eng-latn-10mb-after-ppt-shuff-dyck-100mb-ckpt500_seed3407/checkpoint-9000/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.010782276632785,
  "eval_steps": 3000,
  "global_step": 9000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 4.958112335205078,
      "epoch": 0.0005615769079575448,
      "grad_norm": 13.5625,
      "learning_rate": 2e-06,
      "loss": 14.2842,
      "mean_token_accuracy": 0.0,
      "num_tokens": 11659.0,
      "step": 5
    },
    {
      "entropy": 4.9424741744995115,
      "epoch": 0.0011231538159150896,
      "grad_norm": 13.5625,
      "learning_rate": 4.5e-06,
      "loss": 14.1707,
      "mean_token_accuracy": 0.00011547344038262964,
      "num_tokens": 21794.0,
      "step": 10
    },
    {
      "entropy": 4.973624753952026,
      "epoch": 0.0016847307238726343,
      "grad_norm": 13.6875,
      "learning_rate": 7e-06,
      "loss": 14.149,
      "mean_token_accuracy": 6.915629492141307e-05,
      "num_tokens": 33206.0,
      "step": 15
    },
    {
      "entropy": 4.9998602867126465,
      "epoch": 0.0022463076318301792,
      "grad_norm": 15.6875,
      "learning_rate": 9.5e-06,
      "loss": 14.0249,
      "mean_token_accuracy": 0.00014555996167473496,
      "num_tokens": 44487.0,
      "step": 20
    },
    {
      "entropy": 5.110756301879883,
      "epoch": 0.002807884539787724,
      "grad_norm": 17.875,
      "learning_rate": 1.2e-05,
      "loss": 13.6677,
      "mean_token_accuracy": 7.246377062983811e-05,
      "num_tokens": 56062.0,
      "step": 25
    },
    {
      "entropy": 5.332080888748169,
      "epoch": 0.0033694614477452686,
      "grad_norm": 20.875,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 13.1848,
      "mean_token_accuracy": 0.0,
      "num_tokens": 67453.0,
      "step": 30
    },
    {
      "entropy": 6.396451139450074,
      "epoch": 0.003931038355702813,
      "grad_norm": 25.375,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 12.2399,
      "mean_token_accuracy": 6.105006323195994e-05,
      "num_tokens": 78527.0,
      "step": 35
    },
    {
      "entropy": 9.540253734588623,
      "epoch": 0.0044926152636603585,
      "grad_norm": 4.125,
      "learning_rate": 1.95e-05,
      "loss": 11.0354,
      "mean_token_accuracy": 5.87199057918042e-05,
      "num_tokens": 89267.0,
      "step": 40
    },
    {
      "entropy": 10.682201671600343,
      "epoch": 0.005054192171617903,
      "grad_norm": 3.25,
      "learning_rate": 2.2e-05,
      "loss": 10.738,
      "mean_token_accuracy": 0.007576325320405886,
      "num_tokens": 102035.0,
      "step": 45
    },
    {
      "entropy": 10.73370714187622,
      "epoch": 0.005615769079575448,
      "grad_norm": 3.0,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 10.5603,
      "mean_token_accuracy": 0.015592619962990283,
      "num_tokens": 112734.0,
      "step": 50
    },
    {
      "entropy": 10.730180168151856,
      "epoch": 0.006177345987532993,
      "grad_norm": 2.359375,
      "learning_rate": 2.7e-05,
      "loss": 10.3452,
      "mean_token_accuracy": 0.01546380277723074,
      "num_tokens": 123854.0,
      "step": 55
    },
    {
      "entropy": 10.69598331451416,
      "epoch": 0.006738922895490537,
      "grad_norm": 2.25,
      "learning_rate": 2.95e-05,
      "loss": 10.1223,
      "mean_token_accuracy": 0.01719926903024316,
      "num_tokens": 133444.0,
      "step": 60
    },
    {
      "entropy": 10.701205444335937,
      "epoch": 0.007300499803448083,
      "grad_norm": 2.109375,
      "learning_rate": 3.2e-05,
      "loss": 9.9707,
      "mean_token_accuracy": 0.01521061807870865,
      "num_tokens": 144678.0,
      "step": 65
    },
    {
      "entropy": 10.72805700302124,
      "epoch": 0.007862076711405626,
      "grad_norm": 2.15625,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 9.8757,
      "mean_token_accuracy": 0.02308740708976984,
      "num_tokens": 155866.0,
      "step": 70
    },
    {
      "entropy": 10.729030323028564,
      "epoch": 0.008423653619363172,
      "grad_norm": 1.9765625,
      "learning_rate": 3.7e-05,
      "loss": 9.783,
      "mean_token_accuracy": 0.04029188398271799,
      "num_tokens": 166965.0,
      "step": 75
    },
    {
      "entropy": 10.717912292480468,
      "epoch": 0.008985230527320717,
      "grad_norm": 1.953125,
      "learning_rate": 3.95e-05,
      "loss": 9.6967,
      "mean_token_accuracy": 0.03837036974728107,
      "num_tokens": 177423.0,
      "step": 80
    },
    {
      "entropy": 10.714910125732422,
      "epoch": 0.00954680743527826,
      "grad_norm": 1.8984375,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 9.6005,
      "mean_token_accuracy": 0.039133700914680956,
      "num_tokens": 188687.0,
      "step": 85
    },
    {
      "entropy": 10.715478420257568,
      "epoch": 0.010108384343235806,
      "grad_norm": 1.875,
      "learning_rate": 4.45e-05,
      "loss": 9.5344,
      "mean_token_accuracy": 0.04223253391683102,
      "num_tokens": 200173.0,
      "step": 90
    },
    {
      "entropy": 10.7022141456604,
      "epoch": 0.010669961251193351,
      "grad_norm": 1.9375,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 9.4043,
      "mean_token_accuracy": 0.038442989625036714,
      "num_tokens": 210877.0,
      "step": 95
    },
    {
      "entropy": 10.681341648101807,
      "epoch": 0.011231538159150897,
      "grad_norm": 1.8984375,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 9.2957,
      "mean_token_accuracy": 0.03458300586789846,
      "num_tokens": 221770.0,
      "step": 100
    },
    {
      "entropy": 10.660041809082031,
      "epoch": 0.01179311506710844,
      "grad_norm": 1.8515625,
      "learning_rate": 5.2e-05,
      "loss": 9.172,
      "mean_token_accuracy": 0.03827377744019032,
      "num_tokens": 232140.0,
      "step": 105
    },
    {
      "entropy": 10.632814598083495,
      "epoch": 0.012354691975065986,
      "grad_norm": 1.734375,
      "learning_rate": 5.45e-05,
      "loss": 9.1093,
      "mean_token_accuracy": 0.041411877423524854,
      "num_tokens": 244186.0,
      "step": 110
    },
    {
      "entropy": 10.598613166809082,
      "epoch": 0.012916268883023531,
      "grad_norm": 1.75,
      "learning_rate": 5.7e-05,
      "loss": 8.9603,
      "mean_token_accuracy": 0.04186610747128725,
      "num_tokens": 254805.0,
      "step": 115
    },
    {
      "entropy": 10.550034332275391,
      "epoch": 0.013477845790981074,
      "grad_norm": 1.8125,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 8.7951,
      "mean_token_accuracy": 0.050945664569735524,
      "num_tokens": 265055.0,
      "step": 120
    },
    {
      "entropy": 10.497636890411377,
      "epoch": 0.01403942269893862,
      "grad_norm": 1.71875,
      "learning_rate": 6.2e-05,
      "loss": 8.6784,
      "mean_token_accuracy": 0.05012401342391968,
      "num_tokens": 275850.0,
      "step": 125
    },
    {
      "entropy": 10.439189529418945,
      "epoch": 0.014600999606896165,
      "grad_norm": 1.53125,
      "learning_rate": 6.450000000000001e-05,
      "loss": 8.5813,
      "mean_token_accuracy": 0.04501255042850971,
      "num_tokens": 286802.0,
      "step": 130
    },
    {
      "entropy": 10.362538623809815,
      "epoch": 0.015162576514853709,
      "grad_norm": 1.578125,
      "learning_rate": 6.7e-05,
      "loss": 8.5087,
      "mean_token_accuracy": 0.04473419226706028,
      "num_tokens": 297070.0,
      "step": 135
    },
    {
      "entropy": 10.26766414642334,
      "epoch": 0.015724153422811252,
      "grad_norm": 1.53125,
      "learning_rate": 6.950000000000001e-05,
      "loss": 8.4016,
      "mean_token_accuracy": 0.0456852950155735,
      "num_tokens": 309617.0,
      "step": 140
    },
    {
      "entropy": 10.18057041168213,
      "epoch": 0.0162857303307688,
      "grad_norm": 1.515625,
      "learning_rate": 7.2e-05,
      "loss": 8.2802,
      "mean_token_accuracy": 0.04733948037028313,
      "num_tokens": 321310.0,
      "step": 145
    },
    {
      "entropy": 10.039380741119384,
      "epoch": 0.016847307238726343,
      "grad_norm": 1.3984375,
      "learning_rate": 7.45e-05,
      "loss": 8.1192,
      "mean_token_accuracy": 0.05164117477834225,
      "num_tokens": 332243.0,
      "step": 150
    },
    {
      "entropy": 9.87264642715454,
      "epoch": 0.017408884146683887,
      "grad_norm": 1.1171875,
      "learning_rate": 7.7e-05,
      "loss": 8.0322,
      "mean_token_accuracy": 0.05423316806554794,
      "num_tokens": 342847.0,
      "step": 155
    },
    {
      "entropy": 9.697960758209229,
      "epoch": 0.017970461054641434,
      "grad_norm": 1.2109375,
      "learning_rate": 7.950000000000001e-05,
      "loss": 7.8819,
      "mean_token_accuracy": 0.054126177728176114,
      "num_tokens": 353570.0,
      "step": 160
    },
    {
      "entropy": 9.482976245880128,
      "epoch": 0.018532037962598977,
      "grad_norm": 1.140625,
      "learning_rate": 8.2e-05,
      "loss": 7.7722,
      "mean_token_accuracy": 0.05961498692631721,
      "num_tokens": 364330.0,
      "step": 165
    },
    {
      "entropy": 9.207988929748534,
      "epoch": 0.01909361487055652,
      "grad_norm": 1.25,
      "learning_rate": 8.450000000000001e-05,
      "loss": 7.7247,
      "mean_token_accuracy": 0.05897176414728165,
      "num_tokens": 374930.0,
      "step": 170
    },
    {
      "entropy": 8.946532917022704,
      "epoch": 0.019655191778514068,
      "grad_norm": 0.8828125,
      "learning_rate": 8.7e-05,
      "loss": 7.6508,
      "mean_token_accuracy": 0.05788501463830471,
      "num_tokens": 384690.0,
      "step": 175
    },
    {
      "entropy": 8.74663782119751,
      "epoch": 0.020216768686471612,
      "grad_norm": 0.8515625,
      "learning_rate": 8.95e-05,
      "loss": 7.6041,
      "mean_token_accuracy": 0.05669615678489208,
      "num_tokens": 395753.0,
      "step": 180
    },
    {
      "entropy": 8.500962829589843,
      "epoch": 0.020778345594429155,
      "grad_norm": 0.75390625,
      "learning_rate": 9.2e-05,
      "loss": 7.5318,
      "mean_token_accuracy": 0.06435371935367584,
      "num_tokens": 407049.0,
      "step": 185
    },
    {
      "entropy": 8.297107601165772,
      "epoch": 0.021339922502386702,
      "grad_norm": 0.8046875,
      "learning_rate": 9.45e-05,
      "loss": 7.5052,
      "mean_token_accuracy": 0.07400269955396652,
      "num_tokens": 417942.0,
      "step": 190
    },
    {
      "entropy": 8.161279296875,
      "epoch": 0.021901499410344246,
      "grad_norm": 0.76953125,
      "learning_rate": 9.7e-05,
      "loss": 7.5143,
      "mean_token_accuracy": 0.06796157322824001,
      "num_tokens": 428151.0,
      "step": 195
    },
    {
      "entropy": 8.059497261047364,
      "epoch": 0.022463076318301793,
      "grad_norm": 0.7109375,
      "learning_rate": 9.95e-05,
      "loss": 7.5599,
      "mean_token_accuracy": 0.06278462670743465,
      "num_tokens": 439837.0,
      "step": 200
    },
    {
      "entropy": 8.035380172729493,
      "epoch": 0.023024653226259337,
      "grad_norm": 0.62109375,
      "learning_rate": 0.000102,
      "loss": 7.4149,
      "mean_token_accuracy": 0.0609861858189106,
      "num_tokens": 450357.0,
      "step": 205
    },
    {
      "entropy": 7.987471294403076,
      "epoch": 0.02358623013421688,
      "grad_norm": 0.734375,
      "learning_rate": 0.00010449999999999999,
      "loss": 7.4119,
      "mean_token_accuracy": 0.06853476613759994,
      "num_tokens": 461174.0,
      "step": 210
    },
    {
      "entropy": 7.993591213226319,
      "epoch": 0.024147807042174427,
      "grad_norm": 1.3125,
      "learning_rate": 0.000107,
      "loss": 7.4641,
      "mean_token_accuracy": 0.0716362364590168,
      "num_tokens": 472785.0,
      "step": 215
    },
    {
      "entropy": 7.965004348754883,
      "epoch": 0.02470938395013197,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0001095,
      "loss": 7.4134,
      "mean_token_accuracy": 0.07565828561782836,
      "num_tokens": 483248.0,
      "step": 220
    },
    {
      "entropy": 7.868141603469849,
      "epoch": 0.025270960858089515,
      "grad_norm": 1.109375,
      "learning_rate": 0.000112,
      "loss": 7.3137,
      "mean_token_accuracy": 0.0714982945472002,
      "num_tokens": 494690.0,
      "step": 225
    },
    {
      "entropy": 7.927273750305176,
      "epoch": 0.025832537766047062,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0001145,
      "loss": 7.4905,
      "mean_token_accuracy": 0.07402925454080105,
      "num_tokens": 506192.0,
      "step": 230
    },
    {
      "entropy": 7.935389471054077,
      "epoch": 0.026394114674004605,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.3528,
      "mean_token_accuracy": 0.0740197867155075,
      "num_tokens": 517152.0,
      "step": 235
    },
    {
      "entropy": 7.856435060501099,
      "epoch": 0.02695569158196215,
      "grad_norm": 0.8046875,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.2736,
      "mean_token_accuracy": 0.08407649844884872,
      "num_tokens": 527836.0,
      "step": 240
    },
    {
      "entropy": 7.854199695587158,
      "epoch": 0.027517268489919696,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000122,
      "loss": 7.3984,
      "mean_token_accuracy": 0.07311258018016815,
      "num_tokens": 539491.0,
      "step": 245
    },
    {
      "entropy": 7.888166761398315,
      "epoch": 0.02807884539787724,
      "grad_norm": 1.1875,
      "learning_rate": 0.0001245,
      "loss": 7.3729,
      "mean_token_accuracy": 0.08043631911277771,
      "num_tokens": 550816.0,
      "step": 250
    },
    {
      "entropy": 7.793630981445313,
      "epoch": 0.028640422305834783,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000127,
      "loss": 7.3921,
      "mean_token_accuracy": 0.0777372658252716,
      "num_tokens": 561566.0,
      "step": 255
    },
    {
      "entropy": 7.83606104850769,
      "epoch": 0.02920199921379233,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0001295,
      "loss": 7.323,
      "mean_token_accuracy": 0.07814248651266098,
      "num_tokens": 573032.0,
      "step": 260
    },
    {
      "entropy": 7.867829418182373,
      "epoch": 0.029763576121749874,
      "grad_norm": 0.81640625,
      "learning_rate": 0.000132,
      "loss": 7.2565,
      "mean_token_accuracy": 0.08499897569417954,
      "num_tokens": 584140.0,
      "step": 265
    },
    {
      "entropy": 7.7688757419586185,
      "epoch": 0.030325153029707418,
      "grad_norm": 0.90625,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.3227,
      "mean_token_accuracy": 0.08433490246534348,
      "num_tokens": 595097.0,
      "step": 270
    },
    {
      "entropy": 7.861677742004394,
      "epoch": 0.030886729937664965,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00013700000000000002,
      "loss": 7.3152,
      "mean_token_accuracy": 0.08390655070543289,
      "num_tokens": 606797.0,
      "step": 275
    },
    {
      "entropy": 7.725890064239502,
      "epoch": 0.031448306845622505,
      "grad_norm": 1.484375,
      "learning_rate": 0.0001395,
      "loss": 7.2734,
      "mean_token_accuracy": 0.08379090949892998,
      "num_tokens": 618187.0,
      "step": 280
    },
    {
      "entropy": 7.801093626022339,
      "epoch": 0.032009883753580055,
      "grad_norm": 2.625,
      "learning_rate": 0.00014199999999999998,
      "loss": 7.2249,
      "mean_token_accuracy": 0.0867884710431099,
      "num_tokens": 629829.0,
      "step": 285
    },
    {
      "entropy": 7.699099493026734,
      "epoch": 0.0325714606615376,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0001445,
      "loss": 7.2257,
      "mean_token_accuracy": 0.09084633439779281,
      "num_tokens": 640914.0,
      "step": 290
    },
    {
      "entropy": 7.680918836593628,
      "epoch": 0.03313303756949514,
      "grad_norm": 1.234375,
      "learning_rate": 0.000147,
      "loss": 7.2187,
      "mean_token_accuracy": 0.08423622846603393,
      "num_tokens": 653057.0,
      "step": 295
    },
    {
      "entropy": 7.726095199584961,
      "epoch": 0.033694614477452686,
      "grad_norm": 1.09375,
      "learning_rate": 0.0001495,
      "loss": 7.1693,
      "mean_token_accuracy": 0.08970561251044273,
      "num_tokens": 663437.0,
      "step": 300
    },
    {
      "entropy": 7.676259231567383,
      "epoch": 0.03425619138541023,
      "grad_norm": 1.046875,
      "learning_rate": 0.000152,
      "loss": 7.2165,
      "mean_token_accuracy": 0.08746050000190735,
      "num_tokens": 675416.0,
      "step": 305
    },
    {
      "entropy": 7.61683201789856,
      "epoch": 0.03481776829336777,
      "grad_norm": 1.6875,
      "learning_rate": 0.00015450000000000001,
      "loss": 7.0378,
      "mean_token_accuracy": 0.09221994504332542,
      "num_tokens": 686633.0,
      "step": 310
    },
    {
      "entropy": 7.593777084350586,
      "epoch": 0.035379345201325324,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000157,
      "loss": 7.1128,
      "mean_token_accuracy": 0.09131084829568863,
      "num_tokens": 698065.0,
      "step": 315
    },
    {
      "entropy": 7.737735366821289,
      "epoch": 0.03594092210928287,
      "grad_norm": 1.4375,
      "learning_rate": 0.0001595,
      "loss": 7.1167,
      "mean_token_accuracy": 0.09759770259261132,
      "num_tokens": 708652.0,
      "step": 320
    },
    {
      "entropy": 7.677718257904052,
      "epoch": 0.03650249901724041,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000162,
      "loss": 7.2492,
      "mean_token_accuracy": 0.08909616991877556,
      "num_tokens": 720199.0,
      "step": 325
    },
    {
      "entropy": 7.6231176376342775,
      "epoch": 0.037064075925197955,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00016450000000000001,
      "loss": 7.2337,
      "mean_token_accuracy": 0.09042609483003616,
      "num_tokens": 732964.0,
      "step": 330
    },
    {
      "entropy": 7.56429443359375,
      "epoch": 0.0376256528331555,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00016700000000000002,
      "loss": 7.0063,
      "mean_token_accuracy": 0.09593740776181221,
      "num_tokens": 743528.0,
      "step": 335
    },
    {
      "entropy": 7.644824886322022,
      "epoch": 0.03818722974111304,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00016950000000000003,
      "loss": 7.0692,
      "mean_token_accuracy": 0.09663922786712646,
      "num_tokens": 754365.0,
      "step": 340
    },
    {
      "entropy": 7.584464263916016,
      "epoch": 0.03874880664907059,
      "grad_norm": 1.34375,
      "learning_rate": 0.00017199999999999998,
      "loss": 6.9887,
      "mean_token_accuracy": 0.09488176926970482,
      "num_tokens": 766094.0,
      "step": 345
    },
    {
      "entropy": 7.501060867309571,
      "epoch": 0.039310383557028136,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00017449999999999999,
      "loss": 7.039,
      "mean_token_accuracy": 0.09812524616718292,
      "num_tokens": 777275.0,
      "step": 350
    },
    {
      "entropy": 7.562764501571655,
      "epoch": 0.03987196046498568,
      "grad_norm": 1.0625,
      "learning_rate": 0.000177,
      "loss": 7.085,
      "mean_token_accuracy": 0.1003342442214489,
      "num_tokens": 789436.0,
      "step": 355
    },
    {
      "entropy": 7.534398317337036,
      "epoch": 0.040433537372943223,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0001795,
      "loss": 7.1145,
      "mean_token_accuracy": 0.0983487956225872,
      "num_tokens": 800164.0,
      "step": 360
    },
    {
      "entropy": 7.422997093200683,
      "epoch": 0.04099511428090077,
      "grad_norm": 1.03125,
      "learning_rate": 0.000182,
      "loss": 6.9878,
      "mean_token_accuracy": 0.09883297383785247,
      "num_tokens": 811797.0,
      "step": 365
    },
    {
      "entropy": 7.490043687820434,
      "epoch": 0.04155669118885831,
      "grad_norm": 1.46875,
      "learning_rate": 0.0001845,
      "loss": 6.9941,
      "mean_token_accuracy": 0.10256326273083687,
      "num_tokens": 824158.0,
      "step": 370
    },
    {
      "entropy": 7.519336462020874,
      "epoch": 0.04211826809681586,
      "grad_norm": 1.5859375,
      "learning_rate": 0.000187,
      "loss": 6.9989,
      "mean_token_accuracy": 0.09889804199337959,
      "num_tokens": 835803.0,
      "step": 375
    },
    {
      "entropy": 7.509277868270874,
      "epoch": 0.042679845004773405,
      "grad_norm": 0.9375,
      "learning_rate": 0.0001895,
      "loss": 7.0624,
      "mean_token_accuracy": 0.09996038749814033,
      "num_tokens": 848106.0,
      "step": 380
    },
    {
      "entropy": 7.411794948577881,
      "epoch": 0.04324142191273095,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000192,
      "loss": 7.0267,
      "mean_token_accuracy": 0.09782384037971496,
      "num_tokens": 859101.0,
      "step": 385
    },
    {
      "entropy": 7.444434595108032,
      "epoch": 0.04380299882068849,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0001945,
      "loss": 6.9103,
      "mean_token_accuracy": 0.1043923944234848,
      "num_tokens": 870309.0,
      "step": 390
    },
    {
      "entropy": 7.441433763504028,
      "epoch": 0.044364575728646036,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00019700000000000002,
      "loss": 6.9829,
      "mean_token_accuracy": 0.10260054618120193,
      "num_tokens": 881515.0,
      "step": 395
    },
    {
      "entropy": 7.445163917541504,
      "epoch": 0.044926152636603586,
      "grad_norm": 1.46875,
      "learning_rate": 0.00019950000000000002,
      "loss": 6.9395,
      "mean_token_accuracy": 0.10442078933119774,
      "num_tokens": 892129.0,
      "step": 400
    },
    {
      "entropy": 7.414169979095459,
      "epoch": 0.04548772954456113,
      "grad_norm": 1.1875,
      "learning_rate": 0.000202,
      "loss": 6.9523,
      "mean_token_accuracy": 0.10221787467598915,
      "num_tokens": 903186.0,
      "step": 405
    },
    {
      "entropy": 7.402553081512451,
      "epoch": 0.046049306452518673,
      "grad_norm": 1.3125,
      "learning_rate": 0.00020449999999999998,
      "loss": 6.9065,
      "mean_token_accuracy": 0.10986820012331008,
      "num_tokens": 913471.0,
      "step": 410
    },
    {
      "entropy": 7.422558689117432,
      "epoch": 0.04661088336047622,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000207,
      "loss": 6.9043,
      "mean_token_accuracy": 0.10425306111574173,
      "num_tokens": 924283.0,
      "step": 415
    },
    {
      "entropy": 7.360500431060791,
      "epoch": 0.04717246026843376,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002095,
      "loss": 6.9377,
      "mean_token_accuracy": 0.108520770072937,
      "num_tokens": 934975.0,
      "step": 420
    },
    {
      "entropy": 7.330597305297852,
      "epoch": 0.047734037176391304,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000212,
      "loss": 6.8123,
      "mean_token_accuracy": 0.11223476156592369,
      "num_tokens": 945587.0,
      "step": 425
    },
    {
      "entropy": 7.300155162811279,
      "epoch": 0.048295614084348855,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002145,
      "loss": 6.8755,
      "mean_token_accuracy": 0.10690404027700424,
      "num_tokens": 955935.0,
      "step": 430
    },
    {
      "entropy": 7.336331605911255,
      "epoch": 0.0488571909923064,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00021700000000000002,
      "loss": 6.8399,
      "mean_token_accuracy": 0.10845198333263398,
      "num_tokens": 967960.0,
      "step": 435
    },
    {
      "entropy": 7.356525516510009,
      "epoch": 0.04941876790026394,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002195,
      "loss": 6.8741,
      "mean_token_accuracy": 0.11335771456360817,
      "num_tokens": 979680.0,
      "step": 440
    },
    {
      "entropy": 7.2876073837280275,
      "epoch": 0.049980344808221486,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000222,
      "loss": 6.936,
      "mean_token_accuracy": 0.10629516914486885,
      "num_tokens": 990600.0,
      "step": 445
    },
    {
      "entropy": 7.374933099746704,
      "epoch": 0.05054192171617903,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002245,
      "loss": 6.9701,
      "mean_token_accuracy": 0.11168414130806922,
      "num_tokens": 1002331.0,
      "step": 450
    },
    {
      "entropy": 7.251807832717896,
      "epoch": 0.05110349862413657,
      "grad_norm": 1.421875,
      "learning_rate": 0.00022700000000000002,
      "loss": 6.776,
      "mean_token_accuracy": 0.10971713438630104,
      "num_tokens": 1013473.0,
      "step": 455
    },
    {
      "entropy": 7.274680042266846,
      "epoch": 0.051665075532094124,
      "grad_norm": 1.03125,
      "learning_rate": 0.00022950000000000002,
      "loss": 6.8667,
      "mean_token_accuracy": 0.10595520064234734,
      "num_tokens": 1024434.0,
      "step": 460
    },
    {
      "entropy": 7.303326034545899,
      "epoch": 0.05222665244005167,
      "grad_norm": 1.203125,
      "learning_rate": 0.00023200000000000003,
      "loss": 6.8918,
      "mean_token_accuracy": 0.10734717026352883,
      "num_tokens": 1035452.0,
      "step": 465
    },
    {
      "entropy": 7.202190494537353,
      "epoch": 0.05278822934800921,
      "grad_norm": 1.578125,
      "learning_rate": 0.00023449999999999998,
      "loss": 6.7608,
      "mean_token_accuracy": 0.10601013004779816,
      "num_tokens": 1046213.0,
      "step": 470
    },
    {
      "entropy": 7.248915481567383,
      "epoch": 0.053349806255966754,
      "grad_norm": 1.328125,
      "learning_rate": 0.000237,
      "loss": 6.7926,
      "mean_token_accuracy": 0.10846332460641861,
      "num_tokens": 1057469.0,
      "step": 475
    },
    {
      "entropy": 7.389639139175415,
      "epoch": 0.0539113831639243,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002395,
      "loss": 6.9242,
      "mean_token_accuracy": 0.11156258061528206,
      "num_tokens": 1069543.0,
      "step": 480
    },
    {
      "entropy": 7.164505100250244,
      "epoch": 0.05447296007188184,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000242,
      "loss": 6.7856,
      "mean_token_accuracy": 0.1143708311021328,
      "num_tokens": 1080476.0,
      "step": 485
    },
    {
      "entropy": 7.269598531723022,
      "epoch": 0.05503453697983939,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002445,
      "loss": 6.9008,
      "mean_token_accuracy": 0.11152186244726181,
      "num_tokens": 1091700.0,
      "step": 490
    },
    {
      "entropy": 7.293481063842774,
      "epoch": 0.055596113887796936,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000247,
      "loss": 6.7849,
      "mean_token_accuracy": 0.1107654757797718,
      "num_tokens": 1103056.0,
      "step": 495
    },
    {
      "entropy": 7.240784788131714,
      "epoch": 0.05615769079575448,
      "grad_norm": 1.375,
      "learning_rate": 0.0002495,
      "loss": 6.8439,
      "mean_token_accuracy": 0.11618472486734391,
      "num_tokens": 1114836.0,
      "step": 500
    },
    {
      "entropy": 7.114566326141357,
      "epoch": 0.05671926770371202,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000252,
      "loss": 6.7378,
      "mean_token_accuracy": 0.11399558708071708,
      "num_tokens": 1125993.0,
      "step": 505
    },
    {
      "entropy": 7.239473915100097,
      "epoch": 0.05728084461166957,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002545,
      "loss": 6.8169,
      "mean_token_accuracy": 0.11478226408362388,
      "num_tokens": 1136719.0,
      "step": 510
    },
    {
      "entropy": 7.177135276794433,
      "epoch": 0.05784242151962711,
      "grad_norm": 1.40625,
      "learning_rate": 0.000257,
      "loss": 6.6864,
      "mean_token_accuracy": 0.11818328648805618,
      "num_tokens": 1146947.0,
      "step": 515
    },
    {
      "entropy": 7.201584386825561,
      "epoch": 0.05840399842758466,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002595,
      "loss": 6.8165,
      "mean_token_accuracy": 0.11641246154904365,
      "num_tokens": 1157974.0,
      "step": 520
    },
    {
      "entropy": 7.081555986404419,
      "epoch": 0.058965575335542204,
      "grad_norm": 1.421875,
      "learning_rate": 0.000262,
      "loss": 6.7415,
      "mean_token_accuracy": 0.1143609032034874,
      "num_tokens": 1168573.0,
      "step": 525
    },
    {
      "entropy": 7.127064418792725,
      "epoch": 0.05952715224349975,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00026450000000000003,
      "loss": 6.656,
      "mean_token_accuracy": 0.11710669472813606,
      "num_tokens": 1178446.0,
      "step": 530
    },
    {
      "entropy": 7.138171005249023,
      "epoch": 0.06008872915145729,
      "grad_norm": 1.15625,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.6753,
      "mean_token_accuracy": 0.11334134191274643,
      "num_tokens": 1190558.0,
      "step": 535
    },
    {
      "entropy": 7.144500207901001,
      "epoch": 0.060650306059414835,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00026950000000000005,
      "loss": 6.8331,
      "mean_token_accuracy": 0.11418996006250381,
      "num_tokens": 1201929.0,
      "step": 540
    },
    {
      "entropy": 7.027964735031128,
      "epoch": 0.06121188296737238,
      "grad_norm": 1.0,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.666,
      "mean_token_accuracy": 0.12040585204958916,
      "num_tokens": 1213155.0,
      "step": 545
    },
    {
      "entropy": 7.229724979400634,
      "epoch": 0.06177345987532993,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002745,
      "loss": 6.7297,
      "mean_token_accuracy": 0.11603076979517937,
      "num_tokens": 1223961.0,
      "step": 550
    },
    {
      "entropy": 7.008941268920898,
      "epoch": 0.06233503678328747,
      "grad_norm": 1.125,
      "learning_rate": 0.000277,
      "loss": 6.6331,
      "mean_token_accuracy": 0.12098302766680717,
      "num_tokens": 1233888.0,
      "step": 555
    },
    {
      "entropy": 6.994447660446167,
      "epoch": 0.06289661369124501,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002795,
      "loss": 6.7383,
      "mean_token_accuracy": 0.11173689067363739,
      "num_tokens": 1244405.0,
      "step": 560
    },
    {
      "entropy": 7.166880035400391,
      "epoch": 0.06345819059920256,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.682,
      "mean_token_accuracy": 0.11965646147727967,
      "num_tokens": 1257099.0,
      "step": 565
    },
    {
      "entropy": 7.068090343475342,
      "epoch": 0.06401976750716011,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002845,
      "loss": 6.7839,
      "mean_token_accuracy": 0.12307375743985176,
      "num_tokens": 1267965.0,
      "step": 570
    },
    {
      "entropy": 7.09981951713562,
      "epoch": 0.06458134441511765,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000287,
      "loss": 6.6873,
      "mean_token_accuracy": 0.11978518962860107,
      "num_tokens": 1280075.0,
      "step": 575
    },
    {
      "entropy": 7.102661514282227,
      "epoch": 0.0651429213230752,
      "grad_norm": 1.375,
      "learning_rate": 0.0002895,
      "loss": 6.6995,
      "mean_token_accuracy": 0.12262161523103714,
      "num_tokens": 1290574.0,
      "step": 580
    },
    {
      "entropy": 7.051478481292724,
      "epoch": 0.06570449823103273,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000292,
      "loss": 6.6635,
      "mean_token_accuracy": 0.12353429943323135,
      "num_tokens": 1302680.0,
      "step": 585
    },
    {
      "entropy": 7.008921670913696,
      "epoch": 0.06626607513899029,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002945,
      "loss": 6.6518,
      "mean_token_accuracy": 0.12372544482350349,
      "num_tokens": 1314577.0,
      "step": 590
    },
    {
      "entropy": 7.094518613815308,
      "epoch": 0.06682765204694784,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000297,
      "loss": 6.7729,
      "mean_token_accuracy": 0.1234229065477848,
      "num_tokens": 1326560.0,
      "step": 595
    },
    {
      "entropy": 7.065892362594605,
      "epoch": 0.06738922895490537,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002995,
      "loss": 6.8373,
      "mean_token_accuracy": 0.11262822896242142,
      "num_tokens": 1339105.0,
      "step": 600
    },
    {
      "entropy": 6.947021389007569,
      "epoch": 0.06795080586286292,
      "grad_norm": 1.140625,
      "learning_rate": 0.000302,
      "loss": 6.5099,
      "mean_token_accuracy": 0.1292965017259121,
      "num_tokens": 1349914.0,
      "step": 605
    },
    {
      "entropy": 7.016496849060059,
      "epoch": 0.06851238277082046,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003045,
      "loss": 6.714,
      "mean_token_accuracy": 0.12088268026709556,
      "num_tokens": 1361700.0,
      "step": 610
    },
    {
      "entropy": 7.005446863174439,
      "epoch": 0.06907395967877801,
      "grad_norm": 1.4375,
      "learning_rate": 0.000307,
      "loss": 6.7339,
      "mean_token_accuracy": 0.12360861897468567,
      "num_tokens": 1373103.0,
      "step": 615
    },
    {
      "entropy": 6.974551010131836,
      "epoch": 0.06963553658673555,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0003095,
      "loss": 6.5487,
      "mean_token_accuracy": 0.12374898120760917,
      "num_tokens": 1383742.0,
      "step": 620
    },
    {
      "entropy": 6.948500871658325,
      "epoch": 0.0701971134946931,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000312,
      "loss": 6.6968,
      "mean_token_accuracy": 0.11977056413888931,
      "num_tokens": 1396656.0,
      "step": 625
    },
    {
      "entropy": 6.922845840454102,
      "epoch": 0.07075869040265065,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003145,
      "loss": 6.5605,
      "mean_token_accuracy": 0.13307109400629996,
      "num_tokens": 1406709.0,
      "step": 630
    },
    {
      "entropy": 7.017679023742676,
      "epoch": 0.07132026731060818,
      "grad_norm": 1.296875,
      "learning_rate": 0.000317,
      "loss": 6.6002,
      "mean_token_accuracy": 0.12924449220299722,
      "num_tokens": 1417793.0,
      "step": 635
    },
    {
      "entropy": 7.013447713851929,
      "epoch": 0.07188184421856574,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003195,
      "loss": 6.678,
      "mean_token_accuracy": 0.12412039786577225,
      "num_tokens": 1429093.0,
      "step": 640
    },
    {
      "entropy": 6.820969581604004,
      "epoch": 0.07244342112652327,
      "grad_norm": 0.984375,
      "learning_rate": 0.000322,
      "loss": 6.5974,
      "mean_token_accuracy": 0.1283337689936161,
      "num_tokens": 1440907.0,
      "step": 645
    },
    {
      "entropy": 6.997115421295166,
      "epoch": 0.07300499803448082,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.6672,
      "mean_token_accuracy": 0.12072568610310555,
      "num_tokens": 1452089.0,
      "step": 650
    },
    {
      "entropy": 6.984156370162964,
      "epoch": 0.07356657494243837,
      "grad_norm": 1.296875,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.674,
      "mean_token_accuracy": 0.12165596857666969,
      "num_tokens": 1464378.0,
      "step": 655
    },
    {
      "entropy": 6.951331949234008,
      "epoch": 0.07412815185039591,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.6099,
      "mean_token_accuracy": 0.12271932810544968,
      "num_tokens": 1476678.0,
      "step": 660
    },
    {
      "entropy": 6.843834352493286,
      "epoch": 0.07468972875835346,
      "grad_norm": 1.078125,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.4578,
      "mean_token_accuracy": 0.12957773357629776,
      "num_tokens": 1486808.0,
      "step": 665
    },
    {
      "entropy": 6.973170900344849,
      "epoch": 0.075251305666311,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.7373,
      "mean_token_accuracy": 0.12345995232462884,
      "num_tokens": 1499281.0,
      "step": 670
    },
    {
      "entropy": 7.001640558242798,
      "epoch": 0.07581288257426855,
      "grad_norm": 1.515625,
      "learning_rate": 0.000337,
      "loss": 6.6501,
      "mean_token_accuracy": 0.12579731568694114,
      "num_tokens": 1510202.0,
      "step": 675
    },
    {
      "entropy": 6.91340651512146,
      "epoch": 0.07637445948222608,
      "grad_norm": 1.125,
      "learning_rate": 0.0003395,
      "loss": 6.5045,
      "mean_token_accuracy": 0.13105906471610068,
      "num_tokens": 1521492.0,
      "step": 680
    },
    {
      "entropy": 6.800015640258789,
      "epoch": 0.07693603639018363,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000342,
      "loss": 6.5206,
      "mean_token_accuracy": 0.13716215938329696,
      "num_tokens": 1532856.0,
      "step": 685
    },
    {
      "entropy": 6.814711189270019,
      "epoch": 0.07749761329814119,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.5356,
      "mean_token_accuracy": 0.12902490720152854,
      "num_tokens": 1543408.0,
      "step": 690
    },
    {
      "entropy": 6.86055703163147,
      "epoch": 0.07805919020609872,
      "grad_norm": 1.421875,
      "learning_rate": 0.000347,
      "loss": 6.5525,
      "mean_token_accuracy": 0.12741926535964013,
      "num_tokens": 1555474.0,
      "step": 695
    },
    {
      "entropy": 6.898800182342529,
      "epoch": 0.07862076711405627,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003495,
      "loss": 6.6414,
      "mean_token_accuracy": 0.1282556474208832,
      "num_tokens": 1567484.0,
      "step": 700
    },
    {
      "entropy": 6.824853563308716,
      "epoch": 0.07918234402201381,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000352,
      "loss": 6.4866,
      "mean_token_accuracy": 0.1352921135723591,
      "num_tokens": 1578204.0,
      "step": 705
    },
    {
      "entropy": 6.884771060943604,
      "epoch": 0.07974392092997136,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003545,
      "loss": 6.4897,
      "mean_token_accuracy": 0.1282044917345047,
      "num_tokens": 1589012.0,
      "step": 710
    },
    {
      "entropy": 6.743309259414673,
      "epoch": 0.08030549783792891,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000357,
      "loss": 6.496,
      "mean_token_accuracy": 0.12868643477559089,
      "num_tokens": 1601099.0,
      "step": 715
    },
    {
      "entropy": 6.910572957992554,
      "epoch": 0.08086707474588645,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003595,
      "loss": 6.6518,
      "mean_token_accuracy": 0.1232347123324871,
      "num_tokens": 1612814.0,
      "step": 720
    },
    {
      "entropy": 6.851531744003296,
      "epoch": 0.081428651653844,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000362,
      "loss": 6.543,
      "mean_token_accuracy": 0.12722352966666223,
      "num_tokens": 1624953.0,
      "step": 725
    },
    {
      "entropy": 6.843756008148193,
      "epoch": 0.08199022856180153,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003645,
      "loss": 6.5685,
      "mean_token_accuracy": 0.12934383600950242,
      "num_tokens": 1635673.0,
      "step": 730
    },
    {
      "entropy": 6.824569320678711,
      "epoch": 0.08255180546975908,
      "grad_norm": 1.21875,
      "learning_rate": 0.000367,
      "loss": 6.486,
      "mean_token_accuracy": 0.135271368175745,
      "num_tokens": 1647282.0,
      "step": 735
    },
    {
      "entropy": 6.793824338912964,
      "epoch": 0.08311338237771662,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0003695,
      "loss": 6.4985,
      "mean_token_accuracy": 0.13493574112653733,
      "num_tokens": 1658426.0,
      "step": 740
    },
    {
      "entropy": 6.795789480209351,
      "epoch": 0.08367495928567417,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000372,
      "loss": 6.5264,
      "mean_token_accuracy": 0.130235705524683,
      "num_tokens": 1669958.0,
      "step": 745
    },
    {
      "entropy": 6.745253849029541,
      "epoch": 0.08423653619363172,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003745,
      "loss": 6.5678,
      "mean_token_accuracy": 0.1286076597869396,
      "num_tokens": 1682409.0,
      "step": 750
    },
    {
      "entropy": 6.817635822296142,
      "epoch": 0.08479811310158926,
      "grad_norm": 1.1875,
      "learning_rate": 0.000377,
      "loss": 6.4985,
      "mean_token_accuracy": 0.1403331384062767,
      "num_tokens": 1692409.0,
      "step": 755
    },
    {
      "entropy": 6.841214847564697,
      "epoch": 0.08535969000954681,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003795,
      "loss": 6.5853,
      "mean_token_accuracy": 0.12864278703927995,
      "num_tokens": 1705018.0,
      "step": 760
    },
    {
      "entropy": 6.764056587219239,
      "epoch": 0.08592126691750435,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000382,
      "loss": 6.5082,
      "mean_token_accuracy": 0.12993483915925025,
      "num_tokens": 1715625.0,
      "step": 765
    },
    {
      "entropy": 6.736582279205322,
      "epoch": 0.0864828438254619,
      "grad_norm": 1.125,
      "learning_rate": 0.0003845,
      "loss": 6.5812,
      "mean_token_accuracy": 0.12060019001364708,
      "num_tokens": 1727045.0,
      "step": 770
    },
    {
      "entropy": 6.8794553756713865,
      "epoch": 0.08704442073341945,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.6304,
      "mean_token_accuracy": 0.12890330255031585,
      "num_tokens": 1737374.0,
      "step": 775
    },
    {
      "entropy": 6.838842105865479,
      "epoch": 0.08760599764137698,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.6398,
      "mean_token_accuracy": 0.13158344849944115,
      "num_tokens": 1747788.0,
      "step": 780
    },
    {
      "entropy": 6.650408744812012,
      "epoch": 0.08816757454933453,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.3406,
      "mean_token_accuracy": 0.14368992894887925,
      "num_tokens": 1757769.0,
      "step": 785
    },
    {
      "entropy": 6.700500535964966,
      "epoch": 0.08872915145729207,
      "grad_norm": 1.359375,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.4506,
      "mean_token_accuracy": 0.12652962952852248,
      "num_tokens": 1769246.0,
      "step": 790
    },
    {
      "entropy": 6.819548749923706,
      "epoch": 0.08929072836524962,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.5743,
      "mean_token_accuracy": 0.1305612899363041,
      "num_tokens": 1781306.0,
      "step": 795
    },
    {
      "entropy": 6.736283588409424,
      "epoch": 0.08985230527320717,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003995,
      "loss": 6.4786,
      "mean_token_accuracy": 0.13247107341885567,
      "num_tokens": 1792888.0,
      "step": 800
    },
    {
      "entropy": 6.741839790344239,
      "epoch": 0.09041388218116471,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000402,
      "loss": 6.4969,
      "mean_token_accuracy": 0.13426485359668733,
      "num_tokens": 1802938.0,
      "step": 805
    },
    {
      "entropy": 6.803853511810303,
      "epoch": 0.09097545908912226,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004045,
      "loss": 6.5178,
      "mean_token_accuracy": 0.13269057348370553,
      "num_tokens": 1814738.0,
      "step": 810
    },
    {
      "entropy": 6.784767770767212,
      "epoch": 0.0915370359970798,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.5774,
      "mean_token_accuracy": 0.13045511841773988,
      "num_tokens": 1825100.0,
      "step": 815
    },
    {
      "entropy": 6.586725234985352,
      "epoch": 0.09209861290503735,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004095,
      "loss": 6.4276,
      "mean_token_accuracy": 0.13769749775528908,
      "num_tokens": 1835879.0,
      "step": 820
    },
    {
      "entropy": 6.844768095016479,
      "epoch": 0.09266018981299488,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000412,
      "loss": 6.5658,
      "mean_token_accuracy": 0.12576449438929557,
      "num_tokens": 1848352.0,
      "step": 825
    },
    {
      "entropy": 6.820197343826294,
      "epoch": 0.09322176672095243,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004145,
      "loss": 6.5146,
      "mean_token_accuracy": 0.13076762184500695,
      "num_tokens": 1859687.0,
      "step": 830
    },
    {
      "entropy": 6.698688888549805,
      "epoch": 0.09378334362890998,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000417,
      "loss": 6.5072,
      "mean_token_accuracy": 0.1330545090138912,
      "num_tokens": 1871438.0,
      "step": 835
    },
    {
      "entropy": 6.674928617477417,
      "epoch": 0.09434492053686752,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004195,
      "loss": 6.3932,
      "mean_token_accuracy": 0.13914670795202255,
      "num_tokens": 1881519.0,
      "step": 840
    },
    {
      "entropy": 6.793725776672363,
      "epoch": 0.09490649744482507,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000422,
      "loss": 6.56,
      "mean_token_accuracy": 0.13139583691954612,
      "num_tokens": 1892428.0,
      "step": 845
    },
    {
      "entropy": 6.715428733825684,
      "epoch": 0.09546807435278261,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004245,
      "loss": 6.4517,
      "mean_token_accuracy": 0.13642125204205513,
      "num_tokens": 1903760.0,
      "step": 850
    },
    {
      "entropy": 6.728550386428833,
      "epoch": 0.09602965126074016,
      "grad_norm": 0.953125,
      "learning_rate": 0.000427,
      "loss": 6.5196,
      "mean_token_accuracy": 0.1250755563378334,
      "num_tokens": 1915430.0,
      "step": 855
    },
    {
      "entropy": 6.690254163742066,
      "epoch": 0.09659122816869771,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004295,
      "loss": 6.4753,
      "mean_token_accuracy": 0.12775300145149232,
      "num_tokens": 1925709.0,
      "step": 860
    },
    {
      "entropy": 6.814657306671142,
      "epoch": 0.09715280507665525,
      "grad_norm": 0.984375,
      "learning_rate": 0.000432,
      "loss": 6.5115,
      "mean_token_accuracy": 0.133428718149662,
      "num_tokens": 1938148.0,
      "step": 865
    },
    {
      "entropy": 6.454801988601685,
      "epoch": 0.0977143819846128,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004345,
      "loss": 6.3948,
      "mean_token_accuracy": 0.13895176202058793,
      "num_tokens": 1949028.0,
      "step": 870
    },
    {
      "entropy": 6.735638570785523,
      "epoch": 0.09827595889257033,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000437,
      "loss": 6.452,
      "mean_token_accuracy": 0.1338992938399315,
      "num_tokens": 1960353.0,
      "step": 875
    },
    {
      "entropy": 6.726486015319824,
      "epoch": 0.09883753580052788,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004395,
      "loss": 6.5459,
      "mean_token_accuracy": 0.12802935019135475,
      "num_tokens": 1972335.0,
      "step": 880
    },
    {
      "entropy": 6.657772254943848,
      "epoch": 0.09939911270848542,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000442,
      "loss": 6.3937,
      "mean_token_accuracy": 0.13370347768068314,
      "num_tokens": 1982851.0,
      "step": 885
    },
    {
      "entropy": 6.599719619750976,
      "epoch": 0.09996068961644297,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004445,
      "loss": 6.3824,
      "mean_token_accuracy": 0.13738489374518395,
      "num_tokens": 1993931.0,
      "step": 890
    },
    {
      "entropy": 6.713817548751831,
      "epoch": 0.10052226652440052,
      "grad_norm": 0.90625,
      "learning_rate": 0.000447,
      "loss": 6.5012,
      "mean_token_accuracy": 0.13419413417577744,
      "num_tokens": 2004838.0,
      "step": 895
    },
    {
      "entropy": 6.532304430007935,
      "epoch": 0.10108384343235806,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.3051,
      "mean_token_accuracy": 0.13725042045116426,
      "num_tokens": 2015620.0,
      "step": 900
    },
    {
      "entropy": 6.6708457469940186,
      "epoch": 0.10164542034031561,
      "grad_norm": 1.125,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.4426,
      "mean_token_accuracy": 0.13907212615013123,
      "num_tokens": 2027489.0,
      "step": 905
    },
    {
      "entropy": 6.573997211456299,
      "epoch": 0.10220699724827315,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.3063,
      "mean_token_accuracy": 0.14410469308495522,
      "num_tokens": 2038475.0,
      "step": 910
    },
    {
      "entropy": 6.5359595775604244,
      "epoch": 0.1027685741562307,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.3129,
      "mean_token_accuracy": 0.1398226350545883,
      "num_tokens": 2048938.0,
      "step": 915
    },
    {
      "entropy": 6.542961549758911,
      "epoch": 0.10333015106418825,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.4021,
      "mean_token_accuracy": 0.1346525728702545,
      "num_tokens": 2060372.0,
      "step": 920
    },
    {
      "entropy": 6.594799661636353,
      "epoch": 0.10389172797214578,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000462,
      "loss": 6.3928,
      "mean_token_accuracy": 0.13547346442937852,
      "num_tokens": 2071558.0,
      "step": 925
    },
    {
      "entropy": 6.629229116439819,
      "epoch": 0.10445330488010333,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004645,
      "loss": 6.4373,
      "mean_token_accuracy": 0.13637368977069855,
      "num_tokens": 2082350.0,
      "step": 930
    },
    {
      "entropy": 6.613821411132813,
      "epoch": 0.10501488178806087,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000467,
      "loss": 6.4865,
      "mean_token_accuracy": 0.13623699843883513,
      "num_tokens": 2093448.0,
      "step": 935
    },
    {
      "entropy": 6.735397148132324,
      "epoch": 0.10557645869601842,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004695,
      "loss": 6.5141,
      "mean_token_accuracy": 0.13197903707623482,
      "num_tokens": 2104171.0,
      "step": 940
    },
    {
      "entropy": 6.624809694290161,
      "epoch": 0.10613803560397596,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000472,
      "loss": 6.3949,
      "mean_token_accuracy": 0.13641267418861389,
      "num_tokens": 2114496.0,
      "step": 945
    },
    {
      "entropy": 6.655260181427002,
      "epoch": 0.10669961251193351,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004745,
      "loss": 6.4007,
      "mean_token_accuracy": 0.13775101825594901,
      "num_tokens": 2125930.0,
      "step": 950
    },
    {
      "entropy": 6.571065235137939,
      "epoch": 0.10726118941989106,
      "grad_norm": 1.171875,
      "learning_rate": 0.000477,
      "loss": 6.357,
      "mean_token_accuracy": 0.14058973491191865,
      "num_tokens": 2136452.0,
      "step": 955
    },
    {
      "entropy": 6.6056571960449215,
      "epoch": 0.1078227663278486,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004795,
      "loss": 6.4536,
      "mean_token_accuracy": 0.13291025757789612,
      "num_tokens": 2148233.0,
      "step": 960
    },
    {
      "entropy": 6.473015594482422,
      "epoch": 0.10838434323580615,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000482,
      "loss": 6.2405,
      "mean_token_accuracy": 0.14449345842003822,
      "num_tokens": 2159475.0,
      "step": 965
    },
    {
      "entropy": 6.528903770446777,
      "epoch": 0.10894592014376368,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004845,
      "loss": 6.3596,
      "mean_token_accuracy": 0.13835327625274657,
      "num_tokens": 2170042.0,
      "step": 970
    },
    {
      "entropy": 6.592466688156128,
      "epoch": 0.10950749705172123,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000487,
      "loss": 6.3605,
      "mean_token_accuracy": 0.13567833229899406,
      "num_tokens": 2182234.0,
      "step": 975
    },
    {
      "entropy": 6.484764099121094,
      "epoch": 0.11006907395967878,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004895,
      "loss": 6.3535,
      "mean_token_accuracy": 0.14025619626045227,
      "num_tokens": 2192477.0,
      "step": 980
    },
    {
      "entropy": 6.500120496749878,
      "epoch": 0.11063065086763632,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000492,
      "loss": 6.3405,
      "mean_token_accuracy": 0.13677674606442453,
      "num_tokens": 2203082.0,
      "step": 985
    },
    {
      "entropy": 6.521809720993042,
      "epoch": 0.11119222777559387,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004945,
      "loss": 6.3143,
      "mean_token_accuracy": 0.13907622545957565,
      "num_tokens": 2213684.0,
      "step": 990
    },
    {
      "entropy": 6.506216716766358,
      "epoch": 0.11175380468355141,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000497,
      "loss": 6.3189,
      "mean_token_accuracy": 0.13710450679063796,
      "num_tokens": 2225524.0,
      "step": 995
    },
    {
      "entropy": 6.529984045028686,
      "epoch": 0.11231538159150896,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004995,
      "loss": 6.2864,
      "mean_token_accuracy": 0.1366613119840622,
      "num_tokens": 2236004.0,
      "step": 1000
    },
    {
      "entropy": 6.607108402252197,
      "epoch": 0.1128769584994665,
      "grad_norm": 1.09375,
      "learning_rate": 0.000499999997707492,
      "loss": 6.4427,
      "mean_token_accuracy": 0.12697047293186187,
      "num_tokens": 2247730.0,
      "step": 1005
    },
    {
      "entropy": 6.58273491859436,
      "epoch": 0.11343853540742405,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999999883941787,
      "loss": 6.3966,
      "mean_token_accuracy": 0.14159852117300034,
      "num_tokens": 2260132.0,
      "step": 1010
    },
    {
      "entropy": 6.481048250198365,
      "epoch": 0.1140001123153816,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999999719167785,
      "loss": 6.3627,
      "mean_token_accuracy": 0.13624601066112518,
      "num_tokens": 2270060.0,
      "step": 1015
    },
    {
      "entropy": 6.499448776245117,
      "epoch": 0.11456168922333913,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004999999482752917,
      "loss": 6.3727,
      "mean_token_accuracy": 0.13926395475864412,
      "num_tokens": 2281568.0,
      "step": 1020
    },
    {
      "entropy": 6.562040138244629,
      "epoch": 0.11512326613129668,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999999174697195,
      "loss": 6.4069,
      "mean_token_accuracy": 0.12946545705199242,
      "num_tokens": 2293325.0,
      "step": 1025
    },
    {
      "entropy": 6.611707639694214,
      "epoch": 0.11568484303925422,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999998795000627,
      "loss": 6.4062,
      "mean_token_accuracy": 0.14011984914541245,
      "num_tokens": 2304780.0,
      "step": 1030
    },
    {
      "entropy": 6.517575311660766,
      "epoch": 0.11624641994721177,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999998343663223,
      "loss": 6.3312,
      "mean_token_accuracy": 0.136796210706234,
      "num_tokens": 2315999.0,
      "step": 1035
    },
    {
      "entropy": 6.508590507507324,
      "epoch": 0.11680799685516932,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999997820685,
      "loss": 6.478,
      "mean_token_accuracy": 0.13875503316521645,
      "num_tokens": 2326695.0,
      "step": 1040
    },
    {
      "entropy": 6.5880430221557615,
      "epoch": 0.11736957376312686,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004999997226065974,
      "loss": 6.3187,
      "mean_token_accuracy": 0.14242062866687774,
      "num_tokens": 2336925.0,
      "step": 1045
    },
    {
      "entropy": 6.4472400665283205,
      "epoch": 0.11793115067108441,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999996559806163,
      "loss": 6.2145,
      "mean_token_accuracy": 0.14515327513217927,
      "num_tokens": 2347431.0,
      "step": 1050
    },
    {
      "entropy": 6.449374628067017,
      "epoch": 0.11849272757904195,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999995821905589,
      "loss": 6.3244,
      "mean_token_accuracy": 0.1428348071873188,
      "num_tokens": 2359894.0,
      "step": 1055
    },
    {
      "entropy": 6.484475946426391,
      "epoch": 0.1190543044869995,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999995012364276,
      "loss": 6.3513,
      "mean_token_accuracy": 0.1399578034877777,
      "num_tokens": 2371181.0,
      "step": 1060
    },
    {
      "entropy": 6.558645248413086,
      "epoch": 0.11961588139495703,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004999994131182248,
      "loss": 6.2539,
      "mean_token_accuracy": 0.14039175882935523,
      "num_tokens": 2381531.0,
      "step": 1065
    },
    {
      "entropy": 6.4977819442749025,
      "epoch": 0.12017745830291458,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999993178359535,
      "loss": 6.3618,
      "mean_token_accuracy": 0.14046622514724733,
      "num_tokens": 2391573.0,
      "step": 1070
    },
    {
      "entropy": 6.531288480758667,
      "epoch": 0.12073903521087213,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999992153896166,
      "loss": 6.3495,
      "mean_token_accuracy": 0.13780387118458748,
      "num_tokens": 2402822.0,
      "step": 1075
    },
    {
      "entropy": 6.468347263336182,
      "epoch": 0.12130061211882967,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004999991057792175,
      "loss": 6.2611,
      "mean_token_accuracy": 0.13793368339538575,
      "num_tokens": 2414322.0,
      "step": 1080
    },
    {
      "entropy": 6.4399701118469235,
      "epoch": 0.12186218902678722,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999989890047596,
      "loss": 6.4342,
      "mean_token_accuracy": 0.14464614540338516,
      "num_tokens": 2426560.0,
      "step": 1085
    },
    {
      "entropy": 6.6257377624511715,
      "epoch": 0.12242376593474476,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004999988650662467,
      "loss": 6.3628,
      "mean_token_accuracy": 0.1392559789121151,
      "num_tokens": 2437723.0,
      "step": 1090
    },
    {
      "entropy": 6.458425664901734,
      "epoch": 0.12298534284270231,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999987339636825,
      "loss": 6.3052,
      "mean_token_accuracy": 0.14131897166371346,
      "num_tokens": 2448497.0,
      "step": 1095
    },
    {
      "entropy": 6.3388848304748535,
      "epoch": 0.12354691975065986,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999985956970713,
      "loss": 6.166,
      "mean_token_accuracy": 0.15202731192111968,
      "num_tokens": 2459089.0,
      "step": 1100
    },
    {
      "entropy": 6.451615047454834,
      "epoch": 0.1241084966586174,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999984502664178,
      "loss": 6.2702,
      "mean_token_accuracy": 0.14179657995700837,
      "num_tokens": 2469882.0,
      "step": 1105
    },
    {
      "entropy": 6.480589246749878,
      "epoch": 0.12467007356657495,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999982976717261,
      "loss": 6.2999,
      "mean_token_accuracy": 0.1432420939207077,
      "num_tokens": 2482120.0,
      "step": 1110
    },
    {
      "entropy": 6.46314582824707,
      "epoch": 0.1252316504745325,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999981379130016,
      "loss": 6.2935,
      "mean_token_accuracy": 0.14398316591978072,
      "num_tokens": 2492626.0,
      "step": 1115
    },
    {
      "entropy": 6.512280225753784,
      "epoch": 0.12579322738249002,
      "grad_norm": 0.890625,
      "learning_rate": 0.000499997970990249,
      "loss": 6.3556,
      "mean_token_accuracy": 0.14139315783977507,
      "num_tokens": 2506347.0,
      "step": 1120
    },
    {
      "entropy": 6.345871210098267,
      "epoch": 0.12635480429044757,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999977969034737,
      "loss": 6.2423,
      "mean_token_accuracy": 0.14657723680138587,
      "num_tokens": 2516584.0,
      "step": 1125
    },
    {
      "entropy": 6.479627084732056,
      "epoch": 0.12691638119840512,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999976156526814,
      "loss": 6.279,
      "mean_token_accuracy": 0.14619525671005248,
      "num_tokens": 2526731.0,
      "step": 1130
    },
    {
      "entropy": 6.3882972240448,
      "epoch": 0.12747795810636267,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999974272378776,
      "loss": 6.1877,
      "mean_token_accuracy": 0.14657506570219994,
      "num_tokens": 2537180.0,
      "step": 1135
    },
    {
      "entropy": 6.418006896972656,
      "epoch": 0.12803953501432022,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999972316590685,
      "loss": 6.2725,
      "mean_token_accuracy": 0.14524131938815116,
      "num_tokens": 2547617.0,
      "step": 1140
    },
    {
      "entropy": 6.410860061645508,
      "epoch": 0.12860111192227774,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999970289162604,
      "loss": 6.1483,
      "mean_token_accuracy": 0.1508580230176449,
      "num_tokens": 2557743.0,
      "step": 1145
    },
    {
      "entropy": 6.349888610839844,
      "epoch": 0.1291626888302353,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999968190094595,
      "loss": 6.2487,
      "mean_token_accuracy": 0.1476031497120857,
      "num_tokens": 2568351.0,
      "step": 1150
    },
    {
      "entropy": 6.430012226104736,
      "epoch": 0.12972426573819285,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999966019386727,
      "loss": 6.3033,
      "mean_token_accuracy": 0.14469023644924164,
      "num_tokens": 2579703.0,
      "step": 1155
    },
    {
      "entropy": 6.4646430015563965,
      "epoch": 0.1302858426461504,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999963777039067,
      "loss": 6.2755,
      "mean_token_accuracy": 0.1433379366993904,
      "num_tokens": 2591197.0,
      "step": 1160
    },
    {
      "entropy": 6.39684362411499,
      "epoch": 0.13084741955410795,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499996146305169,
