565
    },
    {
      "entropy": 6.2823975563049315,
      "epoch": 1.3352567801500288,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048330372685258526,
      "loss": 5.8972,
      "mean_token_accuracy": 0.18463776111602784,
      "num_tokens": 29324908.0,
      "step": 11570
    },
    {
      "entropy": 6.192406511306762,
      "epoch": 1.3358338141950374,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004832881242872799,
      "loss": 5.8917,
      "mean_token_accuracy": 0.1834719270467758,
      "num_tokens": 29337755.0,
      "step": 11575
    },
    {
      "entropy": 6.202840042114258,
      "epoch": 1.336410848240046,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004832725147163069,
      "loss": 5.9366,
      "mean_token_accuracy": 0.18037618100643157,
      "num_tokens": 29352148.0,
      "step": 11580
    },
    {
      "entropy": 6.323207044601441,
      "epoch": 1.336987882285055,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00048325689814019134,
      "loss": 5.8911,
      "mean_token_accuracy": 0.18111444562673568,
      "num_tokens": 29365714.0,
      "step": 11585
    },
    {
      "entropy": 6.233119821548462,
      "epoch": 1.3375649163300634,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004832412745594585,
      "loss": 5.8602,
      "mean_token_accuracy": 0.18742570877075196,
      "num_tokens": 29377699.0,
      "step": 11590
    },
    {
      "entropy": 6.2675621509552,
      "epoch": 1.3381419503750722,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048322564397463376,
      "loss": 5.8817,
      "mean_token_accuracy": 0.17823225408792495,
      "num_tokens": 29390293.0,
      "step": 11595
    },
    {
      "entropy": 6.222278308868408,
      "epoch": 1.3387189844200809,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048321000638624296,
      "loss": 5.8674,
      "mean_token_accuracy": 0.19048573672771454,
      "num_tokens": 29403148.0,
      "step": 11600
    },
    {
      "entropy": 6.273978853225708,
      "epoch": 1.3392960184650895,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000483194361794812,
      "loss": 5.9112,
      "mean_token_accuracy": 0.183873550593853,
      "num_tokens": 29416457.0,
      "step": 11605
    },
    {
      "entropy": 6.311326551437378,
      "epoch": 1.3398730525100981,
      "grad_norm": 1.21875,
      "learning_rate": 0.000483178710200867,
      "loss": 5.9675,
      "mean_token_accuracy": 0.1749234139919281,
      "num_tokens": 29429657.0,
      "step": 11610
    },
    {
      "entropy": 6.245604801177978,
      "epoch": 1.3404500865551068,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004831630516049346,
      "loss": 5.8668,
      "mean_token_accuracy": 0.18419479876756667,
      "num_tokens": 29442517.0,
      "step": 11615
    },
    {
      "entropy": 6.211596155166626,
      "epoch": 1.3410271206001154,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004831473860075414,
      "loss": 5.8064,
      "mean_token_accuracy": 0.18939921706914903,
      "num_tokens": 29455360.0,
      "step": 11620
    },
    {
      "entropy": 6.224331903457641,
      "epoch": 1.341604154645124,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048313171340921417,
      "loss": 5.9279,
      "mean_token_accuracy": 0.18605845868587495,
      "num_tokens": 29469321.0,
      "step": 11625
    },
    {
      "entropy": 6.300617504119873,
      "epoch": 1.3421811886901327,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048311603381048025,
      "loss": 5.8777,
      "mean_token_accuracy": 0.19063641726970673,
      "num_tokens": 29481688.0,
      "step": 11630
    },
    {
      "entropy": 6.272624778747558,
      "epoch": 1.3427582227351413,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004831003472118668,
      "loss": 5.9167,
      "mean_token_accuracy": 0.18469232618808745,
      "num_tokens": 29493468.0,
      "step": 11635
    },
    {
      "entropy": 6.291437482833862,
      "epoch": 1.34333525678015,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004830846536139016,
      "loss": 5.9793,
      "mean_token_accuracy": 0.18145175874233246,
      "num_tokens": 29506254.0,
      "step": 11640
    },
    {
      "entropy": 6.3019349575042725,
      "epoch": 1.3439122908251586,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004830689530171124,
      "loss": 5.9616,
      "mean_token_accuracy": 0.1827969804406166,
      "num_tokens": 29518838.0,
      "step": 11645
    },
    {
      "entropy": 6.229454898834229,
      "epoch": 1.3444893248701673,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004830532454220273,
      "loss": 5.9009,
      "mean_token_accuracy": 0.1789653018116951,
      "num_tokens": 29531252.0,
      "step": 11650
    },
    {
      "entropy": 6.329381656646729,
      "epoch": 1.345066358915176,
      "grad_norm": 0.84375,
      "learning_rate": 0.00048303753082917474,
      "loss": 5.999,
      "mean_token_accuracy": 0.1740694984793663,
      "num_tokens": 29545125.0,
      "step": 11655
    },
    {
      "entropy": 6.332653999328613,
      "epoch": 1.3456433929601848,
      "grad_norm": 1.125,
      "learning_rate": 0.00048302180923908306,
      "loss": 5.9507,
      "mean_token_accuracy": 0.17851236909627916,
      "num_tokens": 29558011.0,
      "step": 11660
    },
    {
      "entropy": 6.246814346313476,
      "epoch": 1.3462204270051932,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048300608065228126,
      "loss": 5.8889,
      "mean_token_accuracy": 0.18362092971801758,
      "num_tokens": 29570864.0,
      "step": 11665
    },
    {
      "entropy": 6.268451309204101,
      "epoch": 1.346797461050202,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048299034506929815,
      "loss": 5.8875,
      "mean_token_accuracy": 0.17999668866395951,
      "num_tokens": 29583098.0,
      "step": 11670
    },
    {
      "entropy": 6.188910436630249,
      "epoch": 1.3473744950952107,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048297460249066315,
      "loss": 5.8204,
      "mean_token_accuracy": 0.19384068846702576,
      "num_tokens": 29594980.0,
      "step": 11675
    },
    {
      "entropy": 6.244320869445801,
      "epoch": 1.3479515291402193,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004829588529169057,
      "loss": 5.9027,
      "mean_token_accuracy": 0.17963840514421464,
      "num_tokens": 29608424.0,
      "step": 11680
    },
    {
      "entropy": 6.266217184066773,
      "epoch": 1.348528563185228,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004829430963485555,
      "loss": 5.8302,
      "mean_token_accuracy": 0.18554893732070923,
      "num_tokens": 29621784.0,
      "step": 11685
    },
    {
      "entropy": 6.0988260269165036,
      "epoch": 1.3491055972302366,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004829273327861426,
      "loss": 5.7779,
      "mean_token_accuracy": 0.1852966859936714,
      "num_tokens": 29633875.0,
      "step": 11690
    },
    {
      "entropy": 6.16610746383667,
      "epoch": 1.3496826312752452,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004829115622301971,
      "loss": 5.7714,
      "mean_token_accuracy": 0.18935371488332747,
      "num_tokens": 29647510.0,
      "step": 11695
    },
    {
      "entropy": 6.283204221725464,
      "epoch": 1.3502596653202539,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048289578468124956,
      "loss": 5.9526,
      "mean_token_accuracy": 0.1727623924612999,
      "num_tokens": 29659916.0,
      "step": 11700
    },
    {
      "entropy": 6.293156337738037,
      "epoch": 1.3508366993652625,
      "grad_norm": 0.81640625,
      "learning_rate": 0.00048288000013983046,
      "loss": 5.9829,
      "mean_token_accuracy": 0.17646577209234238,
      "num_tokens": 29673628.0,
      "step": 11705
    },
    {
      "entropy": 6.2947096824646,
      "epoch": 1.3514137334102712,
      "grad_norm": 1.0,
      "learning_rate": 0.00048286420860647086,
      "loss": 5.9344,
      "mean_token_accuracy": 0.18376769721508027,
      "num_tokens": 29685245.0,
      "step": 11710
    },
    {
      "entropy": 6.226466846466065,
      "epoch": 1.3519907674552798,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048284841008170185,
      "loss": 5.8636,
      "mean_token_accuracy": 0.18923982828855515,
      "num_tokens": 29698205.0,
      "step": 11715
    },
    {
      "entropy": 6.244970703125,
      "epoch": 1.3525678015002884,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00048283260456605487,
      "loss": 5.9515,
      "mean_token_accuracy": 0.17597863227128982,
      "num_tokens": 29712060.0,
      "step": 11720
    },
    {
      "entropy": 6.266519069671631,
      "epoch": 1.3531448355452973,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004828167920600614,
      "loss": 5.9076,
      "mean_token_accuracy": 0.17571613192558289,
      "num_tokens": 29724452.0,
      "step": 11725
    },
    {
      "entropy": 6.280622100830078,
      "epoch": 1.3537218695903057,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004828009725642534,
      "loss": 5.8361,
      "mean_token_accuracy": 0.1847669005393982,
      "num_tokens": 29735754.0,
      "step": 11730
    },
    {
      "entropy": 6.191987037658691,
      "epoch": 1.3542989036353146,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004827851460791628,
      "loss": 5.8305,
      "mean_token_accuracy": 0.18259759843349457,
      "num_tokens": 29747918.0,
      "step": 11735
    },
    {
      "entropy": 6.268384408950806,
      "epoch": 1.3548759376803232,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048276931260532213,
      "loss": 5.917,
      "mean_token_accuracy": 0.18099725395441055,
      "num_tokens": 29760152.0,
      "step": 11740
    },
    {
      "entropy": 6.257396793365478,
      "epoch": 1.3554529717253319,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004827534721432639,
      "loss": 5.892,
      "mean_token_accuracy": 0.1828221321105957,
      "num_tokens": 29771677.0,
      "step": 11745
    },
    {
      "entropy": 6.283788108825684,
      "epoch": 1.3560300057703405,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004827376246935207,
      "loss": 5.9123,
      "mean_token_accuracy": 0.18449607342481614,
      "num_tokens": 29784523.0,
      "step": 11750
    },
    {
      "entropy": 6.189620733261108,
      "epoch": 1.3566070398153491,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004827217702566257,
      "loss": 5.8161,
      "mean_token_accuracy": 0.1914103537797928,
      "num_tokens": 29798181.0,
      "step": 11755
    },
    {
      "entropy": 6.250170421600342,
      "epoch": 1.3571840738603578,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048270590883311217,
      "loss": 6.0158,
      "mean_token_accuracy": 0.17536882162094117,
      "num_tokens": 29810530.0,
      "step": 11760
    },
    {
      "entropy": 6.284173202514649,
      "epoch": 1.3577611079053664,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048269004042351363,
      "loss": 5.9157,
      "mean_token_accuracy": 0.1859867602586746,
      "num_tokens": 29823929.0,
      "step": 11765
    },
    {
      "entropy": 6.295540142059326,
      "epoch": 1.358338141950375,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004826741650283637,
      "loss": 5.9418,
      "mean_token_accuracy": 0.17446097135543823,
      "num_tokens": 29836405.0,
      "step": 11770
    },
    {
      "entropy": 6.270499753952026,
      "epoch": 1.3589151759953837,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004826582826481963,
      "loss": 5.8639,
      "mean_token_accuracy": 0.18237340599298477,
      "num_tokens": 29849392.0,
      "step": 11775
    },
    {
      "entropy": 6.273336029052734,
      "epoch": 1.3594922100403923,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004826423932835458,
      "loss": 5.8806,
      "mean_token_accuracy": 0.1895483747124672,
      "num_tokens": 29862431.0,
      "step": 11780
    },
    {
      "entropy": 6.209116411209107,
      "epoch": 1.360069244085401,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048262649693494653,
      "loss": 5.8943,
      "mean_token_accuracy": 0.19191052317619323,
      "num_tokens": 29875628.0,
      "step": 11785
    },
    {
      "entropy": 6.300378656387329,
      "epoch": 1.3606462781304096,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004826105936029332,
      "loss": 5.9918,
      "mean_token_accuracy": 0.17364499270915984,
      "num_tokens": 29888399.0,
      "step": 11790
    },
    {
      "entropy": 6.289611387252807,
      "epoch": 1.3612233121754183,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004825946832880406,
      "loss": 5.9271,
      "mean_token_accuracy": 0.18047730177640914,
      "num_tokens": 29900776.0,
      "step": 11795
    },
    {
      "entropy": 6.262292814254761,
      "epoch": 1.3618003462204271,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048257876599080404,
      "loss": 5.9415,
      "mean_token_accuracy": 0.17547063827514647,
      "num_tokens": 29913764.0,
      "step": 11800
    },
    {
      "entropy": 6.256836652755737,
      "epoch": 1.3623773802654355,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048256284171175874,
      "loss": 5.8783,
      "mean_token_accuracy": 0.17941998839378356,
      "num_tokens": 29925884.0,
      "step": 11805
    },
    {
      "entropy": 6.1477972030639645,
      "epoch": 1.3629544143104444,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048254691045144035,
      "loss": 5.7926,
      "mean_token_accuracy": 0.19762682169675827,
      "num_tokens": 29938151.0,
      "step": 11810
    },
    {
      "entropy": 6.249617195129394,
      "epoch": 1.363531448355453,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004825309722103848,
      "loss": 5.8629,
      "mean_token_accuracy": 0.18313466012477875,
      "num_tokens": 29950696.0,
      "step": 11815
    },
    {
      "entropy": 6.29201488494873,
      "epoch": 1.3641084824004617,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000482515026989128,
      "loss": 5.9664,
      "mean_token_accuracy": 0.17421681731939315,
      "num_tokens": 29963295.0,
      "step": 11820
    },
    {
      "entropy": 6.2213856220245365,
      "epoch": 1.3646855164454703,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048249907478820634,
      "loss": 5.8854,
      "mean_token_accuracy": 0.18519256561994552,
      "num_tokens": 29976709.0,
      "step": 11825
    },
    {
      "entropy": 6.211252689361572,
      "epoch": 1.365262550490479,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048248311560815637,
      "loss": 5.8422,
      "mean_token_accuracy": 0.18762611150741576,
      "num_tokens": 29989341.0,
      "step": 11830
    },
    {
      "entropy": 6.311557674407959,
      "epoch": 1.3658395845354876,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004824671494495149,
      "loss": 5.9223,
      "mean_token_accuracy": 0.17833828777074814,
      "num_tokens": 30001630.0,
      "step": 11835
    },
    {
      "entropy": 6.344333982467651,
      "epoch": 1.3664166185804962,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004824511763128189,
      "loss": 5.972,
      "mean_token_accuracy": 0.17396059185266494,
      "num_tokens": 30014629.0,
      "step": 11840
    },
    {
      "entropy": 6.19668550491333,
      "epoch": 1.3669936526255049,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048243519619860567,
      "loss": 5.8595,
      "mean_token_accuracy": 0.18858882039785385,
      "num_tokens": 30025410.0,
      "step": 11845
    },
    {
      "entropy": 6.274438428878784,
      "epoch": 1.3675706866705135,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004824192091074126,
      "loss": 5.879,
      "mean_token_accuracy": 0.18532084673643112,
      "num_tokens": 30036955.0,
      "step": 11850
    },
    {
      "entropy": 6.300392484664917,
      "epoch": 1.3681477207155222,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004824032150397775,
      "loss": 5.9053,
      "mean_token_accuracy": 0.1761827975511551,
      "num_tokens": 30049426.0,
      "step": 11855
    },
    {
      "entropy": 6.25205135345459,
      "epoch": 1.3687247547605308,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048238721399623824,
      "loss": 5.9286,
      "mean_token_accuracy": 0.18302511125802995,
      "num_tokens": 30061139.0,
      "step": 11860
    },
    {
      "entropy": 6.287611770629883,
      "epoch": 1.3693017888055397,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00048237120597733316,
      "loss": 5.971,
      "mean_token_accuracy": 0.17356999069452286,
      "num_tokens": 30074307.0,
      "step": 11865
    },
    {
      "entropy": 6.253595066070557,
      "epoch": 1.369878822850548,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004823551909836005,
      "loss": 5.8402,
      "mean_token_accuracy": 0.18354870975017548,
      "num_tokens": 30086610.0,
      "step": 11870
    },
    {
      "entropy": 6.225833892822266,
      "epoch": 1.370455856895557,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048233916901557896,
      "loss": 5.9659,
      "mean_token_accuracy": 0.1795847535133362,
      "num_tokens": 30098971.0,
      "step": 11875
    },
    {
      "entropy": 6.324833536148072,
      "epoch": 1.3710328909405656,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048232314007380753,
      "loss": 5.9213,
      "mean_token_accuracy": 0.175594761967659,
      "num_tokens": 30110521.0,
      "step": 11880
    },
    {
      "entropy": 6.216272592544556,
      "epoch": 1.3716099249855742,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00048230710415882524,
      "loss": 5.8382,
      "mean_token_accuracy": 0.18911719769239427,
      "num_tokens": 30123687.0,
      "step": 11885
    },
    {
      "entropy": 6.160957002639771,
      "epoch": 1.3721869590305829,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048229106127117144,
      "loss": 5.848,
      "mean_token_accuracy": 0.18616175800561904,
      "num_tokens": 30137396.0,
      "step": 11890
    },
    {
      "entropy": 6.306479835510254,
      "epoch": 1.3727639930755915,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004822750114113858,
      "loss": 5.9742,
      "mean_token_accuracy": 0.17882277220487594,
      "num_tokens": 30149791.0,
      "step": 11895
    },
    {
      "entropy": 6.185193109512329,
      "epoch": 1.3733410271206001,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004822589545800081,
      "loss": 5.8004,
      "mean_token_accuracy": 0.19617358595132828,
      "num_tokens": 30162829.0,
      "step": 11900
    },
    {
      "entropy": 6.252997541427613,
      "epoch": 1.3739180611656088,
      "grad_norm": 0.875,
      "learning_rate": 0.0004822428907775784,
      "loss": 5.8487,
      "mean_token_accuracy": 0.18423136919736863,
      "num_tokens": 30174476.0,
      "step": 11905
    },
    {
      "entropy": 6.167435216903686,
      "epoch": 1.3744950952106174,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00048222682000463704,
      "loss": 5.8404,
      "mean_token_accuracy": 0.18709428310394288,
      "num_tokens": 30186573.0,
      "step": 11910
    },
    {
      "entropy": 6.268580055236816,
      "epoch": 1.375072129255626,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004822107422617245,
      "loss": 5.8788,
      "mean_token_accuracy": 0.18761981278657913,
      "num_tokens": 30198482.0,
      "step": 11915
    },
    {
      "entropy": 6.282097339630127,
      "epoch": 1.3756491633006347,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048219465754938156,
      "loss": 5.8994,
      "mean_token_accuracy": 0.17917088717222213,
      "num_tokens": 30211206.0,
      "step": 11920
    },
    {
      "entropy": 6.230522203445434,
      "epoch": 1.3762261973456433,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048217856586814926,
      "loss": 5.9308,
      "mean_token_accuracy": 0.1744114264845848,
      "num_tokens": 30224250.0,
      "step": 11925
    },
    {
      "entropy": 6.256836414337158,
      "epoch": 1.376803231390652,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048216246721856875,
      "loss": 5.9193,
      "mean_token_accuracy": 0.1849127158522606,
      "num_tokens": 30236212.0,
      "step": 11930
    },
    {
      "entropy": 6.339182376861572,
      "epoch": 1.3773802654356606,
      "grad_norm": 0.82421875,
      "learning_rate": 0.00048214636160118164,
      "loss": 6.0078,
      "mean_token_accuracy": 0.17261070907115936,
      "num_tokens": 30250681.0,
      "step": 11935
    },
    {
      "entropy": 6.29536919593811,
      "epoch": 1.3779572994806695,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004821302490165295,
      "loss": 6.0344,
      "mean_token_accuracy": 0.16762229800224304,
      "num_tokens": 30264659.0,
      "step": 11940
    },
    {
      "entropy": 6.281734800338745,
      "epoch": 1.378534333525678,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004821141294651544,
      "loss": 5.862,
      "mean_token_accuracy": 0.1831535965204239,
      "num_tokens": 30278821.0,
      "step": 11945
    },
    {
      "entropy": 6.267481803894043,
      "epoch": 1.3791113675706868,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048209800294759836,
      "loss": 5.9446,
      "mean_token_accuracy": 0.17433976233005524,
      "num_tokens": 30290502.0,
      "step": 11950
    },
    {
      "entropy": 6.231668090820312,
      "epoch": 1.3796884016156954,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004820818694644039,
      "loss": 5.8584,
      "mean_token_accuracy": 0.18093544542789458,
      "num_tokens": 30304499.0,
      "step": 11955
    },
    {
      "entropy": 6.290231847763062,
      "epoch": 1.380265435660704,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048206572901611364,
      "loss": 5.8813,
      "mean_token_accuracy": 0.18734426498413087,
      "num_tokens": 30315435.0,
      "step": 11960
    },
    {
      "entropy": 6.282363510131836,
      "epoch": 1.3808424697057127,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048204958160327034,
      "loss": 5.9168,
      "mean_token_accuracy": 0.17745988219976425,
      "num_tokens": 30327502.0,
      "step": 11965
    },
    {
      "entropy": 6.214523839950561,
      "epoch": 1.3814195037507213,
      "grad_norm": 0.82421875,
      "learning_rate": 0.00048203342722641726,
      "loss": 5.8902,
      "mean_token_accuracy": 0.18722237944602965,
      "num_tokens": 30341193.0,
      "step": 11970
    },
    {
      "entropy": 6.206170892715454,
      "epoch": 1.38199653779573,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048201726588609776,
      "loss": 5.8989,
      "mean_token_accuracy": 0.17957627326250075,
      "num_tokens": 30352899.0,
      "step": 11975
    },
    {
      "entropy": 6.309755849838257,
      "epoch": 1.3825735718407386,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00048200109758285534,
      "loss": 5.92,
      "mean_token_accuracy": 0.18512837439775467,
      "num_tokens": 30366672.0,
      "step": 11980
    },
    {
      "entropy": 6.313425588607788,
      "epoch": 1.3831506058857472,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004819849223172337,
      "loss": 5.8994,
      "mean_token_accuracy": 0.1798256203532219,
      "num_tokens": 30379124.0,
      "step": 11985
    },
    {
      "entropy": 6.1657603740692135,
      "epoch": 1.3837276399307559,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048196874008977716,
      "loss": 5.791,
      "mean_token_accuracy": 0.1955113008618355,
      "num_tokens": 30391974.0,
      "step": 11990
    },
    {
      "entropy": 6.2148096561431885,
      "epoch": 1.3843046739757645,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004819525509010298,
      "loss": 5.8255,
      "mean_token_accuracy": 0.18269359171390534,
      "num_tokens": 30404711.0,
      "step": 11995
    },
    {
      "entropy": 6.264818954467773,
      "epoch": 1.3848817080207732,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004819363547515361,
      "loss": 5.8531,
      "mean_token_accuracy": 0.18904300183057784,
      "num_tokens": 30418500.0,
      "step": 12000
    },
    {
      "epoch": 1.3848817080207732,
      "eval_entropy": 6.094769026970775,
      "eval_loss": 5.968371868133545,
      "eval_mean_token_accuracy": 0.1853566014181154,
      "eval_num_tokens": 30418500.0,
      "eval_runtime": 17.6769,
      "eval_samples_per_second": 1591.686,
      "eval_steps_per_second": 198.961,
      "step": 12000
    },
    {
      "entropy": 6.286858463287354,
      "epoch": 1.385458742065782,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000481920151641841,
      "loss": 5.9413,
      "mean_token_accuracy": 0.17997430860996247,
      "num_tokens": 30430851.0,
      "step": 12005
    },
    {
      "entropy": 6.188632154464722,
      "epoch": 1.3860357761107904,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048190394157248935,
      "loss": 5.7864,
      "mean_token_accuracy": 0.1978558659553528,
      "num_tokens": 30442885.0,
      "step": 12010
    },
    {
      "entropy": 6.232287979125976,
      "epoch": 1.3866128101557993,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004818877245440264,
      "loss": 5.9598,
      "mean_token_accuracy": 0.17828823179006575,
      "num_tokens": 30456145.0,
      "step": 12015
    },
    {
      "entropy": 6.29456524848938,
      "epoch": 1.3871898442008077,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048187150055699763,
      "loss": 5.9468,
      "mean_token_accuracy": 0.17825163006782532,
      "num_tokens": 30468222.0,
      "step": 12020
    },
    {
      "entropy": 6.270632410049439,
      "epoch": 1.3877668782458166,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004818552696119487,
      "loss": 5.8855,
      "mean_token_accuracy": 0.1807767853140831,
      "num_tokens": 30481713.0,
      "step": 12025
    },
    {
      "entropy": 6.0479803562164305,
      "epoch": 1.3883439122908252,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004818390317094255,
      "loss": 5.6917,
      "mean_token_accuracy": 0.19636378139257432,
      "num_tokens": 30495622.0,
      "step": 12030
    },
    {
      "entropy": 6.276599979400634,
      "epoch": 1.3889209463358339,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004818227868499742,
      "loss": 5.9081,
      "mean_token_accuracy": 0.18413619101047515,
      "num_tokens": 30507959.0,
      "step": 12035
    },
    {
      "entropy": 6.255237483978272,
      "epoch": 1.3894979803808425,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004818065350341412,
      "loss": 5.9403,
      "mean_token_accuracy": 0.17912512719631196,
      "num_tokens": 30520814.0,
      "step": 12040
    },
    {
      "entropy": 6.350230693817139,
      "epoch": 1.3900750144258511,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048179027626247325,
      "loss": 5.9362,
      "mean_token_accuracy": 0.18090790510177612,
      "num_tokens": 30533761.0,
      "step": 12045
    },
    {
      "entropy": 6.261089086532593,
      "epoch": 1.3906520484708598,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004817740105355169,
      "loss": 5.8643,
      "mean_token_accuracy": 0.18540456444025039,
      "num_tokens": 30546131.0,
      "step": 12050
    },
    {
      "entropy": 6.196657752990722,
      "epoch": 1.3912290825158684,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048175773785381947,
      "loss": 5.828,
      "mean_token_accuracy": 0.1884249582886696,
      "num_tokens": 30559867.0,
      "step": 12055
    },
    {
      "entropy": 6.23419976234436,
      "epoch": 1.391806116560877,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048174145821792833,
      "loss": 5.8281,
      "mean_token_accuracy": 0.18950849622488022,
      "num_tokens": 30572089.0,
      "step": 12060
    },
    {
      "entropy": 6.230244112014771,
      "epoch": 1.3923831506058857,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004817251716283908,
      "loss": 5.8881,
      "mean_token_accuracy": 0.18826514035463332,
      "num_tokens": 30585579.0,
      "step": 12065
    },
    {
      "entropy": 6.232949924468994,
      "epoch": 1.3929601846508943,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004817088780857548,
      "loss": 5.9293,
      "mean_token_accuracy": 0.17915522754192353,
      "num_tokens": 30597560.0,
      "step": 12070
    },
    {
      "entropy": 6.245428800582886,
      "epoch": 1.393537218695903,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048169257759056845,
      "loss": 5.8309,
      "mean_token_accuracy": 0.1850240483880043,
      "num_tokens": 30609323.0,
      "step": 12075
    },
    {
      "entropy": 6.24797739982605,
      "epoch": 1.3941142527409118,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048167627014337994,
      "loss": 5.9319,
      "mean_token_accuracy": 0.17799312025308608,
      "num_tokens": 30621437.0,
      "step": 12080
    },
    {
      "entropy": 6.251001977920533,
      "epoch": 1.3946912867859202,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048165995574473764,
      "loss": 5.8909,
      "mean_token_accuracy": 0.18886545151472092,
      "num_tokens": 30634656.0,
      "step": 12085
    },
    {
      "entropy": 6.28310079574585,
      "epoch": 1.395268320830929,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048164363439519043,
      "loss": 5.8645,
      "mean_token_accuracy": 0.1891761913895607,
      "num_tokens": 30645456.0,
      "step": 12090
    },
    {
      "entropy": 6.157786703109741,
      "epoch": 1.3958453548759377,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004816273060952873,
      "loss": 5.8454,
      "mean_token_accuracy": 0.19033508002758026,
      "num_tokens": 30658093.0,
      "step": 12095
    },
    {
      "entropy": 6.211349678039551,
      "epoch": 1.3964223889209464,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048161097084557726,
      "loss": 5.8783,
      "mean_token_accuracy": 0.18408869504928588,
      "num_tokens": 30670301.0,
      "step": 12100
    },
    {
      "entropy": 6.275969552993774,
      "epoch": 1.396999422965955,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048159462864660993,
      "loss": 5.8541,
      "mean_token_accuracy": 0.1819340929389,
      "num_tokens": 30682512.0,
      "step": 12105
    },
    {
      "entropy": 6.208045244216919,
      "epoch": 1.3975764570109637,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004815782794989348,
      "loss": 5.786,
      "mean_token_accuracy": 0.1936233624815941,
      "num_tokens": 30693677.0,
      "step": 12110
    },
    {
      "entropy": 6.22023720741272,
      "epoch": 1.3981534910559723,
      "grad_norm": 0.875,
      "learning_rate": 0.0004815619234031019,
      "loss": 5.8202,
      "mean_token_accuracy": 0.1801977202296257,
      "num_tokens": 30706994.0,
      "step": 12115
    },
    {
      "entropy": 6.209871244430542,
      "epoch": 1.398730525100981,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004815455603596613,
      "loss": 5.7879,
      "mean_token_accuracy": 0.19214085340499878,
      "num_tokens": 30721164.0,
      "step": 12120
    },
    {
      "entropy": 6.152595138549804,
      "epoch": 1.3993075591459896,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004815291903691634,
      "loss": 5.7781,
      "mean_token_accuracy": 0.18941596001386643,
      "num_tokens": 30734337.0,
      "step": 12125
    },
    {
      "entropy": 6.277353286743164,
      "epoch": 1.3998845931909982,
      "grad_norm": 0.828125,
      "learning_rate": 0.00048151281343215873,
      "loss": 5.9368,
      "mean_token_accuracy": 0.1785549521446228,
      "num_tokens": 30747565.0,
      "step": 12130
    },
    {
      "entropy": 6.26048583984375,
      "epoch": 1.4004616272360069,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004814964295491982,
      "loss": 5.8795,
      "mean_token_accuracy": 0.18632088601589203,
      "num_tokens": 30759285.0,
      "step": 12135
    },
    {
      "entropy": 6.259620618820191,
      "epoch": 1.4010386612810155,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048148003872083286,
      "loss": 5.9448,
      "mean_token_accuracy": 0.18301836401224136,
      "num_tokens": 30772390.0,
      "step": 12140
    },
    {
      "entropy": 6.233786344528198,
      "epoch": 1.4016156953260244,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048146364094761384,
      "loss": 5.8503,
      "mean_token_accuracy": 0.1865824893116951,
      "num_tokens": 30784845.0,
      "step": 12145
    },
    {
      "entropy": 6.321959495544434,
      "epoch": 1.4021927293710328,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048144723623009297,
      "loss": 5.8951,
      "mean_token_accuracy": 0.18258907943964003,
      "num_tokens": 30796705.0,
      "step": 12150
    },
    {
      "entropy": 6.229117345809937,
      "epoch": 1.4027697634160416,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004814308245688218,
      "loss": 5.902,
      "mean_token_accuracy": 0.18317533135414124,
      "num_tokens": 30809143.0,
      "step": 12155
    },
    {
      "entropy": 6.268697118759155,
      "epoch": 1.40334679746105,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048141440596435235,
      "loss": 5.9108,
      "mean_token_accuracy": 0.18186790496110916,
      "num_tokens": 30821519.0,
      "step": 12160
    },
    {
      "entropy": 6.107213115692138,
      "epoch": 1.403923831506059,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004813979804172369,
      "loss": 5.7672,
      "mean_token_accuracy": 0.19262839555740358,
      "num_tokens": 30835310.0,
      "step": 12165
    },
    {
      "entropy": 6.183963251113892,
      "epoch": 1.4045008655510676,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048138154792802795,
      "loss": 5.8479,
      "mean_token_accuracy": 0.18649512976408006,
      "num_tokens": 30847966.0,
      "step": 12170
    },
    {
      "entropy": 6.328726482391358,
      "epoch": 1.4050778995960762,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004813651084972781,
      "loss": 5.9034,
      "mean_token_accuracy": 0.1764056295156479,
      "num_tokens": 30860732.0,
      "step": 12175
    },
    {
      "entropy": 6.2920372009277346,
      "epoch": 1.4056549336410848,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048134866212554036,
      "loss": 5.9399,
      "mean_token_accuracy": 0.18320793956518172,
      "num_tokens": 30874846.0,
      "step": 12180
    },
    {
      "entropy": 6.225037527084351,
      "epoch": 1.4062319676860935,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004813322088133679,
      "loss": 5.799,
      "mean_token_accuracy": 0.19521696418523787,
      "num_tokens": 30888470.0,
      "step": 12185
    },
    {
      "entropy": 6.318968343734741,
      "epoch": 1.4068090017311021,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048131574856131407,
      "loss": 5.939,
      "mean_token_accuracy": 0.1787115827202797,
      "num_tokens": 30901124.0,
      "step": 12190
    },
    {
      "entropy": 6.247391176223755,
      "epoch": 1.4073860357761108,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004812992813699324,
      "loss": 5.9271,
      "mean_token_accuracy": 0.1790681302547455,
      "num_tokens": 30912533.0,
      "step": 12195
    },
    {
      "entropy": 6.201996278762818,
      "epoch": 1.4079630698211194,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000481282807239777,
      "loss": 5.8735,
      "mean_token_accuracy": 0.18215512484312057,
      "num_tokens": 30925070.0,
      "step": 12200
    },
    {
      "entropy": 6.294028997421265,
      "epoch": 1.408540103866128,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004812663261714019,
      "loss": 5.8908,
      "mean_token_accuracy": 0.18230621367692948,
      "num_tokens": 30938305.0,
      "step": 12205
    },
    {
      "entropy": 6.3103162288665775,
      "epoch": 1.4091171379111367,
      "grad_norm": 1.0,
      "learning_rate": 0.0004812498381653613,
      "loss": 5.9383,
      "mean_token_accuracy": 0.1770524188876152,
      "num_tokens": 30951020.0,
      "step": 12210
    },
    {
      "entropy": 6.275585699081421,
      "epoch": 1.4096941719561453,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004812333432222098,
      "loss": 5.8438,
      "mean_token_accuracy": 0.18392059057950974,
      "num_tokens": 30963712.0,
      "step": 12215
    },
    {
      "entropy": 6.278170347213745,
      "epoch": 1.4102712060011542,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004812168413425023,
      "loss": 5.811,
      "mean_token_accuracy": 0.18996177911758422,
      "num_tokens": 30975859.0,
      "step": 12220
    },
    {
      "entropy": 6.252748584747314,
      "epoch": 1.4108482400461626,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048120033252679374,
      "loss": 5.9326,
      "mean_token_accuracy": 0.17918262481689454,
      "num_tokens": 30989100.0,
      "step": 12225
    },
    {
      "entropy": 6.210408449172974,
      "epoch": 1.4114252740911715,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048118381677563946,
      "loss": 5.9139,
      "mean_token_accuracy": 0.18106505423784255,
      "num_tokens": 31001921.0,
      "step": 12230
    },
    {
      "entropy": 6.321139717102051,
      "epoch": 1.41200230813618,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004811672940895949,
      "loss": 5.9403,
      "mean_token_accuracy": 0.17640805691480638,
      "num_tokens": 31015642.0,
      "step": 12235
    },
    {
      "entropy": 6.274531555175781,
      "epoch": 1.4125793421811887,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004811507644692158,
      "loss": 5.8736,
      "mean_token_accuracy": 0.18296072036027908,
      "num_tokens": 31028964.0,
      "step": 12240
    },
    {
      "entropy": 6.301026773452759,
      "epoch": 1.4131563762261974,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004811342279150581,
      "loss": 5.935,
      "mean_token_accuracy": 0.18449933975934982,
      "num_tokens": 31040567.0,
      "step": 12245
    },
    {
      "entropy": 6.311332130432129,
      "epoch": 1.413733410271206,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004811176844276781,
      "loss": 5.9634,
      "mean_token_accuracy": 0.17758635878562928,
      "num_tokens": 31053005.0,
      "step": 12250
    },
    {
      "entropy": 6.25804386138916,
      "epoch": 1.4143104443162147,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004811011340076322,
      "loss": 5.885,
      "mean_token_accuracy": 0.180134516954422,
      "num_tokens": 31065971.0,
      "step": 12255
    },
    {
      "entropy": 6.215049409866333,
      "epoch": 1.4148874783612233,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048108457665547695,
      "loss": 5.8115,
      "mean_token_accuracy": 0.19384694695472718,
      "num_tokens": 31078762.0,
      "step": 12260
    },
    {
      "entropy": 6.276061105728149,
      "epoch": 1.415464512406232,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004810680123717694,
      "loss": 5.8447,
      "mean_token_accuracy": 0.18466615676879883,
      "num_tokens": 31091462.0,
      "step": 12265
    },
    {
      "entropy": 6.258430671691895,
      "epoch": 1.4160415464512406,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004810514411570666,
      "loss": 5.8768,
      "mean_token_accuracy": 0.18574267625808716,
      "num_tokens": 31104610.0,
      "step": 12270
    },
    {
      "entropy": 6.162780046463013,
      "epoch": 1.4166185804962492,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004810348630119259,
      "loss": 5.9055,
      "mean_token_accuracy": 0.1891510084271431,
      "num_tokens": 31117125.0,
      "step": 12275
    },
    {
      "entropy": 6.211641502380371,
      "epoch": 1.4171956145412579,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048101827793690493,
      "loss": 5.8653,
      "mean_token_accuracy": 0.18667034059762955,
      "num_tokens": 31129703.0,
      "step": 12280
    },
    {
      "entropy": 6.264873695373535,
      "epoch": 1.4177726485862667,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004810016859325615,
      "loss": 5.8749,
      "mean_token_accuracy": 0.18419784009456636,
      "num_tokens": 31142127.0,
      "step": 12285
    },
    {
      "entropy": 6.22588210105896,
      "epoch": 1.4183496826312751,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004809850869994537,
      "loss": 5.8644,
      "mean_token_accuracy": 0.1881739690899849,
      "num_tokens": 31154665.0,
      "step": 12290
    },
    {
      "entropy": 6.238256788253784,
      "epoch": 1.418926716676284,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004809684811381398,
      "loss": 5.8913,
      "mean_token_accuracy": 0.1853290930390358,
      "num_tokens": 31167004.0,
      "step": 12295
    },
    {
      "entropy": 6.267100667953491,
      "epoch": 1.4195037507212924,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004809518683491785,
      "loss": 5.8955,
      "mean_token_accuracy": 0.1826624408364296,
      "num_tokens": 31178979.0,
      "step": 12300
    },
    {
      "entropy": 6.286066913604737,
      "epoch": 1.4200807847663013,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048093524863312823,
      "loss": 5.9379,
      "mean_token_accuracy": 0.17354432940483094,
      "num_tokens": 31190877.0,
      "step": 12305
    },
    {
      "entropy": 6.315721464157105,
      "epoch": 1.42065781881131,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004809186219905482,
      "loss": 5.8956,
      "mean_token_accuracy": 0.1830655261874199,
      "num_tokens": 31202988.0,
      "step": 12310
    },
    {
      "entropy": 6.287948131561279,
      "epoch": 1.4212348528563186,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004809019884219976,
      "loss": 5.9113,
      "mean_token_accuracy": 0.1821136936545372,
      "num_tokens": 31215436.0,
      "step": 12315
    },
    {
      "entropy": 6.272874116897583,
      "epoch": 1.4218118869013272,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048088534792803595,
      "loss": 5.8744,
      "mean_token_accuracy": 0.1773661717772484,
      "num_tokens": 31228260.0,
      "step": 12320
    },
    {
      "entropy": 6.270263051986694,
      "epoch": 1.4223889209463358,
      "grad_norm": 0.828125,
      "learning_rate": 0.00048086870050922286,
      "loss": 5.9444,
      "mean_token_accuracy": 0.1788017526268959,
      "num_tokens": 31242496.0,
      "step": 12325
    },
    {
      "entropy": 6.278579807281494,
      "epoch": 1.4229659549913445,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048085204616611833,
      "loss": 5.8856,
      "mean_token_accuracy": 0.18055897653102876,
      "num_tokens": 31254375.0,
      "step": 12330
    },
    {
      "entropy": 6.277070140838623,
      "epoch": 1.4235429890363531,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048083538489928246,
      "loss": 6.0035,
      "mean_token_accuracy": 0.17097550332546235,
      "num_tokens": 31266590.0,
      "step": 12335
    },
    {
      "entropy": 6.367906475067139,
      "epoch": 1.4241200230813618,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004808187167092757,
      "loss": 5.9897,
      "mean_token_accuracy": 0.1780717208981514,
      "num_tokens": 31278928.0,
      "step": 12340
    },
    {
      "entropy": 6.16786823272705,
      "epoch": 1.4246970571263704,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004808020415966586,
      "loss": 5.7425,
      "mean_token_accuracy": 0.19216306656599044,
      "num_tokens": 31292144.0,
      "step": 12345
    },
    {
      "entropy": 6.251178455352783,
      "epoch": 1.425274091171379,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000480785359561992,
      "loss": 5.8492,
      "mean_token_accuracy": 0.1862394005060196,
      "num_tokens": 31303690.0,
      "step": 12350
    },
    {
      "entropy": 6.231591415405274,
      "epoch": 1.4258511252163877,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048076867060583704,
      "loss": 5.8648,
      "mean_token_accuracy": 0.18773231506347657,
      "num_tokens": 31315433.0,
      "step": 12355
    },
    {
      "entropy": 6.211475467681884,
      "epoch": 1.4264281592613965,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004807519747287551,
      "loss": 5.8756,
      "mean_token_accuracy": 0.19038024842739104,
      "num_tokens": 31327483.0,
      "step": 12360
    },
    {
      "entropy": 6.267287349700927,
      "epoch": 1.427005193306405,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004807352719313078,
      "loss": 5.9346,
      "mean_token_accuracy": 0.184219029545784,
      "num_tokens": 31339167.0,
      "step": 12365
    },
    {
      "entropy": 6.232132863998413,
      "epoch": 1.4275822273514138,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004807185622140567,
      "loss": 5.8239,
      "mean_token_accuracy": 0.18949489593505858,
      "num_tokens": 31352476.0,
      "step": 12370
    },
    {
      "entropy": 6.198150587081909,
      "epoch": 1.4281592613964225,
      "grad_norm": 0.875,
      "learning_rate": 0.00048070184557756396,
      "loss": 5.8413,
      "mean_token_accuracy": 0.19509464502334595,
      "num_tokens": 31364286.0,
      "step": 12375
    },
    {
      "entropy": 6.2374444007873535,
      "epoch": 1.428736295441431,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048068512202239177,
      "loss": 5.861,
      "mean_token_accuracy": 0.18188831359148025,
      "num_tokens": 31378141.0,
      "step": 12380
    },
    {
      "entropy": 6.265792560577393,
      "epoch": 1.4293133294864397,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004806683915491028,
      "loss": 5.8618,
      "mean_token_accuracy": 0.19368935078382493,
      "num_tokens": 31390738.0,
      "step": 12385
    },
    {
      "entropy": 6.186821746826172,
      "epoch": 1.4298903635314484,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004806516541582596,
      "loss": 5.8641,
      "mean_token_accuracy": 0.17950132936239244,
      "num_tokens": 31402289.0,
      "step": 12390
    },
    {
      "entropy": 6.267706727981567,
      "epoch": 1.430467397576457,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048063490985042506,
      "loss": 5.9237,
      "mean_token_accuracy": 0.18528805524110795,
      "num_tokens": 31416029.0,
      "step": 12395
    },
    {
      "entropy": 6.1704497814178465,
      "epoch": 1.4310444316214657,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004806181586261626,
      "loss": 5.8595,
      "mean_token_accuracy": 0.1901737704873085,
      "num_tokens": 31428980.0,
      "step": 12400
    },
    {
      "entropy": 6.245457315444947,
      "epoch": 1.4316214656664743,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048060140048603544,
      "loss": 5.8551,
      "mean_token_accuracy": 0.18665973246097564,
      "num_tokens": 31441424.0,
      "step": 12405
    },
    {
      "entropy": 6.306221961975098,
      "epoch": 1.432198499711483,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004805846354306073,
      "loss": 5.8501,
      "mean_token_accuracy": 0.18400493562221526,
      "num_tokens": 31453968.0,
      "step": 12410
    },
    {
      "entropy": 6.247371530532837,
      "epoch": 1.4327755337564916,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048056786346044214,
      "loss": 5.9388,
      "mean_token_accuracy": 0.18178324401378632,
      "num_tokens": 31466385.0,
      "step": 12415
    },
    {
      "entropy": 6.296180772781372,
      "epoch": 1.4333525678015002,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048055108457610395,
      "loss": 5.9276,
      "mean_token_accuracy": 0.17859654128551483,
      "num_tokens": 31478432.0,
      "step": 12420
    },
    {
      "entropy": 6.311766242980957,
      "epoch": 1.4339296018465089,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004805342987781571,
      "loss": 5.9348,
      "mean_token_accuracy": 0.1754424348473549,
      "num_tokens": 31490435.0,
      "step": 12425
    },
    {
      "entropy": 6.23893723487854,
      "epoch": 1.4345066358915175,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004805175060671663,
      "loss": 5.8704,
      "mean_token_accuracy": 0.18845874965190887,
      "num_tokens": 31502573.0,
      "step": 12430
    },
    {
      "entropy": 6.24181342124939,
      "epoch": 1.4350836699365264,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004805007064436962,
      "loss": 5.9264,
      "mean_token_accuracy": 0.18014222532510757,
      "num_tokens": 31516071.0,
      "step": 12435
    },
    {
      "entropy": 6.267311763763428,
      "epoch": 1.4356607039815348,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004804838999083119,
      "loss": 5.8606,
      "mean_token_accuracy": 0.18422294706106185,
      "num_tokens": 31528895.0,
      "step": 12440
    },
    {
      "entropy": 6.342143917083741,
      "epoch": 1.4362377380265436,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004804670864615787,
      "loss": 5.9862,
      "mean_token_accuracy": 0.17318262606859208,
      "num_tokens": 31541759.0,
      "step": 12445
    },
    {
      "entropy": 6.278418684005738,
      "epoch": 1.4368147720715523,
      "grad_norm": 0.81640625,
      "learning_rate": 0.00048045026610406223,
      "loss": 5.9275,
      "mean_token_accuracy": 0.18075491338968278,
      "num_tokens": 31554249.0,
      "step": 12450
    },
    {
      "entropy": 6.334468650817871,
      "epoch": 1.437391806116561,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000480433438836328,
      "loss": 5.9596,
      "mean_token_accuracy": 0.17709464728832244,
      "num_tokens": 31567673.0,
      "step": 12455
    },
    {
      "entropy": 6.247079658508301,
      "epoch": 1.4379688401615696,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048041660465894206,
      "loss": 5.8085,
      "mean_token_accuracy": 0.19189121574163437,
      "num_tokens": 31581209.0,
      "step": 12460
    },
    {
      "entropy": 6.26020884513855,
      "epoch": 1.4385458742065782,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004803997635724707,
      "loss": 5.8963,
      "mean_token_accuracy": 0.18268535435199737,
      "num_tokens": 31593873.0,
      "step": 12465
    },
    {
      "entropy": 6.229708194732666,
      "epoch": 1.4391229082515868,
      "grad_norm": 0.875,
      "learning_rate": 0.0004803829155774804,
      "loss": 5.8594,
      "mean_token_accuracy": 0.18309808522462845,
      "num_tokens": 31605942.0,
      "step": 12470
    },
    {
      "entropy": 6.295551300048828,
      "epoch": 1.4396999422965955,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004803660606745377,
      "loss": 5.884,
      "mean_token_accuracy": 0.17954100519418717,
      "num_tokens": 31617648.0,
      "step": 12475
    },
    {
      "entropy": 6.207191038131714,
      "epoch": 1.4402769763416041,
      "grad_norm": 0.875,
      "learning_rate": 0.00048034919886420953,
      "loss": 5.8489,
      "mean_token_accuracy": 0.1791909545660019,
      "num_tokens": 31630774.0,
      "step": 12480
    },
    {
      "entropy": 6.291480588912964,
      "epoch": 1.4408540103866128,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048033233014706304,
      "loss": 5.8673,
      "mean_token_accuracy": 0.18329765796661376,
      "num_tokens": 31642888.0,
      "step": 12485
    },
    {
      "entropy": 6.212289428710937,
      "epoch": 1.4414310444316214,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048031545452366565,
      "loss": 5.8368,
      "mean_token_accuracy": 0.185288143157959,
      "num_tokens": 31655180.0,
      "step": 12490
    },
    {
      "entropy": 6.219779539108276,
      "epoch": 1.44200807847663,
      "grad_norm": 0.84375,
      "learning_rate": 0.00048029857199458493,
      "loss": 5.8953,
      "mean_token_accuracy": 0.18484043926000596,
      "num_tokens": 31667452.0,
      "step": 12495
    },
    {
      "entropy": 6.349998426437378,
      "epoch": 1.442585112521639,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048028168256038873,
      "loss": 5.8862,
      "mean_token_accuracy": 0.18908318430185317,
      "num_tokens": 31680995.0,
      "step": 12500
    },
    {
      "entropy": 6.285687685012817,
      "epoch": 1.4431621465666473,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048026478622164513,
      "loss": 5.948,
      "mean_token_accuracy": 0.1787041276693344,
      "num_tokens": 31693604.0,
      "step": 12505
    },
    {
      "entropy": 6.276877927780151,
      "epoch": 1.4437391806116562,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048024788297892234,
      "loss": 5.8997,
      "mean_token_accuracy": 0.18149978816509246,
      "num_tokens": 31706370.0,
      "step": 12510
    },
    {
      "entropy": 6.279643583297729,
      "epoch": 1.4443162146566648,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000480230972832789,
      "loss": 5.8614,
      "mean_token_accuracy": 0.17905326932668686,
      "num_tokens": 31719265.0,
      "step": 12515
    },
    {
      "entropy": 6.1697581768035885,
      "epoch": 1.4448932487016735,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048021405578381384,
      "loss": 5.7492,
      "mean_token_accuracy": 0.196394906938076,
      "num_tokens": 31732587.0,
      "step": 12520
    },
    {
      "entropy": 6.234063959121704,
      "epoch": 1.445470282746682,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004801971318325659,
      "loss": 5.8513,
      "mean_token_accuracy": 0.18114555180072783,
      "num_tokens": 31744664.0,
      "step": 12525
    },
    {
      "entropy": 6.27386965751648,
      "epoch": 1.4460473167916907,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004801802009796142,
      "loss": 5.8588,
      "mean_token_accuracy": 0.18989744186401367,
      "num_tokens": 31757814.0,
      "step": 12530
    },
    {
      "entropy": 6.20626974105835,
      "epoch": 1.4466243508366994,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004801632632255285,
      "loss": 5.8206,
      "mean_token_accuracy": 0.1883770138025284,
      "num_tokens": 31770575.0,
      "step": 12535
    },
    {
      "entropy": 6.221070718765259,
      "epoch": 1.447201384881708,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004801463185708783,
      "loss": 5.9257,
      "mean_token_accuracy": 0.18196507543325424,
      "num_tokens": 31782746.0,
      "step": 12540
    },
    {
      "entropy": 6.291515588760376,
      "epoch": 1.4477784189267167,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048012936701623363,
      "loss": 5.8165,
      "mean_token_accuracy": 0.19014340341091157,
      "num_tokens": 31794459.0,
      "step": 12545
    },
    {
      "entropy": 6.185211420059204,
      "epoch": 1.4483554529717253,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048011240856216456,
      "loss": 5.7949,
      "mean_token_accuracy": 0.1866045281291008,
      "num_tokens": 31806205.0,
      "step": 12550
    },
    {
      "entropy": 6.212912702560425,
      "epoch": 1.448932487016734,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048009544320924154,
      "loss": 5.9065,
      "mean_token_accuracy": 0.1799776643514633,
      "num_tokens": 31818590.0,
      "step": 12555
    },
    {
      "entropy": 6.302042293548584,
      "epoch": 1.4495095210617426,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004800784709580351,
      "loss": 5.8359,
      "mean_token_accuracy": 0.1804344967007637,
      "num_tokens": 31830702.0,
      "step": 12560
    },
    {
      "entropy": 6.213062763214111,
      "epoch": 1.4500865551067512,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004800614918091161,
      "loss": 5.8117,
      "mean_token_accuracy": 0.18469692766666412,
      "num_tokens": 31844326.0,
      "step": 12565
    },
    {
      "entropy": 6.23402771949768,
      "epoch": 1.4506635891517599,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004800445057630558,
      "loss": 5.8748,
      "mean_token_accuracy": 0.17864386290311812,
      "num_tokens": 31857153.0,
      "step": 12570
    },
    {
      "entropy": 6.241481590270996,
      "epoch": 1.4512406231967687,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004800275128204254,
      "loss": 5.8267,
      "mean_token_accuracy": 0.1919524073600769,
      "num_tokens": 31870589.0,
      "step": 12575
    },
    {
      "entropy": 6.2238446235656735,
      "epoch": 1.4518176572417771,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048001051298179637,
      "loss": 5.8186,
      "mean_token_accuracy": 0.18940580487251282,
      "num_tokens": 31882867.0,
      "step": 12580
    },
    {
      "entropy": 6.2409850597381595,
      "epoch": 1.452394691286786,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004799935062477407,
      "loss": 5.8757,
      "mean_token_accuracy": 0.17769681960344313,
      "num_tokens": 31895582.0,
      "step": 12585
    },
    {
      "entropy": 6.234495401382446,
      "epoch": 1.4529717253317946,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047997649261883013,
      "loss": 5.865,
      "mean_token_accuracy": 0.18466940820217131,
      "num_tokens": 31907873.0,
      "step": 12590
    },
    {
      "entropy": 6.225005006790161,
      "epoch": 1.4535487593768033,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004799594720956371,
      "loss": 5.7666,
      "mean_token_accuracy": 0.19158240854740144,
      "num_tokens": 31920459.0,
      "step": 12595
    },
    {
      "entropy": 6.255098819732666,
      "epoch": 1.454125793421812,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047994244467873407,
      "loss": 5.9023,
      "mean_token_accuracy": 0.18102422058582307,
      "num_tokens": 31934707.0,
      "step": 12600
    },
    {
      "entropy": 6.304951238632202,
      "epoch": 1.4547028274668206,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047992541036869364,
      "loss": 5.9922,
      "mean_token_accuracy": 0.17874653041362762,
      "num_tokens": 31946891.0,
      "step": 12605
    },
    {
      "entropy": 6.268550777435303,
      "epoch": 1.4552798615118292,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004799083691660889,
      "loss": 5.8508,
      "mean_token_accuracy": 0.18193005174398422,
      "num_tokens": 31960300.0,
      "step": 12610
    },
    {
      "entropy": 6.256899785995484,
      "epoch": 1.4558568955568378,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004798913210714929,
      "loss": 5.8787,
      "mean_token_accuracy": 0.1840341404080391,
      "num_tokens": 31973565.0,
      "step": 12615
    },
    {
      "entropy": 6.228112363815308,
      "epoch": 1.4564339296018465,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047987426608547915,
      "loss": 5.9043,
      "mean_token_accuracy": 0.18167776465415955,
      "num_tokens": 31986256.0,
      "step": 12620
    },
    {
      "entropy": 6.259927558898926,
      "epoch": 1.4570109636468551,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004798572042086212,
      "loss": 5.8556,
      "mean_token_accuracy": 0.18698290437459947,
      "num_tokens": 31999034.0,
      "step": 12625
    },
    {
      "entropy": 6.285195827484131,
      "epoch": 1.4575879976918638,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047984013544149286,
      "loss": 5.9258,
      "mean_token_accuracy": 0.18081545680761338,
      "num_tokens": 32010213.0,
      "step": 12630
    },
    {
      "entropy": 6.303173446655274,
      "epoch": 1.4581650317368724,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047982305978466836,
      "loss": 5.9385,
      "mean_token_accuracy": 0.18032837361097337,
      "num_tokens": 32022516.0,
      "step": 12635
    },
    {
      "entropy": 6.314466714859009,
      "epoch": 1.4587420657818813,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047980597723872205,
      "loss": 5.8852,
      "mean_token_accuracy": 0.18110508024692534,
      "num_tokens": 32035644.0,
      "step": 12640
    },
    {
      "entropy": 6.276114892959595,
      "epoch": 1.4593190998268897,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004797888878042283,
      "loss": 5.95,
      "mean_token_accuracy": 0.18677132576704025,
      "num_tokens": 32049547.0,
      "step": 12645
    },
    {
      "entropy": 6.190760564804077,
      "epoch": 1.4598961338718985,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047977179148176217,
      "loss": 5.8463,
      "mean_token_accuracy": 0.1874154508113861,
      "num_tokens": 32062071.0,
      "step": 12650
    },
    {
      "entropy": 6.2850230693817135,
      "epoch": 1.4604731679169072,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004797546882718985,
      "loss": 5.9396,
      "mean_token_accuracy": 0.1715143322944641,
      "num_tokens": 32073964.0,
      "step": 12655
    },
    {
      "entropy": 6.290622997283935,
      "epoch": 1.4610502019619158,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047973757817521256,
      "loss": 5.889,
      "mean_token_accuracy": 0.18512476831674576,
      "num_tokens": 32087151.0,
      "step": 12660
    },
    {
      "entropy": 6.2918517112731935,
      "epoch": 1.4616272360069245,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004797204611922799,
      "loss": 5.8756,
      "mean_token_accuracy": 0.18420816510915755,
      "num_tokens": 32099072.0,
      "step": 12665
    },
    {
      "entropy": 6.149755048751831,
      "epoch": 1.462204270051933,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047970333732367626,
      "loss": 5.8025,
      "mean_token_accuracy": 0.18933116942644118,
      "num_tokens": 32112182.0,
      "step": 12670
    },
    {
      "entropy": 6.172571229934692,
      "epoch": 1.4627813040969417,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047968620656997754,
      "loss": 5.697,
      "mean_token_accuracy": 0.1965687707066536,
      "num_tokens": 32124067.0,
      "step": 12675
    },
    {
      "entropy": 6.264682626724243,
      "epoch": 1.4633583381419504,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047966906893175994,
      "loss": 5.8784,
      "mean_token_accuracy": 0.18597973138093948,
      "num_tokens": 32136983.0,
      "step": 12680
    },
    {
      "entropy": 6.265399074554443,
      "epoch": 1.463935372186959,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004796519244095998,
      "loss": 5.8528,
      "mean_token_accuracy": 0.1869005709886551,
      "num_tokens": 32150103.0,
      "step": 12685
    },
    {
      "entropy": 6.2796302318573,
      "epoch": 1.4645124062319677,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047963477300407394,
      "loss": 5.9146,
      "mean_token_accuracy": 0.17439359724521636,
      "num_tokens": 32162647.0,
      "step": 12690
    },
    {
      "entropy": 6.216573810577392,
      "epoch": 1.4650894402769763,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047961761471575903,
      "loss": 5.7881,
      "mean_token_accuracy": 0.19099143296480178,
      "num_tokens": 32174707.0,
      "step": 12695
    },
    {
      "entropy": 6.17914514541626,
      "epoch": 1.465666474321985,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047960044954523237,
      "loss": 5.841,
      "mean_token_accuracy": 0.19434589445590972,
      "num_tokens": 32187784.0,
      "step": 12700
    },
    {
      "entropy": 6.315832614898682,
      "epoch": 1.4662435083669936,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00047958327749307117,
      "loss": 5.9423,
      "mean_token_accuracy": 0.18215253055095673,
      "num_tokens": 32200300.0,
      "step": 12705
    },
    {
      "entropy": 6.276373481750488,
      "epoch": 1.4668205424120022,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047956609855985305,
      "loss": 5.8541,
      "mean_token_accuracy": 0.18482187539339065,
      "num_tokens": 32212836.0,
      "step": 12710
    },
    {
      "entropy": 6.097634124755859,
      "epoch": 1.467397576457011,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004795489127461559,
      "loss": 5.6718,
      "mean_token_accuracy": 0.20180321037769317,
      "num_tokens": 32226230.0,
      "step": 12715
    },
    {
      "entropy": 6.275608491897583,
      "epoch": 1.4679746105020195,
      "grad_norm": 0.859375,
      "learning_rate": 0.00047953172005255756,
      "loss": 5.9383,
      "mean_token_accuracy": 0.182850244641304,
      "num_tokens": 32238994.0,
      "step": 12720
    },
    {
      "entropy": 6.2543822765350345,
      "epoch": 1.4685516445470284,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004795145204796365,
      "loss": 5.9015,
      "mean_token_accuracy": 0.1806575983762741,
      "num_tokens": 32250983.0,
      "step": 12725
    },
    {
      "entropy": 6.243729972839356,
      "epoch": 1.469128678592037,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004794973140279711,
      "loss": 5.8948,
      "mean_token_accuracy": 0.1784157082438469,
      "num_tokens": 32263839.0,
      "step": 12730
    },
    {
      "entropy": 6.280362319946289,
      "epoch": 1.4697057126370456,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047948010069814005,
      "loss": 5.8531,
      "mean_token_accuracy": 0.18112268596887587,
      "num_tokens": 32275539.0,
      "step": 12735
    },
    {
      "entropy": 6.245168733596802,
      "epoch": 1.4702827466820543,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004794628804907225,
      "loss": 5.8418,
      "mean_token_accuracy": 0.1846123605966568,
      "num_tokens": 32287421.0,
      "step": 12740
    },
    {
      "entropy": 6.235720252990722,
      "epoch": 1.470859780727063,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047944565340629755,
      "loss": 5.9316,
      "mean_token_accuracy": 0.17319110184907913,
      "num_tokens": 32300246.0,
      "step": 12745
    },
    {
      "entropy": 6.261264419555664,
      "epoch": 1.4714368147720716,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047942841944544453,
      "loss": 5.9044,
      "mean_token_accuracy": 0.1840224340558052,
      "num_tokens": 32311957.0,
      "step": 12750
    },
    {
      "entropy": 6.2243188381195065,
      "epoch": 1.4720138488170802,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004794111786087431,
      "loss": 5.8487,
      "mean_token_accuracy": 0.18722997307777406,
      "num_tokens": 32324710.0,
      "step": 12755
    },
    {
      "entropy": 6.24069504737854,
      "epoch": 1.4725908828620888,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004793939308967733,
      "loss": 5.9185,
      "mean_token_accuracy": 0.18768858760595322,
      "num_tokens": 32337252.0,
      "step": 12760
    },
    {
      "entropy": 6.366286373138427,
      "epoch": 1.4731679169070975,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004793766763101152,
      "loss": 5.8658,
      "mean_token_accuracy": 0.18304926604032518,
      "num_tokens": 32348957.0,
      "step": 12765
    },
    {
      "entropy": 6.271961307525634,
      "epoch": 1.4737449509521061,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00047935941484934905,
      "loss": 5.9254,
      "mean_token_accuracy": 0.17551231682300567,
      "num_tokens": 32362395.0,
      "step": 12770
    },
    {
      "entropy": 6.224658823013305,
      "epoch": 1.4743219849971148,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047934214651505547,
      "loss": 5.8244,
      "mean_token_accuracy": 0.18603911101818085,
      "num_tokens": 32375962.0,
      "step": 12775
    },
    {
      "entropy": 6.228481912612915,
      "epoch": 1.4748990190421236,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047932487130781533,
      "loss": 5.9266,
      "mean_token_accuracy": 0.17757474184036254,
      "num_tokens": 32387457.0,
      "step": 12780
    },
    {
      "entropy": 6.2456066608428955,
      "epoch": 1.475476053087132,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004793075892282097,
      "loss": 5.7948,
      "mean_token_accuracy": 0.18939292430877686,
      "num_tokens": 32400263.0,
      "step": 12785
    },
    {
      "entropy": 6.196566820144653,
      "epoch": 1.476053087132141,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004792903002768197,
      "loss": 5.88,
      "mean_token_accuracy": 0.1863648399710655,
      "num_tokens": 32412552.0,
      "step": 12790
    },
    {
      "entropy": 6.239522266387939,
      "epoch": 1.4766301211771495,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047927300445422687,
      "loss": 5.9273,
      "mean_token_accuracy": 0.18083892166614532,
      "num_tokens": 32424922.0,
      "step": 12795
    },
    {
      "entropy": 6.3519017696380615,
      "epoch": 1.4772071552221582,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004792557017610131,
      "loss": 5.9048,
      "mean_token_accuracy": 0.1829479455947876,
      "num_tokens": 32438164.0,
      "step": 12800
    },
    {
      "entropy": 6.257491540908814,
      "epoch": 1.4777841892671668,
      "grad_norm": 0.82421875,
      "learning_rate": 0.00047923839219776027,
      "loss": 5.7946,
      "mean_token_accuracy": 0.19040709137916564,
      "num_tokens": 32450718.0,
      "step": 12805
    },
    {
      "entropy": 6.214017868041992,
      "epoch": 1.4783612233121755,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004792210757650506,
      "loss": 5.801,
      "mean_token_accuracy": 0.1917333036661148,
      "num_tokens": 32464562.0,
      "step": 12810
    },
    {
      "entropy": 6.277191591262818,
      "epoch": 1.478938257357184,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047920375246346636,
      "loss": 5.9225,
      "mean_token_accuracy": 0.1778622269630432,
      "num_tokens": 32476707.0,
      "step": 12815
    },
    {
      "entropy": 6.297028541564941,
      "epoch": 1.4795152914021927,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047918642229359044,
      "loss": 5.9128,
      "mean_token_accuracy": 0.1832444354891777,
      "num_tokens": 32489979.0,
      "step": 12820
    },
    {
      "entropy": 6.2684142112731935,
      "epoch": 1.4800923254472014,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004791690852560056,
      "loss": 5.9225,
      "mean_token_accuracy": 0.18176766932010652,
      "num_tokens": 32502479.0,
      "step": 12825
    },
    {
      "entropy": 6.334451627731323,
      "epoch": 1.48066935949221,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000479151741351295,
      "loss": 5.8837,
      "mean_token_accuracy": 0.17994625121355057,
      "num_tokens": 32515075.0,
      "step": 12830
    },
    {
      "entropy": 6.245875501632691,
      "epoch": 1.4812463935372187,
      "grad_norm": 0.890625,
      "learning_rate": 0.000479134390580042,
      "loss": 5.8908,
      "mean_token_accuracy": 0.18218020349740982,
      "num_tokens": 32527647.0,
      "step": 12835
    },
    {
      "entropy": 6.315085697174072,
      "epoch": 1.4818234275822273,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047911703294283015,
      "loss": 5.893,
      "mean_token_accuracy": 0.18358839005231858,
      "num_tokens": 32539040.0,
      "step": 12840
    },
    {
      "entropy": 6.270166254043579,
      "epoch": 1.482400461627236,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047909966844024334,
      "loss": 5.8641,
      "mean_token_accuracy": 0.18732004165649413,
      "num_tokens": 32551657.0,
      "step": 12845
    },
    {
      "entropy": 6.286794996261596,
      "epoch": 1.4829774956722446,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00047908229707286547,
      "loss": 5.9952,
      "mean_token_accuracy": 0.17457903772592545,
      "num_tokens": 32564796.0,
      "step": 12850
    },
    {
      "entropy": 6.307313537597656,
      "epoch": 1.4835545297172534,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000479064918841281,
      "loss": 5.9408,
      "mean_token_accuracy": 0.17565248012542725,
      "num_tokens": 32576808.0,
      "step": 12855
    },
    {
      "entropy": 6.274211502075195,
      "epoch": 1.4841315637622619,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00047904753374607427,
      "loss": 5.9076,
      "mean_token_accuracy": 0.1847905546426773,
      "num_tokens": 32590855.0,
      "step": 12860
    },
    {
      "entropy": 6.323662185668946,
      "epoch": 1.4847085978072707,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047903014178783015,
      "loss": 5.9391,
      "mean_token_accuracy": 0.18114969730377198,
      "num_tokens": 32603505.0,
      "step": 12865
    },
    {
      "entropy": 6.231605434417725,
      "epoch": 1.4852856318522794,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004790127429671335,
      "loss": 5.8714,
      "mean_token_accuracy": 0.18963616490364074,
      "num_tokens": 32616366.0,
      "step": 12870
    },
    {
      "entropy": 6.272336816787719,
      "epoch": 1.485862665897288,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004789953372845697,
      "loss": 5.9698,
      "mean_token_accuracy": 0.17649832665920256,
      "num_tokens": 32628087.0,
      "step": 12875
    },
    {
      "entropy": 6.326451826095581,
      "epoch": 1.4864396999422966,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000478977924740724,
      "loss": 5.9331,
      "mean_token_accuracy": 0.17960140258073806,
      "num_tokens": 32642048.0,
      "step": 12880
    },
    {
      "entropy": 6.27869987487793,
      "epoch": 1.4870167339873053,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004789605053361821,
      "loss": 5.8562,
      "mean_token_accuracy": 0.1833032861351967,
      "num_tokens": 32653353.0,
      "step": 12885
    },
    {
      "entropy": 6.262209081649781,
      "epoch": 1.487593768032314,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004789430790715299,
      "loss": 5.9568,
      "mean_token_accuracy": 0.17719827294349672,
      "num_tokens": 32666733.0,
      "step": 12890
    },
    {
      "entropy": 6.301059150695801,
      "epoch": 1.4881708020773226,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047892564594735355,
      "loss": 5.8342,
      "mean_token_accuracy": 0.18444768339395523,
      "num_tokens": 32678666.0,
      "step": 12895
    },
    {
      "entropy": 6.259384965896606,
      "epoch": 1.4887478361223312,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047890820596423943,
      "loss": 5.9658,
      "mean_token_accuracy": 0.17606347054243088,
      "num_tokens": 32691622.0,
      "step": 12900
    },
    {
      "entropy": 6.326651382446289,
      "epoch": 1.4893248701673398,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000478890759122774,
      "loss": 5.9869,
      "mean_token_accuracy": 0.17802775651216507,
      "num_tokens": 32704469.0,
      "step": 12905
    },
    {
      "entropy": 6.282664012908936,
      "epoch": 1.4899019042123485,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004788733054235443,
      "loss": 5.8215,
      "mean_token_accuracy": 0.19414397031068803,
      "num_tokens": 32717119.0,
      "step": 12910
    },
    {
      "entropy": 6.119835138320923,
      "epoch": 1.4904789382573571,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047885584486713717,
      "loss": 5.7169,
      "mean_token_accuracy": 0.20453998297452927,
      "num_tokens": 32729687.0,
      "step": 12915
    },
    {
      "entropy": 6.295299911499024,
      "epoch": 1.491055972302366,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004788383774541399,
      "loss": 5.8896,
      "mean_token_accuracy": 0.1866808980703354,
      "num_tokens": 32741348.0,
      "step": 12920
    },
    {
      "entropy": 6.3022480487823485,
      "epoch": 1.4916330063473744,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004788209031851402,
      "loss": 5.9432,
      "mean_token_accuracy": 0.1811997190117836,
      "num_tokens": 32754665.0,
      "step": 12925
    },
    {
      "entropy": 6.2591334819793705,
      "epoch": 1.4922100403923833,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004788034220607256,
      "loss": 5.8275,
      "mean_token_accuracy": 0.18650948107242585,
      "num_tokens": 32767294.0,
      "step": 12930
    },
    {
      "entropy": 6.288479375839233,
      "epoch": 1.4927870744373917,
      "grad_norm": 0.875,
      "learning_rate": 0.00047878593408148405,
      "loss": 5.8586,
      "mean_token_accuracy": 0.1771461546421051,
      "num_tokens": 32779612.0,
      "step": 12935
    },
    {
      "entropy": 6.184573554992676,
      "epoch": 1.4933641084824005,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047876843924800393,
      "loss": 5.8075,
      "mean_token_accuracy": 0.18673338294029235,
      "num_tokens": 32791826.0,
      "step": 12940
    },
    {
      "entropy": 6.181496095657349,
      "epoch": 1.4939411425274092,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004787509375608735,
      "loss": 5.7439,
      "mean_token_accuracy": 0.20124246627092363,
      "num_tokens": 32805044.0,
      "step": 12945
    },
    {
      "entropy": 6.256733751296997,
      "epoch": 1.4945181765724178,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047873342902068157,
      "loss": 5.9011,
      "mean_token_accuracy": 0.18064064383506775,
      "num_tokens": 32816797.0,
      "step": 12950
    },
    {
      "entropy": 6.180535745620728,
      "epoch": 1.4950952106174265,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047871591362801694,
      "loss": 5.7638,
      "mean_token_accuracy": 0.19532770663499832,
      "num_tokens": 32828670.0,
      "step": 12955
    },
    {
      "entropy": 6.2887355327606205,
      "epoch": 1.495672244662435,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004786983913834687,
      "loss": 5.9582,
      "mean_token_accuracy": 0.1730899006128311,
      "num_tokens": 32841362.0,
      "step": 12960
    },
    {
      "entropy": 6.2650800228118895,
      "epoch": 1.4962492787074437,
      "grad_norm": 0.875,
      "learning_rate": 0.00047868086228762633,
      "loss": 5.8534,
      "mean_token_accuracy": 0.1818915158510208,
      "num_tokens": 32853974.0,
      "step": 12965
    },
    {
      "entropy": 6.21506028175354,
      "epoch": 1.4968263127524524,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047866332634107926,
      "loss": 5.8141,
      "mean_token_accuracy": 0.18658534288406373,
      "num_tokens": 32866683.0,
      "step": 12970
    },
    {
      "entropy": 6.257719087600708,
      "epoch": 1.497403346797461,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047864578354441743,
      "loss": 5.9065,
      "mean_token_accuracy": 0.18656679838895798,
      "num_tokens": 32879294.0,
      "step": 12975
    },
    {
      "entropy": 6.105445861816406,
      "epoch": 1.4979803808424696,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004786282338982308,
      "loss": 5.7767,
      "mean_token_accuracy": 0.19036744385957718,
      "num_tokens": 32891763.0,
      "step": 12980
    },
    {
      "entropy": 6.260914993286133,
      "epoch": 1.4985574148874783,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004786106774031096,
      "loss": 5.9026,
      "mean_token_accuracy": 0.17895008772611617,
      "num_tokens": 32903813.0,
      "step": 12985
    },
    {
      "entropy": 6.287360525131225,
      "epoch": 1.499134448932487,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004785931140596445,
      "loss": 5.8891,
      "mean_token_accuracy": 0.18722523152828216,
      "num_tokens": 32916847.0,
      "step": 12990
    },
    {
      "entropy": 6.234089374542236,
      "epoch": 1.4997114829774958,
      "grad_norm": 1.4375,
      "learning_rate": 0.00047857554386842606,
      "loss": 5.7573,
      "mean_token_accuracy": 0.19252282083034516,
      "num_tokens": 32929727.0,
      "step": 12995
    },
    {
      "entropy": 6.298026657104492,
      "epoch": 1.5002885170225042,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047855796683004534,
      "loss": 5.8771,
      "mean_token_accuracy": 0.17949557304382324,
      "num_tokens": 32942140.0,
      "step": 13000
    },
    {
      "entropy": 6.2665022850036625,
      "epoch": 1.500865551067513,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047854038294509356,
      "loss": 5.895,
      "mean_token_accuracy": 0.18179695904254914,
      "num_tokens": 32954622.0,
      "step": 13005
    },
    {
      "entropy": 6.249915838241577,
      "epoch": 1.5014425851125215,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004785227922141621,
      "loss": 5.8673,
      "mean_token_accuracy": 0.19787512719631195,
      "num_tokens": 32968118.0,
      "step": 13010
    },
    {
      "entropy": 6.30619330406189,
      "epoch": 1.5020196191575304,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047850519463784263,
      "loss": 5.8996,
      "mean_token_accuracy": 0.18216092735528946,
      "num_tokens": 32981275.0,
      "step": 13015
    },
    {
      "entropy": 6.299751853942871,
      "epoch": 1.502596653202539,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047848759021672693,
      "loss": 5.8707,
      "mean_token_accuracy": 0.1776880994439125,
      "num_tokens": 32995583.0,
      "step": 13020
    },
    {
      "entropy": 6.272617959976197,
      "epoch": 1.5031736872475476,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004784699789514073,
      "loss": 5.8751,
      "mean_token_accuracy": 0.1838984951376915,
      "num_tokens": 33008533.0,
      "step": 13025
    },
    {
      "entropy": 6.3040906429290775,
      "epoch": 1.5037507212925563,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000478452360842476,
      "loss": 5.9538,
      "mean_token_accuracy": 0.17843431681394578,
      "num_tokens": 33023507.0,
      "step": 13030
    },
    {
      "entropy": 6.262208557128906,
      "epoch": 1.504327755337565,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047843473589052557,
      "loss": 5.7962,
      "mean_token_accuracy": 0.18935810327529906,
      "num_tokens": 33035655.0,
      "step": 13035
    },
    {
      "entropy": 6.296089744567871,
      "epoch": 1.5049047893825735,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047841710409614893,
      "loss": 5.9034,
      "mean_token_accuracy": 0.18391674458980561,
      "num_tokens": 33047537.0,
      "step": 13040
    },
    {
      "entropy": 6.267186117172241,
      "epoch": 1.5054818234275822,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004783994654599389,
      "loss": 5.8741,
      "mean_token_accuracy": 0.18756560534238814,
      "num_tokens": 33060171.0,
      "step": 13045
    },
    {
      "entropy": 6.287681770324707,
      "epoch": 1.5060588574725908,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047838181998248897,
      "loss": 5.9377,
      "mean_token_accuracy": 0.18129630982875825,
      "num_tokens": 33073444.0,
      "step": 13050
    },
    {
      "entropy": 6.272368288040161,
      "epoch": 1.5066358915175995,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004783641676643925,
      "loss": 5.9148,
      "mean_token_accuracy": 0.17808404713869094,
      "num_tokens": 33086618.0,
      "step": 13055
    },
    {
      "entropy": 6.269512939453125,
      "epoch": 1.5072129255626083,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047834650850624334,
      "loss": 5.8223,
      "mean_token_accuracy": 0.18233495056629181,
      "num_tokens": 33098434.0,
      "step": 13060
    },
    {
      "entropy": 6.2975743293762205,
      "epoch": 1.5077899596076167,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004783288425086353,
      "loss": 5.8653,
      "mean_token_accuracy": 0.17588122934103012,
      "num_tokens": 33110253.0,
      "step": 13065
    },
    {
      "entropy": 6.25357494354248,
      "epoch": 1.5083669936526256,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004783111696721627,
      "loss": 5.895,
      "mean_token_accuracy": 0.18561285585165024,
      "num_tokens": 33122426.0,
      "step": 13070
    },
    {
      "entropy": 6.281819677352905,
      "epoch": 1.508944027697634,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004782934899974199,
      "loss": 5.7969,
      "mean_token_accuracy": 0.1863382488489151,
      "num_tokens": 33136609.0,
      "step": 13075
    },
    {
      "entropy": 6.271950101852417,
      "epoch": 1.5095210617426429,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047827580348500147,
      "loss": 5.9376,
      "mean_token_accuracy": 0.1765441879630089,
      "num_tokens": 33150351.0,
      "step": 13080
    },
    {
      "entropy": 6.281335973739624,
      "epoch": 1.5100980957876513,
      "grad_norm": 0.875,
      "learning_rate": 0.00047825811013550246,
      "loss": 5.8692,
      "mean_token_accuracy": 0.18624221086502074,
      "num_tokens": 33163256.0,
      "step": 13085
    },
    {
      "entropy": 6.315213537216186,
      "epoch": 1.5106751298326602,
      "grad_norm": 0.81640625,
      "learning_rate": 0.00047824040994951786,
      "loss": 5.923,
      "mean_token_accuracy": 0.18030782490968705,
      "num_tokens": 33177356.0,
      "step": 13090
    },
    {
      "entropy": 6.209644746780396,
      "epoch": 1.5112521638776688,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004782227029276429,
      "loss": 5.7956,
      "mean_token_accuracy": 0.19449451565742493,
      "num_tokens": 33189827.0,
      "step": 13095
    },
    {
      "entropy": 6.147610235214233,
      "epoch": 1.5118291979226774,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047820498907047345,
      "loss": 5.7136,
      "mean_token_accuracy": 0.19102597534656524,
      "num_tokens": 33202889.0,
      "step": 13100
    },
    {
      "entropy": 6.237274312973023,
      "epoch": 1.512406231967686,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000478187268378605,
      "loss": 5.7666,
      "mean_token_accuracy": 0.19528148472309112,
      "num_tokens": 33214973.0,
      "step": 13105
    },
    {
      "entropy": 6.284470319747925,
      "epoch": 1.5129832660126947,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047816954085263375,
      "loss": 5.8777,
      "mean_token_accuracy": 0.1867276817560196,
      "num_tokens": 33227515.0,
      "step": 13110
    },
    {
      "entropy": 6.277976608276367,
      "epoch": 1.5135603000577034,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004781518064931559,
      "loss": 5.865,
      "mean_token_accuracy": 0.19404051899909974,
      "num_tokens": 33240858.0,
      "step": 13115
    },
    {
      "entropy": 6.2066364765167235,