      "loss": 6.2434,
      "mean_token_accuracy": 0.14241863116621972,
      "num_tokens": 2601836.0,
      "step": 1165
    },
    {
      "entropy": 6.336419486999512,
      "epoch": 0.13140899646206547,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999959077424665,
      "loss": 6.1961,
      "mean_token_accuracy": 0.14609644562005997,
      "num_tokens": 2613513.0,
      "step": 1170
    },
    {
      "entropy": 6.367645883560181,
      "epoch": 0.13197057337002302,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999956620158072,
      "loss": 6.2193,
      "mean_token_accuracy": 0.14458296820521355,
      "num_tokens": 2625246.0,
      "step": 1175
    },
    {
      "entropy": 6.357092189788818,
      "epoch": 0.13253215027798057,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999954091251987,
      "loss": 6.216,
      "mean_token_accuracy": 0.14300859868526458,
      "num_tokens": 2636375.0,
      "step": 1180
    },
    {
      "entropy": 6.38755168914795,
      "epoch": 0.13309372718593812,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999951490706491,
      "loss": 6.3047,
      "mean_token_accuracy": 0.13991767317056655,
      "num_tokens": 2648586.0,
      "step": 1185
    },
    {
      "entropy": 6.442468166351318,
      "epoch": 0.13365530409389567,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999948818521667,
      "loss": 6.2964,
      "mean_token_accuracy": 0.15023437291383743,
      "num_tokens": 2660847.0,
      "step": 1190
    },
    {
      "entropy": 6.39590106010437,
      "epoch": 0.1342168810018532,
      "grad_norm": 1.0625,
      "learning_rate": 0.00049999460746976,
      "loss": 6.2197,
      "mean_token_accuracy": 0.14705628603696824,
      "num_tokens": 2671160.0,
      "step": 1195
    },
    {
      "entropy": 6.4918781280517575,
      "epoch": 0.13477845790981074,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999943259234377,
      "loss": 6.3445,
      "mean_token_accuracy": 0.14184095710515976,
      "num_tokens": 2681622.0,
      "step": 1200
    },
    {
      "entropy": 6.2997137069702145,
      "epoch": 0.1353400348177683,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004999940372132089,
      "loss": 6.1308,
      "mean_token_accuracy": 0.15328069925308227,
      "num_tokens": 2691503.0,
      "step": 1205
    },
    {
      "entropy": 6.358626222610473,
      "epoch": 0.13590161172572585,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999937413390826,
      "loss": 6.2375,
      "mean_token_accuracy": 0.14330296590924263,
      "num_tokens": 2702207.0,
      "step": 1210
    },
    {
      "entropy": 6.419915342330933,
      "epoch": 0.1364631886336834,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999934383010683,
      "loss": 6.2285,
      "mean_token_accuracy": 0.14443148225545882,
      "num_tokens": 2713178.0,
      "step": 1215
    },
    {
      "entropy": 6.428704166412354,
      "epoch": 0.13702476554164092,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999931280991757,
      "loss": 6.2391,
      "mean_token_accuracy": 0.14996800050139428,
      "num_tokens": 2723717.0,
      "step": 1220
    },
    {
      "entropy": 6.407614850997925,
      "epoch": 0.13758634244959847,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999928107334146,
      "loss": 6.3047,
      "mean_token_accuracy": 0.13595292270183562,
      "num_tokens": 2735669.0,
      "step": 1225
    },
    {
      "entropy": 6.339469146728516,
      "epoch": 0.13814791935755602,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999924862037952,
      "loss": 6.2031,
      "mean_token_accuracy": 0.1468040868639946,
      "num_tokens": 2746087.0,
      "step": 1230
    },
    {
      "entropy": 6.324771356582642,
      "epoch": 0.13870949626551357,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999921545103278,
      "loss": 6.2104,
      "mean_token_accuracy": 0.14297769591212273,
      "num_tokens": 2757143.0,
      "step": 1235
    },
    {
      "entropy": 6.488545560836792,
      "epoch": 0.1392710731734711,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999918156530229,
      "loss": 6.3189,
      "mean_token_accuracy": 0.14068616107106208,
      "num_tokens": 2769340.0,
      "step": 1240
    },
    {
      "entropy": 6.36818494796753,
      "epoch": 0.13983265008142864,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999914696318913,
      "loss": 6.2234,
      "mean_token_accuracy": 0.1445622220635414,
      "num_tokens": 2779514.0,
      "step": 1245
    },
    {
      "entropy": 6.3950879096984865,
      "epoch": 0.1403942269893862,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999911164469441,
      "loss": 6.1746,
      "mean_token_accuracy": 0.15131438672542571,
      "num_tokens": 2790801.0,
      "step": 1250
    },
    {
      "entropy": 6.352122497558594,
      "epoch": 0.14095580389734375,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999907560981925,
      "loss": 6.1722,
      "mean_token_accuracy": 0.1425118625164032,
      "num_tokens": 2801082.0,
      "step": 1255
    },
    {
      "entropy": 6.327898740768433,
      "epoch": 0.1415173808053013,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999903885856479,
      "loss": 6.1828,
      "mean_token_accuracy": 0.14815054833889008,
      "num_tokens": 2812719.0,
      "step": 1260
    },
    {
      "entropy": 6.260138082504272,
      "epoch": 0.14207895771325882,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499990013909322,
      "loss": 6.1789,
      "mean_token_accuracy": 0.14185225889086722,
      "num_tokens": 2823766.0,
      "step": 1265
    },
    {
      "entropy": 6.355242538452148,
      "epoch": 0.14264053462121637,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499989632069227,
      "loss": 6.2622,
      "mean_token_accuracy": 0.14508824944496154,
      "num_tokens": 2834801.0,
      "step": 1270
    },
    {
      "entropy": 6.289164733886719,
      "epoch": 0.14320211152917392,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999892430653747,
      "loss": 6.0878,
      "mean_token_accuracy": 0.14808941408991813,
      "num_tokens": 2845769.0,
      "step": 1275
    },
    {
      "entropy": 6.268469953536988,
      "epoch": 0.14376368843713147,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999888468977776,
      "loss": 6.1299,
      "mean_token_accuracy": 0.15243081003427505,
      "num_tokens": 2856001.0,
      "step": 1280
    },
    {
      "entropy": 6.349359560012817,
      "epoch": 0.14432526534508902,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999884435664485,
      "loss": 6.1636,
      "mean_token_accuracy": 0.15014599561691283,
      "num_tokens": 2867703.0,
      "step": 1285
    },
    {
      "entropy": 6.199705266952515,
      "epoch": 0.14488684225304654,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999880330714,
      "loss": 6.1076,
      "mean_token_accuracy": 0.14883481562137604,
      "num_tokens": 2878671.0,
      "step": 1290
    },
    {
      "entropy": 6.326551866531372,
      "epoch": 0.1454484191610041,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999876154126453,
      "loss": 6.2189,
      "mean_token_accuracy": 0.14116846397519112,
      "num_tokens": 2890137.0,
      "step": 1295
    },
    {
      "entropy": 6.333295202255249,
      "epoch": 0.14600999606896164,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999871905901977,
      "loss": 6.1345,
      "mean_token_accuracy": 0.14464831352233887,
      "num_tokens": 2900344.0,
      "step": 1300
    },
    {
      "entropy": 6.1990049362182615,
      "epoch": 0.1465715729769192,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999867586040706,
      "loss": 6.1449,
      "mean_token_accuracy": 0.14359352886676788,
      "num_tokens": 2912270.0,
      "step": 1305
    },
    {
      "entropy": 6.442420291900635,
      "epoch": 0.14713314988487675,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999863194542779,
      "loss": 6.2245,
      "mean_token_accuracy": 0.13818560764193535,
      "num_tokens": 2924677.0,
      "step": 1310
    },
    {
      "entropy": 6.208629751205445,
      "epoch": 0.14769472679283427,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999858731408336,
      "loss": 6.1435,
      "mean_token_accuracy": 0.15086931586265565,
      "num_tokens": 2935794.0,
      "step": 1315
    },
    {
      "entropy": 6.351964473724365,
      "epoch": 0.14825630370079182,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999854196637518,
      "loss": 6.2405,
      "mean_token_accuracy": 0.1423012413084507,
      "num_tokens": 2945958.0,
      "step": 1320
    },
    {
      "entropy": 6.388653326034546,
      "epoch": 0.14881788060874937,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999849590230469,
      "loss": 6.2022,
      "mean_token_accuracy": 0.1418309211730957,
      "num_tokens": 2957536.0,
      "step": 1325
    },
    {
      "entropy": 6.350034427642822,
      "epoch": 0.14937945751670692,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999844912187336,
      "loss": 6.1859,
      "mean_token_accuracy": 0.14834656044840813,
      "num_tokens": 2967787.0,
      "step": 1330
    },
    {
      "entropy": 6.309407663345337,
      "epoch": 0.14994103442466447,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499984016250827,
      "loss": 6.2062,
      "mean_token_accuracy": 0.1401480630040169,
      "num_tokens": 2980522.0,
      "step": 1335
    },
    {
      "entropy": 6.190045595169067,
      "epoch": 0.150502611332622,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499983534119342,
      "loss": 6.1524,
      "mean_token_accuracy": 0.14754045605659485,
      "num_tokens": 2990497.0,
      "step": 1340
    },
    {
      "entropy": 6.4408313751220705,
      "epoch": 0.15106418824057954,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004999830448242941,
      "loss": 6.2346,
      "mean_token_accuracy": 0.1441300056874752,
      "num_tokens": 3002931.0,
      "step": 1345
    },
    {
      "entropy": 6.302226638793945,
      "epoch": 0.1516257651485371,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999825483656986,
      "loss": 6.1602,
      "mean_token_accuracy": 0.15221837013959885,
      "num_tokens": 3014137.0,
      "step": 1350
    },
    {
      "entropy": 6.390075826644898,
      "epoch": 0.15218734205649465,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999820447435716,
      "loss": 6.3144,
      "mean_token_accuracy": 0.1362255647778511,
      "num_tokens": 3026145.0,
      "step": 1355
    },
    {
      "entropy": 6.2819986820220945,
      "epoch": 0.15274891896445217,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999815339579291,
      "loss": 6.166,
      "mean_token_accuracy": 0.14926141500473022,
      "num_tokens": 3037068.0,
      "step": 1360
    },
    {
      "entropy": 6.3687601566314695,
      "epoch": 0.15331049587240972,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999810160087872,
      "loss": 6.1057,
      "mean_token_accuracy": 0.14911322370171548,
      "num_tokens": 3047836.0,
      "step": 1365
    },
    {
      "entropy": 6.260298824310302,
      "epoch": 0.15387207278036727,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999804908961626,
      "loss": 6.228,
      "mean_token_accuracy": 0.14651088342070578,
      "num_tokens": 3060309.0,
      "step": 1370
    },
    {
      "entropy": 6.318568372726441,
      "epoch": 0.15443364968832482,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999799586200719,
      "loss": 6.1992,
      "mean_token_accuracy": 0.14661515429615973,
      "num_tokens": 3071789.0,
      "step": 1375
    },
    {
      "entropy": 6.3008849143981935,
      "epoch": 0.15499522659628237,
      "grad_norm": 1.078125,
      "learning_rate": 0.000499979419180532,
      "loss": 6.1147,
      "mean_token_accuracy": 0.1510265827178955,
      "num_tokens": 3081756.0,
      "step": 1380
    },
    {
      "entropy": 6.315201616287231,
      "epoch": 0.1555568035042399,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999788725775601,
      "loss": 6.2108,
      "mean_token_accuracy": 0.14352906197309495,
      "num_tokens": 3093456.0,
      "step": 1385
    },
    {
      "entropy": 6.364352989196777,
      "epoch": 0.15611838041219744,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999783188111737,
      "loss": 6.1434,
      "mean_token_accuracy": 0.15190799832344054,
      "num_tokens": 3104565.0,
      "step": 1390
    },
    {
      "entropy": 6.244580602645874,
      "epoch": 0.156679957320155,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999777578813904,
      "loss": 6.1651,
      "mean_token_accuracy": 0.14751930832862853,
      "num_tokens": 3115577.0,
      "step": 1395
    },
    {
      "entropy": 6.268232011795044,
      "epoch": 0.15724153422811254,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499977189788228,
      "loss": 6.0583,
      "mean_token_accuracy": 0.15186556801199913,
      "num_tokens": 3126019.0,
      "step": 1400
    },
    {
      "entropy": 6.308783960342407,
      "epoch": 0.1578031111360701,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999766145317046,
      "loss": 6.2291,
      "mean_token_accuracy": 0.13733314722776413,
      "num_tokens": 3136191.0,
      "step": 1405
    },
    {
      "entropy": 6.364164161682129,
      "epoch": 0.15836468804402762,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999760321118385,
      "loss": 6.0879,
      "mean_token_accuracy": 0.1505395717918873,
      "num_tokens": 3147737.0,
      "step": 1410
    },
    {
      "entropy": 6.208480882644653,
      "epoch": 0.15892626495198517,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999754425286483,
      "loss": 6.1558,
      "mean_token_accuracy": 0.14820881485939025,
      "num_tokens": 3158494.0,
      "step": 1415
    },
    {
      "entropy": 6.3466572761535645,
      "epoch": 0.15948784185994272,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999748457821529,
      "loss": 6.183,
      "mean_token_accuracy": 0.1488707736134529,
      "num_tokens": 3170550.0,
      "step": 1420
    },
    {
      "entropy": 6.248150491714478,
      "epoch": 0.16004941876790027,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000499974241872371,
      "loss": 6.1577,
      "mean_token_accuracy": 0.15534067451953887,
      "num_tokens": 3182860.0,
      "step": 1425
    },
    {
      "entropy": 6.263169813156128,
      "epoch": 0.16061099567585782,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999736307993221,
      "loss": 6.1687,
      "mean_token_accuracy": 0.1433172792196274,
      "num_tokens": 3193807.0,
      "step": 1430
    },
    {
      "entropy": 6.20141077041626,
      "epoch": 0.16117257258381534,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999730125630255,
      "loss": 6.0323,
      "mean_token_accuracy": 0.14806009754538535,
      "num_tokens": 3203959.0,
      "step": 1435
    },
    {
      "entropy": 6.278918790817261,
      "epoch": 0.1617341494917729,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999723871635008,
      "loss": 6.0481,
      "mean_token_accuracy": 0.15965506583452224,
      "num_tokens": 3215416.0,
      "step": 1440
    },
    {
      "entropy": 6.265646123886109,
      "epoch": 0.16229572639973044,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999717546007681,
      "loss": 6.2083,
      "mean_token_accuracy": 0.14182574525475503,
      "num_tokens": 3226938.0,
      "step": 1445
    },
    {
      "entropy": 6.198927593231201,
      "epoch": 0.162857303307688,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999711148748476,
      "loss": 6.0113,
      "mean_token_accuracy": 0.15725158154964447,
      "num_tokens": 3238406.0,
      "step": 1450
    },
    {
      "entropy": 6.139306259155274,
      "epoch": 0.16341888021564555,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999704679857594,
      "loss": 6.0335,
      "mean_token_accuracy": 0.15332483649253845,
      "num_tokens": 3249695.0,
      "step": 1455
    },
    {
      "entropy": 6.430513477325439,
      "epoch": 0.16398045712360307,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999698139335243,
      "loss": 6.1709,
      "mean_token_accuracy": 0.15339794158935546,
      "num_tokens": 3259656.0,
      "step": 1460
    },
    {
      "entropy": 6.138534164428711,
      "epoch": 0.16454203403156062,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999691527181632,
      "loss": 6.0103,
      "mean_token_accuracy": 0.15821679681539536,
      "num_tokens": 3269712.0,
      "step": 1465
    },
    {
      "entropy": 6.10054726600647,
      "epoch": 0.16510361093951817,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999684843396968,
      "loss": 6.0043,
      "mean_token_accuracy": 0.1602013364434242,
      "num_tokens": 3280343.0,
      "step": 1470
    },
    {
      "entropy": 6.187516689300537,
      "epoch": 0.16566518784747572,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999678087981467,
      "loss": 6.0398,
      "mean_token_accuracy": 0.1516653686761856,
      "num_tokens": 3291649.0,
      "step": 1475
    },
    {
      "entropy": 6.330076074600219,
      "epoch": 0.16622676475543324,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999671260935343,
      "loss": 6.2204,
      "mean_token_accuracy": 0.15116648524999618,
      "num_tokens": 3304190.0,
      "step": 1480
    },
    {
      "entropy": 6.137814521789551,
      "epoch": 0.1667883416633908,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999664362258814,
      "loss": 5.945,
      "mean_token_accuracy": 0.15754275172948837,
      "num_tokens": 3313966.0,
      "step": 1485
    },
    {
      "entropy": 6.239295911788941,
      "epoch": 0.16734991857134834,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999657391952099,
      "loss": 6.1924,
      "mean_token_accuracy": 0.14727808460593222,
      "num_tokens": 3326017.0,
      "step": 1490
    },
    {
      "entropy": 6.257507610321045,
      "epoch": 0.1679114954793059,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000499965035001542,
      "loss": 6.0906,
      "mean_token_accuracy": 0.14670320898294448,
      "num_tokens": 3338264.0,
      "step": 1495
    },
    {
      "entropy": 6.237432384490967,
      "epoch": 0.16847307238726345,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999643236449001,
      "loss": 6.1368,
      "mean_token_accuracy": 0.1596009373664856,
      "num_tokens": 3348901.0,
      "step": 1500
    },
    {
      "entropy": 6.235905361175537,
      "epoch": 0.16903464929522097,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999636051253069,
      "loss": 6.0665,
      "mean_token_accuracy": 0.15173731446266175,
      "num_tokens": 3358441.0,
      "step": 1505
    },
    {
      "entropy": 6.294967842102051,
      "epoch": 0.16959622620317852,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999628794427854,
      "loss": 6.1443,
      "mean_token_accuracy": 0.15502254366874696,
      "num_tokens": 3369334.0,
      "step": 1510
    },
    {
      "entropy": 6.230111169815063,
      "epoch": 0.17015780311113607,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999621465973583,
      "loss": 6.1516,
      "mean_token_accuracy": 0.1490892805159092,
      "num_tokens": 3380698.0,
      "step": 1515
    },
    {
      "entropy": 6.401267576217651,
      "epoch": 0.17071938001909362,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999614065890495,
      "loss": 6.1586,
      "mean_token_accuracy": 0.14243442714214324,
      "num_tokens": 3391061.0,
      "step": 1520
    },
    {
      "entropy": 6.223486804962159,
      "epoch": 0.17128095692705117,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004999606594178821,
      "loss": 6.2143,
      "mean_token_accuracy": 0.14877415746450423,
      "num_tokens": 3403983.0,
      "step": 1525
    },
    {
      "entropy": 6.3232080936431885,
      "epoch": 0.1718425338350087,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00049995990508388,
      "loss": 6.1398,
      "mean_token_accuracy": 0.1495580181479454,
      "num_tokens": 3413838.0,
      "step": 1530
    },
    {
      "entropy": 6.128076457977295,
      "epoch": 0.17240411074296624,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999591435870674,
      "loss": 5.9598,
      "mean_token_accuracy": 0.15457928329706191,
      "num_tokens": 3423978.0,
      "step": 1535
    },
    {
      "entropy": 6.287269401550293,
      "epoch": 0.1729656876509238,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999583749274684,
      "loss": 6.1415,
      "mean_token_accuracy": 0.15106903314590453,
      "num_tokens": 3433416.0,
      "step": 1540
    },
    {
      "entropy": 6.143144655227661,
      "epoch": 0.17352726455888134,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999575991051074,
      "loss": 5.9773,
      "mean_token_accuracy": 0.15837069898843764,
      "num_tokens": 3444983.0,
      "step": 1545
    },
    {
      "entropy": 6.186607074737549,
      "epoch": 0.1740888414668389,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999568161200093,
      "loss": 5.9651,
      "mean_token_accuracy": 0.15607634782791138,
      "num_tokens": 3455808.0,
      "step": 1550
    },
    {
      "entropy": 6.141094207763672,
      "epoch": 0.17465041837479642,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999560259721989,
      "loss": 6.1172,
      "mean_token_accuracy": 0.1467524640262127,
      "num_tokens": 3467503.0,
      "step": 1555
    },
    {
      "entropy": 6.361791515350342,
      "epoch": 0.17521199528275397,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999552286617014,
      "loss": 6.3016,
      "mean_token_accuracy": 0.14429757595062256,
      "num_tokens": 3479941.0,
      "step": 1560
    },
    {
      "entropy": 6.129152393341064,
      "epoch": 0.17577357219071152,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999544241885421,
      "loss": 5.9961,
      "mean_token_accuracy": 0.158864426612854,
      "num_tokens": 3491069.0,
      "step": 1565
    },
    {
      "entropy": 6.282902526855469,
      "epoch": 0.17633514909866907,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999536125527467,
      "loss": 6.1721,
      "mean_token_accuracy": 0.14678503796458245,
      "num_tokens": 3502065.0,
      "step": 1570
    },
    {
      "entropy": 6.311311912536621,
      "epoch": 0.17689672600662662,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499952793754341,
      "loss": 6.122,
      "mean_token_accuracy": 0.1542977884411812,
      "num_tokens": 3514228.0,
      "step": 1575
    },
    {
      "entropy": 6.200154161453247,
      "epoch": 0.17745830291458414,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999519677933512,
      "loss": 6.0557,
      "mean_token_accuracy": 0.15597216486930848,
      "num_tokens": 3525801.0,
      "step": 1580
    },
    {
      "entropy": 6.209293603897095,
      "epoch": 0.1780198798225417,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999511346698035,
      "loss": 6.0008,
      "mean_token_accuracy": 0.15367855429649352,
      "num_tokens": 3536431.0,
      "step": 1585
    },
    {
      "entropy": 6.126080894470215,
      "epoch": 0.17858145673049924,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999502943837244,
      "loss": 6.1057,
      "mean_token_accuracy": 0.14980130195617675,
      "num_tokens": 3549098.0,
      "step": 1590
    },
    {
      "entropy": 6.306281661987304,
      "epoch": 0.1791430336384568,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999494469351406,
      "loss": 6.0725,
      "mean_token_accuracy": 0.1504937268793583,
      "num_tokens": 3560532.0,
      "step": 1595
    },
    {
      "entropy": 6.089444303512574,
      "epoch": 0.17970461054641435,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999485923240793,
      "loss": 6.0566,
      "mean_token_accuracy": 0.15513607040047644,
      "num_tokens": 3572175.0,
      "step": 1600
    },
    {
      "entropy": 6.255811595916748,
      "epoch": 0.18026618745437187,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999477305505675,
      "loss": 5.9836,
      "mean_token_accuracy": 0.17298887223005294,
      "num_tokens": 3582862.0,
      "step": 1605
    },
    {
      "entropy": 6.210167360305786,
      "epoch": 0.18082776436232942,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999468616146328,
      "loss": 6.1756,
      "mean_token_accuracy": 0.1523563966155052,
      "num_tokens": 3594685.0,
      "step": 1610
    },
    {
      "entropy": 6.104140567779541,
      "epoch": 0.18138934127028697,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999459855163026,
      "loss": 5.9615,
      "mean_token_accuracy": 0.16202931553125383,
      "num_tokens": 3604847.0,
      "step": 1615
    },
    {
      "entropy": 6.271286487579346,
      "epoch": 0.18195091817824452,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499945102255605,
      "loss": 6.0582,
      "mean_token_accuracy": 0.15297539979219438,
      "num_tokens": 3615549.0,
      "step": 1620
    },
    {
      "entropy": 6.103758001327515,
      "epoch": 0.18251249508620204,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999442118325682,
      "loss": 5.9676,
      "mean_token_accuracy": 0.1630454882979393,
      "num_tokens": 3626300.0,
      "step": 1625
    },
    {
      "entropy": 6.127223777770996,
      "epoch": 0.1830740719941596,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004999433142472205,
      "loss": 5.9958,
      "mean_token_accuracy": 0.15685569494962692,
      "num_tokens": 3637832.0,
      "step": 1630
    },
    {
      "entropy": 6.177744150161743,
      "epoch": 0.18363564890211714,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999424094995902,
      "loss": 6.0401,
      "mean_token_accuracy": 0.14781517460942267,
      "num_tokens": 3649179.0,
      "step": 1635
    },
    {
      "entropy": 6.183869218826294,
      "epoch": 0.1841972258100747,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999414975897066,
      "loss": 6.072,
      "mean_token_accuracy": 0.14898845106363295,
      "num_tokens": 3660011.0,
      "step": 1640
    },
    {
      "entropy": 6.244912004470825,
      "epoch": 0.18475880271803224,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999405785175983,
      "loss": 6.1505,
      "mean_token_accuracy": 0.14797591269016266,
      "num_tokens": 3672791.0,
      "step": 1645
    },
    {
      "entropy": 6.1082813262939455,
      "epoch": 0.18532037962598977,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999396522832949,
      "loss": 5.971,
      "mean_token_accuracy": 0.15111488699913025,
      "num_tokens": 3684709.0,
      "step": 1650
    },
    {
      "entropy": 6.10607328414917,
      "epoch": 0.18588195653394732,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999387188868256,
      "loss": 5.9402,
      "mean_token_accuracy": 0.1563703805208206,
      "num_tokens": 3694675.0,
      "step": 1655
    },
    {
      "entropy": 6.061730670928955,
      "epoch": 0.18644353344190487,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999377783282203,
      "loss": 5.9483,
      "mean_token_accuracy": 0.15275903791189194,
      "num_tokens": 3706384.0,
      "step": 1660
    },
    {
      "entropy": 6.269464302062988,
      "epoch": 0.18700511034986242,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999368306075089,
      "loss": 6.0202,
      "mean_token_accuracy": 0.15324059426784514,
      "num_tokens": 3717419.0,
      "step": 1665
    },
    {
      "entropy": 6.155914354324341,
      "epoch": 0.18756668725781997,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999358757247215,
      "loss": 6.0503,
      "mean_token_accuracy": 0.1521863892674446,
      "num_tokens": 3727306.0,
      "step": 1670
    },
    {
      "entropy": 6.2568072319030765,
      "epoch": 0.1881282641657775,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999349136798886,
      "loss": 6.1685,
      "mean_token_accuracy": 0.14885527342557908,
      "num_tokens": 3739476.0,
      "step": 1675
    },
    {
      "entropy": 6.100805997848511,
      "epoch": 0.18868984107373504,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999339444730409,
      "loss": 5.933,
      "mean_token_accuracy": 0.15855167657136918,
      "num_tokens": 3751035.0,
      "step": 1680
    },
    {
      "entropy": 6.170141315460205,
      "epoch": 0.1892514179816926,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999329681042092,
      "loss": 6.0249,
      "mean_token_accuracy": 0.1590510666370392,
      "num_tokens": 3762881.0,
      "step": 1685
    },
    {
      "entropy": 6.138400840759277,
      "epoch": 0.18981299488965014,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999319845734243,
      "loss": 6.044,
      "mean_token_accuracy": 0.15163577049970628,
      "num_tokens": 3774180.0,
      "step": 1690
    },
    {
      "entropy": 6.174570465087891,
      "epoch": 0.1903745717976077,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999309938807181,
      "loss": 6.0303,
      "mean_token_accuracy": 0.1562027633190155,
      "num_tokens": 3784864.0,
      "step": 1695
    },
    {
      "entropy": 6.1436539649963375,
      "epoch": 0.19093614870556522,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999299960261216,
      "loss": 6.0096,
      "mean_token_accuracy": 0.15322427898645402,
      "num_tokens": 3797693.0,
      "step": 1700
    },
    {
      "entropy": 6.201740980148315,
      "epoch": 0.19149772561352277,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999289910096669,
      "loss": 6.0241,
      "mean_token_accuracy": 0.15492813885211945,
      "num_tokens": 3809379.0,
      "step": 1705
    },
    {
      "entropy": 6.070269393920898,
      "epoch": 0.19205930252148032,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999279788313858,
      "loss": 5.8698,
      "mean_token_accuracy": 0.16566905826330186,
      "num_tokens": 3819473.0,
      "step": 1710
    },
    {
      "entropy": 5.977046871185303,
      "epoch": 0.19262087942943787,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999269594913107,
      "loss": 5.8641,
      "mean_token_accuracy": 0.1651919439435005,
      "num_tokens": 3831037.0,
      "step": 1715
    },
    {
      "entropy": 6.101950931549072,
      "epoch": 0.19318245633739542,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999259329894739,
      "loss": 6.0162,
      "mean_token_accuracy": 0.1505047082901001,
      "num_tokens": 3841767.0,
      "step": 1720
    },
    {
      "entropy": 6.011256885528565,
      "epoch": 0.19374403324535294,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999248993259081,
      "loss": 5.8642,
      "mean_token_accuracy": 0.16040490120649337,
      "num_tokens": 3851629.0,
      "step": 1725
    },
    {
      "entropy": 6.124179744720459,
      "epoch": 0.1943056101533105,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999238585006463,
      "loss": 5.9383,
      "mean_token_accuracy": 0.15869803428649903,
      "num_tokens": 3863247.0,
      "step": 1730
    },
    {
      "entropy": 6.08483362197876,
      "epoch": 0.19486718706126804,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999228105137218,
      "loss": 5.9182,
      "mean_token_accuracy": 0.15641646683216096,
      "num_tokens": 3873949.0,
      "step": 1735
    },
    {
      "entropy": 6.134666872024536,
      "epoch": 0.1954287639692256,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999217553651676,
      "loss": 6.0202,
      "mean_token_accuracy": 0.1586867719888687,
      "num_tokens": 3884992.0,
      "step": 1740
    },
    {
      "entropy": 6.211376428604126,
      "epoch": 0.19599034087718312,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999206930550175,
      "loss": 6.1354,
      "mean_token_accuracy": 0.15906748920679092,
      "num_tokens": 3897029.0,
      "step": 1745
    },
    {
      "entropy": 6.0789470195770265,
      "epoch": 0.19655191778514067,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999196235833054,
      "loss": 5.9818,
      "mean_token_accuracy": 0.15585362017154694,
      "num_tokens": 3909510.0,
      "step": 1750
    },
    {
      "entropy": 6.10754165649414,
      "epoch": 0.19711349469309822,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999185469500651,
      "loss": 5.8919,
      "mean_token_accuracy": 0.1578446224331856,
      "num_tokens": 3920593.0,
      "step": 1755
    },
    {
      "entropy": 6.169074296951294,
      "epoch": 0.19767507160105577,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499917463155331,
      "loss": 6.0256,
      "mean_token_accuracy": 0.1516920581459999,
      "num_tokens": 3932215.0,
      "step": 1760
    },
    {
      "entropy": 6.106430435180664,
      "epoch": 0.19823664850901332,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999163721991378,
      "loss": 5.9855,
      "mean_token_accuracy": 0.1580495446920395,
      "num_tokens": 3943876.0,
      "step": 1765
    },
    {
      "entropy": 6.031008052825928,
      "epoch": 0.19879822541697084,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00049991527408152,
      "loss": 5.9635,
      "mean_token_accuracy": 0.15963634550571443,
      "num_tokens": 3954533.0,
      "step": 1770
    },
    {
      "entropy": 6.156386280059815,
      "epoch": 0.1993598023249284,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999141688025127,
      "loss": 6.0307,
      "mean_token_accuracy": 0.16017568558454515,
      "num_tokens": 3965722.0,
      "step": 1775
    },
    {
      "entropy": 6.099786853790283,
      "epoch": 0.19992137923288594,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499913056362151,
      "loss": 5.9427,
      "mean_token_accuracy": 0.15914792269468309,
      "num_tokens": 3976737.0,
      "step": 1780
    },
    {
      "entropy": 6.102410745620728,
      "epoch": 0.2004829561408435,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999119367604703,
      "loss": 6.0445,
      "mean_token_accuracy": 0.1580652192234993,
      "num_tokens": 3986102.0,
      "step": 1785
    },
    {
      "entropy": 6.172822618484497,
      "epoch": 0.20104453304880104,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999108099975064,
      "loss": 5.995,
      "mean_token_accuracy": 0.1587173655629158,
      "num_tokens": 3996534.0,
      "step": 1790
    },
    {
      "entropy": 6.050444316864014,
      "epoch": 0.20160610995675857,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499909676073295,
      "loss": 5.9112,
      "mean_token_accuracy": 0.15411415696144104,
      "num_tokens": 4006947.0,
      "step": 1795
    },
    {
      "entropy": 6.077579498291016,
      "epoch": 0.20216768686471612,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999085349878723,
      "loss": 5.9524,
      "mean_token_accuracy": 0.16510123014450073,
      "num_tokens": 4017723.0,
      "step": 1800
    },
    {
      "entropy": 6.076634263992309,
      "epoch": 0.20272926377267367,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999073867412746,
      "loss": 6.0289,
      "mean_token_accuracy": 0.15287718251347543,
      "num_tokens": 4028347.0,
      "step": 1805
    },
    {
      "entropy": 6.067677736282349,
      "epoch": 0.20329084068063122,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999062313335385,
      "loss": 5.8889,
      "mean_token_accuracy": 0.1635470509529114,
      "num_tokens": 4039648.0,
      "step": 1810
    },
    {
      "entropy": 6.1201819896698,
      "epoch": 0.20385241758858877,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999050687647008,
      "loss": 5.9439,
      "mean_token_accuracy": 0.15459056794643403,
      "num_tokens": 4050786.0,
      "step": 1815
    },
    {
      "entropy": 6.141393518447876,
      "epoch": 0.2044139944965463,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999038990347984,
      "loss": 5.9858,
      "mean_token_accuracy": 0.16119525581598282,
      "num_tokens": 4061184.0,
      "step": 1820
    },
    {
      "entropy": 6.207761764526367,
      "epoch": 0.20497557140450384,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999027221438688,
      "loss": 6.1171,
      "mean_token_accuracy": 0.15378198772668839,
      "num_tokens": 4073607.0,
      "step": 1825
    },
    {
      "entropy": 6.042503595352173,
      "epoch": 0.2055371483124614,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499901538091949,
      "loss": 5.9276,
      "mean_token_accuracy": 0.15577400028705596,
      "num_tokens": 4083924.0,
      "step": 1830
    },
    {
      "entropy": 6.11335186958313,
      "epoch": 0.20609872522041894,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999003468790773,
      "loss": 5.9258,
      "mean_token_accuracy": 0.15859550833702088,
      "num_tokens": 4094083.0,
      "step": 1835
    },
    {
      "entropy": 6.091872358322144,
      "epoch": 0.2066603021283765,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998991485052911,
      "loss": 5.961,
      "mean_token_accuracy": 0.1666373997926712,
      "num_tokens": 4105124.0,
      "step": 1840
    },
    {
      "entropy": 6.17498631477356,
      "epoch": 0.20722187903633402,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499897942970629,
      "loss": 6.1215,
      "mean_token_accuracy": 0.1484576180577278,
      "num_tokens": 4116792.0,
      "step": 1845
    },
    {
      "entropy": 6.292929363250733,
      "epoch": 0.20778345594429157,
      "grad_norm": 0.875,
      "learning_rate": 0.000499896730275129,
      "loss": 6.1197,
      "mean_token_accuracy": 0.14851790070533752,
      "num_tokens": 4127441.0,
      "step": 1850
    },
    {
      "entropy": 6.086349821090698,
      "epoch": 0.20834503285224912,
      "grad_norm": 0.890625,
      "learning_rate": 0.00049989551041883,
      "loss": 6.0809,
      "mean_token_accuracy": 0.15333859026432037,
      "num_tokens": 4138696.0,
      "step": 1855
    },
    {
      "entropy": 6.135968160629273,
      "epoch": 0.20890660976020667,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998942834017707,
      "loss": 5.9533,
      "mean_token_accuracy": 0.15621326714754105,
      "num_tokens": 4151254.0,
      "step": 1860
    },
    {
      "entropy": 5.997674369812012,
      "epoch": 0.2094681866681642,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998930492239901,
      "loss": 5.9177,
      "mean_token_accuracy": 0.15729122012853622,
      "num_tokens": 4162046.0,
      "step": 1865
    },
    {
      "entropy": 6.080434846878052,
      "epoch": 0.21002976357612174,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998918078855277,
      "loss": 5.8692,
      "mean_token_accuracy": 0.16310780197381974,
      "num_tokens": 4172409.0,
      "step": 1870
    },
    {
      "entropy": 6.069913625717163,
      "epoch": 0.2105913404840793,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004998905593864227,
      "loss": 6.0241,
      "mean_token_accuracy": 0.16471444964408874,
      "num_tokens": 4182949.0,
      "step": 1875
    },
    {
      "entropy": 6.122161340713501,
      "epoch": 0.21115291739203684,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998893037267153,
      "loss": 6.0011,
      "mean_token_accuracy": 0.15665073543787003,
      "num_tokens": 4194164.0,
      "step": 1880
    },
    {
      "entropy": 6.155545330047607,
      "epoch": 0.2117144942999944,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998880409064452,
      "loss": 5.9416,
      "mean_token_accuracy": 0.1604766234755516,
      "num_tokens": 4206056.0,
      "step": 1885
    },
    {
      "entropy": 5.998275089263916,
      "epoch": 0.21227607120795192,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998867709256527,
      "loss": 6.0195,
      "mean_token_accuracy": 0.1543448567390442,
      "num_tokens": 4216660.0,
      "step": 1890
    },
    {
      "entropy": 6.106033754348755,
      "epoch": 0.21283764811590947,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000499885493784378,
      "loss": 5.8928,
      "mean_token_accuracy": 0.1672811269760132,
      "num_tokens": 4228495.0,
      "step": 1895
    },
    {
      "entropy": 6.029328441619873,
      "epoch": 0.21339922502386702,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004998842094826621,
      "loss": 5.9064,
      "mean_token_accuracy": 0.15779903531074524,
      "num_tokens": 4238560.0,
      "step": 1900
    },
    {
      "entropy": 6.210692977905273,
      "epoch": 0.21396080193182457,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998829180205458,
      "loss": 6.1026,
      "mean_token_accuracy": 0.15292961820960044,
      "num_tokens": 4250553.0,
      "step": 1905
    },
    {
      "entropy": 6.1289863109588625,
      "epoch": 0.21452237883978212,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998816193980701,
      "loss": 6.0255,
      "mean_token_accuracy": 0.15285986065864562,
      "num_tokens": 4262050.0,
      "step": 1910
    },
    {
      "entropy": 6.189714097976685,
      "epoch": 0.21508395574773964,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998803136152764,
      "loss": 5.9742,
      "mean_token_accuracy": 0.15248434990644455,
      "num_tokens": 4273019.0,
      "step": 1915
    },
    {
      "entropy": 6.069318389892578,
      "epoch": 0.2156455326556972,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998790006722063,
      "loss": 5.8657,
      "mean_token_accuracy": 0.16439094692468642,
      "num_tokens": 4283585.0,
      "step": 1920
    },
    {
      "entropy": 6.011609268188477,
      "epoch": 0.21620710956365474,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998776805689016,
      "loss": 5.9419,
      "mean_token_accuracy": 0.16155797690153123,
      "num_tokens": 4294905.0,
      "step": 1925
    },
    {
      "entropy": 6.21407265663147,
      "epoch": 0.2167686864716123,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998763533054045,
      "loss": 6.0508,
      "mean_token_accuracy": 0.1579398825764656,
      "num_tokens": 4305926.0,
      "step": 1930
    },
    {
      "entropy": 5.9719137191772464,
      "epoch": 0.21733026337956984,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998750188817568,
      "loss": 6.024,
      "mean_token_accuracy": 0.1534085601568222,
      "num_tokens": 4318297.0,
      "step": 1935
    },
    {
      "entropy": 6.118768739700317,
      "epoch": 0.21789184028752737,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998736772980015,
      "loss": 5.8435,
      "mean_token_accuracy": 0.1607772260904312,
      "num_tokens": 4328795.0,
      "step": 1940
    },
    {
      "entropy": 5.974378633499145,
      "epoch": 0.21845341719548492,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499872328554181,
      "loss": 5.8598,
      "mean_token_accuracy": 0.16153356730937957,
      "num_tokens": 4340279.0,
      "step": 1945
    },
    {
      "entropy": 6.151049709320068,
      "epoch": 0.21901499410344247,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998709726503383,
      "loss": 6.0365,
      "mean_token_accuracy": 0.15092452690005304,
      "num_tokens": 4352157.0,
      "step": 1950
    },
    {
      "entropy": 6.07116470336914,
      "epoch": 0.21957657101140002,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998696095865168,
      "loss": 5.8785,
      "mean_token_accuracy": 0.16288214921951294,
      "num_tokens": 4361909.0,
      "step": 1955
    },
    {
      "entropy": 5.853420877456665,
      "epoch": 0.22013814791935757,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998682393627595,
      "loss": 5.7873,
      "mean_token_accuracy": 0.1609992802143097,
      "num_tokens": 4372177.0,
      "step": 1960
    },
    {
      "entropy": 6.081985902786255,
      "epoch": 0.2206997248273151,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998668619791104,
      "loss": 5.8496,
      "mean_token_accuracy": 0.16345320641994476,
      "num_tokens": 4382872.0,
      "step": 1965
    },
    {
      "entropy": 6.015959978103638,
      "epoch": 0.22126130173527264,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998654774356131,
      "loss": 5.942,
      "mean_token_accuracy": 0.15784113109111786,
      "num_tokens": 4393623.0,
      "step": 1970
    },
    {
      "entropy": 6.030070972442627,
      "epoch": 0.2218228786432302,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004998640857323118,
      "loss": 5.8703,
      "mean_token_accuracy": 0.15540042966604234,
      "num_tokens": 4404215.0,
      "step": 1975
    },
    {
      "entropy": 6.12239670753479,
      "epoch": 0.22238445555118774,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998626868692509,
      "loss": 6.1297,
      "mean_token_accuracy": 0.15531746223568915,
      "num_tokens": 4415978.0,
      "step": 1980
    },
    {
      "entropy": 6.103342962265015,
      "epoch": 0.2229460324591453,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998612808464748,
      "loss": 5.8857,
      "mean_token_accuracy": 0.1632543236017227,
      "num_tokens": 4427060.0,
      "step": 1985
    },
    {
      "entropy": 6.013435125350952,
      "epoch": 0.22350760936710282,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998598676640283,
      "loss": 5.9535,
      "mean_token_accuracy": 0.15278794541954993,
      "num_tokens": 4438145.0,
      "step": 1990
    },
    {
      "entropy": 6.176746988296509,
      "epoch": 0.22406918627506037,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998584473219563,
      "loss": 6.0253,
      "mean_token_accuracy": 0.16004208326339722,
      "num_tokens": 4449324.0,
      "step": 1995
    },
    {
      "entropy": 5.992343759536743,
      "epoch": 0.22463076318301792,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998570198203043,
      "loss": 5.7561,
      "mean_token_accuracy": 0.1650302827358246,
      "num_tokens": 4459728.0,
      "step": 2000
    },
    {
      "entropy": 6.014521694183349,
      "epoch": 0.22519234009097547,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998555851591175,
      "loss": 5.9692,
      "mean_token_accuracy": 0.161544968187809,
      "num_tokens": 4471164.0,
      "step": 2005
    },
    {
      "entropy": 6.107461452484131,
      "epoch": 0.225753916998933,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004998541433384416,
      "loss": 6.0283,
      "mean_token_accuracy": 0.16004153192043305,
      "num_tokens": 4483349.0,
      "step": 2010
    },
    {
      "entropy": 6.108623361587524,
      "epoch": 0.22631549390689054,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998526943583226,
      "loss": 5.9901,
      "mean_token_accuracy": 0.15583523511886596,
      "num_tokens": 4495311.0,
      "step": 2015
    },
    {
      "entropy": 6.01443190574646,
      "epoch": 0.2268770708148481,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004998512382188066,
      "loss": 5.9136,
      "mean_token_accuracy": 0.16296836510300636,
      "num_tokens": 4507017.0,
      "step": 2020
    },
    {
      "entropy": 6.0663471698760985,
      "epoch": 0.22743864772280564,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004998497749199401,
      "loss": 5.9618,
      "mean_token_accuracy": 0.15825531929731368,
      "num_tokens": 4517746.0,
      "step": 2025
    },
    {
      "entropy": 6.141182708740234,
      "epoch": 0.2280002246307632,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998483044617694,
      "loss": 6.0486,
      "mean_token_accuracy": 0.15095944553613663,
      "num_tokens": 4529149.0,
      "step": 2030
    },
    {
      "entropy": 6.051037979125977,
      "epoch": 0.22856180153872072,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998468268443416,
      "loss": 5.9215,
      "mean_token_accuracy": 0.1605049952864647,
      "num_tokens": 4540630.0,
      "step": 2035
    },
    {
      "entropy": 5.922482109069824,
      "epoch": 0.22912337844667827,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998453420677034,
      "loss": 5.8671,
      "mean_token_accuracy": 0.1568058595061302,
      "num_tokens": 4552877.0,
      "step": 2040
    },
    {
      "entropy": 6.109559631347656,
      "epoch": 0.22968495535463582,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998438501319025,
      "loss": 5.9557,
      "mean_token_accuracy": 0.16576798111200333,
      "num_tokens": 4563634.0,
      "step": 2045
    },
    {
      "entropy": 6.006631994247437,
      "epoch": 0.23024653226259337,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998423510369862,
      "loss": 5.8817,
      "mean_token_accuracy": 0.16414335668087005,
      "num_tokens": 4574007.0,
      "step": 2050
    },
    {
      "entropy": 6.055240154266357,
      "epoch": 0.23080810917055092,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998408447830022,
      "loss": 5.9052,
      "mean_token_accuracy": 0.1573481395840645,
      "num_tokens": 4584146.0,
      "step": 2055
    },
    {
      "entropy": 6.047984170913696,
      "epoch": 0.23136968607850844,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998393313699985,
      "loss": 5.9022,
      "mean_token_accuracy": 0.15559675097465514,
      "num_tokens": 4594839.0,
      "step": 2060
    },
    {
      "entropy": 6.17945294380188,
      "epoch": 0.231931262986466,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998378107980233,
      "loss": 6.0713,
      "mean_token_accuracy": 0.15039688497781753,
      "num_tokens": 4608229.0,
      "step": 2065
    },
    {
      "entropy": 6.033313131332397,
      "epoch": 0.23249283989442354,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998362830671249,
      "loss": 5.8768,
      "mean_token_accuracy": 0.1581628516316414,
      "num_tokens": 4619206.0,
      "step": 2070
    },
    {
      "entropy": 5.999654817581177,
      "epoch": 0.2330544168023811,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998347481773522,
      "loss": 5.9064,
      "mean_token_accuracy": 0.16410871893167495,
      "num_tokens": 4628966.0,
      "step": 2075
    },
    {
      "entropy": 6.044484996795655,
      "epoch": 0.23361599371033864,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998332061287538,
      "loss": 5.7844,
      "mean_token_accuracy": 0.1657311201095581,
      "num_tokens": 4639935.0,
      "step": 2080
    },
    {
      "entropy": 6.110540771484375,
      "epoch": 0.23417757061829617,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499831656921379,
      "loss": 6.0312,
      "mean_token_accuracy": 0.15397130995988845,
      "num_tokens": 4651998.0,
      "step": 2085
    },
    {
      "entropy": 6.098618364334106,
      "epoch": 0.23473914752625372,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499830100555277,
      "loss": 6.0119,
      "mean_token_accuracy": 0.15659218579530715,
      "num_tokens": 4664002.0,
      "step": 2090
    },
    {
      "entropy": 6.0605607509613035,
      "epoch": 0.23530072443421127,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998285370304974,
      "loss": 5.8904,
      "mean_token_accuracy": 0.1646180883049965,
      "num_tokens": 4675343.0,
      "step": 2095
    },
    {
      "entropy": 6.067137908935547,
      "epoch": 0.23586230134216882,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00049982696634709,
      "loss": 5.9434,
      "mean_token_accuracy": 0.16149791181087494,
      "num_tokens": 4686410.0,
      "step": 2100
    },
    {
      "entropy": 6.0088635921478275,
      "epoch": 0.23642387825012637,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004998253885051048,
      "loss": 5.9631,
      "mean_token_accuracy": 0.15998635888099672,
      "num_tokens": 4697768.0,
      "step": 2105
    },
    {
      "entropy": 6.186365938186645,
      "epoch": 0.2369854551580839,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499823803504592,
      "loss": 5.9753,
      "mean_token_accuracy": 0.15657732859253884,
      "num_tokens": 4709933.0,
      "step": 2110
    },
    {
      "entropy": 6.0428260326385494,
      "epoch": 0.23754703206604144,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998222113456022,
      "loss": 6.0053,
      "mean_token_accuracy": 0.15467410236597062,
      "num_tokens": 4721385.0,
      "step": 2115
    },
    {
      "entropy": 5.994955587387085,
      "epoch": 0.238108608973999,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998206120281858,
      "loss": 5.829,
      "mean_token_accuracy": 0.166511407494545,
      "num_tokens": 4732562.0,
      "step": 2120
    },
    {
      "entropy": 6.013138246536255,
      "epoch": 0.23867018588195654,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499819005552394,
      "loss": 5.8411,
      "mean_token_accuracy": 0.16182543337345123,
      "num_tokens": 4742794.0,
      "step": 2125
    },
    {
      "entropy": 5.936925029754638,
      "epoch": 0.23923176278991407,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004998173919182779,
      "loss": 5.8084,
      "mean_token_accuracy": 0.166377691924572,
      "num_tokens": 4754903.0,
      "step": 2130
    },
    {
      "entropy": 6.010518169403076,
      "epoch": 0.23979333969787162,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499815771125889,
      "loss": 5.8531,
      "mean_token_accuracy": 0.1603682741522789,
      "num_tokens": 4766165.0,
      "step": 2135
    },
    {
      "entropy": 6.016250658035278,
      "epoch": 0.24035491660582917,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998141431752785,
      "loss": 5.8637,
      "mean_token_accuracy": 0.1660080596804619,
      "num_tokens": 4776551.0,
      "step": 2140
    },
    {
      "entropy": 6.132618141174317,
      "epoch": 0.24091649351378672,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998125080664987,
      "loss": 6.0549,
      "mean_token_accuracy": 0.15564174950122833,
      "num_tokens": 4788344.0,
      "step": 2145
    },
    {
      "entropy": 6.058242464065552,
      "epoch": 0.24147807042174427,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998108657996012,
      "loss": 5.859,
      "mean_token_accuracy": 0.15423323065042496,
      "num_tokens": 4799314.0,
      "step": 2150
    },
    {
      "entropy": 5.994201755523681,
      "epoch": 0.2420396473297018,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998092163746387,
      "loss": 5.9066,
      "mean_token_accuracy": 0.16130903661251067,
      "num_tokens": 4809588.0,
      "step": 2155
    },
    {
      "entropy": 5.960495901107788,
      "epoch": 0.24260122423765934,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998075597916636,
      "loss": 5.7183,
      "mean_token_accuracy": 0.16946181207895278,
      "num_tokens": 4820076.0,
      "step": 2160
    },
    {
      "entropy": 6.063351583480835,
      "epoch": 0.2431628011456169,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998058960507286,
      "loss": 5.9823,
      "mean_token_accuracy": 0.15857519656419755,
      "num_tokens": 4831504.0,
      "step": 2165
    },
    {
      "entropy": 5.972692489624023,
      "epoch": 0.24372437805357444,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998042251518866,
      "loss": 5.8399,
      "mean_token_accuracy": 0.16298711746931077,
      "num_tokens": 4842826.0,
      "step": 2170
    },
    {
      "entropy": 5.924139308929443,
      "epoch": 0.244285954961532,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998025470951908,
      "loss": 5.8275,
      "mean_token_accuracy": 0.1663692355155945,
      "num_tokens": 4853279.0,
      "step": 2175
    },
    {
      "entropy": 6.073525381088257,
      "epoch": 0.24484753186948952,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998008618806949,
      "loss": 5.8844,
      "mean_token_accuracy": 0.16937669962644578,
      "num_tokens": 4863332.0,
      "step": 2180
    },
    {
      "entropy": 6.015661668777466,
      "epoch": 0.24540910877744707,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997991695084523,
      "loss": 6.0104,
      "mean_token_accuracy": 0.15071545988321305,
      "num_tokens": 4874217.0,
      "step": 2185
    },
    {
      "entropy": 6.107934236526489,
      "epoch": 0.24597068568540462,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997974699785169,
      "loss": 5.8694,
      "mean_token_accuracy": 0.16703585684299468,
      "num_tokens": 4885322.0,
      "step": 2190
    },
    {
      "entropy": 5.851417207717896,
      "epoch": 0.24653226259336217,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997957632909429,
      "loss": 5.8075,
      "mean_token_accuracy": 0.16415573805570602,
      "num_tokens": 4895546.0,
      "step": 2195
    },
    {
      "entropy": 6.211975765228272,
      "epoch": 0.24709383950131972,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997940494457846,
      "loss": 6.0196,
      "mean_token_accuracy": 0.1556576743721962,
      "num_tokens": 4906957.0,
      "step": 2200
    },
    {
      "entropy": 6.0884904861450195,
      "epoch": 0.24765541640927724,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004997923284430967,
      "loss": 5.9478,
      "mean_token_accuracy": 0.1524835154414177,
      "num_tokens": 4917965.0,
      "step": 2205
    },
    {
      "entropy": 5.958709621429444,
      "epoch": 0.2482169933172348,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997906002829338,
      "loss": 5.8409,
      "mean_token_accuracy": 0.16658256202936172,
      "num_tokens": 4928585.0,
      "step": 2210
    },
    {
      "entropy": 5.983714294433594,
      "epoch": 0.24877857022519234,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000499788864965351,
      "loss": 5.8507,
      "mean_token_accuracy": 0.15959688723087312,
      "num_tokens": 4939818.0,
      "step": 2215
    },
    {
      "entropy": 6.051799154281616,
      "epoch": 0.2493401471331499,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997871224904036,
      "loss": 5.9405,
      "mean_token_accuracy": 0.15943082571029663,
      "num_tokens": 4951752.0,
      "step": 2220
    },
    {
      "entropy": 5.910864162445068,
      "epoch": 0.24990172404110744,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997853728581468,
      "loss": 5.7656,
      "mean_token_accuracy": 0.1712748274207115,
      "num_tokens": 4961846.0,
      "step": 2225
    },
    {
      "entropy": 5.999048137664795,
      "epoch": 0.250463300949065,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997836160686368,
      "loss": 5.9254,
      "mean_token_accuracy": 0.1582569435238838,
      "num_tokens": 4973693.0,
      "step": 2230
    },
    {
      "entropy": 6.025012540817261,
      "epoch": 0.25102487785702254,
      "grad_norm": 0.875,
      "learning_rate": 0.0004997818521219291,
      "loss": 5.8269,
      "mean_token_accuracy": 0.16620030850172043,
      "num_tokens": 4986156.0,
      "step": 2235
    },
    {
      "entropy": 5.984764194488525,
      "epoch": 0.25158645476498004,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997800810180803,
      "loss": 5.9095,
      "mean_token_accuracy": 0.16009719669818878,
      "num_tokens": 4997615.0,
      "step": 2240
    },
    {
      "entropy": 6.036053133010864,
      "epoch": 0.2521480316729376,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997783027571464,
      "loss": 5.8567,
      "mean_token_accuracy": 0.16056618392467498,
      "num_tokens": 5007655.0,
      "step": 2245
    },
    {
      "entropy": 5.977190017700195,
      "epoch": 0.25270960858089514,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997765173391841,
      "loss": 5.9233,
      "mean_token_accuracy": 0.1614665985107422,
      "num_tokens": 5018933.0,
      "step": 2250
    },
    {
      "entropy": 6.129662370681762,
      "epoch": 0.2532711854888527,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997747247642505,
      "loss": 5.9087,
      "mean_token_accuracy": 0.16192703396081926,
      "num_tokens": 5030495.0,
      "step": 2255
    },
    {
      "entropy": 5.987244319915772,
      "epoch": 0.25383276239681024,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997729250324023,
      "loss": 5.9617,
      "mean_token_accuracy": 0.15674430280923843,
      "num_tokens": 5041553.0,
      "step": 2260
    },
    {
      "entropy": 6.0051435947418215,
      "epoch": 0.2543943393047678,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004997711181436971,
      "loss": 5.9078,
      "mean_token_accuracy": 0.16368893831968306,
      "num_tokens": 5052312.0,
      "step": 2265
    },
    {
      "entropy": 6.128569221496582,
      "epoch": 0.25495591621272534,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004997693040981922,
      "loss": 6.085,
      "mean_token_accuracy": 0.1522883377969265,
      "num_tokens": 5063445.0,
      "step": 2270
    },
    {
      "entropy": 6.1037023067474365,
      "epoch": 0.2555174931206829,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997674828959456,
      "loss": 5.8583,
      "mean_token_accuracy": 0.1592344179749489,
      "num_tokens": 5073052.0,
      "step": 2275
    },
    {
      "entropy": 5.964641094207764,
      "epoch": 0.25607907002864044,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004997656545370151,
      "loss": 5.9127,
      "mean_token_accuracy": 0.16215995848178863,
      "num_tokens": 5084174.0,
      "step": 2280
    },
    {
      "entropy": 6.082666969299316,
      "epoch": 0.256640646936598,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499763819021459,
      "loss": 5.8785,
      "mean_token_accuracy": 0.16361337155103683,
      "num_tokens": 5094106.0,
      "step": 2285
    },
    {
      "entropy": 6.075915050506592,
      "epoch": 0.2572022238445555,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004997619763493357,
      "loss": 5.9702,
      "mean_token_accuracy": 0.16184501498937606,
      "num_tokens": 5106963.0,
      "step": 2290
    },
    {
      "entropy": 5.9209068775177,
      "epoch": 0.25776380075251304,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000499760126520704,
      "loss": 5.7578,
      "mean_token_accuracy": 0.16060159504413604,