      "epoch": 1.514137334102712,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000478134065300768,
      "loss": 5.814,
      "mean_token_accuracy": 0.18544560819864273,
      "num_tokens": 33253234.0,
      "step": 13120
    },
    {
      "entropy": 6.265790319442749,
      "epoch": 1.5147143681477209,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004781163172760667,
      "loss": 5.868,
      "mean_token_accuracy": 0.17961650490760803,
      "num_tokens": 33265567.0,
      "step": 13125
    },
    {
      "entropy": 6.312713670730591,
      "epoch": 1.5152914021927293,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047809856241964893,
      "loss": 5.9157,
      "mean_token_accuracy": 0.1825166329741478,
      "num_tokens": 33278327.0,
      "step": 13130
    },
    {
      "entropy": 6.253669023513794,
      "epoch": 1.5158684362377381,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004780808007321119,
      "loss": 5.8751,
      "mean_token_accuracy": 0.18207056373357772,
      "num_tokens": 33291432.0,
      "step": 13135
    },
    {
      "entropy": 6.24861216545105,
      "epoch": 1.5164454702827466,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004780630322140531,
      "loss": 5.9236,
      "mean_token_accuracy": 0.18289124518632888,
      "num_tokens": 33304613.0,
      "step": 13140
    },
    {
      "entropy": 6.193721532821655,
      "epoch": 1.5170225043277554,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047804525686607,
      "loss": 5.7866,
      "mean_token_accuracy": 0.191617052257061,
      "num_tokens": 33317258.0,
      "step": 13145
    },
    {
      "entropy": 6.158862638473511,
      "epoch": 1.5175995383727638,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004780274746887606,
      "loss": 5.8259,
      "mean_token_accuracy": 0.18823667317628862,
      "num_tokens": 33329925.0,
      "step": 13150
    },
    {
      "entropy": 6.211736822128296,
      "epoch": 1.5181765724177727,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047800968568272284,
      "loss": 5.8135,
      "mean_token_accuracy": 0.19084462523460388,
      "num_tokens": 33343383.0,
      "step": 13155
    },
    {
      "entropy": 6.342063331604004,
      "epoch": 1.5187536064627813,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004779918898485551,
      "loss": 5.9389,
      "mean_token_accuracy": 0.1728406384587288,
      "num_tokens": 33355960.0,
      "step": 13160
    },
    {
      "entropy": 6.279878425598144,
      "epoch": 1.51933064050779,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047797408718685614,
      "loss": 5.8723,
      "mean_token_accuracy": 0.1856397956609726,
      "num_tokens": 33368130.0,
      "step": 13165
    },
    {
      "entropy": 6.26394944190979,
      "epoch": 1.5199076745527986,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047795627769822447,
      "loss": 5.9102,
      "mean_token_accuracy": 0.18491228222846984,
      "num_tokens": 33381548.0,
      "step": 13170
    },
    {
      "entropy": 6.253977060317993,
      "epoch": 1.5204847085978073,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047793846138325925,
      "loss": 5.8437,
      "mean_token_accuracy": 0.18650737404823303,
      "num_tokens": 33393866.0,
      "step": 13175
    },
    {
      "entropy": 6.282749080657959,
      "epoch": 1.521061742642816,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004779206382425598,
      "loss": 5.8853,
      "mean_token_accuracy": 0.19021858125925065,
      "num_tokens": 33405683.0,
      "step": 13180
    },
    {
      "entropy": 6.188709163665772,
      "epoch": 1.5216387766878245,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004779028082767253,
      "loss": 5.8336,
      "mean_token_accuracy": 0.1903842031955719,
      "num_tokens": 33417049.0,
      "step": 13185
    },
    {
      "entropy": 6.19663553237915,
      "epoch": 1.5222158107328332,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004778849714863557,
      "loss": 5.818,
      "mean_token_accuracy": 0.1941642716526985,
      "num_tokens": 33430106.0,
      "step": 13190
    },
    {
      "entropy": 6.229453086853027,
      "epoch": 1.5227928447778418,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004778671278720508,
      "loss": 5.8436,
      "mean_token_accuracy": 0.19037562906742095,
      "num_tokens": 33443337.0,
      "step": 13195
    },
    {
      "entropy": 6.173184394836426,
      "epoch": 1.5233698788228507,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004778492774344109,
      "loss": 5.8402,
      "mean_token_accuracy": 0.1905431404709816,
      "num_tokens": 33458198.0,
      "step": 13200
    },
    {
      "entropy": 6.1361161231994625,
      "epoch": 1.523946912867859,
      "grad_norm": 1.125,
      "learning_rate": 0.0004778314201740363,
      "loss": 5.6543,
      "mean_token_accuracy": 0.20741771310567855,
      "num_tokens": 33470169.0,
      "step": 13205
    },
    {
      "entropy": 6.263641309738159,
      "epoch": 1.524523946912868,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00047781355609152764,
      "loss": 5.9281,
      "mean_token_accuracy": 0.18706061244010924,
      "num_tokens": 33482568.0,
      "step": 13210
    },
    {
      "entropy": 6.227369165420532,
      "epoch": 1.5251009809578764,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004777956851874858,
      "loss": 5.8319,
      "mean_token_accuracy": 0.18869829922914505,
      "num_tokens": 33494092.0,
      "step": 13215
    },
    {
      "entropy": 6.318797826766968,
      "epoch": 1.5256780150028852,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047777780746251176,
      "loss": 5.8455,
      "mean_token_accuracy": 0.187092162668705,
      "num_tokens": 33506961.0,
      "step": 13220
    },
    {
      "entropy": 6.220376491546631,
      "epoch": 1.5262550490478937,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047775992291720687,
      "loss": 5.8747,
      "mean_token_accuracy": 0.18272871673107147,
      "num_tokens": 33519741.0,
      "step": 13225
    },
    {
      "entropy": 6.262176370620727,
      "epoch": 1.5268320830929025,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004777420315521728,
      "loss": 5.902,
      "mean_token_accuracy": 0.18136719912290572,
      "num_tokens": 33532965.0,
      "step": 13230
    },
    {
      "entropy": 6.2160125255584715,
      "epoch": 1.5274091171379112,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004777241333680111,
      "loss": 5.7799,
      "mean_token_accuracy": 0.19514340609312059,
      "num_tokens": 33545979.0,
      "step": 13235
    },
    {
      "entropy": 6.24297513961792,
      "epoch": 1.5279861511829198,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004777062283653239,
      "loss": 5.9098,
      "mean_token_accuracy": 0.18114701062440872,
      "num_tokens": 33558148.0,
      "step": 13240
    },
    {
      "entropy": 6.323754549026489,
      "epoch": 1.5285631852279284,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047768831654471333,
      "loss": 5.8499,
      "mean_token_accuracy": 0.19892722368240356,
      "num_tokens": 33570532.0,
      "step": 13245
    },
    {
      "entropy": 6.222899389266968,
      "epoch": 1.529140219272937,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047767039790678204,
      "loss": 5.8538,
      "mean_token_accuracy": 0.18385592550039292,
      "num_tokens": 33582902.0,
      "step": 13250
    },
    {
      "entropy": 6.201241111755371,
      "epoch": 1.5297172533179457,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047765247245213255,
      "loss": 5.7753,
      "mean_token_accuracy": 0.18826358169317245,
      "num_tokens": 33594765.0,
      "step": 13255
    },
    {
      "entropy": 6.269039964675903,
      "epoch": 1.5302942873629544,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004776345401813678,
      "loss": 5.8816,
      "mean_token_accuracy": 0.18214312493801116,
      "num_tokens": 33606999.0,
      "step": 13260
    },
    {
      "entropy": 6.277407455444336,
      "epoch": 1.5308713214079632,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047761660109509095,
      "loss": 5.8922,
      "mean_token_accuracy": 0.1835222512483597,
      "num_tokens": 33619331.0,
      "step": 13265
    },
    {
      "entropy": 6.246214199066162,
      "epoch": 1.5314483554529716,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004775986551939054,
      "loss": 5.8447,
      "mean_token_accuracy": 0.1858637139201164,
      "num_tokens": 33632009.0,
      "step": 13270
    },
    {
      "entropy": 6.24043116569519,
      "epoch": 1.5320253894979805,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00047758070247841465,
      "loss": 5.8261,
      "mean_token_accuracy": 0.18926349878311158,
      "num_tokens": 33644468.0,
      "step": 13275
    },
    {
      "entropy": 6.24837212562561,
      "epoch": 1.532602423542989,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00047756274294922266,
      "loss": 5.7428,
      "mean_token_accuracy": 0.20056139528751374,
      "num_tokens": 33657067.0,
      "step": 13280
    },
    {
      "entropy": 6.1877374172210695,
      "epoch": 1.5331794575879978,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004775447766069334,
      "loss": 5.815,
      "mean_token_accuracy": 0.1999603897333145,
      "num_tokens": 33669397.0,
      "step": 13285
    },
    {
      "entropy": 6.2190862655639645,
      "epoch": 1.5337564916330062,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047752680345215115,
      "loss": 5.8101,
      "mean_token_accuracy": 0.19042769074440002,
      "num_tokens": 33680997.0,
      "step": 13290
    },
    {
      "entropy": 6.231052017211914,
      "epoch": 1.534333525678015,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004775088234854805,
      "loss": 5.8126,
      "mean_token_accuracy": 0.18846020996570587,
      "num_tokens": 33693407.0,
      "step": 13295
    },
    {
      "entropy": 6.205740976333618,
      "epoch": 1.5349105597230237,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004774908367075262,
      "loss": 5.7643,
      "mean_token_accuracy": 0.2003784492611885,
      "num_tokens": 33705051.0,
      "step": 13300
    },
    {
      "entropy": 6.322617673873902,
      "epoch": 1.5354875937680323,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004774728431188931,
      "loss": 5.9165,
      "mean_token_accuracy": 0.17964526116847992,
      "num_tokens": 33717289.0,
      "step": 13305
    },
    {
      "entropy": 6.223282432556152,
      "epoch": 1.536064627813041,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047745484272018664,
      "loss": 5.8194,
      "mean_token_accuracy": 0.19722916185855865,
      "num_tokens": 33728494.0,
      "step": 13310
    },
    {
      "entropy": 6.249355602264404,
      "epoch": 1.5366416618580496,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004774368355120119,
      "loss": 5.9061,
      "mean_token_accuracy": 0.18718182742595674,
      "num_tokens": 33742065.0,
      "step": 13315
    },
    {
      "entropy": 6.222546768188477,
      "epoch": 1.5372186959030583,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000477418821494975,
      "loss": 5.7841,
      "mean_token_accuracy": 0.19122046381235122,
      "num_tokens": 33754312.0,
      "step": 13320
    },
    {
      "entropy": 6.1582118511199955,
      "epoch": 1.537795729948067,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004774008006696814,
      "loss": 5.8746,
      "mean_token_accuracy": 0.18873933851718902,
      "num_tokens": 33766938.0,
      "step": 13325
    },
    {
      "entropy": 6.318648099899292,
      "epoch": 1.5383727639930755,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004773827730367375,
      "loss": 5.8616,
      "mean_token_accuracy": 0.18600038439035416,
      "num_tokens": 33778494.0,
      "step": 13330
    },
    {
      "entropy": 6.209322786331176,
      "epoch": 1.5389497980380842,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047736473859674955,
      "loss": 5.7402,
      "mean_token_accuracy": 0.19402769953012466,
      "num_tokens": 33789932.0,
      "step": 13335
    },
    {
      "entropy": 6.17824158668518,
      "epoch": 1.539526832083093,
      "grad_norm": 0.859375,
      "learning_rate": 0.00047734669735032404,
      "loss": 5.8778,
      "mean_token_accuracy": 0.18440619707107545,
      "num_tokens": 33803552.0,
      "step": 13340
    },
    {
      "entropy": 6.2184912204742435,
      "epoch": 1.5401038661281015,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047732864929806796,
      "loss": 5.7919,
      "mean_token_accuracy": 0.19423859417438508,
      "num_tokens": 33816108.0,
      "step": 13345
    },
    {
      "entropy": 6.324096965789795,
      "epoch": 1.5406809001731103,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004773105944405883,
      "loss": 5.9157,
      "mean_token_accuracy": 0.18384704142808914,
      "num_tokens": 33829004.0,
      "step": 13350
    },
    {
      "entropy": 6.279313278198242,
      "epoch": 1.5412579342181187,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047729253277849226,
      "loss": 5.8603,
      "mean_token_accuracy": 0.18620263934135436,
      "num_tokens": 33840889.0,
      "step": 13355
    },
    {
      "entropy": 6.237594413757324,
      "epoch": 1.5418349682631276,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047727446431238734,
      "loss": 5.8495,
      "mean_token_accuracy": 0.1846572205424309,
      "num_tokens": 33855443.0,
      "step": 13360
    },
    {
      "entropy": 6.237881326675415,
      "epoch": 1.542412002308136,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004772563890428813,
      "loss": 5.8574,
      "mean_token_accuracy": 0.1851681649684906,
      "num_tokens": 33867365.0,
      "step": 13365
    },
    {
      "entropy": 6.26565899848938,
      "epoch": 1.5429890363531449,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004772383069705821,
      "loss": 5.8192,
      "mean_token_accuracy": 0.1881095588207245,
      "num_tokens": 33880438.0,
      "step": 13370
    },
    {
      "entropy": 6.24500675201416,
      "epoch": 1.5435660703981535,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004772202180960978,
      "loss": 5.8569,
      "mean_token_accuracy": 0.18066115081310272,
      "num_tokens": 33894812.0,
      "step": 13375
    },
    {
      "entropy": 6.266554307937622,
      "epoch": 1.5441431044431622,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047720212242003703,
      "loss": 5.8019,
      "mean_token_accuracy": 0.18672728389501572,
      "num_tokens": 33907619.0,
      "step": 13380
    },
    {
      "entropy": 6.238787841796875,
      "epoch": 1.5447201384881708,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047718401994300825,
      "loss": 5.8481,
      "mean_token_accuracy": 0.18345305174589158,
      "num_tokens": 33921109.0,
      "step": 13385
    },
    {
      "entropy": 6.24917516708374,
      "epoch": 1.5452971725331794,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047716591066562043,
      "loss": 5.8523,
      "mean_token_accuracy": 0.1810479074716568,
      "num_tokens": 33932874.0,
      "step": 13390
    },
    {
      "entropy": 6.288971090316773,
      "epoch": 1.545874206578188,
      "grad_norm": 0.8046875,
      "learning_rate": 0.00047714779458848255,
      "loss": 5.9029,
      "mean_token_accuracy": 0.1866563454270363,
      "num_tokens": 33945291.0,
      "step": 13395
    },
    {
      "entropy": 6.185093688964844,
      "epoch": 1.5464512406231967,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004771296717122041,
      "loss": 5.7211,
      "mean_token_accuracy": 0.19764399528503418,
      "num_tokens": 33958180.0,
      "step": 13400
    },
    {
      "entropy": 6.339099407196045,
      "epoch": 1.5470282746682056,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004771115420373945,
      "loss": 5.9641,
      "mean_token_accuracy": 0.16993046849966048,
      "num_tokens": 33971428.0,
      "step": 13405
    },
    {
      "entropy": 6.261291027069092,
      "epoch": 1.547605308713214,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047709340556466366,
      "loss": 5.8224,
      "mean_token_accuracy": 0.1897300213575363,
      "num_tokens": 33984464.0,
      "step": 13410
    },
    {
      "entropy": 6.1890631198883055,
      "epoch": 1.5481823427582229,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047707526229462144,
      "loss": 5.8218,
      "mean_token_accuracy": 0.18644251972436904,
      "num_tokens": 33997290.0,
      "step": 13415
    },
    {
      "entropy": 6.2182745933532715,
      "epoch": 1.5487593768032313,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047705711222787816,
      "loss": 5.8806,
      "mean_token_accuracy": 0.18388465493917466,
      "num_tokens": 34009249.0,
      "step": 13420
    },
    {
      "entropy": 6.25205717086792,
      "epoch": 1.5493364108482401,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00047703895536504423,
      "loss": 5.8159,
      "mean_token_accuracy": 0.18559204787015915,
      "num_tokens": 34022583.0,
      "step": 13425
    },
    {
      "entropy": 6.348111009597778,
      "epoch": 1.5499134448932486,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004770207917067305,
      "loss": 5.8522,
      "mean_token_accuracy": 0.1904935583472252,
      "num_tokens": 34036987.0,
      "step": 13430
    },
    {
      "entropy": 6.234028434753418,
      "epoch": 1.5504904789382574,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047700262125354765,
      "loss": 5.8977,
      "mean_token_accuracy": 0.18367722779512405,
      "num_tokens": 34050385.0,
      "step": 13435
    },
    {
      "entropy": 6.237712478637695,
      "epoch": 1.551067512983266,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047698444400610707,
      "loss": 5.8425,
      "mean_token_accuracy": 0.18808901757001878,
      "num_tokens": 34064006.0,
      "step": 13440
    },
    {
      "entropy": 6.250394725799561,
      "epoch": 1.5516445470282747,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047696625996502,
      "loss": 5.8815,
      "mean_token_accuracy": 0.18497573882341384,
      "num_tokens": 34075994.0,
      "step": 13445
    },
    {
      "entropy": 6.282526874542237,
      "epoch": 1.5522215810732833,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047694806913089815,
      "loss": 5.9045,
      "mean_token_accuracy": 0.17532371133565902,
      "num_tokens": 34087430.0,
      "step": 13450
    },
    {
      "entropy": 6.290673065185547,
      "epoch": 1.552798615118292,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004769298715043533,
      "loss": 5.8613,
      "mean_token_accuracy": 0.1850669115781784,
      "num_tokens": 34099951.0,
      "step": 13455
    },
    {
      "entropy": 6.233760786056519,
      "epoch": 1.5533756491633006,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004769116670859975,
      "loss": 5.8344,
      "mean_token_accuracy": 0.18648161441087724,
      "num_tokens": 34112048.0,
      "step": 13460
    },
    {
      "entropy": 6.323914813995361,
      "epoch": 1.5539526832083093,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047689345587644314,
      "loss": 5.9422,
      "mean_token_accuracy": 0.18842453211545945,
      "num_tokens": 34124820.0,
      "step": 13465
    },
    {
      "entropy": 6.161574268341065,
      "epoch": 1.554529717253318,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047687523787630256,
      "loss": 5.7362,
      "mean_token_accuracy": 0.19258127957582474,
      "num_tokens": 34137792.0,
      "step": 13470
    },
    {
      "entropy": 6.32023253440857,
      "epoch": 1.5551067512983265,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004768570130861887,
      "loss": 5.9345,
      "mean_token_accuracy": 0.1876098096370697,
      "num_tokens": 34149709.0,
      "step": 13475
    },
    {
      "entropy": 6.32637209892273,
      "epoch": 1.5556837853433354,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004768387815067144,
      "loss": 6.0034,
      "mean_token_accuracy": 0.1782930850982666,
      "num_tokens": 34163958.0,
      "step": 13480
    },
    {
      "entropy": 6.225791501998901,
      "epoch": 1.5562608193883438,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047682054313849304,
      "loss": 5.8474,
      "mean_token_accuracy": 0.18833521753549576,
      "num_tokens": 34176349.0,
      "step": 13485
    },
    {
      "entropy": 6.259426069259644,
      "epoch": 1.5568378534333527,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004768022979821379,
      "loss": 5.9238,
      "mean_token_accuracy": 0.17510734498500824,
      "num_tokens": 34188000.0,
      "step": 13490
    },
    {
      "entropy": 6.290373373031616,
      "epoch": 1.557414887478361,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004767840460382628,
      "loss": 5.8185,
      "mean_token_accuracy": 0.18521863967180252,
      "num_tokens": 34199675.0,
      "step": 13495
    },
    {
      "entropy": 6.278991556167602,
      "epoch": 1.55799192152337,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004767657873074815,
      "loss": 5.8116,
      "mean_token_accuracy": 0.19116545766592025,
      "num_tokens": 34212443.0,
      "step": 13500
    },
    {
      "entropy": 6.306239366531372,
      "epoch": 1.5585689555683784,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004767475217904081,
      "loss": 5.895,
      "mean_token_accuracy": 0.18091119527816774,
      "num_tokens": 34224664.0,
      "step": 13505
    },
    {
      "entropy": 6.108881855010987,
      "epoch": 1.5591459896133872,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047672924948765705,
      "loss": 5.7348,
      "mean_token_accuracy": 0.1988232597708702,
      "num_tokens": 34236418.0,
      "step": 13510
    },
    {
      "entropy": 6.273750591278076,
      "epoch": 1.5597230236583959,
      "grad_norm": 0.8125,
      "learning_rate": 0.00047671097039984293,
      "loss": 5.8232,
      "mean_token_accuracy": 0.18797826021909714,
      "num_tokens": 34249508.0,
      "step": 13515
    },
    {
      "entropy": 6.284932804107666,
      "epoch": 1.5603000577034045,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047669268452758053,
      "loss": 5.8736,
      "mean_token_accuracy": 0.19025820046663283,
      "num_tokens": 34263077.0,
      "step": 13520
    },
    {
      "entropy": 6.226687383651734,
      "epoch": 1.5608770917484132,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00047667439187148476,
      "loss": 5.746,
      "mean_token_accuracy": 0.19800482094287872,
      "num_tokens": 34276026.0,
      "step": 13525
    },
    {
      "entropy": 6.279954624176026,
      "epoch": 1.5614541257934218,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004766560924321711,
      "loss": 5.9137,
      "mean_token_accuracy": 0.176610004901886,
      "num_tokens": 34288398.0,
      "step": 13530
    },
    {
      "entropy": 6.289290523529052,
      "epoch": 1.5620311598384304,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047663778621025496,
      "loss": 5.8365,
      "mean_token_accuracy": 0.1856300339102745,
      "num_tokens": 34301064.0,
      "step": 13535
    },
    {
      "entropy": 6.285242891311645,
      "epoch": 1.562608193883439,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004766194732063519,
      "loss": 5.8561,
      "mean_token_accuracy": 0.18800656646490096,
      "num_tokens": 34312959.0,
      "step": 13540
    },
    {
      "entropy": 6.223040294647217,
      "epoch": 1.563185227928448,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00047660115342107814,
      "loss": 5.7966,
      "mean_token_accuracy": 0.18968226611614228,
      "num_tokens": 34325851.0,
      "step": 13545
    },
    {
      "entropy": 6.274968862533569,
      "epoch": 1.5637622619734564,
      "grad_norm": 0.875,
      "learning_rate": 0.00047658282685504973,
      "loss": 5.8144,
      "mean_token_accuracy": 0.18771258145570754,
      "num_tokens": 34338193.0,
      "step": 13550
    },
    {
      "entropy": 6.31854214668274,
      "epoch": 1.5643392960184652,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000476564493508883,
      "loss": 5.9123,
      "mean_token_accuracy": 0.1746668204665184,
      "num_tokens": 34350876.0,
      "step": 13555
    },
    {
      "entropy": 6.183593654632569,
      "epoch": 1.5649163300634736,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047654615338319466,
      "loss": 5.7979,
      "mean_token_accuracy": 0.1907200276851654,
      "num_tokens": 34363614.0,
      "step": 13560
    },
    {
      "entropy": 6.156565475463867,
      "epoch": 1.5654933641084825,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004765278064786016,
      "loss": 5.8547,
      "mean_token_accuracy": 0.19591960906982422,
      "num_tokens": 34376238.0,
      "step": 13565
    },
    {
      "entropy": 6.2628155708312985,
      "epoch": 1.566070398153491,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047650945279572085,
      "loss": 5.8353,
      "mean_token_accuracy": 0.19010327905416488,
      "num_tokens": 34389141.0,
      "step": 13570
    },
    {
      "entropy": 6.2319074153900145,
      "epoch": 1.5666474321984998,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004764910923351698,
      "loss": 5.7939,
      "mean_token_accuracy": 0.18960819393396378,
      "num_tokens": 34400844.0,
      "step": 13575
    },
    {
      "entropy": 6.147141647338867,
      "epoch": 1.5672244662435084,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047647272509756584,
      "loss": 5.7658,
      "mean_token_accuracy": 0.18838662803173065,
      "num_tokens": 34413558.0,
      "step": 13580
    },
    {
      "entropy": 6.2264729022979735,
      "epoch": 1.567801500288517,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004764543510835269,
      "loss": 5.8356,
      "mean_token_accuracy": 0.18928266018629075,
      "num_tokens": 34425189.0,
      "step": 13585
    },
    {
      "entropy": 6.242620420455933,
      "epoch": 1.5683785343335257,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000476435970293671,
      "loss": 5.7415,
      "mean_token_accuracy": 0.20520604401826859,
      "num_tokens": 34438649.0,
      "step": 13590
    },
    {
      "entropy": 6.205072546005249,
      "epoch": 1.5689555683785343,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047641758272861626,
      "loss": 5.771,
      "mean_token_accuracy": 0.18849036246538162,
      "num_tokens": 34451159.0,
      "step": 13595
    },
    {
      "entropy": 6.306575298309326,
      "epoch": 1.569532602423543,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004763991883889811,
      "loss": 5.857,
      "mean_token_accuracy": 0.18365900963544846,
      "num_tokens": 34462821.0,
      "step": 13600
    },
    {
      "entropy": 6.275446271896362,
      "epoch": 1.5701096364685516,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004763807872753843,
      "loss": 5.8675,
      "mean_token_accuracy": 0.18376044780015946,
      "num_tokens": 34474845.0,
      "step": 13605
    },
    {
      "entropy": 6.268782281875611,
      "epoch": 1.5706866705135603,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047636237938844484,
      "loss": 5.8814,
      "mean_token_accuracy": 0.1838698387145996,
      "num_tokens": 34488501.0,
      "step": 13610
    },
    {
      "entropy": 6.297438383102417,
      "epoch": 1.571263704558569,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004763439647287817,
      "loss": 5.8702,
      "mean_token_accuracy": 0.1816619634628296,
      "num_tokens": 34500760.0,
      "step": 13615
    },
    {
      "entropy": 6.238699054718017,
      "epoch": 1.5718407386035778,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004763255432970144,
      "loss": 5.8006,
      "mean_token_accuracy": 0.18592008501291274,
      "num_tokens": 34513857.0,
      "step": 13620
    },
    {
      "entropy": 6.256166076660156,
      "epoch": 1.5724177726485862,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047630711509376235,
      "loss": 5.8083,
      "mean_token_accuracy": 0.19307002574205398,
      "num_tokens": 34526091.0,
      "step": 13625
    },
    {
      "entropy": 6.281120157241821,
      "epoch": 1.572994806693595,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004762886801196455,
      "loss": 5.8687,
      "mean_token_accuracy": 0.18041985780000686,
      "num_tokens": 34539353.0,
      "step": 13630
    },
    {
      "entropy": 6.1506201267242435,
      "epoch": 1.5735718407386035,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047627023837528386,
      "loss": 5.738,
      "mean_token_accuracy": 0.19625691175460816,
      "num_tokens": 34551633.0,
      "step": 13635
    },
    {
      "entropy": 6.261639833450317,
      "epoch": 1.5741488747836123,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047625178986129775,
      "loss": 5.8779,
      "mean_token_accuracy": 0.19015572518110274,
      "num_tokens": 34563108.0,
      "step": 13640
    },
    {
      "entropy": 6.304862546920776,
      "epoch": 1.5747259088286207,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004762333345783078,
      "loss": 5.9546,
      "mean_token_accuracy": 0.18061692714691163,
      "num_tokens": 34576014.0,
      "step": 13645
    },
    {
      "entropy": 6.308314466476441,
      "epoch": 1.5753029428736296,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00047621487252693436,
      "loss": 5.8884,
      "mean_token_accuracy": 0.18400309830904008,
      "num_tokens": 34590025.0,
      "step": 13650
    },
    {
      "entropy": 6.301814174652099,
      "epoch": 1.5758799769186382,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047619640370779876,
      "loss": 5.8264,
      "mean_token_accuracy": 0.1879364013671875,
      "num_tokens": 34601627.0,
      "step": 13655
    },
    {
      "entropy": 6.214540767669678,
      "epoch": 1.5764570109636469,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047617792812152207,
      "loss": 5.7806,
      "mean_token_accuracy": 0.1942703977227211,
      "num_tokens": 34613712.0,
      "step": 13660
    },
    {
      "entropy": 6.289804124832154,
      "epoch": 1.5770340450086555,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004761594457687256,
      "loss": 5.9117,
      "mean_token_accuracy": 0.17855042666196824,
      "num_tokens": 34625861.0,
      "step": 13665
    },
    {
      "entropy": 6.350810384750366,
      "epoch": 1.5776110790536642,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004761409566500313,
      "loss": 5.9667,
      "mean_token_accuracy": 0.17831951528787612,
      "num_tokens": 34638457.0,
      "step": 13670
    },
    {
      "entropy": 6.248854923248291,
      "epoch": 1.5781881130986728,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00047612246076606066,
      "loss": 5.8251,
      "mean_token_accuracy": 0.1864610120654106,
      "num_tokens": 34652037.0,
      "step": 13675
    },
    {
      "entropy": 6.292226791381836,
      "epoch": 1.5787651471436814,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047610395811743594,
      "loss": 5.8388,
      "mean_token_accuracy": 0.18664174824953078,
      "num_tokens": 34664902.0,
      "step": 13680
    },
    {
      "entropy": 6.294170618057251,
      "epoch": 1.5793421811886903,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047608544870477956,
      "loss": 5.8544,
      "mean_token_accuracy": 0.18502934873104096,
      "num_tokens": 34676772.0,
      "step": 13685
    },
    {
      "entropy": 6.196129941940308,
      "epoch": 1.5799192152336987,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047606693252871395,
      "loss": 5.8733,
      "mean_token_accuracy": 0.19259540289640426,
      "num_tokens": 34690013.0,
      "step": 13690
    },
    {
      "entropy": 6.249138736724854,
      "epoch": 1.5804962492787076,
      "grad_norm": 0.921875,
      "learning_rate": 0.000476048409589862,
      "loss": 5.8291,
      "mean_token_accuracy": 0.1854974612593651,
      "num_tokens": 34702855.0,
      "step": 13695
    },
    {
      "entropy": 6.278041362762451,
      "epoch": 1.581073283323716,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004760298798888466,
      "loss": 5.905,
      "mean_token_accuracy": 0.18881870657205582,
      "num_tokens": 34716926.0,
      "step": 13700
    },
    {
      "entropy": 6.328575277328492,
      "epoch": 1.5816503173687249,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004760113434262911,
      "loss": 5.9069,
      "mean_token_accuracy": 0.17344938218593597,
      "num_tokens": 34729528.0,
      "step": 13705
    },
    {
      "entropy": 6.30189962387085,
      "epoch": 1.5822273514137333,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047599280020281894,
      "loss": 5.9606,
      "mean_token_accuracy": 0.18056074678897857,
      "num_tokens": 34742016.0,
      "step": 13710
    },
    {
      "entropy": 6.255933570861816,
      "epoch": 1.5828043854587421,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047597425021905364,
      "loss": 5.7832,
      "mean_token_accuracy": 0.1908837854862213,
      "num_tokens": 34754670.0,
      "step": 13715
    },
    {
      "entropy": 6.209108114242554,
      "epoch": 1.5833814195037508,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004759556934756194,
      "loss": 5.7859,
      "mean_token_accuracy": 0.19513892382383347,
      "num_tokens": 34767411.0,
      "step": 13720
    },
    {
      "entropy": 6.266744232177734,
      "epoch": 1.5839584535487594,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047593712997314017,
      "loss": 5.8148,
      "mean_token_accuracy": 0.18857741206884385,
      "num_tokens": 34780498.0,
      "step": 13725
    },
    {
      "entropy": 6.272948980331421,
      "epoch": 1.584535487593768,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00047591855971224035,
      "loss": 5.874,
      "mean_token_accuracy": 0.1850752890110016,
      "num_tokens": 34792918.0,
      "step": 13730
    },
    {
      "entropy": 6.206170129776001,
      "epoch": 1.5851125216387767,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004758999826935446,
      "loss": 5.8148,
      "mean_token_accuracy": 0.18377465158700942,
      "num_tokens": 34805631.0,
      "step": 13735
    },
    {
      "entropy": 6.317245721817017,
      "epoch": 1.5856895556837853,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004758813989176778,
      "loss": 5.9543,
      "mean_token_accuracy": 0.17631225883960724,
      "num_tokens": 34818790.0,
      "step": 13740
    },
    {
      "entropy": 6.219623231887818,
      "epoch": 1.586266589728794,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047586280838526483,
      "loss": 5.76,
      "mean_token_accuracy": 0.196747325360775,
      "num_tokens": 34831701.0,
      "step": 13745
    },
    {
      "entropy": 6.245278549194336,
      "epoch": 1.5868436237738026,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004758442110969312,
      "loss": 5.8441,
      "mean_token_accuracy": 0.187895467877388,
      "num_tokens": 34843628.0,
      "step": 13750
    },
    {
      "entropy": 6.257695913314819,
      "epoch": 1.5874206578188113,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004758256070533022,
      "loss": 5.8182,
      "mean_token_accuracy": 0.1816932126879692,
      "num_tokens": 34856327.0,
      "step": 13755
    },
    {
      "entropy": 6.266289138793946,
      "epoch": 1.5879976918638201,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004758069962550037,
      "loss": 5.8388,
      "mean_token_accuracy": 0.1833254799246788,
      "num_tokens": 34869973.0,
      "step": 13760
    },
    {
      "entropy": 6.166522741317749,
      "epoch": 1.5885747259088285,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00047578837870266156,
      "loss": 5.8146,
      "mean_token_accuracy": 0.20029770731925964,
      "num_tokens": 34884633.0,
      "step": 13765
    },
    {
      "entropy": 6.279387521743774,
      "epoch": 1.5891517599538374,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047576975439690206,
      "loss": 5.7988,
      "mean_token_accuracy": 0.19542221873998641,
      "num_tokens": 34897364.0,
      "step": 13770
    },
    {
      "entropy": 6.172811460494995,
      "epoch": 1.5897287939988458,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047575112333835164,
      "loss": 5.7366,
      "mean_token_accuracy": 0.19509423673152923,
      "num_tokens": 34909789.0,
      "step": 13775
    },
    {
      "entropy": 6.235009431838989,
      "epoch": 1.5903058280438547,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047573248552763684,
      "loss": 5.8017,
      "mean_token_accuracy": 0.18663742691278457,
      "num_tokens": 34923127.0,
      "step": 13780
    },
    {
      "entropy": 6.315503215789795,
      "epoch": 1.590882862088863,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047571384096538474,
      "loss": 5.897,
      "mean_token_accuracy": 0.18442563265562056,
      "num_tokens": 34935710.0,
      "step": 13785
    },
    {
      "entropy": 6.156521940231324,
      "epoch": 1.591459896133872,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004756951896522222,
      "loss": 5.7116,
      "mean_token_accuracy": 0.18641358166933059,
      "num_tokens": 34947566.0,
      "step": 13790
    },
    {
      "entropy": 6.279241418838501,
      "epoch": 1.5920369301788806,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004756765315887766,
      "loss": 5.8255,
      "mean_token_accuracy": 0.18567173629999162,
      "num_tokens": 34960008.0,
      "step": 13795
    },
    {
      "entropy": 6.233031320571899,
      "epoch": 1.5926139642238892,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004756578667756756,
      "loss": 5.7736,
      "mean_token_accuracy": 0.19006529152393342,
      "num_tokens": 34972299.0,
      "step": 13800
    },
    {
      "entropy": 6.250890064239502,
      "epoch": 1.5931909982688979,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004756391952135469,
      "loss": 5.8551,
      "mean_token_accuracy": 0.18757143914699553,
      "num_tokens": 34984252.0,
      "step": 13805
    },
    {
      "entropy": 6.297386407852173,
      "epoch": 1.5937680323139065,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047562051690301855,
      "loss": 5.8396,
      "mean_token_accuracy": 0.1920837566256523,
      "num_tokens": 34997711.0,
      "step": 13810
    },
    {
      "entropy": 6.287389135360717,
      "epoch": 1.5943450663589152,
      "grad_norm": 0.828125,
      "learning_rate": 0.00047560183184471873,
      "loss": 5.8235,
      "mean_token_accuracy": 0.1908173903822899,
      "num_tokens": 35010440.0,
      "step": 13815
    },
    {
      "entropy": 6.180474519729614,
      "epoch": 1.5949221004039238,
      "grad_norm": 0.90625,
      "learning_rate": 0.000475583140039276,
      "loss": 5.7412,
      "mean_token_accuracy": 0.19224562197923661,
      "num_tokens": 35022493.0,
      "step": 13820
    },
    {
      "entropy": 6.2107096195220945,
      "epoch": 1.5954991344489324,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00047556444148731897,
      "loss": 5.8058,
      "mean_token_accuracy": 0.191436767578125,
      "num_tokens": 35035756.0,
      "step": 13825
    },
    {
      "entropy": 6.283245420455932,
      "epoch": 1.596076168493941,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004755457361894766,
      "loss": 5.8294,
      "mean_token_accuracy": 0.19067344516515733,
      "num_tokens": 35046946.0,
      "step": 13830
    },
    {
      "entropy": 6.282238578796386,
      "epoch": 1.59665320253895,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047552702414637793,
      "loss": 5.8078,
      "mean_token_accuracy": 0.17738350331783295,
      "num_tokens": 35060803.0,
      "step": 13835
    },
    {
      "entropy": 6.21819543838501,
      "epoch": 1.5972302365839584,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047550830535865245,
      "loss": 5.7592,
      "mean_token_accuracy": 0.18955118656158448,
      "num_tokens": 35072834.0,
      "step": 13840
    },
    {
      "entropy": 6.317784261703491,
      "epoch": 1.5978072706289672,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047548957982692974,
      "loss": 5.9147,
      "mean_token_accuracy": 0.19114601314067842,
      "num_tokens": 35084872.0,
      "step": 13845
    },
    {
      "entropy": 6.290470886230469,
      "epoch": 1.5983843046739756,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004754708475518396,
      "loss": 5.9259,
      "mean_token_accuracy": 0.1835696965456009,
      "num_tokens": 35098070.0,
      "step": 13850
    },
    {
      "entropy": 6.284690713882446,
      "epoch": 1.5989613387189845,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047545210853401214,
      "loss": 5.9136,
      "mean_token_accuracy": 0.1826196640729904,
      "num_tokens": 35111614.0,
      "step": 13855
    },
    {
      "entropy": 6.336691093444824,
      "epoch": 1.599538372763993,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047543336277407753,
      "loss": 5.9181,
      "mean_token_accuracy": 0.17983610183000565,
      "num_tokens": 35124061.0,
      "step": 13860
    },
    {
      "entropy": 6.252399635314942,
      "epoch": 1.6001154068090018,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047541461027266624,
      "loss": 5.7755,
      "mean_token_accuracy": 0.1871828705072403,
      "num_tokens": 35135853.0,
      "step": 13865
    },
    {
      "entropy": 6.2618138790130615,
      "epoch": 1.6006924408540104,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004753958510304091,
      "loss": 5.8086,
      "mean_token_accuracy": 0.18253249824047088,
      "num_tokens": 35147811.0,
      "step": 13870
    },
    {
      "entropy": 6.262500810623169,
      "epoch": 1.601269474899019,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00047537708504793714,
      "loss": 5.8513,
      "mean_token_accuracy": 0.19067204594612122,
      "num_tokens": 35161738.0,
      "step": 13875
    },
    {
      "entropy": 6.209204816818238,
      "epoch": 1.6018465089440277,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00047535831232588146,
      "loss": 5.8214,
      "mean_token_accuracy": 0.19552301466464997,
      "num_tokens": 35174777.0,
      "step": 13880
    },
    {
      "entropy": 6.268672370910645,
      "epoch": 1.6024235429890363,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047533953286487345,
      "loss": 5.8999,
      "mean_token_accuracy": 0.18508652448654175,
      "num_tokens": 35186814.0,
      "step": 13885
    },
    {
      "entropy": 6.266240739822388,
      "epoch": 1.603000577034045,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004753207466655447,
      "loss": 5.7987,
      "mean_token_accuracy": 0.1863708809018135,
      "num_tokens": 35199229.0,
      "step": 13890
    },
    {
      "entropy": 6.18424620628357,
      "epoch": 1.6035776110790536,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004753019537285273,
      "loss": 5.8129,
      "mean_token_accuracy": 0.18863163590431214,
      "num_tokens": 35212110.0,
      "step": 13895
    },
    {
      "entropy": 6.306322956085205,
      "epoch": 1.6041546451240625,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047528315405445296,
      "loss": 5.9304,
      "mean_token_accuracy": 0.1795900583267212,
      "num_tokens": 35226203.0,
      "step": 13900
    },
    {
      "entropy": 6.241583347320557,
      "epoch": 1.604731679169071,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047526434764395435,
      "loss": 5.8003,
      "mean_token_accuracy": 0.18944022357463836,
      "num_tokens": 35238537.0,
      "step": 13905
    },
    {
      "entropy": 6.220981693267822,
      "epoch": 1.6053087132140798,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047524553449766386,
      "loss": 5.8124,
      "mean_token_accuracy": 0.19010360538959503,
      "num_tokens": 35251444.0,
      "step": 13910
    },
    {
      "entropy": 6.274720191955566,
      "epoch": 1.6058857472590882,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00047522671461621433,
      "loss": 5.8222,
      "mean_token_accuracy": 0.19289374202489853,
      "num_tokens": 35264159.0,
      "step": 13915
    },
    {
      "entropy": 6.301994943618775,
      "epoch": 1.606462781304097,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004752078880002386,
      "loss": 5.8559,
      "mean_token_accuracy": 0.1832738384604454,
      "num_tokens": 35277475.0,
      "step": 13920
    },
    {
      "entropy": 6.257745456695557,
      "epoch": 1.6070398153491054,
      "grad_norm": 0.78125,
      "learning_rate": 0.00047518905465037005,
      "loss": 5.7422,
      "mean_token_accuracy": 0.1939549922943115,
      "num_tokens": 35290739.0,
      "step": 13925
    },
    {
      "entropy": 6.182768154144287,
      "epoch": 1.6076168493941143,
      "grad_norm": 0.859375,
      "learning_rate": 0.00047517021456724214,
      "loss": 5.8388,
      "mean_token_accuracy": 0.18307804465293884,
      "num_tokens": 35303053.0,
      "step": 13930
    },
    {
      "entropy": 6.240389156341553,
      "epoch": 1.608193883439123,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047515136775148843,
      "loss": 5.8248,
      "mean_token_accuracy": 0.1935925528407097,
      "num_tokens": 35316105.0,
      "step": 13935
    },
    {
      "entropy": 6.30060887336731,
      "epoch": 1.6087709174841316,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004751325142037429,
      "loss": 5.8379,
      "mean_token_accuracy": 0.187453256547451,
      "num_tokens": 35329144.0,
      "step": 13940
    },
    {
      "entropy": 6.1760862350463865,
      "epoch": 1.6093479515291402,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047511365392463974,
      "loss": 5.7002,
      "mean_token_accuracy": 0.19486261308193206,
      "num_tokens": 35342447.0,
      "step": 13945
    },
    {
      "entropy": 6.313580656051636,
      "epoch": 1.6099249855741489,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047509478691481317,
      "loss": 5.8407,
      "mean_token_accuracy": 0.1903296336531639,
      "num_tokens": 35354073.0,
      "step": 13950
    },
    {
      "entropy": 6.2513103008270265,
      "epoch": 1.6105020196191575,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047507591317489783,
      "loss": 5.7867,
      "mean_token_accuracy": 0.19361701011657714,
      "num_tokens": 35367286.0,
      "step": 13955
    },
    {
      "entropy": 6.221856594085693,
      "epoch": 1.6110790536641661,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004750570327055286,
      "loss": 5.8479,
      "mean_token_accuracy": 0.19463634341955185,
      "num_tokens": 35379262.0,
      "step": 13960
    },
    {
      "entropy": 6.240395164489746,
      "epoch": 1.6116560877091748,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00047503814550734034,
      "loss": 5.7635,
      "mean_token_accuracy": 0.19315397292375563,
      "num_tokens": 35392423.0,
      "step": 13965
    },
    {
      "entropy": 6.289327716827392,
      "epoch": 1.6122331217541834,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004750192515809685,
      "loss": 5.8906,
      "mean_token_accuracy": 0.18289685845375062,
      "num_tokens": 35405007.0,
      "step": 13970
    },
    {
      "entropy": 6.206415891647339,
      "epoch": 1.6128101557991923,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047500035092704843,
      "loss": 5.7802,
      "mean_token_accuracy": 0.19542228132486344,
      "num_tokens": 35418181.0,
      "step": 13975
    },
    {
      "entropy": 6.238216876983643,
      "epoch": 1.6133871898442007,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004749814435462159,
      "loss": 5.7301,
      "mean_token_accuracy": 0.19913457781076432,
      "num_tokens": 35432495.0,
      "step": 13980
    },
    {
      "entropy": 6.206194639205933,
      "epoch": 1.6139642238892096,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004749625294391069,
      "loss": 5.7792,
      "mean_token_accuracy": 0.20099081993103027,
      "num_tokens": 35446327.0,
      "step": 13985
    },
    {
      "entropy": 6.168758678436279,
      "epoch": 1.614541257934218,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00047494360860635755,
      "loss": 5.8103,
      "mean_token_accuracy": 0.19489537924528122,
      "num_tokens": 35459964.0,
      "step": 13990
    },
    {
      "entropy": 6.240785598754883,
      "epoch": 1.6151182919792268,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004749246810486042,
      "loss": 5.7909,
      "mean_token_accuracy": 0.18760397285223007,
      "num_tokens": 35471704.0,
      "step": 13995
    },
    {
      "entropy": 6.281989240646363,
      "epoch": 1.6156953260242353,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004749057467664836,
      "loss": 5.8725,
      "mean_token_accuracy": 0.18600586652755738,
      "num_tokens": 35483772.0,
      "step": 14000
    },
    {
      "entropy": 6.2468711853027346,
      "epoch": 1.6162723600692441,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047488680576063247,
      "loss": 5.7693,
      "mean_token_accuracy": 0.20100188106298447,
      "num_tokens": 35497114.0,
      "step": 14005
    },
    {
      "entropy": 6.262406492233277,
      "epoch": 1.6168493941142528,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004748678580316878,
      "loss": 5.8674,
      "mean_token_accuracy": 0.17838740050792695,
      "num_tokens": 35509988.0,
      "step": 14010
    },
    {
      "entropy": 6.265255975723266,
      "epoch": 1.6174264281592614,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047484890358028714,
      "loss": 5.8181,
      "mean_token_accuracy": 0.19672373831272125,
      "num_tokens": 35522613.0,
      "step": 14015
    },
    {
      "entropy": 6.229588556289673,
      "epoch": 1.61800346220427,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004748299424070678,
      "loss": 5.7973,
      "mean_token_accuracy": 0.18971120566129684,
      "num_tokens": 35534618.0,
      "step": 14020
    },
    {
      "entropy": 6.20670599937439,
      "epoch": 1.6185804962492787,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004748109745126677,
      "loss": 5.7901,
      "mean_token_accuracy": 0.18894937485456467,
      "num_tokens": 35547461.0,
      "step": 14025
    },
    {
      "entropy": 6.291375732421875,
      "epoch": 1.6191575302942873,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047479199989772464,
      "loss": 5.7931,
      "mean_token_accuracy": 0.19160605520009993,
      "num_tokens": 35560026.0,
      "step": 14030
    },
    {
      "entropy": 6.249812459945678,
      "epoch": 1.619734564339296,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000474773018562877,
      "loss": 5.7607,
      "mean_token_accuracy": 0.18875414580106736,
      "num_tokens": 35571700.0,
      "step": 14035
    },
    {
      "entropy": 6.2016417503356935,
      "epoch": 1.6203115983843048,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047475403050876297,
      "loss": 5.8097,
      "mean_token_accuracy": 0.19091146290302277,
      "num_tokens": 35585366.0,
      "step": 14040
    },
    {
      "entropy": 6.299675703048706,
      "epoch": 1.6208886324293132,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004747350357360214,
      "loss": 5.9019,
      "mean_token_accuracy": 0.18595071732997895,
      "num_tokens": 35597809.0,
      "step": 14045
    },
    {
      "entropy": 6.300312948226929,
      "epoch": 1.621465666474322,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004747160342452912,
      "loss": 5.8565,
      "mean_token_accuracy": 0.18761351108551025,
      "num_tokens": 35610410.0,
      "step": 14050
    },
    {
      "entropy": 6.226084470748901,
      "epoch": 1.6220427005193305,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047469702603721134,
      "loss": 5.8282,
      "mean_token_accuracy": 0.18603773713111876,
      "num_tokens": 35624247.0,
      "step": 14055
    },
    {
      "entropy": 6.1976320266723635,
      "epoch": 1.6226197345643394,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004746780111124212,
      "loss": 5.762,
      "mean_token_accuracy": 0.18759202659130098,
      "num_tokens": 35636658.0,
      "step": 14060
    },
    {
      "entropy": 6.324990224838257,
      "epoch": 1.6231967686093478,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047465898947156033,
      "loss": 5.8929,
      "mean_token_accuracy": 0.18402266055345534,
      "num_tokens": 35649179.0,
      "step": 14065
    },
    {
      "entropy": 6.220640850067139,
      "epoch": 1.6237738026543567,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047463996111526854,
      "loss": 5.8193,
      "mean_token_accuracy": 0.1952624127268791,
      "num_tokens": 35661709.0,
      "step": 14070
    },
    {
      "entropy": 6.236063861846924,
      "epoch": 1.6243508366993653,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047462092604418576,
      "loss": 5.7389,
      "mean_token_accuracy": 0.19360555112361907,
      "num_tokens": 35673337.0,
      "step": 14075
    },
    {
      "entropy": 6.282483959197998,
      "epoch": 1.624927870744374,
      "grad_norm": 0.7890625,
      "learning_rate": 0.00047460188425895236,
      "loss": 5.9647,
      "mean_token_accuracy": 0.18415629416704177,
      "num_tokens": 35688726.0,
      "step": 14080
    },
    {
      "entropy": 6.309821939468383,
      "epoch": 1.6255049047893826,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00047458283576020874,
      "loss": 5.9004,
      "mean_token_accuracy": 0.1806803300976753,
      "num_tokens": 35702514.0,
      "step": 14085
    },
    {
      "entropy": 6.324135208129883,
      "epoch": 1.6260819388343912,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00047456378054859554,
      "loss": 5.9053,
      "mean_token_accuracy": 0.1814290851354599,
      "num_tokens": 35715153.0,
      "step": 14090
    },
    {
      "entropy": 6.226585912704468,
      "epoch": 1.6266589728793999,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047454471862475375,
      "loss": 5.9194,
      "mean_token_accuracy": 0.19234729558229446,
      "num_tokens": 35725893.0,
      "step": 14095
    },
    {
      "entropy": 6.161426734924317,
      "epoch": 1.6272360069244085,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047452564998932446,
      "loss": 5.7387,
      "mean_token_accuracy": 0.20032327473163605,
      "num_tokens": 35738308.0,
      "step": 14100
    },
    {
      "entropy": 6.28751220703125,
      "epoch": 1.6278130409694171,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000474506574642949,
      "loss": 5.8991,
      "mean_token_accuracy": 0.18265990316867828,
      "num_tokens": 35750538.0,
      "step": 14105
    },
    {
      "entropy": 6.285173988342285,
      "epoch": 1.6283900750144258,
      "grad_norm": 0.84375,
      "learning_rate": 0.000474487492586269,
      "loss": 5.8817,
      "mean_token_accuracy": 0.18699338883161545,
      "num_tokens": 35763974.0,
      "step": 14110
    },
    {
      "entropy": 6.2182127952575685,
      "epoch": 1.6289671090594346,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004744684038199263,
      "loss": 5.8087,
      "mean_token_accuracy": 0.20204776525497437,
      "num_tokens": 35775967.0,
      "step": 14115
    },
    {
      "entropy": 6.2255031108856205,
      "epoch": 1.629544143104443,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047444930834456293,
      "loss": 5.7229,
      "mean_token_accuracy": 0.20009350329637526,
      "num_tokens": 35788231.0,
      "step": 14120
    },
    {
      "entropy": 6.273572874069214,
      "epoch": 1.630121177149452,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004744302061608212,
      "loss": 5.8569,
      "mean_token_accuracy": 0.18506935536861419,
      "num_tokens": 35801752.0,
      "step": 14125
    },
    {
      "entropy": 6.258766078948975,
      "epoch": 1.6306982111944603,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047441109726934345,
      "loss": 5.8067,
      "mean_token_accuracy": 0.18814273327589034,
      "num_tokens": 35814849.0,
      "step": 14130
    },
    {
      "entropy": 6.2243156909942625,
      "epoch": 1.6312752452394692,
      "grad_norm": 0.84375,
      "learning_rate": 0.00047439198167077256,
      "loss": 5.7987,
      "mean_token_accuracy": 0.18756136149168015,
      "num_tokens": 35827839.0,
      "step": 14135
    },
    {
      "entropy": 6.236831903457642,
      "epoch": 1.6318522792844776,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004743728593657514,
      "loss": 5.7696,
      "mean_token_accuracy": 0.19610755294561386,
      "num_tokens": 35839683.0,
      "step": 14140
    },
    {
      "entropy": 6.19918909072876,
      "epoch": 1.6324293133294865,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004743537303549231,
      "loss": 5.7473,
      "mean_token_accuracy": 0.20063692033290864,
      "num_tokens": 35853523.0,
      "step": 14145
    },
    {
      "entropy": 6.210893821716309,
      "epoch": 1.6330063473744951,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004743345946389311,
      "loss": 5.8518,
      "mean_token_accuracy": 0.19142117202281952,
      "num_tokens": 35865189.0,
      "step": 14150
    },
    {
      "entropy": 6.295489263534546,
      "epoch": 1.6335833814195038,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047431545221841907,
      "loss": 5.8938,
      "mean_token_accuracy": 0.18685827851295472,
      "num_tokens": 35877352.0,
      "step": 14155
    },
    {
      "entropy": 6.2898290157318115,
      "epoch": 1.6341604154645124,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047429630309403086,
      "loss": 5.8479,
      "mean_token_accuracy": 0.18105699270963668,
      "num_tokens": 35890030.0,
      "step": 14160
    },
    {
      "entropy": 6.241196632385254,
      "epoch": 1.634737449509521,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047427714726641044,
      "loss": 5.7174,
      "mean_token_accuracy": 0.19796631932258607,
      "num_tokens": 35901822.0,
      "step": 14165
    },
    {
      "entropy": 6.208907890319824,
      "epoch": 1.6353144835545297,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047425798473620215,
      "loss": 5.7875,
      "mean_token_accuracy": 0.19444545954465867,
      "num_tokens": 35914553.0,
      "step": 14170
    },
    {
      "entropy": 6.239507102966309,
      "epoch": 1.6358915175995383,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004742388155040505,
      "loss": 5.8447,
      "mean_token_accuracy": 0.1936868041753769,
      "num_tokens": 35928536.0,
      "step": 14175
    },
    {
      "entropy": 6.2511533260345455,
      "epoch": 1.6364685516445472,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047421963957060026,
      "loss": 5.8622,
      "mean_token_accuracy": 0.18557486087083816,
      "num_tokens": 35941860.0,
      "step": 14180
    },
    {
      "entropy": 6.28944034576416,
      "epoch": 1.6370455856895556,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004742004569364964,
      "loss": 5.8582,
      "mean_token_accuracy": 0.18758884817361832,
      "num_tokens": 35957126.0,
      "step": 14185
    },
    {
      "entropy": 6.236316394805908,
      "epoch": 1.6376226197345645,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047418126760238416,
      "loss": 5.7751,
      "mean_token_accuracy": 0.19353905320167542,
      "num_tokens": 35969871.0,
      "step": 14190
    },
    {
      "entropy": 6.305867004394531,
      "epoch": 1.6381996537795729,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047416207156890887,
      "loss": 5.9674,
      "mean_token_accuracy": 0.1786883741617203,
      "num_tokens": 35983341.0,
      "step": 14195
    },
    {
      "entropy": 6.267291593551636,
      "epoch": 1.6387766878245817,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004741428688367164,
      "loss": 5.8469,
      "mean_token_accuracy": 0.18889210522174835,
      "num_tokens": 35995306.0,
      "step": 14200
    },
    {
      "entropy": 6.257507753372193,
      "epoch": 1.6393537218695902,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00047412365940645225,
      "loss": 5.8224,
      "mean_token_accuracy": 0.19164760112762452,
      "num_tokens": 36007424.0,
      "step": 14205
    },
    {
      "entropy": 6.259178352355957,
      "epoch": 1.639930755914599,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047410444327876286,
      "loss": 5.8823,
      "mean_token_accuracy": 0.19056559801101686,
      "num_tokens": 36020752.0,
      "step": 14210
    },
    {
      "entropy": 6.286505031585693,
      "epoch": 1.6405077899596077,
      "grad_norm": 0.875,
      "learning_rate": 0.00047408522045429426,
      "loss": 5.8329,
      "mean_token_accuracy": 0.18520755618810653,
      "num_tokens": 36031802.0,
      "step": 14215
    },
    {
      "entropy": 6.240025520324707,
      "epoch": 1.6410848240046163,
      "grad_norm": 0.875,
      "learning_rate": 0.0004740659909336933,
      "loss": 5.8865,
      "mean_token_accuracy": 0.18477454036474228,
      "num_tokens": 36044784.0,
      "step": 14220
    },
    {
      "entropy": 6.2177042961120605,
      "epoch": 1.641661858049625,
      "grad_norm": 0.875,
      "learning_rate": 0.00047404675471760655,
      "loss": 5.7621,
      "mean_token_accuracy": 0.18875966221094131,
      "num_tokens": 36057301.0,
      "step": 14225
    },
    {
      "entropy": 6.292890739440918,
      "epoch": 1.6422388920946336,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004740275118066811,
      "loss": 5.8269,
      "mean_token_accuracy": 0.18487808406352996,
      "num_tokens": 36070301.0,
      "step": 14230
    },
    {
      "entropy": 6.266335487365723,
      "epoch": 1.6428159261396422,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00047400826220156416,
      "loss": 5.8133,
      "mean_token_accuracy": 0.19563194811344148,
      "num_tokens": 36083122.0,
      "step": 14235
    },
    {
      "entropy": 6.28100552558899,
      "epoch": 1.6433929601846509,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047398900590290313,
      "loss": 5.9105,
      "mean_token_accuracy": 0.177164751291275,
      "num_tokens": 36095888.0,
      "step": 14240
    },
    {
      "entropy": 6.296646404266357,
      "epoch": 1.6439699942296595,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047396974291134575,
      "loss": 5.8361,
      "mean_token_accuracy": 0.18737712055444716,
      "num_tokens": 36108865.0,
      "step": 14245
    },
    {
      "entropy": 6.32221155166626,
      "epoch": 1.6445470282746681,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047395047322754,
      "loss": 5.8743,
      "mean_token_accuracy": 0.1833844855427742,
      "num_tokens": 36121318.0,
      "step": 14250
    },
    {
      "entropy": 6.308139848709106,
      "epoch": 1.645124062319677,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047393119685213374,
      "loss": 5.9116,
      "mean_token_accuracy": 0.18478037565946578,
      "num_tokens": 36134097.0,
      "step": 14255
    },
    {
      "entropy": 6.191109561920166,
      "epoch": 1.6457010963646854,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004739119137857756,
      "loss": 5.8016,
      "mean_token_accuracy": 0.18747840225696563,
      "num_tokens": 36146382.0,
      "step": 14260
    },
    {
      "entropy": 6.350604915618897,
      "epoch": 1.6462781304096943,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047389262402911404,
      "loss": 5.8878,
      "mean_token_accuracy": 0.18004481941461564,
      "num_tokens": 36159323.0,
      "step": 14265
    },
    {
      "entropy": 6.274691820144653,
      "epoch": 1.6468551644547027,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047387332758279784,
      "loss": 5.7763,
      "mean_token_accuracy": 0.19494640529155732,
      "num_tokens": 36171545.0,
      "step": 14270
    },
    {
      "entropy": 6.240463924407959,
      "epoch": 1.6474321984997116,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047385402444747606,
      "loss": 5.7949,
      "mean_token_accuracy": 0.18744019120931626,
      "num_tokens": 36184175.0,
      "step": 14275
    },
    {
      "entropy": 6.261882734298706,
      "epoch": 1.64800923254472,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047383471462379803,
      "loss": 5.7641,
      "mean_token_accuracy": 0.1878871962428093,
      "num_tokens": 36196444.0,
      "step": 14280
    },
    {
      "entropy": 6.212790632247925,
      "epoch": 1.6485862665897288,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000473815398112413,
      "loss": 5.7628,
      "mean_token_accuracy": 0.19088119864463807,
      "num_tokens": 36208899.0,
      "step": 14285
    },
    {
      "entropy": 6.232113885879516,
      "epoch": 1.6491633006347375,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004737960749139708,
      "loss": 5.8208,
      "mean_token_accuracy": 0.19247862696647644,
      "num_tokens": 36223039.0,
      "step": 14290
    },
    {
      "entropy": 6.354962253570557,
      "epoch": 1.6497403346797461,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004737767450291215,
      "loss": 5.9272,
      "mean_token_accuracy": 0.1800476238131523,
      "num_tokens": 36234917.0,
      "step": 14295
    },
    {
      "entropy": 6.208756732940674,
      "epoch": 1.6503173687247548,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047375740845851506,
      "loss": 5.7823,
      "mean_token_accuracy": 0.19674338698387145,
      "num_tokens": 36246791.0,
      "step": 14300
    },
    {
      "entropy": 6.328754091262818,
      "epoch": 1.6508944027697634,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004737380652028019,
      "loss": 5.9018,
      "mean_token_accuracy": 0.18426549732685088,
      "num_tokens": 36259666.0,
      "step": 14305
    },
    {
      "entropy": 6.207915735244751,
      "epoch": 1.651471436814772,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047371871526263263,
      "loss": 5.7838,
      "mean_token_accuracy": 0.18885438591241838,
      "num_tokens": 36272208.0,
      "step": 14310
    },
    {
      "entropy": 6.248191976547242,
      "epoch": 1.6520484708597807,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004736993586386581,
      "loss": 5.8202,
      "mean_token_accuracy": 0.1872250661253929,
      "num_tokens": 36285108.0,
      "step": 14315
    },
    {
      "entropy": 6.218618535995484,
      "epoch": 1.6526255049047895,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047367999533152934,
      "loss": 5.8324,
      "mean_token_accuracy": 0.19356549978256227,
      "num_tokens": 36297228.0,
      "step": 14320
    },
    {
      "entropy": 6.282768297195434,
      "epoch": 1.653202538949798,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047366062534189756,
      "loss": 5.829,
      "mean_token_accuracy": 0.1872438296675682,
      "num_tokens": 36310291.0,
      "step": 14325
    },
    {
      "entropy": 6.273576021194458,
      "epoch": 1.6537795729948068,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047364124867041446,
      "loss": 5.7708,
      "mean_token_accuracy": 0.19236373901367188,
      "num_tokens": 36322932.0,
      "step": 14330
    },
    {
      "entropy": 6.256026792526245,
      "epoch": 1.6543566070398152,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047362186531773156,
      "loss": 5.846,
      "mean_token_accuracy": 0.1908264175057411,
      "num_tokens": 36334896.0,
      "step": 14335
    },
    {
      "entropy": 6.29288125038147,
      "epoch": 1.654933641084824,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004736024752845008,
      "loss": 5.8613,
      "mean_token_accuracy": 0.1872571364045143,
      "num_tokens": 36347866.0,
      "step": 14340
    },
    {
      "entropy": 6.2074216365814205,
      "epoch": 1.6555106751298325,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004735830785713746,
      "loss": 5.8334,
      "mean_token_accuracy": 0.19257211536169053,
      "num_tokens": 36360930.0,
      "step": 14345
    },
    {
      "entropy": 6.272889566421509,
      "epoch": 1.6560877091748414,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004735636751790051,
      "loss": 5.7861,
      "mean_token_accuracy": 0.19864102452993393,
      "num_tokens": 36373798.0,
      "step": 14350
    },
    {
      "entropy": 6.265282011032104,
      "epoch": 1.65666474321985,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000473544265108045,
      "loss": 5.8089,
      "mean_token_accuracy": 0.19314082860946655,
      "num_tokens": 36387654.0,
      "step": 14355
    },
    {
      "entropy": 6.223183250427246,
      "epoch": 1.6572417772648587,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00047352484835914716,
      "loss": 5.8637,
      "mean_token_accuracy": 0.18828810453414918,
      "num_tokens": 36399963.0,
      "step": 14360
    },
    {
      "entropy": 6.338477897644043,
      "epoch": 1.6578188113098673,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004735054249329647,
      "loss": 5.9302,
      "mean_token_accuracy": 0.18202648758888246,
      "num_tokens": 36412397.0,
      "step": 14365
    },
    {
      "entropy": 6.286065149307251,
      "epoch": 1.658395845354876,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047348599483015087,
      "loss": 5.7952,
      "mean_token_accuracy": 0.18482748419046402,
      "num_tokens": 36425180.0,
      "step": 14370
    },
    {
      "entropy": 6.189918184280396,
      "epoch": 1.6589728793998846,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047346655805135916,
      "loss": 5.8173,
      "mean_token_accuracy": 0.19728542268276214,
      "num_tokens": 36437447.0,
      "step": 14375
    },
    {
      "entropy": 6.228664207458496,
      "epoch": 1.6595499134448932,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004734471145972434,
      "loss": 5.7484,
      "mean_token_accuracy": 0.19263490438461303,
      "num_tokens": 36449778.0,
      "step": 14380
    },
    {
      "entropy": 6.210476684570312,
      "epoch": 1.6601269474899019,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047342766446845753,
      "loss": 5.7449,
      "mean_token_accuracy": 0.2004138171672821,
      "num_tokens": 36461759.0,
      "step": 14385
    },
    {
      "entropy": 6.29231014251709,
      "epoch": 1.6607039815349105,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004734082076656557,
      "loss": 5.8754,
      "mean_token_accuracy": 0.17757974714040756,
      "num_tokens": 36474613.0,
      "step": 14390
    },
    {
      "entropy": 6.329931640625,
      "epoch": 1.6612810155799194,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047338874418949235,
      "loss": 5.8804,
      "mean_token_accuracy": 0.18555438071489333,
      "num_tokens": 36486569.0,
      "step": 14395
    },
    {
      "entropy": 6.295885038375855,
      "epoch": 1.6618580496249278,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00047336927404062213,
      "loss": 5.8831,
      "mean_token_accuracy": 0.18584659099578857,
      "num_tokens": 36500144.0,
      "step": 14400
    },
    {
      "entropy": 6.306648683547974,
      "epoch": 1.6624350836699366,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047334979721969995,
      "loss": 5.8762,
      "mean_token_accuracy": 0.1820911541581154,
      "num_tokens": 36512665.0,
      "step": 14405
    },
    {
      "entropy": 6.251386308670044,
      "epoch": 1.663012117714945,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004733303137273808,
      "loss": 5.7975,
      "mean_token_accuracy": 0.19215971380472183,
      "num_tokens": 36525876.0,
      "step": 14410
    },
    {
      "entropy": 6.223016309738159,
      "epoch": 1.663589151759954,
      "grad_norm": 0.875,
      "learning_rate": 0.00047331082356432015,
      "loss": 5.7508,
      "mean_token_accuracy": 0.1980337008833885,
      "num_tokens": 36538253.0,
      "step": 14415
    },
    {
      "entropy": 6.248967027664184,
      "epoch": 1.6641661858049623,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004732913267311734,
      "loss": 5.8377,
      "mean_token_accuracy": 0.193584606051445,
      "num_tokens": 36550303.0,
      "step": 14420
    },
    {
      "entropy": 6.170731735229492,
      "epoch": 1.6647432198499712,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004732718232285964,
      "loss": 5.7316,
      "mean_token_accuracy": 0.19512959867715834,
      "num_tokens": 36564835.0,
      "step": 14425
    },
    {
      "entropy": 6.307872104644775,
      "epoch": 1.6653202538949798,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00047325231305724507,
      "loss": 5.8991,
      "mean_token_accuracy": 0.18057988435029984,
      "num_tokens": 36577648.0,
      "step": 14430
    },
    {
      "entropy": 6.221223497390747,
      "epoch": 1.6658972879399885,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004732327962177757,
      "loss": 5.7842,
      "mean_token_accuracy": 0.1905799388885498,
      "num_tokens": 36590451.0,
      "step": 14435
    },
    {
      "entropy": 6.3358324527740475,
      "epoch": 1.6664743219849971,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004732132727108447,
      "loss": 5.853,
      "mean_token_accuracy": 0.18064933121204377,
      "num_tokens": 36602767.0,
      "step": 14440
    },
    {
      "entropy": 6.254924774169922,
      "epoch": 1.6670513560300058,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047319374253710874,
      "loss": 5.7936,
      "mean_token_accuracy": 0.1912146896123886,
      "num_tokens": 36614938.0,
      "step": 14445
    },
    {
      "entropy": 6.207167339324951,
      "epoch": 1.6676283900750144,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004731742056972247,
      "loss": 5.7531,
      "mean_token_accuracy": 0.19748119711875917,
      "num_tokens": 36626880.0,
      "step": 14450
    },
    {
      "entropy": 6.19822473526001,
      "epoch": 1.668205424120023,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004731546621918497,
      "loss": 5.8086,
      "mean_token_accuracy": 0.19121013283729554,
      "num_tokens": 36640729.0,
      "step": 14455
    },
    {
      "entropy": 6.23019232749939,
      "epoch": 1.668782458165032,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000473135112021641,
      "loss": 5.773,
      "mean_token_accuracy": 0.20230078101158142,
      "num_tokens": 36656359.0,
      "step": 14460
    },
    {
      "entropy": 6.262730979919434,
      "epoch": 1.6693594922100403,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047311555518725617,
      "loss": 5.7849,
      "mean_token_accuracy": 0.1985212653875351,
      "num_tokens": 36667985.0,
      "step": 14465
    },
    {
      "entropy": 6.3092857837677006,
      "epoch": 1.6699365262550492,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047309599168935323,
      "loss": 5.9379,
      "mean_token_accuracy": 0.17907729297876357,
      "num_tokens": 36681477.0,
      "step": 14470
    },
    {
      "entropy": 6.286537981033325,
      "epoch": 1.6705135603000576,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047307642152858993,
      "loss": 5.8269,
      "mean_token_accuracy": 0.19062384963035583,
      "num_tokens": 36695532.0,
      "step": 14475
    },
    {
      "entropy": 6.273736524581909,
      "epoch": 1.6710905943450665,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047305684470562453,
      "loss": 5.8306,
      "mean_token_accuracy": 0.18725011199712754,
      "num_tokens": 36708914.0,
      "step": 14480
    },
    {
      "entropy": 6.2877583503723145,
      "epoch": 1.6716676283900749,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004730372612211156,
      "loss": 5.9162,
      "mean_token_accuracy": 0.18136829435825347,
      "num_tokens": 36720849.0,
      "step": 14485
    },
    {
      "entropy": 6.25967812538147,
      "epoch": 1.6722446624350837,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047301767107572174,
      "loss": 5.795,
      "mean_token_accuracy": 0.19021519869565964,
      "num_tokens": 36733786.0,
      "step": 14490
    },
    {
      "entropy": 6.270547676086426,
      "epoch": 1.6728216964800924,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004729980742701018,
      "loss": 5.8469,
      "mean_token_accuracy": 0.18713072538375855,
      "num_tokens": 36746101.0,
      "step": 14495
    },
    {
      "entropy": 6.229681301116943,
      "epoch": 1.673398730525101,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004729784708049151,
      "loss": 5.7565,
      "mean_token_accuracy": 0.19194569140672685,
      "num_tokens": 36758644.0,
      "step": 14500
    },
    {
      "entropy": 6.3293468952178955,
      "epoch": 1.6739757645701097,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004729588606808209,
      "loss": 5.922,
      "mean_token_accuracy": 0.18240256905555724,
      "num_tokens": 36770301.0,
      "step": 14505
    },
    {
      "entropy": 6.251408433914184,
      "epoch": 1.6745527986151183,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047293924389847864,
      "loss": 5.8092,
      "mean_token_accuracy": 0.1863787889480591,
      "num_tokens": 36782337.0,
      "step": 14510
    },
    {
      "entropy": 6.261877107620239,
      "epoch": 1.675129832660127,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004729196204585483,
      "loss": 5.8487,
      "mean_token_accuracy": 0.1870492786169052,
      "num_tokens": 36793817.0,
      "step": 14515
    },
    {
      "entropy": 6.239163875579834,
      "epoch": 1.6757068667051356,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047289999036168983,
      "loss": 5.8138,
      "mean_token_accuracy": 0.1873685672879219,
      "num_tokens": 36806577.0,
      "step": 14520
    },
    {
      "entropy": 6.317363834381103,
      "epoch": 1.6762839007501442,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004728803536085634,
      "loss": 5.8209,
      "mean_token_accuracy": 0.1926939979195595,
      "num_tokens": 36819896.0,
      "step": 14525
    },
    {
      "entropy": 6.217675065994262,
      "epoch": 1.6768609347951529,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047286071019982974,
      "loss": 5.8159,
      "mean_token_accuracy": 0.19518922865390778,
      "num_tokens": 36831995.0,
      "step": 14530
    },
    {
      "entropy": 6.262294721603394,
      "epoch": 1.6774379688401617,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047284106013614926,
      "loss": 5.7824,
      "mean_token_accuracy": 0.1857282966375351,
      "num_tokens": 36844452.0,
      "step": 14535
    },
    {
      "entropy": 6.268368673324585,
      "epoch": 1.6780150028851701,
      "grad_norm": 0.890625,
      "learning_rate": 0.000472821403418183,
      "loss": 5.7491,
      "mean_token_accuracy": 0.19941954612731932,
      "num_tokens": 36857578.0,
      "step": 14540
    },
    {
      "entropy": 6.242843246459961,
      "epoch": 1.678592036930179,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004728017400465921,
      "loss": 5.8144,
      "mean_token_accuracy": 0.1899886816740036,
      "num_tokens": 36870406.0,
      "step": 14545
    },
    {
      "entropy": 6.229114103317261,
      "epoch": 1.6791690709751874,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047278207002203796,
      "loss": 5.7698,
      "mean_token_accuracy": 0.18690193444490433,
      "num_tokens": 36881149.0,
      "step": 14550
    },
    {
      "entropy": 6.244127607345581,
      "epoch": 1.6797461050201963,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047276239334518216,
      "loss": 5.7917,
      "mean_token_accuracy": 0.19791980981826782,
      "num_tokens": 36893175.0,
      "step": 14555
    },
    {
      "entropy": 6.18767032623291,
      "epoch": 1.6803231390652047,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047274271001668646,
      "loss": 5.7426,
      "mean_token_accuracy": 0.19336573630571366,
      "num_tokens": 36904993.0,
      "step": 14560
    },
    {
      "entropy": 6.263120508193969,
      "epoch": 1.6809001731102136,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00047272302003721286,
      "loss": 5.8542,
      "mean_token_accuracy": 0.18332159221172334,
      "num_tokens": 36918597.0,
      "step": 14565
    },
    {
      "entropy": 6.282705068588257,
      "epoch": 1.6814772071552222,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047270332340742377,
      "loss": 5.8447,
      "mean_token_accuracy": 0.18700587004423141,
      "num_tokens": 36931404.0,
      "step": 14570
    },
    {
      "entropy": 6.290522336959839,
      "epoch": 1.6820542412002308,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047268362012798157,
      "loss": 5.9126,
      "mean_token_accuracy": 0.18323498368263244,
      "num_tokens": 36942657.0,
      "step": 14575
    },
    {
      "entropy": 6.251866197586059,
      "epoch": 1.6826312752452395,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004726639101995491,
      "loss": 5.8545,
      "mean_token_accuracy": 0.1905171111226082,
      "num_tokens": 36956130.0,
      "step": 14580
    },
    {
      "entropy": 6.282032155990601,
      "epoch": 1.6832083092902481,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047264419362278906,
      "loss": 5.8515,
      "mean_token_accuracy": 0.19287431687116624,
      "num_tokens": 36968865.0,
      "step": 14585
    },
    {
      "entropy": 6.320575857162476,
      "epoch": 1.6837853433352568,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047262447039836484,
      "loss": 5.9019,
      "mean_token_accuracy": 0.19080299288034439,
      "num_tokens": 36980466.0,
      "step": 14590
    },
    {
      "entropy": 6.239609956741333,
      "epoch": 1.6843623773802654,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047260474052693963,
      "loss": 5.8548,
      "mean_token_accuracy": 0.17984795868396758,
      "num_tokens": 36992263.0,
      "step": 14595
    },
    {
      "entropy": 6.323089647293091,
      "epoch": 1.684939411425274,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047258500400917724,
      "loss": 5.8668,
      "mean_token_accuracy": 0.18974658995866775,
      "num_tokens": 37003422.0,
      "step": 14600
    },
    {
      "entropy": 6.269136571884156,
      "epoch": 1.6855164454702827,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047256526084574137,
      "loss": 5.7713,
      "mean_token_accuracy": 0.19197449535131456,
      "num_tokens": 37015064.0,
      "step": 14605
    },
    {
      "entropy": 6.211493110656738,
      "epoch": 1.6860934795152915,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047254551103729597,
      "loss": 5.7396,
      "mean_token_accuracy": 0.1953820765018463,
      "num_tokens": 37027750.0,
      "step": 14610
    },
    {
      "entropy": 6.273235511779785,
      "epoch": 1.6866705135603,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004725257545845055,
      "loss": 5.8296,
      "mean_token_accuracy": 0.18275714963674544,
      "num_tokens": 37039501.0,
      "step": 14615
    },
    {
      "entropy": 6.322292232513428,
      "epoch": 1.6872475476053088,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047250599148803443,
      "loss": 5.8875,
      "mean_token_accuracy": 0.18079376518726348,
      "num_tokens": 37051179.0,
      "step": 14620
    },
    {
      "entropy": 6.242916440963745,
      "epoch": 1.6878245816503172,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047248622174854746,
      "loss": 5.8163,
      "mean_token_accuracy": 0.1894673928618431,
      "num_tokens": 37064098.0,
      "step": 14625
    },
    {
      "entropy": 6.287539100646972,
      "epoch": 1.688401615695326,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004724664453667094,
      "loss": 5.8929,
      "mean_token_accuracy": 0.17877427637577056,
      "num_tokens": 37076315.0,
      "step": 14630
    },
    {
      "entropy": 6.3396220207214355,
      "epoch": 1.6889786497403345,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004724466623431857,
      "loss": 5.9268,
      "mean_token_accuracy": 0.186344937980175,
      "num_tokens": 37088354.0,
      "step": 14635
    },
    {
      "entropy": 6.266402864456177,
      "epoch": 1.6895556837853434,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004724268726786415,
      "loss": 5.8364,
      "mean_token_accuracy": 0.19499343037605285,
      "num_tokens": 37101296.0,
      "step": 14640
    },
    {
      "entropy": 6.282308292388916,
      "epoch": 1.690132717830352,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004724070763737424,
      "loss": 5.8903,
      "mean_token_accuracy": 0.18634043633937836,
      "num_tokens": 37114095.0,
      "step": 14645
    },
    {
      "entropy": 6.27353801727295,
      "epoch": 1.6907097518753607,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004723872734291545,
      "loss": 5.7706,
      "mean_token_accuracy": 0.19720648378133773,
      "num_tokens": 37125963.0,
      "step": 14650
    },
    {
      "entropy": 6.269825410842896,
      "epoch": 1.6912867859203693,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047236746384554364,
      "loss": 5.8779,
      "mean_token_accuracy": 0.1874593272805214,
      "num_tokens": 37139210.0,
      "step": 14655
    },
    {
      "entropy": 6.117925691604614,
      "epoch": 1.691863819965378,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004723476476235762,
      "loss": 5.7075,
      "mean_token_accuracy": 0.198999385535717,
      "num_tokens": 37151592.0,
      "step": 14660
    },
    {
      "entropy": 6.258327388763428,
      "epoch": 1.6924408540103866,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004723278247639186,
      "loss": 5.8465,
      "mean_token_accuracy": 0.18850377053022385,
      "num_tokens": 37164722.0,
      "step": 14665
    },
    {
      "entropy": 6.213457918167114,
      "epoch": 1.6930178880553952,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004723079952672377,
      "loss": 5.7466,
      "mean_token_accuracy": 0.1984286665916443,
      "num_tokens": 37177188.0,
      "step": 14670
    },
    {
      "entropy": 6.268677997589111,
      "epoch": 1.693594922100404,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047228815913420035,
      "loss": 5.8071,
      "mean_token_accuracy": 0.1969089224934578,