      "num_tokens": 5119476.0,
      "step": 2295
    },
    {
      "entropy": 6.066161727905273,
      "epoch": 0.2583253776604706,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004997582695356225,
      "loss": 6.0127,
      "mean_token_accuracy": 0.15276381820440293,
      "num_tokens": 5132169.0,
      "step": 2300
    },
    {
      "entropy": 6.004144668579102,
      "epoch": 0.25888695456842814,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004997564053941507,
      "loss": 5.8593,
      "mean_token_accuracy": 0.1646599918603897,
      "num_tokens": 5143354.0,
      "step": 2305
    },
    {
      "entropy": 5.958068227767944,
      "epoch": 0.2594485314763857,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004997545340963478,
      "loss": 5.9369,
      "mean_token_accuracy": 0.15578429996967316,
      "num_tokens": 5154624.0,
      "step": 2310
    },
    {
      "entropy": 6.065716075897217,
      "epoch": 0.26001010838434324,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997526556422733,
      "loss": 5.8178,
      "mean_token_accuracy": 0.1637489452958107,
      "num_tokens": 5165349.0,
      "step": 2315
    },
    {
      "entropy": 5.9429624557495115,
      "epoch": 0.2605716852923008,
      "grad_norm": 1.125,
      "learning_rate": 0.0004997507700319872,
      "loss": 5.9019,
      "mean_token_accuracy": 0.16330261379480362,
      "num_tokens": 5175874.0,
      "step": 2320
    },
    {
      "entropy": 5.980010223388672,
      "epoch": 0.26113326220025834,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997488772655492,
      "loss": 5.7911,
      "mean_token_accuracy": 0.16592907160520554,
      "num_tokens": 5186591.0,
      "step": 2325
    },
    {
      "entropy": 5.940133285522461,
      "epoch": 0.2616948391082159,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997469773430199,
      "loss": 5.8994,
      "mean_token_accuracy": 0.16807617843151093,
      "num_tokens": 5197307.0,
      "step": 2330
    },
    {
      "entropy": 5.951906204223633,
      "epoch": 0.2622564160161734,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004997450702644596,
      "loss": 5.8541,
      "mean_token_accuracy": 0.16079187989234925,
      "num_tokens": 5209868.0,
      "step": 2335
    },
    {
      "entropy": 6.027046060562133,
      "epoch": 0.26281799292413094,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004997431560299292,
      "loss": 5.8703,
      "mean_token_accuracy": 0.16482030898332595,
      "num_tokens": 5220365.0,
      "step": 2340
    },
    {
      "entropy": 6.010868310928345,
      "epoch": 0.2633795698320885,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004997412346394894,
      "loss": 5.7749,
      "mean_token_accuracy": 0.16639859229326248,
      "num_tokens": 5231575.0,
      "step": 2345
    },
    {
      "entropy": 6.0061742782592775,
      "epoch": 0.26394114674004604,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997393060932017,
      "loss": 5.8708,
      "mean_token_accuracy": 0.1627236321568489,
      "num_tokens": 5241852.0,
      "step": 2350
    },
    {
      "entropy": 6.04571762084961,
      "epoch": 0.2645027236480036,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004997373703911273,
      "loss": 5.8988,
      "mean_token_accuracy": 0.16516432911157608,
      "num_tokens": 5254536.0,
      "step": 2355
    },
    {
      "entropy": 6.020670461654663,
      "epoch": 0.26506430055596114,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997354275333277,
      "loss": 5.8305,
      "mean_token_accuracy": 0.16114060282707215,
      "num_tokens": 5264946.0,
      "step": 2360
    },
    {
      "entropy": 5.751336193084716,
      "epoch": 0.2656258774639187,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499733477519865,
      "loss": 5.6596,
      "mean_token_accuracy": 0.1632534980773926,
      "num_tokens": 5275527.0,
      "step": 2365
    },
    {
      "entropy": 6.002241468429565,
      "epoch": 0.26618745437187624,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997315203508013,
      "loss": 5.8389,
      "mean_token_accuracy": 0.16308278292417527,
      "num_tokens": 5285471.0,
      "step": 2370
    },
    {
      "entropy": 5.865285634994507,
      "epoch": 0.2667490312798338,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997295560261988,
      "loss": 5.7524,
      "mean_token_accuracy": 0.16390506327152252,
      "num_tokens": 5295743.0,
      "step": 2375
    },
    {
      "entropy": 6.007153558731079,
      "epoch": 0.26731060818779134,
      "grad_norm": 0.96875,
      "learning_rate": 0.00049972758454612,
      "loss": 5.8958,
      "mean_token_accuracy": 0.15694423466920854,
      "num_tokens": 5307963.0,
      "step": 2380
    },
    {
      "entropy": 5.956030988693238,
      "epoch": 0.26787218509574884,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997256059106279,
      "loss": 5.7866,
      "mean_token_accuracy": 0.16542109698057175,
      "num_tokens": 5318394.0,
      "step": 2385
    },
    {
      "entropy": 5.967977046966553,
      "epoch": 0.2684337620037064,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004997236201197853,
      "loss": 6.0107,
      "mean_token_accuracy": 0.15119242966175078,
      "num_tokens": 5331728.0,
      "step": 2390
    },
    {
      "entropy": 6.002740478515625,
      "epoch": 0.26899533891166394,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997216271736554,
      "loss": 5.9046,
      "mean_token_accuracy": 0.1701723799109459,
      "num_tokens": 5342825.0,
      "step": 2395
    },
    {
      "entropy": 6.016122913360595,
      "epoch": 0.2695569158196215,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997196270723018,
      "loss": 5.8094,
      "mean_token_accuracy": 0.16391160041093827,
      "num_tokens": 5354232.0,
      "step": 2400
    },
    {
      "entropy": 5.991284513473511,
      "epoch": 0.27011849272757904,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997176198157881,
      "loss": 5.8986,
      "mean_token_accuracy": 0.15782748833298682,
      "num_tokens": 5364807.0,
      "step": 2405
    },
    {
      "entropy": 5.94521279335022,
      "epoch": 0.2706800696355366,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997156054041781,
      "loss": 5.8784,
      "mean_token_accuracy": 0.15805881172418595,
      "num_tokens": 5375593.0,
      "step": 2410
    },
    {
      "entropy": 5.9776763916015625,
      "epoch": 0.27124164654349414,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004997135838375362,
      "loss": 5.8305,
      "mean_token_accuracy": 0.16881559789180756,
      "num_tokens": 5386740.0,
      "step": 2415
    },
    {
      "entropy": 5.918038845062256,
      "epoch": 0.2718032234514517,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997115551159267,
      "loss": 5.8129,
      "mean_token_accuracy": 0.16638245284557343,
      "num_tokens": 5397732.0,
      "step": 2420
    },
    {
      "entropy": 6.003497171401977,
      "epoch": 0.27236480035940924,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499709519239414,
      "loss": 5.8656,
      "mean_token_accuracy": 0.15994038432836533,
      "num_tokens": 5409392.0,
      "step": 2425
    },
    {
      "entropy": 5.84049072265625,
      "epoch": 0.2729263772673668,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004997074762080632,
      "loss": 5.8188,
      "mean_token_accuracy": 0.17128721624612808,
      "num_tokens": 5420903.0,
      "step": 2430
    },
    {
      "entropy": 5.895452499389648,
      "epoch": 0.2734879541753243,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000499705426021939,
      "loss": 5.7535,
      "mean_token_accuracy": 0.16452100574970246,
      "num_tokens": 5432631.0,
      "step": 2435
    },
    {
      "entropy": 5.933747625350952,
      "epoch": 0.27404953108328184,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499703368681107,
      "loss": 5.7259,
      "mean_token_accuracy": 0.1722749501466751,
      "num_tokens": 5443463.0,
      "step": 2440
    },
    {
      "entropy": 6.0137842178344725,
      "epoch": 0.2746111079912394,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004997013041856326,
      "loss": 6.0335,
      "mean_token_accuracy": 0.16273341327905655,
      "num_tokens": 5455903.0,
      "step": 2445
    },
    {
      "entropy": 6.063531303405762,
      "epoch": 0.27517268489919694,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004996992325355815,
      "loss": 5.8611,
      "mean_token_accuracy": 0.1601864755153656,
      "num_tokens": 5467201.0,
      "step": 2450
    },
    {
      "entropy": 5.9214695453643795,
      "epoch": 0.2757342618071545,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004996971537310198,
      "loss": 5.7394,
      "mean_token_accuracy": 0.1641487091779709,
      "num_tokens": 5478109.0,
      "step": 2455
    },
    {
      "entropy": 5.962538623809815,
      "epoch": 0.27629583871511204,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004996950677720133,
      "loss": 5.8707,
      "mean_token_accuracy": 0.1609448343515396,
      "num_tokens": 5489621.0,
      "step": 2460
    },
    {
      "entropy": 5.942273902893066,
      "epoch": 0.2768574156230696,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000499692974658629,
      "loss": 5.8163,
      "mean_token_accuracy": 0.1653575286269188,
      "num_tokens": 5500606.0,
      "step": 2465
    },
    {
      "entropy": 5.997242736816406,
      "epoch": 0.27741899253102714,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004996908743909331,
      "loss": 5.7714,
      "mean_token_accuracy": 0.17135910093784332,
      "num_tokens": 5510768.0,
      "step": 2470
    },
    {
      "entropy": 5.871461677551269,
      "epoch": 0.2779805694389847,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996887669689926,
      "loss": 5.7834,
      "mean_token_accuracy": 0.16663116961717606,
      "num_tokens": 5522210.0,
      "step": 2475
    },
    {
      "entropy": 5.993370628356933,
      "epoch": 0.2785421463469422,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996866523928747,
      "loss": 5.906,
      "mean_token_accuracy": 0.1565551370382309,
      "num_tokens": 5533055.0,
      "step": 2480
    },
    {
      "entropy": 6.032248830795288,
      "epoch": 0.27910372325489974,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004996845306626467,
      "loss": 5.9126,
      "mean_token_accuracy": 0.16118640154600145,
      "num_tokens": 5544345.0,
      "step": 2485
    },
    {
      "entropy": 5.952008962631226,
      "epoch": 0.2796653001628573,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996824017783759,
      "loss": 5.8219,
      "mean_token_accuracy": 0.16080627143383025,
      "num_tokens": 5556734.0,
      "step": 2490
    },
    {
      "entropy": 5.994874143600464,
      "epoch": 0.28022687707081484,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004996802657401304,
      "loss": 5.8645,
      "mean_token_accuracy": 0.16483049243688583,
      "num_tokens": 5566874.0,
      "step": 2495
    },
    {
      "entropy": 5.973210859298706,
      "epoch": 0.2807884539787724,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004996781225479781,
      "loss": 5.8653,
      "mean_token_accuracy": 0.16056755185127258,
      "num_tokens": 5578812.0,
      "step": 2500
    },
    {
      "entropy": 5.982892560958862,
      "epoch": 0.28135003088672994,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004996759722019872,
      "loss": 5.9063,
      "mean_token_accuracy": 0.16154668629169464,
      "num_tokens": 5590021.0,
      "step": 2505
    },
    {
      "entropy": 5.968021821975708,
      "epoch": 0.2819116077946875,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004996738147022262,
      "loss": 5.9167,
      "mean_token_accuracy": 0.15916191339492797,
      "num_tokens": 5600921.0,
      "step": 2510
    },
    {
      "entropy": 5.988654041290284,
      "epoch": 0.28247318470264504,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004996716500487638,
      "loss": 5.842,
      "mean_token_accuracy": 0.16069669425487518,
      "num_tokens": 5610615.0,
      "step": 2515
    },
    {
      "entropy": 5.831103801727295,
      "epoch": 0.2830347616106026,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499669478241669,
      "loss": 5.7641,
      "mean_token_accuracy": 0.16576093584299087,
      "num_tokens": 5621596.0,
      "step": 2520
    },
    {
      "entropy": 6.091034746170044,
      "epoch": 0.28359633851856014,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004996672992810108,
      "loss": 5.8097,
      "mean_token_accuracy": 0.17065599262714387,
      "num_tokens": 5633047.0,
      "step": 2525
    },
    {
      "entropy": 5.871144771575928,
      "epoch": 0.28415791542651764,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004996651131668587,
      "loss": 5.7919,
      "mean_token_accuracy": 0.16313879787921906,
      "num_tokens": 5644970.0,
      "step": 2530
    },
    {
      "entropy": 5.932213592529297,
      "epoch": 0.2847194923344752,
      "grad_norm": 0.875,
      "learning_rate": 0.0004996629198992823,
      "loss": 5.912,
      "mean_token_accuracy": 0.1626256823539734,
      "num_tokens": 5656631.0,
      "step": 2535
    },
    {
      "entropy": 5.9674906730651855,
      "epoch": 0.28528106924243274,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004996607194783513,
      "loss": 5.8584,
      "mean_token_accuracy": 0.1576921448111534,
      "num_tokens": 5668808.0,
      "step": 2540
    },
    {
      "entropy": 5.915722131729126,
      "epoch": 0.2858426461503903,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004996585119041359,
      "loss": 5.7941,
      "mean_token_accuracy": 0.16423020511865616,
      "num_tokens": 5679493.0,
      "step": 2545
    },
    {
      "entropy": 5.967693042755127,
      "epoch": 0.28640422305834784,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996562971767063,
      "loss": 5.8304,
      "mean_token_accuracy": 0.16942446082830429,
      "num_tokens": 5689879.0,
      "step": 2550
    },
    {
      "entropy": 6.001515483856201,
      "epoch": 0.2869657999663054,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499654075296133,
      "loss": 5.9249,
      "mean_token_accuracy": 0.16562376022338868,
      "num_tokens": 5699934.0,
      "step": 2555
    },
    {
      "entropy": 5.961040449142456,
      "epoch": 0.28752737687426294,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499651846262487,
      "loss": 5.8402,
      "mean_token_accuracy": 0.16511016041040422,
      "num_tokens": 5710171.0,
      "step": 2560
    },
    {
      "entropy": 5.93832802772522,
      "epoch": 0.2880889537822205,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996496100758389,
      "loss": 5.8211,
      "mean_token_accuracy": 0.15841380059719085,
      "num_tokens": 5721190.0,
      "step": 2565
    },
    {
      "entropy": 6.034723091125488,
      "epoch": 0.28865053069017804,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996473667362604,
      "loss": 5.9317,
      "mean_token_accuracy": 0.16139888912439346,
      "num_tokens": 5733706.0,
      "step": 2570
    },
    {
      "entropy": 5.9140746116638185,
      "epoch": 0.28921210759813554,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004996451162438224,
      "loss": 5.8885,
      "mean_token_accuracy": 0.1655304715037346,
      "num_tokens": 5745712.0,
      "step": 2575
    },
    {
      "entropy": 5.83914361000061,
      "epoch": 0.2897736845060931,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996428585985967,
      "loss": 5.7097,
      "mean_token_accuracy": 0.1763711154460907,
      "num_tokens": 5757915.0,
      "step": 2580
    },
    {
      "entropy": 5.975746059417725,
      "epoch": 0.29033526141405064,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004996405938006553,
      "loss": 5.8639,
      "mean_token_accuracy": 0.16630038470029831,
      "num_tokens": 5768788.0,
      "step": 2585
    },
    {
      "entropy": 5.969762325286865,
      "epoch": 0.2908968383220082,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996383218500703,
      "loss": 5.7839,
      "mean_token_accuracy": 0.16176656484603882,
      "num_tokens": 5779806.0,
      "step": 2590
    },
    {
      "entropy": 5.840362787246704,
      "epoch": 0.29145841522996574,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499636042746914,
      "loss": 5.754,
      "mean_token_accuracy": 0.16268186420202255,
      "num_tokens": 5790177.0,
      "step": 2595
    },
    {
      "entropy": 5.988736343383789,
      "epoch": 0.2920199921379233,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499633756491259,
      "loss": 5.7853,
      "mean_token_accuracy": 0.16794030368328094,
      "num_tokens": 5800251.0,
      "step": 2600
    },
    {
      "entropy": 5.903754186630249,
      "epoch": 0.29258156904588084,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004996314630831781,
      "loss": 5.8661,
      "mean_token_accuracy": 0.16568554937839508,
      "num_tokens": 5812210.0,
      "step": 2605
    },
    {
      "entropy": 6.010973787307739,
      "epoch": 0.2931431459538384,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004996291625227443,
      "loss": 5.8473,
      "mean_token_accuracy": 0.1710708886384964,
      "num_tokens": 5823280.0,
      "step": 2610
    },
    {
      "entropy": 6.021354675292969,
      "epoch": 0.29370472286179594,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004996268548100307,
      "loss": 5.9222,
      "mean_token_accuracy": 0.15569793581962585,
      "num_tokens": 5835076.0,
      "step": 2615
    },
    {
      "entropy": 5.831182050704956,
      "epoch": 0.2942662997697535,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499624539945111,
      "loss": 5.7202,
      "mean_token_accuracy": 0.17435432225465775,
      "num_tokens": 5845376.0,
      "step": 2620
    },
    {
      "entropy": 5.840474462509155,
      "epoch": 0.294827876677711,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004996222179280589,
      "loss": 5.7454,
      "mean_token_accuracy": 0.1622859939932823,
      "num_tokens": 5857124.0,
      "step": 2625
    },
    {
      "entropy": 6.0213315963745115,
      "epoch": 0.29538945358566854,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004996198887589481,
      "loss": 5.846,
      "mean_token_accuracy": 0.1605096936225891,
      "num_tokens": 5868681.0,
      "step": 2630
    },
    {
      "entropy": 5.971142911911011,
      "epoch": 0.2959510304936261,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004996175524378531,
      "loss": 5.8464,
      "mean_token_accuracy": 0.165427128970623,
      "num_tokens": 5880532.0,
      "step": 2635
    },
    {
      "entropy": 5.9712464809417725,
      "epoch": 0.29651260740158364,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499615208964848,
      "loss": 5.7886,
      "mean_token_accuracy": 0.17320181131362916,
      "num_tokens": 5891582.0,
      "step": 2640
    },
    {
      "entropy": 5.880084705352783,
      "epoch": 0.2970741843095412,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004996128583400077,
      "loss": 5.7977,
      "mean_token_accuracy": 0.16468533724546433,
      "num_tokens": 5903543.0,
      "step": 2645
    },
    {
      "entropy": 5.907572174072266,
      "epoch": 0.29763576121749874,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004996105005634066,
      "loss": 5.7894,
      "mean_token_accuracy": 0.1727600485086441,
      "num_tokens": 5913140.0,
      "step": 2650
    },
    {
      "entropy": 5.902091121673584,
      "epoch": 0.2981973381254563,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996081356351202,
      "loss": 5.7472,
      "mean_token_accuracy": 0.1698393240571022,
      "num_tokens": 5924010.0,
      "step": 2655
    },
    {
      "entropy": 5.959113073348999,
      "epoch": 0.29875891503341384,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996057635552237,
      "loss": 5.9362,
      "mean_token_accuracy": 0.16425639539957046,
      "num_tokens": 5936246.0,
      "step": 2660
    },
    {
      "entropy": 5.979541826248169,
      "epoch": 0.2993204919413714,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996033843237925,
      "loss": 5.8277,
      "mean_token_accuracy": 0.15435630530118943,
      "num_tokens": 5947594.0,
      "step": 2665
    },
    {
      "entropy": 5.950691604614258,
      "epoch": 0.29988206884932894,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996009979409025,
      "loss": 5.7607,
      "mean_token_accuracy": 0.1573063001036644,
      "num_tokens": 5957660.0,
      "step": 2670
    },
    {
      "entropy": 5.775152015686035,
      "epoch": 0.30044364575728644,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004995986044066295,
      "loss": 5.6079,
      "mean_token_accuracy": 0.1789610669016838,
      "num_tokens": 5968881.0,
      "step": 2675
    },
    {
      "entropy": 5.896370124816895,
      "epoch": 0.301005222665244,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00049959620372105,
      "loss": 5.7695,
      "mean_token_accuracy": 0.16645503789186478,
      "num_tokens": 5980032.0,
      "step": 2680
    },
    {
      "entropy": 5.799341154098511,
      "epoch": 0.30156679957320154,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004995937958842401,
      "loss": 5.7523,
      "mean_token_accuracy": 0.16536690443754196,
      "num_tokens": 5991781.0,
      "step": 2685
    },
    {
      "entropy": 5.909103488922119,
      "epoch": 0.3021283764811591,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004995913808962767,
      "loss": 5.8017,
      "mean_token_accuracy": 0.16248384714126587,
      "num_tokens": 6002303.0,
      "step": 2690
    },
    {
      "entropy": 6.048180866241455,
      "epoch": 0.30268995338911664,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004995889587572366,
      "loss": 5.8939,
      "mean_token_accuracy": 0.16326897740364074,
      "num_tokens": 6011523.0,
      "step": 2695
    },
    {
      "entropy": 5.917131185531616,
      "epoch": 0.3032515302970742,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000499586529467197,
      "loss": 5.699,
      "mean_token_accuracy": 0.16807254999876023,
      "num_tokens": 6022169.0,
      "step": 2700
    },
    {
      "entropy": 5.779902648925781,
      "epoch": 0.30381310720503174,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995840930262351,
      "loss": 5.7567,
      "mean_token_accuracy": 0.1646862246096134,
      "num_tokens": 6033393.0,
      "step": 2705
    },
    {
      "entropy": 5.978853797912597,
      "epoch": 0.3043746841129893,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004995816494344287,
      "loss": 5.8127,
      "mean_token_accuracy": 0.16036659330129624,
      "num_tokens": 6045365.0,
      "step": 2710
    },
    {
      "entropy": 5.882379770278931,
      "epoch": 0.30493626102094684,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004995791986918554,
      "loss": 5.8456,
      "mean_token_accuracy": 0.1643943414092064,
      "num_tokens": 6057482.0,
      "step": 2715
    },
    {
      "entropy": 5.924764823913574,
      "epoch": 0.30549783792890434,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995767407985933,
      "loss": 5.748,
      "mean_token_accuracy": 0.16873836368322373,
      "num_tokens": 6068194.0,
      "step": 2720
    },
    {
      "entropy": 5.872408819198609,
      "epoch": 0.3060594148368619,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004995742757547207,
      "loss": 5.7969,
      "mean_token_accuracy": 0.16693959385156631,
      "num_tokens": 6080540.0,
      "step": 2725
    },
    {
      "entropy": 5.886364841461182,
      "epoch": 0.30662099174481944,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000499571803560316,
      "loss": 5.8463,
      "mean_token_accuracy": 0.16102577671408652,
      "num_tokens": 6092464.0,
      "step": 2730
    },
    {
      "entropy": 5.98257966041565,
      "epoch": 0.307182568652777,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995693242154581,
      "loss": 5.802,
      "mean_token_accuracy": 0.16416530013084413,
      "num_tokens": 6103242.0,
      "step": 2735
    },
    {
      "entropy": 5.849971914291382,
      "epoch": 0.30774414556073454,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004995668377202258,
      "loss": 5.8183,
      "mean_token_accuracy": 0.1640830472111702,
      "num_tokens": 6116383.0,
      "step": 2740
    },
    {
      "entropy": 5.84266676902771,
      "epoch": 0.3083057224686921,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004995643440746982,
      "loss": 5.6394,
      "mean_token_accuracy": 0.17457838654518126,
      "num_tokens": 6126883.0,
      "step": 2745
    },
    {
      "entropy": 5.919090747833252,
      "epoch": 0.30886729937664964,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004995618432789548,
      "loss": 5.8257,
      "mean_token_accuracy": 0.15941525995731354,
      "num_tokens": 6139613.0,
      "step": 2750
    },
    {
      "entropy": 5.950811958312988,
      "epoch": 0.3094288762846072,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004995593353330753,
      "loss": 5.8907,
      "mean_token_accuracy": 0.16074301451444625,
      "num_tokens": 6151950.0,
      "step": 2755
    },
    {
      "entropy": 5.862967920303345,
      "epoch": 0.30999045319256474,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995568202371393,
      "loss": 5.6848,
      "mean_token_accuracy": 0.17329418808221816,
      "num_tokens": 6161563.0,
      "step": 2760
    },
    {
      "entropy": 5.823839712142944,
      "epoch": 0.3105520301005223,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995542979912272,
      "loss": 5.7043,
      "mean_token_accuracy": 0.17179031819105148,
      "num_tokens": 6170966.0,
      "step": 2765
    },
    {
      "entropy": 5.878972959518433,
      "epoch": 0.3111136070084798,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004995517685954192,
      "loss": 5.7693,
      "mean_token_accuracy": 0.16634240001440048,
      "num_tokens": 6182563.0,
      "step": 2770
    },
    {
      "entropy": 5.939313697814941,
      "epoch": 0.31167518391643734,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004995492320497958,
      "loss": 5.8706,
      "mean_token_accuracy": 0.1600440576672554,
      "num_tokens": 6194495.0,
      "step": 2775
    },
    {
      "entropy": 5.873399639129639,
      "epoch": 0.3122367608243949,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004995466883544376,
      "loss": 5.7662,
      "mean_token_accuracy": 0.1647852584719658,
      "num_tokens": 6206594.0,
      "step": 2780
    },
    {
      "entropy": 5.954919242858887,
      "epoch": 0.31279833773235244,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004995441375094258,
      "loss": 5.7954,
      "mean_token_accuracy": 0.16247922629117967,
      "num_tokens": 6218371.0,
      "step": 2785
    },
    {
      "entropy": 5.913018321990966,
      "epoch": 0.31335991464031,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004995415795148417,
      "loss": 5.7292,
      "mean_token_accuracy": 0.1699868232011795,
      "num_tokens": 6228723.0,
      "step": 2790
    },
    {
      "entropy": 5.919421529769897,
      "epoch": 0.31392149154826754,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004995390143707665,
      "loss": 5.9119,
      "mean_token_accuracy": 0.16231555193662645,
      "num_tokens": 6240535.0,
      "step": 2795
    },
    {
      "entropy": 5.8778153419494625,
      "epoch": 0.3144830684562251,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995364420772821,
      "loss": 5.6162,
      "mean_token_accuracy": 0.17505984455347062,
      "num_tokens": 6250643.0,
      "step": 2800
    },
    {
      "entropy": 5.861969089508056,
      "epoch": 0.31504464536418264,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995338626344702,
      "loss": 5.7474,
      "mean_token_accuracy": 0.16888705044984817,
      "num_tokens": 6259569.0,
      "step": 2805
    },
    {
      "entropy": 5.898400115966797,
      "epoch": 0.3156062222721402,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499531276042413,
      "loss": 5.8918,
      "mean_token_accuracy": 0.170860131084919,
      "num_tokens": 6271108.0,
      "step": 2810
    },
    {
      "entropy": 5.952102947235107,
      "epoch": 0.31616779918009774,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004995286823011929,
      "loss": 5.7256,
      "mean_token_accuracy": 0.16026092171669007,
      "num_tokens": 6282352.0,
      "step": 2815
    },
    {
      "entropy": 5.886004495620727,
      "epoch": 0.31672937608805524,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004995260814108926,
      "loss": 5.851,
      "mean_token_accuracy": 0.1604265809059143,
      "num_tokens": 6294946.0,
      "step": 2820
    },
    {
      "entropy": 5.816281795501709,
      "epoch": 0.3172909529960128,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004995234733715947,
      "loss": 5.6743,
      "mean_token_accuracy": 0.1721860095858574,
      "num_tokens": 6305552.0,
      "step": 2825
    },
    {
      "entropy": 5.833727693557739,
      "epoch": 0.31785252990397034,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004995208581833824,
      "loss": 5.73,
      "mean_token_accuracy": 0.17419680058956147,
      "num_tokens": 6315815.0,
      "step": 2830
    },
    {
      "entropy": 5.898771047592163,
      "epoch": 0.3184141068119279,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995182358463388,
      "loss": 5.78,
      "mean_token_accuracy": 0.16381437554955483,
      "num_tokens": 6327384.0,
      "step": 2835
    },
    {
      "entropy": 5.888091230392456,
      "epoch": 0.31897568371988544,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004995156063605475,
      "loss": 5.7537,
      "mean_token_accuracy": 0.162344291806221,
      "num_tokens": 6338083.0,
      "step": 2840
    },
    {
      "entropy": 5.935994815826416,
      "epoch": 0.319537260627843,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004995129697260922,
      "loss": 5.7551,
      "mean_token_accuracy": 0.16952051520347594,
      "num_tokens": 6349058.0,
      "step": 2845
    },
    {
      "entropy": 5.8725059032440186,
      "epoch": 0.32009883753580054,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004995103259430569,
      "loss": 5.7565,
      "mean_token_accuracy": 0.1690447896718979,
      "num_tokens": 6359655.0,
      "step": 2850
    },
    {
      "entropy": 5.961487817764282,
      "epoch": 0.3206604144437581,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004995076750115258,
      "loss": 5.8847,
      "mean_token_accuracy": 0.16322531849145888,
      "num_tokens": 6372088.0,
      "step": 2855
    },
    {
      "entropy": 5.895509099960327,
      "epoch": 0.32122199135171564,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004995050169315832,
      "loss": 5.7935,
      "mean_token_accuracy": 0.16238415837287903,
      "num_tokens": 6382946.0,
      "step": 2860
    },
    {
      "entropy": 5.883610486984253,
      "epoch": 0.32178356825967314,
      "grad_norm": 0.875,
      "learning_rate": 0.0004995023517033138,
      "loss": 5.7896,
      "mean_token_accuracy": 0.16411769539117813,
      "num_tokens": 6392591.0,
      "step": 2865
    },
    {
      "entropy": 5.826980686187744,
      "epoch": 0.3223451451676307,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004994996793268024,
      "loss": 5.7277,
      "mean_token_accuracy": 0.17058148384094238,
      "num_tokens": 6403159.0,
      "step": 2870
    },
    {
      "entropy": 5.929961633682251,
      "epoch": 0.32290672207558824,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004994969998021342,
      "loss": 5.7521,
      "mean_token_accuracy": 0.16720252335071564,
      "num_tokens": 6414232.0,
      "step": 2875
    },
    {
      "entropy": 5.839997386932373,
      "epoch": 0.3234682989835458,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004994943131293944,
      "loss": 5.723,
      "mean_token_accuracy": 0.1778748333454132,
      "num_tokens": 6425762.0,
      "step": 2880
    },
    {
      "entropy": 5.903552103042602,
      "epoch": 0.32402987589150334,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004994916193086687,
      "loss": 5.7644,
      "mean_token_accuracy": 0.16310417652130127,
      "num_tokens": 6439077.0,
      "step": 2885
    },
    {
      "entropy": 5.938978385925293,
      "epoch": 0.3245914527994609,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004994889183400428,
      "loss": 5.8324,
      "mean_token_accuracy": 0.16593091040849686,
      "num_tokens": 6450425.0,
      "step": 2890
    },
    {
      "entropy": 5.865291404724121,
      "epoch": 0.32515302970741844,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004994862102236025,
      "loss": 5.8155,
      "mean_token_accuracy": 0.16792695373296737,
      "num_tokens": 6461347.0,
      "step": 2895
    },
    {
      "entropy": 5.897014474868774,
      "epoch": 0.325714606615376,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004994834949594342,
      "loss": 5.7877,
      "mean_token_accuracy": 0.1707422986626625,
      "num_tokens": 6472428.0,
      "step": 2900
    },
    {
      "entropy": 5.895409822463989,
      "epoch": 0.32627618352333354,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004994807725476246,
      "loss": 5.7553,
      "mean_token_accuracy": 0.17379966527223586,
      "num_tokens": 6483393.0,
      "step": 2905
    },
    {
      "entropy": 5.895529937744141,
      "epoch": 0.3268377604312911,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004994780429882599,
      "loss": 5.7343,
      "mean_token_accuracy": 0.1717408314347267,
      "num_tokens": 6493520.0,
      "step": 2910
    },
    {
      "entropy": 5.796739149093628,
      "epoch": 0.3273993373392486,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004994753062814274,
      "loss": 5.6676,
      "mean_token_accuracy": 0.1763666331768036,
      "num_tokens": 6505282.0,
      "step": 2915
    },
    {
      "entropy": 5.960649013519287,
      "epoch": 0.32796091424720614,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004994725624272139,
      "loss": 5.8475,
      "mean_token_accuracy": 0.17001553177833556,
      "num_tokens": 6516808.0,
      "step": 2920
    },
    {
      "entropy": 5.907238531112671,
      "epoch": 0.3285224911551637,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004994698114257071,
      "loss": 5.7925,
      "mean_token_accuracy": 0.16663295030593872,
      "num_tokens": 6529184.0,
      "step": 2925
    },
    {
      "entropy": 5.828442096710205,
      "epoch": 0.32908406806312124,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004994670532769944,
      "loss": 5.6616,
      "mean_token_accuracy": 0.1743689388036728,
      "num_tokens": 6539909.0,
      "step": 2930
    },
    {
      "entropy": 5.829112958908081,
      "epoch": 0.3296456449710788,
      "grad_norm": 0.875,
      "learning_rate": 0.0004994642879811637,
      "loss": 5.7021,
      "mean_token_accuracy": 0.17460112124681473,
      "num_tokens": 6550870.0,
      "step": 2935
    },
    {
      "entropy": 5.865769577026367,
      "epoch": 0.33020722187903634,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000499461515538303,
      "loss": 5.764,
      "mean_token_accuracy": 0.1730511799454689,
      "num_tokens": 6561784.0,
      "step": 2940
    },
    {
      "entropy": 5.788119268417359,
      "epoch": 0.3307687987869939,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004994587359485006,
      "loss": 5.7077,
      "mean_token_accuracy": 0.1722067967057228,
      "num_tokens": 6573422.0,
      "step": 2945
    },
    {
      "entropy": 5.982589244842529,
      "epoch": 0.33133037569495144,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000499455949211845,
      "loss": 5.7757,
      "mean_token_accuracy": 0.17219349592924119,
      "num_tokens": 6584565.0,
      "step": 2950
    },
    {
      "entropy": 5.886816024780273,
      "epoch": 0.331891952602909,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004994531553284248,
      "loss": 5.7037,
      "mean_token_accuracy": 0.16870585829019547,
      "num_tokens": 6594238.0,
      "step": 2955
    },
    {
      "entropy": 5.75544548034668,
      "epoch": 0.3324535295108665,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004994503542983293,
      "loss": 5.7631,
      "mean_token_accuracy": 0.16708966344594955,
      "num_tokens": 6606309.0,
      "step": 2960
    },
    {
      "entropy": 6.004780960083008,
      "epoch": 0.33301510641882404,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004994475461216472,
      "loss": 5.8607,
      "mean_token_accuracy": 0.1612526834011078,
      "num_tokens": 6617916.0,
      "step": 2965
    },
    {
      "entropy": 5.856878280639648,
      "epoch": 0.3335766833267816,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004994447307984684,
      "loss": 5.7077,
      "mean_token_accuracy": 0.16824225783348085,
      "num_tokens": 6627951.0,
      "step": 2970
    },
    {
      "entropy": 5.817630767822266,
      "epoch": 0.33413826023473914,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004994419083288823,
      "loss": 5.7693,
      "mean_token_accuracy": 0.16544559448957444,
      "num_tokens": 6639285.0,
      "step": 2975
    },
    {
      "entropy": 5.849265623092651,
      "epoch": 0.3346998371426967,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994390787129787,
      "loss": 5.6435,
      "mean_token_accuracy": 0.17561201453208924,
      "num_tokens": 6649973.0,
      "step": 2980
    },
    {
      "entropy": 5.782076692581176,
      "epoch": 0.33526141405065424,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004994362419508478,
      "loss": 5.6531,
      "mean_token_accuracy": 0.17298476547002792,
      "num_tokens": 6660625.0,
      "step": 2985
    },
    {
      "entropy": 5.8877051830291744,
      "epoch": 0.3358229909586118,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00049943339804258,
      "loss": 5.7379,
      "mean_token_accuracy": 0.16518658101558686,
      "num_tokens": 6673191.0,
      "step": 2990
    },
    {
      "entropy": 5.85368161201477,
      "epoch": 0.33638456786656934,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004994305469882657,
      "loss": 5.7308,
      "mean_token_accuracy": 0.17128600776195527,
      "num_tokens": 6683993.0,
      "step": 2995
    },
    {
      "entropy": 5.846784543991089,
      "epoch": 0.3369461447745269,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004994276887879957,
      "loss": 5.7637,
      "mean_token_accuracy": 0.17115221917629242,
      "num_tokens": 6694881.0,
      "step": 3000
    },
    {
      "epoch": 0.3369461447745269,
      "eval_entropy": 5.67483484951036,
      "eval_loss": 5.763563632965088,
      "eval_mean_token_accuracy": 0.17509971401080784,
      "eval_num_tokens": 6694881.0,
      "eval_runtime": 18.4605,
      "eval_samples_per_second": 1679.968,
      "eval_steps_per_second": 210.016,
      "step": 3000
    },
    {
      "entropy": 5.875175809860229,
      "epoch": 0.33750772168248444,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004994248234418611,
      "loss": 5.7534,
      "mean_token_accuracy": 0.17117161750793458,
      "num_tokens": 6706648.0,
      "step": 3005
    },
    {
      "entropy": 5.860319662094116,
      "epoch": 0.33806929859044194,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499421950949953,
      "loss": 5.7721,
      "mean_token_accuracy": 0.1703120931982994,
      "num_tokens": 6716858.0,
      "step": 3010
    },
    {
      "entropy": 5.823848390579224,
      "epoch": 0.3386308754983995,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499419071312363,
      "loss": 5.7206,
      "mean_token_accuracy": 0.17155161648988723,
      "num_tokens": 6726558.0,
      "step": 3015
    },
    {
      "entropy": 5.874156332015991,
      "epoch": 0.33919245240635704,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994161845291828,
      "loss": 5.7208,
      "mean_token_accuracy": 0.16561077386140824,
      "num_tokens": 6737819.0,
      "step": 3020
    },
    {
      "entropy": 5.815466213226318,
      "epoch": 0.3397540293143146,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004994132906005041,
      "loss": 5.6749,
      "mean_token_accuracy": 0.16978734135627746,
      "num_tokens": 6748593.0,
      "step": 3025
    },
    {
      "entropy": 5.768634176254272,
      "epoch": 0.34031560622227214,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004994103895264193,
      "loss": 5.7696,
      "mean_token_accuracy": 0.1597857430577278,
      "num_tokens": 6760631.0,
      "step": 3030
    },
    {
      "entropy": 5.9401463031768795,
      "epoch": 0.3408771831302297,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004994074813070207,
      "loss": 5.6809,
      "mean_token_accuracy": 0.16422876566648484,
      "num_tokens": 6772318.0,
      "step": 3035
    },
    {
      "entropy": 5.749310636520386,
      "epoch": 0.34143876003818724,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004994045659424007,
      "loss": 5.7261,
      "mean_token_accuracy": 0.1704739198088646,
      "num_tokens": 6782684.0,
      "step": 3040
    },
    {
      "entropy": 5.868061590194702,
      "epoch": 0.3420003369461448,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004994016434326524,
      "loss": 5.5936,
      "mean_token_accuracy": 0.17995783537626267,
      "num_tokens": 6793895.0,
      "step": 3045
    },
    {
      "entropy": 5.702157974243164,
      "epoch": 0.34256191385410234,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004993987137778687,
      "loss": 5.6668,
      "mean_token_accuracy": 0.17065583020448685,
      "num_tokens": 6804794.0,
      "step": 3050
    },
    {
      "entropy": 5.9270241260528564,
      "epoch": 0.3431234907620599,
      "grad_norm": 0.875,
      "learning_rate": 0.000499395776978143,
      "loss": 5.7889,
      "mean_token_accuracy": 0.16845911741256714,
      "num_tokens": 6815976.0,
      "step": 3055
    },
    {
      "entropy": 5.864881563186645,
      "epoch": 0.3436850676700174,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004993928330335686,
      "loss": 5.7659,
      "mean_token_accuracy": 0.1645062267780304,
      "num_tokens": 6828027.0,
      "step": 3060
    },
    {
      "entropy": 5.8461141109466555,
      "epoch": 0.34424664457797494,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004993898819442395,
      "loss": 5.7456,
      "mean_token_accuracy": 0.16880060285329818,
      "num_tokens": 6839195.0,
      "step": 3065
    },
    {
      "entropy": 5.71008448600769,
      "epoch": 0.3448082214859325,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004993869237102494,
      "loss": 5.6111,
      "mean_token_accuracy": 0.17087177485227584,
      "num_tokens": 6851731.0,
      "step": 3070
    },
    {
      "entropy": 5.900761651992798,
      "epoch": 0.34536979839389004,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004993839583316927,
      "loss": 5.654,
      "mean_token_accuracy": 0.17333684116601944,
      "num_tokens": 6862916.0,
      "step": 3075
    },
    {
      "entropy": 5.906678771972656,
      "epoch": 0.3459313753018476,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004993809858086638,
      "loss": 5.8352,
      "mean_token_accuracy": 0.16391761153936385,
      "num_tokens": 6876003.0,
      "step": 3080
    },
    {
      "entropy": 5.827907466888428,
      "epoch": 0.34649295220980514,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004993780061412571,
      "loss": 5.81,
      "mean_token_accuracy": 0.16843279749155043,
      "num_tokens": 6886040.0,
      "step": 3085
    },
    {
      "entropy": 5.860449409484863,
      "epoch": 0.3470545291177627,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004993750193295678,
      "loss": 5.726,
      "mean_token_accuracy": 0.17591039836406708,
      "num_tokens": 6896758.0,
      "step": 3090
    },
    {
      "entropy": 5.86633791923523,
      "epoch": 0.34761610602572024,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004993720253736909,
      "loss": 5.7568,
      "mean_token_accuracy": 0.17040549218654633,
      "num_tokens": 6907287.0,
      "step": 3095
    },
    {
      "entropy": 5.875124502182007,
      "epoch": 0.3481776829336778,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004993690242737216,
      "loss": 5.7851,
      "mean_token_accuracy": 0.1637913167476654,
      "num_tokens": 6918992.0,
      "step": 3100
    },
    {
      "entropy": 5.905808973312378,
      "epoch": 0.3487392598416353,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004993660160297555,
      "loss": 5.7882,
      "mean_token_accuracy": 0.1670123279094696,
      "num_tokens": 6931493.0,
      "step": 3105
    },
    {
      "entropy": 5.829394721984864,
      "epoch": 0.34930083674959284,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004993630006418885,
      "loss": 5.6375,
      "mean_token_accuracy": 0.17146755903959274,
      "num_tokens": 6941807.0,
      "step": 3110
    },
    {
      "entropy": 5.915087652206421,
      "epoch": 0.3498624136575504,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004993599781102166,
      "loss": 5.8224,
      "mean_token_accuracy": 0.16491484344005586,
      "num_tokens": 6953436.0,
      "step": 3115
    },
    {
      "entropy": 6.0053098678588865,
      "epoch": 0.35042399056550794,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993569484348359,
      "loss": 5.9436,
      "mean_token_accuracy": 0.16106159016489982,
      "num_tokens": 6965247.0,
      "step": 3120
    },
    {
      "entropy": 5.80642032623291,
      "epoch": 0.3509855674734655,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004993539116158429,
      "loss": 5.725,
      "mean_token_accuracy": 0.17518459707498552,
      "num_tokens": 6976605.0,
      "step": 3125
    },
    {
      "entropy": 5.791266775131225,
      "epoch": 0.35154714438142304,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004993508676533346,
      "loss": 5.6675,
      "mean_token_accuracy": 0.18179810494184495,
      "num_tokens": 6986205.0,
      "step": 3130
    },
    {
      "entropy": 5.788539028167724,
      "epoch": 0.3521087212893806,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004993478165474074,
      "loss": 5.6471,
      "mean_token_accuracy": 0.17194326221942902,
      "num_tokens": 6997699.0,
      "step": 3135
    },
    {
      "entropy": 5.7256981372833256,
      "epoch": 0.35267029819733814,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004993447582981588,
      "loss": 5.6691,
      "mean_token_accuracy": 0.16925414353609086,
      "num_tokens": 7008968.0,
      "step": 3140
    },
    {
      "entropy": 5.9573891162872314,
      "epoch": 0.3532318751052957,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499341692905686,
      "loss": 5.6991,
      "mean_token_accuracy": 0.16565983295440673,
      "num_tokens": 7020633.0,
      "step": 3145
    },
    {
      "entropy": 5.7820827007293705,
      "epoch": 0.35379345201325324,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004993386203700868,
      "loss": 5.6082,
      "mean_token_accuracy": 0.17894810289144517,
      "num_tokens": 7031187.0,
      "step": 3150
    },
    {
      "entropy": 5.832602310180664,
      "epoch": 0.35435502892121074,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004993355406914589,
      "loss": 5.8346,
      "mean_token_accuracy": 0.16269235461950302,
      "num_tokens": 7042662.0,
      "step": 3155
    },
    {
      "entropy": 5.915209007263184,
      "epoch": 0.3549166058291683,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004993324538699003,
      "loss": 5.8188,
      "mean_token_accuracy": 0.17167404145002366,
      "num_tokens": 7054297.0,
      "step": 3160
    },
    {
      "entropy": 5.9110935688018795,
      "epoch": 0.35547818273712584,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004993293599055094,
      "loss": 5.6766,
      "mean_token_accuracy": 0.17038068771362305,
      "num_tokens": 7064961.0,
      "step": 3165
    },
    {
      "entropy": 5.87855715751648,
      "epoch": 0.3560397596450834,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004993262587983848,
      "loss": 5.8245,
      "mean_token_accuracy": 0.16553898453712462,
      "num_tokens": 7076857.0,
      "step": 3170
    },
    {
      "entropy": 5.852333307266235,
      "epoch": 0.35660133655304094,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499323150548625,
      "loss": 5.6905,
      "mean_token_accuracy": 0.17079026848077775,
      "num_tokens": 7088926.0,
      "step": 3175
    },
    {
      "entropy": 5.737040853500366,
      "epoch": 0.3571629134609985,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004993200351563291,
      "loss": 5.6139,
      "mean_token_accuracy": 0.1790690004825592,
      "num_tokens": 7099478.0,
      "step": 3180
    },
    {
      "entropy": 5.782344150543213,
      "epoch": 0.35772449036895604,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004993169126215962,
      "loss": 5.6644,
      "mean_token_accuracy": 0.17297754138708116,
      "num_tokens": 7111776.0,
      "step": 3185
    },
    {
      "entropy": 5.957902288436889,
      "epoch": 0.3582860672769136,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004993137829445259,
      "loss": 5.7537,
      "mean_token_accuracy": 0.17062186896800996,
      "num_tokens": 7123084.0,
      "step": 3190
    },
    {
      "entropy": 5.793644762039184,
      "epoch": 0.35884764418487114,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004993106461252177,
      "loss": 5.7197,
      "mean_token_accuracy": 0.17108529657125474,
      "num_tokens": 7134973.0,
      "step": 3195
    },
    {
      "entropy": 5.753469848632813,
      "epoch": 0.3594092210928287,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004993075021637716,
      "loss": 5.7386,
      "mean_token_accuracy": 0.16747153848409652,
      "num_tokens": 7147227.0,
      "step": 3200
    },
    {
      "entropy": 5.9279557228088375,
      "epoch": 0.3599707980007862,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004993043510602875,
      "loss": 5.7355,
      "mean_token_accuracy": 0.16783691793680192,
      "num_tokens": 7158201.0,
      "step": 3205
    },
    {
      "entropy": 5.787822914123535,
      "epoch": 0.36053237490874374,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000499301192814866,
      "loss": 5.6274,
      "mean_token_accuracy": 0.17954721450805664,
      "num_tokens": 7168909.0,
      "step": 3210
    },
    {
      "entropy": 5.87915940284729,
      "epoch": 0.3610939518167013,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004992980274276074,
      "loss": 5.7837,
      "mean_token_accuracy": 0.16720784157514573,
      "num_tokens": 7179877.0,
      "step": 3215
    },
    {
      "entropy": 5.808961343765259,
      "epoch": 0.36165552872465884,
      "grad_norm": 1.0,
      "learning_rate": 0.0004992948548986128,
      "loss": 5.7001,
      "mean_token_accuracy": 0.17544217705726622,
      "num_tokens": 7190823.0,
      "step": 3220
    },
    {
      "entropy": 5.812773990631103,
      "epoch": 0.3622171056326164,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004992916752279829,
      "loss": 5.6393,
      "mean_token_accuracy": 0.1736850306391716,
      "num_tokens": 7201700.0,
      "step": 3225
    },
    {
      "entropy": 5.819822072982788,
      "epoch": 0.36277868254057394,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000499288488415819,
      "loss": 5.6785,
      "mean_token_accuracy": 0.1758155643939972,
      "num_tokens": 7213618.0,
      "step": 3230
    },
    {
      "entropy": 5.923905992507935,
      "epoch": 0.3633402594485315,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004992852944622229,
      "loss": 5.8616,
      "mean_token_accuracy": 0.1677390828728676,
      "num_tokens": 7226207.0,
      "step": 3235
    },
    {
      "entropy": 5.804302787780761,
      "epoch": 0.36390183635648904,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004992820933672958,
      "loss": 5.708,
      "mean_token_accuracy": 0.17419291138648987,
      "num_tokens": 7237545.0,
      "step": 3240
    },
    {
      "entropy": 5.774352645874023,
      "epoch": 0.3644634132644466,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004992788851311399,
      "loss": 5.6843,
      "mean_token_accuracy": 0.17325382083654403,
      "num_tokens": 7247799.0,
      "step": 3245
    },
    {
      "entropy": 5.793538188934326,
      "epoch": 0.3650249901724041,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004992756697538574,
      "loss": 5.6709,
      "mean_token_accuracy": 0.16672472804784774,
      "num_tokens": 7260229.0,
      "step": 3250
    },
    {
      "entropy": 5.85179386138916,
      "epoch": 0.36558656708036164,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004992724472355505,
      "loss": 5.7755,
      "mean_token_accuracy": 0.16767250150442123,
      "num_tokens": 7271457.0,
      "step": 3255
    },
    {
      "entropy": 5.798571062088013,
      "epoch": 0.3661481439883192,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004992692175763219,
      "loss": 5.6623,
      "mean_token_accuracy": 0.1711890771985054,
      "num_tokens": 7282124.0,
      "step": 3260
    },
    {
      "entropy": 5.885174179077149,
      "epoch": 0.36670972089627674,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004992659807762745,
      "loss": 5.7771,
      "mean_token_accuracy": 0.1681201785802841,
      "num_tokens": 7292981.0,
      "step": 3265
    },
    {
      "entropy": 6.040387916564941,
      "epoch": 0.3672712978042343,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004992627368355112,
      "loss": 5.9261,
      "mean_token_accuracy": 0.1669810563325882,
      "num_tokens": 7304591.0,
      "step": 3270
    },
    {
      "entropy": 5.865202951431274,
      "epoch": 0.36783287471219184,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004992594857541355,
      "loss": 5.6676,
      "mean_token_accuracy": 0.17515869438648224,
      "num_tokens": 7315918.0,
      "step": 3275
    },
    {
      "entropy": 5.827836322784424,
      "epoch": 0.3683944516201494,
      "grad_norm": 0.875,
      "learning_rate": 0.0004992562275322506,
      "loss": 5.774,
      "mean_token_accuracy": 0.16934425830841066,
      "num_tokens": 7327725.0,
      "step": 3280
    },
    {
      "entropy": 5.960494947433472,
      "epoch": 0.36895602852810694,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004992529621699606,
      "loss": 5.7771,
      "mean_token_accuracy": 0.1653957635164261,
      "num_tokens": 7339787.0,
      "step": 3285
    },
    {
      "entropy": 5.870577096939087,
      "epoch": 0.3695176054360645,
      "grad_norm": 0.875,
      "learning_rate": 0.0004992496896673692,
      "loss": 5.7626,
      "mean_token_accuracy": 0.16915116757154464,
      "num_tokens": 7351284.0,
      "step": 3290
    },
    {
      "entropy": 5.762804365158081,
      "epoch": 0.37007918234402204,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004992464100245809,
      "loss": 5.5773,
      "mean_token_accuracy": 0.17599964439868926,
      "num_tokens": 7362105.0,
      "step": 3295
    },
    {
      "entropy": 5.795164585113525,
      "epoch": 0.37064075925197953,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004992431232416997,
      "loss": 5.6802,
      "mean_token_accuracy": 0.17545291483402253,
      "num_tokens": 7375204.0,
      "step": 3300
    },
    {
      "entropy": 5.989410543441773,
      "epoch": 0.3712023361599371,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004992398293188307,
      "loss": 5.8362,
      "mean_token_accuracy": 0.16860273480415344,
      "num_tokens": 7385853.0,
      "step": 3305
    },
    {
      "entropy": 5.787294244766235,
      "epoch": 0.37176391306789464,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004992365282560784,
      "loss": 5.6916,
      "mean_token_accuracy": 0.16280674636363984,
      "num_tokens": 7397459.0,
      "step": 3310
    },
    {
      "entropy": 5.812543725967407,
      "epoch": 0.3723254899758522,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992332200535483,
      "loss": 5.6792,
      "mean_token_accuracy": 0.16694144606590272,
      "num_tokens": 7408996.0,
      "step": 3315
    },
    {
      "entropy": 5.887885236740113,
      "epoch": 0.37288706688380974,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004992299047113452,
      "loss": 5.7823,
      "mean_token_accuracy": 0.16049795746803283,
      "num_tokens": 7419938.0,
      "step": 3320
    },
    {
      "entropy": 5.880823612213135,
      "epoch": 0.3734486437917673,
      "grad_norm": 0.875,
      "learning_rate": 0.0004992265822295752,
      "loss": 5.7849,
      "mean_token_accuracy": 0.165161195397377,
      "num_tokens": 7430181.0,
      "step": 3325
    },
    {
      "entropy": 5.8815710067749025,
      "epoch": 0.37401022069972484,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004992232526083438,
      "loss": 5.7325,
      "mean_token_accuracy": 0.17060090452432633,
      "num_tokens": 7440774.0,
      "step": 3330
    },
    {
      "entropy": 5.858735418319702,
      "epoch": 0.3745717976076824,
      "grad_norm": 0.828125,
      "learning_rate": 0.000499219915847757,
      "loss": 5.6612,
      "mean_token_accuracy": 0.17302175462245942,
      "num_tokens": 7451744.0,
      "step": 3335
    },
    {
      "entropy": 5.715199279785156,
      "epoch": 0.37513337451563994,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004992165719479211,
      "loss": 5.7042,
      "mean_token_accuracy": 0.1720335766673088,
      "num_tokens": 7463731.0,
      "step": 3340
    },
    {
      "entropy": 5.869427442550659,
      "epoch": 0.37569495142359743,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004992132209089427,
      "loss": 5.6973,
      "mean_token_accuracy": 0.17209299504756928,
      "num_tokens": 7474392.0,
      "step": 3345
    },
    {
      "entropy": 5.7452164649963375,
      "epoch": 0.376256528331555,
      "grad_norm": 0.875,
      "learning_rate": 0.0004992098627309283,
      "loss": 5.626,
      "mean_token_accuracy": 0.1825716570019722,
      "num_tokens": 7484860.0,
      "step": 3350
    },
    {
      "entropy": 5.778187274932861,
      "epoch": 0.37681810523951254,
      "grad_norm": 0.890625,
      "learning_rate": 0.000499206497413985,
      "loss": 5.6684,
      "mean_token_accuracy": 0.17513581067323686,
      "num_tokens": 7495877.0,
      "step": 3355
    },
    {
      "entropy": 5.856958818435669,
      "epoch": 0.3773796821474701,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004992031249582197,
      "loss": 5.7367,
      "mean_token_accuracy": 0.17677867114543916,
      "num_tokens": 7506879.0,
      "step": 3360
    },
    {
      "entropy": 5.777524995803833,
      "epoch": 0.37794125905542764,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004991997453637401,
      "loss": 5.5729,
      "mean_token_accuracy": 0.17503220438957215,
      "num_tokens": 7517405.0,
      "step": 3365
    },
    {
      "entropy": 5.763284540176391,
      "epoch": 0.3785028359633852,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004991963586306536,
      "loss": 5.6492,
      "mean_token_accuracy": 0.17201928347349166,
      "num_tokens": 7528832.0,
      "step": 3370
    },
    {
      "entropy": 5.818809223175049,
      "epoch": 0.37906441287134274,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004991929647590682,
      "loss": 5.6242,
      "mean_token_accuracy": 0.17255856096744537,
      "num_tokens": 7541048.0,
      "step": 3375
    },
    {
      "entropy": 5.834233093261719,
      "epoch": 0.3796259897793003,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004991895637490917,
      "loss": 5.6547,
      "mean_token_accuracy": 0.17642329931259154,
      "num_tokens": 7551849.0,
      "step": 3380
    },
    {
      "entropy": 5.7006148338317875,
      "epoch": 0.38018756668725784,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004991861556008326,
      "loss": 5.6741,
      "mean_token_accuracy": 0.1750218689441681,
      "num_tokens": 7563148.0,
      "step": 3385
    },
    {
      "entropy": 5.93495659828186,
      "epoch": 0.3807491435952154,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004991827403143992,
      "loss": 5.8557,
      "mean_token_accuracy": 0.16593129485845565,
      "num_tokens": 7575119.0,
      "step": 3390
    },
    {
      "entropy": 5.93589768409729,
      "epoch": 0.3813107205031729,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004991793178899006,
      "loss": 5.7073,
      "mean_token_accuracy": 0.17126083225011826,
      "num_tokens": 7585899.0,
      "step": 3395
    },
    {
      "entropy": 5.836880254745483,
      "epoch": 0.38187229741113043,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004991758883274455,
      "loss": 5.6039,
      "mean_token_accuracy": 0.17948321551084517,
      "num_tokens": 7598938.0,
      "step": 3400
    },
    {
      "entropy": 5.653288316726685,
      "epoch": 0.382433874319088,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004991724516271432,
      "loss": 5.6793,
      "mean_token_accuracy": 0.1676811695098877,
      "num_tokens": 7610087.0,
      "step": 3405
    },
    {
      "entropy": 5.965075922012329,
      "epoch": 0.38299545122704554,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004991690077891032,
      "loss": 5.8081,
      "mean_token_accuracy": 0.16608484089374542,
      "num_tokens": 7622186.0,
      "step": 3410
    },
    {
      "entropy": 5.877589464187622,
      "epoch": 0.3835570281350031,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004991655568134348,