      "num_tokens": 37189702.0,
      "step": 14675
    },
    {
      "entropy": 6.260214328765869,
      "epoch": 1.6941719561454125,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004722683163654738,
      "loss": 5.7964,
      "mean_token_accuracy": 0.18863973766565323,
      "num_tokens": 37202327.0,
      "step": 14680
    },
    {
      "entropy": 6.238330125808716,
      "epoch": 1.6947489901904214,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004722484669617254,
      "loss": 5.7869,
      "mean_token_accuracy": 0.18888368159532548,
      "num_tokens": 37215557.0,
      "step": 14685
    },
    {
      "entropy": 6.2546703815460205,
      "epoch": 1.6953260242354298,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047222861092362277,
      "loss": 5.8265,
      "mean_token_accuracy": 0.19578233361244202,
      "num_tokens": 37228863.0,
      "step": 14690
    },
    {
      "entropy": 6.2640172958374025,
      "epoch": 1.6959030582804386,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047220874825183387,
      "loss": 5.8714,
      "mean_token_accuracy": 0.18507293462753296,
      "num_tokens": 37241591.0,
      "step": 14695
    },
    {
      "entropy": 6.284923553466797,
      "epoch": 1.696480092325447,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047218887894702656,
      "loss": 5.9105,
      "mean_token_accuracy": 0.1883197844028473,
      "num_tokens": 37254440.0,
      "step": 14700
    },
    {
      "entropy": 6.241726684570312,
      "epoch": 1.697057126370456,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004721690030098693,
      "loss": 5.787,
      "mean_token_accuracy": 0.18930340856313704,
      "num_tokens": 37266803.0,
      "step": 14705
    },
    {
      "entropy": 6.301064491271973,
      "epoch": 1.6976341604154646,
      "grad_norm": 1.0,
      "learning_rate": 0.0004721491204410305,
      "loss": 5.8819,
      "mean_token_accuracy": 0.18446458727121354,
      "num_tokens": 37280339.0,
      "step": 14710
    },
    {
      "entropy": 6.217402124404908,
      "epoch": 1.6982111944604732,
      "grad_norm": 0.84375,
      "learning_rate": 0.00047212923124117915,
      "loss": 5.7659,
      "mean_token_accuracy": 0.18945713192224503,
      "num_tokens": 37293816.0,
      "step": 14715
    },
    {
      "entropy": 6.288745021820068,
      "epoch": 1.6987882285054818,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004721093354109839,
      "loss": 5.832,
      "mean_token_accuracy": 0.18953413516283035,
      "num_tokens": 37306642.0,
      "step": 14720
    },
    {
      "entropy": 6.285134601593017,
      "epoch": 1.6993652625504905,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047208943295111406,
      "loss": 5.833,
      "mean_token_accuracy": 0.19792227298021317,
      "num_tokens": 37320153.0,
      "step": 14725
    },
    {
      "entropy": 6.299769496917724,
      "epoch": 1.6999422965954991,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047206952386223905,
      "loss": 5.8871,
      "mean_token_accuracy": 0.18497230261564254,
      "num_tokens": 37333660.0,
      "step": 14730
    },
    {
      "entropy": 6.234125995635987,
      "epoch": 1.7005193306405078,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004720496081450285,
      "loss": 5.7968,
      "mean_token_accuracy": 0.18866164982318878,
      "num_tokens": 37346589.0,
      "step": 14735
    },
    {
      "entropy": 6.24196720123291,
      "epoch": 1.7010963646855164,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047202968580015224,
      "loss": 5.7725,
      "mean_token_accuracy": 0.19921280592679977,
      "num_tokens": 37359309.0,
      "step": 14740
    },
    {
      "entropy": 6.296859264373779,
      "epoch": 1.701673398730525,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047200975682828045,
      "loss": 5.8887,
      "mean_token_accuracy": 0.19096756875514984,
      "num_tokens": 37372547.0,
      "step": 14745
    },
    {
      "entropy": 6.267851257324219,
      "epoch": 1.702250432775534,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004719898212300832,
      "loss": 5.827,
      "mean_token_accuracy": 0.18974183201789857,
      "num_tokens": 37384987.0,
      "step": 14750
    },
    {
      "entropy": 6.30603928565979,
      "epoch": 1.7028274668205423,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047196987900623134,
      "loss": 5.8513,
      "mean_token_accuracy": 0.1876941904425621,
      "num_tokens": 37397065.0,
      "step": 14755
    },
    {
      "entropy": 6.2352031707763675,
      "epoch": 1.7034045008655512,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00047194993015739527,
      "loss": 5.7869,
      "mean_token_accuracy": 0.19300023317337037,
      "num_tokens": 37409467.0,
      "step": 14760
    },
    {
      "entropy": 6.297355651855469,
      "epoch": 1.7039815349105596,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047192997468424624,
      "loss": 5.7901,
      "mean_token_accuracy": 0.1993768110871315,
      "num_tokens": 37422268.0,
      "step": 14765
    },
    {
      "entropy": 6.239738988876343,
      "epoch": 1.7045585689555685,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004719100125874553,
      "loss": 5.8058,
      "mean_token_accuracy": 0.195662422478199,
      "num_tokens": 37434637.0,
      "step": 14770
    },
    {
      "entropy": 6.226162433624268,
      "epoch": 1.7051356030005769,
      "grad_norm": 0.875,
      "learning_rate": 0.0004718900438676939,
      "loss": 5.7855,
      "mean_token_accuracy": 0.19193764328956603,
      "num_tokens": 37447683.0,
      "step": 14775
    },
    {
      "entropy": 6.296966457366944,
      "epoch": 1.7057126370455857,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004718700685256337,
      "loss": 5.8867,
      "mean_token_accuracy": 0.18528142124414443,
      "num_tokens": 37459960.0,
      "step": 14780
    },
    {
      "entropy": 6.314924478530884,
      "epoch": 1.7062896710905944,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004718500865619465,
      "loss": 5.9514,
      "mean_token_accuracy": 0.1803431436419487,
      "num_tokens": 37471262.0,
      "step": 14785
    },
    {
      "entropy": 6.23435959815979,
      "epoch": 1.706866705135603,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004718300979773044,
      "loss": 5.7567,
      "mean_token_accuracy": 0.18967181146144868,
      "num_tokens": 37484843.0,
      "step": 14790
    },
    {
      "entropy": 6.358298444747925,
      "epoch": 1.7074437391806117,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047181010277237977,
      "loss": 5.8041,
      "mean_token_accuracy": 0.19069833159446717,
      "num_tokens": 37497197.0,
      "step": 14795
    },
    {
      "entropy": 6.312958860397339,
      "epoch": 1.7080207732256203,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004717901009478451,
      "loss": 5.7773,
      "mean_token_accuracy": 0.18386179208755493,
      "num_tokens": 37508812.0,
      "step": 14800
    },
    {
      "entropy": 6.22099232673645,
      "epoch": 1.708597807270629,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047177009250437313,
      "loss": 5.8216,
      "mean_token_accuracy": 0.18812813311815263,
      "num_tokens": 37520861.0,
      "step": 14805
    },
    {
      "entropy": 6.294021987915039,
      "epoch": 1.7091748413156376,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047175007744263683,
      "loss": 5.8382,
      "mean_token_accuracy": 0.18528886884450912,
      "num_tokens": 37533011.0,
      "step": 14810
    },
    {
      "entropy": 6.156536293029785,
      "epoch": 1.7097518753606464,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047173005576330935,
      "loss": 5.7228,
      "mean_token_accuracy": 0.19585810601711273,
      "num_tokens": 37546613.0,
      "step": 14815
    },
    {
      "entropy": 6.239349985122681,
      "epoch": 1.7103289094056549,
      "grad_norm": 0.875,
      "learning_rate": 0.00047171002746706424,
      "loss": 5.7191,
      "mean_token_accuracy": 0.19512592256069183,
      "num_tokens": 37559429.0,
      "step": 14820
    },
    {
      "entropy": 6.302822017669678,
      "epoch": 1.7109059434506637,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047168999255457506,
      "loss": 5.8638,
      "mean_token_accuracy": 0.18944178223609925,
      "num_tokens": 37571530.0,
      "step": 14825
    },
    {
      "entropy": 6.287614345550537,
      "epoch": 1.7114829774956721,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047166995102651565,
      "loss": 5.8561,
      "mean_token_accuracy": 0.18424619287252425,
      "num_tokens": 37584653.0,
      "step": 14830
    },
    {
      "entropy": 6.2718182563781735,
      "epoch": 1.712060011540681,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004716499028835601,
      "loss": 5.8353,
      "mean_token_accuracy": 0.19543803185224534,
      "num_tokens": 37597361.0,
      "step": 14835
    },
    {
      "entropy": 6.283855772018432,
      "epoch": 1.7126370455856894,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004716298481263828,
      "loss": 5.8647,
      "mean_token_accuracy": 0.19398764967918397,
      "num_tokens": 37611525.0,
      "step": 14840
    },
    {
      "entropy": 6.299528169631958,
      "epoch": 1.7132140796306983,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004716097867556583,
      "loss": 5.8392,
      "mean_token_accuracy": 0.18878912478685378,
      "num_tokens": 37624012.0,
      "step": 14845
    },
    {
      "entropy": 6.325048494338989,
      "epoch": 1.713791113675707,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047158971877206114,
      "loss": 5.854,
      "mean_token_accuracy": 0.18885803818702698,
      "num_tokens": 37636877.0,
      "step": 14850
    },
    {
      "entropy": 6.277114772796631,
      "epoch": 1.7143681477207156,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004715696441762665,
      "loss": 5.8095,
      "mean_token_accuracy": 0.18925647288560868,
      "num_tokens": 37649791.0,
      "step": 14855
    },
    {
      "entropy": 6.242626142501831,
      "epoch": 1.7149451817657242,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047154956296894954,
      "loss": 5.8318,
      "mean_token_accuracy": 0.18954112827777864,
      "num_tokens": 37661699.0,
      "step": 14860
    },
    {
      "entropy": 6.267629480361938,
      "epoch": 1.7155222158107328,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004715294751507856,
      "loss": 5.8439,
      "mean_token_accuracy": 0.18587643057107925,
      "num_tokens": 37674498.0,
      "step": 14865
    },
    {
      "entropy": 6.299939393997192,
      "epoch": 1.7160992498557415,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004715093807224505,
      "loss": 5.805,
      "mean_token_accuracy": 0.1898824855685234,
      "num_tokens": 37687196.0,
      "step": 14870
    },
    {
      "entropy": 6.188147687911988,
      "epoch": 1.71667628390075,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047148927968462,
      "loss": 5.7089,
      "mean_token_accuracy": 0.19724368900060654,
      "num_tokens": 37698888.0,
      "step": 14875
    },
    {
      "entropy": 6.229661989212036,
      "epoch": 1.7172533179457588,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047146917203797,
      "loss": 5.8471,
      "mean_token_accuracy": 0.18993326872587205,
      "num_tokens": 37711560.0,
      "step": 14880
    },
    {
      "entropy": 6.192458200454712,
      "epoch": 1.7178303519907674,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004714490577831771,
      "loss": 5.7647,
      "mean_token_accuracy": 0.19304184317588807,
      "num_tokens": 37723846.0,
      "step": 14885
    },
    {
      "entropy": 6.221444892883301,
      "epoch": 1.7184073860357763,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004714289369209177,
      "loss": 5.744,
      "mean_token_accuracy": 0.1923162430524826,
      "num_tokens": 37735535.0,
      "step": 14890
    },
    {
      "entropy": 6.233892726898193,
      "epoch": 1.7189844200807847,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047140880945186863,
      "loss": 5.8074,
      "mean_token_accuracy": 0.1940958708524704,
      "num_tokens": 37748015.0,
      "step": 14895
    },
    {
      "entropy": 6.187597465515137,
      "epoch": 1.7195614541257935,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047138867537670676,
      "loss": 5.7789,
      "mean_token_accuracy": 0.19544515311717986,
      "num_tokens": 37760278.0,
      "step": 14900
    },
    {
      "entropy": 6.3203342914581295,
      "epoch": 1.720138488170802,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047136853469610933,
      "loss": 5.8401,
      "mean_token_accuracy": 0.19449251890182495,
      "num_tokens": 37772274.0,
      "step": 14905
    },
    {
      "entropy": 6.198821783065796,
      "epoch": 1.7207155222158108,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047134838741075383,
      "loss": 5.7946,
      "mean_token_accuracy": 0.19303936809301375,
      "num_tokens": 37785919.0,
      "step": 14910
    },
    {
      "entropy": 6.218606758117676,
      "epoch": 1.7212925562608192,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047132823352131787,
      "loss": 5.7496,
      "mean_token_accuracy": 0.1953893095254898,
      "num_tokens": 37797770.0,
      "step": 14915
    },
    {
      "entropy": 6.233460998535156,
      "epoch": 1.721869590305828,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004713080730284793,
      "loss": 5.7987,
      "mean_token_accuracy": 0.1913621410727501,
      "num_tokens": 37809962.0,
      "step": 14920
    },
    {
      "entropy": 6.316199207305909,
      "epoch": 1.7224466243508367,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047128790593291617,
      "loss": 5.8462,
      "mean_token_accuracy": 0.18395719826221466,
      "num_tokens": 37822345.0,
      "step": 14925
    },
    {
      "entropy": 6.233913898468018,
      "epoch": 1.7230236583958454,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00047126773223530677,
      "loss": 5.7906,
      "mean_token_accuracy": 0.1922600522637367,
      "num_tokens": 37835568.0,
      "step": 14930
    },
    {
      "entropy": 6.306128454208374,
      "epoch": 1.723600692440854,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047124755193632975,
      "loss": 5.8479,
      "mean_token_accuracy": 0.19030786752700807,
      "num_tokens": 37849356.0,
      "step": 14935
    },
    {
      "entropy": 6.223813581466675,
      "epoch": 1.7241777264858626,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047122736503666377,
      "loss": 5.7571,
      "mean_token_accuracy": 0.19723167270421982,
      "num_tokens": 37861970.0,
      "step": 14940
    },
    {
      "entropy": 6.303425025939942,
      "epoch": 1.7247547605308713,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004712071715369878,
      "loss": 5.8567,
      "mean_token_accuracy": 0.18961762636899948,
      "num_tokens": 37874066.0,
      "step": 14945
    },
    {
      "entropy": 6.259584188461304,
      "epoch": 1.72533179457588,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004711869714379812,
      "loss": 5.8191,
      "mean_token_accuracy": 0.18672791868448257,
      "num_tokens": 37885529.0,
      "step": 14950
    },
    {
      "entropy": 6.362218761444092,
      "epoch": 1.7259088286208888,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004711667647403232,
      "loss": 5.844,
      "mean_token_accuracy": 0.18382576256990432,
      "num_tokens": 37899517.0,
      "step": 14955
    },
    {
      "entropy": 6.260833072662353,
      "epoch": 1.7264858626658972,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00047114655144469354,
      "loss": 5.7932,
      "mean_token_accuracy": 0.19213725477457047,
      "num_tokens": 37912586.0,
      "step": 14960
    },
    {
      "entropy": 6.269883394241333,
      "epoch": 1.727062896710906,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047112633155177203,
      "loss": 5.8666,
      "mean_token_accuracy": 0.18703035712242128,
      "num_tokens": 37924559.0,
      "step": 14965
    },
    {
      "entropy": 6.1387838363647464,
      "epoch": 1.7276399307559145,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004711061050622388,
      "loss": 5.6776,
      "mean_token_accuracy": 0.20505414307117462,
      "num_tokens": 37937470.0,
      "step": 14970
    },
    {
      "entropy": 6.282984828948974,
      "epoch": 1.7282169648009233,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047108587197677404,
      "loss": 5.8082,
      "mean_token_accuracy": 0.19098254889249802,
      "num_tokens": 37949502.0,
      "step": 14975
    },
    {
      "entropy": 6.2145274639129635,
      "epoch": 1.7287939988459318,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047106563229605845,
      "loss": 5.7964,
      "mean_token_accuracy": 0.1975490093231201,
      "num_tokens": 37962298.0,
      "step": 14980
    },
    {
      "entropy": 6.2187957763671875,
      "epoch": 1.7293710328909406,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047104538602077276,
      "loss": 5.7104,
      "mean_token_accuracy": 0.2040121152997017,
      "num_tokens": 37975614.0,
      "step": 14985
    },
    {
      "entropy": 6.22262225151062,
      "epoch": 1.7299480669359493,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004710251331515978,
      "loss": 5.7285,
      "mean_token_accuracy": 0.2045437291264534,
      "num_tokens": 37990117.0,
      "step": 14990
    },
    {
      "entropy": 6.264737224578857,
      "epoch": 1.730525100980958,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047100487368921485,
      "loss": 5.8019,
      "mean_token_accuracy": 0.1892474740743637,
      "num_tokens": 38003654.0,
      "step": 14995
    },
    {
      "entropy": 6.303586721420288,
      "epoch": 1.7311021350259665,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004709846076343055,
      "loss": 5.857,
      "mean_token_accuracy": 0.19145375043153762,
      "num_tokens": 38016299.0,
      "step": 15000
    },
    {
      "epoch": 1.7311021350259665,
      "eval_entropy": 6.034497247082556,
      "eval_loss": 5.8745269775390625,
      "eval_mean_token_accuracy": 0.19555749452466614,
      "eval_num_tokens": 38016299.0,
      "eval_runtime": 17.8379,
      "eval_samples_per_second": 1577.313,
      "eval_steps_per_second": 197.164,
      "step": 15000
    },
    {
      "entropy": 6.228104972839356,
      "epoch": 1.7316791690709752,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047096433498755103,
      "loss": 5.8407,
      "mean_token_accuracy": 0.19088809788227082,
      "num_tokens": 38027793.0,
      "step": 15005
    },
    {
      "entropy": 6.307938766479492,
      "epoch": 1.7322562031159838,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047094405574963364,
      "loss": 5.8067,
      "mean_token_accuracy": 0.18836195319890975,
      "num_tokens": 38040167.0,
      "step": 15010
    },
    {
      "entropy": 6.245943212509156,
      "epoch": 1.7328332371609925,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047092376992123516,
      "loss": 5.7559,
      "mean_token_accuracy": 0.19289580434560777,
      "num_tokens": 38053334.0,
      "step": 15015
    },
    {
      "entropy": 6.215775823593139,
      "epoch": 1.733410271206001,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047090347750303803,
      "loss": 5.7877,
      "mean_token_accuracy": 0.19255453497171401,
      "num_tokens": 38065831.0,
      "step": 15020
    },
    {
      "entropy": 6.203187704086304,
      "epoch": 1.7339873052510097,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004708831784957247,
      "loss": 5.8302,
      "mean_token_accuracy": 0.1962748259305954,
      "num_tokens": 38079298.0,
      "step": 15025
    },
    {
      "entropy": 6.326677131652832,
      "epoch": 1.7345643392960186,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004708628728999781,
      "loss": 5.9035,
      "mean_token_accuracy": 0.18355834484100342,
      "num_tokens": 38092473.0,
      "step": 15030
    },
    {
      "entropy": 6.276353979110718,
      "epoch": 1.735141373341027,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004708425607164809,
      "loss": 5.7804,
      "mean_token_accuracy": 0.1939054995775223,
      "num_tokens": 38106148.0,
      "step": 15035
    },
    {
      "entropy": 6.274563217163086,
      "epoch": 1.7357184073860359,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004708222419459165,
      "loss": 5.8649,
      "mean_token_accuracy": 0.1868964046239853,
      "num_tokens": 38119061.0,
      "step": 15040
    },
    {
      "entropy": 6.2413183689117435,
      "epoch": 1.7362954414310443,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004708019165889683,
      "loss": 5.8236,
      "mean_token_accuracy": 0.19083871692419052,
      "num_tokens": 38131390.0,
      "step": 15045
    },
    {
      "entropy": 6.291907548904419,
      "epoch": 1.7368724754760532,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004707815846463199,
      "loss": 5.7802,
      "mean_token_accuracy": 0.19039222002029418,
      "num_tokens": 38145393.0,
      "step": 15050
    },
    {
      "entropy": 6.355773305892944,
      "epoch": 1.7374495095210616,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004707612461186552,
      "loss": 5.8892,
      "mean_token_accuracy": 0.18643365949392318,
      "num_tokens": 38158341.0,
      "step": 15055
    },
    {
      "entropy": 6.334320497512818,
      "epoch": 1.7380265435660704,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047074090100665805,
      "loss": 5.8725,
      "mean_token_accuracy": 0.1892070069909096,
      "num_tokens": 38170796.0,
      "step": 15060
    },
    {
      "entropy": 6.2609837532043455,
      "epoch": 1.738603577611079,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047072054931101306,
      "loss": 5.8224,
      "mean_token_accuracy": 0.18834739476442336,
      "num_tokens": 38184280.0,
      "step": 15065
    },
    {
      "entropy": 6.293169260025024,
      "epoch": 1.7391806116560877,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004707001910324046,
      "loss": 5.8228,
      "mean_token_accuracy": 0.18142785727977753,
      "num_tokens": 38195709.0,
      "step": 15070
    },
    {
      "entropy": 6.298284530639648,
      "epoch": 1.7397576457010964,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047067982617151737,
      "loss": 5.8628,
      "mean_token_accuracy": 0.19391675442457199,
      "num_tokens": 38208904.0,
      "step": 15075
    },
    {
      "entropy": 6.24472861289978,
      "epoch": 1.740334679746105,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004706594547290365,
      "loss": 5.8432,
      "mean_token_accuracy": 0.18700127005577089,
      "num_tokens": 38222827.0,
      "step": 15080
    },
    {
      "entropy": 6.323432493209839,
      "epoch": 1.7409117137911136,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047063907670564695,
      "loss": 5.836,
      "mean_token_accuracy": 0.19232436120510102,
      "num_tokens": 38235972.0,
      "step": 15085
    },
    {
      "entropy": 6.2099555969238285,
      "epoch": 1.7414887478361223,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004706186921020342,
      "loss": 5.6976,
      "mean_token_accuracy": 0.20087052285671234,
      "num_tokens": 38247998.0,
      "step": 15090
    },
    {
      "entropy": 6.247817182540894,
      "epoch": 1.7420657818811311,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047059830091888407,
      "loss": 5.901,
      "mean_token_accuracy": 0.18259721547365187,
      "num_tokens": 38260460.0,
      "step": 15095
    },
    {
      "entropy": 6.264999103546143,
      "epoch": 1.7426428159261396,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004705779031568821,
      "loss": 5.7604,
      "mean_token_accuracy": 0.19740022718906403,
      "num_tokens": 38271829.0,
      "step": 15100
    },
    {
      "entropy": 6.202706289291382,
      "epoch": 1.7432198499711484,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047055749881671463,
      "loss": 5.7532,
      "mean_token_accuracy": 0.19828899800777436,
      "num_tokens": 38283602.0,
      "step": 15105
    },
    {
      "entropy": 6.2953267097473145,
      "epoch": 1.7437968840161568,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004705370878990677,
      "loss": 5.8496,
      "mean_token_accuracy": 0.18788864016532897,
      "num_tokens": 38296112.0,
      "step": 15110
    },
    {
      "entropy": 6.277995252609253,
      "epoch": 1.7443739180611657,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047051667040462806,
      "loss": 5.8088,
      "mean_token_accuracy": 0.19381272047758102,
      "num_tokens": 38309159.0,
      "step": 15115
    },
    {
      "entropy": 6.247989797592163,
      "epoch": 1.7449509521061741,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047049624633408224,
      "loss": 5.7721,
      "mean_token_accuracy": 0.1949317291378975,
      "num_tokens": 38321468.0,
      "step": 15120
    },
    {
      "entropy": 6.2564249515533445,
      "epoch": 1.745527986151183,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00047047581568811724,
      "loss": 5.78,
      "mean_token_accuracy": 0.18629155904054642,
      "num_tokens": 38333866.0,
      "step": 15125
    },
    {
      "entropy": 6.289843225479126,
      "epoch": 1.7461050201961916,
      "grad_norm": 0.875,
      "learning_rate": 0.00047045537846742043,
      "loss": 5.8624,
      "mean_token_accuracy": 0.18235021382570266,
      "num_tokens": 38345965.0,
      "step": 15130
    },
    {
      "entropy": 6.261213397979736,
      "epoch": 1.7466820542412003,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000470434934672679,
      "loss": 5.7693,
      "mean_token_accuracy": 0.1883448153734207,
      "num_tokens": 38358536.0,
      "step": 15135
    },
    {
      "entropy": 6.249023294448852,
      "epoch": 1.747259088286209,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047041448430458054,
      "loss": 5.7499,
      "mean_token_accuracy": 0.19873300641775132,
      "num_tokens": 38371209.0,
      "step": 15140
    },
    {
      "entropy": 6.252534437179565,
      "epoch": 1.7478361223312175,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047039402736381305,
      "loss": 5.8386,
      "mean_token_accuracy": 0.18851133286952973,
      "num_tokens": 38384270.0,
      "step": 15145
    },
    {
      "entropy": 6.165564060211182,
      "epoch": 1.7484131563762262,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004703735638510645,
      "loss": 5.6877,
      "mean_token_accuracy": 0.2077319160103798,
      "num_tokens": 38396328.0,
      "step": 15150
    },
    {
      "entropy": 6.30913233757019,
      "epoch": 1.7489901904212348,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004703530937670232,
      "loss": 5.8555,
      "mean_token_accuracy": 0.1889762833714485,
      "num_tokens": 38408460.0,
      "step": 15155
    },
    {
      "entropy": 6.254319429397583,
      "epoch": 1.7495672244662435,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004703326171123776,
      "loss": 5.8001,
      "mean_token_accuracy": 0.19340113550424576,
      "num_tokens": 38421211.0,
      "step": 15160
    },
    {
      "entropy": 6.232342100143432,
      "epoch": 1.750144258511252,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004703121338878164,
      "loss": 5.8814,
      "mean_token_accuracy": 0.18773785829544068,
      "num_tokens": 38433720.0,
      "step": 15165
    },
    {
      "entropy": 6.289814710617065,
      "epoch": 1.750721292556261,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004702916440940286,
      "loss": 5.8029,
      "mean_token_accuracy": 0.1900806337594986,
      "num_tokens": 38445776.0,
      "step": 15170
    },
    {
      "entropy": 6.265978097915649,
      "epoch": 1.7512983266012694,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004702711477317034,
      "loss": 5.7608,
      "mean_token_accuracy": 0.1927257478237152,
      "num_tokens": 38457218.0,
      "step": 15175
    },
    {
      "entropy": 6.236439561843872,
      "epoch": 1.7518753606462782,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004702506448015301,
      "loss": 5.8559,
      "mean_token_accuracy": 0.18266476541757584,
      "num_tokens": 38469944.0,
      "step": 15180
    },
    {
      "entropy": 6.284494161605835,
      "epoch": 1.7524523946912867,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047023013530419843,
      "loss": 5.8303,
      "mean_token_accuracy": 0.18842567205429078,
      "num_tokens": 38482683.0,
      "step": 15185
    },
    {
      "entropy": 6.228808736801147,
      "epoch": 1.7530294287362955,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004702096192403981,
      "loss": 5.764,
      "mean_token_accuracy": 0.19846258461475372,
      "num_tokens": 38494652.0,
      "step": 15190
    },
    {
      "entropy": 6.259058475494385,
      "epoch": 1.753606462781304,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004701890966108192,
      "loss": 5.8425,
      "mean_token_accuracy": 0.19090690314769745,
      "num_tokens": 38506567.0,
      "step": 15195
    },
    {
      "entropy": 6.278779649734497,
      "epoch": 1.7541834968263128,
      "grad_norm": 0.96875,
      "learning_rate": 0.000470168567416152,
      "loss": 5.7748,
      "mean_token_accuracy": 0.19407596588134765,
      "num_tokens": 38519399.0,
      "step": 15200
    },
    {
      "entropy": 6.159409141540527,
      "epoch": 1.7547605308713214,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004701480316570869,
      "loss": 5.696,
      "mean_token_accuracy": 0.2083926558494568,
      "num_tokens": 38533100.0,
      "step": 15205
    },
    {
      "entropy": 6.259949779510498,
      "epoch": 1.75533756491633,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004701274893343147,
      "loss": 5.8168,
      "mean_token_accuracy": 0.1888844296336174,
      "num_tokens": 38545412.0,
      "step": 15210
    },
    {
      "entropy": 6.217299556732177,
      "epoch": 1.7559145989613387,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047010694044852643,
      "loss": 5.7535,
      "mean_token_accuracy": 0.1941909596323967,
      "num_tokens": 38557428.0,
      "step": 15215
    },
    {
      "entropy": 6.2496740341186525,
      "epoch": 1.7564916330063474,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000470086385000413,
      "loss": 5.9049,
      "mean_token_accuracy": 0.19202267080545427,
      "num_tokens": 38569656.0,
      "step": 15220
    },
    {
      "entropy": 6.288322067260742,
      "epoch": 1.757068667051356,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004700658229906661,
      "loss": 5.8537,
      "mean_token_accuracy": 0.18653422445058823,
      "num_tokens": 38582280.0,
      "step": 15225
    },
    {
      "entropy": 6.272194242477417,
      "epoch": 1.7576457010963646,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047004525441997694,
      "loss": 5.7558,
      "mean_token_accuracy": 0.1996623009443283,
      "num_tokens": 38594578.0,
      "step": 15230
    },
    {
      "entropy": 6.25399489402771,
      "epoch": 1.7582227351413735,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004700246792890376,
      "loss": 5.7939,
      "mean_token_accuracy": 0.19056859314441682,
      "num_tokens": 38606803.0,
      "step": 15235
    },
    {
      "entropy": 6.270346736907959,
      "epoch": 1.758799769186382,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004700040975985401,
      "loss": 5.7612,
      "mean_token_accuracy": 0.19472208470106125,
      "num_tokens": 38618851.0,
      "step": 15240
    },
    {
      "entropy": 6.230273580551147,
      "epoch": 1.7593768032313908,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046998350934917654,
      "loss": 5.7417,
      "mean_token_accuracy": 0.1918221652507782,
      "num_tokens": 38630505.0,
      "step": 15245
    },
    {
      "entropy": 6.296421861648559,
      "epoch": 1.7599538372763992,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046996291454163956,
      "loss": 5.7677,
      "mean_token_accuracy": 0.19856485575437546,
      "num_tokens": 38642601.0,
      "step": 15250
    },
    {
      "entropy": 6.223921823501587,
      "epoch": 1.760530871321408,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004699423131766218,
      "loss": 5.7606,
      "mean_token_accuracy": 0.19583483040332794,
      "num_tokens": 38655713.0,
      "step": 15255
    },
    {
      "entropy": 6.311620426177979,
      "epoch": 1.7611079053664165,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004699217052548161,
      "loss": 5.8851,
      "mean_token_accuracy": 0.18551360219717025,
      "num_tokens": 38667909.0,
      "step": 15260
    },
    {
      "entropy": 6.321383905410767,
      "epoch": 1.7616849394114253,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046990109077691577,
      "loss": 5.9014,
      "mean_token_accuracy": 0.1860878586769104,
      "num_tokens": 38681729.0,
      "step": 15265
    },
    {
      "entropy": 6.3360082626342775,
      "epoch": 1.762261973456434,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046988046974361406,
      "loss": 5.8637,
      "mean_token_accuracy": 0.18773409277200698,
      "num_tokens": 38694957.0,
      "step": 15270
    },
    {
      "entropy": 6.250478982925415,
      "epoch": 1.7628390075014426,
      "grad_norm": 1.0,
      "learning_rate": 0.0004698598421556045,
      "loss": 5.8352,
      "mean_token_accuracy": 0.19254247099161148,
      "num_tokens": 38707914.0,
      "step": 15275
    },
    {
      "entropy": 6.280478572845459,
      "epoch": 1.7634160415464513,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004698392080135809,
      "loss": 5.8107,
      "mean_token_accuracy": 0.1899206355214119,
      "num_tokens": 38719011.0,
      "step": 15280
    },
    {
      "entropy": 6.237488317489624,
      "epoch": 1.76399307559146,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004698185673182374,
      "loss": 5.8021,
      "mean_token_accuracy": 0.19203453958034516,
      "num_tokens": 38731473.0,
      "step": 15285
    },
    {
      "entropy": 6.218995141983032,
      "epoch": 1.7645701096364685,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046979792007026817,
      "loss": 5.827,
      "mean_token_accuracy": 0.19166237115859985,
      "num_tokens": 38745061.0,
      "step": 15290
    },
    {
      "entropy": 6.2922694206237795,
      "epoch": 1.7651471436814772,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004697772662703676,
      "loss": 5.8056,
      "mean_token_accuracy": 0.19527062475681306,
      "num_tokens": 38757980.0,
      "step": 15295
    },
    {
      "entropy": 6.281692743301392,
      "epoch": 1.7657241777264858,
      "grad_norm": 0.90625,
      "learning_rate": 0.00046975660591923047,
      "loss": 5.9212,
      "mean_token_accuracy": 0.18119412064552307,
      "num_tokens": 38771918.0,
      "step": 15300
    },
    {
      "entropy": 6.281959009170532,
      "epoch": 1.7663012117714945,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004697359390175516,
      "loss": 5.7844,
      "mean_token_accuracy": 0.1961324095726013,
      "num_tokens": 38783702.0,
      "step": 15305
    },
    {
      "entropy": 6.172072839736939,
      "epoch": 1.7668782458165033,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046971526556602625,
      "loss": 5.7571,
      "mean_token_accuracy": 0.19817993938922882,
      "num_tokens": 38797574.0,
      "step": 15310
    },
    {
      "entropy": 6.234071397781372,
      "epoch": 1.7674552798615117,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004696945855653495,
      "loss": 5.6968,
      "mean_token_accuracy": 0.20038567036390303,
      "num_tokens": 38810302.0,
      "step": 15315
    },
    {
      "entropy": 6.184006690979004,
      "epoch": 1.7680323139065206,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004696738990162172,
      "loss": 5.6698,
      "mean_token_accuracy": 0.19754522293806076,
      "num_tokens": 38823968.0,
      "step": 15320
    },
    {
      "entropy": 6.316234827041626,
      "epoch": 1.768609347951529,
      "grad_norm": 0.953125,
      "learning_rate": 0.000469653205919325,
      "loss": 5.8141,
      "mean_token_accuracy": 0.1844889536499977,
      "num_tokens": 38837641.0,
      "step": 15325
    },
    {
      "entropy": 6.250605201721191,
      "epoch": 1.7691863819965379,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046963250627536884,
      "loss": 5.7665,
      "mean_token_accuracy": 0.190290230512619,
      "num_tokens": 38849556.0,
      "step": 15330
    },
    {
      "entropy": 6.24680724143982,
      "epoch": 1.7697634160415463,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004696118000850451,
      "loss": 5.8171,
      "mean_token_accuracy": 0.1876034691929817,
      "num_tokens": 38863276.0,
      "step": 15335
    },
    {
      "entropy": 6.251680707931518,
      "epoch": 1.7703404500865552,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046959108734905003,
      "loss": 5.807,
      "mean_token_accuracy": 0.18602417409420013,
      "num_tokens": 38876230.0,
      "step": 15340
    },
    {
      "entropy": 6.336206483840942,
      "epoch": 1.7709174841315638,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046957036806808045,
      "loss": 5.9067,
      "mean_token_accuracy": 0.18851778507232667,
      "num_tokens": 38891269.0,
      "step": 15345
    },
    {
      "entropy": 6.310498523712158,
      "epoch": 1.7714945181765724,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004695496422428332,
      "loss": 5.8508,
      "mean_token_accuracy": 0.1881686896085739,
      "num_tokens": 38904081.0,
      "step": 15350
    },
    {
      "entropy": 6.256695795059204,
      "epoch": 1.772071552221581,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004695289098740054,
      "loss": 5.791,
      "mean_token_accuracy": 0.1901915416121483,
      "num_tokens": 38916422.0,
      "step": 15355
    },
    {
      "entropy": 6.342411088943481,
      "epoch": 1.7726485862665897,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004695081709622943,
      "loss": 5.8552,
      "mean_token_accuracy": 0.18733129948377608,
      "num_tokens": 38928554.0,
      "step": 15360
    },
    {
      "entropy": 6.252917623519897,
      "epoch": 1.7732256203115984,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046948742550839744,
      "loss": 5.7533,
      "mean_token_accuracy": 0.19529881477355956,
      "num_tokens": 38942252.0,
      "step": 15365
    },
    {
      "entropy": 6.263192510604858,
      "epoch": 1.773802654356607,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004694666735130127,
      "loss": 5.8502,
      "mean_token_accuracy": 0.19108292013406752,
      "num_tokens": 38955604.0,
      "step": 15370
    },
    {
      "entropy": 6.249915981292725,
      "epoch": 1.7743796884016156,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000469445914976838,
      "loss": 5.8136,
      "mean_token_accuracy": 0.19352486878633499,
      "num_tokens": 38968439.0,
      "step": 15375
    },
    {
      "entropy": 6.291030073165894,
      "epoch": 1.7749567224466243,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004694251499005715,
      "loss": 5.8479,
      "mean_token_accuracy": 0.18742929697036742,
      "num_tokens": 38980251.0,
      "step": 15380
    },
    {
      "entropy": 6.268551588058472,
      "epoch": 1.7755337564916331,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004694043782849116,
      "loss": 5.8248,
      "mean_token_accuracy": 0.1958874374628067,
      "num_tokens": 38992242.0,
      "step": 15385
    },
    {
      "entropy": 6.240305805206299,
      "epoch": 1.7761107905366416,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00046938360013055715,
      "loss": 5.7743,
      "mean_token_accuracy": 0.1861726388335228,
      "num_tokens": 39004372.0,
      "step": 15390
    },
    {
      "entropy": 6.254604387283325,
      "epoch": 1.7766878245816504,
      "grad_norm": 1.0,
      "learning_rate": 0.00046936281543820673,
      "loss": 5.804,
      "mean_token_accuracy": 0.19045973122119902,
      "num_tokens": 39015928.0,
      "step": 15395
    },
    {
      "entropy": 6.231979465484619,
      "epoch": 1.7772648586266588,
      "grad_norm": 0.890625,
      "learning_rate": 0.00046934202420855967,
      "loss": 5.7782,
      "mean_token_accuracy": 0.1973385602235794,
      "num_tokens": 39028948.0,
      "step": 15400
    },
    {
      "entropy": 6.215716361999512,
      "epoch": 1.7778418926716677,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046932122644231507,
      "loss": 5.7861,
      "mean_token_accuracy": 0.1937277927994728,
      "num_tokens": 39042486.0,
      "step": 15405
    },
    {
      "entropy": 6.290436315536499,
      "epoch": 1.7784189267166761,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00046930042214017256,
      "loss": 5.9017,
      "mean_token_accuracy": 0.1885281980037689,
      "num_tokens": 39054660.0,
      "step": 15410
    },
    {
      "entropy": 6.281246900558472,
      "epoch": 1.778995960761685,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004692796113028319,
      "loss": 5.8428,
      "mean_token_accuracy": 0.18685497790575029,
      "num_tokens": 39067272.0,
      "step": 15415
    },
    {
      "entropy": 6.3292577266693115,
      "epoch": 1.7795729948066936,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000469258793930993,
      "loss": 5.8221,
      "mean_token_accuracy": 0.18773300647735597,
      "num_tokens": 39080146.0,
      "step": 15420
    },
    {
      "entropy": 6.209461498260498,
      "epoch": 1.7801500288517023,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046923797002535605,
      "loss": 5.6914,
      "mean_token_accuracy": 0.20140144526958464,
      "num_tokens": 39093093.0,
      "step": 15425
    },
    {
      "entropy": 6.285262775421143,
      "epoch": 1.780727062896711,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004692171395866214,
      "loss": 5.8214,
      "mean_token_accuracy": 0.19057723134756088,
      "num_tokens": 39105239.0,
      "step": 15430
    },
    {
      "entropy": 6.280012893676758,
      "epoch": 1.7813040969417195,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046919630261548986,
      "loss": 5.8336,
      "mean_token_accuracy": 0.1861838296055794,
      "num_tokens": 39118265.0,
      "step": 15435
    },
    {
      "entropy": 6.261084938049317,
      "epoch": 1.7818811309867282,
      "grad_norm": 0.90625,
      "learning_rate": 0.00046917545911266207,
      "loss": 5.8698,
      "mean_token_accuracy": 0.17991481274366378,
      "num_tokens": 39131999.0,
      "step": 15440
    },
    {
      "entropy": 6.301594400405884,
      "epoch": 1.7824581650317368,
      "grad_norm": 1.0,
      "learning_rate": 0.00046915460907883923,
      "loss": 5.8299,
      "mean_token_accuracy": 0.19017376005649567,
      "num_tokens": 39145529.0,
      "step": 15445
    },
    {
      "entropy": 6.187797403335571,
      "epoch": 1.7830351990767457,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046913375251472246,
      "loss": 5.6316,
      "mean_token_accuracy": 0.20047187954187393,
      "num_tokens": 39157736.0,
      "step": 15450
    },
    {
      "entropy": 6.275732517242432,
      "epoch": 1.783612233121754,
      "grad_norm": 0.90625,
      "learning_rate": 0.00046911288942101345,
      "loss": 5.8668,
      "mean_token_accuracy": 0.18608424663543702,
      "num_tokens": 39169126.0,
      "step": 15455
    },
    {
      "entropy": 6.359842920303345,
      "epoch": 1.784189267166763,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004690920197984138,
      "loss": 5.8967,
      "mean_token_accuracy": 0.18641961514949798,
      "num_tokens": 39183190.0,
      "step": 15460
    },
    {
      "entropy": 6.208134603500366,
      "epoch": 1.7847663012117714,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004690711436476254,
      "loss": 5.7374,
      "mean_token_accuracy": 0.1964043363928795,
      "num_tokens": 39194874.0,
      "step": 15465
    },
    {
      "entropy": 6.211254072189331,
      "epoch": 1.7853433352567802,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00046905026096935056,
      "loss": 5.8556,
      "mean_token_accuracy": 0.19062747210264205,
      "num_tokens": 39208373.0,
      "step": 15470
    },
    {
      "entropy": 6.2817164897918705,
      "epoch": 1.7859203693017887,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004690293717642916,
      "loss": 5.7955,
      "mean_token_accuracy": 0.19003399163484574,
      "num_tokens": 39219848.0,
      "step": 15475
    },
    {
      "entropy": 6.2621817111969,
      "epoch": 1.7864974033467975,
      "grad_norm": 1.015625,
      "learning_rate": 0.000469008476033151,
      "loss": 5.809,
      "mean_token_accuracy": 0.1933301717042923,
      "num_tokens": 39232737.0,
      "step": 15480
    },
    {
      "entropy": 6.310494613647461,
      "epoch": 1.7870744373918062,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046898757377663176,
      "loss": 5.9337,
      "mean_token_accuracy": 0.1848430022597313,
      "num_tokens": 39245220.0,
      "step": 15485
    },
    {
      "entropy": 6.324825954437256,
      "epoch": 1.7876514714368148,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046896666499543683,
      "loss": 5.8665,
      "mean_token_accuracy": 0.19146413654088973,
      "num_tokens": 39257828.0,
      "step": 15490
    },
    {
      "entropy": 6.317997646331787,
      "epoch": 1.7882285054818234,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046894574969026946,
      "loss": 5.8738,
      "mean_token_accuracy": 0.18459345251321793,
      "num_tokens": 39270679.0,
      "step": 15495
    },
    {
      "entropy": 6.310544061660766,
      "epoch": 1.788805539526832,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046892482786183313,
      "loss": 5.8295,
      "mean_token_accuracy": 0.18798051178455352,
      "num_tokens": 39282634.0,
      "step": 15500
    },
    {
      "entropy": 6.280200624465943,
      "epoch": 1.7893825735718407,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046890389951083155,
      "loss": 5.7838,
      "mean_token_accuracy": 0.19883925914764405,
      "num_tokens": 39295383.0,
      "step": 15505
    },
    {
      "entropy": 6.2420878410339355,
      "epoch": 1.7899596076168494,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046888296463796857,
      "loss": 5.7837,
      "mean_token_accuracy": 0.19070164412260054,
      "num_tokens": 39308636.0,
      "step": 15510
    },
    {
      "entropy": 6.304009962081909,
      "epoch": 1.790536641661858,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004688620232439485,
      "loss": 5.9604,
      "mean_token_accuracy": 0.1819470539689064,
      "num_tokens": 39321330.0,
      "step": 15515
    },
    {
      "entropy": 6.373122596740723,
      "epoch": 1.7911136757068666,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046884107532947547,
      "loss": 5.9176,
      "mean_token_accuracy": 0.17946398705244065,
      "num_tokens": 39334920.0,
      "step": 15520
    },
    {
      "entropy": 6.282447290420532,
      "epoch": 1.7916907097518755,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046882012089525415,
      "loss": 5.82,
      "mean_token_accuracy": 0.19539097249507903,
      "num_tokens": 39347827.0,
      "step": 15525
    },
    {
      "entropy": 6.1816854000091555,
      "epoch": 1.792267743796884,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004687991599419895,
      "loss": 5.7609,
      "mean_token_accuracy": 0.1982062876224518,
      "num_tokens": 39359732.0,
      "step": 15530
    },
    {
      "entropy": 6.108600234985351,
      "epoch": 1.7928447778418928,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046877819247038624,
      "loss": 5.6484,
      "mean_token_accuracy": 0.2097707748413086,
      "num_tokens": 39372056.0,
      "step": 15535
    },
    {
      "entropy": 6.274241018295288,
      "epoch": 1.7934218118869012,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004687572184811497,
      "loss": 5.7912,
      "mean_token_accuracy": 0.1940767228603363,
      "num_tokens": 39383302.0,
      "step": 15540
    },
    {
      "entropy": 6.216291427612305,
      "epoch": 1.79399884593191,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046873623797498545,
      "loss": 5.7726,
      "mean_token_accuracy": 0.1980548918247223,
      "num_tokens": 39395666.0,
      "step": 15545
    },
    {
      "entropy": 6.2581483840942385,
      "epoch": 1.7945758799769185,
      "grad_norm": 0.90625,
      "learning_rate": 0.000468715250952599,
      "loss": 5.7369,
      "mean_token_accuracy": 0.1998322919011116,
      "num_tokens": 39407192.0,
      "step": 15550
    },
    {
      "entropy": 6.353096628189087,
      "epoch": 1.7951529140219273,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046869425741469635,
      "loss": 5.9304,
      "mean_token_accuracy": 0.18556698113679887,
      "num_tokens": 39418331.0,
      "step": 15555
    },
    {
      "entropy": 6.293079042434693,
      "epoch": 1.795729948066936,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004686732573619835,
      "loss": 5.8223,
      "mean_token_accuracy": 0.18931469768285752,
      "num_tokens": 39431789.0,
      "step": 15560
    },
    {
      "entropy": 6.319036769866943,
      "epoch": 1.7963069821119446,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004686522507951669,
      "loss": 5.7839,
      "mean_token_accuracy": 0.18964634090662003,
      "num_tokens": 39444516.0,
      "step": 15565
    },
    {
      "entropy": 6.2102165699005125,
      "epoch": 1.7968840161569533,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004686312377149531,
      "loss": 5.7381,
      "mean_token_accuracy": 0.20061405897140502,
      "num_tokens": 39456490.0,
      "step": 15570
    },
    {
      "entropy": 6.202072334289551,
      "epoch": 1.797461050201962,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046861021812204874,
      "loss": 5.787,
      "mean_token_accuracy": 0.19272930324077606,
      "num_tokens": 39468232.0,
      "step": 15575
    },
    {
      "entropy": 6.297913503646851,
      "epoch": 1.7980380842469705,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046858919201716085,
      "loss": 5.7822,
      "mean_token_accuracy": 0.19338036328554153,
      "num_tokens": 39481037.0,
      "step": 15580
    },
    {
      "entropy": 6.134289121627807,
      "epoch": 1.7986151182919792,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004685681594009966,
      "loss": 5.6709,
      "mean_token_accuracy": 0.198665289580822,
      "num_tokens": 39493872.0,
      "step": 15585
    },
    {
      "entropy": 6.27271876335144,
      "epoch": 1.799192152336988,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046854712027426357,
      "loss": 5.8375,
      "mean_token_accuracy": 0.18383905291557312,
      "num_tokens": 39505997.0,
      "step": 15590
    },
    {
      "entropy": 6.348964786529541,
      "epoch": 1.7997691863819965,
      "grad_norm": 0.90625,
      "learning_rate": 0.00046852607463766923,
      "loss": 5.9208,
      "mean_token_accuracy": 0.18758784532546996,
      "num_tokens": 39519114.0,
      "step": 15595
    },
    {
      "entropy": 6.25947675704956,
      "epoch": 1.8003462204270053,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004685050224919215,
      "loss": 5.7587,
      "mean_token_accuracy": 0.19867537021636963,
      "num_tokens": 39530629.0,
      "step": 15600
    },
    {
      "entropy": 6.17420220375061,
      "epoch": 1.8009232544720137,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046848396383772844,
      "loss": 5.7718,
      "mean_token_accuracy": 0.19239344894886018,
      "num_tokens": 39543549.0,
      "step": 15605
    },
    {
      "entropy": 6.298850965499878,
      "epoch": 1.8015002885170226,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004684628986757984,
      "loss": 5.8404,
      "mean_token_accuracy": 0.18453985303640366,
      "num_tokens": 39555487.0,
      "step": 15610
    },
    {
      "entropy": 6.272164440155029,
      "epoch": 1.802077322562031,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004684418270068398,
      "loss": 5.7887,
      "mean_token_accuracy": 0.19623170495033265,
      "num_tokens": 39568594.0,
      "step": 15615
    },
    {
      "entropy": 6.145430374145508,
      "epoch": 1.8026543566070399,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004684207488315615,
      "loss": 5.6999,
      "mean_token_accuracy": 0.1937992200255394,
      "num_tokens": 39582069.0,
      "step": 15620
    },
    {
      "entropy": 6.19322714805603,
      "epoch": 1.8032313906520485,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004683996641506724,
      "loss": 5.8276,
      "mean_token_accuracy": 0.18662248998880387,
      "num_tokens": 39595316.0,
      "step": 15625
    },
    {
      "entropy": 6.254276371002197,
      "epoch": 1.8038084246970572,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046837857296488163,
      "loss": 5.7211,
      "mean_token_accuracy": 0.1971890762448311,
      "num_tokens": 39609363.0,
      "step": 15630
    },
    {
      "entropy": 6.261358594894409,
      "epoch": 1.8043854587420658,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046835747527489857,
      "loss": 5.8248,
      "mean_token_accuracy": 0.18931077271699906,
      "num_tokens": 39621597.0,
      "step": 15635
    },
    {
      "entropy": 6.315931272506714,
      "epoch": 1.8049624927870744,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000468336371081433,
      "loss": 5.8699,
      "mean_token_accuracy": 0.18564362972974777,
      "num_tokens": 39633913.0,
      "step": 15640
    },
    {
      "entropy": 6.273999452590942,
      "epoch": 1.805539526832083,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046831526038519444,
      "loss": 5.8175,
      "mean_token_accuracy": 0.19879835098981857,
      "num_tokens": 39646616.0,
      "step": 15645
    },
    {
      "entropy": 6.293560028076172,
      "epoch": 1.8061165608770917,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004682941431868933,
      "loss": 5.8388,
      "mean_token_accuracy": 0.18527151197195052,
      "num_tokens": 39659636.0,
      "step": 15650
    },
    {
      "entropy": 6.304013776779175,
      "epoch": 1.8066935949221004,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046827301948723963,
      "loss": 5.7741,
      "mean_token_accuracy": 0.19347509890794753,
      "num_tokens": 39672824.0,
      "step": 15655
    },
    {
      "entropy": 6.252510118484497,
      "epoch": 1.807270628967109,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046825188928694393,
      "loss": 5.7927,
      "mean_token_accuracy": 0.19260887801647186,
      "num_tokens": 39684988.0,
      "step": 15660
    },
    {
      "entropy": 6.287731885910034,
      "epoch": 1.8078476630121179,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004682307525867169,
      "loss": 5.7964,
      "mean_token_accuracy": 0.19256947487592696,
      "num_tokens": 39697313.0,
      "step": 15665
    },
    {
      "entropy": 6.250778245925903,
      "epoch": 1.8084246970571263,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004682096093872695,
      "loss": 5.7892,
      "mean_token_accuracy": 0.1953515961766243,
      "num_tokens": 39710005.0,
      "step": 15670
    },
    {
      "entropy": 6.181289434432983,
      "epoch": 1.8090017311021351,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046818845968931284,
      "loss": 5.7789,
      "mean_token_accuracy": 0.19419307112693787,
      "num_tokens": 39724495.0,
      "step": 15675
    },
    {
      "entropy": 6.2674795627594,
      "epoch": 1.8095787651471436,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046816730349355843,
      "loss": 5.8031,
      "mean_token_accuracy": 0.19015991538763047,
      "num_tokens": 39736978.0,
      "step": 15680
    },
    {
      "entropy": 6.314813184738159,
      "epoch": 1.8101557991921524,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046814614080071756,
      "loss": 5.8847,
      "mean_token_accuracy": 0.18165763914585115,
      "num_tokens": 39748694.0,
      "step": 15685
    },
    {
      "entropy": 6.266888761520386,
      "epoch": 1.8107328332371608,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046812497161150224,
      "loss": 5.8098,
      "mean_token_accuracy": 0.1898970752954483,
      "num_tokens": 39760388.0,
      "step": 15690
    },
    {
      "entropy": 6.247245740890503,
      "epoch": 1.8113098672821697,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046810379592662445,
      "loss": 5.7601,
      "mean_token_accuracy": 0.1992327407002449,
      "num_tokens": 39773424.0,
      "step": 15695
    },
    {
      "entropy": 6.286348009109497,
      "epoch": 1.8118869013271783,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00046808261374679637,
      "loss": 5.7645,
      "mean_token_accuracy": 0.19223696142435073,
      "num_tokens": 39786472.0,
      "step": 15700
    },
    {
      "entropy": 6.22647762298584,
      "epoch": 1.812463935372187,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004680614250727305,
      "loss": 5.7583,
      "mean_token_accuracy": 0.19804392457008363,
      "num_tokens": 39799968.0,
      "step": 15705
    },
    {
      "entropy": 6.247883605957031,
      "epoch": 1.8130409694171956,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004680402299051395,
      "loss": 5.7829,
      "mean_token_accuracy": 0.19796701967716218,
      "num_tokens": 39812332.0,
      "step": 15710
    },
    {
      "entropy": 6.316992330551147,
      "epoch": 1.8136180034622043,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004680190282447363,
      "loss": 5.844,
      "mean_token_accuracy": 0.19067827016115188,
      "num_tokens": 39825545.0,
      "step": 15715
    },
    {
      "entropy": 6.289811468124389,
      "epoch": 1.814195037507213,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004679978200922339,
      "loss": 5.9298,
      "mean_token_accuracy": 0.17935722917318345,
      "num_tokens": 39839691.0,
      "step": 15720
    },
    {
      "entropy": 6.2180251121521,
      "epoch": 1.8147720715522215,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004679766054483457,
      "loss": 5.7081,
      "mean_token_accuracy": 0.1986555814743042,
      "num_tokens": 39852225.0,
      "step": 15725
    },
    {
      "entropy": 6.221176958084106,
      "epoch": 1.8153491055972304,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004679553843137852,
      "loss": 5.6938,
      "mean_token_accuracy": 0.1960545673966408,
      "num_tokens": 39866451.0,
      "step": 15730
    },
    {
      "entropy": 6.256980085372925,
      "epoch": 1.8159261396422388,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046793415668926625,
      "loss": 5.803,
      "mean_token_accuracy": 0.19141706079244614,
      "num_tokens": 39878564.0,
      "step": 15735
    },
    {
      "entropy": 6.257886457443237,
      "epoch": 1.8165031736872477,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004679129225755028,
      "loss": 5.8067,
      "mean_token_accuracy": 0.19265975058078766,
      "num_tokens": 39890588.0,
      "step": 15740
    },
    {
      "entropy": 6.227653932571411,
      "epoch": 1.817080207732256,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000467891681973209,
      "loss": 5.709,
      "mean_token_accuracy": 0.20228794664144517,
      "num_tokens": 39901868.0,
      "step": 15745
    },
    {
      "entropy": 6.310026836395264,
      "epoch": 1.817657241777265,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046787043488309926,
      "loss": 5.7621,
      "mean_token_accuracy": 0.20052236020565034,
      "num_tokens": 39914475.0,
      "step": 15750
    },
    {
      "entropy": 6.064589881896973,
      "epoch": 1.8182342758222734,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004678491813058882,
      "loss": 5.5357,
      "mean_token_accuracy": 0.21241525709629058,
      "num_tokens": 39927488.0,
      "step": 15755
    },
    {
      "entropy": 6.294579410552979,
      "epoch": 1.8188113098672822,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004678279212422908,
      "loss": 5.8917,
      "mean_token_accuracy": 0.18239615112543106,
      "num_tokens": 39940753.0,
      "step": 15760
    },
    {
      "entropy": 6.278861427307129,
      "epoch": 1.8193883439122909,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004678066546930221,
      "loss": 5.7953,
      "mean_token_accuracy": 0.1926351636648178,
      "num_tokens": 39954779.0,
      "step": 15765
    },
    {
      "entropy": 6.281329202651977,
      "epoch": 1.8199653779572995,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046778538165879725,
      "loss": 5.7613,
      "mean_token_accuracy": 0.19292855709791185,
      "num_tokens": 39966962.0,
      "step": 15770
    },
    {
      "entropy": 6.244415378570556,
      "epoch": 1.8205424120023082,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046776410214033185,
      "loss": 5.7729,
      "mean_token_accuracy": 0.19418783932924272,
      "num_tokens": 39978590.0,
      "step": 15775
    },
    {
      "entropy": 6.198941993713379,
      "epoch": 1.8211194460473168,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004677428161383417,
      "loss": 5.7296,
      "mean_token_accuracy": 0.2055260419845581,
      "num_tokens": 39990929.0,
      "step": 15780
    },
    {
      "entropy": 6.2898965835571286,
      "epoch": 1.8216964800923254,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004677215236535426,
      "loss": 5.879,
      "mean_token_accuracy": 0.18635733127593995,
      "num_tokens": 40003301.0,
      "step": 15785
    },
    {
      "entropy": 6.337834692001342,
      "epoch": 1.822273514137334,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004677002246866508,
      "loss": 5.8132,
      "mean_token_accuracy": 0.19050079435110093,
      "num_tokens": 40016325.0,
      "step": 15790
    },
    {
      "entropy": 6.260081577301025,
      "epoch": 1.8228505481823427,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046767891923838264,
      "loss": 5.7418,
      "mean_token_accuracy": 0.20360271334648133,
      "num_tokens": 40030666.0,
      "step": 15795
    },
    {
      "entropy": 6.283757257461548,
      "epoch": 1.8234275822273514,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004676576073094548,
      "loss": 5.7753,
      "mean_token_accuracy": 0.19135726988315582,
      "num_tokens": 40041192.0,
      "step": 15800
    },
    {
      "entropy": 6.304760599136353,
      "epoch": 1.8240046162723602,
      "grad_norm": 0.859375,
      "learning_rate": 0.00046763628890058396,
      "loss": 5.7753,
      "mean_token_accuracy": 0.1929359644651413,
      "num_tokens": 40053620.0,
      "step": 15805
    },
    {
      "entropy": 6.239361381530761,
      "epoch": 1.8245816503173686,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046761496401248734,
      "loss": 5.7208,
      "mean_token_accuracy": 0.20190890729427338,
      "num_tokens": 40066268.0,
      "step": 15810
    },
    {
      "entropy": 6.270322465896607,
      "epoch": 1.8251586843623775,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000467593632645882,
      "loss": 5.8571,
      "mean_token_accuracy": 0.1878486007452011,
      "num_tokens": 40079251.0,
      "step": 15815
    },
    {
      "entropy": 6.276887702941894,
      "epoch": 1.825735718407386,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004675722948014855,
      "loss": 5.8363,
      "mean_token_accuracy": 0.19715673625469207,
      "num_tokens": 40093778.0,
      "step": 15820
    },
    {
      "entropy": 6.2018979549407955,
      "epoch": 1.8263127524523948,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046755095048001556,
      "loss": 5.6588,
      "mean_token_accuracy": 0.2037507936358452,
      "num_tokens": 40106604.0,
      "step": 15825
    },
    {
      "entropy": 6.24299840927124,
      "epoch": 1.8268897864974032,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004675295996821899,
      "loss": 5.7967,
      "mean_token_accuracy": 0.19352024048566818,
      "num_tokens": 40119841.0,
      "step": 15830
    },
    {
      "entropy": 6.315410518646241,
      "epoch": 1.827466820542412,
      "grad_norm": 0.921875,
      "learning_rate": 0.000467508242408727,
      "loss": 5.791,
      "mean_token_accuracy": 0.19541407823562623,
      "num_tokens": 40132790.0,
      "step": 15835
    },
    {
      "entropy": 6.284898710250855,
      "epoch": 1.8280438545874207,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004674868786603448,
      "loss": 5.8011,
      "mean_token_accuracy": 0.1885874092578888,
      "num_tokens": 40145175.0,
      "step": 15840
    },
    {
      "entropy": 6.289483070373535,
      "epoch": 1.8286208886324293,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004674655084377622,
      "loss": 5.8318,
      "mean_token_accuracy": 0.19365091025829315,
      "num_tokens": 40158311.0,
      "step": 15845
    },
    {
      "entropy": 6.295757722854614,
      "epoch": 1.829197922677438,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004674441317416978,
      "loss": 5.8664,
      "mean_token_accuracy": 0.18632291704416276,
      "num_tokens": 40170486.0,
      "step": 15850
    },
    {
      "entropy": 6.220132112503052,
      "epoch": 1.8297749567224466,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046742274857287057,
      "loss": 5.8753,
      "mean_token_accuracy": 0.19006639569997788,
      "num_tokens": 40184065.0,
      "step": 15855
    },
    {
      "entropy": 6.295844459533692,
      "epoch": 1.8303519907674553,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004674013589319998,
      "loss": 5.8035,
      "mean_token_accuracy": 0.19630544781684875,
      "num_tokens": 40196387.0,
      "step": 15860
    },
    {
      "entropy": 6.377473545074463,
      "epoch": 1.830929024812464,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004673799628198049,
      "loss": 5.9929,
      "mean_token_accuracy": 0.17197505980730057,
      "num_tokens": 40208983.0,
      "step": 15865
    },
    {
      "entropy": 6.366978168487549,
      "epoch": 1.8315060588574728,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046735856023700546,
      "loss": 5.8979,
      "mean_token_accuracy": 0.18480385094881058,
      "num_tokens": 40223094.0,
      "step": 15870
    },
    {
      "entropy": 6.229592132568359,
      "epoch": 1.8320830929024812,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004673371511843215,
      "loss": 5.7752,
      "mean_token_accuracy": 0.19192079454660416,
      "num_tokens": 40235027.0,
      "step": 15875
    },
    {
      "entropy": 6.24800968170166,
      "epoch": 1.83266012694749,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004673157356624729,
      "loss": 5.7701,
      "mean_token_accuracy": 0.20041902363300323,
      "num_tokens": 40247070.0,
      "step": 15880
    },
    {
      "entropy": 6.174242067337036,
      "epoch": 1.8332371609924984,
      "grad_norm": 1.0,
      "learning_rate": 0.0004672943136721801,
      "loss": 5.7521,
      "mean_token_accuracy": 0.19722027331590652,
      "num_tokens": 40259691.0,
      "step": 15885
    },
    {
      "entropy": 6.294105291366577,
      "epoch": 1.8338141950375073,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004672728852141636,
      "loss": 5.834,
      "mean_token_accuracy": 0.18209927082061766,
      "num_tokens": 40272068.0,
      "step": 15890
    },
    {
      "entropy": 6.269830751419067,
      "epoch": 1.8343912290825157,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046725145028914404,
      "loss": 5.7755,
      "mean_token_accuracy": 0.1917776495218277,
      "num_tokens": 40285224.0,
      "step": 15895
    },
    {
      "entropy": 6.309124088287353,
      "epoch": 1.8349682631275246,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004672300088978425,
      "loss": 5.9048,
      "mean_token_accuracy": 0.18510963767766953,
      "num_tokens": 40297280.0,
      "step": 15900
    },
    {
      "entropy": 6.232507228851318,
      "epoch": 1.8355452971725332,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004672085610409801,
      "loss": 5.7888,
      "mean_token_accuracy": 0.19095606207847596,
      "num_tokens": 40309097.0,
      "step": 15905
    },
    {
      "entropy": 6.234492778778076,
      "epoch": 1.8361223312175419,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046718710671927815,
      "loss": 5.7669,
      "mean_token_accuracy": 0.19423119723796844,
      "num_tokens": 40321665.0,
      "step": 15910
    },
    {
      "entropy": 6.303921556472778,
      "epoch": 1.8366993652625505,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046716564593345843,
      "loss": 5.8233,
      "mean_token_accuracy": 0.19307274967432023,
      "num_tokens": 40333889.0,
      "step": 15915
    },
    {
      "entropy": 6.27899055480957,
      "epoch": 1.8372763993075591,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00046714417868424265,
      "loss": 5.7946,
      "mean_token_accuracy": 0.18963310569524766,
      "num_tokens": 40346448.0,
      "step": 15920
    },
    {
      "entropy": 6.278658103942871,
      "epoch": 1.8378534333525678,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046712270497235284,
      "loss": 5.7561,
      "mean_token_accuracy": 0.18691599667072295,
      "num_tokens": 40359033.0,
      "step": 15925
    },
    {
      "entropy": 6.2421876907348635,
      "epoch": 1.8384304673975764,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004671012247985112,
      "loss": 5.7686,
      "mean_token_accuracy": 0.1965618833899498,
      "num_tokens": 40372598.0,
      "step": 15930
    },
    {
      "entropy": 6.321458148956299,
      "epoch": 1.839007501442585,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046707973816344044,
      "loss": 5.8376,
      "mean_token_accuracy": 0.19177174866199492,
      "num_tokens": 40385009.0,
      "step": 15935
    },
    {
      "entropy": 6.289776754379273,
      "epoch": 1.8395845354875937,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046705824506786304,
      "loss": 5.8443,
      "mean_token_accuracy": 0.19071316868066787,
      "num_tokens": 40396368.0,
      "step": 15940
    },
    {
      "entropy": 6.213792848587036,
      "epoch": 1.8401615695326026,
      "grad_norm": 1.0,
      "learning_rate": 0.00046703674551250193,
      "loss": 5.6852,
      "mean_token_accuracy": 0.2025594562292099,
      "num_tokens": 40408084.0,
      "step": 15945
    },
    {
      "entropy": 6.233068752288818,
      "epoch": 1.840738603577611,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004670152394980803,
      "loss": 5.7494,
      "mean_token_accuracy": 0.19568323940038682,
      "num_tokens": 40420290.0,
      "step": 15950
    },
    {
      "entropy": 6.241335010528564,
      "epoch": 1.8413156376226198,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004669937270253214,
      "loss": 5.7978,
      "mean_token_accuracy": 0.19388476014137268,
      "num_tokens": 40434561.0,
      "step": 15955
    },
    {
      "entropy": 6.25380277633667,
      "epoch": 1.8418926716676283,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000466972208094949,
      "loss": 5.7139,
      "mean_token_accuracy": 0.19614969193935394,
      "num_tokens": 40447065.0,
      "step": 15960
    },
    {
      "entropy": 6.292102527618408,
      "epoch": 1.8424697057126371,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046695068270768665,
      "loss": 5.8383,
      "mean_token_accuracy": 0.18852407038211821,
      "num_tokens": 40460441.0,
      "step": 15965
    },
    {
      "entropy": 6.2078187465667725,
      "epoch": 1.8430467397576455,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046692915086425846,
      "loss": 5.8058,
      "mean_token_accuracy": 0.1946534588932991,
      "num_tokens": 40473381.0,
      "step": 15970
    },
    {
      "entropy": 6.315789699554443,
      "epoch": 1.8436237738026544,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046690761256538857,
      "loss": 5.8635,
      "mean_token_accuracy": 0.18588996380567552,
      "num_tokens": 40486346.0,
      "step": 15975
    },
    {
      "entropy": 6.1990807056427,
      "epoch": 1.844200807847663,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004668860678118015,
      "loss": 5.6974,
      "mean_token_accuracy": 0.20170274525880813,
      "num_tokens": 40497660.0,
      "step": 15980
    },
    {
      "entropy": 6.297496700286866,
      "epoch": 1.8447778418926717,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046686451660422185,
      "loss": 5.8569,
      "mean_token_accuracy": 0.19411776959896088,
      "num_tokens": 40510351.0,
      "step": 15985
    },
    {
      "entropy": 6.286705780029297,
      "epoch": 1.8453548759376803,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046684295894337455,
      "loss": 5.796,
      "mean_token_accuracy": 0.19226273596286775,
      "num_tokens": 40524051.0,
      "step": 15990
    },
    {
      "entropy": 6.265786600112915,
      "epoch": 1.845931909982689,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004668213948299845,
      "loss": 5.8278,
      "mean_token_accuracy": 0.19208566546440126,
      "num_tokens": 40536502.0,
      "step": 15995
    },
    {
      "entropy": 6.270235157012939,
      "epoch": 1.8465089440276976,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004667998242647772,
      "loss": 5.7611,
      "mean_token_accuracy": 0.1958916425704956,
      "num_tokens": 40549383.0,
      "step": 16000
    },
    {
      "entropy": 6.222903251647949,
      "epoch": 1.8470859780727062,
      "grad_norm": 0.953125,
      "learning_rate": 0.000466778247248478,
      "loss": 5.8196,
      "mean_token_accuracy": 0.1900719404220581,
      "num_tokens": 40562125.0,
      "step": 16005
    },
    {
      "entropy": 6.263120031356811,
      "epoch": 1.847663012117715,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046675666378181275,
      "loss": 5.8581,
      "mean_token_accuracy": 0.18920047134160994,
      "num_tokens": 40576293.0,
      "step": 16010
    },
    {
      "entropy": 6.2205424308776855,
      "epoch": 1.8482400461627235,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004667350738655074,
      "loss": 5.6909,
      "mean_token_accuracy": 0.1982554703950882,
      "num_tokens": 40589009.0,
      "step": 16015
    },
    {
      "entropy": 6.3415224075317385,
      "epoch": 1.8488170802077324,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046671347750028806,
      "loss": 5.8776,
      "mean_token_accuracy": 0.18460484892129897,
      "num_tokens": 40601826.0,
      "step": 16020
    },
    {
      "entropy": 6.282042121887207,
      "epoch": 1.8493941142527408,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004666918746868811,
      "loss": 5.8501,
      "mean_token_accuracy": 0.1977418288588524,
      "num_tokens": 40613453.0,
      "step": 16025
    },
    {
      "entropy": 6.29360876083374,
      "epoch": 1.8499711482977497,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004666702654260133,
      "loss": 5.8108,
      "mean_token_accuracy": 0.1901460438966751,
      "num_tokens": 40626205.0,
      "step": 16030
    },
    {
      "entropy": 6.246698617935181,
      "epoch": 1.850548182342758,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00046664864971841113,
      "loss": 5.7368,
      "mean_token_accuracy": 0.19478611797094345,
      "num_tokens": 40637853.0,
      "step": 16035
    },
    {
      "entropy": 6.239678764343262,
      "epoch": 1.851125216387767,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046662702756480195,
      "loss": 5.7885,
      "mean_token_accuracy": 0.19378144592046737,
      "num_tokens": 40651059.0,
      "step": 16040
    },
    {
      "entropy": 6.287972402572632,
      "epoch": 1.8517022504327756,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00046660539896591286,
      "loss": 5.8344,
      "mean_token_accuracy": 0.1872550815343857,
      "num_tokens": 40663434.0,
      "step": 16045
    },
    {
      "entropy": 6.26961579322815,
      "epoch": 1.8522792844777842,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004665837639224713,
      "loss": 5.7939,
      "mean_token_accuracy": 0.19220657348632814,
      "num_tokens": 40675448.0,
      "step": 16050
    },
    {
      "entropy": 6.301541948318482,
      "epoch": 1.8528563185227929,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046656212243520505,
      "loss": 5.8897,
      "mean_token_accuracy": 0.19003831148147582,
      "num_tokens": 40687203.0,
      "step": 16055
    },
    {
      "entropy": 6.279691362380982,
      "epoch": 1.8534333525678015,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046654047450484193,
      "loss": 5.8203,
      "mean_token_accuracy": 0.19006477445363998,
      "num_tokens": 40700057.0,
      "step": 16060
    },
    {
      "entropy": 6.287166261672974,
      "epoch": 1.8540103866128101,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004665188201321102,
      "loss": 5.8211,
      "mean_token_accuracy": 0.19234730750322343,
      "num_tokens": 40714252.0,
      "step": 16065
    },
    {
      "entropy": 6.1973330020904545,
      "epoch": 1.8545874206578188,
      "grad_norm": 0.890625,
      "learning_rate": 0.00046649715931773795,
      "loss": 5.7527,
      "mean_token_accuracy": 0.20028307288885117,
      "num_tokens": 40727666.0,
      "step": 16070
    },
    {
      "entropy": 6.23803391456604,
      "epoch": 1.8551644547028274,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000466475492062454,
      "loss": 5.7644,
      "mean_token_accuracy": 0.19588567912578583,
      "num_tokens": 40740340.0,
      "step": 16075
    },
    {
      "entropy": 6.252159643173218,
      "epoch": 1.855741488747836,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046645381836698684,
      "loss": 5.7433,
      "mean_token_accuracy": 0.19701393842697143,
      "num_tokens": 40752275.0,
      "step": 16080
    },
    {
      "entropy": 6.27236499786377,
      "epoch": 1.856318522792845,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004664321382320657,
      "loss": 5.8192,
      "mean_token_accuracy": 0.18893344700336456,
      "num_tokens": 40765163.0,
      "step": 16085
    },
    {
      "entropy": 6.30366358757019,
      "epoch": 1.8568955568378533,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046641045165841965,
      "loss": 5.8072,
      "mean_token_accuracy": 0.19026349037885665,
      "num_tokens": 40778306.0,
      "step": 16090
    },
    {
      "entropy": 6.30154275894165,
      "epoch": 1.8574725908828622,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046638875864677814,
      "loss": 5.8204,
      "mean_token_accuracy": 0.1877804219722748,
      "num_tokens": 40791409.0,
      "step": 16095
    },
    {
      "entropy": 6.304695463180542,
      "epoch": 1.8580496249278706,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004663670591978708,
      "loss": 5.8536,
      "mean_token_accuracy": 0.1913789391517639,
      "num_tokens": 40803899.0,
      "step": 16100
    },
    {
      "entropy": 6.212089157104492,
      "epoch": 1.8586266589728795,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004663453533124275,
      "loss": 5.7714,
      "mean_token_accuracy": 0.19893707782030107,
      "num_tokens": 40815796.0,
      "step": 16105
    },
    {
      "entropy": 6.231546068191529,
      "epoch": 1.859203693017888,
      "grad_norm": 1.125,
      "learning_rate": 0.0004663236409911783,
      "loss": 5.7737,
      "mean_token_accuracy": 0.20179929286241532,
      "num_tokens": 40829275.0,
      "step": 16110
    },
    {
      "entropy": 6.259246158599853,
      "epoch": 1.8597807270628968,
      "grad_norm": 0.828125,
      "learning_rate": 0.00046630192223485345,
      "loss": 5.7975,
      "mean_token_accuracy": 0.19373729377985,
      "num_tokens": 40841582.0,
      "step": 16115
    },
    {
      "entropy": 6.266847467422485,
      "epoch": 1.8603577611079054,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046628019704418345,
      "loss": 5.7726,
      "mean_token_accuracy": 0.19129831939935685,
      "num_tokens": 40853416.0,
      "step": 16120
    },
    {
      "entropy": 6.078700160980224,
      "epoch": 1.860934795152914,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000466258465419899,
      "loss": 5.6211,
      "mean_token_accuracy": 0.2052396982908249,
      "num_tokens": 40865728.0,
      "step": 16125
    },
    {
      "entropy": 6.194612264633179,
      "epoch": 1.8615118291979227,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004662367273627312,
      "loss": 5.7783,
      "mean_token_accuracy": 0.19038672149181365,
      "num_tokens": 40878222.0,
      "step": 16130
    },
    {
      "entropy": 6.330041170120239,
      "epoch": 1.8620888632429313,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000466214982873411,
      "loss": 5.8671,
      "mean_token_accuracy": 0.18892959356307984,
      "num_tokens": 40890238.0,
      "step": 16135
    },
    {
      "entropy": 6.251346015930176,
      "epoch": 1.86266589728794,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046619323195266975,
      "loss": 5.8587,
      "mean_token_accuracy": 0.20042849630117415,
      "num_tokens": 40903258.0,
      "step": 16140
    },
    {
      "entropy": 6.311374092102051,
      "epoch": 1.8632429313329486,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004661714746012392,
      "loss": 5.76,
      "mean_token_accuracy": 0.18999213725328445,
      "num_tokens": 40916620.0,
      "step": 16145
    },
    {
      "entropy": 6.24119553565979,
      "epoch": 1.8638199653779572,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000466149710819851,
      "loss": 5.8053,
      "mean_token_accuracy": 0.19558262825012207,
      "num_tokens": 40928608.0,
      "step": 16150
    },
    {
      "entropy": 6.230431652069091,
      "epoch": 1.8643969994229659,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004661279406092373,
      "loss": 5.7438,
      "mean_token_accuracy": 0.20112016946077346,
      "num_tokens": 40940543.0,
      "step": 16155
    },
    {
      "entropy": 6.281244945526123,
      "epoch": 1.8649740334679747,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046610616397013027,
      "loss": 5.8243,
      "mean_token_accuracy": 0.18685097992420197,
      "num_tokens": 40953774.0,
      "step": 16160
    },
    {
      "entropy": 6.3247040748596195,
      "epoch": 1.8655510675129832,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004660843809032623,
      "loss": 5.8172,
      "mean_token_accuracy": 0.19491468518972396,
      "num_tokens": 40965673.0,
      "step": 16165
    },
    {
      "entropy": 6.126169013977051,
      "epoch": 1.866128101557992,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004660625914093661,
      "loss": 5.6933,
      "mean_token_accuracy": 0.2072587490081787,
      "num_tokens": 40979301.0,
      "step": 16170
    },
    {
      "entropy": 6.349891614913941,
      "epoch": 1.8667051356030004,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004660407954891745,
      "loss": 5.8222,
      "mean_token_accuracy": 0.1901684135198593,
      "num_tokens": 40991475.0,
      "step": 16175
    },
    {
      "entropy": 6.244673442840576,
      "epoch": 1.8672821696480093,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004660189931434207,
      "loss": 5.7847,
      "mean_token_accuracy": 0.19604144096374512,
      "num_tokens": 41005592.0,
      "step": 16180
    },
    {
      "entropy": 6.285253524780273,
      "epoch": 1.8678592036930177,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004659971843728379,
      "loss": 5.8204,
      "mean_token_accuracy": 0.18606774955987931,
      "num_tokens": 41017812.0,
      "step": 16185
    },
    {
      "entropy": 6.324299669265747,
      "epoch": 1.8684362377380266,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004659753691781597,
      "loss": 5.8046,
      "mean_token_accuracy": 0.19704865515232087,
      "num_tokens": 41029560.0,
      "step": 16190
    },
    {
      "entropy": 6.251793289184571,
      "epoch": 1.8690132717830352,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004659535475601198,
      "loss": 5.7606,
      "mean_token_accuracy": 0.19139352291822434,
      "num_tokens": 41043219.0,
      "step": 16195
    },
    {
      "entropy": 6.174185037612915,
      "epoch": 1.8695903058280439,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046593171951945226,
      "loss": 5.6552,
      "mean_token_accuracy": 0.20583807080984115,
      "num_tokens": 41056314.0,
      "step": 16200
    },
    {
      "entropy": 6.266889190673828,
      "epoch": 1.8701673398730525,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046590988505689114,
      "loss": 5.7733,
      "mean_token_accuracy": 0.19769065529108049,
      "num_tokens": 41068760.0,
      "step": 16205
    },
    {
      "entropy": 6.287529134750367,
      "epoch": 1.8707443739180611,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046588804417317084,
      "loss": 5.8436,
      "mean_token_accuracy": 0.19444874972105025,
      "num_tokens": 41080867.0,
      "step": 16210
    },
    {
      "entropy": 6.204365253448486,
      "epoch": 1.8713214079630698,
      "grad_norm": 0.890625,
      "learning_rate": 0.00046586619686902597,
      "loss": 5.7709,
      "mean_token_accuracy": 0.19827110767364503,
      "num_tokens": 41093573.0,
      "step": 16215
    },
    {
      "entropy": 6.217724275588989,
      "epoch": 1.8718984420080784,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046584434314519144,
      "loss": 5.694,
      "mean_token_accuracy": 0.19997957199811936,
      "num_tokens": 41106381.0,
      "step": 16220
    },
    {
      "entropy": 6.26164140701294,
      "epoch": 1.8724754760530873,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004658224830024022,
      "loss": 5.728,
      "mean_token_accuracy": 0.20802415758371354,
      "num_tokens": 41119791.0,
      "step": 16225
    },
    {
      "entropy": 6.230882883071899,
      "epoch": 1.8730525100980957,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004658006164413935,
      "loss": 5.7942,
      "mean_token_accuracy": 0.1913125365972519,
      "num_tokens": 41132934.0,
      "step": 16230
    },
    {
      "entropy": 6.190240573883057,
      "epoch": 1.8736295441431046,
      "grad_norm": 1.125,
      "learning_rate": 0.0004657787434629009,
      "loss": 5.6569,
      "mean_token_accuracy": 0.2046666920185089,
      "num_tokens": 41144751.0,
      "step": 16235
    },
    {
      "entropy": 6.232805871963501,
      "epoch": 1.874206578188113,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046575686406765993,
      "loss": 5.7886,
      "mean_token_accuracy": 0.18829830586910248,
      "num_tokens": 41157563.0,
      "step": 16240
    },
    {
      "entropy": 6.271267032623291,
      "epoch": 1.8747836122331218,
      "grad_norm": 1.0,
      "learning_rate": 0.00046573497825640664,
      "loss": 5.764,
      "mean_token_accuracy": 0.19409054964780809,
      "num_tokens": 41170130.0,
      "step": 16245
    },
    {
      "entropy": 6.270219421386718,
      "epoch": 1.8753606462781303,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004657130860298771,
      "loss": 5.7554,
      "mean_token_accuracy": 0.1939103677868843,
      "num_tokens": 41182091.0,
      "step": 16250
    },
    {
      "entropy": 6.180555009841919,
      "epoch": 1.8759376803231391,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004656911873888077,
      "loss": 5.7241,
      "mean_token_accuracy": 0.19950637072324753,
      "num_tokens": 41194429.0,
      "step": 16255
    },
    {
      "entropy": 6.294121837615966,
      "epoch": 1.8765147143681478,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004656692823339349,
      "loss": 5.7976,
      "mean_token_accuracy": 0.18713220953941345,
      "num_tokens": 41207732.0,
      "step": 16260
    },
    {
      "entropy": 6.321100950241089,
      "epoch": 1.8770917484131564,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004656473708659955,
      "loss": 5.8591,
      "mean_token_accuracy": 0.1865515112876892,
      "num_tokens": 41220973.0,
      "step": 16265
    },
    {
      "entropy": 6.281434440612793,
      "epoch": 1.877668782458165,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046562545298572646,
      "loss": 5.8271,
      "mean_token_accuracy": 0.1906524419784546,
      "num_tokens": 41232590.0,
      "step": 16270
    },
    {
      "entropy": 6.193849182128906,
      "epoch": 1.8782458165031737,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000465603528693865,
      "loss": 5.7408,
      "mean_token_accuracy": 0.19555010348558427,
      "num_tokens": 41244806.0,
      "step": 16275
    },
    {
      "entropy": 6.307599925994873,
      "epoch": 1.8788228505481823,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046558159799114853,
      "loss": 5.8758,
      "mean_token_accuracy": 0.1852497234940529,
      "num_tokens": 41257818.0,
      "step": 16280
    },
    {
      "entropy": 6.166018962860107,
      "epoch": 1.879399884593191,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004655596608783147,
      "loss": 5.6867,
      "mean_token_accuracy": 0.20225792676210402,
      "num_tokens": 41270396.0,
      "step": 16285
    },
    {
      "entropy": 6.224357843399048,
      "epoch": 1.8799769186381996,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046553771735610136,
      "loss": 5.7784,
      "mean_token_accuracy": 0.19106739461421968,
      "num_tokens": 41282647.0,
      "step": 16290
    },
    {
      "entropy": 6.234634685516357,
      "epoch": 1.8805539526832082,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004655157674252465,
      "loss": 5.7616,
      "mean_token_accuracy": 0.20386153012514113,
      "num_tokens": 41294995.0,
      "step": 16295
    },
    {
      "entropy": 6.24113187789917,
      "epoch": 1.881130986728217,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046549381108648843,
      "loss": 5.8192,
      "mean_token_accuracy": 0.19326353073120117,
      "num_tokens": 41307420.0,
      "step": 16300
    },
    {
      "entropy": 6.267973327636719,
      "epoch": 1.8817080207732255,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004654718483405656,
      "loss": 5.783,
      "mean_token_accuracy": 0.19391203075647354,
      "num_tokens": 41320145.0,
      "step": 16305
    },
    {
      "entropy": 6.327183198928833,
      "epoch": 1.8822850548182344,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046544987918821685,
      "loss": 5.8477,
      "mean_token_accuracy": 0.1883450359106064,
      "num_tokens": 41331740.0,
      "step": 16310
    },
    {
      "entropy": 6.217634773254394,
      "epoch": 1.8828620888632428,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000465427903630181,
      "loss": 5.8338,
      "mean_token_accuracy": 0.18636228740215302,
      "num_tokens": 41344120.0,
      "step": 16315
    },
    {
      "entropy": 6.228198909759522,
      "epoch": 1.8834391229082517,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004654059216671972,
      "loss": 5.7214,
      "mean_token_accuracy": 0.19977904707193375,
      "num_tokens": 41357272.0,
      "step": 16320
    },
    {
      "entropy": 6.3132682800292965,
      "epoch": 1.88401615695326,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004653839333000048,
      "loss": 5.863,
      "mean_token_accuracy": 0.18509012460708618,
      "num_tokens": 41370557.0,
      "step": 16325
    },
    {
      "entropy": 6.218054914474488,
      "epoch": 1.884593190998269,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046536193852934334,
      "loss": 5.7418,
      "mean_token_accuracy": 0.1964620590209961,
      "num_tokens": 41381678.0,
      "step": 16330
    },
    {
      "entropy": 6.264736747741699,
      "epoch": 1.8851702250432776,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046533993735595273,
      "loss": 5.7841,
      "mean_token_accuracy": 0.1914903923869133,
      "num_tokens": 41395025.0,
      "step": 16335
    },
    {
      "entropy": 6.345235013961792,
      "epoch": 1.8857472590882862,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046531792978057277,
      "loss": 5.8586,
      "mean_token_accuracy": 0.18877815157175065,
      "num_tokens": 41407159.0,
      "step": 16340
    },
    {
      "entropy": 6.217168474197388,
      "epoch": 1.8863242931332949,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004652959158039438,
      "loss": 5.7367,
      "mean_token_accuracy": 0.19592882990837096,
      "num_tokens": 41420120.0,
      "step": 16345
    },
    {
      "entropy": 6.244872045516968,
      "epoch": 1.8869013271783035,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004652738954268062,
      "loss": 5.7477,
      "mean_token_accuracy": 0.19732989519834518,
      "num_tokens": 41433637.0,
      "step": 16350
    },
    {
      "entropy": 6.305624675750733,
      "epoch": 1.8874783612233121,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046525186864990073,
      "loss": 5.8655,
      "mean_token_accuracy": 0.19049603193998338,
      "num_tokens": 41447547.0,
      "step": 16355
    },
    {
      "entropy": 6.239545249938965,
      "epoch": 1.8880553952683208,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046522983547396806,
      "loss": 5.7159,
      "mean_token_accuracy": 0.2017817273736,
      "num_tokens": 41460906.0,
      "step": 16360
    },
    {
      "entropy": 6.270957994461059,
      "epoch": 1.8886324293133296,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004652077958997494,
      "loss": 5.7405,
      "mean_token_accuracy": 0.19422893822193146,
      "num_tokens": 41473928.0,
      "step": 16365
    },
    {
      "entropy": 6.252617835998535,
      "epoch": 1.889209463358338,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046518574992798604,
      "loss": 5.8579,
      "mean_token_accuracy": 0.19025357216596603,
      "num_tokens": 41487201.0,
      "step": 16370
    },
    {
      "entropy": 6.21436038017273,
      "epoch": 1.889786497403347,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046516369755941945,
      "loss": 5.7496,
      "mean_token_accuracy": 0.19222482442855834,
      "num_tokens": 41498697.0,
      "step": 16375
    },
    {
      "entropy": 6.321563673019409,
      "epoch": 1.8903635314483553,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004651416387947914,
      "loss": 5.8369,
      "mean_token_accuracy": 0.18715517371892929,
      "num_tokens": 41510533.0,
      "step": 16380
    },
    {
      "entropy": 6.266793584823608,
      "epoch": 1.8909405654933642,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004651195736348437,
      "loss": 5.7761,
      "mean_token_accuracy": 0.18927032947540284,
      "num_tokens": 41522753.0,
      "step": 16385
    },
    {
      "entropy": 6.27957935333252,
      "epoch": 1.8915175995383726,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004650975020803186,
      "loss": 5.7786,
      "mean_token_accuracy": 0.19329847693443297,
      "num_tokens": 41534700.0,
      "step": 16390
    },
    {
      "entropy": 6.248805046081543,
      "epoch": 1.8920946335833815,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004650754241319584,
      "loss": 5.7795,
      "mean_token_accuracy": 0.19324369430541993,
      "num_tokens": 41548126.0,
      "step": 16395
    },
    {
      "entropy": 6.273646306991577,
      "epoch": 1.8926716676283901,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046505333979050573,
      "loss": 5.7664,
      "mean_token_accuracy": 0.19396275281906128,
      "num_tokens": 41560423.0,
      "step": 16400
    },
    {
      "entropy": 6.288725233078003,
      "epoch": 1.8932487016733988,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004650312490567035,
      "loss": 5.8552,
      "mean_token_accuracy": 0.18827454298734664,
      "num_tokens": 41573714.0,
      "step": 16405
    },
    {
      "entropy": 6.363057041168213,
      "epoch": 1.8938257357184074,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004650091519312945,
      "loss": 5.8243,
      "mean_token_accuracy": 0.19138666689395906,
      "num_tokens": 41586785.0,
      "step": 16410
    },
    {
      "entropy": 6.2953544616699215,
      "epoch": 1.894402769763416,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046498704841502203,
      "loss": 5.8316,
      "mean_token_accuracy": 0.19538988471031188,
      "num_tokens": 41599001.0,
      "step": 16415
    },
    {
      "entropy": 6.205738067626953,
      "epoch": 1.8949798038084247,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004649649385086296,
      "loss": 5.8291,
      "mean_token_accuracy": 0.1931207522749901,
      "num_tokens": 41611876.0,
      "step": 16420
    },
    {
      "entropy": 6.222410345077515,
      "epoch": 1.8955568378534333,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004649428222128608,
      "loss": 5.7543,
      "mean_token_accuracy": 0.1988398790359497,
      "num_tokens": 41624074.0,
      "step": 16425
    },
    {
      "entropy": 6.343634176254272,
      "epoch": 1.896133871898442,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046492069952845953,
      "loss": 5.8703,
      "mean_token_accuracy": 0.1901182934641838,
      "num_tokens": 41638063.0,
      "step": 16430
    },
    {
      "entropy": 6.306716203689575,
      "epoch": 1.8967109059434506,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046489857045617,
      "loss": 5.7884,
      "mean_token_accuracy": 0.20012305229902266,
      "num_tokens": 41651123.0,
      "step": 16435
    },
    {
      "entropy": 6.13146710395813,
      "epoch": 1.8972879399884595,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046487643499673625,
      "loss": 5.6678,
      "mean_token_accuracy": 0.2059904009103775,
      "num_tokens": 41664025.0,
      "step": 16440
    },
    {
      "entropy": 6.236382246017456,
      "epoch": 1.8978649740334679,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004648542931509029,
      "loss": 5.7405,
      "mean_token_accuracy": 0.19478911459445952,
      "num_tokens": 41675740.0,
      "step": 16445
    },
    {
      "entropy": 6.2932414531707765,
      "epoch": 1.8984420080784767,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004648321449194148,
      "loss": 5.8224,
      "mean_token_accuracy": 0.18835417181253433,
      "num_tokens": 41688103.0,
      "step": 16450
    },
    {
      "entropy": 6.303207683563232,
      "epoch": 1.8990190421234852,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00046480999030301673,
      "loss": 5.7905,
      "mean_token_accuracy": 0.188765050470829,
      "num_tokens": 41701569.0,
      "step": 16455
    },
    {
      "entropy": 6.251394844055175,
      "epoch": 1.899596076168494,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00046478782930245395,
      "loss": 5.7285,
      "mean_token_accuracy": 0.19941480308771134,
      "num_tokens": 41713373.0,
      "step": 16460
    },
    {
      "entropy": 6.294210433959961,
      "epoch": 1.9001731102135024,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046476566191847176,
      "loss": 5.8518,
      "mean_token_accuracy": 0.190013287961483,
      "num_tokens": 41725550.0,
      "step": 16465
    },
    {
      "entropy": 6.279415464401245,
      "epoch": 1.9007501442585113,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004647434881518159,
      "loss": 5.7656,
      "mean_token_accuracy": 0.20348167270421982,
      "num_tokens": 41739064.0,
      "step": 16470
    },
    {
      "entropy": 6.286113214492798,
      "epoch": 1.90132717830352,
      "grad_norm": 0.890625,
      "learning_rate": 0.00046472130800323194,
      "loss": 5.8353,
      "mean_token_accuracy": 0.20325555503368378,
      "num_tokens": 41752614.0,
      "step": 16475
    },
    {
      "entropy": 6.273328876495361,
      "epoch": 1.9019042123485286,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004646991214734661,
      "loss": 5.8441,
      "mean_token_accuracy": 0.1906109169125557,
      "num_tokens": 41766458.0,
      "step": 16480
    },
    {
      "entropy": 6.132444763183594,
      "epoch": 1.9024812463935372,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004646769285632645,
      "loss": 5.7315,
      "mean_token_accuracy": 0.2030339390039444,
      "num_tokens": 41778742.0,
      "step": 16485
    },
    {
      "entropy": 6.34202823638916,
      "epoch": 1.9030582804385459,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004646547292733737,
      "loss": 5.8375,
      "mean_token_accuracy": 0.18939153850078583,
      "num_tokens": 41793498.0,
      "step": 16490
    },
    {
      "entropy": 6.3038512706756595,
      "epoch": 1.9036353144835545,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004646325236045402,
      "loss": 5.8306,
      "mean_token_accuracy": 0.19345833361148834,
      "num_tokens": 41807042.0,
      "step": 16495
    },
    {
      "entropy": 6.266147136688232,
      "epoch": 1.9042123485285631,
      "grad_norm": 1.0,
      "learning_rate": 0.000464610311557511,
      "loss": 5.7938,
      "mean_token_accuracy": 0.20254878401756288,
      "num_tokens": 41818197.0,
      "step": 16500
    },
    {
      "entropy": 6.267581033706665,
      "epoch": 1.904789382573572,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004645880931330331,
      "loss": 5.829,
      "mean_token_accuracy": 0.19266870617866516,
      "num_tokens": 41830183.0,
      "step": 16505
    },
    {
      "entropy": 6.340952157974243,
      "epoch": 1.9053664166185804,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004645658683318539,
      "loss": 5.8326,
      "mean_token_accuracy": 0.19312754273414612,
      "num_tokens": 41843626.0,
      "step": 16510
    },
    {
      "entropy": 6.338138580322266,
      "epoch": 1.9059434506635893,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004645436371547209,
      "loss": 5.858,
      "mean_token_accuracy": 0.19040029793977736,
      "num_tokens": 41855047.0,
      "step": 16515
    },
    {
      "entropy": 6.294572353363037,
      "epoch": 1.9065204847085977,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046452139960238186,
      "loss": 5.8549,
      "mean_token_accuracy": 0.1923055097460747,
      "num_tokens": 41868480.0,
      "step": 16520
    },
    {
      "entropy": 6.302013254165649,
      "epoch": 1.9070975187536066,
      "grad_norm": 0.81640625,
      "learning_rate": 0.00046449915567558467,
      "loss": 5.838,
      "mean_token_accuracy": 0.18735503703355788,
      "num_tokens": 41882267.0,
      "step": 16525
    },
    {
      "entropy": 6.307902240753174,
      "epoch": 1.907674552798615,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004644769053750775,
      "loss": 5.8901,
      "mean_token_accuracy": 0.18574007451534272,
      "num_tokens": 41894270.0,
      "step": 16530
    },
    {
      "entropy": 6.195050573348999,
      "epoch": 1.9082515868436238,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004644546487016087,
      "loss": 5.7305,
      "mean_token_accuracy": 0.20084942877292633,
      "num_tokens": 41908283.0,
      "step": 16535
    },
    {
      "entropy": 6.389201211929321,
      "epoch": 1.9088286208886325,
      "grad_norm": 0.890625,
      "learning_rate": 0.00046443238565592687,
      "loss": 5.8934,
      "mean_token_accuracy": 0.1849999874830246,
      "num_tokens": 41921411.0,
      "step": 16540
    },
    {
      "entropy": 6.23441801071167,
      "epoch": 1.9094056549336411,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046441011623878087,
      "loss": 5.734,
      "mean_token_accuracy": 0.19902278631925582,
      "num_tokens": 41933221.0,
      "step": 16545
    },
    {
      "entropy": 6.2158966064453125,
      "epoch": 1.9099826889786498,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004643878404509197,
      "loss": 5.7557,
      "mean_token_accuracy": 0.1935444727540016,
      "num_tokens": 41946497.0,
      "step": 16550
    },
    {
      "entropy": 6.321931266784668,
      "epoch": 1.9105597230236584,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046436555829309264,
      "loss": 5.8318,
      "mean_token_accuracy": 0.17710322588682176,
      "num_tokens": 41959118.0,
      "step": 16555
    },
    {
      "entropy": 6.373388051986694,
      "epoch": 1.911136757068667,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000464343269766049,
      "loss": 5.8206,
      "mean_token_accuracy": 0.19518220871686937,
      "num_tokens": 41972185.0,
      "step": 16560
    },
    {
      "entropy": 6.277178049087524,
      "epoch": 1.9117137911136757,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046432097487053857,
      "loss": 5.7574,
      "mean_token_accuracy": 0.19397285729646682,
      "num_tokens": 41984841.0,
      "step": 16565
    },
    {
      "entropy": 6.177239418029785,
      "epoch": 1.9122908251586843,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046429867360731124,
      "loss": 5.7611,
      "mean_token_accuracy": 0.19423114955425264,
      "num_tokens": 41998462.0,
      "step": 16570
    },
    {
      "entropy": 6.255329704284668,
      "epoch": 1.912867859203693,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046427636597711695,
      "loss": 5.796,
      "mean_token_accuracy": 0.1969738319516182,
      "num_tokens": 42011300.0,
      "step": 16575
    },
    {
      "entropy": 6.32501254081726,
      "epoch": 1.9134448932487018,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046425405198070624,
      "loss": 5.8089,
      "mean_token_accuracy": 0.19235641062259673,
      "num_tokens": 42022334.0,
      "step": 16580
    },
    {
      "entropy": 6.2773723125457765,
      "epoch": 1.9140219272937102,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046423173161882944,
      "loss": 5.7622,
      "mean_token_accuracy": 0.19717597216367722,
      "num_tokens": 42035495.0,
      "step": 16585
    },
    {
      "entropy": 6.27013373374939,
      "epoch": 1.914598961338719,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046420940489223735,
      "loss": 5.8043,
      "mean_token_accuracy": 0.18957719206809998,
      "num_tokens": 42046836.0,
      "step": 16590
    },
    {
      "entropy": 6.209165477752686,
      "epoch": 1.9151759953837275,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004641870718016809,
      "loss": 5.7425,
      "mean_token_accuracy": 0.1928727835416794,
      "num_tokens": 42058828.0,
      "step": 16595
    },
    {
      "entropy": 6.1998231410980225,
      "epoch": 1.9157530294287364,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004641647323479113,
      "loss": 5.7546,
      "mean_token_accuracy": 0.19812162667512895,
      "num_tokens": 42070549.0,
      "step": 16600
    },
    {
      "entropy": 6.255869674682617,
      "epoch": 1.9163300634737448,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004641423865316798,
      "loss": 5.749,
      "mean_token_accuracy": 0.19926544576883315,
      "num_tokens": 42082328.0,
      "step": 16605
    },
    {
      "entropy": 6.232963275909424,
      "epoch": 1.9169070975187537,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004641200343537381,
      "loss": 5.7409,
      "mean_token_accuracy": 0.2012409433722496,
      "num_tokens": 42094765.0,
      "step": 16610
    },
    {
      "entropy": 6.290447092056274,
      "epoch": 1.9174841315637623,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000464097675814838,
      "loss": 5.7919,
      "mean_token_accuracy": 0.18661585599184036,
      "num_tokens": 42107418.0,
      "step": 16615
    },
    {
      "entropy": 6.306303024291992,
      "epoch": 1.918061165608771,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004640753109157316,
      "loss": 5.8139,
      "mean_token_accuracy": 0.19126466512680054,
      "num_tokens": 42120623.0,
      "step": 16620
    },
    {
      "entropy": 6.18929648399353,
      "epoch": 1.9186381996537796,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00046405293965717093,
      "loss": 5.7089,
      "mean_token_accuracy": 0.20124325305223464,
      "num_tokens": 42133168.0,
      "step": 16625
    },
    {
      "entropy": 6.169099807739258,
      "epoch": 1.9192152336987882,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004640305620399086,
      "loss": 5.7148,
      "mean_token_accuracy": 0.20203870981931688,
      "num_tokens": 42146512.0,
      "step": 16630
    },
    {
      "entropy": 6.231920576095581,
      "epoch": 1.9197922677437969,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004640081780646971,
      "loss": 5.7675,
      "mean_token_accuracy": 0.19690223634243012,
      "num_tokens": 42158815.0,
      "step": 16635
    },
    {
      "entropy": 6.272224807739258,
      "epoch": 1.9203693017888055,
      "grad_norm": 1.0,
      "learning_rate": 0.00046398578773228954,
      "loss": 5.7875,
      "mean_token_accuracy": 0.19522379636764525,
      "num_tokens": 42170614.0,
      "step": 16640
    },
    {
      "entropy": 6.282483196258545,
      "epoch": 1.9209463358338144,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046396339104343886,
      "loss": 5.7787,
      "mean_token_accuracy": 0.19406996220350264,
      "num_tokens": 42183758.0,
      "step": 16645
    },
    {
      "entropy": 6.279674959182739,
      "epoch": 1.9215233698788228,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004639409879988984,
      "loss": 5.8156,
      "mean_token_accuracy": 0.1937737062573433,
      "num_tokens": 42196111.0,
      "step": 16650
    },
    {
      "entropy": 6.28812952041626,
      "epoch": 1.9221004039238316,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004639185785994216,
      "loss": 5.8222,
      "mean_token_accuracy": 0.19681738168001175,
      "num_tokens": 42209245.0,
      "step": 16655
    },
    {
      "entropy": 6.312433099746704,
      "epoch": 1.92267743796884,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00046389616284576226,
      "loss": 5.801,
      "mean_token_accuracy": 0.19765992611646652,
      "num_tokens": 42222356.0,
      "step": 16660
    },
    {
      "entropy": 6.157405948638916,
      "epoch": 1.923254472013849,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046387374073867435,
      "loss": 5.6649,
      "mean_token_accuracy": 0.21139516085386276,
      "num_tokens": 42236469.0,
      "step": 16665
    },
    {
      "entropy": 6.310583829879761,
      "epoch": 1.9238315060588573,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046385131227891197,
      "loss": 5.8022,
      "mean_token_accuracy": 0.19045525491237641,
      "num_tokens": 42248794.0,
      "step": 16670
    },
    {
      "entropy": 6.194639730453491,
      "epoch": 1.9244085401038662,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004638288774672295,
      "loss": 5.6942,
      "mean_token_accuracy": 0.20971496999263764,
      "num_tokens": 42262099.0,
      "step": 16675
    },
    {
      "entropy": 6.126167440414429,
      "epoch": 1.9249855741488748,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004638064363043816,
      "loss": 5.6732,
      "mean_token_accuracy": 0.1996750056743622,
      "num_tokens": 42274319.0,
      "step": 16680
    },
    {
      "entropy": 6.221785831451416,
      "epoch": 1.9255626081938835,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046378398879112293,
      "loss": 5.7311,
      "mean_token_accuracy": 0.1956358715891838,
      "num_tokens": 42285029.0,
      "step": 16685
    },
    {
      "entropy": 6.316143035888672,
      "epoch": 1.9261396422388921,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004637615349282086,
      "loss": 5.7984,
      "mean_token_accuracy": 0.20264261811971665,
      "num_tokens": 42297909.0,
      "step": 16690
    },
    {
      "entropy": 6.266218471527099,
      "epoch": 1.9267166762839008,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004637390747163938,
      "loss": 5.806,
      "mean_token_accuracy": 0.1904374435544014,
      "num_tokens": 42309918.0,
      "step": 16695
    },
    {
      "entropy": 6.261489152908325,
      "epoch": 1.9272937103289094,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004637166081564339,
      "loss": 5.7826,
      "mean_token_accuracy": 0.1931398630142212,
      "num_tokens": 42322112.0,
      "step": 16700
    },
    {
      "entropy": 6.208802175521851,
      "epoch": 1.927870744373918,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046369413524908473,
      "loss": 5.6888,
      "mean_token_accuracy": 0.20140590965747834,
      "num_tokens": 42335074.0,
      "step": 16705
    },
    {
      "entropy": 6.211992931365967,
      "epoch": 1.9284477784189267,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046367165599510207,
      "loss": 5.7483,
      "mean_token_accuracy": 0.20345504581928253,
      "num_tokens": 42348760.0,
      "step": 16710
    },
    {
      "entropy": 6.284917974472046,
      "epoch": 1.9290248124639353,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004636491703952419,
      "loss": 5.7721,
      "mean_token_accuracy": 0.18829681128263473,
      "num_tokens": 42361745.0,
      "step": 16715
    },
    {
      "entropy": 6.296495819091797,
      "epoch": 1.9296018465089442,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046362667845026057,
      "loss": 5.781,
      "mean_token_accuracy": 0.19478481709957124,
      "num_tokens": 42375267.0,
      "step": 16720
    },
    {
      "entropy": 6.283175277709961,
      "epoch": 1.9301788805539526,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046360418016091467,
      "loss": 5.7746,
      "mean_token_accuracy": 0.1930742859840393,
      "num_tokens": 42387589.0,
      "step": 16725
    },
    {
      "entropy": 6.3217850685119625,
      "epoch": 1.9307559145989615,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046358167552796085,
      "loss": 5.8076,
      "mean_token_accuracy": 0.19299133270978927,
      "num_tokens": 42400144.0,
      "step": 16730
    },
    {
      "entropy": 6.277649831771851,
      "epoch": 1.9313329486439699,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046355916455215597,
      "loss": 5.8507,
      "mean_token_accuracy": 0.18077297359704972,
      "num_tokens": 42412529.0,
      "step": 16735
    },
    {
      "entropy": 6.299275207519531,
      "epoch": 1.9319099826889787,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004635366472342573,
      "loss": 5.824,
      "mean_token_accuracy": 0.19157508611679078,
      "num_tokens": 42424155.0,
      "step": 16740
    },
    {
      "entropy": 6.258406352996826,
      "epoch": 1.9324870167339872,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004635141235750222,
      "loss": 5.7375,
      "mean_token_accuracy": 0.19815794229507447,
      "num_tokens": 42436299.0,
      "step": 16745
    },
    {
      "entropy": 6.192769861221313,
      "epoch": 1.933064050778996,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046349159357520815,
      "loss": 5.7399,
      "mean_token_accuracy": 0.20777881592512132,
      "num_tokens": 42450059.0,
      "step": 16750
    },
    {
      "entropy": 6.270134973526001,
      "epoch": 1.9336410848240047,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000463469057235573,
      "loss": 5.7622,
      "mean_token_accuracy": 0.1948457181453705,
      "num_tokens": 42463403.0,
      "step": 16755
    },
    {
      "entropy": 6.329901123046875,
      "epoch": 1.9342181188690133,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046344651455687476,
      "loss": 5.8252,
      "mean_token_accuracy": 0.19359085857868194,
      "num_tokens": 42476115.0,
      "step": 16760
    },
    {
      "entropy": 6.254556035995483,
      "epoch": 1.934795152914022,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046342396553987155,
      "loss": 5.8332,
      "mean_token_accuracy": 0.19167038500308992,
      "num_tokens": 42487992.0,
      "step": 16765
    },
    {
      "entropy": 6.173498678207397,
      "epoch": 1.9353721869590306,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046340141018532186,
      "loss": 5.7047,
      "mean_token_accuracy": 0.20131428688764572,
      "num_tokens": 42499961.0,
      "step": 16770
    },
    {
      "entropy": 6.285065174102783,
      "epoch": 1.9359492210040392,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004633788484939843,
      "loss": 5.8332,
      "mean_token_accuracy": 0.19102754145860673,
      "num_tokens": 42511432.0,
      "step": 16775
    },
    {
      "entropy": 6.2204570293426515,
      "epoch": 1.9365262550490479,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046335628046661776,
      "loss": 5.7386,
      "mean_token_accuracy": 0.18399754613637925,
      "num_tokens": 42523904.0,
      "step": 16780
    },
    {
      "entropy": 6.33710503578186,
      "epoch": 1.9371032890940567,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046333370610398135,
      "loss": 5.7958,
      "mean_token_accuracy": 0.19457192569971085,
      "num_tokens": 42537625.0,
      "step": 16785
    },
    {
      "entropy": 6.220838832855224,
      "epoch": 1.9376803231390651,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004633111254068342,
      "loss": 5.768,
      "mean_token_accuracy": 0.1984094962477684,
      "num_tokens": 42551374.0,
      "step": 16790
    },
    {
      "entropy": 6.140050983428955,
      "epoch": 1.938257357184074,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004632885383759359,
      "loss": 5.6405,
      "mean_token_accuracy": 0.22144414633512496,
      "num_tokens": 42564325.0,
      "step": 16795
    },
    {
      "entropy": 6.229633569717407,
      "epoch": 1.9388343912290824,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046326594501204624,
      "loss": 5.7656,
      "mean_token_accuracy": 0.19468972682952881,
      "num_tokens": 42576353.0,
      "step": 16800
    },
    {
      "entropy": 6.150761222839355,
      "epoch": 1.9394114252740913,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000463243345315925,
      "loss": 5.661,
      "mean_token_accuracy": 0.20656837821006774,
      "num_tokens": 42589199.0,
      "step": 16805
    },
    {
      "entropy": 6.251586532592773,
      "epoch": 1.9399884593190997,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046322073928833224,
      "loss": 5.8357,
      "mean_token_accuracy": 0.19115519374608994,
      "num_tokens": 42602145.0,
      "step": 16810
    },
    {
      "entropy": 6.296150207519531,
      "epoch": 1.9405654933641086,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004631981269300285,
      "loss": 5.8127,
      "mean_token_accuracy": 0.19709358960390091,
      "num_tokens": 42615234.0,
      "step": 16815
    },
    {
      "entropy": 6.24841628074646,
      "epoch": 1.9411425274091172,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004631755082417742,
      "loss": 5.7655,
      "mean_token_accuracy": 0.1966732695698738,
      "num_tokens": 42626515.0,
      "step": 16820
    },
    {
      "entropy": 6.299662637710571,
      "epoch": 1.9417195614541258,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004631528832243302,
      "loss": 5.7783,
      "mean_token_accuracy": 0.19628046303987504,
      "num_tokens": 42638217.0,
      "step": 16825
    },
    {
      "entropy": 6.202601671218872,
      "epoch": 1.9422965954991345,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046313025187845735,
      "loss": 5.7027,
      "mean_token_accuracy": 0.19806416630744933,
      "num_tokens": 42650070.0,
      "step": 16830
    },
    {
      "entropy": 6.259328842163086,
      "epoch": 1.942873629544143,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046310761420491705,
      "loss": 5.7789,
      "mean_token_accuracy": 0.19775720834732055,
      "num_tokens": 42662849.0,
      "step": 16835
    },
    {
      "entropy": 6.264229249954224,
      "epoch": 1.9434506635891518,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004630849702044705,
      "loss": 5.7631,
      "mean_token_accuracy": 0.19413664042949677,
      "num_tokens": 42676001.0,
      "step": 16840
    },
    {
      "entropy": 6.299164772033691,
      "epoch": 1.9440276976341604,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046306231987787937,
      "loss": 5.7644,
      "mean_token_accuracy": 0.195182004570961,
      "num_tokens": 42688115.0,
      "step": 16845
    },
    {
      "entropy": 6.1574790477752686,
      "epoch": 1.944604731679169,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046303966322590556,
      "loss": 5.7248,
      "mean_token_accuracy": 0.1994083568453789,
      "num_tokens": 42700123.0,
      "step": 16850
    },
    {
      "entropy": 6.316443395614624,
      "epoch": 1.9451817657241777,
      "grad_norm": 1.0,
      "learning_rate": 0.0004630170002493111,
      "loss": 5.8277,
      "mean_token_accuracy": 0.19929512739181518,
      "num_tokens": 42713942.0,
      "step": 16855
    },
    {
      "entropy": 6.297909116744995,
      "epoch": 1.9457587997691865,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046299433094885826,
      "loss": 5.8108,
      "mean_token_accuracy": 0.19474846571683885,
      "num_tokens": 42726596.0,
      "step": 16860
    },
    {
      "entropy": 6.300844144821167,
      "epoch": 1.946335833814195,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046297165532530944,
      "loss": 5.8618,
      "mean_token_accuracy": 0.1863092616200447,
      "num_tokens": 42738169.0,
      "step": 16865
    },
    {
      "entropy": 6.384887266159057,
      "epoch": 1.9469128678592038,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004629489733794274,
      "loss": 5.7982,
      "mean_token_accuracy": 0.19159150421619414,
      "num_tokens": 42751262.0,
      "step": 16870
    },
    {
      "entropy": 6.25150351524353,
      "epoch": 1.9474899019042122,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000462926285111975,
      "loss": 5.7467,
      "mean_token_accuracy": 0.1960839718580246,
      "num_tokens": 42763366.0,
      "step": 16875
    },
    {
      "entropy": 6.216044139862061,
      "epoch": 1.948066935949221,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046290359052371535,
      "loss": 5.7761,
      "mean_token_accuracy": 0.19175081551074982,
      "num_tokens": 42775801.0,
      "step": 16880
    },
    {
      "entropy": 6.370111799240112,
      "epoch": 1.9486439699942295,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004628808896154117,
      "loss": 5.8137,
      "mean_token_accuracy": 0.19441015273332596,
      "num_tokens": 42788610.0,
      "step": 16885
    },
    {
      "entropy": 6.359094047546387,
      "epoch": 1.9492210040392384,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004628581823878277,
      "loss": 5.8937,
      "mean_token_accuracy": 0.19397713840007783,
      "num_tokens": 42801517.0,
      "step": 16890
    },
    {
      "entropy": 6.233689689636231,
      "epoch": 1.949798038084247,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000462835468841727,
      "loss": 5.7297,
      "mean_token_accuracy": 0.19526649564504622,
      "num_tokens": 42814007.0,
      "step": 16895
    },
    {
      "entropy": 6.184755134582519,
      "epoch": 1.9503750721292556,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004628127489778737,
      "loss": 5.657,
      "mean_token_accuracy": 0.19246759712696077,
      "num_tokens": 42826960.0,
      "step": 16900
    },
    {
      "entropy": 6.2426739692687985,
      "epoch": 1.9509521061742643,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004627900227970318,
      "loss": 5.7502,
      "mean_token_accuracy": 0.19220670610666274,
      "num_tokens": 42838950.0,
      "step": 16905
    },
    {
      "entropy": 6.262869167327881,
      "epoch": 1.951529140219273,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046276729029996576,
      "loss": 5.8203,
      "mean_token_accuracy": 0.19222410917282104,
      "num_tokens": 42853470.0,
      "step": 16910
    },
    {
      "entropy": 6.256709098815918,
      "epoch": 1.9521061742642816,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046274455148744025,
      "loss": 5.7102,
      "mean_token_accuracy": 0.19570399820804596,
      "num_tokens": 42865946.0,
      "step": 16915
    },
    {
      "entropy": 6.278725433349609,
      "epoch": 1.9526832083092902,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046272180636022,
      "loss": 5.8593,
      "mean_token_accuracy": 0.183206607401371,
      "num_tokens": 42878248.0,
      "step": 16920
    },
    {
      "entropy": 6.1847998142242435,
      "epoch": 1.953260242354299,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046269905491907,
      "loss": 5.7124,
      "mean_token_accuracy": 0.1982881873846054,
      "num_tokens": 42890750.0,
      "step": 16925
    },
    {
      "entropy": 6.234466505050659,
      "epoch": 1.9538372763993075,
      "grad_norm": 0.875,
      "learning_rate": 0.0004626762971647555,
      "loss": 5.7192,
      "mean_token_accuracy": 0.19522606432437897,
      "num_tokens": 42904737.0,
      "step": 16930
    },
    {
      "entropy": 6.276740026473999,
      "epoch": 1.9544143104443163,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046265353309804205,
      "loss": 5.8369,
      "mean_token_accuracy": 0.1870988979935646,
      "num_tokens": 42917374.0,
      "step": 16935
    },
    {
      "entropy": 6.244482851028442,
      "epoch": 1.9549913444893248,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004626307627196952,
      "loss": 5.7421,
      "mean_token_accuracy": 0.1940678209066391,
      "num_tokens": 42930068.0,
      "step": 16940
    },
    {
      "entropy": 6.314016914367675,
      "epoch": 1.9555683785343336,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004626079860304808,
      "loss": 5.8575,
      "mean_token_accuracy": 0.1865780085325241,
      "num_tokens": 42944014.0,
      "step": 16945
    },
    {
      "entropy": 6.299301958084106,
      "epoch": 1.956145412579342,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046258520303116496,
      "loss": 5.7928,
      "mean_token_accuracy": 0.18596257269382477,
      "num_tokens": 42956765.0,
      "step": 16950
    },
    {
      "entropy": 6.234346199035644,
      "epoch": 1.956722446624351,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004625624137225141,
      "loss": 5.7513,
      "mean_token_accuracy": 0.2019268348813057,
      "num_tokens": 42970661.0,
      "step": 16955
    },
    {
      "entropy": 6.232459449768067,
      "epoch": 1.9572994806693593,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004625396181052945,
      "loss": 5.7576,
      "mean_token_accuracy": 0.1912333607673645,
      "num_tokens": 42983359.0,
      "step": 16960
    },
    {
      "entropy": 6.3312461376190186,
      "epoch": 1.9578765147143682,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046251681618027316,
      "loss": 5.8495,
      "mean_token_accuracy": 0.18995251804590224,
      "num_tokens": 42995650.0,
      "step": 16965
    },
    {
      "entropy": 6.277827978134155,
      "epoch": 1.9584535487593768,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004624940079482167,
      "loss": 5.8043,
      "mean_token_accuracy": 0.19214609265327454,
      "num_tokens": 43009161.0,
      "step": 16970
    },
    {
      "entropy": 6.276555776596069,
      "epoch": 1.9590305828043855,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046247119340989254,
      "loss": 5.7857,
      "mean_token_accuracy": 0.20107332468032837,
      "num_tokens": 43022314.0,
      "step": 16975
    },
    {
      "entropy": 6.265553283691406,
      "epoch": 1.959607616849394,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004624483725660678,
      "loss": 5.7358,
      "mean_token_accuracy": 0.19886453002691268,
      "num_tokens": 43034458.0,
      "step": 16980
    },
    {
      "entropy": 6.2213709354400635,
      "epoch": 1.9601846508944027,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004624255454175102,
      "loss": 5.6912,
      "mean_token_accuracy": 0.2044117584824562,
      "num_tokens": 43047341.0,
      "step": 16985
    },
    {
      "entropy": 6.269777011871338,
      "epoch": 1.9607616849394114,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004624027119649875,
      "loss": 5.7845,
      "mean_token_accuracy": 0.19658008664846421,
      "num_tokens": 43059363.0,
      "step": 16990
    },
    {
      "entropy": 6.235811376571656,
      "epoch": 1.96133871898442,
      "grad_norm": 1.0,
      "learning_rate": 0.00046237987220926766,
      "loss": 5.7597,
      "mean_token_accuracy": 0.19717204123735427,
      "num_tokens": 43070574.0,
      "step": 16995
    },
    {
      "entropy": 6.228890132904053,
      "epoch": 1.9619157530294289,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046235702615111886,
      "loss": 5.8142,
      "mean_token_accuracy": 0.19661410599946977,
      "num_tokens": 43082582.0,
      "step": 17000
    },
    {
      "entropy": 6.293433904647827,
      "epoch": 1.9624927870744373,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004623341737913096,
      "loss": 5.8151,
      "mean_token_accuracy": 0.1928779140114784,
      "num_tokens": 43096808.0,
      "step": 17005
    },
    {
      "entropy": 6.317771339416504,
      "epoch": 1.9630698211194462,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004623113151306085,
      "loss": 5.8554,
      "mean_token_accuracy": 0.18486102819442748,
      "num_tokens": 43108977.0,
      "step": 17010
    },
    {
      "entropy": 6.296752166748047,
      "epoch": 1.9636468551644546,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046228845016978425,
      "loss": 5.9085,
      "mean_token_accuracy": 0.182223142683506,
      "num_tokens": 43121135.0,
      "step": 17015
    },
    {
      "entropy": 6.206674098968506,
      "epoch": 1.9642238892094634,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004622655789096061,
      "loss": 5.7209,
      "mean_token_accuracy": 0.2048693850636482,
      "num_tokens": 43132955.0,
      "step": 17020
    },
    {
      "entropy": 6.290903091430664,
      "epoch": 1.9648009232544719,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046224270135084315,
      "loss": 5.707,
      "mean_token_accuracy": 0.2017136797308922,
      "num_tokens": 43145029.0,
      "step": 17025
    },
    {
      "entropy": 6.2341893196105955,
      "epoch": 1.9653779572994807,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046221981749426503,
      "loss": 5.7826,
      "mean_token_accuracy": 0.1951350137591362,
      "num_tokens": 43157594.0,
      "step": 17030
    },
    {
      "entropy": 6.304322814941406,
      "epoch": 1.9659549913444894,
      "grad_norm": 0.859375,
      "learning_rate": 0.00046219692734064124,
      "loss": 5.8241,
      "mean_token_accuracy": 0.18392543643712997,
      "num_tokens": 43169922.0,
      "step": 17035
    },
    {
      "entropy": 6.301657819747925,
      "epoch": 1.966532025389498,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004621740308907419,
      "loss": 5.7264,
      "mean_token_accuracy": 0.19587419033050538,
      "num_tokens": 43183151.0,
      "step": 17040
    },
    {
      "entropy": 6.241607666015625,
      "epoch": 1.9671090594345066,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004621511281453369,
      "loss": 5.7061,
      "mean_token_accuracy": 0.19622530937194824,
      "num_tokens": 43195542.0,
      "step": 17045
    },
    {
      "entropy": 6.292487525939942,
      "epoch": 1.9676860934795153,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004621282191051967,
      "loss": 5.8022,
      "mean_token_accuracy": 0.1928291514515877,
      "num_tokens": 43208560.0,
      "step": 17050
    },
    {
      "entropy": 6.141060924530029,
      "epoch": 1.968263127524524,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004621053037710918,
      "loss": 5.5891,
      "mean_token_accuracy": 0.20298729985952377,
      "num_tokens": 43220959.0,
      "step": 17055
    },
    {
      "entropy": 6.2530725479125975,
      "epoch": 1.9688401615695326,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000462082382143793,
      "loss": 5.7991,
      "mean_token_accuracy": 0.18837548792362213,
      "num_tokens": 43231458.0,
      "step": 17060
    },
    {
      "entropy": 6.274489641189575,
      "epoch": 1.9694171956145412,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046205945422407113,
      "loss": 5.7617,
      "mean_token_accuracy": 0.1958347886800766,
      "num_tokens": 43243712.0,
      "step": 17065
    },
    {
      "entropy": 6.254511880874634,
      "epoch": 1.9699942296595498,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00046203652001269754,
      "loss": 5.7635,
      "mean_token_accuracy": 0.1965350791811943,
      "num_tokens": 43256931.0,
      "step": 17070
    },
    {
      "entropy": 6.290057897567749,
      "epoch": 1.9705712637045587,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046201357951044337,
      "loss": 5.8105,
      "mean_token_accuracy": 0.18815018981695175,
      "num_tokens": 43270420.0,
      "step": 17075
    },
    {
      "entropy": 6.337242889404297,
      "epoch": 1.9711482977495671,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046199063271808047,
      "loss": 5.8193,
      "mean_token_accuracy": 0.190450718998909,
      "num_tokens": 43283183.0,
      "step": 17080
    },
    {
      "entropy": 6.210440301895142,
      "epoch": 1.971725331794576,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004619676796363804,
      "loss": 5.6624,
      "mean_token_accuracy": 0.21262233704328537,
      "num_tokens": 43295608.0,
      "step": 17085
    },
    {
      "entropy": 6.212130880355835,
      "epoch": 1.9723023658395844,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046194472026611546,
      "loss": 5.8031,
      "mean_token_accuracy": 0.19070430994033813,
      "num_tokens": 43307189.0,
      "step": 17090
    },
    {
      "entropy": 6.249095296859741,
      "epoch": 1.9728793998845933,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004619217546080576,
      "loss": 5.7753,
      "mean_token_accuracy": 0.19518021494150162,
      "num_tokens": 43319729.0,
      "step": 17095
    },
    {
      "entropy": 6.166251182556152,
      "epoch": 1.9734564339296017,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004618987826629794,
      "loss": 5.7386,
      "mean_token_accuracy": 0.20338575541973114,
      "num_tokens": 43331481.0,
      "step": 17100
    },
    {
      "entropy": 6.225678968429565,
      "epoch": 1.9740334679746105,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046187580443165344,
      "loss": 5.7607,
      "mean_token_accuracy": 0.1990882486104965,
      "num_tokens": 43344815.0,
      "step": 17105
    },
    {
      "entropy": 6.336755752563477,
      "epoch": 1.9746105020196192,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004618528199148527,
      "loss": 5.8489,
      "mean_token_accuracy": 0.18714679032564163,
      "num_tokens": 43358825.0,
      "step": 17110
    },
    {
      "entropy": 6.274466180801392,
      "epoch": 1.9751875360646278,
      "grad_norm": 0.890625,
      "learning_rate": 0.00046182982911335016,
      "loss": 5.7683,
      "mean_token_accuracy": 0.19329881072044372,
      "num_tokens": 43372536.0,
      "step": 17115
    },
    {
      "entropy": 6.339049053192139,
      "epoch": 1.9757645701096365,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046180683202791905,
      "loss": 5.8618,
      "mean_token_accuracy": 0.18405837267637254,
      "num_tokens": 43384658.0,
      "step": 17120
    },
    {
      "entropy": 6.316912841796875,
      "epoch": 1.976341604154645,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000461783828659333,
      "loss": 5.8373,
      "mean_token_accuracy": 0.1914736732840538,
      "num_tokens": 43396532.0,
      "step": 17125
    },
    {
      "entropy": 6.324210977554321,
      "epoch": 1.9769186381996537,
      "grad_norm": 1.0,
      "learning_rate": 0.00046176081900836565,
      "loss": 5.8721,
      "mean_token_accuracy": 0.18451156169176103,
      "num_tokens": 43409811.0,
      "step": 17130
    },
    {
      "entropy": 6.3312663555145265,
      "epoch": 1.9774956722446624,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046173780307579086,
      "loss": 5.8037,
      "mean_token_accuracy": 0.18460054248571395,
      "num_tokens": 43422835.0,
      "step": 17135
    },
    {
      "entropy": 6.25698881149292,
      "epoch": 1.9780727062896712,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004617147808623829,
      "loss": 5.7887,
      "mean_token_accuracy": 0.2014346867799759,
      "num_tokens": 43434418.0,
      "step": 17140
    },
    {
      "entropy": 6.315674638748169,
      "epoch": 1.9786497403346797,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046169175236891605,
      "loss": 5.7761,
      "mean_token_accuracy": 0.19286873936653137,
      "num_tokens": 43447447.0,
      "step": 17145
    },
    {
      "entropy": 6.242588758468628,
      "epoch": 1.9792267743796885,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004616687175961648,
      "loss": 5.7597,
      "mean_token_accuracy": 0.19495011419057845,
      "num_tokens": 43461205.0,
      "step": 17150
    },
    {
      "entropy": 6.275207042694092,
      "epoch": 1.979803808424697,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004616456765449039,
      "loss": 5.756,
      "mean_token_accuracy": 0.1966785728931427,
      "num_tokens": 43473838.0,
      "step": 17155
    },
    {
      "entropy": 6.2823442935943605,
      "epoch": 1.9803808424697058,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046162262921590845,
      "loss": 5.8553,
      "mean_token_accuracy": 0.1864657461643219,
      "num_tokens": 43485661.0,
      "step": 17160
    },
    {
      "entropy": 6.316865730285644,
      "epoch": 1.9809578765147142,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004615995756099535,
      "loss": 5.7865,
      "mean_token_accuracy": 0.19358981102705003,
      "num_tokens": 43499219.0,
      "step": 17165
    },
    {
      "entropy": 6.26077880859375,
      "epoch": 1.981534910559723,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004615765157278146,
      "loss": 5.7604,
      "mean_token_accuracy": 0.19489420503377913,
      "num_tokens": 43512588.0,
      "step": 17170
    },
    {
      "entropy": 6.308348321914673,
      "epoch": 1.9821119446047317,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046155344957026726,
      "loss": 5.8343,
      "mean_token_accuracy": 0.19233404397964476,
      "num_tokens": 43525419.0,
      "step": 17175
    },
    {
      "entropy": 6.304824161529541,
      "epoch": 1.9826889786497404,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004615303771380873,
      "loss": 5.7552,
      "mean_token_accuracy": 0.1940368741750717,
      "num_tokens": 43540244.0,
      "step": 17180
    },
    {
      "entropy": 6.280411434173584,
      "epoch": 1.983266012694749,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046150729843205077,
      "loss": 5.8079,
      "mean_token_accuracy": 0.18945251703262328,
      "num_tokens": 43552108.0,
      "step": 17185
    },
    {
      "entropy": 6.3281933784484865,
      "epoch": 1.9838430467397576,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046148421345293384,
      "loss": 5.7824,
      "mean_token_accuracy": 0.1947069361805916,
      "num_tokens": 43563146.0,
      "step": 17190
    },
    {
      "entropy": 6.17760443687439,
      "epoch": 1.9844200807847663,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004614611222015131,
      "loss": 5.7913,
      "mean_token_accuracy": 0.19198838621377945,
      "num_tokens": 43577188.0,
      "step": 17195
    },
    {
      "entropy": 6.228463840484619,
      "epoch": 1.984997114829775,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046143802467856507,
      "loss": 5.7198,
      "mean_token_accuracy": 0.201992067694664,
      "num_tokens": 43590366.0,
      "step": 17200
    },
    {
      "entropy": 6.3014179229736325,
      "epoch": 1.9855741488747836,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046141492088486673,
      "loss": 5.7817,
      "mean_token_accuracy": 0.18913347721099855,
      "num_tokens": 43603830.0,
      "step": 17205
    },
    {
      "entropy": 6.1208446502685545,
      "epoch": 1.9861511829197922,
      "grad_norm": 1.0,
      "learning_rate": 0.0004613918108211951,
      "loss": 5.6112,
      "mean_token_accuracy": 0.20270614475011825,
      "num_tokens": 43617730.0,
      "step": 17210
    },
    {
      "entropy": 6.288542985916138,
      "epoch": 1.986728216964801,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004613686944883275,
      "loss": 5.7995,
      "mean_token_accuracy": 0.1900461956858635,
      "num_tokens": 43630420.0,
      "step": 17215
    },
    {
      "entropy": 6.295857286453247,
      "epoch": 1.9873052510098095,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046134557188704146,
      "loss": 5.83,
      "mean_token_accuracy": 0.19219378829002381,
      "num_tokens": 43643089.0,
      "step": 17220
    },
    {
      "entropy": 6.22946138381958,
      "epoch": 1.9878822850548183,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046132244301811466,
      "loss": 5.7653,
      "mean_token_accuracy": 0.200616355240345,
      "num_tokens": 43656803.0,
      "step": 17225
    },
    {
      "entropy": 6.261227607727051,
      "epoch": 1.9884593190998268,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046129930788232497,
      "loss": 5.7089,
      "mean_token_accuracy": 0.20484155416488647,
      "num_tokens": 43669163.0,
      "step": 17230
    },
    {
      "entropy": 6.2542376041412355,
      "epoch": 1.9890363531448356,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046127616648045073,
      "loss": 5.8291,
      "mean_token_accuracy": 0.19284798353910446,
      "num_tokens": 43680724.0,
      "step": 17235
    },
    {
      "entropy": 6.285570096969605,
      "epoch": 1.989613387189844,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046125301881327007,
      "loss": 5.8342,
      "mean_token_accuracy": 0.19420784413814546,
      "num_tokens": 43694823.0,
      "step": 17240
    },
    {
      "entropy": 6.327510738372803,
      "epoch": 1.990190421234853,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046122986488156167,
      "loss": 5.8383,
      "mean_token_accuracy": 0.1897768497467041,
      "num_tokens": 43707406.0,
      "step": 17245
    },
    {
      "entropy": 6.27443904876709,
      "epoch": 1.9907674552798615,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046120670468610426,
      "loss": 5.7689,
      "mean_token_accuracy": 0.1933450698852539,
      "num_tokens": 43719663.0,
      "step": 17250
    },
    {
      "entropy": 6.23111252784729,
      "epoch": 1.9913444893248702,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046118353822767683,
      "loss": 5.7687,
      "mean_token_accuracy": 0.19529650509357452,
      "num_tokens": 43732212.0,
      "step": 17255
    },
    {
      "entropy": 6.357538509368896,
      "epoch": 1.9919215233698788,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004611603655070586,
      "loss": 5.818,
      "mean_token_accuracy": 0.18879110217094422,
      "num_tokens": 43743686.0,
      "step": 17260
    },
    {
      "entropy": 6.242610073089599,
      "epoch": 1.9924985574148875,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046113718652502896,
      "loss": 5.7168,
      "mean_token_accuracy": 0.20119693726301194,
      "num_tokens": 43756578.0,
      "step": 17265
    },
    {
      "entropy": 6.279796886444092,
      "epoch": 1.993075591459896,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004611140012823675,
      "loss": 5.8114,
      "mean_token_accuracy": 0.19276085942983628,
      "num_tokens": 43770054.0,
      "step": 17270
    },
    {
      "entropy": 6.297373199462891,
      "epoch": 1.9936526255049047,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046109080977985395,
      "loss": 5.6997,
      "mean_token_accuracy": 0.21232107132673264,
      "num_tokens": 43782669.0,
      "step": 17275
    },
    {
      "entropy": 6.254519128799439,
      "epoch": 1.9942296595499136,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046106761201826854,
      "loss": 5.6932,
      "mean_token_accuracy": 0.19854396879673003,
      "num_tokens": 43795330.0,
      "step": 17280
    },
    {
      "entropy": 6.238685178756714,
      "epoch": 1.994806693594922,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046104440799839145,
      "loss": 5.7791,
      "mean_token_accuracy": 0.20067429393529893,
      "num_tokens": 43808380.0,
      "step": 17285
    },
    {
      "entropy": 6.278558349609375,
      "epoch": 1.9953837276399309,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000461021197721003,
      "loss": 5.7892,
      "mean_token_accuracy": 0.1948169767856598,
      "num_tokens": 43821346.0,
      "step": 17290
    },
    {
      "entropy": 6.284453344345093,
      "epoch": 1.9959607616849393,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046099798118688407,
      "loss": 5.7684,
      "mean_token_accuracy": 0.19073053747415541,
      "num_tokens": 43832989.0,
      "step": 17295
    },
    {
      "entropy": 6.295646905899048,
      "epoch": 1.9965377957299482,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004609747583968154,
      "loss": 5.7746,
      "mean_token_accuracy": 0.20367611199617386,
      "num_tokens": 43846607.0,
      "step": 17300
    },
    {
      "entropy": 6.1778778553009035,
      "epoch": 1.9971148297749566,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004609515293515781,
      "loss": 5.7022,
      "mean_token_accuracy": 0.20777408331632613,
      "num_tokens": 43860244.0,
      "step": 17305
    },
    {
      "entropy": 6.283413410186768,
      "epoch": 1.9976918638199654,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004609282940519535,
      "loss": 5.8292,
      "mean_token_accuracy": 0.20035456717014313,
      "num_tokens": 43874243.0,
      "step": 17310
    },
    {
      "entropy": 6.218620777130127,
      "epoch": 1.998268897864974,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004609050524987231,
      "loss": 5.7203,
      "mean_token_accuracy": 0.20023976713418962,
      "num_tokens": 43887241.0,
      "step": 17315
    },
    {
      "entropy": 6.290399074554443,
      "epoch": 1.9988459319099827,
      "grad_norm": 1.0,
      "learning_rate": 0.0004608818046926686,
      "loss": 5.7791,
      "mean_token_accuracy": 0.19466456025838852,
      "num_tokens": 43901006.0,
      "step": 17320
    },
    {
      "entropy": 6.300667715072632,
      "epoch": 1.9994229659549914,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004608585506345719,
      "loss": 5.7927,
      "mean_token_accuracy": 0.19142765998840333,
      "num_tokens": 43913167.0,
      "step": 17325
    },
    {
      "entropy": 6.272757291793823,
      "epoch": 2.0,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004608352903252152,
      "loss": 5.8252,
      "mean_token_accuracy": 0.19059846848249434,
      "num_tokens": 43926234.0,
      "step": 17330
    },
    {
      "entropy": 6.219476318359375,
      "epoch": 2.000577034045009,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046081202376538084,
      "loss": 5.648,
      "mean_token_accuracy": 0.20688224136829375,
      "num_tokens": 43939047.0,
      "step": 17335
    },
    {
      "entropy": 6.215475940704346,
      "epoch": 2.0011540680900173,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004607887509558513,
      "loss": 5.655,
      "mean_token_accuracy": 0.2025853857398033,
      "num_tokens": 43951649.0,
      "step": 17340
    },
    {
      "entropy": 6.267815589904785,
      "epoch": 2.001731102135026,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004607654718974094,
      "loss": 5.6846,
      "mean_token_accuracy": 0.20141282975673674,
      "num_tokens": 43964140.0,
      "step": 17345
    },
    {
      "entropy": 6.261360454559326,
      "epoch": 2.0023081361800346,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004607421865908382,
      "loss": 5.7337,
      "mean_token_accuracy": 0.19530714452266693,
      "num_tokens": 43976189.0,
      "step": 17350
    },
    {
      "entropy": 6.3559654712677,
      "epoch": 2.0028851702250434,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004607188950369207,
      "loss": 5.7604,
      "mean_token_accuracy": 0.2000871330499649,
      "num_tokens": 43990920.0,
      "step": 17355
    },
    {
      "entropy": 6.272830533981323,
      "epoch": 2.003462204270052,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004606955972364405,
      "loss": 5.6324,
      "mean_token_accuracy": 0.20784922391176225,
      "num_tokens": 44003033.0,
      "step": 17360
    },
    {
      "entropy": 6.283031034469604,
      "epoch": 2.0040392383150607,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004606722931901812,
      "loss": 5.727,
      "mean_token_accuracy": 0.19508080333471298,
      "num_tokens": 44014731.0,
      "step": 17365
    },
    {
      "entropy": 6.189108562469483,
      "epoch": 2.004616272360069,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004606489828989264,
      "loss": 5.6384,
      "mean_token_accuracy": 0.1996154323220253,
      "num_tokens": 44026076.0,
      "step": 17370
    },
    {
      "entropy": 6.287853956222534,
      "epoch": 2.005193306405078,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004606256663634604,
      "loss": 5.7866,
      "mean_token_accuracy": 0.19353626817464828,
      "num_tokens": 44038244.0,
      "step": 17375
    },
    {
      "entropy": 6.3365833282470705,
      "epoch": 2.0057703404500864,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004606023435845672,
      "loss": 5.7009,
      "mean_token_accuracy": 0.20100168287754058,
      "num_tokens": 44051541.0,
      "step": 17380
    },
    {
      "entropy": 6.243878507614136,
      "epoch": 2.0063473744950953,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004605790145630314,
      "loss": 5.6578,
      "mean_token_accuracy": 0.20186127573251725,
      "num_tokens": 44064670.0,
      "step": 17385
    },
    {
      "entropy": 6.152621841430664,
      "epoch": 2.0069244085401037,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004605556792996377,
      "loss": 5.5767,
      "mean_token_accuracy": 0.20817508697509765,
      "num_tokens": 44077140.0,
      "step": 17390
    },
    {
      "entropy": 6.2445197105407715,
      "epoch": 2.0075014425851125,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004605323377951709,
      "loss": 5.7278,
      "mean_token_accuracy": 0.21122053861618043,
      "num_tokens": 44091641.0,
      "step": 17395
    },
    {
      "entropy": 6.22586989402771,
      "epoch": 2.008078476630121,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000460508990050416,
      "loss": 5.6972,
      "mean_token_accuracy": 0.20626843273639678,
      "num_tokens": 44104003.0,
      "step": 17400
    },
    {
      "entropy": 6.260751581192016,
      "epoch": 2.00865551067513,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004604856360661584,
      "loss": 5.748,
      "mean_token_accuracy": 0.1947092279791832,
      "num_tokens": 44116735.0,
      "step": 17405
    },
    {
      "entropy": 6.229497385025025,
      "epoch": 2.0092325447201387,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004604622758431835,
      "loss": 5.6968,
      "mean_token_accuracy": 0.2012970045208931,
      "num_tokens": 44129842.0,
      "step": 17410
    },
    {
      "entropy": 6.293037605285645,
      "epoch": 2.009809578765147,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046043890938227724,
      "loss": 5.6693,
      "mean_token_accuracy": 0.19783441126346588,
      "num_tokens": 44141800.0,
      "step": 17415
    },
    {
      "entropy": 6.1824850082397464,
      "epoch": 2.010386612810156,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046041553668422526,
      "loss": 5.6083,
      "mean_token_accuracy": 0.20880649238824844,
      "num_tokens": 44155705.0,
      "step": 17420
    },
    {
      "entropy": 6.234085273742676,
      "epoch": 2.0109636468551644,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046039215774981376,
      "loss": 5.6924,
      "mean_token_accuracy": 0.1968440905213356,
      "num_tokens": 44167780.0,
      "step": 17425
    },
    {
      "entropy": 6.2297539710998535,
      "epoch": 2.0115406809001732,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046036877257982917,
      "loss": 5.6747,
      "mean_token_accuracy": 0.20557420402765275,
      "num_tokens": 44179990.0,
      "step": 17430
    },
    {
      "entropy": 6.280623197555542,
      "epoch": 2.0121177149451817,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000460345381175058,
      "loss": 5.7069,
      "mean_token_accuracy": 0.20191171616315842,
      "num_tokens": 44192608.0,
      "step": 17435
    },
    {
      "entropy": 6.262015247344971,
      "epoch": 2.0126947489901905,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000460321983536287,
      "loss": 5.6907,
      "mean_token_accuracy": 0.20276592671871185,
      "num_tokens": 44206017.0,
      "step": 17440
    },
    {
      "entropy": 6.166565227508545,
      "epoch": 2.013271783035199,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046029857966430315,
      "loss": 5.6516,
      "mean_token_accuracy": 0.1965866968035698,
      "num_tokens": 44219028.0,
      "step": 17445
    },
    {
      "entropy": 6.296320533752441,
      "epoch": 2.013848817080208,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046027516955989364,
      "loss": 5.7126,
      "mean_token_accuracy": 0.20423431396484376,
      "num_tokens": 44230902.0,
      "step": 17450
    },
    {
      "entropy": 6.295812129974365,
      "epoch": 2.014425851125216,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046025175322384577,
      "loss": 5.7048,
      "mean_token_accuracy": 0.19583599716424943,
      "num_tokens": 44243108.0,
      "step": 17455
    },
    {
      "entropy": 6.285852098464966,
      "epoch": 2.015002885170225,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046022833065694727,
      "loss": 5.7902,
      "mean_token_accuracy": 0.19244285225868224,
      "num_tokens": 44255442.0,
      "step": 17460
    },
    {
      "entropy": 6.240371036529541,
      "epoch": 2.0155799192152335,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046020490185998575,
      "loss": 5.6765,
      "mean_token_accuracy": 0.20701410323381425,
      "num_tokens": 44269180.0,
      "step": 17465
    },
    {
      "entropy": 6.345157480239868,
      "epoch": 2.0161569532602424,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004601814668337495,
      "loss": 5.7749,
      "mean_token_accuracy": 0.19330511540174483,
      "num_tokens": 44280905.0,
      "step": 17470
    },
    {
      "entropy": 6.3587547779083256,
      "epoch": 2.016733987305251,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046015802557902654,
      "loss": 5.7463,
      "mean_token_accuracy": 0.191445192694664,
      "num_tokens": 44292672.0,
      "step": 17475
    },
    {
      "entropy": 6.202437829971314,
      "epoch": 2.0173110213502596,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046013457809660537,
      "loss": 5.5716,
      "mean_token_accuracy": 0.2128385618329048,
      "num_tokens": 44305605.0,
      "step": 17480
    },
    {
      "entropy": 6.276415586471558,
      "epoch": 2.0178880553952685,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046011112438727454,
      "loss": 5.7422,
      "mean_token_accuracy": 0.19434767216444016,
      "num_tokens": 44318792.0,
      "step": 17485
    },
    {
      "entropy": 6.2118316173553465,
      "epoch": 2.018465089440277,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004600876644518231,
      "loss": 5.6889,
      "mean_token_accuracy": 0.20440880209207535,
      "num_tokens": 44331878.0,
      "step": 17490
    },
    {
      "entropy": 6.321443748474121,
      "epoch": 2.0190421234852858,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046006419829104,
      "loss": 5.8271,
      "mean_token_accuracy": 0.1947207272052765,
      "num_tokens": 44345305.0,
      "step": 17495
    },
    {
      "entropy": 6.259106302261353,
      "epoch": 2.019619157530294,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004600407259057145,
      "loss": 5.6333,
      "mean_token_accuracy": 0.19805167615413666,
      "num_tokens": 44358512.0,
      "step": 17500
    },
    {
      "entropy": 6.2571320056915285,
      "epoch": 2.020196191575303,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004600172472966361,
      "loss": 5.691,
      "mean_token_accuracy": 0.2088336616754532,
      "num_tokens": 44371139.0,
      "step": 17505
    },
    {
      "entropy": 6.275822544097901,
      "epoch": 2.0207732256203115,
      "grad_norm": 0.890625,
      "learning_rate": 0.00045999376246459446,
      "loss": 5.7424,
      "mean_token_accuracy": 0.1876313179731369,
      "num_tokens": 44383984.0,
      "step": 17510
    },
    {
      "entropy": 6.317080020904541,
      "epoch": 2.0213502596653203,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004599702714103795,
      "loss": 5.7332,
      "mean_token_accuracy": 0.20072898864746094,
      "num_tokens": 44396481.0,
      "step": 17515
    },
    {
      "entropy": 6.271301698684693,
      "epoch": 2.0219272937103288,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004599467741347814,
      "loss": 5.7403,
      "mean_token_accuracy": 0.20137519985437394,
      "num_tokens": 44410843.0,
      "step": 17520
    },
    {
      "entropy": 6.294673204421997,
      "epoch": 2.0225043277553376,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004599232706385904,
      "loss": 5.7404,
      "mean_token_accuracy": 0.19582299292087554,
      "num_tokens": 44423797.0,
      "step": 17525
    },
    {
      "entropy": 6.261082506179809,
      "epoch": 2.023081361800346,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000459899760922597,
      "loss": 5.7253,
      "mean_token_accuracy": 0.19836447536945342,
      "num_tokens": 44436408.0,
      "step": 17530
    },
    {
      "entropy": 6.229238510131836,
      "epoch": 2.023658395845355,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004598762449875921,
      "loss": 5.5688,
      "mean_token_accuracy": 0.21689133942127228,
      "num_tokens": 44449145.0,
      "step": 17535
    },
    {
      "entropy": 6.253745603561401,
      "epoch": 2.0242354298903633,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004598527228343665,
      "loss": 5.6634,
      "mean_token_accuracy": 0.20183916836977006,
      "num_tokens": 44462945.0,
      "step": 17540
    },
    {
      "entropy": 6.203923940658569,
      "epoch": 2.024812463935372,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004598291944637112,
      "loss": 5.6874,
      "mean_token_accuracy": 0.19929226785898208,
      "num_tokens": 44474478.0,
      "step": 17545
    },
    {
      "entropy": 6.130711555480957,
      "epoch": 2.025389497980381,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045980565987641797,
      "loss": 5.5295,
      "mean_token_accuracy": 0.2191875919699669,
      "num_tokens": 44488418.0,
      "step": 17550
    },
    {
      "entropy": 6.316514348983764,
      "epoch": 2.0259665320253895,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045978211907327804,
      "loss": 5.6901,
      "mean_token_accuracy": 0.19926541894674302,
      "num_tokens": 44501424.0,
      "step": 17555
    },
    {
      "entropy": 6.292569828033447,
      "epoch": 2.0265435660703983,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004597585720550834,
      "loss": 5.7869,
      "mean_token_accuracy": 0.19016227424144744,
      "num_tokens": 44513137.0,
      "step": 17560
    },
    {
      "entropy": 6.2519642353057865,
      "epoch": 2.0271206001154067,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000459735018822626,
      "loss": 5.6344,
      "mean_token_accuracy": 0.20431955754756928,
      "num_tokens": 44525195.0,
      "step": 17565
    },
    {
      "entropy": 6.279894256591797,
      "epoch": 2.0276976341604156,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045971145937669794,
      "loss": 5.7923,
      "mean_token_accuracy": 0.19178905189037324,
      "num_tokens": 44537599.0,
      "step": 17570
    },
    {
      "entropy": 6.284214019775391,
      "epoch": 2.028274668205424,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004596878937180917,
      "loss": 5.6913,
      "mean_token_accuracy": 0.19432979971170425,
      "num_tokens": 44549384.0,
      "step": 17575
    },
    {
      "entropy": 6.2086451053619385,
      "epoch": 2.028851702250433,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004596643218475999,
      "loss": 5.6273,
      "mean_token_accuracy": 0.20383896678686142,
      "num_tokens": 44561162.0,
      "step": 17580
    },
    {
      "entropy": 6.25622034072876,
      "epoch": 2.0294287362954413,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00045964074376601534,
      "loss": 5.7043,
      "mean_token_accuracy": 0.20383191406726836,
      "num_tokens": 44576047.0,
      "step": 17585
    },
    {
      "entropy": 6.24154314994812,
      "epoch": 2.03000577034045,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045961715947413106,
      "loss": 5.7048,
      "mean_token_accuracy": 0.2030588984489441,
      "num_tokens": 44589506.0,
      "step": 17590
    },
    {
      "entropy": 6.24237551689148,
      "epoch": 2.0305828043854586,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004595935689727404,
      "loss": 5.6528,
      "mean_token_accuracy": 0.20691690295934678,
      "num_tokens": 44604483.0,
      "step": 17595
    },
    {
      "entropy": 6.257435989379883,
      "epoch": 2.0311598384304674,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004595699722626367,
      "loss": 5.6843,
      "mean_token_accuracy": 0.20579712241888046,
      "num_tokens": 44617696.0,
      "step": 17600
    },
    {
      "entropy": 6.165670394897461,
      "epoch": 2.031736872475476,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045954636934461367,
      "loss": 5.5716,
      "mean_token_accuracy": 0.216947278380394,
      "num_tokens": 44631482.0,
      "step": 17605
    },
    {
      "entropy": 6.1732042789459225,
      "epoch": 2.0323139065204847,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004595227602194652,
      "loss": 5.6347,
      "mean_token_accuracy": 0.21314602941274643,
      "num_tokens": 44645544.0,
      "step": 17610
    },
    {
      "entropy": 6.202894639968872,
      "epoch": 2.0328909405654936,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004594991448879853,
      "loss": 5.674,
      "mean_token_accuracy": 0.19939693063497543,
      "num_tokens": 44658869.0,
      "step": 17615
    },
    {
      "entropy": 6.290826225280762,
      "epoch": 2.033467974610502,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004594755233509683,
      "loss": 5.7555,
      "mean_token_accuracy": 0.19476155191659927,
      "num_tokens": 44670507.0,
      "step": 17620
    },
    {
      "entropy": 6.199102783203125,
      "epoch": 2.034045008655511,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045945189560920875,
      "loss": 5.6087,
      "mean_token_accuracy": 0.2108069032430649,
      "num_tokens": 44684063.0,
      "step": 17625
    },
    {
      "entropy": 6.15113582611084,
      "epoch": 2.0346220427005193,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004594282616635013,
      "loss": 5.495,
      "mean_token_accuracy": 0.2290724828839302,
      "num_tokens": 44698818.0,
      "step": 17630
    },
    {
      "entropy": 6.188174867630005,
      "epoch": 2.035199076745528,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004594046215146409,
      "loss": 5.6884,
      "mean_token_accuracy": 0.2025702878832817,
      "num_tokens": 44711401.0,
      "step": 17635
    },
    {
      "entropy": 6.29749608039856,
      "epoch": 2.0357761107905366,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045938097516342257,
      "loss": 5.6866,
      "mean_token_accuracy": 0.20159076005220414,
      "num_tokens": 44723141.0,
      "step": 17640
    },
    {
      "entropy": 6.370982265472412,
      "epoch": 2.0363531448355454,
      "grad_norm": 0.921875,
      "learning_rate": 0.00045935732261064184,
      "loss": 5.7691,
      "mean_token_accuracy": 0.1923420622944832,
      "num_tokens": 44736180.0,
      "step": 17645
    },
    {
      "entropy": 6.182464170455932,
      "epoch": 2.036930178880554,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00045933366385709405,
      "loss": 5.6843,
      "mean_token_accuracy": 0.19790453165769578,
      "num_tokens": 44749453.0,
      "step": 17650
    },
    {
      "entropy": 6.245594072341919,
      "epoch": 2.0375072129255627,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004593099989035751,
      "loss": 5.7212,
      "mean_token_accuracy": 0.19962759912014008,
      "num_tokens": 44762609.0,
      "step": 17655
    },
    {
      "entropy": 6.23505973815918,
      "epoch": 2.038084246970571,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004592863277508809,
      "loss": 5.6025,
      "mean_token_accuracy": 0.20563398599624633,
      "num_tokens": 44776239.0,
      "step": 17660
    },
    {
      "entropy": 6.171493864059448,
      "epoch": 2.03866128101558,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004592626503998076,
      "loss": 5.5854,
      "mean_token_accuracy": 0.208074751496315,
      "num_tokens": 44788979.0,
      "step": 17665
    },
    {
      "entropy": 6.291809034347534,
      "epoch": 2.0392383150605884,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004592389668511515,
      "loss": 5.7705,
      "mean_token_accuracy": 0.19543236643075942,
      "num_tokens": 44801715.0,
      "step": 17670
    },
    {
      "entropy": 6.213357019424438,
      "epoch": 2.0398153491055973,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004592152771057093,
      "loss": 5.7012,
      "mean_token_accuracy": 0.20022368431091309,
      "num_tokens": 44815004.0,
      "step": 17675
    },
    {
      "entropy": 6.324912071228027,
      "epoch": 2.0403923831506057,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00045919158116427785,
      "loss": 5.749,
      "mean_token_accuracy": 0.20140667259693146,
      "num_tokens": 44827933.0,
      "step": 17680
    },
    {
      "entropy": 6.220606565475464,
      "epoch": 2.0409694171956145,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045916787902765396,
      "loss": 5.6796,
      "mean_token_accuracy": 0.20727723687887192,
      "num_tokens": 44840967.0,
      "step": 17685
    },
    {
      "entropy": 6.221808004379272,
      "epoch": 2.0415464512406234,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004591441706966349,
      "loss": 5.7262,
      "mean_token_accuracy": 0.19526759386062623,
      "num_tokens": 44855456.0,
      "step": 17690
    },
    {
      "entropy": 6.246689701080323,
      "epoch": 2.042123485285632,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004591204561720183,
      "loss": 5.5924,
      "mean_token_accuracy": 0.2040194973349571,
      "num_tokens": 44867607.0,
      "step": 17695
    },
    {
      "entropy": 6.2509807586669925,
      "epoch": 2.0427005193306407,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004590967354546015,
      "loss": 5.6947,
      "mean_token_accuracy": 0.19814493209123613,
      "num_tokens": 44880505.0,
      "step": 17700
    },
    {
      "entropy": 6.259732437133789,
      "epoch": 2.043277553375649,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004590730085451824,
      "loss": 5.6982,
      "mean_token_accuracy": 0.2000792771577835,
      "num_tokens": 44894007.0,
      "step": 17705
    },
    {
      "entropy": 6.251287984848022,
      "epoch": 2.043854587420658,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00045904927544455914,
      "loss": 5.7258,
      "mean_token_accuracy": 0.19729716032743455,
      "num_tokens": 44907710.0,
      "step": 17710
    },
    {
      "entropy": 6.358564853668213,
      "epoch": 2.0444316214656664,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045902553615353005,
      "loss": 5.7577,
      "mean_token_accuracy": 0.19966264218091964,
      "num_tokens": 44921485.0,
      "step": 17715
    },
    {
      "entropy": 6.28543791770935,
      "epoch": 2.0450086555106752,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004590017906728933,
      "loss": 5.7388,
      "mean_token_accuracy": 0.20060048252344131,
      "num_tokens": 44934266.0,
      "step": 17720
    },
    {
      "entropy": 6.293681669235229,
      "epoch": 2.0455856895556837,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045897803900344774,
      "loss": 5.7742,
      "mean_token_accuracy": 0.190099273622036,
      "num_tokens": 44946828.0,
      "step": 17725
    },
    {
      "entropy": 6.255872583389282,
      "epoch": 2.0461627236006925,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004589542811459923,
      "loss": 5.6265,
      "mean_token_accuracy": 0.21555765420198442,
      "num_tokens": 44959381.0,
      "step": 17730
    },
    {
      "entropy": 6.212204885482788,
      "epoch": 2.046739757645701,
      "grad_norm": 1.0,
      "learning_rate": 0.0004589305171013259,
      "loss": 5.6714,
      "mean_token_accuracy": 0.20380066484212875,
      "num_tokens": 44973232.0,
      "step": 17735
    },
    {
      "entropy": 6.312522315979004,
      "epoch": 2.04731679169071,
      "grad_norm": 0.9375,
      "learning_rate": 0.000458906746870248,
      "loss": 5.7444,
      "mean_token_accuracy": 0.19396522045135497,
      "num_tokens": 44987007.0,
      "step": 17740
    },
    {
      "entropy": 6.231363010406494,
      "epoch": 2.047893825735718,
      "grad_norm": 0.921875,
      "learning_rate": 0.000458882970453558,
      "loss": 5.6035,
      "mean_token_accuracy": 0.20363103598356247,
      "num_tokens": 44999358.0,
      "step": 17745
    },
    {
      "entropy": 6.291912031173706,
      "epoch": 2.048470859780727,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004588591878520556,
      "loss": 5.7011,
      "mean_token_accuracy": 0.19990004301071168,
      "num_tokens": 45011737.0,
      "step": 17750
    },
    {
      "entropy": 6.269243764877319,
      "epoch": 2.049047893825736,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004588353990665407,
      "loss": 5.7271,
      "mean_token_accuracy": 0.1984774187207222,
      "num_tokens": 45024305.0,
      "step": 17755
    },
    {
      "entropy": 6.261275339126587,
      "epoch": 2.0496249278707444,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004588116040978136,
      "loss": 5.7289,
      "mean_token_accuracy": 0.19115320295095445,
      "num_tokens": 45036208.0,
      "step": 17760
    },
    {
      "entropy": 6.26155834197998,
      "epoch": 2.050201961915753,
      "grad_norm": 0.921875,
      "learning_rate": 0.00045878780294667437,
      "loss": 5.6959,
      "mean_token_accuracy": 0.1992887869477272,
      "num_tokens": 45048106.0,
      "step": 17765
    },
    {
      "entropy": 6.2821893215179445,
      "epoch": 2.0507789959607616,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004587639956139237,
      "loss": 5.7027,
      "mean_token_accuracy": 0.1980034351348877,
      "num_tokens": 45061938.0,
      "step": 17770
    },
    {
      "entropy": 6.329940938949585,
      "epoch": 2.0513560300057705,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004587401821003624,
      "loss": 5.7865,
      "mean_token_accuracy": 0.19354816377162934,
      "num_tokens": 45074534.0,
      "step": 17775
    },
    {
      "entropy": 6.259947299957275,
      "epoch": 2.051933064050779,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045871636240679133,
      "loss": 5.7578,
      "mean_token_accuracy": 0.1910829320549965,
      "num_tokens": 45086332.0,
      "step": 17780
    },
    {
      "entropy": 6.2283759117126465,
      "epoch": 2.0525100980957878,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004586925365340117,
      "loss": 5.6788,
      "mean_token_accuracy": 0.19686616212129593,
      "num_tokens": 45098559.0,
      "step": 17785