      "loss": 5.752,
      "mean_token_accuracy": 0.16870230734348296,
      "num_tokens": 7633501.0,
      "step": 3415
    },
    {
      "entropy": 5.751620483398438,
      "epoch": 0.38411860504296064,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004991620987002483,
      "loss": 5.6748,
      "mean_token_accuracy": 0.17229740172624589,
      "num_tokens": 7643656.0,
      "step": 3420
    },
    {
      "entropy": 5.880866575241089,
      "epoch": 0.3846801819509182,
      "grad_norm": 0.875,
      "learning_rate": 0.0004991586334496536,
      "loss": 5.6853,
      "mean_token_accuracy": 0.17358008176088333,
      "num_tokens": 7654565.0,
      "step": 3425
    },
    {
      "entropy": 5.795616102218628,
      "epoch": 0.38524175885887574,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004991551610617611,
      "loss": 5.6088,
      "mean_token_accuracy": 0.18103413432836532,
      "num_tokens": 7664725.0,
      "step": 3430
    },
    {
      "entropy": 5.694900178909302,
      "epoch": 0.3858033357668333,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004991516815366812,
      "loss": 5.7008,
      "mean_token_accuracy": 0.1738528475165367,
      "num_tokens": 7674772.0,
      "step": 3435
    },
    {
      "entropy": 5.961912012100219,
      "epoch": 0.38636491267479084,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499148194874525,
      "loss": 5.7588,
      "mean_token_accuracy": 0.1647901713848114,
      "num_tokens": 7686488.0,
      "step": 3440
    },
    {
      "entropy": 5.830978870391846,
      "epoch": 0.38692648958274833,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004991447010754032,
      "loss": 5.7418,
      "mean_token_accuracy": 0.17187318801879883,
      "num_tokens": 7698842.0,
      "step": 3445
    },
    {
      "entropy": 5.76347599029541,
      "epoch": 0.3874880664907059,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004991412001394273,
      "loss": 5.5267,
      "mean_token_accuracy": 0.17366759181022645,
      "num_tokens": 7709255.0,
      "step": 3450
    },
    {
      "entropy": 5.6753857135772705,
      "epoch": 0.38804964339866344,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991376920667086,
      "loss": 5.5818,
      "mean_token_accuracy": 0.17516542375087737,
      "num_tokens": 7720182.0,
      "step": 3455
    },
    {
      "entropy": 5.744015836715699,
      "epoch": 0.388611220306621,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004991341768573589,
      "loss": 5.6227,
      "mean_token_accuracy": 0.18164215385913848,
      "num_tokens": 7730944.0,
      "step": 3460
    },
    {
      "entropy": 5.835626792907715,
      "epoch": 0.38917279721457854,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00049913065451149,
      "loss": 5.6871,
      "mean_token_accuracy": 0.1743626058101654,
      "num_tokens": 7741686.0,
      "step": 3465
    },
    {
      "entropy": 5.803769063949585,
      "epoch": 0.3897343741225361,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004991271250292141,
      "loss": 5.6823,
      "mean_token_accuracy": 0.17172736823558807,
      "num_tokens": 7752245.0,
      "step": 3470
    },
    {
      "entropy": 5.721233749389649,
      "epoch": 0.39029595103049364,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004991235884106437,
      "loss": 5.659,
      "mean_token_accuracy": 0.16859646886587143,
      "num_tokens": 7762622.0,
      "step": 3475
    },
    {
      "entropy": 5.9113177299499515,
      "epoch": 0.3908575279384512,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004991200446558913,
      "loss": 5.7763,
      "mean_token_accuracy": 0.17350095212459565,
      "num_tokens": 7773550.0,
      "step": 3480
    },
    {
      "entropy": 5.766241502761841,
      "epoch": 0.39141910484640874,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004991164937650698,
      "loss": 5.6612,
      "mean_token_accuracy": 0.16828784644603728,
      "num_tokens": 7784988.0,
      "step": 3485
    },
    {
      "entropy": 5.8275816440582275,
      "epoch": 0.39198068175436623,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004991129357382922,
      "loss": 5.732,
      "mean_token_accuracy": 0.17502650618553162,
      "num_tokens": 7797706.0,
      "step": 3490
    },
    {
      "entropy": 5.831998872756958,
      "epoch": 0.3925422586623238,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004991093705756718,
      "loss": 5.6227,
      "mean_token_accuracy": 0.17694617360830306,
      "num_tokens": 7808960.0,
      "step": 3495
    },
    {
      "entropy": 5.709803771972656,
      "epoch": 0.39310383557028133,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004991057982773221,
      "loss": 5.6035,
      "mean_token_accuracy": 0.181922022998333,
      "num_tokens": 7819271.0,
      "step": 3500
    },
    {
      "entropy": 5.7892396450042725,
      "epoch": 0.3936654124782389,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004991022188433569,
      "loss": 5.7072,
      "mean_token_accuracy": 0.1694733813405037,
      "num_tokens": 7832257.0,
      "step": 3505
    },
    {
      "entropy": 5.911541509628296,
      "epoch": 0.39422698938619644,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00049909863227389,
      "loss": 5.7313,
      "mean_token_accuracy": 0.165920627117157,
      "num_tokens": 7842684.0,
      "step": 3510
    },
    {
      "entropy": 5.805456018447876,
      "epoch": 0.394788566294154,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004990950385690358,
      "loss": 5.7275,
      "mean_token_accuracy": 0.168575157225132,
      "num_tokens": 7855031.0,
      "step": 3515
    },
    {
      "entropy": 5.725608158111572,
      "epoch": 0.39535014320211154,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004990914377289088,
      "loss": 5.5467,
      "mean_token_accuracy": 0.17219574004411697,
      "num_tokens": 7865291.0,
      "step": 3520
    },
    {
      "entropy": 5.697873544692993,
      "epoch": 0.3959117201100691,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004990878297536233,
      "loss": 5.516,
      "mean_token_accuracy": 0.17715096175670625,
      "num_tokens": 7876401.0,
      "step": 3525
    },
    {
      "entropy": 5.7096973896026615,
      "epoch": 0.39647329701802664,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004990842146432946,
      "loss": 5.56,
      "mean_token_accuracy": 0.17544865012168884,
      "num_tokens": 7887787.0,
      "step": 3530
    },
    {
      "entropy": 5.745721673965454,
      "epoch": 0.3970348739259842,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004990805923980375,
      "loss": 5.5525,
      "mean_token_accuracy": 0.17566726952791215,
      "num_tokens": 7898550.0,
      "step": 3535
    },
    {
      "entropy": 5.739796924591064,
      "epoch": 0.3975964508339417,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004990769630179673,
      "loss": 5.7174,
      "mean_token_accuracy": 0.1733885794878006,
      "num_tokens": 7909254.0,
      "step": 3540
    },
    {
      "entropy": 5.902084589004517,
      "epoch": 0.39815802774189923,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004990733265032,
      "loss": 5.7334,
      "mean_token_accuracy": 0.17753101885318756,
      "num_tokens": 7920164.0,
      "step": 3545
    },
    {
      "entropy": 5.760942506790161,
      "epoch": 0.3987196046498568,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004990696828538508,
      "loss": 5.6809,
      "mean_token_accuracy": 0.1701009228825569,
      "num_tokens": 7931390.0,
      "step": 3550
    },
    {
      "entropy": 5.846822118759155,
      "epoch": 0.39928118155781434,
      "grad_norm": 0.828125,
      "learning_rate": 0.000499066032070036,
      "loss": 5.6915,
      "mean_token_accuracy": 0.17291721999645232,
      "num_tokens": 7943765.0,
      "step": 3555
    },
    {
      "entropy": 5.725397300720215,
      "epoch": 0.3998427584657719,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004990623741518719,
      "loss": 5.5307,
      "mean_token_accuracy": 0.17921568602323532,
      "num_tokens": 7954621.0,
      "step": 3560
    },
    {
      "entropy": 5.7358253479003904,
      "epoch": 0.40040433537372944,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000499058709099475,
      "loss": 5.7402,
      "mean_token_accuracy": 0.1680094376206398,
      "num_tokens": 7966063.0,
      "step": 3565
    },
    {
      "entropy": 5.892997169494629,
      "epoch": 0.400965912281687,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004990550369129618,
      "loss": 5.6401,
      "mean_token_accuracy": 0.17921878546476364,
      "num_tokens": 7977391.0,
      "step": 3570
    },
    {
      "entropy": 5.8006116390228275,
      "epoch": 0.40152748918964454,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004990513575924494,
      "loss": 5.7605,
      "mean_token_accuracy": 0.16944273859262465,
      "num_tokens": 7988786.0,
      "step": 3575
    },
    {
      "entropy": 5.800322484970093,
      "epoch": 0.4020890660976021,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004990476711380548,
      "loss": 5.5581,
      "mean_token_accuracy": 0.17370157539844513,
      "num_tokens": 8000848.0,
      "step": 3580
    },
    {
      "entropy": 5.7842577457427975,
      "epoch": 0.40265064300555964,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004990439775498954,
      "loss": 5.7083,
      "mean_token_accuracy": 0.16895742416381837,
      "num_tokens": 8012621.0,
      "step": 3585
    },
    {
      "entropy": 5.924812698364258,
      "epoch": 0.40321221991351713,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499040276828089,
      "loss": 5.7746,
      "mean_token_accuracy": 0.16589070111513138,
      "num_tokens": 8023599.0,
      "step": 3590
    },
    {
      "entropy": 5.685869359970093,
      "epoch": 0.4037737968214747,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004990365689727533,
      "loss": 5.5652,
      "mean_token_accuracy": 0.17779057174921037,
      "num_tokens": 8037151.0,
      "step": 3595
    },
    {
      "entropy": 5.762771415710449,
      "epoch": 0.40433537372943223,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004990328539840062,
      "loss": 5.7688,
      "mean_token_accuracy": 0.17146938890218735,
      "num_tokens": 8048606.0,
      "step": 3600
    },
    {
      "entropy": 5.743241453170777,
      "epoch": 0.4048969506373898,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004990291318619663,
      "loss": 5.5658,
      "mean_token_accuracy": 0.18698559552431107,
      "num_tokens": 8058430.0,
      "step": 3605
    },
    {
      "entropy": 5.710903930664062,
      "epoch": 0.40545852754534734,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004990254026067518,
      "loss": 5.5738,
      "mean_token_accuracy": 0.17726856023073195,
      "num_tokens": 8069554.0,
      "step": 3610
    },
    {
      "entropy": 5.852778959274292,
      "epoch": 0.4060201044533049,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004990216662184817,
      "loss": 5.724,
      "mean_token_accuracy": 0.16802725791931153,
      "num_tokens": 8082131.0,
      "step": 3615
    },
    {
      "entropy": 5.758425712585449,
      "epoch": 0.40658168136126244,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004990179226972749,
      "loss": 5.5652,
      "mean_token_accuracy": 0.17815440595149995,
      "num_tokens": 8093454.0,
      "step": 3620
    },
    {
      "entropy": 5.861598491668701,
      "epoch": 0.40714325826922,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004990141720432504,
      "loss": 5.8162,
      "mean_token_accuracy": 0.1576846346259117,
      "num_tokens": 8105502.0,
      "step": 3625
    },
    {
      "entropy": 5.796506643295288,
      "epoch": 0.40770483517717754,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004990104142565279,
      "loss": 5.6967,
      "mean_token_accuracy": 0.16818222850561143,
      "num_tokens": 8116524.0,
      "step": 3630
    },
    {
      "entropy": 5.830021142959595,
      "epoch": 0.40826641208513503,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499006649337227,
      "loss": 5.6719,
      "mean_token_accuracy": 0.17779282480478287,
      "num_tokens": 8126949.0,
      "step": 3635
    },
    {
      "entropy": 5.786426734924317,
      "epoch": 0.4088279889930926,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004990028772854674,
      "loss": 5.6246,
      "mean_token_accuracy": 0.17908453643321992,
      "num_tokens": 8137346.0,
      "step": 3640
    },
    {
      "entropy": 5.695159387588501,
      "epoch": 0.40938956590105013,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004989990981013693,
      "loss": 5.6553,
      "mean_token_accuracy": 0.16744260787963866,
      "num_tokens": 8148499.0,
      "step": 3645
    },
    {
      "entropy": 5.859077310562133,
      "epoch": 0.4099511428090077,
      "grad_norm": 1.0,
      "learning_rate": 0.000498995311785053,
      "loss": 5.6979,
      "mean_token_accuracy": 0.17695163786411286,
      "num_tokens": 8159645.0,
      "step": 3650
    },
    {
      "entropy": 5.725503015518188,
      "epoch": 0.41051271971696524,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004989915183366392,
      "loss": 5.5713,
      "mean_token_accuracy": 0.1773812249302864,
      "num_tokens": 8171033.0,
      "step": 3655
    },
    {
      "entropy": 5.7521998405456545,
      "epoch": 0.4110742966249228,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004989877177562485,
      "loss": 5.6448,
      "mean_token_accuracy": 0.17140540927648545,
      "num_tokens": 8181582.0,
      "step": 3660
    },
    {
      "entropy": 5.767964839935303,
      "epoch": 0.41163587353288034,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000498983910044002,
      "loss": 5.5475,
      "mean_token_accuracy": 0.17604551315307618,
      "num_tokens": 8192432.0,
      "step": 3665
    },
    {
      "entropy": 5.835826826095581,
      "epoch": 0.4121974504408379,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004989800952000209,
      "loss": 5.7276,
      "mean_token_accuracy": 0.17614867240190507,
      "num_tokens": 8203995.0,
      "step": 3670
    },
    {
      "entropy": 5.735976648330689,
      "epoch": 0.41275902734879544,
      "grad_norm": 0.875,
      "learning_rate": 0.0004989762732244268,
      "loss": 5.5651,
      "mean_token_accuracy": 0.17965177595615386,
      "num_tokens": 8214323.0,
      "step": 3675
    },
    {
      "entropy": 5.6943518161773685,
      "epoch": 0.413320604256753,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004989724441173413,
      "loss": 5.6338,
      "mean_token_accuracy": 0.1703054964542389,
      "num_tokens": 8226285.0,
      "step": 3680
    },
    {
      "entropy": 5.888356161117554,
      "epoch": 0.4138821811647105,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004989686078788862,
      "loss": 5.7957,
      "mean_token_accuracy": 0.16576283574104309,
      "num_tokens": 8237961.0,
      "step": 3685
    },
    {
      "entropy": 5.7234046936035154,
      "epoch": 0.41444375807266803,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004989647645091839,
      "loss": 5.5544,
      "mean_token_accuracy": 0.1777231752872467,
      "num_tokens": 8248463.0,
      "step": 3690
    },
    {
      "entropy": 5.77109923362732,
      "epoch": 0.4150053349806256,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004989609140083565,
      "loss": 5.6881,
      "mean_token_accuracy": 0.1743547424674034,
      "num_tokens": 8260342.0,
      "step": 3695
    },
    {
      "entropy": 5.835375022888184,
      "epoch": 0.41556691188858313,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004989570563765269,
      "loss": 5.7808,
      "mean_token_accuracy": 0.17079318016767503,
      "num_tokens": 8272055.0,
      "step": 3700
    },
    {
      "entropy": 5.932144737243652,
      "epoch": 0.4161284887965407,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004989531916138176,
      "loss": 5.7344,
      "mean_token_accuracy": 0.17019769847393035,
      "num_tokens": 8282129.0,
      "step": 3705
    },
    {
      "entropy": 5.836971998214722,
      "epoch": 0.41669006570449824,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004989493197203519,
      "loss": 5.7324,
      "mean_token_accuracy": 0.170888352394104,
      "num_tokens": 8294663.0,
      "step": 3710
    },
    {
      "entropy": 5.749900007247925,
      "epoch": 0.4172516426124558,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000498945440696253,
      "loss": 5.547,
      "mean_token_accuracy": 0.18026671707630157,
      "num_tokens": 8305882.0,
      "step": 3715
    },
    {
      "entropy": 5.718420171737671,
      "epoch": 0.41781321952041334,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004989415545416444,
      "loss": 5.6551,
      "mean_token_accuracy": 0.17579069584608079,
      "num_tokens": 8316604.0,
      "step": 3720
    },
    {
      "entropy": 5.898751163482666,
      "epoch": 0.4183747964283709,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004989376612566498,
      "loss": 5.7414,
      "mean_token_accuracy": 0.16762872338294982,
      "num_tokens": 8328126.0,
      "step": 3725
    },
    {
      "entropy": 5.744472789764404,
      "epoch": 0.4189363733363284,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004989337608413933,
      "loss": 5.6581,
      "mean_token_accuracy": 0.17621897161006927,
      "num_tokens": 8339050.0,
      "step": 3730
    },
    {
      "entropy": 5.707731151580811,
      "epoch": 0.41949795024428593,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004989298532959989,
      "loss": 5.4984,
      "mean_token_accuracy": 0.1825895607471466,
      "num_tokens": 8350216.0,
      "step": 3735
    },
    {
      "entropy": 5.771092891693115,
      "epoch": 0.4200595271522435,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004989259386205912,
      "loss": 5.6545,
      "mean_token_accuracy": 0.1750851973891258,
      "num_tokens": 8362311.0,
      "step": 3740
    },
    {
      "entropy": 5.860318231582641,
      "epoch": 0.42062110406020103,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004989220168152948,
      "loss": 5.8135,
      "mean_token_accuracy": 0.1725629597902298,
      "num_tokens": 8373751.0,
      "step": 3745
    },
    {
      "entropy": 5.8404439926147464,
      "epoch": 0.4211826809681586,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004989180878802346,
      "loss": 5.6464,
      "mean_token_accuracy": 0.17549248933792114,
      "num_tokens": 8384932.0,
      "step": 3750
    },
    {
      "entropy": 5.704224920272827,
      "epoch": 0.42174425787611614,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004989141518155354,
      "loss": 5.6048,
      "mean_token_accuracy": 0.17422576248645782,
      "num_tokens": 8394973.0,
      "step": 3755
    },
    {
      "entropy": 5.786598777770996,
      "epoch": 0.4223058347840737,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004989102086213229,
      "loss": 5.6805,
      "mean_token_accuracy": 0.17264141142368317,
      "num_tokens": 8406590.0,
      "step": 3760
    },
    {
      "entropy": 5.745992517471313,
      "epoch": 0.42286741169203124,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004989062582977225,
      "loss": 5.6136,
      "mean_token_accuracy": 0.1738652318716049,
      "num_tokens": 8417127.0,
      "step": 3765
    },
    {
      "entropy": 5.810067939758301,
      "epoch": 0.4234289885999888,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004989023008448599,
      "loss": 5.7003,
      "mean_token_accuracy": 0.17736413925886155,
      "num_tokens": 8429442.0,
      "step": 3770
    },
    {
      "entropy": 5.7441330909729,
      "epoch": 0.42399056550794634,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004988983362628614,
      "loss": 5.6321,
      "mean_token_accuracy": 0.174859656393528,
      "num_tokens": 8439375.0,
      "step": 3775
    },
    {
      "entropy": 5.743936538696289,
      "epoch": 0.42455214241590383,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004988943645518528,
      "loss": 5.6393,
      "mean_token_accuracy": 0.175520221889019,
      "num_tokens": 8449817.0,
      "step": 3780
    },
    {
      "entropy": 5.709410619735718,
      "epoch": 0.4251137193238614,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004988903857119611,
      "loss": 5.5375,
      "mean_token_accuracy": 0.17987922430038453,
      "num_tokens": 8460096.0,
      "step": 3785
    },
    {
      "entropy": 5.787642288208008,
      "epoch": 0.42567529623181893,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004988863997433125,
      "loss": 5.6421,
      "mean_token_accuracy": 0.17529507577419282,
      "num_tokens": 8470241.0,
      "step": 3790
    },
    {
      "entropy": 5.787087154388428,
      "epoch": 0.4262368731397765,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004988824066460341,
      "loss": 5.6727,
      "mean_token_accuracy": 0.16719111651182175,
      "num_tokens": 8481438.0,
      "step": 3795
    },
    {
      "entropy": 5.83219690322876,
      "epoch": 0.42679845004773403,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000498878406420253,
      "loss": 5.7146,
      "mean_token_accuracy": 0.169234099984169,
      "num_tokens": 8492588.0,
      "step": 3800
    },
    {
      "entropy": 5.74571533203125,
      "epoch": 0.4273600269556916,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004988743990660967,
      "loss": 5.6422,
      "mean_token_accuracy": 0.18158345967531203,
      "num_tokens": 8502298.0,
      "step": 3805
    },
    {
      "entropy": 5.6950747013092045,
      "epoch": 0.42792160386364914,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004988703845836927,
      "loss": 5.4966,
      "mean_token_accuracy": 0.17507661432027816,
      "num_tokens": 8513507.0,
      "step": 3810
    },
    {
      "entropy": 5.781854581832886,
      "epoch": 0.4284831807716067,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004988663629731687,
      "loss": 5.593,
      "mean_token_accuracy": 0.18122898489236833,
      "num_tokens": 8523888.0,
      "step": 3815
    },
    {
      "entropy": 5.767497730255127,
      "epoch": 0.42904475767956424,
      "grad_norm": 0.984375,
      "learning_rate": 0.000498862334234653,
      "loss": 5.5472,
      "mean_token_accuracy": 0.18372734934091567,
      "num_tokens": 8534032.0,
      "step": 3820
    },
    {
      "entropy": 5.699436330795288,
      "epoch": 0.4296063345875218,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004988582983682737,
      "loss": 5.6755,
      "mean_token_accuracy": 0.17637507766485214,
      "num_tokens": 8545544.0,
      "step": 3825
    },
    {
      "entropy": 5.766273307800293,
      "epoch": 0.4301679114954793,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004988542553741594,
      "loss": 5.6444,
      "mean_token_accuracy": 0.175729963183403,
      "num_tokens": 8555428.0,
      "step": 3830
    },
    {
      "entropy": 5.707126522064209,
      "epoch": 0.43072948840343683,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004988502052524388,
      "loss": 5.5462,
      "mean_token_accuracy": 0.17771149128675462,
      "num_tokens": 8566456.0,
      "step": 3835
    },
    {
      "entropy": 5.716542482376099,
      "epoch": 0.4312910653113944,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004988461480032408,
      "loss": 5.541,
      "mean_token_accuracy": 0.18008371144533158,
      "num_tokens": 8578125.0,
      "step": 3840
    },
    {
      "entropy": 5.762414979934692,
      "epoch": 0.43185264221935193,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004988420836266946,
      "loss": 5.6128,
      "mean_token_accuracy": 0.16972145438194275,
      "num_tokens": 8589530.0,
      "step": 3845
    },
    {
      "entropy": 5.798019933700561,
      "epoch": 0.4324142191273095,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004988380121229298,
      "loss": 5.7228,
      "mean_token_accuracy": 0.17366854399442672,
      "num_tokens": 8601655.0,
      "step": 3850
    },
    {
      "entropy": 5.7458666324615475,
      "epoch": 0.43297579603526704,
      "grad_norm": 0.875,
      "learning_rate": 0.0004988339334920757,
      "loss": 5.633,
      "mean_token_accuracy": 0.17678244262933732,
      "num_tokens": 8613355.0,
      "step": 3855
    },
    {
      "entropy": 5.736477994918824,
      "epoch": 0.4335373729432246,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004988298477342625,
      "loss": 5.572,
      "mean_token_accuracy": 0.1793879896402359,
      "num_tokens": 8624154.0,
      "step": 3860
    },
    {
      "entropy": 5.742373609542847,
      "epoch": 0.43409894985118214,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00049882575484962,
      "loss": 5.6516,
      "mean_token_accuracy": 0.17551479935646058,
      "num_tokens": 8633888.0,
      "step": 3865
    },
    {
      "entropy": 5.810504198074341,
      "epoch": 0.4346605267591397,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004988216548382787,
      "loss": 5.627,
      "mean_token_accuracy": 0.17445796430110933,
      "num_tokens": 8644965.0,
      "step": 3870
    },
    {
      "entropy": 5.759542989730835,
      "epoch": 0.4352221036670972,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004988175477003691,
      "loss": 5.5735,
      "mean_token_accuracy": 0.17949347347021102,
      "num_tokens": 8655483.0,
      "step": 3875
    },
    {
      "entropy": 5.740917587280274,
      "epoch": 0.43578368057505473,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000498813433436022,
      "loss": 5.5556,
      "mean_token_accuracy": 0.17531621158123017,
      "num_tokens": 8666443.0,
      "step": 3880
    },
    {
      "entropy": 5.802572917938233,
      "epoch": 0.4363452574830123,
      "grad_norm": 1.0,
      "learning_rate": 0.0004988093120453684,
      "loss": 5.6173,
      "mean_token_accuracy": 0.18023148626089097,
      "num_tokens": 8677198.0,
      "step": 3885
    },
    {
      "entropy": 5.7109557628631595,
      "epoch": 0.43690683439096983,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004988051835285394,
      "loss": 5.6753,
      "mean_token_accuracy": 0.170777590572834,
      "num_tokens": 8688303.0,
      "step": 3890
    },
    {
      "entropy": 5.801574659347534,
      "epoch": 0.4374684112989274,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004988010478856667,
      "loss": 5.622,
      "mean_token_accuracy": 0.1717620387673378,
      "num_tokens": 8700326.0,
      "step": 3895
    },
    {
      "entropy": 5.655847024917603,
      "epoch": 0.43802998820688493,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004987969051168816,
      "loss": 5.583,
      "mean_token_accuracy": 0.17746342718601227,
      "num_tokens": 8711101.0,
      "step": 3900
    },
    {
      "entropy": 5.7517682075500485,
      "epoch": 0.4385915651148425,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004987927552223165,
      "loss": 5.7132,
      "mean_token_accuracy": 0.1727053388953209,
      "num_tokens": 8721751.0,
      "step": 3905
    },
    {
      "entropy": 5.84909954071045,
      "epoch": 0.43915314202280004,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004987885982021032,
      "loss": 5.8075,
      "mean_token_accuracy": 0.16774704456329345,
      "num_tokens": 8733550.0,
      "step": 3910
    },
    {
      "entropy": 5.786842679977417,
      "epoch": 0.4397147189307576,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004987844340563741,
      "loss": 5.5948,
      "mean_token_accuracy": 0.17783285081386566,
      "num_tokens": 8744218.0,
      "step": 3915
    },
    {
      "entropy": 5.718330192565918,
      "epoch": 0.44027629583871514,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004987802627852619,
      "loss": 5.6256,
      "mean_token_accuracy": 0.17466724067926406,
      "num_tokens": 8756531.0,
      "step": 3920
    },
    {
      "entropy": 5.917486333847046,
      "epoch": 0.44083787274667263,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004987760843888992,
      "loss": 5.6763,
      "mean_token_accuracy": 0.1780080035328865,
      "num_tokens": 8768114.0,
      "step": 3925
    },
    {
      "entropy": 5.719629764556885,
      "epoch": 0.4413994496546302,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004987718988674193,
      "loss": 5.6774,
      "mean_token_accuracy": 0.17154836356639863,
      "num_tokens": 8780010.0,
      "step": 3930
    },
    {
      "entropy": 5.801930952072143,
      "epoch": 0.44196102656258773,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004987677062209552,
      "loss": 5.6291,
      "mean_token_accuracy": 0.16851936429738998,
      "num_tokens": 8792183.0,
      "step": 3935
    },
    {
      "entropy": 5.818271064758301,
      "epoch": 0.4425226034705453,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004987635064496407,
      "loss": 5.7151,
      "mean_token_accuracy": 0.17167862504720688,
      "num_tokens": 8804513.0,
      "step": 3940
    },
    {
      "entropy": 5.7347640037536625,
      "epoch": 0.44308418037850283,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004987592995536093,
      "loss": 5.6189,
      "mean_token_accuracy": 0.17910695523023606,
      "num_tokens": 8814918.0,
      "step": 3945
    },
    {
      "entropy": 5.744248485565185,
      "epoch": 0.4436457572864604,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004987550855329951,
      "loss": 5.5629,
      "mean_token_accuracy": 0.17823148667812347,
      "num_tokens": 8826176.0,
      "step": 3950
    },
    {
      "entropy": 5.649262762069702,
      "epoch": 0.44420733419441794,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004987508643879322,
      "loss": 5.6038,
      "mean_token_accuracy": 0.1767020508646965,
      "num_tokens": 8837242.0,
      "step": 3955
    },
    {
      "entropy": 5.837818336486817,
      "epoch": 0.4447689111023755,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004987466361185549,
      "loss": 5.5786,
      "mean_token_accuracy": 0.1753518432378769,
      "num_tokens": 8848972.0,
      "step": 3960
    },
    {
      "entropy": 5.650309371948242,
      "epoch": 0.44533048801033304,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004987424007249979,
      "loss": 5.5458,
      "mean_token_accuracy": 0.17371795773506166,
      "num_tokens": 8861108.0,
      "step": 3965
    },
    {
      "entropy": 5.754533195495606,
      "epoch": 0.4458920649182906,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004987381582073962,
      "loss": 5.6485,
      "mean_token_accuracy": 0.1748553916811943,
      "num_tokens": 8872161.0,
      "step": 3970
    },
    {
      "entropy": 5.947879314422607,
      "epoch": 0.4464536418262481,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004987339085658848,
      "loss": 5.7171,
      "mean_token_accuracy": 0.1774493247270584,
      "num_tokens": 8882497.0,
      "step": 3975
    },
    {
      "entropy": 5.564437675476074,
      "epoch": 0.44701521873420563,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004987296518005988,
      "loss": 5.4548,
      "mean_token_accuracy": 0.1823514148592949,
      "num_tokens": 8893155.0,
      "step": 3980
    },
    {
      "entropy": 5.674736356735229,
      "epoch": 0.4475767956421632,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000498725387911674,
      "loss": 5.5355,
      "mean_token_accuracy": 0.1841505229473114,
      "num_tokens": 8904478.0,
      "step": 3985
    },
    {
      "entropy": 5.707546997070312,
      "epoch": 0.44813837255012073,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004987211168992461,
      "loss": 5.6388,
      "mean_token_accuracy": 0.1715686649084091,
      "num_tokens": 8915470.0,
      "step": 3990
    },
    {
      "entropy": 5.79236364364624,
      "epoch": 0.4486999494580783,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000498716838763451,
      "loss": 5.5795,
      "mean_token_accuracy": 0.17429860830307006,
      "num_tokens": 8925644.0,
      "step": 3995
    },
    {
      "entropy": 5.7518922805786135,
      "epoch": 0.44926152636603583,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000498712553504425,
      "loss": 5.6189,
      "mean_token_accuracy": 0.1779789984226227,
      "num_tokens": 8937000.0,
      "step": 4000
    },
    {
      "entropy": 5.736722755432129,
      "epoch": 0.4498231032739934,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004987082611223046,
      "loss": 5.6095,
      "mean_token_accuracy": 0.17029931470751764,
      "num_tokens": 8948258.0,
      "step": 4005
    },
    {
      "entropy": 5.774406385421753,
      "epoch": 0.45038468018195094,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004987039616172265,
      "loss": 5.6054,
      "mean_token_accuracy": 0.17519907802343368,
      "num_tokens": 8960107.0,
      "step": 4010
    },
    {
      "entropy": 5.765909576416016,
      "epoch": 0.4509462570899085,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004986996549893273,
      "loss": 5.5994,
      "mean_token_accuracy": 0.1765146180987358,
      "num_tokens": 8969612.0,
      "step": 4015
    },
    {
      "entropy": 5.806485176086426,
      "epoch": 0.451507833997866,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004986953412387445,
      "loss": 5.7071,
      "mean_token_accuracy": 0.16734818518161773,
      "num_tokens": 8980970.0,
      "step": 4020
    },
    {
      "entropy": 5.805275297164917,
      "epoch": 0.45206941090582353,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004986910203656151,
      "loss": 5.621,
      "mean_token_accuracy": 0.17632754147052765,
      "num_tokens": 8992218.0,
      "step": 4025
    },
    {
      "entropy": 5.640167713165283,
      "epoch": 0.4526309878137811,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004986866923700769,
      "loss": 5.5727,
      "mean_token_accuracy": 0.1780338168144226,
      "num_tokens": 9003737.0,
      "step": 4030
    },
    {
      "entropy": 5.752987957000732,
      "epoch": 0.45319256472173863,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004986823572522677,
      "loss": 5.6354,
      "mean_token_accuracy": 0.1733468621969223,
      "num_tokens": 9015427.0,
      "step": 4035
    },
    {
      "entropy": 5.740896224975586,
      "epoch": 0.4537541416296962,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004986780150123254,
      "loss": 5.646,
      "mean_token_accuracy": 0.17190898954868317,
      "num_tokens": 9025003.0,
      "step": 4040
    },
    {
      "entropy": 5.7495654106140135,
      "epoch": 0.45431571853765373,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004986736656503885,
      "loss": 5.5925,
      "mean_token_accuracy": 0.17648608535528182,
      "num_tokens": 9036074.0,
      "step": 4045
    },
    {
      "entropy": 5.848660945892334,
      "epoch": 0.4548772954456113,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004986693091665952,
      "loss": 5.6039,
      "mean_token_accuracy": 0.17320455014705657,
      "num_tokens": 9046878.0,
      "step": 4050
    },
    {
      "entropy": 5.745128726959228,
      "epoch": 0.45543887235356884,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004986649455610845,
      "loss": 5.6469,
      "mean_token_accuracy": 0.1768433302640915,
      "num_tokens": 9058366.0,
      "step": 4055
    },
    {
      "entropy": 5.7796296119689945,
      "epoch": 0.4560004492615264,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000498660574833995,
      "loss": 5.6861,
      "mean_token_accuracy": 0.17641495764255524,
      "num_tokens": 9069542.0,
      "step": 4060
    },
    {
      "entropy": 5.856907892227173,
      "epoch": 0.45656202616948394,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004986561969854661,
      "loss": 5.8471,
      "mean_token_accuracy": 0.16817901730537416,
      "num_tokens": 9081219.0,
      "step": 4065
    },
    {
      "entropy": 5.793590927124024,
      "epoch": 0.45712360307744143,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004986518120156373,
      "loss": 5.5449,
      "mean_token_accuracy": 0.18378487527370452,
      "num_tokens": 9092101.0,
      "step": 4070
    },
    {
      "entropy": 5.727591800689697,
      "epoch": 0.457685179985399,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000498647419924648,
      "loss": 5.6005,
      "mean_token_accuracy": 0.17107009142637253,
      "num_tokens": 9101900.0,
      "step": 4075
    },
    {
      "entropy": 5.828117609024048,
      "epoch": 0.45824675689335653,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000498643020712638,
      "loss": 5.7318,
      "mean_token_accuracy": 0.17306118160486222,
      "num_tokens": 9114369.0,
      "step": 4080
    },
    {
      "entropy": 5.69248948097229,
      "epoch": 0.4588083338013141,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004986386143797475,
      "loss": 5.5293,
      "mean_token_accuracy": 0.17788425236940383,
      "num_tokens": 9126548.0,
      "step": 4085
    },
    {
      "entropy": 5.782763481140137,
      "epoch": 0.45936991070927163,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004986342009261169,
      "loss": 5.7209,
      "mean_token_accuracy": 0.1756329655647278,
      "num_tokens": 9138070.0,
      "step": 4090
    },
    {
      "entropy": 5.806504344940185,
      "epoch": 0.4599314876172292,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004986297803518865,
      "loss": 5.6438,
      "mean_token_accuracy": 0.17369131743907928,
      "num_tokens": 9149489.0,
      "step": 4095
    },
    {
      "entropy": 5.753714275360108,
      "epoch": 0.46049306452518673,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004986253526571971,
      "loss": 5.6317,
      "mean_token_accuracy": 0.1789081260561943,
      "num_tokens": 9160363.0,
      "step": 4100
    },
    {
      "entropy": 5.773330402374268,
      "epoch": 0.4610546414331443,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004986209178421899,
      "loss": 5.6845,
      "mean_token_accuracy": 0.17110240757465361,
      "num_tokens": 9171619.0,
      "step": 4105
    },
    {
      "entropy": 5.707543563842774,
      "epoch": 0.46161621834110184,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000498616475907006,
      "loss": 5.5299,
      "mean_token_accuracy": 0.18105072379112244,
      "num_tokens": 9182828.0,
      "step": 4110
    },
    {
      "entropy": 5.8059014797210695,
      "epoch": 0.46217779524905933,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004986120268517865,
      "loss": 5.6934,
      "mean_token_accuracy": 0.17877480536699294,
      "num_tokens": 9195249.0,
      "step": 4115
    },
    {
      "entropy": 5.6697998046875,
      "epoch": 0.4627393721570169,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004986075706766736,
      "loss": 5.5034,
      "mean_token_accuracy": 0.1782999262213707,
      "num_tokens": 9205770.0,
      "step": 4120
    },
    {
      "entropy": 5.6768879890441895,
      "epoch": 0.46330094906497443,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004986031073818088,
      "loss": 5.5156,
      "mean_token_accuracy": 0.17851390838623046,
      "num_tokens": 9216367.0,
      "step": 4125
    },
    {
      "entropy": 5.743587303161621,
      "epoch": 0.463862525972932,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004985986369673343,
      "loss": 5.6053,
      "mean_token_accuracy": 0.17105733305215837,
      "num_tokens": 9228524.0,
      "step": 4130
    },
    {
      "entropy": 5.73607234954834,
      "epoch": 0.46442410288088953,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004985941594333926,
      "loss": 5.5852,
      "mean_token_accuracy": 0.18082618117332458,
      "num_tokens": 9238907.0,
      "step": 4135
    },
    {
      "entropy": 5.781908941268921,
      "epoch": 0.4649856797888471,
      "grad_norm": 0.8125,
      "learning_rate": 0.000498589674780126,
      "loss": 5.7066,
      "mean_token_accuracy": 0.17598383128643036,
      "num_tokens": 9250491.0,
      "step": 4140
    },
    {
      "entropy": 5.737407779693603,
      "epoch": 0.46554725669680463,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004985851830076775,
      "loss": 5.5149,
      "mean_token_accuracy": 0.18421683758497237,
      "num_tokens": 9260984.0,
      "step": 4145
    },
    {
      "entropy": 5.783055973052979,
      "epoch": 0.4661088336047622,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004985806841161901,
      "loss": 5.6661,
      "mean_token_accuracy": 0.17464132010936737,
      "num_tokens": 9272416.0,
      "step": 4150
    },
    {
      "entropy": 5.792357873916626,
      "epoch": 0.46667041051271974,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000498576178105807,
      "loss": 5.6466,
      "mean_token_accuracy": 0.17614581137895585,
      "num_tokens": 9284513.0,
      "step": 4155
    },
    {
      "entropy": 5.76757025718689,
      "epoch": 0.4672319874206773,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004985716649766717,
      "loss": 5.5566,
      "mean_token_accuracy": 0.18276074379682541,
      "num_tokens": 9293422.0,
      "step": 4160
    },
    {
      "entropy": 5.686552810668945,
      "epoch": 0.4677935643286348,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004985671447289277,
      "loss": 5.5305,
      "mean_token_accuracy": 0.1756766617298126,
      "num_tokens": 9304735.0,
      "step": 4165
    },
    {
      "entropy": 5.692088985443116,
      "epoch": 0.46835514123659233,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004985626173627194,
      "loss": 5.65,
      "mean_token_accuracy": 0.17546037435531617,
      "num_tokens": 9315659.0,
      "step": 4170
    },
    {
      "entropy": 5.70388445854187,
      "epoch": 0.4689167181445499,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004985580828781904,
      "loss": 5.6523,
      "mean_token_accuracy": 0.17618295401334763,
      "num_tokens": 9327106.0,
      "step": 4175
    },
    {
      "entropy": 5.873208284378052,
      "epoch": 0.46947829505250743,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004985535412754855,
      "loss": 5.6937,
      "mean_token_accuracy": 0.17542646080255508,
      "num_tokens": 9338762.0,
      "step": 4180
    },
    {
      "entropy": 5.750082731246948,
      "epoch": 0.470039871960465,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000498548992554749,
      "loss": 5.5635,
      "mean_token_accuracy": 0.17716280668973922,
      "num_tokens": 9349495.0,
      "step": 4185
    },
    {
      "entropy": 5.819986200332641,
      "epoch": 0.47060144886842253,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004985444367161259,
      "loss": 5.6616,
      "mean_token_accuracy": 0.18115428537130357,
      "num_tokens": 9360830.0,
      "step": 4190
    },
    {
      "entropy": 5.644582891464234,
      "epoch": 0.4711630257763801,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004985398737597614,
      "loss": 5.5503,
      "mean_token_accuracy": 0.1805099591612816,
      "num_tokens": 9371668.0,
      "step": 4195
    },
    {
      "entropy": 5.771477174758911,
      "epoch": 0.47172460268433763,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004985353036858004,
      "loss": 5.6782,
      "mean_token_accuracy": 0.16847940683364868,
      "num_tokens": 9383437.0,
      "step": 4200
    },
    {
      "entropy": 5.68648886680603,
      "epoch": 0.4722861795922952,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004985307264943887,
      "loss": 5.4987,
      "mean_token_accuracy": 0.18313559144735336,
      "num_tokens": 9394597.0,
      "step": 4205
    },
    {
      "entropy": 5.5720500469207765,
      "epoch": 0.47284775650025274,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004985261421856721,
      "loss": 5.4521,
      "mean_token_accuracy": 0.179040589928627,
      "num_tokens": 9404322.0,
      "step": 4210
    },
    {
      "entropy": 5.821425914764404,
      "epoch": 0.47340933340821023,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004985215507597963,
      "loss": 5.8046,
      "mean_token_accuracy": 0.16372974067926407,
      "num_tokens": 9416740.0,
      "step": 4215
    },
    {
      "entropy": 5.786899042129517,
      "epoch": 0.4739709103161678,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004985169522169076,
      "loss": 5.6012,
      "mean_token_accuracy": 0.1774366691708565,
      "num_tokens": 9428617.0,
      "step": 4220
    },
    {
      "entropy": 5.6100372791290285,
      "epoch": 0.47453248722412533,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004985123465571526,
      "loss": 5.563,
      "mean_token_accuracy": 0.17956954836845399,
      "num_tokens": 9439686.0,
      "step": 4225
    },
    {
      "entropy": 5.741409254074097,
      "epoch": 0.4750940641320829,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004985077337806777,
      "loss": 5.5565,
      "mean_token_accuracy": 0.18088674247264863,
      "num_tokens": 9450378.0,
      "step": 4230
    },
    {
      "entropy": 5.765839576721191,
      "epoch": 0.47565564104004043,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004985031138876299,
      "loss": 5.5764,
      "mean_token_accuracy": 0.17544191628694533,
      "num_tokens": 9461659.0,
      "step": 4235
    },
    {
      "entropy": 5.72921576499939,
      "epoch": 0.476217217947998,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004984984868781563,
      "loss": 5.6168,
      "mean_token_accuracy": 0.1789289742708206,
      "num_tokens": 9472528.0,
      "step": 4240
    },
    {
      "entropy": 5.738912105560303,
      "epoch": 0.47677879485595553,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004984938527524041,
      "loss": 5.5508,
      "mean_token_accuracy": 0.18738911896944047,
      "num_tokens": 9483388.0,
      "step": 4245
    },
    {
      "entropy": 5.543483304977417,
      "epoch": 0.4773403717639131,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000498489211510521,
      "loss": 5.4031,
      "mean_token_accuracy": 0.1883103370666504,
      "num_tokens": 9495178.0,
      "step": 4250
    },
    {
      "entropy": 5.686541748046875,
      "epoch": 0.47790194867187064,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004984845631526547,
      "loss": 5.6335,
      "mean_token_accuracy": 0.17493624985218048,
      "num_tokens": 9506226.0,
      "step": 4255
    },
    {
      "entropy": 5.812045621871948,
      "epoch": 0.47846352557982813,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004984799076789534,
      "loss": 5.5445,
      "mean_token_accuracy": 0.17952838987112046,
      "num_tokens": 9517671.0,
      "step": 4260
    },
    {
      "entropy": 5.657545375823974,
      "epoch": 0.4790251024877857,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004984752450895649,
      "loss": 5.6246,
      "mean_token_accuracy": 0.18217895328998565,
      "num_tokens": 9527637.0,
      "step": 4265
    },
    {
      "entropy": 5.784671020507813,
      "epoch": 0.47958667939574323,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000498470575384638,
      "loss": 5.7009,
      "mean_token_accuracy": 0.17047436237335206,
      "num_tokens": 9539686.0,
      "step": 4270
    },
    {
      "entropy": 5.795252799987793,
      "epoch": 0.4801482563037008,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004984658985643214,
      "loss": 5.6578,
      "mean_token_accuracy": 0.18040469735860826,
      "num_tokens": 9550775.0,
      "step": 4275
    },
    {
      "entropy": 5.7318836688995365,
      "epoch": 0.48070983321165833,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004984612146287638,
      "loss": 5.5701,
      "mean_token_accuracy": 0.1788272365927696,
      "num_tokens": 9561760.0,
      "step": 4280
    },
    {
      "entropy": 5.626402378082275,
      "epoch": 0.4812714101196159,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004984565235781147,
      "loss": 5.6173,
      "mean_token_accuracy": 0.17459286004304886,
      "num_tokens": 9573899.0,
      "step": 4285
    },
    {
      "entropy": 5.718928289413452,
      "epoch": 0.48183298702757343,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000498451825412523,
      "loss": 5.5664,
      "mean_token_accuracy": 0.17680693566799163,
      "num_tokens": 9584336.0,
      "step": 4290
    },
    {
      "entropy": 5.864633417129516,
      "epoch": 0.482394563935531,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004984471201321386,
      "loss": 5.7493,
      "mean_token_accuracy": 0.16751802265644072,
      "num_tokens": 9596614.0,
      "step": 4295
    },
    {
      "entropy": 5.745125102996826,
      "epoch": 0.48295614084348854,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004984424077371112,
      "loss": 5.5983,
      "mean_token_accuracy": 0.1794119358062744,
      "num_tokens": 9606796.0,
      "step": 4300
    },
    {
      "entropy": 5.7320390224456785,
      "epoch": 0.4835177177514461,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004984376882275909,
      "loss": 5.5783,
      "mean_token_accuracy": 0.17826914340257644,
      "num_tokens": 9617705.0,
      "step": 4305
    },
    {
      "entropy": 5.663240814208985,
      "epoch": 0.4840792946594036,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004984329616037279,
      "loss": 5.5983,
      "mean_token_accuracy": 0.18012045025825502,
      "num_tokens": 9628098.0,
      "step": 4310
    },
    {
      "entropy": 5.709856700897217,
      "epoch": 0.48464087156736113,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004984282278656727,
      "loss": 5.5547,
      "mean_token_accuracy": 0.1855262354016304,
      "num_tokens": 9637697.0,
      "step": 4315
    },
    {
      "entropy": 5.668587398529053,
      "epoch": 0.4852024484753187,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004984234870135762,
      "loss": 5.4405,
      "mean_token_accuracy": 0.18603906631469727,
      "num_tokens": 9648221.0,
      "step": 4320
    },
    {
      "entropy": 5.774236965179443,
      "epoch": 0.48576402538327623,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004984187390475891,
      "loss": 5.6361,
      "mean_token_accuracy": 0.17187175750732422,
      "num_tokens": 9660182.0,
      "step": 4325
    },
    {
      "entropy": 5.699810218811035,
      "epoch": 0.4863256022912338,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004984139839678628,
      "loss": 5.6456,
      "mean_token_accuracy": 0.17410355508327485,
      "num_tokens": 9672121.0,
      "step": 4330
    },
    {
      "entropy": 5.793808364868164,
      "epoch": 0.48688717919919133,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004984092217745486,
      "loss": 5.546,
      "mean_token_accuracy": 0.1762766033411026,
      "num_tokens": 9682651.0,
      "step": 4335
    },
    {
      "entropy": 5.762834453582764,
      "epoch": 0.4874487561071489,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004984044524677981,
      "loss": 5.6997,
      "mean_token_accuracy": 0.17066227197647094,
      "num_tokens": 9693989.0,
      "step": 4340
    },
    {
      "entropy": 5.630291748046875,
      "epoch": 0.48801033301510643,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004983996760477632,
      "loss": 5.548,
      "mean_token_accuracy": 0.18355241864919664,
      "num_tokens": 9703802.0,
      "step": 4345
    },
    {
      "entropy": 5.724739265441895,
      "epoch": 0.488571909923064,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000498394892514596,
      "loss": 5.5065,
      "mean_token_accuracy": 0.1915830433368683,
      "num_tokens": 9714287.0,
      "step": 4350
    },
    {
      "entropy": 5.667872381210327,
      "epoch": 0.48913348683102154,
      "grad_norm": 0.875,
      "learning_rate": 0.0004983901018684486,
      "loss": 5.5215,
      "mean_token_accuracy": 0.18638120591640472,
      "num_tokens": 9724819.0,
      "step": 4355
    },
    {
      "entropy": 5.688709402084351,
      "epoch": 0.48969506373897903,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004983853041094739,
      "loss": 5.6674,
      "mean_token_accuracy": 0.17144156247377396,
      "num_tokens": 9736326.0,
      "step": 4360
    },
    {
      "entropy": 5.817929601669311,
      "epoch": 0.4902566406469366,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004983804992378243,
      "loss": 5.6862,
      "mean_token_accuracy": 0.17721305191516876,
      "num_tokens": 9747883.0,
      "step": 4365
    },
    {
      "entropy": 5.693543004989624,
      "epoch": 0.49081821755489413,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004983756872536532,
      "loss": 5.5489,
      "mean_token_accuracy": 0.17930855304002763,
      "num_tokens": 9758191.0,
      "step": 4370
    },
    {
      "entropy": 5.790836381912231,
      "epoch": 0.4913797944628517,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004983708681571134,
      "loss": 5.6408,
      "mean_token_accuracy": 0.17361029833555222,
      "num_tokens": 9768852.0,
      "step": 4375
    },
    {
      "entropy": 5.772389030456543,
      "epoch": 0.49194137137080923,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004983660419483586,
      "loss": 5.5875,
      "mean_token_accuracy": 0.17166981995105743,
      "num_tokens": 9779033.0,
      "step": 4380
    },
    {
      "entropy": 5.676150321960449,
      "epoch": 0.4925029482787668,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004983612086275424,
      "loss": 5.5073,
      "mean_token_accuracy": 0.18141463547945022,
      "num_tokens": 9790777.0,
      "step": 4385
    },
    {
      "entropy": 5.8726073741912845,
      "epoch": 0.49306452518672433,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004983563681948187,
      "loss": 5.7722,
      "mean_token_accuracy": 0.1676598995923996,
      "num_tokens": 9802305.0,
      "step": 4390
    },
    {
      "entropy": 5.717857122421265,
      "epoch": 0.4936261020946819,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004983515206503416,
      "loss": 5.571,
      "mean_token_accuracy": 0.17534351348876953,
      "num_tokens": 9813590.0,
      "step": 4395
    },
    {
      "entropy": 5.715846014022827,
      "epoch": 0.49418767900263944,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004983466659942655,
      "loss": 5.6561,
      "mean_token_accuracy": 0.17254330813884736,
      "num_tokens": 9824080.0,
      "step": 4400
    },
    {
      "entropy": 5.755474662780761,
      "epoch": 0.49474925591059693,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004983418042267449,
      "loss": 5.5404,
      "mean_token_accuracy": 0.17523749321699142,
      "num_tokens": 9835441.0,
      "step": 4405
    },
    {
      "entropy": 5.661392021179199,
      "epoch": 0.4953108328185545,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004983369353479345,
      "loss": 5.5278,
      "mean_token_accuracy": 0.18364492952823638,
      "num_tokens": 9846750.0,
      "step": 4410
    },
    {
      "entropy": 5.759102821350098,
      "epoch": 0.49587240972651203,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004983320593579895,
      "loss": 5.6891,
      "mean_token_accuracy": 0.1710416004061699,
      "num_tokens": 9857987.0,
      "step": 4415
    },
    {
      "entropy": 5.727979898452759,
      "epoch": 0.4964339866344696,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004983271762570653,
      "loss": 5.6164,
      "mean_token_accuracy": 0.18061216920614243,
      "num_tokens": 9869736.0,
      "step": 4420
    },
    {
      "entropy": 5.760137033462525,
      "epoch": 0.49699556354242713,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000498322286045317,
      "loss": 5.5905,
      "mean_token_accuracy": 0.17386518716812133,
      "num_tokens": 9881610.0,
      "step": 4425
    },
    {
      "entropy": 5.716667127609253,
      "epoch": 0.4975571404503847,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004983173887229006,
      "loss": 5.5809,
      "mean_token_accuracy": 0.17728231251239776,
      "num_tokens": 9893203.0,
      "step": 4430
    },
    {
      "entropy": 5.704579782485962,
      "epoch": 0.49811871735834223,
      "grad_norm": 0.875,
      "learning_rate": 0.0004983124842899719,
      "loss": 5.5098,
      "mean_token_accuracy": 0.17457224279642106,
      "num_tokens": 9904392.0,
      "step": 4435
    },
    {
      "entropy": 5.576231861114502,
      "epoch": 0.4986802942662998,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004983075727466871,
      "loss": 5.502,
      "mean_token_accuracy": 0.1825874388217926,
      "num_tokens": 9916184.0,
      "step": 4440
    },
    {
      "entropy": 5.748888397216797,
      "epoch": 0.49924187117425733,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004983026540932026,
      "loss": 5.6144,
      "mean_token_accuracy": 0.18264617770910263,
      "num_tokens": 9927516.0,
      "step": 4445
    },
    {
      "entropy": 5.730844163894654,
      "epoch": 0.4998034480822149,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004982977283296749,
      "loss": 5.6342,
      "mean_token_accuracy": 0.17622260451316835,
      "num_tokens": 9939182.0,
      "step": 4450
    },
    {
      "entropy": 5.78140287399292,
      "epoch": 0.5003650249901724,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000498292795456261,
      "loss": 5.6597,
      "mean_token_accuracy": 0.17473997920751572,
      "num_tokens": 9950382.0,
      "step": 4455
    },
    {
      "entropy": 5.77541675567627,
      "epoch": 0.50092660189813,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004982878554731178,
      "loss": 5.6498,
      "mean_token_accuracy": 0.17531245350837707,
      "num_tokens": 9961053.0,
      "step": 4460
    },
    {
      "entropy": 5.674372911453247,
      "epoch": 0.5014881788060875,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004982829083804028,
      "loss": 5.5558,
      "mean_token_accuracy": 0.18165653198957443,
      "num_tokens": 9971313.0,
      "step": 4465
    },
    {
      "entropy": 5.70073709487915,
      "epoch": 0.5020497557140451,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004982779541782733,
      "loss": 5.5745,
      "mean_token_accuracy": 0.19134376794099808,
      "num_tokens": 9981257.0,
      "step": 4470
    },
    {
      "entropy": 5.696200275421143,
      "epoch": 0.5026113326220026,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004982729928668872,
      "loss": 5.4999,
      "mean_token_accuracy": 0.17655672281980514,
      "num_tokens": 9991860.0,
      "step": 4475
    },
    {
      "entropy": 5.6428797245025635,
      "epoch": 0.5031729095299601,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004982680244464024,
      "loss": 5.6021,
      "mean_token_accuracy": 0.17707193940877913,
      "num_tokens": 10002320.0,
      "step": 4480
    },
    {
      "entropy": 5.801257133483887,
      "epoch": 0.5037344864379176,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004982630489169771,
      "loss": 5.6731,
      "mean_token_accuracy": 0.1757950156927109,
      "num_tokens": 10013187.0,
      "step": 4485
    },
    {
      "entropy": 5.641104221343994,
      "epoch": 0.5042960633458752,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004982580662787697,
      "loss": 5.4191,
      "mean_token_accuracy": 0.18904485553503036,
      "num_tokens": 10024268.0,
      "step": 4490
    },
    {
      "entropy": 5.7485558032989506,
      "epoch": 0.5048576402538327,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004982530765319389,
      "loss": 5.6806,
      "mean_token_accuracy": 0.17672700956463813,
      "num_tokens": 10034738.0,
      "step": 4495
    },
    {
      "entropy": 5.825581836700439,
      "epoch": 0.5054192171617903,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004982480796766436,
      "loss": 5.6377,
      "mean_token_accuracy": 0.17643290162086486,
      "num_tokens": 10045666.0,
      "step": 4500
    },
    {
      "entropy": 5.684539318084717,
      "epoch": 0.5059807940697478,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004982430757130429,
      "loss": 5.5049,
      "mean_token_accuracy": 0.18175057023763658,
      "num_tokens": 10057084.0,
      "step": 4505
    },
    {
      "entropy": 5.664234256744384,
      "epoch": 0.5065423709777054,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004982380646412959,
      "loss": 5.4959,
      "mean_token_accuracy": 0.17929004728794098,
      "num_tokens": 10066754.0,
      "step": 4510
    },
    {
      "entropy": 5.648970460891723,
      "epoch": 0.5071039478856629,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004982330464615625,
      "loss": 5.4936,
      "mean_token_accuracy": 0.18434620052576065,
      "num_tokens": 10078139.0,
      "step": 4515
    },
    {
      "entropy": 5.698085880279541,
      "epoch": 0.5076655247936205,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004982280211740023,
      "loss": 5.5572,
      "mean_token_accuracy": 0.1727925255894661,
      "num_tokens": 10089652.0,
      "step": 4520
    },
    {
      "entropy": 5.559485292434692,
      "epoch": 0.508227101701578,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004982229887787753,
      "loss": 5.3908,
      "mean_token_accuracy": 0.18712979853153228,
      "num_tokens": 10100642.0,
      "step": 4525
    },
    {
      "entropy": 5.743833398818969,
      "epoch": 0.5087886786095356,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004982179492760418,
      "loss": 5.6274,
      "mean_token_accuracy": 0.177151058614254,
      "num_tokens": 10112036.0,
      "step": 4530
    },
    {
      "entropy": 5.626854944229126,
      "epoch": 0.5093502555174931,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004982129026659621,
      "loss": 5.4873,
      "mean_token_accuracy": 0.18291615396738053,
      "num_tokens": 10121653.0,
      "step": 4535
    },