    },
    {
      "entropy": 6.305904865264893,
      "epoch": 2.053087132140796,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004586687044828247,
      "loss": 5.683,
      "mean_token_accuracy": 0.200490702688694,
      "num_tokens": 45109742.0,
      "step": 17790
    },
    {
      "entropy": 6.24164776802063,
      "epoch": 2.053664166185805,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004586448662540322,
      "loss": 5.7321,
      "mean_token_accuracy": 0.19761016517877578,
      "num_tokens": 45122438.0,
      "step": 17795
    },
    {
      "entropy": 6.2409178733825685,
      "epoch": 2.0542412002308135,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004586210218484358,
      "loss": 5.6764,
      "mean_token_accuracy": 0.20628942251205445,
      "num_tokens": 45136499.0,
      "step": 17800
    },
    {
      "entropy": 6.276852989196778,
      "epoch": 2.0548182342758223,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045859717126683745,
      "loss": 5.734,
      "mean_token_accuracy": 0.19557081609964372,
      "num_tokens": 45149313.0,
      "step": 17805
    },
    {
      "entropy": 6.236595916748047,
      "epoch": 2.0553952683208307,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045857331451003953,
      "loss": 5.7414,
      "mean_token_accuracy": 0.19966294318437577,
      "num_tokens": 45162258.0,
      "step": 17810
    },
    {
      "entropy": 6.279610729217529,
      "epoch": 2.0559723023658396,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045854945157884435,
      "loss": 5.6731,
      "mean_token_accuracy": 0.2031253084540367,
      "num_tokens": 45174868.0,
      "step": 17815
    },
    {
      "entropy": 6.273764562606812,
      "epoch": 2.056549336410848,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045852558247405455,
      "loss": 5.7434,
      "mean_token_accuracy": 0.19454240798950195,
      "num_tokens": 45185743.0,
      "step": 17820
    },
    {
      "entropy": 6.17339825630188,
      "epoch": 2.057126370455857,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045850170719647287,
      "loss": 5.6332,
      "mean_token_accuracy": 0.20456925183534622,
      "num_tokens": 45197950.0,
      "step": 17825
    },
    {
      "entropy": 6.200180864334106,
      "epoch": 2.0577034045008658,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00045847782574690254,
      "loss": 5.6772,
      "mean_token_accuracy": 0.20495675057172774,
      "num_tokens": 45210721.0,
      "step": 17830
    },
    {
      "entropy": 6.3373496532440186,
      "epoch": 2.058280438545874,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045845393812614664,
      "loss": 5.7409,
      "mean_token_accuracy": 0.20105002224445342,
      "num_tokens": 45222896.0,
      "step": 17835
    },
    {
      "entropy": 6.216390705108642,
      "epoch": 2.058857472590883,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004584300443350086,
      "loss": 5.5883,
      "mean_token_accuracy": 0.20855611711740493,
      "num_tokens": 45235787.0,
      "step": 17840
    },
    {
      "entropy": 6.205613040924073,
      "epoch": 2.0594345066358914,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004584061443742922,
      "loss": 5.617,
      "mean_token_accuracy": 0.19994848519563674,
      "num_tokens": 45247675.0,
      "step": 17845
    },
    {
      "entropy": 6.26388897895813,
      "epoch": 2.0600115406809003,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045838223824480124,
      "loss": 5.695,
      "mean_token_accuracy": 0.2007700815796852,
      "num_tokens": 45259786.0,
      "step": 17850
    },
    {
      "entropy": 6.2275309562683105,
      "epoch": 2.0605885747259087,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004583583259473397,
      "loss": 5.7209,
      "mean_token_accuracy": 0.19771451950073243,
      "num_tokens": 45273069.0,
      "step": 17855
    },
    {
      "entropy": 6.336365127563477,
      "epoch": 2.0611656087709176,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045833440748271203,
      "loss": 5.7938,
      "mean_token_accuracy": 0.19805472940206528,
      "num_tokens": 45285593.0,
      "step": 17860
    },
    {
      "entropy": 6.179568910598755,
      "epoch": 2.061742642815926,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045831048285172266,
      "loss": 5.6655,
      "mean_token_accuracy": 0.1977894425392151,
      "num_tokens": 45297938.0,
      "step": 17865
    },
    {
      "entropy": 6.15509672164917,
      "epoch": 2.062319676860935,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004582865520551762,
      "loss": 5.5354,
      "mean_token_accuracy": 0.21566729843616486,
      "num_tokens": 45311149.0,
      "step": 17870
    },
    {
      "entropy": 6.2976806640625,
      "epoch": 2.0628967109059433,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045826261509387755,
      "loss": 5.7415,
      "mean_token_accuracy": 0.1998835876584053,
      "num_tokens": 45324169.0,
      "step": 17875
    },
    {
      "entropy": 6.271880674362182,
      "epoch": 2.063473744950952,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045823867196863197,
      "loss": 5.7528,
      "mean_token_accuracy": 0.20079977214336395,
      "num_tokens": 45336724.0,
      "step": 17880
    },
    {
      "entropy": 6.29028959274292,
      "epoch": 2.0640507789959606,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004582147226802446,
      "loss": 5.65,
      "mean_token_accuracy": 0.20694242417812347,
      "num_tokens": 45349045.0,
      "step": 17885
    },
    {
      "entropy": 6.254119920730591,
      "epoch": 2.0646278130409694,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004581907672295211,
      "loss": 5.7369,
      "mean_token_accuracy": 0.19902952313423156,
      "num_tokens": 45360732.0,
      "step": 17890
    },
    {
      "entropy": 6.220662784576416,
      "epoch": 2.065204847085978,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045816680561726716,
      "loss": 5.7472,
      "mean_token_accuracy": 0.18771279901266097,
      "num_tokens": 45372538.0,
      "step": 17895
    },
    {
      "entropy": 6.2814429759979244,
      "epoch": 2.0657818811309867,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004581428378442886,
      "loss": 5.6965,
      "mean_token_accuracy": 0.19907506108283995,
      "num_tokens": 45385240.0,
      "step": 17900
    },
    {
      "entropy": 6.256937837600708,
      "epoch": 2.0663589151759956,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00045811886391139173,
      "loss": 5.7163,
      "mean_token_accuracy": 0.19756720811128617,
      "num_tokens": 45398863.0,
      "step": 17905
    },
    {
      "entropy": 6.265989923477173,
      "epoch": 2.066935949221004,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045809488381938284,
      "loss": 5.7833,
      "mean_token_accuracy": 0.18860110491514206,
      "num_tokens": 45412870.0,
      "step": 17910
    },
    {
      "entropy": 6.281545639038086,
      "epoch": 2.067512983266013,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004580708975690684,
      "loss": 5.6418,
      "mean_token_accuracy": 0.20227479040622712,
      "num_tokens": 45424314.0,
      "step": 17915
    },
    {
      "entropy": 6.2972393989562985,
      "epoch": 2.0680900173110213,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004580469051612554,
      "loss": 5.6795,
      "mean_token_accuracy": 0.20801949501037598,
      "num_tokens": 45437403.0,
      "step": 17920
    },
    {
      "entropy": 6.212697219848633,
      "epoch": 2.06866705135603,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045802290659675057,
      "loss": 5.667,
      "mean_token_accuracy": 0.19779070019721984,
      "num_tokens": 45451239.0,
      "step": 17925
    },
    {
      "entropy": 6.316596603393554,
      "epoch": 2.0692440854010385,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00045799890187636123,
      "loss": 5.7607,
      "mean_token_accuracy": 0.20335497707128525,
      "num_tokens": 45463157.0,
      "step": 17930
    },
    {
      "entropy": 6.254085540771484,
      "epoch": 2.0698211194460474,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045797489100089464,
      "loss": 5.6966,
      "mean_token_accuracy": 0.20371705442667007,
      "num_tokens": 45475774.0,
      "step": 17935
    },
    {
      "entropy": 6.334724950790405,
      "epoch": 2.070398153491056,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004579508739711585,
      "loss": 5.7973,
      "mean_token_accuracy": 0.19609928876161575,
      "num_tokens": 45489292.0,
      "step": 17940
    },
    {
      "entropy": 6.2905802726745605,
      "epoch": 2.0709751875360647,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045792685078796075,
      "loss": 5.7213,
      "mean_token_accuracy": 0.19594995975494384,
      "num_tokens": 45501021.0,
      "step": 17945
    },
    {
      "entropy": 6.250385046005249,
      "epoch": 2.071552221581073,
      "grad_norm": 1.03125,
      "learning_rate": 0.000457902821452109,
      "loss": 5.7107,
      "mean_token_accuracy": 0.20061303526163102,
      "num_tokens": 45513369.0,
      "step": 17950
    },
    {
      "entropy": 6.231299638748169,
      "epoch": 2.072129255626082,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045787878596441193,
      "loss": 5.6273,
      "mean_token_accuracy": 0.20198728144168854,
      "num_tokens": 45525077.0,
      "step": 17955
    },
    {
      "entropy": 6.403262901306152,
      "epoch": 2.0727062896710904,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004578547443256776,
      "loss": 5.7873,
      "mean_token_accuracy": 0.18789468258619307,
      "num_tokens": 45538008.0,
      "step": 17960
    },
    {
      "entropy": 6.286366891860962,
      "epoch": 2.0732833237160992,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004578306965367148,
      "loss": 5.6904,
      "mean_token_accuracy": 0.19396430104970933,
      "num_tokens": 45550797.0,
      "step": 17965
    },
    {
      "entropy": 6.282064914703369,
      "epoch": 2.073860357761108,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045780664259833235,
      "loss": 5.6846,
      "mean_token_accuracy": 0.1975371405482292,
      "num_tokens": 45564146.0,
      "step": 17970
    },
    {
      "entropy": 6.1940206527709964,
      "epoch": 2.0744373918061165,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00045778258251133924,
      "loss": 5.6654,
      "mean_token_accuracy": 0.20849392265081407,
      "num_tokens": 45577246.0,
      "step": 17975
    },
    {
      "entropy": 6.277230739593506,
      "epoch": 2.0750144258511254,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045775851627654474,
      "loss": 5.724,
      "mean_token_accuracy": 0.19564638286828995,
      "num_tokens": 45589042.0,
      "step": 17980
    },
    {
      "entropy": 6.277932405471802,
      "epoch": 2.075591459896134,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004577344438947584,
      "loss": 5.667,
      "mean_token_accuracy": 0.2090568095445633,
      "num_tokens": 45602472.0,
      "step": 17985
    },
    {
      "entropy": 6.400613594055176,
      "epoch": 2.0761684939411427,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045771036536678984,
      "loss": 5.7897,
      "mean_token_accuracy": 0.18762761056423188,
      "num_tokens": 45614208.0,
      "step": 17990
    },
    {
      "entropy": 6.15055103302002,
      "epoch": 2.076745527986151,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00045768628069344885,
      "loss": 5.6672,
      "mean_token_accuracy": 0.20891030579805375,
      "num_tokens": 45626396.0,
      "step": 17995
    },
    {
      "entropy": 6.227409791946411,
      "epoch": 2.07732256203116,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004576621898755455,
      "loss": 5.6688,
      "mean_token_accuracy": 0.1943758860230446,
      "num_tokens": 45640482.0,
      "step": 18000
    },
    {
      "epoch": 2.07732256203116,
      "eval_entropy": 6.0897077121007905,
      "eval_loss": 5.826183319091797,
      "eval_mean_token_accuracy": 0.20169892374564885,
      "eval_num_tokens": 45640482.0,
      "eval_runtime": 17.8044,
      "eval_samples_per_second": 1580.283,
      "eval_steps_per_second": 197.535,
      "step": 18000
    },
    {
      "entropy": 6.331846237182617,
      "epoch": 2.0778995960761684,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00045763809291389024,
      "loss": 5.8134,
      "mean_token_accuracy": 0.1936564639210701,
      "num_tokens": 45654414.0,
      "step": 18005
    },
    {
      "entropy": 6.157422399520874,
      "epoch": 2.0784766301211772,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004576139898092933,
      "loss": 5.616,
      "mean_token_accuracy": 0.20844219923019408,
      "num_tokens": 45667429.0,
      "step": 18010
    },
    {
      "entropy": 6.119392538070679,
      "epoch": 2.0790536641661856,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004575898805625657,
      "loss": 5.6429,
      "mean_token_accuracy": 0.20009539276361465,
      "num_tokens": 45680269.0,
      "step": 18015
    },
    {
      "entropy": 6.371662235260009,
      "epoch": 2.0796306982111945,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045756576517451804,
      "loss": 5.7295,
      "mean_token_accuracy": 0.20457745790481568,
      "num_tokens": 45693950.0,
      "step": 18020
    },
    {
      "entropy": 6.285341453552246,
      "epoch": 2.080207732256203,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045754164364596156,
      "loss": 5.6794,
      "mean_token_accuracy": 0.20406908094882964,
      "num_tokens": 45706656.0,
      "step": 18025
    },
    {
      "entropy": 6.233389234542846,
      "epoch": 2.080784766301212,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004575175159777076,
      "loss": 5.6219,
      "mean_token_accuracy": 0.20810570269823075,
      "num_tokens": 45718894.0,
      "step": 18030
    },
    {
      "entropy": 6.295356416702271,
      "epoch": 2.0813618003462206,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004574933821705676,
      "loss": 5.7296,
      "mean_token_accuracy": 0.19933123588562013,
      "num_tokens": 45731556.0,
      "step": 18035
    },
    {
      "entropy": 6.218157768249512,
      "epoch": 2.081938834391229,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004574692422253534,
      "loss": 5.6678,
      "mean_token_accuracy": 0.20097315758466722,
      "num_tokens": 45743618.0,
      "step": 18040
    },
    {
      "entropy": 6.222191858291626,
      "epoch": 2.082515868436238,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045744509614287687,
      "loss": 5.6925,
      "mean_token_accuracy": 0.2045382335782051,
      "num_tokens": 45757086.0,
      "step": 18045
    },
    {
      "entropy": 6.225840759277344,
      "epoch": 2.0830929024812463,
      "grad_norm": 1.125,
      "learning_rate": 0.0004574209439239501,
      "loss": 5.6582,
      "mean_token_accuracy": 0.2001233920454979,
      "num_tokens": 45770153.0,
      "step": 18050
    },
    {
      "entropy": 6.27244381904602,
      "epoch": 2.083669936526255,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045739678556938563,
      "loss": 5.6557,
      "mean_token_accuracy": 0.2014635756611824,
      "num_tokens": 45783356.0,
      "step": 18055
    },
    {
      "entropy": 6.2805821895599365,
      "epoch": 2.0842469705712636,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045737262107999575,
      "loss": 5.6809,
      "mean_token_accuracy": 0.2069031909108162,
      "num_tokens": 45795557.0,
      "step": 18060
    },
    {
      "entropy": 6.266357040405273,
      "epoch": 2.0848240046162725,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004573484504565934,
      "loss": 5.7455,
      "mean_token_accuracy": 0.20099477469921112,
      "num_tokens": 45808011.0,
      "step": 18065
    },
    {
      "entropy": 6.222839117050171,
      "epoch": 2.085401038661281,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004573242736999915,
      "loss": 5.6695,
      "mean_token_accuracy": 0.20446588844060898,
      "num_tokens": 45820514.0,
      "step": 18070
    },
    {
      "entropy": 6.270034885406494,
      "epoch": 2.0859780727062898,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00045730009081100314,
      "loss": 5.7396,
      "mean_token_accuracy": 0.19203182458877563,
      "num_tokens": 45833062.0,
      "step": 18075
    },
    {
      "entropy": 6.267688465118408,
      "epoch": 2.086555106751298,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045727590179044195,
      "loss": 5.6736,
      "mean_token_accuracy": 0.20693072825670242,
      "num_tokens": 45845723.0,
      "step": 18080
    },
    {
      "entropy": 6.301341438293457,
      "epoch": 2.087132140796307,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004572517066391211,
      "loss": 5.7538,
      "mean_token_accuracy": 0.2029941499233246,
      "num_tokens": 45857779.0,
      "step": 18085
    },
    {
      "entropy": 6.206268405914306,
      "epoch": 2.0877091748413155,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045722750535785484,
      "loss": 5.6995,
      "mean_token_accuracy": 0.20098061114549637,
      "num_tokens": 45871321.0,
      "step": 18090
    },
    {
      "entropy": 6.280726146697998,
      "epoch": 2.0882862088863243,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00045720329794745685,
      "loss": 5.7099,
      "mean_token_accuracy": 0.20572860389947892,
      "num_tokens": 45884048.0,
      "step": 18095
    },
    {
      "entropy": 6.277518796920776,
      "epoch": 2.0888632429313327,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004571790844087415,
      "loss": 5.7691,
      "mean_token_accuracy": 0.1944665029644966,
      "num_tokens": 45895986.0,
      "step": 18100
    },
    {
      "entropy": 6.310266637802124,
      "epoch": 2.0894402769763416,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004571548647425231,
      "loss": 5.7941,
      "mean_token_accuracy": 0.19742656201124192,
      "num_tokens": 45909120.0,
      "step": 18105
    },
    {
      "entropy": 6.251747703552246,
      "epoch": 2.0900173110213505,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004571306389496164,
      "loss": 5.6278,
      "mean_token_accuracy": 0.21161530166864395,
      "num_tokens": 45921164.0,
      "step": 18110
    },
    {
      "entropy": 6.156686496734619,
      "epoch": 2.090594345066359,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045710640703083594,
      "loss": 5.5586,
      "mean_token_accuracy": 0.2104140818119049,
      "num_tokens": 45933037.0,
      "step": 18115
    },
    {
      "entropy": 6.260765743255615,
      "epoch": 2.0911713791113677,
      "grad_norm": 0.875,
      "learning_rate": 0.00045708216898699707,
      "loss": 5.6803,
      "mean_token_accuracy": 0.2000722736120224,
      "num_tokens": 45945381.0,
      "step": 18120
    },
    {
      "entropy": 6.2144571304321286,
      "epoch": 2.091748413156376,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045705792481891483,
      "loss": 5.5671,
      "mean_token_accuracy": 0.2117965266108513,
      "num_tokens": 45959446.0,
      "step": 18125
    },
    {
      "entropy": 6.314937591552734,
      "epoch": 2.092325447201385,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00045703367452740477,
      "loss": 5.8216,
      "mean_token_accuracy": 0.19025691598653793,
      "num_tokens": 45971773.0,
      "step": 18130
    },
    {
      "entropy": 6.238446998596191,
      "epoch": 2.0929024812463934,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045700941811328247,
      "loss": 5.6701,
      "mean_token_accuracy": 0.1970654010772705,
      "num_tokens": 45983543.0,
      "step": 18135
    },
    {
      "entropy": 6.2596104621887205,
      "epoch": 2.0934795152914023,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00045698515557736374,
      "loss": 5.6492,
      "mean_token_accuracy": 0.20044853836297988,
      "num_tokens": 45995231.0,
      "step": 18140
    },
    {
      "entropy": 6.273843717575073,
      "epoch": 2.0940565493364107,
      "grad_norm": 0.90625,
      "learning_rate": 0.00045696088692046477,
      "loss": 5.7108,
      "mean_token_accuracy": 0.19593378454446791,
      "num_tokens": 46009414.0,
      "step": 18145
    },
    {
      "entropy": 6.278366613388061,
      "epoch": 2.0946335833814196,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045693661214340174,
      "loss": 5.684,
      "mean_token_accuracy": 0.19920516461133958,
      "num_tokens": 46020781.0,
      "step": 18150
    },
    {
      "entropy": 6.293325519561767,
      "epoch": 2.095210617426428,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045691233124699117,
      "loss": 5.7107,
      "mean_token_accuracy": 0.19906881600618362,
      "num_tokens": 46032479.0,
      "step": 18155
    },
    {
      "entropy": 6.201935338973999,
      "epoch": 2.095787651471437,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004568880442320497,
      "loss": 5.6714,
      "mean_token_accuracy": 0.2054503530263901,
      "num_tokens": 46044837.0,
      "step": 18160
    },
    {
      "entropy": 6.2733629703521725,
      "epoch": 2.0963646855164453,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045686375109939417,
      "loss": 5.7859,
      "mean_token_accuracy": 0.19492304623126983,
      "num_tokens": 46056194.0,
      "step": 18165
    },
    {
      "entropy": 6.23380823135376,
      "epoch": 2.096941719561454,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004568394518498417,
      "loss": 5.6609,
      "mean_token_accuracy": 0.20163753926753997,
      "num_tokens": 46069536.0,
      "step": 18170
    },
    {
      "entropy": 6.123551940917968,
      "epoch": 2.0975187536064626,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045681514648420965,
      "loss": 5.5757,
      "mean_token_accuracy": 0.21342228651046752,
      "num_tokens": 46082616.0,
      "step": 18175
    },
    {
      "entropy": 6.215941476821899,
      "epoch": 2.0980957876514714,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004567908350033154,
      "loss": 5.6719,
      "mean_token_accuracy": 0.19653585255146028,
      "num_tokens": 46094839.0,
      "step": 18180
    },
    {
      "entropy": 6.258538055419922,
      "epoch": 2.0986728216964803,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004567665174079767,
      "loss": 5.7458,
      "mean_token_accuracy": 0.1929170683026314,
      "num_tokens": 46106475.0,
      "step": 18185
    },
    {
      "entropy": 6.231928110122681,
      "epoch": 2.0992498557414887,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045674219369901153,
      "loss": 5.6436,
      "mean_token_accuracy": 0.20075293481349946,
      "num_tokens": 46118922.0,
      "step": 18190
    },
    {
      "entropy": 6.324683618545532,
      "epoch": 2.0998268897864976,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004567178638772379,
      "loss": 5.804,
      "mean_token_accuracy": 0.19216287285089492,
      "num_tokens": 46131848.0,
      "step": 18195
    },
    {
      "entropy": 6.206061458587646,
      "epoch": 2.100403923831506,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004566935279434742,
      "loss": 5.657,
      "mean_token_accuracy": 0.2077864482998848,
      "num_tokens": 46145130.0,
      "step": 18200
    },
    {
      "entropy": 6.215090370178222,
      "epoch": 2.100980957876515,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045666918589853895,
      "loss": 5.5926,
      "mean_token_accuracy": 0.20910869985818864,
      "num_tokens": 46160301.0,
      "step": 18205
    },
    {
      "entropy": 6.320216131210327,
      "epoch": 2.1015579919215233,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00045664483774325094,
      "loss": 5.7151,
      "mean_token_accuracy": 0.19806645363569259,
      "num_tokens": 46171650.0,
      "step": 18210
    },
    {
      "entropy": 6.288967704772949,
      "epoch": 2.102135025966532,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004566204834784289,
      "loss": 5.6948,
      "mean_token_accuracy": 0.2075858861207962,
      "num_tokens": 46183302.0,
      "step": 18215
    },
    {
      "entropy": 6.282423305511474,
      "epoch": 2.1027120600115405,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045659612310489225,
      "loss": 5.6944,
      "mean_token_accuracy": 0.19787320047616958,
      "num_tokens": 46195459.0,
      "step": 18220
    },
    {
      "entropy": 6.167770147323608,
      "epoch": 2.1032890940565494,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045657175662346014,
      "loss": 5.6158,
      "mean_token_accuracy": 0.20401471853256226,
      "num_tokens": 46207969.0,
      "step": 18225
    },
    {
      "entropy": 6.2769927978515625,
      "epoch": 2.103866128101558,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004565473840349522,
      "loss": 5.7407,
      "mean_token_accuracy": 0.19622113555669785,
      "num_tokens": 46219976.0,
      "step": 18230
    },
    {
      "entropy": 6.211724138259887,
      "epoch": 2.1044431621465667,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045652300534018816,
      "loss": 5.6433,
      "mean_token_accuracy": 0.20081702321767808,
      "num_tokens": 46232172.0,
      "step": 18235
    },
    {
      "entropy": 6.299686765670776,
      "epoch": 2.105020196191575,
      "grad_norm": 1.0625,
      "learning_rate": 0.000456498620539988,
      "loss": 5.787,
      "mean_token_accuracy": 0.19355374276638032,
      "num_tokens": 46244288.0,
      "step": 18240
    },
    {
      "entropy": 6.283264112472534,
      "epoch": 2.105597230236584,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004564742296351719,
      "loss": 5.7273,
      "mean_token_accuracy": 0.19863540828227996,
      "num_tokens": 46256464.0,
      "step": 18245
    },
    {
      "entropy": 6.344845628738403,
      "epoch": 2.106174264281593,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045644983262656014,
      "loss": 5.8087,
      "mean_token_accuracy": 0.19721813052892684,
      "num_tokens": 46270356.0,
      "step": 18250
    },
    {
      "entropy": 6.227012586593628,
      "epoch": 2.1067512983266012,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004564254295149735,
      "loss": 5.6783,
      "mean_token_accuracy": 0.20330015569925308,
      "num_tokens": 46282990.0,
      "step": 18255
    },
    {
      "entropy": 6.257531023025512,
      "epoch": 2.10732833237161,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045640102030123264,
      "loss": 5.6564,
      "mean_token_accuracy": 0.19855670481920243,
      "num_tokens": 46294665.0,
      "step": 18260
    },
    {
      "entropy": 6.246777105331421,
      "epoch": 2.1079053664166185,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045637660498615865,
      "loss": 5.6721,
      "mean_token_accuracy": 0.1997729331254959,
      "num_tokens": 46306950.0,
      "step": 18265
    },
    {
      "entropy": 6.318729400634766,
      "epoch": 2.1084824004616274,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004563521835705725,
      "loss": 5.8027,
      "mean_token_accuracy": 0.19635479301214218,
      "num_tokens": 46320475.0,
      "step": 18270
    },
    {
      "entropy": 6.303743505477906,
      "epoch": 2.109059434506636,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00045632775605529587,
      "loss": 5.7185,
      "mean_token_accuracy": 0.19810597151517867,
      "num_tokens": 46332624.0,
      "step": 18275
    },
    {
      "entropy": 6.269377613067627,
      "epoch": 2.1096364685516447,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004563033224411501,
      "loss": 5.6899,
      "mean_token_accuracy": 0.20409863740205764,
      "num_tokens": 46344257.0,
      "step": 18280
    },
    {
      "entropy": 6.213299512863159,
      "epoch": 2.110213502596653,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004562788827289572,
      "loss": 5.6706,
      "mean_token_accuracy": 0.1971700981259346,
      "num_tokens": 46356526.0,
      "step": 18285
    },
    {
      "entropy": 6.285840082168579,
      "epoch": 2.110790536641662,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045625443691953914,
      "loss": 5.7161,
      "mean_token_accuracy": 0.2019614666700363,
      "num_tokens": 46368439.0,
      "step": 18290
    },
    {
      "entropy": 6.265392827987671,
      "epoch": 2.1113675706866704,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004562299850137181,
      "loss": 5.7212,
      "mean_token_accuracy": 0.20125612765550613,
      "num_tokens": 46381627.0,
      "step": 18295
    },
    {
      "entropy": 6.250712490081787,
      "epoch": 2.111944604731679,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004562055270123166,
      "loss": 5.7071,
      "mean_token_accuracy": 0.19654446691274643,
      "num_tokens": 46393869.0,
      "step": 18300
    },
    {
      "entropy": 6.2484911441802975,
      "epoch": 2.1125216387766876,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045618106291615715,
      "loss": 5.6767,
      "mean_token_accuracy": 0.20133116543293,
      "num_tokens": 46405911.0,
      "step": 18305
    },
    {
      "entropy": 6.25935640335083,
      "epoch": 2.1130986728216965,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004561565927260627,
      "loss": 5.7414,
      "mean_token_accuracy": 0.19686522632837294,
      "num_tokens": 46418258.0,
      "step": 18310
    },
    {
      "entropy": 6.349966621398925,
      "epoch": 2.1136757068667054,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004561321164428561,
      "loss": 5.7997,
      "mean_token_accuracy": 0.19227427542209624,
      "num_tokens": 46431211.0,
      "step": 18315
    },
    {
      "entropy": 6.3598590850830075,
      "epoch": 2.114252740911714,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004561076340673609,
      "loss": 5.8386,
      "mean_token_accuracy": 0.18303107619285583,
      "num_tokens": 46445062.0,
      "step": 18320
    },
    {
      "entropy": 6.327579498291016,
      "epoch": 2.1148297749567226,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004560831456004003,
      "loss": 5.7968,
      "mean_token_accuracy": 0.19026097655296326,
      "num_tokens": 46457251.0,
      "step": 18325
    },
    {
      "entropy": 6.233110046386718,
      "epoch": 2.115406809001731,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004560586510427981,
      "loss": 5.6071,
      "mean_token_accuracy": 0.2059030994772911,
      "num_tokens": 46471973.0,
      "step": 18330
    },
    {
      "entropy": 6.2864134311676025,
      "epoch": 2.11598384304674,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004560341503953781,
      "loss": 5.6776,
      "mean_token_accuracy": 0.1983066439628601,
      "num_tokens": 46484363.0,
      "step": 18335
    },
    {
      "entropy": 6.277480840682983,
      "epoch": 2.1165608770917483,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004560096436589643,
      "loss": 5.6965,
      "mean_token_accuracy": 0.19985044449567796,
      "num_tokens": 46496033.0,
      "step": 18340
    },
    {
      "entropy": 6.237420654296875,
      "epoch": 2.117137911136757,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045598513083438115,
      "loss": 5.5873,
      "mean_token_accuracy": 0.2068473607301712,
      "num_tokens": 46507826.0,
      "step": 18345
    },
    {
      "entropy": 6.203905439376831,
      "epoch": 2.1177149451817656,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045596061192245297,
      "loss": 5.6887,
      "mean_token_accuracy": 0.2012255847454071,
      "num_tokens": 46519697.0,
      "step": 18350
    },
    {
      "entropy": 6.281008720397949,
      "epoch": 2.1182919792267745,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004559360869240045,
      "loss": 5.7172,
      "mean_token_accuracy": 0.19613824188709258,
      "num_tokens": 46532486.0,
      "step": 18355
    },
    {
      "entropy": 6.319827461242676,
      "epoch": 2.118869013271783,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004559115558398606,
      "loss": 5.7389,
      "mean_token_accuracy": 0.19992063492536544,
      "num_tokens": 46543985.0,
      "step": 18360
    },
    {
      "entropy": 6.261981344223022,
      "epoch": 2.1194460473167918,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004558870186708465,
      "loss": 5.6693,
      "mean_token_accuracy": 0.20575758814811707,
      "num_tokens": 46556326.0,
      "step": 18365
    },
    {
      "entropy": 6.249263191223145,
      "epoch": 2.1200230813618,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045586247541778724,
      "loss": 5.7492,
      "mean_token_accuracy": 0.19440328627824782,
      "num_tokens": 46567955.0,
      "step": 18370
    },
    {
      "entropy": 6.2992654800415036,
      "epoch": 2.120600115406809,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004558379260815085,
      "loss": 5.6621,
      "mean_token_accuracy": 0.19913120120763778,
      "num_tokens": 46580581.0,
      "step": 18375
    },
    {
      "entropy": 6.294353771209717,
      "epoch": 2.1211771494518175,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004558133706628359,
      "loss": 5.7208,
      "mean_token_accuracy": 0.19804624617099761,
      "num_tokens": 46592503.0,
      "step": 18380
    },
    {
      "entropy": 6.224205017089844,
      "epoch": 2.1217541834968263,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045578880916259554,
      "loss": 5.6122,
      "mean_token_accuracy": 0.2135900393128395,
      "num_tokens": 46605542.0,
      "step": 18385
    },
    {
      "entropy": 6.239010572433472,
      "epoch": 2.122331217541835,
      "grad_norm": 1.125,
      "learning_rate": 0.0004557642415816132,
      "loss": 5.6354,
      "mean_token_accuracy": 0.20222847610712053,
      "num_tokens": 46618826.0,
      "step": 18390
    },
    {
      "entropy": 6.246191692352295,
      "epoch": 2.1229082515868436,
      "grad_norm": 1.0,
      "learning_rate": 0.0004557396679207155,
      "loss": 5.6871,
      "mean_token_accuracy": 0.19225419461727142,
      "num_tokens": 46630295.0,
      "step": 18395
    },
    {
      "entropy": 6.2538454055786135,
      "epoch": 2.1234852856318525,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045571508818072887,
      "loss": 5.7018,
      "mean_token_accuracy": 0.1993197590112686,
      "num_tokens": 46642605.0,
      "step": 18400
    },
    {
      "entropy": 6.328603744506836,
      "epoch": 2.124062319676861,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004556905023624799,
      "loss": 5.7912,
      "mean_token_accuracy": 0.19132236391305923,
      "num_tokens": 46656411.0,
      "step": 18405
    },
    {
      "entropy": 6.329950571060181,
      "epoch": 2.1246393537218697,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045566591046679577,
      "loss": 5.7552,
      "mean_token_accuracy": 0.1973779872059822,
      "num_tokens": 46668231.0,
      "step": 18410
    },
    {
      "entropy": 6.318213891983032,
      "epoch": 2.125216387766878,
      "grad_norm": 0.921875,
      "learning_rate": 0.00045564131249450343,
      "loss": 5.7601,
      "mean_token_accuracy": 0.19535335004329682,
      "num_tokens": 46681448.0,
      "step": 18415
    },
    {
      "entropy": 6.256973600387573,
      "epoch": 2.125793421811887,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004556167084464302,
      "loss": 5.7004,
      "mean_token_accuracy": 0.19549612253904342,
      "num_tokens": 46695043.0,
      "step": 18420
    },
    {
      "entropy": 6.32691011428833,
      "epoch": 2.1263704558568954,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004555920983234038,
      "loss": 5.725,
      "mean_token_accuracy": 0.19680947810411453,
      "num_tokens": 46707099.0,
      "step": 18425
    },
    {
      "entropy": 6.210586261749268,
      "epoch": 2.1269474899019043,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045556748212625183,
      "loss": 5.6289,
      "mean_token_accuracy": 0.20531365126371384,
      "num_tokens": 46720913.0,
      "step": 18430
    },
    {
      "entropy": 6.284554433822632,
      "epoch": 2.1275245239469127,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004555428598558023,
      "loss": 5.715,
      "mean_token_accuracy": 0.1985595241189003,
      "num_tokens": 46733405.0,
      "step": 18435
    },
    {
      "entropy": 6.321535968780518,
      "epoch": 2.1281015579919216,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004555182315128833,
      "loss": 5.7635,
      "mean_token_accuracy": 0.19226474016904832,
      "num_tokens": 46745588.0,
      "step": 18440
    },
    {
      "entropy": 6.332419300079346,
      "epoch": 2.12867859203693,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004554935970983234,
      "loss": 5.8084,
      "mean_token_accuracy": 0.18648878633975982,
      "num_tokens": 46758091.0,
      "step": 18445
    },
    {
      "entropy": 6.3149045467376705,
      "epoch": 2.129255626081939,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004554689566129509,
      "loss": 5.7576,
      "mean_token_accuracy": 0.19343363493680954,
      "num_tokens": 46770966.0,
      "step": 18450
    },
    {
      "entropy": 6.251780557632446,
      "epoch": 2.1298326601269473,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045544431005759467,
      "loss": 5.7207,
      "mean_token_accuracy": 0.19936630129814148,
      "num_tokens": 46783947.0,
      "step": 18455
    },
    {
      "entropy": 6.188159322738647,
      "epoch": 2.130409694171956,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00045541965743308376,
      "loss": 5.5755,
      "mean_token_accuracy": 0.20447495430707932,
      "num_tokens": 46796985.0,
      "step": 18460
    },
    {
      "entropy": 6.266006422042847,
      "epoch": 2.130986728216965,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004553949987402473,
      "loss": 5.66,
      "mean_token_accuracy": 0.20018266886472702,
      "num_tokens": 46809397.0,
      "step": 18465
    },
    {
      "entropy": 6.224253225326538,
      "epoch": 2.1315637622619734,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004553703339799146,
      "loss": 5.6545,
      "mean_token_accuracy": 0.21134542375802995,
      "num_tokens": 46822336.0,
      "step": 18470
    },
    {
      "entropy": 6.21381630897522,
      "epoch": 2.1321407963069823,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004553456631529154,
      "loss": 5.6733,
      "mean_token_accuracy": 0.20699447840452195,
      "num_tokens": 46835048.0,
      "step": 18475
    },
    {
      "entropy": 6.275369358062744,
      "epoch": 2.1327178303519907,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004553209862600794,
      "loss": 5.7869,
      "mean_token_accuracy": 0.19285908341407776,
      "num_tokens": 46847295.0,
      "step": 18480
    },
    {
      "entropy": 6.244692802429199,
      "epoch": 2.1332948643969996,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004552963033022365,
      "loss": 5.645,
      "mean_token_accuracy": 0.20026969015598298,
      "num_tokens": 46859373.0,
      "step": 18485
    },
    {
      "entropy": 6.27680230140686,
      "epoch": 2.133871898442008,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045527161428021706,
      "loss": 5.6406,
      "mean_token_accuracy": 0.20426035225391387,
      "num_tokens": 46871307.0,
      "step": 18490
    },
    {
      "entropy": 6.2386716365814205,
      "epoch": 2.134448932487017,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004552469191948514,
      "loss": 5.655,
      "mean_token_accuracy": 0.2093985363841057,
      "num_tokens": 46883577.0,
      "step": 18495
    },
    {
      "entropy": 6.2179876327514645,
      "epoch": 2.1350259665320253,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004552222180469702,
      "loss": 5.6721,
      "mean_token_accuracy": 0.20348166525363923,
      "num_tokens": 46896413.0,
      "step": 18500
    },
    {
      "entropy": 6.278118705749511,
      "epoch": 2.135603000577034,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045519751083740413,
      "loss": 5.7186,
      "mean_token_accuracy": 0.19281302392482758,
      "num_tokens": 46907225.0,
      "step": 18505
    },
    {
      "entropy": 6.24754900932312,
      "epoch": 2.1361800346220425,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045517279756698435,
      "loss": 5.6125,
      "mean_token_accuracy": 0.20307822227478028,
      "num_tokens": 46919852.0,
      "step": 18510
    },
    {
      "entropy": 6.303329086303711,
      "epoch": 2.1367570686670514,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045514807823654203,
      "loss": 5.7676,
      "mean_token_accuracy": 0.19219275265932084,
      "num_tokens": 46932803.0,
      "step": 18515
    },
    {
      "entropy": 6.235187530517578,
      "epoch": 2.13733410271206,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004551233528469086,
      "loss": 5.7275,
      "mean_token_accuracy": 0.19320896863937378,
      "num_tokens": 46945340.0,
      "step": 18520
    },
    {
      "entropy": 6.2784827709197994,
      "epoch": 2.1379111367570687,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045509862139891553,
      "loss": 5.7399,
      "mean_token_accuracy": 0.19213842749595642,
      "num_tokens": 46957837.0,
      "step": 18525
    },
    {
      "entropy": 6.289545774459839,
      "epoch": 2.138488170802077,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00045507388389339496,
      "loss": 5.6882,
      "mean_token_accuracy": 0.20365771204233168,
      "num_tokens": 46971106.0,
      "step": 18530
    },
    {
      "entropy": 6.2903299808502195,
      "epoch": 2.139065204847086,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045504914033117866,
      "loss": 5.7183,
      "mean_token_accuracy": 0.1933821439743042,
      "num_tokens": 46982897.0,
      "step": 18535
    },
    {
      "entropy": 6.2963423252105715,
      "epoch": 2.139642238892095,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045502439071309897,
      "loss": 5.7418,
      "mean_token_accuracy": 0.19283353984355928,
      "num_tokens": 46995812.0,
      "step": 18540
    },
    {
      "entropy": 6.357389354705811,
      "epoch": 2.1402192729371032,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045499963503998835,
      "loss": 5.7585,
      "mean_token_accuracy": 0.1926417037844658,
      "num_tokens": 47008092.0,
      "step": 18545
    },
    {
      "entropy": 6.1804078102111815,
      "epoch": 2.140796306982112,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004549748733126794,
      "loss": 5.5459,
      "mean_token_accuracy": 0.20961364656686782,
      "num_tokens": 47020870.0,
      "step": 18550
    },
    {
      "entropy": 6.165998220443726,
      "epoch": 2.1413733410271205,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000454950105532005,
      "loss": 5.6052,
      "mean_token_accuracy": 0.20679046660661698,
      "num_tokens": 47033051.0,
      "step": 18555
    },
    {
      "entropy": 6.234206008911133,
      "epoch": 2.1419503750721294,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045492533169879816,
      "loss": 5.6704,
      "mean_token_accuracy": 0.20061807781457902,
      "num_tokens": 47044868.0,
      "step": 18560
    },
    {
      "entropy": 6.272685432434082,
      "epoch": 2.142527409117138,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00045490055181389216,
      "loss": 5.6508,
      "mean_token_accuracy": 0.20505160093307495,
      "num_tokens": 47057050.0,
      "step": 18565
    },
    {
      "entropy": 6.229650449752808,
      "epoch": 2.1431044431621467,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045487576587812046,
      "loss": 5.6393,
      "mean_token_accuracy": 0.20420267581939697,
      "num_tokens": 47070338.0,
      "step": 18570
    },
    {
      "entropy": 6.1334577083587645,
      "epoch": 2.143681477207155,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045485097389231675,
      "loss": 5.5567,
      "mean_token_accuracy": 0.21689492613077163,
      "num_tokens": 47082958.0,
      "step": 18575
    },
    {
      "entropy": 6.212721538543701,
      "epoch": 2.144258511252164,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004548261758573149,
      "loss": 5.6793,
      "mean_token_accuracy": 0.2072567641735077,
      "num_tokens": 47096803.0,
      "step": 18580
    },
    {
      "entropy": 6.2943220138549805,
      "epoch": 2.1448355452971724,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004548013717739489,
      "loss": 5.7012,
      "mean_token_accuracy": 0.19397093951702118,
      "num_tokens": 47109858.0,
      "step": 18585
    },
    {
      "entropy": 6.337107133865357,
      "epoch": 2.145412579342181,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004547765616430531,
      "loss": 5.7234,
      "mean_token_accuracy": 0.20031313002109527,
      "num_tokens": 47123664.0,
      "step": 18590
    },
    {
      "entropy": 6.21878080368042,
      "epoch": 2.14598961338719,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004547517454654619,
      "loss": 5.6832,
      "mean_token_accuracy": 0.20109350979328156,
      "num_tokens": 47135860.0,
      "step": 18595
    },
    {
      "entropy": 6.259080266952514,
      "epoch": 2.1465666474321985,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045472692324201005,
      "loss": 5.7273,
      "mean_token_accuracy": 0.19150826334953308,
      "num_tokens": 47147571.0,
      "step": 18600
    },
    {
      "entropy": 6.284874486923218,
      "epoch": 2.1471436814772074,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045470209497353243,
      "loss": 5.6594,
      "mean_token_accuracy": 0.20161578059196472,
      "num_tokens": 47158947.0,
      "step": 18605
    },
    {
      "entropy": 6.167023706436157,
      "epoch": 2.1477207155222158,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004546772606608641,
      "loss": 5.62,
      "mean_token_accuracy": 0.20948661863803864,
      "num_tokens": 47172666.0,
      "step": 18610
    },
    {
      "entropy": 6.161411571502685,
      "epoch": 2.1482977495672246,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004546524203048404,
      "loss": 5.6099,
      "mean_token_accuracy": 0.2125784933567047,
      "num_tokens": 47185664.0,
      "step": 18615
    },
    {
      "entropy": 6.377475166320801,
      "epoch": 2.148874783612233,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004546275739062967,
      "loss": 5.8006,
      "mean_token_accuracy": 0.18675549179315568,
      "num_tokens": 47198325.0,
      "step": 18620
    },
    {
      "entropy": 6.210091257095337,
      "epoch": 2.149451817657242,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004546027214660688,
      "loss": 5.6331,
      "mean_token_accuracy": 0.2095345064997673,
      "num_tokens": 47211676.0,
      "step": 18625
    },
    {
      "entropy": 6.26652979850769,
      "epoch": 2.1500288517022503,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004545778629849926,
      "loss": 5.7183,
      "mean_token_accuracy": 0.19484145641326905,
      "num_tokens": 47223956.0,
      "step": 18630
    },
    {
      "entropy": 6.249640083312988,
      "epoch": 2.150605885747259,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004545529984639042,
      "loss": 5.6889,
      "mean_token_accuracy": 0.20277153104543685,
      "num_tokens": 47236806.0,
      "step": 18635
    },
    {
      "entropy": 6.222601652145386,
      "epoch": 2.1511829197922676,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004545281279036398,
      "loss": 5.6724,
      "mean_token_accuracy": 0.19764189273118973,
      "num_tokens": 47249750.0,
      "step": 18640
    },
    {
      "entropy": 6.221406412124634,
      "epoch": 2.1517599538372765,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004545032513050361,
      "loss": 5.66,
      "mean_token_accuracy": 0.2100660040974617,
      "num_tokens": 47261480.0,
      "step": 18645
    },
    {
      "entropy": 6.270481395721435,
      "epoch": 2.152336987882285,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004544783686689296,
      "loss": 5.7242,
      "mean_token_accuracy": 0.20075965225696563,
      "num_tokens": 47275025.0,
      "step": 18650
    },
    {
      "entropy": 6.2308391571044925,
      "epoch": 2.1529140219272938,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045445347999615736,
      "loss": 5.6805,
      "mean_token_accuracy": 0.20471232533454894,
      "num_tokens": 47287966.0,
      "step": 18655
    },
    {
      "entropy": 6.274382448196411,
      "epoch": 2.153491055972302,
      "grad_norm": 1.0,
      "learning_rate": 0.00045442858528755653,
      "loss": 5.6877,
      "mean_token_accuracy": 0.20347724705934525,
      "num_tokens": 47299883.0,
      "step": 18660
    },
    {
      "entropy": 6.266181802749633,
      "epoch": 2.154068090017311,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045440368454396435,
      "loss": 5.7312,
      "mean_token_accuracy": 0.1956154316663742,
      "num_tokens": 47311704.0,
      "step": 18665
    },
    {
      "entropy": 6.299458312988281,
      "epoch": 2.15464512406232,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004543787777662183,
      "loss": 5.8052,
      "mean_token_accuracy": 0.19705142229795455,
      "num_tokens": 47325663.0,
      "step": 18670
    },
    {
      "entropy": 6.335583066940307,
      "epoch": 2.1552221581073283,
      "grad_norm": 0.921875,
      "learning_rate": 0.00045435386495515617,
      "loss": 5.7962,
      "mean_token_accuracy": 0.1926809921860695,
      "num_tokens": 47339349.0,
      "step": 18675
    },
    {
      "entropy": 6.326417589187622,
      "epoch": 2.155799192152337,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004543289461116159,
      "loss": 5.8342,
      "mean_token_accuracy": 0.193597774207592,
      "num_tokens": 47352245.0,
      "step": 18680
    },
    {
      "entropy": 6.275993919372558,
      "epoch": 2.1563762261973456,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004543040212364356,
      "loss": 5.6839,
      "mean_token_accuracy": 0.19787552058696747,
      "num_tokens": 47364564.0,
      "step": 18685
    },
    {
      "entropy": 6.245632362365723,
      "epoch": 2.1569532602423545,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004542790903304536,
      "loss": 5.6335,
      "mean_token_accuracy": 0.20095878690481186,
      "num_tokens": 47377156.0,
      "step": 18690
    },
    {
      "entropy": 6.250470781326294,
      "epoch": 2.157530294287363,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004542541533945085,
      "loss": 5.6701,
      "mean_token_accuracy": 0.20707043260335922,
      "num_tokens": 47390003.0,
      "step": 18695
    },
    {
      "entropy": 6.330975008010864,
      "epoch": 2.1581073283323717,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045422921042943885,
      "loss": 5.7901,
      "mean_token_accuracy": 0.19620831310749054,
      "num_tokens": 47403651.0,
      "step": 18700
    },
    {
      "entropy": 6.297650861740112,
      "epoch": 2.15868436237738,
      "grad_norm": 1.046875,
      "learning_rate": 0.000454204261436084,
      "loss": 5.7019,
      "mean_token_accuracy": 0.19565952122211455,
      "num_tokens": 47415481.0,
      "step": 18705
    },
    {
      "entropy": 6.25938310623169,
      "epoch": 2.159261396422389,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045417930641528255,
      "loss": 5.7431,
      "mean_token_accuracy": 0.19842022508382798,
      "num_tokens": 47428303.0,
      "step": 18710
    },
    {
      "entropy": 6.239545297622681,
      "epoch": 2.1598384304673974,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045415434536787424,
      "loss": 5.6255,
      "mean_token_accuracy": 0.20707356184720993,
      "num_tokens": 47441270.0,
      "step": 18715
    },
    {
      "entropy": 6.253909969329834,
      "epoch": 2.1604154645124063,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004541293782946985,
      "loss": 5.636,
      "mean_token_accuracy": 0.20728814899921416,
      "num_tokens": 47454870.0,
      "step": 18720
    },
    {
      "entropy": 6.273397827148438,
      "epoch": 2.1609924985574147,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004541044051965952,
      "loss": 5.6776,
      "mean_token_accuracy": 0.2051711544394493,
      "num_tokens": 47466099.0,
      "step": 18725
    },
    {
      "entropy": 6.26264910697937,
      "epoch": 2.1615695326024236,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004540794260744041,
      "loss": 5.6038,
      "mean_token_accuracy": 0.2155757650732994,
      "num_tokens": 47477206.0,
      "step": 18730
    },
    {
      "entropy": 6.058430147171021,
      "epoch": 2.162146566647432,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004540544409289655,
      "loss": 5.5478,
      "mean_token_accuracy": 0.20786997377872468,
      "num_tokens": 47489500.0,
      "step": 18735
    },
    {
      "entropy": 6.3412182331085205,
      "epoch": 2.162723600692441,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004540294497611197,
      "loss": 5.7582,
      "mean_token_accuracy": 0.19162163585424424,
      "num_tokens": 47502090.0,
      "step": 18740
    },
    {
      "entropy": 6.382848024368286,
      "epoch": 2.1633006347374497,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00045400445257170725,
      "loss": 5.839,
      "mean_token_accuracy": 0.19052740633487703,
      "num_tokens": 47515496.0,
      "step": 18745
    },
    {
      "entropy": 6.282927322387695,
      "epoch": 2.163877668782458,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004539794493615689,
      "loss": 5.683,
      "mean_token_accuracy": 0.1982918366789818,
      "num_tokens": 47529174.0,
      "step": 18750
    },
    {
      "entropy": 6.27329797744751,
      "epoch": 2.164454702827467,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004539544401315458,
      "loss": 5.7143,
      "mean_token_accuracy": 0.19837197810411453,
      "num_tokens": 47542495.0,
      "step": 18755
    },
    {
      "entropy": 6.279916143417358,
      "epoch": 2.1650317368724754,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000453929424882479,
      "loss": 5.7066,
      "mean_token_accuracy": 0.19956169426441192,
      "num_tokens": 47554936.0,
      "step": 18760
    },
    {
      "entropy": 6.379324626922608,
      "epoch": 2.1656087709174843,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045390440361520987,
      "loss": 5.7876,
      "mean_token_accuracy": 0.19110388308763504,
      "num_tokens": 47568168.0,
      "step": 18765
    },
    {
      "entropy": 6.2624505996704105,
      "epoch": 2.1661858049624927,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004538793763305799,
      "loss": 5.6334,
      "mean_token_accuracy": 0.20755986124277115,
      "num_tokens": 47581944.0,
      "step": 18770
    },
    {
      "entropy": 6.239258623123169,
      "epoch": 2.1667628390075016,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045385434302943104,
      "loss": 5.7278,
      "mean_token_accuracy": 0.19743309617042543,
      "num_tokens": 47595554.0,
      "step": 18775
    },
    {
      "entropy": 6.363761234283447,
      "epoch": 2.16733987305251,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004538293037126052,
      "loss": 5.8393,
      "mean_token_accuracy": 0.187453293800354,
      "num_tokens": 47608516.0,
      "step": 18780
    },
    {
      "entropy": 6.268542051315308,
      "epoch": 2.167916907097519,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045380425838094444,
      "loss": 5.7128,
      "mean_token_accuracy": 0.19627011120319365,
      "num_tokens": 47620100.0,
      "step": 18785
    },
    {
      "entropy": 6.21559624671936,
      "epoch": 2.1684939411425272,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00045377920703529133,
      "loss": 5.651,
      "mean_token_accuracy": 0.2086021736264229,
      "num_tokens": 47632982.0,
      "step": 18790
    },
    {
      "entropy": 6.255552577972412,
      "epoch": 2.169070975187536,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004537541496764885,
      "loss": 5.656,
      "mean_token_accuracy": 0.19968265295028687,
      "num_tokens": 47645967.0,
      "step": 18795
    },
    {
      "entropy": 6.1886743068695065,
      "epoch": 2.1696480092325445,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004537290863053786,
      "loss": 5.6349,
      "mean_token_accuracy": 0.20659561157226564,
      "num_tokens": 47659392.0,
      "step": 18800
    },
    {
      "entropy": 6.320357608795166,
      "epoch": 2.1702250432775534,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045370401692280455,
      "loss": 5.7337,
      "mean_token_accuracy": 0.19001376926898955,
      "num_tokens": 47672516.0,
      "step": 18805
    },
    {
      "entropy": 6.304880905151367,
      "epoch": 2.170802077322562,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045367894152960976,
      "loss": 5.7336,
      "mean_token_accuracy": 0.19413152784109117,
      "num_tokens": 47684638.0,
      "step": 18810
    },
    {
      "entropy": 6.263280200958252,
      "epoch": 2.1713791113675707,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00045365386012663744,
      "loss": 5.6962,
      "mean_token_accuracy": 0.20471349358558655,
      "num_tokens": 47697573.0,
      "step": 18815
    },
    {
      "entropy": 6.204102802276611,
      "epoch": 2.1719561454125795,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045362877271473137,
      "loss": 5.6786,
      "mean_token_accuracy": 0.20395950078964234,
      "num_tokens": 47709120.0,
      "step": 18820
    },
    {
      "entropy": 6.2735084056854244,
      "epoch": 2.172533179457588,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045360367929473517,
      "loss": 5.728,
      "mean_token_accuracy": 0.1982928603887558,
      "num_tokens": 47722279.0,
      "step": 18825
    },
    {
      "entropy": 6.2566743850708,
      "epoch": 2.173110213502597,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000453578579867493,
      "loss": 5.6567,
      "mean_token_accuracy": 0.20497894883155823,
      "num_tokens": 47735069.0,
      "step": 18830
    },
    {
      "entropy": 6.263159894943238,
      "epoch": 2.1736872475476052,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00045355347443384896,
      "loss": 5.7372,
      "mean_token_accuracy": 0.19475683867931365,
      "num_tokens": 47748370.0,
      "step": 18835
    },
    {
      "entropy": 6.199267721176147,
      "epoch": 2.174264281592614,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045352836299464755,
      "loss": 5.6305,
      "mean_token_accuracy": 0.20588551163673402,
      "num_tokens": 47761509.0,
      "step": 18840
    },
    {
      "entropy": 6.157886695861817,
      "epoch": 2.1748413156376225,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004535032455507333,
      "loss": 5.5665,
      "mean_token_accuracy": 0.21668360829353334,
      "num_tokens": 47774551.0,
      "step": 18845
    },
    {
      "entropy": 6.33867883682251,
      "epoch": 2.1754183496826314,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045347812210295107,
      "loss": 5.7499,
      "mean_token_accuracy": 0.19671245962381362,
      "num_tokens": 47788063.0,
      "step": 18850
    },
    {
      "entropy": 6.344557857513427,
      "epoch": 2.17599538372764,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045345299265214583,
      "loss": 5.7522,
      "mean_token_accuracy": 0.19014041125774384,
      "num_tokens": 47799781.0,
      "step": 18855
    },
    {
      "entropy": 6.222022199630738,
      "epoch": 2.1765724177726486,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045342785719916284,
      "loss": 5.6951,
      "mean_token_accuracy": 0.19823031276464462,
      "num_tokens": 47812066.0,
      "step": 18860
    },
    {
      "entropy": 6.224159574508667,
      "epoch": 2.177149451817657,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045340271574484755,
      "loss": 5.7939,
      "mean_token_accuracy": 0.19934281706809998,
      "num_tokens": 47824943.0,
      "step": 18865
    },
    {
      "entropy": 6.302858924865722,
      "epoch": 2.177726485862666,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004533775682900454,
      "loss": 5.7213,
      "mean_token_accuracy": 0.20161155611276627,
      "num_tokens": 47836942.0,
      "step": 18870
    },
    {
      "entropy": 6.318057823181152,
      "epoch": 2.1783035199076743,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004533524148356025,
      "loss": 5.7503,
      "mean_token_accuracy": 0.19378005117177963,
      "num_tokens": 47849895.0,
      "step": 18875
    },
    {
      "entropy": 6.254288196563721,
      "epoch": 2.178880553952683,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004533272553823646,
      "loss": 5.6242,
      "mean_token_accuracy": 0.20286617577075958,
      "num_tokens": 47861452.0,
      "step": 18880
    },
    {
      "entropy": 6.2657733917236325,
      "epoch": 2.179457587997692,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045330208993117816,
      "loss": 5.743,
      "mean_token_accuracy": 0.19137451946735382,
      "num_tokens": 47873913.0,
      "step": 18885
    },
    {
      "entropy": 6.298262071609497,
      "epoch": 2.1800346220427005,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004532769184828894,
      "loss": 5.7181,
      "mean_token_accuracy": 0.19667167365550994,
      "num_tokens": 47887986.0,
      "step": 18890
    },
    {
      "entropy": 6.320073223114013,
      "epoch": 2.1806116560877093,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004532517410383451,
      "loss": 5.7579,
      "mean_token_accuracy": 0.19366970509290696,
      "num_tokens": 47899272.0,
      "step": 18895
    },
    {
      "entropy": 6.236899375915527,
      "epoch": 2.1811886901327178,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000453226557598392,
      "loss": 5.5996,
      "mean_token_accuracy": 0.2143673375248909,
      "num_tokens": 47911097.0,
      "step": 18900
    },
    {
      "entropy": 6.24728217124939,
      "epoch": 2.1817657241777266,
      "grad_norm": 1.0,
      "learning_rate": 0.0004532013681638771,
      "loss": 5.7533,
      "mean_token_accuracy": 0.19475191682577134,
      "num_tokens": 47924034.0,
      "step": 18905
    },
    {
      "entropy": 6.258606100082398,
      "epoch": 2.182342758222735,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004531761727356478,
      "loss": 5.6315,
      "mean_token_accuracy": 0.21130064278841018,
      "num_tokens": 47937589.0,
      "step": 18910
    },
    {
      "entropy": 6.335956001281739,
      "epoch": 2.182919792267744,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004531509713145514,
      "loss": 5.7684,
      "mean_token_accuracy": 0.19314059466123581,
      "num_tokens": 47951880.0,
      "step": 18915
    },
    {
      "entropy": 6.275867605209351,
      "epoch": 2.1834968263127523,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00045312576390143557,
      "loss": 5.6743,
      "mean_token_accuracy": 0.2059902861714363,
      "num_tokens": 47964599.0,
      "step": 18920
    },
    {
      "entropy": 6.2859495162963865,
      "epoch": 2.184073860357761,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045310055049714815,
      "loss": 5.7614,
      "mean_token_accuracy": 0.19627155214548112,
      "num_tokens": 47976879.0,
      "step": 18925
    },
    {
      "entropy": 6.315849494934082,
      "epoch": 2.1846508944027696,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045307533110253726,
      "loss": 5.7669,
      "mean_token_accuracy": 0.1897657260298729,
      "num_tokens": 47990462.0,
      "step": 18930
    },
    {
      "entropy": 6.325391864776611,
      "epoch": 2.1852279284477785,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045305010571845096,
      "loss": 5.7009,
      "mean_token_accuracy": 0.20379158109426498,
      "num_tokens": 48002924.0,
      "step": 18935
    },
    {
      "entropy": 6.271603202819824,
      "epoch": 2.185804962492787,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004530248743457378,
      "loss": 5.7269,
      "mean_token_accuracy": 0.19308781772851943,
      "num_tokens": 48013826.0,
      "step": 18940
    },
    {
      "entropy": 6.239746904373169,
      "epoch": 2.1863819965377957,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004529996369852465,
      "loss": 5.6825,
      "mean_token_accuracy": 0.20461236089468002,
      "num_tokens": 48027087.0,
      "step": 18945
    },
    {
      "entropy": 6.305883312225342,
      "epoch": 2.1869590305828046,
      "grad_norm": 0.921875,
      "learning_rate": 0.00045297439363782576,
      "loss": 5.7432,
      "mean_token_accuracy": 0.19815013706684112,
      "num_tokens": 48041157.0,
      "step": 18950
    },
    {
      "entropy": 6.2592040538787845,
      "epoch": 2.187536064627813,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004529491443043247,
      "loss": 5.6538,
      "mean_token_accuracy": 0.19850932359695433,
      "num_tokens": 48054168.0,
      "step": 18955
    },
    {
      "entropy": 6.262095880508423,
      "epoch": 2.188113098672822,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004529238889855926,
      "loss": 5.6899,
      "mean_token_accuracy": 0.19947659373283386,
      "num_tokens": 48066826.0,
      "step": 18960
    },
    {
      "entropy": 6.259306478500366,
      "epoch": 2.1886901327178303,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004528986276824789,
      "loss": 5.7517,
      "mean_token_accuracy": 0.19551298171281814,
      "num_tokens": 48079788.0,
      "step": 18965
    },
    {
      "entropy": 6.275222730636597,
      "epoch": 2.189267166762839,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004528733603958331,
      "loss": 5.6748,
      "mean_token_accuracy": 0.19502945840358735,
      "num_tokens": 48091517.0,
      "step": 18970
    },
    {
      "entropy": 6.22862434387207,
      "epoch": 2.1898442008078476,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004528480871265053,
      "loss": 5.6476,
      "mean_token_accuracy": 0.20176999270915985,
      "num_tokens": 48104964.0,
      "step": 18975
    },
    {
      "entropy": 6.287322521209717,
      "epoch": 2.1904212348528564,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045282280787534537,
      "loss": 5.7169,
      "mean_token_accuracy": 0.2030135288834572,
      "num_tokens": 48118232.0,
      "step": 18980
    },
    {
      "entropy": 6.311247777938843,
      "epoch": 2.190998268897865,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004527975226432036,
      "loss": 5.7794,
      "mean_token_accuracy": 0.19234614968299865,
      "num_tokens": 48131320.0,
      "step": 18985
    },
    {
      "entropy": 6.28266716003418,
      "epoch": 2.1915753029428737,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045277223143093057,
      "loss": 5.7129,
      "mean_token_accuracy": 0.20204475075006484,
      "num_tokens": 48143432.0,
      "step": 18990
    },
    {
      "entropy": 6.28989577293396,
      "epoch": 2.192152336987882,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045274693423937674,
      "loss": 5.689,
      "mean_token_accuracy": 0.20814001113176345,
      "num_tokens": 48158205.0,
      "step": 18995
    },
    {
      "entropy": 6.254542303085327,
      "epoch": 2.192729371032891,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045272163106939314,
      "loss": 5.7014,
      "mean_token_accuracy": 0.19329966604709625,
      "num_tokens": 48169225.0,
      "step": 19000
    },
    {
      "entropy": 6.345037841796875,
      "epoch": 2.1933064050778994,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00045269632192183076,
      "loss": 5.8283,
      "mean_token_accuracy": 0.18812476992607116,
      "num_tokens": 48182715.0,
      "step": 19005
    },
    {
      "entropy": 6.242325735092163,
      "epoch": 2.1938834391229083,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00045267100679754075,
      "loss": 5.6497,
      "mean_token_accuracy": 0.20813382863998414,
      "num_tokens": 48195817.0,
      "step": 19010
    },
    {
      "entropy": 6.226392078399658,
      "epoch": 2.1944604731679167,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004526456856973748,
      "loss": 5.6771,
      "mean_token_accuracy": 0.2018361508846283,
      "num_tokens": 48208347.0,
      "step": 19015
    },
    {
      "entropy": 6.264702939987183,
      "epoch": 2.1950375072129256,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004526203586221844,
      "loss": 5.6808,
      "mean_token_accuracy": 0.19835537672042847,
      "num_tokens": 48221202.0,
      "step": 19020
    },
    {
      "entropy": 6.231685447692871,
      "epoch": 2.1956145412579344,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045259502557282145,
      "loss": 5.6367,
      "mean_token_accuracy": 0.20124043375253678,
      "num_tokens": 48233471.0,
      "step": 19025
    },
    {
      "entropy": 6.249116897583008,
      "epoch": 2.196191575302943,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004525696865501381,
      "loss": 5.6714,
      "mean_token_accuracy": 0.19930269718170165,
      "num_tokens": 48246685.0,
      "step": 19030
    },
    {
      "entropy": 6.260441780090332,
      "epoch": 2.1967686093479517,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004525443415549865,
      "loss": 5.6645,
      "mean_token_accuracy": 0.20589246451854706,
      "num_tokens": 48259942.0,
      "step": 19035
    },
    {
      "entropy": 6.236582612991333,
      "epoch": 2.19734564339296,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045251899058821915,
      "loss": 5.6435,
      "mean_token_accuracy": 0.2073597177863121,
      "num_tokens": 48272628.0,
      "step": 19040
    },
    {
      "entropy": 6.298931646347046,
      "epoch": 2.197922677437969,
      "grad_norm": 1.0,
      "learning_rate": 0.0004524936336506887,
      "loss": 5.6392,
      "mean_token_accuracy": 0.2070186272263527,
      "num_tokens": 48286463.0,
      "step": 19045
    },
    {
      "entropy": 6.2622260570526125,
      "epoch": 2.1984997114829774,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004524682707432482,
      "loss": 5.7438,
      "mean_token_accuracy": 0.2015097215771675,
      "num_tokens": 48299074.0,
      "step": 19050
    },
    {
      "entropy": 6.2962483882904055,
      "epoch": 2.1990767455279863,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045244290186675034,
      "loss": 5.7796,
      "mean_token_accuracy": 0.20110249668359756,
      "num_tokens": 48312419.0,
      "step": 19055
    },
    {
      "entropy": 6.360716342926025,
      "epoch": 2.1996537795729947,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004524175270220489,
      "loss": 5.7297,
      "mean_token_accuracy": 0.1973346948623657,
      "num_tokens": 48325085.0,
      "step": 19060
    },
    {
      "entropy": 6.366532516479492,
      "epoch": 2.2002308136180035,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045239214620999683,
      "loss": 5.7745,
      "mean_token_accuracy": 0.19607491046190262,
      "num_tokens": 48337461.0,
      "step": 19065
    },
    {
      "entropy": 6.1924303531646725,
      "epoch": 2.200807847663012,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004523667594314481,
      "loss": 5.7253,
      "mean_token_accuracy": 0.20713589042425157,
      "num_tokens": 48351851.0,
      "step": 19070
    },
    {
      "entropy": 6.297838592529297,
      "epoch": 2.201384881708021,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045234136668725655,
      "loss": 5.7181,
      "mean_token_accuracy": 0.19961626678705216,
      "num_tokens": 48363443.0,
      "step": 19075
    },
    {
      "entropy": 6.234776020050049,
      "epoch": 2.2019619157530292,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045231596797827616,
      "loss": 5.673,
      "mean_token_accuracy": 0.19805619567632676,
      "num_tokens": 48376757.0,
      "step": 19080
    },
    {
      "entropy": 6.215353727340698,
      "epoch": 2.202538949798038,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045229056330536134,
      "loss": 5.5809,
      "mean_token_accuracy": 0.20904551446437836,
      "num_tokens": 48390127.0,
      "step": 19085
    },
    {
      "entropy": 6.3331140041351315,
      "epoch": 2.2031159838430465,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045226515266936644,
      "loss": 5.7145,
      "mean_token_accuracy": 0.1929883301258087,
      "num_tokens": 48401760.0,
      "step": 19090
    },
    {
      "entropy": 6.277010488510132,
      "epoch": 2.2036930178880554,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004522397360711462,
      "loss": 5.7673,
      "mean_token_accuracy": 0.1950174793601036,
      "num_tokens": 48414291.0,
      "step": 19095
    },
    {
      "entropy": 6.303528642654419,
      "epoch": 2.2042700519330642,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004522143135115555,
      "loss": 5.6861,
      "mean_token_accuracy": 0.2072443500161171,
      "num_tokens": 48428818.0,
      "step": 19100
    },
    {
      "entropy": 6.300106239318848,
      "epoch": 2.2048470859780727,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045218888499144933,
      "loss": 5.7484,
      "mean_token_accuracy": 0.19432809352874755,
      "num_tokens": 48442425.0,
      "step": 19105
    },
    {
      "entropy": 6.181846666336059,
      "epoch": 2.2054241200230815,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045216345051168314,
      "loss": 5.6606,
      "mean_token_accuracy": 0.19863310903310777,
      "num_tokens": 48454814.0,
      "step": 19110
    },
    {
      "entropy": 6.311408042907715,
      "epoch": 2.20600115406809,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004521380100731122,
      "loss": 5.624,
      "mean_token_accuracy": 0.20406401306390762,
      "num_tokens": 48467211.0,
      "step": 19115
    },
    {
      "entropy": 6.256623363494873,
      "epoch": 2.206578188113099,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045211256367659234,
      "loss": 5.6106,
      "mean_token_accuracy": 0.21176437437534332,
      "num_tokens": 48479252.0,
      "step": 19120
    },
    {
      "entropy": 6.19311490058899,
      "epoch": 2.207155222158107,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004520871113229793,
      "loss": 5.6675,
      "mean_token_accuracy": 0.2024013876914978,
      "num_tokens": 48491227.0,
      "step": 19125
    },
    {
      "entropy": 6.297635459899903,
      "epoch": 2.207732256203116,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045206165301312927,
      "loss": 5.7736,
      "mean_token_accuracy": 0.2019822984933853,
      "num_tokens": 48503374.0,
      "step": 19130
    },
    {
      "entropy": 6.248849391937256,
      "epoch": 2.2083092902481245,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004520361887478985,
      "loss": 5.6847,
      "mean_token_accuracy": 0.1993539184331894,
      "num_tokens": 48515294.0,
      "step": 19135
    },
    {
      "entropy": 6.2712489604949955,
      "epoch": 2.2088863242931334,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004520107185281435,
      "loss": 5.7568,
      "mean_token_accuracy": 0.19303477853536605,
      "num_tokens": 48528429.0,
      "step": 19140
    },
    {
      "entropy": 6.264723396301269,
      "epoch": 2.209463358338142,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004519852423547208,
      "loss": 5.6535,
      "mean_token_accuracy": 0.20044994354248047,
      "num_tokens": 48541077.0,
      "step": 19145
    },
    {
      "entropy": 6.228554391860962,
      "epoch": 2.2100403923831506,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004519597602284875,
      "loss": 5.6608,
      "mean_token_accuracy": 0.21175492256879808,
      "num_tokens": 48555338.0,
      "step": 19150
    },
    {
      "entropy": 6.3012443542480465,
      "epoch": 2.210617426428159,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004519342721503005,
      "loss": 5.776,
      "mean_token_accuracy": 0.19323104619979858,
      "num_tokens": 48568374.0,
      "step": 19155
    },
    {
      "entropy": 6.279902362823487,
      "epoch": 2.211194460473168,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004519087781210171,
      "loss": 5.6885,
      "mean_token_accuracy": 0.20026946663856507,
      "num_tokens": 48582412.0,
      "step": 19160
    },
    {
      "entropy": 6.298382902145386,
      "epoch": 2.2117714945181763,
      "grad_norm": 0.96875,
      "learning_rate": 0.000451883278141495,
      "loss": 5.6964,
      "mean_token_accuracy": 0.19915968477725982,
      "num_tokens": 48594942.0,
      "step": 19165
    },
    {
      "entropy": 6.221723794937134,
      "epoch": 2.212348528563185,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045185777221259157,
      "loss": 5.5346,
      "mean_token_accuracy": 0.22117967009544373,
      "num_tokens": 48605748.0,
      "step": 19170
    },
    {
      "entropy": 6.275056982040406,
      "epoch": 2.212925562608194,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004518322603351648,
      "loss": 5.8214,
      "mean_token_accuracy": 0.19259049445390702,
      "num_tokens": 48617894.0,
      "step": 19175
    },
    {
      "entropy": 6.295895004272461,
      "epoch": 2.2135025966532025,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045180674251007287,
      "loss": 5.698,
      "mean_token_accuracy": 0.19850409477949144,
      "num_tokens": 48631237.0,
      "step": 19180
    },
    {
      "entropy": 6.245034027099609,
      "epoch": 2.2140796306982113,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045178121873817395,
      "loss": 5.7293,
      "mean_token_accuracy": 0.19948506653308867,
      "num_tokens": 48643218.0,
      "step": 19185
    },
    {
      "entropy": 6.293656969070435,
      "epoch": 2.2146566647432198,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004517556890203265,
      "loss": 5.7456,
      "mean_token_accuracy": 0.19517271667718888,
      "num_tokens": 48655281.0,
      "step": 19190
    },
    {
      "entropy": 6.336243963241577,
      "epoch": 2.2152336987882286,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045173015335738925,
      "loss": 5.7686,
      "mean_token_accuracy": 0.1993575870990753,
      "num_tokens": 48669301.0,
      "step": 19195
    },
    {
      "entropy": 6.25295934677124,
      "epoch": 2.215810732833237,
      "grad_norm": 0.90625,
      "learning_rate": 0.00045170461175022116,
      "loss": 5.682,
      "mean_token_accuracy": 0.20224205851554872,
      "num_tokens": 48681966.0,
      "step": 19200
    },
    {
      "entropy": 6.258906126022339,
      "epoch": 2.216387766878246,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004516790641996812,
      "loss": 5.7306,
      "mean_token_accuracy": 0.1944676771759987,
      "num_tokens": 48695414.0,
      "step": 19205
    },
    {
      "entropy": 6.323549842834472,
      "epoch": 2.2169648009232543,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004516535107066286,
      "loss": 5.7688,
      "mean_token_accuracy": 0.19341218769550322,
      "num_tokens": 48707953.0,
      "step": 19210
    },
    {
      "entropy": 6.18471212387085,
      "epoch": 2.217541834968263,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045162795127192296,
      "loss": 5.6442,
      "mean_token_accuracy": 0.21179237365722656,
      "num_tokens": 48720524.0,
      "step": 19215
    },
    {
      "entropy": 6.286249494552612,
      "epoch": 2.2181188690132716,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000451602385896424,
      "loss": 5.7294,
      "mean_token_accuracy": 0.1984877899289131,
      "num_tokens": 48732544.0,
      "step": 19220
    },
    {
      "entropy": 6.372013425827026,
      "epoch": 2.2186959030582805,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045157681458099145,
      "loss": 5.8077,
      "mean_token_accuracy": 0.19115511178970337,
      "num_tokens": 48745520.0,
      "step": 19225
    },
    {
      "entropy": 6.272190237045288,
      "epoch": 2.2192729371032893,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004515512373264854,
      "loss": 5.7775,
      "mean_token_accuracy": 0.192821903526783,
      "num_tokens": 48757442.0,
      "step": 19230
    },
    {
      "entropy": 6.2458030700683596,
      "epoch": 2.2198499711482977,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045152565413376623,
      "loss": 5.6928,
      "mean_token_accuracy": 0.20217442214488984,
      "num_tokens": 48771351.0,
      "step": 19235
    },
    {
      "entropy": 6.2228117942810055,
      "epoch": 2.2204270051933066,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004515000650036944,
      "loss": 5.7112,
      "mean_token_accuracy": 0.2047142803668976,
      "num_tokens": 48783897.0,
      "step": 19240
    },
    {
      "entropy": 6.218664884567261,
      "epoch": 2.221004039238315,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004514744699371305,
      "loss": 5.6037,
      "mean_token_accuracy": 0.21483812779188155,
      "num_tokens": 48796334.0,
      "step": 19245
    },
    {
      "entropy": 6.296231651306153,
      "epoch": 2.221581073283324,
      "grad_norm": 0.921875,
      "learning_rate": 0.00045144886893493547,
      "loss": 5.6723,
      "mean_token_accuracy": 0.2011644199490547,
      "num_tokens": 48809762.0,
      "step": 19250
    },
    {
      "entropy": 6.1894958972930905,
      "epoch": 2.2221581073283323,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045142326199797047,
      "loss": 5.6413,
      "mean_token_accuracy": 0.20891234129667283,
      "num_tokens": 48822038.0,
      "step": 19255
    },
    {
      "entropy": 6.251821279525757,
      "epoch": 2.222735141373341,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00045139764912709656,
      "loss": 5.7117,
      "mean_token_accuracy": 0.19587273746728898,
      "num_tokens": 48835284.0,
      "step": 19260
    },
    {
      "entropy": 6.23588080406189,
      "epoch": 2.2233121754183496,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004513720303231754,
      "loss": 5.6297,
      "mean_token_accuracy": 0.2087322324514389,
      "num_tokens": 48848673.0,
      "step": 19265
    },
    {
      "entropy": 6.292097187042236,
      "epoch": 2.2238892094633584,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045134640558706864,
      "loss": 5.7712,
      "mean_token_accuracy": 0.19155097901821136,
      "num_tokens": 48861071.0,
      "step": 19270
    },
    {
      "entropy": 6.325938653945923,
      "epoch": 2.224466243508367,
      "grad_norm": 0.921875,
      "learning_rate": 0.000451320774919638,
      "loss": 5.8036,
      "mean_token_accuracy": 0.1967108130455017,
      "num_tokens": 48873471.0,
      "step": 19275
    },
    {
      "entropy": 6.253922796249389,
      "epoch": 2.2250432775533757,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045129513832174587,
      "loss": 5.7008,
      "mean_token_accuracy": 0.20379954129457473,
      "num_tokens": 48885565.0,
      "step": 19280
    },
    {
      "entropy": 6.268419313430786,
      "epoch": 2.225620311598384,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045126949579425414,
      "loss": 5.6781,
      "mean_token_accuracy": 0.20051735788583755,
      "num_tokens": 48897785.0,
      "step": 19285
    },
    {
      "entropy": 6.330311822891235,
      "epoch": 2.226197345643393,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045124384733802563,
      "loss": 5.7636,
      "mean_token_accuracy": 0.190012226998806,
      "num_tokens": 48910202.0,
      "step": 19290
    },
    {
      "entropy": 6.180883502960205,
      "epoch": 2.2267743796884014,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004512181929539228,
      "loss": 5.6024,
      "mean_token_accuracy": 0.20691357105970382,
      "num_tokens": 48923643.0,
      "step": 19295
    },
    {
      "entropy": 6.318269300460815,
      "epoch": 2.2273514137334103,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045119253264280855,
      "loss": 5.7237,
      "mean_token_accuracy": 0.19523371756076813,
      "num_tokens": 48936513.0,
      "step": 19300
    },
    {
      "entropy": 6.233939552307129,
      "epoch": 2.227928447778419,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00045116686640554607,
      "loss": 5.7036,
      "mean_token_accuracy": 0.19331514239311218,
      "num_tokens": 48950082.0,
      "step": 19305
    },
    {
      "entropy": 6.243971109390259,
      "epoch": 2.2285054818234276,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00045114119424299843,
      "loss": 5.6649,
      "mean_token_accuracy": 0.20551566332578658,
      "num_tokens": 48961773.0,
      "step": 19310
    },
    {
      "entropy": 6.251380729675293,
      "epoch": 2.2290825158684364,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004511155161560293,
      "loss": 5.7071,
      "mean_token_accuracy": 0.2024727553129196,
      "num_tokens": 48974171.0,
      "step": 19315
    },
    {
      "entropy": 6.307606267929077,
      "epoch": 2.229659549913445,
      "grad_norm": 0.890625,
      "learning_rate": 0.00045108983214550225,
      "loss": 5.7324,
      "mean_token_accuracy": 0.1924608677625656,
      "num_tokens": 48987913.0,
      "step": 19320
    },
    {
      "entropy": 6.251869487762451,
      "epoch": 2.2302365839584537,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004510641422122811,
      "loss": 5.6611,
      "mean_token_accuracy": 0.19807603508234023,
      "num_tokens": 49001189.0,
      "step": 19325
    },
    {
      "entropy": 6.229005193710327,
      "epoch": 2.230813618003462,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00045103844635723005,
      "loss": 5.6058,
      "mean_token_accuracy": 0.20438386499881744,
      "num_tokens": 49013827.0,
      "step": 19330
    },
    {
      "entropy": 6.237604427337646,
      "epoch": 2.231390652048471,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045101274458121335,
      "loss": 5.6522,
      "mean_token_accuracy": 0.20513474494218825,
      "num_tokens": 49024939.0,
      "step": 19335
    },
    {
      "entropy": 6.289460372924805,
      "epoch": 2.2319676860934794,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00045098703688509537,
      "loss": 5.7463,
      "mean_token_accuracy": 0.19357877522706984,
      "num_tokens": 49037707.0,
      "step": 19340
    },
    {
      "entropy": 6.310045146942139,
      "epoch": 2.2325447201384883,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004509613232697408,
      "loss": 5.6369,
      "mean_token_accuracy": 0.2065887466073036,
      "num_tokens": 49051408.0,
      "step": 19345
    },
    {
      "entropy": 6.236478471755982,
      "epoch": 2.2331217541834967,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004509356037360145,
      "loss": 5.6113,
      "mean_token_accuracy": 0.2069323867559433,
      "num_tokens": 49063664.0,
      "step": 19350
    },
    {
      "entropy": 6.216299152374267,
      "epoch": 2.2336987882285055,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004509098782847816,
      "loss": 5.6737,
      "mean_token_accuracy": 0.20115928202867508,
      "num_tokens": 49076321.0,
      "step": 19355
    },
    {
      "entropy": 6.3085630416870115,
      "epoch": 2.234275822273514,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004508841469169073,
      "loss": 5.746,
      "mean_token_accuracy": 0.19430652707815171,
      "num_tokens": 49088794.0,
      "step": 19360
    },
    {
      "entropy": 6.359749794006348,
      "epoch": 2.234852856318523,
      "grad_norm": 1.0,
      "learning_rate": 0.00045085840963325716,
      "loss": 5.772,
      "mean_token_accuracy": 0.19175889492034912,
      "num_tokens": 49099986.0,
      "step": 19365
    },
    {
      "entropy": 6.19860463142395,
      "epoch": 2.2354298903635312,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004508326664346967,
      "loss": 5.6517,
      "mean_token_accuracy": 0.20582344233989716,
      "num_tokens": 49112811.0,
      "step": 19370
    },
    {
      "entropy": 6.198754024505615,
      "epoch": 2.23600692440854,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045080691732209176,
      "loss": 5.6776,
      "mean_token_accuracy": 0.20600303262472153,
      "num_tokens": 49124893.0,
      "step": 19375
    },
    {
      "entropy": 6.212229537963867,
      "epoch": 2.236583958453549,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045078116229630864,
      "loss": 5.6842,
      "mean_token_accuracy": 0.1988367110490799,
      "num_tokens": 49137239.0,
      "step": 19380
    },
    {
      "entropy": 6.341038846969605,
      "epoch": 2.2371609924985574,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045075540135821343,
      "loss": 5.7378,
      "mean_token_accuracy": 0.20010823756456375,
      "num_tokens": 49149876.0,
      "step": 19385
    },
    {
      "entropy": 6.308857870101929,
      "epoch": 2.2377380265435662,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004507296345086725,
      "loss": 5.6959,
      "mean_token_accuracy": 0.19724998623132706,
      "num_tokens": 49161599.0,
      "step": 19390
    },
    {
      "entropy": 6.281613540649414,
      "epoch": 2.2383150605885747,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004507038617485526,
      "loss": 5.7662,
      "mean_token_accuracy": 0.19550470858812333,
      "num_tokens": 49174278.0,
      "step": 19395
    },
    {
      "entropy": 6.296798896789551,
      "epoch": 2.2388920946335835,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045067808307872064,
      "loss": 5.6914,
      "mean_token_accuracy": 0.20638294667005538,
      "num_tokens": 49187922.0,
      "step": 19400
    },
    {
      "entropy": 6.309092903137207,
      "epoch": 2.239469128678592,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045065229850004365,
      "loss": 5.7194,
      "mean_token_accuracy": 0.19664821028709412,
      "num_tokens": 49200109.0,
      "step": 19405
    },
    {
      "entropy": 6.236276721954345,
      "epoch": 2.240046162723601,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004506265080133888,
      "loss": 5.7373,
      "mean_token_accuracy": 0.19400642514228822,
      "num_tokens": 49213192.0,
      "step": 19410
    },
    {
      "entropy": 6.315651941299438,
      "epoch": 2.240623196768609,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00045060071161962364,
      "loss": 5.7776,
      "mean_token_accuracy": 0.1890144631266594,
      "num_tokens": 49225439.0,
      "step": 19415
    },
    {
      "entropy": 6.258051013946533,
      "epoch": 2.241200230813618,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004505749093196158,
      "loss": 5.7244,
      "mean_token_accuracy": 0.20273986905813218,
      "num_tokens": 49240445.0,
      "step": 19420
    },
    {
      "entropy": 6.233918046951294,
      "epoch": 2.2417772648586265,
      "grad_norm": 1.015625,
      "learning_rate": 0.000450549101114233,
      "loss": 5.6702,
      "mean_token_accuracy": 0.19776078909635544,
      "num_tokens": 49253909.0,
      "step": 19425
    },
    {
      "entropy": 6.212040805816651,
      "epoch": 2.2423542989036354,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004505232870043434,
      "loss": 5.6195,
      "mean_token_accuracy": 0.20423374474048614,
      "num_tokens": 49266816.0,
      "step": 19430
    },
    {
      "entropy": 6.343074607849121,
      "epoch": 2.2429313329486438,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004504974669908152,
      "loss": 5.6879,
      "mean_token_accuracy": 0.2002343326807022,
      "num_tokens": 49279561.0,
      "step": 19435
    },
    {
      "entropy": 6.289524841308594,
      "epoch": 2.2435083669936526,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045047164107451696,
      "loss": 5.7861,
      "mean_token_accuracy": 0.19451519399881362,
      "num_tokens": 49292448.0,
      "step": 19440
    },
    {
      "entropy": 6.308961629867554,
      "epoch": 2.244085401038661,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004504458092563172,
      "loss": 5.7131,
      "mean_token_accuracy": 0.1998423829674721,
      "num_tokens": 49305078.0,
      "step": 19445
    },
    {
      "entropy": 6.2683573246002195,
      "epoch": 2.24466243508367,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045041997153708475,
      "loss": 5.6617,
      "mean_token_accuracy": 0.20156909376382828,
      "num_tokens": 49317353.0,
      "step": 19450
    },
    {
      "entropy": 6.2303284168243405,
      "epoch": 2.2452394691286788,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045039412791768877,
      "loss": 5.6986,
      "mean_token_accuracy": 0.2047604203224182,
      "num_tokens": 49329730.0,
      "step": 19455
    },
    {
      "entropy": 6.286963653564453,
      "epoch": 2.245816503173687,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045036827839899833,
      "loss": 5.6891,
      "mean_token_accuracy": 0.19605113863945006,
      "num_tokens": 49342656.0,
      "step": 19460
    },
    {
      "entropy": 6.18856110572815,
      "epoch": 2.246393537218696,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00045034242298188303,
      "loss": 5.5974,
      "mean_token_accuracy": 0.2118084564805031,
      "num_tokens": 49355216.0,
      "step": 19465
    },
    {
      "entropy": 6.250526714324951,
      "epoch": 2.2469705712637045,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004503165616672124,
      "loss": 5.6559,
      "mean_token_accuracy": 0.2012261062860489,
      "num_tokens": 49367897.0,
      "step": 19470
    },
    {
      "entropy": 6.270518684387207,
      "epoch": 2.2475476053087133,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004502906944558563,
      "loss": 5.661,
      "mean_token_accuracy": 0.19900657385587692,
      "num_tokens": 49380509.0,
      "step": 19475
    },
    {
      "entropy": 6.3380677700042725,
      "epoch": 2.2481246393537218,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004502648213486848,
      "loss": 5.8224,
      "mean_token_accuracy": 0.19129932522773743,
      "num_tokens": 49393832.0,
      "step": 19480
    },
    {
      "entropy": 6.323802518844604,
      "epoch": 2.2487016733987306,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045023894234656807,
      "loss": 5.7648,
      "mean_token_accuracy": 0.19538202732801438,
      "num_tokens": 49406568.0,
      "step": 19485
    },
    {
      "entropy": 6.273510837554932,
      "epoch": 2.249278707443739,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004502130574503766,
      "loss": 5.7049,
      "mean_token_accuracy": 0.19761135429143906,
      "num_tokens": 49419171.0,
      "step": 19490
    },
    {
      "entropy": 6.226946878433227,
      "epoch": 2.249855741488748,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004501871666609809,
      "loss": 5.6385,
      "mean_token_accuracy": 0.21178994327783585,
      "num_tokens": 49431177.0,
      "step": 19495
    },
    {
      "entropy": 6.207280921936035,
      "epoch": 2.2504327755337563,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004501612699792519,
      "loss": 5.6018,
      "mean_token_accuracy": 0.2038458079099655,
      "num_tokens": 49443403.0,
      "step": 19500
    },
    {
      "entropy": 6.311682033538818,
      "epoch": 2.251009809578765,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004501353674060606,
      "loss": 5.7853,
      "mean_token_accuracy": 0.1852878600358963,
      "num_tokens": 49456338.0,
      "step": 19505
    },
    {
      "entropy": 6.230403757095337,
      "epoch": 2.251586843623774,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004501094589422782,
      "loss": 5.6182,
      "mean_token_accuracy": 0.21316218078136445,
      "num_tokens": 49470946.0,
      "step": 19510
    },
    {
      "entropy": 6.246678447723388,
      "epoch": 2.2521638776687825,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045008354458877614,
      "loss": 5.665,
      "mean_token_accuracy": 0.20493778437376023,
      "num_tokens": 49483174.0,
      "step": 19515
    },
    {
      "entropy": 6.2527224063873295,
      "epoch": 2.252740911713791,
      "grad_norm": 1.0,
      "learning_rate": 0.000450057624346426,
      "loss": 5.6908,
      "mean_token_accuracy": 0.20189860761165618,
      "num_tokens": 49495099.0,
      "step": 19520
    },
    {
      "entropy": 6.257351112365723,
      "epoch": 2.2533179457587997,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045003169821609967,
      "loss": 5.6418,
      "mean_token_accuracy": 0.1989719659090042,
      "num_tokens": 49507200.0,
      "step": 19525
    },
    {
      "entropy": 6.262548446655273,
      "epoch": 2.2538949798038086,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045000576619866914,
      "loss": 5.7633,
      "mean_token_accuracy": 0.19742924422025682,
      "num_tokens": 49520093.0,
      "step": 19530
    },
    {
      "entropy": 6.2344118595123295,
      "epoch": 2.254472013848817,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044997982829500657,
      "loss": 5.6889,
      "mean_token_accuracy": 0.20035693943500518,
      "num_tokens": 49532190.0,
      "step": 19535
    },
    {
      "entropy": 6.226594877243042,
      "epoch": 2.255049047893826,
      "grad_norm": 0.921875,
      "learning_rate": 0.00044995388450598436,
      "loss": 5.6616,
      "mean_token_accuracy": 0.20882945209741594,
      "num_tokens": 49544468.0,
      "step": 19540
    },
    {
      "entropy": 6.185228109359741,
      "epoch": 2.2556260819388343,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004499279348324751,
      "loss": 5.6292,
      "mean_token_accuracy": 0.20441656261682511,
      "num_tokens": 49556753.0,
      "step": 19545
    },
    {
      "entropy": 6.316551685333252,
      "epoch": 2.256203115983843,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044990197927535173,
      "loss": 5.6984,
      "mean_token_accuracy": 0.20404880940914155,
      "num_tokens": 49567756.0,
      "step": 19550
    },
    {
      "entropy": 6.255786752700805,
      "epoch": 2.2567801500288516,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044987601783548703,
      "loss": 5.6875,
      "mean_token_accuracy": 0.2036540225148201,
      "num_tokens": 49579996.0,
      "step": 19555
    },
    {
      "entropy": 6.289878511428833,
      "epoch": 2.2573571840738604,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004498500505137545,
      "loss": 5.734,
      "mean_token_accuracy": 0.20217235386371613,
      "num_tokens": 49593407.0,
      "step": 19560
    },
    {
      "entropy": 6.3424074172973635,
      "epoch": 2.257934218118869,
      "grad_norm": 1.0,
      "learning_rate": 0.0004498240773110273,
      "loss": 5.743,
      "mean_token_accuracy": 0.19768321216106416,
      "num_tokens": 49606231.0,
      "step": 19565
    },
    {
      "entropy": 6.327963733673096,
      "epoch": 2.2585112521638777,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004497980982281791,
      "loss": 5.7184,
      "mean_token_accuracy": 0.1943321332335472,
      "num_tokens": 49616860.0,
      "step": 19570
    },
    {
      "entropy": 6.215011167526245,
      "epoch": 2.259088286208886,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004497721132660837,
      "loss": 5.625,
      "mean_token_accuracy": 0.2035585656762123,
      "num_tokens": 49630415.0,
      "step": 19575
    },
    {
      "entropy": 6.238117361068726,
      "epoch": 2.259665320253895,
      "grad_norm": 0.921875,
      "learning_rate": 0.00044974612242561516,
      "loss": 5.7071,
      "mean_token_accuracy": 0.19392564594745637,
      "num_tokens": 49643173.0,
      "step": 19580
    },
    {
      "entropy": 6.368253517150879,
      "epoch": 2.260242354298904,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004497201257076475,
      "loss": 5.7588,
      "mean_token_accuracy": 0.19694225192070008,
      "num_tokens": 49656670.0,
      "step": 19585
    },
    {
      "entropy": 6.267241716384888,
      "epoch": 2.2608193883439123,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004496941231130552,
      "loss": 5.709,
      "mean_token_accuracy": 0.19851476401090623,
      "num_tokens": 49669572.0,
      "step": 19590
    },
    {
      "entropy": 6.286822700500489,
      "epoch": 2.261396422388921,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004496681146427129,
      "loss": 5.7328,
      "mean_token_accuracy": 0.2006918102502823,
      "num_tokens": 49682253.0,
      "step": 19595
    },
    {
      "entropy": 6.241611909866333,
      "epoch": 2.2619734564339296,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00044964210029749523,
      "loss": 5.6095,
      "mean_token_accuracy": 0.21050963252782823,
      "num_tokens": 49695284.0,
      "step": 19600
    },
    {
      "entropy": 6.233285427093506,
      "epoch": 2.2625504904789384,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044961608007827736,
      "loss": 5.6929,
      "mean_token_accuracy": 0.19603051096200944,
      "num_tokens": 49708039.0,
      "step": 19605
    },
    {
      "entropy": 6.25356092453003,
      "epoch": 2.263127524523947,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004495900539859344,
      "loss": 5.6476,
      "mean_token_accuracy": 0.21425366699695586,
      "num_tokens": 49722182.0,
      "step": 19610
    },
    {
      "entropy": 6.3008325576782225,
      "epoch": 2.2637045585689557,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044956402202134157,
      "loss": 5.698,
      "mean_token_accuracy": 0.2030737057328224,
      "num_tokens": 49733875.0,
      "step": 19615
    },
    {
      "entropy": 6.085651683807373,
      "epoch": 2.264281592613964,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044953798418537465,
      "loss": 5.5034,
      "mean_token_accuracy": 0.21815079301595688,
      "num_tokens": 49746701.0,
      "step": 19620
    },
    {
      "entropy": 6.194913244247436,
      "epoch": 2.264858626658973,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004495119404789093,
      "loss": 5.6109,
      "mean_token_accuracy": 0.2007535830140114,
      "num_tokens": 49760636.0,
      "step": 19625
    },
    {
      "entropy": 6.331758117675781,
      "epoch": 2.2654356607039814,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004494858909028216,
      "loss": 5.7249,
      "mean_token_accuracy": 0.19530639350414275,
      "num_tokens": 49773655.0,
      "step": 19630
    },
    {
      "entropy": 6.270000410079956,
      "epoch": 2.2660126947489903,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004494598354579875,
      "loss": 5.7602,
      "mean_token_accuracy": 0.1903955489397049,
      "num_tokens": 49787575.0,
      "step": 19635
    },
    {
      "entropy": 6.2434831142425535,
      "epoch": 2.2665897287939987,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004494337741452836,
      "loss": 5.7161,
      "mean_token_accuracy": 0.19572321474552154,
      "num_tokens": 49802021.0,
      "step": 19640
    },
    {
      "entropy": 6.318602561950684,
      "epoch": 2.2671667628390075,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004494077069655863,
      "loss": 5.6602,
      "mean_token_accuracy": 0.20091204196214676,
      "num_tokens": 49813640.0,
      "step": 19645
    },
    {
      "entropy": 6.293531608581543,
      "epoch": 2.267743796884016,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004493816339197724,
      "loss": 5.7719,
      "mean_token_accuracy": 0.1956032022833824,
      "num_tokens": 49826318.0,
      "step": 19650
    },
    {
      "entropy": 6.255597972869873,
      "epoch": 2.268320830929025,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044935555500871897,
      "loss": 5.7271,
      "mean_token_accuracy": 0.19798053801059723,
      "num_tokens": 49839278.0,
      "step": 19655
    },
    {
      "entropy": 6.321634101867676,
      "epoch": 2.2688978649740337,
      "grad_norm": 1.0,
      "learning_rate": 0.000449329470233303,
      "loss": 5.7563,
      "mean_token_accuracy": 0.1946269005537033,
      "num_tokens": 49851417.0,
      "step": 19660
    },
    {
      "entropy": 6.279968643188477,
      "epoch": 2.269474899019042,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044930337959440183,
      "loss": 5.648,
      "mean_token_accuracy": 0.2031923606991768,
      "num_tokens": 49863797.0,
      "step": 19665
    },
    {
      "entropy": 6.26726016998291,
      "epoch": 2.270051933064051,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004492772830928931,
      "loss": 5.6988,
      "mean_token_accuracy": 0.19734385758638381,
      "num_tokens": 49877767.0,
      "step": 19670
    },
    {
      "entropy": 6.295455980300903,
      "epoch": 2.2706289671090594,
      "grad_norm": 1.0,
      "learning_rate": 0.00044925118072965456,
      "loss": 5.7132,
      "mean_token_accuracy": 0.20181403011083604,
      "num_tokens": 49890301.0,
      "step": 19675
    },
    {
      "entropy": 6.25642352104187,
      "epoch": 2.2712060011540682,
      "grad_norm": 0.90625,
      "learning_rate": 0.000449225072505564,
      "loss": 5.7153,
      "mean_token_accuracy": 0.2006752610206604,
      "num_tokens": 49902095.0,
      "step": 19680
    },
    {
      "entropy": 6.307287263870239,
      "epoch": 2.2717830351990767,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044919895842149976,
      "loss": 5.7282,
      "mean_token_accuracy": 0.19565233141183852,
      "num_tokens": 49914764.0,
      "step": 19685
    },
    {
      "entropy": 6.254042530059815,
      "epoch": 2.2723600692440855,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044917283847834005,
      "loss": 5.6757,
      "mean_token_accuracy": 0.20074882358312607,
      "num_tokens": 49927729.0,
      "step": 19690
    },
    {
      "entropy": 6.259390258789063,
      "epoch": 2.272937103289094,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004491467126769635,
      "loss": 5.6741,
      "mean_token_accuracy": 0.2071201264858246,
      "num_tokens": 49940017.0,
      "step": 19695
    },
    {
      "entropy": 6.308801794052124,
      "epoch": 2.273514137334103,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044912058101824866,
      "loss": 5.7805,
      "mean_token_accuracy": 0.19596584737300873,
      "num_tokens": 49951894.0,
      "step": 19700
    },
    {
      "entropy": 6.169390439987183,
      "epoch": 2.274091171379111,
      "grad_norm": 1.0,
      "learning_rate": 0.00044909444350307463,
      "loss": 5.5492,
      "mean_token_accuracy": 0.2183400422334671,
      "num_tokens": 49965152.0,
      "step": 19705
    },
    {
      "entropy": 6.247048091888428,
      "epoch": 2.27466820542412,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004490683001323205,
      "loss": 5.7076,
      "mean_token_accuracy": 0.1969557896256447,
      "num_tokens": 49978713.0,
      "step": 19710
    },
    {
      "entropy": 6.283925628662109,
      "epoch": 2.2752452394691285,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044904215090686554,
      "loss": 5.7397,
      "mean_token_accuracy": 0.1911838099360466,
      "num_tokens": 49990090.0,
      "step": 19715
    },
    {
      "entropy": 6.302199125289917,
      "epoch": 2.2758222735141374,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044901599582758926,
      "loss": 5.7276,
      "mean_token_accuracy": 0.20146616399288178,
      "num_tokens": 50002205.0,
      "step": 19720
    },
    {
      "entropy": 6.212772417068481,
      "epoch": 2.2763993075591458,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044898983489537143,
      "loss": 5.6228,
      "mean_token_accuracy": 0.20220943987369538,
      "num_tokens": 50014274.0,
      "step": 19725
    },
    {
      "entropy": 6.306591510772705,
      "epoch": 2.2769763416041546,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000448963668111092,
      "loss": 5.8135,
      "mean_token_accuracy": 0.18814051896333694,
      "num_tokens": 50026789.0,
      "step": 19730
    },
    {
      "entropy": 6.3241089344024655,
      "epoch": 2.2775533756491635,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00044893749547563085,
      "loss": 5.6928,
      "mean_token_accuracy": 0.20344648510217667,
      "num_tokens": 50039615.0,
      "step": 19735
    },
    {
      "entropy": 6.256423139572144,
      "epoch": 2.278130409694172,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044891131698986846,
      "loss": 5.6959,
      "mean_token_accuracy": 0.20017684400081634,
      "num_tokens": 50051799.0,
      "step": 19740
    },
    {
      "entropy": 6.257806253433228,
      "epoch": 2.2787074437391808,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004488851326546854,
      "loss": 5.7068,
      "mean_token_accuracy": 0.2040300875902176,
      "num_tokens": 50064774.0,
      "step": 19745
    },
    {
      "entropy": 6.328410863876343,
      "epoch": 2.279284477784189,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004488589424709622,
      "loss": 5.773,
      "mean_token_accuracy": 0.19195556193590163,
      "num_tokens": 50076631.0,
      "step": 19750
    },
    {
      "entropy": 6.238110780715942,
      "epoch": 2.279861511829198,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004488327464395799,
      "loss": 5.6039,
      "mean_token_accuracy": 0.20832854956388475,
      "num_tokens": 50089493.0,
      "step": 19755
    },
    {
      "entropy": 6.27990756034851,
      "epoch": 2.2804385458742065,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004488065445614195,
      "loss": 5.6937,
      "mean_token_accuracy": 0.20207191109657288,
      "num_tokens": 50101412.0,
      "step": 19760
    },
    {
      "entropy": 6.296133375167846,
      "epoch": 2.2810155799192153,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004487803368373623,
      "loss": 5.6869,
      "mean_token_accuracy": 0.20134456157684327,
      "num_tokens": 50113634.0,
      "step": 19765
    },
    {
      "entropy": 6.273417901992798,
      "epoch": 2.2815926139642237,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004487541232682898,
      "loss": 5.7141,
      "mean_token_accuracy": 0.1970142975449562,
      "num_tokens": 50125771.0,
      "step": 19770
    },
    {
      "entropy": 6.278803586959839,
      "epoch": 2.2821696480092326,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004487279038550836,
      "loss": 5.7043,
      "mean_token_accuracy": 0.2009200781583786,
      "num_tokens": 50138260.0,
      "step": 19775
    },
    {
      "entropy": 6.2721024513244625,
      "epoch": 2.282746682054241,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004487016785986258,
      "loss": 5.6979,
      "mean_token_accuracy": 0.20853228121995926,
      "num_tokens": 50150813.0,
      "step": 19780
    },
    {
      "entropy": 6.233594274520874,
      "epoch": 2.28332371609925,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044867544749979817,
      "loss": 5.6395,
      "mean_token_accuracy": 0.20502638667821885,
      "num_tokens": 50163775.0,
      "step": 19785
    },
    {
      "entropy": 6.265291547775268,
      "epoch": 2.2839007501442588,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044864921055948317,
      "loss": 5.6576,
      "mean_token_accuracy": 0.1998529314994812,
      "num_tokens": 50175697.0,
      "step": 19790
    },
    {
      "entropy": 6.281085014343262,
      "epoch": 2.284477784189267,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044862296777856326,
      "loss": 5.7384,
      "mean_token_accuracy": 0.20307090878486633,
      "num_tokens": 50187826.0,
      "step": 19795
    },
    {
      "entropy": 6.313817167282105,
      "epoch": 2.2850548182342756,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004485967191579211,
      "loss": 5.7388,
      "mean_token_accuracy": 0.2009593978524208,
      "num_tokens": 50201611.0,
      "step": 19800
    },
    {
      "entropy": 6.293828868865967,
      "epoch": 2.2856318522792844,
      "grad_norm": 1.0,
      "learning_rate": 0.0004485704646984394,
      "loss": 5.6692,
      "mean_token_accuracy": 0.19924700409173965,
      "num_tokens": 50213341.0,
      "step": 19805
    },
    {
      "entropy": 6.2259186744689945,
      "epoch": 2.2862088863242933,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004485442044010015,
      "loss": 5.6911,
      "mean_token_accuracy": 0.1970667615532875,
      "num_tokens": 50225165.0,
      "step": 19810
    },
    {
      "entropy": 6.233582162857056,
      "epoch": 2.2867859203693017,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004485179382664904,
      "loss": 5.599,
      "mean_token_accuracy": 0.20216548442840576,
      "num_tokens": 50238207.0,
      "step": 19815
    },
    {
      "entropy": 6.301874208450317,
      "epoch": 2.2873629544143106,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004484916662957896,
      "loss": 5.6052,
      "mean_token_accuracy": 0.20451566129922866,
      "num_tokens": 50250347.0,
      "step": 19820
    },
    {
      "entropy": 6.238548040390015,
      "epoch": 2.287939988459319,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004484653884897829,
      "loss": 5.7187,
      "mean_token_accuracy": 0.19722494035959243,
      "num_tokens": 50263487.0,
      "step": 19825
    },
    {
      "entropy": 6.215698003768921,
      "epoch": 2.288517022504328,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044843910484935394,
      "loss": 5.6326,
      "mean_token_accuracy": 0.20198740810155869,
      "num_tokens": 50275020.0,
      "step": 19830
    },
    {
      "entropy": 6.2797911167144775,
      "epoch": 2.2890940565493363,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004484128153753868,
      "loss": 5.7391,
      "mean_token_accuracy": 0.20075925886631013,
      "num_tokens": 50287888.0,
      "step": 19835
    },
    {
      "entropy": 6.313848495483398,
      "epoch": 2.289671090594345,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00044838652006876583,
      "loss": 5.72,
      "mean_token_accuracy": 0.19894591271877288,
      "num_tokens": 50301271.0,
      "step": 19840
    },
    {
      "entropy": 6.304997444152832,
      "epoch": 2.2902481246393536,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00044836021893037537,
      "loss": 5.7464,
      "mean_token_accuracy": 0.20201318860054016,
      "num_tokens": 50312707.0,
      "step": 19845
    },
    {
      "entropy": 6.277853107452392,
      "epoch": 2.2908251586843624,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004483339119611001,
      "loss": 5.7809,
      "mean_token_accuracy": 0.19293169677257538,
      "num_tokens": 50325399.0,
      "step": 19850
    },
    {
      "entropy": 6.279010677337647,
      "epoch": 2.291402192729371,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044830759916182476,
      "loss": 5.6746,
      "mean_token_accuracy": 0.19894415885210037,
      "num_tokens": 50338424.0,
      "step": 19855
    },
    {
      "entropy": 6.239544439315796,
      "epoch": 2.2919792267743797,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004482812805334344,
      "loss": 5.6426,
      "mean_token_accuracy": 0.19835784137248993,
      "num_tokens": 50351183.0,
      "step": 19860
    },
    {
      "entropy": 6.261954116821289,
      "epoch": 2.2925562608193886,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004482549560768142,
      "loss": 5.72,
      "mean_token_accuracy": 0.1977445363998413,
      "num_tokens": 50364223.0,
      "step": 19865
    },
    {
      "entropy": 6.307745361328125,
      "epoch": 2.293133294864397,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004482286257928497,
      "loss": 5.729,
      "mean_token_accuracy": 0.19250791668891906,
      "num_tokens": 50376867.0,
      "step": 19870
    },
    {
      "entropy": 6.296713733673096,
      "epoch": 2.293710328909406,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004482022896824263,
      "loss": 5.6537,
      "mean_token_accuracy": 0.20242024809122086,
      "num_tokens": 50389180.0,
      "step": 19875
    },
    {
      "entropy": 6.25498399734497,
      "epoch": 2.2942873629544143,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044817594774643004,
      "loss": 5.7172,
      "mean_token_accuracy": 0.2000199407339096,
      "num_tokens": 50401286.0,
      "step": 19880
    },
    {
      "entropy": 6.247745323181152,
      "epoch": 2.294864396999423,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044814959998574675,
      "loss": 5.7423,
      "mean_token_accuracy": 0.19842909425497054,
      "num_tokens": 50414215.0,
      "step": 19885
    },
    {
      "entropy": 6.307348585128784,
      "epoch": 2.2954414310444315,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044812324640126265,
      "loss": 5.6405,
      "mean_token_accuracy": 0.20206768363714217,
      "num_tokens": 50425672.0,
      "step": 19890
    },
    {
      "entropy": 6.291373062133789,
      "epoch": 2.2960184650894404,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004480968869938642,
      "loss": 5.7125,
      "mean_token_accuracy": 0.1984001338481903,
      "num_tokens": 50438578.0,
      "step": 19895
    },
    {
      "entropy": 6.244321870803833,
      "epoch": 2.296595499134449,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044807052176443793,
      "loss": 5.6533,
      "mean_token_accuracy": 0.2029689520597458,
      "num_tokens": 50451762.0,
      "step": 19900
    },
    {
      "entropy": 6.27501335144043,
      "epoch": 2.2971725331794577,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044804415071387067,
      "loss": 5.6743,
      "mean_token_accuracy": 0.20273153632879257,
      "num_tokens": 50464064.0,
      "step": 19905
    },
    {
      "entropy": 6.282047367095947,
      "epoch": 2.297749567224466,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004480177738430492,
      "loss": 5.7115,
      "mean_token_accuracy": 0.1993885740637779,
      "num_tokens": 50476843.0,
      "step": 19910
    },
    {
      "entropy": 6.2680799007415775,
      "epoch": 2.298326601269475,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044799139115286104,
      "loss": 5.6651,
      "mean_token_accuracy": 0.1962010845541954,
      "num_tokens": 50488209.0,
      "step": 19915
    },
    {
      "entropy": 6.321750783920288,
      "epoch": 2.2989036353144834,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004479650026441933,
      "loss": 5.7311,
      "mean_token_accuracy": 0.19435369074344636,
      "num_tokens": 50500292.0,
      "step": 19920
    },
    {
      "entropy": 6.3289227962493895,
      "epoch": 2.2994806693594922,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044793860831793356,
      "loss": 5.7237,
      "mean_token_accuracy": 0.1999133050441742,
      "num_tokens": 50512563.0,
      "step": 19925
    },
    {
      "entropy": 6.322342205047607,
      "epoch": 2.3000577034045007,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044791220817496963,
      "loss": 5.6868,
      "mean_token_accuracy": 0.20065617710351943,
      "num_tokens": 50525614.0,
      "step": 19930
    },
    {
      "entropy": 6.2436662197113035,
      "epoch": 2.3006347374495095,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004478858022161895,
      "loss": 5.585,
      "mean_token_accuracy": 0.20625341236591338,
      "num_tokens": 50537387.0,
      "step": 19935
    },
    {
      "entropy": 6.210444450378418,
      "epoch": 2.3012117714945184,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004478593904424813,
      "loss": 5.6607,
      "mean_token_accuracy": 0.2027381181716919,
      "num_tokens": 50549588.0,
      "step": 19940
    },
    {
      "entropy": 6.296148824691772,
      "epoch": 2.301788805539527,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004478329728547334,
      "loss": 5.8402,
      "mean_token_accuracy": 0.18521838039159774,
      "num_tokens": 50562276.0,
      "step": 19945
    },
    {
      "entropy": 6.305137872695923,
      "epoch": 2.3023658395845357,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044780654945383424,
      "loss": 5.7156,
      "mean_token_accuracy": 0.19944151043891906,
      "num_tokens": 50575650.0,
      "step": 19950
    },
    {
      "entropy": 6.315970945358276,
      "epoch": 2.302942873629544,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044778012024067267,
      "loss": 5.7711,
      "mean_token_accuracy": 0.18741468489170074,
      "num_tokens": 50587611.0,
      "step": 19955
    },
    {
      "entropy": 6.297332143783569,
      "epoch": 2.303519907674553,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004477536852161376,
      "loss": 5.7047,
      "mean_token_accuracy": 0.1925889655947685,
      "num_tokens": 50601001.0,
      "step": 19960
    },
    {
      "entropy": 6.341242933273316,
      "epoch": 2.3040969417195614,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004477272443811181,
      "loss": 5.7355,
      "mean_token_accuracy": 0.1954132691025734,
      "num_tokens": 50613539.0,
      "step": 19965
    },
    {
      "entropy": 6.211915349960327,
      "epoch": 2.3046739757645702,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004477007977365035,
      "loss": 5.5991,
      "mean_token_accuracy": 0.20108244121074675,
      "num_tokens": 50626536.0,
      "step": 19970
    },
    {
      "entropy": 6.268716096878052,
      "epoch": 2.3052510098095786,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004476743452831834,
      "loss": 5.6753,
      "mean_token_accuracy": 0.19727351367473603,
      "num_tokens": 50638617.0,
      "step": 19975
    },
    {
      "entropy": 6.320518207550049,
      "epoch": 2.3058280438545875,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044764788702204747,
      "loss": 5.7535,
      "mean_token_accuracy": 0.19708998650312423,
      "num_tokens": 50651779.0,
      "step": 19980
    },
    {
      "entropy": 6.254730319976806,
      "epoch": 2.306405077899596,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004476214229539856,
      "loss": 5.6438,
      "mean_token_accuracy": 0.2050560235977173,
      "num_tokens": 50665205.0,
      "step": 19985
    },
    {
      "entropy": 6.32389464378357,
      "epoch": 2.306982111944605,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004475949530798879,
      "loss": 5.7355,
      "mean_token_accuracy": 0.19657264947891234,
      "num_tokens": 50677621.0,
      "step": 19990
    },
    {
      "entropy": 6.302538108825684,
      "epoch": 2.307559145989613,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044756847740064475,
      "loss": 5.7568,
      "mean_token_accuracy": 0.19099175930023193,
      "num_tokens": 50689544.0,
      "step": 19995
    },
    {
      "entropy": 6.3180803775787355,
      "epoch": 2.308136180034622,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004475419959171465,
      "loss": 5.692,
      "mean_token_accuracy": 0.204245026409626,
      "num_tokens": 50702733.0,
      "step": 20000
    },
    {
      "entropy": 6.257320261001587,
      "epoch": 2.3087132140796305,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000447515508630284,
      "loss": 5.6611,
      "mean_token_accuracy": 0.20171004980802537,
      "num_tokens": 50716296.0,
      "step": 20005
    },
    {
      "entropy": 6.233843994140625,
      "epoch": 2.3092902481246393,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044748901554094806,
      "loss": 5.6172,
      "mean_token_accuracy": 0.20525160431861877,
      "num_tokens": 50729170.0,
      "step": 20010
    },
    {
      "entropy": 6.291717863082885,
      "epoch": 2.309867282169648,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004474625166500297,
      "loss": 5.6872,
      "mean_token_accuracy": 0.20180599987506867,
      "num_tokens": 50741400.0,
      "step": 20015
    },
    {
      "entropy": 6.283157777786255,
      "epoch": 2.3104443162146566,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044743601195842026,
      "loss": 5.7208,
      "mean_token_accuracy": 0.19231373220682144,
      "num_tokens": 50755231.0,
      "step": 20020
    },
    {
      "entropy": 6.318492126464844,
      "epoch": 2.3110213502596655,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044740950146701127,
      "loss": 5.7499,
      "mean_token_accuracy": 0.1937901794910431,
      "num_tokens": 50768044.0,
      "step": 20025
    },
    {
      "entropy": 6.338865041732788,
      "epoch": 2.311598384304674,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004473829851766943,
      "loss": 5.7737,
      "mean_token_accuracy": 0.18798430263996124,
      "num_tokens": 50780693.0,
      "step": 20030
    },
    {
      "entropy": 6.263759326934815,
      "epoch": 2.3121754183496828,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004473564630883612,
      "loss": 5.6374,
      "mean_token_accuracy": 0.20116141587495803,
      "num_tokens": 50792829.0,
      "step": 20035
    },
    {
      "entropy": 6.204412269592285,
      "epoch": 2.312752452394691,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004473299352029042,
      "loss": 5.616,
      "mean_token_accuracy": 0.20639759600162505,
      "num_tokens": 50806045.0,
      "step": 20040
    },
    {
      "entropy": 6.241959762573242,
      "epoch": 2.3133294864397,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004473034015212154,
      "loss": 5.7087,
      "mean_token_accuracy": 0.2030839815735817,
      "num_tokens": 50818789.0,
      "step": 20045
    },
    {
      "entropy": 6.351341724395752,
      "epoch": 2.3139065204847085,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004472768620441872,
      "loss": 5.7615,
      "mean_token_accuracy": 0.1937183603644371,
      "num_tokens": 50830681.0,
      "step": 20050
    },
    {
      "entropy": 6.231671953201294,
      "epoch": 2.3144835545297173,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00044725031677271234,
      "loss": 5.6369,
      "mean_token_accuracy": 0.2064118891954422,
      "num_tokens": 50843162.0,
      "step": 20055
    },
    {
      "entropy": 6.186645603179931,
      "epoch": 2.3150605885747257,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004472237657076837,
      "loss": 5.6871,
      "mean_token_accuracy": 0.20377921164035798,
      "num_tokens": 50855339.0,
      "step": 20060
    },
    {
      "entropy": 6.333691072463989,
      "epoch": 2.3156376226197346,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004471972088499942,
      "loss": 5.7499,
      "mean_token_accuracy": 0.19801906049251555,
      "num_tokens": 50868885.0,
      "step": 20065
    },
    {
      "entropy": 6.25673656463623,
      "epoch": 2.3162146566647435,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004471706462005371,
      "loss": 5.6603,
      "mean_token_accuracy": 0.20235869139432908,
      "num_tokens": 50881027.0,
      "step": 20070
    },
    {
      "entropy": 6.223751640319824,
      "epoch": 2.316791690709752,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004471440777602059,
      "loss": 5.6324,
      "mean_token_accuracy": 0.20869249999523162,
      "num_tokens": 50893717.0,
      "step": 20075
    },
    {
      "entropy": 6.32664384841919,
      "epoch": 2.3173687247547603,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044711750352989407,
      "loss": 5.7508,
      "mean_token_accuracy": 0.19796873033046722,
      "num_tokens": 50904965.0,
      "step": 20080
    },
    {
      "entropy": 6.271141862869262,
      "epoch": 2.317945758799769,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004470909235104956,
      "loss": 5.7287,
      "mean_token_accuracy": 0.2037908047437668,
      "num_tokens": 50917804.0,
      "step": 20085
    },
    {
      "entropy": 6.232886028289795,
      "epoch": 2.318522792844778,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004470643377029043,
      "loss": 5.6957,
      "mean_token_accuracy": 0.19649168103933334,
      "num_tokens": 50931124.0,
      "step": 20090
    },
    {
      "entropy": 6.243575572967529,
      "epoch": 2.3190998268897864,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004470377461080145,
      "loss": 5.6507,
      "mean_token_accuracy": 0.2034371703863144,
      "num_tokens": 50942983.0,
      "step": 20095
    },
    {
      "entropy": 6.31448884010315,
      "epoch": 2.3196768609347953,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004470111487267206,
      "loss": 5.7817,
      "mean_token_accuracy": 0.19622427225112915,
      "num_tokens": 50956282.0,
      "step": 20100
    },
    {
      "entropy": 6.365381908416748,
      "epoch": 2.3202538949798037,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004469845455599171,
      "loss": 5.7463,
      "mean_token_accuracy": 0.19253993928432464,
      "num_tokens": 50968112.0,
      "step": 20105
    },
    {
      "entropy": 6.164116525650025,
      "epoch": 2.3208309290248126,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004469579366084989,
      "loss": 5.5126,
      "mean_token_accuracy": 0.21525044590234757,
      "num_tokens": 50980377.0,
      "step": 20110
    },
    {
      "entropy": 6.294187116622925,
      "epoch": 2.321407963069821,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00044693132187336094,
      "loss": 5.7256,
      "mean_token_accuracy": 0.19777166247367858,
      "num_tokens": 50993371.0,
      "step": 20115
    },
    {
      "entropy": 6.309091472625733,
      "epoch": 2.32198499711483,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004469047013553983,
      "loss": 5.7302,
      "mean_token_accuracy": 0.1962558463215828,
      "num_tokens": 51004409.0,
      "step": 20120
    },
    {
      "entropy": 6.194480562210083,
      "epoch": 2.3225620311598383,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00044687807505550646,
      "loss": 5.5688,
      "mean_token_accuracy": 0.21419651210308074,
      "num_tokens": 51017890.0,
      "step": 20125
    },
    {
      "entropy": 6.298191499710083,
      "epoch": 2.323139065204847,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044685144297458096,
      "loss": 5.7122,
      "mean_token_accuracy": 0.19731850326061248,
      "num_tokens": 51029734.0,
      "step": 20130
    },
    {
      "entropy": 6.277208948135376,
      "epoch": 2.3237160992498556,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044682480511351754,
      "loss": 5.6803,
      "mean_token_accuracy": 0.20674144923686982,
      "num_tokens": 51041876.0,
      "step": 20135
    },
    {
      "entropy": 6.244891357421875,
      "epoch": 2.3242931332948644,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004467981614732121,
      "loss": 5.706,
      "mean_token_accuracy": 0.19904226660728455,
      "num_tokens": 51054603.0,
      "step": 20140
    },
    {
      "entropy": 6.287162208557129,
      "epoch": 2.3248701673398733,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044677151205456086,
      "loss": 5.6687,
      "mean_token_accuracy": 0.20424564331769943,
      "num_tokens": 51067283.0,
      "step": 20145
    },
    {
      "entropy": 6.290923690795898,
      "epoch": 2.3254472013848817,
      "grad_norm": 1.125,
      "learning_rate": 0.0004467448568584601,
      "loss": 5.7577,
      "mean_token_accuracy": 0.19339394867420195,
      "num_tokens": 51079841.0,
      "step": 20150
    },
    {
      "entropy": 6.168074464797973,
      "epoch": 2.32602423542989,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004467181958858064,
      "loss": 5.5699,
      "mean_token_accuracy": 0.21571094393730164,
      "num_tokens": 51091873.0,
      "step": 20155
    },
    {
      "entropy": 6.317538452148438,
      "epoch": 2.326601269474899,
      "grad_norm": 1.0,
      "learning_rate": 0.0004466915291374965,
      "loss": 5.7323,
      "mean_token_accuracy": 0.19878080040216445,
      "num_tokens": 51104955.0,
      "step": 20160
    },
    {
      "entropy": 6.324763774871826,
      "epoch": 2.327178303519908,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004466648566144273,
      "loss": 5.7337,
      "mean_token_accuracy": 0.197710981965065,
      "num_tokens": 51117699.0,
      "step": 20165
    },
    {
      "entropy": 6.270043134689331,
      "epoch": 2.3277553375649163,
      "grad_norm": 1.09375,
      "learning_rate": 0.000446638178317496,
      "loss": 5.6704,
      "mean_token_accuracy": 0.20942886620759965,
      "num_tokens": 51130322.0,
      "step": 20170
    },
    {
      "entropy": 6.294571208953857,
      "epoch": 2.328332371609925,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044661149424759974,
      "loss": 5.753,
      "mean_token_accuracy": 0.19835832267999648,
      "num_tokens": 51144685.0,
      "step": 20175
    },
    {
      "entropy": 6.332891178131104,
      "epoch": 2.3289094056549335,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004465848044056361,
      "loss": 5.7316,
      "mean_token_accuracy": 0.1950002521276474,
      "num_tokens": 51157653.0,
      "step": 20180
    },
    {
      "entropy": 6.327544116973877,
      "epoch": 2.3294864396999424,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004465581087925028,
      "loss": 5.7268,
      "mean_token_accuracy": 0.1993822678923607,
      "num_tokens": 51171314.0,
      "step": 20185
    },
    {
      "entropy": 6.215159940719604,
      "epoch": 2.330063473744951,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004465314074090978,
      "loss": 5.611,
      "mean_token_accuracy": 0.20631832033395767,
      "num_tokens": 51184261.0,
      "step": 20190
    },
    {
      "entropy": 6.33493218421936,
      "epoch": 2.3306405077899597,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044650470025631904,
      "loss": 5.8089,
      "mean_token_accuracy": 0.19799033552408218,
      "num_tokens": 51195908.0,
      "step": 20195
    },
    {
      "entropy": 6.3337541103363035,
      "epoch": 2.331217541834968,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004464779873350649,
      "loss": 5.7497,
      "mean_token_accuracy": 0.19071601033210756,
      "num_tokens": 51208275.0,
      "step": 20200
    },
    {
      "entropy": 6.294582366943359,
      "epoch": 2.331794575879977,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004464512686462339,
      "loss": 5.6822,
      "mean_token_accuracy": 0.1966435730457306,
      "num_tokens": 51221331.0,
      "step": 20205
    },
    {
      "entropy": 6.242805004119873,
      "epoch": 2.3323716099249854,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044642454419072455,
      "loss": 5.6757,
      "mean_token_accuracy": 0.20696126967668532,
      "num_tokens": 51233875.0,
      "step": 20210
    },
    {
      "entropy": 6.287078094482422,
      "epoch": 2.3329486439699942,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004463978139694358,
      "loss": 5.7171,
      "mean_token_accuracy": 0.20364162176847458,
      "num_tokens": 51246397.0,
      "step": 20215
    },
    {
      "entropy": 6.280299711227417,
      "epoch": 2.333525678015003,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044637107798326675,
      "loss": 5.696,
      "mean_token_accuracy": 0.19791121035814285,
      "num_tokens": 51260823.0,
      "step": 20220
    },
    {
      "entropy": 6.272793006896973,
      "epoch": 2.3341027120600115,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004463443362331166,
      "loss": 5.6366,
      "mean_token_accuracy": 0.20347680002450944,
      "num_tokens": 51273093.0,
      "step": 20225
    },
    {
      "entropy": 6.293143081665039,
      "epoch": 2.3346797461050204,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044631758871988486,
      "loss": 5.6875,
      "mean_token_accuracy": 0.20134591162204743,
      "num_tokens": 51284971.0,
      "step": 20230
    },
    {
      "entropy": 6.306660747528076,
      "epoch": 2.335256780150029,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004462908354444711,
      "loss": 5.7799,
      "mean_token_accuracy": 0.19994009137153626,
      "num_tokens": 51296989.0,
      "step": 20235
    },
    {
      "entropy": 6.321559953689575,
      "epoch": 2.3358338141950377,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004462640764077751,
      "loss": 5.7514,
      "mean_token_accuracy": 0.1948893427848816,
      "num_tokens": 51309022.0,
      "step": 20240
    },
    {
      "entropy": 6.292261934280395,
      "epoch": 2.336410848240046,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004462373116106971,
      "loss": 5.7936,
      "mean_token_accuracy": 0.1958243042230606,
      "num_tokens": 51322481.0,
      "step": 20245
    },
    {
      "entropy": 6.361962890625,
      "epoch": 2.336987882285055,
      "grad_norm": 1.0,
      "learning_rate": 0.00044621054105413704,
      "loss": 5.74,
      "mean_token_accuracy": 0.19552091509103775,
      "num_tokens": 51334499.0,
      "step": 20250
    },
    {
      "entropy": 6.3130964756011965,
      "epoch": 2.3375649163300634,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044618376473899555,
      "loss": 5.678,
      "mean_token_accuracy": 0.19823378473520278,
      "num_tokens": 51346489.0,
      "step": 20255
    },
    {
      "entropy": 6.351144552230835,
      "epoch": 2.338141950375072,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004461569826661732,
      "loss": 5.8002,
      "mean_token_accuracy": 0.19324805289506913,
      "num_tokens": 51359624.0,
      "step": 20260
    },
    {
      "entropy": 6.328731489181519,
      "epoch": 2.3387189844200806,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004461301948365707,
      "loss": 5.7211,
      "mean_token_accuracy": 0.20283400267362595,
      "num_tokens": 51373768.0,
      "step": 20265
    },
    {
      "entropy": 6.251941585540772,
      "epoch": 2.3392960184650895,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004461034012510891,
      "loss": 5.7548,
      "mean_token_accuracy": 0.18799879550933837,
      "num_tokens": 51386160.0,
      "step": 20270
    },
    {
      "entropy": 6.34953761100769,
      "epoch": 2.339873052510098,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044607660191062963,
      "loss": 5.7411,
      "mean_token_accuracy": 0.20084796398878096,
      "num_tokens": 51397579.0,
      "step": 20275
    },
    {
      "entropy": 6.275950193405151,
      "epoch": 2.340450086555107,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044604979681609364,
      "loss": 5.686,
      "mean_token_accuracy": 0.19774765521287918,
      "num_tokens": 51409925.0,
      "step": 20280
    },
    {
      "entropy": 6.277930068969726,
      "epoch": 2.341027120600115,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044602298596838264,
      "loss": 5.7592,
      "mean_token_accuracy": 0.19631763100624083,
      "num_tokens": 51422840.0,
      "step": 20285
    },
    {
      "entropy": 6.300139570236206,
      "epoch": 2.341604154645124,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044599616936839853,
      "loss": 5.6665,
      "mean_token_accuracy": 0.199966599047184,
      "num_tokens": 51434626.0,
      "step": 20290
    },
    {
      "entropy": 6.29100227355957,
      "epoch": 2.342181188690133,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004459693470170432,
      "loss": 5.6871,
      "mean_token_accuracy": 0.20537949353456497,
      "num_tokens": 51446358.0,
      "step": 20295
    },
    {
      "entropy": 6.257119655609131,
      "epoch": 2.3427582227351413,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004459425189152187,
      "loss": 5.7404,
      "mean_token_accuracy": 0.19182835519313812,
      "num_tokens": 51459029.0,
      "step": 20300
    },
    {
      "entropy": 6.297420406341553,
      "epoch": 2.34333525678015,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044591568506382757,
      "loss": 5.6606,
      "mean_token_accuracy": 0.20428209900856018,
      "num_tokens": 51471250.0,
      "step": 20305
    },
    {
      "entropy": 6.290306806564331,
      "epoch": 2.3439122908251586,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044588884546377226,
      "loss": 5.7222,
      "mean_token_accuracy": 0.19511672854423523,
      "num_tokens": 51482456.0,
      "step": 20310
    },
    {
      "entropy": 6.17057728767395,
      "epoch": 2.3444893248701675,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004458620001159555,
      "loss": 5.6251,
      "mean_token_accuracy": 0.20548744797706603,
      "num_tokens": 51495413.0,
      "step": 20315
    },
    {
      "entropy": 6.290595006942749,
      "epoch": 2.345066358915176,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004458351490212803,
      "loss": 5.715,
      "mean_token_accuracy": 0.1964409753680229,
      "num_tokens": 51507068.0,
      "step": 20320
    },
    {
      "entropy": 6.1644501209259035,
      "epoch": 2.3456433929601848,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044580829218064964,
      "loss": 5.57,
      "mean_token_accuracy": 0.20574390143156052,
      "num_tokens": 51519553.0,
      "step": 20325
    },
    {
      "entropy": 6.285444021224976,
      "epoch": 2.346220427005193,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000445781429594967,
      "loss": 5.7126,
      "mean_token_accuracy": 0.19945957958698274,
      "num_tokens": 51531027.0,
      "step": 20330
    },
    {
      "entropy": 6.266679477691651,
      "epoch": 2.346797461050202,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004457545612651357,
      "loss": 5.6732,
      "mean_token_accuracy": 0.19885026067495346,
      "num_tokens": 51543883.0,
      "step": 20335
    },
    {
      "entropy": 6.278856897354126,
      "epoch": 2.3473744950952105,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044572768719205964,
      "loss": 5.6573,
      "mean_token_accuracy": 0.2038573905825615,
      "num_tokens": 51557514.0,
      "step": 20340
    },
    {
      "entropy": 6.274388360977173,
      "epoch": 2.3479515291402193,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044570080737664264,
      "loss": 5.7115,
      "mean_token_accuracy": 0.200784595310688,
      "num_tokens": 51569990.0,
      "step": 20345
    },
    {
      "entropy": 6.349345588684082,
      "epoch": 2.3485285631852277,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044567392181978874,
      "loss": 5.7171,
      "mean_token_accuracy": 0.20142400711774827,
      "num_tokens": 51581794.0,
      "step": 20350
    },
    {
      "entropy": 6.277327346801758,
      "epoch": 2.3491055972302366,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044564703052240223,
      "loss": 5.6434,
      "mean_token_accuracy": 0.2014608860015869,
      "num_tokens": 51594121.0,
      "step": 20355
    },
    {
      "entropy": 6.248101377487183,
      "epoch": 2.349682631275245,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004456201334853876,
      "loss": 5.6431,
      "mean_token_accuracy": 0.20842165499925613,
      "num_tokens": 51606418.0,
      "step": 20360
    },
    {
      "entropy": 6.285100364685059,
      "epoch": 2.350259665320254,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044559323070964956,
      "loss": 5.7463,
      "mean_token_accuracy": 0.19385244250297545,
      "num_tokens": 51619529.0,
      "step": 20365
    },
    {
      "entropy": 6.294569206237793,
      "epoch": 2.3508366993652627,
      "grad_norm": 1.0,
      "learning_rate": 0.000445566322196093,
      "loss": 5.7809,
      "mean_token_accuracy": 0.19219572693109513,
      "num_tokens": 51631724.0,
      "step": 20370
    },
    {
      "entropy": 6.33821473121643,
      "epoch": 2.351413733410271,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004455394079456228,
      "loss": 5.7187,
      "mean_token_accuracy": 0.1950804263353348,
      "num_tokens": 51643908.0,
      "step": 20375
    },
    {
      "entropy": 6.348220539093018,
      "epoch": 2.35199076745528,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044551248795914446,
      "loss": 5.7911,
      "mean_token_accuracy": 0.19523229002952575,
      "num_tokens": 51654873.0,
      "step": 20380
    },
    {
      "entropy": 6.246188354492188,
      "epoch": 2.3525678015002884,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004454855622375632,
      "loss": 5.6663,
      "mean_token_accuracy": 0.2012071952223778,
      "num_tokens": 51667765.0,
      "step": 20385
    },
    {
      "entropy": 6.248934125900268,
      "epoch": 2.3531448355452973,
      "grad_norm": 1.0,
      "learning_rate": 0.0004454586307817848,
      "loss": 5.7081,
      "mean_token_accuracy": 0.1975087583065033,
      "num_tokens": 51680950.0,
      "step": 20390
    },
    {
      "entropy": 6.257642316818237,
      "epoch": 2.3537218695903057,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000445431693592715,
      "loss": 5.6126,
      "mean_token_accuracy": 0.21779688000679015,
      "num_tokens": 51696437.0,
      "step": 20395
    },
    {
      "entropy": 6.2939142227172855,
      "epoch": 2.3542989036353146,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004454047506712598,
      "loss": 5.7275,
      "mean_token_accuracy": 0.20075519979000092,
      "num_tokens": 51709142.0,
      "step": 20400
    },
    {
      "entropy": 6.219584274291992,
      "epoch": 2.354875937680323,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044537780201832545,
      "loss": 5.6139,
      "mean_token_accuracy": 0.2038278639316559,
      "num_tokens": 51721197.0,
      "step": 20405
    },
    {
      "entropy": 6.3291013717651365,
      "epoch": 2.355452971725332,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004453508476348184,
      "loss": 5.7099,
      "mean_token_accuracy": 0.20175200551748276,
      "num_tokens": 51734921.0,
      "step": 20410
    },
    {
      "entropy": 6.28833794593811,
      "epoch": 2.3560300057703403,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004453238875216452,
      "loss": 5.6937,
      "mean_token_accuracy": 0.19974484592676162,
      "num_tokens": 51747143.0,
      "step": 20415
    },
    {
      "entropy": 6.293047666549683,
      "epoch": 2.356607039815349,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004452969216797127,
      "loss": 5.7237,
      "mean_token_accuracy": 0.18916042298078536,
      "num_tokens": 51760203.0,
      "step": 20420
    },
    {
      "entropy": 6.306429862976074,
      "epoch": 2.357184073860358,
      "grad_norm": 1.125,
      "learning_rate": 0.0004452699501099277,
      "loss": 5.7041,
      "mean_token_accuracy": 0.20085110366344452,
      "num_tokens": 51772287.0,
      "step": 20425
    },
    {
      "entropy": 6.2686573505401615,
      "epoch": 2.3577611079053664,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044524297281319766,
      "loss": 5.7589,
      "mean_token_accuracy": 0.19329400658607482,
      "num_tokens": 51783893.0,
      "step": 20430
    },
    {
      "entropy": 6.3595240116119385,
      "epoch": 2.358338141950375,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00044521598979042963,
      "loss": 5.8107,
      "mean_token_accuracy": 0.19836225509643554,
      "num_tokens": 51795482.0,
      "step": 20435
    },
    {
      "entropy": 6.317958402633667,
      "epoch": 2.3589151759953837,
      "grad_norm": 0.890625,
      "learning_rate": 0.00044518900104253154,
      "loss": 5.6933,
      "mean_token_accuracy": 0.20123952478170395,
      "num_tokens": 51810306.0,
      "step": 20440
    },
    {
      "entropy": 6.3059063911437985,
      "epoch": 2.3594922100403926,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004451620065704108,
      "loss": 5.7448,
      "mean_token_accuracy": 0.19207025468349456,
      "num_tokens": 51824297.0,
      "step": 20445
    },
    {
      "entropy": 6.311263751983643,
      "epoch": 2.360069244085401,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044513500637497546,
      "loss": 5.6528,
      "mean_token_accuracy": 0.20024437308311463,
      "num_tokens": 51835870.0,
      "step": 20450
    },
    {
      "entropy": 6.298462200164795,
      "epoch": 2.36064627813041,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044510800045713373,
      "loss": 5.7368,
      "mean_token_accuracy": 0.2015553206205368,
      "num_tokens": 51847235.0,
      "step": 20455
    },
    {
      "entropy": 6.318539953231811,
      "epoch": 2.3612233121754183,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044508098881779396,
      "loss": 5.7125,
      "mean_token_accuracy": 0.19752871990203857,
      "num_tokens": 51859192.0,
      "step": 20460
    },
    {
      "entropy": 6.21719970703125,
      "epoch": 2.361800346220427,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004450539714578645,
      "loss": 5.5952,
      "mean_token_accuracy": 0.21614652574062349,
      "num_tokens": 51872463.0,
      "step": 20465
    },
    {
      "entropy": 6.289052248001099,
      "epoch": 2.3623773802654355,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004450269483782543,
      "loss": 5.7496,
      "mean_token_accuracy": 0.19656335711479186,
      "num_tokens": 51886320.0,
      "step": 20470
    },
    {
      "entropy": 6.331766796112061,
      "epoch": 2.3629544143104444,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004449999195798721,
      "loss": 5.726,
      "mean_token_accuracy": 0.1946369394659996,
      "num_tokens": 51898670.0,
      "step": 20475
    },
    {
      "entropy": 6.34332594871521,
      "epoch": 2.363531448355453,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044497288506362706,
      "loss": 5.7809,
      "mean_token_accuracy": 0.19378641098737717,
      "num_tokens": 51911510.0,
      "step": 20480
    },
    {
      "entropy": 6.293490362167359,
      "epoch": 2.3641084824004617,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004449458448304284,
      "loss": 5.6837,
      "mean_token_accuracy": 0.21060922592878342,
      "num_tokens": 51923322.0,
      "step": 20485
    },
    {
      "entropy": 6.233779668807983,
      "epoch": 2.36468551644547,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044491879888118574,
      "loss": 5.5798,
      "mean_token_accuracy": 0.20787028670310975,
      "num_tokens": 51935010.0,
      "step": 20490
    },
    {
      "entropy": 6.254586029052734,
      "epoch": 2.365262550490479,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004448917472168087,
      "loss": 5.5933,
      "mean_token_accuracy": 0.21142445504665375,
      "num_tokens": 51949083.0,
      "step": 20495
    },
    {
      "entropy": 6.243372535705566,
      "epoch": 2.365839584535488,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044486468983820707,
      "loss": 5.7045,
      "mean_token_accuracy": 0.19030468314886093,
      "num_tokens": 51962150.0,
      "step": 20500
    },
    {
      "entropy": 6.326984834671021,
      "epoch": 2.3664166185804962,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000444837626746291,
      "loss": 5.7371,
      "mean_token_accuracy": 0.19505570232868194,
      "num_tokens": 51974867.0,
      "step": 20505
    },
    {
      "entropy": 6.300310182571411,
      "epoch": 2.366993652625505,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004448105579419707,
      "loss": 5.6631,
      "mean_token_accuracy": 0.1996551051735878,
      "num_tokens": 51986654.0,
      "step": 20510
    },
    {
      "entropy": 6.235426664352417,
      "epoch": 2.3675706866705135,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004447834834261567,
      "loss": 5.6946,
      "mean_token_accuracy": 0.1977281779050827,
      "num_tokens": 51999402.0,
      "step": 20515
    },
    {
      "entropy": 6.344518995285034,
      "epoch": 2.3681477207155224,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004447564031997595,
      "loss": 5.743,
      "mean_token_accuracy": 0.19702544808387756,
      "num_tokens": 52013583.0,
      "step": 20520
    },
    {
      "entropy": 6.335379648208618,
      "epoch": 2.368724754760531,
      "grad_norm": 0.875,
      "learning_rate": 0.00044472931726369,
      "loss": 5.7604,
      "mean_token_accuracy": 0.19680674225091935,
      "num_tokens": 52026892.0,
      "step": 20525
    },
    {
      "entropy": 6.214617729187012,
      "epoch": 2.3693017888055397,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044470222561885926,
      "loss": 5.6276,
      "mean_token_accuracy": 0.21170947700738907,
      "num_tokens": 52039837.0,
      "step": 20530
    },
    {
      "entropy": 6.265171241760254,
      "epoch": 2.369878822850548,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044467512826617845,
      "loss": 5.6615,
      "mean_token_accuracy": 0.20665282756090164,
      "num_tokens": 52054079.0,
      "step": 20535
    },
    {
      "entropy": 6.328473281860352,
      "epoch": 2.370455856895557,
      "grad_norm": 0.90625,
      "learning_rate": 0.00044464802520655897,
      "loss": 5.7046,
      "mean_token_accuracy": 0.19810098856687547,
      "num_tokens": 52066895.0,
      "step": 20540
    },
    {
      "entropy": 6.2767518043518065,
      "epoch": 2.3710328909405654,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004446209164409124,
      "loss": 5.7095,
      "mean_token_accuracy": 0.19833730459213256,
      "num_tokens": 52078983.0,
      "step": 20545
    },
    {
      "entropy": 6.259795045852661,
      "epoch": 2.371609924985574,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004445938019701506,
      "loss": 5.7595,
      "mean_token_accuracy": 0.19253990054130554,
      "num_tokens": 52091692.0,
      "step": 20550
    },
    {
      "entropy": 6.386502122879028,
      "epoch": 2.3721869590305826,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004445666817951855,
      "loss": 5.6967,
      "mean_token_accuracy": 0.20134469717741013,
      "num_tokens": 52103010.0,
      "step": 20555
    },
    {
      "entropy": 6.234955930709839,
      "epoch": 2.3727639930755915,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004445395559169293,
      "loss": 5.6785,
      "mean_token_accuracy": 0.20398869514465331,
      "num_tokens": 52115185.0,
      "step": 20560
    },
    {
      "entropy": 6.27038049697876,
      "epoch": 2.3733410271206,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004445124243362943,
      "loss": 5.6802,
      "mean_token_accuracy": 0.20006603598594666,
      "num_tokens": 52127713.0,
      "step": 20565
    },
    {
      "entropy": 6.218042707443237,
      "epoch": 2.3739180611656088,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004444852870541932,
      "loss": 5.6539,
      "mean_token_accuracy": 0.20626647472381593,
      "num_tokens": 52141127.0,
      "step": 20570
    },
    {
      "entropy": 6.1550905227661135,
      "epoch": 2.3744950952106176,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004444581440715386,
      "loss": 5.6539,
      "mean_token_accuracy": 0.21168829947710038,
      "num_tokens": 52154063.0,
      "step": 20575
    },
    {
      "entropy": 6.322919273376465,
      "epoch": 2.375072129255626,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044443099538924347,
      "loss": 5.7224,
      "mean_token_accuracy": 0.1993056207895279,
      "num_tokens": 52165605.0,
      "step": 20580
    },
    {
      "entropy": 6.319612836837768,
      "epoch": 2.375649163300635,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004444038410082211,
      "loss": 5.6718,
      "mean_token_accuracy": 0.20211448073387145,
      "num_tokens": 52178739.0,
      "step": 20585
    },
    {
      "entropy": 6.373740863800049,
      "epoch": 2.3762261973456433,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004443766809293846,
      "loss": 5.7989,
      "mean_token_accuracy": 0.1939915508031845,
      "num_tokens": 52192056.0,
      "step": 20590
    },
    {
      "entropy": 6.271612691879272,
      "epoch": 2.376803231390652,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004443495151536475,
      "loss": 5.626,
      "mean_token_accuracy": 0.20548810362815856,
      "num_tokens": 52204791.0,
      "step": 20595
    },
    {
      "entropy": 6.259437990188599,
      "epoch": 2.3773802654356606,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004443223436819237,
      "loss": 5.6803,
      "mean_token_accuracy": 0.20553783625364302,
      "num_tokens": 52216929.0,
      "step": 20600
    },
    {
      "entropy": 6.336847543716431,
      "epoch": 2.3779572994806695,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00044429516651512687,
      "loss": 5.821,
      "mean_token_accuracy": 0.19546782672405244,
      "num_tokens": 52228946.0,
      "step": 20605
    },
    {
      "entropy": 6.279832124710083,
      "epoch": 2.378534333525678,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044426798365417133,
      "loss": 5.6502,
      "mean_token_accuracy": 0.20633373856544496,
      "num_tokens": 52242114.0,
      "step": 20610
    },
    {
      "entropy": 6.254376459121704,
      "epoch": 2.3791113675706868,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044424079509997104,
      "loss": 5.7392,
      "mean_token_accuracy": 0.19593341946601867,
      "num_tokens": 52255434.0,
      "step": 20615
    },
    {
      "entropy": 6.329301977157593,
      "epoch": 2.379688401615695,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004442136008534409,
      "loss": 5.7653,
      "mean_token_accuracy": 0.19942591041326524,
      "num_tokens": 52268774.0,
      "step": 20620
    },
    {
      "entropy": 6.291130065917969,
      "epoch": 2.380265435660704,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00044418640091549525,
      "loss": 5.697,
      "mean_token_accuracy": 0.20560376346111298,
      "num_tokens": 52282046.0,
      "step": 20625
    },
    {
      "entropy": 6.267010927200317,
      "epoch": 2.3808424697057124,
      "grad_norm": 0.9375,
      "learning_rate": 0.000444159195287049,
      "loss": 5.693,
      "mean_token_accuracy": 0.1991726741194725,
      "num_tokens": 52295266.0,
      "step": 20630
    },
    {
      "entropy": 6.241264581680298,
      "epoch": 2.3814195037507213,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004441319839690173,
      "loss": 5.6182,
      "mean_token_accuracy": 0.20492142736911773,
      "num_tokens": 52307530.0,
      "step": 20635
    },
    {
      "entropy": 6.215575265884399,
      "epoch": 2.3819965377957297,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004441047669623153,
      "loss": 5.6055,
      "mean_token_accuracy": 0.20616517066955567,
      "num_tokens": 52319461.0,
      "step": 20640
    },
    {
      "entropy": 6.244829320907593,
      "epoch": 2.3825735718407386,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044407754426785845,
      "loss": 5.6606,
      "mean_token_accuracy": 0.19798144996166228,
      "num_tokens": 52332163.0,
      "step": 20645
    },
    {
      "entropy": 6.362397289276123,
      "epoch": 2.3831506058857475,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004440503158865625,
      "loss": 5.8311,
      "mean_token_accuracy": 0.19148696959018707,
      "num_tokens": 52344970.0,
      "step": 20650
    },
    {
      "entropy": 6.328934240341186,
      "epoch": 2.383727639930756,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044402308181934295,
      "loss": 5.7411,
      "mean_token_accuracy": 0.2022900328040123,
      "num_tokens": 52358378.0,
      "step": 20655
    },
    {
      "entropy": 6.306194639205932,
      "epoch": 2.3843046739757647,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004439958420671162,
      "loss": 5.6939,
      "mean_token_accuracy": 0.19589851945638656,
      "num_tokens": 52369335.0,
      "step": 20660
    },
    {
      "entropy": 6.263232707977295,
      "epoch": 2.384881708020773,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044396859663079814,
      "loss": 5.6673,
      "mean_token_accuracy": 0.20530399680137634,
      "num_tokens": 52381573.0,
      "step": 20665
    },
    {
      "entropy": 6.260594177246094,
      "epoch": 2.385458742065782,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044394134551130533,
      "loss": 5.6475,
      "mean_token_accuracy": 0.20745763331651687,
      "num_tokens": 52395316.0,
      "step": 20670
    },
    {
      "entropy": 6.347339916229248,
      "epoch": 2.3860357761107904,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004439140887095542,
      "loss": 5.6514,
      "mean_token_accuracy": 0.1984178066253662,
      "num_tokens": 52407751.0,
      "step": 20675
    },
    {
      "entropy": 6.237292814254761,
      "epoch": 2.3866128101557993,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044388682622646165,
      "loss": 5.6992,
      "mean_token_accuracy": 0.1995289295911789,
      "num_tokens": 52420880.0,
      "step": 20680
    },
    {
      "entropy": 6.214630889892578,
      "epoch": 2.3871898442008077,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004438595580629446,
      "loss": 5.6858,
      "mean_token_accuracy": 0.2032725393772125,
      "num_tokens": 52433600.0,
      "step": 20685
    },
    {
      "entropy": 6.2300440788269045,
      "epoch": 2.3877668782458166,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004438322842199202,
      "loss": 5.6064,
      "mean_token_accuracy": 0.21170931607484816,
      "num_tokens": 52446983.0,
      "step": 20690
    },
    {
      "entropy": 6.342708492279053,
      "epoch": 2.388343912290825,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004438050046983057,
      "loss": 5.7063,
      "mean_token_accuracy": 0.1992245852947235,
      "num_tokens": 52460668.0,
      "step": 20695
    },
    {
      "entropy": 6.284821844100952,
      "epoch": 2.388920946335834,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044377771949901876,
      "loss": 5.6732,
      "mean_token_accuracy": 0.202784663438797,
      "num_tokens": 52472838.0,
      "step": 20700
    },
    {
      "entropy": 6.3212975025177,
      "epoch": 2.3894979803808427,
      "grad_norm": 0.90625,
      "learning_rate": 0.00044375042862297703,
      "loss": 5.7671,
      "mean_token_accuracy": 0.1938607633113861,
      "num_tokens": 52485412.0,
      "step": 20705
    },
    {
      "entropy": 6.316295433044433,
      "epoch": 2.390075014425851,
      "grad_norm": 0.921875,
      "learning_rate": 0.00044372313207109856,
      "loss": 5.7376,
      "mean_token_accuracy": 0.19882224947214128,
      "num_tokens": 52497440.0,
      "step": 20710
    },
    {
      "entropy": 6.296503782272339,
      "epoch": 2.3906520484708595,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044369582984430127,
      "loss": 5.7211,
      "mean_token_accuracy": 0.19760092198848725,
      "num_tokens": 52510549.0,
      "step": 20715
    },
    {
      "entropy": 6.213027620315552,
      "epoch": 2.3912290825158684,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044366852194350354,
      "loss": 5.6466,
      "mean_token_accuracy": 0.20739204734563826,
      "num_tokens": 52522542.0,
      "step": 20720
    },
    {
      "entropy": 6.173247194290161,
      "epoch": 2.3918061165608773,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004436412083696239,
      "loss": 5.5335,
      "mean_token_accuracy": 0.2131787970662117,
      "num_tokens": 52535607.0,
      "step": 20725
    },
    {
      "entropy": 6.309634351730347,
      "epoch": 2.3923831506058857,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044361388912358095,
      "loss": 5.6506,
      "mean_token_accuracy": 0.2048327788710594,
      "num_tokens": 52548515.0,
      "step": 20730
    },
    {
      "entropy": 6.346760988235474,
      "epoch": 2.3929601846508946,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004435865642062936,
      "loss": 5.7199,
      "mean_token_accuracy": 0.19563523083925247,
      "num_tokens": 52561142.0,
      "step": 20735
    },
    {
      "entropy": 6.276459789276123,
      "epoch": 2.393537218695903,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004435592336186809,
      "loss": 5.6696,
      "mean_token_accuracy": 0.20301832705736161,
      "num_tokens": 52573355.0,
      "step": 20740
    },
    {
      "entropy": 6.341160440444947,
      "epoch": 2.394114252740912,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004435318973616622,
      "loss": 5.7768,
      "mean_token_accuracy": 0.1882468953728676,
      "num_tokens": 52586353.0,
      "step": 20745
    },
    {
      "entropy": 6.302660131454468,
      "epoch": 2.3946912867859202,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044350455543615665,
      "loss": 5.7045,
      "mean_token_accuracy": 0.19920828193426132,
      "num_tokens": 52597673.0,
      "step": 20750
    },
    {
      "entropy": 6.262860679626465,
      "epoch": 2.395268320830929,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004434772078430843,
      "loss": 5.7308,
      "mean_token_accuracy": 0.1938989371061325,
      "num_tokens": 52609165.0,
      "step": 20755
    },
    {
      "entropy": 6.283085918426513,
      "epoch": 2.3958453548759375,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004434498545833646,
      "loss": 5.7655,
      "mean_token_accuracy": 0.19652896076440812,
      "num_tokens": 52623335.0,
      "step": 20760
    },
    {
      "entropy": 6.28039813041687,
      "epoch": 2.3964223889209464,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004434224956579177,
      "loss": 5.6278,
      "mean_token_accuracy": 0.21030324101448059,
      "num_tokens": 52635754.0,
      "step": 20765
    },
    {
      "entropy": 6.240499353408813,
      "epoch": 2.396999422965955,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004433951310676639,
      "loss": 5.7082,
      "mean_token_accuracy": 0.20202796459197997,
      "num_tokens": 52649095.0,
      "step": 20770
    },
    {
      "entropy": 6.246279907226563,
      "epoch": 2.3975764570109637,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044336776081352347,
      "loss": 5.6403,
      "mean_token_accuracy": 0.20732228010892867,
      "num_tokens": 52661421.0,
      "step": 20775
    },
    {
      "entropy": 6.185015630722046,
      "epoch": 2.3981534910559725,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00044334038489641706,
      "loss": 5.5748,
      "mean_token_accuracy": 0.2054827481508255,
      "num_tokens": 52674865.0,
      "step": 20780
    },
    {
      "entropy": 6.280226802825927,
      "epoch": 2.398730525100981,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044331300331726544,
      "loss": 5.7325,
      "mean_token_accuracy": 0.19248102605342865,
      "num_tokens": 52687241.0,
      "step": 20785
    },
    {
      "entropy": 6.314509010314941,
      "epoch": 2.3993075591459894,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044328561607698947,
      "loss": 5.7369,
      "mean_token_accuracy": 0.1893750622868538,
      "num_tokens": 52699567.0,
      "step": 20790
    },
    {
      "entropy": 6.2687211513519285,
      "epoch": 2.3998845931909982,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004432582231765105,
      "loss": 5.6636,
      "mean_token_accuracy": 0.2000647783279419,
      "num_tokens": 52711607.0,
      "step": 20795
    },
    {
      "entropy": 6.2142681121826175,
      "epoch": 2.400461627236007,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044323082461674974,
      "loss": 5.5873,
      "mean_token_accuracy": 0.20686126202344896,
      "num_tokens": 52724256.0,
      "step": 20800
    },
    {
      "entropy": 6.269721746444702,
      "epoch": 2.4010386612810155,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004432034203986287,
      "loss": 5.6777,
      "mean_token_accuracy": 0.21000705510377884,
      "num_tokens": 52736356.0,
      "step": 20805
    },
    {
      "entropy": 6.2595006942749025,
      "epoch": 2.4016156953260244,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004431760105230693,
      "loss": 5.6708,
      "mean_token_accuracy": 0.1999936357140541,
      "num_tokens": 52748623.0,
      "step": 20810
    },
    {
      "entropy": 6.3321715831756595,
      "epoch": 2.402192729371033,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00044314859499099325,
      "loss": 5.7043,
      "mean_token_accuracy": 0.19771352410316467,
      "num_tokens": 52761479.0,
      "step": 20815
    },
    {
      "entropy": 6.2657386302948,
      "epoch": 2.4027697634160416,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044312117380332274,
      "loss": 5.6647,
      "mean_token_accuracy": 0.21245635747909547,
      "num_tokens": 52774681.0,
      "step": 20820
    },
    {
      "entropy": 6.2982886791229244,
      "epoch": 2.40334679746105,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044309374696098,
      "loss": 5.6935,
      "mean_token_accuracy": 0.20492263436317443,
      "num_tokens": 52787495.0,
      "step": 20825
    },
    {
      "entropy": 6.288589859008789,
      "epoch": 2.403923831506059,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004430663144648876,
      "loss": 5.7186,
      "mean_token_accuracy": 0.20073367804288864,
      "num_tokens": 52799958.0,
      "step": 20830
    },
    {
      "entropy": 6.337665653228759,
      "epoch": 2.4045008655510673,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044303887631596823,
      "loss": 5.7004,
      "mean_token_accuracy": 0.20305854827165604,
      "num_tokens": 52812033.0,
      "step": 20835
    },
    {
      "entropy": 6.307322835922241,
      "epoch": 2.405077899596076,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004430114325151447,
      "loss": 5.7223,
      "mean_token_accuracy": 0.19516057819128035,
      "num_tokens": 52824269.0,
      "step": 20840
    },
    {
      "entropy": 6.318175411224365,
      "epoch": 2.4056549336410846,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004429839830633401,
      "loss": 5.709,
      "mean_token_accuracy": 0.2010985031723976,
      "num_tokens": 52837180.0,
      "step": 20845
    },
    {
      "entropy": 6.345485782623291,
      "epoch": 2.4062319676860935,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004429565279614777,
      "loss": 5.7248,
      "mean_token_accuracy": 0.18853194862604142,
      "num_tokens": 52849751.0,
      "step": 20850
    },
    {
      "entropy": 6.288502979278564,
      "epoch": 2.4068090017311023,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00044292906721048094,
      "loss": 5.6719,
      "mean_token_accuracy": 0.19948452115058898,
      "num_tokens": 52862902.0,
      "step": 20855
    },
    {
      "entropy": 6.258730030059814,
      "epoch": 2.4073860357761108,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004429016008112733,
      "loss": 5.6583,
      "mean_token_accuracy": 0.20170366615056992,
      "num_tokens": 52876104.0,
      "step": 20860
    },
    {
      "entropy": 6.296816110610962,
      "epoch": 2.4079630698211196,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004428741287647788,
      "loss": 5.7247,
      "mean_token_accuracy": 0.20037238895893097,
      "num_tokens": 52889173.0,
      "step": 20865
    },
    {
      "entropy": 6.257041263580322,
      "epoch": 2.408540103866128,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044284665107192136,
      "loss": 5.6146,
      "mean_token_accuracy": 0.20975644439458846,
      "num_tokens": 52902064.0,
      "step": 20870
    },
    {
      "entropy": 6.274162626266479,
      "epoch": 2.409117137911137,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004428191677336251,
      "loss": 5.7151,
      "mean_token_accuracy": 0.19541309475898744,
      "num_tokens": 52913895.0,
      "step": 20875
    },
    {
      "entropy": 6.285509204864502,
      "epoch": 2.4096941719561453,
      "grad_norm": 1.0,
      "learning_rate": 0.0004427916787508146,
      "loss": 5.6636,
      "mean_token_accuracy": 0.20441102534532546,
      "num_tokens": 52925589.0,
      "step": 20880
    },
    {
      "entropy": 6.3110274314880375,
      "epoch": 2.410271206001154,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004427641841244144,
      "loss": 5.6706,
      "mean_token_accuracy": 0.21215699166059493,
      "num_tokens": 52940519.0,
      "step": 20885
    },
    {
      "entropy": 6.246319007873535,
      "epoch": 2.4108482400461626,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000442736683855349,
      "loss": 5.6584,
      "mean_token_accuracy": 0.1991742581129074,
      "num_tokens": 52954205.0,
      "step": 20890
    },
    {
      "entropy": 6.245836877822876,
      "epoch": 2.4114252740911715,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004427091779445437,
      "loss": 5.6658,
      "mean_token_accuracy": 0.20347630530595778,
      "num_tokens": 52967949.0,
      "step": 20895
    },
    {
      "entropy": 6.228832769393921,
      "epoch": 2.41200230813618,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004426816663929235,