    {
      "entropy": 5.699587345123291,
      "epoch": 0.5099118324254507,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004982078489486972,
      "loss": 5.5895,
      "mean_token_accuracy": 0.17425881922245026,
      "num_tokens": 10132271.0,
      "step": 4540
    },
    {
      "entropy": 5.600173854827881,
      "epoch": 0.5104734093334082,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004982027881244076,
      "loss": 5.4829,
      "mean_token_accuracy": 0.17858110815286637,
      "num_tokens": 10143411.0,
      "step": 4545
    },
    {
      "entropy": 5.78210072517395,
      "epoch": 0.5110349862413658,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004981977201932549,
      "loss": 5.5789,
      "mean_token_accuracy": 0.1853160157799721,
      "num_tokens": 10154338.0,
      "step": 4550
    },
    {
      "entropy": 5.739124059677124,
      "epoch": 0.5115965631493233,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004981926451554001,
      "loss": 5.6201,
      "mean_token_accuracy": 0.1795254245400429,
      "num_tokens": 10164810.0,
      "step": 4555
    },
    {
      "entropy": 5.687288093566894,
      "epoch": 0.5121581400572809,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004981875630110048,
      "loss": 5.5265,
      "mean_token_accuracy": 0.18270367085933686,
      "num_tokens": 10176323.0,
      "step": 4560
    },
    {
      "entropy": 5.685714817047119,
      "epoch": 0.5127197169652384,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004981824737602311,
      "loss": 5.5554,
      "mean_token_accuracy": 0.17915245592594148,
      "num_tokens": 10187110.0,
      "step": 4565
    },
    {
      "entropy": 5.655750560760498,
      "epoch": 0.513281293873196,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004981773774032409,
      "loss": 5.5313,
      "mean_token_accuracy": 0.17791365683078766,
      "num_tokens": 10199099.0,
      "step": 4570
    },
    {
      "entropy": 5.736877298355102,
      "epoch": 0.5138428707811534,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004981722739401962,
      "loss": 5.5151,
      "mean_token_accuracy": 0.18200454413890838,
      "num_tokens": 10209726.0,
      "step": 4575
    },
    {
      "entropy": 5.76546688079834,
      "epoch": 0.514404447689111,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00049816716337126,
      "loss": 5.6138,
      "mean_token_accuracy": 0.17597122490406036,
      "num_tokens": 10220858.0,
      "step": 4580
    },
    {
      "entropy": 5.542819976806641,
      "epoch": 0.5149660245970685,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004981620456965946,
      "loss": 5.4297,
      "mean_token_accuracy": 0.18689085096120833,
      "num_tokens": 10232250.0,
      "step": 4585
    },
    {
      "entropy": 5.814619684219361,
      "epoch": 0.5155276015050261,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004981569209163633,
      "loss": 5.6249,
      "mean_token_accuracy": 0.17199626117944716,
      "num_tokens": 10244584.0,
      "step": 4590
    },
    {
      "entropy": 5.745997428894043,
      "epoch": 0.5160891784129836,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004981517890307288,
      "loss": 5.5128,
      "mean_token_accuracy": 0.18016383796930313,
      "num_tokens": 10255926.0,
      "step": 4595
    },
    {
      "entropy": 5.691600227355957,
      "epoch": 0.5166507553209412,
      "grad_norm": 0.921875,
      "learning_rate": 0.000498146650039855,
      "loss": 5.5905,
      "mean_token_accuracy": 0.1809764713048935,
      "num_tokens": 10266691.0,
      "step": 4600
    },
    {
      "entropy": 5.761801528930664,
      "epoch": 0.5172123322288987,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004981415039439054,
      "loss": 5.6628,
      "mean_token_accuracy": 0.1747693821787834,
      "num_tokens": 10277744.0,
      "step": 4605
    },
    {
      "entropy": 5.642522096633911,
      "epoch": 0.5177739091368563,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004981363507430436,
      "loss": 5.4894,
      "mean_token_accuracy": 0.18373941481113434,
      "num_tokens": 10288792.0,
      "step": 4610
    },
    {
      "entropy": 5.740934944152832,
      "epoch": 0.5183354860448138,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004981311904374339,
      "loss": 5.542,
      "mean_token_accuracy": 0.18077210038900376,
      "num_tokens": 10299569.0,
      "step": 4615
    },
    {
      "entropy": 5.724869060516357,
      "epoch": 0.5188970629527714,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004981260230272405,
      "loss": 5.5363,
      "mean_token_accuracy": 0.18323881179094315,
      "num_tokens": 10309742.0,
      "step": 4620
    },
    {
      "entropy": 5.730915212631226,
      "epoch": 0.5194586398607289,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004981208485126281,
      "loss": 5.5761,
      "mean_token_accuracy": 0.1752224564552307,
      "num_tokens": 10320603.0,
      "step": 4625
    },
    {
      "entropy": 5.647926568984985,
      "epoch": 0.5200202167686865,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004981156668937613,
      "loss": 5.5424,
      "mean_token_accuracy": 0.18151523917913437,
      "num_tokens": 10331498.0,
      "step": 4630
    },
    {
      "entropy": 5.689660978317261,
      "epoch": 0.520581793676644,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004981104781708051,
      "loss": 5.576,
      "mean_token_accuracy": 0.18089094907045364,
      "num_tokens": 10342002.0,
      "step": 4635
    },
    {
      "entropy": 5.7177896976470945,
      "epoch": 0.5211433705846016,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004981052823439247,
      "loss": 5.535,
      "mean_token_accuracy": 0.17622677385807037,
      "num_tokens": 10351797.0,
      "step": 4640
    },
    {
      "entropy": 5.641194248199463,
      "epoch": 0.5217049474925591,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004981000794132858,
      "loss": 5.4592,
      "mean_token_accuracy": 0.18417613357305526,
      "num_tokens": 10362538.0,
      "step": 4645
    },
    {
      "entropy": 5.667370271682739,
      "epoch": 0.5222665244005167,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004980948693790536,
      "loss": 5.4788,
      "mean_token_accuracy": 0.18743957132101058,
      "num_tokens": 10373365.0,
      "step": 4650
    },
    {
      "entropy": 5.6157654285430905,
      "epoch": 0.5228281013084742,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004980896522413943,
      "loss": 5.4732,
      "mean_token_accuracy": 0.17729185819625853,
      "num_tokens": 10385087.0,
      "step": 4655
    },
    {
      "entropy": 5.690910053253174,
      "epoch": 0.5233896782164318,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004980844280004739,
      "loss": 5.599,
      "mean_token_accuracy": 0.17634045630693435,
      "num_tokens": 10396014.0,
      "step": 4660
    },
    {
      "entropy": 5.659593391418457,
      "epoch": 0.5239512551243893,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004980791966564589,
      "loss": 5.518,
      "mean_token_accuracy": 0.17636214047670365,
      "num_tokens": 10407844.0,
      "step": 4665
    },
    {
      "entropy": 5.6255475044250485,
      "epoch": 0.5245128320323468,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004980739582095158,
      "loss": 5.5871,
      "mean_token_accuracy": 0.175868359208107,
      "num_tokens": 10419892.0,
      "step": 4670
    },
    {
      "entropy": 5.720735740661621,
      "epoch": 0.5250744089403043,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004980687126598111,
      "loss": 5.5521,
      "mean_token_accuracy": 0.17343892008066178,
      "num_tokens": 10430686.0,
      "step": 4675
    },
    {
      "entropy": 5.689608860015869,
      "epoch": 0.5256359858482619,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004980634600075123,
      "loss": 5.5682,
      "mean_token_accuracy": 0.1766488343477249,
      "num_tokens": 10443785.0,
      "step": 4680
    },
    {
      "entropy": 5.653850984573364,
      "epoch": 0.5261975627562194,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004980582002527863,
      "loss": 5.4821,
      "mean_token_accuracy": 0.18440356552600862,
      "num_tokens": 10454889.0,
      "step": 4685
    },
    {
      "entropy": 5.6599023818969725,
      "epoch": 0.526759139664177,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004980529333958008,
      "loss": 5.6349,
      "mean_token_accuracy": 0.17384447157382965,
      "num_tokens": 10466525.0,
      "step": 4690
    },
    {
      "entropy": 5.801600790023803,
      "epoch": 0.5273207165721345,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004980476594367233,
      "loss": 5.587,
      "mean_token_accuracy": 0.17820865511894227,
      "num_tokens": 10476329.0,
      "step": 4695
    },
    {
      "entropy": 5.756714868545532,
      "epoch": 0.5278822934800921,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004980423783757219,
      "loss": 5.6403,
      "mean_token_accuracy": 0.17722613215446473,
      "num_tokens": 10487178.0,
      "step": 4700
    },
    {
      "entropy": 5.745976305007934,
      "epoch": 0.5284438703880496,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004980370902129646,
      "loss": 5.5836,
      "mean_token_accuracy": 0.16918132901191713,
      "num_tokens": 10499293.0,
      "step": 4705
    },
    {
      "entropy": 5.822352600097656,
      "epoch": 0.5290054472960072,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00049803179494862,
      "loss": 5.7616,
      "mean_token_accuracy": 0.1672948680818081,
      "num_tokens": 10511762.0,
      "step": 4710
    },
    {
      "entropy": 5.696107244491577,
      "epoch": 0.5295670242039647,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004980264925828563,
      "loss": 5.4851,
      "mean_token_accuracy": 0.1872537076473236,
      "num_tokens": 10521725.0,
      "step": 4715
    },
    {
      "entropy": 5.631313133239746,
      "epoch": 0.5301286011119223,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004980211831158427,
      "loss": 5.4453,
      "mean_token_accuracy": 0.17969469875097274,
      "num_tokens": 10531764.0,
      "step": 4720
    },
    {
      "entropy": 5.718774509429932,
      "epoch": 0.5306901780198798,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004980158665477481,
      "loss": 5.5188,
      "mean_token_accuracy": 0.18248597532510757,
      "num_tokens": 10542127.0,
      "step": 4725
    },
    {
      "entropy": 5.741027212142944,
      "epoch": 0.5312517549278374,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004980105428787419,
      "loss": 5.604,
      "mean_token_accuracy": 0.17546663731336593,
      "num_tokens": 10552580.0,
      "step": 4730
    },
    {
      "entropy": 5.704824066162109,
      "epoch": 0.5318133318357949,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004980052121089934,
      "loss": 5.5418,
      "mean_token_accuracy": 0.17995575517416001,
      "num_tokens": 10562961.0,
      "step": 4735
    },
    {
      "entropy": 5.792839479446411,
      "epoch": 0.5323749087437525,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004979998742386725,
      "loss": 5.5362,
      "mean_token_accuracy": 0.183878792822361,
      "num_tokens": 10573210.0,
      "step": 4740
    },
    {
      "entropy": 5.572492694854736,
      "epoch": 0.53293648565171,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004979945292679492,
      "loss": 5.5081,
      "mean_token_accuracy": 0.18159353882074356,
      "num_tokens": 10583878.0,
      "step": 4745
    },
    {
      "entropy": 5.730389022827149,
      "epoch": 0.5334980625596676,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004979891771969935,
      "loss": 5.5817,
      "mean_token_accuracy": 0.17595281153917314,
      "num_tokens": 10593420.0,
      "step": 4750
    },
    {
      "entropy": 5.712625503540039,
      "epoch": 0.5340596394676251,
      "grad_norm": 0.90625,
      "learning_rate": 0.000497983818025976,
      "loss": 5.5908,
      "mean_token_accuracy": 0.1781250759959221,
      "num_tokens": 10604213.0,
      "step": 4755
    },
    {
      "entropy": 5.715660333633423,
      "epoch": 0.5346212163755827,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004979784517550671,
      "loss": 5.5752,
      "mean_token_accuracy": 0.17217695862054824,
      "num_tokens": 10615502.0,
      "step": 4760
    },
    {
      "entropy": 5.733752822875976,
      "epoch": 0.5351827932835401,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004979730783844379,
      "loss": 5.6099,
      "mean_token_accuracy": 0.17149658054113387,
      "num_tokens": 10627589.0,
      "step": 4765
    },
    {
      "entropy": 5.628414821624756,
      "epoch": 0.5357443701914977,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004979676979142595,
      "loss": 5.4685,
      "mean_token_accuracy": 0.1829899251461029,
      "num_tokens": 10638365.0,
      "step": 4770
    },
    {
      "entropy": 5.685724830627441,
      "epoch": 0.5363059470994552,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000497962310344703,
      "loss": 5.5559,
      "mean_token_accuracy": 0.18290282487869264,
      "num_tokens": 10650664.0,
      "step": 4775
    },
    {
      "entropy": 5.768656587600708,
      "epoch": 0.5368675240074128,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00049795691567594,
      "loss": 5.6069,
      "mean_token_accuracy": 0.1871672421693802,
      "num_tokens": 10662418.0,
      "step": 4780
    },
    {
      "entropy": 5.684687328338623,
      "epoch": 0.5374291009153703,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004979515139081426,
      "loss": 5.4991,
      "mean_token_accuracy": 0.18115153759717942,
      "num_tokens": 10672793.0,
      "step": 4785
    },
    {
      "entropy": 5.626856088638306,
      "epoch": 0.5379906778233279,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004979461050414822,
      "loss": 5.5393,
      "mean_token_accuracy": 0.19159039705991746,
      "num_tokens": 10684082.0,
      "step": 4790
    },
    {
      "entropy": 5.615609836578369,
      "epoch": 0.5385522547312854,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004979406890761315,
      "loss": 5.4737,
      "mean_token_accuracy": 0.18325647413730622,
      "num_tokens": 10694490.0,
      "step": 4795
    },
    {
      "entropy": 5.67169976234436,
      "epoch": 0.539113831639243,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004979352660122629,
      "loss": 5.4928,
      "mean_token_accuracy": 0.18378038257360457,
      "num_tokens": 10704731.0,
      "step": 4800
    },
    {
      "entropy": 5.677918195724487,
      "epoch": 0.5396754085472005,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004979298358500488,
      "loss": 5.5071,
      "mean_token_accuracy": 0.1856495887041092,
      "num_tokens": 10715735.0,
      "step": 4805
    },
    {
      "entropy": 5.595276594161987,
      "epoch": 0.5402369854551581,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004979243985896624,
      "loss": 5.481,
      "mean_token_accuracy": 0.18190116286277772,
      "num_tokens": 10726951.0,
      "step": 4810
    },
    {
      "entropy": 5.655351066589356,
      "epoch": 0.5407985623631156,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004979189542312765,
      "loss": 5.5361,
      "mean_token_accuracy": 0.17646143734455108,
      "num_tokens": 10737609.0,
      "step": 4815
    },
    {
      "entropy": 5.660394239425659,
      "epoch": 0.5413601392710732,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004979135027750649,
      "loss": 5.4333,
      "mean_token_accuracy": 0.19128430932760238,
      "num_tokens": 10748957.0,
      "step": 4820
    },
    {
      "entropy": 5.691059350967407,
      "epoch": 0.5419217161790307,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004979080442212007,
      "loss": 5.6163,
      "mean_token_accuracy": 0.17857613712549208,
      "num_tokens": 10759398.0,
      "step": 4825
    },
    {
      "entropy": 5.749173212051391,
      "epoch": 0.5424832930869883,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004979025785698581,
      "loss": 5.5964,
      "mean_token_accuracy": 0.17563450634479522,
      "num_tokens": 10770436.0,
      "step": 4830
    },
    {
      "entropy": 5.636345958709716,
      "epoch": 0.5430448699949458,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004978971058212108,
      "loss": 5.435,
      "mean_token_accuracy": 0.19062792807817458,
      "num_tokens": 10781160.0,
      "step": 4835
    },
    {
      "entropy": 5.71708607673645,
      "epoch": 0.5436064469029034,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004978916259754331,
      "loss": 5.5662,
      "mean_token_accuracy": 0.18546935468912124,
      "num_tokens": 10793170.0,
      "step": 4840
    },
    {
      "entropy": 5.614717149734497,
      "epoch": 0.5441680238108609,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004978861390326998,
      "loss": 5.5002,
      "mean_token_accuracy": 0.18158200830221177,
      "num_tokens": 10803515.0,
      "step": 4845
    },
    {
      "entropy": 5.714852380752563,
      "epoch": 0.5447296007188185,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004978806449931854,
      "loss": 5.5319,
      "mean_token_accuracy": 0.1809350073337555,
      "num_tokens": 10814508.0,
      "step": 4850
    },
    {
      "entropy": 5.708811235427857,
      "epoch": 0.545291177626776,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004978751438570647,
      "loss": 5.6413,
      "mean_token_accuracy": 0.17780861258506775,
      "num_tokens": 10825924.0,
      "step": 4855
    },
    {
      "entropy": 5.7226824283599855,
      "epoch": 0.5458527545347336,
      "grad_norm": 0.875,
      "learning_rate": 0.0004978696356245129,
      "loss": 5.6158,
      "mean_token_accuracy": 0.17809754461050034,
      "num_tokens": 10837854.0,
      "step": 4860
    },
    {
      "entropy": 5.679769134521484,
      "epoch": 0.546414331442691,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004978641202957057,
      "loss": 5.5141,
      "mean_token_accuracy": 0.1806867852807045,
      "num_tokens": 10848279.0,
      "step": 4865
    },
    {
      "entropy": 5.6330842018127445,
      "epoch": 0.5469759083506486,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004978585978708183,
      "loss": 5.5304,
      "mean_token_accuracy": 0.1749546006321907,
      "num_tokens": 10859615.0,
      "step": 4870
    },
    {
      "entropy": 5.749248266220093,
      "epoch": 0.5475374852586061,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004978530683500267,
      "loss": 5.4767,
      "mean_token_accuracy": 0.18210446536540986,
      "num_tokens": 10871388.0,
      "step": 4875
    },
    {
      "entropy": 5.6507909297943115,
      "epoch": 0.5480990621665637,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004978475317335071,
      "loss": 5.5511,
      "mean_token_accuracy": 0.17665998339653016,
      "num_tokens": 10882411.0,
      "step": 4880
    },
    {
      "entropy": 5.679306602478027,
      "epoch": 0.5486606390745212,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004978419880214356,
      "loss": 5.6578,
      "mean_token_accuracy": 0.17885621637105942,
      "num_tokens": 10894116.0,
      "step": 4885
    },
    {
      "entropy": 5.610882949829102,
      "epoch": 0.5492222159824788,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004978364372139887,
      "loss": 5.4483,
      "mean_token_accuracy": 0.1855167880654335,
      "num_tokens": 10904514.0,
      "step": 4890
    },
    {
      "entropy": 5.616874313354492,
      "epoch": 0.5497837928904363,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004978308793113434,
      "loss": 5.4779,
      "mean_token_accuracy": 0.18350235372781754,
      "num_tokens": 10915602.0,
      "step": 4895
    },
    {
      "entropy": 5.634255838394165,
      "epoch": 0.5503453697983939,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004978253143136762,
      "loss": 5.4631,
      "mean_token_accuracy": 0.18849995285272597,
      "num_tokens": 10926172.0,
      "step": 4900
    },
    {
      "entropy": 5.724899578094482,
      "epoch": 0.5509069467063514,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004978197422211647,
      "loss": 5.5876,
      "mean_token_accuracy": 0.17352376878261566,
      "num_tokens": 10938435.0,
      "step": 4905
    },
    {
      "entropy": 5.616976499557495,
      "epoch": 0.551468523614309,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004978141630339863,
      "loss": 5.4772,
      "mean_token_accuracy": 0.18257201462984085,
      "num_tokens": 10949118.0,
      "step": 4910
    },
    {
      "entropy": 5.687697410583496,
      "epoch": 0.5520301005222665,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004978085767523184,
      "loss": 5.5252,
      "mean_token_accuracy": 0.17709917426109315,
      "num_tokens": 10960120.0,
      "step": 4915
    },
    {
      "entropy": 5.678229141235351,
      "epoch": 0.5525916774302241,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000497802983376339,
      "loss": 5.4931,
      "mean_token_accuracy": 0.18287136703729628,
      "num_tokens": 10971210.0,
      "step": 4920
    },
    {
      "entropy": 5.636813402175903,
      "epoch": 0.5531532543381816,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004977973829062263,
      "loss": 5.4865,
      "mean_token_accuracy": 0.19063847362995148,
      "num_tokens": 10983446.0,
      "step": 4925
    },
    {
      "entropy": 5.726073265075684,
      "epoch": 0.5537148312461392,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004977917753421585,
      "loss": 5.6099,
      "mean_token_accuracy": 0.17679833322763444,
      "num_tokens": 10994776.0,
      "step": 4930
    },
    {
      "entropy": 5.701025438308716,
      "epoch": 0.5542764081540967,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000497786160684314,
      "loss": 5.5196,
      "mean_token_accuracy": 0.1810809165239334,
      "num_tokens": 11005280.0,
      "step": 4935
    },
    {
      "entropy": 5.615891456604004,
      "epoch": 0.5548379850620543,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004977805389328719,
      "loss": 5.5004,
      "mean_token_accuracy": 0.17875374257564544,
      "num_tokens": 11017509.0,
      "step": 4940
    },
    {
      "entropy": 5.698488521575928,
      "epoch": 0.5553995619700118,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000497774910088011,
      "loss": 5.5246,
      "mean_token_accuracy": 0.18177431970834732,
      "num_tokens": 11029087.0,
      "step": 4945
    },
    {
      "entropy": 5.690251684188842,
      "epoch": 0.5559611388779694,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004977692741499106,
      "loss": 5.579,
      "mean_token_accuracy": 0.17716402411460877,
      "num_tokens": 11041144.0,
      "step": 4950
    },
    {
      "entropy": 5.825561285018921,
      "epoch": 0.5565227157859269,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00049776363111875,
      "loss": 5.6489,
      "mean_token_accuracy": 0.17067064493894576,
      "num_tokens": 11054050.0,
      "step": 4955
    },
    {
      "entropy": 5.6383764266967775,
      "epoch": 0.5570842926938844,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004977579809947091,
      "loss": 5.4667,
      "mean_token_accuracy": 0.18543528914451599,
      "num_tokens": 11064257.0,
      "step": 4960
    },
    {
      "entropy": 5.746443605422973,
      "epoch": 0.5576458696018419,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004977523237779676,
      "loss": 5.5888,
      "mean_token_accuracy": 0.17980761528015138,
      "num_tokens": 11075568.0,
      "step": 4965
    },
    {
      "entropy": 5.673269939422608,
      "epoch": 0.5582074465097995,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004977466594687057,
      "loss": 5.5265,
      "mean_token_accuracy": 0.1836976781487465,
      "num_tokens": 11087078.0,
      "step": 4970
    },
    {
      "entropy": 5.806062984466553,
      "epoch": 0.558769023417757,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004977409880671039,
      "loss": 5.6728,
      "mean_token_accuracy": 0.1747068002820015,
      "num_tokens": 11100655.0,
      "step": 4975
    },
    {
      "entropy": 5.747005605697632,
      "epoch": 0.5593306003257146,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004977353095733427,
      "loss": 5.5826,
      "mean_token_accuracy": 0.17522813081741334,
      "num_tokens": 11110675.0,
      "step": 4980
    },
    {
      "entropy": 5.672699785232544,
      "epoch": 0.5598921772336721,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004977296239876026,
      "loss": 5.5659,
      "mean_token_accuracy": 0.17682826668024063,
      "num_tokens": 11122337.0,
      "step": 4985
    },
    {
      "entropy": 5.760096597671509,
      "epoch": 0.5604537541416297,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000497723931310065,
      "loss": 5.4863,
      "mean_token_accuracy": 0.1837332382798195,
      "num_tokens": 11132672.0,
      "step": 4990
    },
    {
      "entropy": 5.668780374526977,
      "epoch": 0.5610153310495872,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004977182315409111,
      "loss": 5.5144,
      "mean_token_accuracy": 0.183090141415596,
      "num_tokens": 11143740.0,
      "step": 4995
    },
    {
      "entropy": 5.596275424957275,
      "epoch": 0.5615769079575448,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004977125246803224,
      "loss": 5.4677,
      "mean_token_accuracy": 0.18188604712486267,
      "num_tokens": 11153722.0,
      "step": 5000
    },
    {
      "entropy": 5.728806972503662,
      "epoch": 0.5621384848655023,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004977068107284803,
      "loss": 5.5914,
      "mean_token_accuracy": 0.17763902395963668,
      "num_tokens": 11165042.0,
      "step": 5005
    },
    {
      "entropy": 5.693523120880127,
      "epoch": 0.5627000617734599,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004977010896855672,
      "loss": 5.4635,
      "mean_token_accuracy": 0.18478105068206788,
      "num_tokens": 11176834.0,
      "step": 5010
    },
    {
      "entropy": 5.666048431396485,
      "epoch": 0.5632616386814174,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004976953615517649,
      "loss": 5.5462,
      "mean_token_accuracy": 0.18114150762557985,
      "num_tokens": 11188442.0,
      "step": 5015
    },
    {
      "entropy": 5.611933422088623,
      "epoch": 0.563823215589375,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000497689626327256,
      "loss": 5.4453,
      "mean_token_accuracy": 0.18681304156780243,
      "num_tokens": 11197855.0,
      "step": 5020
    },
    {
      "entropy": 5.666592693328857,
      "epoch": 0.5643847924973325,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000497683884012223,
      "loss": 5.5547,
      "mean_token_accuracy": 0.1762824296951294,
      "num_tokens": 11209300.0,
      "step": 5025
    },
    {
      "entropy": 5.6833336353302,
      "epoch": 0.5649463694052901,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004976781346068487,
      "loss": 5.5584,
      "mean_token_accuracy": 0.18144589960575103,
      "num_tokens": 11222175.0,
      "step": 5030
    },
    {
      "entropy": 5.742627143859863,
      "epoch": 0.5655079463132476,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004976723781113163,
      "loss": 5.5444,
      "mean_token_accuracy": 0.17976466566324234,
      "num_tokens": 11233277.0,
      "step": 5035
    },
    {
      "entropy": 5.721432399749756,
      "epoch": 0.5660695232212052,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004976666145258089,
      "loss": 5.5756,
      "mean_token_accuracy": 0.17921737134456633,
      "num_tokens": 11245241.0,
      "step": 5040
    },
    {
      "entropy": 5.691900682449341,
      "epoch": 0.5666311001291627,
      "grad_norm": 0.875,
      "learning_rate": 0.0004976608438505103,
      "loss": 5.6022,
      "mean_token_accuracy": 0.17975134253501893,
      "num_tokens": 11257359.0,
      "step": 5045
    },
    {
      "entropy": 5.670284128189087,
      "epoch": 0.5671926770371203,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004976550660856041,
      "loss": 5.5047,
      "mean_token_accuracy": 0.18505486249923705,
      "num_tokens": 11270651.0,
      "step": 5050
    },
    {
      "entropy": 5.653756237030029,
      "epoch": 0.5677542539450777,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000497649281231274,
      "loss": 5.5035,
      "mean_token_accuracy": 0.18090778738260269,
      "num_tokens": 11280164.0,
      "step": 5055
    },
    {
      "entropy": 5.640340995788574,
      "epoch": 0.5683158308530353,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004976434892877046,
      "loss": 5.4388,
      "mean_token_accuracy": 0.18042992055416107,
      "num_tokens": 11290156.0,
      "step": 5060
    },
    {
      "entropy": 5.681944036483765,
      "epoch": 0.5688774077609928,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004976376902550802,
      "loss": 5.505,
      "mean_token_accuracy": 0.18570900708436966,
      "num_tokens": 11301929.0,
      "step": 5065
    },
    {
      "entropy": 5.583669805526734,
      "epoch": 0.5694389846689504,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004976318841335854,
      "loss": 5.4294,
      "mean_token_accuracy": 0.1863689035177231,
      "num_tokens": 11312805.0,
      "step": 5070
    },
    {
      "entropy": 5.513946628570556,
      "epoch": 0.5700005615769079,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004976260709234049,
      "loss": 5.3341,
      "mean_token_accuracy": 0.18819766789674758,
      "num_tokens": 11322316.0,
      "step": 5075
    },
    {
      "entropy": 5.599709844589233,
      "epoch": 0.5705621384848655,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000497620250624724,
      "loss": 5.5137,
      "mean_token_accuracy": 0.18599254339933396,
      "num_tokens": 11332994.0,
      "step": 5080
    },
    {
      "entropy": 5.668393278121949,
      "epoch": 0.571123715392823,
      "grad_norm": 0.828125,
      "learning_rate": 0.000497614423237728,
      "loss": 5.4962,
      "mean_token_accuracy": 0.1800137221813202,
      "num_tokens": 11344355.0,
      "step": 5085
    },
    {
      "entropy": 5.6857749938964846,
      "epoch": 0.5716852923007806,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0004976085887626026,
      "loss": 5.5214,
      "mean_token_accuracy": 0.17533157467842103,
      "num_tokens": 11356275.0,
      "step": 5090
    },
    {
      "entropy": 5.641538953781128,
      "epoch": 0.5722468692087381,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004976027471995331,
      "loss": 5.5218,
      "mean_token_accuracy": 0.17730026245117186,
      "num_tokens": 11368624.0,
      "step": 5095
    },
    {
      "entropy": 5.766763496398926,
      "epoch": 0.5728084461166957,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000497596898548706,
      "loss": 5.6375,
      "mean_token_accuracy": 0.17876659482717513,
      "num_tokens": 11380520.0,
      "step": 5100
    },
    {
      "entropy": 5.634265899658203,
      "epoch": 0.5733700230246532,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004975910428103072,
      "loss": 5.5184,
      "mean_token_accuracy": 0.18526270240545273,
      "num_tokens": 11392293.0,
      "step": 5105
    },
    {
      "entropy": 5.665168237686157,
      "epoch": 0.5739315999326108,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004975851799845233,
      "loss": 5.5767,
      "mean_token_accuracy": 0.17941045314073562,
      "num_tokens": 11403866.0,
      "step": 5110
    },
    {
      "entropy": 5.703632116317749,
      "epoch": 0.5744931768405683,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000497579310071541,
      "loss": 5.4263,
      "mean_token_accuracy": 0.1896760493516922,
      "num_tokens": 11415334.0,
      "step": 5115
    },
    {
      "entropy": 5.596745872497559,
      "epoch": 0.5750547537485259,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004975734330715471,
      "loss": 5.568,
      "mean_token_accuracy": 0.1806974560022354,
      "num_tokens": 11426840.0,
      "step": 5120
    },
    {
      "entropy": 5.774216032028198,
      "epoch": 0.5756163306564834,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004975675489847287,
      "loss": 5.6955,
      "mean_token_accuracy": 0.17582051008939742,
      "num_tokens": 11438497.0,
      "step": 5125
    },
    {
      "entropy": 5.711315393447876,
      "epoch": 0.576177907564441,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004975616578112734,
      "loss": 5.4432,
      "mean_token_accuracy": 0.18717486560344695,
      "num_tokens": 11448705.0,
      "step": 5130
    },
    {
      "entropy": 5.759217739105225,
      "epoch": 0.5767394844723985,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004975557595513684,
      "loss": 5.6959,
      "mean_token_accuracy": 0.16969826221466064,
      "num_tokens": 11460610.0,
      "step": 5135
    },
    {
      "entropy": 5.661580038070679,
      "epoch": 0.5773010613803561,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004975498542052018,
      "loss": 5.4536,
      "mean_token_accuracy": 0.1856856554746628,
      "num_tokens": 11471170.0,
      "step": 5140
    },
    {
      "entropy": 5.648734855651855,
      "epoch": 0.5778626382883136,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004975439417729615,
      "loss": 5.4445,
      "mean_token_accuracy": 0.18800076693296433,
      "num_tokens": 11482090.0,
      "step": 5145
    },
    {
      "entropy": 5.613851642608642,
      "epoch": 0.5784242151962711,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004975380222548358,
      "loss": 5.5317,
      "mean_token_accuracy": 0.1817023813724518,
      "num_tokens": 11493650.0,
      "step": 5150
    },
    {
      "entropy": 5.674195671081543,
      "epoch": 0.5789857921042286,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004975320956510131,
      "loss": 5.5208,
      "mean_token_accuracy": 0.17385974973440171,
      "num_tokens": 11505903.0,
      "step": 5155
    },
    {
      "entropy": 5.736664152145385,
      "epoch": 0.5795473690121862,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004975261619616822,
      "loss": 5.5556,
      "mean_token_accuracy": 0.17978273332118988,
      "num_tokens": 11517194.0,
      "step": 5160
    },
    {
      "entropy": 5.679175329208374,
      "epoch": 0.5801089459201437,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004975202211870321,
      "loss": 5.5723,
      "mean_token_accuracy": 0.1829549714922905,
      "num_tokens": 11528683.0,
      "step": 5165
    },
    {
      "entropy": 5.686576080322266,
      "epoch": 0.5806705228281013,
      "grad_norm": 0.875,
      "learning_rate": 0.0004975142733272518,
      "loss": 5.5981,
      "mean_token_accuracy": 0.18052947670221328,
      "num_tokens": 11541940.0,
      "step": 5170
    },
    {
      "entropy": 5.634933853149414,
      "epoch": 0.5812320997360588,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004975083183825306,
      "loss": 5.4736,
      "mean_token_accuracy": 0.18062784969806672,
      "num_tokens": 11553045.0,
      "step": 5175
    },
    {
      "entropy": 5.601237106323242,
      "epoch": 0.5817936766440164,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004975023563530584,
      "loss": 5.4531,
      "mean_token_accuracy": 0.18912421315908431,
      "num_tokens": 11564401.0,
      "step": 5180
    },
    {
      "entropy": 5.677291345596314,
      "epoch": 0.5823552535519739,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004974963872390249,
      "loss": 5.4939,
      "mean_token_accuracy": 0.18018419444561004,
      "num_tokens": 11576424.0,
      "step": 5185
    },
    {
      "entropy": 5.700665283203125,
      "epoch": 0.5829168304599315,
      "grad_norm": 0.8125,
      "learning_rate": 0.00049749041104062,
      "loss": 5.4979,
      "mean_token_accuracy": 0.18000409454107286,
      "num_tokens": 11587853.0,
      "step": 5190
    },
    {
      "entropy": 5.682964229583741,
      "epoch": 0.583478407367889,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004974844277580343,
      "loss": 5.5001,
      "mean_token_accuracy": 0.18551741987466813,
      "num_tokens": 11597938.0,
      "step": 5195
    },
    {
      "entropy": 5.791597557067871,
      "epoch": 0.5840399842758466,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004974784373914579,
      "loss": 5.6626,
      "mean_token_accuracy": 0.17738450616598128,
      "num_tokens": 11610077.0,
      "step": 5200
    },
    {
      "entropy": 5.612270212173462,
      "epoch": 0.5846015611838041,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004974724399410818,
      "loss": 5.5296,
      "mean_token_accuracy": 0.18512929677963258,
      "num_tokens": 11622079.0,
      "step": 5205
    },
    {
      "entropy": 5.593129062652588,
      "epoch": 0.5851631380917617,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004974664354070971,
      "loss": 5.4511,
      "mean_token_accuracy": 0.18932920545339585,
      "num_tokens": 11633047.0,
      "step": 5210
    },
    {
      "entropy": 5.62644624710083,
      "epoch": 0.5857247149997192,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004974604237896947,
      "loss": 5.4037,
      "mean_token_accuracy": 0.18625760376453399,
      "num_tokens": 11643858.0,
      "step": 5215
    },
    {
      "entropy": 5.565522241592407,
      "epoch": 0.5862862919076768,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004974544050890662,
      "loss": 5.4315,
      "mean_token_accuracy": 0.18860581368207932,
      "num_tokens": 11654550.0,
      "step": 5220
    },
    {
      "entropy": 5.585755157470703,
      "epoch": 0.5868478688156343,
      "grad_norm": 0.859375,
      "learning_rate": 0.000497448379305403,
      "loss": 5.4558,
      "mean_token_accuracy": 0.1839885249733925,
      "num_tokens": 11665787.0,
      "step": 5225
    },
    {
      "entropy": 5.714945077896118,
      "epoch": 0.5874094457235919,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004974423464388973,
      "loss": 5.586,
      "mean_token_accuracy": 0.17909320145845414,
      "num_tokens": 11677717.0,
      "step": 5230
    },
    {
      "entropy": 5.658526468276977,
      "epoch": 0.5879710226315494,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000497436306489741,
      "loss": 5.4642,
      "mean_token_accuracy": 0.18368464857339858,
      "num_tokens": 11687714.0,
      "step": 5235
    },
    {
      "entropy": 5.590728521347046,
      "epoch": 0.588532599539507,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004974302594581264,
      "loss": 5.3657,
      "mean_token_accuracy": 0.19021448791027068,
      "num_tokens": 11698777.0,
      "step": 5240
    },
    {
      "entropy": 5.620222473144532,
      "epoch": 0.5890941764474645,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004974242053442461,
      "loss": 5.4614,
      "mean_token_accuracy": 0.18774818629026413,
      "num_tokens": 11710922.0,
      "step": 5245
    },
    {
      "entropy": 5.650982141494751,
      "epoch": 0.589655753355422,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004974181441482927,
      "loss": 5.5099,
      "mean_token_accuracy": 0.18162333220243454,
      "num_tokens": 11721450.0,
      "step": 5250
    },
    {
      "entropy": 5.573250198364258,
      "epoch": 0.5902173302633795,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004974120758704595,
      "loss": 5.4274,
      "mean_token_accuracy": 0.1852564662694931,
      "num_tokens": 11732458.0,
      "step": 5255
    },
    {
      "entropy": 5.578911018371582,
      "epoch": 0.5907789071713371,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004974060005109394,
      "loss": 5.3866,
      "mean_token_accuracy": 0.19007213711738585,
      "num_tokens": 11742649.0,
      "step": 5260
    },
    {
      "entropy": 5.617689704895019,
      "epoch": 0.5913404840792946,
      "grad_norm": 0.890625,
      "learning_rate": 0.000497399918069926,
      "loss": 5.4233,
      "mean_token_accuracy": 0.18863955438137053,
      "num_tokens": 11753811.0,
      "step": 5265
    },
    {
      "entropy": 5.661263370513916,
      "epoch": 0.5919020609872522,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004973938285476131,
      "loss": 5.5362,
      "mean_token_accuracy": 0.1776759296655655,
      "num_tokens": 11763965.0,
      "step": 5270
    },
    {
      "entropy": 5.678827333450317,
      "epoch": 0.5924636378952097,
      "grad_norm": 0.875,
      "learning_rate": 0.0004973877319441943,
      "loss": 5.5265,
      "mean_token_accuracy": 0.18722453862428665,
      "num_tokens": 11774323.0,
      "step": 5275
    },
    {
      "entropy": 5.673748731613159,
      "epoch": 0.5930252148031673,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004973816282598639,
      "loss": 5.4969,
      "mean_token_accuracy": 0.18301146030426024,
      "num_tokens": 11786157.0,
      "step": 5280
    },
    {
      "entropy": 5.64190354347229,
      "epoch": 0.5935867917111248,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004973755174948162,
      "loss": 5.4679,
      "mean_token_accuracy": 0.18977839052677153,
      "num_tokens": 11797344.0,
      "step": 5285
    },
    {
      "entropy": 5.510661315917969,
      "epoch": 0.5941483686190824,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004973693996492458,
      "loss": 5.4122,
      "mean_token_accuracy": 0.18232263326644899,
      "num_tokens": 11808149.0,
      "step": 5290
    },
    {
      "entropy": 5.632110357284546,
      "epoch": 0.5947099455270399,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004973632747233475,
      "loss": 5.4315,
      "mean_token_accuracy": 0.18891571164131166,
      "num_tokens": 11820203.0,
      "step": 5295
    },
    {
      "entropy": 5.680429220199585,
      "epoch": 0.5952715224349975,
      "grad_norm": 0.875,
      "learning_rate": 0.0004973571427173163,
      "loss": 5.5328,
      "mean_token_accuracy": 0.18260134756565094,
      "num_tokens": 11830172.0,
      "step": 5300
    },
    {
      "entropy": 5.667116737365722,
      "epoch": 0.595833099342955,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004973510036313473,
      "loss": 5.4756,
      "mean_token_accuracy": 0.1869452863931656,
      "num_tokens": 11840203.0,
      "step": 5305
    },
    {
      "entropy": 5.64485239982605,
      "epoch": 0.5963946762509126,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004973448574656362,
      "loss": 5.4228,
      "mean_token_accuracy": 0.18376448154449462,
      "num_tokens": 11851002.0,
      "step": 5310
    },
    {
      "entropy": 5.601604366302491,
      "epoch": 0.5969562531588701,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004973387042203786,
      "loss": 5.4332,
      "mean_token_accuracy": 0.18251495212316513,
      "num_tokens": 11861559.0,
      "step": 5315
    },
    {
      "entropy": 5.611603164672852,
      "epoch": 0.5975178300668277,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004973325438957704,
      "loss": 5.5169,
      "mean_token_accuracy": 0.1836209237575531,
      "num_tokens": 11872535.0,
      "step": 5320
    },
    {
      "entropy": 5.731847190856934,
      "epoch": 0.5980794069747852,
      "grad_norm": 0.875,
      "learning_rate": 0.0004973263764920079,
      "loss": 5.5284,
      "mean_token_accuracy": 0.18352710008621215,
      "num_tokens": 11884308.0,
      "step": 5325
    },
    {
      "entropy": 5.7274538516998295,
      "epoch": 0.5986409838827428,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004973202020092871,
      "loss": 5.5931,
      "mean_token_accuracy": 0.17703742533922195,
      "num_tokens": 11896136.0,
      "step": 5330
    },
    {
      "entropy": 5.641735029220581,
      "epoch": 0.5992025607907003,
      "grad_norm": 0.9375,
      "learning_rate": 0.000497314020447805,
      "loss": 5.411,
      "mean_token_accuracy": 0.1960025116801262,
      "num_tokens": 11906497.0,
      "step": 5335
    },
    {
      "entropy": 5.6923949241638185,
      "epoch": 0.5997641376986579,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004973078318077583,
      "loss": 5.5416,
      "mean_token_accuracy": 0.17766277939081193,
      "num_tokens": 11917872.0,
      "step": 5340
    },
    {
      "entropy": 5.711421394348145,
      "epoch": 0.6003257146066153,
      "grad_norm": 0.921875,
      "learning_rate": 0.000497301636089344,
      "loss": 5.5655,
      "mean_token_accuracy": 0.1786518082022667,
      "num_tokens": 11929727.0,
      "step": 5345
    },
    {
      "entropy": 5.652766180038452,
      "epoch": 0.6008872915145729,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004972954332927592,
      "loss": 5.3497,
      "mean_token_accuracy": 0.1854843944311142,
      "num_tokens": 11939996.0,
      "step": 5350
    },
    {
      "entropy": 5.585961151123047,
      "epoch": 0.6014488684225304,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004972892234182018,
      "loss": 5.4457,
      "mean_token_accuracy": 0.18749601542949676,
      "num_tokens": 11950683.0,
      "step": 5355
    },
    {
      "entropy": 5.538220310211182,
      "epoch": 0.602010445330488,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004972830064658692,
      "loss": 5.4325,
      "mean_token_accuracy": 0.18115904331207275,
      "num_tokens": 11962447.0,
      "step": 5360
    },
    {
      "entropy": 5.707037019729614,
      "epoch": 0.6025720222384455,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004972767824359596,
      "loss": 5.522,
      "mean_token_accuracy": 0.18260627686977388,
      "num_tokens": 11972740.0,
      "step": 5365
    },
    {
      "entropy": 5.652205228805542,
      "epoch": 0.6031335991464031,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004972705513286709,
      "loss": 5.407,
      "mean_token_accuracy": 0.18472368270158768,
      "num_tokens": 11983292.0,
      "step": 5370
    },
    {
      "entropy": 5.692061424255371,
      "epoch": 0.6036951760543606,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004972643131442017,
      "loss": 5.5937,
      "mean_token_accuracy": 0.17778845131397247,
      "num_tokens": 11996384.0,
      "step": 5375
    },
    {
      "entropy": 5.715888023376465,
      "epoch": 0.6042567529623182,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004972580678827506,
      "loss": 5.5318,
      "mean_token_accuracy": 0.19098808318376542,
      "num_tokens": 12006938.0,
      "step": 5380
    },
    {
      "entropy": 5.542979669570923,
      "epoch": 0.6048183298702757,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004972518155445163,
      "loss": 5.3636,
      "mean_token_accuracy": 0.18590497523546218,
      "num_tokens": 12018729.0,
      "step": 5385
    },
    {
      "entropy": 5.792513847351074,
      "epoch": 0.6053799067782333,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004972455561296982,
      "loss": 5.7131,
      "mean_token_accuracy": 0.17683746367692948,
      "num_tokens": 12029625.0,
      "step": 5390
    },
    {
      "entropy": 5.6201722621917725,
      "epoch": 0.6059414836861908,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004972392896384952,
      "loss": 5.4164,
      "mean_token_accuracy": 0.18248458951711655,
      "num_tokens": 12040243.0,
      "step": 5395
    },
    {
      "entropy": 5.720900344848633,
      "epoch": 0.6065030605941484,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004972330160711071,
      "loss": 5.7045,
      "mean_token_accuracy": 0.17212155759334563,
      "num_tokens": 12051961.0,
      "step": 5400
    },
    {
      "entropy": 5.716148376464844,
      "epoch": 0.6070646375021059,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004972267354277336,
      "loss": 5.5617,
      "mean_token_accuracy": 0.18395792543888093,
      "num_tokens": 12063969.0,
      "step": 5405
    },
    {
      "entropy": 5.677671957015991,
      "epoch": 0.6076262144100635,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004972204477085746,
      "loss": 5.4424,
      "mean_token_accuracy": 0.18653500825166702,
      "num_tokens": 12075682.0,
      "step": 5410
    },
    {
      "entropy": 5.717992162704467,
      "epoch": 0.608187791318021,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004972141529138303,
      "loss": 5.5725,
      "mean_token_accuracy": 0.18072832524776458,
      "num_tokens": 12086595.0,
      "step": 5415
    },
    {
      "entropy": 5.652366399765015,
      "epoch": 0.6087493682259786,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004972078510437013,
      "loss": 5.4478,
      "mean_token_accuracy": 0.18351133912801743,
      "num_tokens": 12099421.0,
      "step": 5420
    },
    {
      "entropy": 5.688243627548218,
      "epoch": 0.6093109451339361,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004972015420983882,
      "loss": 5.5789,
      "mean_token_accuracy": 0.18049831837415695,
      "num_tokens": 12109680.0,
      "step": 5425
    },
    {
      "entropy": 5.59382586479187,
      "epoch": 0.6098725220418937,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004971952260780915,
      "loss": 5.4402,
      "mean_token_accuracy": 0.1922355830669403,
      "num_tokens": 12121196.0,
      "step": 5430
    },
    {
      "entropy": 5.60436520576477,
      "epoch": 0.6104340989498512,
      "grad_norm": 0.796875,
      "learning_rate": 0.000497188902983013,
      "loss": 5.3938,
      "mean_token_accuracy": 0.18242020308971404,
      "num_tokens": 12132262.0,
      "step": 5435
    },
    {
      "entropy": 5.695270586013794,
      "epoch": 0.6109956758578087,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004971825728133535,
      "loss": 5.5711,
      "mean_token_accuracy": 0.18374685794115067,
      "num_tokens": 12143747.0,
      "step": 5440
    },
    {
      "entropy": 5.801350402832031,
      "epoch": 0.6115572527657662,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004971762355693146,
      "loss": 5.6598,
      "mean_token_accuracy": 0.17659540474414825,
      "num_tokens": 12155887.0,
      "step": 5445
    },
    {
      "entropy": 5.595334148406982,
      "epoch": 0.6121188296737238,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004971698912510983,
      "loss": 5.4522,
      "mean_token_accuracy": 0.1790104940533638,
      "num_tokens": 12166718.0,
      "step": 5450
    },
    {
      "entropy": 5.698393392562866,
      "epoch": 0.6126804065816813,
      "grad_norm": 0.75,
      "learning_rate": 0.0004971635398589066,
      "loss": 5.5895,
      "mean_token_accuracy": 0.18182177245616912,
      "num_tokens": 12178540.0,
      "step": 5455
    },
    {
      "entropy": 5.821663808822632,
      "epoch": 0.6132419834896389,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004971571813929415,
      "loss": 5.6333,
      "mean_token_accuracy": 0.17573781311511993,
      "num_tokens": 12189820.0,
      "step": 5460
    },
    {
      "entropy": 5.648117637634277,
      "epoch": 0.6138035603975964,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004971508158534056,
      "loss": 5.5348,
      "mean_token_accuracy": 0.1826782926917076,
      "num_tokens": 12201679.0,
      "step": 5465
    },
    {
      "entropy": 5.64471583366394,
      "epoch": 0.614365137305554,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004971444432405015,
      "loss": 5.5409,
      "mean_token_accuracy": 0.1761327564716339,
      "num_tokens": 12213793.0,
      "step": 5470
    },
    {
      "entropy": 5.7703358173370365,
      "epoch": 0.6149267142135115,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004971380635544323,
      "loss": 5.6236,
      "mean_token_accuracy": 0.18235935866832734,
      "num_tokens": 12224519.0,
      "step": 5475
    },
    {
      "entropy": 5.810359096527099,
      "epoch": 0.6154882911214691,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004971316767954009,
      "loss": 5.603,
      "mean_token_accuracy": 0.1789076015353203,
      "num_tokens": 12236401.0,
      "step": 5480
    },
    {
      "entropy": 5.678476762771607,
      "epoch": 0.6160498680294266,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004971252829636108,
      "loss": 5.6391,
      "mean_token_accuracy": 0.17531829327344894,
      "num_tokens": 12247670.0,
      "step": 5485
    },
    {
      "entropy": 5.641606616973877,
      "epoch": 0.6166114449373842,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004971188820592656,
      "loss": 5.5054,
      "mean_token_accuracy": 0.1811036080121994,
      "num_tokens": 12258877.0,
      "step": 5490
    },
    {
      "entropy": 5.6840931415557865,
      "epoch": 0.6171730218453417,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000497112474082569,
      "loss": 5.4868,
      "mean_token_accuracy": 0.1805300682783127,
      "num_tokens": 12268507.0,
      "step": 5495
    },
    {
      "entropy": 5.6497025966644285,
      "epoch": 0.6177345987532993,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000497106059033725,
      "loss": 5.4745,
      "mean_token_accuracy": 0.17916177362203597,
      "num_tokens": 12280420.0,
      "step": 5500
    },
    {
      "entropy": 5.595959234237671,
      "epoch": 0.6182961756612568,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004970996369129381,
      "loss": 5.47,
      "mean_token_accuracy": 0.1789115697145462,
      "num_tokens": 12290876.0,
      "step": 5505
    },
    {
      "entropy": 5.616681003570557,
      "epoch": 0.6188577525692144,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004970932077204125,
      "loss": 5.3739,
      "mean_token_accuracy": 0.19202976375818254,
      "num_tokens": 12301831.0,
      "step": 5510
    },
    {
      "entropy": 5.578046464920044,
      "epoch": 0.6194193294771719,
      "grad_norm": 0.9375,
      "learning_rate": 0.000497086771456353,
      "loss": 5.4092,
      "mean_token_accuracy": 0.18497390747070314,
      "num_tokens": 12312678.0,
      "step": 5515
    },
    {
      "entropy": 5.707816553115845,
      "epoch": 0.6199809063851295,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004970803281209646,
      "loss": 5.5873,
      "mean_token_accuracy": 0.1760278731584549,
      "num_tokens": 12322905.0,
      "step": 5520
    },
    {
      "entropy": 5.635919761657715,
      "epoch": 0.620542483293087,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004970738777144524,
      "loss": 5.4468,
      "mean_token_accuracy": 0.18116901367902755,
      "num_tokens": 12333091.0,
      "step": 5525
    },
    {
      "entropy": 5.661448192596436,
      "epoch": 0.6211040602010446,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004970674202370218,
      "loss": 5.4355,
      "mean_token_accuracy": 0.18937963992357254,
      "num_tokens": 12346310.0,
      "step": 5530
    },
    {
      "entropy": 5.5523984909057615,
      "epoch": 0.621665637109002,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004970609556888785,
      "loss": 5.3682,
      "mean_token_accuracy": 0.18358201831579207,
      "num_tokens": 12357138.0,
      "step": 5535
    },
    {
      "entropy": 5.698223304748535,
      "epoch": 0.6222272140169596,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004970544840702282,
      "loss": 5.4631,
      "mean_token_accuracy": 0.18379671722650529,
      "num_tokens": 12367382.0,
      "step": 5540
    },
    {
      "entropy": 5.583384656906128,
      "epoch": 0.6227887909249171,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004970480053812771,
      "loss": 5.4413,
      "mean_token_accuracy": 0.18021013885736464,
      "num_tokens": 12380357.0,
      "step": 5545
    },
    {
      "entropy": 5.719734191894531,
      "epoch": 0.6233503678328747,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004970415196222312,
      "loss": 5.6194,
      "mean_token_accuracy": 0.17542009055614471,
      "num_tokens": 12392187.0,
      "step": 5550
    },
    {
      "entropy": 5.688438844680786,
      "epoch": 0.6239119447408322,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004970350267932973,
      "loss": 5.506,
      "mean_token_accuracy": 0.18586140424013137,
      "num_tokens": 12402924.0,
      "step": 5555
    },
    {
      "entropy": 5.718354654312134,
      "epoch": 0.6244735216487898,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000497028526894682,
      "loss": 5.6159,
      "mean_token_accuracy": 0.17857843786478042,
      "num_tokens": 12413554.0,
      "step": 5560
    },
    {
      "entropy": 5.649330759048462,
      "epoch": 0.6250350985567473,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004970220199265923,
      "loss": 5.5065,
      "mean_token_accuracy": 0.1867228090763092,
      "num_tokens": 12424525.0,
      "step": 5565
    },
    {
      "entropy": 5.614630699157715,
      "epoch": 0.6255966754647049,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004970155058892354,
      "loss": 5.5027,
      "mean_token_accuracy": 0.18657971173524857,
      "num_tokens": 12436405.0,
      "step": 5570
    },
    {
      "entropy": 5.719746494293213,
      "epoch": 0.6261582523726624,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004970089847828187,
      "loss": 5.5689,
      "mean_token_accuracy": 0.1784456565976143,
      "num_tokens": 12446909.0,
      "step": 5575
    },
    {
      "entropy": 5.705295562744141,
      "epoch": 0.62671982928062,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004970024566075497,
      "loss": 5.5364,
      "mean_token_accuracy": 0.1828928768634796,
      "num_tokens": 12458739.0,
      "step": 5580
    },
    {
      "entropy": 5.609265422821045,
      "epoch": 0.6272814061885775,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004969959213636365,
      "loss": 5.4823,
      "mean_token_accuracy": 0.18272341936826705,
      "num_tokens": 12469529.0,
      "step": 5585
    },
    {
      "entropy": 5.734248256683349,
      "epoch": 0.6278429830965351,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004969893790512871,
      "loss": 5.4995,
      "mean_token_accuracy": 0.18193261921405793,
      "num_tokens": 12480067.0,
      "step": 5590
    },
    {
      "entropy": 5.64008641242981,
      "epoch": 0.6284045600044926,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004969828296707097,
      "loss": 5.3917,
      "mean_token_accuracy": 0.1919093370437622,
      "num_tokens": 12491043.0,
      "step": 5595
    },
    {
      "entropy": 5.528051137924194,
      "epoch": 0.6289661369124502,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004969762732221129,
      "loss": 5.3256,
      "mean_token_accuracy": 0.18950122892856597,
      "num_tokens": 12501852.0,
      "step": 5600
    },
    {
      "entropy": 5.666510152816772,
      "epoch": 0.6295277138204077,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004969697097057054,
      "loss": 5.6363,
      "mean_token_accuracy": 0.17815123796463012,
      "num_tokens": 12512696.0,
      "step": 5605
    },
    {
      "entropy": 5.713888835906983,
      "epoch": 0.6300892907283653,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004969631391216964,
      "loss": 5.4869,
      "mean_token_accuracy": 0.18827647268772124,
      "num_tokens": 12523643.0,
      "step": 5610
    },
    {
      "entropy": 5.638426303863525,
      "epoch": 0.6306508676363228,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004969565614702948,
      "loss": 5.4586,
      "mean_token_accuracy": 0.18541952967643738,
      "num_tokens": 12534827.0,
      "step": 5615
    },
    {
      "entropy": 5.526660060882568,
      "epoch": 0.6312124445442804,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004969499767517104,
      "loss": 5.3259,
      "mean_token_accuracy": 0.19049309492111205,
      "num_tokens": 12545977.0,
      "step": 5620
    },
    {
      "entropy": 5.535588264465332,
      "epoch": 0.6317740214522379,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004969433849661525,
      "loss": 5.4065,
      "mean_token_accuracy": 0.18236255049705505,
      "num_tokens": 12555890.0,
      "step": 5625
    },
    {
      "entropy": 5.6138678073883055,
      "epoch": 0.6323355983601955,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004969367861138313,
      "loss": 5.4525,
      "mean_token_accuracy": 0.1870817169547081,
      "num_tokens": 12566412.0,
      "step": 5630
    },
    {
      "entropy": 5.6797936916351315,
      "epoch": 0.6328971752681529,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004969301801949565,
      "loss": 5.4972,
      "mean_token_accuracy": 0.18230903148651123,
      "num_tokens": 12577175.0,
      "step": 5635
    },
    {
      "entropy": 5.583798217773437,
      "epoch": 0.6334587521761105,
      "grad_norm": 0.84375,
      "learning_rate": 0.000496923567209739,
      "loss": 5.4185,
      "mean_token_accuracy": 0.18198846876621247,
      "num_tokens": 12588760.0,
      "step": 5640
    },
    {
      "entropy": 5.615443754196167,
      "epoch": 0.634020329084068,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004969169471583889,
      "loss": 5.4868,
      "mean_token_accuracy": 0.18632501363754272,
      "num_tokens": 12599899.0,
      "step": 5645
    },
    {
      "entropy": 5.692332220077515,
      "epoch": 0.6345819059920256,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004969103200411171,
      "loss": 5.4479,
      "mean_token_accuracy": 0.19541003108024596,
      "num_tokens": 12610836.0,
      "step": 5650
    },
    {
      "entropy": 5.589169692993164,
      "epoch": 0.6351434828999831,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004969036858581346,
      "loss": 5.4161,
      "mean_token_accuracy": 0.19111810475587845,
      "num_tokens": 12621855.0,
      "step": 5655
    },
    {
      "entropy": 5.661271619796753,
      "epoch": 0.6357050598079407,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004968970446096527,
      "loss": 5.5076,
      "mean_token_accuracy": 0.18335456550121307,
      "num_tokens": 12632964.0,
      "step": 5660
    },
    {
      "entropy": 5.546975088119507,
      "epoch": 0.6362666367158982,
      "grad_norm": 0.84375,
      "learning_rate": 0.000496890396295883,
      "loss": 5.3536,
      "mean_token_accuracy": 0.1898905813694,
      "num_tokens": 12643717.0,
      "step": 5665
    },
    {
      "entropy": 5.6402575969696045,
      "epoch": 0.6368282136238558,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004968837409170368,
      "loss": 5.4717,
      "mean_token_accuracy": 0.1824107810854912,
      "num_tokens": 12654491.0,
      "step": 5670
    },
    {
      "entropy": 5.616565608978272,
      "epoch": 0.6373897905318133,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004968770784733263,
      "loss": 5.5173,
      "mean_token_accuracy": 0.18348828107118606,
      "num_tokens": 12666128.0,
      "step": 5675
    },
    {
      "entropy": 5.714895868301392,
      "epoch": 0.6379513674397709,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004968704089649636,
      "loss": 5.5828,
      "mean_token_accuracy": 0.18119271397590636,
      "num_tokens": 12677957.0,
      "step": 5680
    },
    {
      "entropy": 5.664091491699219,
      "epoch": 0.6385129443477284,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004968637323921611,
      "loss": 5.4404,
      "mean_token_accuracy": 0.1860332727432251,
      "num_tokens": 12687964.0,
      "step": 5685
    },
    {
      "entropy": 5.537417697906494,
      "epoch": 0.639074521255686,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004968570487551312,
      "loss": 5.4294,
      "mean_token_accuracy": 0.18320983052253723,
      "num_tokens": 12698805.0,
      "step": 5690
    },
    {
      "entropy": 5.623385000228882,
      "epoch": 0.6396360981636435,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004968503580540869,
      "loss": 5.4446,
      "mean_token_accuracy": 0.18390532433986664,
      "num_tokens": 12709942.0,
      "step": 5695
    },
    {
      "entropy": 5.76184663772583,
      "epoch": 0.6401976750716011,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004968436602892411,
      "loss": 5.5208,
      "mean_token_accuracy": 0.18527029901742936,
      "num_tokens": 12721856.0,
      "step": 5700
    },
    {
      "entropy": 5.569256067276001,
      "epoch": 0.6407592519795586,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004968369554608071,
      "loss": 5.3544,
      "mean_token_accuracy": 0.18365223854780197,
      "num_tokens": 12732619.0,
      "step": 5705
    },
    {
      "entropy": 5.610283374786377,
      "epoch": 0.6413208288875162,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004968302435689985,
      "loss": 5.4529,
      "mean_token_accuracy": 0.18736341893672942,
      "num_tokens": 12743478.0,
      "step": 5710
    },
    {
      "entropy": 5.606819677352905,
      "epoch": 0.6418824057954737,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004968235246140288,
      "loss": 5.3918,
      "mean_token_accuracy": 0.18309973329305648,
      "num_tokens": 12754761.0,
      "step": 5715
    },
    {
      "entropy": 5.688323307037353,
      "epoch": 0.6424439827034313,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004968167985961121,
      "loss": 5.4935,
      "mean_token_accuracy": 0.18603483140468596,
      "num_tokens": 12765803.0,
      "step": 5720
    },
    {
      "entropy": 5.647583913803101,
      "epoch": 0.6430055596113888,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004968100655154625,
      "loss": 5.5249,
      "mean_token_accuracy": 0.1832832470536232,
      "num_tokens": 12777882.0,
      "step": 5725
    },
    {
      "entropy": 5.675435543060303,
      "epoch": 0.6435671365193463,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004968033253722943,
      "loss": 5.5392,
      "mean_token_accuracy": 0.17872941941022874,
      "num_tokens": 12788268.0,
      "step": 5730
    },
    {
      "entropy": 5.622819519042968,
      "epoch": 0.6441287134273038,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004967965781668223,
      "loss": 5.5232,
      "mean_token_accuracy": 0.18471589386463166,
      "num_tokens": 12800506.0,
      "step": 5735
    },
    {
      "entropy": 5.625365304946899,
      "epoch": 0.6446902903352614,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004967898238992613,
      "loss": 5.4948,
      "mean_token_accuracy": 0.18443508297204972,
      "num_tokens": 12813040.0,
      "step": 5740
    },
    {
      "entropy": 5.721566677093506,
      "epoch": 0.6452518672432189,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004967830625698262,
      "loss": 5.5295,
      "mean_token_accuracy": 0.1862854167819023,
      "num_tokens": 12825534.0,
      "step": 5745
    },
    {
      "entropy": 5.5815990447998045,
      "epoch": 0.6458134441511765,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004967762941787323,
      "loss": 5.4311,
      "mean_token_accuracy": 0.18513077795505523,
      "num_tokens": 12835882.0,
      "step": 5750
    },
    {
      "entropy": 5.627452182769775,
      "epoch": 0.646375021059134,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004967695187261952,
      "loss": 5.6073,
      "mean_token_accuracy": 0.1751113072037697,
      "num_tokens": 12847663.0,
      "step": 5755
    },
    {
      "entropy": 5.839111089706421,
      "epoch": 0.6469365979670916,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004967627362124306,
      "loss": 5.7341,
      "mean_token_accuracy": 0.17808868288993834,
      "num_tokens": 12860789.0,
      "step": 5760
    },
    {
      "entropy": 5.681763696670532,
      "epoch": 0.6474981748750491,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004967559466376546,
      "loss": 5.4811,
      "mean_token_accuracy": 0.18174173831939697,
      "num_tokens": 12872509.0,
      "step": 5765
    },
    {
      "entropy": 5.576981925964356,
      "epoch": 0.6480597517830067,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004967491500020831,
      "loss": 5.4693,
      "mean_token_accuracy": 0.18398525267839433,
      "num_tokens": 12884321.0,
      "step": 5770
    },
    {
      "entropy": 5.507149696350098,
      "epoch": 0.6486213286909642,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004967423463059328,
      "loss": 5.3747,
      "mean_token_accuracy": 0.19095166623592377,
      "num_tokens": 12894559.0,
      "step": 5775
    },
    {
      "entropy": 5.585085010528564,
      "epoch": 0.6491829055989218,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004967355355494201,
      "loss": 5.474,
      "mean_token_accuracy": 0.18762808591127395,
      "num_tokens": 12904934.0,
      "step": 5780
    },
    {
      "entropy": 5.757120370864868,
      "epoch": 0.6497444825068793,
      "grad_norm": 0.890625,
      "learning_rate": 0.000496728717732762,
      "loss": 5.5776,
      "mean_token_accuracy": 0.1789213702082634,
      "num_tokens": 12916551.0,
      "step": 5785
    },
    {
      "entropy": 5.660614824295044,
      "epoch": 0.6503060594148369,
      "grad_norm": 0.875,
      "learning_rate": 0.0004967218928561755,
      "loss": 5.474,
      "mean_token_accuracy": 0.18255353420972825,
      "num_tokens": 12928097.0,
      "step": 5790
    },
    {
      "entropy": 5.587710380554199,
      "epoch": 0.6508676363227944,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004967150609198779,
      "loss": 5.3655,
      "mean_token_accuracy": 0.19426413774490356,
      "num_tokens": 12939161.0,
      "step": 5795
    },
    {
      "entropy": 5.529241323471069,
      "epoch": 0.651429213230752,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004967082219240869,
      "loss": 5.408,
      "mean_token_accuracy": 0.1858175128698349,
      "num_tokens": 12950862.0,
      "step": 5800
    },
    {
      "entropy": 5.644560480117798,
      "epoch": 0.6519907901387095,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004967013758690199,
      "loss": 5.4523,
      "mean_token_accuracy": 0.18228911608457565,
      "num_tokens": 12962333.0,
      "step": 5805
    },
    {
      "entropy": 5.6136664867401125,
      "epoch": 0.6525523670466671,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004966945227548953,
      "loss": 5.4248,
      "mean_token_accuracy": 0.18576802015304567,
      "num_tokens": 12973136.0,
      "step": 5810
    },
    {
      "entropy": 5.635822057723999,
      "epoch": 0.6531139439546246,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004966876625819309,
      "loss": 5.4649,
      "mean_token_accuracy": 0.1858026549220085,
      "num_tokens": 12983602.0,
      "step": 5815
    },
    {
      "entropy": 5.618609476089477,
      "epoch": 0.6536755208625822,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004966807953503456,
      "loss": 5.3849,
      "mean_token_accuracy": 0.191538205742836,
      "num_tokens": 12994129.0,
      "step": 5820
    },
    {
      "entropy": 5.609147596359253,
      "epoch": 0.6542370977705396,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004966739210603578,
      "loss": 5.44,
      "mean_token_accuracy": 0.18199084401130677,
      "num_tokens": 13005071.0,
      "step": 5825
    },
    {
      "entropy": 5.719333600997925,
      "epoch": 0.6547986746784972,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004966670397121862,
      "loss": 5.6166,
      "mean_token_accuracy": 0.18030202090740205,
      "num_tokens": 13016202.0,
      "step": 5830
    },
    {
      "entropy": 5.697877311706543,
      "epoch": 0.6553602515864547,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004966601513060502,
      "loss": 5.6002,
      "mean_token_accuracy": 0.1707790046930313,
      "num_tokens": 13028468.0,
      "step": 5835
    },
    {
      "entropy": 5.664613342285156,
      "epoch": 0.6559218284944123,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000496653255842169,
      "loss": 5.4898,
      "mean_token_accuracy": 0.18285662829875945,
      "num_tokens": 13039667.0,
      "step": 5840
    },
    {
      "entropy": 5.634717416763306,
      "epoch": 0.6564834054023698,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004966463533207621,
      "loss": 5.5443,
      "mean_token_accuracy": 0.17505890876054764,
      "num_tokens": 13051583.0,
      "step": 5845
    },
    {
      "entropy": 5.6710968017578125,
      "epoch": 0.6570449823103274,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004966394437420493,
      "loss": 5.4938,
      "mean_token_accuracy": 0.18007268905639648,
      "num_tokens": 13062401.0,
      "step": 5850
    },
    {
      "entropy": 5.549707126617432,
      "epoch": 0.6576065592182849,
      "grad_norm": 0.875,
      "learning_rate": 0.0004966325271062508,
      "loss": 5.3819,
      "mean_token_accuracy": 0.1912681207060814,
      "num_tokens": 13074219.0,
      "step": 5855
    },
    {
      "entropy": 5.688369846343994,
      "epoch": 0.6581681361262425,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004966256034135866,
      "loss": 5.5685,
      "mean_token_accuracy": 0.1818229228258133,
      "num_tokens": 13086073.0,
      "step": 5860
    },
    {
      "entropy": 5.674010515213013,
      "epoch": 0.6587297130342,
      "grad_norm": 0.875,
      "learning_rate": 0.0004966186726642772,
      "loss": 5.4881,
      "mean_token_accuracy": 0.17973924726247786,
      "num_tokens": 13096340.0,
      "step": 5865
    },
    {
      "entropy": 5.592394208908081,
      "epoch": 0.6592912899421576,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004966117348585434,
      "loss": 5.4616,
      "mean_token_accuracy": 0.18663387447595597,
      "num_tokens": 13107273.0,
      "step": 5870
    },
    {
      "entropy": 5.579114198684692,
      "epoch": 0.6598528668501151,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004966047899966059,
      "loss": 5.4325,
      "mean_token_accuracy": 0.18317066729068757,
      "num_tokens": 13117388.0,
      "step": 5875
    },
    {
      "entropy": 5.629977655410767,
      "epoch": 0.6604144437580727,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000496597838078686,
      "loss": 5.5449,
      "mean_token_accuracy": 0.18076352626085282,
      "num_tokens": 13128720.0,
      "step": 5880
    },
    {
      "entropy": 5.673728704452515,
      "epoch": 0.6609760206660302,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000496590879105005,
      "loss": 5.4676,
      "mean_token_accuracy": 0.18354215174913407,
      "num_tokens": 13139405.0,
      "step": 5885
    },
    {
      "entropy": 5.636421537399292,
      "epoch": 0.6615375975739878,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004965839130757845,
      "loss": 5.5428,
      "mean_token_accuracy": 0.18387479782104493,
      "num_tokens": 13151408.0,
      "step": 5890
    },
    {
      "entropy": 5.668087720870972,
      "epoch": 0.6620991744819453,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004965769399912462,
      "loss": 5.4475,
      "mean_token_accuracy": 0.1862862527370453,
      "num_tokens": 13162592.0,
      "step": 5895
    },
    {
      "entropy": 5.661455202102661,
      "epoch": 0.6626607513899029,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004965699598516122,
      "loss": 5.4971,
      "mean_token_accuracy": 0.18750617504119874,
      "num_tokens": 13173040.0,
      "step": 5900
    },
    {
      "entropy": 5.624880647659301,
      "epoch": 0.6632223282978604,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004965629726571047,
      "loss": 5.5644,
      "mean_token_accuracy": 0.18215004950761796,
      "num_tokens": 13185475.0,
      "step": 5905
    },
    {
      "entropy": 5.646581649780273,
      "epoch": 0.663783905205818,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004965559784079463,
      "loss": 5.4266,
      "mean_token_accuracy": 0.19297462105751037,
      "num_tokens": 13196481.0,
      "step": 5910
    },
    {
      "entropy": 5.5824360847473145,
      "epoch": 0.6643454821137755,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004965489771043596,
      "loss": 5.3139,
      "mean_token_accuracy": 0.19192130714654923,
      "num_tokens": 13207947.0,
      "step": 5915
    },
    {
      "entropy": 5.49367733001709,
      "epoch": 0.664907059021733,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004965419687465676,
      "loss": 5.363,
      "mean_token_accuracy": 0.20135646611452102,
      "num_tokens": 13219392.0,
      "step": 5920
    },
    {
      "entropy": 5.539564180374145,
      "epoch": 0.6654686359296905,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004965349533347934,
      "loss": 5.4093,
      "mean_token_accuracy": 0.19332722276449205,
      "num_tokens": 13231556.0,
      "step": 5925
    },
    {
      "entropy": 5.6006489276885985,
      "epoch": 0.6660302128376481,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004965279308692602,
      "loss": 5.3613,
      "mean_token_accuracy": 0.18630005568265914,
      "num_tokens": 13241287.0,
      "step": 5930
    },
    {
      "entropy": 5.566129541397094,
      "epoch": 0.6665917897456056,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004965209013501919,
      "loss": 5.3882,
      "mean_token_accuracy": 0.1785157650709152,
      "num_tokens": 13254120.0,
      "step": 5935
    },
    {
      "entropy": 5.6418122291564945,
      "epoch": 0.6671533666535632,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004965138647778122,
      "loss": 5.479,
      "mean_token_accuracy": 0.1813420072197914,
      "num_tokens": 13265745.0,
      "step": 5940
    },
    {
      "entropy": 5.741047620773315,
      "epoch": 0.6677149435615207,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000496506821152345,
      "loss": 5.5227,
      "mean_token_accuracy": 0.18122239857912065,
      "num_tokens": 13277791.0,
      "step": 5945
    },
    {
      "entropy": 5.565553379058838,
      "epoch": 0.6682765204694783,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004964997704740147,
      "loss": 5.4042,
      "mean_token_accuracy": 0.18556176573038102,
      "num_tokens": 13288078.0,
      "step": 5950
    },
    {
      "entropy": 5.585549449920654,
      "epoch": 0.6688380973774358,
      "grad_norm": 0.78515625,
      "learning_rate": 0.000496492712743046,
      "loss": 5.4356,
      "mean_token_accuracy": 0.18918974101543426,
      "num_tokens": 13300547.0,
      "step": 5955
    },
    {
      "entropy": 5.648870992660522,
      "epoch": 0.6693996742853934,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004964856479596632,
      "loss": 5.4988,
      "mean_token_accuracy": 0.18634683936834334,
      "num_tokens": 13310339.0,
      "step": 5960
    },
    {
      "entropy": 5.70676703453064,
      "epoch": 0.6699612511933509,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004964785761240915,
      "loss": 5.4684,
      "mean_token_accuracy": 0.1832331269979477,
      "num_tokens": 13321294.0,
      "step": 5965
    },
    {
      "entropy": 5.659217786788941,
      "epoch": 0.6705228281013085,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004964714972365562,
      "loss": 5.5019,
      "mean_token_accuracy": 0.18534504622220993,
      "num_tokens": 13332608.0,
      "step": 5970
    },
    {
      "entropy": 5.624098920822144,
      "epoch": 0.671084405009266,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004964644112972824,
      "loss": 5.4483,
      "mean_token_accuracy": 0.19581855982542037,
      "num_tokens": 13344169.0,
      "step": 5975
    },
    {
      "entropy": 5.517456293106079,
      "epoch": 0.6716459819172236,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004964573183064959,
      "loss": 5.4071,
      "mean_token_accuracy": 0.18526545614004136,
      "num_tokens": 13355713.0,
      "step": 5980
    },
    {
      "entropy": 5.667885208129883,
      "epoch": 0.6722075588251811,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004964502182644225,
      "loss": 5.4951,
      "mean_token_accuracy": 0.18628591895103455,
      "num_tokens": 13366644.0,
      "step": 5985
    },
    {
      "entropy": 5.644480562210083,
      "epoch": 0.6727691357331387,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004964431111712883,
      "loss": 5.4678,
      "mean_token_accuracy": 0.18084282875061036,
      "num_tokens": 13378125.0,
      "step": 5990
    },
    {
      "entropy": 5.6003001689910885,
      "epoch": 0.6733307126410962,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004964359970273195,
      "loss": 5.5043,
      "mean_token_accuracy": 0.1825743645429611,
      "num_tokens": 13389681.0,
      "step": 5995
    },
    {
      "entropy": 5.728780269622803,
      "epoch": 0.6738922895490538,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004964288758327426,
      "loss": 5.509,
      "mean_token_accuracy": 0.1827269658446312,
      "num_tokens": 13399814.0,
      "step": 6000
    },
    {
      "epoch": 0.6738922895490538,
      "eval_entropy": 5.471465606187649,
      "eval_loss": 5.484393119812012,
      "eval_mean_token_accuracy": 0.1914711880555527,
      "eval_num_tokens": 13399814.0,
      "eval_runtime": 18.2619,
      "eval_samples_per_second": 1698.237,
      "eval_steps_per_second": 212.3,
      "step": 6000
    },
    {
      "entropy": 5.686691474914551,
      "epoch": 0.6744538664570113,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004964217475877846,
      "loss": 5.5245,
      "mean_token_accuracy": 0.18244125843048095,
      "num_tokens": 13410614.0,
      "step": 6005
    },
    {
      "entropy": 5.609695959091186,
      "epoch": 0.6750154433649689,
      "grad_norm": 0.984375,
      "learning_rate": 0.000496414612292672,
      "loss": 5.5263,
      "mean_token_accuracy": 0.18162218034267424,
      "num_tokens": 13421392.0,
      "step": 6010
    },
    {
      "entropy": 5.691805219650268,
      "epoch": 0.6755770202729264,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004964074699476325,
      "loss": 5.563,
      "mean_token_accuracy": 0.18037434816360473,
      "num_tokens": 13432004.0,
      "step": 6015
    },
    {
      "entropy": 5.662922954559326,
      "epoch": 0.6761385971808839,
      "grad_norm": 0.875,
      "learning_rate": 0.0004964003205528933,
      "loss": 5.4041,
      "mean_token_accuracy": 0.19028606861829758,
      "num_tokens": 13442450.0,
      "step": 6020
    },
    {
      "entropy": 5.617747402191162,
      "epoch": 0.6767001740888414,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004963931641086819,
      "loss": 5.4521,
      "mean_token_accuracy": 0.18333385586738588,
      "num_tokens": 13453848.0,
      "step": 6025
    },
    {
      "entropy": 5.565561103820801,
      "epoch": 0.677261750996799,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004963860006152262,
      "loss": 5.4563,
      "mean_token_accuracy": 0.19009874016046524,
      "num_tokens": 13465119.0,
      "step": 6030
    },
    {
      "entropy": 5.574697637557984,
      "epoch": 0.6778233279047565,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004963788300727546,
      "loss": 5.4296,
      "mean_token_accuracy": 0.18604847639799119,
      "num_tokens": 13475937.0,
      "step": 6035
    },
    {
      "entropy": 5.6277101039886475,
      "epoch": 0.6783849048127141,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004963716524814951,
      "loss": 5.4266,
      "mean_token_accuracy": 0.18910760134458543,
      "num_tokens": 13487430.0,
      "step": 6040
    },
    {
      "entropy": 5.657799100875854,
      "epoch": 0.6789464817206716,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004963644678416763,
      "loss": 5.3994,
      "mean_token_accuracy": 0.188941989839077,
      "num_tokens": 13499631.0,
      "step": 6045
    },
    {
      "entropy": 5.597229385375977,
      "epoch": 0.6795080586286292,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000496357276153527,
      "loss": 5.484,
      "mean_token_accuracy": 0.18789518475532532,
      "num_tokens": 13511753.0,
      "step": 6050
    },
    {
      "entropy": 5.663161516189575,
      "epoch": 0.6800696355365867,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004963500774172763,
      "loss": 5.5563,
      "mean_token_accuracy": 0.1755910784006119,
      "num_tokens": 13524228.0,
      "step": 6055
    },
    {
      "entropy": 5.727927207946777,
      "epoch": 0.6806312124445443,
      "grad_norm": 0.81640625,
      "learning_rate": 0.000496342871633153,
      "loss": 5.5677,
      "mean_token_accuracy": 0.18308880180120468,
      "num_tokens": 13535700.0,
      "step": 6060
    },
    {
      "entropy": 5.646354818344117,
      "epoch": 0.6811927893525018,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004963356588013871,
      "loss": 5.5457,
      "mean_token_accuracy": 0.1795308217406273,
      "num_tokens": 13547337.0,
      "step": 6065
    },
    {
      "entropy": 5.550048112869263,
      "epoch": 0.6817543662604594,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004963284389222078,
      "loss": 5.399,
      "mean_token_accuracy": 0.1841079279780388,
      "num_tokens": 13558350.0,
      "step": 6070
    },
    {
      "entropy": 5.648842239379883,
      "epoch": 0.6823159431684169,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004963212119958453,
      "loss": 5.4401,
      "mean_token_accuracy": 0.18983326256275176,
      "num_tokens": 13569230.0,
      "step": 6075
    },
    {
      "entropy": 5.699696397781372,
      "epoch": 0.6828775200763745,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004963139780225295,
      "loss": 5.5281,
      "mean_token_accuracy": 0.18517981469631195,
      "num_tokens": 13580502.0,
      "step": 6080
    },
    {
      "entropy": 5.628199863433838,
      "epoch": 0.683439096984332,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004963067370024909,
      "loss": 5.4405,
      "mean_token_accuracy": 0.19121507555246353,
      "num_tokens": 13591524.0,
      "step": 6085
    },
    {
      "entropy": 5.6541832447052,
      "epoch": 0.6840006738922896,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004962994889359598,
      "loss": 5.4742,
      "mean_token_accuracy": 0.19287072420120238,
      "num_tokens": 13602432.0,
      "step": 6090
    },
    {
      "entropy": 5.550453233718872,
      "epoch": 0.6845622508002471,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004962922338231673,
      "loss": 5.3987,
      "mean_token_accuracy": 0.1944218397140503,
      "num_tokens": 13613052.0,
      "step": 6095
    },
    {
      "entropy": 5.5798016548156735,
      "epoch": 0.6851238277082047,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004962849716643441,
      "loss": 5.4311,
      "mean_token_accuracy": 0.18726307451725005,
      "num_tokens": 13623961.0,
      "step": 6100
    },
    {
      "entropy": 5.565670490264893,
      "epoch": 0.6856854046161622,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004962777024597216,
      "loss": 5.4166,
      "mean_token_accuracy": 0.18439457565546036,
      "num_tokens": 13635285.0,
      "step": 6105
    },
    {
      "entropy": 5.727707719802856,
      "epoch": 0.6862469815241198,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004962704262095314,
      "loss": 5.5558,
      "mean_token_accuracy": 0.18315457850694655,
      "num_tokens": 13647720.0,
      "step": 6110
    },
    {
      "entropy": 5.559944772720337,
      "epoch": 0.6868085584320772,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004962631429140048,
      "loss": 5.421,
      "mean_token_accuracy": 0.18251441270112992,
      "num_tokens": 13659726.0,
      "step": 6115
    },
    {
      "entropy": 5.6885453224182125,
      "epoch": 0.6873701353400348,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000496255852573374,
      "loss": 5.6023,
      "mean_token_accuracy": 0.17781501561403273,
      "num_tokens": 13670332.0,
      "step": 6120
    },
    {
      "entropy": 5.763287734985352,
      "epoch": 0.6879317122479923,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004962485551878711,
      "loss": 5.5508,
      "mean_token_accuracy": 0.17652425169944763,
      "num_tokens": 13682419.0,
      "step": 6125
    },
    {
      "entropy": 5.519194316864014,
      "epoch": 0.6884932891559499,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004962412507577283,
      "loss": 5.3013,
      "mean_token_accuracy": 0.1899576961994171,
      "num_tokens": 13692698.0,
      "step": 6130
    },
    {
      "entropy": 5.5489198684692385,
      "epoch": 0.6890548660639074,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004962339392831783,
      "loss": 5.501,
      "mean_token_accuracy": 0.18194376975297927,
      "num_tokens": 13704791.0,
      "step": 6135
    },
    {
      "entropy": 5.637999105453491,
      "epoch": 0.689616442971865,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000496226620764454,
      "loss": 5.4171,
      "mean_token_accuracy": 0.1895674079656601,
      "num_tokens": 13715860.0,
      "step": 6140
    },
    {
      "entropy": 5.592366313934326,
      "epoch": 0.6901780198798225,
      "grad_norm": 0.953125,
      "learning_rate": 0.000496219295201788,
      "loss": 5.3898,
      "mean_token_accuracy": 0.19269133359193802,
      "num_tokens": 13727029.0,
      "step": 6145
    },
    {
      "entropy": 5.568031597137451,
      "epoch": 0.6907395967877801,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004962119625954141,
      "loss": 5.3954,
      "mean_token_accuracy": 0.1959039807319641,
      "num_tokens": 13737168.0,
      "step": 6150
    },
    {
      "entropy": 5.675641822814941,
      "epoch": 0.6913011736957376,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004962046229455653,
      "loss": 5.5447,
      "mean_token_accuracy": 0.17651770412921905,
      "num_tokens": 13749828.0,
      "step": 6155
    },
    {
      "entropy": 5.58080563545227,
      "epoch": 0.6918627506036952,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004961972762524757,
      "loss": 5.3447,
      "mean_token_accuracy": 0.18753199130296708,
      "num_tokens": 13761617.0,
      "step": 6160
    },
    {
      "entropy": 5.62925353050232,
      "epoch": 0.6924243275116527,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000496189922516379,
      "loss": 5.4227,
      "mean_token_accuracy": 0.1916441112756729,
      "num_tokens": 13772721.0,
      "step": 6165
    },
    {
      "entropy": 5.549926996231079,
      "epoch": 0.6929859044196103,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004961825617375093,
      "loss": 5.395,
      "mean_token_accuracy": 0.183025561273098,
      "num_tokens": 13783215.0,
      "step": 6170
    },
    {
      "entropy": 5.551498174667358,
      "epoch": 0.6935474813275678,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004961751939161011,
      "loss": 5.3594,
      "mean_token_accuracy": 0.19442307204008102,
      "num_tokens": 13793496.0,
      "step": 6175
    },
    {
      "entropy": 5.493255615234375,
      "epoch": 0.6941090582355254,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000496167819052389,
      "loss": 5.2816,
      "mean_token_accuracy": 0.18362998515367507,
      "num_tokens": 13802818.0,
      "step": 6180
    },
    {
      "entropy": 5.749206972122193,
      "epoch": 0.6946706351434829,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004961604371466077,
      "loss": 5.5724,
      "mean_token_accuracy": 0.1818404197692871,
      "num_tokens": 13814449.0,
      "step": 6185
    },
    {
      "entropy": 5.632386827468872,
      "epoch": 0.6952322120514405,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004961530481989923,
      "loss": 5.5194,
      "mean_token_accuracy": 0.18550000339746475,
      "num_tokens": 13825369.0,
      "step": 6190
    },
    {
      "entropy": 5.565806579589844,
      "epoch": 0.695793788959398,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004961456522097782,
      "loss": 5.3664,
      "mean_token_accuracy": 0.18206513822078704,
      "num_tokens": 13837992.0,
      "step": 6195
    },
    {
      "entropy": 5.646046733856201,
      "epoch": 0.6963553658673556,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004961382491792006,
      "loss": 5.4624,
      "mean_token_accuracy": 0.18312713503837585,
      "num_tokens": 13849771.0,
      "step": 6200
    },
    {
      "entropy": 5.621421480178833,
      "epoch": 0.6969169427753131,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004961308391074953,
      "loss": 5.4505,
      "mean_token_accuracy": 0.1850816324353218,
      "num_tokens": 13861997.0,
      "step": 6205
    },
    {
      "entropy": 5.701976299285889,
      "epoch": 0.6974785196832706,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004961234219948986,
      "loss": 5.5396,
      "mean_token_accuracy": 0.18531987220048904,
      "num_tokens": 13872972.0,
      "step": 6210
    },
    {
      "entropy": 5.539809465408325,
      "epoch": 0.6980400965912281,
      "grad_norm": 0.875,
      "learning_rate": 0.0004961159978416461,
      "loss": 5.3187,
      "mean_token_accuracy": 0.1881298005580902,
      "num_tokens": 13883139.0,
      "step": 6215
    },
    {
      "entropy": 5.603868722915649,
      "epoch": 0.6986016734991857,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004961085666479746,
      "loss": 5.4777,
      "mean_token_accuracy": 0.191552098095417,
      "num_tokens": 13894326.0,
      "step": 6220
    },
    {
      "entropy": 5.667512464523315,
      "epoch": 0.6991632504071432,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004961011284141206,
      "loss": 5.5046,
      "mean_token_accuracy": 0.18769108653068542,
      "num_tokens": 13905525.0,
      "step": 6225
    },
    {
      "entropy": 5.56863842010498,
      "epoch": 0.6997248273151008,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004960936831403209,
      "loss": 5.4244,
      "mean_token_accuracy": 0.18609517365694045,
      "num_tokens": 13915973.0,
      "step": 6230
    },
    {
      "entropy": 5.53609824180603,
      "epoch": 0.7002864042230583,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004960862308268126,
      "loss": 5.2595,
      "mean_token_accuracy": 0.19248360395431519,
      "num_tokens": 13926380.0,
      "step": 6235
    },
    {
      "entropy": 5.62336368560791,
      "epoch": 0.7008479811310159,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004960787714738329,
      "loss": 5.4272,
      "mean_token_accuracy": 0.18464776128530502,
      "num_tokens": 13938523.0,
      "step": 6240
    },
    {
      "entropy": 5.575710582733154,
      "epoch": 0.7014095580389734,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004960713050816194,
      "loss": 5.4245,
      "mean_token_accuracy": 0.1909477174282074,
      "num_tokens": 13949343.0,
      "step": 6245
    },
    {
      "entropy": 5.6513666152954105,
      "epoch": 0.701971134946931,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004960638316504098,
      "loss": 5.5401,
      "mean_token_accuracy": 0.18556744158267974,
      "num_tokens": 13960056.0,
      "step": 6250
    },
    {
      "entropy": 5.673279094696045,
      "epoch": 0.7025327118548885,
      "grad_norm": 0.875,
      "learning_rate": 0.0004960563511804421,
      "loss": 5.4909,
      "mean_token_accuracy": 0.18654337674379348,
      "num_tokens": 13971932.0,
      "step": 6255
    },
    {
      "entropy": 5.684124517440796,
      "epoch": 0.7030942887628461,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004960488636719545,
      "loss": 5.5129,
      "mean_token_accuracy": 0.1847507417201996,
      "num_tokens": 13981992.0,
      "step": 6260
    },
    {
      "entropy": 5.672705411911011,
      "epoch": 0.7036558656708036,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004960413691251852,
      "loss": 5.4479,
      "mean_token_accuracy": 0.193498632311821,
      "num_tokens": 13992215.0,
      "step": 6265
    },
    {
      "entropy": 5.581198787689209,
      "epoch": 0.7042174425787612,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004960338675403731,
      "loss": 5.3434,
      "mean_token_accuracy": 0.20130388885736467,
      "num_tokens": 14002689.0,
      "step": 6270
    },
    {
      "entropy": 5.586406135559082,
      "epoch": 0.7047790194867187,
      "grad_norm": 0.734375,
      "learning_rate": 0.0004960263589177568,
      "loss": 5.4722,
      "mean_token_accuracy": 0.18611990213394164,
      "num_tokens": 14014951.0,
      "step": 6275
    },
    {
      "entropy": 5.607772779464722,
      "epoch": 0.7053405963946763,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004960188432575757,
      "loss": 5.4162,
      "mean_token_accuracy": 0.18997807651758195,
      "num_tokens": 14025502.0,
      "step": 6280
    },
    {
      "entropy": 5.664870929718018,
      "epoch": 0.7059021733026338,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004960113205600688,
      "loss": 5.4573,
      "mean_token_accuracy": 0.19363315105438234,
      "num_tokens": 14036576.0,
      "step": 6285
    },
    {
      "entropy": 5.643691778182983,
      "epoch": 0.7064637502105914,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004960037908254757,
      "loss": 5.4989,
      "mean_token_accuracy": 0.1870731920003891,
      "num_tokens": 14046493.0,
      "step": 6290
    },
    {
      "entropy": 5.640169382095337,
      "epoch": 0.7070253271185489,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004959962540540361,
      "loss": 5.5311,
      "mean_token_accuracy": 0.18734958469867707,
      "num_tokens": 14057297.0,
      "step": 6295
    },
    {
      "entropy": 5.553561687469482,
      "epoch": 0.7075869040265065,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004959887102459903,
      "loss": 5.3188,
      "mean_token_accuracy": 0.1909932941198349,
      "num_tokens": 14068947.0,
      "step": 6300
    },
    {
      "entropy": 5.6068792819976805,
      "epoch": 0.7081484809344639,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004959811594015781,
      "loss": 5.4183,
      "mean_token_accuracy": 0.19013651013374328,
      "num_tokens": 14078979.0,
      "step": 6305
    },
    {
      "entropy": 5.550307703018189,
      "epoch": 0.7087100578424215,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004959736015210401,
      "loss": 5.3562,
      "mean_token_accuracy": 0.19337991178035735,
      "num_tokens": 14089785.0,
      "step": 6310
    },
    {
      "entropy": 5.576477193832398,
      "epoch": 0.709271634750379,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004959660366046169,
      "loss": 5.4504,
      "mean_token_accuracy": 0.18256163150072097,
      "num_tokens": 14099921.0,
      "step": 6315
    },
    {
      "entropy": 5.54937424659729,
      "epoch": 0.7098332116583366,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004959584646525494,
      "loss": 5.3454,
      "mean_token_accuracy": 0.1859096884727478,
      "num_tokens": 14110965.0,
      "step": 6320
    },
    {
      "entropy": 5.649898386001587,
      "epoch": 0.7103947885662941,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004959508856650788,
      "loss": 5.4806,
      "mean_token_accuracy": 0.18888039886951447,
      "num_tokens": 14122409.0,
      "step": 6325
    },
    {
      "entropy": 5.620231246948242,
      "epoch": 0.7109563654742517,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004959432996424463,
      "loss": 5.4286,
      "mean_token_accuracy": 0.1915147602558136,
      "num_tokens": 14134017.0,
      "step": 6330
    },
    {
      "entropy": 5.621440410614014,
      "epoch": 0.7115179423822092,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004959357065848932,
      "loss": 5.5444,
      "mean_token_accuracy": 0.18655117154121398,
      "num_tokens": 14144843.0,
      "step": 6335
    },
    {
      "entropy": 5.640246200561523,
      "epoch": 0.7120795192901668,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004959281064926618,
      "loss": 5.409,
      "mean_token_accuracy": 0.18741687536239623,
      "num_tokens": 14155944.0,
      "step": 6340
    },
    {
      "entropy": 5.616680335998535,
      "epoch": 0.7126410961981243,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004959204993659936,
      "loss": 5.4793,
      "mean_token_accuracy": 0.18680408149957656,
      "num_tokens": 14167064.0,
      "step": 6345
    },
    {
      "entropy": 5.662240076065063,
      "epoch": 0.7132026731060819,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004959128852051311,
      "loss": 5.4567,
      "mean_token_accuracy": 0.1829783409833908,
      "num_tokens": 14178583.0,
      "step": 6350
    },
    {
      "entropy": 5.565739488601684,
      "epoch": 0.7137642500140394,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004959052640103167,
      "loss": 5.3587,
      "mean_token_accuracy": 0.18854238390922545,
      "num_tokens": 14188752.0,
      "step": 6355
    },
    {
      "entropy": 5.558682250976562,
      "epoch": 0.714325826921997,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000495897635781793,
      "loss": 5.3557,
      "mean_token_accuracy": 0.18964823782444,
      "num_tokens": 14199962.0,
      "step": 6360
    },
    {
      "entropy": 5.579587507247925,
      "epoch": 0.7148874038299545,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004958900005198029,
      "loss": 5.3877,
      "mean_token_accuracy": 0.18765462636947633,
      "num_tokens": 14209877.0,
      "step": 6365
    },
    {
      "entropy": 5.544109201431274,
      "epoch": 0.7154489807379121,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004958823582245894,
      "loss": 5.43,
      "mean_token_accuracy": 0.1922684833407402,
      "num_tokens": 14220672.0,
      "step": 6370
    },
    {
      "entropy": 5.724985074996948,
      "epoch": 0.7160105576458696,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004958747088963961,
      "loss": 5.4258,
      "mean_token_accuracy": 0.19170248359441758,
      "num_tokens": 14232348.0,
      "step": 6375
    },
    {
      "entropy": 5.623469591140747,
      "epoch": 0.7165721345538272,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004958670525354663,
      "loss": 5.4318,
      "mean_token_accuracy": 0.1807127982378006,
      "num_tokens": 14243985.0,
      "step": 6380
    },
    {
      "entropy": 5.50121693611145,
      "epoch": 0.7171337114617847,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000495859389142044,
      "loss": 5.3218,
      "mean_token_accuracy": 0.19600646644830705,
      "num_tokens": 14253548.0,
      "step": 6385
    },
    {
      "entropy": 5.665089845657349,
      "epoch": 0.7176952883697423,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000495851718716373,
      "loss": 5.4681,
      "mean_token_accuracy": 0.1878298193216324,
      "num_tokens": 14264757.0,
      "step": 6390
    },
    {
      "entropy": 5.531696939468384,
      "epoch": 0.7182568652776998,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004958440412586977,
      "loss": 5.3724,
      "mean_token_accuracy": 0.1982964903116226,
      "num_tokens": 14275946.0,
      "step": 6395
    },
    {
      "entropy": 5.542787885665893,
      "epoch": 0.7188184421856574,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004958363567692624,
      "loss": 5.4548,
      "mean_token_accuracy": 0.1816986545920372,
      "num_tokens": 14287006.0,
      "step": 6400
    },
    {
      "entropy": 5.664372253417969,
      "epoch": 0.7193800190936148,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004958286652483118,
      "loss": 5.4712,
      "mean_token_accuracy": 0.1816631406545639,
      "num_tokens": 14299328.0,
      "step": 6405
    },
    {
      "entropy": 5.672094917297363,
      "epoch": 0.7199415960015724,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004958209666960909,
      "loss": 5.4655,
      "mean_token_accuracy": 0.1899723470211029,
      "num_tokens": 14310048.0,
      "step": 6410
    },
    {
      "entropy": 5.558601570129395,
      "epoch": 0.7205031729095299,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004958132611128448,
      "loss": 5.3942,
      "mean_token_accuracy": 0.19261456727981568,
      "num_tokens": 14321107.0,
      "step": 6415
    },
    {
      "entropy": 5.551626920700073,
      "epoch": 0.7210647498174875,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004958055484988188,
      "loss": 5.3141,
      "mean_token_accuracy": 0.19180237203836442,
      "num_tokens": 14332561.0,
      "step": 6420
    },
    {
      "entropy": 5.685160827636719,
      "epoch": 0.721626326725445,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004957978288542585,
      "loss": 5.5132,
      "mean_token_accuracy": 0.1826438695192337,
      "num_tokens": 14343238.0,
      "step": 6425
    },
    {
      "entropy": 5.612111663818359,
      "epoch": 0.7221879036334026,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004957901021794098,
      "loss": 5.4023,
      "mean_token_accuracy": 0.1889569118618965,
      "num_tokens": 14356067.0,
      "step": 6430
    },
    {
      "entropy": 5.581905031204224,
      "epoch": 0.7227494805413601,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004957823684745184,
      "loss": 5.4262,
      "mean_token_accuracy": 0.1919057235121727,
      "num_tokens": 14367302.0,
      "step": 6435
    },
    {
      "entropy": 5.549056053161621,
      "epoch": 0.7233110574493177,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004957746277398309,
      "loss": 5.3541,
      "mean_token_accuracy": 0.1938706859946251,
      "num_tokens": 14377535.0,
      "step": 6440
    },
    {
      "entropy": 5.623479318618775,
      "epoch": 0.7238726343572752,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004957668799755937,
      "loss": 5.4956,
      "mean_token_accuracy": 0.18477451503276826,
      "num_tokens": 14389493.0,
      "step": 6445
    },
    {
      "entropy": 5.566183853149414,
      "epoch": 0.7244342112652328,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004957591251820532,
      "loss": 5.4018,
      "mean_token_accuracy": 0.19189279079437255,
      "num_tokens": 14401564.0,
      "step": 6450
    },
    {
      "entropy": 5.645642328262329,
      "epoch": 0.7249957881731903,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004957513633594567,
      "loss": 5.4639,
      "mean_token_accuracy": 0.1894807681441307,
      "num_tokens": 14412123.0,
      "step": 6455
    },
    {
      "entropy": 5.619349193572998,
      "epoch": 0.7255573650811479,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004957435945080512,
      "loss": 5.4633,
      "mean_token_accuracy": 0.1876132532954216,
      "num_tokens": 14423614.0,
      "step": 6460
    },
    {
      "entropy": 5.62836480140686,
      "epoch": 0.7261189419891054,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004957358186280839,
      "loss": 5.4343,
      "mean_token_accuracy": 0.1854957312345505,
      "num_tokens": 14434830.0,
      "step": 6465
    },
    {
      "entropy": 5.616745090484619,
      "epoch": 0.726680518897063,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004957280357198026,
      "loss": 5.4757,
      "mean_token_accuracy": 0.1899498999118805,
      "num_tokens": 14445015.0,
      "step": 6470
    },
    {
      "entropy": 5.629921817779541,
      "epoch": 0.7272420958050205,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004957202457834551,
      "loss": 5.5575,
      "mean_token_accuracy": 0.17569052577018737,
      "num_tokens": 14455809.0,
      "step": 6475
    },
    {
      "entropy": 5.654147005081176,
      "epoch": 0.7278036727129781,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004957124488192893,
      "loss": 5.3375,
      "mean_token_accuracy": 0.19723852276802062,
      "num_tokens": 14466565.0,
      "step": 6480
    },
    {
      "entropy": 5.6605278015136715,
      "epoch": 0.7283652496209356,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004957046448275535,
      "loss": 5.434,
      "mean_token_accuracy": 0.19561199098825455,
      "num_tokens": 14477383.0,
      "step": 6485
    },
    {
      "entropy": 5.559695434570313,
      "epoch": 0.7289268265288932,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004956968338084961,
      "loss": 5.4199,
      "mean_token_accuracy": 0.18627320379018783,
      "num_tokens": 14489968.0,
      "step": 6490
    },
    {
      "entropy": 5.625599384307861,
      "epoch": 0.7294884034368507,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004956890157623661,
      "loss": 5.4631,
      "mean_token_accuracy": 0.18891706317663193,
      "num_tokens": 14500808.0,
      "step": 6495
    },
    {
      "entropy": 5.5972761631011965,
      "epoch": 0.7300499803448082,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004956811906894122,
      "loss": 5.4764,
      "mean_token_accuracy": 0.18980774581432341,
      "num_tokens": 14511444.0,
      "step": 6500
    },
    {
      "entropy": 5.624223899841309,
      "epoch": 0.7306115572527657,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004956733585898836,
      "loss": 5.3913,
      "mean_token_accuracy": 0.1891003057360649,
      "num_tokens": 14522562.0,
      "step": 6505
    },
    {
      "entropy": 5.575834798812866,
      "epoch": 0.7311731341607233,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004956655194640297,
      "loss": 5.4241,
      "mean_token_accuracy": 0.18901626169681549,
      "num_tokens": 14533497.0,
      "step": 6510
    },
    {
      "entropy": 5.614347839355469,
      "epoch": 0.7317347110686808,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004956576733121,
      "loss": 5.4463,
      "mean_token_accuracy": 0.1829809218645096,
      "num_tokens": 14545637.0,
      "step": 6515
    },
    {
      "entropy": 5.7005633354187015,
      "epoch": 0.7322962879766384,
      "grad_norm": 0.875,
      "learning_rate": 0.0004956498201343444,
      "loss": 5.4554,
      "mean_token_accuracy": 0.18328080624341964,
      "num_tokens": 14556826.0,
      "step": 6520
    },
    {
      "entropy": 5.548034763336181,
      "epoch": 0.7328578648845959,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000495641959931013,
      "loss": 5.3919,
      "mean_token_accuracy": 0.18504128307104112,
      "num_tokens": 14567315.0,
      "step": 6525
    },
    {
      "entropy": 5.600523328781128,
      "epoch": 0.7334194417925535,
      "grad_norm": 0.875,
      "learning_rate": 0.0004956340927023561,
      "loss": 5.4488,
      "mean_token_accuracy": 0.18702426701784133,
      "num_tokens": 14577685.0,
      "step": 6530
    },
    {
      "entropy": 5.6270569324493405,
      "epoch": 0.733981018700511,
      "grad_norm": 0.84375,
      "learning_rate": 0.000495626218448624,
      "loss": 5.4079,
      "mean_token_accuracy": 0.1909566566348076,
      "num_tokens": 14588885.0,
      "step": 6535
    },
    {
      "entropy": 5.63847131729126,
      "epoch": 0.7345425956084686,
      "grad_norm": 0.875,
      "learning_rate": 0.0004956183371700676,
      "loss": 5.437,
      "mean_token_accuracy": 0.19199780076742173,
      "num_tokens": 14599774.0,
      "step": 6540
    },
    {
      "entropy": 5.692119169235229,
      "epoch": 0.7351041725164261,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004956104488669377,
      "loss": 5.4657,
      "mean_token_accuracy": 0.1846656620502472,
      "num_tokens": 14612101.0,
      "step": 6545
    },
    {
      "entropy": 5.732361507415772,
      "epoch": 0.7356657494243837,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004956025535394856,
      "loss": 5.539,
      "mean_token_accuracy": 0.17391334027051925,
      "num_tokens": 14622151.0,
      "step": 6550
    },
    {
      "entropy": 5.641064310073853,
      "epoch": 0.7362273263323412,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004955946511879626,
      "loss": 5.4371,
      "mean_token_accuracy": 0.1776665896177292,
      "num_tokens": 14634631.0,
      "step": 6555
    },
    {
      "entropy": 5.589716958999634,
      "epoch": 0.7367889032402988,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004955867418126205,
      "loss": 5.3936,
      "mean_token_accuracy": 0.18580004423856736,
      "num_tokens": 14645429.0,
      "step": 6560
    },
    {
      "entropy": 5.569635725021362,
      "epoch": 0.7373504801482563,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004955788254137108,
      "loss": 5.4005,
      "mean_token_accuracy": 0.1898912414908409,
      "num_tokens": 14657309.0,
      "step": 6565
    },
    {
      "entropy": 5.547045755386352,
      "epoch": 0.7379120570562139,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004955709019914857,
      "loss": 5.3237,
      "mean_token_accuracy": 0.18987113237380981,
      "num_tokens": 14668578.0,
      "step": 6570
    },
    {
      "entropy": 5.637281179428101,
      "epoch": 0.7384736339641714,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004955629715461977,
      "loss": 5.5469,
      "mean_token_accuracy": 0.18562143445014953,
      "num_tokens": 14679120.0,
      "step": 6575
    },
    {
      "entropy": 5.692667865753174,
      "epoch": 0.739035210872129,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004955550340780991,
      "loss": 5.4928,
      "mean_token_accuracy": 0.18449171483516694,
      "num_tokens": 14690849.0,
      "step": 6580
    },
    {
      "entropy": 5.660857343673706,
      "epoch": 0.7395967877800865,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004955470895874427,
      "loss": 5.4461,
      "mean_token_accuracy": 0.18612099289894105,
      "num_tokens": 14702431.0,
      "step": 6585
    },
    {
      "entropy": 5.527232503890991,
      "epoch": 0.7401583646880441,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004955391380744813,
      "loss": 5.3571,
      "mean_token_accuracy": 0.19393749833106994,
      "num_tokens": 14714271.0,
      "step": 6590
    },
    {
      "entropy": 5.547136354446411,
      "epoch": 0.7407199415960015,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004955311795394684,
      "loss": 5.4949,
      "mean_token_accuracy": 0.18664878308773042,
      "num_tokens": 14724796.0,
      "step": 6595
    },
    {
      "entropy": 5.622562074661255,
      "epoch": 0.7412815185039591,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000495523213982657,
      "loss": 5.4174,
      "mean_token_accuracy": 0.1868535026907921,
      "num_tokens": 14735331.0,
      "step": 6600
    },
    {
      "entropy": 5.706984138488769,
      "epoch": 0.7418430954119166,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004955152414043011,
      "loss": 5.4726,
      "mean_token_accuracy": 0.1846792370080948,
      "num_tokens": 14748595.0,
      "step": 6605
    },
    {
      "entropy": 5.671131229400634,
      "epoch": 0.7424046723198742,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004955072618046543,
      "loss": 5.4031,
      "mean_token_accuracy": 0.18740653097629548,
      "num_tokens": 14759501.0,
      "step": 6610
    },
    {
      "entropy": 5.522915506362915,
      "epoch": 0.7429662492278317,
      "grad_norm": 0.875,
      "learning_rate": 0.0004954992751839708,
      "loss": 5.451,
      "mean_token_accuracy": 0.1917654827237129,
      "num_tokens": 14769752.0,
      "step": 6615
    },
    {
      "entropy": 5.600664186477661,
      "epoch": 0.7435278261357893,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004954912815425048,
      "loss": 5.4,
      "mean_token_accuracy": 0.18774126023054122,
      "num_tokens": 14781166.0,
      "step": 6620
    },
    {
      "entropy": 5.571253871917724,
      "epoch": 0.7440894030437468,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000495483280880511,
      "loss": 5.2935,
      "mean_token_accuracy": 0.1929462119936943,
      "num_tokens": 14790536.0,
      "step": 6625
    },
    {
      "entropy": 5.647788190841675,
      "epoch": 0.7446509799517044,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000495475273198244,
      "loss": 5.511,
      "mean_token_accuracy": 0.18383040130138398,
      "num_tokens": 14801673.0,
      "step": 6630
    },
    {
      "entropy": 5.598875665664673,
      "epoch": 0.7452125568596619,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004954672584959588,
      "loss": 5.4075,
      "mean_token_accuracy": 0.185679093003273,
      "num_tokens": 14812345.0,
      "step": 6635
    },
    {
      "entropy": 5.693237352371216,
      "epoch": 0.7457741337676195,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004954592367739104,
      "loss": 5.5748,
      "mean_token_accuracy": 0.18231677860021592,
      "num_tokens": 14823620.0,
      "step": 6640
    },
    {
      "entropy": 5.614204454421997,
      "epoch": 0.746335710675577,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004954512080323546,
      "loss": 5.3788,
      "mean_token_accuracy": 0.1952967420220375,
      "num_tokens": 14833487.0,
      "step": 6645
    },
    {
      "entropy": 5.4528480052948,
      "epoch": 0.7468972875835346,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004954431722715468,
      "loss": 5.2549,
      "mean_token_accuracy": 0.19585105329751967,
      "num_tokens": 14844421.0,
      "step": 6650
    },
    {
      "entropy": 5.539920473098755,
      "epoch": 0.7474588644914921,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004954351294917428,
      "loss": 5.422,
      "mean_token_accuracy": 0.1950678840279579,
      "num_tokens": 14855525.0,
      "step": 6655
    },
    {
      "entropy": 5.734260225296021,
      "epoch": 0.7480204413994497,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004954270796931989,
      "loss": 5.4802,
      "mean_token_accuracy": 0.1867149919271469,
      "num_tokens": 14866717.0,
      "step": 6660
    },
    {
      "entropy": 5.566936492919922,
      "epoch": 0.7485820183074072,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004954190228761712,
      "loss": 5.3928,
      "mean_token_accuracy": 0.1848048210144043,
      "num_tokens": 14877372.0,
      "step": 6665
    },
    {
      "entropy": 5.556979513168335,
      "epoch": 0.7491435952153648,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004954109590409164,
      "loss": 5.4363,
      "mean_token_accuracy": 0.18893857151269913,
      "num_tokens": 14888618.0,
      "step": 6670
    },
    {
      "entropy": 5.573990201950073,
      "epoch": 0.7497051721233223,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004954028881876911,
      "loss": 5.4466,
      "mean_token_accuracy": 0.18674829751253127,
      "num_tokens": 14900421.0,
      "step": 6675
    },
    {
      "entropy": 5.595251989364624,
      "epoch": 0.7502667490312799,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004953948103167524,
      "loss": 5.4254,
      "mean_token_accuracy": 0.18429961055517197,
      "num_tokens": 14910634.0,
      "step": 6680
    },
    {
      "entropy": 5.682254457473755,
      "epoch": 0.7508283259392374,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004953867254283575,
      "loss": 5.4796,
      "mean_token_accuracy": 0.1812393471598625,
      "num_tokens": 14921408.0,
      "step": 6685
    },
    {
      "entropy": 5.602823925018311,
      "epoch": 0.7513899028471949,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004953786335227636,
      "loss": 5.4347,
      "mean_token_accuracy": 0.18957145810127257,
      "num_tokens": 14933435.0,
      "step": 6690
    },
    {
      "entropy": 5.56487889289856,
      "epoch": 0.7519514797551524,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004953705346002287,
      "loss": 5.4672,
      "mean_token_accuracy": 0.1836774691939354,
      "num_tokens": 14946488.0,
      "step": 6695
    },
    {
      "entropy": 5.620325899124145,
      "epoch": 0.75251305666311,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004953624286610106,
      "loss": 5.3474,
      "mean_token_accuracy": 0.18888706415891648,
      "num_tokens": 14958053.0,
      "step": 6700
    },
    {
      "entropy": 5.602153968811035,
      "epoch": 0.7530746335710675,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004953543157053672,
      "loss": 5.4076,
      "mean_token_accuracy": 0.185575632750988,
      "num_tokens": 14968922.0,
      "step": 6705
    },
    {
      "entropy": 5.606420373916626,
      "epoch": 0.7536362104790251,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004953461957335569,
      "loss": 5.4234,
      "mean_token_accuracy": 0.19019713550806044,
      "num_tokens": 14979541.0,
      "step": 6710
    },
    {
      "entropy": 5.6142607688903805,
      "epoch": 0.7541977873869826,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004953380687458382,
      "loss": 5.4493,
      "mean_token_accuracy": 0.18471300601959229,
      "num_tokens": 14990009.0,
      "step": 6715
    },
    {
      "entropy": 5.579521751403808,
      "epoch": 0.7547593642949402,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004953299347424701,
      "loss": 5.4377,
      "mean_token_accuracy": 0.18548462837934493,
      "num_tokens": 15001712.0,
      "step": 6720
    },
    {
      "entropy": 5.651700496673584,
      "epoch": 0.7553209412028977,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004953217937237114,
      "loss": 5.3858,
      "mean_token_accuracy": 0.19591145664453508,
      "num_tokens": 15012574.0,
      "step": 6725
    },
    {
      "entropy": 5.516970777511597,
      "epoch": 0.7558825181108553,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004953136456898212,
      "loss": 5.4035,
      "mean_token_accuracy": 0.1902242347598076,
      "num_tokens": 15023034.0,
      "step": 6730
    },
    {
      "entropy": 5.6495749950408936,
      "epoch": 0.7564440950188128,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004953054906410593,
      "loss": 5.4422,
      "mean_token_accuracy": 0.18694359213113784,
      "num_tokens": 15033925.0,
      "step": 6735
    },
    {
      "entropy": 5.635606193542481,
      "epoch": 0.7570056719267704,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004952973285776848,
      "loss": 5.4311,
      "mean_token_accuracy": 0.1911318153142929,
      "num_tokens": 15044947.0,
      "step": 6740
    },
    {
      "entropy": 5.617737531661987,
      "epoch": 0.7575672488347279,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004952891594999582,
      "loss": 5.3497,
      "mean_token_accuracy": 0.19404886066913604,
      "num_tokens": 15054784.0,
      "step": 6745
    },
    {
      "entropy": 5.601634979248047,
      "epoch": 0.7581288257426855,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004952809834081392,
      "loss": 5.4227,
      "mean_token_accuracy": 0.1814880058169365,
      "num_tokens": 15066272.0,
      "step": 6750
    },
    {
      "entropy": 5.633253955841065,
      "epoch": 0.758690402650643,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004952728003024883,
      "loss": 5.4895,
      "mean_token_accuracy": 0.18502201586961747,
      "num_tokens": 15079504.0,
      "step": 6755
    },
    {
      "entropy": 5.607704830169678,
      "epoch": 0.7592519795586006,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004952646101832659,
      "loss": 5.4318,
      "mean_token_accuracy": 0.19325195997953415,
      "num_tokens": 15089623.0,
      "step": 6760
    },
    {
      "entropy": 5.553632593154907,
      "epoch": 0.7598135564665581,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004952564130507331,
      "loss": 5.3411,
      "mean_token_accuracy": 0.18783064931631088,
      "num_tokens": 15100816.0,
      "step": 6765
    },
    {
      "entropy": 5.562135124206543,
      "epoch": 0.7603751333745157,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004952482089051506,
      "loss": 5.4344,
      "mean_token_accuracy": 0.19225283116102218,
      "num_tokens": 15111697.0,