      "loss": 5.5809,
      "mean_token_accuracy": 0.2109801337122917,
      "num_tokens": 52980125.0,
      "step": 20900
    },
    {
      "entropy": 6.219481706619263,
      "epoch": 2.4125793421811887,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00044265414920141366,
      "loss": 5.5889,
      "mean_token_accuracy": 0.21946226060390472,
      "num_tokens": 52992721.0,
      "step": 20905
    },
    {
      "entropy": 6.279675054550171,
      "epoch": 2.413156376226197,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044262662637093987,
      "loss": 5.7303,
      "mean_token_accuracy": 0.19758040606975555,
      "num_tokens": 53004278.0,
      "step": 20910
    },
    {
      "entropy": 6.2898626804351805,
      "epoch": 2.413733410271206,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00044259909790242776,
      "loss": 5.6601,
      "mean_token_accuracy": 0.2013942554593086,
      "num_tokens": 53018290.0,
      "step": 20915
    },
    {
      "entropy": 6.258230447769165,
      "epoch": 2.4143104443162144,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004425715637968032,
      "loss": 5.5841,
      "mean_token_accuracy": 0.2142757162451744,
      "num_tokens": 53031458.0,
      "step": 20920
    },
    {
      "entropy": 6.2622472763061525,
      "epoch": 2.4148874783612233,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004425440240549923,
      "loss": 5.6577,
      "mean_token_accuracy": 0.19613652527332306,
      "num_tokens": 53042894.0,
      "step": 20925
    },
    {
      "entropy": 6.331469345092773,
      "epoch": 2.415464512406232,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044251647867792147,
      "loss": 5.7883,
      "mean_token_accuracy": 0.20005614310503006,
      "num_tokens": 53056206.0,
      "step": 20930
    },
    {
      "entropy": 6.289569234848022,
      "epoch": 2.4160415464512406,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044248892766651706,
      "loss": 5.6673,
      "mean_token_accuracy": 0.20417799204587936,
      "num_tokens": 53069249.0,
      "step": 20935
    },
    {
      "entropy": 6.298072719573975,
      "epoch": 2.4166185804962494,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004424613710217057,
      "loss": 5.7249,
      "mean_token_accuracy": 0.1984723836183548,
      "num_tokens": 53080716.0,
      "step": 20940
    },
    {
      "entropy": 6.301897239685059,
      "epoch": 2.417195614541258,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044243380874441437,
      "loss": 5.7325,
      "mean_token_accuracy": 0.19592752903699875,
      "num_tokens": 53093671.0,
      "step": 20945
    },
    {
      "entropy": 6.332162237167358,
      "epoch": 2.4177726485862667,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044240624083557,
      "loss": 5.7376,
      "mean_token_accuracy": 0.1967240661382675,
      "num_tokens": 53106057.0,
      "step": 20950
    },
    {
      "entropy": 6.27943902015686,
      "epoch": 2.418349682631275,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044237866729609994,
      "loss": 5.6402,
      "mean_token_accuracy": 0.20447174608707427,
      "num_tokens": 53118902.0,
      "step": 20955
    },
    {
      "entropy": 6.250809955596924,
      "epoch": 2.418926716676284,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004423510881269315,
      "loss": 5.7597,
      "mean_token_accuracy": 0.1952386900782585,
      "num_tokens": 53133084.0,
      "step": 20960
    },
    {
      "entropy": 6.274754571914673,
      "epoch": 2.4195037507212924,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004423235033289924,
      "loss": 5.5788,
      "mean_token_accuracy": 0.20877631902694702,
      "num_tokens": 53144877.0,
      "step": 20965
    },
    {
      "entropy": 6.301543951034546,
      "epoch": 2.4200807847663013,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004422959129032103,
      "loss": 5.6073,
      "mean_token_accuracy": 0.21382453739643098,
      "num_tokens": 53158328.0,
      "step": 20970
    },
    {
      "entropy": 6.232356452941895,
      "epoch": 2.4206578188113097,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044226831685051333,
      "loss": 5.7,
      "mean_token_accuracy": 0.1896368682384491,
      "num_tokens": 53169769.0,
      "step": 20975
    },
    {
      "entropy": 6.309162950515747,
      "epoch": 2.4212348528563186,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004422407151718296,
      "loss": 5.6835,
      "mean_token_accuracy": 0.20633565485477448,
      "num_tokens": 53182686.0,
      "step": 20980
    },
    {
      "entropy": 6.2930186748504635,
      "epoch": 2.4218118869013274,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044221310786808746,
      "loss": 5.701,
      "mean_token_accuracy": 0.2023796021938324,
      "num_tokens": 53195720.0,
      "step": 20985
    },
    {
      "entropy": 6.184375429153443,
      "epoch": 2.422388920946336,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004421854949402155,
      "loss": 5.582,
      "mean_token_accuracy": 0.21068751960992813,
      "num_tokens": 53209068.0,
      "step": 20990
    },
    {
      "entropy": 6.227423000335693,
      "epoch": 2.4229659549913443,
      "grad_norm": 1.0,
      "learning_rate": 0.00044215787638914245,
      "loss": 5.6562,
      "mean_token_accuracy": 0.19896408021450043,
      "num_tokens": 53220739.0,
      "step": 20995
    },
    {
      "entropy": 6.236641263961792,
      "epoch": 2.423542989036353,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004421302522157973,
      "loss": 5.7268,
      "mean_token_accuracy": 0.19308724701404573,
      "num_tokens": 53232633.0,
      "step": 21000
    },
    {
      "epoch": 2.423542989036353,
      "eval_entropy": 6.0818525100249206,
      "eval_loss": 5.786560535430908,
      "eval_mean_token_accuracy": 0.20466101162595304,
      "eval_num_tokens": 53232633.0,
      "eval_runtime": 17.6874,
      "eval_samples_per_second": 1590.733,
      "eval_steps_per_second": 198.842,
      "step": 21000
    },
    {
      "entropy": 6.262138319015503,
      "epoch": 2.424120023081362,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004421026224211091,
      "loss": 5.5961,
      "mean_token_accuracy": 0.20974094569683074,
      "num_tokens": 53246202.0,
      "step": 21005
    },
    {
      "entropy": 6.293935060501099,
      "epoch": 2.4246970571263704,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044207498700600724,
      "loss": 5.704,
      "mean_token_accuracy": 0.19473184496164322,
      "num_tokens": 53258415.0,
      "step": 21010
    },
    {
      "entropy": 6.3245786190032955,
      "epoch": 2.4252740911713793,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044204734597142115,
      "loss": 5.7575,
      "mean_token_accuracy": 0.1926491841673851,
      "num_tokens": 53271348.0,
      "step": 21015
    },
    {
      "entropy": 6.274426651000977,
      "epoch": 2.4258511252163877,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044201969931828057,
      "loss": 5.7145,
      "mean_token_accuracy": 0.19771473556756974,
      "num_tokens": 53285573.0,
      "step": 21020
    },
    {
      "entropy": 6.261736059188843,
      "epoch": 2.4264281592613965,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004419920470475154,
      "loss": 5.6506,
      "mean_token_accuracy": 0.2060353621840477,
      "num_tokens": 53297163.0,
      "step": 21025
    },
    {
      "entropy": 6.255591344833374,
      "epoch": 2.427005193306405,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004419643891600556,
      "loss": 5.6603,
      "mean_token_accuracy": 0.19760891050100327,
      "num_tokens": 53311119.0,
      "step": 21030
    },
    {
      "entropy": 6.333131599426269,
      "epoch": 2.427582227351414,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00044193672565683153,
      "loss": 5.7562,
      "mean_token_accuracy": 0.19083672016859055,
      "num_tokens": 53322753.0,
      "step": 21035
    },
    {
      "entropy": 6.3285027027130125,
      "epoch": 2.4281592613964222,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004419090565387736,
      "loss": 5.7805,
      "mean_token_accuracy": 0.20390615463256836,
      "num_tokens": 53335333.0,
      "step": 21040
    },
    {
      "entropy": 6.215524625778198,
      "epoch": 2.428736295441431,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044188138180681255,
      "loss": 5.5817,
      "mean_token_accuracy": 0.20926980525255204,
      "num_tokens": 53348500.0,
      "step": 21045
    },
    {
      "entropy": 6.177030420303344,
      "epoch": 2.4293133294864395,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000441853701461879,
      "loss": 5.605,
      "mean_token_accuracy": 0.21050911843776704,
      "num_tokens": 53360708.0,
      "step": 21050
    },
    {
      "entropy": 6.304535913467407,
      "epoch": 2.4298903635314484,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044182601550490414,
      "loss": 5.6801,
      "mean_token_accuracy": 0.21321144104003906,
      "num_tokens": 53374496.0,
      "step": 21055
    },
    {
      "entropy": 6.302006816864013,
      "epoch": 2.4304673975764572,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004417983239368192,
      "loss": 5.6674,
      "mean_token_accuracy": 0.20267537534236907,
      "num_tokens": 53387301.0,
      "step": 21060
    },
    {
      "entropy": 6.323476982116699,
      "epoch": 2.4310444316214657,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044177062675855534,
      "loss": 5.8036,
      "mean_token_accuracy": 0.19215258955955505,
      "num_tokens": 53399913.0,
      "step": 21065
    },
    {
      "entropy": 6.238678598403931,
      "epoch": 2.431621465666474,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004417429239710444,
      "loss": 5.6757,
      "mean_token_accuracy": 0.20352690070867538,
      "num_tokens": 53415108.0,
      "step": 21070
    },
    {
      "entropy": 6.3369687557220455,
      "epoch": 2.432198499711483,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004417152155752179,
      "loss": 5.7148,
      "mean_token_accuracy": 0.19939194917678832,
      "num_tokens": 53427163.0,
      "step": 21075
    },
    {
      "entropy": 6.279685544967651,
      "epoch": 2.432775533756492,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004416875015720081,
      "loss": 5.6831,
      "mean_token_accuracy": 0.20299756973981858,
      "num_tokens": 53439712.0,
      "step": 21080
    },
    {
      "entropy": 6.269053745269775,
      "epoch": 2.4333525678015,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004416597819623469,
      "loss": 5.6574,
      "mean_token_accuracy": 0.204183629155159,
      "num_tokens": 53453581.0,
      "step": 21085
    },
    {
      "entropy": 6.301329278945923,
      "epoch": 2.433929601846509,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044163205674716666,
      "loss": 5.6265,
      "mean_token_accuracy": 0.20538261234760286,
      "num_tokens": 53465805.0,
      "step": 21090
    },
    {
      "entropy": 6.333134555816651,
      "epoch": 2.4345066358915175,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044160432592740015,
      "loss": 5.6982,
      "mean_token_accuracy": 0.2017306625843048,
      "num_tokens": 53479023.0,
      "step": 21095
    },
    {
      "entropy": 6.319257545471191,
      "epoch": 2.4350836699365264,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044157658950397965,
      "loss": 5.6867,
      "mean_token_accuracy": 0.20240473747253418,
      "num_tokens": 53492547.0,
      "step": 21100
    },
    {
      "entropy": 6.297816228866577,
      "epoch": 2.435660703981535,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044154884747783844,
      "loss": 5.7428,
      "mean_token_accuracy": 0.2016036555171013,
      "num_tokens": 53505827.0,
      "step": 21105
    },
    {
      "entropy": 6.332560014724732,
      "epoch": 2.4362377380265436,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044152109984990954,
      "loss": 5.7801,
      "mean_token_accuracy": 0.20456285923719406,
      "num_tokens": 53518452.0,
      "step": 21110
    },
    {
      "entropy": 6.256117963790894,
      "epoch": 2.436814772071552,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000441493346621126,
      "loss": 5.6137,
      "mean_token_accuracy": 0.2107243537902832,
      "num_tokens": 53531489.0,
      "step": 21115
    },
    {
      "entropy": 6.269166564941406,
      "epoch": 2.437391806116561,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044146558779242155,
      "loss": 5.6437,
      "mean_token_accuracy": 0.21201954334974288,
      "num_tokens": 53544874.0,
      "step": 21120
    },
    {
      "entropy": 6.314460802078247,
      "epoch": 2.4379688401615693,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004414378233647298,
      "loss": 5.7117,
      "mean_token_accuracy": 0.20049450993537904,
      "num_tokens": 53557233.0,
      "step": 21125
    },
    {
      "entropy": 6.248986911773682,
      "epoch": 2.438545874206578,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044141005333898437,
      "loss": 5.6692,
      "mean_token_accuracy": 0.20350752770900726,
      "num_tokens": 53570012.0,
      "step": 21130
    },
    {
      "entropy": 6.32624135017395,
      "epoch": 2.439122908251587,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004413822777161196,
      "loss": 5.7312,
      "mean_token_accuracy": 0.2043614700436592,
      "num_tokens": 53582413.0,
      "step": 21135
    },
    {
      "entropy": 6.2847898483276365,
      "epoch": 2.4396999422965955,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004413544964970695,
      "loss": 5.712,
      "mean_token_accuracy": 0.1968155786395073,
      "num_tokens": 53595136.0,
      "step": 21140
    },
    {
      "entropy": 6.373248147964477,
      "epoch": 2.4402769763416043,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004413267096827686,
      "loss": 5.7886,
      "mean_token_accuracy": 0.1904096484184265,
      "num_tokens": 53607219.0,
      "step": 21145
    },
    {
      "entropy": 6.226694774627686,
      "epoch": 2.4408540103866128,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004412989172741514,
      "loss": 5.6105,
      "mean_token_accuracy": 0.20633069723844527,
      "num_tokens": 53620510.0,
      "step": 21150
    },
    {
      "entropy": 6.256235408782959,
      "epoch": 2.4414310444316216,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044127111927215267,
      "loss": 5.6832,
      "mean_token_accuracy": 0.20127054005861283,
      "num_tokens": 53634036.0,
      "step": 21155
    },
    {
      "entropy": 6.2259890079498295,
      "epoch": 2.44200807847663,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00044124331567770746,
      "loss": 5.6524,
      "mean_token_accuracy": 0.20224977433681487,
      "num_tokens": 53647472.0,
      "step": 21160
    },
    {
      "entropy": 6.356272840499878,
      "epoch": 2.442585112521639,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004412155064917509,
      "loss": 5.7529,
      "mean_token_accuracy": 0.1898324191570282,
      "num_tokens": 53660602.0,
      "step": 21165
    },
    {
      "entropy": 6.33570351600647,
      "epoch": 2.4431621465666473,
      "grad_norm": 1.0,
      "learning_rate": 0.00044118769171521836,
      "loss": 5.67,
      "mean_token_accuracy": 0.20335320085287095,
      "num_tokens": 53672685.0,
      "step": 21170
    },
    {
      "entropy": 6.242777585983276,
      "epoch": 2.443739180611656,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044115987134904543,
      "loss": 5.6418,
      "mean_token_accuracy": 0.20441045463085175,
      "num_tokens": 53685810.0,
      "step": 21175
    },
    {
      "entropy": 6.312706565856933,
      "epoch": 2.4443162146566646,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044113204539416776,
      "loss": 5.7063,
      "mean_token_accuracy": 0.19710328727960585,
      "num_tokens": 53698894.0,
      "step": 21180
    },
    {
      "entropy": 6.335492849349976,
      "epoch": 2.4448932487016735,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004411042138515212,
      "loss": 5.7505,
      "mean_token_accuracy": 0.19698887765407563,
      "num_tokens": 53711837.0,
      "step": 21185
    },
    {
      "entropy": 6.309172773361206,
      "epoch": 2.445470282746682,
      "grad_norm": 1.0,
      "learning_rate": 0.0004410763767220419,
      "loss": 5.7075,
      "mean_token_accuracy": 0.20506821274757386,
      "num_tokens": 53724154.0,
      "step": 21190
    },
    {
      "entropy": 6.332321405410767,
      "epoch": 2.4460473167916907,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004410485340066662,
      "loss": 5.7366,
      "mean_token_accuracy": 0.19712792634963988,
      "num_tokens": 53737083.0,
      "step": 21195
    },
    {
      "entropy": 6.278997325897217,
      "epoch": 2.446624350836699,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004410206857063305,
      "loss": 5.6376,
      "mean_token_accuracy": 0.20256400108337402,
      "num_tokens": 53750263.0,
      "step": 21200
    },
    {
      "entropy": 6.309938526153564,
      "epoch": 2.447201384881708,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004409928318219715,
      "loss": 5.7353,
      "mean_token_accuracy": 0.19984076023101807,
      "num_tokens": 53764029.0,
      "step": 21205
    },
    {
      "entropy": 6.303734588623047,
      "epoch": 2.447778418926717,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004409649723545262,
      "loss": 5.7416,
      "mean_token_accuracy": 0.19316499084234237,
      "num_tokens": 53777280.0,
      "step": 21210
    },
    {
      "entropy": 6.3440220832824705,
      "epoch": 2.4483554529717253,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004409371073049313,
      "loss": 5.6729,
      "mean_token_accuracy": 0.20029273629188538,
      "num_tokens": 53790622.0,
      "step": 21215
    },
    {
      "entropy": 6.301061582565308,
      "epoch": 2.448932487016734,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004409092366741243,
      "loss": 5.6858,
      "mean_token_accuracy": 0.20370455086231232,
      "num_tokens": 53803714.0,
      "step": 21220
    },
    {
      "entropy": 6.335336208343506,
      "epoch": 2.4495095210617426,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004408813604630425,
      "loss": 5.7043,
      "mean_token_accuracy": 0.20059310644865036,
      "num_tokens": 53815984.0,
      "step": 21225
    },
    {
      "entropy": 6.226279592514038,
      "epoch": 2.4500865551067514,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004408534786726236,
      "loss": 5.63,
      "mean_token_accuracy": 0.20272864550352096,
      "num_tokens": 53827591.0,
      "step": 21230
    },
    {
      "entropy": 6.2872631549835205,
      "epoch": 2.45066358915176,
      "grad_norm": 1.0,
      "learning_rate": 0.0004408255913038053,
      "loss": 5.7152,
      "mean_token_accuracy": 0.2004805952310562,
      "num_tokens": 53839730.0,
      "step": 21235
    },
    {
      "entropy": 6.293054008483887,
      "epoch": 2.4512406231967687,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004407976983575256,
      "loss": 5.6939,
      "mean_token_accuracy": 0.20715493857860565,
      "num_tokens": 53852753.0,
      "step": 21240
    },
    {
      "entropy": 6.3600085258483885,
      "epoch": 2.451817657241777,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004407697998347227,
      "loss": 5.7384,
      "mean_token_accuracy": 0.19790853708982467,
      "num_tokens": 53866092.0,
      "step": 21245
    },
    {
      "entropy": 6.3393693447113035,
      "epoch": 2.452394691286786,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004407418957363349,
      "loss": 5.7108,
      "mean_token_accuracy": 0.19982357919216157,
      "num_tokens": 53878065.0,
      "step": 21250
    },
    {
      "entropy": 6.264062166213989,
      "epoch": 2.4529717253317944,
      "grad_norm": 1.0,
      "learning_rate": 0.00044071398606330075,
      "loss": 5.6508,
      "mean_token_accuracy": 0.20337645560503007,
      "num_tokens": 53890164.0,
      "step": 21255
    },
    {
      "entropy": 6.2856817722320555,
      "epoch": 2.4535487593768033,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000440686070816559,
      "loss": 5.7268,
      "mean_token_accuracy": 0.19435691684484482,
      "num_tokens": 53901329.0,
      "step": 21260
    },
    {
      "entropy": 6.297254133224487,
      "epoch": 2.4541257934218117,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004406581499970486,
      "loss": 5.7507,
      "mean_token_accuracy": 0.2013178750872612,
      "num_tokens": 53912891.0,
      "step": 21265
    },
    {
      "entropy": 6.300068998336792,
      "epoch": 2.4547028274668206,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004406302236057086,
      "loss": 5.692,
      "mean_token_accuracy": 0.2031162664294243,
      "num_tokens": 53925320.0,
      "step": 21270
    },
    {
      "entropy": 6.268418979644776,
      "epoch": 2.455279861511829,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044060229164347826,
      "loss": 5.663,
      "mean_token_accuracy": 0.21043100953102112,
      "num_tokens": 53938648.0,
      "step": 21275
    },
    {
      "entropy": 6.309641647338867,
      "epoch": 2.455856895556838,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044057435411129714,
      "loss": 5.7111,
      "mean_token_accuracy": 0.19734565019607545,
      "num_tokens": 53951098.0,
      "step": 21280
    },
    {
      "entropy": 6.350348711013794,
      "epoch": 2.4564339296018467,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044054641101010485,
      "loss": 5.7566,
      "mean_token_accuracy": 0.19389044493436813,
      "num_tokens": 53963623.0,
      "step": 21285
    },
    {
      "entropy": 6.352903509140015,
      "epoch": 2.457010963646855,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00044051846234084127,
      "loss": 5.7946,
      "mean_token_accuracy": 0.19069381654262543,
      "num_tokens": 53975814.0,
      "step": 21290
    },
    {
      "entropy": 6.363426780700683,
      "epoch": 2.457587997691864,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004404905081044464,
      "loss": 5.661,
      "mean_token_accuracy": 0.20619637817144393,
      "num_tokens": 53987804.0,
      "step": 21295
    },
    {
      "entropy": 6.308203411102295,
      "epoch": 2.4581650317368724,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004404625483018605,
      "loss": 5.7361,
      "mean_token_accuracy": 0.2047600656747818,
      "num_tokens": 54000295.0,
      "step": 21300
    },
    {
      "entropy": 6.271647691726685,
      "epoch": 2.4587420657818813,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000440434582934024,
      "loss": 5.6509,
      "mean_token_accuracy": 0.20455532073974608,
      "num_tokens": 54011937.0,
      "step": 21305
    },
    {
      "entropy": 6.2994678020477295,
      "epoch": 2.4593190998268897,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044040661200187746,
      "loss": 5.71,
      "mean_token_accuracy": 0.20766065567731856,
      "num_tokens": 54024129.0,
      "step": 21310
    },
    {
      "entropy": 6.155481147766113,
      "epoch": 2.4598961338718985,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00044037863550636173,
      "loss": 5.5263,
      "mean_token_accuracy": 0.21638084948062897,
      "num_tokens": 54037208.0,
      "step": 21315
    },
    {
      "entropy": 6.215900516510009,
      "epoch": 2.460473167916907,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044035065344841766,
      "loss": 5.7211,
      "mean_token_accuracy": 0.2056300774216652,
      "num_tokens": 54049253.0,
      "step": 21320
    },
    {
      "entropy": 6.369457244873047,
      "epoch": 2.461050201961916,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044032266582898655,
      "loss": 5.7514,
      "mean_token_accuracy": 0.19624644070863723,
      "num_tokens": 54060891.0,
      "step": 21325
    },
    {
      "entropy": 6.273269748687744,
      "epoch": 2.4616272360069242,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004402946726490097,
      "loss": 5.6961,
      "mean_token_accuracy": 0.20423348248004913,
      "num_tokens": 54074000.0,
      "step": 21330
    },
    {
      "entropy": 6.2106773376464846,
      "epoch": 2.462204270051933,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044026667390942867,
      "loss": 5.5408,
      "mean_token_accuracy": 0.2100485548377037,
      "num_tokens": 54087005.0,
      "step": 21335
    },
    {
      "entropy": 6.210607624053955,
      "epoch": 2.462781304096942,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004402386696111851,
      "loss": 5.5945,
      "mean_token_accuracy": 0.206695619225502,
      "num_tokens": 54098850.0,
      "step": 21340
    },
    {
      "entropy": 6.123465490341187,
      "epoch": 2.4633583381419504,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044021065975522096,
      "loss": 5.5279,
      "mean_token_accuracy": 0.21508188992738725,
      "num_tokens": 54112079.0,
      "step": 21345
    },
    {
      "entropy": 6.225316905975342,
      "epoch": 2.463935372186959,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004401826443424784,
      "loss": 5.7298,
      "mean_token_accuracy": 0.19987537562847138,
      "num_tokens": 54124473.0,
      "step": 21350
    },
    {
      "entropy": 6.334025907516479,
      "epoch": 2.4645124062319677,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044015462337389954,
      "loss": 5.7402,
      "mean_token_accuracy": 0.1886001318693161,
      "num_tokens": 54137367.0,
      "step": 21355
    },
    {
      "entropy": 6.329929685592651,
      "epoch": 2.4650894402769765,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044012659685042704,
      "loss": 5.6732,
      "mean_token_accuracy": 0.2010939121246338,
      "num_tokens": 54150047.0,
      "step": 21360
    },
    {
      "entropy": 6.312292623519897,
      "epoch": 2.465666474321985,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004400985647730034,
      "loss": 5.6487,
      "mean_token_accuracy": 0.1995648890733719,
      "num_tokens": 54162728.0,
      "step": 21365
    },
    {
      "entropy": 6.2892578125,
      "epoch": 2.466243508366994,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00044007052714257165,
      "loss": 5.6376,
      "mean_token_accuracy": 0.20172197073698045,
      "num_tokens": 54175116.0,
      "step": 21370
    },
    {
      "entropy": 6.345024824142456,
      "epoch": 2.466820542412002,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004400424839600747,
      "loss": 5.8509,
      "mean_token_accuracy": 0.1897601529955864,
      "num_tokens": 54187721.0,
      "step": 21375
    },
    {
      "entropy": 6.303437995910644,
      "epoch": 2.467397576457011,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044001443522645575,
      "loss": 5.7211,
      "mean_token_accuracy": 0.20107627511024476,
      "num_tokens": 54200879.0,
      "step": 21380
    },
    {
      "entropy": 6.325482273101807,
      "epoch": 2.4679746105020195,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004399863809426582,
      "loss": 5.8245,
      "mean_token_accuracy": 0.19167621284723282,
      "num_tokens": 54212679.0,
      "step": 21385
    },
    {
      "entropy": 6.4041656970977785,
      "epoch": 2.4685516445470284,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043995832110962576,
      "loss": 5.8115,
      "mean_token_accuracy": 0.1952131360769272,
      "num_tokens": 54226332.0,
      "step": 21390
    },
    {
      "entropy": 6.2952347755432125,
      "epoch": 2.4691286785920368,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00043993025572830203,
      "loss": 5.7062,
      "mean_token_accuracy": 0.1985270231962204,
      "num_tokens": 54239819.0,
      "step": 21395
    },
    {
      "entropy": 6.326632928848267,
      "epoch": 2.4697057126370456,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043990218479963115,
      "loss": 5.7373,
      "mean_token_accuracy": 0.19882996678352355,
      "num_tokens": 54251760.0,
      "step": 21400
    },
    {
      "entropy": 6.3600182056427,
      "epoch": 2.470282746682054,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043987410832455715,
      "loss": 5.7265,
      "mean_token_accuracy": 0.2009703993797302,
      "num_tokens": 54264411.0,
      "step": 21405
    },
    {
      "entropy": 6.355018854141235,
      "epoch": 2.470859780727063,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043984602630402447,
      "loss": 5.7714,
      "mean_token_accuracy": 0.1956922635436058,
      "num_tokens": 54275644.0,
      "step": 21410
    },
    {
      "entropy": 6.173326301574707,
      "epoch": 2.4714368147720718,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043981793873897756,
      "loss": 5.5995,
      "mean_token_accuracy": 0.21567746698856355,
      "num_tokens": 54287806.0,
      "step": 21415
    },
    {
      "entropy": 6.196595811843872,
      "epoch": 2.47201384881708,
      "grad_norm": 1.0,
      "learning_rate": 0.0004397898456303612,
      "loss": 5.6127,
      "mean_token_accuracy": 0.21363929957151412,
      "num_tokens": 54301379.0,
      "step": 21420
    },
    {
      "entropy": 6.2627510070800785,
      "epoch": 2.472590882862089,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043976174697912015,
      "loss": 5.6514,
      "mean_token_accuracy": 0.20803814977407456,
      "num_tokens": 54313376.0,
      "step": 21425
    },
    {
      "entropy": 6.230313682556153,
      "epoch": 2.4731679169070975,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004397336427861996,
      "loss": 5.5967,
      "mean_token_accuracy": 0.20610441118478776,
      "num_tokens": 54326838.0,
      "step": 21430
    },
    {
      "entropy": 6.36318826675415,
      "epoch": 2.4737449509521063,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004397055330525448,
      "loss": 5.7415,
      "mean_token_accuracy": 0.19543064683675765,
      "num_tokens": 54339771.0,
      "step": 21435
    },
    {
      "entropy": 6.260603713989258,
      "epoch": 2.4743219849971148,
      "grad_norm": 1.0,
      "learning_rate": 0.0004396774177791012,
      "loss": 5.6045,
      "mean_token_accuracy": 0.20064403265714645,
      "num_tokens": 54352962.0,
      "step": 21440
    },
    {
      "entropy": 6.228683567047119,
      "epoch": 2.4748990190421236,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00043964929696681444,
      "loss": 5.693,
      "mean_token_accuracy": 0.19991200268268586,
      "num_tokens": 54364779.0,
      "step": 21445
    },
    {
      "entropy": 6.234091663360596,
      "epoch": 2.475476053087132,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004396211706166305,
      "loss": 5.6184,
      "mean_token_accuracy": 0.20524659752845764,
      "num_tokens": 54376561.0,
      "step": 21450
    },
    {
      "entropy": 6.356166172027588,
      "epoch": 2.476053087132141,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004395930387294951,
      "loss": 5.721,
      "mean_token_accuracy": 0.2022240102291107,
      "num_tokens": 54389321.0,
      "step": 21455
    },
    {
      "entropy": 6.331257343292236,
      "epoch": 2.4766301211771493,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004395649013063546,
      "loss": 5.7582,
      "mean_token_accuracy": 0.19299271404743196,
      "num_tokens": 54402234.0,
      "step": 21460
    },
    {
      "entropy": 6.257222414016724,
      "epoch": 2.477207155222158,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004395367583481554,
      "loss": 5.6611,
      "mean_token_accuracy": 0.2078343704342842,
      "num_tokens": 54414863.0,
      "step": 21465
    },
    {
      "entropy": 6.3436931610107425,
      "epoch": 2.4777841892671666,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043950860985584423,
      "loss": 5.6439,
      "mean_token_accuracy": 0.20503326654434204,
      "num_tokens": 54425972.0,
      "step": 21470
    },
    {
      "entropy": 6.333819103240967,
      "epoch": 2.4783612233121755,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004394804558303675,
      "loss": 5.7452,
      "mean_token_accuracy": 0.19452296942472458,
      "num_tokens": 54438871.0,
      "step": 21475
    },
    {
      "entropy": 6.250410413742065,
      "epoch": 2.478938257357184,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004394522962726724,
      "loss": 5.724,
      "mean_token_accuracy": 0.19858806878328322,
      "num_tokens": 54451877.0,
      "step": 21480
    },
    {
      "entropy": 6.291604232788086,
      "epoch": 2.4795152914021927,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00043942413118370604,
      "loss": 5.6408,
      "mean_token_accuracy": 0.20981614887714387,
      "num_tokens": 54464126.0,
      "step": 21485
    },
    {
      "entropy": 6.335710048675537,
      "epoch": 2.4800923254472016,
      "grad_norm": 1.0,
      "learning_rate": 0.0004393959605644157,
      "loss": 5.7625,
      "mean_token_accuracy": 0.19183712750673293,
      "num_tokens": 54476569.0,
      "step": 21490
    },
    {
      "entropy": 6.264228677749633,
      "epoch": 2.48066935949221,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004393677844157488,
      "loss": 5.636,
      "mean_token_accuracy": 0.20429372489452363,
      "num_tokens": 54488719.0,
      "step": 21495
    },
    {
      "entropy": 6.2780975818634035,
      "epoch": 2.481246393537219,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004393396027386532,
      "loss": 5.6559,
      "mean_token_accuracy": 0.20973989218473435,
      "num_tokens": 54501324.0,
      "step": 21500
    },
    {
      "entropy": 6.255328559875489,
      "epoch": 2.4818234275822273,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043931141553407667,
      "loss": 5.6975,
      "mean_token_accuracy": 0.19942784309387207,
      "num_tokens": 54514369.0,
      "step": 21505
    },
    {
      "entropy": 6.301109886169433,
      "epoch": 2.482400461627236,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004392832228029672,
      "loss": 5.6931,
      "mean_token_accuracy": 0.20155691504478454,
      "num_tokens": 54527368.0,
      "step": 21510
    },
    {
      "entropy": 6.363375186920166,
      "epoch": 2.4829774956722446,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00043925502454627327,
      "loss": 5.7635,
      "mean_token_accuracy": 0.1982794240117073,
      "num_tokens": 54540160.0,
      "step": 21515
    },
    {
      "entropy": 6.2854323387146,
      "epoch": 2.4835545297172534,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004392268207649431,
      "loss": 5.6554,
      "mean_token_accuracy": 0.21472620218992233,
      "num_tokens": 54552718.0,
      "step": 21520
    },
    {
      "entropy": 6.3139947891235355,
      "epoch": 2.484131563762262,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043919861145992546,
      "loss": 5.7426,
      "mean_token_accuracy": 0.19724248945713044,
      "num_tokens": 54565065.0,
      "step": 21525
    },
    {
      "entropy": 6.363930940628052,
      "epoch": 2.4847085978072707,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000439170396632169,
      "loss": 5.7299,
      "mean_token_accuracy": 0.19429460912942886,
      "num_tokens": 54577608.0,
      "step": 21530
    },
    {
      "entropy": 6.224084424972534,
      "epoch": 2.485285631852279,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004391421762826229,
      "loss": 5.5992,
      "mean_token_accuracy": 0.20149815380573272,
      "num_tokens": 54590410.0,
      "step": 21535
    },
    {
      "entropy": 6.286251354217529,
      "epoch": 2.485862665897288,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004391139504122362,
      "loss": 5.7186,
      "mean_token_accuracy": 0.2033986493945122,
      "num_tokens": 54603231.0,
      "step": 21540
    },
    {
      "entropy": 6.34293532371521,
      "epoch": 2.4864396999422964,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043908571902195827,
      "loss": 5.743,
      "mean_token_accuracy": 0.20485443025827407,
      "num_tokens": 54614666.0,
      "step": 21545
    },
    {
      "entropy": 6.350252246856689,
      "epoch": 2.4870167339873053,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043905748211273864,
      "loss": 5.7609,
      "mean_token_accuracy": 0.20120716094970703,
      "num_tokens": 54627715.0,
      "step": 21550
    },
    {
      "entropy": 6.343413877487182,
      "epoch": 2.4875937680323137,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043902923968552714,
      "loss": 5.6626,
      "mean_token_accuracy": 0.20358366370201111,
      "num_tokens": 54638876.0,
      "step": 21555
    },
    {
      "entropy": 6.257961750030518,
      "epoch": 2.4881708020773226,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004390009917412737,
      "loss": 5.64,
      "mean_token_accuracy": 0.21005151569843292,
      "num_tokens": 54650733.0,
      "step": 21560
    },
    {
      "entropy": 6.2547454833984375,
      "epoch": 2.4887478361223314,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043897273828092824,
      "loss": 5.6837,
      "mean_token_accuracy": 0.2011951357126236,
      "num_tokens": 54662333.0,
      "step": 21565
    },
    {
      "entropy": 6.298535251617432,
      "epoch": 2.48932487016734,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004389444793054413,
      "loss": 5.6857,
      "mean_token_accuracy": 0.2024449571967125,
      "num_tokens": 54673629.0,
      "step": 21570
    },
    {
      "entropy": 6.271597862243652,
      "epoch": 2.4899019042123487,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043891621481576324,
      "loss": 5.663,
      "mean_token_accuracy": 0.20982982814311982,
      "num_tokens": 54686876.0,
      "step": 21575
    },
    {
      "entropy": 6.2440308094024655,
      "epoch": 2.490478938257357,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004388879448128446,
      "loss": 5.5978,
      "mean_token_accuracy": 0.2143885925412178,
      "num_tokens": 54699410.0,
      "step": 21580
    },
    {
      "entropy": 6.26668906211853,
      "epoch": 2.491055972302366,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043885966929763635,
      "loss": 5.7288,
      "mean_token_accuracy": 0.20049264132976533,
      "num_tokens": 54711773.0,
      "step": 21585
    },
    {
      "entropy": 6.320494937896728,
      "epoch": 2.4916330063473744,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043883138827108964,
      "loss": 5.7187,
      "mean_token_accuracy": 0.20028091073036194,
      "num_tokens": 54724819.0,
      "step": 21590
    },
    {
      "entropy": 6.2549858570098875,
      "epoch": 2.4922100403923833,
      "grad_norm": 1.125,
      "learning_rate": 0.0004388031017341555,
      "loss": 5.6183,
      "mean_token_accuracy": 0.20552769303321838,
      "num_tokens": 54737139.0,
      "step": 21595
    },
    {
      "entropy": 6.102586460113526,
      "epoch": 2.4927870744373917,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043877480968778536,
      "loss": 5.5085,
      "mean_token_accuracy": 0.21529248207807541,
      "num_tokens": 54749792.0,
      "step": 21600
    },
    {
      "entropy": 6.277773189544678,
      "epoch": 2.4933641084824005,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004387465121329309,
      "loss": 5.6689,
      "mean_token_accuracy": 0.20449299067258836,
      "num_tokens": 54762464.0,
      "step": 21605
    },
    {
      "entropy": 6.307204914093018,
      "epoch": 2.493941142527409,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043871820907054375,
      "loss": 5.6794,
      "mean_token_accuracy": 0.20920268446207047,
      "num_tokens": 54775955.0,
      "step": 21610
    },
    {
      "entropy": 6.263454627990723,
      "epoch": 2.494518176572418,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043868990050157617,
      "loss": 5.6546,
      "mean_token_accuracy": 0.20594882369041442,
      "num_tokens": 54787953.0,
      "step": 21615
    },
    {
      "entropy": 6.246506690979004,
      "epoch": 2.4950952106174267,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004386615864269799,
      "loss": 5.6857,
      "mean_token_accuracy": 0.2027042806148529,
      "num_tokens": 54802749.0,
      "step": 21620
    },
    {
      "entropy": 6.294088220596313,
      "epoch": 2.495672244662435,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043863326684770746,
      "loss": 5.6041,
      "mean_token_accuracy": 0.20658694505691527,
      "num_tokens": 54814362.0,
      "step": 21625
    },
    {
      "entropy": 6.289592933654785,
      "epoch": 2.4962492787074435,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004386049417647115,
      "loss": 5.6772,
      "mean_token_accuracy": 0.20580474585294722,
      "num_tokens": 54826739.0,
      "step": 21630
    },
    {
      "entropy": 6.21641354560852,
      "epoch": 2.4968263127524524,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004385766111789445,
      "loss": 5.6376,
      "mean_token_accuracy": 0.20230588763952256,
      "num_tokens": 54840431.0,
      "step": 21635
    },
    {
      "entropy": 6.174298524856567,
      "epoch": 2.4974033467974612,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004385482750913595,
      "loss": 5.5335,
      "mean_token_accuracy": 0.215751776099205,
      "num_tokens": 54852853.0,
      "step": 21640
    },
    {
      "entropy": 6.2415771484375,
      "epoch": 2.4979803808424696,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004385199335029094,
      "loss": 5.6516,
      "mean_token_accuracy": 0.20901017189025878,
      "num_tokens": 54865221.0,
      "step": 21645
    },
    {
      "entropy": 6.188607120513916,
      "epoch": 2.4985574148874785,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043849158641454767,
      "loss": 5.589,
      "mean_token_accuracy": 0.20466280430555345,
      "num_tokens": 54877466.0,
      "step": 21650
    },
    {
      "entropy": 6.226304912567139,
      "epoch": 2.499134448932487,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043846323382722765,
      "loss": 5.6924,
      "mean_token_accuracy": 0.2046091765165329,
      "num_tokens": 54890965.0,
      "step": 21655
    },
    {
      "entropy": 6.326944875717163,
      "epoch": 2.499711482977496,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00043843487574190285,
      "loss": 5.7314,
      "mean_token_accuracy": 0.19542012214660645,
      "num_tokens": 54903807.0,
      "step": 21660
    },
    {
      "entropy": 6.1907285213470455,
      "epoch": 2.500288517022504,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043840651215952726,
      "loss": 5.5955,
      "mean_token_accuracy": 0.2120587036013603,
      "num_tokens": 54916144.0,
      "step": 21665
    },
    {
      "entropy": 6.288946485519409,
      "epoch": 2.500865551067513,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043837814308105477,
      "loss": 5.7295,
      "mean_token_accuracy": 0.20089642107486724,
      "num_tokens": 54928927.0,
      "step": 21670
    },
    {
      "entropy": 6.361982965469361,
      "epoch": 2.5014425851125215,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043834976850743964,
      "loss": 5.717,
      "mean_token_accuracy": 0.19631865322589875,
      "num_tokens": 54940423.0,
      "step": 21675
    },
    {
      "entropy": 6.227539443969727,
      "epoch": 2.5020196191575304,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004383213884396361,
      "loss": 5.6404,
      "mean_token_accuracy": 0.20412955433130264,
      "num_tokens": 54954634.0,
      "step": 21680
    },
    {
      "entropy": 6.225951147079468,
      "epoch": 2.5025966532025388,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004382930028785989,
      "loss": 5.6773,
      "mean_token_accuracy": 0.1998716801404953,
      "num_tokens": 54967092.0,
      "step": 21685
    },
    {
      "entropy": 6.227726316452026,
      "epoch": 2.5031736872475476,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004382646118252826,
      "loss": 5.6732,
      "mean_token_accuracy": 0.20380735546350479,
      "num_tokens": 54979421.0,
      "step": 21690
    },
    {
      "entropy": 6.358871936798096,
      "epoch": 2.5037507212925565,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043823621528064216,
      "loss": 5.8034,
      "mean_token_accuracy": 0.18986698240041733,
      "num_tokens": 54992743.0,
      "step": 21695
    },
    {
      "entropy": 6.342610502243042,
      "epoch": 2.504327755337565,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043820781324563276,
      "loss": 5.723,
      "mean_token_accuracy": 0.1946076661348343,
      "num_tokens": 55004948.0,
      "step": 21700
    },
    {
      "entropy": 6.3717139720916744,
      "epoch": 2.5049047893825733,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00043817940572120963,
      "loss": 5.7342,
      "mean_token_accuracy": 0.1978147804737091,
      "num_tokens": 55017645.0,
      "step": 21705
    },
    {
      "entropy": 6.214679431915283,
      "epoch": 2.505481823427582,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043815099270832814,
      "loss": 5.6308,
      "mean_token_accuracy": 0.20393857806921006,
      "num_tokens": 55029466.0,
      "step": 21710
    },
    {
      "entropy": 6.282233142852784,
      "epoch": 2.506058857472591,
      "grad_norm": 1.0,
      "learning_rate": 0.00043812257420794415,
      "loss": 5.6899,
      "mean_token_accuracy": 0.20416367053985596,
      "num_tokens": 55042473.0,
      "step": 21715
    },
    {
      "entropy": 6.33713583946228,
      "epoch": 2.5066358915175995,
      "grad_norm": 0.953125,
      "learning_rate": 0.00043809415022101335,
      "loss": 5.7206,
      "mean_token_accuracy": 0.19984404891729354,
      "num_tokens": 55055984.0,
      "step": 21720
    },
    {
      "entropy": 6.220205450057984,
      "epoch": 2.5072129255626083,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004380657207484919,
      "loss": 5.5965,
      "mean_token_accuracy": 0.21211631298065187,
      "num_tokens": 55068945.0,
      "step": 21725
    },
    {
      "entropy": 6.3032379150390625,
      "epoch": 2.5077899596076167,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043803728579133586,
      "loss": 5.6612,
      "mean_token_accuracy": 0.20086454898118972,
      "num_tokens": 55082284.0,
      "step": 21730
    },
    {
      "entropy": 6.282631683349609,
      "epoch": 2.5083669936526256,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004380088453505017,
      "loss": 5.6718,
      "mean_token_accuracy": 0.19950748682022096,
      "num_tokens": 55093779.0,
      "step": 21735
    },
    {
      "entropy": 6.353281497955322,
      "epoch": 2.508944027697634,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00043798039942694603,
      "loss": 5.7614,
      "mean_token_accuracy": 0.19599814414978028,
      "num_tokens": 55106499.0,
      "step": 21740
    },
    {
      "entropy": 6.211190843582154,
      "epoch": 2.509521061742643,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043795194802162557,
      "loss": 5.5272,
      "mean_token_accuracy": 0.20966080129146575,
      "num_tokens": 55119644.0,
      "step": 21745
    },
    {
      "entropy": 6.248804187774658,
      "epoch": 2.5100980957876513,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004379234911354973,
      "loss": 5.644,
      "mean_token_accuracy": 0.20803987681865693,
      "num_tokens": 55131987.0,
      "step": 21750
    },
    {
      "entropy": 6.277766275405884,
      "epoch": 2.51067512983266,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043789502876951834,
      "loss": 5.6609,
      "mean_token_accuracy": 0.20360077619552613,
      "num_tokens": 55144228.0,
      "step": 21755
    },
    {
      "entropy": 6.268870162963867,
      "epoch": 2.5112521638776686,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004378665609246459,
      "loss": 5.7035,
      "mean_token_accuracy": 0.20455182790756227,
      "num_tokens": 55156166.0,
      "step": 21760
    },
    {
      "entropy": 6.332865953445435,
      "epoch": 2.5118291979226774,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043783808760183764,
      "loss": 5.6952,
      "mean_token_accuracy": 0.1992311269044876,
      "num_tokens": 55167373.0,
      "step": 21765
    },
    {
      "entropy": 6.416872453689575,
      "epoch": 2.5124062319676863,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043780960880205125,
      "loss": 5.7189,
      "mean_token_accuracy": 0.20526453852653503,
      "num_tokens": 55181480.0,
      "step": 21770
    },
    {
      "entropy": 6.316271209716797,
      "epoch": 2.5129832660126947,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043778112452624445,
      "loss": 5.6928,
      "mean_token_accuracy": 0.19676965475082397,
      "num_tokens": 55193402.0,
      "step": 21775
    },
    {
      "entropy": 6.177290916442871,
      "epoch": 2.513560300057703,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043775263477537546,
      "loss": 5.6114,
      "mean_token_accuracy": 0.20639609396457673,
      "num_tokens": 55206265.0,
      "step": 21780
    },
    {
      "entropy": 6.333525800704956,
      "epoch": 2.514137334102712,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004377241395504024,
      "loss": 5.7522,
      "mean_token_accuracy": 0.1994367480278015,
      "num_tokens": 55218161.0,
      "step": 21785
    },
    {
      "entropy": 6.327044296264648,
      "epoch": 2.514714368147721,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004376956388522837,
      "loss": 5.6735,
      "mean_token_accuracy": 0.2044881671667099,
      "num_tokens": 55231180.0,
      "step": 21790
    },
    {
      "entropy": 6.249074935913086,
      "epoch": 2.5152914021927293,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004376671326819781,
      "loss": 5.7113,
      "mean_token_accuracy": 0.19978694766759872,
      "num_tokens": 55244124.0,
      "step": 21795
    },
    {
      "entropy": 6.314837026596069,
      "epoch": 2.515868436237738,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043763862104044424,
      "loss": 5.7265,
      "mean_token_accuracy": 0.19820688366889955,
      "num_tokens": 55256424.0,
      "step": 21800
    },
    {
      "entropy": 6.322049570083618,
      "epoch": 2.5164454702827466,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043761010392864114,
      "loss": 5.7082,
      "mean_token_accuracy": 0.20131746232509612,
      "num_tokens": 55268683.0,
      "step": 21805
    },
    {
      "entropy": 6.354723834991455,
      "epoch": 2.5170225043277554,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00043758158134752793,
      "loss": 5.7891,
      "mean_token_accuracy": 0.19384117275476456,
      "num_tokens": 55281277.0,
      "step": 21810
    },
    {
      "entropy": 6.307178163528443,
      "epoch": 2.517599538372764,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004375530532980642,
      "loss": 5.6763,
      "mean_token_accuracy": 0.19684169590473174,
      "num_tokens": 55293565.0,
      "step": 21815
    },
    {
      "entropy": 6.282255172729492,
      "epoch": 2.5181765724177727,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000437524519781209,
      "loss": 5.6879,
      "mean_token_accuracy": 0.19828752875328065,
      "num_tokens": 55306189.0,
      "step": 21820
    },
    {
      "entropy": 6.2387855052948,
      "epoch": 2.5187536064627816,
      "grad_norm": 1.125,
      "learning_rate": 0.0004374959807979224,
      "loss": 5.608,
      "mean_token_accuracy": 0.20779090821743013,
      "num_tokens": 55319165.0,
      "step": 21825
    },
    {
      "entropy": 6.266111087799072,
      "epoch": 2.51933064050779,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004374674363491642,
      "loss": 5.6123,
      "mean_token_accuracy": 0.20428169518709183,
      "num_tokens": 55331890.0,
      "step": 21830
    },
    {
      "entropy": 6.2766142845153805,
      "epoch": 2.5199076745527984,
      "grad_norm": 0.890625,
      "learning_rate": 0.00043743888643589453,
      "loss": 5.7264,
      "mean_token_accuracy": 0.19554169327020646,
      "num_tokens": 55344789.0,
      "step": 21835
    },
    {
      "entropy": 6.351927661895752,
      "epoch": 2.5204847085978073,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043741033105907355,
      "loss": 5.715,
      "mean_token_accuracy": 0.1972411334514618,
      "num_tokens": 55357184.0,
      "step": 21840
    },
    {
      "entropy": 6.331570243835449,
      "epoch": 2.521061742642816,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004373817702196618,
      "loss": 5.7197,
      "mean_token_accuracy": 0.20283523350954055,
      "num_tokens": 55369890.0,
      "step": 21845
    },
    {
      "entropy": 6.247327995300293,
      "epoch": 2.5216387766878245,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043735320391861986,
      "loss": 5.644,
      "mean_token_accuracy": 0.19415172040462494,
      "num_tokens": 55383191.0,
      "step": 21850
    },
    {
      "entropy": 6.250979804992676,
      "epoch": 2.522215810732833,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004373246321569085,
      "loss": 5.6031,
      "mean_token_accuracy": 0.20793095380067825,
      "num_tokens": 55396329.0,
      "step": 21855
    },
    {
      "entropy": 6.28544135093689,
      "epoch": 2.522792844777842,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004372960549354888,
      "loss": 5.7207,
      "mean_token_accuracy": 0.19757884591817856,
      "num_tokens": 55407977.0,
      "step": 21860
    },
    {
      "entropy": 6.267713785171509,
      "epoch": 2.5233698788228507,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00043726747225532195,
      "loss": 5.6772,
      "mean_token_accuracy": 0.20953007638454438,
      "num_tokens": 55419869.0,
      "step": 21865
    },
    {
      "entropy": 6.244552087783814,
      "epoch": 2.523946912867859,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004372388841173692,
      "loss": 5.626,
      "mean_token_accuracy": 0.20370595902204514,
      "num_tokens": 55432129.0,
      "step": 21870
    },
    {
      "entropy": 6.271215915679932,
      "epoch": 2.524523946912868,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004372102905225922,
      "loss": 5.6731,
      "mean_token_accuracy": 0.1960826426744461,
      "num_tokens": 55446415.0,
      "step": 21875
    },
    {
      "entropy": 6.334109163284301,
      "epoch": 2.5251009809578764,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043718169147195275,
      "loss": 5.6835,
      "mean_token_accuracy": 0.20029303580522537,
      "num_tokens": 55458395.0,
      "step": 21880
    },
    {
      "entropy": 6.276555299758911,
      "epoch": 2.5256780150028852,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00043715308696641255,
      "loss": 5.6867,
      "mean_token_accuracy": 0.20147744417190552,
      "num_tokens": 55470875.0,
      "step": 21885
    },
    {
      "entropy": 6.242990303039551,
      "epoch": 2.5262550490478937,
      "grad_norm": 1.0,
      "learning_rate": 0.0004371244770069338,
      "loss": 5.7209,
      "mean_token_accuracy": 0.19794623255729676,
      "num_tokens": 55483307.0,
      "step": 21890
    },
    {
      "entropy": 6.345460367202759,
      "epoch": 2.5268320830929025,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043709586159447884,
      "loss": 5.7506,
      "mean_token_accuracy": 0.20299078971147538,
      "num_tokens": 55496353.0,
      "step": 21895
    },
    {
      "entropy": 6.334553146362305,
      "epoch": 2.5274091171379114,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043706724073001017,
      "loss": 5.7421,
      "mean_token_accuracy": 0.2000865250825882,
      "num_tokens": 55509354.0,
      "step": 21900
    },
    {
      "entropy": 6.297677373886108,
      "epoch": 2.52798615118292,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004370386144144902,
      "loss": 5.6515,
      "mean_token_accuracy": 0.20190757811069487,
      "num_tokens": 55521860.0,
      "step": 21905
    },
    {
      "entropy": 6.194522047042847,
      "epoch": 2.5285631852279282,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004370099826488821,
      "loss": 5.5301,
      "mean_token_accuracy": 0.21408119946718215,
      "num_tokens": 55534335.0,
      "step": 21910
    },
    {
      "entropy": 6.249485206604004,
      "epoch": 2.529140219272937,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004369813454341486,
      "loss": 5.6222,
      "mean_token_accuracy": 0.21350693106651306,
      "num_tokens": 55547282.0,
      "step": 21915
    },
    {
      "entropy": 6.229116058349609,
      "epoch": 2.529717253317946,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00043695270277125303,
      "loss": 5.6093,
      "mean_token_accuracy": 0.20875063687562942,
      "num_tokens": 55560419.0,
      "step": 21920
    },
    {
      "entropy": 6.291175365447998,
      "epoch": 2.5302942873629544,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004369240546611587,
      "loss": 5.7471,
      "mean_token_accuracy": 0.20174427926540375,
      "num_tokens": 55573067.0,
      "step": 21925
    },
    {
      "entropy": 6.249997568130493,
      "epoch": 2.5308713214079632,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043689540110482933,
      "loss": 5.626,
      "mean_token_accuracy": 0.20564635396003722,
      "num_tokens": 55586753.0,
      "step": 21930
    },
    {
      "entropy": 6.305003547668457,
      "epoch": 2.5314483554529716,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004368667421032285,
      "loss": 5.7678,
      "mean_token_accuracy": 0.19207421094179153,
      "num_tokens": 55599571.0,
      "step": 21935
    },
    {
      "entropy": 6.359702396392822,
      "epoch": 2.5320253894979805,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043683807765732013,
      "loss": 5.7422,
      "mean_token_accuracy": 0.19449844360351562,
      "num_tokens": 55611094.0,
      "step": 21940
    },
    {
      "entropy": 6.33309268951416,
      "epoch": 2.532602423542989,
      "grad_norm": 1.0,
      "learning_rate": 0.00043680940776806847,
      "loss": 5.7328,
      "mean_token_accuracy": 0.20769111961126327,
      "num_tokens": 55625872.0,
      "step": 21945
    },
    {
      "entropy": 6.34452714920044,
      "epoch": 2.533179457587998,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004367807324364378,
      "loss": 5.7268,
      "mean_token_accuracy": 0.2009881004691124,
      "num_tokens": 55638833.0,
      "step": 21950
    },
    {
      "entropy": 6.342330503463745,
      "epoch": 2.533756491633006,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043675205166339247,
      "loss": 5.7188,
      "mean_token_accuracy": 0.19964758157730103,
      "num_tokens": 55651851.0,
      "step": 21955
    },
    {
      "entropy": 6.245658826828003,
      "epoch": 2.534333525678015,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004367233654498973,
      "loss": 5.6017,
      "mean_token_accuracy": 0.21168553084135056,
      "num_tokens": 55664919.0,
      "step": 21960
    },
    {
      "entropy": 6.253633260726929,
      "epoch": 2.5349105597230235,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004366946737969171,
      "loss": 5.6637,
      "mean_token_accuracy": 0.20350811928510665,
      "num_tokens": 55677179.0,
      "step": 21965
    },
    {
      "entropy": 6.31117467880249,
      "epoch": 2.5354875937680323,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004366659767054168,
      "loss": 5.6593,
      "mean_token_accuracy": 0.20240916609764098,
      "num_tokens": 55690300.0,
      "step": 21970
    },
    {
      "entropy": 6.307062196731567,
      "epoch": 2.536064627813041,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043663727417636166,
      "loss": 5.6915,
      "mean_token_accuracy": 0.20666760355234146,
      "num_tokens": 55703395.0,
      "step": 21975
    },
    {
      "entropy": 6.203319025039673,
      "epoch": 2.5366416618580496,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004366085662107171,
      "loss": 5.6607,
      "mean_token_accuracy": 0.20390082597732545,
      "num_tokens": 55717008.0,
      "step": 21980
    },
    {
      "entropy": 6.28852915763855,
      "epoch": 2.537218695903058,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043657985280944875,
      "loss": 5.6499,
      "mean_token_accuracy": 0.2063484862446785,
      "num_tokens": 55729982.0,
      "step": 21985
    },
    {
      "entropy": 6.330930519104004,
      "epoch": 2.537795729948067,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004365511339735223,
      "loss": 5.688,
      "mean_token_accuracy": 0.2060350388288498,
      "num_tokens": 55742035.0,
      "step": 21990
    },
    {
      "entropy": 6.292611074447632,
      "epoch": 2.5383727639930758,
      "grad_norm": 1.0,
      "learning_rate": 0.0004365224097039036,
      "loss": 5.7958,
      "mean_token_accuracy": 0.1925215393304825,
      "num_tokens": 55755974.0,
      "step": 21995
    },
    {
      "entropy": 6.2156596183776855,
      "epoch": 2.538949798038084,
      "grad_norm": 1.015625,
      "learning_rate": 0.000436493680001559,
      "loss": 5.5673,
      "mean_token_accuracy": 0.21436562687158583,
      "num_tokens": 55767893.0,
      "step": 22000
    },
    {
      "entropy": 6.386861085891724,
      "epoch": 2.539526832083093,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043646494486745464,
      "loss": 5.7381,
      "mean_token_accuracy": 0.20206450521945954,
      "num_tokens": 55780530.0,
      "step": 22005
    },
    {
      "entropy": 6.250147962570191,
      "epoch": 2.5401038661281015,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004364362043025571,
      "loss": 5.6368,
      "mean_token_accuracy": 0.20497675687074662,
      "num_tokens": 55791957.0,
      "step": 22010
    },
    {
      "entropy": 6.197332286834717,
      "epoch": 2.5406809001731103,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043640745830783297,
      "loss": 5.6572,
      "mean_token_accuracy": 0.20558187663555144,
      "num_tokens": 55804123.0,
      "step": 22015
    },
    {
      "entropy": 6.260128736495972,
      "epoch": 2.5412579342181187,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004363787068842491,
      "loss": 5.6045,
      "mean_token_accuracy": 0.20882643312215804,
      "num_tokens": 55816555.0,
      "step": 22020
    },
    {
      "entropy": 6.323992776870727,
      "epoch": 2.5418349682631276,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004363499500327727,
      "loss": 5.7295,
      "mean_token_accuracy": 0.20008298009634018,
      "num_tokens": 55829206.0,
      "step": 22025
    },
    {
      "entropy": 6.234121513366699,
      "epoch": 2.542412002308136,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00043632118775437085,
      "loss": 5.621,
      "mean_token_accuracy": 0.2042347326874733,
      "num_tokens": 55841722.0,
      "step": 22030
    },
    {
      "entropy": 6.349940538406372,
      "epoch": 2.542989036353145,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00043629242005001096,
      "loss": 5.6704,
      "mean_token_accuracy": 0.2028152033686638,
      "num_tokens": 55854624.0,
      "step": 22035
    },
    {
      "entropy": 6.271018838882446,
      "epoch": 2.5435660703981533,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004362636469206607,
      "loss": 5.6998,
      "mean_token_accuracy": 0.198223277926445,
      "num_tokens": 55867030.0,
      "step": 22040
    },
    {
      "entropy": 6.192486667633057,
      "epoch": 2.544143104443162,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004362348683672877,
      "loss": 5.5599,
      "mean_token_accuracy": 0.21324284225702286,
      "num_tokens": 55879837.0,
      "step": 22045
    },
    {
      "entropy": 6.310703277587891,
      "epoch": 2.544720138488171,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004362060843908601,
      "loss": 5.6477,
      "mean_token_accuracy": 0.2001812294125557,
      "num_tokens": 55892022.0,
      "step": 22050
    },
    {
      "entropy": 6.319234800338745,
      "epoch": 2.5452971725331794,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004361772949923458,
      "loss": 5.6158,
      "mean_token_accuracy": 0.20543404072523117,
      "num_tokens": 55904535.0,
      "step": 22055
    },
    {
      "entropy": 6.28077392578125,
      "epoch": 2.545874206578188,
      "grad_norm": 1.0,
      "learning_rate": 0.0004361485001727132,
      "loss": 5.7347,
      "mean_token_accuracy": 0.1960916504263878,
      "num_tokens": 55916372.0,
      "step": 22060
    },
    {
      "entropy": 6.276385259628296,
      "epoch": 2.5464512406231967,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000436119699932931,
      "loss": 5.6732,
      "mean_token_accuracy": 0.20117932856082915,
      "num_tokens": 55930563.0,
      "step": 22065
    },
    {
      "entropy": 6.2770373821258545,
      "epoch": 2.5470282746682056,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00043609089427396763,
      "loss": 5.62,
      "mean_token_accuracy": 0.2042050302028656,
      "num_tokens": 55942432.0,
      "step": 22070
    },
    {
      "entropy": 6.29171404838562,
      "epoch": 2.547605308713214,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043606208319679215,
      "loss": 5.7171,
      "mean_token_accuracy": 0.2000429630279541,
      "num_tokens": 55954547.0,
      "step": 22075
    },
    {
      "entropy": 6.331845140457153,
      "epoch": 2.548182342758223,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004360332667023734,
      "loss": 5.7403,
      "mean_token_accuracy": 0.19935042709112166,
      "num_tokens": 55967103.0,
      "step": 22080
    },
    {
      "entropy": 6.278891897201538,
      "epoch": 2.5487593768032313,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043600444479168084,
      "loss": 5.6884,
      "mean_token_accuracy": 0.20107759684324264,
      "num_tokens": 55980236.0,
      "step": 22085
    },
    {
      "entropy": 6.2847668647766115,
      "epoch": 2.54933641084824,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004359756174656836,
      "loss": 5.7005,
      "mean_token_accuracy": 0.20582483112812042,
      "num_tokens": 55993015.0,
      "step": 22090
    },
    {
      "entropy": 6.333690500259399,
      "epoch": 2.5499134448932486,
      "grad_norm": 1.0,
      "learning_rate": 0.00043594678472535153,
      "loss": 5.6929,
      "mean_token_accuracy": 0.20122768580913544,
      "num_tokens": 56006057.0,
      "step": 22095
    },
    {
      "entropy": 6.285467529296875,
      "epoch": 2.5504904789382574,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043591794657165425,
      "loss": 5.6217,
      "mean_token_accuracy": 0.2088159829378128,
      "num_tokens": 56018450.0,
      "step": 22100
    },
    {
      "entropy": 6.280207729339599,
      "epoch": 2.5510675129832663,
      "grad_norm": 1.0,
      "learning_rate": 0.0004358891030055617,
      "loss": 5.6911,
      "mean_token_accuracy": 0.2005801171064377,
      "num_tokens": 56031711.0,
      "step": 22105
    },
    {
      "entropy": 6.2584391117095945,
      "epoch": 2.5516445470282747,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004358602540280443,
      "loss": 5.6578,
      "mean_token_accuracy": 0.20714095383882522,
      "num_tokens": 56042988.0,
      "step": 22110
    },
    {
      "entropy": 6.300547695159912,
      "epoch": 2.552221581073283,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043583139964007203,
      "loss": 5.7274,
      "mean_token_accuracy": 0.1969756230711937,
      "num_tokens": 56056927.0,
      "step": 22115
    },
    {
      "entropy": 6.291201734542847,
      "epoch": 2.552798615118292,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043580253984261554,
      "loss": 5.7025,
      "mean_token_accuracy": 0.1952582359313965,
      "num_tokens": 56069148.0,
      "step": 22120
    },
    {
      "entropy": 6.339542627334595,
      "epoch": 2.553375649163301,
      "grad_norm": 1.0,
      "learning_rate": 0.00043577367463664546,
      "loss": 5.7437,
      "mean_token_accuracy": 0.19363578408956528,
      "num_tokens": 56080043.0,
      "step": 22125
    },
    {
      "entropy": 6.199355888366699,
      "epoch": 2.5539526832083093,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004357448040231327,
      "loss": 5.5803,
      "mean_token_accuracy": 0.20987774431705475,
      "num_tokens": 56092723.0,
      "step": 22130
    },
    {
      "entropy": 6.191482782363892,
      "epoch": 2.5545297172533177,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004357159280030483,
      "loss": 5.6184,
      "mean_token_accuracy": 0.20691514760255814,
      "num_tokens": 56103808.0,
      "step": 22135
    },
    {
      "entropy": 6.246099185943604,
      "epoch": 2.5551067512983265,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004356870465773635,
      "loss": 5.6281,
      "mean_token_accuracy": 0.2107979789376259,
      "num_tokens": 56117272.0,
      "step": 22140
    },
    {
      "entropy": 6.288285207748413,
      "epoch": 2.5556837853433354,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043565815974704974,
      "loss": 5.6504,
      "mean_token_accuracy": 0.2035933956503868,
      "num_tokens": 56129752.0,
      "step": 22145
    },
    {
      "entropy": 6.31098575592041,
      "epoch": 2.556260819388344,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043562926751307857,
      "loss": 5.7558,
      "mean_token_accuracy": 0.19477622658014299,
      "num_tokens": 56142819.0,
      "step": 22150
    },
    {
      "entropy": 6.361627721786499,
      "epoch": 2.5568378534333527,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043560036987642177,
      "loss": 5.7578,
      "mean_token_accuracy": 0.20238527655601501,
      "num_tokens": 56155142.0,
      "step": 22155
    },
    {
      "entropy": 6.317470264434815,
      "epoch": 2.557414887478361,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043557146683805134,
      "loss": 5.737,
      "mean_token_accuracy": 0.19400882869958877,
      "num_tokens": 56167217.0,
      "step": 22160
    },
    {
      "entropy": 6.268490982055664,
      "epoch": 2.55799192152337,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004355425583989393,
      "loss": 5.645,
      "mean_token_accuracy": 0.20740866661071777,
      "num_tokens": 56180210.0,
      "step": 22165
    },
    {
      "entropy": 6.3303436756134035,
      "epoch": 2.5585689555683784,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043551364456005816,
      "loss": 5.753,
      "mean_token_accuracy": 0.19600175321102142,
      "num_tokens": 56192288.0,
      "step": 22170
    },
    {
      "entropy": 6.296503019332886,
      "epoch": 2.5591459896133872,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043548472532238015,
      "loss": 5.6296,
      "mean_token_accuracy": 0.20567398518323898,
      "num_tokens": 56204973.0,
      "step": 22175
    },
    {
      "entropy": 6.213656806945801,
      "epoch": 2.559723023658396,
      "grad_norm": 1.0,
      "learning_rate": 0.00043545580068687836,
      "loss": 5.5642,
      "mean_token_accuracy": 0.2136990562081337,
      "num_tokens": 56217649.0,
      "step": 22180
    },
    {
      "entropy": 6.295787668228149,
      "epoch": 2.5603000577034045,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004354268706545252,
      "loss": 5.6487,
      "mean_token_accuracy": 0.20341018587350845,
      "num_tokens": 56229007.0,
      "step": 22185
    },
    {
      "entropy": 6.193588399887085,
      "epoch": 2.560877091748413,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00043539793522629405,
      "loss": 5.5939,
      "mean_token_accuracy": 0.21116895824670792,
      "num_tokens": 56241779.0,
      "step": 22190
    },
    {
      "entropy": 6.287145948410034,
      "epoch": 2.561454125793422,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043536899440315796,
      "loss": 5.7093,
      "mean_token_accuracy": 0.20702861994504929,
      "num_tokens": 56253130.0,
      "step": 22195
    },
    {
      "entropy": 6.411194801330566,
      "epoch": 2.5620311598384307,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004353400481860904,
      "loss": 5.7696,
      "mean_token_accuracy": 0.19092954248189925,
      "num_tokens": 56264423.0,
      "step": 22200
    },
    {
      "entropy": 6.294098281860352,
      "epoch": 2.562608193883439,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043531109657606504,
      "loss": 5.7257,
      "mean_token_accuracy": 0.19906480610370636,
      "num_tokens": 56276622.0,
      "step": 22205
    },
    {
      "entropy": 6.275125503540039,
      "epoch": 2.563185227928448,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004352821395740555,
      "loss": 5.6804,
      "mean_token_accuracy": 0.20354412943124772,
      "num_tokens": 56289957.0,
      "step": 22210
    },
    {
      "entropy": 6.323467540740967,
      "epoch": 2.5637622619734564,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004352531771810359,
      "loss": 5.7165,
      "mean_token_accuracy": 0.2032003879547119,
      "num_tokens": 56302157.0,
      "step": 22215
    },
    {
      "entropy": 6.280712223052978,
      "epoch": 2.564339296018465,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004352242093979802,
      "loss": 5.5608,
      "mean_token_accuracy": 0.20886647701263428,
      "num_tokens": 56314735.0,
      "step": 22220
    },
    {
      "entropy": 6.271326065063477,
      "epoch": 2.5649163300634736,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004351952362258628,
      "loss": 5.6641,
      "mean_token_accuracy": 0.20041901767253875,
      "num_tokens": 56327077.0,
      "step": 22225
    },
    {
      "entropy": 6.297061252593994,
      "epoch": 2.5654933641084825,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043516625766565825,
      "loss": 5.7124,
      "mean_token_accuracy": 0.20260732620954514,
      "num_tokens": 56339942.0,
      "step": 22230
    },
    {
      "entropy": 6.303098535537719,
      "epoch": 2.566070398153491,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004351372737183412,
      "loss": 5.6722,
      "mean_token_accuracy": 0.20310757756233216,
      "num_tokens": 56352020.0,
      "step": 22235
    },
    {
      "entropy": 6.338209772109986,
      "epoch": 2.5666474321985,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004351082843848865,
      "loss": 5.7721,
      "mean_token_accuracy": 0.19831058531999587,
      "num_tokens": 56364794.0,
      "step": 22240
    },
    {
      "entropy": 6.322439193725586,
      "epoch": 2.567224466243508,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043507928966626916,
      "loss": 5.7414,
      "mean_token_accuracy": 0.20152952522039413,
      "num_tokens": 56376552.0,
      "step": 22245
    },
    {
      "entropy": 6.253318357467651,
      "epoch": 2.567801500288517,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004350502895634644,
      "loss": 5.6658,
      "mean_token_accuracy": 0.20396380126476288,
      "num_tokens": 56390021.0,
      "step": 22250
    },
    {
      "entropy": 6.237610769271851,
      "epoch": 2.568378534333526,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004350212840774476,
      "loss": 5.583,
      "mean_token_accuracy": 0.20588896721601485,
      "num_tokens": 56402311.0,
      "step": 22255
    },
    {
      "entropy": 6.301930618286133,
      "epoch": 2.5689555683785343,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00043499227320919456,
      "loss": 5.6735,
      "mean_token_accuracy": 0.2079094812273979,
      "num_tokens": 56415031.0,
      "step": 22260
    },
    {
      "entropy": 6.235403251647949,
      "epoch": 2.5695326024235428,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00043496325695968076,
      "loss": 5.6194,
      "mean_token_accuracy": 0.20982007980346679,
      "num_tokens": 56428823.0,
      "step": 22265
    },
    {
      "entropy": 6.260178899765014,
      "epoch": 2.5701096364685516,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004349342353298823,
      "loss": 5.6957,
      "mean_token_accuracy": 0.19989748150110245,
      "num_tokens": 56441875.0,
      "step": 22270
    },
    {
      "entropy": 6.3290478706359865,
      "epoch": 2.5706866705135605,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004349052083207753,
      "loss": 5.7253,
      "mean_token_accuracy": 0.19897918701171874,
      "num_tokens": 56454511.0,
      "step": 22275
    },
    {
      "entropy": 6.265756034851075,
      "epoch": 2.571263704558569,
      "grad_norm": 1.015625,
      "learning_rate": 0.000434876175933336,
      "loss": 5.6364,
      "mean_token_accuracy": 0.20455872416496276,
      "num_tokens": 56465201.0,
      "step": 22280
    },
    {
      "entropy": 6.3022034645080565,
      "epoch": 2.5718407386035778,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043484713816854093,
      "loss": 5.6941,
      "mean_token_accuracy": 0.2012573316693306,
      "num_tokens": 56477753.0,
      "step": 22285
    },
    {
      "entropy": 6.259252643585205,
      "epoch": 2.572417772648586,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004348180950273668,
      "loss": 5.6155,
      "mean_token_accuracy": 0.20587154030799865,
      "num_tokens": 56491254.0,
      "step": 22290
    },
    {
      "entropy": 6.211175203323364,
      "epoch": 2.572994806693595,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004347890465107905,
      "loss": 5.554,
      "mean_token_accuracy": 0.21337357461452483,
      "num_tokens": 56503837.0,
      "step": 22295
    },
    {
      "entropy": 6.257611989974976,
      "epoch": 2.5735718407386035,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043475999261978894,
      "loss": 5.6515,
      "mean_token_accuracy": 0.19793593287467956,
      "num_tokens": 56516624.0,
      "step": 22300
    },
    {
      "entropy": 6.276579999923706,
      "epoch": 2.5741488747836123,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004347309333553393,
      "loss": 5.7409,
      "mean_token_accuracy": 0.19274209439754486,
      "num_tokens": 56529523.0,
      "step": 22305
    },
    {
      "entropy": 6.363157606124878,
      "epoch": 2.5747259088286207,
      "grad_norm": 1.8359375,
      "learning_rate": 0.00043470186871841915,
      "loss": 5.6978,
      "mean_token_accuracy": 0.19099193811416626,
      "num_tokens": 56543110.0,
      "step": 22310
    },
    {
      "entropy": 6.31586389541626,
      "epoch": 2.5753029428736296,
      "grad_norm": 1.046875,
      "learning_rate": 0.000434672798710006,
      "loss": 5.6587,
      "mean_token_accuracy": 0.20200337767601012,
      "num_tokens": 56555190.0,
      "step": 22315
    },
    {
      "entropy": 6.282614707946777,
      "epoch": 2.575879976918638,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043464372333107756,
      "loss": 5.7024,
      "mean_token_accuracy": 0.20311116576194763,
      "num_tokens": 56567705.0,
      "step": 22320
    },
    {
      "entropy": 6.311984395980835,
      "epoch": 2.576457010963647,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043461464258261174,
      "loss": 5.7023,
      "mean_token_accuracy": 0.19981039613485335,
      "num_tokens": 56579905.0,
      "step": 22325
    },
    {
      "entropy": 6.301828384399414,
      "epoch": 2.5770340450086557,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004345855564655867,
      "loss": 5.6945,
      "mean_token_accuracy": 0.20406469106674194,
      "num_tokens": 56593883.0,
      "step": 22330
    },
    {
      "entropy": 6.263874912261963,
      "epoch": 2.577611079053664,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043455646498098075,
      "loss": 5.6736,
      "mean_token_accuracy": 0.20208864212036132,
      "num_tokens": 56606342.0,
      "step": 22335
    },
    {
      "entropy": 6.226379299163819,
      "epoch": 2.5781881130986726,
      "grad_norm": 1.125,
      "learning_rate": 0.00043452736812977236,
      "loss": 5.5355,
      "mean_token_accuracy": 0.21760296821594238,
      "num_tokens": 56618302.0,
      "step": 22340
    },
    {
      "entropy": 6.287581300735473,
      "epoch": 2.5787651471436814,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004344982659129402,
      "loss": 5.7258,
      "mean_token_accuracy": 0.19926684647798537,
      "num_tokens": 56630199.0,
      "step": 22345
    },
    {
      "entropy": 6.292850017547607,
      "epoch": 2.5793421811886903,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004344691583314631,
      "loss": 5.696,
      "mean_token_accuracy": 0.19728785753250122,
      "num_tokens": 56642020.0,
      "step": 22350
    },
    {
      "entropy": 6.286031293869018,
      "epoch": 2.5799192152336987,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043444004538632005,
      "loss": 5.686,
      "mean_token_accuracy": 0.20830992609262466,
      "num_tokens": 56653779.0,
      "step": 22355
    },
    {
      "entropy": 6.3310089111328125,
      "epoch": 2.5804962492787076,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043441092707849024,
      "loss": 5.7146,
      "mean_token_accuracy": 0.19985073804855347,
      "num_tokens": 56665695.0,
      "step": 22360
    },
    {
      "entropy": 6.294081974029541,
      "epoch": 2.581073283323716,
      "grad_norm": 0.953125,
      "learning_rate": 0.00043438180340895315,
      "loss": 5.7076,
      "mean_token_accuracy": 0.19552246928215028,
      "num_tokens": 56678294.0,
      "step": 22365
    },
    {
      "entropy": 6.2895012378692625,
      "epoch": 2.581650317368725,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043435267437868825,
      "loss": 5.6707,
      "mean_token_accuracy": 0.20251239687204362,
      "num_tokens": 56690624.0,
      "step": 22370
    },
    {
      "entropy": 6.3160481452941895,
      "epoch": 2.5822273514137333,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004343235399886753,
      "loss": 5.6656,
      "mean_token_accuracy": 0.20556568801403047,
      "num_tokens": 56703879.0,
      "step": 22375
    },
    {
      "entropy": 6.305972003936768,
      "epoch": 2.582804385458742,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004342944002398942,
      "loss": 5.7364,
      "mean_token_accuracy": 0.199286550283432,
      "num_tokens": 56716957.0,
      "step": 22380
    },
    {
      "entropy": 6.291890287399292,
      "epoch": 2.583381419503751,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004342652551333251,
      "loss": 5.6415,
      "mean_token_accuracy": 0.20715494006872176,
      "num_tokens": 56728597.0,
      "step": 22385
    },
    {
      "entropy": 6.333414936065674,
      "epoch": 2.5839584535487594,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004342361046699482,
      "loss": 5.6893,
      "mean_token_accuracy": 0.2042376235127449,
      "num_tokens": 56740133.0,
      "step": 22390
    },
    {
      "entropy": 6.355626153945923,
      "epoch": 2.584535487593768,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004342069488507441,
      "loss": 5.7374,
      "mean_token_accuracy": 0.19570707976818086,
      "num_tokens": 56752976.0,
      "step": 22395
    },
    {
      "entropy": 6.310667085647583,
      "epoch": 2.5851125216387767,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004341777876766933,
      "loss": 5.7539,
      "mean_token_accuracy": 0.1990024581551552,
      "num_tokens": 56765569.0,
      "step": 22400
    },
    {
      "entropy": 6.228538942337036,
      "epoch": 2.5856895556837856,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004341486211487767,
      "loss": 5.5656,
      "mean_token_accuracy": 0.22085673362016678,
      "num_tokens": 56779293.0,
      "step": 22405
    },
    {
      "entropy": 6.2322581768035885,
      "epoch": 2.586266589728794,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004341194492679753,
      "loss": 5.6236,
      "mean_token_accuracy": 0.2050124377012253,
      "num_tokens": 56792106.0,
      "step": 22410
    },
    {
      "entropy": 6.251925706863403,
      "epoch": 2.5868436237738024,
      "grad_norm": 1.5625,
      "learning_rate": 0.00043409027203527016,
      "loss": 5.6166,
      "mean_token_accuracy": 0.2062378093600273,
      "num_tokens": 56805132.0,
      "step": 22415
    },
    {
      "entropy": 6.153520154953003,
      "epoch": 2.5874206578188113,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043406108945164283,
      "loss": 5.5173,
      "mean_token_accuracy": 0.21563137024641038,
      "num_tokens": 56817044.0,
      "step": 22420
    },
    {
      "entropy": 6.3026123046875,
      "epoch": 2.58799769186382,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043403190151807474,
      "loss": 5.7615,
      "mean_token_accuracy": 0.19751387387514113,
      "num_tokens": 56831278.0,
      "step": 22425
    },
    {
      "entropy": 6.241002368927002,
      "epoch": 2.5885747259088285,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004340027082355477,
      "loss": 5.6687,
      "mean_token_accuracy": 0.2015869364142418,
      "num_tokens": 56842462.0,
      "step": 22430
    },
    {
      "entropy": 6.2867003917694095,
      "epoch": 2.5891517599538374,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004339735096050434,
      "loss": 5.687,
      "mean_token_accuracy": 0.20024093240499496,
      "num_tokens": 56855601.0,
      "step": 22435
    },
    {
      "entropy": 6.262297534942627,
      "epoch": 2.589728793998846,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004339443056275443,
      "loss": 5.592,
      "mean_token_accuracy": 0.2052015095949173,
      "num_tokens": 56868077.0,
      "step": 22440
    },
    {
      "entropy": 6.305109977722168,
      "epoch": 2.5903058280438547,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043391509630403224,
      "loss": 5.7574,
      "mean_token_accuracy": 0.19968690276145934,
      "num_tokens": 56881625.0,
      "step": 22445
    },
    {
      "entropy": 6.248378849029541,
      "epoch": 2.590882862088863,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00043388588163549,
      "loss": 5.5937,
      "mean_token_accuracy": 0.21829177886247636,
      "num_tokens": 56894726.0,
      "step": 22450
    },
    {