      "step": 6770
    },
    {
      "entropy": 5.597704935073852,
      "epoch": 0.7609367102824732,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004952399977467796,
      "loss": 5.5048,
      "mean_token_accuracy": 0.18445664942264556,
      "num_tokens": 15122869.0,
      "step": 6775
    },
    {
      "entropy": 5.527349233627319,
      "epoch": 0.7614982871904308,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004952317795758817,
      "loss": 5.306,
      "mean_token_accuracy": 0.18894683569669724,
      "num_tokens": 15134709.0,
      "step": 6780
    },
    {
      "entropy": 5.534115743637085,
      "epoch": 0.7620598640983883,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004952235543927185,
      "loss": 5.3438,
      "mean_token_accuracy": 0.19631075859069824,
      "num_tokens": 15145319.0,
      "step": 6785
    },
    {
      "entropy": 5.551209545135498,
      "epoch": 0.7626214410063458,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000495215322197552,
      "loss": 5.3635,
      "mean_token_accuracy": 0.18809704035520552,
      "num_tokens": 15156463.0,
      "step": 6790
    },
    {
      "entropy": 5.668431520462036,
      "epoch": 0.7631830179143033,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004952070829906442,
      "loss": 5.5203,
      "mean_token_accuracy": 0.17649743407964708,
      "num_tokens": 15167618.0,
      "step": 6795
    },
    {
      "entropy": 5.6212465286254885,
      "epoch": 0.7637445948222609,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004951988367722574,
      "loss": 5.4444,
      "mean_token_accuracy": 0.18950096219778062,
      "num_tokens": 15179312.0,
      "step": 6800
    },
    {
      "entropy": 5.531707763671875,
      "epoch": 0.7643061717302184,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004951905835426544,
      "loss": 5.3959,
      "mean_token_accuracy": 0.1927101954817772,
      "num_tokens": 15191370.0,
      "step": 6805
    },
    {
      "entropy": 5.624406147003174,
      "epoch": 0.764867748638176,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004951823233020976,
      "loss": 5.366,
      "mean_token_accuracy": 0.19002669900655747,
      "num_tokens": 15201060.0,
      "step": 6810
    },
    {
      "entropy": 5.580497217178345,
      "epoch": 0.7654293255461335,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004951740560508504,
      "loss": 5.3019,
      "mean_token_accuracy": 0.1920270413160324,
      "num_tokens": 15212186.0,
      "step": 6815
    },
    {
      "entropy": 5.518378591537475,
      "epoch": 0.7659909024540911,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004951657817891758,
      "loss": 5.32,
      "mean_token_accuracy": 0.1960553213953972,
      "num_tokens": 15223301.0,
      "step": 6820
    },
    {
      "entropy": 5.560190439224243,
      "epoch": 0.7665524793620486,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004951575005173373,
      "loss": 5.4608,
      "mean_token_accuracy": 0.18384651392698287,
      "num_tokens": 15234475.0,
      "step": 6825
    },
    {
      "entropy": 5.612673902511597,
      "epoch": 0.7671140562700062,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004951492122355986,
      "loss": 5.4411,
      "mean_token_accuracy": 0.18663634955883027,
      "num_tokens": 15245877.0,
      "step": 6830
    },
    {
      "entropy": 5.629204654693604,
      "epoch": 0.7676756331779637,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004951409169442236,
      "loss": 5.4162,
      "mean_token_accuracy": 0.18725284188985825,
      "num_tokens": 15257375.0,
      "step": 6835
    },
    {
      "entropy": 5.639363956451416,
      "epoch": 0.7682372100859213,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004951326146434765,
      "loss": 5.439,
      "mean_token_accuracy": 0.19170643240213395,
      "num_tokens": 15268079.0,
      "step": 6840
    },
    {
      "entropy": 5.575444936752319,
      "epoch": 0.7687987869938788,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004951243053336214,
      "loss": 5.4263,
      "mean_token_accuracy": 0.1878459632396698,
      "num_tokens": 15278678.0,
      "step": 6845
    },
    {
      "entropy": 5.652875518798828,
      "epoch": 0.7693603639018364,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004951159890149232,
      "loss": 5.5369,
      "mean_token_accuracy": 0.18432751893997193,
      "num_tokens": 15290080.0,
      "step": 6850
    },
    {
      "entropy": 5.6735540390014645,
      "epoch": 0.7699219408097939,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004951076656876466,
      "loss": 5.4536,
      "mean_token_accuracy": 0.1816029131412506,
      "num_tokens": 15300602.0,
      "step": 6855
    },
    {
      "entropy": 5.62719874382019,
      "epoch": 0.7704835177177515,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004950993353520564,
      "loss": 5.4511,
      "mean_token_accuracy": 0.19249038845300676,
      "num_tokens": 15312823.0,
      "step": 6860
    },
    {
      "entropy": 5.631157922744751,
      "epoch": 0.771045094625709,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000495090998008418,
      "loss": 5.4635,
      "mean_token_accuracy": 0.1833692356944084,
      "num_tokens": 15323870.0,
      "step": 6865
    },
    {
      "entropy": 5.543543195724487,
      "epoch": 0.7716066715336666,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000495082653656997,
      "loss": 5.3636,
      "mean_token_accuracy": 0.19439253807067872,
      "num_tokens": 15334887.0,
      "step": 6870
    },
    {
      "entropy": 5.629412364959717,
      "epoch": 0.7721682484416241,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004950743022980587,
      "loss": 5.4599,
      "mean_token_accuracy": 0.1855699747800827,
      "num_tokens": 15345871.0,
      "step": 6875
    },
    {
      "entropy": 5.573194169998169,
      "epoch": 0.7727298253495817,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004950659439318696,
      "loss": 5.3172,
      "mean_token_accuracy": 0.1956945091485977,
      "num_tokens": 15356530.0,
      "step": 6880
    },
    {
      "entropy": 5.601297807693482,
      "epoch": 0.7732914022575391,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004950575785586952,
      "loss": 5.3829,
      "mean_token_accuracy": 0.19240699112415313,
      "num_tokens": 15368756.0,
      "step": 6885
    },
    {
      "entropy": 5.649917840957642,
      "epoch": 0.7738529791654967,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004950492061788023,
      "loss": 5.4682,
      "mean_token_accuracy": 0.19151344299316406,
      "num_tokens": 15379450.0,
      "step": 6890
    },
    {
      "entropy": 5.519516181945801,
      "epoch": 0.7744145560734542,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004950408267924573,
      "loss": 5.377,
      "mean_token_accuracy": 0.18265025466680526,
      "num_tokens": 15389501.0,
      "step": 6895
    },
    {
      "entropy": 5.44174280166626,
      "epoch": 0.7749761329814118,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004950324403999271,
      "loss": 5.377,
      "mean_token_accuracy": 0.1893697515130043,
      "num_tokens": 15400608.0,
      "step": 6900
    },
    {
      "entropy": 5.68777813911438,
      "epoch": 0.7755377098893693,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004950240470014785,
      "loss": 5.4729,
      "mean_token_accuracy": 0.18293607831001282,
      "num_tokens": 15411709.0,
      "step": 6905
    },
    {
      "entropy": 5.562110757827758,
      "epoch": 0.7760992867973269,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004950156465973791,
      "loss": 5.3218,
      "mean_token_accuracy": 0.19510839581489564,
      "num_tokens": 15422292.0,
      "step": 6910
    },
    {
      "entropy": 5.587017011642456,
      "epoch": 0.7766608637052844,
      "grad_norm": 0.890625,
      "learning_rate": 0.000495007239187896,
      "loss": 5.4768,
      "mean_token_accuracy": 0.1842653051018715,
      "num_tokens": 15434439.0,
      "step": 6915
    },
    {
      "entropy": 5.574909353256226,
      "epoch": 0.777222440613242,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004949988247732972,
      "loss": 5.4154,
      "mean_token_accuracy": 0.19177116006612777,
      "num_tokens": 15445480.0,
      "step": 6920
    },
    {
      "entropy": 5.533121061325073,
      "epoch": 0.7777840175211995,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004949904033538504,
      "loss": 5.4006,
      "mean_token_accuracy": 0.18785960972309113,
      "num_tokens": 15456960.0,
      "step": 6925
    },
    {
      "entropy": 5.543678426742554,
      "epoch": 0.7783455944291571,
      "grad_norm": 0.859375,
      "learning_rate": 0.000494981974929824,
      "loss": 5.404,
      "mean_token_accuracy": 0.18626379370689392,
      "num_tokens": 15468169.0,
      "step": 6930
    },
    {
      "entropy": 5.614240550994873,
      "epoch": 0.7789071713371146,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000494973539501486,
      "loss": 5.373,
      "mean_token_accuracy": 0.1923552170395851,
      "num_tokens": 15478740.0,
      "step": 6935
    },
    {
      "entropy": 5.6082438945770265,
      "epoch": 0.7794687482450722,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004949650970691052,
      "loss": 5.4753,
      "mean_token_accuracy": 0.18252963423728943,
      "num_tokens": 15489705.0,
      "step": 6940
    },
    {
      "entropy": 5.517890501022339,
      "epoch": 0.7800303251530297,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004949566476329505,
      "loss": 5.318,
      "mean_token_accuracy": 0.1861863002181053,
      "num_tokens": 15500549.0,
      "step": 6945
    },
    {
      "entropy": 5.681436491012573,
      "epoch": 0.7805919020609873,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004949481911932908,
      "loss": 5.4546,
      "mean_token_accuracy": 0.18514592051506043,
      "num_tokens": 15511124.0,
      "step": 6950
    },
    {
      "entropy": 5.569234991073609,
      "epoch": 0.7811534789689448,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004949397277503954,
      "loss": 5.3654,
      "mean_token_accuracy": 0.19057340174913406,
      "num_tokens": 15521810.0,
      "step": 6955
    },
    {
      "entropy": 5.567523193359375,
      "epoch": 0.7817150558769024,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004949312573045338,
      "loss": 5.4108,
      "mean_token_accuracy": 0.18870485574007034,
      "num_tokens": 15533108.0,
      "step": 6960
    },
    {
      "entropy": 5.597011947631836,
      "epoch": 0.7822766327848599,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004949227798559757,
      "loss": 5.4324,
      "mean_token_accuracy": 0.1902460500597954,
      "num_tokens": 15544841.0,
      "step": 6965
    },
    {
      "entropy": 5.6039785861969,
      "epoch": 0.7828382096928175,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004949142954049909,
      "loss": 5.4191,
      "mean_token_accuracy": 0.18016085475683213,
      "num_tokens": 15556297.0,
      "step": 6970
    },
    {
      "entropy": 5.574081945419311,
      "epoch": 0.783399786600775,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004949058039518496,
      "loss": 5.3091,
      "mean_token_accuracy": 0.19485378861427308,
      "num_tokens": 15566607.0,
      "step": 6975
    },
    {
      "entropy": 5.607875156402588,
      "epoch": 0.7839613635087325,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004948973054968222,
      "loss": 5.4862,
      "mean_token_accuracy": 0.186459019780159,
      "num_tokens": 15577435.0,
      "step": 6980
    },
    {
      "entropy": 5.608172130584717,
      "epoch": 0.78452294041669,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004948888000401794,
      "loss": 5.385,
      "mean_token_accuracy": 0.19240848571062089,
      "num_tokens": 15587844.0,
      "step": 6985
    },
    {
      "entropy": 5.639331436157226,
      "epoch": 0.7850845173246476,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004948802875821918,
      "loss": 5.4426,
      "mean_token_accuracy": 0.19013611972332,
      "num_tokens": 15599586.0,
      "step": 6990
    },
    {
      "entropy": 5.55504002571106,
      "epoch": 0.7856460942326051,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004948717681231308,
      "loss": 5.3361,
      "mean_token_accuracy": 0.1892746388912201,
      "num_tokens": 15610838.0,
      "step": 6995
    },
    {
      "entropy": 5.617995309829712,
      "epoch": 0.7862076711405627,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004948632416632672,
      "loss": 5.4141,
      "mean_token_accuracy": 0.19103100746870041,
      "num_tokens": 15622652.0,
      "step": 7000
    },
    {
      "entropy": 5.669953107833862,
      "epoch": 0.7867692480485202,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004948547082028728,
      "loss": 5.5065,
      "mean_token_accuracy": 0.182385590672493,
      "num_tokens": 15634280.0,
      "step": 7005
    },
    {
      "entropy": 5.536403465270996,
      "epoch": 0.7873308249564778,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004948461677422193,
      "loss": 5.2456,
      "mean_token_accuracy": 0.20017145276069642,
      "num_tokens": 15644088.0,
      "step": 7010
    },
    {
      "entropy": 5.559831428527832,
      "epoch": 0.7878924018644353,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004948376202815785,
      "loss": 5.489,
      "mean_token_accuracy": 0.18129177689552306,
      "num_tokens": 15654968.0,
      "step": 7015
    },
    {
      "entropy": 5.581280612945557,
      "epoch": 0.7884539787723929,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004948290658212225,
      "loss": 5.392,
      "mean_token_accuracy": 0.19254007637500764,
      "num_tokens": 15666758.0,
      "step": 7020
    },
    {
      "entropy": 5.4905070781707765,
      "epoch": 0.7890155556803504,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004948205043614239,
      "loss": 5.2514,
      "mean_token_accuracy": 0.1979697033762932,
      "num_tokens": 15677295.0,
      "step": 7025
    },
    {
      "entropy": 5.5794915676116945,
      "epoch": 0.789577132588308,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004948119359024551,
      "loss": 5.4545,
      "mean_token_accuracy": 0.19262483417987825,
      "num_tokens": 15689508.0,
      "step": 7030
    },
    {
      "entropy": 5.606967878341675,
      "epoch": 0.7901387094962655,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004948033604445891,
      "loss": 5.4019,
      "mean_token_accuracy": 0.19027716517448426,
      "num_tokens": 15701289.0,
      "step": 7035
    },
    {
      "entropy": 5.524088621139526,
      "epoch": 0.7907002864042231,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004947947779880988,
      "loss": 5.3618,
      "mean_token_accuracy": 0.18586177378892899,
      "num_tokens": 15711579.0,
      "step": 7040
    },
    {
      "entropy": 5.608291244506836,
      "epoch": 0.7912618633121806,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004947861885332576,
      "loss": 5.4441,
      "mean_token_accuracy": 0.19303490072488785,
      "num_tokens": 15722629.0,
      "step": 7045
    },
    {
      "entropy": 5.587085437774658,
      "epoch": 0.7918234402201382,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004947775920803388,
      "loss": 5.3711,
      "mean_token_accuracy": 0.18527416586875917,
      "num_tokens": 15735818.0,
      "step": 7050
    },
    {
      "entropy": 5.5413893699646,
      "epoch": 0.7923850171280957,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004947689886296162,
      "loss": 5.4359,
      "mean_token_accuracy": 0.18212716281414032,
      "num_tokens": 15746914.0,
      "step": 7055
    },
    {
      "entropy": 5.564472103118897,
      "epoch": 0.7929465940360533,
      "grad_norm": 1.078125,
      "learning_rate": 0.000494760378181364,
      "loss": 5.3403,
      "mean_token_accuracy": 0.18668182939291,
      "num_tokens": 15757597.0,
      "step": 7060
    },
    {
      "entropy": 5.4883537769317625,
      "epoch": 0.7935081709440108,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004947517607358559,
      "loss": 5.2671,
      "mean_token_accuracy": 0.19818475097417831,
      "num_tokens": 15768860.0,
      "step": 7065
    },
    {
      "entropy": 5.583630514144898,
      "epoch": 0.7940697478519684,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004947431362933666,
      "loss": 5.3977,
      "mean_token_accuracy": 0.18761927038431167,
      "num_tokens": 15779476.0,
      "step": 7070
    },
    {
      "entropy": 5.600676393508911,
      "epoch": 0.7946313247599258,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004947345048541706,
      "loss": 5.4337,
      "mean_token_accuracy": 0.18490423411130905,
      "num_tokens": 15792142.0,
      "step": 7075
    },
    {
      "entropy": 5.586075925827027,
      "epoch": 0.7951929016678834,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004947258664185427,
      "loss": 5.4109,
      "mean_token_accuracy": 0.1896215096116066,
      "num_tokens": 15802307.0,
      "step": 7080
    },
    {
      "entropy": 5.5321457862854,
      "epoch": 0.7957544785758409,
      "grad_norm": 0.875,
      "learning_rate": 0.0004947172209867581,
      "loss": 5.3702,
      "mean_token_accuracy": 0.18866233229637147,
      "num_tokens": 15812986.0,
      "step": 7085
    },
    {
      "entropy": 5.5097345352172855,
      "epoch": 0.7963160554837985,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004947085685590919,
      "loss": 5.3294,
      "mean_token_accuracy": 0.1948045462369919,
      "num_tokens": 15823938.0,
      "step": 7090
    },
    {
      "entropy": 5.507440233230591,
      "epoch": 0.796877632391756,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004946999091358196,
      "loss": 5.3,
      "mean_token_accuracy": 0.1948933109641075,
      "num_tokens": 15835813.0,
      "step": 7095
    },
    {
      "entropy": 5.603211259841919,
      "epoch": 0.7974392092997136,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000494691242717217,
      "loss": 5.3854,
      "mean_token_accuracy": 0.19351608008146287,
      "num_tokens": 15846203.0,
      "step": 7100
    },
    {
      "entropy": 5.5829263687133786,
      "epoch": 0.7980007862076711,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004946825693035601,
      "loss": 5.3383,
      "mean_token_accuracy": 0.19789962321519852,
      "num_tokens": 15856919.0,
      "step": 7105
    },
    {
      "entropy": 5.59893913269043,
      "epoch": 0.7985623631156287,
      "grad_norm": 0.9375,
      "learning_rate": 0.000494673888895125,
      "loss": 5.4612,
      "mean_token_accuracy": 0.18726717680692673,
      "num_tokens": 15867149.0,
      "step": 7110
    },
    {
      "entropy": 5.527942371368408,
      "epoch": 0.7991239400235862,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004946652014921881,
      "loss": 5.3452,
      "mean_token_accuracy": 0.18912626951932907,
      "num_tokens": 15877229.0,
      "step": 7115
    },
    {
      "entropy": 5.565203142166138,
      "epoch": 0.7996855169315438,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004946565070950259,
      "loss": 5.328,
      "mean_token_accuracy": 0.190609410405159,
      "num_tokens": 15888799.0,
      "step": 7120
    },
    {
      "entropy": 5.486018657684326,
      "epoch": 0.8002470938395013,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004946478057039153,
      "loss": 5.3278,
      "mean_token_accuracy": 0.20010972768068314,
      "num_tokens": 15900144.0,
      "step": 7125
    },
    {
      "entropy": 5.5885766506195065,
      "epoch": 0.8008086707474589,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004946390973191335,
      "loss": 5.5039,
      "mean_token_accuracy": 0.18320483565330506,
      "num_tokens": 15912030.0,
      "step": 7130
    },
    {
      "entropy": 5.607905292510987,
      "epoch": 0.8013702476554164,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004946303819409575,
      "loss": 5.3655,
      "mean_token_accuracy": 0.18694779723882676,
      "num_tokens": 15923606.0,
      "step": 7135
    },
    {
      "entropy": 5.576479768753051,
      "epoch": 0.801931824563374,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004946216595696652,
      "loss": 5.3716,
      "mean_token_accuracy": 0.19240496903657914,
      "num_tokens": 15935077.0,
      "step": 7140
    },
    {
      "entropy": 5.479331302642822,
      "epoch": 0.8024934014713315,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004946129302055338,
      "loss": 5.2902,
      "mean_token_accuracy": 0.19725685715675353,
      "num_tokens": 15945331.0,
      "step": 7145
    },
    {
      "entropy": 5.529646873474121,
      "epoch": 0.8030549783792891,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004946041938488417,
      "loss": 5.3955,
      "mean_token_accuracy": 0.18659822642803192,
      "num_tokens": 15957031.0,
      "step": 7150
    },
    {
      "entropy": 5.622544622421264,
      "epoch": 0.8036165552872466,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004945954504998668,
      "loss": 5.428,
      "mean_token_accuracy": 0.19067881852388383,
      "num_tokens": 15967545.0,
      "step": 7155
    },
    {
      "entropy": 5.706559801101685,
      "epoch": 0.8041781321952042,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004945867001588877,
      "loss": 5.4755,
      "mean_token_accuracy": 0.18083505630493163,
      "num_tokens": 15978700.0,
      "step": 7160
    },
    {
      "entropy": 5.548195266723633,
      "epoch": 0.8047397091031617,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004945779428261827,
      "loss": 5.3602,
      "mean_token_accuracy": 0.1963795468211174,
      "num_tokens": 15989712.0,
      "step": 7165
    },
    {
      "entropy": 5.676448059082031,
      "epoch": 0.8053012860111193,
      "grad_norm": 0.77734375,
      "learning_rate": 0.000494569178502031,
      "loss": 5.5668,
      "mean_token_accuracy": 0.18384641110897065,
      "num_tokens": 16002380.0,
      "step": 7170
    },
    {
      "entropy": 5.607323741912841,
      "epoch": 0.8058628629190767,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004945604071867114,
      "loss": 5.3396,
      "mean_token_accuracy": 0.18998024761676788,
      "num_tokens": 16013749.0,
      "step": 7175
    },
    {
      "entropy": 5.6413187980651855,
      "epoch": 0.8064244398270343,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004945516288805035,
      "loss": 5.5276,
      "mean_token_accuracy": 0.18483836948871613,
      "num_tokens": 16025131.0,
      "step": 7180
    },
    {
      "entropy": 5.608022546768188,
      "epoch": 0.8069860167349918,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004945428435836864,
      "loss": 5.3923,
      "mean_token_accuracy": 0.1946142703294754,
      "num_tokens": 16037269.0,
      "step": 7185
    },
    {
      "entropy": 5.703870010375977,
      "epoch": 0.8075475936429494,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004945340512965401,
      "loss": 5.5825,
      "mean_token_accuracy": 0.18042129874229432,
      "num_tokens": 16050992.0,
      "step": 7190
    },
    {
      "entropy": 5.604102420806885,
      "epoch": 0.8081091705509069,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004945252520193443,
      "loss": 5.4016,
      "mean_token_accuracy": 0.18522187769412995,
      "num_tokens": 16062309.0,
      "step": 7195
    },
    {
      "entropy": 5.650782060623169,
      "epoch": 0.8086707474588645,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004945164457523795,
      "loss": 5.4421,
      "mean_token_accuracy": 0.18486936241388321,
      "num_tokens": 16073079.0,
      "step": 7200
    },
    {
      "entropy": 5.558680629730224,
      "epoch": 0.809232324366822,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004945076324959258,
      "loss": 5.3378,
      "mean_token_accuracy": 0.19475983828306198,
      "num_tokens": 16082885.0,
      "step": 7205
    },
    {
      "entropy": 5.4442063331604,
      "epoch": 0.8097939012747796,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004944988122502641,
      "loss": 5.2999,
      "mean_token_accuracy": 0.19786840975284575,
      "num_tokens": 16094173.0,
      "step": 7210
    },
    {
      "entropy": 5.567581558227539,
      "epoch": 0.8103554781827371,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000494489985015675,
      "loss": 5.3518,
      "mean_token_accuracy": 0.19348257333040236,
      "num_tokens": 16106351.0,
      "step": 7215
    },
    {
      "entropy": 5.48266372680664,
      "epoch": 0.8109170550906947,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004944811507924396,
      "loss": 5.3307,
      "mean_token_accuracy": 0.19099057018756865,
      "num_tokens": 16116892.0,
      "step": 7220
    },
    {
      "entropy": 5.617967224121093,
      "epoch": 0.8114786319986522,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004944723095808393,
      "loss": 5.4659,
      "mean_token_accuracy": 0.1819067344069481,
      "num_tokens": 16128316.0,
      "step": 7225
    },
    {
      "entropy": 5.588520336151123,
      "epoch": 0.8120402089066098,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004944634613811555,
      "loss": 5.3091,
      "mean_token_accuracy": 0.194782093167305,
      "num_tokens": 16139245.0,
      "step": 7230
    },
    {
      "entropy": 5.562133026123047,
      "epoch": 0.8126017858145673,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00049445460619367,
      "loss": 5.3598,
      "mean_token_accuracy": 0.19668255150318145,
      "num_tokens": 16150893.0,
      "step": 7235
    },
    {
      "entropy": 5.590014028549194,
      "epoch": 0.8131633627225249,
      "grad_norm": 1.75,
      "learning_rate": 0.0004944457440186648,
      "loss": 5.4728,
      "mean_token_accuracy": 0.172024929523468,
      "num_tokens": 16163430.0,
      "step": 7240
    },
    {
      "entropy": 5.661122465133667,
      "epoch": 0.8137249396304824,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004944368748564219,
      "loss": 5.3863,
      "mean_token_accuracy": 0.1914934352040291,
      "num_tokens": 16173954.0,
      "step": 7245
    },
    {
      "entropy": 5.580222320556641,
      "epoch": 0.81428651653844,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004944279987072238,
      "loss": 5.4347,
      "mean_token_accuracy": 0.1816001445055008,
      "num_tokens": 16185491.0,
      "step": 7250
    },
    {
      "entropy": 5.500250291824341,
      "epoch": 0.8148480934463975,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004944191155713531,
      "loss": 5.2908,
      "mean_token_accuracy": 0.2001260355114937,
      "num_tokens": 16195301.0,
      "step": 7255
    },
    {
      "entropy": 5.587153100967408,
      "epoch": 0.8154096703543551,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004944102254490928,
      "loss": 5.4167,
      "mean_token_accuracy": 0.18557130247354509,
      "num_tokens": 16206620.0,
      "step": 7260
    },
    {
      "entropy": 5.594261312484742,
      "epoch": 0.8159712472623126,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004944013283407257,
      "loss": 5.4898,
      "mean_token_accuracy": 0.18437268137931823,
      "num_tokens": 16218693.0,
      "step": 7265
    },
    {
      "entropy": 5.621704483032227,
      "epoch": 0.8165328241702701,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004943924242465352,
      "loss": 5.3871,
      "mean_token_accuracy": 0.19240337908267974,
      "num_tokens": 16229124.0,
      "step": 7270
    },
    {
      "entropy": 5.519367122650147,
      "epoch": 0.8170944010782276,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004943835131668049,
      "loss": 5.3529,
      "mean_token_accuracy": 0.18895668983459474,
      "num_tokens": 16239999.0,
      "step": 7275
    },
    {
      "entropy": 5.560688734054565,
      "epoch": 0.8176559779861852,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004943745951018184,
      "loss": 5.4404,
      "mean_token_accuracy": 0.19284715354442597,
      "num_tokens": 16251027.0,
      "step": 7280
    },
    {
      "entropy": 5.627780294418335,
      "epoch": 0.8182175548941427,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004943656700518597,
      "loss": 5.3913,
      "mean_token_accuracy": 0.19192791879177093,
      "num_tokens": 16262924.0,
      "step": 7285
    },
    {
      "entropy": 5.650020503997803,
      "epoch": 0.8187791318021003,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004943567380172129,
      "loss": 5.4833,
      "mean_token_accuracy": 0.18830612152814866,
      "num_tokens": 16274481.0,
      "step": 7290
    },
    {
      "entropy": 5.556411933898926,
      "epoch": 0.8193407087100578,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004943477989981627,
      "loss": 5.4246,
      "mean_token_accuracy": 0.19418288320302962,
      "num_tokens": 16285214.0,
      "step": 7295
    },
    {
      "entropy": 5.603423595428467,
      "epoch": 0.8199022856180154,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004943388529949933,
      "loss": 5.4864,
      "mean_token_accuracy": 0.18698683381080627,
      "num_tokens": 16296077.0,
      "step": 7300
    },
    {
      "entropy": 5.655724811553955,
      "epoch": 0.8204638625259729,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004943299000079899,
      "loss": 5.3929,
      "mean_token_accuracy": 0.19048387706279754,
      "num_tokens": 16307514.0,
      "step": 7305
    },
    {
      "entropy": 5.667319250106812,
      "epoch": 0.8210254394339305,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004943209400374373,
      "loss": 5.3998,
      "mean_token_accuracy": 0.1893719330430031,
      "num_tokens": 16318734.0,
      "step": 7310
    },
    {
      "entropy": 5.522929906845093,
      "epoch": 0.821587016341888,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000494311973083621,
      "loss": 5.4352,
      "mean_token_accuracy": 0.18904135078191758,
      "num_tokens": 16329057.0,
      "step": 7315
    },
    {
      "entropy": 5.529419946670532,
      "epoch": 0.8221485932498456,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004943029991468264,
      "loss": 5.3488,
      "mean_token_accuracy": 0.19056213796138763,
      "num_tokens": 16339646.0,
      "step": 7320
    },
    {
      "entropy": 5.61848931312561,
      "epoch": 0.8227101701578031,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004942940182273393,
      "loss": 5.4254,
      "mean_token_accuracy": 0.18892553448677063,
      "num_tokens": 16349783.0,
      "step": 7325
    },
    {
      "entropy": 5.631988906860352,
      "epoch": 0.8232717470657607,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004942850303254455,
      "loss": 5.3568,
      "mean_token_accuracy": 0.19152336418628693,
      "num_tokens": 16361078.0,
      "step": 7330
    },
    {
      "entropy": 5.528772020339966,
      "epoch": 0.8238333239737182,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004942760354414313,
      "loss": 5.3713,
      "mean_token_accuracy": 0.19279086142778395,
      "num_tokens": 16371308.0,
      "step": 7335
    },
    {
      "entropy": 5.575279951095581,
      "epoch": 0.8243949008816758,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004942670335755831,
      "loss": 5.4126,
      "mean_token_accuracy": 0.19368164837360383,
      "num_tokens": 16382682.0,
      "step": 7340
    },
    {
      "entropy": 5.651817083358765,
      "epoch": 0.8249564777896333,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004942580247281876,
      "loss": 5.4323,
      "mean_token_accuracy": 0.1876862168312073,
      "num_tokens": 16393715.0,
      "step": 7345
    },
    {
      "entropy": 5.517585277557373,
      "epoch": 0.8255180546975909,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004942490088995315,
      "loss": 5.4221,
      "mean_token_accuracy": 0.1912041798233986,
      "num_tokens": 16406332.0,
      "step": 7350
    },
    {
      "entropy": 5.53419861793518,
      "epoch": 0.8260796316055484,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004942399860899018,
      "loss": 5.3395,
      "mean_token_accuracy": 0.19024117290973663,
      "num_tokens": 16417702.0,
      "step": 7355
    },
    {
      "entropy": 5.584663152694702,
      "epoch": 0.826641208513506,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004942309562995862,
      "loss": 5.3957,
      "mean_token_accuracy": 0.18683216124773025,
      "num_tokens": 16428497.0,
      "step": 7360
    },
    {
      "entropy": 5.588092088699341,
      "epoch": 0.8272027854214634,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004942219195288716,
      "loss": 5.3819,
      "mean_token_accuracy": 0.1921641230583191,
      "num_tokens": 16440637.0,
      "step": 7365
    },
    {
      "entropy": 5.585617923736573,
      "epoch": 0.827764362329421,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004942128757780463,
      "loss": 5.3984,
      "mean_token_accuracy": 0.18151830583810807,
      "num_tokens": 16453213.0,
      "step": 7370
    },
    {
      "entropy": 5.56573519706726,
      "epoch": 0.8283259392373785,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004942038250473979,
      "loss": 5.4205,
      "mean_token_accuracy": 0.19382362961769103,
      "num_tokens": 16464345.0,
      "step": 7375
    },
    {
      "entropy": 5.5302290439605715,
      "epoch": 0.8288875161453361,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004941947673372147,
      "loss": 5.3031,
      "mean_token_accuracy": 0.19457138180732728,
      "num_tokens": 16475399.0,
      "step": 7380
    },
    {
      "entropy": 5.5569110870361325,
      "epoch": 0.8294490930532936,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004941857026477851,
      "loss": 5.2821,
      "mean_token_accuracy": 0.19275529235601424,
      "num_tokens": 16485439.0,
      "step": 7385
    },
    {
      "entropy": 5.519453668594361,
      "epoch": 0.8300106699612512,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004941766309793978,
      "loss": 5.2893,
      "mean_token_accuracy": 0.19429101645946503,
      "num_tokens": 16496765.0,
      "step": 7390
    },
    {
      "entropy": 5.508600568771362,
      "epoch": 0.8305722468692087,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004941675523323414,
      "loss": 5.3277,
      "mean_token_accuracy": 0.1900573968887329,
      "num_tokens": 16509199.0,
      "step": 7395
    },
    {
      "entropy": 5.569557428359985,
      "epoch": 0.8311338237771663,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004941584667069052,
      "loss": 5.4763,
      "mean_token_accuracy": 0.1870730921626091,
      "num_tokens": 16520073.0,
      "step": 7400
    },
    {
      "entropy": 5.512719821929932,
      "epoch": 0.8316954006851238,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004941493741033784,
      "loss": 5.3476,
      "mean_token_accuracy": 0.1921248510479927,
      "num_tokens": 16530919.0,
      "step": 7405
    },
    {
      "entropy": 5.571261548995972,
      "epoch": 0.8322569775930814,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004941402745220507,
      "loss": 5.3724,
      "mean_token_accuracy": 0.19373850673437118,
      "num_tokens": 16542591.0,
      "step": 7410
    },
    {
      "entropy": 5.621003484725952,
      "epoch": 0.8328185545010389,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004941311679632115,
      "loss": 5.3875,
      "mean_token_accuracy": 0.1910845473408699,
      "num_tokens": 16554984.0,
      "step": 7415
    },
    {
      "entropy": 5.525754785537719,
      "epoch": 0.8333801314089965,
      "grad_norm": 0.828125,
      "learning_rate": 0.000494122054427151,
      "loss": 5.2809,
      "mean_token_accuracy": 0.19670228511095048,
      "num_tokens": 16566598.0,
      "step": 7420
    },
    {
      "entropy": 5.638074016571045,
      "epoch": 0.833941708316954,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004941129339141592,
      "loss": 5.4152,
      "mean_token_accuracy": 0.18603952676057817,
      "num_tokens": 16577868.0,
      "step": 7425
    },
    {
      "entropy": 5.556251049041748,
      "epoch": 0.8345032852249116,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004941038064245267,
      "loss": 5.3817,
      "mean_token_accuracy": 0.18709893375635148,
      "num_tokens": 16587891.0,
      "step": 7430
    },
    {
      "entropy": 5.54441728591919,
      "epoch": 0.8350648621328691,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000494094671958544,
      "loss": 5.3361,
      "mean_token_accuracy": 0.1987432137131691,
      "num_tokens": 16599924.0,
      "step": 7435
    },
    {
      "entropy": 5.500008678436279,
      "epoch": 0.8356264390408267,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004940855305165019,
      "loss": 5.28,
      "mean_token_accuracy": 0.1929803043603897,
      "num_tokens": 16609958.0,
      "step": 7440
    },
    {
      "entropy": 5.5318121910095215,
      "epoch": 0.8361880159487842,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004940763820986916,
      "loss": 5.4068,
      "mean_token_accuracy": 0.19022383987903596,
      "num_tokens": 16621159.0,
      "step": 7445
    },
    {
      "entropy": 5.674335241317749,
      "epoch": 0.8367495928567418,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004940672267054044,
      "loss": 5.4979,
      "mean_token_accuracy": 0.1903561159968376,
      "num_tokens": 16632263.0,
      "step": 7450
    },
    {
      "entropy": 5.553600835800171,
      "epoch": 0.8373111697646993,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004940580643369316,
      "loss": 5.3774,
      "mean_token_accuracy": 0.18678880631923675,
      "num_tokens": 16644117.0,
      "step": 7455
    },
    {
      "entropy": 5.434643268585205,
      "epoch": 0.8378727466726568,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004940488949935651,
      "loss": 5.3186,
      "mean_token_accuracy": 0.1955294579267502,
      "num_tokens": 16654529.0,
      "step": 7460
    },
    {
      "entropy": 5.540984773635865,
      "epoch": 0.8384343235806143,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004940397186755968,
      "loss": 5.3204,
      "mean_token_accuracy": 0.19437023103237153,
      "num_tokens": 16666191.0,
      "step": 7465
    },
    {
      "entropy": 5.614701128005981,
      "epoch": 0.8389959004885719,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004940305353833191,
      "loss": 5.3055,
      "mean_token_accuracy": 0.19504969716072082,
      "num_tokens": 16677252.0,
      "step": 7470
    },
    {
      "entropy": 5.453291702270508,
      "epoch": 0.8395574773965294,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000494021345117024,
      "loss": 5.3598,
      "mean_token_accuracy": 0.18967387676239014,
      "num_tokens": 16689487.0,
      "step": 7475
    },
    {
      "entropy": 5.519382190704346,
      "epoch": 0.840119054304487,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004940121478770044,
      "loss": 5.2558,
      "mean_token_accuracy": 0.19846489876508713,
      "num_tokens": 16699955.0,
      "step": 7480
    },
    {
      "entropy": 5.459865999221802,
      "epoch": 0.8406806312124445,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004940029436635531,
      "loss": 5.3143,
      "mean_token_accuracy": 0.188406366109848,
      "num_tokens": 16711071.0,
      "step": 7485
    },
    {
      "entropy": 5.468415307998657,
      "epoch": 0.8412422081204021,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004939937324769631,
      "loss": 5.314,
      "mean_token_accuracy": 0.19810989499092102,
      "num_tokens": 16722115.0,
      "step": 7490
    },
    {
      "entropy": 5.555672359466553,
      "epoch": 0.8418037850283596,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004939845143175276,
      "loss": 5.3625,
      "mean_token_accuracy": 0.1840437799692154,
      "num_tokens": 16733146.0,
      "step": 7495
    },
    {
      "entropy": 5.517715787887573,
      "epoch": 0.8423653619363172,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004939752891855404,
      "loss": 5.3357,
      "mean_token_accuracy": 0.19065668880939485,
      "num_tokens": 16745543.0,
      "step": 7500
    },
    {
      "entropy": 5.5761387825012205,
      "epoch": 0.8429269388442747,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004939660570812951,
      "loss": 5.3967,
      "mean_token_accuracy": 0.19377561956644057,
      "num_tokens": 16755782.0,
      "step": 7505
    },
    {
      "entropy": 5.602105569839478,
      "epoch": 0.8434885157522323,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004939568180050855,
      "loss": 5.4241,
      "mean_token_accuracy": 0.19132550209760665,
      "num_tokens": 16767509.0,
      "step": 7510
    },
    {
      "entropy": 5.524803829193115,
      "epoch": 0.8440500926601898,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004939475719572059,
      "loss": 5.3013,
      "mean_token_accuracy": 0.19385440349578859,
      "num_tokens": 16779257.0,
      "step": 7515
    },
    {
      "entropy": 5.516928243637085,
      "epoch": 0.8446116695681474,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004939383189379507,
      "loss": 5.3332,
      "mean_token_accuracy": 0.19200672209262848,
      "num_tokens": 16790769.0,
      "step": 7520
    },
    {
      "entropy": 5.579675102233887,
      "epoch": 0.8451732464761049,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004939290589476145,
      "loss": 5.3378,
      "mean_token_accuracy": 0.19205561578273772,
      "num_tokens": 16800054.0,
      "step": 7525
    },
    {
      "entropy": 5.642947816848755,
      "epoch": 0.8457348233840625,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004939197919864922,
      "loss": 5.4314,
      "mean_token_accuracy": 0.1880166009068489,
      "num_tokens": 16811485.0,
      "step": 7530
    },
    {
      "entropy": 5.631079387664795,
      "epoch": 0.84629640029202,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004939105180548788,
      "loss": 5.4866,
      "mean_token_accuracy": 0.18868340104818343,
      "num_tokens": 16822683.0,
      "step": 7535
    },
    {
      "entropy": 5.609653091430664,
      "epoch": 0.8468579771999776,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004939012371530696,
      "loss": 5.3564,
      "mean_token_accuracy": 0.1905887544155121,
      "num_tokens": 16834200.0,
      "step": 7540
    },
    {
      "entropy": 5.609530782699585,
      "epoch": 0.8474195541079351,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004938919492813601,
      "loss": 5.4015,
      "mean_token_accuracy": 0.18654782176017762,
      "num_tokens": 16846399.0,
      "step": 7545
    },
    {
      "entropy": 5.661482143402099,
      "epoch": 0.8479811310158927,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004938826544400461,
      "loss": 5.5441,
      "mean_token_accuracy": 0.18139485865831376,
      "num_tokens": 16858322.0,
      "step": 7550
    },
    {
      "entropy": 5.58749680519104,
      "epoch": 0.8485427079238502,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004938733526294234,
      "loss": 5.3694,
      "mean_token_accuracy": 0.19260751008987426,
      "num_tokens": 16869537.0,
      "step": 7555
    },
    {
      "entropy": 5.664776372909546,
      "epoch": 0.8491042848318077,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004938640438497883,
      "loss": 5.4069,
      "mean_token_accuracy": 0.19095445424318314,
      "num_tokens": 16881566.0,
      "step": 7560
    },
    {
      "entropy": 5.501281452178955,
      "epoch": 0.8496658617397652,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004938547281014372,
      "loss": 5.3331,
      "mean_token_accuracy": 0.19538736939430237,
      "num_tokens": 16892959.0,
      "step": 7565
    },
    {
      "entropy": 5.569608402252197,
      "epoch": 0.8502274386477228,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004938454053846666,
      "loss": 5.3952,
      "mean_token_accuracy": 0.18582664132118226,
      "num_tokens": 16904809.0,
      "step": 7570
    },
    {
      "entropy": 5.60516128540039,
      "epoch": 0.8507890155556803,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004938360756997735,
      "loss": 5.4049,
      "mean_token_accuracy": 0.18384134769439697,
      "num_tokens": 16916969.0,
      "step": 7575
    },
    {
      "entropy": 5.509972047805786,
      "epoch": 0.8513505924636379,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004938267390470547,
      "loss": 5.2925,
      "mean_token_accuracy": 0.1963014855980873,
      "num_tokens": 16928253.0,
      "step": 7580
    },
    {
      "entropy": 5.528853464126587,
      "epoch": 0.8519121693715954,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004938173954268079,
      "loss": 5.3514,
      "mean_token_accuracy": 0.1856411725282669,
      "num_tokens": 16939875.0,
      "step": 7585
    },
    {
      "entropy": 5.687437248229981,
      "epoch": 0.852473746279553,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004938080448393302,
      "loss": 5.4596,
      "mean_token_accuracy": 0.1921010732650757,
      "num_tokens": 16950173.0,
      "step": 7590
    },
    {
      "entropy": 5.532755947113037,
      "epoch": 0.8530353231875105,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004937986872849195,
      "loss": 5.3872,
      "mean_token_accuracy": 0.19412214308977127,
      "num_tokens": 16960933.0,
      "step": 7595
    },
    {
      "entropy": 5.550678634643555,
      "epoch": 0.8535969000954681,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004937893227638738,
      "loss": 5.2888,
      "mean_token_accuracy": 0.19146729558706282,
      "num_tokens": 16972284.0,
      "step": 7600
    },
    {
      "entropy": 5.535526704788208,
      "epoch": 0.8541584770034256,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004937799512764912,
      "loss": 5.3117,
      "mean_token_accuracy": 0.19900417625904082,
      "num_tokens": 16983541.0,
      "step": 7605
    },
    {
      "entropy": 5.586095476150513,
      "epoch": 0.8547200539113832,
      "grad_norm": 0.9375,
      "learning_rate": 0.00049377057282307,
      "loss": 5.4539,
      "mean_token_accuracy": 0.18137775957584382,
      "num_tokens": 16993773.0,
      "step": 7610
    },
    {
      "entropy": 5.684417200088501,
      "epoch": 0.8552816308193407,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000493761187403909,
      "loss": 5.4271,
      "mean_token_accuracy": 0.18703560531139374,
      "num_tokens": 17005287.0,
      "step": 7615
    },
    {
      "entropy": 5.5849937915802,
      "epoch": 0.8558432077272983,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000493751795019307,
      "loss": 5.3677,
      "mean_token_accuracy": 0.19125597327947616,
      "num_tokens": 17015664.0,
      "step": 7620
    },
    {
      "entropy": 5.5015625,
      "epoch": 0.8564047846352558,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004937423956695629,
      "loss": 5.278,
      "mean_token_accuracy": 0.19442010968923568,
      "num_tokens": 17026876.0,
      "step": 7625
    },
    {
      "entropy": 5.514821767807007,
      "epoch": 0.8569663615432134,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004937329893549762,
      "loss": 5.3138,
      "mean_token_accuracy": 0.20338367074728012,
      "num_tokens": 17037321.0,
      "step": 7630
    },
    {
      "entropy": 5.557236289978027,
      "epoch": 0.8575279384511709,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004937235760758462,
      "loss": 5.4103,
      "mean_token_accuracy": 0.1925960212945938,
      "num_tokens": 17047506.0,
      "step": 7635
    },
    {
      "entropy": 5.6168313980102536,
      "epoch": 0.8580895153591285,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004937141558324726,
      "loss": 5.4388,
      "mean_token_accuracy": 0.1877633735537529,
      "num_tokens": 17059710.0,
      "step": 7640
    },
    {
      "entropy": 5.627493476867675,
      "epoch": 0.858651092267086,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004937047286251556,
      "loss": 5.4613,
      "mean_token_accuracy": 0.1798863798379898,
      "num_tokens": 17070956.0,
      "step": 7645
    },
    {
      "entropy": 5.585094022750854,
      "epoch": 0.8592126691750436,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004936952944541952,
      "loss": 5.4173,
      "mean_token_accuracy": 0.1886191412806511,
      "num_tokens": 17081404.0,
      "step": 7650
    },
    {
      "entropy": 5.664061737060547,
      "epoch": 0.859774246083001,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004936858533198917,
      "loss": 5.4388,
      "mean_token_accuracy": 0.18900777399539948,
      "num_tokens": 17092692.0,
      "step": 7655
    },
    {
      "entropy": 5.534651756286621,
      "epoch": 0.8603358229909586,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000493676405222546,
      "loss": 5.284,
      "mean_token_accuracy": 0.1939801588654518,
      "num_tokens": 17103730.0,
      "step": 7660
    },
    {
      "entropy": 5.578586006164551,
      "epoch": 0.8608973998989161,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004936669501624586,
      "loss": 5.3678,
      "mean_token_accuracy": 0.18749492913484572,
      "num_tokens": 17115086.0,
      "step": 7665
    },
    {
      "entropy": 5.58275260925293,
      "epoch": 0.8614589768068737,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004936574881399308,
      "loss": 5.3733,
      "mean_token_accuracy": 0.1900656431913376,
      "num_tokens": 17125928.0,
      "step": 7670
    },
    {
      "entropy": 5.572338962554932,
      "epoch": 0.8620205537148312,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004936480191552637,
      "loss": 5.3881,
      "mean_token_accuracy": 0.18689943552017213,
      "num_tokens": 17137328.0,
      "step": 7675
    },
    {
      "entropy": 5.574598073959351,
      "epoch": 0.8625821306227888,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004936385432087589,
      "loss": 5.3649,
      "mean_token_accuracy": 0.19075820297002793,
      "num_tokens": 17148274.0,
      "step": 7680
    },
    {
      "entropy": 5.5741301536560055,
      "epoch": 0.8631437075307463,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004936290603007181,
      "loss": 5.4113,
      "mean_token_accuracy": 0.18898654133081436,
      "num_tokens": 17158834.0,
      "step": 7685
    },
    {
      "entropy": 5.510143423080445,
      "epoch": 0.8637052844387039,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004936195704314432,
      "loss": 5.2595,
      "mean_token_accuracy": 0.19425098448991776,
      "num_tokens": 17169276.0,
      "step": 7690
    },
    {
      "entropy": 5.525685501098633,
      "epoch": 0.8642668613466614,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004936100736012363,
      "loss": 5.3131,
      "mean_token_accuracy": 0.19559844732284545,
      "num_tokens": 17180642.0,
      "step": 7695
    },
    {
      "entropy": 5.553886985778808,
      "epoch": 0.864828438254619,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004936005698104,
      "loss": 5.3706,
      "mean_token_accuracy": 0.1983370766043663,
      "num_tokens": 17192477.0,
      "step": 7700
    },
    {
      "entropy": 5.637523126602173,
      "epoch": 0.8653900151625765,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004935910590592367,
      "loss": 5.5051,
      "mean_token_accuracy": 0.1851526230573654,
      "num_tokens": 17202653.0,
      "step": 7705
    },
    {
      "entropy": 5.464889812469482,
      "epoch": 0.8659515920705341,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004935815413480494,
      "loss": 5.2497,
      "mean_token_accuracy": 0.19569826424121856,
      "num_tokens": 17212607.0,
      "step": 7710
    },
    {
      "entropy": 5.542541742324829,
      "epoch": 0.8665131689784916,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000493572016677141,
      "loss": 5.3872,
      "mean_token_accuracy": 0.19178338795900346,
      "num_tokens": 17223701.0,
      "step": 7715
    },
    {
      "entropy": 5.588152503967285,
      "epoch": 0.8670747458864492,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004935624850468148,
      "loss": 5.2939,
      "mean_token_accuracy": 0.19530390948057175,
      "num_tokens": 17234878.0,
      "step": 7720
    },
    {
      "entropy": 5.5534826755523685,
      "epoch": 0.8676363227944067,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004935529464573742,
      "loss": 5.3674,
      "mean_token_accuracy": 0.19480834007263184,
      "num_tokens": 17245813.0,
      "step": 7725
    },
    {
      "entropy": 5.567597150802612,
      "epoch": 0.8681978997023643,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004935434009091232,
      "loss": 5.4477,
      "mean_token_accuracy": 0.18925633281469345,
      "num_tokens": 17257378.0,
      "step": 7730
    },
    {
      "entropy": 5.535342264175415,
      "epoch": 0.8687594766103218,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004935338484023655,
      "loss": 5.3668,
      "mean_token_accuracy": 0.19339324682950973,
      "num_tokens": 17269108.0,
      "step": 7735
    },
    {
      "entropy": 5.591677808761597,
      "epoch": 0.8693210535182794,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004935242889374052,
      "loss": 5.3397,
      "mean_token_accuracy": 0.1855355754494667,
      "num_tokens": 17280643.0,
      "step": 7740
    },
    {
      "entropy": 5.6471106052398685,
      "epoch": 0.8698826304262369,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004935147225145469,
      "loss": 5.4151,
      "mean_token_accuracy": 0.19462792426347733,
      "num_tokens": 17291236.0,
      "step": 7745
    },
    {
      "entropy": 5.6206165790557865,
      "epoch": 0.8704442073341944,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004935051491340951,
      "loss": 5.4723,
      "mean_token_accuracy": 0.18134547919034957,
      "num_tokens": 17303547.0,
      "step": 7750
    },
    {
      "entropy": 5.653502893447876,
      "epoch": 0.8710057842421519,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004934955687963546,
      "loss": 5.4523,
      "mean_token_accuracy": 0.19064424633979798,
      "num_tokens": 17314803.0,
      "step": 7755
    },
    {
      "entropy": 5.535930919647217,
      "epoch": 0.8715673611501095,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004934859815016304,
      "loss": 5.3237,
      "mean_token_accuracy": 0.18816096633672713,
      "num_tokens": 17326009.0,
      "step": 7760
    },
    {
      "entropy": 5.634857511520385,
      "epoch": 0.872128938058067,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004934763872502278,
      "loss": 5.3351,
      "mean_token_accuracy": 0.19198465198278428,
      "num_tokens": 17337078.0,
      "step": 7765
    },
    {
      "entropy": 5.574754953384399,
      "epoch": 0.8726905149660246,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004934667860424523,
      "loss": 5.457,
      "mean_token_accuracy": 0.18738923370838165,
      "num_tokens": 17349266.0,
      "step": 7770
    },
    {
      "entropy": 5.426212310791016,
      "epoch": 0.8732520918739821,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004934571778786096,
      "loss": 5.2548,
      "mean_token_accuracy": 0.19445228427648545,
      "num_tokens": 17360327.0,
      "step": 7775
    },
    {
      "entropy": 5.668573904037475,
      "epoch": 0.8738136687819397,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004934475627590057,
      "loss": 5.4021,
      "mean_token_accuracy": 0.18578940033912658,
      "num_tokens": 17372727.0,
      "step": 7780
    },
    {
      "entropy": 5.648229169845581,
      "epoch": 0.8743752456898972,
      "grad_norm": 0.875,
      "learning_rate": 0.0004934379406839466,
      "loss": 5.4682,
      "mean_token_accuracy": 0.18564943671226503,
      "num_tokens": 17384160.0,
      "step": 7785
    },
    {
      "entropy": 5.565241289138794,
      "epoch": 0.8749368225978548,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004934283116537387,
      "loss": 5.3994,
      "mean_token_accuracy": 0.18572111129760743,
      "num_tokens": 17396009.0,
      "step": 7790
    },
    {
      "entropy": 5.59627890586853,
      "epoch": 0.8754983995058123,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004934186756686888,
      "loss": 5.4002,
      "mean_token_accuracy": 0.19138868749141694,
      "num_tokens": 17406903.0,
      "step": 7795
    },
    {
      "entropy": 5.518985652923584,
      "epoch": 0.8760599764137699,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004934090327291034,
      "loss": 5.2569,
      "mean_token_accuracy": 0.19830069988965987,
      "num_tokens": 17417694.0,
      "step": 7800
    },
    {
      "entropy": 5.536848449707032,
      "epoch": 0.8766215533217274,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004933993828352898,
      "loss": 5.4955,
      "mean_token_accuracy": 0.1933062881231308,
      "num_tokens": 17428731.0,
      "step": 7805
    },
    {
      "entropy": 5.5544257164001465,
      "epoch": 0.877183130229685,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004933897259875551,
      "loss": 5.2355,
      "mean_token_accuracy": 0.19531145840883254,
      "num_tokens": 17439242.0,
      "step": 7810
    },
    {
      "entropy": 5.610169696807861,
      "epoch": 0.8777447071376425,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004933800621862069,
      "loss": 5.3638,
      "mean_token_accuracy": 0.18832424581050872,
      "num_tokens": 17449980.0,
      "step": 7815
    },
    {
      "entropy": 5.63561520576477,
      "epoch": 0.8783062840456001,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004933703914315527,
      "loss": 5.4485,
      "mean_token_accuracy": 0.18203552216291427,
      "num_tokens": 17464140.0,
      "step": 7820
    },
    {
      "entropy": 5.655234003067017,
      "epoch": 0.8788678609535576,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004933607137239006,
      "loss": 5.4593,
      "mean_token_accuracy": 0.18870255947113038,
      "num_tokens": 17476246.0,
      "step": 7825
    },
    {
      "entropy": 5.620034646987915,
      "epoch": 0.8794294378615152,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004933510290635587,
      "loss": 5.447,
      "mean_token_accuracy": 0.17544399052858353,
      "num_tokens": 17488169.0,
      "step": 7830
    },
    {
      "entropy": 5.523562431335449,
      "epoch": 0.8799910147694727,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004933413374508353,
      "loss": 5.2925,
      "mean_token_accuracy": 0.19425535351037979,
      "num_tokens": 17498465.0,
      "step": 7835
    },
    {
      "entropy": 5.578102016448975,
      "epoch": 0.8805525916774303,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004933316388860391,
      "loss": 5.2907,
      "mean_token_accuracy": 0.20331917256116866,
      "num_tokens": 17509217.0,
      "step": 7840
    },
    {
      "entropy": 5.5390363216400145,
      "epoch": 0.8811141685853877,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004933219333694787,
      "loss": 5.3235,
      "mean_token_accuracy": 0.1949213981628418,
      "num_tokens": 17519851.0,
      "step": 7845
    },
    {
      "entropy": 5.512254047393799,
      "epoch": 0.8816757454933453,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004933122209014634,
      "loss": 5.3714,
      "mean_token_accuracy": 0.1975010797381401,
      "num_tokens": 17532144.0,
      "step": 7850
    },
    {
      "entropy": 5.577784252166748,
      "epoch": 0.8822373224013028,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004933025014823022,
      "loss": 5.2905,
      "mean_token_accuracy": 0.19637045562267302,
      "num_tokens": 17542929.0,
      "step": 7855
    },
    {
      "entropy": 5.636116981506348,
      "epoch": 0.8827988993092604,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004932927751123048,
      "loss": 5.4722,
      "mean_token_accuracy": 0.18297981917858125,
      "num_tokens": 17554931.0,
      "step": 7860
    },
    {
      "entropy": 5.49958381652832,
      "epoch": 0.8833604762172179,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004932830417917808,
      "loss": 5.2782,
      "mean_token_accuracy": 0.19264158755540847,
      "num_tokens": 17566026.0,
      "step": 7865
    },
    {
      "entropy": 5.536040830612182,
      "epoch": 0.8839220531251755,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00049327330152104,
      "loss": 5.3824,
      "mean_token_accuracy": 0.1927091062068939,
      "num_tokens": 17578396.0,
      "step": 7870
    },
    {
      "entropy": 5.612289619445801,
      "epoch": 0.884483630033133,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004932635543003927,
      "loss": 5.3768,
      "mean_token_accuracy": 0.19341448992490767,
      "num_tokens": 17590559.0,
      "step": 7875
    },
    {
      "entropy": 5.537471771240234,
      "epoch": 0.8850452069410906,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004932538001301491,
      "loss": 5.3416,
      "mean_token_accuracy": 0.19170386642217635,
      "num_tokens": 17600833.0,
      "step": 7880
    },
    {
      "entropy": 5.525249481201172,
      "epoch": 0.8856067838490481,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004932440390106199,
      "loss": 5.3269,
      "mean_token_accuracy": 0.193048195540905,
      "num_tokens": 17612166.0,
      "step": 7885
    },
    {
      "entropy": 5.630873298645019,
      "epoch": 0.8861683607570057,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004932342709421159,
      "loss": 5.4388,
      "mean_token_accuracy": 0.18851923793554307,
      "num_tokens": 17623070.0,
      "step": 7890
    },
    {
      "entropy": 5.633476686477661,
      "epoch": 0.8867299376649632,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000493224495924948,
      "loss": 5.5164,
      "mean_token_accuracy": 0.1864357993006706,
      "num_tokens": 17633319.0,
      "step": 7895
    },
    {
      "entropy": 5.560423278808594,
      "epoch": 0.8872915145729208,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004932147139594276,
      "loss": 5.3248,
      "mean_token_accuracy": 0.19333451688289643,
      "num_tokens": 17643818.0,
      "step": 7900
    },
    {
      "entropy": 5.525871324539184,
      "epoch": 0.8878530914808783,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004932049250458659,
      "loss": 5.3066,
      "mean_token_accuracy": 0.19413001835346222,
      "num_tokens": 17655665.0,
      "step": 7905
    },
    {
      "entropy": 5.529631090164185,
      "epoch": 0.8884146683888359,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004931951291845749,
      "loss": 5.2367,
      "mean_token_accuracy": 0.20333325266838073,
      "num_tokens": 17666200.0,
      "step": 7910
    },
    {
      "entropy": 5.529998302459717,
      "epoch": 0.8889762452967934,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004931853263758663,
      "loss": 5.4274,
      "mean_token_accuracy": 0.18334907591342925,
      "num_tokens": 17677299.0,
      "step": 7915
    },
    {
      "entropy": 5.5925451755523685,
      "epoch": 0.889537822204751,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004931755166200523,
      "loss": 5.304,
      "mean_token_accuracy": 0.20149568915367128,
      "num_tokens": 17687870.0,
      "step": 7920
    },
    {
      "entropy": 5.504533290863037,
      "epoch": 0.8900993991127085,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004931656999174452,
      "loss": 5.3262,
      "mean_token_accuracy": 0.18994814306497573,
      "num_tokens": 17698853.0,
      "step": 7925
    },
    {
      "entropy": 5.556023359298706,
      "epoch": 0.8906609760206661,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004931558762683577,
      "loss": 5.3744,
      "mean_token_accuracy": 0.19724411517381668,
      "num_tokens": 17709897.0,
      "step": 7930
    },
    {
      "entropy": 5.576913976669312,
      "epoch": 0.8912225529286236,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004931460456731024,
      "loss": 5.3564,
      "mean_token_accuracy": 0.19360800087451935,
      "num_tokens": 17720974.0,
      "step": 7935
    },
    {
      "entropy": 5.665620470046997,
      "epoch": 0.8917841298365812,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004931362081319925,
      "loss": 5.5085,
      "mean_token_accuracy": 0.18557514995336533,
      "num_tokens": 17734449.0,
      "step": 7940
    },
    {
      "entropy": 5.472063589096069,
      "epoch": 0.8923457067445386,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004931263636453409,
      "loss": 5.2675,
      "mean_token_accuracy": 0.19221799522638322,
      "num_tokens": 17744735.0,
      "step": 7945
    },
    {
      "entropy": 5.483690786361694,
      "epoch": 0.8929072836524962,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004931165122134616,
      "loss": 5.2939,
      "mean_token_accuracy": 0.1950702428817749,
      "num_tokens": 17755641.0,
      "step": 7950
    },
    {
      "entropy": 5.579967927932739,
      "epoch": 0.8934688605604537,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004931066538366677,
      "loss": 5.3935,
      "mean_token_accuracy": 0.18724148869514465,
      "num_tokens": 17767621.0,
      "step": 7955
    },
    {
      "entropy": 5.713730430603027,
      "epoch": 0.8940304374684113,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004930967885152734,
      "loss": 5.4742,
      "mean_token_accuracy": 0.18756558746099472,
      "num_tokens": 17779507.0,
      "step": 7960
    },
    {
      "entropy": 5.543541288375854,
      "epoch": 0.8945920143763688,
      "grad_norm": 0.875,
      "learning_rate": 0.0004930869162495929,
      "loss": 5.3232,
      "mean_token_accuracy": 0.1917298048734665,
      "num_tokens": 17790371.0,
      "step": 7965
    },
    {
      "entropy": 5.535320043563843,
      "epoch": 0.8951535912843264,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004930770370399403,
      "loss": 5.357,
      "mean_token_accuracy": 0.1943719580769539,
      "num_tokens": 17800903.0,
      "step": 7970
    },
    {
      "entropy": 5.663268947601319,
      "epoch": 0.8957151681922839,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004930671508866303,
      "loss": 5.536,
      "mean_token_accuracy": 0.17450857907533646,
      "num_tokens": 17812053.0,
      "step": 7975
    },
    {
      "entropy": 5.529007863998413,
      "epoch": 0.8962767451002415,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004930572577899776,
      "loss": 5.2556,
      "mean_token_accuracy": 0.19764071255922316,
      "num_tokens": 17822113.0,
      "step": 7980
    },
    {
      "entropy": 5.5263481616973875,
      "epoch": 0.896838322008199,
      "grad_norm": 0.875,
      "learning_rate": 0.0004930473577502972,
      "loss": 5.3101,
      "mean_token_accuracy": 0.19586232900619507,
      "num_tokens": 17832741.0,
      "step": 7985
    },
    {
      "entropy": 5.424175262451172,
      "epoch": 0.8973998989161566,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004930374507679044,
      "loss": 5.2408,
      "mean_token_accuracy": 0.19129867404699324,
      "num_tokens": 17842968.0,
      "step": 7990
    },
    {
      "entropy": 5.654350709915161,
      "epoch": 0.8979614758241141,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004930275368431146,
      "loss": 5.4952,
      "mean_token_accuracy": 0.18171659708023072,
      "num_tokens": 17854295.0,
      "step": 7995
    },
    {
      "entropy": 5.672578620910644,
      "epoch": 0.8985230527320717,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004930176159762435,
      "loss": 5.426,
      "mean_token_accuracy": 0.19333248883485793,
      "num_tokens": 17865597.0,
      "step": 8000
    },
    {
      "entropy": 5.475166606903076,
      "epoch": 0.8990846296400292,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004930076881676069,
      "loss": 5.2193,
      "mean_token_accuracy": 0.19720959663391113,
      "num_tokens": 17877405.0,
      "step": 8005
    },
    {
      "entropy": 5.657891702651978,
      "epoch": 0.8996462065479868,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004929977534175211,
      "loss": 5.3982,
      "mean_token_accuracy": 0.18937858939170837,
      "num_tokens": 17889273.0,
      "step": 8010
    },
    {
      "entropy": 5.49162917137146,
      "epoch": 0.9002077834559443,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004929878117263021,
      "loss": 5.3097,
      "mean_token_accuracy": 0.19807998538017274,
      "num_tokens": 17899761.0,
      "step": 8015
    },
    {
      "entropy": 5.511668682098389,
      "epoch": 0.9007693603639019,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004929778630942668,
      "loss": 5.324,
      "mean_token_accuracy": 0.19051908999681472,
      "num_tokens": 17910607.0,
      "step": 8020
    },
    {
      "entropy": 5.582989263534546,
      "epoch": 0.9013309372718594,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004929679075217317,
      "loss": 5.3744,
      "mean_token_accuracy": 0.18642621040344237,
      "num_tokens": 17922538.0,
      "step": 8025
    },
    {
      "entropy": 5.547327280044556,
      "epoch": 0.901892514179817,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004929579450090139,
      "loss": 5.4106,
      "mean_token_accuracy": 0.1927017793059349,
      "num_tokens": 17933313.0,
      "step": 8030
    },
    {
      "entropy": 5.652021217346191,
      "epoch": 0.9024540910877745,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004929479755564305,
      "loss": 5.4893,
      "mean_token_accuracy": 0.18322046995162963,
      "num_tokens": 17944893.0,
      "step": 8035
    },
    {
      "entropy": 5.57831711769104,
      "epoch": 0.903015667995732,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004929379991642991,
      "loss": 5.3637,
      "mean_token_accuracy": 0.19140897691249847,
      "num_tokens": 17955019.0,
      "step": 8040
    },
    {
      "entropy": 5.674903345108032,
      "epoch": 0.9035772449036895,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004929280158329373,
      "loss": 5.3985,
      "mean_token_accuracy": 0.1890375629067421,
      "num_tokens": 17965793.0,
      "step": 8045
    },
    {
      "entropy": 5.531117105484009,
      "epoch": 0.9041388218116471,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000492918025562663,
      "loss": 5.2364,
      "mean_token_accuracy": 0.19421636909246445,
      "num_tokens": 17975709.0,
      "step": 8050
    },
    {
      "entropy": 5.6074751853942875,
      "epoch": 0.9047003987196046,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004929080283537942,
      "loss": 5.4638,
      "mean_token_accuracy": 0.18565727323293685,
      "num_tokens": 17985405.0,
      "step": 8055
    },
    {
      "entropy": 5.618918800354004,
      "epoch": 0.9052619756275622,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004928980242066492,
      "loss": 5.4797,
      "mean_token_accuracy": 0.18802008032798767,
      "num_tokens": 17997105.0,
      "step": 8060
    },
    {
      "entropy": 5.596008920669556,
      "epoch": 0.9058235525355197,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004928880131215467,
      "loss": 5.3806,
      "mean_token_accuracy": 0.19388099163770675,
      "num_tokens": 18007914.0,
      "step": 8065
    },
    {
      "entropy": 5.528563833236694,
      "epoch": 0.9063851294434773,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004928779950988053,
      "loss": 5.3799,
      "mean_token_accuracy": 0.1898500069975853,
      "num_tokens": 18018328.0,
      "step": 8070
    },
    {
      "entropy": 5.589106130599975,
      "epoch": 0.9069467063514348,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000492867970138744,
      "loss": 5.3385,
      "mean_token_accuracy": 0.1902944937348366,
      "num_tokens": 18028568.0,
      "step": 8075
    },
    {
      "entropy": 5.599682855606079,
      "epoch": 0.9075082832593924,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004928579382416821,
      "loss": 5.3492,
      "mean_token_accuracy": 0.1922268256545067,
      "num_tokens": 18038655.0,
      "step": 8080
    },
    {
      "entropy": 5.513388538360596,
      "epoch": 0.9080698601673499,
      "grad_norm": 0.921875,
      "learning_rate": 0.000492847899407939,
      "loss": 5.2925,
      "mean_token_accuracy": 0.19455196261405944,
      "num_tokens": 18049716.0,
      "step": 8085
    },
    {
      "entropy": 5.474845504760742,
      "epoch": 0.9086314370753075,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004928378536378341,
      "loss": 5.2758,
      "mean_token_accuracy": 0.19619437605142592,
      "num_tokens": 18062530.0,
      "step": 8090
    },
    {
      "entropy": 5.540895414352417,
      "epoch": 0.909193013983265,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004928278009316877,
      "loss": 5.379,
      "mean_token_accuracy": 0.1869027018547058,
      "num_tokens": 18073076.0,
      "step": 8095
    },
    {
      "entropy": 5.596837520599365,
      "epoch": 0.9097545908912226,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004928177412898195,
      "loss": 5.312,
      "mean_token_accuracy": 0.18975626975297927,
      "num_tokens": 18083759.0,
      "step": 8100
    },
    {
      "entropy": 5.491744184494019,
      "epoch": 0.9103161677991801,
      "grad_norm": 0.875,
      "learning_rate": 0.0004928076747125499,
      "loss": 5.2652,
      "mean_token_accuracy": 0.19359455555677413,
      "num_tokens": 18095980.0,
      "step": 8105
    },
    {
      "entropy": 5.557895565032959,
      "epoch": 0.9108777447071377,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004927976012001995,
      "loss": 5.3629,
      "mean_token_accuracy": 0.19761702716350554,
      "num_tokens": 18106672.0,
      "step": 8110
    },
    {
      "entropy": 5.516842794418335,
      "epoch": 0.9114393216150952,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004927875207530889,
      "loss": 5.3748,
      "mean_token_accuracy": 0.18991637825965882,
      "num_tokens": 18117481.0,
      "step": 8115
    },
    {
      "entropy": 5.541539812088013,
      "epoch": 0.9120008985230528,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004927774333715394,
      "loss": 5.3223,
      "mean_token_accuracy": 0.1943124935030937,
      "num_tokens": 18127602.0,
      "step": 8120
    },
    {
      "entropy": 5.621410274505616,
      "epoch": 0.9125624754310103,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004927673390558718,
      "loss": 5.3833,
      "mean_token_accuracy": 0.1924299031496048,
      "num_tokens": 18140109.0,
      "step": 8125
    },
    {
      "entropy": 5.507348871231079,
      "epoch": 0.9131240523389679,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004927572378064077,
      "loss": 5.2535,
      "mean_token_accuracy": 0.19700659364461898,
      "num_tokens": 18150271.0,
      "step": 8130
    },
    {
      "entropy": 5.566203784942627,
      "epoch": 0.9136856292469253,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004927471296234687,
      "loss": 5.3769,
      "mean_token_accuracy": 0.19731779396533966,
      "num_tokens": 18160746.0,
      "step": 8135
    },
    {
      "entropy": 5.6196905136108395,
      "epoch": 0.9142472061548829,
      "grad_norm": 0.875,
      "learning_rate": 0.0004927370145073767,
      "loss": 5.3677,
      "mean_token_accuracy": 0.1905166283249855,
      "num_tokens": 18172271.0,
      "step": 8140
    },
    {
      "entropy": 5.537528467178345,
      "epoch": 0.9148087830628404,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004927268924584536,
      "loss": 5.3126,
      "mean_token_accuracy": 0.19049279540777206,
      "num_tokens": 18183116.0,
      "step": 8145
    },
    {
      "entropy": 5.5165931224823,
      "epoch": 0.915370359970798,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004927167634770219,
      "loss": 5.2657,
      "mean_token_accuracy": 0.20448511242866516,
      "num_tokens": 18195147.0,
      "step": 8150
    },
    {
      "entropy": 5.531662893295288,
      "epoch": 0.9159319368787555,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000492706627563404,
      "loss": 5.3307,
      "mean_token_accuracy": 0.19220539331436157,
      "num_tokens": 18206217.0,
      "step": 8155
    },
    {
      "entropy": 5.548134613037109,
      "epoch": 0.9164935137867131,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004926964847179225,
      "loss": 5.3377,
      "mean_token_accuracy": 0.18856058418750762,
      "num_tokens": 18217434.0,
      "step": 8160
    },
    {
      "entropy": 5.59363431930542,
      "epoch": 0.9170550906946706,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004926863349409006,
      "loss": 5.3458,
      "mean_token_accuracy": 0.1884212538599968,
      "num_tokens": 18229277.0,
      "step": 8165
    },
    {
      "entropy": 5.512733125686646,
      "epoch": 0.9176166676026282,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004926761782326613,
      "loss": 5.2131,
      "mean_token_accuracy": 0.19836175590753555,
      "num_tokens": 18240320.0,
      "step": 8170
    },
    {
      "entropy": 5.512052154541015,
      "epoch": 0.9181782445105857,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004926660145935282,
      "loss": 5.2555,
      "mean_token_accuracy": 0.20000111013650895,
      "num_tokens": 18250380.0,
      "step": 8175
    },
    {
      "entropy": 5.5467030048370365,
      "epoch": 0.9187398214185433,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004926558440238247,
      "loss": 5.4236,
      "mean_token_accuracy": 0.19226550310850143,
      "num_tokens": 18261703.0,
      "step": 8180
    },
    {
      "entropy": 5.5692695617675785,
      "epoch": 0.9193013983265008,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004926456665238746,
      "loss": 5.3638,
      "mean_token_accuracy": 0.1893811494112015,
      "num_tokens": 18272472.0,
      "step": 8185
    },
    {
      "entropy": 5.666653347015381,
      "epoch": 0.9198629752344584,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004926354820940023,
      "loss": 5.4401,
      "mean_token_accuracy": 0.18993701338768004,
      "num_tokens": 18282168.0,
      "step": 8190
    },
    {
      "entropy": 5.669243097305298,
      "epoch": 0.9204245521424159,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004926252907345317,
      "loss": 5.4371,
      "mean_token_accuracy": 0.18197981417179107,
      "num_tokens": 18293750.0,
      "step": 8195
    },
    {
      "entropy": 5.551488780975342,
      "epoch": 0.9209861290503735,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004926150924457874,
      "loss": 5.3437,
      "mean_token_accuracy": 0.1958581879734993,
      "num_tokens": 18304568.0,
      "step": 8200
    },
    {
      "entropy": 5.504587841033936,
      "epoch": 0.921547705958331,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004926048872280943,
      "loss": 5.2913,
      "mean_token_accuracy": 0.19624826759099961,
      "num_tokens": 18315129.0,
      "step": 8205
    },
    {
      "entropy": 5.486863994598389,
      "epoch": 0.9221092828662886,
      "grad_norm": 0.96875,
      "learning_rate": 0.000492594675081777,
      "loss": 5.2496,
      "mean_token_accuracy": 0.1928882509469986,
      "num_tokens": 18324628.0,
      "step": 8210
    },
    {
      "entropy": 5.424786043167114,
      "epoch": 0.9226708597742461,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000492584456007161,
      "loss": 5.2687,
      "mean_token_accuracy": 0.20118917673826217,
      "num_tokens": 18334899.0,
      "step": 8215
    },
    {
      "entropy": 5.56889910697937,
      "epoch": 0.9232324366822037,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004925742300045716,
      "loss": 5.2767,
      "mean_token_accuracy": 0.1952001690864563,
      "num_tokens": 18345719.0,
      "step": 8220
    },
    {
      "entropy": 5.598507738113403,
      "epoch": 0.9237940135901612,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004925639970743341,
      "loss": 5.3519,
      "mean_token_accuracy": 0.19173127114772798,
      "num_tokens": 18355814.0,
      "step": 8225
    },
    {
      "entropy": 5.597356796264648,
      "epoch": 0.9243555904981187,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004925537572167746,
      "loss": 5.4328,
      "mean_token_accuracy": 0.1876605972647667,
      "num_tokens": 18366746.0,
      "step": 8230
    },
    {
      "entropy": 5.512669420242309,
      "epoch": 0.9249171674060762,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004925435104322192,
      "loss": 5.2528,
      "mean_token_accuracy": 0.1983408436179161,
      "num_tokens": 18375846.0,
      "step": 8235
    },
    {
      "entropy": 5.488119220733642,
      "epoch": 0.9254787443140338,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000492533256720994,
      "loss": 5.3727,
      "mean_token_accuracy": 0.19273528903722764,
      "num_tokens": 18386850.0,
      "step": 8240
    },
    {
      "entropy": 5.524377059936524,
      "epoch": 0.9260403212219913,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004925229960834256,
      "loss": 5.2618,
      "mean_token_accuracy": 0.19778251349925996,
      "num_tokens": 18396347.0,
      "step": 8245
    },
    {
      "entropy": 5.531424474716187,
      "epoch": 0.9266018981299489,
      "grad_norm": 1.0,
      "learning_rate": 0.0004925127285198407,
      "loss": 5.3062,
      "mean_token_accuracy": 0.19472510516643524,
      "num_tokens": 18406583.0,
      "step": 8250
    },
    {
      "entropy": 5.62885217666626,
      "epoch": 0.9271634750379064,
      "grad_norm": 0.90625,
      "learning_rate": 0.000492502454030566,
      "loss": 5.4293,
      "mean_token_accuracy": 0.1894119054079056,
      "num_tokens": 18418586.0,
      "step": 8255
    },
    {
      "entropy": 5.563667106628418,
      "epoch": 0.927725051945864,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004924921726159288,
      "loss": 5.2609,
      "mean_token_accuracy": 0.19322708249092102,
      "num_tokens": 18429752.0,
      "step": 8260
    },
    {
      "entropy": 5.542651796340943,
      "epoch": 0.9282866288538215,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004924818842762565,
      "loss": 5.3518,
      "mean_token_accuracy": 0.18900565207004547,
      "num_tokens": 18440663.0,
      "step": 8265
    },
    {
      "entropy": 5.4974071979522705,
      "epoch": 0.9288482057617791,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004924715890118768,
      "loss": 5.2987,
      "mean_token_accuracy": 0.19703351110219955,
      "num_tokens": 18451497.0,
      "step": 8270
    },
    {
      "entropy": 5.570805358886719,
      "epoch": 0.9294097826697366,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004924612868231173,
      "loss": 5.2954,
      "mean_token_accuracy": 0.18560000360012055,
      "num_tokens": 18462794.0,
      "step": 8275
    },
    {
      "entropy": 5.528308010101318,
      "epoch": 0.9299713595776942,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004924509777103059,
      "loss": 5.2825,
      "mean_token_accuracy": 0.19653030335903168,
      "num_tokens": 18472973.0,
      "step": 8280
    },
    {
      "entropy": 5.562204837799072,
      "epoch": 0.9305329364856517,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004924406616737712,
      "loss": 5.4457,
      "mean_token_accuracy": 0.1930410623550415,
      "num_tokens": 18484451.0,
      "step": 8285
    },
    {
      "entropy": 5.623847484588623,
      "epoch": 0.9310945133936093,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004924303387138415,
      "loss": 5.3037,
      "mean_token_accuracy": 0.20219796150922775,
      "num_tokens": 18496412.0,
      "step": 8290
    },
    {
      "entropy": 5.568887042999267,
      "epoch": 0.9316560903015668,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004924200088308455,
      "loss": 5.3345,
      "mean_token_accuracy": 0.19731771647930146,
      "num_tokens": 18508547.0,
      "step": 8295
    },
    {
      "entropy": 5.489627313613892,
      "epoch": 0.9322176672095244,
      "grad_norm": 0.96875,
      "learning_rate": 0.000492409672025112,
      "loss": 5.366,
      "mean_token_accuracy": 0.1914461597800255,
      "num_tokens": 18518954.0,
      "step": 8300
    },
    {
      "entropy": 5.5001569271087645,
      "epoch": 0.9327792441174819,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004923993282969703,
      "loss": 5.2211,
      "mean_token_accuracy": 0.19981079995632173,
      "num_tokens": 18530444.0,
      "step": 8305
    },
    {
      "entropy": 5.583559036254883,
      "epoch": 0.9333408210254395,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004923889776467495,
      "loss": 5.2891,
      "mean_token_accuracy": 0.19373121559619905,
      "num_tokens": 18541612.0,
      "step": 8310
    },
    {
      "entropy": 5.5619526386260985,
      "epoch": 0.933902397933397,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004923786200747795,
      "loss": 5.446,
      "mean_token_accuracy": 0.185287968814373,
      "num_tokens": 18553316.0,
      "step": 8315
    },
    {
      "entropy": 5.48585786819458,
      "epoch": 0.9344639748413546,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00049236825558139,
      "loss": 5.325,
      "mean_token_accuracy": 0.1941107466816902,
      "num_tokens": 18564468.0,
      "step": 8320
    },
    {
      "entropy": 5.491370820999146,
      "epoch": 0.9350255517493121,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004923578841669108,
      "loss": 5.2518,
      "mean_token_accuracy": 0.20075540542602538,
      "num_tokens": 18575536.0,
      "step": 8325
    },
    {
      "entropy": 5.5512049198150635,
      "epoch": 0.9355871286572696,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004923475058316723,
      "loss": 5.3735,
      "mean_token_accuracy": 0.18168114870786667,
      "num_tokens": 18586906.0,
      "step": 8330
    },
    {
      "entropy": 5.621929168701172,
      "epoch": 0.9361487055652271,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004923371205760049,
      "loss": 5.3459,
      "mean_token_accuracy": 0.19497404545545577,
      "num_tokens": 18599064.0,
      "step": 8335
    },
    {
      "entropy": 5.51338152885437,
      "epoch": 0.9367102824731847,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004923267284002394,
      "loss": 5.3881,
      "mean_token_accuracy": 0.1887581616640091,
      "num_tokens": 18610057.0,
      "step": 8340
    },
    {
      "entropy": 5.461229038238526,
      "epoch": 0.9372718593811422,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004923163293047065,
      "loss": 5.2078,
      "mean_token_accuracy": 0.19492341876029967,
      "num_tokens": 18620008.0,
      "step": 8345
    },
    {
      "entropy": 5.47933292388916,
      "epoch": 0.9378334362890998,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004923059232897374,
      "loss": 5.2121,
      "mean_token_accuracy": 0.1944200277328491,
      "num_tokens": 18630599.0,
      "step": 8350
    },
    {
      "entropy": 5.530019998550415,
      "epoch": 0.9383950131970573,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004922955103556633,
      "loss": 5.3928,
      "mean_token_accuracy": 0.1933053806424141,
      "num_tokens": 18643898.0,
      "step": 8355
    },
    {
      "entropy": 5.588576030731201,
      "epoch": 0.9389565901050149,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000492285090502816,
      "loss": 5.3894,
      "mean_token_accuracy": 0.1873052164912224,
      "num_tokens": 18654760.0,
      "step": 8360
    },
    {
      "entropy": 5.598599529266357,
      "epoch": 0.9395181670129724,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004922746637315272,
      "loss": 5.336,
      "mean_token_accuracy": 0.19302780032157899,
      "num_tokens": 18666036.0,
      "step": 8365
    },
    {
      "entropy": 5.575030755996704,
      "epoch": 0.94007974392093,
      "grad_norm": 0.875,
      "learning_rate": 0.0004922642300421288,
      "loss": 5.3792,
      "mean_token_accuracy": 0.19365965873003005,
      "num_tokens": 18677806.0,
      "step": 8370
    },
    {
      "entropy": 5.603202867507934,
      "epoch": 0.9406413208288875,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000492253789434953,
      "loss": 5.5135,
      "mean_token_accuracy": 0.18439876586198806,
      "num_tokens": 18688971.0,
      "step": 8375
    },
    {
      "entropy": 5.550760078430176,
      "epoch": 0.9412028977368451,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004922433419103323,
      "loss": 5.3778,
      "mean_token_accuracy": 0.18816610425710678,
      "num_tokens": 18700153.0,
      "step": 8380
    },
    {
      "entropy": 5.711612224578857,
      "epoch": 0.9417644746448026,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004922328874685993,
      "loss": 5.4274,
      "mean_token_accuracy": 0.18841125816106796,
      "num_tokens": 18711935.0,
      "step": 8385
    },
    {
      "entropy": 5.443498992919922,
      "epoch": 0.9423260515527602,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004922224261100869,
      "loss": 5.2485,
      "mean_token_accuracy": 0.19647979438304902,
      "num_tokens": 18722694.0,
      "step": 8390
    },
    {
      "entropy": 5.55060510635376,
      "epoch": 0.9428876284607177,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004922119578351283,
      "loss": 5.4049,
      "mean_token_accuracy": 0.19630158692598343,
      "num_tokens": 18733443.0,
      "step": 8395
    },
    {
      "entropy": 5.630517768859863,
      "epoch": 0.9434492053686753,
      "grad_norm": 0.875,
      "learning_rate": 0.0004922014826440566,
      "loss": 5.3036,
      "mean_token_accuracy": 0.19102171212434768,
      "num_tokens": 18743965.0,
      "step": 8400
    },
    {
      "entropy": 5.549164772033691,
      "epoch": 0.9440107822766328,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004921910005372056,
      "loss": 5.3935,
      "mean_token_accuracy": 0.19468172788619995,
      "num_tokens": 18755712.0,
      "step": 8405
    },
    {
      "entropy": 5.54606294631958,
      "epoch": 0.9445723591845904,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004921805115149088,
      "loss": 5.3116,
      "mean_token_accuracy": 0.1930754855275154,
      "num_tokens": 18766756.0,
      "step": 8410
    },
    {
      "entropy": 5.5654528617858885,
      "epoch": 0.9451339360925479,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004921700155775002,
      "loss": 5.4188,
      "mean_token_accuracy": 0.1847954958677292,
      "num_tokens": 18776932.0,
      "step": 8415
    },
    {
      "entropy": 5.660779285430908,
      "epoch": 0.9456955130005055,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004921595127253142,
      "loss": 5.3492,
      "mean_token_accuracy": 0.18943994790315627,
      "num_tokens": 18787197.0,
      "step": 8420
    },
    {
      "entropy": 5.522884225845337,
      "epoch": 0.9462570899084629,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004921490029586852,
      "loss": 5.3051,
      "mean_token_accuracy": 0.19406166672706604,
      "num_tokens": 18797861.0,
      "step": 8425
    },
    {
      "entropy": 5.507755756378174,
      "epoch": 0.9468186668164205,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004921384862779475,
      "loss": 5.2915,
      "mean_token_accuracy": 0.19445022493600844,
      "num_tokens": 18809111.0,
      "step": 8430
    },
    {
      "entropy": 5.5553669929504395,
      "epoch": 0.947380243724378,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004921279626834363,
      "loss": 5.3131,
      "mean_token_accuracy": 0.20412296801805496,
      "num_tokens": 18820236.0,
      "step": 8435
    },
    {
      "entropy": 5.514404058456421,
      "epoch": 0.9479418206323356,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004921174321754865,
      "loss": 5.2971,
      "mean_token_accuracy": 0.19629298150539398,
      "num_tokens": 18831942.0,
      "step": 8440
    },
    {
      "entropy": 5.56217999458313,
      "epoch": 0.9485033975402931,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004921068947544335,
      "loss": 5.339,
      "mean_token_accuracy": 0.19513231813907622,
      "num_tokens": 18843271.0,
      "step": 8445
    },
    {
      "entropy": 5.518742799758911,
      "epoch": 0.9490649744482507,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004920963504206128,
      "loss": 5.3849,
      "mean_token_accuracy": 0.19368332177400588,
      "num_tokens": 18856285.0,
      "step": 8450
    },
    {
      "entropy": 5.623357963562012,
      "epoch": 0.9496265513562082,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00049208579917436,
      "loss": 5.3904,
      "mean_token_accuracy": 0.1901460736989975,
      "num_tokens": 18867203.0,
      "step": 8455
    },
    {
      "entropy": 5.608232164382935,
      "epoch": 0.9501881282641658,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004920752410160113,
      "loss": 5.4474,
      "mean_token_accuracy": 0.1828993320465088,
      "num_tokens": 18878598.0,
      "step": 8460
    },
    {
      "entropy": 5.6119218349456785,
      "epoch": 0.9507497051721233,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004920646759459026,
      "loss": 5.3567,
      "mean_token_accuracy": 0.19444518238306047,
      "num_tokens": 18890233.0,
      "step": 8465
    },
    {
      "entropy": 5.5893336772918705,
      "epoch": 0.9513112820800809,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004920541039643705,
      "loss": 5.4334,
      "mean_token_accuracy": 0.18779987692832947,
      "num_tokens": 18900362.0,
      "step": 8470
    },
    {
      "entropy": 5.5757238388061525,
      "epoch": 0.9518728589880384,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004920435250717515,
      "loss": 5.3625,
      "mean_token_accuracy": 0.19813914149999617,
      "num_tokens": 18911924.0,
      "step": 8475
    },
    {
      "entropy": 5.534209823608398,
      "epoch": 0.952434435895996,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004920329392683826,
      "loss": 5.3825,
      "mean_token_accuracy": 0.18825122267007827,
      "num_tokens": 18923931.0,
      "step": 8480
    },
    {
      "entropy": 5.538607883453369,
      "epoch": 0.9529960128039535,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004920223465546006,
      "loss": 5.2831,
      "mean_token_accuracy": 0.19365378767251967,
      "num_tokens": 18935859.0,
      "step": 8485
    },
    {
      "entropy": 5.561859846115112,
      "epoch": 0.9535575897119111,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004920117469307429,
      "loss": 5.4123,
      "mean_token_accuracy": 0.1889643505215645,
      "num_tokens": 18946939.0,
      "step": 8490
    },
    {
      "entropy": 5.628473615646362,
      "epoch": 0.9541191666198686,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004920011403971471,
      "loss": 5.3836,
      "mean_token_accuracy": 0.18897086381912231,
      "num_tokens": 18958067.0,
      "step": 8495
    },
    {
      "entropy": 5.582294082641601,
      "epoch": 0.9546807435278262,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004919905269541508,
      "loss": 5.2919,
      "mean_token_accuracy": 0.19774951487779618,
      "num_tokens": 18967920.0,
      "step": 8500
    },
    {
      "entropy": 5.5259764194488525,
      "epoch": 0.9552423204357837,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000491979906602092,
      "loss": 5.3104,
      "mean_token_accuracy": 0.199825456738472,
      "num_tokens": 18978244.0,
      "step": 8505
    },
    {
      "entropy": 5.562439489364624,
      "epoch": 0.9558038973437413,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004919692793413089,
      "loss": 5.2908,
      "mean_token_accuracy": 0.19242303520441056,
      "num_tokens": 18988464.0,
      "step": 8510
    },
    {
      "entropy": 5.487201261520386,
      "epoch": 0.9563654742516988,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004919586451721396,
      "loss": 5.2425,
      "mean_token_accuracy": 0.19650277346372605,
      "num_tokens": 18998318.0,
      "step": 8515
    },
    {
      "entropy": 5.488058042526245,
      "epoch": 0.9569270511596563,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004919480040949229,
      "loss": 5.23,
      "mean_token_accuracy": 0.19894272089004517,
      "num_tokens": 19009802.0,
      "step": 8520
    },
    {
      "entropy": 5.61924524307251,
      "epoch": 0.9574886280676138,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004919373561099977,
      "loss": 5.4464,
      "mean_token_accuracy": 0.19334100931882858,
      "num_tokens": 19020174.0,
      "step": 8525
    },
    {
      "entropy": 5.553637933731079,
      "epoch": 0.9580502049755714,
      "grad_norm": 0.84375,
      "learning_rate": 0.000491926701217703,
      "loss": 5.365,
      "mean_token_accuracy": 0.18908390998840333,
      "num_tokens": 19031505.0,
      "step": 8530
    },
    {
      "entropy": 5.536982488632202,
      "epoch": 0.9586117818835289,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004919160394183779,
      "loss": 5.2761,
      "mean_token_accuracy": 0.2001358225941658,
      "num_tokens": 19042052.0,
      "step": 8535
    },
    {
      "entropy": 5.696752119064331,
      "epoch": 0.9591733587914865,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000491905370712362,
      "loss": 5.4774,
      "mean_token_accuracy": 0.18816607743501662,
      "num_tokens": 19052050.0,
      "step": 8540
    },
    {
      "entropy": 5.62021427154541,
      "epoch": 0.959734935699444,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004918946950999949,
      "loss": 5.3878,
      "mean_token_accuracy": 0.19243552088737487,
      "num_tokens": 19062053.0,
      "step": 8545
    },
    {
      "entropy": 5.485058116912842,
      "epoch": 0.9602965126074016,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004918840125816167,
      "loss": 5.3242,
      "mean_token_accuracy": 0.1906524658203125,
      "num_tokens": 19073095.0,
      "step": 8550
    },
    {
      "entropy": 5.562147617340088,
      "epoch": 0.9608580895153591,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004918733231575673,
      "loss": 5.3881,
      "mean_token_accuracy": 0.1923004448413849,
      "num_tokens": 19084521.0,
      "step": 8555
    },
    {
      "entropy": 5.5889898300170895,
      "epoch": 0.9614196664233167,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004918626268281874,
      "loss": 5.3682,
      "mean_token_accuracy": 0.19235495626926422,
      "num_tokens": 19096372.0,
      "step": 8560
    },
    {
      "entropy": 5.675411415100098,
      "epoch": 0.9619812433312742,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004918519235938171,
      "loss": 5.4788,
      "mean_token_accuracy": 0.18506063222885133,
      "num_tokens": 19108535.0,
      "step": 8565
    },
    {
      "entropy": 5.657005357742309,
      "epoch": 0.9625428202392318,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004918412134547976,
      "loss": 5.3371,
      "mean_token_accuracy": 0.19821767657995223,
      "num_tokens": 19119480.0,
      "step": 8570
    },
    {
      "entropy": 5.511168909072876,
      "epoch": 0.9631043971471893,
      "grad_norm": 1.0,
      "learning_rate": 0.0004918304964114696,
      "loss": 5.3272,
      "mean_token_accuracy": 0.19723112583160402,
      "num_tokens": 19131168.0,
      "step": 8575
    },
    {
      "entropy": 5.5765660285949705,
      "epoch": 0.9636659740551469,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004918197724641747,
      "loss": 5.4787,
      "mean_token_accuracy": 0.18758818954229356,
      "num_tokens": 19142454.0,
      "step": 8580
    },
    {
      "entropy": 5.586180114746094,
      "epoch": 0.9642275509631044,
      "grad_norm": 0.76171875,
      "learning_rate": 0.000491809041613254,
      "loss": 5.3292,
      "mean_token_accuracy": 0.19211253374814988,
      "num_tokens": 19154682.0,
      "step": 8585
    },
    {
      "entropy": 5.539849281311035,
      "epoch": 0.964789127871062,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004917983038590495,
      "loss": 5.2647,
      "mean_token_accuracy": 0.19301576763391495,
      "num_tokens": 19165190.0,
      "step": 8590
    },
    {
      "entropy": 5.5794212341308596,
      "epoch": 0.9653507047790195,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004917875592019028,
      "loss": 5.397,
      "mean_token_accuracy": 0.19465365409851074,
      "num_tokens": 19176075.0,
      "step": 8595
    },
    {
      "entropy": 5.5599894523620605,
      "epoch": 0.9659122816869771,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004917768076421561,
      "loss": 5.3951,
      "mean_token_accuracy": 0.190712970495224,
      "num_tokens": 19188450.0,
      "step": 8600
    },
    {
      "entropy": 5.503190326690674,
      "epoch": 0.9664738585949346,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004917660491801518,
      "loss": 5.2607,
      "mean_token_accuracy": 0.2057673826813698,
      "num_tokens": 19201254.0,
      "step": 8605
    },
    {
      "entropy": 5.483463287353516,
      "epoch": 0.9670354355028922,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004917552838162324,
      "loss": 5.3466,
      "mean_token_accuracy": 0.1964012160897255,
      "num_tokens": 19214062.0,
      "step": 8610
    },
    {
      "entropy": 5.54677095413208,
      "epoch": 0.9675970124108496,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004917445115507408,
      "loss": 5.2518,
      "mean_token_accuracy": 0.19913691133260727,
      "num_tokens": 19224844.0,
      "step": 8615
    },
    {
      "entropy": 5.5595475196838375,
      "epoch": 0.9681585893188072,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004917337323840198,
      "loss": 5.4114,
      "mean_token_accuracy": 0.18910901844501496,
      "num_tokens": 19237347.0,
      "step": 8620
    },
    {
      "entropy": 5.566824340820313,
      "epoch": 0.9687201662267647,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004917229463164128,
      "loss": 5.3016,
      "mean_token_accuracy": 0.19430307149887086,
      "num_tokens": 19248743.0,
      "step": 8625
    },
    {
      "entropy": 5.574544525146484,
      "epoch": 0.9692817431347223,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004917121533482631,
      "loss": 5.2939,
      "mean_token_accuracy": 0.1953691765666008,
      "num_tokens": 19259904.0,
      "step": 8630
    },
    {
      "entropy": 5.490644359588623,
      "epoch": 0.9698433200426798,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004917013534799144,
      "loss": 5.2321,
      "mean_token_accuracy": 0.19915363937616348,
      "num_tokens": 19271388.0,
      "step": 8635
    },
    {
      "entropy": 5.503468179702759,
      "epoch": 0.9704048969506374,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004916905467117104,
      "loss": 5.3656,
      "mean_token_accuracy": 0.19418666064739226,
      "num_tokens": 19282790.0,
      "step": 8640
    },
    {
      "entropy": 5.5520261287689205,
      "epoch": 0.9709664738585949,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004916797330439955,
      "loss": 5.273,
      "mean_token_accuracy": 0.19777465164661406,
      "num_tokens": 19293389.0,
      "step": 8645
    },
    {
      "entropy": 5.530950260162354,
      "epoch": 0.9715280507665525,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004916689124771138,
      "loss": 5.3382,
      "mean_token_accuracy": 0.1914778858423233,
      "num_tokens": 19304593.0,
      "step": 8650
    },
    {
      "entropy": 5.533031558990478,
      "epoch": 0.97208962767451,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004916580850114099,
      "loss": 5.258,
      "mean_token_accuracy": 0.19529124051332475,
      "num_tokens": 19317388.0,
      "step": 8655
    },
    {
      "entropy": 5.579355764389038,
      "epoch": 0.9726512045824676,
      "grad_norm": 0.875,
      "learning_rate": 0.0004916472506472286,
      "loss": 5.3341,
      "mean_token_accuracy": 0.19594714790582657,
      "num_tokens": 19329228.0,
      "step": 8660
    },
    {
      "entropy": 5.563479232788086,
      "epoch": 0.9732127814904251,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004916364093849147,
      "loss": 5.3048,
      "mean_token_accuracy": 0.20031571686267852,
      "num_tokens": 19339421.0,
      "step": 8665
    },
    {
      "entropy": 5.433558607101441,
      "epoch": 0.9737743583983827,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004916255612248136,
      "loss": 5.201,
      "mean_token_accuracy": 0.19635034948587418,
      "num_tokens": 19349618.0,
      "step": 8670
    },
    {
      "entropy": 5.412133502960205,
      "epoch": 0.9743359353063402,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004916147061672706,
      "loss": 5.2964,
      "mean_token_accuracy": 0.19377315640449524,
      "num_tokens": 19361324.0,
      "step": 8675
    },
    {
      "entropy": 5.594657707214355,
      "epoch": 0.9748975122142978,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004916038442126314,
      "loss": 5.3056,
      "mean_token_accuracy": 0.1933646932244301,
      "num_tokens": 19373631.0,
      "step": 8680
    },
    {
      "entropy": 5.535374784469605,
      "epoch": 0.9754590891222553,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004915929753612416,
      "loss": 5.2712,
      "mean_token_accuracy": 0.1887732043862343,
      "num_tokens": 19384682.0,
      "step": 8685
    },
    {
      "entropy": 5.587987232208252,
      "epoch": 0.9760206660302129,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004915820996134477,
      "loss": 5.3746,
      "mean_token_accuracy": 0.1901007518172264,
      "num_tokens": 19396344.0,
      "step": 8690
    },
    {
      "entropy": 5.537419605255127,
      "epoch": 0.9765822429381704,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004915712169695956,
      "loss": 5.2472,
      "mean_token_accuracy": 0.19150758683681487,
      "num_tokens": 19406921.0,
      "step": 8695
    },
    {
      "entropy": 5.524228048324585,
      "epoch": 0.977143819846128,
      "grad_norm": 0.875,
      "learning_rate": 0.000491560327430032,
      "loss": 5.3525,
      "mean_token_accuracy": 0.19139362126588821,
      "num_tokens": 19416851.0,
      "step": 8700
    },
    {
      "entropy": 5.548289632797241,
      "epoch": 0.9777053967540855,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004915494309951035,
      "loss": 5.3293,
      "mean_token_accuracy": 0.18962969034910201,
      "num_tokens": 19428180.0,
      "step": 8705
    },
    {
      "entropy": 5.647749185562134,
      "epoch": 0.9782669736620431,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004915385276651573,
      "loss": 5.3755,
      "mean_token_accuracy": 0.19601155817508698,
      "num_tokens": 19439575.0,
      "step": 8710
    },
    {
      "entropy": 5.576963901519775,
      "epoch": 0.9788285505700005,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004915276174405403,
      "loss": 5.3513,
      "mean_token_accuracy": 0.19250843226909636,
      "num_tokens": 19452074.0,
      "step": 8715
    },
    {
      "entropy": 5.470134925842285,
      "epoch": 0.9793901274779581,
      "grad_norm": 0.875,
      "learning_rate": 0.0004915167003215999,
      "loss": 5.2674,
      "mean_token_accuracy": 0.1982989028096199,
      "num_tokens": 19463085.0,
      "step": 8720
    },
    {
      "entropy": 5.672174739837646,
      "epoch": 0.9799517043859156,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004915057763086838,
      "loss": 5.4308,
      "mean_token_accuracy": 0.18489108979701996,
      "num_tokens": 19472726.0,
      "step": 8725
    },
    {
      "entropy": 5.500839471817017,
      "epoch": 0.9805132812938732,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00049149484540214,
      "loss": 5.2709,
      "mean_token_accuracy": 0.19282296150922776,
      "num_tokens": 19482275.0,
      "step": 8730
    },
    {
      "entropy": 5.540094566345215,
      "epoch": 0.9810748582018307,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004914839076023161,
      "loss": 5.3673,
      "mean_token_accuracy": 0.1930101215839386,
      "num_tokens": 19492782.0,
      "step": 8735
    },
    {
      "entropy": 5.481459474563598,
      "epoch": 0.9816364351097883,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004914729629095609,
      "loss": 5.2441,
      "mean_token_accuracy": 0.19420044869184494,
      "num_tokens": 19505709.0,
      "step": 8740
    },
    {
      "entropy": 5.552941751480103,
      "epoch": 0.9821980120177458,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004914620113242224,
      "loss": 5.3416,
      "mean_token_accuracy": 0.19303423017263413,
      "num_tokens": 19516269.0,
      "step": 8745
    },
    {
      "entropy": 5.560787487030029,
      "epoch": 0.9827595889257034,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004914510528466495,
      "loss": 5.304,
      "mean_token_accuracy": 0.19139294028282167,
      "num_tokens": 19526786.0,
      "step": 8750
    },
    {
      "entropy": 5.532483434677124,
      "epoch": 0.9833211658336609,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004914400874771913,
      "loss": 5.3226,
      "mean_token_accuracy": 0.19039182215929032,
      "num_tokens": 19537865.0,
      "step": 8755
    },
    {
      "entropy": 5.454063606262207,
      "epoch": 0.9838827427416185,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004914291152161967,
      "loss": 5.2872,
      "mean_token_accuracy": 0.1969689294695854,
      "num_tokens": 19548320.0,
      "step": 8760
    },
    {
      "entropy": 5.55257215499878,
      "epoch": 0.984444319649576,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000491418136064015,
      "loss": 5.2831,
      "mean_token_accuracy": 0.19893567860126496,
      "num_tokens": 19558856.0,
      "step": 8765
    },
    {
      "entropy": 5.567751789093018,
      "epoch": 0.9850058965575336,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000491407150020996,
      "loss": 5.3729,
      "mean_token_accuracy": 0.19709732085466386,
      "num_tokens": 19570792.0,
      "step": 8770
    },
    {
      "entropy": 5.556698894500732,
      "epoch": 0.9855674734654911,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004913961570874895,
      "loss": 5.2904,
      "mean_token_accuracy": 0.19880957901477814,
      "num_tokens": 19581982.0,
      "step": 8775
    },
    {
      "entropy": 5.497059345245361,
      "epoch": 0.9861290503734487,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004913851572638453,
      "loss": 5.2543,
      "mean_token_accuracy": 0.19287339150905608,
      "num_tokens": 19592882.0,
      "step": 8780
    },
    {
      "entropy": 5.601714849472046,
      "epoch": 0.9866906272814062,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004913741505504138,
      "loss": 5.4643,
      "mean_token_accuracy": 0.18941612541675568,
      "num_tokens": 19605967.0,
      "step": 8785
    },
    {
      "entropy": 5.597661256790161,
      "epoch": 0.9872522041893638,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004913631369475454,
      "loss": 5.3059,
      "mean_token_accuracy": 0.19688683599233628,
      "num_tokens": 19616494.0,
      "step": 8790
    },
    {
      "entropy": 5.591172122955323,
      "epoch": 0.9878137810973213,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004913521164555909,
      "loss": 5.3998,
      "mean_token_accuracy": 0.19313349723815917,
      "num_tokens": 19627013.0,
      "step": 8795
    },
    {
      "entropy": 5.552135467529297,
      "epoch": 0.9883753580052789,
      "grad_norm": 0.78125,
      "learning_rate": 0.000491341089074901,
      "loss": 5.2942,
      "mean_token_accuracy": 0.19946253299713135,
      "num_tokens": 19639099.0,
      "step": 8800
    },
    {
      "entropy": 5.4626716613769535,
      "epoch": 0.9889369349132364,
      "grad_norm": 0.921875,
      "learning_rate": 0.000491330054805827,
      "loss": 5.2779,
      "mean_token_accuracy": 0.19900939166545867,
      "num_tokens": 19650650.0,
      "step": 8805
    },
    {
      "entropy": 5.557640838623047,
      "epoch": 0.9894985118211939,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004913190136487201,
      "loss": 5.3083,
      "mean_token_accuracy": 0.19384468495845794,
      "num_tokens": 19662869.0,
      "step": 8810
    },
    {
      "entropy": 5.52273964881897,
      "epoch": 0.9900600887291514,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004913079656039318,
      "loss": 5.32,
      "mean_token_accuracy": 0.19614305794239045,
      "num_tokens": 19673232.0,
      "step": 8815
    },
    {
      "entropy": 5.534802961349487,
      "epoch": 0.990621665637109,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004912969106718142,
      "loss": 5.3288,
      "mean_token_accuracy": 0.19278744012117385,
      "num_tokens": 19685069.0,
      "step": 8820
    },
    {
      "entropy": 5.579485177993774,
      "epoch": 0.9911832425450665,
      "grad_norm": 0.81640625,
      "learning_rate": 0.000491285848852719,
      "loss": 5.4265,
      "mean_token_accuracy": 0.1862275630235672,
      "num_tokens": 19697018.0,
      "step": 8825
    },
    {
      "entropy": 5.572660207748413,
      "epoch": 0.9917448194530241,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004912747801469984,
      "loss": 5.2834,
      "mean_token_accuracy": 0.2001495510339737,
      "num_tokens": 19707922.0,
      "step": 8830
    },
    {
      "entropy": 5.4422228813171385,
      "epoch": 0.9923063963609816,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000491263704555005,
      "loss": 5.2398,
      "mean_token_accuracy": 0.1928875058889389,
      "num_tokens": 19718693.0,
      "step": 8835
    },
    {
      "entropy": 5.50144739151001,
      "epoch": 0.9928679732689392,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004912526220770913,
      "loss": 5.3073,
      "mean_token_accuracy": 0.19468493163585662,
      "num_tokens": 19730846.0,
      "step": 8840
    },
    {
      "entropy": 5.530048656463623,
      "epoch": 0.9934295501768967,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004912415327136103,
      "loss": 5.3248,
      "mean_token_accuracy": 0.19607166051864625,
      "num_tokens": 19742099.0,
      "step": 8845
    },
    {
      "entropy": 5.5552326202392575,
      "epoch": 0.9939911270848543,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000491230436464915,
      "loss": 5.3195,
      "mean_token_accuracy": 0.1953793615102768,
      "num_tokens": 19753216.0,
      "step": 8850
    },
    {
      "entropy": 5.52514271736145,
      "epoch": 0.9945527039928118,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004912193333313589,
      "loss": 5.2624,
      "mean_token_accuracy": 0.19465669095516205,
      "num_tokens": 19765316.0,
      "step": 8855
    },
    {
      "entropy": 5.516343784332276,
      "epoch": 0.9951142809007694,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004912082233132951,
      "loss": 5.2933,
      "mean_token_accuracy": 0.19606266766786576,
      "num_tokens": 19776208.0,
      "step": 8860
    },
    {
      "entropy": 5.58620285987854,
      "epoch": 0.9956758578087269,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004911971064110778,
      "loss": 5.3218,
      "mean_token_accuracy": 0.19771239757537842,
      "num_tokens": 19788065.0,
      "step": 8865
    },
    {
      "entropy": 5.532779359817505,
      "epoch": 0.9962374347166845,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004911859826250607,
      "loss": 5.2601,
      "mean_token_accuracy": 0.19397482573986052,
      "num_tokens": 19799388.0,
      "step": 8870
    },
    {
      "entropy": 5.54878830909729,
      "epoch": 0.996799011624642,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004911748519555981,
      "loss": 5.3411,
      "mean_token_accuracy": 0.18842808604240419,
      "num_tokens": 19809933.0,
      "step": 8875
    },
    {
      "entropy": 5.663855743408203,
      "epoch": 0.9973605885325996,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004911637144030444,
      "loss": 5.4326,
      "mean_token_accuracy": 0.19064835011959075,
      "num_tokens": 19821566.0,
      "step": 8880
    },
    {
      "entropy": 5.607615995407104,
      "epoch": 0.9979221654405571,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004911525699677541,
      "loss": 5.3856,
      "mean_token_accuracy": 0.1939131423830986,
      "num_tokens": 19833727.0,
      "step": 8885
    },
    {
      "entropy": 5.601259803771972,
      "epoch": 0.9984837423485147,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004911414186500822,
      "loss": 5.3924,
      "mean_token_accuracy": 0.19136462807655336,
      "num_tokens": 19845176.0,
      "step": 8890
    },
    {
      "entropy": 5.526238584518433,
      "epoch": 0.9990453192564722,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004911302604503837,
      "loss": 5.2998,
      "mean_token_accuracy": 0.18740231841802596,
      "num_tokens": 19857528.0,
      "step": 8895
    },
    {
      "entropy": 5.548251390457153,
      "epoch": 0.9996068961644298,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004911190953690139,
      "loss": 5.2376,
      "mean_token_accuracy": 0.19450028836727143,
      "num_tokens": 19868232.0,
      "step": 8900
    },
    {
      "entropy": 5.5173755751715765,
      "epoch": 1.0001123153815914,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004911079234063284,
      "loss": 5.1763,
      "mean_token_accuracy": 0.19815027879344094,
      "num_tokens": 19877881.0,
      "step": 8905
    },
    {
      "entropy": 5.441170883178711,
      "epoch": 1.000673892289549,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004910967445626826,
      "loss": 5.2635,
      "mean_token_accuracy": 0.2042294844985008,
      "num_tokens": 19888398.0,
      "step": 8910
    },
    {
      "entropy": 5.588966321945191,
      "epoch": 1.0012354691975065,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004910855588384327,
      "loss": 5.2331,
      "mean_token_accuracy": 0.19393816590309143,
      "num_tokens": 19899661.0,
      "step": 8915
    },
    {
      "entropy": 5.6033649921417235,
      "epoch": 1.0017970461054642,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004910743662339348,
      "loss": 5.291,
      "mean_token_accuracy": 0.19550175964832306,
      "num_tokens": 19911073.0,
      "step": 8920
    },
    {
      "entropy": 5.497482061386108,
      "epoch": 1.0023586230134216,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004910631667495452,
      "loss": 5.2196,
      "mean_token_accuracy": 0.2009853392839432,
      "num_tokens": 19922567.0,
      "step": 8925
    },
    {
      "entropy": 5.576870584487915,
      "epoch": 1.0029201999213793,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004910519603856207,
      "loss": 5.3047,
      "mean_token_accuracy": 0.19970787167549134,
      "num_tokens": 19934547.0,
      "step": 8930
    },
    {
      "entropy": 5.541706085205078,
      "epoch": 1.0034817768293367,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004910407471425179,
      "loss": 5.3166,
      "mean_token_accuracy": 0.19714968353509904,
      "num_tokens": 19947659.0,
      "step": 8935
    },
    {
      "entropy": 5.618759393692017,
      "epoch": 1.0040433537372944,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000491029527020594,
      "loss": 5.223,
      "mean_token_accuracy": 0.19711460173130035,
      "num_tokens": 19958150.0,
      "step": 8940
    },
    {
      "entropy": 5.522732162475586,
      "epoch": 1.0046049306452518,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004910183000202061,
      "loss": 5.2108,
      "mean_token_accuracy": 0.20407613962888718,
      "num_tokens": 19968173.0,
      "step": 8945
    },
    {
      "entropy": 5.476001024246216,
      "epoch": 1.0051665075532095,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004910070661417119,
      "loss": 5.2624,
      "mean_token_accuracy": 0.1943003013730049,
      "num_tokens": 19978543.0,
      "step": 8950
    },
    {
      "entropy": 5.5286647319793705,
      "epoch": 1.005728084461167,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004909958253854687,
      "loss": 5.2288,
      "mean_token_accuracy": 0.1909499555826187,
      "num_tokens": 19989586.0,
      "step": 8955
    },
    {
      "entropy": 5.525811767578125,
      "epoch": 1.0062896613691246,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004909845777518347,
      "loss": 5.2384,
      "mean_token_accuracy": 0.19470058232545853,
      "num_tokens": 20000524.0,
      "step": 8960
    },
    {
      "entropy": 5.568044757843017,
      "epoch": 1.006851238277082,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004909733232411681,
      "loss": 5.3456,
      "mean_token_accuracy": 0.19450613856315613,
      "num_tokens": 20012935.0,
      "step": 8965
    },
    {
      "entropy": 5.568811702728271,
      "epoch": 1.0074128151850397,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004909620618538271,
      "loss": 5.2843,
      "mean_token_accuracy": 0.19141610115766525,
      "num_tokens": 20024435.0,
      "step": 8970
    },
    {
      "entropy": 5.573111915588379,
      "epoch": 1.007974392092997,
      "grad_norm": 0.875,
      "learning_rate": 0.0004909507935901702,
      "loss": 5.2894,
      "mean_token_accuracy": 0.19018115252256393,
      "num_tokens": 20036017.0,
      "step": 8975
    },
    {
      "entropy": 5.558777904510498,
      "epoch": 1.0085359690009548,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004909395184505564,
      "loss": 5.3071,
      "mean_token_accuracy": 0.19670409858226776,
      "num_tokens": 20047216.0,
      "step": 8980
    },
    {
      "entropy": 5.542899894714355,
      "epoch": 1.0090975459089122,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004909282364353445,
      "loss": 5.1705,
      "mean_token_accuracy": 0.19697398990392684,
      "num_tokens": 20057439.0,
      "step": 8985
    },
    {
      "entropy": 5.554143285751342,
      "epoch": 1.0096591228168699,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004909169475448938,
      "loss": 5.2987,
      "mean_token_accuracy": 0.197182796895504,
      "num_tokens": 20069602.0,
      "step": 8990
    },
    {
      "entropy": 5.472641181945801,
      "epoch": 1.0102206997248273,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004909056517795638,
      "loss": 5.2777,
      "mean_token_accuracy": 0.19489063173532487,
      "num_tokens": 20081366.0,
      "step": 8995
    },
    {
      "entropy": 5.558881044387817,
      "epoch": 1.010782276632785,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004908943491397141,
      "loss": 5.2624,
      "mean_token_accuracy": 0.19739680290222167,
      "num_tokens": 20091384.0,
      "step": 9000
    },
    {
      "epoch": 1.010782276632785,
      "eval_entropy": 5.379588173257957,
      "eval_loss": 5.343245983123779,
      "eval_mean_token_accuracy": 0.20053441575297168,
      "eval_num_tokens": 20091384.0,
      "eval_runtime": 18.4214,
      "eval_samples_per_second": 1683.529,
      "eval_steps_per_second": 210.461,
      "step": 9000
    }
  ],
  "logging_steps": 5,
  "max_steps": 89030,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 3000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4665094071091200.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}