      "entropy": 6.274321413040161,
      "epoch": 2.591459896133872,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043385666162289995,
      "loss": 5.7557,
      "mean_token_accuracy": 0.20218839198350907,
      "num_tokens": 56907709.0,
      "step": 22455
    },
    {
      "entropy": 6.225597953796386,
      "epoch": 2.592036930178881,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00043382743626724506,
      "loss": 5.6195,
      "mean_token_accuracy": 0.20132321566343309,
      "num_tokens": 56921289.0,
      "step": 22460
    },
    {
      "entropy": 6.295007419586182,
      "epoch": 2.5926139642238892,
      "grad_norm": 1.0,
      "learning_rate": 0.0004337982055695082,
      "loss": 5.7012,
      "mean_token_accuracy": 0.20052748769521714,
      "num_tokens": 56932738.0,
      "step": 22465
    },
    {
      "entropy": 6.24111852645874,
      "epoch": 2.5931909982688977,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004337689695306726,
      "loss": 5.6435,
      "mean_token_accuracy": 0.20369818806648254,
      "num_tokens": 56945944.0,
      "step": 22470
    },
    {
      "entropy": 6.365180969238281,
      "epoch": 2.5937680323139065,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004337397281517215,
      "loss": 5.7179,
      "mean_token_accuracy": 0.2079303115606308,
      "num_tokens": 56959253.0,
      "step": 22475
    },
    {
      "entropy": 6.317916440963745,
      "epoch": 2.5943450663589154,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004337104814336386,
      "loss": 5.7597,
      "mean_token_accuracy": 0.19659004360437393,
      "num_tokens": 56971983.0,
      "step": 22480
    },
    {
      "entropy": 6.31353497505188,
      "epoch": 2.594922100403924,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004336812293774075,
      "loss": 5.7668,
      "mean_token_accuracy": 0.19597004503011703,
      "num_tokens": 56986246.0,
      "step": 22485
    },
    {
      "entropy": 6.2919048309326175,
      "epoch": 2.595499134448932,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043365197198401195,
      "loss": 5.6642,
      "mean_token_accuracy": 0.20688205510377883,
      "num_tokens": 56997465.0,
      "step": 22490
    },
    {
      "entropy": 6.357540702819824,
      "epoch": 2.596076168493941,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004336227092544362,
      "loss": 5.7589,
      "mean_token_accuracy": 0.18876300752162933,
      "num_tokens": 57010593.0,
      "step": 22495
    },
    {
      "entropy": 6.26652455329895,
      "epoch": 2.59665320253895,
      "grad_norm": 1.0,
      "learning_rate": 0.0004335934411896644,
      "loss": 5.6455,
      "mean_token_accuracy": 0.20344152599573134,
      "num_tokens": 57023516.0,
      "step": 22500
    },
    {
      "entropy": 6.302176237106323,
      "epoch": 2.5972302365839584,
      "grad_norm": 1.015625,
      "learning_rate": 0.000433564167790681,
      "loss": 5.7641,
      "mean_token_accuracy": 0.20127285420894622,
      "num_tokens": 57036316.0,
      "step": 22505
    },
    {
      "entropy": 6.3343854427337645,
      "epoch": 2.597807270628967,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004335348890584706,
      "loss": 5.6471,
      "mean_token_accuracy": 0.20793196558952332,
      "num_tokens": 57049757.0,
      "step": 22510
    },
    {
      "entropy": 6.321691656112671,
      "epoch": 2.5983843046739756,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043350560499401793,
      "loss": 5.6652,
      "mean_token_accuracy": 0.20056487172842025,
      "num_tokens": 57062569.0,
      "step": 22515
    },
    {
      "entropy": 6.3389874458312985,
      "epoch": 2.5989613387189845,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004334763155983079,
      "loss": 5.7106,
      "mean_token_accuracy": 0.2043031558394432,
      "num_tokens": 57075039.0,
      "step": 22520
    },
    {
      "entropy": 6.319142580032349,
      "epoch": 2.599538372763993,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043344702087232584,
      "loss": 5.7096,
      "mean_token_accuracy": 0.1980224519968033,
      "num_tokens": 57087068.0,
      "step": 22525
    },
    {
      "entropy": 6.278364133834839,
      "epoch": 2.6001154068090018,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004334177208170569,
      "loss": 5.7171,
      "mean_token_accuracy": 0.20176517516374587,
      "num_tokens": 57099997.0,
      "step": 22530
    },
    {
      "entropy": 6.312618160247803,
      "epoch": 2.6006924408540106,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043338841543348654,
      "loss": 5.6848,
      "mean_token_accuracy": 0.19499486535787583,
      "num_tokens": 57112375.0,
      "step": 22535
    },
    {
      "entropy": 6.202201080322266,
      "epoch": 2.601269474899019,
      "grad_norm": 1.296875,
      "learning_rate": 0.00043335910472260057,
      "loss": 5.5282,
      "mean_token_accuracy": 0.20783442556858062,
      "num_tokens": 57124968.0,
      "step": 22540
    },
    {
      "entropy": 6.3078688621521,
      "epoch": 2.6018465089440275,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043332978868538483,
      "loss": 5.7794,
      "mean_token_accuracy": 0.19118839055299758,
      "num_tokens": 57137226.0,
      "step": 22545
    },
    {
      "entropy": 6.2260149955749515,
      "epoch": 2.6024235429890363,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043330046732282515,
      "loss": 5.6319,
      "mean_token_accuracy": 0.21023293882608413,
      "num_tokens": 57149305.0,
      "step": 22550
    },
    {
      "entropy": 6.313079595565796,
      "epoch": 2.603000577034045,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00043327114063590797,
      "loss": 5.6696,
      "mean_token_accuracy": 0.20654285103082656,
      "num_tokens": 57161972.0,
      "step": 22555
    },
    {
      "entropy": 6.28898286819458,
      "epoch": 2.6035776110790536,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004332418086256196,
      "loss": 5.7126,
      "mean_token_accuracy": 0.1968448504805565,
      "num_tokens": 57173644.0,
      "step": 22560
    },
    {
      "entropy": 6.319484567642212,
      "epoch": 2.6041546451240625,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004332124712929466,
      "loss": 5.6615,
      "mean_token_accuracy": 0.2030421257019043,
      "num_tokens": 57186828.0,
      "step": 22565
    },
    {
      "entropy": 6.242533588409424,
      "epoch": 2.604731679169071,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043318312863887566,
      "loss": 5.7,
      "mean_token_accuracy": 0.2012163519859314,
      "num_tokens": 57198976.0,
      "step": 22570
    },
    {
      "entropy": 6.29468264579773,
      "epoch": 2.6053087132140798,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043315378066439384,
      "loss": 5.7502,
      "mean_token_accuracy": 0.1933215007185936,
      "num_tokens": 57211397.0,
      "step": 22575
    },
    {
      "entropy": 6.3332743644714355,
      "epoch": 2.605885747259088,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004331244273704881,
      "loss": 5.7137,
      "mean_token_accuracy": 0.19941538870334624,
      "num_tokens": 57223959.0,
      "step": 22580
    },
    {
      "entropy": 6.393739795684814,
      "epoch": 2.606462781304097,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004330950687581458,
      "loss": 5.8112,
      "mean_token_accuracy": 0.18732271939516068,
      "num_tokens": 57236725.0,
      "step": 22585
    },
    {
      "entropy": 6.402460813522339,
      "epoch": 2.6070398153491054,
      "grad_norm": 0.921875,
      "learning_rate": 0.00043306570482835443,
      "loss": 5.7674,
      "mean_token_accuracy": 0.1931520640850067,
      "num_tokens": 57249818.0,
      "step": 22590
    },
    {
      "entropy": 6.243662977218628,
      "epoch": 2.6076168493941143,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004330363355821016,
      "loss": 5.6476,
      "mean_token_accuracy": 0.20524583756923676,
      "num_tokens": 57262479.0,
      "step": 22595
    },
    {
      "entropy": 6.2619117259979244,
      "epoch": 2.6081938834391227,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004330069610203751,
      "loss": 5.6513,
      "mean_token_accuracy": 0.20647508203983306,
      "num_tokens": 57274823.0,
      "step": 22600
    },
    {
      "entropy": 6.261155509948731,
      "epoch": 2.6087709174841316,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004329775811441629,
      "loss": 5.6608,
      "mean_token_accuracy": 0.2037465527653694,
      "num_tokens": 57287471.0,
      "step": 22605
    },
    {
      "entropy": 6.257754707336426,
      "epoch": 2.6093479515291405,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043294819595445324,
      "loss": 5.6418,
      "mean_token_accuracy": 0.20889206528663634,
      "num_tokens": 57301394.0,
      "step": 22610
    },
    {
      "entropy": 6.302092742919922,
      "epoch": 2.609924985574149,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004329188054522345,
      "loss": 5.67,
      "mean_token_accuracy": 0.19910663664340972,
      "num_tokens": 57313480.0,
      "step": 22615
    },
    {
      "entropy": 6.250244283676148,
      "epoch": 2.6105020196191573,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004328894096384952,
      "loss": 5.5924,
      "mean_token_accuracy": 0.2063554048538208,
      "num_tokens": 57326079.0,
      "step": 22620
    },
    {
      "entropy": 6.281641960144043,
      "epoch": 2.611079053664166,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000432860008514224,
      "loss": 5.6698,
      "mean_token_accuracy": 0.209816275537014,
      "num_tokens": 57338741.0,
      "step": 22625
    },
    {
      "entropy": 6.343434906005859,
      "epoch": 2.611656087709175,
      "grad_norm": 1.0,
      "learning_rate": 0.0004328306020804099,
      "loss": 5.7303,
      "mean_token_accuracy": 0.19731898754835128,
      "num_tokens": 57350868.0,
      "step": 22630
    },
    {
      "entropy": 6.289169836044311,
      "epoch": 2.6122331217541834,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043280119033804184,
      "loss": 5.722,
      "mean_token_accuracy": 0.1942977026104927,
      "num_tokens": 57364234.0,
      "step": 22635
    },
    {
      "entropy": 6.1695497035980225,
      "epoch": 2.6128101557991923,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043277177328810914,
      "loss": 5.4644,
      "mean_token_accuracy": 0.20502195954322816,
      "num_tokens": 57376481.0,
      "step": 22640
    },
    {
      "entropy": 6.249718379974365,
      "epoch": 2.6133871898442007,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004327423509316012,
      "loss": 5.6153,
      "mean_token_accuracy": 0.20974057167768478,
      "num_tokens": 57388550.0,
      "step": 22645
    },
    {
      "entropy": 6.267832279205322,
      "epoch": 2.6139642238892096,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043271292326950764,
      "loss": 5.6569,
      "mean_token_accuracy": 0.21232789307832717,
      "num_tokens": 57400099.0,
      "step": 22650
    },
    {
      "entropy": 6.24877896308899,
      "epoch": 2.614541257934218,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004326834903028182,
      "loss": 5.6366,
      "mean_token_accuracy": 0.2092032790184021,
      "num_tokens": 57414039.0,
      "step": 22655
    },
    {
      "entropy": 6.2636476993560795,
      "epoch": 2.615118291979227,
      "grad_norm": 1.03125,
      "learning_rate": 0.000432654052032523,
      "loss": 5.6265,
      "mean_token_accuracy": 0.20537174642086028,
      "num_tokens": 57426449.0,
      "step": 22660
    },
    {
      "entropy": 6.27854290008545,
      "epoch": 2.6156953260242353,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000432624608459612,
      "loss": 5.6669,
      "mean_token_accuracy": 0.20295319557189942,
      "num_tokens": 57440122.0,
      "step": 22665
    },
    {
      "entropy": 6.277081108093261,
      "epoch": 2.616272360069244,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004325951595850756,
      "loss": 5.6495,
      "mean_token_accuracy": 0.20576129704713822,
      "num_tokens": 57452368.0,
      "step": 22670
    },
    {
      "entropy": 6.290420961380005,
      "epoch": 2.6168493941142525,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00043256570540990425,
      "loss": 5.7069,
      "mean_token_accuracy": 0.2156609296798706,
      "num_tokens": 57465634.0,
      "step": 22675
    },
    {
      "entropy": 6.30668535232544,
      "epoch": 2.6174264281592614,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004325362459350888,
      "loss": 5.6866,
      "mean_token_accuracy": 0.20049384236335754,
      "num_tokens": 57477137.0,
      "step": 22680
    },
    {
      "entropy": 6.342857360839844,
      "epoch": 2.6180034622042703,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043250678116161985,
      "loss": 5.7318,
      "mean_token_accuracy": 0.19208012223243714,
      "num_tokens": 57490496.0,
      "step": 22685
    },
    {
      "entropy": 6.350476217269898,
      "epoch": 2.6185804962492787,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043247731109048856,
      "loss": 5.6806,
      "mean_token_accuracy": 0.20622063130140306,
      "num_tokens": 57503111.0,
      "step": 22690
    },
    {
      "entropy": 6.29325647354126,
      "epoch": 2.619157530294287,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004324478357226861,
      "loss": 5.7025,
      "mean_token_accuracy": 0.20451032668352126,
      "num_tokens": 57515735.0,
      "step": 22695
    },
    {
      "entropy": 6.284058523178101,
      "epoch": 2.619734564339296,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043241835505920405,
      "loss": 5.6888,
      "mean_token_accuracy": 0.20425134897232056,
      "num_tokens": 57528102.0,
      "step": 22700
    },
    {
      "entropy": 6.229598140716552,
      "epoch": 2.620311598384305,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004323888691010337,
      "loss": 5.6896,
      "mean_token_accuracy": 0.20290261507034302,
      "num_tokens": 57542347.0,
      "step": 22705
    },
    {
      "entropy": 6.247469568252564,
      "epoch": 2.6208886324293132,
      "grad_norm": 0.953125,
      "learning_rate": 0.00043235937784916694,
      "loss": 5.6388,
      "mean_token_accuracy": 0.20144413262605668,
      "num_tokens": 57554882.0,
      "step": 22710
    },
    {
      "entropy": 6.349951267242432,
      "epoch": 2.621465666474322,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043232988130459567,
      "loss": 5.7584,
      "mean_token_accuracy": 0.19770350009202958,
      "num_tokens": 57568201.0,
      "step": 22715
    },
    {
      "entropy": 6.325543403625488,
      "epoch": 2.6220427005193305,
      "grad_norm": 0.953125,
      "learning_rate": 0.000432300379468312,
      "loss": 5.7121,
      "mean_token_accuracy": 0.19676875472068786,
      "num_tokens": 57581684.0,
      "step": 22720
    },
    {
      "entropy": 6.340763902664184,
      "epoch": 2.6226197345643394,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004322708723413082,
      "loss": 5.7898,
      "mean_token_accuracy": 0.190242138504982,
      "num_tokens": 57596314.0,
      "step": 22725
    },
    {
      "entropy": 6.302534294128418,
      "epoch": 2.623196768609348,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043224135992457667,
      "loss": 5.6748,
      "mean_token_accuracy": 0.2043718859553337,
      "num_tokens": 57607322.0,
      "step": 22730
    },
    {
      "entropy": 6.27992091178894,
      "epoch": 2.6237738026543567,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004322118422191102,
      "loss": 5.6265,
      "mean_token_accuracy": 0.20416467040777206,
      "num_tokens": 57620232.0,
      "step": 22735
    },
    {
      "entropy": 6.294591951370239,
      "epoch": 2.6243508366993655,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043218231922590144,
      "loss": 5.6926,
      "mean_token_accuracy": 0.19955601990222932,
      "num_tokens": 57633656.0,
      "step": 22740
    },
    {
      "entropy": 6.365624475479126,
      "epoch": 2.624927870744374,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004321527909459435,
      "loss": 5.7734,
      "mean_token_accuracy": 0.19128601998090744,
      "num_tokens": 57645380.0,
      "step": 22745
    },
    {
      "entropy": 6.334028053283691,
      "epoch": 2.6255049047893824,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043212325738022946,
      "loss": 5.6965,
      "mean_token_accuracy": 0.19699230492115022,
      "num_tokens": 57659104.0,
      "step": 22750
    },
    {
      "entropy": 6.239507246017456,
      "epoch": 2.6260819388343912,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004320937185297527,
      "loss": 5.6565,
      "mean_token_accuracy": 0.20263622403144838,
      "num_tokens": 57672476.0,
      "step": 22755
    },
    {
      "entropy": 6.309638118743896,
      "epoch": 2.6266589728794,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043206417439550684,
      "loss": 5.6885,
      "mean_token_accuracy": 0.20456310659646987,
      "num_tokens": 57685248.0,
      "step": 22760
    },
    {
      "entropy": 6.309205102920532,
      "epoch": 2.6272360069244085,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043203462497848534,
      "loss": 5.713,
      "mean_token_accuracy": 0.2035017043352127,
      "num_tokens": 57697704.0,
      "step": 22765
    },
    {
      "entropy": 6.239285516738891,
      "epoch": 2.627813040969417,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043200507027968236,
      "loss": 5.6207,
      "mean_token_accuracy": 0.21212908029556274,
      "num_tokens": 57709966.0,
      "step": 22770
    },
    {
      "entropy": 6.320084047317505,
      "epoch": 2.628390075014426,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004319755103000918,
      "loss": 5.6107,
      "mean_token_accuracy": 0.20686028450727462,
      "num_tokens": 57722359.0,
      "step": 22775
    },
    {
      "entropy": 6.349383401870727,
      "epoch": 2.6289671090594346,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004319459450407079,
      "loss": 5.7201,
      "mean_token_accuracy": 0.1995735287666321,
      "num_tokens": 57733736.0,
      "step": 22780
    },
    {
      "entropy": 6.295598030090332,
      "epoch": 2.629544143104443,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043191637450252514,
      "loss": 5.7197,
      "mean_token_accuracy": 0.19351973086595536,
      "num_tokens": 57745981.0,
      "step": 22785
    },
    {
      "entropy": 6.135420227050782,
      "epoch": 2.630121177149452,
      "grad_norm": 1.0,
      "learning_rate": 0.000431886798686538,
      "loss": 5.5215,
      "mean_token_accuracy": 0.2145640268921852,
      "num_tokens": 57758324.0,
      "step": 22790
    },
    {
      "entropy": 6.257735109329223,
      "epoch": 2.6306982111944603,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004318572175937414,
      "loss": 5.6708,
      "mean_token_accuracy": 0.2024206444621086,
      "num_tokens": 57770852.0,
      "step": 22795
    },
    {
      "entropy": 6.316752290725708,
      "epoch": 2.631275245239469,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004318276312251301,
      "loss": 5.7024,
      "mean_token_accuracy": 0.20170061141252518,
      "num_tokens": 57783951.0,
      "step": 22800
    },
    {
      "entropy": 6.32301516532898,
      "epoch": 2.6318522792844776,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004317980395816994,
      "loss": 5.6394,
      "mean_token_accuracy": 0.20439137667417526,
      "num_tokens": 57797071.0,
      "step": 22805
    },
    {
      "entropy": 6.203200435638427,
      "epoch": 2.6324293133294865,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004317684426644445,
      "loss": 5.6219,
      "mean_token_accuracy": 0.21139226257801055,
      "num_tokens": 57808747.0,
      "step": 22810
    },
    {
      "entropy": 6.229562425613404,
      "epoch": 2.6330063473744953,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043173884047436093,
      "loss": 5.6367,
      "mean_token_accuracy": 0.20920082181692123,
      "num_tokens": 57821349.0,
      "step": 22815
    },
    {
      "entropy": 6.302713012695312,
      "epoch": 2.6335833814195038,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004317092330124443,
      "loss": 5.7513,
      "mean_token_accuracy": 0.19796638786792756,
      "num_tokens": 57834228.0,
      "step": 22820
    },
    {
      "entropy": 6.297818613052368,
      "epoch": 2.634160415464512,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043167962027969043,
      "loss": 5.5761,
      "mean_token_accuracy": 0.20789769887924195,
      "num_tokens": 57847577.0,
      "step": 22825
    },
    {
      "entropy": 6.194095277786255,
      "epoch": 2.634737449509521,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004316500022770954,
      "loss": 5.6233,
      "mean_token_accuracy": 0.20715264081954957,
      "num_tokens": 57860306.0,
      "step": 22830
    },
    {
      "entropy": 6.302250909805298,
      "epoch": 2.63531448355453,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043162037900565533,
      "loss": 5.6925,
      "mean_token_accuracy": 0.20176810622215272,
      "num_tokens": 57872820.0,
      "step": 22835
    },
    {
      "entropy": 6.341640329360962,
      "epoch": 2.6358915175995383,
      "grad_norm": 0.9375,
      "learning_rate": 0.00043159075046636666,
      "loss": 5.7356,
      "mean_token_accuracy": 0.1944164901971817,
      "num_tokens": 57885939.0,
      "step": 22840
    },
    {
      "entropy": 6.256580924987793,
      "epoch": 2.636468551644547,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004315611166602258,
      "loss": 5.597,
      "mean_token_accuracy": 0.2059169664978981,
      "num_tokens": 57898837.0,
      "step": 22845
    },
    {
      "entropy": 6.373171615600586,
      "epoch": 2.6370455856895556,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00043153147758822957,
      "loss": 5.771,
      "mean_token_accuracy": 0.19661803245544435,
      "num_tokens": 57910822.0,
      "step": 22850
    },
    {
      "entropy": 6.291859817504883,
      "epoch": 2.6376226197345645,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004315018332513749,
      "loss": 5.6871,
      "mean_token_accuracy": 0.1979863837361336,
      "num_tokens": 57923441.0,
      "step": 22855
    },
    {
      "entropy": 6.216495704650879,
      "epoch": 2.638199653779573,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004314721836506587,
      "loss": 5.5556,
      "mean_token_accuracy": 0.20789090394973755,
      "num_tokens": 57936540.0,
      "step": 22860
    },
    {
      "entropy": 6.3213215351104735,
      "epoch": 2.6387766878245817,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00043144252878707845,
      "loss": 5.6982,
      "mean_token_accuracy": 0.20722079426050186,
      "num_tokens": 57949758.0,
      "step": 22865
    },
    {
      "entropy": 6.314129734039307,
      "epoch": 2.63935372186959,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004314128686616314,
      "loss": 5.684,
      "mean_token_accuracy": 0.20580807626247405,
      "num_tokens": 57962523.0,
      "step": 22870
    },
    {
      "entropy": 6.23826003074646,
      "epoch": 2.639930755914599,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004313832032753152,
      "loss": 5.6693,
      "mean_token_accuracy": 0.20674923360347747,
      "num_tokens": 57976294.0,
      "step": 22875
    },
    {
      "entropy": 6.2344598293304445,
      "epoch": 2.6405077899596074,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00043135353262912764,
      "loss": 5.6314,
      "mean_token_accuracy": 0.206416517496109,
      "num_tokens": 57989214.0,
      "step": 22880
    },
    {
      "entropy": 6.355399703979492,
      "epoch": 2.6410848240046163,
      "grad_norm": 0.921875,
      "learning_rate": 0.00043132385672406664,
      "loss": 5.8316,
      "mean_token_accuracy": 0.1829749509692192,
      "num_tokens": 58003008.0,
      "step": 22885
    },
    {
      "entropy": 6.382001638412476,
      "epoch": 2.641661858049625,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004312941755611305,
      "loss": 5.7222,
      "mean_token_accuracy": 0.20114904940128325,
      "num_tokens": 58014911.0,
      "step": 22890
    },
    {
      "entropy": 6.27892804145813,
      "epoch": 2.6422388920946336,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043126448914131724,
      "loss": 5.7355,
      "mean_token_accuracy": 0.20032203197479248,
      "num_tokens": 58027636.0,
      "step": 22895
    },
    {
      "entropy": 6.284564781188965,
      "epoch": 2.642815926139642,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004312347974656256,
      "loss": 5.6885,
      "mean_token_accuracy": 0.20665575116872786,
      "num_tokens": 58039896.0,
      "step": 22900
    },
    {
      "entropy": 6.291294956207276,
      "epoch": 2.643392960184651,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004312051005350541,
      "loss": 5.6324,
      "mean_token_accuracy": 0.20280472785234452,
      "num_tokens": 58052979.0,
      "step": 22905
    },
    {
      "entropy": 6.2824150085449215,
      "epoch": 2.6439699942296597,
      "grad_norm": 1.0,
      "learning_rate": 0.00043117539835060164,
      "loss": 5.7194,
      "mean_token_accuracy": 0.19531827569007873,
      "num_tokens": 58065185.0,
      "step": 22910
    },
    {
      "entropy": 6.2464892864227295,
      "epoch": 2.644547028274668,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004311456909132673,
      "loss": 5.7153,
      "mean_token_accuracy": 0.20522386580705643,
      "num_tokens": 58078836.0,
      "step": 22915
    },
    {
      "entropy": 6.306563806533814,
      "epoch": 2.645124062319677,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004311159782240502,
      "loss": 5.6578,
      "mean_token_accuracy": 0.20077867209911346,
      "num_tokens": 58091846.0,
      "step": 22920
    },
    {
      "entropy": 6.34948205947876,
      "epoch": 2.6457010963646854,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004310862602839496,
      "loss": 5.6907,
      "mean_token_accuracy": 0.20606040805578232,
      "num_tokens": 58105840.0,
      "step": 22925
    },
    {
      "entropy": 6.305211973190308,
      "epoch": 2.6462781304096943,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004310565370939653,
      "loss": 5.6868,
      "mean_token_accuracy": 0.2038790062069893,
      "num_tokens": 58117654.0,
      "step": 22930
    },
    {
      "entropy": 6.303582811355591,
      "epoch": 2.6468551644547027,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043102680865509687,
      "loss": 5.6227,
      "mean_token_accuracy": 0.21243294775485994,
      "num_tokens": 58130005.0,
      "step": 22935
    },
    {
      "entropy": 6.204463911056519,
      "epoch": 2.6474321984997116,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004309970749683442,
      "loss": 5.5976,
      "mean_token_accuracy": 0.21068568378686905,
      "num_tokens": 58142813.0,
      "step": 22940
    },
    {
      "entropy": 6.301526260375977,
      "epoch": 2.64800923254472,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00043096733603470737,
      "loss": 5.6641,
      "mean_token_accuracy": 0.20202164351940155,
      "num_tokens": 58156545.0,
      "step": 22945
    },
    {
      "entropy": 6.3644733905792235,
      "epoch": 2.648586266589729,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00043093759185518677,
      "loss": 5.7337,
      "mean_token_accuracy": 0.20150115191936493,
      "num_tokens": 58169680.0,
      "step": 22950
    },
    {
      "entropy": 6.343415021896362,
      "epoch": 2.6491633006347373,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00043090784243078264,
      "loss": 5.7481,
      "mean_token_accuracy": 0.1947374701499939,
      "num_tokens": 58181916.0,
      "step": 22955
    },
    {
      "entropy": 6.287647342681884,
      "epoch": 2.649740334679746,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004308780877624957,
      "loss": 5.6859,
      "mean_token_accuracy": 0.19545212686061858,
      "num_tokens": 58194152.0,
      "step": 22960
    },
    {
      "entropy": 6.325816583633423,
      "epoch": 2.650317368724755,
      "grad_norm": 1.0,
      "learning_rate": 0.0004308483278513267,
      "loss": 5.7255,
      "mean_token_accuracy": 0.19536166191101073,
      "num_tokens": 58206120.0,
      "step": 22965
    },
    {
      "entropy": 6.358600616455078,
      "epoch": 2.6508944027697634,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043081856269827656,
      "loss": 5.7296,
      "mean_token_accuracy": 0.19455128759145737,
      "num_tokens": 58219530.0,
      "step": 22970
    },
    {
      "entropy": 6.284403324127197,
      "epoch": 2.651471436814772,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004307887923043465,
      "loss": 5.6688,
      "mean_token_accuracy": 0.20879181623458862,
      "num_tokens": 58231448.0,
      "step": 22975
    },
    {
      "entropy": 6.302872371673584,
      "epoch": 2.6520484708597807,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004307590166705379,
      "loss": 5.6106,
      "mean_token_accuracy": 0.21469044983386992,
      "num_tokens": 58243974.0,
      "step": 22980
    },
    {
      "entropy": 6.298255777359008,
      "epoch": 2.6526255049047895,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043072923579785203,
      "loss": 5.7313,
      "mean_token_accuracy": 0.18597004413604737,
      "num_tokens": 58257038.0,
      "step": 22985
    },
    {
      "entropy": 6.292060279846192,
      "epoch": 2.653202538949798,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004306994496872907,
      "loss": 5.5906,
      "mean_token_accuracy": 0.21229056715965272,
      "num_tokens": 58270494.0,
      "step": 22990
    },
    {
      "entropy": 6.354970407485962,
      "epoch": 2.653779572994807,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043066965833985567,
      "loss": 5.765,
      "mean_token_accuracy": 0.2001862868666649,
      "num_tokens": 58284032.0,
      "step": 22995
    },
    {
      "entropy": 6.254119634628296,
      "epoch": 2.6543566070398152,
      "grad_norm": 1.0,
      "learning_rate": 0.00043063986175654906,
      "loss": 5.5856,
      "mean_token_accuracy": 0.2100608915090561,
      "num_tokens": 58296659.0,
      "step": 23000
    },
    {
      "entropy": 6.253914546966553,
      "epoch": 2.654933641084824,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043061005993837306,
      "loss": 5.6539,
      "mean_token_accuracy": 0.2053273007273674,
      "num_tokens": 58308576.0,
      "step": 23005
    },
    {
      "entropy": 6.268325996398926,
      "epoch": 2.6555106751298325,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043058025288632995,
      "loss": 5.6801,
      "mean_token_accuracy": 0.2065323367714882,
      "num_tokens": 58321713.0,
      "step": 23010
    },
    {
      "entropy": 6.376092767715454,
      "epoch": 2.6560877091748414,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043055044060142224,
      "loss": 5.768,
      "mean_token_accuracy": 0.19516690373420714,
      "num_tokens": 58333498.0,
      "step": 23015
    },
    {
      "entropy": 6.316672515869141,
      "epoch": 2.6566647432198502,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043052062308465277,
      "loss": 5.6456,
      "mean_token_accuracy": 0.20738409757614135,
      "num_tokens": 58344399.0,
      "step": 23020
    },
    {
      "entropy": 6.190460872650147,
      "epoch": 2.6572417772648587,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004304908003370244,
      "loss": 5.5831,
      "mean_token_accuracy": 0.21945423036813735,
      "num_tokens": 58356695.0,
      "step": 23025
    },
    {
      "entropy": 6.244876766204834,
      "epoch": 2.657818811309867,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00043046097235954024,
      "loss": 5.6835,
      "mean_token_accuracy": 0.2036574274301529,
      "num_tokens": 58369712.0,
      "step": 23030
    },
    {
      "entropy": 6.381704998016358,
      "epoch": 2.658395845354876,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043043113915320345,
      "loss": 5.7247,
      "mean_token_accuracy": 0.19422842413187028,
      "num_tokens": 58382887.0,
      "step": 23035
    },
    {
      "entropy": 6.147567796707153,
      "epoch": 2.658972879399885,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043040130071901747,
      "loss": 5.5193,
      "mean_token_accuracy": 0.21892310827970504,
      "num_tokens": 58396374.0,
      "step": 23040
    },
    {
      "entropy": 6.250977468490601,
      "epoch": 2.659549913444893,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00043037145705798605,
      "loss": 5.6421,
      "mean_token_accuracy": 0.20711357295513153,
      "num_tokens": 58410589.0,
      "step": 23045
    },
    {
      "entropy": 6.26227707862854,
      "epoch": 2.6601269474899016,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043034160817111275,
      "loss": 5.6836,
      "mean_token_accuracy": 0.20642994940280915,
      "num_tokens": 58423433.0,
      "step": 23050
    },
    {
      "entropy": 6.334927129745483,
      "epoch": 2.6607039815349105,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043031175405940157,
      "loss": 5.7079,
      "mean_token_accuracy": 0.19939419478178025,
      "num_tokens": 58435244.0,
      "step": 23055
    },
    {
      "entropy": 6.388784122467041,
      "epoch": 2.6612810155799194,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004302818947238568,
      "loss": 5.7163,
      "mean_token_accuracy": 0.2018528774380684,
      "num_tokens": 58447240.0,
      "step": 23060
    },
    {
      "entropy": 6.292679357528686,
      "epoch": 2.661858049624928,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043025203016548263,
      "loss": 5.6799,
      "mean_token_accuracy": 0.21016372740268707,
      "num_tokens": 58460385.0,
      "step": 23065
    },
    {
      "entropy": 6.282902622222901,
      "epoch": 2.6624350836699366,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043022216038528347,
      "loss": 5.6611,
      "mean_token_accuracy": 0.20241563469171525,
      "num_tokens": 58473257.0,
      "step": 23070
    },
    {
      "entropy": 6.335480451583862,
      "epoch": 2.663012117714945,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043019228538426416,
      "loss": 5.708,
      "mean_token_accuracy": 0.19989538341760635,
      "num_tokens": 58486173.0,
      "step": 23075
    },
    {
      "entropy": 6.257171726226806,
      "epoch": 2.663589151759954,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043016240516342933,
      "loss": 5.5918,
      "mean_token_accuracy": 0.20949176847934722,
      "num_tokens": 58499271.0,
      "step": 23080
    },
    {
      "entropy": 6.273191308975219,
      "epoch": 2.6641661858049623,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043013251972378404,
      "loss": 5.7889,
      "mean_token_accuracy": 0.19261036664247513,
      "num_tokens": 58511751.0,
      "step": 23085
    },
    {
      "entropy": 6.305680561065674,
      "epoch": 2.664743219849971,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00043010262906633355,
      "loss": 5.6488,
      "mean_token_accuracy": 0.2047525018453598,
      "num_tokens": 58525855.0,
      "step": 23090
    },
    {
      "entropy": 6.316273307800293,
      "epoch": 2.66532025389498,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043007273319208326,
      "loss": 5.7233,
      "mean_token_accuracy": 0.20027858912944793,
      "num_tokens": 58537672.0,
      "step": 23095
    },
    {
      "entropy": 6.345172357559204,
      "epoch": 2.6658972879399885,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004300428321020385,
      "loss": 5.75,
      "mean_token_accuracy": 0.19163197427988052,
      "num_tokens": 58550228.0,
      "step": 23100
    },
    {
      "entropy": 6.364157342910767,
      "epoch": 2.666474321984997,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043001292579720514,
      "loss": 5.7027,
      "mean_token_accuracy": 0.20050170868635178,
      "num_tokens": 58562960.0,
      "step": 23105
    },
    {
      "entropy": 6.284440326690674,
      "epoch": 2.6670513560300058,
      "grad_norm": 0.984375,
      "learning_rate": 0.000429983014278589,
      "loss": 5.6724,
      "mean_token_accuracy": 0.2027953490614891,
      "num_tokens": 58575406.0,
      "step": 23110
    },
    {
      "entropy": 6.274879598617554,
      "epoch": 2.6676283900750146,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004299530975471962,
      "loss": 5.6737,
      "mean_token_accuracy": 0.20379206985235215,
      "num_tokens": 58588038.0,
      "step": 23115
    },
    {
      "entropy": 6.348702621459961,
      "epoch": 2.668205424120023,
      "grad_norm": 1.0,
      "learning_rate": 0.0004299231756040329,
      "loss": 5.7539,
      "mean_token_accuracy": 0.19494491517543794,
      "num_tokens": 58600477.0,
      "step": 23120
    },
    {
      "entropy": 6.3002846240997314,
      "epoch": 2.668782458165032,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004298932484501055,
      "loss": 5.645,
      "mean_token_accuracy": 0.20147651731967925,
      "num_tokens": 58613219.0,
      "step": 23125
    },
    {
      "entropy": 6.252772188186645,
      "epoch": 2.6693594922100403,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042986331608642076,
      "loss": 5.6449,
      "mean_token_accuracy": 0.21267254650592804,
      "num_tokens": 58625754.0,
      "step": 23130
    },
    {
      "entropy": 6.289946889877319,
      "epoch": 2.669936526255049,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00042983337851398525,
      "loss": 5.6581,
      "mean_token_accuracy": 0.20351698696613313,
      "num_tokens": 58637707.0,
      "step": 23135
    },
    {
      "entropy": 6.377548027038574,
      "epoch": 2.6705135603000576,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042980343573380595,
      "loss": 5.7601,
      "mean_token_accuracy": 0.19556596130132675,
      "num_tokens": 58650723.0,
      "step": 23140
    },
    {
      "entropy": 6.308612966537476,
      "epoch": 2.6710905943450665,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042977348774688996,
      "loss": 5.7181,
      "mean_token_accuracy": 0.20351560711860656,
      "num_tokens": 58664347.0,
      "step": 23145
    },
    {
      "entropy": 6.258338117599488,
      "epoch": 2.671667628390075,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004297435345542445,
      "loss": 5.6428,
      "mean_token_accuracy": 0.20043806582689286,
      "num_tokens": 58675562.0,
      "step": 23150
    },
    {
      "entropy": 6.2886011600494385,
      "epoch": 2.6722446624350837,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004297135761568773,
      "loss": 5.6675,
      "mean_token_accuracy": 0.20482282489538192,
      "num_tokens": 58689160.0,
      "step": 23155
    },
    {
      "entropy": 6.257562446594238,
      "epoch": 2.672821696480092,
      "grad_norm": 1.0,
      "learning_rate": 0.00042968361255579573,
      "loss": 5.6794,
      "mean_token_accuracy": 0.19902031272649764,
      "num_tokens": 58701699.0,
      "step": 23160
    },
    {
      "entropy": 6.322961807250977,
      "epoch": 2.673398730525101,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042965364375200765,
      "loss": 5.7092,
      "mean_token_accuracy": 0.1982409566640854,
      "num_tokens": 58713399.0,
      "step": 23165
    },
    {
      "entropy": 6.317610025405884,
      "epoch": 2.67397576457011,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004296236697465212,
      "loss": 5.6409,
      "mean_token_accuracy": 0.20403126925230025,
      "num_tokens": 58726985.0,
      "step": 23170
    },
    {
      "entropy": 6.281086444854736,
      "epoch": 2.6745527986151183,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004295936905403444,
      "loss": 5.6622,
      "mean_token_accuracy": 0.20791808217763902,
      "num_tokens": 58739932.0,
      "step": 23175
    },
    {
      "entropy": 6.282163333892822,
      "epoch": 2.6751298326601267,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004295637061344855,
      "loss": 5.6056,
      "mean_token_accuracy": 0.20617732852697374,
      "num_tokens": 58751745.0,
      "step": 23180
    },
    {
      "entropy": 6.254665279388428,
      "epoch": 2.6757068667051356,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004295337165299532,
      "loss": 5.6141,
      "mean_token_accuracy": 0.21312341541051866,
      "num_tokens": 58763872.0,
      "step": 23185
    },
    {
      "entropy": 6.2862225532531735,
      "epoch": 2.6762839007501444,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042950372172775614,
      "loss": 5.6952,
      "mean_token_accuracy": 0.2031521439552307,
      "num_tokens": 58776564.0,
      "step": 23190
    },
    {
      "entropy": 6.265504837036133,
      "epoch": 2.676860934795153,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000429473721728903,
      "loss": 5.625,
      "mean_token_accuracy": 0.20872161835432052,
      "num_tokens": 58789527.0,
      "step": 23195
    },
    {
      "entropy": 6.356201171875,
      "epoch": 2.6774379688401617,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004294437165344031,
      "loss": 5.7847,
      "mean_token_accuracy": 0.19497882425785065,
      "num_tokens": 58803345.0,
      "step": 23200
    },
    {
      "entropy": 6.283417701721191,
      "epoch": 2.67801500288517,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042941370614526554,
      "loss": 5.6963,
      "mean_token_accuracy": 0.20109895914793013,
      "num_tokens": 58815099.0,
      "step": 23205
    },
    {
      "entropy": 6.3835978507995605,
      "epoch": 2.678592036930179,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004293836905624996,
      "loss": 5.6666,
      "mean_token_accuracy": 0.20106339305639268,
      "num_tokens": 58827761.0,
      "step": 23210
    },
    {
      "entropy": 6.194533014297486,
      "epoch": 2.6791690709751874,
      "grad_norm": 0.890625,
      "learning_rate": 0.00042935366978711487,
      "loss": 5.5956,
      "mean_token_accuracy": 0.21061190515756606,
      "num_tokens": 58841119.0,
      "step": 23215
    },
    {
      "entropy": 6.296024417877197,
      "epoch": 2.6797461050201963,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042932364382012125,
      "loss": 5.6842,
      "mean_token_accuracy": 0.20348697453737258,
      "num_tokens": 58853997.0,
      "step": 23220
    },
    {
      "entropy": 6.310422039031982,
      "epoch": 2.6803231390652047,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042929361266252843,
      "loss": 5.7562,
      "mean_token_accuracy": 0.19881097972393036,
      "num_tokens": 58864909.0,
      "step": 23225
    },
    {
      "entropy": 6.237200784683227,
      "epoch": 2.6809001731102136,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004292635763153466,
      "loss": 5.6645,
      "mean_token_accuracy": 0.20698654949665068,
      "num_tokens": 58878371.0,
      "step": 23230
    },
    {
      "entropy": 6.344329357147217,
      "epoch": 2.681477207155222,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042923353477958607,
      "loss": 5.7701,
      "mean_token_accuracy": 0.19743875712156295,
      "num_tokens": 58889200.0,
      "step": 23235
    },
    {
      "entropy": 6.362970209121704,
      "epoch": 2.682054241200231,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042920348805625716,
      "loss": 5.7873,
      "mean_token_accuracy": 0.19367658942937852,
      "num_tokens": 58901324.0,
      "step": 23240
    },
    {
      "entropy": 6.275134420394897,
      "epoch": 2.6826312752452397,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00042917343614637057,
      "loss": 5.7297,
      "mean_token_accuracy": 0.20267629474401475,
      "num_tokens": 58916175.0,
      "step": 23245
    },
    {
      "entropy": 6.365994501113891,
      "epoch": 2.683208309290248,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042914337905093695,
      "loss": 5.705,
      "mean_token_accuracy": 0.20217215418815612,
      "num_tokens": 58928186.0,
      "step": 23250
    },
    {
      "entropy": 6.370102262496948,
      "epoch": 2.6837853433352565,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004291133167709674,
      "loss": 5.7285,
      "mean_token_accuracy": 0.19601055532693862,
      "num_tokens": 58940652.0,
      "step": 23255
    },
    {
      "entropy": 6.19692587852478,
      "epoch": 2.6843623773802654,
      "grad_norm": 1.0,
      "learning_rate": 0.00042908324930747297,
      "loss": 5.6333,
      "mean_token_accuracy": 0.2109960526227951,
      "num_tokens": 58954413.0,
      "step": 23260
    },
    {
      "entropy": 6.328251647949219,
      "epoch": 2.6849394114252743,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042905317666146496,
      "loss": 5.6753,
      "mean_token_accuracy": 0.19735278189182281,
      "num_tokens": 58967198.0,
      "step": 23265
    },
    {
      "entropy": 6.374035978317261,
      "epoch": 2.6855164454702827,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042902309883395497,
      "loss": 5.7367,
      "mean_token_accuracy": 0.2000562697649002,
      "num_tokens": 58980007.0,
      "step": 23270
    },
    {
      "entropy": 6.267512559890747,
      "epoch": 2.6860934795152915,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004289930158259545,
      "loss": 5.5995,
      "mean_token_accuracy": 0.21268263906240464,
      "num_tokens": 58994365.0,
      "step": 23275
    },
    {
      "entropy": 6.283232879638672,
      "epoch": 2.6866705135603,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042896292763847544,
      "loss": 5.6626,
      "mean_token_accuracy": 0.19449129551649094,
      "num_tokens": 59006596.0,
      "step": 23280
    },
    {
      "entropy": 6.314735174179077,
      "epoch": 2.687247547605309,
      "grad_norm": 1.0,
      "learning_rate": 0.00042893283427252984,
      "loss": 5.6951,
      "mean_token_accuracy": 0.20233504176139833,
      "num_tokens": 59021077.0,
      "step": 23285
    },
    {
      "entropy": 6.307542371749878,
      "epoch": 2.6878245816503172,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004289027357291297,
      "loss": 5.7478,
      "mean_token_accuracy": 0.20082065612077712,
      "num_tokens": 59033627.0,
      "step": 23290
    },
    {
      "entropy": 6.2515106201171875,
      "epoch": 2.688401615695326,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004288726320092876,
      "loss": 5.6305,
      "mean_token_accuracy": 0.209030944108963,
      "num_tokens": 59045865.0,
      "step": 23295
    },
    {
      "entropy": 6.401576995849609,
      "epoch": 2.6889786497403345,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004288425231140159,
      "loss": 5.7733,
      "mean_token_accuracy": 0.1993776738643646,
      "num_tokens": 59056683.0,
      "step": 23300
    },
    {
      "entropy": 6.337088775634766,
      "epoch": 2.6895556837853434,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00042881240904432737,
      "loss": 5.7242,
      "mean_token_accuracy": 0.20440036207437515,
      "num_tokens": 59069946.0,
      "step": 23305
    },
    {
      "entropy": 6.124974870681763,
      "epoch": 2.690132717830352,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004287822898012348,
      "loss": 5.5449,
      "mean_token_accuracy": 0.2136279374361038,
      "num_tokens": 59081697.0,
      "step": 23310
    },
    {
      "entropy": 6.348212671279907,
      "epoch": 2.6907097518753607,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004287521653857514,
      "loss": 5.7545,
      "mean_token_accuracy": 0.1979800969362259,
      "num_tokens": 59093818.0,
      "step": 23315
    },
    {
      "entropy": 6.367703628540039,
      "epoch": 2.6912867859203695,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042872203579889025,
      "loss": 5.7084,
      "mean_token_accuracy": 0.20282455831766127,
      "num_tokens": 59107811.0,
      "step": 23320
    },
    {
      "entropy": 6.272344779968262,
      "epoch": 2.691863819965378,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042869190104166476,
      "loss": 5.7016,
      "mean_token_accuracy": 0.20074718296527863,
      "num_tokens": 59120761.0,
      "step": 23325
    },
    {
      "entropy": 6.334949064254761,
      "epoch": 2.6924408540103864,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004286617611150886,
      "loss": 5.7261,
      "mean_token_accuracy": 0.20262394100427628,
      "num_tokens": 59133822.0,
      "step": 23330
    },
    {
      "entropy": 6.27693395614624,
      "epoch": 2.693017888055395,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042863161602017536,
      "loss": 5.6538,
      "mean_token_accuracy": 0.2082969292998314,
      "num_tokens": 59146101.0,
      "step": 23335
    },
    {
      "entropy": 6.376944732666016,
      "epoch": 2.693594922100404,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004286014657579391,
      "loss": 5.7478,
      "mean_token_accuracy": 0.19225982576608658,
      "num_tokens": 59158416.0,
      "step": 23340
    },
    {
      "entropy": 6.288856554031372,
      "epoch": 2.6941719561454125,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004285713103293938,
      "loss": 5.613,
      "mean_token_accuracy": 0.20414792597293854,
      "num_tokens": 59171056.0,
      "step": 23345
    },
    {
      "entropy": 6.3021856307983395,
      "epoch": 2.6947489901904214,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004285411497355537,
      "loss": 5.6779,
      "mean_token_accuracy": 0.20204017609357833,
      "num_tokens": 59183015.0,
      "step": 23350
    },
    {
      "entropy": 6.267050743103027,
      "epoch": 2.6953260242354298,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042851098397743334,
      "loss": 5.6632,
      "mean_token_accuracy": 0.202258338034153,
      "num_tokens": 59195821.0,
      "step": 23355
    },
    {
      "entropy": 6.261898851394653,
      "epoch": 2.6959030582804386,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004284808130560473,
      "loss": 5.5771,
      "mean_token_accuracy": 0.20885568857192993,
      "num_tokens": 59209578.0,
      "step": 23360
    },
    {
      "entropy": 6.255591201782226,
      "epoch": 2.696480092325447,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042845063697241034,
      "loss": 5.6892,
      "mean_token_accuracy": 0.1987489342689514,
      "num_tokens": 59221691.0,
      "step": 23365
    },
    {
      "entropy": 6.254851484298706,
      "epoch": 2.697057126370456,
      "grad_norm": 1.0,
      "learning_rate": 0.0004284204557275374,
      "loss": 5.637,
      "mean_token_accuracy": 0.20686128735542297,
      "num_tokens": 59233820.0,
      "step": 23370
    },
    {
      "entropy": 6.301329278945923,
      "epoch": 2.6976341604154648,
      "grad_norm": 1.0,
      "learning_rate": 0.0004283902693224437,
      "loss": 5.6198,
      "mean_token_accuracy": 0.21021140962839127,
      "num_tokens": 59245065.0,
      "step": 23375
    },
    {
      "entropy": 6.290903472900391,
      "epoch": 2.698211194460473,
      "grad_norm": 0.9375,
      "learning_rate": 0.00042836007775814446,
      "loss": 5.7205,
      "mean_token_accuracy": 0.19731257855892181,
      "num_tokens": 59258711.0,
      "step": 23380
    },
    {
      "entropy": 6.2723784923553465,
      "epoch": 2.6987882285054816,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004283298810356551,
      "loss": 5.6651,
      "mean_token_accuracy": 0.21167835891246795,
      "num_tokens": 59272324.0,
      "step": 23385
    },
    {
      "entropy": 6.228175973892212,
      "epoch": 2.6993652625504905,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004282996791559914,
      "loss": 5.6627,
      "mean_token_accuracy": 0.2084403395652771,
      "num_tokens": 59285860.0,
      "step": 23390
    },
    {
      "entropy": 6.306097888946534,
      "epoch": 2.6999422965954993,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042826947212016924,
      "loss": 5.6059,
      "mean_token_accuracy": 0.20925479084253312,
      "num_tokens": 59298839.0,
      "step": 23395
    },
    {
      "entropy": 6.313662958145142,
      "epoch": 2.7005193306405078,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004282392599292045,
      "loss": 5.7131,
      "mean_token_accuracy": 0.198974372446537,
      "num_tokens": 59311804.0,
      "step": 23400
    },
    {
      "entropy": 6.348927068710327,
      "epoch": 2.701096364685516,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004282090425841133,
      "loss": 5.6422,
      "mean_token_accuracy": 0.20031845718622207,
      "num_tokens": 59323856.0,
      "step": 23405
    },
    {
      "entropy": 6.274404430389405,
      "epoch": 2.701673398730525,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004281788200859121,
      "loss": 5.6383,
      "mean_token_accuracy": 0.20689753741025924,
      "num_tokens": 59336633.0,
      "step": 23410
    },
    {
      "entropy": 6.276021242141724,
      "epoch": 2.702250432775534,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042814859243561727,
      "loss": 5.6674,
      "mean_token_accuracy": 0.2030237302184105,
      "num_tokens": 59349152.0,
      "step": 23415
    },
    {
      "entropy": 6.351522016525268,
      "epoch": 2.7028274668205423,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042811835963424573,
      "loss": 5.7586,
      "mean_token_accuracy": 0.19291130900382997,
      "num_tokens": 59361937.0,
      "step": 23420
    },
    {
      "entropy": 6.2556380271911625,
      "epoch": 2.703404500865551,
      "grad_norm": 1.0,
      "learning_rate": 0.00042808812168281413,
      "loss": 5.6444,
      "mean_token_accuracy": 0.20091213285923004,
      "num_tokens": 59373252.0,
      "step": 23425
    },
    {
      "entropy": 6.238467693328857,
      "epoch": 2.7039815349105596,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042805787858233966,
      "loss": 5.6623,
      "mean_token_accuracy": 0.20646535307168962,
      "num_tokens": 59385929.0,
      "step": 23430
    },
    {
      "entropy": 6.300511169433594,
      "epoch": 2.7045585689555685,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004280276303338395,
      "loss": 5.6739,
      "mean_token_accuracy": 0.207147715985775,
      "num_tokens": 59399020.0,
      "step": 23435
    },
    {
      "entropy": 6.277471113204956,
      "epoch": 2.705135603000577,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000427997376938331,
      "loss": 5.6808,
      "mean_token_accuracy": 0.20572642832994462,
      "num_tokens": 59410926.0,
      "step": 23440
    },
    {
      "entropy": 6.3607253074646,
      "epoch": 2.7057126370455857,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004279671183968317,
      "loss": 5.7647,
      "mean_token_accuracy": 0.19241357147693633,
      "num_tokens": 59422944.0,
      "step": 23445
    },
    {
      "entropy": 6.3600225925445555,
      "epoch": 2.7062896710905946,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004279368547103593,
      "loss": 5.766,
      "mean_token_accuracy": 0.19837311804294586,
      "num_tokens": 59435361.0,
      "step": 23450
    },
    {
      "entropy": 6.3106097221374515,
      "epoch": 2.706866705135603,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004279065858799318,
      "loss": 5.6523,
      "mean_token_accuracy": 0.19905702024698257,
      "num_tokens": 59446964.0,
      "step": 23455
    },
    {
      "entropy": 6.360878133773804,
      "epoch": 2.7074437391806114,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042787631190656725,
      "loss": 5.7202,
      "mean_token_accuracy": 0.20271656811237335,
      "num_tokens": 59460120.0,
      "step": 23460
    },
    {
      "entropy": 6.385726833343506,
      "epoch": 2.7080207732256203,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00042784603279128386,
      "loss": 5.7791,
      "mean_token_accuracy": 0.1957649528980255,
      "num_tokens": 59472149.0,
      "step": 23465
    },
    {
      "entropy": 6.290087127685547,
      "epoch": 2.708597807270629,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004278157485351001,
      "loss": 5.71,
      "mean_token_accuracy": 0.1939320832490921,
      "num_tokens": 59485268.0,
      "step": 23470
    },
    {
      "entropy": 6.361720895767212,
      "epoch": 2.7091748413156376,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042778545913903454,
      "loss": 5.7208,
      "mean_token_accuracy": 0.19970663785934448,
      "num_tokens": 59497473.0,
      "step": 23475
    },
    {
      "entropy": 6.281435537338257,
      "epoch": 2.7097518753606464,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042775516460410594,
      "loss": 5.6042,
      "mean_token_accuracy": 0.20734673291444777,
      "num_tokens": 59511349.0,
      "step": 23480
    },
    {
      "entropy": 6.29406852722168,
      "epoch": 2.710328909405655,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004277248649313333,
      "loss": 5.7379,
      "mean_token_accuracy": 0.19508339315652848,
      "num_tokens": 59524984.0,
      "step": 23485
    },
    {
      "entropy": 6.2247504711151125,
      "epoch": 2.7109059434506637,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004276945601217357,
      "loss": 5.6063,
      "mean_token_accuracy": 0.21221801340579988,
      "num_tokens": 59537857.0,
      "step": 23490
    },
    {
      "entropy": 6.309620428085327,
      "epoch": 2.711482977495672,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042766425017633236,
      "loss": 5.6587,
      "mean_token_accuracy": 0.20382095873355865,
      "num_tokens": 59549640.0,
      "step": 23495
    },
    {
      "entropy": 6.314326667785645,
      "epoch": 2.712060011540681,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042763393509614284,
      "loss": 5.6174,
      "mean_token_accuracy": 0.20749787837266923,
      "num_tokens": 59562231.0,
      "step": 23500
    },
    {
      "entropy": 6.240929079055786,
      "epoch": 2.7126370455856894,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004276036148821867,
      "loss": 5.6638,
      "mean_token_accuracy": 0.2059198021888733,
      "num_tokens": 59574989.0,
      "step": 23505
    },
    {
      "entropy": 6.296467065811157,
      "epoch": 2.7132140796306983,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042757328953548385,
      "loss": 5.6452,
      "mean_token_accuracy": 0.21158193349838256,
      "num_tokens": 59587647.0,
      "step": 23510
    },
    {
      "entropy": 6.304464721679688,
      "epoch": 2.7137911136757067,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004275429590570541,
      "loss": 5.6466,
      "mean_token_accuracy": 0.21240080744028092,
      "num_tokens": 59600643.0,
      "step": 23515
    },
    {
      "entropy": 6.279160356521606,
      "epoch": 2.7143681477207156,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00042751262344791776,
      "loss": 5.6754,
      "mean_token_accuracy": 0.20605522096157075,
      "num_tokens": 59612727.0,
      "step": 23520
    },
    {
      "entropy": 6.240147829055786,
      "epoch": 2.7149451817657244,
      "grad_norm": 0.984375,
      "learning_rate": 0.000427482282709095,
      "loss": 5.6492,
      "mean_token_accuracy": 0.2071886420249939,
      "num_tokens": 59624490.0,
      "step": 23525
    },
    {
      "entropy": 6.340973615646362,
      "epoch": 2.715522215810733,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004274519368416065,
      "loss": 5.6879,
      "mean_token_accuracy": 0.20833470523357392,
      "num_tokens": 59636626.0,
      "step": 23530
    },
    {
      "entropy": 6.301939916610718,
      "epoch": 2.7160992498557412,
      "grad_norm": 1.0,
      "learning_rate": 0.0004274215858464727,
      "loss": 5.6915,
      "mean_token_accuracy": 0.19857121258974075,
      "num_tokens": 59648544.0,
      "step": 23535
    },
    {
      "entropy": 6.268226480484008,
      "epoch": 2.71667628390075,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00042739122972471473,
      "loss": 5.6197,
      "mean_token_accuracy": 0.213410684466362,
      "num_tokens": 59661761.0,
      "step": 23540
    },
    {
      "entropy": 6.285334300994873,
      "epoch": 2.717253317945759,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004273608684773534,
      "loss": 5.6951,
      "mean_token_accuracy": 0.20214398801326752,
      "num_tokens": 59674371.0,
      "step": 23545
    },
    {
      "entropy": 6.356138515472412,
      "epoch": 2.7178303519907674,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042733050210540984,
      "loss": 5.7331,
      "mean_token_accuracy": 0.20313535630702972,
      "num_tokens": 59687385.0,
      "step": 23550
    },
    {
      "entropy": 6.308844518661499,
      "epoch": 2.7184073860357763,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004273001306099056,
      "loss": 5.7191,
      "mean_token_accuracy": 0.2007930815219879,
      "num_tokens": 59700698.0,
      "step": 23555
    },
    {
      "entropy": 6.27632474899292,
      "epoch": 2.7189844200807847,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004272697539918621,
      "loss": 5.6145,
      "mean_token_accuracy": 0.20595595091581345,
      "num_tokens": 59712644.0,
      "step": 23560
    },
    {
      "entropy": 6.257412052154541,
      "epoch": 2.7195614541257935,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004272393722523011,
      "loss": 5.6125,
      "mean_token_accuracy": 0.2060173287987709,
      "num_tokens": 59725739.0,
      "step": 23565
    },
    {
      "entropy": 6.312278604507446,
      "epoch": 2.720138488170802,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042720898539224436,
      "loss": 5.6664,
      "mean_token_accuracy": 0.2027875453233719,
      "num_tokens": 59737693.0,
      "step": 23570
    },
    {
      "entropy": 6.267192220687866,
      "epoch": 2.720715522215811,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004271785934127141,
      "loss": 5.7142,
      "mean_token_accuracy": 0.19817878305912018,
      "num_tokens": 59750011.0,
      "step": 23575
    },
    {
      "entropy": 6.283588600158692,
      "epoch": 2.7212925562608192,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042714819631473234,
      "loss": 5.6217,
      "mean_token_accuracy": 0.20717367976903917,
      "num_tokens": 59762217.0,
      "step": 23580
    },
    {
      "entropy": 6.35259518623352,
      "epoch": 2.721869590305828,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042711779409932164,
      "loss": 5.6734,
      "mean_token_accuracy": 0.20504896491765975,
      "num_tokens": 59774194.0,
      "step": 23585
    },
    {
      "entropy": 6.285326623916626,
      "epoch": 2.7224466243508365,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042708738676750446,
      "loss": 5.7128,
      "mean_token_accuracy": 0.20431385189294815,
      "num_tokens": 59785444.0,
      "step": 23590
    },
    {
      "entropy": 6.283195877075196,
      "epoch": 2.7230236583958454,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004270569743203036,
      "loss": 5.6151,
      "mean_token_accuracy": 0.20313784182071687,
      "num_tokens": 59799166.0,
      "step": 23595
    },
    {
      "entropy": 6.297642087936401,
      "epoch": 2.7236006924408542,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004270265567587419,
      "loss": 5.7499,
      "mean_token_accuracy": 0.19959780126810073,
      "num_tokens": 59811678.0,
      "step": 23600
    },
    {
      "entropy": 6.414148664474487,
      "epoch": 2.7241777264858626,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00042699613408384246,
      "loss": 5.7535,
      "mean_token_accuracy": 0.19851747155189514,
      "num_tokens": 59824373.0,
      "step": 23605
    },
    {
      "entropy": 6.32968487739563,
      "epoch": 2.724754760530871,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004269657062966286,
      "loss": 5.7031,
      "mean_token_accuracy": 0.2034940391778946,
      "num_tokens": 59836834.0,
      "step": 23610
    },
    {
      "entropy": 6.2941139221191404,
      "epoch": 2.72533179457588,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004269352733981238,
      "loss": 5.7161,
      "mean_token_accuracy": 0.2070586934685707,
      "num_tokens": 59848510.0,
      "step": 23615
    },
    {
      "entropy": 6.276827955245972,
      "epoch": 2.725908828620889,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042690483538935137,
      "loss": 5.684,
      "mean_token_accuracy": 0.20437561720609665,
      "num_tokens": 59861153.0,
      "step": 23620
    },
    {
      "entropy": 6.311435174942017,
      "epoch": 2.726485862665897,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004268743922713354,
      "loss": 5.7559,
      "mean_token_accuracy": 0.1990489736199379,
      "num_tokens": 59873329.0,
      "step": 23625
    },
    {
      "entropy": 6.288034105300904,
      "epoch": 2.727062896710906,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042684394404509957,
      "loss": 5.6713,
      "mean_token_accuracy": 0.19572677314281464,
      "num_tokens": 59886143.0,
      "step": 23630
    },
    {
      "entropy": 6.380124616622925,
      "epoch": 2.7276399307559145,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042681349071166814,
      "loss": 5.6729,
      "mean_token_accuracy": 0.2073363333940506,
      "num_tokens": 59897837.0,
      "step": 23635
    },
    {
      "entropy": 6.27218074798584,
      "epoch": 2.7282169648009233,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042678303227206536,
      "loss": 5.6332,
      "mean_token_accuracy": 0.20563837438821791,
      "num_tokens": 59910700.0,
      "step": 23640
    },
    {
      "entropy": 6.309125995635986,
      "epoch": 2.7287939988459318,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004267525687273157,
      "loss": 5.7341,
      "mean_token_accuracy": 0.20194521844387053,
      "num_tokens": 59922469.0,
      "step": 23645
    },
    {
      "entropy": 6.309378337860108,
      "epoch": 2.7293710328909406,
      "grad_norm": 0.9375,
      "learning_rate": 0.00042672210007844383,
      "loss": 5.6712,
      "mean_token_accuracy": 0.19990531653165816,
      "num_tokens": 59935115.0,
      "step": 23650
    },
    {
      "entropy": 6.3465064525604244,
      "epoch": 2.7299480669359495,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042669162632647434,
      "loss": 5.7323,
      "mean_token_accuracy": 0.20108094811439514,
      "num_tokens": 59948649.0,
      "step": 23655
    },
    {
      "entropy": 6.253906965255737,
      "epoch": 2.730525100980958,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042666114747243243,
      "loss": 5.6866,
      "mean_token_accuracy": 0.20712722092866898,
      "num_tokens": 59961713.0,
      "step": 23660
    },
    {
      "entropy": 6.345532989501953,
      "epoch": 2.7311021350259663,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004266306635173432,
      "loss": 5.7019,
      "mean_token_accuracy": 0.2038264602422714,
      "num_tokens": 59974597.0,
      "step": 23665
    },
    {
      "entropy": 6.330751228332519,
      "epoch": 2.731679169070975,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004266001744622319,
      "loss": 5.7051,
      "mean_token_accuracy": 0.19917648881673813,
      "num_tokens": 59987481.0,
      "step": 23670
    },
    {
      "entropy": 6.320524883270264,
      "epoch": 2.732256203115984,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000426569680308124,
      "loss": 5.6617,
      "mean_token_accuracy": 0.20108617842197418,
      "num_tokens": 59999464.0,
      "step": 23675
    },
    {
      "entropy": 6.330943155288696,
      "epoch": 2.7328332371609925,
      "grad_norm": 0.9375,
      "learning_rate": 0.00042653918105604524,
      "loss": 5.7279,
      "mean_token_accuracy": 0.20274415016174316,
      "num_tokens": 60012221.0,
      "step": 23680
    },
    {
      "entropy": 6.272569179534912,
      "epoch": 2.733410271206001,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00042650867670702123,
      "loss": 5.5912,
      "mean_token_accuracy": 0.20528029203414916,
      "num_tokens": 60025331.0,
      "step": 23685
    },
    {
      "entropy": 6.198603534698487,
      "epoch": 2.7339873052510097,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004264781672620783,
      "loss": 5.5045,
      "mean_token_accuracy": 0.22077662497758865,
      "num_tokens": 60038147.0,
      "step": 23690
    },
    {
      "entropy": 6.293131256103516,
      "epoch": 2.7345643392960186,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004264476527222425,
      "loss": 5.7239,
      "mean_token_accuracy": 0.20074425935745238,
      "num_tokens": 60051333.0,
      "step": 23695
    },
    {
      "entropy": 6.2726915836334225,
      "epoch": 2.735141373341027,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004264171330885401,
      "loss": 5.649,
      "mean_token_accuracy": 0.19944679886102676,
      "num_tokens": 60062629.0,
      "step": 23700
    },
    {
      "entropy": 6.321165704727173,
      "epoch": 2.735718407386036,
      "grad_norm": 1.0,
      "learning_rate": 0.00042638660836199756,
      "loss": 5.7308,
      "mean_token_accuracy": 0.20399460792541504,
      "num_tokens": 60075809.0,
      "step": 23705
    },
    {
      "entropy": 6.268091058731079,
      "epoch": 2.7362954414310443,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042635607854364166,
      "loss": 5.634,
      "mean_token_accuracy": 0.20696480125188826,
      "num_tokens": 60089599.0,
      "step": 23710
    },
    {
      "entropy": 6.277140951156616,
      "epoch": 2.736872475476053,
      "grad_norm": 0.9375,
      "learning_rate": 0.00042632554363449926,
      "loss": 5.6743,
      "mean_token_accuracy": 0.20150331109762193,
      "num_tokens": 60102511.0,
      "step": 23715
    },
    {
      "entropy": 6.301269817352295,
      "epoch": 2.7374495095210616,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042629500363559734,
      "loss": 5.6551,
      "mean_token_accuracy": 0.21141407936811446,
      "num_tokens": 60115572.0,
      "step": 23720
    },
    {
      "entropy": 6.347796440124512,
      "epoch": 2.7380265435660704,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042626445854796326,
      "loss": 5.7825,
      "mean_token_accuracy": 0.19758963882923125,
      "num_tokens": 60128121.0,
      "step": 23725
    },
    {
      "entropy": 6.302715826034546,
      "epoch": 2.7386035776110793,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004262339083726241,
      "loss": 5.6951,
      "mean_token_accuracy": 0.20136351734399796,
      "num_tokens": 60141246.0,
      "step": 23730
    },
    {
      "entropy": 6.2851673603057865,
      "epoch": 2.7391806116560877,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004262033531106076,
      "loss": 5.6439,
      "mean_token_accuracy": 0.2097453087568283,
      "num_tokens": 60153698.0,
      "step": 23735
    },
    {
      "entropy": 6.196819734573364,
      "epoch": 2.739757645701096,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042617279276294145,
      "loss": 5.5858,
      "mean_token_accuracy": 0.21945101022720337,
      "num_tokens": 60166354.0,
      "step": 23740
    },
    {
      "entropy": 6.332563829421997,
      "epoch": 2.740334679746105,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004261422273306535,
      "loss": 5.7121,
      "mean_token_accuracy": 0.20087351202964782,
      "num_tokens": 60179753.0,
      "step": 23745
    },
    {
      "entropy": 6.21227388381958,
      "epoch": 2.740911713791114,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004261116568147718,
      "loss": 5.6197,
      "mean_token_accuracy": 0.2115341156721115,
      "num_tokens": 60194231.0,
      "step": 23750
    },
    {
      "entropy": 6.333675384521484,
      "epoch": 2.7414887478361223,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004260810812163246,
      "loss": 5.6423,
      "mean_token_accuracy": 0.2046241655945778,
      "num_tokens": 60206518.0,
      "step": 23755
    },
    {
      "entropy": 6.287317371368408,
      "epoch": 2.742065781881131,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042605050053634027,
      "loss": 5.6773,
      "mean_token_accuracy": 0.19769222438335418,
      "num_tokens": 60217610.0,
      "step": 23760
    },
    {
      "entropy": 6.282694101333618,
      "epoch": 2.7426428159261396,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004260199147758474,
      "loss": 5.7025,
      "mean_token_accuracy": 0.20385900884866714,
      "num_tokens": 60230306.0,
      "step": 23765
    },
    {
      "entropy": 6.323690700531006,
      "epoch": 2.7432198499711484,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004259893239358747,
      "loss": 5.6079,
      "mean_token_accuracy": 0.2116149604320526,
      "num_tokens": 60242555.0,
      "step": 23770
    },
    {
      "entropy": 6.262629461288452,
      "epoch": 2.743796884016157,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042595872801745106,
      "loss": 5.6393,
      "mean_token_accuracy": 0.20700765252113343,
      "num_tokens": 60254866.0,
      "step": 23775
    },
    {
      "entropy": 6.20080304145813,
      "epoch": 2.7443739180611657,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004259281270216056,
      "loss": 5.5796,
      "mean_token_accuracy": 0.2140206888318062,
      "num_tokens": 60267562.0,
      "step": 23780
    },
    {
      "entropy": 6.2951442241668705,
      "epoch": 2.744950952106174,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00042589752094936763,
      "loss": 5.6301,
      "mean_token_accuracy": 0.21043893098831176,
      "num_tokens": 60281579.0,
      "step": 23785
    },
    {
      "entropy": 6.33000864982605,
      "epoch": 2.745527986151183,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004258669098017664,
      "loss": 5.6913,
      "mean_token_accuracy": 0.2015838161110878,
      "num_tokens": 60293019.0,
      "step": 23790
    },
    {
      "entropy": 6.280864620208741,
      "epoch": 2.7461050201961914,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042583629357983164,
      "loss": 5.6766,
      "mean_token_accuracy": 0.19887039065361023,
      "num_tokens": 60304942.0,
      "step": 23795
    },
    {
      "entropy": 6.3675275325775145,
      "epoch": 2.7466820542412003,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042580567228459303,
      "loss": 5.7755,
      "mean_token_accuracy": 0.19905489981174468,
      "num_tokens": 60316275.0,
      "step": 23800
    },
    {
      "entropy": 6.3462159633636475,
      "epoch": 2.747259088286209,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004257750459170806,
      "loss": 5.722,
      "mean_token_accuracy": 0.20228638350963593,
      "num_tokens": 60328305.0,
      "step": 23805
    },
    {
      "entropy": 6.248806715011597,
      "epoch": 2.7478361223312175,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00042574441447832436,
      "loss": 5.6674,
      "mean_token_accuracy": 0.2019027888774872,
      "num_tokens": 60341307.0,
      "step": 23810
    },
    {
      "entropy": 6.257761001586914,
      "epoch": 2.748413156376226,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004257137779693546,
      "loss": 5.6457,
      "mean_token_accuracy": 0.20007436275482177,
      "num_tokens": 60352243.0,
      "step": 23815
    },
    {
      "entropy": 6.375108909606934,
      "epoch": 2.748990190421235,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004256831363912018,
      "loss": 5.7146,
      "mean_token_accuracy": 0.2022201955318451,
      "num_tokens": 60365128.0,
      "step": 23820
    },
    {
      "entropy": 6.3249894142150875,
      "epoch": 2.7495672244662437,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042565248974489654,
      "loss": 5.68,
      "mean_token_accuracy": 0.1965864673256874,
      "num_tokens": 60376901.0,
      "step": 23825
    },
    {
      "entropy": 6.239331245422363,
      "epoch": 2.750144258511252,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004256218380314697,
      "loss": 5.5974,
      "mean_token_accuracy": 0.21113515198230742,
      "num_tokens": 60388357.0,
      "step": 23830
    },
    {
      "entropy": 6.284210252761841,
      "epoch": 2.750721292556261,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004255911812519521,
      "loss": 5.6803,
      "mean_token_accuracy": 0.2017245188355446,
      "num_tokens": 60400495.0,
      "step": 23835
    },
    {
      "entropy": 6.345863962173462,
      "epoch": 2.7512983266012694,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004255605194073749,
      "loss": 5.6968,
      "mean_token_accuracy": 0.19924532026052474,
      "num_tokens": 60413884.0,
      "step": 23840
    },
    {
      "entropy": 6.26440954208374,
      "epoch": 2.7518753606462782,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004255298524987695,
      "loss": 5.6415,
      "mean_token_accuracy": 0.19940197914838792,
      "num_tokens": 60426745.0,
      "step": 23845
    },
    {
      "entropy": 6.283579778671265,
      "epoch": 2.7524523946912867,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004254991805271672,
      "loss": 5.6925,
      "mean_token_accuracy": 0.19681255370378495,
      "num_tokens": 60440679.0,
      "step": 23850
    },
    {
      "entropy": 6.312639856338501,
      "epoch": 2.7530294287362955,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042546850349359976,
      "loss": 5.6625,
      "mean_token_accuracy": 0.20387524962425232,
      "num_tokens": 60453517.0,
      "step": 23855
    },
    {
      "entropy": 6.300760126113891,
      "epoch": 2.753606462781304,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042543782139909894,
      "loss": 5.6848,
      "mean_token_accuracy": 0.20168583542108537,
      "num_tokens": 60466475.0,
      "step": 23860
    },
    {
      "entropy": 6.262047004699707,
      "epoch": 2.754183496826313,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042540713424469667,
      "loss": 5.5726,
      "mean_token_accuracy": 0.2065932884812355,
      "num_tokens": 60479204.0,
      "step": 23865
    },
    {
      "entropy": 6.23803973197937,
      "epoch": 2.7547605308713212,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042537644203142523,
      "loss": 5.6409,
      "mean_token_accuracy": 0.20556975305080413,
      "num_tokens": 60491970.0,
      "step": 23870
    },
    {
      "entropy": 6.323281574249267,
      "epoch": 2.75533756491633,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00042534574476031683,
      "loss": 5.7608,
      "mean_token_accuracy": 0.19690536707639694,
      "num_tokens": 60505083.0,
      "step": 23875
    },
    {
      "entropy": 6.319398403167725,
      "epoch": 2.755914598961339,
      "grad_norm": 1.015625,
      "learning_rate": 0.000425315042432404,
      "loss": 5.7035,
      "mean_token_accuracy": 0.20399057269096374,
      "num_tokens": 60517884.0,
      "step": 23880
    },
    {
      "entropy": 6.100912141799927,
      "epoch": 2.7564916330063474,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00042528433504871935,
      "loss": 5.4766,
      "mean_token_accuracy": 0.21970414668321608,
      "num_tokens": 60530824.0,
      "step": 23885
    },
    {
      "entropy": 6.232281970977783,
      "epoch": 2.757068667051356,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004252536226102958,
      "loss": 5.6017,
      "mean_token_accuracy": 0.2118530049920082,
      "num_tokens": 60544992.0,
      "step": 23890
    },
    {
      "entropy": 6.367205429077148,
      "epoch": 2.7576457010963646,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004252229051181662,
      "loss": 5.732,
      "mean_token_accuracy": 0.2028604805469513,
      "num_tokens": 60558280.0,
      "step": 23895
    },
    {
      "entropy": 6.196542739868164,
      "epoch": 2.7582227351413735,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042519218257336383,
      "loss": 5.6248,
      "mean_token_accuracy": 0.20808978378772736,
      "num_tokens": 60571292.0,
      "step": 23900
    },
    {
      "entropy": 6.270206069946289,
      "epoch": 2.758799769186382,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042516145497692204,
      "loss": 5.6571,
      "mean_token_accuracy": 0.1989253744482994,
      "num_tokens": 60583605.0,
      "step": 23905
    },
    {
      "entropy": 6.3446009159088135,
      "epoch": 2.759376803231391,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042513072232987426,
      "loss": 5.7004,
      "mean_token_accuracy": 0.20143622606992723,
      "num_tokens": 60596332.0,
      "step": 23910
    },
    {
      "entropy": 6.292921352386474,
      "epoch": 2.759953837276399,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004250999846332543,
      "loss": 5.6785,
      "mean_token_accuracy": 0.20390901118516921,
      "num_tokens": 60608398.0,
      "step": 23915
    },
    {
      "entropy": 6.289727783203125,
      "epoch": 2.760530871321408,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042506924188809574,
      "loss": 5.7117,
      "mean_token_accuracy": 0.19905742555856704,
      "num_tokens": 60621250.0,
      "step": 23920
    },
    {
      "entropy": 6.3303979396820065,
      "epoch": 2.7611079053664165,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000425038494095433,
      "loss": 5.6848,
      "mean_token_accuracy": 0.19703512787818908,
      "num_tokens": 60633933.0,
      "step": 23925
    },
    {
      "entropy": 6.324948263168335,
      "epoch": 2.7616849394114253,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042500774125629986,
      "loss": 5.7157,
      "mean_token_accuracy": 0.20182085931301116,
      "num_tokens": 60645856.0,
      "step": 23930
    },
    {
      "entropy": 6.290975522994995,
      "epoch": 2.762261973456434,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004249769833717309,
      "loss": 5.6703,
      "mean_token_accuracy": 0.20467734187841416,
      "num_tokens": 60658490.0,
      "step": 23935
    },
    {
      "entropy": 6.263008165359497,
      "epoch": 2.7628390075014426,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042494622044276066,
      "loss": 5.6172,
      "mean_token_accuracy": 0.20724271535873412,
      "num_tokens": 60671639.0,
      "step": 23940
    },
    {
      "entropy": 6.31686429977417,
      "epoch": 2.763416041546451,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004249154524704237,
      "loss": 5.725,
      "mean_token_accuracy": 0.20578849464654922,
      "num_tokens": 60683915.0,
      "step": 23945
    },
    {
      "entropy": 6.213581705093384,
      "epoch": 2.76399307559146,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000424884679455755,
      "loss": 5.627,
      "mean_token_accuracy": 0.21153458952903748,
      "num_tokens": 60695871.0,
      "step": 23950
    },
    {
      "entropy": 6.399197435379028,
      "epoch": 2.7645701096364688,
      "grad_norm": 0.9375,
      "learning_rate": 0.00042485390139978955,
      "loss": 5.7563,
      "mean_token_accuracy": 0.19963719546794892,
      "num_tokens": 60709029.0,
      "step": 23955
    },
    {
      "entropy": 6.339520692825317,
      "epoch": 2.765147143681477,
      "grad_norm": 1.0,
      "learning_rate": 0.0004248231183035626,
      "loss": 5.6591,
      "mean_token_accuracy": 0.20782131105661392,
      "num_tokens": 60721639.0,
      "step": 23960
    },
    {
      "entropy": 6.324173545837402,
      "epoch": 2.7657241777264856,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004247923301681095,
      "loss": 5.6534,
      "mean_token_accuracy": 0.21259555369615554,
      "num_tokens": 60733702.0,
      "step": 23965
    },
    {
      "entropy": 6.327179336547852,
      "epoch": 2.7663012117714945,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042476153699446567,
      "loss": 5.7507,
      "mean_token_accuracy": 0.19707633256912233,
      "num_tokens": 60746521.0,
      "step": 23970
    },
    {
      "entropy": 6.325180530548096,
      "epoch": 2.7668782458165033,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042473073878366695,
      "loss": 5.6967,
      "mean_token_accuracy": 0.19857559055089952,
      "num_tokens": 60760091.0,
      "step": 23975
    },
    {
      "entropy": 6.284106636047364,
      "epoch": 2.7674552798615117,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004246999355367493,
      "loss": 5.6066,
      "mean_token_accuracy": 0.20383976697921752,
      "num_tokens": 60773021.0,
      "step": 23980
    },
    {
      "entropy": 6.272068738937378,
      "epoch": 2.7680323139065206,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042466912725474865,
      "loss": 5.6655,
      "mean_token_accuracy": 0.20759907513856887,
      "num_tokens": 60786057.0,
      "step": 23985
    },
    {
      "entropy": 6.317358446121216,
      "epoch": 2.768609347951529,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042463831393870123,
      "loss": 5.7334,
      "mean_token_accuracy": 0.20188882648944856,
      "num_tokens": 60799097.0,
      "step": 23990
    },
    {
      "entropy": 6.367629623413086,
      "epoch": 2.769186381996538,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042460749558964346,
      "loss": 5.7451,
      "mean_token_accuracy": 0.19779258817434311,
      "num_tokens": 60811404.0,
      "step": 23995
    },
    {
      "entropy": 6.2804468154907225,
      "epoch": 2.7697634160415463,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004245766722086118,
      "loss": 5.6206,
      "mean_token_accuracy": 0.21009254455566406,
      "num_tokens": 60823210.0,
      "step": 24000
    },
    {
      "epoch": 2.7697634160415463,
      "eval_entropy": 6.060240020526889,
      "eval_loss": 5.763261318206787,
      "eval_mean_token_accuracy": 0.20716380341074364,
      "eval_num_tokens": 60823210.0,
      "eval_runtime": 17.6656,
      "eval_samples_per_second": 1592.704,
      "eval_steps_per_second": 199.088,
      "step": 24000
    }
  ],
  "logging_steps": 5,
  "max_steps": 86650,
  "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": 1.330108238610432e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}